u.s. natural gas availability: gas supply through the year 2000

8/14/2019 U.S. Natural Gas Availability: Gas Supply Through the Year 2000

http://slidepdf.com/reader/full/us-natural-gas-availability-gas-supply-through-the-year-2000 1/249

U.S. Natural Gas Availability: Gas SupplyThrough the Year 2000

February 1985

NTIS order #PB84-114909



Recommended Citation:

U.S. Natural Gas Availability: Gas Supply Through the Year 2000 (Washington, DC: U.S.Congress, Office of Technology Assessment, OTA-E-245, February 1985).

Library of Congress Catalog Card Number 85-600507

For sa le b y the Superintend ent o f Doc ume nts

U.S. Government Printing Office, Washington, DC 20402



Foreword

This report responds to a request from the House Com mitte e on Energy andCommerce and its Subcommittee on Fossil and Synthetic Fuels to conduct a study ofthe domestic (Lower 48 States onshore and offshore) natural gas availability over thenext few de c ad es. OTA wa s asked to examine bo th co nventiona l and unco nventiona lsources of natural gas, review current estimates of resource bases and production poten-tials, and examine key technical issues that will affect the future development of those

source s. This request wa s sup ported by the Sub c om mitte e on Energy Resea rch a ndDeve lop me nt of the Sena te C om mittee o n Energy a nd Natural Resourc es. Two interimrepo rts we re relea sed in c onjunc tion with the reque st: a Tec hnica l Me mo ra nd um onU.S. Natural Gas Availability: Conventional Gas Supply Through the Year 2000, a nda Staff Mem orandum on The Effects of Decontrol on 0/d Gas Recovery. This rep ortcombines the results of the earlier reports with considerable additional material onthe unconventional gas sources to provide a comprehensive survey of the natural gassource s ava ilab le to the United Sta tes during the next few d ec ades.

Nationa l c onc erns about g as supplies have ea sed c onsiderab ly with the c urrentsurplus deliverability and recent improvements in the rate of additions to the Nation’sp rove d reserves. The Co ngress’ pe rc ep tion of the Nation’ s long-term g as sup p ly p ros-pects will still, however, play an important role in its consideration of several impor-

ta nt p olicy initia tives. These initia tives includ e p ressure for rep ea l of c onstraints on ga susage embodied in the Powerplant and Industrial Fuel Use Act of 1978, continuingatte mp ts to d ec ontrol natural ga s pric es, and initiatives to supp ort the a dd ed use ofnatural gas in p owe r generation a s pa rt of a strate gy to red uce ac id rain.

In addition, arguments about the appropriate Federal role in research on unconven-

tional gas sources hinges in large part on Congress’ concerns about the need for newsource s to supp lem ent the United State s’ trad itiona l do me stic g as source s. OTA ho pe sthat this report will help to clarify the continuing arguments about gas’ potential long-te rm role in future U.S. ene rgy c onsump tion a nd assist the C ongress in formulat ingeffective natural gas and energy policies.

OTA rec eived sub sta ntial help from ma ny orga niza tions and ind ividu als during thec ourse o f this stud y. We w ould like to thank the p rojec t’ s c ontrac tors, who p rep a red

c ritica l bac kground a na lyses, the p rojec t’ s advisory panel and the w orkshop pa rtici-pants, who provided guidance and extensive crit ical reviews, and the many additionalreviewers who gave their time to ensure the accuracy of this report.

. . .Il l



U.S. Natural Gas Availability Panel

William Vogely,Department of Mineral Economics,

Marc CooperEnergy Director

Consume r Fed eration o f Ame rica

Lloyd ElkinsPetroleum ConsultantTulsa, OK

Ed EricksonProfessor of Economics and BusinessNorth Carolina Sta te University

Daniel GrubbVice President o f Ga s Sup p lyNatural Gas Pipeline Co. of America

John HaunProfessor of GeologyColorad o Sc hoo l of Mines

Don Kash

George Lynn Cross Research Professor ofPolitic a l Sc ienc e, a nd Resea rch Fellow

Sc ienc e a nd Pub lic Policy ProgramUniversity of Oklahoma

Chairman Pennsylvania State University

Lawrence MossIndependent ConsultantEstes Park, CA

Roy E. RoadiferChief GeologistMobil Oil Corp.

John Sc han zSenior Specialist in Energy Resource PolicyCongressional Research Service

Benjamin Sc hlesing erIndependent ConsultantBethesda , MD

John C . Sha rerDirectorNatural Ga s Sup p ly Resea rch

Gas Research Institute

John WeyantDeputy Director

Energy Modeling ForumStanford University

Harry C. KentDirec tor, Pote ntial Gas Agenc y



OTA U.S. Natural Gas Availability Project Staff

Lionel S. Johns, Assistant Director, OTA Energy, Materials, and International Security Division

Richard E. Rowberg, Energy and Materials Program Manager

Steven E. Plotkin, Project Director

Julia C. Crowley, Unconventional/ Gas Sources

Vykie G. Whipple, Gas Imports, Devonian Shale Gas, Coal Seam Methane

David Strom, Computer Supply Models

Administrative Staff

Lillian Quigg Edna Sa und ers

Other Contributors

Joseph Riva, Congressional Research Service

Kathryn Van Wyk, Editor (Part I)

Contractors and Consultants

Richa rd Nehring , Co lorad o Springs, COJensen Associates, Inc., Fairfax, VA

Jame s W. Crafton, Go lden, COJohnston & Associates, Haymarket VA



AcknowledgmentsMany other people contributed to this report by providing information, advice, and substantive reviews

of d raft ma terials. OTA wo uld like to spec ially thank Lawrenc e Drew o f the U.S. Geo log ica l Survey, Rob ertKalisch of the American Gas Association, Wallace Howes of the Natural Gas Pipeline Company of America,Vello Kuuskraa of Lewin & Associates, and Charles Matthews of the Shell Oil Co. for their exceptional contribu-tions to this report. We would like to acknowledge the

Dayne Adams,El Paso Natural Gas

Ovid BakerMobil Research & Development Corp.

Porter BrownColumbia Gas Transmission Co.

David Burnett,Chandler & Associates

Paul ChingShell Oil Co.

Tim ConsidineBank of America

Tod d Dosc herDoscher’s Group, Inc.

Everett EhrlichCongressional Budget Office

Scott FarrowCarnegie Mellon University

William Fishe r, Robert FinleyUniversity of Texas at Austin

Michael German, Jeffrey WingenrothAmerican Gas Association

Gail HarrisonWexler, Reyno lds, Ha rrison , & Sc hule

Robert A. Hefner, IllThe GHK Companies

Steve Holditch

Texas A&M UniversityC. J. Holland, Jr.Sun Exploration & Production Co.

Howard Kilchrist, Webster Gray,Wayne Thom pson, Vic tor Zab el

Federal Energy Regulatory Commission

Charles KomarMorga ntow n Energy Tec hnolog y Cente r,

U.S. Department of Energy

valuable contributions of the following individuals:

Charles MankinOklahoma Geological Survey

Richard O’Neill, Lucio D’Andrea, Phil ShambaughEnergy Information Administration,U.S. Department of Energy

Gary PaglianoCongressional Research Service

Dudley Rice, Gordon Dolton, Charles SpencerU.S. Geological Survey, Denver, CO

David Root, John Houghton, Richard Meyer,Terry Offield

U.S. Geological Survey, Reston, VA

R.J. SaucierShell Western E&P, Inc.

John Sche unema yer

University of DelawareJ. N. Sommerfrucht, David White, Bill James,

Robert Megill, Warren Strickland, W. R. Finger,Exxon Co., U.S.A.

Ralph VeatchAmoco Production Co.

W. R. WatkinsDeGolyer & MacNaughton

Steve Watson, Steve Minihan, Richard JefferisU.S. Department of Energy

James WhiteU.S. Department of Energy

Ralph WillardDowell Division of Dow Chemical Co.

Thomas Woods, Daniel Dreyfus, Joseph RosenbergGas Research Institute



Workshop on Unconventional Natural Gas Resources

Dayne AdamsDivision of Reservoir EngineeringEl Paso Natural Gas

Ovid BakerMobil Research & Development Corp.

David Burnett

Chandler & AssociatesMarc CooperConsumer Fed eration o f Am eric a

Ed EricksonDepartment of Economics and BusinessNorth C arolina Sta te University

John Haun

Colorad o Sc hoo l of Mines

Dona ld KashSc ience and pub lic Polic y Prog ramUniversity of Oklahoma

Cha rles Kom arMorga ntow n Energy Tec hnology Ce nter

Vello KuuskraaLew in & Assoc iate s

John McCormickEnvironm enta l Policy Institute

Joseph RosenbergGa s Research Institute

Charles SpencerOil and Gas Resource BranchU.S. Geological Survey

James R. White

Department of Energy

Ralph Willard

Dowe!l Division of Dow Chemical Co.

vii



Chapter 1

Introduction and Overview



introduc tion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Projections of Future production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

The Ga s Resourc e Base. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Natural Gas and Energy Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

How This Report is Orga nized . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

FIGURE

3

5

6

79

10

Figure No. Page

1. Natural Gas production, Additions to Reserves, andTotal Reserves of the U.S. Lower 48 States . . . . . . . . . . . . . . . . . . . . . . . . . . . 4



4 q U.S. Natural Gas Availability: Gas Supply Through the Year 2000

Figure 1.

300

280

260

240

220

200

180

—Natural Gas Production, Additions to Reserves, and Total Reserves of the U.S. Lower 48 States

Years

30

1 5 1 -

20

‘ t

Production

I

Additions

1950 1955 1960 1965 1970 1975 1980 1985

Years

SOURCE American Gas Association, The Gas Energy Supply Outlook: 1983-2000, October 1983.

low leve l of p rove d reserves (whic h increa ses thevolatility of p rod uc tion); 2) substa ntial cha nges

in gas prices and demand; 3) rapid advances intechnology and the subsequent emergence ofnew and relatively unproven gas supply regionssuc h a s the Western Overthrust Belt; and 4) thepotential shift of production from the familiar con-ventiona l sources to the less fam ilia r unco nven-tional sources of gas.

In rea c tion to this c ha ng ing outloo k for U.S.

natural gas supply, the House Committee on—.——‘The current proved reserves in the Lower 48 States would very

quickly be depleted without a constant influx of new reserves. Afailure to add substantially to reserves would soon be followed by

a major decline In gas production.

Energy and C omm erce a nd its Subc om mittee onFossil and Synthetic Fuels, supported by the Sub-committee on Energy Research and Developmentof the Sena te Co mm ittee o n Energy and NaturalResource s, asked OTA to c ond uc t a study of do-mestic (Lower 48 States onshore and offshore) nat-

ural gas availability over the next few decades. Thestudy wa s to exam ine b oth c onventional and un-conventional sources of natural gas, review cur-rent estimates of resource bases and productionpotentials, and examine key technical issues thatwill affect the future development of those

sources.

This rep ort presents the results of OTA’ s studyof U.S. na tural g as ava ilab ility,



Ch. 1—lntroduction and Overview q 5

OVERVIEW

There is a d istinc t possibility tha t the UnitedStates’ traditional domestic sources of natural gas,with only modest help from supplemental sources,will be sufficient to maintain present levels of gas

usag e for the next few d ec ad es. In fac t, the p ro- jection of only moderate rates of decline in pro-duc tion from these trad itiona l source s, ba rringany collapse of drilling activity, appears to havea growing c onsensus amo ng gas ana lysts. OTAconcludes, however, that this is not the onlyplausible future course for U.S. natural gas avail-ability; it is also quite possible that gas suppliesfrom current sources will decline considerablymore sharply over the next few d ec ad es thanpredicted by most recent forecasts. In otherwo rds, the unce rtainty of both the future p rod uc-tion a nd tota l rec overab le resources of na tural

ga s is still high , and the range of p lausible ga s“futures” is greater than is generally acknowl-ed ge d. Thus, co mp lace nc y ab out U.S. natural gasavailabil ity over the next few decades would bean error. If the United States wants to be confi-dent about the availability of a continued highlevel of gas supply for the next few decades, itmust ensure that it can gain access to significantnew sources of supply in case its traditional supply

turns downward.

In a dd ition to “ co nventional” ga s obta inab leat current prices with available technology, the

United States has substantial gas resources thatcan come into production only with improvedrecovery technologies or higher gas prices, or

both, This gas resides in tight gas reservoirs, inDevonian sha les, in c oa l sea ms, and in ha rder-

to-produce conventional-type gas reservoirs.Also, there are substantial possibilities for ex-

pa nde d ga s imp orts. For each of these alterna-tives, the magnitude of the recoverable resourcebase and, for the new sources, the time requiredto reach high production (or import) levels are dif-ficult to assess. Given the uncertainties aboutfuture production, reliance on only one or two of

the alternatives might expose the United States tofuture gas supply shortages. Instead, a diversifieddevelopment strategy that allows for access to allpotential gas sources appears most desirable. OTA

believes that the probability of obtaining adequate

gas supplies is high if such a strategy is pursued.

Conventional and Unconventional Gas

Conventional natural gas is gas that is recover-able using technology that is either currently avail-

able or is a modest extension of current technol-ogy, at prices similar to or slightly higher thantoday’s. Virtua lly a ll of the resource estima tes of“ rem a ining rec ove rab le U.S. g as resource s” -

including the well-known resource estimate pub-lished by the U.S. Geological Survey 9 –are meantto be estimates of conventional gas only.10 Con-seq uently, OTA’ s ana lysis of future c onventiona lgas supplies excludes gas whose recovery de-pends on new technologies that are not readilyforeseeable extensions of existing technologies,or well head prices much higher than today ’s. Inaddition, in order to focus on technical rather

than market uncertainties in the supply of con-ventional gas, the analysis assumes that demandfor gas is high enough that exploration and pro-duction are not curtailed because of soft markets.Consequently, “pessimistic” scenarios examinedin the analysis of conventional supply reflect onlypessimism about technical prospects for gas dis-covery and production and do not reflect the pos-sibility that low gas demand may drive down ex-

ploratory drilling, discovery rates, and production.

Also, the a nalysis c anno t a c c ount for any effectsthat higher gas prices and advancing technologymay have on expansion of conventional produc-

tion into deeper and more hostile waters andother ge ologica lly c onventional b ut (c urrently)uneconomic formations.

Unconventional gas is produced from reservoirs that are different in geologic character from conventional gas reservoirs, and requires higher gas

prices or significant advances in production tech-nology—or both—for its economic recovery.Thus, in exa mining un c onven tiona l ga s sup p lies,OTA has relaxed the p ric e a nd te c hnolog y as-sumptions adopted for evaluating conventional

9G. L. Dolton, et al., Estimates of Uncl/sco\+ered Reco\erab/e Con-ven t iona l Resources of Oi/ and Gas in the Urr\ted States, U.S. Geo-

logical Sur\ey Circular 860, 1981.10However, It IS ce~a i n that these estimates do i nc I ude some gas

that might be characterized as unconventional. Because some pro-

duction of tight sands gas and gas from Devon Ian shales ha~

occurred In the pa~t and occurs today, the houndary between ‘ ‘con-

k en t i ona l ’ and ‘ ‘unconventional” is ambiguous.



6 qU.S. Natural Gas Availability: Gas Supply Through the Year 2000

gas supply. Unconventional gas sources includegas from Iow-permeability11 sandstone and lime-stone formations (so-called “tight gas”), Devo-nian shales, coal seams, and geopressurized aqui-fers. Recently, gas vented from deep within theEarth (“d eep source g as” ) and ga s hydrates–ga s

trapped with water in an ice-like state—have beenad ded to the “ unconvent ional” c atego ry. (Seebox A for definitions of the individual unconven-tional sources.) Of these six unconventionalsources, three—tight gas, Devonian shale gas, andcoal seam methane—generally are considered to

be the most likely to play a significant role in U.S.gas sup p ly within the next 20 to 30 yea rs. OTA

I I permeabil i ty is a Measure of how easily Iiqu ids and gases flow

through porous rock. Thus, low-permeability rock is rock through

which liquids and gases may flow only with difficulty.

has chosen to focus its analysis on these three

sources.

Projections of Future Production

OTA finds that, even if the uncertainties about

future gas prices and markets could somehow beeliminated, the technical and geological uncertain-ties associated with the gas resource base and ex-

ploration process are too great to allow a reliableconsensus to be established about a single “mostlikely” estimate of future annual gas production.For the production of conventional gas, a credi-ble range for Lower 48 State production levels in

the year 2000 is 9 trillion to 19 trillion cubic feetper year (TCF), assuming gas demand remains high

and gas prices do not soar. This range encom-



Ch. 1—Introduction and Overview q 7

passes drastically different conceptions of the role

of conventional natural gas in the United States’future energy supply. The w id th of the rangereflects uncertainties about the magnitude andcharacter of the remaining conventional resource,the appropriate interpretation and extrapolation

of past discovery trends, and the production ratespossible from reserves not yet discovered.

Projections of unconventional/ gas production

suffer from many of the geologic uncertaintiesassociated with projections of the conventional re-source, but in a more severe form because therehas been less information-gathering and becausemeasurement of geologic parameters generally ismore difficult in the unconventional fields. I n ad-dit ion, these projections are confounded by dif-f iculty in forecasting the rate of development ofnewly emerging production technologies and bythe lack of an extensive production history to pro-

vide guidance about the shape of future produc-

t ion. Furthermore, because of the technologicalrisks, the pa c e o f deve lop ment o f the unc onven-tional gas sources will be particularly sensitive toga s p rices and to the a va ilab ility of low er risk pro-spects for conventional gas production.

In light of these uncertainties, OTA believes that

the year 2000 production of unconventional tightgas, over and above production in areas being de-veloped today, could range anywhere from 1 to4 TCF/yr or perhaps even higher, depending ongas prices, conventional gas production, the pace

of tight gas research programs, and the outcomeof numerous geological and technological uncer-tainties. Similarly, new Devonian shale produc-tion could range from negligible amounts to about1.0 to 1.5 TCF/yr by the year 2000. At this time,the technical uncertainties associated with pro-duc ing c oal seam methane a re too great to pro-vide a usefuI estimate of future production fromthis resource.

In ad d ition to do me stic Low er 48 prod uc tion,

the United States can supplement its gas supplywith pipeline i reports from Canada and Mexico,

and impo rts of lique fied na tural ga s (LNG) froma variety of suppliers. Also, there is the possibil-ity of pipeline or LNG deliveries of natural gasfrom Alaska. OTA ha s examined ava ilab le esti-mates of future pipeline gas imports and Alaskangas delive ries to the U.S. Lowe r 48 Sta tes; the se

estimates range from 1 to about 6 TCF/yr by 2000.LNG imports are even less certain, and were not

projected. Factors affecting gas imports includethe U.S. domestic gas supply balance, gas prices,and the export policies and energy supply situa-tions of the exporting nations.

The separate projections of conventional gas

production, unconventional gas production, andgas imports cannot simply be added together toyield a projection of total U.S. natural gas avail-ability, because the projections have differentbaseline assumptions and because the projectionsare not independent—the gas volume attained byany one of the sources affects and is affected bythe vo lume a tta ined by the others. The na tureof this interdependence is that neither the“ low s’ ’ -the p essimistic e stima te s–no r the“highs’ ’-the optimistic estimates–are likely tooc c ur tog ether in the sam e sc ena rio.

Year 2000

Resource production, TCF/yr Conventional gas . . . . . . . . . . 9-19Tig ht g a s . . . . . . . . . . . . . . . . . 1 -4+

Devonian shale gas . . . . . . . . 0-1.5Coal sea m m etha ne . . . . . . . . unknownimports and Alaskan gas . . . . 1-6+

The Gas Resource Base

Uncertainties about the magnitude and char-acter of the conventional and unconventional gasresource bases are an important source of uncer-tainty in projections of future gas production.12

For conventional gas, the critical areas of uncer-tainty inc lude the:

role of small gasfields, which up to now haveprovided an extremely small share of cumu-lative production and proved reserves;po tential of dee p onshore ga s, and ga s infrontier areas such as the Eastern and West-ern Overthrust Belts, the eastern Gulf ofMexico, the Georges Bank, and elsewhere;potential for obtaining substantial quantitiesof a dditional ga s from olde r ga sfields; andthe possibility of finding large quantities ofga s in stra tigraph ic traps13 bypassed by oldexploration methods:

I zTh I ~ I $ not a trl~,lc] I qa tement. It the resource base were suH-

clently large, u ncertal nty about the rnajorlt y of It stl II might not at-

tect projections of near-term production, which presumably will

draw trom the best-understood portion of the base.I JSrrC2tlgrcqph;c traps: t ra Ps, 1.e., geologic barriers that ‘ ‘trap’ gas

and 011 and allow, them to accumulate, formed by gradual chan~et

i n the permeahl Iity of sedimentary layers rather than by (more ea$l Iy

detected) abrupt structural shifts and deformation of the layers.




In ad d ition, there rem ain cha lleng ing q uestionsabout the appropriate interpretation of past trendsin ga s d isc overy and the ir usefu lness in projec t-ing remaining discoveries. Accounting for theseuncertainties, OTA concludes that a reasonablerange for the remaining conventional gas resource

in the U.S. Lower 48 States is 430 to 900 TCF. Be-c ause the d efinition of “ c onventional” includesprice and technology constraints, however, thisrange is conservative; it excludes gas in deep-water offshore areas, in deep onshore formations,and in sma ll fields that w ill be a dd ed to the e c o-nom ica lly rec ove rable resource base by futuregas price increases and technological advances.

Three unconventional source s are exam inedin this report. Of the three, tight gas ge nerally is

c onside red to ha ve the largest reco verable re-

source base. Because tight formations often occur

in basins that have undergone much convention-al gas development, and because considerabledevelopment of relatively t ight gas has alreadyoccurred, there is a substantial base of informa-tion from w hic h to project the tight g as resource.The primary a reas of rema ining unc ertainty a sso-ciated with the tight gas resource base magnitude

include:

q

q

q

the volume of recoverable gas present in theNorthern Great Plains and in the numeroustight gas basins that are unexplored or lightlyexp lored;the a bility of well stimula tion tec hnolog iesto allow produc tion from low-pe rmea bility“ lenses” (sma ll, disc ontinuous reservo irsthat occur in large, thick formations) that arenot penetrated directly by the wellbore(without suc h “remote ” p rod uc tion, much

of the tight resource w ill not be ec onom i-cally recoverable except at extremely highgas prices); andour future ability to create very long frac-tures at low costs.

In addition, the size of the tight gas recoverableresource is quite sensitive to gas prices and to re-

quired rates of return. Given the uncertaintiesassociated with all the above factors, the tight gas

140f the several existing estimates of the tight gas resource, on IY

one–the study by the National Petroleum Council—has attempted

to estimate the resources in the unexplored/lightly explored basins.

recoverable resource will most likely be in therange of 100 to 400 TCF. A more optimistic—butstill plausible—view held by some industry sci-

entists would raise the high end of the range bya few hund red TCF.

Devonian shale gas, like tight g as, has a c on-siderab le history of p rod uc tion, but its deve lop -ment has been more constrained both in area andin technological innovation, and there is less in-formation available for a reliable estimate of re-coverable resources. In particular, until recently,resource appraisers did not have the reservoirmodeling capabilities that are available to asses-sors of tight gas resources. primary areas of uncer-tainty include: basic geological/resource charac-teristics of areas outside the current limitedde velopme nt area; the po tential rec overy effi-c ienc y a vailab le with new stimulation tec hnol-

og ies and imp roved drilling pa tterns; and the p o-tential for ec onom ic rec overy from areas that d onot have well-developed natural fracture systems.Given these uncertainties, a moderately conserv-ative range for the Devonian shale recoverable re-source is 20 to 50 TCF for the Appalachian Basin,

15

with a reasonable potential for up to 80 or 100

TCF with high gas prices and successful technol-ogy development. in addition, there is some po-tential to add considerably to the recoverable re-source if a m ea ns is found to p rod uc e from theunfrac tured pa rt of the shale.

Coal seam methane has not bee n extensivelydeveloped to date, although there are small de-velopment efforts in the Black Warrior Basin inAlabama, the San Juan Basin in New Mexico, and

elsewhere. It is the Ieast understood of the threeresources, with important uncertainties associ-ated with the basic characteristics of the coal re-source, the gas production mechanisms, and thepossibilities for and characteristics of new stimula-tion methods. In addition, access to the gas resid-ing in shallow, minable coal seams is hamperedby c onc erns ab out ow nership o f the ga s and thepossibility that well stimulation (fracturing) will

da ma ge the integrity of the rock overlying thecoal seam, adversely affecting mine safety. Ex-

I sL~ss is known a b o u t the two other Devonian shale basins, the

Michigan and Illinois Basins. However, recoverable resources ap-

pear !ikely to be considerably smaller than those in the Appalachian

Basin,



Ch. 1—Introduction and Overview 9

isting resource estimates are crude, and althoughit seems likely that the recoverable resource willbe a t lea st a few multip les of 10 TCF, the range

of possible resource values probably extends, atleast spec ulat ively, up into the 100s of TCF a t high

ga s p ric es.

Recoverable Resource resource, TCF Conventional gas . . . . . . . . . . 430-900+Tight gas . . . . . . . . . . . . . . . . . 100-400+Devonian shale gas . . . . . . . . 20-100+Coal seam methane . . . . . . . . 20-200 +

Natural Gas and Energy Policy

The resource a nd p rod uc tion estimate s haveimplications for national energy policy. One ofthe more important is that any policy that wouldtend to restrict U.S. gas availability to “con-ventional” gas supplies—e.g., a policy that re-

stricted gas prices to below market levels and thusdiscouraged technology development–wouldstrengthen the possibility that natural gas could

be in short supply by the 1990s. This is be c ausethe lower end of the range of year 2000 produc-tion potential for conventional gas is 9 TCF/ yr,far below expected gas requirements. A will-ingness to let gas prices seek a market level andan active encouragement of technology develop-ment and the exploitation of new gas sourceswould make it more likely that any shortfall ofconventional gas could be made up by alternativegas sources.

The to ta l rec ove rab le ga s resource ba se w ill re-spond to price increases and technology ad-vances in a number of ways. First, as notedabove, the boundaries—and thus the magni-tude—of the conventional gas resource base willexpand with higher prices and improved technol-ogy. These b ound aries a re d efined b y ma ximumwater depth, minimum exploitable field size,

maximum feasible drilling depth as a function offield size and geology, minimum “pay” (gas-bea ring) thickness, and so on.

16Not on ly will for-

me rly unec ono mic fields and reservoirs now be

developed, but measures will be taken to increasega s reco very from fields and reservoirs who se d e-

16An Ongol ng OTA study, Tec hnology fo r De~’eloping Offshore

0/ / and Ga s Resources In Hos/1/e Environments, is examining one

of these ‘ ‘bound aries. ”

velopment would have been less intensive underthe old conditions. In fact, because gas prices insome fields have been controlled at below-marketrates, gas recovery could be increased merely by

allow ing g as from these fields to obta in tod ay’sfree m arket p ric es. OTA calculated the potential

increase in recoverable gas from decontrol of theprice-controlled fields, over and above increasesalready programmed into existing legislation, tobe 19 to 38 TCF.17

Sec ond , the magnitude of the unconventionalgas recoverable resource base will increase sub-stantially with higher gas prices and advances inrecovery technology. For example, current studiesimply that doubling gas prices from today’s levelswould approximately double the recoverable

tight gas resource with present technology. Sim-ilarly, solving the numerous remaining technicalproblems associated with the unconventionalresources—improving logging techniques, ex-panding effective fracture lengths and increasingfracture efficiencies, developing accurate reser-voir simulation models, and so on—will allow

higher recovery efficiencies and open up mored iffic ult areas to c om me rc ial exp loitat ion. (Seebox B for a list o f tec hnica l req uirem ents for de -veloping the unconventional gas resources.) Withtight gas, which ha s alrea dy see n c onsiderab lecommercial exploitation and technology devel-opment, further technology development stillholds the promise of expanding the recoverable

resource by 40 percent or more. With Devonianshale gas and coal bed methane, further technol-ogy development holds the promise of evenlarger g a ins.

Another important conclusion is that, given thehigh risks and long Ieadtimes necessary to estab-lish new sources of supply, the United Statesshould place a high premium on providing an earlywarning of any impending shifts in gas supply.

Comprehensive data collection and gas supplyana lysis c apab ilities exist outside of the Fed era l

Government, for example, in organizations suchas the Ame ric an Ga s Assoc ia tion a nd Ga s Re-

sea rch Institute. The p ersp ec tives of these o rga -nizations and the uses to which they put theirforecasts may be quite dissimilar to the perspec-

‘ ~At a market price of $3,50/JMCF (1 983 dollars),

3 8 - 7 4 2 0 - 8 5 - 2



10 q U.S. Natural Gas Availability: Gas Supply Through the year 2000

tive and required uses of the Government, how-

ever. Consequently, Congress may not wish torely solely on such organizations for warningsabout impending supply problems, Inside theFederal Government, the Energy Information Ad-ministration plays the critical role in collectingand analyzing natural gas supply statistics. Main-

HOW THIS REPORT IS

The rema inde r of the rep ort is orga nized asfollows:

@ Part l—Co nventiona l Ga s Supplies:

– Chapter 2: Summary--Avai labi l i ty of Con-

–

ventional Gas Supplies summarizes OTA’ smajor findings about future production ofconventional gas, the magnitude of the con-ventional gas resource base, and the poten-tia l for ga s imp orts to t he Lower 48 Sta tes. —

– Chapter 3: Natural Gas Basics presents abrief review of basic natural gas terminologyand concepts.

– Chapter 4: The Conventional Natural Gas

tenance of ElA’s capabilities in this area, as well

as protection of i ts independence from the De-partment of Energy policymaking apparatusshould be considered a high priority by thosevaluing an independent warning system for futuresupply problems.

ORGANIZED

ua tes a num ber of c ritica l resourc e issues,and p resents OTA’ s c onc lusions about t hema gnitude of the rema ining resource b ase.Chapter 5: Conventional Gas Production Po-

tential describes four approaches used byOTA to evaluate the ga s produc tion p oten-tial to the yea r 2000, and p resents OTA’ sc onc lusions ab out t his po tential.Chapter 6: Gas imports–Overview brieflyreviews the prospects for gas imports to theLow er 48 Sta tes–liquefied ga s imports andpipeline imports from Alaska, Canada, andMexico.

Resource Base reviews resource assessment q Part n-Unc onvent iona l Gas Supp lies:methodologies, describes and critiques sev- – Chapter 7: Introduction and Summary– eral specific gas resource assessments, eval- Availability of Unconventional Gas Supplies



Ch. 1—introduction and Overview . 11

briefly defines the unconventional gas re-sources, describes the definitional problemof separating unconventional from conven-tional gas, and introduces the major uncer-ta inties in de fining the resource base a ndproduction potential, then describes thetechnologies for producing unconventionalgas and summ arizes OTA’ s findings about

the size of the resource base and the futureproduction potential for this gas.

– Chapter 8: Tight Gas describes the tight gasresource and the technology necessary toexploit it, and discusses and evaluates projections of the resource base and future pro-

duct ion. – Chapter 9: Gas From Devonian Shales a nd

Chapter 10: Coalbed Methane dup l i ca te

c hap ter 8 for these tw o resources.



Part I

Conventional Gas Supplies*

q The ma teria l in Pa rt I is based on : U.S. Natural Gas Availability: Conventional Gas Supply Through

the Year 2000—A Technical Memorandum (Washington, DC: U.S. Co ngress, Office of Tec hno logyAssessme nt, OTA-TM-E-12, Sep temb er 1983). At the time o f pub lic a tion o f the tec hnica l me mo ran-dum, data on natural gas production and reserves were available through 1981. Currently, data areavailable through 1983. A portion of the material in Part I has been revised and updated to take ac-c ount of this new d ata . How ever, the two a dd itiona l years of da ta g enerally do no t cha nge the c onc lu-sions of Part I as originally presented. A small exception is the raising of the lower bound of year1990 production potential from 13 to 14 TCF.



Chapter 2

Summary: Availability of

Conventional Gas Supplies



Contents

Page

Introduc tion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Major Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Natural Gas Prod uc tion Pote ntial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18Unc erta inty 1: The Ga s Resource Base . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19Unc erta inty: Inte rpreta tion a nd Extrapola tion o f Disc overy Trend s . . . . . . . . 23Unc erta inty 3: Prod uc tion From Prove d Reserves—The R/ P Ra tio . . . . . . . . . . 24Summa ry of Assumptions and Co nd itions Und erlying OTA’ s Projec tions . . . . 25

TABLES

Table No.

1. Ga s Prod uc tion Forec asts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. Alternative Estimates of Remaining Conventional Natural Gas

Resourc es in the U.S. Lower 48 Sta te s. . . . . . . . . . . . . . . . . . . . . . . . .3. Bases for OTA’ s Projec tions of Natura l Gas Prod uc tion—Baseline

Assump tions: Good Ma rket C ond itions, Rea dily Foreseeab le Tec hnolog y . . .

FIGURE

Figure No.

2. Alternative Concepts of the Natural Gas Production CycleIf Rema ining Resourc es = 400 TCF . . . . . . . . . . . . . . . . . . . . . . . . . . .

Page

. . . . . . . 19

. . . . . . . 19

. . . 25

Page

. . . . . . . 20



Chapter 2

Summary: Availability of Conventional Gas Supplies

INTRODUCTION

Within the last 5 years or so, the general per-

c ep tion a bo ut the outloo k for future U.S. ga s sup-plies has moved from pessimism to considerableop timism. The pessimism wa s based partly o n

short-term prob lems, suc h a s period ic reg ionalshortages, and partly on disturbing long-termtrends, such as the declining finding rate for newgasfields and, since the late 1960s, the ominousand apparently unstoppable decline of provedreserves. The new op timism is based on severalfactors, including the gas “bubble” caused bydeclining gas demand coupled with high gas de-liverab ility, the reb ound of reserve a dd itions tolevels which e xc eed ed prod uc tion in 1978 and1981, and continuing optimistic estimates of do-mestic gas resources by the U.S. Geological Sur-vey (USGS) and the industry-based Potential GasCommittee (PGC).

What does this apparent change in the outlookfor U.S. natural gas supply me an? Ca n w e no wc ount on na tural ga s to p lay a ma jor, pe rhap seven expanded role in satisfying U.S. energy re-quirements, or is the seeming turnabout only atemporary respite from a continuing decline in

gas reserve levels and, soon to follow, a decline

in gas production capabilities?

Part I of th is rep ort p resents OTA’ s assessment

of the future prospects for the discovery and pro-duction of conventional natural gas in the Lower48 Sta te s. This assessme nt exa mines the gas re-source base and future production potentialunder the following conditions:

q

q

q

wellhead prices are assumed not to changesubstantially from today’s levels in real terms,new technologies that are not rea dily fore-seeable extensions of existing technology are

not considered, anddemand is assumed to be high enough toavoid reductions in production-potential dueto curtailment of investments in explorationand product ion

Part I also summarizes the prospects for addi-tional conventional supplies to the Lower 48 frompipeline imports from Canada and Mexico, Alas-kan ga s, liquefied natu ra l ga s (LNG) imp orts, andsynthetic g as from c oa l.

MAJOR FINDINGS

Certain technical uncertainties-primarily

those associated with incomplete geological un-derstanding, alternative interpretations of pastdiscovery trends, and difficulties in projectinglikely patterns of future gas discoveries—are sosubstantial that by themselves they prevent areliable estimation of the remaining recoverablegas resource and the likely year 2000 produc-tion rate. Even after ignoring the potential for sig-nif ic ant c hang es in g as pric es and tec hnology in

the future, OTA could not narrow its range ofestimates of resources and future productionbeyond a factor of 2 from the lowest to the high-est estimate. Inclusion of uncertainties associatedwith changing ga s prices and m arket d ema nd a nd

the continuing evolution of gas exploration andproduction technology would undoubtedlywiden the range still further.

Spec ific findings of the stud y are as follows:

q Current proved reserves in the Lower 48States will supply only a few trillion cubic feet(TCF) pe r yea r of p rod uc tion by the yea r2000. All other domestic production must

come from gas which has not yet been iden-

tified by drilling.qThere is no c onvinc ing b asis for the c om mo n

argument that the a rea of the Lowe r 48 State sis so inte nsively explored and its geo logy is

so w ell know n tha t there is a sub sta ntial co n-

17



18 . U.S. Natural Gas Availability: Gas Supply Through the Year 2000

sensus on the magnitude of the gas resourceba se. Plausible estimate s of t he amo unt o f

remaining conventional natural gas in theLow er 48 Sta tes that is rec ove rable und erpresent and easily foreseeable technologicaland economic conditions range from 430 to

900 TCF. At the lowe r end of th is rang e, p ro-duction in the year 2000 will be seriouslyconstrained by the magnitude of the re-source base.

qAssureing market conditions generally favor-able to gas exploration and production andno radical changes in technology or gaspric es, plausible estima tes of the yea r 2000

production potential of conventional natu-ral ga s in the Lowe r 48 Sta tes rang e from 9to 19 TCF/ yr. In 1990, produc tion is likelyto be a nywhere from 14 to 20 TCF/ yr.

qBecause it is unclear whether the recent

surge in the rate of additions to proved gas

q

q

reserves’ is sustainable, the range of plausi-ble annual reserve additions is wide even forthe nea r future. The range for the Low er 48Sta tes for 1987 and beyo nd is from 7 or 8TCF/ yr up to 16 or 17 TCF/ yr, assum ing tha tthe current excess of gas production capac-

ity c ea ses and m arket co nd itions improve.The rate a t which g as ca n be withdraw n from

proved reserves, or R/P (reserves-to-produc-tion) ratio, may range from 7.0 to 9.5 as anational average by the year 2000, further

ad ding to the d iffic ulty of p rojec ting futureproduction potential.An imp ortant source of unc ertainty in eva l-

uating past discovery trends is the lack ofpublicly available, unambiguous, disaggre-ga te d ata ab out ga s disc overies.

‘The 1981 addition was about 21 TCF v. about 10 TCF/yr or less

for 1969-77.

NATURAL GAS PRODUCTION POTENTIAL

OTA finds insufficient evidence on which tobase either an optimistic or a pessimistic outlookfor conventional domestic gas production.Given market conditions generally favorable togas exploration and production, but assumingthat real gas prices do not rise well above today’slevels, the production of natural gas from con-ventional sources within the Lower 48 States

could range from 9 to 19 TCF/yr by the year2000. Simila rly, p roduc tion in the yea r 1990c ou ld rang e from 14 to 20 TCF/ yr. Current an-nua l p rod uc tion is ab out 17 to 18 TCF/ yr and ac -tual produc tion capacity is p rob ably a t least 1 or2 TCF/yr higher. These ranges do not include gasfrom pipeline or LNG imports, synthetic gas from

coal or other materials, or gas from unconven-tional sources that a re not p rod ucing to da y. Theydo inc lude ga s from low-pe rmeability reservoirsthat is currently economically recoverable, even

thoug h this ga s is borde rline c onve ntional a ndmight be considered unconventional by some as-sessors.

OTA’ s wide range for plausible levels of c on-ventional gas production in the Lower 48 Statesin the year 2000 is in sharp contrast to the rela-

tively narrow range displayed in publicly avail-a b le forec asts. Ta b le 1 presents the summa rizedresults of 20 separate forecasts from oil com-panies, other private institutions and individuals,and Government agencies. A striking feature ofthis group of forecasts is that 13 of the 15 fore-

c asts that project a year 2000 prod uction levelfall within 77 to 75 TCF/yr. This high leve l ofag reeme nt for a product ion rate two d ec ad es inthe future is made all the more unusual by theprobability that there are substantive differencesin the baseline assumptions used by the variousforec asters. The h igh leve l of a greement m ight,however, reflect the probability that the forecasts

are not all independent, original estimates; somemay simply be averages of other forecasts, reflect-

ing the “conventional wisdom, ” and some mayhave been influenced by others that precededthem.

The w ide rang e in OTA’ s projec tion o f futuregas production reflects the existing high degreeof uncertainty about:

1. the magnitude and character of the gas re-source base;



Ch. 2—Summary: Availability of Conventional Gas Supplies . 19

Table 1 .—Gas Production Forecasts (In trillion cubic feet)

GovernmentOil companies Other p riva te agenc ies Average O TA

1985Lowest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.0 15,5 16.5 — —Ave rag e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.7 17.1 17.3 17.9 —

Highest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.5 18.3 18.0— —

1990Lowest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.9 13.6 14.3 — 14 Average . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.1 15.4 15.1 16.7 —Highest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.8 17.7 15.5 — 20

2000Lowest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.9 11.6 12.8 9Average . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

—

13.5 12.2 13.1 13.1 —Highest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.6 13.5 13.5 — 19

Number of individual forecasts . . . . . . . . . . . . . . . . . 9 6 5 — —

NOTE: All forecasts calculate gas on ’’dry’’ basis at standard temperature and pressure, Some forecasts include unconventional sources of supply, such as tight sandsand Devonian shales; others include only conventional sources.

SOURCE Office of Technology Assessment, based on data in Jensen Associates, Inc., “Understanding Natural Gas Supply in the U.S.,’’ contractor report to the Officeof Technology Assessment, April 1983

2. the a pp rop riate interpretation a nd extrap -elation of past trends in natural gas discov-ery, and ;

3. the rapidity with which gas in proved re-

serves can be produced, expressed as thereserves-to-production ratio.

The first two sou rce s of unc erta inty are insep a ra-ble; the magnitude and character of the resourcebase have played–and will continue to play– an important role in shaping trends in gas dis-c ove ry, and these trend s in turn provide imp or-tant clues to gauging the remaining resourcebase. Consequently, uncertainties in trend inter-pretation automatically contribute to uncertain-ties in resource assessment, and resource un-certainties in turn complicate the process ofprojecting future discovery trends. Similarly, esti-mating future R/P ratios will depend on project-ing discovery trends and understanding the char-

ac ter of the rem a ining resource s.

Eac h of the three sourc es of unc ertainty will bediscussed in turn.

Uncertainty 1: The Gas Resource Base

Many individuals and organizations have pub-lished assessme nts of the na tural g as resourcesof the Lower 48 Sta tes. Tab le 2 presents sevensuch estimates of the gas resources that remainedin the Low er 48 a t the beg inning of 1983. They

Table 2.—Alternative Estimates of RemainingConventional Natural Gas Resourcesa in theU.S. Lower 48 States (as of Jan. 1, 1983)

TrillionSource b (publication date) cubic feet

Hubbert (1980) . . . . . . . . . . . . . . . . . . . 244RAND Corp. (1981). . . . . . . . . . . . . . . . 283Bromberg/Hartigan (1975). . . . . . . . . . 340Shell (1984) . . . . . . . . . . . . . . . . . . . . . . 525Wiorkowsky (1975). . . . . . . . . . . . . . . . 663US. Geological Survey (1981) . . . . . . 774Potential Gas Committee (1983) . . . . 916

aThe term “resources” includes proved reserves, expected growth of existingfields, and undiscovered recoverable resources. In all but the Hubbert estimate,the term does not include gas not recoverable by current or readily foreseeable

btechnology nor gas not recoverable at price/cost ratios similar to today ’s.

In most cases, the sources for these estimates were assessments of either theultimately recoverable resource or the undiscovered resource base, Theestimates shown are derived by subtracting cumulative production fromestimates of ultimately recoverable resource or by adding proved reserves andexpected growth of known fields to estimates of the undiscovered resource,Where ranges of resource estimates are given by the source, the estimate inthis table is based on the mean value,

SOURCE: Office of Technology Assessment, 1983

rang e from Hub bert’ s 244 TCF to the PGC ’s 916TCF.

The resource estimate s at high a nd low endsof the range in table 2 have quite different

messages for gas production forecasters. At theupper end , the USGS and PGC estima tes imp ly

that gas production in this century will be rela-tively unconstrained because of the resource basemagnitude—although this does not rule out thepo ssib ility tha t p rod uc tion ma y be sha rply con-

strained by the character of the remaining re-




sources. 2 In contrast, estimates at the lowerend—Hubbert, RAND, and Bromberg/Hartigan—imp ly a serious resource c onstraint. If these esti-mates are correct, gas production will decline

substantially by the year 2000 (see fig, 2). There-fore, selec tion of a “ best” resource estimate , or

narrowing of the range, could conceivably haveprofound implications for expectations of futuregas production.

Som e o f the d ifferences in the estimate s ma yme rely b e the result of d ifferenc es in b aseline a s-sumptions or boundaries. For example, variousassessments may use different assumptions abouteconomic conditions and the state of exploration

2For ~xample, by their location, depth, degree Of contamination,

and size distribution of fields and reservoirs.

and recovery technology and may have differentge og rap hic al bo unda ry c ond itions. They ma y ormay not include areas currently inaccessible todevelopment, gas from portions of tight sands orother “ unco nventional” sources that are p res-ently recoverable, or nonmethane components

of the gas. Finally, assessments may differ in theirdefinitions of the degree of certainty that shouldbe attached to the estimate. Unfortunately, many

assessments do not fully specify their assumptionsand definitions, nor is it always clear what effectsthese assumptions have on the resource esti-mates. Consequently, it is not possible to “nor-malize” the various estimates so that they are fullycomparable. s —

3Th is does not imply, of course, that some normalization cannot

be accomplished. For example, PGC has incorporated into its re-

Figure 2.-Alternative Concepts of the Natural Gas Production Cycle lf Remaining Resources = 400 TCF(conventional gas only)

Curves A and B represent two conceptions of how the produc-tion cycle for natural gas in the Lower 48 might be completed ifthe remaining resources as of December 31, 1982, were 400TCF, Curve B represents a future where explorers and pro-

ducers work extremely hard to maintain production at highlevels for as long as possible. But even in this situation, pro-duction drops drastically by the year 2000. Curves can bedrawn that would allow production to remain high through2000 while obeying a resource constraint of 400 TCF, but thesecurves would not be realistic. A substantial portion of the re-maining resource cannot be produced quickly enoughbecause it is located in small fields, which require many ex-ploratory wells and many years to find, or in low permeabilityreservoirs, which typically release small quantities of gas overlong periods of time.

1900 1920 1940 1960 1980 2000 2020 2040 2060 2080 2

Year

00



Ch. 2—Summary: Availability of Conventional Gas Supplies q 21

It is OTA’ s op inion , how eve r, tha t “ normaliza -tion” of the various estimates would not elimi-nate the major differences between them. OTAfinds no convincing basis for the common argu-

ment that the area of the Lower 48 States is sointensively explored and its geology is so well

known that there is a substantial consensus onthe magnitude of the gas resource base.

Instead, there are several substantive resourcebase issues that remain unresolved. Among themore important of these are:

The Use of Past Discovery Trends

The extrap olation o f p a st trend s in the d isc ov-ery of na tural gas ha s genera lly led to p essimis-tic estimates of the magnitude of the gas resourcebase. For example, of the resource base assess-me nts exam ined b y OTA, three of the four that

used trend extrap olation tec hnique s arrived atestima tes tha t we re at least 400 TCF below theUSGS med ian estimate. Ac c ep tanc e o f d isc overytrend extrapolation as a valid method of resourcebase assessment, therefore, can yield conclusions

about the magnitude of the resource base thatare radically different from those that result fromusing o the r assessme nt m ethods.

The va lid ity of using p ast d isc overy trend s toestimate the magnitude of the resource base de-

pends on whether the trends are affected moreby the nature of the resource b ase tha n by the

general economic and regulatory climate of thetimes. Resource “optimists” argue that the disap-pointing trends in gas discovery of the past fewdecades have resulted from controlled gas prices,high levels of p rove d reserves, and limited ma r-kets that until recently gave little incentive for

high-risk or high-cost drilling. They argue that ex-trapolation of these trends is invalid because theeconomic and regulatory conditions that createdthe trends have changed. Resource “pessimists”argue that the trends are driven mainly by a de-pleting resource base and are affected only

——source estimate quantities of presently recoverable gas in tight res-

ervoirs, whereas both RAND and USGS have tended to exclude

this gas from their resource estimates. Consequently, equalizing

the conventional/u nconventional boundaries of the assessments

should reduce the differences between PGC’S estimate and those

of USGS and RAND.

minimally by economic and regulatory condi-

tions; therefore, extrapolation is valid.

In a dd ition to this basic issue, othe r questions

ha ve a risen o ver the valid ity and interpreta tionof resource estimates based on extrapolation ofpast trends. For example, the accuracy of early

records of gas discovery and production is ques-tionable; thus, trend analyses cannot accuratelyincorporate the entire discovery and productionhistory. Also, the precise economic, technologi-cal, geographic, and geologic boundaries of these

estimate s are d iffic ult to d efine.

The Potential of Small Fields

Although fields that contain less than 60 billionc ubic feet (BCF) of g as have p layed a minor rolein gas production, some analysts believe thatsmall fields will have a major role in the future.

The d ifferenc e b etw een op timistic a nd pessimisticestimates of the future role of small fields maybe 100 TCF or more. In OTA’ s judg me nt, the ar-

gum ents on b oth sid es are b ased p rima rily on un-proven statistical models of field size distributionsand on economic tradeoffs that are highly sensi-tive to gas prices. Only time and further explora-tion will settle this issue.

New Gas From Old Fields

There a re sha rp d isag reeme nts about t he e x-

tent to which the resources recoverable fromolder producing fields may respond to price

inc reases. The m ec ha nisms to inc rease the “ ulti-mately recoverable resources” of these fieldsmight include delaying well abandonments, ei-ther by reworking damaged wells or by merelycontinuing operation to lower abandonmentpressures, drilling a t sma ller sp a c ing to loc at e ga spockets that otherwise would not be drained,fracturing the reservoir rock to allow recovery

from low permea b ility p ortions of fields, and ex-panding development to formerly subeconomicportions of fields. Currently, estimates of the po-tential increase in recoverable resources rangefrom a few TCF to about 50 TCF.

OTA ha s exa mined this issue in some deta il.4

We concluded that the potential additional re-

4See Office of Technology Assessment, ‘‘Staff Memorandum on

the Effects of Decontrol on Old Gas Recovery, ” February 1984.




serves that may be obtained from older fields, ifga s in these fields c an ob ta in a m arket pric e o f$3.50 to $4.50/MM Btu, is about 43 to 65 TCF.Because the NGPA maintains low controlledpric es in m any o lde r fields, the likely reserve in-

crease with no legislation changes or other new

incentives is somewhat lower: 20 to 35 TCF, withsome possibility of additional reserves from the

current price incentives for “stripper” (low pro-duction) wells and production enhancementme a sures. The a b ove estima tes includ e reservesfrom de layed ab and onme nt, drilling at sma llerspa c ing (infill drilling), a nd frac turing and othe rwell stimulation measures, but do not include re-serves that might be available from expanding de-velopment onto formerly subeconomic portionsof fields.

The Potential of Frontier Areas,Including Deep Gas

Although a ll resource analysts c onsider a rea s

such as the deep-water Gulf of Mexico, the deepAnadarko Basin, and the Western Overthrust Beltto have considerable gas potential, considerabledisagreement exists over the actual amount of re-

coverable resources in these areas. Recent indica-tions of engineering problems and rapid pressuredeclines from deep wells in the Anadarko, cou-pled with price declines from previous very highlevels, ra ise d oub ts about w hethe r much of thisarea’s gas resource will be part of the (currently)

ec ono mica lly rec ove rab le resource . Som e ofthese d oub ts must be temp ered, howe ver, by re-cent declines in drilling costs and the continuedad vance of de ep drilling tec hnology a nd e xper-tise. In the Overthrust Belt, doubts about the mag-nitude of the resource center on the significanceof the failure of exp lorers to find a gian t field o verthe past 3 to 4 years. Also, areas such as the east-ern Gulf of M exic o, the Southea st G eo rgia Em-bayment, the Georges Bank, and the BaltimoreCa nyon have b een expensive failures thus far,a nd the ir event ua l c on tribution to sa tisfying U.S.energy requirements is unknown.

Estimates of the recoverable resource poten-tial in the frontier area s va ry by up to 100 TCFor mo re (t he USGS and PGC d iffer by ne arly 30TCF in the ir assessme nts of the easte rn Gulf ofMexico, alone).

The Potential of Stratigraphic Traps

Stratigrap hic tra ps are b arriers to p etroleum mi-gration formed by gradual changes in the perme-ability of sedimentary layers rather than by abruptstructural shifts and deformation of the Iayers. Be-cause the structural traps are easier to locate, they

have been the primary targets for exploration.Som e explorers p red ict tha t large resources re-main to be found in “mature” areas in subtlestratigraphic traps. Although this issue is not set-t led , the op timistic argume nt is we akened by o b-servations that nume rous stratigrap hic trap s have been found in the Permian Basin and elsewhereand tha t the e xtensive d rilling in area s tha t ap -

pear to have good prospects for stratigraphic trapsshould have uncovered most of the larger traps,whic h g ene rally a re extensive in area . Thoug h itmay appear more likely than not that most of therem a ining und isc ove red trap s will be sma ll in vol-ume, a possibility exists that larger fields may haveremained hidden because of the less effective ex-p loration m etho ds used in the p ast a nd drilling

that, while extensive, might have clustered in thewrong p lac es or bee n too shallow.

In addition to these five issues, a level of uncer-tainty is ever present in the process of estimat-ing the quantity of a resource that cannot bemeasured direc tly prior to its ac tua l production.

The p resenc e of ec onom ic ally rec overab le c on-centrations of natural gas requires an unbroken

chain of events or conditions, the presence orabsence of which generally cannot be measureddirectly. First, adequate amounts of organic ma-terial and suitable temperature and pressure con-d itions for gas forma tion a nd preservation m ust

be p resent. Sec ond , the ga s must be free tomigrate , and third , an a deq ua te reservoir mustbe available in the path of migration to containthe ga s. Finally, there must b e a mec hanism totrap the gas, and the trap must remain un-bleached until the gas is discovered and pro-d uc ed . These source s of unc ertainty ac c ount forthe various manifestations of risk in natural gas

development–the large number of dry holesdrilled during exploration, the often huge dif-ferences in bids for leases, the multimillion dol-lar failures of many of the leased areas, and the

continuing disagreements over the size of theremaining resource.



Ch. 2—Summary: Availability of Convention/ Gas Supplies . 23

OTA to ok into a c c ount these ge neral issues, aswe ll as spec ific p rob lems with individua l assess-ments, in arriving at a plausible range for theamo unt o f rem a ining gas resourc es. In OTA’ s

judgment, a reasonable range for the amountof the remaining conventional natural gas in theU.S. Lower 48 that is recoverable under present

and easily foreseeable technological and eco-nomic conditions is 430 to 900 TCF as of De-

cember 1982. This range is somewhat narrowerthan the range displayed in table 2, because OTAconsiders the low end of the range of resourceestimates in the table to be overly pessimistic.How eve r, the g ene ral implic a tion of OTA’ s rangeis similar to the implication of the range in thetable: The uncertainty in estimating the remain-

ing recoverable gas resource is too high to de-

termine whether or not the resource base mag-nitude will constrain gas production in this

century. On the other hand, even the more op-timistic resource estimates imply that conven-tional gas production must decline sharply by theyear 2020 or before unless technological ad-vances and/or sharp increases in gas prices addsubstantial quantities of gas to the “economicallyrecoverable” category.

Uncertainty 2: Interpretation andExtrapolation of Discovery Trends

The key to projec ting ga s produc tion p otentialto the year 2000 is the successful prediction of

future discovery trends and of additions to provedreserve s. This foc us on the d isc overy proc ess isnecessary because gas that is already discovered,that is, gas in proved reserves, will be of dimin-ishing importance to production as we moveinto the 1990s. Assuming a constant R/P ratio of8.0, the current proved reserves of about 169TCF in the Lower 48 will provide only 2 TCF tototal production by the year 2000. All other pro-duction must come from gas added to provedreserves by the discovery of new fields, the dis-

covery of additional reservoirs in known fields(“new pool discoveries”), the expansion of the

areas of known reservoirs (“extensions”), andthe reserve changes due to new information orchanged economics or technology (“revi-sions’’). s

5This last category of reserve additions may be negative.

In addition to the effects of resource baseuncertainty, interpretation and extrapolation ofdisc overy trend s are ha mp ered b y a variety ofother prob lems. These inc lude :

Inadequate Discovery Indicators

The interpreta tion a nd e xtrap olation of trendsfor projecting future reserve additions require theavailability of discovery “indicator s,” such asfind ing ra tes for new field wildc a ts, that c an b e

interpreted in a relatively unambiguous fashion.OTA found tha t essentially all indicators avail-able from public data that describe the naturalgas discovery process have ambiguous interpre-tations bec ause the d ata are highly ag gregatedand are dependent on a wide variety of factors.For example, the “exploration” whose success

is be ing measured by a find ing ra te a c tua lly in-c lude s seve ral kind s of e xploratory d rilling, from

high-risk, high-return drilling that searches forgiant fields in new geologic horizons, to low-risk,low-return drilling that clusters around a new

strike or redrills already explored areas that havegrown more attractive with price increases. Be-cause the proportions of different varieties of

exploratory drilling may change substantiallywith changing market conditions, interpretingtrends in finding rates and other indicators ofexploration success is difficult. This is especiallytrue if the data are highly aggregated geograph-ically.

Uncertainty About the Future Growthof New Fields

At least three-quarters of past additions to

proved ga s reserves have c ome from the d isc ov-ery process that follows the discovery of newfields. This sec ond ary disc overy p roc ess see ksnew reservoirs in the field and the expa nsion ofknown boundaries of already discovered reser-voirs. The extent to which rec ently found fieldsand future fields will grow in the same manneras fields found in the p ast is c ritica l to future re-serve levels and thus to fu ture p roduc tion. The re

has been speculation that the decline in findinggian t fields–which require ma ny yea rs and d is-covery wells to develop fully—and the additionto the reserve b a se o f inc reasing num be rs of ve rysmall fields will lead to significant declines in fieldgrowth. If the new fields discovered in the past



24 q U.S. Natura l Gas Availab ility: Ga s Supply Through the Year 2000

few years do not grow at near-historic levels,

then reserve additions due to new pool discov-eries and extensions will decline substantiallyfrom recent levels, even if new field discoveriescan stay at their present higher rate. OTA be-lieves that such a decline in field growth is

plausible, but verification requires additionalanalysis at the individual field level and con-tinued observation of field growth trends.

Difficulties in Interpreting the Recent Surgein Reserve Additions

After the decade 1969-78, during which addi-tions to gas reserves in the Lower 48 States aver-aged less than 10 TCF/yr,6 reserve additions havesurged to over 20 TCF7 in 1981 and over 17 TCFin 1982. This surge has bee n the c enterpiec e o farguments for future high production levels.

In OTA’ s judg me nt, it is not clear whether ornot the recent high rates of additions to proved

gas reserves are sustainable, even if drilling ratesrebound to the levels achieved before the recentslump. For exa mp le, 13.5 TCF of the to ta l 1981

additions came from secondary discoveries, thatis, extensions and new pool discoveries. Nor-mally, such a surge in secondary discoverieswouId be preceded a few years earlier by an in-crease in new field discoveries, because recentlydiscovered fields provide the most promising tar-get areas for secondary discoveries. However, thenumb er of new fields d isc ove red in the 5 years

be fore 1981 did not seem high e nough to b e theprimary cause of 1981 high secondary discov-eries. Alternative o r ad ditional c auses of t he re-cent increases in secondary discoveries could in-clude: an acceleration in the normal pace of fieldgrowth (e.g., growth tha t normally might oc c urover a 20-year period instead is achieved in 5yea rs, yielding a short-term inc rease in “ p er yea r”reserve additions followed by a dropoff in later

years); the rap id de velopm ent o f a limited inven-tory of low -risk drilling p rospe c ts tha t ha d be enidentified in p rior years bu t ignored be c ause o f

unfavorable economic conditions; and a sub-

stantially increased growth potential for the cur-

rent (and future) inventory of discovered fields

because of the expansion of recoverable re-sources with higher prices and improved explora-tion and p rod uc tion tec hnolog y. The first twoc auses wo uId imp ly tha t sec ond ary disc ove rieswill decline sharply in the near future as the

limited inventory of prospects is used up; the thirdc ause implies that high levels of sec ond ary d is-c overies might b e susta inab le. In fa c t, it is likely

that all three causes played a role in the recentsurge, but their relative share is uncertain.

Simila rly, it is no t c lea r to w ha t extent rec enthigher reported rates of new field discoveries arecaused by any (or all) of the following factors: anincreased willingness of explorers to go afterriskier p rospec ts; the e xploita tion o f a limited in-vento ry o f low-risk p rospec ts identified by p astexploration; an increase in the numb er of ec o-

nomically viable fields, caused by improved tech-nology a nd higher pric es; and rec ent c hang es inreserve reporting methodologies. s

OTA p rojecte d a plausible range of future a d-d itions to Low er 48 gas reserves by trying to a c -count for uncertainties about the resource basemagnitude, the resource characteristics mostlikely to affect the discovery process, and the ac-

tual causes of past and recent discovery trends.OTA c onc lude d tha t, unde r the assumed de ma nd

/price/technology conditions, multiyear averagelevels of total reserve additions could range from7 to 8 TCF/yr to 16 to 17 TCF/yr or higher by

1987. Projected average values for individualc om pone nts of reserve ad d itions a re:

New field discoveries . . . . . . . . . . . . . . . . . 1.5-3.5 TCF/yrExtensions and new pool discoveries , . .6.0-11.0 TCF/yr

Revisions. . . . . . . . . . . . . . , . . . . . . . . . . . . 0-+2.0 T CF / y r

Uncertainty 3: Production FromProved Reserves-The RIP Ratio

The reserves-to-p roduc tion (R/ P) ratio reflec tsthe rate at which gas is being withdrawn from dis-covered reservoirs; consequently, it represents

the a nalytic al link betw een projec tions of new

6As reported by the American Gas Association.7As reported by the Energy Information Administration. The Amer-

ican Gas Association, the major source of reserve data prior to 1977,no longer publishes detailed information on reserve additions.

BThe American Gas Association reported U.S. reserve additions

until 1979. The Energy Information Administration began report-

ing reserve additions in 1977 using a different data collection and

analysis procedure, and modified this procedure in 1979.



Ch. 2—Summary: Availability of Conventional Gas Supplies q 25

d isc ove ries and forec asts of g as produc tion. There

are very large differences in R/P ratios from fieldto field, depending on the age, geology, location,and contract terms of the gas production. OTAprojects that the aggregate average R/P ratio forthe Lower 48 may range from 7.0 to 9.5 by the

year 2000,assuming that economic conditions aregenerally favorable to production (in other words,in contrast to to da y’s ga s “ bubb le” ). The R/ P ratio

in 1981 was 9.0, the result of a long and relativelysteady decline from a level of 30 in 1946.9

Although the R/P ratio is sensitive to economicfactors, such as actual and expected gas pricesand interest rates, technical factors will also playan impo rtant role in de termining this ra tio in thefuture. Gas in low-permeability reservoirs willplay an increasing role in reserves, tending to push

up R/ P levels. The impo rta nc e of o ffshore d eve l-opment will affect national R/P levels because off-

9The recent cutbacks in gas production because of weak demand

have temporarily raised R/P even though production capacity has

not necessarily fallen.

Table 3.—Bases for OTA’s Projections of Natural Gas

shore fields have typically been exploited veryquickly. As more and more gas is produced infrontier areas with very high drilling costs, diffi-cult tradeoffs will have to be made between thedesire for rapid production and the costs ofd rilling additional d eve lop ment w ells. The ra te

of adding new reserves—which itself is highlyuncertain-will determine the average age of theUnited Sta tes’ p roduc ing fields, an imp ortant fa c -tor in production rates. Uncertainty in these fac-tors makes it difficult to predict whether future

av erag e R/P leve ls will inc rease o r de c rea se fromtod ay's leve l.

Summary of AssumptionsConditions Underlying

OTA’s Projections

and

Tab le 3 summa rizes the assumptions a nd c on -

ditions that lead to the low and high ends ofOTA’s p rojec tion for c onve ntional ga s prod uc -tion fo r the Lowe r 48 Sta tes in the yea r 2000.

Production-Baseline Assumptions: Good MarketConditions, Readily Foreseeable Technology -

9 TCF/yr in 2000 19 TCF/yr In 2000

1. Magnitude of remaining resources: 430 TCF

2. Character of remaining resources: Remaining exploration playsa are only of moderate size;

few surprises. Some major potential remaining infrontier areas but deep resource is disappointing. Small

fields are only a minor source of additional gasbecause of economics and/or smaller numbers than astraight-line extrapolation would predict. Resource instratigraphic traps is disappointing; remaining growthof old fields is moderate.

3. Causes of past trends in gas discovery: Magnitude and character of the resource base were the

primary causes.

4. Meaning of recent surge in reserve additions: A temporary response to higher prices, drilling a backlog

of easy but formerly marginal prospects—notsustainable. Possibly also caused by a change inreporting practices.

5. Projected rate of future annual reserve additions: Total: Declines to 7.5 TCF by 1987:

New field discoveries: 1.5 TCFExtensions and new pool discoveries: 6.0 TCF by 1987Revisions: O

6. R/P Ratios: 9.5 by 2000, predicated on lower permeability reserve

additions, difficult production conditions.

900 TCF

High potential for major new exploration plays. Deepresource is both plentiful and economically accessible.Small fields may play an important role, but many largefields still remain. Resource remaining in mature areas,much of it in subtle stratigraphic traps, is substantial.Remaining growth of old fields is high.

Artificially low prices and rigid regulation were asimportant as the resource base.

An indication of a real turnabout in gas discovery; theopening up of major new exploration horizons, readilysustainable if exploratory drilling revives.

Maintained at 16.5 TCF or above for the next fewdecades:

3.5 TCF11.0 TCF+ 2.0 TCF

7.0 by 2000, predicated on high demand coupled withgenerally favorable physical conditions.

aPlay —.An exploratory campaign based on a cohesive geologic idea.

SOURCE Office of Technology Assessment, 1983.

3 8 - 7 4 2 0 - 8 5 - 3




OTHER SOURCES OF LOWER 48 SUPPLY

Aside from domestic conventional gas produc-tion, g as c onsume rs in the Low er 48 Sta tes ma yhave access to other sources of supply, includingproduction from so-cal led unconventional

sources (tight sands, coal beds, gas in geopres-surized aquifers, and Devonian shales); pipelineimports from Alaska, Canada, and Mexico; LNGimports from a variety of gas-producing nationsthroughout the world; and synthetic natural gasfrom c oa l and b iom ass. The p ote ntial sup p ly fromunc onventiona l source s is d isc ussed in Pa rt II ofthis rep ort. OTA p reviously d isc ussed synthe tic

natural gas in a report released in 1982.10

The United Sta tes c urrently imports about 0.8TCF of g a s pe r year, most of it from C an ad a . Be-c ause e ac h of the four sources of g as imp ort po-tential have substantial and accessible resources,

imports could theoretically satisfy a major por-tion of U.S. gas requirements later in this cen-

tury and beyond. However, each of the import

l o /nc rea5& Automobile Fuel Eficiency and Synthetic Fuels: Alter-

natives for Reducing 0;/ Imports (Washington, D. C.: U.S. Congress,

Office of Technology Assessment, OTA-E-185, September 1982).

sources, like future domestic production, is sub-ject to considerable uncertainty. High trans-portation costs are a particular problem forAlaskan gas and LNG, creating the need, at a min-

imum, to accept wellhead prices substantiallybelow equivalent oil prices. Similar problems ex-ist for Canadian and Mexican gas. Canadian andMe xic an expo rts to the United Sta tes must alsocompete with the uncertain future requirementsof their own domestic gas users. Based on avail-ab le stud ies, the expected import potential from

Canada, Alaska, and Mexico may range from 1.0

to 5.7 TCF11 in the year 2000, with Canada be-ing the most certain large contributor. LNG im-ports are even less predictable. Finally, OTA pro-

jects synthetic natural gas from coal to range fromO to 1.6 TCF/ yr by t he yea r 2000.

I I This estimate is derived by adding individual estimates for each

of the three import categories. This simple addition may overstate

the range somewhat because the categories are not independent.

For example, high levels of Canadian gas imports might serve to

dampen prospects for Alaskan gas, so it is not certain that the “high”

Canadian and Alaskan projections are compatible with each other.



Chapter 3

Natural Gas Basics



Chapter 3

Natural Gas Basics

This c ha pter b riefly describes basic theo ries and how discoveries are reported. Words in boldfacete rmino logy used within the rep ort-i. e., it briefly are d efined in Appendix D: Glossary. Readers fa-

describes what natural gas is, how it is formed, miliar with basic terminology and concepts of nat-

how it is found and subsequently produced, and ural gas supply may wish to skip this section.

WHAT IS NATURAL GAS?

As its name implies, na tural g as is a na turally dustry. Associated heavier hydrocarbons such as

occurring mixture of hydrocarbon and nonhydro- ethane, propane, and butane and impurities such

carbon gases found in subsurface reservoirs as water, hydroge n sulfide , and nitrog en o c c urwithin the Earth’s crust. Methane (CH4, a light with the methane. If the concentrations of thesehydrocarbon, is the primary constituent of natu- other constituents render the gas unmarketable,ral gas and of principal interest to the energy in- they must be removed prior to use.

HOW DOES NATURAL GAS FORM?

There is no universa lly ac c ep ted explana tionof how natural gas formed. Most hydrocarbon de-posits of significant size have been discovered insedimentary basins, however, and generally arethought to have originated from the decay andalteration of organic matter. Hundreds of millionsof years ago, seas that covered a large portionof the land exposed today were inhabited by tinyplants and animals that, upon dying, sank to the

bo ttom and were b uried unde r layers of sed i-ment. I n a rea s of rap id sedimentation, organicdecay was accompanied by high pressures andtemperatures which, over millions of years, ef-fectively “cooked” the organic material into pe-

troleum (oil and natural gas). Hydrocarbons mayalso have been formed by other processes: bythe anaerobic (without oxygen) digestion oforganic materials by bacteria at relatively lowtemperatures, and inorganically by the chemicaltransformation of carbon compounds at highpressures and temperatures deep within the

Ea rth. The frac tion of na tural ga s tha t is b ac teria-generated gas, usually called biogenic gas, is atissue, but estima tes rang e to a s high a s 20 per-

cent or more. ] As for the “deep earth gas, ” a major controversy exists as to whether or not thisgas even exists. According to arguments byThoma s Go ld a nd Steven Sot er of C ornell Univer-sity, most naturally occurring methane is of in-

orga nic o rigin and va st reserves rema in to b e d is-c overed at de pths be low 15,000 ft.2

Tem perature a nd p ressure c ond itions play a

c ritic a l role in de termining the p hysic a l sta te o fthe hydrocarbons that result. Natural gas may befound at all depths, but it originated mostly inrocks subjected to particularly high temperaturesand p ressures ove r long period s of t ime. It g en-

erally is the only hydrocarbon present at depthsbeyond 16,000 ft. Liquid hydrocarbons occur atsha llow er dep ths, from about 2,500 to 16,000 ft,

] D. D. Rice and G. E. Claypool, “Generation, Accumulation, andResource Potential of Biogenic Gas, ” AAPG f?u/ /et in, vol. 65, No.

1, January 1981.2

T. Gold and S. Soter, “Abiogenic Methane and the Origin ofPetroleum, ” Energy Exploration and Exploitation (city, st: Graham

& Trotman, Ltd., 1982).

29




where lower temperatures are characteristic.3

Most crude oil is found between 6,500 and 9,000ft, with light hydrocarbon liquids occurring atdepths greater than 9,500 ft.4 Biogenic gas is usu-. — -— —

3H. Douglas Klemme, Geotherrna/ Gradients, F/eat How, and Hy-

droca rbon Rec overy. Petroleum and Globa l Tec tonics (Princeton

and Landon: Princeton University Press, 1975), p. 260. Cited inJensen Associates, Inc., Understa nding Natu ral Gas Supp ly in the

U. S., contractor report to OTA, April 1983.4B. P. Tissot and D. H. Welte, Petroleum Formation and Occu r -

ally found at d ep ths of a few hund red to a fewthousand feet, be c ause it is formed at the lowtemperatures that accompany shallow burial andrarely is generated at depths greater than 3,000ft .5

rence (New York: Springer-Verlag, 1978), p. 202. Cited in Jensen

Associates, Inc., op. cit.

sRice and Claypool, op. cit.

WHERE IS NATURAL GAS FOUND?

Petroleum accumulations occur as reservoirsor pools—not in caverns or large holes in a rockma ss but in the m inute p ore spa c es be twee n thepa rticles tha t c om po se the roc k. The g reate r theamount of pore space in the rock (porosity), th e

larger the qua ntity of ga s or oil that m ay b e c on-ta ined w ithin it. Poo ls often oc c ur toge ther in afield, and multiple fields in similar geologic envi-ronm ents co nstitute a province.

Gas occurs separate from (nonassociated gas)and tog ether with oil. When to ge ther, it oc c ursin solution w ith the oil (dissolved gas) or, whenno m ore ga s c an be held in solution und er thepressure and temperature conditions of the reser-vo ir, as free gas (associated gas) in a g as ca p.

The sea rc h for hydrocarbo n ac c umulations is

narrow ed by the req uireme nt for the p resenc eof orga nic ma terial in the sed iment a t the timeof burial. Sedimentary basins are the areas mostlikely to have contained the organic-rich rocks–source rocks—required for petroleum formation.Sed imenta ry roc ks c om po se a bo ut 75 pe rc ent o fthe exposed rocks at the surface, but onlys per-c ent o f the Ea rth’s crust (o ute r 10 miles) . Theknown oil- and ga s-bea ring a rea s in the UnitedSta tes a re identified in figu re 3.

Commercial petroleum accumulations are notusua lly found in the source roc k. Source roc ks

are ge nerally of too low a pe rmea bility, mea n-ing the texture of the source rock does not allowpetroleum to flow ea sily throug h the p ores to aproduc ing well. Typica lly, after the pe troleum hasformed , the g ases and fluids (oil and formation

water) migrate from the source b ed to a morepermeable rock, called the “reservoir rock” (thisp roc ess is c alled “ primary migration” ). The fluidsmo ve in the pa th o f least resista nc e (o r highest

permeability) and continue migrating within the

reservoir rock (secondary migration) until an im-pe rme ab le ba rrier is enc ountered, which prohib-its further migration into adjacent or overlyingrock units or formations. The p etroleum the n

migrates further along the barrier to a place of

accumulation, called a trap—usually located atthe h ighest point whe re the reservoir roc k co n-tac ts the mo re imp ermea b le, b a rrier roc k. Thefour requirements for a hydrocarbon accumula-tion–a source rock, reservoir rock, impermeablebarrier rock, and trap–are illustrated pictoriallyin figure 4.

There a re three b a sic typ es of p etroleum trap s:structural, stratigraphic, and combination (seefig. s). Struc tural trap s are formed by ea rthmovements that deform or rupture rock strata,thereby creating favorable locations for hydrocar-bons to accumulate. Such structural features asfaults and anticlines create enclosures that serveas loc i for migrating p etroleum. Stratigraphic trap sare created by permeability and porosity changescharacteristic of the alternating rock layers thatresult from the sedimentation process. In strati-graphic traps, pinched-out beds, sandbars, orreefs serve as reservoirs for migrating petroleum.

Combination traps result from both structural andstratigraphic c ond itions. An examp le o f a c om-

bination trap is one that results from a salt domeintrusion during deposition that alters the thick-ness of the strata deposited.



Ch. 3—Natural Gas Basics 31

Figure 3. —Known Oil- and Gas. Bearing Areas in the Lower 48 States

Figure 4.—Four Requirements for

Petroleum Accumulation

Water layer,

Impermeable barrier:

-Reservoir rock:

organic-rich shale

SOURCE: Office of Technology Assessment



—

32 . U.S. Natural Gas Availability: Gas Supply Through the year 2000

Figure 5.-Trapping Mechanisms

Structural trap–fault

Gasaccumulation

Stratigraphic trap—pinched-out bed

Gas

accumulation

Combination trap—salt intrusion

Gasaccumulation

SOURCE: Office of Technology Assessment.

HOW IS NATURAL GAS DISCOVERED?

Before an understanding of subsurface geologywas acquired or rules of petroleum occurrencewere established, petroleum discoveries werebased on surface seeps, knowledge gained fromwa ter we ll drilling, and luc k. Tod ay the re a re avariety of concepts, exploration methods, and in-

struments available to help geologists locate sub-surface hydrocarbon accumulations.

The typ e o f exploration tec hniques used va riesbetween sites and depends on how much isknown about the area being explored. In areaswhere little is known about the subsurface, recon-naissance techniques—which provide limited in-forma tion o ver a large a rea –are used to ide n-

tify favorab le areas that wa rrant more d eta iledinvestigation. Satellite and high-altitude imagery



Ch. 3—Natural Gas Basics . 33

sometimes reveals large geologic features ortrends that are surface expressions of subsurfacege ologic structure. Ma gnetic and gravity surveysdetect changes in the magnetic or density prop-erties of the Earth’s crust and are also used to in-fer subsurface structure.

Once a promising area has been identified,more detailed, higher resolution exploratory tech-

niques are used to locate individual prospects forthe drill and to project conceptually relatedgroups of prospects, or plays. The seismic reflec-tion method, which measures and interprets thereflections of sound waves off of geologic discon-tinuities, is particularly effective for providingdetailed subsurface information. Drilling is thefinal stage of the exploratory effort and the onlysure way to determine if hydrocarbon-filled reser-voirs exist in the subsurface.

In some basins, drilling may be performed socheaply that predrilling exploration expendituresfor seismic surveys and other analyses are not

justified . These sha llow a reas a re bec om ing in-creasingly scarce, and the role of predrilling ex-ploration analysis is increasing in importance, par-

ticularly in frontier areas. If these high-cost areasare t o b e drilled, o p erato rs must b e relatively surethat the drilling expense is justified.

The de gree of risk involved in drilling d ep end son how much is known about the subsurface atthe drill site. As illustrated in figure 6, a classifica-tion scheme has been established to categorizethe exploratory wells based on their relationshipto kno wn pe troleum d isc ove ries. There a re threebasic kinds of exploratory we ll. A new field wild-cat is a we ll d rilled in sea rc h o f a new field, tha tis, in a ge olog ic structural fea ture or environm ent

Figure 6.—AAPG and API Classification of Wells

SOURCE, Lahee classification of wells, as applied by the Committee on Statistics of Drilling of the American Association of Petroleum Geologists, and the AmericanPetroleum Institute Developed by Frederic H Lahee in 1944.




that has never been proven productive. New fieldwildcats generally have the greatest associatedrisk because they are drilled based on the leastpreexisting knowledge. New pool wildcats–insearch of pools above (shallower), below (deep-er), or outside the a rea l limits of a lrea dy known

poo ls–a re genera lly less risky be c ause the fieldin which they are drilled has been proven pro-ductive. Outpost and extension tests are drilledto determine the bounds of known pools. De-velopment wells are the least risky because theirprimary function is to extract the petroleum fromthe already proven pools; they are not ex-

ploratory wells.

When an exploratory well encounters petro-leum, the quantity of proved reserves is esti-ma ted , and the c om mercial viab ility of the reser-voir evaluated. Proved reserves are determinedby analyzing actual production data or the resultsof c onc lusive forma tion tests. The p rove d areais the a rea tha t has been delinea ted by drilling

and the a djoining a rea not yet d rilled but judge das economically producible based upon available

ge ologic a nd e ngineering d ata . Bec ause of itsconservative nature, the initial estimate of provedreserves based on a field’s discovery well is gen-erally significantly smaller than the quantity of gasultimately recovered from the field. Wells drilledin subsequent years may increase the proved area

of the reservoir or lead to the discovery of addi-tional reservoirs within the field.

Each year, the sum of reserve additions attrib-uted to the three typ es of exp lorato ry wells a rerep orted by the Energy Informa tion Ad ministra -

tion (EIA) as “new field discoveries” (these arethe initial, first year estimates of a new field’sproved reserves), “extensions,” and “new reser-

voir (pool) discoveries in old fields. ” (In thisme mo randum, this last ca teg ory of reserve ad -ditions is called new pool discoveries, for brevity.)Another reporting category, “revisions,” includesthose reserves that are added or subracted be-cause of new information about old fields, for ex-am ple, an indica tion tha t the fields will be draw ndown to lower pressures because of a gas priceincrease, pressure histories during production

that deviate from the expected values, or the useof measures to increase recovery. Another cate-gory, “Net of Corrections and Adjustments,”reports reserve changes from corrections of pre-

vious arithmetic or clerical errors, adjustments topreviously reported gas production volumes, latereporting of reserve additions, and so forth. *Tab le 4 shows the c ha ng es in U.S. ga s reservesfrom 1977 to 1981 as reported by the EIA, U.S.

Crude Oil, Natural Gas and Natural Gas Liquids Reserves–Annual Report.

*EIA began its data series in 1977. The American Gas Associa-

tion and American Petroleum Institute also published reserve sta-

tistics in basically the same format (without a “corrections and ad-

justments” category) from 1966 to 1979, and in a somewhat different

format from 1947 to 1965,

Table 4.–Estimated Total U.S. Proved Reserves of Natural Gas–1977-83

Net of New reservoir Proved Netcorrections Revision Revision Extensions to discoveries New field Total reserves b change from

& adjustments increases decreases old reservoirs in old fields discoveries discoveries Production 12/31 prior year

Year (1) (2) (3) (4) (5) (6) (7) (8) (9) (lo)

Natural gas c

1976 . . — — — — — — — 213,278” — 1977 . . –2od 13,691 15,296 8,129 3,301 –

3,173 14,603 18,843 207,413 –5,8651978 . . 2,429 14,969 15,994 9,582 4,579 3,860 18,021 18,805 208,033 6201979 . . –2,264 16,410 16,629 8,950 2,566 3,188 14,704 19,257 200,997 – 7,0361980 . . 1,201 16,972 15,923 9,357 2,577 2,539 14,473 18,699 199,021 – 1,9761981 . . 1,627 16,412 13,813 10,491 2,998 3,731 17,220 18,737 201,730 2,7091982 . . 2,378 19,795 19,340 8,349 2,687 3,419 14,455 17,506 201,512 –2181983 . . 3,090 17,602 17,617 6,909 1,574 2,965 11,448 15,788 200,247 .1,265NOTE: “Old” means discovered in a prior year. “New” means discovered during the report year.aColumn 4 + Column 5 + Column 6.bPrior year Column 9 + Column 1 + Column 2 – Column 3 + Column 7 – COIU Mn 8cBillion cubic feet, 14,73 psia, 60°F.dConsists only of reported corrections.eBased on following year data only.

SOURCE’ Energy Information Administration, US. Crude Oil, Natural Gas, and Natural Gas Liquids Reserves-1983 Annual Report, DOE/EIA-0216 (83), October 1984.



Ch. 3—Natural Gas Basics q 35

HOW IS NATURAL GAS PRODUCED?

The w ay in which g as is produc ed de pe nds onthe properties of the reservoir rock and whetherthe gas occurs by itself or in association with oil.As illustrated in figure 7, hydrocarbons in the res-

ervoir rock migrate to the producing well becauseof the pressure differential between the reservoirand the well. How readily this migration occursis a function of the pressure of the reservoir andthe permeability of the reservoir rock. When thereservoir rock is of low permeability, the rockma y be artif ic ially frac tured to form p athw ays tothe well b ore. This is ac c om p lished eithe r w ith ex-plosives or by hydraulic means, pumping a fluidunder pressure into the well.

Production can continue as long as there is ade-

quate pressure in the reservoir to propel thehydroca rbo ns tow ard the p rod uc ing w ell. * If ga sis the on ly prop ellant, the reservoir pressure d e-creases as the gas is extracted and is eventuallyno longe r suffic ient to forc e the hydroca rbo nstoward the well. In a water-drive reservoir, waterdisplaces the hydrocarbons from the pores of thereservoir rock, maintaining reservoir pressure dur-ing production and improving the recoverabilityof the hyd roc a rbons. In m ost reservo irs, ga s re-c ove ry is high, generally grea ter than oil rec ov-ery. A “ typica l” rec overy value o f 80 pe rc ent isoften cited, but the basis for this value is not firm,and recovery is certainly less in many reservoirs

u rid er c urrent c ond itions. When g as oc c urs in a s-sociation with oil, it can be reinfected into thereservoir to maintain pressure for maximum oil

recovery. Gas is also reinfected when there areno p ipeline fac ilities availab le to transpo rt it tomarket.

Onc e the raw ga s is prod uce d from the we ll,i t is gathered with production from other nearbywe lls and p roc essed to remo ve na tural gas liq -uids and impurities that could cause problems

*However, the requ irement that the revenues generated by gas

sales at least match the expense of operating the well generally WIII

force an end to production before the flow of gas actually ceases.

Figure 7.—Production Mechanics

Barrier rock

-Reservoir rock

Gas and water drive mechanism

Barrier rock

-Reservoir rock

SOURCE Off Ice of Technology Assessment

in the p ipe line . The gas is the n sent by p ipe lineto local gas utilities who sell it to the end-user.In some instances, such as those involving largeindustrial users, the pipeline will sell directly to

the e nd -user and bypa ss the loc a l gas utility.



Chapter 4

The Natural Gas Resource Base



Page

Resource Base Concepts . . . . . . . . . . . . . . . 40

Approaches to Gas Resource Estimation . . 42Geologic Approaches . . . . . . . . . . . . . . . . 42

Historical Approaches . . . . . . . . . . . . . . . 43Dealing With Uncertainty . . . . . . . . . . . . 46

Comparison and Review of individual

Estima te s . . . . . . . . . . . . . . . . . . . . . . . . . . 47U.S. Ge olog ica l Survey . . . . . . . . . . . . . . 47Potential Gas Committee. . . . . . . . . . . . . 50

RAND/ Nehring . . . . . . . . . . . . . . . . . . . . . 52Hubert. . . . . . . . . . . . . . . . . . . . . . . . . . . 54

.

Reconciling the Different Estimates . . . . . . 56ls There a Consensus? . . . . . . . . . . . . . . . 57

I-low Credible Are the Methods . . . . . . . 60

Are the Physical Implications of theAssessments Plausib le? . . . . . . . . . . . . . 60

A Break With Past Trends? . . . . . . . . . . . 61Som e A lterna tive Explana tions . . . . . . . . 66Conc lusions. . . . . . . . . . . . . . . . . . . . . . . . 77

TABLES

Table No. Page

5. Why It Is Difficult to CompareResource Estimates . . . . . . . . . . . . . . . . 41

6. Alternative Estimates of UltimatelyRecoverable and Remaining Natural

Gas in the United States . . . . . . . . . . . . 497. Comp arison o f Pote ntial Gas Com mittee

Estima tes of Ultima te ly Rec overab le

Ga s Resourc es in the Low er 48 Sta tes. 51

8. Field Disc overy imp lic a tions of USGSCircular 725, Onshore Lower 48Undiscovered Petroleum Resource . . . 54

9. Hubbert’s 1980 Estimates of UltimatelyRecoverable Gas Resources in the

Lower 48 . . . . . . . . . . . . . . . . . . . . . . . . 55IO. The Op timist’ s View of Ga s

Resourc es . . . . . . . . . . . . . . . . . . . . . . . . 58

Table No. Page

11. The Pessimist’ s View of GasResourc es . . . . . . . . . . . . . . . . . . . . . . . . 58

12. Returns to New Field Wildcat Drillingin the Onshore Lowe r 48 Sta tes,

1966-81 . . . . . . . . . . . . . . . . . . . . . . . . . . 6113. Historical Record: Number of New

Gasfield Discoveries Proved After SixYears To Be of Sign ificant Size . . . . . . . 63

14. Oil and Ga s Drilling Suc c ess Ra tes . . . 6915. Potential Recoverable Gas Resources

From New Discoveries in the PermianBasin . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

16. Pote ntial Add itions to “ Old G as”Reserves From Low ered Abandonm entPressures, Well Reworking, Infill

Drilling, and Well Stimulation . . . . . . . 75

FIGURES

Figure No. Page

8. The McKelvey Box . . . . . . . . . . . . . . . . 409. Disc ove ries of Rec ove rab le Natural

Ga s in the Low er 48 Sta tes v.

Cumulative Exploratory Drilling . . . . . . 4410. Probability Distribution for

Undiscovered Recoverable GasResourc es in a Provinc e . . . . . . . . . . . . 47

11. Number of Gasfields Discovered As aPercentage of New Field Wildcats

Drilled, by Field Size Grouping . . . . . . 6412. Number of Gasfields Discovered

per Yea r, by Field Size Group ing. . . . . 6413. Gas Potential, Gas Well Completions,

and Expenditures–1978 . . . . . . . . . . . . 6714. New Field Wildcat Success Rate,

1966-81 . . . . . . . . . . . . . . . . . . . . . . . . . . 6815. Size Distribut ion of Disc overed Oil

and Gas Fields in the Lower 48 States 7016. Known and Projected Size

Distributions of Discovered Oil andGas Fields in the Lower 48 States . . . . 72



Chapter 4

The Natural Gas Resource Base

About the only thing that any estimator can saywith certainty about his (resource) estimate is that

it is w rong . Richard P. SheldonU.S. Geological Survey

The foc us of this report is on U.S. na tura l ga s

availabil ity for the next few decades—and, spe-c if ic ally, on the g as supp ly that c an be provide dby production in the Lower 48 States. Some ana-lysts have claimed that the resource base is notan important constraint to gas supply during thisperiod be c ause the U.S. Geo log ica l Survey

(USGS) estima ted resource rep resents ove r 40years of supply at current production levels,which does not count huge resources of uncon-

ventional gas (e. g., tight sands gas and methanefrom ge op ressurized aquifers) a nd po tential im-ports of liquefied natural gas (LNG) or pipelinegas from Mexico, Alaska, and Canada.

In OTA’ s op inion , the c laim tha t the resourceba se is unimpo rtant to “ midterm” (1 985-2000)

supply is arguable. Most theories of resourcedepletion imply that the “easiest” part of the re-source base—for gas, this would be the largest,most accessible fields—tends to be discoveredand exploited in the early stages of developmentand that d ec lines in d isc overy rate s and produc -

tion will occur well before the “last” resourcesare discovered and extracted. Consequently, theresource estimates of USGS and the even higherestimates of the Potential Gas Committee (PGC)d o not necessarily imply a capability to continuegas production at current levels for decades toc om e. These estimate s indica te tha t we ha vealready produced about 40 percent of the Lower48 gas resource obtainable within the currentprice te c hnolog y reg ime. The rema ining 60 per-cent will be more difficult and more expensiveto find a nd eve ntually extrac t than the a lrea dy

prod uc ed po rtion. The very pessimistic rec entestimates of M. King Hubbert 1 imply that the

United States ma y have produc ed 70 pe rc ent ofa ll the gas it sha ll ever prod uc e in the Low er 48.The l--lub ber-t estima te t hus imp lies that t he UnitedStates may enc ounter an a lmost immed iate drop-off in discoveries and reserve additions, followedshortly thereafter by sharp reductions in gas pro-duction. Even the more optimistic USGS and PGCestimates do not deny the possibility of signifi-cant reductions in supply within this century. *Therefore, an und ersta nd ing of resourc e ba se es-timate s is impo rtant to midte rm a s we ll as long-term planning rega rding na tural g as polic y.

In this sec tion, OTA ha s not a ttem pted to c rea tea new , ind ep endent assessme nt o f U.S. naturalgas resources nor to settle on any existing assess-ment as the “ be st. ” Instead we attempted toaccomplish the following four goals:

1.

2.

3.

4.

To g ive the read er an ide a o f how na tural

gas resource assessments are made.To d esc ribe the prob lem s assoc ia ted withgeneral resource assessment methods andwith particular individual assessments.To d efine the c ontinuing area s of c on-trove rsy abo ut the size a nd c ha rac teristicsof the remaining conventional gas resource

base.To c onve y OTA’ s eva lua tion of the se c on-troversies and of the c red ib ility of som e o fthe mo st w ide ly used assessme nts.

‘M. K. Hubbert, “Techniques ot’ Prediction as Applled to the Pro-

duction of Oil and Gas, ” In Oil and Ga s Supp ly Modeling, 5. 1.

Gass (cd.), National Bureau of Standards Special Publlcatlon 631,

May 1982.

* For a discussion about the production implications of the Hub-

bert, USGS, and PGC assessments, see ch. 5, “Approach Number

4–Graphing the Complete Product Ion Cycle. ”

39



40 q U .S . Natural Gas Availability: Gas Supply Through the Year 2000

RESOURCE BASE CONCEPTS

An important source of difficulty in interpretingand comparing resource base estimates is the fail-ure of the estimator to state and explain preciselythe boundaries of his estimate—his definition of

the resource base—and the failure of the clientto c omp rehend wha t a resource b ase is, or wha ta particular resource base is.

The w ell-know n Mc Kelvey Box (nam ed afte rits orig ina to r, the former direc to r of USGS) is auseful tool in explaining basic resource base con-cepts (see fig. 8). The McKelvey Box classifies re-sources according to their economic feasibilityof rec overy and the ge ologic c ertainty of theiroc c urrenc e. The o uter bound aries of the bo xde fine the tota l amo unt of the ma terial–in this

case, natural gas—remaining within the crust ofthe Ea rth. The to p third of the bo x (the p rop or-

Figure 8.—The McKelvey Box

SOURCE: Adapted from V. E. McKelvey, “Mineral Resource Estimates and Publicpolicy,” American Scientist, vol, 60, No. 1, 1972, pp. 32-40.

tions are not meant to be indicative of magnitude)represents gas that is economically producible atc urren t price s using existing tec hno log y. The m id-dle third represents gas that is expected at some

future time to be producible but is currently noteconomically producible, either because of theabsence of recovery technology or because ofeconomic conditions. The lowe r third representsgas accumulations under such difficult physicalc ond itions that they a re likely never to b e e c o-nom ica lly p rod uc ib le. obviously, our inability toac curately p rojec t future ec onomic co nditionsand future technology developments prevents usfrom knowing where to place the line betweensubec ono mic resource s and “ nonresource s. ”

The left ha lf of the b ox rep resen ts ide ntified re-sources— “resources whose location and quantityare known or are estimated from specific geologice v i d e n c e . The ec onom ic ally rec overab le p or-tion of the identified resources is called “re-serves” in the box, but t his is no t a universa llyac c ep ted de finit ion. (How ever, it is ge nerally a c -cepted that use of the term “reserves” to desig-nate the total recoverable resource is a poorusage of the term. Reserves should always referto g as tha t is in som e sense w ithin the ready in-ventory available for production. ) Proved ormeasured reserves are the most certain portionof the recoverable identified resource, gas which

has been estimated from geologic evidence sup-ported directly by engineering measurements. Anactual physical discovery by drilling is necessaryfor inc lusion within this c a teg ory. The rem a inde rof the rec ove rab le ide ntified resource is som e-wha t po orly defined b ec ause of disag reem entabout what “ ident i f ied” or “discovered” means.To USGS, for example, untapped reservoirs in dis-covered fields belong to the “discovered” re-source, 3 whereas to the PGC, they are “undis-covered. ’

——ZG. I.. Do ] ton , et al., Estimates of Undiscovered Recoverable Con-

ventional Resources of Oil and Gas in the (Jnited States, U.S. Geo-logical Survey Circular 860, 1981.31bid.4Potential Gas Agency, Poten t ia / Supp / y of Natura / Gas in the

United States (as of Dec. 31, 1980), May 1981.



Ch. 4—The Natural Gas Resource Base . 41

A crit ical feature of the components of the re-source base is that they are not static. As the pro-duc tion and disc overy proc ess c ontinues, ga sflows out of reserves and is processed, distrib-uted, and consumed, and other gas moves from“undiscovered” to “ identi f ied” as geologic

knowledge increases. Additionally, improvedtechnology and economics cause gas to movefrom the subeconomic to the economic portionof the resource base. For example, improvementsin offshore drilling technology may allow drilling

in deeper waters and more hostile conditions,opening up new territories to development.Higher ga s pric es ma y allow the d evelopm ent ofsmaller reservoirs that were previously uneco-nomic, or allow known economic reservoirs tobe de veloped more intensively and drained tolower abandonment pressures.

In the history of development of nonrenewable

resources, the process of advancing technologyand knowledge and o f changing ec onomic condi-tions has not always been smooth. Consequently,assessments of nonrenewable resources havetend ed to run in cyc les. The d isc ove ry of re-sources in areas or under geologic conditionswhere they had not been expected or the devel-

opment of new extraction and processing tech-nologies can generate higher estimates of the re-maining resource which may then taper off as thatportion of the resource base is systematicallydepleted. For most resources, analysts assessing

the remaining recoverable materials at the endof each cycle have been convinced that the mostrecent cycle upturn was the last and that resourcede pletion w as imminent. They ha ve b een provenwrong time and again. *

Recognizing this, many resource estimatorshave c onfined their assessments to only a po r-tion of the McKelvey Box, usually the top thirdand a small portion of the middle, subeconomic

*Oil has undergone such cycles of apparent depletion followed

by large new discoveries and drastic upward revisions in resource

estimates. Two other well-known materials that have undergone

similar cycles are uranium and iron ore.

third. In doing so, they explicitly accept the pos-sibil ity that changing economic and technologi-cal conditions could make their recoverable re-source estimates obsolete. Unfortunately, thestated boundaries of the assessments are seldomvery prec ise, and it is not alwa ys c lear that the

estimato rs have c onsistently follow ed their ownspe c ified rules for inc lud ing and exclud ing por-tions of the total physical resource. Furthermore,besides the ambiguity of the boundary definitions,some resource assessments have chosen differentboundaries than the “top third and a small por-tion” indicated above. Hubbert, for example,

c laims to c ap ture the ultimately recoverable re-source—the top two-thirds of the box—in his esti-mate, although he restricts the estimate to “con-ventional” gas and excludes such sources asmetha ne in c oa l sea ms. s

The d ifferenc es in ec onom ic / tec hnologica l

boundary conditions between alternative gas re-source assessments is one of several reasons whycomparisons of assessments must be handledwith c aution. Tab le 5 lists som e o f the c om mo nproblems encountered in comparing estimates.

‘Hubbert, op. cit.

Table 5.—Why It Is Difficult to CompareResource Estimates

q

q

q

q

q

q

q

Geographical areas (or geological limitations, such asdepth) included in the estimate may be different—especially offshore boundaries.Assumptions about economic conditions and the stateof technology may be different. Also, theseassumptions are often poorly defined and appear insome cases to have been applied inconsistently.Some estimates may have included someunconventional resources.Areas that are currently legally inaccessible (e.g.,wilderness areas) may or may not be included.Definitions of “undiscovered” may differ; they may ormay not include undiscovered reservoirs in knownfields.Degree of optimism about estimates (e.g., assignedprobabilities) may differ.Estimates may or may not correct for liquid contentand for impurities.

SOURCE: Office of Technology Assessment, 1983.

38-742 0 -85 _ 4



42 q U.S. /Vatural Gas Availability: Gas Supply Through the Year ,2000

APPROACHES TO GAS RESOURCE ESTIMATION*

Although the extensive literature on oil and gasresource assessme nt identifies a wide va riety ofestimation te c hniques, all of the tec hniques fallinto two basic categories. Geologic approaches

rely on informa tion and assump tions ab out t hephysical nature of the resource: volumes of sedi-mentary rock, numbers of geologic structures,presenc e o f “ source” roc ks, t ime profiles of sub-surface pressure and temperature, and the like.Historical approaches rely on the evaluation andextrapolation of past trends in gas production anddiscovery in the assumption that the size and

c harac ter of the resource ba se, rather than tran-sitory economic conditions and technological de-velopme nts, are the mo st impo rtant fa c tors c on-trolling the discovery and production cycle. If thisassumption is correct, the evidence provided by

the ma nner in which the de velopm ent cyc le ha sunfolded can be used to ascertain the nature ofthe resource base.

Geologic Approaches

Geologic approaches run the gamut from sim-ple–for exam ple, the c ollec tion o f expert geo -log ic o p inion o n the size o f the ove ra ll resourcebase–to complex procedures involving probabil-istic estimates of the geochemical and geologicfactors affecting the formation, migration, and ac-c umulation o f ga s. The m etho ds listed ma y be

used in c omb ination.

in geologic analogy, untested areas are exam-

ined for comparison with known producinga rea s. Com parisons range from simp le eva lua-tions of hydrocarbon source beds or reservoirbeds to evaluation of dozens of factors. Becausethe use o f ana log y is ba sic to a ll geologic andgeochemical understanding, this method in somesense is the basis for all the other methods.

In the Delphi approach, in its simplest form,

each member of a group of geologists evaluatesthe geologic evidence available for an area and

*This section is based largely on U.S. Geological Survey Circular

860, “Estimates of Undiscovered Recoverable Conventional Re-

sources of Oil and Gas in the United States, ” G. L. Dolton, et al.,

1981; and D. A. White and H. M. Gehman, “Methods of Estimat-

ing Oil and Gas Resources, ” AAPG Bulletin, vol. 63, No. 12, De-

cember 1979.

estima tes the a rea ’s potent ial resource s. These in-dividual estimates are then reviewed by thegroup, possibly modified, and then averaged into

a single e stima te. This ap proa c h ma y also b e used

as a tool to assist other resource estimation ap-p roa c hes, as whe n experts a re a sked to jointlyevaluate the hydrocarbon yield of an untestedarea in barrels per acre-foot as an input to a re-source assessment using a volumetric yield ap-proach (see below).

AreaI-yield and volumetric-yield approaches in-volve the estimation of the amounts of hydrocar-

bon per unit area or volume of potentially pro-duc tive rock in a region and the multiplic ationof these estimated yields by the appropriate areaor volume. The yields are g ene ra lly c alc ulated b ygeologic analogy.

Geochemical material balances, elaborations ofthe volumetric-yield approach, attempt to ac-count explicitly for the process of gas generation,migration, and entrapment. Rather than estimat-ing a simp le volumet ric yield, for exam ple, this

approach might estimate the amount of organicmatter in source beds, the fraction converted intohydrocarbons, the fraction actually able to movefrom the source bed s into reservoirs, and finally

the fraction of this amount actually trapped andconcentrated and thus available for extraction.

Field number and size approaches attempt tocount or estimate the number of prospectivefields in the area being evaluated and to estimatetheir success rate and size distribution in orderto yield an overall area resource estimate. Esti-mation methods include actual counting of struc-tural traps by using seismic surveys, extrapola-tion from historic field size distributions (a historicapproach, as discussed below), and calculationof success ratios by geologic analogy. Other levelsof a gg reg ation b esides the field a re a lso used ; playana lyses, for examp le, foc us on group s of fieldsor prospects with several common geologic char-

acteristics.Som e g eneralizations c an b e ma de ab out these

ap proac hes. The simple me thod s that use few fac -to rs to c a lcula te gas resources a ll sha re the riskthat key geologic factors, such as the temperature



Ch. 4—The Natural Gas Resource Base q 43

history of the roc ks, ma y b e left o ut. The c on-verse i s tha t the more complex methods, suchas geochemical mater ia l balances, may assumea h igher leve l o f geo log ic know ledge than cur -r e n t l y e x i s t s . A l t h o u g h t h e b r e a k d o w n o f t h e

resource assessment into several individual com-

ponents appears precise, the uncer ta inty associ -ated with each component is quite large and thepotential for error in the resource estimate is high.For example, incorporat ing factors such as pres-sure and temperature histor ies into resource est i-mat ion a l lows the est imator to account d i rect lyfor the probabi l i ty that petro leum actual ly wasf o r m e d a n d s u r v i v e d . H o w e v e r , b e c a u s e t h e

geology of most areas has changed s igni f icant lyover time, it is diff icult to trace these changes toreconst ruc t the tempera ture and pressures tha tex is ted dur ing the per iods o f hydrocarbon fo r -m a t i on , m ig ra t i on , and accum u la t i on .

The simp ler me tho d s a re m ost useful in th eearly stages of development of a basin when fewdata are available and the r-teed for expert judg-me nt a nd intuition is a t a p ea k. The o b vious d is-advantage, however, is that documentat ion of theest imation process is minimal or, in the case ofthe s imples t De lph i approach, lack ing ent i re ly .The c red ib i li t y o f t he se est im a tes , t hen , r est s

mainly on the reputation of the experts involvedin t he assessm en t o r o f t he sponso r i ng o rga -n i za t i on .

Fina lly , the g eo grap hic al ly d isag grega ted ap -

proaches, such as play analysis, are most usefulwhen considerable explorat ion data are avai lable.Many analysts think highly of these approaches,perhaps because the approaches deal in units thatmost accurately reflect the discovery process andthus allow participants in the resource assessmentto draw most readily on their experience for ge-o log i ca l ana logs .

Historical Approaches

A variety of histor ical approaches to resourcee s t i m a t i o n r e l y o n e x t r a p o l a t i o n o f h i s t o r i c a l

trends in production, reserve addit ions, and dis-covery rates as functions of t ime, number of wellsdrilled, or cumulative feet of exploratory drilling.Som e of the se a p proa c hes lac k expl ic it a ssump -tions about geology and simply search for curves

that achieve the best fit to the data. Others (some of Hubbert ’s approaches) f i rst assumeera l m ode l s o f t he p roduc t i on and d i scoprocess and then adjust the models to fit the

A var iety of formulat ions can lead to an

mate of the resource base. One s imple exa

is shown in figure 9, which plots the rate ofcovery of natural gas, in thousands of cubicper foot of exploratory wel l dr i l led, versus cum u la t i ve f oo tage d r i l l ed . A n exponen t i aother function can be fit to the historical dataextrapolated into the future. After f feet have dri l led, the area under the curve is equal to

total amount of gas discovered up to that poThe to ta l resource ba se c a n then b e e st imby measur ing the area under the curve whhas been extrapolated to the point where ac ove ra b le g a s ha s b ee n loc a ted . Th is p o iassum ed to be :

q

q

when the amount of gas discovered pero f d r i l l i ng fa l l s be low some chosen lolimit, orwhen the cumuIative exploratory footage is

judged high enough to have allowed essen-tially all p rospec tive ac rea ge in the UnitedStates to ha ve b een explored .

Although Hubbert’s estimate of gas resourceswill be reviewe d ind ividually late r, historic a l ap -p roa c hes to g as resource estimat ion as a classhave some common limitations. First, areas thatare not “ mature’ ’-that do not have a substan-tial drilling or discovery history—are not repre-

sented in the historic al da ta b ase a nd c an b e in-c lude d in the assessme nt o nly if one is willing toassume they are part of the development proc-ess of a larger area and are not rea lly inde pe nd-ent. Consequently, Alaska is typically not in-cluded in the historical approaches, and theoffshore areas are sometimes excluded as well.This limitation c an be a prob lem with geo logic

as well as geographic categories; there is someque stion, for examp le, as to w hether dee p ga s(below 15,000 ft) should be included in a “his-

torical” resource estimate,

*Area = j: (amount of gas discovered per foot drilled)

d (cumulative feet drilled).



44 q U.S. Ntural Gas Avaiability: Gas Supply Through the Year 2000

Figure 9.-DiscoverieS of Recoverable Natural Gas in the Lower 48 States v. Cumulative Exploratory Drilling

Expected

Reported

growth

SOURCE: David H Root, USGS.



Ch. 4—The Natural Gas Resource Base 45

Second, since the resource estimates are totallydependent on extrapolations of the historical rec-ord, they depend heavily on the accuracy of thisrecord. In the case of natural gas, this accuracyis p robab ly poo r. Throug h muc h of its d isc overyand production history, gas was usually a byprod-uct o f the sea rc h for and produc tion of oil andin the early years was often considered to be ofvery low va lue a t be st. Much g as wa s fla red or

otherwise wasted, production records were notkept, and gas discoveries often went unreported.

Third, a ll of the se m etho d s sha re th e c om mo nassumption of all trend extrapolations: the futurewill be a reflection of the past. However, the

“past” in the case of gas exploration and devel-opment has had interludes of radical change inthe ec onom ic und erpinnings and Gove rnmentregulation of the industry and, to a certain ex-

tent, in the technology and geologic understand-

ing driving the development process. Conse-que ntly, the historic al a pp roa c hes c onta in theimplicit assumption either that the process ofchange will continue in the same manner in thefuture or that the physical nature of the resourcebase–unchanging except for changes wroughtby d eve lop ment itself–is the ma in forc e d riving

gas development. In the long run, the physicalnature of the resource base is seen as overwhelm-ing the importance of volatile and transitory

events or forces such as Government regulationsand ga s de ma nd a nd p ric e in dete rmining theshape of the development curves. *

Fourth, it is difficult to define the economic,

technologic, geographic, and geologic bound-aries of a resource assessment based on histori-cal trends. For example, data on the development

of U.S. gas resources tracks a steady expansionof geo graphic co verag e of explorat ion and pro-duc tion, an increase o ver time in the d ep th ofwells, and a radical improvement in explorationtechnology. Did historical assessments of the U.S.gas resource done before Anadarko deep drillinginclude or exclude this deep resource? Will an

*In support of this view, it is worth mentioning that neither the

major technical advances in exploration nor the opening of new

territories since World War I I were of sufficient importance to re-

store the 011 or gas discovery rate to pre-war levels; instead, the

discovery rate continued a fairly steady downward drift for several

decades, In seeming disregard of changing conditions and tech-

nology.

assessment based on historical data account fora n ew Ove rthrust Belt type of d eve lopm ent? Tothe extent that the historical curves capture pastchange, can they account for future changes?These q uestions are essentia lly unresolved . Acommon criticism of historical approaches is thatthey do not adequately capture the effect of newtechnologies and other changes. However, thereis little agreement on what they do capture: opin-ions rang e from the full ca pture of future ec o-nomic conditions to the capture only of gas thatwould b e d isc overed and produce d unde r thesocioeconomic conditions of the last severald e c a d e s7 –in other words, from the top two-thirds of the Mc Kelvey Box to only the top third .

It is worth noting that a substantial “sur-

prise’’--e.g., the unexpected discovery of a newgeologic “horizon” —cannot be accurately pre-d ic ted b y a historic a l ap p roa c h. This is be c au se

a true surp rise w ill not ha ve a ffec ted the p revi-ous discovery and production history in any dis-c ernib le ma nner. Therefore, the historic al m eth-od will yield the same resource estimate nomatter how big the surprise turns out to be. (Al-though the geologic approach cannot predictsuch a surprise, it can incorporate its effects im-mediately for future predictions. )

Fifth, although “historical approaches” seek toextrapolate trends that are functions primarily ofthe resource base and are relatively unaffectedby transient economic effects, the available data

may be too aggregated to allow this. Generally, thedata measure processes that are made up of twoor more components, some of which are sensitiveto market conditions. For example, the findingrate of new field wildcats may be used to repre-sent the success of the discovery process. * How-ever, finding rate da ta m ea sure the c omb inedsuc c ess of a t least two qu ite different kind s of e x-p loration . The hig h-risk, high-p a yoff w ild c a ts rep -resent the search for large fields in untried areas

61bld.7R. P. Sheldon, “Estimates of Undiscovered Petroleum Resources

—A Perspective,”U.S. Geological Survey Annual Report, Fiscal Year1978.

*Discovery data generally is preferred over production data I n

a historical approach because the discovery cycle is always a few

years older than the production cycle. Extrapolation to the end of

the cycle consequently is less severe for discovery than for pro-

duction.




and the exploration of older areas ba sed on newge olog ic interpreta tions. The find ing rate of the sewildcats is a critical determinant of the long-termrep lenishmen t of p rove d reserve s. The low-risk,low-payoff wildcats represent the redrilling of old,

formerly unec ono mic a rea s, or the c lustering o f

exploratory drilling around a successful newstrike. Because drilling statistics do not separatenew field wildc ats into d ifferent risk c ateg ories,the data on low-risk, low-payoff drilling, whichis very sensitive to market conditions, dilutes anddistorts the data on the drilling activity most rele-vant to e nsuring the future o f ga s prod uction.

The prob lem o f using a single d a ta series to

measure a process that has two or more dissimilarcomponents becomes more acute as larger andlarger aggregations, geographical and otherwise,are used . Com piling the d ata for individua l prov-inces ma y be useful bec ause, for examp le, ex-

ploratory drilling on a local scale is more likelyto be either high or low risk rather than a com-bination of the two. Thus, a d isag grega ted ap -proach conceivably may be more successful than

a national one in appropriately interpreting im-plications of a changing finding rate. On the otherhand, the reduction in data points may tend toc ause da ta series for sma ll areas to be very er-ratic, and aggregation over larger areas may benecessary to detect long-term trends.

Dealing With Uncertainty

It must seem obvious from past mistakes thatpetroleum resource assessment is a risky business.For example, tracts in the offshore south Atlan-tic shelf were recently leased to industry for mil-lions of dollars (proceeds from the first two sales,lease sales 43 and 56, exceeded $400 million 8)with an industry/Government consensus thatlarge vo lumes of ec onom ic ally rec overab le o il

and gas were present, yet drilling results have thusfar been negative.9 Similarly, expected large fieldsin the Gulf of Alaska ha ve fa iled to ma teria lizeunder the drill. Conversely, drilling since 1975in the Western Overthrust Belt has revealed alarge, previously misunderstood potential for oil

and ga s. Even the c alculation of p roved reserves

‘USGS Open-File Report 82-15, South Atlantic Summary Report

2, May 1982.glbld

is uncertain and in some instances (e.g., Louisiana

a nd Texa s) ha s required extensive c orrec tions inlater years.

A major reason for the risk in resource assess-ment is that the presence of economically recov-erable concentrations of petroleum requires the

completion of an unbroken chain of events, eachof which is difficult to predict. First, adequateamounts of source rock containing organic ma-terial must be present. Second, the temperatureand pressure conditions must remain within arange capable of transforming the organic mat-ter into p etroleum. Third , geo logic c ond itionsmust be right to allow the petroleum, once gen-erated, to migrate. Fourth, permeable and porousroc ks must be in the m igrat ion p a th to serve asa reservoir. Fifth, a geologic structure must bepresent to trap the petroleum so it can ac-cumulate into commercial quantities. Not only

the availability of the required conditions but alsotheir timing are c rit ic al. The p resenc e o f an ad e-quate trap, detectable with seismic or othersearch techniques, does not guarantee that thetrap wa s present a t the time o f pe troleum m igra-tion; if it wa s not, or if the trap wa s brea c hed a tsome time after the petroleum entered the reser-voir, the oil or gas would have escaped andwould probably have reached the surface anddissipated.

Some estimators either (apparently) ignore un-c ertainty or ac knowled ge it only by expressing

their results as an undefined or vaguely definedrange (e.g., “optimistic/pessimistic”). Uncertaintycan be dealt with explicitly and quantitatively inresource estimations, however. Resource esti-ma tes, or the ind ividua l fac tors used in e stimat-ing resources (e.g., volume of sedimentary rock,hydrocarbon yield factor), can be expressed asprob ability func tions instea d of p oint estimate sor ranges. For example, figure 10 illustrates a hy-pothetical probability function for the undiscov-ered recoverable gas resources of a single prov-inc e. The c urve show s the p rob a bility that the reare more than Q undiscovered resources in the

province. * “Probabilistic estimates” such as these

— —*The probability is not 100 percent at Q = O because there is a

finite probability that the province does not have “more than O

resources,“ in a totally unexplored province, this probability of zero

recoverable resources may be quite large.




Figure 10.—Probability Distribution for UndiscoveredRecoverable Gas Resources in a Province

Resources Q, TCF of undiscovered recoverable gas

NOTE: “More than” cumulative distribution function.

SOURCE: David H. Root, USGS.

c anno t be direc tly ad de d (or, in the c ase o f esti-

mates for volumes and yield factors, multiplied)to form aggregate resource estimates, such as an

estimate of total U.S. gas resources. Instead, theyare added statistically; one commonly used tech-nique is called Monte Carlo simulation (see boxC). *——.——“In Monte Carlo simulation, a value is selected at random from

each of the separate probability functions that are the components

of the resource estimate (e. g., for a nationwide assessment, the com-

ponents are the individual province assessments; for a volumetric

COMPARISON AND REVIEW

Although many readers may be aware only ofthe wo rk of USGS and perhaps tha t o f M. KingHub bert, a ssessments of t he U.S. na tura l ga s re-

source base are quite numerous and use a wideva riety of a pp roac hes. Tab le 6 lists som e o f themore recent estimates of the “ult imately recov-erable resource’ ’—the total amount of gas thatwill be prod uc ed . The ta ble a lso show s estima tesof the rec overab le resource rema ining as well asthe resources not yet added to proved reserves.The w ide rang e of m ea n estimates for the rem ain-

ing resources in the Lower 48 States—244 to 916trillion c ub ic fe et (TCF)—implies, in turn, a w iderang e in the o utlook for future ga s produc tion,especially in the longer term.

Although p rob ab ilistic m etho ds a re useful fordisplaying some of the uncertainties associatedwith resource estimation, the lang uage used todescribe the results of these methods is often mis-understood by a lay audience. It is critical to re-member that the accuracy of probabilistic esti-

mates is limited by the extent to which theestimators’ model of the physical universe is acorrect one. In estimates such as those of USGS,the “ 95th pe rc entile” estimate should not be in-terpreted as meaning that there actually is a 95percent probability that the resource base is largertha n th is estima te . It should instead be interpretedto mean only that the assessors, with whateverlimitations their geologic “mindsets” and theirlimited data may impose on them, believe thatthe re is suc h a 95 perce nt p robab ility. This d if-ferenc e m ay seem sub tle, and it c ertainly is notkept secret by the estimators, but it is nevertheless

important.

resource assessment, the components are the volume of sediment-

ary rock and the hydrocarbon yield factor). These values are then

combined arithmetically to form a single point estimate of the re-

source base (for the nationwide assessment, the values from each

province are added; for the volumetric, the values selected for vol-

ume and yield are multiplied). This procedure is repeated many

times, each time producing a new point estimate, until a probabil-ity function for the resource base IS formed.

OF INDIVIDUAL ESTIMATESMany available resource assessments are poorly

doc umented and ca nnot be evaluated. OTA hasreviewed some of the more widely known esti-mates, however, including those of USGS, PGC,the RAND Corp., and M. King Hubbert.

U.S. Geological Survey

Recent estimates of undiscovered gas resourcesby USGS, as presented in 1975 in “ Circu lar725” 10 and more recently in 1981 in “Circular

860,”

11

are probably the most widely used gas10B, M, Miller, e t al., Geo/ogica/ ~SfirTrates O f ufJdi5COVered

Rec ove rable O il and G as Resources in the United Sta tes, ” USGS

Circular 725, 1975.I I Dolton, et a!., Op. cit.



resource e stima tes. The m ost rec ent e stima te usesa Delphi-type approach whereby teams of geolo-g ists a rrive d irec tly a t resource estima tes for in-

dividual petroleum provinces through a subjec-tive assessment of the available geological dataand the results of a variety of estimation ap-proaches (including volumetric, play analysis, andother geologic methods as well as finding-rateana lyses and othe r historic a l metho ds).

The estimates are probabilistic, that is, each ispresented as a curve that shows the probabilitythat the actual resource base is larger than anypartic ular va lue (see fig. 10). Thus, the 95th pe r-centile estimate reflects the USGS assessment thatthere is a 95 percent probability that the actualresource base is at least this large. Because onlythose resources that are virtually certain to existare included, this estimate would be consideredthe pessimistic extreme of the range of estimates.The individu al province estima tes are a d de d sta -tistically, using a Monte Carlo technique, to

ac hieve a national estima te. As de sc ribe d pre-viously (box A), the “high-low” range describedby the 5th and 95th percentiles is narrower than

would be the case if the interdependence of indi-vidual province estimates could be taken into ac-

count. However, the potential problem was de-sc ribed as minor by the expe rts OTA ta lked with,la rge ly b ec ause of USGS’ selec tion o f p rovinc eboundaries.

The USGS assessment is unusual in that indi-vidual probabilistic estimates are available for

each of 137 provinces, providing a very fine levelof detail. Also, detailed information files on indi-vidual provinces are open to the public at USGS’Denver facility. As with most geologic estimates,the USGS estima te is no t me a nt to includ e a ll re-sources that may be recoverable at any time, butis instead limited to t he resources tha t “ will be

recoverable under conditions represented by acontinuation of price-cost relationships and tech-nological trends that prevailed at the time of




Table 6.-Alternative Estimates of Ultimately Recoverable and Remaining Natural Gas in the United States (TCF)

Remaining resources

Ultimately recoverable resourcesnot yet-added to

Publication Remaining resources proved reserves,Estimator date Lower 48 Total U.S. Lower 48, 1983

aLower 48, 1983

b

Mobil . . . . . . . . . . . . . . 1975 — 1,076-1,241-1,456 —

Garrett . . . . . . . . . . . . .

—

1975 — 1,313 —

Wiorkowsky . . . . . . . . .

—

1975 1,221 -1,289-1 ,357a — 595-663-731 421-489-557Bromberg/Hartigan . . . 1975 966

C — 340 166Exxon Attainable . . . . 1976 — 917-1,112-1,577 — —

Shell (1) . . . . . . . . . . . . 1978 946 910-1,075-1,260 320 146Shell (2) . . . . . . . . . . . . 1984 1,150d

1,265d 525d350d

IGT . . . . . . . . . . . . . . . . 1980 — 1,288-1,798 — —

PGC . . . . . . . . . . . . . . . 1983 1,542 1,711 916 742Hubbert (1). . . . . . . . . . 1980 870 — 244 70Hubbert (2). . . . . . . . . . 1980 989’ — 363 189RAND . . . . . . . . . . . . . . 1981 902 989 283 109USGS . . . . . . . . . . . . . . 1981 1,400 1,422-1,541-1,686 774 600

aApproximate cumulative Lower 48 production through 1962 was 631 TCF, of which about 5 TCF is in underground storage. “Remaining resource” is “Lower 48” (ultimatelyrecoverable) column value minus 631 TCF plus 5 TCF.

Lower 48 proved reserves assumed to be 169 TCF on Dec. 31, 1982 (excluding underground storage).cOriginal estimate for onshore gas only. Total arrived at by adding USGS (mean) estimate for ultimately recoverable offshore gas in Lower 48 (235 TCF)‘Estimate includes 52 TCF for additional resources obtainable with old gas decontrol.‘Based on an analysis of finding rates by David Root, USGS.SOURCE: Mobil—J. D. Moody and R, E. Geiger, “Petroleum Resources; How Much Oil and Where,” Technology Review, March/April 1975. Verbal comments by John

Moody at a FPC presentation, Apr. 14, 1975.Garrett-R. W. Garrett, “Average of Some Estimates by Major Oil Companies and Others, 1975, ” oral presentation at Executive Conference of the AmericanGas Association, June 9-11, 1975, cited in Potential Gas Committee, A Comparison of Estimates of Ultimately Recoverable Quantities of Natural Gas In theUnited States, Gas Resource Studies No. 1, Potential Gas Agency, April 1977.Wiorkowski-J J Wiorkowski, Estimation of Oil and Natural Gas Reserves Usirn Historical Data Series: A Critical Review,unpublished manuscript, 1975,cited in J. J Wiorkowski, ‘iEstimating Volumes of Remaining Fossil Fuel Resources: A Critical Review, “ in J. Am, Stat. Assoc., vol. 76, No. 875, September 1961Bromberg/Hartigan-L. Bromberg and J. A Hartigan, Report to the Federal Energy Administration, unpublished manuscript, 1975, cited in Wiorkowski (1981),noted above.Exxon—Exxon Co , U. S. A., Exploration Department, “U.S. Oil and Gas Potential, ” March 1976.Oil and Gas Journal, “Exxon Says U.S. Still Has Vast Potential, ”Mar. 22, 1976,Shell (1)—C. L. Blackburn, Shell Oil Co., “Long-Range Potential of Domestic Oil and Gas,” presented at NAPIA/PIRA Fall Conference, Boca Raton, Fla., Oct.19, 1978, Oil and Gas Journal, “Shell: Alaska Holds 58% of Future U.S. Oil Finds, ” Nov. 20, 1978.Shell (2)—R. A. Rozendal, Convention/ U.S. Oil and Gas Remaining To Be Discovered: Estimates and Methodology Used by Shell Oil Company, draft, Aug.1, 1964, Shell Oil Co.IGT—J. D. Parent A Survey of United Statesand Total World Production, Proved Reserves, and Remaining Recoverable Resources of Fossil Fuels and Uranium,Institute of Gas Technology, Chicago, August 1960, cited in American Gas Association, “Energy Analysis A Comparison of U S. and World Remaining Gasand Oil Resources,” Aug,-7, 1981.PGC– Potential Gas Agency,Potential Supply of Natural Gas in the United States (as of Dec. 31, 1982),Colorado School of Mines, June 1983.Hubbert (1) (2)—M, K. Hubbert, “Techniques of Prediction as Applied to the Production of Oil and Gas,“ in Oil and Gas Supply Modeling, S. I.Gass (cd,),National Bureau of Standards Special Publication 631, May 1982.RAND—R. Nehring with E. R. Van Driest 11, TheDiscovery of Significant Oil and Gas Fields in the United States, R.2654/l -USGS/DOE, RAND Corp ,January 1981,

USGS—G L Dolton, et al., Estimates of Undiscovered Recoverable Convention/ Resources of Oil and Gas in the United States, U.S. Geological Survey Cir-cular 660, 1981

assessme nt (1980).” 12 Co nseq uen tly, resourcesthat are currently in fields that are too small,

under too much water, under geologic conditionsthat are to o d iffic ult, or are otherwise no t ec o-nomica lly rec overab le a re no t reflec ted in thecurrent estimates but could be expected to en-ter the recove rab le resource b ase in the futureif gas prices rise and technology improves signif-icantly.

In contrast to the approach for estimating re-sources in undiscovered fields, USGS calculatedthe rema ining resource s in undisc ove red po olsin known fields and expansion of the proved

‘z Ibid,

areas of knownelation fromgrowth. 13 Fieldgas, and USGS

pools* by using a simple extrap-historical records of gasfieldgrow th is a significa nt source ofc a lc ulated the resource s in this

c a teg ory to b e a bo ut 172 TCF, or ove r one-fifthof the remaining gas resources. Unfortunately,the USGS approa c h to assessing this source isproblematical because the historical growth ratesof known fields have tended to be extremely vari-able, and the characteristics of fields discovered

recently, and calculated by this method to yield

*These resources are called ‘‘inferred reserves’ i n the USGS

assessment and are equivalent to the “probable potential resources”

in the PGC assessment.

131bid, app. F,




the most growth, are quite different from thefields tha t supp lied the h istoric a l da ta . In OTA’ sop inion, there is a signific ant p ote ntial for errorin this approach.

in USGS’ 1975 resource estimate, the econom-

ic boundary of recoverable resources also provedto be a problem; a survey of the assessment teamrevealed considerable differences between theirvarious interpretations of the meaning of theb ound ary de finition .14 Althoug h OTA und ertookno formal survey for the 1981 assessment, infor-mal talks with analysts close to the assessmentprocess lead OTA to believe this problem still ex-ists. For example, several analysts believe that part

of the offshore resource in the USGS assessmentis far too expensive to be developed unless gasprices escalate substantially. If this is correct,

these resources are subeconomic, according to

USGS’s definition , and should no t be inc ludedin the estimate of rec ove rab le resources.

Another potential problem area in the assess-ment is the b ounda ry be tween “ co nventional”a nd “ unc onven tiona l” resou rce s. The USGS esti-ma te is of “ undisc overed rec overab le conven-tional resources (our emphasis)” and excludes“gas in low permeability (’tight’) reservoirs” andothe r so-c a lled unc onve ntiona l resource s.15 Theprecise meaning of the exclusion is unclear, how-ever. In moving towards lower and lower per-meabilities, there is no general consensus aboutwhere “conventional but low permeability reser-voirs” end and “ unconventional ‘t ight’ reser-

voirs” begin, and USGS has not defined a thresh-old value of p ermeab ility to sep arate the two.

Circ ular 860 doe s imp ly, how eve r, tha t som eundiscovered gas in low-permeability reservoirswas excluded from the estimated conventionalresource b ase e ven thoug h the g as c ould c ur-rently be defined as economically recoverable.Consequently, all else being equal, the USGS esti-mate should be expected to be smaller than esti-ma tes that include all economical ly recoverablegas resources.

It also is commonly believed that USGS’ Delphitec hnique, d esc ribed by USGS as relying on re-

Idpersonal communication with John Schanz, Congressional Re-

search Service.~SDolton, et al., Op. cit.

views of the results of a va riety o f app roa c hes,relies p rima rily on the results of vo lumet ric ana l-

ysis. This relianc e o n the vo lumet ric approac h isp rob ab ly d ue to da ta limita tions. The USGS d atabase, although substantial, is generally limited topubl ic data.16 Volumetric analysis has often been

associated with relatively optimistic resourceassessments.

Potential Gas Committee

The e stima tes of “ pot en tial” g a s resou rce s–re-coverable resources that have not been produced

or proved—by the PGC represent the gas indus-try co unte rpo int to t he USGS estima te . *

PGC’s mo st rec ent estima te o f the tota l U.S.pot ent ial resou rce —876 TCF for the end of 198217

–-represents a decrease from the year-end 1980estimate. 18 Because this decrease is balanced by

add itions to p rove d reserves during the p eriod ,the old and new estimates are similar in their es-timates of total ultimately recoverable resources.

The PGC estima tion p roc ed ure is gene rallystructured like a volumetric analysis in that the

PGC analysts separately estimate the volume ofpotential gas-bearing reservoir rock and a yieldfac tor (am ount of ga s pe r volume of roc k) andmultiply the two to arrive at an initial resourceestima te . The a na lysis c om b ines a sp ec ts of o the rgeologic approaches, however. It is also strength-

ened by the sep arate e stima tion of g as po tential

for 11 distinct geographical areas within theLow er 48 Sta tes, for three d istinc t c a tegories ofresource within the areas according to their state

of development, * for offshore and onshore re-———

16G Dojton, USGS, presentation at RAND workshop on esti-

mating U.S. natural gas resources, Washington, DC, Mar. 1-3, 1982.

* PGC is composed of members and observers from gas producers,pipelines, and distribution companies and observers from the Amer-

ican Gas Association, Department of Energy, Gas Research insti-

tute, and other public and private organizations. The actual esti-

mating workgroups consist mainly of industry employees and

consultants, but State geological surveys are well represented, and

some of the groups include personnel from Federal agencies and

from universities.I 7News release, potential Gas Agency, Feb. 261 1983.I Spotential Gas Agency, Potential Supply of Natural Gas in the

United States (as of Dec. 31, 1980), May 1981.*The categories are “probable,” “possible,” and “speculative”

resources. Probable gas results from the growth of known fields,

possible gas is associated with the projection of plays or trends of

a producing formation into a less well-explored area of the same

gec~logic province, and speculative gas is from formations or prov-

inces that have not yet proven to be productive.




sources, for resources ab ove and be low a de pthof 15,000 ft in the onshore portion, and for re-sources ab ove a nd b elow wa ter dep ths of 200meters to a maximum of 1,000 meters offshore.The e stimate s “ inc lud e o nly the na tural ga s re-

source which can be discovered and produced

using current or foreseeable technology andunder the condition that the price/cost ratio willbe favorable.” 19 These c ond itions a re similar tothose adopted by USGS, but what constitutes a“ favo rab le pric e/ c ost ra tio” rem ains unc lea r. Thelarge propo rtion o f de ep resources inc orporatedin the estimate may imply, however, that PGChas included resources that will require pricesabove present market clearing levels.**

The PGC volumet ric estimation proc ed ure isconsiderably more sophisticated than early tech-niques that were based on total volumes of sedi-

mentary rock, In the PGC analysis, the volumesof p ote ntial ga s-bea ring reservoir roc k are e sti-

ma ted by a dd ing up estima tes of individua l trap sand trap sizes where sufficient data is available.According to PGC’s methodology description,20

techniques such as play analysis and field num-be r and size a pp roa c hes are used to c onstruc tan areawide volume estimate based on a varietyof existing geological data. Yield factors (gas vol-

umes/rock volumes) are then calculated by select-ing appropriate analogs from producing areas and

adjusting the yields to account for geochemicalfactors such as the thermal history of the source

rocks. Finally, the a na lysts a re a sked to multip lythe (volume) X (yield) estimates by their assess-ments of the probabilities that traps actually ex-ist and that a n ac tual acc umulation of ga s hasoccurred.

The a na lysts a lso a re asked to sepa rate ly esti-mate “optimistic,” “most likely, ” and “pessi-mistic” volume s of g a s in a m an ner simila r to tha tof the USGS. In contrast to USGS, however, PGCpub lishes only the “ mo st likely” estima tes. Theothe r estimate s a re a pparently used for reviewpurposes only.

———. ——.—.191 bid.

* *Orr the other hand, the actual price requirements for produc-

ing deep gas under tree market conditions are uncertain, and it

is possible that much of PGC’S deep potential is producible at prices

not tar removed from today’s.

‘(’1 bid.

Because PGC publishes only the results of itsanalyses and does not release any internal details

of the resource calculations (except for generalmethodology descriptions), and because it is

essentially a gas industry organization, the cred-ibility of PGC’s resource estimates may be ques-

tioned . In OTA’ s op inion , how ever, the PGC esti-mates should be taken as a serious effort atresource assessme nt b y a na lysts with e xce llen t

ac c ess to e xplora tion d ata . The e stima ting w ork-groups, although composed mostly of industryemp loyees, have a suffic ient num be r of other

participants-and a sufficient divergence of incen-tives within different segments of the industry—to prevent any attempts to subvert the assessmentprocess significantly. Also, the long-term profes-sional history of the organization (since 1966) andthe o versight of the C olorad o Sc hoo l of Minesare substantial arguments for accepting the PGC

estimates as honest reflections of the professional judg ment of the organization.

An advantage of the PGC estimates is that thebasic methodology has been applied, with evolu-tiona ry chang es, for 16 yea rs. Ta b le 7 shows theeight estimates of ultimately recoverable gas re-sourc es in the Low er 48 Sta tes produc ed b y PGCsince 1966. The c onsistenc y o f the se e stima tesis high. In fact, given the advances in technol-ogy and the major additions to the known bound-

aries of conventional gas supply that haveoccurred in the past 16 years, * the mildness of

*For example, the addition of the Western Overthrust Belt duelargely to advances in seismic technology, and the addition of large

amounts of gas from low permeable formations due to advances

in fracturing.

Table 7.—Comparison of Potential Gas CommitteeEstimates of Ultimately Recoverable Gas Resources

in the Lower 48 States

Ultimately recoverableEstimate as of year end resources (in TCF)

1966 . . . . . . . . . . . . . . . . 1,2831968 . . . . . . . . . . . . . . . . 1,4261970 . . . . . . . . . . . . . . . . 1,4981972 . . . . . . . . . . . . . . . . 1,4461976 . . . . . . . . . . . . . . . . 1,396-1,421-1,446

1978 . . . . . . . . . . . . . . . . 1,5501980 . . . . . . . . . . . . . . . . 1,5021982 . . . . . . . . . . . . . . . . 1,542aaApproximate—A portion of the difference between the 1980 and 1982 estimates

is due to discrepancies between the proved reserve values computed by AGA(used for the 1960 calculation) and the EIA (used for the 1982 calculation)

SOURCE Potential Gas Agency, Potential Supply of Natural Gas in the United States (as of Dec. 31, 1980), May 1981, and Potential Gas Agency, newsrelease, Feb. 26, 1983




the upward trend in the estimates over this timeperiod implies a movement toward more con-servative estimates. This conservatism is particu-larly interesting in light of PGC’s resource esti-ma tes be ing a mong the m ost op timistic of themajor assessments.

In its 1982 assessment, PGC attempted to iso-late that p ortion o f the estima ted po tential re-source that occurs in tight formations—tight sandswith permeability levels less than 0.1 millidarcy(conforming to the Federal Energy RegulatoryCommission definition for gas eligible for incen-tive pricing) and Devonian shales. A series ofareaw ide estima tes were produc ed for de pthsab ove and be low 15,000 ft. The “ tight” po rtion

of the U.S. potential gas resource was estimatedto b e a bo ut 20 perce nt of the tota l, or 172 TCF,

This estima te is high ly signific ant fo r tw o rea -

sons. First, it d em onstrate s grap hica lly the long -term growth in the “ult imately recoverable” gasresource base and offers some support to the op-

timistic view that advancing technology can over-

c om e a t least some of the effec ts of resourcede pletion, Sec ond , to the e xtent tha t othe r re-source a ssessors ma y ha ve exc luded tight g asfrom their estimates, it may bring the PGC esti-

mate closer to the “mean” of gas resource esti-mates in table 6. Unfortunately, the definitionsof the boundary conditions of most of the assess-ments in table 6 are not sufficiently clear to ascer-tain whether tight g as that is recoverable under

the PGC boundary conditions were excluded orincluded. A p ossib le exce p tion, how eve r, is the

USGS assessment, whose stated boundary condi-tions ap pe ar to b e m ore restric tive than PGC’ s.It is probable that some of the tight gas includedin the PGC estimate was not included in theUSGS estim a te .

RAND/Nehring

Richard Nehring of the RAND Corp. has pro-

duced an assessment of conventional U.S. oil andgas resources by a method that stresses an evalua-tion of the discovery of significant fields. 21 Theassessment incorporates a variety of approaches:

ZI R. Nehri ng With E. R. Van Driest 11, The ~;SCOVerY Of S;g~;f;-

carrt O i/ and Ga s Fiekfs in th e United States, RAND Corp. ReportR-2654/l-USGS/DOE, January 1981.

1.

2.

3.

4.

To e stima te the g rowth of reserves in knownfields, a combination of methods were used,including extrapolating by historical fieldgrowth factors and by more analytical ap-proaches that used available geologic infor-ma tion and known prod uction prac tic es.

To e stimate the a mo unt of resource rem a in-ing to be discovered in known producing

plays, an approach based on extrapolatinghistoric a l trends wa s used . The key to thisapproach was the establishment of a database containing production and reservevalues, the year of discovery, discoverymethod, trap type, depth, and other data forvirtually every petroleum field discovered inthe United Sta tes by 1975 large r tha n c lassC (10 million to 25 million barrels-of-oil-equivalent). Despite the emphasis on the his-torical record, however, the approach also

incorporates geologic methods based onplay analysis.play analysis was used to estimate the re-

sources in new plays in mature regions.Depending on the availabil ity of data, a va-riety of approaches were used to estimateresources in the frontier (ranging from vol-ume tric an alysis to field num be r and size a p -proaches).

The e stima tes for new p lays in ma ture reg ionsand frontier areas were “ risked ” (i.e., the p rob -ability that there are no recoverable resources in

the p lay is taken into a c c ount ), and the a ssess-ments of undiscovered resources were expressed

as probability distributions in a manner essentiallyidentica l to tha t used by USGS.

The RAND assessment has been criticized be-cause of its alleged failure to define the processby which its massive data base is translated intoresou rce base c onc lusions. In OTA’ s op inion , thedescription of the methodology that appears inthe RAND report is indeed brief and generalizedand gives no specific examples of the assessmentprocess. However, this failure is endemic to re-source assessments as a class. Even the PGC as-sessment, which describes its analytical processin some detail, publishes no backup data and pro-

vides only the sketchiest details of the geologicreasoning behind its regional results. in contrast,the RAND assessment explicitly defines the his-



Ch. 4—The Na tura l Ga s Resource Base q 5 3

torical and geologic reasons for its regional assess-ments and identifies—and argues against—oppos-ing views. This a pp roa c h a llow s a t least a p artialevaluation of the assessment, whereas most

assessments can be evaluated only to the extentof either accepting or rejecting the final estimates.

At the core of Nehring’s argument for his quitepessimistic estimate is the thesis that the geologicpossibilities for finding substantial new oil and gasresources in the United States have been largelyexhausted. Nehring identifies four major hypoth-eses about where significant amounts of oil andga s ma y yet b e fo und –in fields below 15,000 ft

in depth (for natural gas only); in subtle, diffi-

cult-to-detect stratigraphic traps; in small fields;and in frontier areas, including the Eastern a ndWestern Overthrust Belts–and argues againsthigh optimism in each, with the possible excep-tion o f the frontier a rea s. The four hypo theses and

Nehring’s countering arguments are summarizedin box D. A more detailed discussion of these hy-potheses is presented later in this chapter.

A second facet to this argument is that this ex-haustion of geologic possibilities is reflected inthe recent (disappointing) history of exploratorydrilling. Nehring argues that optimistic assess-

Box D.—Rand Assessment’s Arguments Against a Large Undiscovered Oil and Gas Resource Base

Deep Discoveries

. Major argument: Deep sediments are relatively unexplored. The few exploratory wells that have beendrilled have been highly successful.

c RAND rebuttal: Physical and chemical conditions at these depths can be poor for methane stability.Reservoir porosity is often lacking. The area with deep sediments is a small fraction of total prospec-tive sedimentary area. Most of the potentially productive structures in several basins have alreadybeen tested.

Stratigraphic Traps. Major argument: Exploration has focused on structural traps, leavingsignificant opportunities in subtle

stratigraphic traps.q RAND rebuttal: Actually, considerable attention has been paid to stratigraphic traps in the Anadarko,

Permian, and other basins. Aside from the stable interior provinces, multiple stratigraphic traps areunlikely. Because stratigraphically trapped reservoirs tend to be thin, large fields would cover largeareas and would likely have been discovered. Large traps would be vulnerable to breaching and othercauses of petroleum loss.

Very Small Fieldsq Major argument: Because small gas fields were previously subeconomic, their discovery went unre-

ported. Many more small fields exist than indicated by historical experience, and they forma sizablepart of the recoverable gas resource.

q RAND rebuttal: Future reliance on small fields is based on assumption only; there is neither historicalnor geologic argument to back it up. Also, because giant and large fields are two-to-four orders ofmagnitude larger than fields small enough to have been ignored in the past, there would have tobe many tens of thousands of such fields to make any significant difference.

New Frontiersq Major argument: Areas such as Alaska, the offshore Lower 48 States, and the Overthrust Belts have

not been extensively explored and offer the potential for many significant discoveries.. RAND rebuttal: Yes, but the small number of exploratory wells drilledin the Gulf of Alaska, the Outer

Banks of California, the eastern Gulf of Mexico, the Southeast Georgia Embayment, and BaltimoreCanyon are sufficient to severely dampen optimism for these areas. Some very promising areas doremain, however, including the deeper Gulf of Mexico, offshore Ventura Basin, and others.

SOURCE: Office of Technology Assessment, based on R. Nehring, The Discoveryof Significant Oil andGas Fields in the United States, R-2654/1-USGS/OOE, RAND Corp.,January 1981.



54 U.S. Natural Gas Availability: Gas Supply Through the Year 2000

ments simp ly do not b ea r up unde r the w eightof the q uestion, “Is it likely that we will find as

many large fields as this assessment implies mustbe there?” For example, table 8 presents a pro-posed field size d istribution tha t w ould yield an

und isc ove red petroleum (oil plus ga s) resource

eq ua l to tha t p red icted in the 1975 USGS (Cir-c ular 725) onsho re assessment. This d istribut ionwould a lso b e a pp roxima tely equivalent to themore rec ent 1981 (C irc ula r 860) USGS assess-

ment, although the more recent assessment is

slightly more optimistic. in the table, the pro-posed distribution is compared to actual fieldd isc ove ry sta tistic s fo r 1971 through 1978. Thelast column shows how long it would take to findthe necessary number of fields of each size cat-egory if the annual discovery rates of 1971through 1978 continued for the life of the re-source. In Nehring’s opinion, the number of large

fields that would have to be discovered to fulfillthe USGS assessment is too large to be credible.The long “ times of d isc ove ry” in the ta b le a p-pear to reinforce this opinion. Unfortunately,none of the reviewed assessments defined atimeframe for complete discovery of the resourcebase, and an interpretation of the compatibil ityof a particular resource base/discovery rate com-b ina tion is a nything b ut straightforwa rd . Also, thecessation of the American Gas Association’s(AGA) reserve data (particularly reserve additionsfrom new field wild c ats) in 1979 preve nts an ea sycheck on whether post-1978 new field discov-

eries are ahead of discoveries during 1971-78;if they were, an argument could be mad e that

the times in ta ble 8 we re m islea d ingly long b e-

cause the assumed discovery rate was too low.On the other hand, the assumption in table 8 ofa constant annual discovery rate for new gasfieldsover a 50- to 100-year period appears optimistic,even if the assumed rate is a bit low at the be-

g inning o f the period . This is bec ause d isc overyrates per foot drilled appear likely to decline dur-ing this period, and a constant annual discoveryrate thus implies an ever-increasing rate of newfield wildcat drilling in an increasingly hostile and

expensive environment.

One portion of the RAND assessment that nowseems particularly suspect is the median estimatefor field growth. The estimate (67 TCF) wa s onlyab out one-third o f the field g row th estimates ofUSGS and PGC, a seemingly surprising differenceconsidering the substantial amount of geologic

knowledge available. * Recent large reserve ad-

ditions from field growth make it clear that this

estimate was too low. * *

Hubbert

As noted earlier, M. King Hubbert is one of a

considerable number of analysts who have useda historical approach—fitting curves to past trendsin p rod uc tion, reserve g row th, d isc ove ries, and

*However, the recent controversy over the magnitude of addi-

tional gas that might be obtainable from old gas decontrol demon-

strates that the availability of extensive geologic knowledge does

not guarantee agreement over resources present.

* *Nehring acknowledged this problem to OTA in a recent tele-phone conversation.

Table 8.—Field Discovery Implications of USGS Circular 725, Onshore Lower 48 Undiscovered Petroleum Resource

Implied time to findPotential field size

Actual field discoveriesUSGS undiscovered resource,

distribution: USGS — constant annual discoveryField sizea Circular 725 1971-75 1976-78 rate at 1971-78 average (years)

AAAA (>500/>3,000) . . . . . . . . . . . . . . . . 11 0 1 88AAA (200-500/1,200-3,000) . . . . . . . . . . . . 44 0 0 Large but indeterminate

AA (100-200/600-1,200). . . . . . . . . . . . . . 94 7 1 94A (50-100/300-600) . . . . . . . . . . . . . . . . 199 7 3 159B (25-50/150-300) . . . . . . . . . . . . . . . . . 375 15 8 130

C (10-25/60-150) . . . . . . . . . . . . . . . . . . 977 44 22 118D (1-10/6-60) . . . . . . . . . . . . . . . . . . . . . 6,000 455b — 66E (< 1/<6) . . . . . . . . . . . . . . . . . . . . . . . 70,000 3,041

b115

aValues in parenthesis are sizerange in millions of barrels of oil equivalent (mmboe)/billions of cubic feet of gas (BCF).1972-76 Committee on Statistics of Drilling of the American Association of petroleum Geologists.

SOURCE: Office of Technology Assessment, based on R. Nehring, The Discovery of Significant Oil and Gas Fields in the United States, RAND Corp. reportR-2654/l-USGS/DOE, January 1961. Also, personal communication, Richard Nehring.




so forth-to petroleum resource assessment.However, Hubbert’s estimates must be accordedspecial attention. In 1962 Hubbert predicted thatU.S. oil p rod uc tion wo uld p ea k in 1969 and de-c line the rea fter. He the n held his g round in theface of substantial criticism until the peak actually

did o c c ur, only a year late r than he said it would.From that time, his assessments of petroleumtrends and resources have received considerablymore attention and respect.

Hubbert’s mo st rec ent estimate o f the size ofthe natural gas resource base was made in 1980.22He estimates the ultimate cumulative productionof c onventional natural ga s (Q~) for the Lowe r48 Sta tes to be approxima te ly 870 TCF. This isa rem arkab ly low estima te g iven c umulative p ro-duc tion to da te of a bo ut 631 TCF and proved re-serves of about 167 TCF; * * if correct, it leaves only

a bo ut 70 TCF rem a ining to b e a dd ed to reservesfrom the g rowth o f known fields (ca lc ulate d by

USGS to be 172 TCF) and new field d isc overies.In other words, Hubbert’s assessment implies thatthe precipitous declines of the early 1970s in

Lower 48 proved reserves will resume againalmost immediately, with subsequent drastic con-sequences for production rates within only a fewyears.

in his 1980 assessment, Hubbert obtained fiveseparate estimates, using basically three ap-p roa c hes (table 9). In his first app roa c h he d e-

*2Hubbert, op. cit.

* *As of the beginning of 1983.

Table 9.—Hubbert’s 1980 Estimates of UltimatelyRecoverable Gas Resources in the Lower 48

Method of estimation Q m (TCF)

1. Extrapolating the plot of productionrate as a function of cumulative production . 810

2. Estimating the approach of cumulativediscoveries to Qm as time approaches 00 . . 871

3. Finding the equation of cumulativediscoveries versus time . . . . . . . . . . . . . . . . . . 840

4. Using oil resource estimate and

assuming stable gas/oil discovery ratio. . . . . 876-8965. Fitting and extrapolating thecurve of discoveries per 108 feetof exploratory drilling . . . . . . . . . . . . . . . . . . . . 989

SOURCE: Office of Technology Assessment, based on M. K, Hubbert, “Tech-niques of Production as Applied to the Production of Oil and Gas, ”in 0// arrd Gas Supp/y Mode/irrg, S. 1, Gass (cd,), National Bureau ofStandards Special Publication 631, May 1982,

rived equations for the magnitudes and rates ofchange of gas production and discoveries bynoting some simple boundary conditions for theproduction c ycIe* and fitting a second order

equation * * to these conditions. By further

manipulating the equation obtained by this ex-

ercise, Hubbert derived three separate but relatedmethods of estimating Q m, two involving the

curve of cumulative discoveries and one involv-ing production rate as a function of cumulativeproduct ion.

In his second approach Hubbert assumed thatthe ratio of the discoveries of natural gas to thoseof c rude o il will tend to rema in stab le, a llowingthe g as resource ba se to be c alculated as a sim-ple function of the oil resource base.

The third a pp roac h involved e xtra po lating thedeclining finding rate for gas out to the point

where exploratory drilling ceases and taking thearea under the curve, as discussed in the earliersection on historical approaches to resource es-timation (see fig. 8).

Hubbert’s work has been the subject of numer-ous critical appraisals.23 This d isc ussion w ill no tattempt to review the appraisals but will incor-porate some of their key points.

Of Hubbert’s five estimates, the first three in-volve the assump tion tha t the c urves of declin-ing production and proved reserves will be themirror image of the curves of the (increasing) first

portion of the resource development cycle. Thisderives from Hubbert’s satisfaction with the “fit”of the simple quadratic equation he uses to ap-proximate the curve of ~ v. Q. Aside from thecriticism associated with all historical approaches

*Cumulative production Q is zero at the beginning of the cycleand Qm at the end; the production rate ~~ is zero when Q = O

and also when Q = Q ~).

**R = C,Q + C2Q2

zJFor example, L. S. Mayer cites three: D. V. P. Harris, “Con-

ventional Crude Oil Resources of the U. S.: Recent Estimates, Meth-

ods for Estimation and Policy Consideration, ” Makrja/s art~ socje~y

1, 1977; N. Uri, “A Reexamination of the Estimation of Undiscov-ered Oil Resources in the U.S.,” DOE/TM/ES/79-03, 1979, EIA; L.

Mayer, et al., “Modeling the Rates of Domestic Crude Oil Discov-ery and Production, ” report to the EIA, Princeton University, De-

partment of Statistics, 1979. (In comment on j, j. Wiorkowski, “Esti-

mating Volumes of Remaining Fossil Fuel Resources: A Critical

Review, ” j. Am . Stat. Assoc ., September 1981)




–that the future does not have to look like the

past, and more often than not doesn’t-–Hubbertnever explores the possibility that he could

achieve an equal or better fit with a different

equation and thereby calculate a different Q~.Critics have shown, for example, that the re-

source base values obtained from fitting a curveto o il prod uc tion d ata are sensitive to the typ eof curve used, and that Hubbert’s assumed curveis not the best choice.24 Although Hubbert’s

curve for oil discovery is more satisfactory, it maybe that the less mature gas discovery curve is alsoflawed. *

The a ssumption of the fo urth estima te , tha t theratio o f gas d isc overies to oil disc overies will re-

ma in stab le, ap pe ars to b e ve ry we ak. The g rea tmajority (85 percent) of gas discoveries today arenot assoc iated with o il, and it is the c onsensus

of ma ny ge ologists that a large po rtion o f the re-maining gas resource lies below 15,000 ft in a

physical environment hostile to the preservationof oil. A method predicated on stable gas/oil ratioswould appear to guarantee an overly pessimis-tic gas resource base estimate.

in the last estimate , Hubbert fits an expone n-

tial curve to a historic al p lot of finding rate (theultimate volume of gas to be produced from fields

discovered by 108 ft of exploratory drilling) versusc umulative exploratory d rilling, by req uiring thecurve to pass through the last data point and byrequiring the area under the fitted curve to equal

24 E,g., J. j. wiorkowski,1981 ~ “Estimating Volumes of Remain-

i ng Fossil Fuel Resources: A Critical Review, ” ). Arrr. Stat. ASSOC .,

September 1981, vol. 76, No. 875.

*The reasoning here is that the oil discovery curve gives more

satisfactory results than the oil production curve because d iscov-

ery is more advanced in its overall cycle. The less advanced, or

less “mature,” the curve, the less satisfactory will be the results.

the a rea unde r the historic al d ata plot (see fig.8). This estima te ha s severa l serious prob lems.

First the curve does not fit the data because it vir-tually ignores the “form” of the data and con-centrates instead on the last data point.25 Sec ond ,

the estimate is very sensitive to this last data point,

yet the ma gnitude o f the p oint is the sum o f avalue (reported new field wildcat discoveries) thatmay vary with economic conditions* and with

the state of depletion of the resource base plusa sec ond va lue (reserve grow th a fter the initia lreporting period) that is, at best, a gross approx-imation. * * Third, as with the first three estimates,Hubbert makes no attempt to explore the possi-bil ity that he could achieve a better “f it” with adifferent curve. His choice of a negative exponen-tial curve is an assertion made several times butunsupported by reasoning in his text.

An interesting o bserva tion a bout this last esti-ma te is that d espite the fac t that the fitted c urveis well under the trend line of the last several unitsof drilling—an ingredient for an overly conserva-tive estimate—the estimate is considerably higherthan the four other estimates in table 9.

‘25 Harris, 197’7, Op. cit.

*For example, a period of high risk exploratory effort-responding

to economic conditions that favor this sort of activity–will tend

to yield high discovery rates, whereas one of lower risk effort re-

sponding to different conditions generally will yield lower rates.

This is important here because Hubbert’s analysis is dependent on

the finding rate being a function only of the physical resource base

and its state of depletion.

‘ *The procedure used to estimate reserve growth utilizes the aver-age growth rate over many years. However, the year-to-year his-

torical growth rates have tended to be quite volatile, so the aver-

age growth rate for a single year or single period of 10* ft of drilling

is ,~t best a rough approximation. Furthermore, there are reasons

to suspect that the /ong- term trend of reserve growth may now be

turning downwards, causing a further error in an estimate assum-

ing an unchanging trend.

RECONCILING THE DIFFERENT ESTIMATES

Which of these resource assessments are to be 2. How credible are the methods used by thebe lieve d? in a pp roa c hing this question, OTA used

assessors, in the abstract and in actual per-three c riteria: formance?

3. Wha t d o t he d ifferent a ssessme nts imp ly in1. Is there a consensus, or even a “central terms of geology and future discoveries? Are

tendency, “ in the scientific community? these implications credible?




Is There A Consensus?

In OTA’ s judgm ent, the ra nge of o pinion in thescientific community about the size of the natu-ral gas resource is too wide to represent a signif-icant consensus. Not only are there the obvious

divisions along the lines of the various estimates,or simply between “optimistic” and “pessimis-tic, ” there is also an important division betweenscientists who believe i n a particuIar estimate orrange of estimates and those who do not believethat the state of knowledge is adequate enoughto a llow any reliab le estimate to b e ma de. Fur-thermore, some scientists believe that those esti-

Box E.—A Very Brief History

The history of petroleum exploration in general,

mates that invoke current technology and eco-nomic relationships—the great majority—aresimply irrelevant, whether or not they are correctwithin the constraints of these assumptions. Thesescientists believe that both the inexorable ad-vance of technology and rising prices that reflect

resource sc a rc ity w ill c onstantly push o utwa rdsthe boundaries of the recoverable resource base.As noted previously, the history of resource esti-mation in general tends to support this view;

cycles of predictions of scarcity followed byradical upward revisions in resource assessmentsappear to be common for nonrenewable re-sources (see box E). On the other hand, the USGS

of Petroleum Exploration

and exploration for natural gas in particular, hasbeen one of continuous movement toward new discovery horizons and resulting reappraisals of resource

potential. The “movement” encompasses new geologic theories and “ideas,” new exploration and pro-duction technologies, and new geographic areas.

During the first half-century of exploration following Drake’s initialdiscovery in 1859, exploratorydrilling was essentially random drilling, drilling at oil seeps, or drilling in areas where previous strikeshad be en m ad e. Then a succ ession o f ge ologic insights beg an to o pe n up new horizons for exploration:first, the understanding that anticlines, some with surface manifestations, could serve as traps forpetroleum; then, the discovery that petroleum deposits could exist in traps on the flanks of salt domes;next, the recognition of the petroleum potential of sand lenses and stratigraphic traps; and finally, theinsight that petroleum could exist in recoverable quantities underneath thrusting plates, leading to theopening up of the Overthrust Belts to exploration and eventual Iarge, discoveries.

Another discovery “horizon” was the growing sophistication of the tools of the trade: the adventof the gravity meter and magnetometer, allowing the locating of geologic anomalies that might signalthe existence of structural traps; the addition to the explorer’s tool kit of refractive and then reflective

seismology, which permitted the detailed mapping of geologic structures; the introduction of rotarydrilling and advanced drill bits that allowed deeper horizons to be explored; the growing use of fractur-ing technologies, which opened up another geologic horizon in petroleum-bearing rock of low per-meability; and the engineering triumphs of offshore drilling technologies.

At the same time, exploration and development moved into new regions, sometimes driven by thenew technologies (e.g., the continental shelves) or new ideas (e.g., into Texas after realization of theimportance of salt domes) and sometimes driven simply by the need for new supplies and dwindlingprospects in the mature regions. Thus, exploration began in the Appalachian region but moved inex-orably into Ohio and Kansas, into California and the mid-continent region, to the onshore Gulf of Mexico,and spilled out into the Offshore, moved to the Overthrust Belt, and drove to deeper horizons in theAnadarko.

This history of constant movement to new horizons provides grist for the mill of both the resource

optimists and the pessimists. The optimists focus on the seemingly continuous ability of explorationists

to find new g eologic conc epts and to develop new technologies that allow them to expand the petroleumresource base over and over again. The pessimists focus on the questions: Just how long can this goon? How many additional places are there to look? As noted earlier in the section on “Resource Base

Concepts,” this history and the ongoing controversy in the search for petroleum is a paradigm for thedevelopment of many nonrenewable resources.SOURCE Dr. John Schanz, Senior Specialist in Energy Resources Policy, Congressional Research Service.

3 8 - 7 4 2 0 - 8 5 - 5




oil and gas resource estimates of the past dec-ade and a half sustained some very substantialdownward revisions as estimation procedures be-came more sophisticated.

Tables 10 and 11 summ arize som e of the key

argume nts used by the op timists and pessimists

in explaining their positions on the probable sizeof the gas resource base. Because each of the ar-gum ents has me rit, it is ob vious that a n una mb ig-uous answe r to t he q uestion, “ How large is theU.S. gas resource base? “ is no t likely. Selec tionof a “ best” estimate is further co nfused by theobservation that some major disagreements ex-ist even among assessors who appear to have thesame ge neral outlook (see bo x F), and som e o fthe more important disagreements occur in areas

Table 10.—The Optimist’s View of Gas Resources

q

q

q

q

q

b

q

—

Just a few short years ago nobody had heard about theOverthrust Belt and the Tuscaloosa Trend; noweverybody has jumped in. The pessimists have alwaysbeen wrong about resource shortages.Increased prices for gas and better explorationtechniques have opened up a huge new resource insmall fields. Past estimates of the number of smallfields relied on data from a time when a small fieldwas likely to be abandoned as a dry hole.We haven’t been looking for natural gas for more thana few decades, so a mature basin for oil—with littleprospects for significant new finds—isn’t necessarilymature at all for gas. This is especially true becausethe conditions that led to gas are often hostile to theformation and preservation of oil, and thus thepresence of these conditions would have tended tokeep explorers away. A key example of this effect is

the deep gas resource.A good part of the lower finding rates of the recentpast was due to the substantial increase in low-risk,low-yield drilling. The lower rates therefore do notnecessarily imply “resource depletion. ”Most resource estimates—including optimistic onessuch as those of USGS and PGC—represent onlysnapshots in time, reflecting current economics andtechnology. The resource base estimates will tend togrow over time as prices rise and technologyadvances.The decline in proved reserves of the past decade,interpreted by many as a sign of resource depletion,actually represents merely a rational response to highdiscount rates, that is, a reduction in inventory to theminimum amount necessary to sustain production.Recent price increases have opened up a largepotential for new reserves from the growth of olderfields. This new gas will come from closer spaceddrilling, the extention of fields to lower permeabilityareas that were previously uneconomic, the lowering ofabandonment pressures, and well workovers.


Table 11.—The Pessimist’s View of Gas Resource

We have drilled too many holes in the Lower 48 States andtested too many ideas to believe there is much room forbrand new natural gas horizons.If there’s so much gas right here in the Lower 48, why arewe testing the limits of hostile environments in the Arcticand continental slopes?The geologists who make industry’s resource estimates

tend to be the most successful ones, those who have abuilt-in bias toward optimism because of their experience.We have already found most of the “easy,” giant fields.The future is in the smaller reservoirs, and there doesn’tappear to be enough of these to provide the amount ofresources the optimists say is there.The depletion effects apparent in exploratory drillingfinding rates are actually understated because the advanceof exploration technology, by increasing the success rateof exploratory drilling, has tended to hide the onset ofdepletion.The higher resource estimates, when translated into thenumber of fields of various sizes that must be discovered to yield this much gas, look very shaky when comparedto the numbers of these fields that we have actually beendiscovering lately.


where c onsiderab le geo logic d ata exists to a id theresource assessments (and where, consequently,th e most agreement might be expected).

Given what OTA w ould term a lac k of c onsen-sus, is there at least a “central tendency?” Whatis an acceptable range of estimates for the sizeof the recoverable resource base that excludes“unconventional gas”* and gas that cannot beexploited profitably at gas prices in the samerange as today’s and with technology that is wellwithin rea c h in the next few d ec ad es? OTA be -

lieves that a substantial majority of scientists con-cerned about the gas resource base would feelcomfortable somewhere within* * a range that in-cluded Nehring’s estimate as the extremely pes-simistic minimum and the PGC estimate as notquite the ma ximum , but c lose to it. This rangeis about 280 to 915 TCF for the rem aining c on-ventional gas resource (including proved reserves

and the growth of known fields) recoverable, withreadily foreseeable technology and given today’sec ono mics, for the Lowe r 48 Sta tes.

*Gas from very tight formations, geopressurized zones, coal beds,and Devonian shales. However, gas that arguably could be placedin these categories but that is commonly produced today, wouldbe considered conventional.

* *Many wou Id no doubt disagree strongly with values near oneextreme or the other, however.




OTA believes tha t the m inority who m ight like from this end of the rang e. * It should b e a dd ed ,the range extended would consist mainly of those however, that some of those who are consider-who believe that the upper end should be higher. ab ly less op timistic tha n PGC, a nd even USGS,

Furthermo re, OTA suspec ts tha t a tho rough re- —view of the production implications of the lower *As shown in chapter 5, a 280-TCF remaining resource implies

end of the range—as discussed in the next chap-that the year 2000 production of Lower 48 conventional gas, re-

ter—would tend to push many scientists awaycoverable with existing or foreseeable technology and at the cur-

rent cost/price relationships, cannot be much greater than 4 TCF/yr.




a re ma jor oil and ga s prod uc ers—e.g., Exxon* *—who are very familiar with most of the areas thatare supposed to supply the United States with the‘‘optimistic’ levels of new gas discoveries.

How Credible Are the Methods?

How credible the methods are is generally dif-ficult to determine because few resource assess-ments using geologic approaches reveal manydetails of their assessment processes. Generally,more details are available for the assessmentsbased on historical, extrapolative approaches; inad dition, USGS ma kes a va ilab le to the p ublic itsop en-file repo rts and da ta. OTA d id not a ttemp tto review the e xtensive USGS bac kup informa -tion because of time and budget constraints. His-torical approaches have been reviewed in a num-ber of reports,26 and for the mo st p art OTA c hoseto use these reports instead of conducting a totallyindependent review.

In g enera l, OTA is skep tica l of histo ric a l ap -proaches to resource assessment when they arebased on nat ional data and when they are thesole m ea ns of e stima tion. The substa ntial d a taproblems associated with natural gas exploration

(especially during those years when gas was val-ued as little m ore than a byp rod uct of oil pro-duc t ion), the broad range of ac t ivity co vered byany single data series, and the distorting effectsof Government controls are important sourcesof this skepticism.

The mo st imp ortant estima te ba sed stric tly ona historic a l approa c h is Hubbe rt’ s, bec ause hehas gained substantial credibility from his success-ful predictions of declining U.S. oil production. As

d isc ussed ea rlier in this c ha p ter, OTA notes sub -stantial problems with Hubbert’s approach andbelieves tha t h is extremely p essimistic estima te(870 TCF) of ultima tely rec ove ra ble c onve ntionalgas is too low.

Of the assessments using geologic approaches,only the assessme nts of USGS and PGC are re-viewable in any sense because details of the

oth ers a re not p ub lic informa tion. In OTA’ s op in- —.. . - ——

* *OTA has been told informally by Exxon geologists that Exxon’s

most recent internal estimates of the U.S. gas resource base are

considerably below those of USGS and PGC. The major disagree-

ments are with estimates for the Lower 48 onshore gas potential.2 6 F o r example, Wiorkowski, oP ~it.

ion, both assessment processes are serious at-

tem pts to w restle with a m ost d iffic ult problem .One problem with both assessments is the fail-ure to include the detailed assumptions behind,and implications of, the assessment, thus preclud-ing muc h op po rtunity for useful feed ba c k from

those outside the assessment process. The USGS

assessment may also be hampered by lack of ac-c ess to p rop rietary industry da ta ; PGC, o n theother hand, apparently has access to excellentdata but appears to ignore the insight that mightbe ga ined from a na lyses of d isc ove ry trend s (i.e.,the historic approach),

Are the Physical Implications ofthe Assessments Plausible?

Most gas resource assessments do not providedesc riptions of either the direc t p hysic al imp li-cations of their resource estimates (e.g., the num-ber and size of fields implied by the estimate) or,conversely, the initial physical model used to de-rive the estimate. Nevertheless, some physical im-plications can be drawn directly from the esti-ma tes. This is espec ially true w hen the estima tesare separated into components: onshore and off-

shore (quite common), deep and shallow (e.g.,the PGC assessme nt), and individua l reg ions or

even smaller provinces (USGS divides the UnitedSta tes into 137 sep a rate provinces). Conse-quently, it is clear that PGC believes that the deepresource below 15,000 ft represents a massive

source; fully 39 percent of the onshore undiscov-ered resource of the Lower 48 States is projectedto be dee p gas. In a simila r vein, USGS c learlyappears to have given up on the eastern Gulf ofMexico but has great hope–as does PGC–foranothe r “frontier” a rea , the Western OverthrustBelt.

Rathe r than c arrying out a de tailed “ transla-tion” of e ac h a ssessme nt, OTA c hose to e xam-ine two basic physical issues that appear to cutac ross virtua lly all of the assessments. Theseissues, as stated by Nehring,27 are:

q Does the assessment imply a substantialbreak w ith pa st a nd rec ent d isc overy trend sand patterns?

27R. Nehring, The Disco very of Significa nt Oil and Gas Fields in

th e United States, op. cit.




q If the assessment does imply such a break,what is the explanation for it? IS i t credible?

A Break With Past Trends?*

The mo st o bv ious ties betw ee n p a st trend s a nd

the magnitude of the resource base are the anal-yses performed in the “historic approaches” toresource assessment. In general, these ap-proaches have given relatively pessimistic resultswhen used with U.S. gas production and explora-tion da ta . For examp le, a ll four of the estimate s

using pure data-tracking techniques (two by Hub-bert, one each by Wiorkowsky and Bromberg/ Hartigan) in table 6 are below the USGS estimate,with three o f the four at lea st 400 TCF below . I naddition, the RAND estimate, which is at leastpartly dependent on past discovery trends, isnea rly 500 TCF below the USGS estima te s.

This series of pessimistic resource estima tes

based on trend analysis, when coupled with thevery low rates of reserve additions in the Lower48 States from 1968 to 1978 (average yearly AGAreserve a dd itions we re 9.6 TCF v. average p ro-duc tion o f 20.6 TCF/ yr), represent a strong ini-

*Readers interested in past trends In petroleum exploration may

also wish to read Exploration for Oil and Gas in the United States:

An Ana lysis of Trends and Op po rtunity, by John j. Schanz, jr. and

Joseph P. Riva, jr., of the Congressional Research Service (CRS re-

port No. 82-138 S, Sept. 16, 1982).

Table 12.—Returns to New Field Wildcat Drilling in

tial a rgum ent tha t the m ore op timistic resource

estimates do represent a break with past trends,while the pessimistic estimates do not. However,as noted in the discussion of historical approaches

to resource assessment, the available data usedto measure trends in exploratory success (or

trends in other factors that may be used to form judgments about the probable size of the re-source base) tend to measure multiple rather thansingle processes; for example, measures of thesuccess of drilling for new fields are, in fact, meas-

uring a range of activities from the high-risk testingof new geological ideas to the low-risk redrillingof formerly uneconomic dry holes. Consequently,none of these trends can be interpreted in an un-amb iguous ma nner. The d isc ussions in cha p ter5 about the factors that affect the various compo-nents of reserve additions give a sense of the com-plexity of individual trends and of the difficuIties

in interpreting the trends.

Trend s in the d isc ove ry of ne w fields app ea r

likely to b e m ost c losely assoc ia ted with the re-

maining recoverable resource base; these trendsare e xam ined in the following p arag rap hs.

Tab le 12 disp lays the returns to new fieldwildc at d rilling in the o nshore Low er 48 Sta tesfrom 1966 to 1981, The patterns displayed in thetab le d em and c areful de c iphe ring. The g as vol-umes found per successful gas new field wildcat

the Onshore Lower 48 States, 1966-81 (BCF/well)

New field discoveries Percent of new fieldPer new field Per new gasfield discovery wells

Year Per all NFWs discovery well discovery well that find gas

1966 . . . . . . . . . . . . . . . . . 0.46 4.56 18.56 251967 . . . . . . . . . . . . . . . . . 0.42 3.96 11.93 33

1988 . . . . . . . . . . . . . . . . . 0.24 2.66 10.25 27

1989 . . . . . . . . . . . . . . . . . 0.24 2.66 7.47 36

1970 . . . . . . . . . . . . . . . . . 0.29 3.01 8.20 37

1971 . . . . . . . . . . . . . . . . . 0.16 1.67 3.70 45

1972 . . . . . . . . . . . . . . . . . 0.24 2.11 4.46 47

1973 . . . . . . . . . . . . . . . . . 0.34 2.30 3.89 59

1974 . . . . . . . . . . . . . . . . . 0.24 1.60 2.88 561975 . . . . . . . . . . . . . . . . . 0.22 1.47 2.91 51

1976 . . . . . . . . . . . . . . . . . 0.18 1,02 1.85 551977 . . . . . . . . . . . . . . . . . 0.20 (.32)a 1.15 (1 .86) 2.23 (3.61) 52

1978 . . . . . . . . . . . . . . . . . 0.17 (0.36) 1.07 (2.27) 1.96 (4.17) 551979 . . . . . . . . . . . . . . . . . 0.20 (0.26) 1.07 (1 .40) 1.85 (2.48) 58

1980 . . . . . . . . . . . . . . . . . (0,27) (1.37) (2.69) 51

1981 . . . . . . . . . . . . . . . . . (0.34) (1.88) (3.95) 48

aAGA data (EIA data).

SOURCE: R Nehring, “Problems in Natural Gas Reserve, Drilling, and Discovery Date,” contractor report to the Off Ice of Technology Assessment, 1983




show a startling decline during the period, from18.56 billion cubic feet (BCF) per well in 1966to 1.85 BCF per well in 1979 (use of EIA datamoderates this trend somewhat, but the EIA andAGA d ata a re not stric tly c om pa rable). This me ans

that the average field size found by a successful

gas wildcat declined by a factor of 10 during1966-79.

Because the larger fields in a basin are gener-ally found early in the discovery process, asharply declining average field size is often in-terpreted as a sign that the discovery cycle is win-ding down. However, the data shown in the tableare collected from multiple basins, and during thetime period in question, the pattern of gas ex-ploration may have been influenced by increasedgas prices and other factors. For example, it iswidely believed that de libe rate exploration forsma ll gas ta rgets (e. g ., in area s where p ast ex-

ploration identified then-uneconomic gas depos-its) increased sharply during this period. Such anincrease in the willingness of explorationists togo after sma ll targe ts wo uld tend to reduc e fieldsize averages even if high-risk exploration for large

fields maintained a steady success record. Con-

sequently, the decline in a verag e field size m aynot fairly represent the actual condition of theresource base.

The rec ord of returns to w ildc a t d rilling per welldrilled tends to support this view . These returnsper well d rilled have e xhib ited only a slight d e-

cline since 1968; the success rate, which variesfrom a low of 2.3 percent in 1968 to a high of

10.8 percent in 1979, essentially compensates forthe declining field size. In other words, while each

gas wildcat well completed returned far less gasin 1979 than in 1966, the actual number of wildcatwells drilled to find each trillion cubic feet of gasdid not increase very much during this period. Thisrelatively op timistic result should be tem pered ,however, by the observation that the percentageof wildcats aimed deliberately at gas targets prob-

ably increased during this period. Consequently,it is likely that the actual gas-directed effort-as

distinct from the total petroleum-directed effort–that wa s need ed to find a unit of ga s proba blydid increase during the p eriod .

Although the da ta in tab le 12 loo k more o pti-mistic than might ha ve b een initially expec ted ,the history of natural gas development impliesthat, in order to sustain successful levels of re-serve additions for the long-term, efforts must bema de to op en new g eolog ic ho rizons and find

the large fields that are the cornerstone of reservegrowth in later years. Consequently, it is usefulto examine the pattern of discovery of different-sized fields.

The Ame ric an Assoc ia tion of Petroleum Ge olo-gists (AAPG) publishes the primary public recordof the disc ove ry of p etroleum fields by size a ndd isc ove ry yea r, and this rec ord m ay b e used toexam ine pa tterns of d isc ove ry. The rec ord m ustbe used cautiously, however, because AAPG ap-pears to have undercounted the number of fieldsdiscovered. * For example, from 1971 to 1975,AAPG reports only 49 gas discoveries of a size

greater than 60 BCF. In comparison, the RANDdata base reports 141 fields in this size range dur-ing the same time period.28 Consequently, theAAPG data should be examined for trends ratherthan ab solute mag nitude , and even the trend s

may be skewed if undercounting and other prob-lems we re not c onsistent ove r time .

Tab le 13 presents the histo ric a l rec ord o f newgas field discoveries by field size, for 1945-75, ascompiled by AAPG. * * In parallel with the trendsshown in table 12, the percent of significant (sizeclass A through D) gas fields in all gas discoveries

de c rea sed over the 30-yea r period , while the ef-fort required to find a significant field increasedthrough the 1960s but then declined to earlierlevels.

‘The d ata in the tab le c an b e used to examinethe discovery trends of larger fields. Figures 11and 12 show trends in, respectively, the number

*Part of this problem may arise from simple disagreements over

field boundaries; the EIA data base, for example, treats the Hugotonfield as three separate large fields, whereas other analysts might

count it as one. Also, field reserve estimates are not consistent across

data bases.zeR, Nehrlng, prob lems in Natural Gas Reserve, Drilling, and Dis-

covery Data, contractor report to OTA, 1983.* +! The record stops in197s because AAPG classifies fieids as gas

or oil fields only after the passage of 6 years past the discovery

report.




Figure 11.– Number of Gasfields Discovered As a Percentage of New Field Wildcats Drilled,by Field Size Grouping

1.5

1.4

1.3

1.2

1.1

Field sizes (BCF): A >300

150-30060-150

6-60

1955 1960

Year

1965 1970 1974

SOURCE: Office of Technology Assessment, based on data from table 16 in R. R. Johnston, “North American Drilling Activity in 1961,” AAPG Bulletin, vol. 66/1 1, November

1982.

Figure 12.-Number of Gasfields Discovered per Year, by Field Size Grouping25

1960

Year

1970 19741946 1955 1965

SOURCE: Office of Technology Assessment, based on data from table 16 in R. R. Johnston, “North American Drilling Activity in 1981 ,“ AAPG Bulletin, vol. 66/11, November

1982.




economic range of a l imited inventory of knownprospec ts and an a cc eleration of the normal pa ceof field d eve lopm ent. OTA has not see n any c on-vincing a na lyses arguing one side o r the othe r.

As for the Overthrust Belt a nd Ana da rko, thefuture of these areas is uncertain. The OverthrustBelt did p rod uc e som e ve ry large new fields inthe late 1970s (the Whitney Canyon/Carter Creekand East Anschutz Ranch fields appear to haveresource s grea ter tha n 1 TCF) and its potentialis substantial. However, despite continued

searching, no new giant fields have been discov-ered in the past few years. In the Anadarko, the

recent declines in p ric es for deep ga s ma y havemoved some gas from “economic” to “subeco-nomic, ’ al though the earlier superheated mar-ket for this gas and the resulting distortionsin prices and production costs make it difficult

to predict where the ec onomic/ subec onomicboundary might lie in the future. Also, recentengineering difficulties and rapid pressure de-clines in some fields imply that some overesti-

mates may ha ve be en ma de in c alculating re-serves and estima ting resource s.

In c onc lusion, in OTA’ s op inion the gas dis-

covery trends of the past several decades, whilenot supporting the most pessimistic of the re-

cent gas resource estimates, also do not supportthe relatively optimistic estimates of PGC and,possibly, USGS.

Some Alternative Explanations

The (until rec en tly) mo d era te ly pe ssimistic d is-covery trends and optimistic resource base esti-mates can be reconci led by two possiblearguments:

q

q

It is not the resource base but the marketdistortions caused by Government regula-

tions tha t have c aused d isc ove ry trend s tobe disap pointing. Exploratory incentives

have been skewed toward low-risk, low-pa yoff ga s prospe c ts.

The historic a l trend s d o represent the d ep le-tion of traditional sources of natural gas.Now, howe ver, imp roved tec hnology andhigher prices will allow explorers to find largequantities of g as from : —small fields;

—reworking of older fields; —new frontiers, inc luding d ee p ga s; and

—subtle stratigraphic traps.

The Causes of Past Trends

ISit the nature of the remaining resource basethat has been the primary influence on histori-c a l dec lining trend s in new field d isc ove ries, orwa s it instea d the economic and regulatory envi-ronment that provided the controll ing influence?Does the relatively low rate of discovery of largenew gasfields during the last decade and a halfreflect resource depletion, or are these rates anartifact of the erratic price and regulatory historyof natural gas? If gas resources are substantiallydepleted, it appears unlikely that gas finding ratesand discoveries of large new fields will reboundto levels that would sustain high production rates.If the ec ono mic/ reg ula tory history of g as is thec ause, then o ptimism a bo ut future p rod uc tionpo tential may b e w ell founde d, a ssuming thateconomic and regulatory conditions can be madefavorable to the gas discovery process.

The b asic argum ent tha t low finding rate s fornew fields and othe r warning signa ls do not re-f lec t resource de pletion c enters around the ideathat the rigid price controls of the period beforepassage of the Natural Ga s Polic y Ac t of 1978(NGPA) locked drilling into lower cost and risk

areas that do not coincide with where the majorga s p ot en tial reside s. The “ c ulprit” for this is sa id

to b e the m ethod used by the o ld Fed eral Pow erCom mission (FPC) to c alculate allow ab le “ area”and “national” gas prices. FPC assumed that fu-ture e xplorato ry and de velopme nt c osts wouldbe similar to past average costs, and by basingthe allowable price on this assumption, essentiallyguaranteed that drilling would be confined toarea s where c osts were expec ted to b e low.

A p ast p rop onent of this view has be en theAmerican Gas Association. AGA has conducteda series of studies32 comparing total gas well com-p letions to estimate s of g as resource p ote ntial*

in the Outer Co ntinenta l She lf, Alaska, the

JzThe latest is AGA, “Gas Well Drilling Activity and Expenditures

in Relation to Potential Resources, ” Gas Energy Review, vol. 9,

No. 1, January 1981.

*The measure used for “Resource Potential” was PGC’S estimates

of potential supply.




shallow Lower 48 area, and deep (below 15,000ft) horizons. Its September 1979 analysis, whichincludes drilling data through 1977, concludesthat “the drilling data suggested that the declinein proved reserves was not due to a depletion ofgas sources but rather to a lack of economic incen-

tives for drilling under an artificially constrained,regulated environment [emphasis added].”33 Thisconclusion was based on the poor correlation ofgas well completions to gas resource potential de-tected in the study* (see the first two circlecharts in fig. 13). However, a more recent (Janu-ary 1981 ) analysis added a comparison of gas well

expenditures to gas resource potential (third cir-cle chart in fig. 13). Noticing a good correlationof expenditures to resource potential,** AGAom itted the e arlier co nc lusion a nd attributed the

imbalance between drilling and potential to “themuch lower cost-per-well and cost-per-foot f ig-

ures for the shallow, Lower 48 wells. ” 34 The ve ryhigh drilling costs and risks of the high gas po-tential frontier areas necessitate a very cautiousattitude toward drilling, whereas the lower costs

33A(JA, “Drilling Activity and Potential Gas Resources, ” Gas

Energy Review, vol. 7, No. 11, September 1979.

*Of course, an alternate reason for the poor correlation could

be that gas entrepreneurs do not agree with AGA’s view aboutwhere the resource potential lies.

* *Except for Alaska, where lack of a transportation system blocksgasfield development.34AGA, “Gas Well Drilling and Expenditures. . . ,“ op. cit.

in developed onshore areas encourage closelyspaced development drilling and exploratorydrilling for small reservoirs and other marginaltargets.

A corollary to the argument about the effectsof low allowable gas prices is used to explain whythe sharp price increases of the past several yearshave not improved the rate of new field discov-eries. According to this view, drilling priorities willnot immediately be corrected by rising prices be-

cause the long period of controls has created alarge backlog of low-risk, previously marginal ex-ploration p rospe c ts that are now c omm erciallyviable. Until this backlog is reduced, the argu-ment goes, exploratory drilling will stay awayfrom the high-risk, high-payoff wells that couldfind the large fields35 that now only appear to bescarce. Furthermore, because price increases ex-

pand the boundaries of the “economically recov-erable” resource base and thus add to the inven-tory of low-risk prospects, it is claimed that thetrend towa rd low -risk, low -pa yoff d rilling is likelyto continue if prices continue rising.36

Jsjensen Associates, InC., “Early Effects of the Natural Gas Pol-icy Act of 1978 on U.S. Gas Supply, ” report to the Office of Oil

and Natural Gas, U.S. DOE, April 1981.MR p O’Neill, “Issues in Forecasting Conventional Otl and Gas. .

Prod uction, ” in Oi/ and Gas Supp / y Mode / i ng , National Bureau

of Standards Special Publication 631, May 1982.

Figure 13.—Gas Potential, Gas Well Completions, and Expenditures—1978

1,019 TCF 13,306 gas wells $4,978.9 million

Ia

Potential Gas well completions Gas well expendituresNOTE: “Shallow” and “deep” refer to Lower 48 States onshore; “potential” IS based on PGC’s estimates of the undiscovered gas resource

SOURCE “Gas Well Drilling Activity and Expenditures in Relation to Potential Resource,” in Gas Energy Review, vol 9, No. 1 (Arlington, Va.. American Gas Associa-tion, January 1981).



68 . U .S . Natural Gas Availability: Gas Supply Through the Year 2000

High-risk, high-payoff drilling may be expectedto yield low success rates. Consequently, the

sharply improved success ratios of both total ex-ploratory drilling and new field wildcat drillingduring the past decade and a half, shown in table14 and figu re 14, has bee n used to support the

the sis tha t d rilling is skewed towa rd the low-risktargets. The overall success rate of these drillingcategories may be affected by a variety of factors,however, that cannot be separated out. For ex-ample, substantial progress in improving explora-tion techniques and computer technology dur-ing this pe riod undo ubte dly ac ted to increase

success rates, but to an unknown degree. * Also,the success rate is automatically elevated by the

*The exten~ive investigation of the effects of new technology by

the National Petroleum Council in 1965 could find no credible

quantitative measurement of these effects.

decrease in minimum acceptable field sizes andthe gas flow rates associated with increased gasprices; small fields and low-permeability reser-

voirs that in the pa st wo uld have be en c onsid-ered “dry” are now being developed as pro-

d uc ers. Therefore, it is quite c onc eivab le tha t a n

increase in overall success rates could be accom-panied by an increase in high-risk drilling if theother factors affecting success rate were strongenough to overcome the negative effects of theshift in risk,

In a dd ition to argume nts ab out the effec ts of

price controls, some analysts point out that main-tenance of high levels of proved reserves in rela-tionship to production would not be compatiblewith go od business p rac tices. Acc ording to thisargument, high interest rates made it sensible for

Figure 14.– New Field Wildcat Success Rate, 1966=81

1966

r

1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1978 1979

Year of reported discovery

NOTE: Gas success rate data not available after 1975 because gasfields and oilfields are separated out only after a 6-year review by AAPG.

SOURCE: Office of Technology Assessment, based on data from American Petroleum Institute, Quarterly Review of Drilling Statistics




Table 14.-Oil and Gas Drilling Success Rates(discoveries as a percentage of exploratory drilling effort)

Exploratory wells “Wildcats”

Year Completed Total Rate Completed Total Rate

1966 . . . . . . . . . . . . . . 1,894 10,313 18.40/o 635 6,158 10.3 ”/01967 . . . . . . . . . . . . . . 1,518 8,878 17.1 544 5,271 10.31968 . . . . . . . . . . . . . . 1,440 8,879 16.2 442 5,205 8.51989 . . . . . . . . . . . . . . 1,700 9,701 17.5 535 5,956 9.0

1970. . . . . . . . . . . . . . 1,271 7,693 16.5 493 5,069 9.71971 . . . . . . . . . . . . . . 1,088 6,922 15.7 436 4,463 9.71972. . . . . . . . . . . . . . 1,285 7,539 17.0 566 5,086 11.11973. . . . . . . . . . . . . . 1,519 7,466 20.3 701 4,989 14.11974. . . . . . . . . . . . . . 2,009 8,619 23.3 805 5,652 14.2

1975. . . . . . . . . . . . . . 2,143 9,214 23.3 876 6,104 14.41976. . . . . . . . . . . . . . 2,449 9,234 26.5 986 5,840 16.91977. . . . . . . . . . . . . . 2,686 9,961 27.0 1,004 6,101 16.51978. . . . . . . . . . . . . . 2,728 10,677 25.6 983 6,505 15.11979. . . . . . . . . . . . . . 3,024 10,484 28.8 1,162 6,413 18.1

1980. . . . . . . . . . . . . . 3,574 11,916 30.0 1,340 7,034 19.01981 . . . . . . . . . . . . . . 4,585 15,168 30.2 1,423 8,052 17.71982. . . . . . . . . . . . . . 4,847 16,470 29.4 1,400 7,912 17.7SOURCE’ American Petroleum Institute, “Quarterly Review of Drilling Statistics”

gas producers to reduce their standing inven-

tory—i.e., proved reserves—by maximizing de-Iiverability and reducing exploration. Conse-quently, from the drilling Iow point of 1971 to1982, developmental drilling rose by a factor of3.66 (18,929 wells drilled v. 69,330), whereastota l explorato ry d rilling rose b y only a fa c tor of2.38 and new field wildcats rose by only 1.77.

37

Carrying this argument further the economic in-centive to increase reserves will occur only whenthe cost of reducing R/P ratios—of adding to thedeliverability occurrent reserves–outweighs the

c ost of a dd ing new reserves.

Although the a rgument a bo ut the lac k of aneconomic incentive to increase reserves is a fairone, it does not take into account the incentivefor exploration provided by a number of factors,including the perception in the industry that therapid declines in reserve levels were dangerousand should be halted if possible, the continuedprofitab ility of mo st large r ga sfields even a t lowprices, and the former inseparability of gas andoil exploration, which a llow ed ga s d isc ove ry to

benefit from exploration incentives provided by

oi I.

‘“American Petroleum Institute, “Quarterly Re\lew of Drilling Sta-

tlstlc5,‘‘

The a rgum ent a bo ut the real ca use o f the

dow nwa rd trend s of pa st de c ade s is d iffic ult toresolve because the opposing sides are generally

arguing less about the data themselves than abouttheir interpretation. Both sides agree, for exam-ple, that onshore gas exploration has become in-c rea singly oriente d to prospec ts with less “ d ryho le” risk but with sma ller reservoirs with poo rer

p rod uc ing c ha rac teristics. Those a rgu ing for re-source depletion believe, however, that this trendhas oc c urred prima rily b ec ause that is the natureof the remaining resource base; those arguing fo r

a mo re o p timistic view o f resource s argue tha tthe trend reflec ts a na tural market respo nse toea rly controlled pric es, rec ent p rice increases,and high d isc ount rates that favor produc tionover inventory. Undoubtedly, both argumentsare valid to some degree; the problem is in deter-mining the relative imp ortanc e of ea c h.

Potential Major Sources of Additional Gas

Small Fields.–One basic argument revolvesaround the question of whether or not a sizableresource—large enoug h to supp ort c ontinued

high rates of production–lies in fields contain-ing 60 BCF of gas or less. The source of the argu-ment lies in the shape of the field size distribu-tion curve.




Historically, the cumulative number of gas and due to past economics; many small finds were

oil fields are distributed according to size in a too small to be economically developed and con-

manner shown in figure 15. In this figure, the size sequently were reported as dry holes rather than

classes 1 through 20 (on the x axis) are scaled added to the historic a l rec ord a s a c lass D or E

so tha t the up per limit o f size c lass 20 is one-ha lf field. Because pipeline gathering systems are re-

the upper limit of 19, and so on. As shown in the quired in order to d evelop ga sfields no m atte r

figure, the cumulative number of fields increases wha t the field size, and a lso b ec ause the p ric ewith decreasing size class as a geometric series, (pe r unit o f ene rgy) o f gas has historic a lly b ee n

down to about size class 13 (or class D in the lower than that of oil, the minimum field size

AAPG no ta tion), and then rap idly leve ls off. At suitable for development is larger—and thus theleast a portion of this “truncation,” or leveling truncation described above is more severe—foroff, of the field size distribution is undoubtedly ga s tha n for oil. The c rux of the c urrent a rgu me nt

Figure 15.—Slze Distribution of Discovered Oil and Gas Fields in the Lower 48 States

20 19 18 17 16 15 14 13 12 11 10 9 8 7 6

300. ’600150. 300

9.38- 18.754.69- 9.382.34- 4.691.17- 2.34.59- 1.17.29- .59.15- .29

5 4 3




is, simply, what will the shape of the field sizedistribution curve look like when the effects ofhigher gas prices run their course? An importantc orollary to this a rgument is, how expensive will

it be to discover and develop these small fields,and, consequent ly, how many of them can ap-

propriately be included in the recoverable gas re-source base?

Proponents of the thesis that small fields rep-resent a very sizable resource argue that the trendobserved for fields larger than size class D— i.e.,a progressive increase in the number of fields dis-covered in each size class as one moves from thelarger field sizes to the smaller–will be continuedinto the small field sizes below class D once thesefields are made the target of intensive exploratoryeffo rts. This a rgument ma intains tha t the ta iling-off of the curve in figure 15 is almost entirely the

result of economics and that there are no geologicreasons for the drop in the number of very smallfields. Sc heune meyer and Drew ,38 in examiningfield size d istributions in the Gulf o f Me xic o a nd

the Denver Basin and at three depth intervals inthe Permian Basin, show that the “truncationp oint” of the field size d istribut ion mo ves to largerfield sizes when exploration and developmentc osts are higher, whic h wo uld be expec ted if thetruncation were economically determined. Also,they note that the po int mo ved to sma ller fieldsizes after gas prices rose and the minimum profit-able field size b ec am e sma ller.

A straightforward argument against the “smallfields the sis” is tha t estima tes of large resourcesfrom sma ll fields c anno t b e b ased on m ore thanan assumption or extrapolation—because no pe-troleum basin has experienced the intensity of

drilling that would be required to find the postu-late d numb er of sma ll fields. This argument ap -

pears to be a powerful one, but it works equallywe ll ag a inst tho se w ho m ight d eny the p ossibil-ity of large numbers of small fields. It probablyis not possible at this time to estimate crediblythe ultimate numb er of sma ll ga sfields rem ain-ing to be discovered in the United States and theresource s the se fields rep resent.

38j. H. Scheunemeyer and L. j. Drew, “A Procedure to Estimate

the Parent Population of the Size of Oil and Gas Fields as Revealed

by a Study of Economic Truncation, ” Mathematical Geo /ogy , vol.

15 , No. 1, 1983.

A second argument that has been presentedis that, in some basins, the field size truncation

does not appear to be generated by economicsand is more likely to have been caused by geol-ogy—the simple lack of sufficient small fields. Forexample, Nehring 39 identifies subduction and

delta provinces, * that account for more thanone-q ua rter of U.S. oil and gas resource s, as anexample of basins where the number of fields inea c h size c ate go ry b eg ins to d rop at a size levelconsiderably above any historical field size min-imum . Nehring a rgue s tha t only a p ortion of U.S.province s ac t ac c ording to Sc heuneme yer andDrew’s thesis and that there are four distinct

groupings of field size distributions, ranging fromone with a rap id increa se in the num be r of f ieldswith decreasing field size (similar to those dis-c ussed by Sc heuneme yer and Drew), to o ne w itha single pe ak at a bout size c lass D, to one with

little increase in the number of fields at field sizesbelow A or B.

A third argument notes that it takes about 1,000class E fields to equal three class A fields, 40 a ndthat even a sharp increase in the number of smallfields discovered may not be of major significanceto the overall resource base. Figure 16 shows theknown field size distribution, as in figure 15, andtwo projected distributions for the ultimatelyrecoverable resource base—one that assumes adoubling’ of the approximately 24,000 fieldsknown as of 1975, with m ost o f the increase a t

the sma ller sizes, and a sec ond tha t a ssumes amuch larger increase at the smaller field sizes,essentially by assuming that the truncation of thenum ber o f fields a t sma ller sizes is entirely an ef-fect of economics and that the actual number offields continues to increase logarithmically with

S9R. Nehring, The Disco very of Significa nt Oil and Ga s Fields in

the United Sta te s, R-2654/ l -USGS/DOE, RAND Corp., january 1981,

pp. 78-94. Excursus, “The Distribution of Petroleum Resources by

Field Size in the Geologic Provinces of the United States. ”

*Subduction provinces are small, linear basins located along theconverging margins of plates. They account for about 11 percent

of U.S. oil and gas resources in the RAND assessment. The three

largest are the San )oaquin, Los Angeles, and Ventura provinces

on the west coast. Delta provinces are small-to-medium sized,circular-shaped, and derived from major continental drainagecenters. The one producing delta province in the United States is

the Mississippi Delta, which accounts for about 17 percent of U.S.

oil and gas resources in the RAND assessment.GOR. Neh~ing, prob lems in Natural Ga s Re~erve, Drilling, and Dis-

covery DdL?, op. cit.




Figure 16.-Known and Projected Size Distributions of Discovered Oil and GasFields in the Lower 48 States

40,000

30,000

20,000

10,000

5,000

1,000

500

100

50

20

020 19 18 17 16 15 14 13 12

Intervals

11 10

1200-3000600-1200300-600150-30060-150

6-60<6

9 8 7 6

150- 30075- 150

5 4 3 2 1

Field size

decreasing field size. * The first p rojec tion pro-duces 48,000 fields, the second about 115,000.Of critical importance is the difference in re-sources between the two projections, all of whicharises from different assumptions about how theexisting truncation of small fields will “ fill in” w ithfuture discoveries; it is about 7 percent of the total

resource base represented by the second projec-*The projected distribution is drawn by assuming that the num-

ber of fields in each size interval smaller than 100 million BOE (0.6

TCF) is 50 percent greater than the number of fields in the next

larger interval.

tion. Extrapolating to the gas resource base (andassuming the “ c entral tende nc y” range of 902to 1,542 TCF of ultimately recoverable resources),the assumption that the ultimate number of smallgasfields found will be much larger than indicatedby the historical field size distribution–might leadto a n increa se in OTA’ s estima tes of p ote ntial ga s

resources of app roxima te ly 60 to 110 TCF.

A fourth a rgum ent note s tha t the sma ll size ofthe fields makes them only marginally economicat best. For gasfields, especially, many of the fields




in the projec ted distributions ma y not b e e c o-nomic at current and projected gas prices andtherefore may not belong in the recoverable re-source base a t this time. * In partial support ofthis a rgum ent, USGS stud ies of the e ffec t o f ga s

price and other economic variables on recover-

able gas resources in the Permian Basin indicateconsiderable sensitivity of the size of the remain-ing resou rce to t hese va ria b les. Ta b le 15 p resen tsestimates of the amount of exploratory drillingthat could profitably be pursued and the gas re-sources that would be discovered by this drill-ing as a function of wellhead price. If the model

used by the study is correct, the size of therecoverable resource in small fields is sharply sen-

sitive to price (also rate-of-return), although the

sensitivity declines at gas prices above $5 or $6per MCF,

New Gas From Old Fields.–Over the lifetimeof a field, from initial discovery to depletion, es-timates of the field’s ultimately recoverable re-source s ge nerally increa se with time as norma l

de velopme nt p rob es the full extent of the fieldand as improved technology and rising pricesbring subeconomic portions of the field into theeconomically recoverable range. * * Although theeffects of improved technology and prices havelong been acknowledged as crit ical for increas-ing oil recovery, gas recovery rates have longbeen considered to be very high under most con-

*in other words, they are subeconomlc resources In the McKelvey

Bo x (fig. 8).* * Reser\e estimates in some fields WIII decrease with time. Small

fields are generally considered to be more susceptible than large

fields to such reserve “shrink age,”

Table 15.—Potential Recoverable Gas ResourcesFrom New Discoveries in the Permian Basin

(assumed 15 percent of return)

Wellhead price Exploration wells New discoveries$/BOE ($/MMBtu)a drilled (thousands) (TCF)

10 (1.50) . . . . . . . . . . 5 4.9815 (2.40) . . . . . . . . . . 12 9.1720 (3.20) . . . . . . . . . . 18 11.3825 (4.00) . . . . . . . . . . 24 13.02

30 (4.80) . . . . . . . . . . 29 14.1235 (5.60) . . . . . . . . . . 34 15.1340 (6.40) . . . . . . . , . . 38 15.81aDollars per barrel of oil equivalent (dollars per million Btu)

SOURCE Geological Survey Circular 828—Future Supply of 011 and Gas Fromthe Permian Basin of West Texas and Southeastern New Mexico, In.teragency 011 and Gas Supply Project, 1980

ditions and thus somewhat insensitive to priceand technology. * Consequently, increases in re-serve estimates from known gasfields were gen-erally considered to be primarily an effect of thenorma l proce ss of exploring for new p oo ls and

ma pp ing the b ound aries of know n p oo ls. This

view is now be ing c ha llenge d , as reserve add i-t ions are being credited to lowering of the aban-donment pressure of depleting reservoirs, to ex-tension of field boundaries into marginal areas

(with low permeability, thin pay zones, or otherconditions adversely affecting economic recov-ery), to well stimulations and well reworking, andto infill drilling to well spacings lower than theold norm of 640-acre spacing (see box G). Forexam p le, from 1969 to 1979, ultimate rec ove ryin the Hugoton-Panhandle field (discovereda round 1920) in Kansas, Oklahoma , and TexasRailroad Commission District 10 increased from

71.0 to 84.0 TCF, * * and ultimate recovery in theBlanco Basin fields (discovered from 1927 to1950) in the San Juan Basin increased from 15.2to 21.7 TCF.41 Although growth rates of knownfields have va ried c onsiderab ly a c ross d ifferent

geographic areas, these substantial increases inknown recovery from quite old fields are well

beyond w hat might have bee n predicted b y thehistoric data on growth of old fields.

Claims ab out the likely future growth of olde rfields have c rea ted substantial c ontrove rsy. Thereare seve ra l areas of d isagreement. For lowered

abandonment pressures, analysts disagree aboutthe effectiveness of regulatory measures designed

to prevent premature abandonment of old wells —e. g., incentive prices for low-production-ratewe lls c a lled “ stripper” we lls. They a lso d isagreeabout when the wells would be abandoned if thec urrent low p ric es d id no t c ha nge. This is impor-

tant because the lower the “old” abandonmentpressure is, the lower is the volume of additionalgas that can be recovered in response to a higherp ric e. For well stimulation and infill drilling, a ma-

jor disag reem ent c onc erns the e xtent to whichadditional gas production from these measures

*However, the mre of recovery IS extremely wn~{t{~ e to thewtactors, a~ is the econornlc threshold of development for a tleld

* *This field IS not considered a single field by all analyst>, nor

are Its reser~e levels corn pletely agreed on. As noted pre~ IOU \l\’,

the~e are not u n[ ommon pro blem~, espec-lall~, with large ~Ield\.

‘t Ibid.

38-742 0 -85 - 6



74 . U .S . Natural Gas Availability: Gas Supply Through the Year 2000

represents new gas that otherwise would nothave been produced, or whether the productioninstead represents merely an accelerated drain-

ing from the field o f ga s that would e ventuallyhave been produced by the existing network ofwells, without infilling or stimulation. And for all

the measures, there are substantive disagree-ments about the gas prices necessary to stir ma-

jor activity, with producers tending to see theneed for higher prices and buyers tending toblame poor markets for current disinterest in pro-duction enhancement measures.

Another complication in gauging the likelyfuture grow th o f olde r fields is the unc ertainty sur-round ing future gas p ric es in these fields. Aside

from the ever-present difficulty of forecastingfuture ma rket p rice s for ga s, it is not c lear to w ha t

extent the currently low prices in the old fieldswill be allowed to rise to the higher market price,or to some other set of defined prices. Although

the existing legislation —the NGPA—does definea specific set of price revisions in 1985 and 1987,it is not c lear whe ther or not new legislation w illsupersed e the NGPA’ s pric e sc hedule. Also, it isnot p ossible to ac c urate ly project a ll of the p ric eeffects of the NGPA, for several reasons. First, thevolume o f reserves in eac h NGPA p rice c ate go ryis not a c c urate ly known. Sec ond , the eventua lprice attained by reserves decontrolled by NGPAwill depend on unpredictable price negotiationsbe twe en p urcha sers and prod uc ers. Third , theeffects of NGPA provisions that provide higherpric es for gas from enha nce d rec overy are not

known because there are inadequate data on eli-gible reserves and because it is unclear how pro-




ducers and pipelines will respond to the variousbureaucratic provisions in the incentive regu-lations.

OTA ha s ca lc ulate d the p otential reserve ad di-tions available from four of the five sources of“new gas from old fields” listed in box G; we did

not estima te reserves from “ extension o f d rillinginto fo rmerly subec ono mic po rtions of fields. ”

We attempted to account for some of the sub-stantial uncertainties by examining two legislativesc enarios and by expressing the e stima tes as arange . The c alc ulations are explained in ap pe n-d ix A a t the end of this report. The t wo legislativescenarios are:

q Scenario 1: Continuation of the existingNGPA with its partial price decontrol and

price escalation provisions; assume that nonew ga s pric ing legisla tion will be pa ssed .

q Scenario 2: Assume that new pricing legis-lation is passed that allows any new gas fromgrowth of old fields to attain full marketprices without regulatory restraints.

We included only those reserves that are not

“new” reserves as defined by the NGPA; roughlyspeaking, these “old” reserves are associatedwith reservoirs that began to produce gas before

passage of the NGPA. Our conclusions, shownin detail in table 16, are that continuation of theNGPA in its present form (scenario 1 ) could even-tua lly ad d at least 20 to 35 TCF to c umulat ive “ oldgas” reserves if market prices reach $3.50 to$4/MMBtu (1983$), whereas changes in the law

Table 16.—Potential Additions to “Old Gas”Reserves From Lowered Abandonment Pressures,

Well Reworking, Infill Drilling, and Well Stimulation(TCF)

NGPA Marketprices

a

Delayed abandonment andwell reworking . . . . . . . . . . . . 5-13 12-31

Infill drilling . . . . . . . . . . . . . . . . 13-20 25-28Well stimulation . . . . . . . . . . . . 2 6

Total . . . . . . . . . . . . . . . . . . . . 20-35b

43-65aAssumption : All additional reserves can obtain market Price of $3.50 to$4.00/MMBtu. In infill drilling, drainage from original well network to new wellsis assumed not to prevent the necessary change in well spacing rules.

bDoes not include reserves added by stripper incentive, production enhancementincentive,

SOURCE Office of Technology Assessment, “Staff Memorandum on the Effectsof Decontrol on Old Gas Recovery, ” February 1984.

that allowed additional reserves to attain the fullma rket pric e (sc ena rio 2) would yield a bout 43

to 65 TCF a t the same p ric e leve l.

The a bo ve rang es d o not reflec t ab solute min-ima and maxima because some uncertainties

were not accounted for. For example, the range

in scenario 1 does not include any contributionto reserve growth from the NGPA price incen-tives for stripper produc tion and for a va riety of

production-enhancement measures. We had noba sis with which t o q ua ntify the e ffec ts of suc hmeasures, although we do not believe them tobe insignificant. Also, the ranges do not includeadditional reserves that might be added by larger-than-expected growth in reserves from new fieldsdiscovered since 1978 and field growth resultingfrom the development of formerly subeconomicareas of fields. Recent annual reserve additionshave reflected a high level of “extensions” and“ revision s,” the two reserve ca teg ories thatwo uld a bsorb new reserves from this latter typ eof d eve lopm ent. This might b e a signal tha t thereis a substantial potential from this source of fieldgrowth.

The num erica l results reflec t severa l inte rmed i-ate c onc lusions of OTA’ s ana lysis. In p a rtic ular,we discovered that the NGPA allows a substan-tial amount of old gas to attain higher prices. Atleast one-third, and probably more, of U.S. “oldgas” reserves will attain substantial price increasesin 1985 or 1987 under the NGPA, and these in-

c reases will trigger signific ant leve ls of reservegrowth. Also, as a result of extensive discussions with geologists and petroleum engineers, we con-

cluded that infill drilling and well stimulation canadd m od erate ly to gas rec ove ry in most fields.A considerable portion of this infill potential willnot occur until available gas prices reach $3.50to $4/MMBtu, but some fields will be developedat the current infill incentive price of about $2.85/ MMBtu. Unfortunately, we were not able to re-solve d isagreem ents about the va lid ity of alter-nate estimates of abandonment pressures, andc onseq uently the rang es of e xpe c ted reserve

growth reflect both extremes.

New Frontiers, Including Deep Areas.–Eventhough recent exploratory drilling in the frontier

areas has had mixed success and several severedisappointments, considerable areas of untested




or inadequately tested sedimentary rock remainthat ma y hold c onside rab le po tential. Even e x-treme pessimists view areas such as the deep-wa ter Gulf of Me xic o, the Ana darko Basin, andthe Western Overthrust Belt a s ha ving c onside r-

able potential. However, it is also inarguably true

that areas such as the Gulf of Alaska, eastern Gulfof Mexico , the Southe ast Geo rgia Em b aym ent,and the Baltimore Ca nyon have be en e xpensivefailures42 thus far. Unfortunate ly, it is no t e asy to

document the opinions of the major oil compa-nies—who traditionally are leaders in frontier ex-ploration– because few details of their most re-c ent resource assessments a re ava ilab le to thepublic. It is clear, however, that some of the ma-

jors, notably Exxon and Shell, are pessimisticabout the overall Lower 48 potential and theLower 48 onshore frontier areas. Given thespecuIative nature of these resources, the range

of credible estimates of frontier undiscovered gasmust b e c onside red quite wide.

An important part of the controversy over theresource pote ntial of frontier a rea s involves theeconomic viability of the potential resourcesra ther tha n their physica l presenc e. For exam p le,much of the intense deep drilling activity of theearly 1980s in basins such as the Anadarko ap-pe ars to ha ve b een a direc t respo nse to the veryhigh prices for deep gas (as much as three timesthe market-clearing price) resulting from the

price-controlled market. Prices for deep gas andother categories of gas entitled to special in-centive pricing under NGPA have now droppedsharply, and drilling activity has dropped sharplyas well. Consequently, some analysts questionwhether these expensive resources still belongin the economically recoverable resource base.Similar questions have arisen over some of thegas under the deep waters of the continentalslope, now included in the USGS assessment andothers.

A c ounterpoint to p essimism a bo ut the e c o-nomic viability of the deep resource is the recentdrop in drilling costs in response to the dropoff

in ac tivity. Althoug h c osts prob a bly w ill rise whe nthe oversupply in rigs eases, the costs during thedril l ing boom period probably did not reflect the

——‘2R. Nehrlng, “The Discovery of . . .,” op. cit.

eq uilibrium p rod uc tion costs tha t should b e theb asis for de fining the rec ove ra ble resource . Also,deep drillers argue that their growing experienceand the rapid advancement of drill ing technologywill overco me much of the effect of the price drops.

The a pp rop riate plac em ent o f these resources

inside or outside of the recoverable resource baseis complicated by several factors. First, uncer-tainty ab out the p rec ise g eolog ic c ond itions ofthese resources combined with the recent rapidfluctuations in drilling costs create substantialuncertainty about the cost of producing the re-source s using tod ay’s tec hnolog y. Sec ond , the

present hesitancy of the industry to drill for theseresources may not necessarily reflect the re-sources’ lac k of long-term ec onom ic viab ility b utrather the current lac k of gas de ma nd a nd regu-latory unc ertainties about dec ontrol. Third , un-certainty is added by ambiguities in the common

definitions of “recoverable resource base,” someof which, e.g., allow the possibility of techno-logical improvements that are in line with trendsprevailing at the time of the assessment (this is

USGS’s bound ary cond ition). This greatly com -plicates the evaluation of resources whose pro-duction may involve technological difficulties. Be-c ause o f these fa c to rs, in OTA’ s op inion theboundary between economic and subeconomic,and consequently the magnitude of the recov-erable resource base, is not well defined for thefrontier resources.

Stratigraphic Traps.–Over the cycle of gas ex-ploration, structural traps have tended to be themost favored drilling prospects. As possibilitiesfor find ing ne w large struc tures have d ec lined ,many explorers have shifted their strategy towardlocating subtle stratigraphic traps, i.e., potentialreservoirs whose m ain trapp ing m ec ha nism is agradation of the reservoir rock into layers of rockof low p ermeab ility Iaid d own by the sed imen-

tation process. Resource optimists expect to findlarge amounts of resources in these traps.

There a re two ma jor a rgum ents aga inst suc h

expe c ta tions. First, the re have bee n sign ific antpast efforts aimed at finding stratigraphic traps,especially in the Anadarko, Permian, Denver, andPowder River basins.43 Sec ond , it is a rgued : 1)

‘d]l~id




that very large stratigraphic gasfields are unlikelyto have remained undiscovered in the exploredbasins of the Low er 48 Sta tes bec ause o f thef ields’ large areal extent and the very high den-sity of drilling in these basins; and 2) that mostof the stratigraphic traps remaining to be discov-

ered will be sma ll. Nehring44

also cites geologicarguments against the possibility of finding manyla rge new stratigraphic trap s, inc lud ing the vul-

nerability of such traps to degradation or dissipa-tion and Nehring’s contention that the presenceof multiple structural trapping possibilities inbasins outside of the stable interior provinces

ma kes it unlikely that ma ny stra tigra ph ic trap s willexist outside of these provinces, the source ofmo st p ast d isc overies.

These a re strong but not conc lusive a rgum ents.New efforts to locate stratigraphic traps can use

seismic e xp loratory tec hniques not ava ilab le tothe ea rlier effo rts. It is possible, tho ugh spec ula-tive, tha t severa l sizab le trap s tha t were “ invisi-

ble” to earlier techniques could now be located.Simila rly, a rguments about d rilling density a revalid but must be tempered by the depth limita-tions of much of this earlier drilling and theclustering of such drilling around areas consid-ered p rospe c tive by e a rlier stand ards.

Even if the arguments against finding largestratigraphic traps are correct, there remains sig-nif icant uncertainty about the number of smallerfields that might exist and the actual potential forfinding and exploiting these fields–the sameuncertainty that affects assessment of the resourcepote ntial of sma ll fields in g ene ral. Key fac torsaffecting the potential for producing signif icantquantities of gas from these fields include gasprices and reductions in the costs of effective ex-ploration techniques.

Conclusions

Based on the p revious d isc ussion, OTA a c c ep tsthe possibility that discovery trends may have

been sufficiently distorted by past regulatory andeconomic conditions and that sufficient resourcepossibilities exist in small fields, growth of oldfields, and other sources to allow us to accept

the estimates of PGC as possible, but very opti-mistic—a reasonable upper bound to the prob-

a41bid.

able magnitude of the conventional gas resourceba se. On the o ther hand, we c onside r the e x-tremely pessimistic estimate of Hubbert to beunlikely, and to a lesser extent we are also skep-tica l of the RAND estima te. Both c om e ve ry c loseto the analogy of “running into a brick wall.”

Looking ahead to chapter 5, we can see that theHubbert estimate implies a “ c onve ntional ga s”produc tion rate o f ab out 3 or 4 TCF in 2000, ana stou nd ingly low va lue. The RAN D estima te im-plies tha t there will be a t be st only a ha nd ful ofnew exploration p lays in the Low er 48, that thesewill be on ly modera te in size (2 to 10 TCF), and

tha t the re will be no rea lly la rge “ surprises” left;we believe this is possible, a lthoug h q uite p essi-

mistic. However, the RAND assessment appearsto have underestimated the potential for reservegrowth in known fields, and it apparently has ex-c luded som e ga s in low -permea bility reservoirs

tha t is c urrently ec ono mic ally reco verab le. There-fore, we consider a credible lower bound to besomewhat higher than the RAND estimate.

In c onc lusion, our best g uess–and we c hosethis word carefully—is that a reasonable range forthe magnitude of remaining conventional natu-ral gas resources, recoverable under technologi-cal and economic conditions not far removed fromtoday’s, * is about 430 to 900 TCF as of the endof 1982. This is not really a very wide range, giventhe basic uncertainty associated with resource as-

sessment, but it is a wide range with respect to

future p rod uc tion po tential. The tw o en ds of therange have very different implications about howdifficult it is going to be to continue to replenishour current inventory of gas reserves over thenext decade or two, and they have profound im-plic ations ab out w hat the role o f natural ga s inour ene rgy e c ono my w ill be in 2000. OTA b e-

lieves that if the lower end is correct, reserve ad-ditions will fall off drastically within a few years,with prod uc tion ra tes d rop ping in response. On

the other hand, the upper range implies the po-tential for a very positive future for conventionalga s prod uc tion d uring this c entury. The next

chapter explores these production issues ingreater detail.

‘Including gas in low-permeability reservoirs that otherwise sat-

isfies the conditions. This recognizes the ambiguous boundary be-

tween “conventional” and “unconventional” gas in such reservolr~.



Chapter 5

Gas Production Potential



Contents

Page

Approach Number 1–Projecting Trendsin the Individual Components of

Reserve Additions. . . . . . . . . . . . . . . . . . . 82

New Field Discoveries . . . . . . . . . . . . . . . 82Extensions and New Pool Disc overies . . 86Revisions . . . . . . . . . . . . . . . . . . . . . . . . . . 90

Reserves-to-Production Ratio ....., . . . . 93Prod uc tion Scena rios . . . . . . . . . . . . . . . . 96

Approach Number 2—Projecting NewPool Disc ove ries, Extensions, and

Revisions as a Single Growth Factor . . . . 98

Approach Number 3–Region-by-RegionRev iew of Resource s andExp lora to ry Suc c e ss . . . . . . .. . . . .. . . . .100

Approach Number 4-Graphing the

Complete Production Cycle. . .........103

A Range for Future Gas Production .. ...,105

Pub lic and Priva te Sec to r Forec asts ofFuture Gas Production . . . . . . . . . . . . . . .106

TABLESTable No. Page

17. Additions to Lower 48 Natural GasProved Reserves: Extensions and New

Pool Disc ove ries, 1966-83 (BCF) . . . . . 88

18. Summa ry of Projec tions ofComp one nts of Reserve Ad d itions andRaps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

19. Low er 48 Sta tes Natural Ga sProduction and Reserves 1981-2000(in TCF) SCENARIO 1A: Op tim istic

Exploration, Rapid Production . . . . . . . 97

20. Low er 48 Sta tes Natural Ga sProduction and Reserves 1981-2000(in TCF) SCENARIO 1B: PessimisticExplorat ion, Slowed Prod uc tion . . . . . . 97

21. Arguments for the Que stion, “ Will the

Reserve Growth in New Fields Be

Larger o r Sma ller Tha n the Grow thRecorded in Previously DiscoveredField s?” . . . . . . . . . . . . . . . . . . . . . . . . . . 99

22. Alternative Estimates of GrowthFac to rs for Initia l Reserve Estima tes for

Gasfie lds . . . . . . . . . . . . . . . . . . . . . . . . . 99

23. Low er 48 Sta tes Natural Ga sProduction and Reserves, 1982-2000

(in TCF)–Sc ena rio 2A: Ve ry Op timistic 10024. Low er 48 Sta tes Natural Ga s

Prod uc tion a nd Reserves, 1982-2000(in TCF)–Sc ena rio 2B: Pessimistic .. ..101

25. Low er 48 Sta tes Natural Ga sProd uc tion a nd Reserves, 1982-2000(in TCF)–Sc ena rio 2C: Ve ryPessimistic ., . . . . . . . . . . . . . . . . . . . . .101

26. Lower 48 States Natural GasProd uc tion a nd Reserves, 1982-2000(in TC F)–Sc e na rio 3A . . . . . . . . . . . . . .102

27. Lower 48 States Natural GasProduction and Reserves, 1982-2000(in TC F)–Sc e na rio 3B . . . . . . . . . . . . . .102

28. A Comparison of Conventional Lower48 Natura l Ga s Sup p ly Forec asts (TCF) 107

FIGURES

Figure No. Page

17. Natural Gas Production From 1981Lower 48 Proved Reserves . . . . . . . . . . 81

18. Additions to Lower 48 Natural GasProved Reserves: New Field WildcatDisc ove ries, 1966-83 (BCF) . . . . . . . . . . 84

19. Additions to Lower 48 Natural GasProved Reserves: Extensions and NewPool Disc ove ries, 1966-81 (TCF). . . . . . 87

20. Additions to Lower 48 Natural GasProved Reserves: Revisions AS .

Reported, 1966-83 (BCF) . . . . . . . . . . . . 9121. Additions to Lower 48 Natural Gas

Proved Reserves: Revisions AsAmended, 1966-83 (BCF) . . . . . . . . . . . 92

22, Reserve-to-Production Ratios forNatura l Gas in the Lowe r 48 Sta tes . . . 95

23. The Grow th of Yea r-of-Disc overyEstimate s of the Amo unt o f

Recoverable Natural Gas Discoveredin the Lower 48 States . . . . . . . . . . . . . 98

24. Future Production Curves forConventional Natural Gas in theLo w e r 48 Sta te s . . . . . . . . . . . . . . .. . . .103

25. Two Prod uc tion Futures, OneResource Base: Alterna tiveRepresentations of Future Productionof Conventional Natural Gas in theLowe r 48 Sta tes, Based on the USGS(1981) Resource Assessment . ........104



Chapter 5

Gas Production Potential

There are a variety of a lternative a pp roa c hesto e stima ting the future ga s produc tion po tentialof the United Sta tes, inc lud ing the use o f com -

plex computer programs using econometric,process engineering, or system dynamics ap-proaches to model separately the gas exploration,development, and production processes. Al-tho ugh during the c ourse o f this stud y OTA ex-amined several complex models in detail, wehave c hosen to use four relatively simple tec h-niques to projec t future prod uc tion p ote ntial. Thisapproach reflects the high costs of using the com-plex models and some doubts we entertain aboutthe expected accuracy of these models as fore-c asting to ols. These d oub ts do no t nec essa rily ex-tend to the usefulness of the models as policyana lysis tools; often, these m od els offe r the va l-uable ability to test alternative policies undercarefully controlled conditions.

Of the four app roa c hes used by O TA to project the mid- to long-term (1 990 and beyond) pro-duction potential for natural gas in the Lower 48Sta tes, three foc us spe c ifica lly on t he p ote ntialfor continued add itions to U.S. prove d reserves.The addition of new reserves to the U.S. gas sys-tem is the primary determinant of future gas avail-ability. The imp ortanc e of ne w reserves c an beillustrated quite simply by drawing the produc-

tion that would likely result from the failure toadd to reserve levels and reliance instead on cur-rent proved reserves as the sole “inventory” for

production to draw on (see fig. 17). Assuming aconstant reserves-to-production (R/P) ratio of 8,0,beginning in 1984, production would immediate-ly begin to drop with shocking rapidity to about2 trillion c ub ic feet (TCF) by the end of the

century. *

Of the three approaches focusing on continuedadd itions to U.S. proved reserve s, the first p ro-

jects future reserves by examining historical

trends in all components of reserve additions

“Concewably, the [nltlal reduction In production could be slowed

by drllllrrg additional development wells, eftectlvely lowering the

R/P ratio, The end result ot this strategy would be, however, an

even more rapid product ion collapse occurring a few years later

than that shown in fig. 17,

Figure 17.–Natural Gas Production From 1981

s

1982

s

Lower 48 Proved Reserves

s

1985

R/P = 8.0 beginning in 1984

1990 1995 2000

YearSOURCE: Office of Technology Assessment

(new field discoveries, extensions, new pool dis-coveries, and revisions), examining the under-lying causes of the trends, and extrapolating intothe future b ased on O TA’ s expe c tations of futureconditions. I n this extrapolation, we have drawn

heavily on the insights gained in our examina-tion of g as resou rce base a ssessme nts. The sec -

ond ap proac h projects only new field d isc overiesand then applies a “growth factor’ to these dis-coveries based on historical experience with thegrowth o f new fields and OTA’s judg ment ab outhow the g rowth rate ma y have c hang ed . Thethird approach is based on a geologist’s * region-by-region examination of available gas resourcesand past exploratory success. In all three cases,production rates are calculated from reserve databy p rojec ting future levels of the R/ P rat io.

The fo urth a pp roa c h b orrows a m ethod used

by M. King Hubbert in 1956,1 tying future pro-

*joseph P. Rlva, jr., of the Congressional Research Service.

1 Described I n M, K, H ubbert, ‘Tee h nlques of Production as Ap-

plled to the Product Ion of 0!1 and Gas, ” 0/ / and Ga s Supp / y ,A40d-e/ fng , S, 1. Gass (cd.), National Bureau ot ‘jtandards Special Pub-

Iicatlon 631, May 1982.

87




duction directly to available resources by draw-ing freeform p lots of the c omp lete na tural ga sproduction cycle in such a manner that the cu-mulative production conforms to existing re-source base e stima tes—in this c ase, to the esti-ma tes of Hubb ert, the U.S. Geo log ica l Survey

(USGS), and the Potential Gas Committee (PGC).

In ea c h of the four ap proac hes, rang es of p ro-duction potential are estimated based on alter-native assumptions about the magnitude of theresource base, efficiency of the exploratory proc-ess, and other factors.

OTA’ s use o f four ap p roa c hes, and alternativeassumptions within the approaches, reflects ourskept icism o f our own and othe rs’ ab ility to p ro-

jec t future gas prod uc tion ra tes with any p rec i-sion. A “most likely” or “best” projection was

deliberately avoided bec ause w e b elieve tha t

such a projection, beyond 5 years or so into thefuture, would be futile. Our purpose in this sec-tion is to illustrate the plausible range of p ossi-ble future production rates and the general effectson production estimates of different interpreta-

tions of the c auses of p ast trend s and d ifferentassump tions abou t future c ond itions. The first a p -proach is our slight favorite, but only because itslevel of d isag grega tion forc es the a nalyst to de almore explicitly with the underlying causes of pastevents. This app roac h is d isc ussed in the g reat-

est detail.

At the end of the c hap ter, a va riety of ga s pro-duction forecasts by public agencies, privatec om pa nies, and institutions a re p resente d anddiscussed.

APPROACH NUMBER 1–PROJECTING TRENDS IN THEINDIVIDUAL COMPONENTS OF RESERVE ADDITIONS

The first a p proa c h sep arate ly p rojec ts trends inreserve a d ditions from new field d isc ove ries, newpool discoveries and extensions, and revisions.

New Field Discoveries

The discovery of new gasfields represents the

single most important force necessary for buildinga sustainable natural gas supply because a newgasfield not only adds to current reserves but alsoprovides a source of considerably larger additions

to future reserves through field grow th a fter thediscovery year. Reserve additions attributable toextensions and new pool discoveries and, to anextent, to revisions, are all, in fact, the inevita-ble c onseq uenc e o f previous new field d isc ov-eries. Therefo re, if new field d isc overy rates in-crease or decrease, then at some point in the nearfuture, reserve additions from extensions and new

pool discoveries will almost certainly increase or

decrease in a like manner.

Factors Affecting New Field Discoveries

The rate of annual additions to reserves fromnew field discoveries depends on a variety of fac-

tors, but most importantly on:

q

q

The undiscovered resource base. –The phys-ical nature of the resource base—including

the amount of resources remaining to befound, the distribution of field sizes, the loca-tions of fields, the distribution of types of geo-log ica l trap s (mo re or less diffic ult to p inpointwith available exploration techniques), andother physical attributes—is considered bysome to be the single most important deter-minant of future new field discoveries.Exploration technology .-The rapid advanceof exploration technology, for example,computer-aided seismic technology, affectsdrilling success rates and, consequently,ove ra ll d isc ove ry rates, Also, tec hnolog ica l

improvements have op ened up to c omme r-



Ch. 5—Gas Production Potential 83

q

q

q

q

c ial exploitation som e areas whose c omp lexgeology had previously prevented accept-able success rates. Consequently, these im-provements have expanded the recoverableresource base. Development of the WesternOverthrust Belt is an imp ortant exam ple of

this effect.Drilling and production technology .-lm-provements in production technology createan expanding recoverable resource baseand, in turn, an increase in targets for the

drill. For example, massive hydraulic fractur-ing technologies allow exploitation of fieldsin sands of low permeability that previouslywould have bee n subec onomic. improve-ments in o ffshore d rilling tec hnolog y a llow

exploitation of gasfields in deeper and morehostile wa ters.

Current and perceived future gas prices and

other economic variables. -Such variables af-fect the propensity to drill and determinewhere to draw the l ine between a produci-ble w ell and a d ry hole. In som e c ases, thehigher prices allow the use of well stimula-t ion te c hnique s that w ould otherwise b e to oexpensive, allowing successful production tobe a c hieved from a well that wo uld o ther-wise ha ve ha d to o low a flow rate . Addition-ally, the minimum acceptable reservoir sizefor p roduc tion ha s grown sma ller. The rela-tive prices of gas and oil are important alsobecause these will determine whether drill-

ing will be preferentially a imed at targetswhere gas or oil are more likely to be found.Schedules, financial terms, and other aspects

of leasing. –These a lso d etermine the num -ber of attractive targets available for ex-ploratory activity.

industry willingness to take risks.-All of theabove factors and others play a role in deter-mining the propensity of the exploration seg-ment of the industry to assume the risks ofwildcat drilling in unproved areas where

much of the gas resource potential is thought

to exist. Because this type of drilling often

involves hostile environments and large cap-ital requirements, much of this drilling is thedomain of the major integrated oil com-panies and the large independents. Conse-quently, those factors that strongly affect the

cash flow, capital availability, and economicinc entives for this group of c om panies a repa rticularly likely to a ffec t the ind ustry’s pro-pensity for risk-taking.Historic prices and exploratory experience.

–There a lways exists an invento ry of new

field p rospec ts known to explorers throughpa st e xplorato ry a c tivities but u nd rilled or (inthe c ase of “ dry holes”) unc omp leted be-cause of economic conditions or the avail-ability of more promising prospects else-wh ere. The key dete rminant s of the size a ndcharacter of this inventory are past explora-tory expe rienc e and p ric e p rofiles. Thena ture o f the invento ry is, in turn, an imp or-tant determinant of new field discovery ratesin the short term, especially during the

period following a change in price levels orreg ula tory c ontrols.

Historical Variation of New Field Discoveries

During the 14 years of American Gas Associa-tion (AGA) data availability, * the annual additionsof ne w field d isc overies in the Low er 48 Sta tesremained fairly steady, if somewhat cyclic, vary-ing b etw ee n a high of 2.9 TCF and a low o f 1.3TCF. (See fig. 18, which a lso shows the Energy

Information Administration (EIA) data for 1977 to1983.) From 1967, the last ye ar in w hich AGAestimated that total reserve additions exceededproduction, until 1979, the average of new fielddiscoveries was 2.0 TCF. Similarly, nonassociatednew field d isc ove ries we re e qua lly stea dy, witha 14-year averag e o f 1.7 TCF. Conseq uently, newfield discoveries played a surprisingly small di-rec t role in annua l reserve a dditions during this

period; * * they ave rag ed less than 20 perce nt ofall annual additions from new discoveries and ex-tensions and never exceeded 25 percent in anyyear.

Although the reserve additions reported as newfield discoveries remained steady during this

period, the size distribution of the fields dis-covered did not. As shown earlier in table 12, the

*Actually, AGA compiled reserve additions data since 1947, but

only began separately estimating new field discoveries in 1967.

**Clearly they did not play a small indirect role since many ofthe new pool discoveries and extensions in this period represented

development of the fields discovered earlier in the period.




Figure 18.–Additions to Lower 48 Naturai Gas Proved Reserves: New Field Wildcat Discoveries, 1966.83 (BCF)

1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983

Reporting year

SOURCES: Office of Technology Assessment, baaed on data from Energy Information Administration, U.S. Crude Oil, Natural Gas, and Nafural Gas Liquids Reserves— 1983 Annual Report, DOE/EIA.0216 (83), October 1964, and American Petroleum Institute, American Gas Association, and Canadian Petroleum Association, Re-

Serves of Crude 0il, Natural Gas Liquids, and Natural Gas in the United States and Canada as of December 37, 1979, vol. 34, June 1980.

ave rag e size of new ga sfields be c am e c onside r-ably smaller (reported year-of-discovery reservesof 1.85 billion cubic feet [BCF] per successful newfield wildcat in 1979 v. 18.56 BCF per successfulnew field wildcat in 1966). Furthermore, thelower average did not imply only a reduction ind isc ove ries of g iant fields; althoug h this did oc -cur, another change involved a very large in-crease in the number of very small class E fields*brought into production.

Because of the smaller size of newly discoveredfields, a steady expansion of successful explora-tory wells was required just to maintain the ratherlow annual discovery rate of the period. For ex-am ple, co mp letions of new field (ga s) wildc atsin the onshore Lower 48 increased from 126 in1968 to 671 in 1979. Bec ause o f the sub sta ntia limprovements in success rates (see fig. 14) for allnew field wildcat drilling (from 8.5 percent in1968 to 19.0 perce nt in 1980), howe ver, ac tua l

d rilling ra tes d id not have to increase in prop or-tion to the rate of completion. From a low of4,463 wildcat wells in 1971, drilling reached

6,413 wells in 1979 and 8,052 in 1981.

*Class E fields contain less than 6 billion cubic feet of recoverable

gas.

The mo re rec ent (1977 to 1983) EIA ne w field

discovery data (fig. 18) show considerable year-to-year variation w ith no ap pa rent trend asidefrom the rece nt d rop assoc iate d with the c urrentpoor market, and are made even more diff icultto interpret because of the break with the AGAdata series. How eve r, the EIA estima tes of newfield d isc overies we re highe r during 4 of the first

5 years of record than any AGA-recorded discov-

ery rate from 1966 to 1979. Of interest is thesource of these discoveries. Although areas like

the Western Overthrust Belt and deep AnadarkoBasin ha ve b een in the forefront o f med ia c ov-erag e, most new field d isc ove ries c ontinued tocome from more tradit ional gas-producing areas —onshore and offshore (Gulf of Mexico) Loui-siana a nd Texa s. For examp le, during bot h 1980and 1981 these tw o Sta tes p rovide d two -thirds

of the total magnitude of reserve additions fromnew field discoveries in the Lower 48.2

ZEIA, (J. s. cru& Oil, Natural Gas, and Natural Gas Liquids Re-

serves, 1980 and 1981 Annual Repotts, DOEIEIA-0216 (80) and (81),October 1981 and August 1982.



Ch . 5—Gas Production Potential q 85

Implications

The following key issues pe rtaining to new fielddiscoveries remain essentially unresolved:

qCa n the 1968-79 trend i n new field d isc ov-eries—essentially a steady cycling around anaverag e of 2 TCF/ yr or so—be c ontinued we llinto the future? If op timists ab out the g a s re-source po tential of sma ll fields a re c orrec t,a continuing strong exploratory drilling ef-fort should be able to maintain this level fora number of years. If the pessimists aboutsmall fields—and about the remaining re-

source base in general—are correct, reserveadditions from new field disc ove ries mightdrop within a few years.

qDo the highe r EIA e stima tes of 1977-82 newfield discoveries represent an actual increase,

or are they the result of the change in report-

ing methodology? Does the EIA methodol-ogy place more of a newly discovered field’s

ultimate reserves into the first year reserveestimate , leaving less roo m for sec ond ary(extension and new pool wildcat) discover-ies? If the EIA values represent a true increasein new field discovery rates, the sustainabilityof a high rate (p erhaps 3.5 TCF/ yr) of ne wf ield discoveries would seem to depend ei-ther on the availability of new giant fields oron extremely high rates of exploratory drill-ing and the availability of massive numbersof sma ll fields, supported by either or bo th

strong pric e ince ntives and c ontinued im-provements in exploration technologies (es-pecially in terms of lowering the cost of

detailed geological surveys).

The c om p arison of the three ov erla pp ing yea rsof EIA and AGA data in figure 18 is tantalizingbec ause the d ifference betw een the two da ta setsis c onside rab ly sma ller in 1979 tha n in 1977 and1978, and the EIA methodology changed in 1979.Some analysts have chosen to use AGA data un-til either 1978 or 1979, and EIA data thereafter,assureing that the two series are essentially con-tinuous. However, the coincidence between the1979 EIA and AGA estimates for new field dis-co veries may b e a n ac cident; the two da ta setsdiffer considerably for all of the other reserve ad-d ition c ateg ories in 1979.

The fa ilure to resolve the above issues imp lies

that a c red ible range for future ne w field d isc ov-ery rate s would be quite wide. Although d efin-ing the range is a matter for subjective judgment,

OTA wo uld put the rang e a t a bo ut 1.5 to 3.5TCF/yr for the next 10 to 15 years, assuming thatexploratory drilling remains active for the period. *

The range for the ne xt 2 or 3 years should b e na r-rowe r, howe ver, perhaps 2.0 to 3.5 or 2.5 to 3.5TCF/ yr. The rea soning for these jud gme nts is asfollows:

q

q

q

The high e nd of the rang e for the imme di-

ate future is ba sed on the d istribution of ne wfield sizes. Bec ause the rec ent high d isc ov -ery rate did not depend on discoveringgiant fields—notoriously erratic occurrences –but on e mp loying a very large numb er ofexploration te am s to disc over ma ny med i-um-sized and small fields, the physical ability

of the system to m a inta in its rec ent ne w fielddiscovery rates should logically be quite high

unless the ga s “ bub ble’ ’ -and the c urrentslump in drilling and all other exploratoryactivity —conti nues.To o b tain the low er end of the 10-to 15-yearrange, we assumed that the 1970s AGA data

more accurately reflect the likely future andthat continuing resource depletion will leadto p oorer prospe c ts and a slump from the

averag e ‘ of 2.0 TCF/ yr during tha t p eriod .Also, it was assumed that the major reasonsfor the highe r EIA va lues a re m etho dolog i-

ca l and d o not reflec t an actual increase overdiscovery rates reported by AGA. Conse-q uen tly, the 1.5 TCF/ yr reflec ts AG A c onven-tions and probable followup field growth.The d isc ov ery levels ac tua lly recorded by EIA

would be expected to be higher than this val-ue, but the reserve growth caused by exten-

sions and new pool tests would then belower than would be predicted by pre-EIAhistorical experience.

The h igher end of the 10- to 15-yea r rangeassumes that the EIA data accurately reflect amajor upward shift in the finding rate (volume

*Drilling is now in a substantial slump. The ranges of reserve ad-

ditions discussed here would be unrealistically high if the current

“bubble” in gas deliverability and the related difficulties in mar-

keting new gas were to continue.



86 q U.S. Natural Gas Availability: Gas Supply Through the Year’2000

of gas discovered per unit of exploratory ac-tivity). Additionally, it is assumed that con-

tinued improvements in exploration and pro-duction technologies allow further increasesin finding rates and/or that exploratory drill-ing rate s a re inc reased . This end of the rang e

is a ligned with a large resource ba se.

Extensions and New Pool Discoveries

As a lrea dy note d , a new field is ge nerally notsufficiently defined in its year of discovery toallow the “new field discoveries” portion of re-ported reserve additions to represent all or mostof the actual recoverable resource in that f ield.In the yea rs follow ing d isc ove ry, ad d itiona l ex-

ploratory wells are drilled to delineate the full ex-tent of the resources present in the field. Wellstha t p rob e the bo undaries of reservoirs or fields

in order to establish their productive area arecalled extension wells or extension tests. Wellsthat search for additional reservoirs within alreadydiscovered fields are called new pool tests or newp oo l wild c ats. The reserve a d ditions from exten-sion we lls and new poo l wildc a ts rep resent theresults of a secondary or followup discovery proc-

ess for new fields.

Factors That Affect Extensionsand New Pool Discoveries

As with new field discoveries, the major deter-minants of extensions and new pool discoveries

are the mag nitude a nd na ture of the “ target” ( inthis case, not the undiscovered recoverable re-source base, but only that portion of the remain-ing resource associated with discovered fields),the technology available to find the gas, and the

nature of the incentives to drill:

q The target. —The “ resourc e b a se” for exten-sions and new pool discoveries is the inven-tory of discovered but incompletely deline-ated fields. Limited data from the late 1960sand 1970s indicate that the major part of newfield g row th ha s oc c urred within the first 5

years after discovery. Consequently, unlessincentives for gasfield development are lack-

ing, * the magnitude of extensions and new

*The current gas “bubble” provides a disincentive for field de-

velopment.

q

p oo l disc ove ries should be strong ly and posi-tively tied to recent new field discoveries. Ad-ditionally, measures that increase currentnew field discoveries should soon lead to in-creases in extensions and new pool discov-eries as the new fields are further developed.

Aside from the total gas volume repre-

sented by the “target,” that is, the inventoryof discovered fields, the geological charac-te ristics of the fields will a lso p lay a n imp or-tant role in determining future extensionsand new pool discoveries. For example, old-er f ields that were incompletely developedbecause a substantial portion of their in-

ground resource was subeconomic* at thetime of d isc overy are now go od targets fornew exploratory effo rts. The size a nd c om -plexity of newly discovered fields will par-tially determine the relationship between the

initial year-of-discovery reported reservesand the late r extensions and ne w p oo l d is-coveries that signify further development ofthe fields. Because the discovery wells ofsmaller, less complex fields can generally“ prove” a high p erc entage of their total re-source, these fields may offer less oppor-tunity for this later development than was thecase with the generally large, complex fieldsof earlier decades.Tec hnology.-The sam e tec hnologica l fac torsthat affect new field discoveries affect exten-sions and new pool discoveries. Computer-

assisted seismic technology is consideredespecially important in allowing extensionwe lls and new po ol tests to b e d rilled with

high success rates. Fracturing technologies,by opening up previously uneconomic reser-voir margins and tight reservoirs in alreadydiscovered fields, expand the target resourceavailable.

Advancements in exploratory technologyhave other, varied effects, however. For ex-ample, advanced seismic techniques, by of-fering a highly accurate picture of the poten-

tial of new fields shortly after discovery, may

encourage a larger proportion of the ultimaterec overab le g as to be drilled and “ proved”

“Because of the small size or low permeability of the reservoirs

or the low quality of the gas.



Ch. 5—Gas Production Potential q 87

in the initial year-of-discovery, leaving less

room for followup discoveries. Advanced

seismic techniques also may help compressthe remaining field delineation into a shorterspa n of time , lea ding to inc rea ses ove r ex-pe c ted levels in extensions and new po ol

discoveries* for a few years after discoveryof the field, followed by a later decrease inexpected levels of these reserve additions.

Historical Variation of Extensionsand New Pool Discoveries

Figure 19 and table 17 illustrate the variationof e xtensions and new po ol d isc ove ries** in theLower 48 States from 1966 to 1983. Extensions

*That is, increases over the discovery rates projected by usinghistorical data.

* *As noted previously, new pool discoveries are reported as “newreservoir discoveries in old fields” in the AGA and EIA reservereports.

have c onsistently played the ma jor role in tota lreserve a dd itions. After de c lining in the mid t ola te 1960s, they rem a ined stab le a round 6,000BCF/yr from 1969 to 1976 and began to move

upwards thereafter. As with the other categoriesof reserve a dd itions, the shift to EIA d a ta c om -

plicates an interpretation of the past few years.According to that data, however, extensions by

themselves prod uc ed reserve a d d itions of 10 TCFin 1981, eq ua ling or surpa ssing total reserve a d -ditions in most years of the 1970s.

Som e o f the un de rlying c au ses of the se trend sma y be understood by examining the trends in

extensions of individual PGC reporting areas. 3 Ex-tensions tend to b e c onc entrated in only a fewof these areas. Before 1968, field growth (primar-ily in relative ly de ep fields) in the Permian Basin

3From R. Nehring, “Problems in Natural Gas Reserve, Drilling,

and Discovery Data, ” contractor report to OTA, 1983.

Figure 19. -Additions to Lower 48 Natural Gas Proved Reserves: Extensionsand New Pool Discoveries, 1966-83 (BCF)

14,000 f

13,000t

9,000 -

8,000 -

7,000 -

6,000 -

5,000 -

4,000 -

3,000 -

2,000 -

1,000 -

r ) -1970

Year

AGA data EIA data

1975 1980

I

1982 1983

SOURCE: Office of Technology Assessment, based on AGA and EIA data,




Table 17.—Additions to Lower 48 Natural GasProved Reserves: Extensions and New Pool

Discoveries 1966-83 (BCF)

New poolYear Extensions discoveries Total

1966 . ....8,7671967 . ....9,472

1968 . ....7,0371969 . ....5,8001970 . ....6,1461971 . ....6,3751972 . ....6,1541973 . ....5,9311974 . ....5,6931975 . ....5,9261976 . ....5,3371977 . . . . . 6,569 (8,056)

a

1978 . . . . . 6,720 (9,582)1979 . . . . . 7,112 (8,949)1980 . . . . . (9,046)1981 . . . . . (10,485)1982 . . . . . (8,349)1983 . . . . . (6,908)

3,1102,420

1,4262,0433,3633,3613,0961,9701,9521,6491,9942,144 (3,301)1,964 (4,277)1,690 (2,566)

(2,577)(2,994)(3,419)(2,965)

11,87711,892

8,4637,8439,5099,7369,2507,9017,6457,5757,3318,713 (1 1,357)8,684 (13,859)8,802 (1 1,515)

(11,623)(13,429)(11,768)(9,873)

aThe values in parentheses are from EIA data; all other values are AGA reserve

data.SOURCES: Office of Technology Assessment, based on data from Energy Infer.

mation Administration, U.S. Crude Oil, Natural Gas, and Natural Gas Liquids Reserves— 1983 Annual Report, DOE/EIA-0216 (83), October1984, and American Petroleum Institute, American Gas Association,and Canadian Petroleum Association, Reserves of Crude Oil, Natu- ral Gas Liquids, and Natural Gas in the United States and Canada

as of December 31, 1979, vol. 34, June 1980.

provide d a ma jor frac tion of to ta l U.S. exten-

sions—e. g., 43 percent in 1966. A sharp declinein Permian Basin reserve growth in 1968 was theprimary reason for the general decline in exten-sions a t the sa me time. The increa se in extensionsna tionwide , beg inning in 1977, resulted prima r-ily from increases in:

q

q

q

q

Western O verthrust Belt d eve lopm ent;development of the deep Anadarko Basin;tight gas sand development in NortheastTexa s, Arkansas, and Lou isiana ; andTexas gulf co a st de velop me nt, inc luding off-sho re fie lds, the South Texas Lob o Trend ,and tight sands in the Austin Chalk.

Implications

Although the shift in data collection from AGAto EIA complicates interpretation, the sum of ex-tensions and new pool discoveries has apparently

been increasing from about 1976 to the present,In the AGA data, however, the increase only

takes these “followup” discoveries back towardthe levels achieved during the brief surge in newp oo l disc ove ries tha t o c c urred in the ea rly 1970s.

The EIA data show a considerably higher levelof “ follow up” disc overies at ab out the levels thatAGA estimate d for 1966 and 1967.

In order to understand the recent variations inextensions and new pool discoveries, it is gen-erally necessary to track the new field discoveries

that serve as the “ inventory” for the sec ond aryexp loration proc ess. There is no ob vious trendin the national new field discovery pattern (fig.18) that would explain the recent higher level ofsecondary discoveries; AGA new field discoverydata in the period immediately before this appar-

ent surge in secondary discoveries show no simi-lar increase. Consequently, in order to under-stand fully the causes of the recent surge, itprobably is necessary to undertake a detailed ex-amination of data at the level of individual fields.This is beyo nd the sc op e of OTA’ s stud y. How -

ever, some reasonable hypotheses can be fash-

ioned based on the avai lable data.

One possible explanation for the recent in-c rea ses in extensions and new poo l disc ove riesis that the increment over “normal” levels rep-

resents the delayed development of fields discov-ered earlier but not developed for economic rea-sons. The dip in new pool discoveries from about1973-76 (fig. 19), which occurred despite an

earlier period of steady new field discoveries thatnorma lly should ha ve m ainta ined stea dy leve lsof extensions and new pool discoveries, supportsthis explana tion.

Some field-spe c ific da ta also supp ort a “ de -layed development” cause for part of the in-creases. For example, recent extensions in theAustin Chalk fields in southeast Texas appear tobe tied to o ld fields that w ere ma rginally ec o-nomic when discovered and had never under-gone major development before recent price in-creases encouraged a reexamination. Becausethese fields we re no t “ new ,” recent d isc ove rieswere probably recorded as extensions and newpool discoveries, even though there was little inthe way of previously recorded new field dis-covery “inventory” to trace as the statistical causeof these sec ond a ry d isc ove ries.

Similarly, another of the areas providing a sub-stantial fraction of the increased extensions—theWestern over-thrust Belt–probably also followed



Ch. 5—Gas Production Potential . 89

a delayed pattern of development. In this area,there was little incentive to delineate immediatelythe first new fields discovered because there wasno means to transport the gas. Consequently, asubstantial inventory of new fields could havebuilt up until a p oint was rea c hed where it be-

came clear that the area contained sufficient re-serves to justify a p ipe line. At ta ining this leve l ofreserves would have introduced an incentive forfield delineation, and secondary explorationwould have then proceeded to cause a surge inextensions.

An ad ditiona l c ause o f the rec ent higher re-

c orded levels of sec ond ary disc overies c ould b ean ac c eleration in the pa c e o f f ield size d elinea -tion and de velopme nt, Suc h an ac c elerationwould result in the field size growth that in the

past might have been spread out over a 60-year

span being compressed into a shorter timeperiod , with highe r leve ls of a nnua l reserve a d -ditions during this shorter period. Acceleratedfield size g rowth w ould b e a n expec ted c onse-

quence of higher gas prices, although the recentproblems of reduced gas demand would tend tohave the reverse effect, that of slowing down the

pac e of growth.

To summ arize, two possible c auses for rec entincreases in extensions and new pool discoveriesare an accelerated field development pace andthe delayed development of earlier new field dis-coveries whose development was (at least in part)

initially uneconomic. If these are indeed the pri-mary causes of the increases, this has importantimplications for future reserve additions. First, thefaster pa ce of de velopm ent mea ns that feweropportunities for field growth will be availablein the later years of development; this should tendto dec rea se future reserve a dd itions unless therate of new field discoveries increases. Second,unless add itional o ppo rtunities for growth from

olde r fields are availab le, this source of “ inven-tory” for extensions and new poo l disc ove ries isunlikely to allow continuation of the currentlyhigh reported levels of reserve additions. Al-

though continuing technological advances andfuture gas price increases could offer some po-tential for sustaining reserve additions from olderfields, the ac tua l pote ntial for reserve additions

from this source is controversial. * In any case,most of any additional reserve growth from olderfields see ms likely to be a ttributed to infill d rillingand other causes that will be reported as positiverevisions rather than as extensions and new pooldiscoveries.

Recent and future discoveries of new fields stillprovide the primary source of inventory for futureextensions and new pool discoveries. Conse-quently, future reserve additions from extensionsand new pools depend heavily on the meaningof the sha rp ly higher levels of new field discov-eries reported during 5 of the past 7 years by EIA.As d isc ussed in the “ New Fields” sec tion , OTA

suspects that part of the reason why ElA’s com-pilation of new field discoveries is substantiallygreater in magnitude than the levels shown byAGA is tha t the EIA d ata c ap tures som e o f the

reserves that AGA would have reported as sec-ond-year extensions, new pool discoveries, orpositive revisions. If this is c orrec t, the “ growthfactor” that should be applied to ElA’s new fielddiscovery data to account for f ield growth afterthe year of discovery will be smaller than thegrowth fac tor app lica ble to AGA data. For thisrea son, we do not believe that continuation ofhigh levels of e xtensions and new po ol disc ov-eries is probable under current conditions.

Aside from the effec ts of the c hang e in rep ort-ing, there are o ther rea sons to b elieve tha t futurelevels of extensions and new pool discoveries

ma y d rop . First, muc h o f the field g row th in thepast has come from the giant fields that took yearsto develop. A large percentage of recent new field

discoveries, however, are small, class E (less than6 BCF of recoverable gas) fields that will requirelittle additional exploratory drilling past the ini-

t ial wildc at. Sec ond , the suspe c ted ac c elerationin the pace of field development implies thatsome of the development that might in the pasthave taken place in the second year (and thatwould have been reported as extensions and new

pool discoveries) now takes place in the first andwill be reported as part of the “new field dis-

c ove ries” reserve a d d itions. Finally, the high c a p-ital requirements for developing new fields in hos-

*As discussed in ch. 4, “New Gas From Old Fields. ”

38-742 0 -85 - 7




tile environments–an increasing part of the

remaining resource—demand a more thoroughinitial estimate of reserves, possibly leading tolow er (sta tistica l) growth la ter on.

To c onc lude , OTA d oe s not b elieve it is likelythat recent reserve additions from extensions andnew p oo l disc overies of 12 to 14 TCF/ yr will besustained in the future even if the gas “bubble”ends and its negative effects on drilling cease, in-

stea d, w e p rojec t a ra nge of 6 to 11 TCF/ yr a s a naverage over the next 10 to 15 years, except thatfor 1985-86 we p rojec t a ra ng e o f 8 to 12 TCF/ yr.The sole p ossibility of a highe r long-term rate o f

reserve additions from this source lies with the

disc ove ry of seve ral new , co mp lex, super giantgasfields with large growth potentials; however,this possibility appears low.

RevisionsRevisions indicate changes in the volume of

proved reserves that result from new informationgained by drilling and production experience andcorrections made to earlier estimates during thereporting year.

The AGA a nd DOE/ EIA rep orting of revisionsis not identical because EIA has a separate cate-

go ry of “ ad justments and c orrec tions” that in-cludes adjustments for changes in data samples,corrections of reporting errors, inclusion of lateresponses, and other factors. Theoretically, AGA’s

revisions should be equivalent to the sum of ElA’srevisions, adjustments, and corrections. How-ever, the d ata ga thering a nd a nalysis metho ds

used by the two surveys a re rad ica lly different,and their reserve and reserve addition estimatesin the 3 years of overlap do not show good agree-

ment, Consequently, they are not equivalent, al-though in displaying historical trends AGA revi-sions will be compared to EIA revisions plusadjustments and corrections.

Factors Affecting Revisions

Generally, revisions oc c ur bec ause o f unc er-tainty associated with estimating the extent of theunderground reservoir rock within a trap, the por-

osity and permeability of that rock, water satura-tion, pressure, and other physical reservoir char-acteristics that affect the cumulative volume of

production over the life of the reservoir. Revisionstend to be a “catchal l” category of reserve ad-ditions and deletions, and the m any source s ofrevisions are difficuIt to separate out of the data.These sources inc lude:

1.

2.

3.

4.

5.

6.

7.

new knowledg e ga ined b y normal develop-ment drilling and production experience

(e.g., changes in reservoir pressure declinetrends that indicate that earlier estimateswere incorrect);numerica l errors in the origina l com p ilationof reserve estima tes;discoveries for which reporting had beendelayed;development drilling on a closer spacing that

“discovers” new reserves;*changes in production economics that lower

or raise the abandonment pressure of a res-ervoir or that allow or prevent the use ofwell-stimulation techniques that increaserecovery efficiency;knowledge gained from extension tests thatindicate a decrease in the estimated provedarea of a reservoir or field (an increase wouldbe recorded as an extension); andmiscellaneous statistical corrections and ad-

justme nts to the d a ta .

Sources 2, 3, and 7 are considered “Adjustmentsand Corrections” by EIA and are reported sepa-rately.

Historical Variation of Revisions

From 1966 to 1983, net revisions were easilythe most volatile of any of the four types of re-serve ad ditions. In the da ta rep orted b y AGA forthe c ontiguo us 48 Sta tes, revisions va ried from

+6,256 BCF in 1967 to –3,546 BCF in 1973. Inthe EIA d a ta for the same a rea , revisions plus ad - justments and corrections varied from –2,91 1BCF in 1977 to +4,346 BCF in 1981. Conse-quently, the year-to-year changes in revisionswere the primary determinant of the year-to-year

changes in gross reserve additions during the past

16 yea rs. As shown in figure 20, a series of sub-

*New reserves “discovered” by a development well would be

recorded as a revision if the gas is located in a pocket within the

established boundaries of a reservoir yet is physically isolated fromthe reservoir’s main drainage system and would not otherwise be

produced.



Ch. 5—Gas Produc tion Pote ntial 91

Figure 20.–Additions to Lower 48 Natural Gas Proved Reserves: Revisions As Reported, a 1966-83 (BCF)

7,000

6,000

t

-1,000t

1966

r

I I I I I I I I t I I1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979

I I I I1980 1981 1982 1983

‘ tear

aNOTE: EIA plots for revisions + adjustments and Corrections.

SOURCES Office of Technology Assessment, based on data from Energy Information Administration, U S Crude 011, Natural Gas, and Natural Gas Liquids Reserves— 1983 Annual Report, DOE/EIA-0216 (83), October 1984, and American Petroleum Institute, American Gas Association, and Canadian Petroleum Association, Re- serves of Crude Oil, Natural Gas Liquids, and Natural Gas in the United States and Canada as of December 31, 1979, vol 34, June 1980

stantial positive revisions in the mid-1960schanged to net negative revisions in almost everyyear in the 1970s, particularly in the onshore con-tiguous 48 States. As discussed later, understand-ing the role o f the se revisions is importa nt in in-terpreting reserve c hanges during this period .

The largest negative revisions in the 1970s werereported in onshore south Louisiana and TexasRailroad Commission Districts 1, 2, 3, 4, and 6.Tog ethe r, they c ontrib uted a tota l of over 30 TCFand proved to be rema rkably pe rsistent, co ntinu-ing throughout the 1970s in both the AGA and

EIA d a ta. They w ere c onc entrate d in olde r fieldsthat had been producing for one to three dec-ades before the revisions began.

The negative revisions in these six areas appear

to b e c ausally related to a situation that e nc our-aged optimism in reserve calculations. During the

1930s, 1940s, and 1950s, exploration for na turalgas in and adjacent to the gulf coast was highlysuc c essful. As a result, muc h m ore gas wa s d is-covered than could be produced, given the smallsize of the national natural gas market at the time.The tra nsmission c om pa nies, having c ontrac tedfor reserves with a productive capacity substan-tially exceeding what they could market, devel-op ed a system for prorating p rod uc tion a mongop era tors on a basis of rem a ining reserves (i e.,the large r an op erato r’ s reserves, the mo re g asthe transmission c om panies wo uld b uy). This c re-ated a strong incentive for the operators to pro-vide the most optimistic estimates of reserves theycould justify. By 1970, following years of increas-

ing production and gradual depletion, the oper-ators were beginning to realize that reserves wereove rstat ed . The size, timing, and ge og rap hic d is-tribution of the rep orted neg a tive revisions tha t




followed dep ended primarily on when ea ch m a- jor operating company recognized the problemand how they d ec ide d to revise their estima tes

dow nward, choosing to ta ke them al l at onc e orsprea d ing them o ut over seve ra l years.4

ImplicationsAn argument can be made that the historical

record, erratic as it seems, supports the idea of

genera lly po sitive revisions in the long te rm. Thisis based on the view that the large but localizedneg a tive revisions of the 1970s appea r to ha ve

ended . The t rends in rev isions for the a reas out-side the source area for the negative revisionsseem far more positive.5 For example, if the gulfcoast revisions were subtracted from the totalLower 48 revisions, as shown in figure 21, the

qlbid.‘T. j. Woods, “On Natural Gas Trend s,” Gas Research Institute,

1982; R. Nehring, contractor report to OTA, op. cit.

“ am end ed revisions” would a pp ea r to supp orta projection of positive future revisions. On the

other hand, an examination of the sources of revi-sions indica tes tha t extreme c aution should beused in forecasting the direction of futurerevisions.

Of the seven source s of revisions listed p revi-ously, the second and seventh are essentially ran-

dom . The others eithe r will alwa ys yield p ositiverevisions, will always yield negative revisions, orma y have a bias in one direc tion o r the o ther.The first source , drilling and p roduc tion e xpe ri-

ence, would be random if there were no incen-tives to be either pessimistic or optimistic in re-serve calculations. However, the requirement toraise capital for field development or to meet min-imum reserve requirements for a new pipeline

are powerful incentives for optimistic reserve esti-ma tes. A tend enc y tow ard op timistic estima tes

would result in mostly negative revisions fromdrilling and prod uc tion expe rienc e. The large

6,000 -

5,000 -

4,000 -

3,000 -

2,000 -

1,000 -

0 —

-1,000 -

-2,000 -

I

I

Year

aNOTE: EIA plots are for revisions + adjustments and corrections.

SOURCE: Off Ice of Technology Assessment




nega tive revisions of the 1970s in the g ulf co ast

and adjacent provinces appear to have resultedfrom just such a tendency.

The fifth source , c ha nge s in prod uc tion ec o-nomics, also could be random in that gas pricescould rise faster (yielding positive revisions) orslower (negative revisions) than the costs of oper-ating fields and enhancing production. Althoughrigid price controls or the competition of low-

priced alternative fuels could conceivably leadto negative revisions from this source, it seems

more likely that most such revisions would bepositive, especially if gas becomes scarcer. In sup-port of this argument, the growth in reserves at-tributed to well reworking, infill drilling, andlowered abandonment pressures–growth thatwould be reported as positive revisions—is seenby some analysts as an extremely important com-

ponent of future reserve additions (see ch. 4, sec-tion o n “ New Ga s From O ld Fields” ).

Of the remaining sources of revisions, the thirdand fourth will always yield positive revisions, andthe sixth always will yield negative revisions. *

The c onfusing m ix of “ positive,” “ neg at ive,”and “random” sources of revision make it ex-tremely difficult to predict how revisions will

behave in the future. Also, revisions data do notindica te w hic h p revious years’ da ta are b eingrevised. Consequently, it is difficult to know thecauses of past revisions—a necessary prerequisite

for intelligent forecasting. For these reasons, someanalysts disregard revisions entirely in their trendanalyses and implicitly assume they will not bea signific ant c omp onent of future reserve ad -ditions.

A reasonable range of average yearly revisionsfor the next 10 to 15 years appears to be O to 2TCF/ yr, with the p ositive t end enc y b ased on

OTA’ s b elief tha t there m ay be som e sign ific a ntpo tentia l from the g row th of olde r fields due tolowered abandonment pressures, infill drilling,and the like.

*The sixth, knowledge gained from extension tests, yields only

negative revistons because an increase in reserves caused by this

source would be reported as an extension.

Reserves-to-Production Ratio

Because the reserves-to-production ratio, (R/P),

measures the rate at which gas is produced fromdiscovered reservoirs, it represents the analyti-cal link between projections of new discoveries

and forecasts of gas production.

Factors Affecting R/P

At the level of the individual production firm,

the selection of a production rate—and, conse-quently, the selection of the R/P–represents an

ec onom ic trad eoff b etwe en the c ost of drillingadditional wells and installing additional gasgathering and processing facilities (i.e., the costof increasing production), on the one hand, andthe c ost of ho lding reserves in the g round , on theother. Consequently, factors such as explorationand development costs, present and expected fu-

ture gas prices, and interest rates all affect the R/P.For example, increases in current prices will theo-retically lead to faster production, whiIe expec-tations of real increases in future prices can causeproduct ion to be delayed.6

In o il p rod uc tion, it is we ll know n tha t to o fasta production rate—too low an R/P—can cause apremature decline in production and a loss of po-tentially recoverable reserves. For example, in areservoir whose pressure is supplied mainly by

wa ter that d isp lac es the o il as it is p rod uc ed (a“ wa ter-d rive” reservoir), an overly rap id ra te o f

product ion can c ause the enc roac hing wa ter toflow a round less pe rme a ble sec tions of the reser-voir, leaving behind the oil in these sections.When the water reaches the well, the added costsof wa ter sep aration and dispo sal ca n c ause p re-mature abandonment.7

Bec ause g as flows mo re ea sily tha n o il, thereis far more leeway in gas production, and pro-duc tion rates freq uently c an vary over a widerange . There a re, howe ver, the same kinds ofphysical limits to gas production as to oil produc-tion. Althoug h som e loss of ultima tely recove r-able gas from the well may be acceptable to the

bDouglas Boh i and Michael Toman, ‘‘Understanding Nonrenew-

able Resource Supply Behavior, ” Sc ienc e, vol. 219, Feb. 25, 1983.

7P. A. Stockil (cd.), Our /r rduQry Petrcdeurn (London: British Pe-

troleum Co. Ltd., 1977).




producer in exchange for a more rap id p ayba ck(from the higher flow rate), the potential for largelosses will serve to limit flow rates.

Aside from the obvious economic factors andphysical limitations to avoid resource loss, sev-

eral other factors affect R/P:

q

q

q

Technology .–The major technology affect-ing R/P may be rock-fracturing methods. Theuse of massive hydraulic fracturing and otherfracturing techniques can open up low-permeability rock and cause marginal wellswith low flow rates to become rapid pro-duc ers. The a va ilab ility of sop histica tedseismic exploratory techniques has reducedoverall drilling costs—enhancing the incen-tive to d rill ad ditional we lls to e xpand pro-duc tion–by increa sing the suc c ess ra tio; it

also has helped improve the placement of

successful wells to maximize production.Ge olog y.–The ra te o f ga s flow is d irec tly de-pendent on the permeability of the gas reser-voir formation and on its pressure andthickness. Although fracturing can partly

compensate for low permeability, wells intight gas formations generally produce muchmo re slowly than d o w ells in mo re p erme-

able rock because the fractures do not reacha ll of the tight reservoir roc k. Simila rly, gasin deep over-pressure formations will forshort periods of time produce far morerapidly than in shallow, low-pressure forma-

tions; * in fact, the high pressures in such for-mations have caused severe technical prob-

lems in fields such as the Fletcher Field insouthw estern Oklahom a, where we lls and

dril l ing equipment have been destroyed byfailure to control the enormous pressures

built up deep undergrounds Also, field sizedistributions may affect R/P because smallerfields, which will be of increasing importancein future reserve additions, may be producedfaster than large, c om plex fields.

Field Maturity .–Early in a field’s lifetime,

R/Ps are typically very high because the ma-

*However, once the “propping effect” of the gas IJnder pres-

sure is removed by partial production, the permeability of the reser-

voir may be reduced to “tight gas” levels, and production will slow.B’ ‘Fletcher Area U nderscor~ Perils in Deep Gas Reservoirs, ” 0;/

and Gas Journal, Feb. 7, 1983, p. 25.

q

q

q

jor focus is on reserve delineation rather thandevelopment; during this period, pipelineand gas processing capacity may be nonex-istent or minimal and markets may be un-

developed, As pipeline capacity is added andsales contracts signed, the R/P will decrease

rapidly. As the field tends toward depletion,the R/P may rise again as gas pressures dropand as drilling gravitates to the marginal, low-permeability formations. However, becausethe R/P will equal 1.0 in the last year of afield’s production, the R/P will decrease dur-ing the very last years of the field.

Conservation Regulations.–Some gas-produc ing States direc tly regulate p rod uc -tion-relate d va riab les suc h a s we ll spa c ing

and flow ra tes. These reg ulations a re in-tended to promote efficient development ofreserves to prevent loss of ultimate recovery.

Their orig in lies in the d isrup tion c aused bythe discovery of the east Texas field in 1930and the large over supply and resultingwasteful gas production practices thatfollowed. 9

Market Demand.–When the market isdemand-limited (deliverability exceeds de-mand), as it is today, the R/P no longer pro-vides a measure of gas production capacity.Low demand can raise the R/ P.Reserve Requirements.-The substantial cap-ital requirements of gas transmission and dis-tribution systems has led the transmission

and distribution companies as well as Gov-ernment regulatory agencies to pursue long-

term contracts requiring high R/Ps and highreserve requirements for pipeline approvals.These req uirem ents d o no t ap ply, howe ver,to mature areas where pipeline capacity isalready in place.

Historical Variation of R/P*

The ea rly yea rs of g as d isc overy in this c oun-try were marked by lack of a gas distribution net-work, substantial discoveries of gas as a low-

valued or even unwanted byproduc t of o il ex-

9R. E. Megill, An Introduction to Exploration Economics (Tulsa,OK: Petroleum Publishing Co., 1971 ).

*Based on jensen Associates, inc., Understanding Natural Gas

Supply in the U. S., April 1983, contractor report to OTA.



Ch. 5—Gas Production Potential 9 5

ploration, and the e ventual d isc overy of eno r-

mous reserves (e.g., the 1922 discovery of thegiant Hugoton field in Kansas) that overwhelmedexisting de ma nd . The c om b ina tion led to ve ryhigh R/ Ps in the 1920-40 period , follow ed by anera of c ontinued d ec line.

In the early years of the post-world War IIgrowth, as new pipeline systems were con-

structed, previously unproductive proved re-serves we re de velop ed . This ac tivity inc rea sed thelevel of production without adding substantiallyto the volume of p rove d reserves, thus loweringthe R/ P. La te r in the 1960s and 1970s, when thenatural gas market had become supply con-strained, production was again maintained by fur-ther development and a lowering of the R/P. Atthis time, however, the ability to obtain greaterproduction from a given volume of proved re-

serves wa s improved by a g eo grap hica l shift inproduction to the Gulf of Mexico and encouragedby economic changes that favored more rapidextraction rates.

The dec line in the R/ P from 1946 to 1983 forthe Lowe r 48 Sta tes is shown in figure 22. TheAGA data cover the period 1946 through 1979,while DOE/EIA includes only the 7 years from1977 through 1983. *

The AG A da ta show strong yea r-to-year de -c lines in the R/ P over virtua lly the entire 34-yea rperiod of available data. Recently, the rate of de-

cline eased from an average of over 0.8 per yearbetween 1966 and 1974, to an average of 0.5 peryea r during the 1975-79 pe riod . DOE/ EIA e sti-ma ted dry ga s da ta show a further easing o f thedecline rate to about 0.2 per year between 1977and 1981, and an actual increase in R/P thereafter

due to the wea k markets for gas and relativelystrong reserve replacements during 1981-83.

Currently, the lowest R/P for the nonassociatedga s of a ma jor p rod uc ing Sta te is 8.4 in Lou isiana .The low est R/ P for any ge og rap hica l sub d ivisionpublished by the DOE/EIA reserves report wasa 5.0 for the Sta te d om a in of the Texas offshore

* *These displayed ratios are developed using the year-end re-

serves estimate for the year prior to the production period. This

approach to calculating the R/P stems from a belief that produc-

tion in a given year is more likely to be representative of reserves

that are available at the beginning of the year.

Figure 22.— Reserve-to-Production Ratiosfor Natural Gas in the Lower 48 States

25 }

20

10

1945 1950 1955 1960 1965 1970 1975 1980 1985

Year

SOURCES: Office of Technology Assessment, based on data from Energy infor-mation Administration, U.S. Crude Oil, Natural Gas, and Natural Gas

Liquids Reserves— 1983 Annual Report, DOE/El-0216 (63), October1964, and American Petroleum Institute, American Gas Association,and Canadian Petroleum Association, Reserves of Crude Oil, Natu- ral Gas Liquids, and Natural Gas in the United States and Canada

as of December 31, 1979, vol. 34, June 1980,

(for 1982). The total Texas and Louisiana offshore,rep resenting about one -third of Low er 48 Sta teproduction, stood at 6.5 in 1981 and 8.4 in 1983.With the Gulf of Mexic o exclude d, the b alanc eof the Lower 48 States had a 1981 R/P of 9.8 anda 1983 R/P of 11.7. Contrasting strongly with thelow er R/ Ps of the g ulf c oa st wo uld b e tha t of 26.3for the heavily depleted reservoirs of Kansas and

23.6 in Wyoming, where field development fornewly discovered reserves was incomplete in1983.

Implications

These recent exam p les of R/ Ps for d ifferent

area s of the United Sta tes ma y ind ica te tha t theLow er 48 Sta te R/ P c ould mo ve further do wn-ward in future years if gas supplies were foundin areas with combinations of high reservoir per-meabilities and economics that favor extensivefield development. * This is in fac t what happ ene d

throughout the 1970s as the G ulf of Me xic o b e-

*Some opinion exists, however, that some of the lower R/Ps aredue to underreporting of reserves rather than to extremely rapid

production. If true, this might indicate less potential for further

lowering the national R/P.



96. U .S . Natural Gas Availability: Gas Supply Through the Year 2000

came an increasingly large component of the totalsupply. Between 1973 and 1981 the Gulf’s shareof production grew from 20 to 33 percent.

An additional factor that might tend to push the

R/ P dow nwa rds is a c ontinuat ion of c urrent d is-

covery trends towards smaller field sizes. It iswide ly be lieved tha t sma ller fields will be delin-eated, developed, and produced over shorter pe-riods of time than was historically the case withthe mix of field sizes discovered until now.

On the other hand, some factors could causethe R/P to climb upward. Future productiontrends may tend to increase the shares of gas fromtighter, low er p ermea bility reservoirs and othe rsources more expensive to develop, which couldlead to slow rates of production from proved re-serves. For exam ple, bo th the dee p Tusc a loosatrend and the Western Overthrust Belt are ex-

pected to have relatively high R/Ps; field devel-opment and gas processing costs for these areasare too high to allow rap id de pletion a t c urrentgas prices.10 in addition, the R/P might tend toincrease if future reserve a dditions we re belowannual production rates because the production

capability (as a percentage of remaining reserves)of reservoirs tends to decline with their age, * anda rate of reserve additions that is below replace-ment levels will lead to an increasing average agefor U.S. gas reservoirs.11

It is important to note that the balance between

dem and and supply will also p lay a c ritica l rolein determining the R/P. Because the purpose of

IOE F Hardy and c. P. Neill, testimony to the Subcommittee on. .

Fossil and Synthetic Fuels, Committee on Energy and Commerce,

U.S. House of Representatives, June 1, 1981.

*Up to a point. During the last few years of a reservoir’s life, its

R/P must decrease because, in the last year, it will be 1.0. The last

year’s production will use up the entire remaining reserve.

1‘ I bid.

this evaluation is to examine the potential for gassupply if gas is highly sought after, gas produc-tion–and, consequently R/ P—is assum ed to bebased on a supp ly-limited situa tion . * * This situ-ation would tend to intensify the incentives to de-velop fields rapidly and to maximize production

(minimize R/P). Rapid field development shouldnot be expected, however, if the current gas“bubble” of oversupply continues. In this case,field d evelopm ent a nd p rod uction a re likely tobe slowed.

In conclusion, expected R/Ps in 15 to 20 yearsmay range from values below recent (pre-bubble)levels–perhaps 7.0, or even somewhat lower–

to levels slightly higher–perhaps 9.5. Part of

the future trend w ill be c aused by the ge ologicnature of new discoveries and their geographicenvironm ent. These fac tors c a n be m an ip ula ted

somewhat but are more likely to be imposed bythe random success of future exploration. Be-

c au se the R/ P is also strong ly affec ted by the will-ingness to drill development wells and to take

other (expensive) production-enhancing meas-ures, large increases in gas prices would tend todrive the R/ P dow n to its lowe r limit. The lowe rvalue ob viously c an oc c ur only with high ga s de -mand, an assumption of this study. If gas demandwe re p oor, the R/ P c ould exc eed 9.5 for a while.Eventua lly, howe ver, the lac k of explora tion in-centives would move proved reserves back intoba lanc e w ith p rod uc tion requireme nts.

Production Scenarios

Tab le 18 summa rizes the rang es of reserve ad -d itions and R/ Ps p rojec ted for Low er 48 natural

— —* *That is, a situation where additional supplies at prevailing prices

would be easily absorbed,

Table 18.-Summary of Projections of Components of Reserve Additions and R/Ps

New field discoveries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1985-86 2.0-3.5 TCF/yr1987-2000 1.5-3.5 TCF/yr

Extensions and new pool discoveries. . . . . . . . . . . . . . . . . . . . . 1985-86 8.0-12 TCF/yr1987-2000 6.0-11 TCF/yr

Revisions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1985-2000 0-2.0 TCF/yrR/P . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2000 7.0-9.5

Scenario 1A: reserve additions. . . . . . . . . . . . . . . . . . . . . . . . . . . . 1985-86 17.5 TCF/yr1987-2000 16.5 TCF/yr

R/P . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2000 7.0

Scenario 1 B: reserve additions. . . . . . . . . . . . . . . . . . . . . . . . . . . . 1985-86 10.5 TCF/yr1987-2000 7.5 TCF/yr

R/P . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2000 9.5SOURCE Office of Technology Assessment.



Ch. 5—Gas Production Potential 97

ga s d eve lopm ent. Tab les 19 a nd 20 present p ro- bability reserves where production rates are

duc tion and reserves p rojec tions tha t represent limited. Because each projection represents a

the two extrem es of the rang es in ta b le 18. The c onvergenc e o f events of relatively low proba bil-first projection assumes an optimistic exploration ity—e.g., the lowest rates of new field discoveries,

future a nd rap id p rod uction of newly found re- extensions and new pool discoveries, zero revi-

serves-predicated upon high gas prices, high de- sions, and an upturn in R/P-the projections

mand , and a n avoidanc e o f large reserve a dd i- should be viewed as approximately bounding the

tions in low p ermea bility area s tha t a re ha rd to range of production and proved reserve levels,de velop rap id ly. The sec ond p rojec tion a ssume s rather than as identifying likely values.

low finding rates and an increase in low perme-

able 19.–Lower 48 States Natural Gas Production and Reserves 1981.2000 (in TCF)SCENARIO 1A: Optimistic Exploration, Rapid Production

Year Production Reserve additions Proved reserves R/Pa

1981 (actual). . . . . . . . . . . . . . . . . . 18.5 21.5 168.6 9.01982 (actual) . . . . . . . . . . . . . . . . . . 17.2 15.1 166.5 9.81983 (actual) . . . . . . . . . . . . . . . . . . 15.5 15.0 166.0 10.71984 . . . . . . . . . . . . . . . . . . . . . . . . . 16.0 15.0 165.0 10.41985 . . . . . . . . . . . . . . . . . . . . . . . . . 18.0 17.5 164.5 9.21986 . . . . . . . . . . . . . . . . . . . . . . . . . 18.3 17.5 163.7 9.01987 . . . . . . . . . . . . . . . . . . . . . . . . . 18.6 16.5 161.6 8.81988 . . . . . . . . . . . . . . . . . . . . . . . . . 18.8 16.5 159.3 8.6

1989 . . . . . . . . . . . . . . . . . . . . . . . . . 19.0 16,5 156.8 8.41990 . . . . . . . . . . . . . . . . . . . . . . . . . 18,9 16,5 154.4 8.31991 . . . . . . . . . . . . . . . . . . . . . . . . . 18.8 16.5 152.1 8.21992 . . . . . . . . . . . . . . . . , ... , . . . . 18.8 16.5 149.8 8.11993 . . . . . . . . . . . . . . . . . . . ... , >. 18.7 16.5 147.6 8.01994 . . . . . . . . . . . . . . . . . . . . . . . . . 18.9 16.5 145.2 7.81995 ......., . . . . . . . . . . . . . . . . . 18.9 16.5 142.8 7.71996 . . . . . . . . . . . . . . . . . . . . . . . . . 18.8 16.5 140.5 7.61997 . . . . . . . . . . . . . . . . . . . . . . . . . 18.7 16,5 138.3 7.51998 . . . . . . . . . . . . . . . . . . . . . . . . . 18.7 16.5 136.1 7.41999 ......., . . . . . . . . . . . . . . . . . 18.6 16.5 134.0 7.32000 . . . . . . . . . . . . . . . . . . . . . . . . . 18.6 16.5 131.9 7.2

Cumulative production after 1982 = 330,6 = 42%0 of USGS remaining resource.

aR/p calculatedby dividing previous year’s (year end) reserves by production in the listed Year

SOURCE Off Ice of Technology Assessment,

Table 20.– Lower 48 States Natural Gas Production and Reserves 1981.2000 (in TCF)SCENARIO 1 B: Pessimistic Exploration, Slowed Production

Year Production Reserve additions Proved reserves R/Pa

1981 (actual) . . . . . . . . . . . . . . . . . .1982 (actual) . . . . . . . . . . . . . . . . . .1983 (actual). . . . . . . . . . . . . . . . . .1984, . . . . . . . . . . . . . . . . . . . . . . . .1985 . . . . . . . . . . . . . . . . . . . . . . . . .1986 . . . . . . . . . . . . . . . . . . . . . . . . .1987 . . . . . . . . . . . . . . . . . . . . . . . . .1988 . . . . . . . . . . . . . . . . . . . . . . . . .1989 . . . . . . . . . . . . . . . . . . . . . . . . .1990 . . . . . . . . . . . . . . . . . . . . . . . . .1991 . . . . . . . . . . . . . . . . . . . . . . . . .1992 . . . . . . . . . . . . . . . . . . . . . . . . .1993 . . . . . . . . . . . . . . . . . . . . . . . . .1994 . . . . . . . . . . . . . . . . . . . . . . . . .1995 . . . . . . . . . . . . . . . . . . . . . . . . .1996 . . . . . . . . . . . . . . . . . . . . . . . . .1997 . . . . . . . . . . . . . . . . . . . . . . . . .1998 . . . . . . . . . . . . . . . . . . . . . . . . .1999 . . . . . . . . . . . . . . . . . . . . . . . . .2000 . . . . . . . . . . . . . . . . . . . . . . . . .

18.517.215.516.018.017.316.615.614.513.713.012.311.811.310.810.410.19.89.59.3

21.515.115.015.010,510.57.57.57.57.57.57.57.57.57,57.57.57.57.57.5

168.6166.5166.0165.0157.5150.7141.6133.5126.0119.8114.3109.5105.2101.4

98.195.292.690.388.386.5

9.0

9.8

10.7

10.49.29.19.19.19.29.29.29.39.39.3

9.49.49.59.59.5

Cumulative production after 1982 = 236TCF = 310/o of USGS remaining resource.

aR/P calculated by dividlng previous year’s (year end) reserves by production In the listed Year

SOURCE Office of Technology Assessment




APPROACH NUMBER 2–PROJECTING NEW POOL DISCOVERIES,EXTENSIONS, AND REVISIONS AS A SINGLE GROWTH FACTOR

The p rec ed ing a pp roa c h is de signe d to a l lowa projection of future gas reserves based on sep-arate est imates of new f ield discoveries, exten-

s ions, new pool d iscover ies, and revis ions. Anal ternat ive method is to project only new f ie lddiscoveries and apply a “growth factor” to these

discoveries that combines the effects of the otherthree categories of reserve addit ions.

USGS has used a field g row th a pp roa c h to c al-cu la te the amount o f gas remain ing to be d is -covered in the inventory of ident i f ied f ie lds. 12 Intha t app l i ca t ion , a curve was const ruc ted tha tdescribes the reserve growth in initial discoveriesthat occurs after the year of discovery, averagedover a l l d iscovered f ie lds nat ionwide and over

9 of the 14 discovery years where appropr iatedata were available (1966-79). This curve, illus-

trated in figure 23, shows a 60-year growth in re-serves to about four times the initial (discoveryyea r) estima te o f ga s vo lumes d isc overed . Thec urve show s tha t m ost o f this grow th o c c urs inthe first 5 years after the discovery year. In theUSGS calculation, the curve was applied to theinitial discoveries reported in every discovery

yea r, assuming tha t reserve g row th p a tterns ofrec ently disc overed fields wo uld be the sam e a sthe p a tterns of m uc h olde r disc ove ries. The g a svolumes calculated in this manner—gas that is dif-

ficult to classify as discovered or undiscovered–are c a lled “ inferred reserves” by USGS.

This metho d m ay b e extende d to projec t howthe first-year estimates of reserve additions fromfuture new field discoveries will grow in the yearsfollow ing the d isc ove ry yea r. How eve r, ce rta in

adjustments have to be made. First, a growth fac-tor calculated by tracking “initial discoveries”data must be increased if it is to apply directlyto new field discovery data. This is because thediscovery data13 includes not only new field d is-coveries, but also “certain hydrocarbon accumu-lations which are significant from the standpoint

lzu.s. Geological Survey Circular 860, app. F.I JThe data Came from table XIV of Reserves of Crude O;/, NahJ-

ral Gas Liquids, and Natural Gas in the United States and Canada,

VOIS. 21-34, 1966 through 1979, American Gas Association/Ameri-

can Petroleum Institute/Canadian Petroleum Association.

Figure 23.-The Growth of Year-of-DiscoveryEstimates of the Amount of Recoverable Natural

Gas Discovered in the Lower 48 States

Years after discovery

SOURCE: D. H. Root, “Estimation of Inferred Plus Indicated Resevees for theUnited States,” app. F in Estimates of Undiscovered Recoverable Conventional Resources of Oil and Gas in the United States, U.S.Geological Survey Circular 860, 1981.

that advances in exploration technology resultedin the discovery of such reservoirs. 14 Conse-

quently, the year-of-discovery values are largerthan those o f “ new field d isc overies,” a nd thelater expansion is lower because some technol-ogy-based expansions are excluded. Adjusting thecalculated growth factor to account only for

growth of new fields may raise the factor by about20 percent.15

Sec ond , for the method to b e c red ible, the a s-sumption that the historical growth curve willcontinue to be valid must be relaxed somewhat.

Ma ny of the fac tors affec ting the growth of re-coverable reserves in newly discovered fieldshave changed; consequently, it appears likely that

the g row th c urve ha s c hang ed as well. The d e-velopment of a credible forecasting proceduredepends on defining a new curve or family of

curves that logically fit these changed conditions.

Tab le 21 lists the a rgum ents–som e spec ula-t ive—that supp ort an inc rea se or de c rea se (overhistorical levels of field growth) in the ultimatema gn itud e o f reserve g row th in new fields. *

14[bid.

I SR(3be~ pa~kiewicz, Jensen Associates, perSOnal communi-

cation.

*These conditions are the same as those affecting revisions, ex -

tensiclns, and new pool discoveries.



Ch. 5—Gas Production Potential . 99

Table 21 .-Arguments a for the Question,“Will the Reserve Growth in New Fields Be

Larger or Smaller Than the Growth Recordedin Previously Discovered Fields?”

A. New fields will grow more: 1. Recent increases in real gas prices are leading to

greater recovery factors for gasfields—from closerspacing of development wells, extensions into less-

permeable margins of reservoirs, exploitation ofsmaller pools, lowering of abandonment pressures, andreworking of older wells. Together, they increase theultimate recovery (reported cumulative production atfield abandonment).

2. The historical growth factor does not accurately reflectthe actual field growth, The large negative revisionsin onshore south Louisiana and Texas have artificial-ly depressed reported field-growth rates. Becausethese revisions were due to a unique set of circum-stances, they are unlikely to recur, and reported growthrates should increase.

B. New fields will grow less: 1.

2.

3.

4.

5.

Part of the reason that the levels of new field discover-ies reported by EIA were higher than those reportedby AGA during the 3 years the two reports overlapped

is probably that EIA reported reserve additions duringthe discovery year that AGA did not report until the sec-ond year. Therefore, when ElA-reported trends are usedto project future new field discoveries, the growth fac-tor used should be smaller than the historical average,which was derived from AGA data.The historical growth factor was derived from datadeveloped during a time when giant gasfields domi-nated gas reserves. Giant fields with multiple poolstake many years to develop and are generally believedto have greater relative growth than small fields. Pres-ent and future field sizes will be smaller and shouldbe expected to have smaller growth factors and fasterdevelopment.Improvements in seismic and other exploration tech-nology, as well as in reservoir engineering, allow

clearer initial delineation of field boundaries and other

field characteristics and more accurate first-year re-serve estimates. This should leave less room for

growth.

Increased gas prices have led to acceleration of fielddevelopment. Some of the development that might pre-viously have taken place in the second year now takesplace in the first year and is reported as part of the ini-tial new field discovery reserve data.High capital requirements to develop new fields inhostile environments—an increasing feature of today’sresource base—require a more accurate first-yearestimate of reserves, leading to lower “growth” lateron.

aSome of these arguments are speculative. For example, in B.1., OTA has not

determined the cause of the AGA/EIA differences in reported new fielddiscoveries.


USGS’s estimate is not the only available esti-

mate of field growth. Table 22 presents three

other estimates, with ultimate growth ranging

from 3,5 to 6.3 times the initial year-of-discovery

estimate.

In order to use the “growth factor” approach

to project future gas production, Jensen Associ-ate s, Inc ., an OTA c ontrac tor, con structe d a sim-ple mo del that a pplied grow th curves similar tothat in figure 23 to both known fields and to projec ted levels of ne w field d isc ove ries. A grow thcurve that reached a factor of 4.0 in 60 years wasapplied to all pre-1982 discoveries, while curveswith 30-year growth periods were applied to dis-c overies from 1982 on. The p eriod of 30 yea rswa s selecte d to reflec t OTA’s belief tha t the p ac eof f ield d evelopm ent ha s quickened . The c hoiceis a guess because data sufficient to calculate anew timeta ble a re no t a vailab le. The unc ertainty

in the ultimate value for the growth factor isreflected in a range of values from 3.0 to 5.0. InOTA’s opinion, 5.0 represents an optimistic

upper-bound on future growth in new fields.

Tab les 23 throug h 25 present the results ofthree scenarios representing the search for rea-sonable upper- and lower-bounds on future gas

Table 22.—Alternative Estimates of Growth Factorsfor Initial Reserve Estimates for Gasfields

Author Suggested growth factors

1. USGS (Root) (1981) , . 4.0, all fields2. Haun (1981) . . . . . . . . . 4.0, fields younger than 48 years

5.0, fields older than 48 years3. Hubbert (1974) . . . . . . 3,5, all fields4. Marsh (1971) . . . . . . . . 5.0, fields younger than 28 years

6.3, fields older than 28 years1. D. H. Root, “Estimation of Inferred Plus Indicated Reserves for the United

States,” app. F in G L. Dolton, et al , Estimates of Undiscovered RecoverableConventional Resources of Oil and Gas in the United States, U S GeologicalSurvey Circular 660, 1981,

2. J D. Haun, “Future of Petroleum Exploration in the United States,” AAPG Bulletin 656(10), 1981.

3 M. K. Hubbert, “U S, Energy Resources, A Review as of 1972, ” S Res 45, ser.No. 93-40 (92-75), Committee on Interior and Insular Affairs, U S Senate, 1974,cited in Haun, ibid.

4. G. R. Marsh, “HOW Much Oil Are We Really Finding, ” Oil and Gas Joumal, Apr5, 1971, cited in Haun, op cit.




—


production. * For eac h scenario, the “ growth sepa rately by app lying a gas-to-oil production

curve” method ology w as ap plied only to non- ratio of 1.3 MCF per barrel of crude oil to the

assoc ia ted ga s. Assoc ia ted ga s wa s p rojec ted ElA’s 1981 oil production forecast.16

*The production projections in the three tables should be viewed

as slightly pessimistic. This is because they were based on projected – ––

1982 nonassociated reserve additions of 8.7-10.2 TCF, whereas ac - 6U .S. De–p~rtment of Energy, 1981 Annua / Repofi to Congress,

tual 1982 additions were about 14 TCF. VOI. 3, p. 62.

Table 23.–Lower 48 States Natural Gas Production and Reserves, 1982-2000 (in TCF)-Scenario 2A: Very Optimistica

Total gas Nonassociated gas Assoc./dissolvedYear production Production Reserve additions Proved reserve R/P gas

bproduction

1982 . . . . . . . . . . . . . . . . . . . .1983 . . . . . . . . . . . . . . . . . . . .1984 . . . . . . . . . . . . . . . . . . . .1985 . . . . . . . . . . . . . . . . ., . .1986 . . . . . . . . . . . . . . . . . . . .1987 . . . . . . . . . . . . . . . . . . . .1988 ........, . . . . . . . . . . .

1989 . . . . . . . . . . . . . . . . . . . .1990 . . . . . . . . . . . . . . . . . . . .1991 . . . . . . . . . . . . . . . . . . . .1992 . . . . . . . . . . . . . . . . . . . .1993 . . . . . . . . . . . . . . . . . . . .1994 ........, . . . . . . . . . . .1995 . . . . . . . . . . . . . . . . . . . .1996 . . . . . . . . . . . . . . . . . . . .1997 . . . . . . . . . . . . . . . . . . . .1998 . . . . . . . . . . . . . . . . . . . .1999 . . . . . . . . . . . . . . . . . . . .2000 . . . . . . . . . . . . . . . . . . . .

18.718.218.018.018.318.618.9

19.319,519.719.819.819.719.619.319.119.018.818.7

15.715.315.215.215.515.816.2

16.516.817.117.317.317.217.216.916.816.616.516.4

10.211.713.014.815.515.815.9

16.016.016.116.216.216.315.315.415.415.415.515.5

132.8129.2127.0126.6126,7126.6126,3

125.7124.9123.9122.8121.7120.8119.0117.4116.0114.8113.8112.9

8.88.78.58.38.28.07.8

7.67.57.37.27.17.17.07.07.07.07.06.9

3.0

2.9

2.9

2.8

2.8

2.7

2.7

2.7 2.7

2.6

2.5

2.5

2.4

2.4

2.4

2.3

2.3

2.3

Cumulative production after 1982 = 342.4 TCF = 44°/0 USGS remaining resource.Note: Rows and columns may not add exactly due to rounding.aAssumptions: Nonassociated gas new field discovery rate = 3,000 BCF/yr

Growth factor = 5.0Additional growth from price rises for old gas = 1000 BCF/yr from 1985 tO 1995.

bAssociated/dissolved gas–gas found in the same reservoir with oil.

SOURCE: Jensen Associates, Inc., contract submission to the Office of Technology Assessment, 1983.

APPROACH NUMBER 3–REGION-BY-REGION REVIEWOF RESOURCES AND EXPLORATORY SUCCESS**

Using a region-by-region review to project fu- States and his subjective evaluation of their future

ture gas production involves a geologist’s exam- production potential.ination of a variety of factors affecting produc-

Fc]r this approach, the gas resource base wastion in 10 individual regions of the Lower 48

assumed to be a compromise between the assess-

me nts of USGS and PGC . For ea c h reg ion, a re- —. --- .— —

**The analysis described in this section was performed by Joseph source value was selected by examining the fieldP. Riva, Jr., Specialist in Earth Sciences, Congressional Research size a nd num ber imp lic a tions of the two a ssess-Service (CRS). ments and choosing the value that seemed more



Ch. 5—Gas Production Pofenyial q 101

Table 24.–Lower 48 States Natural Gas Production and Reserves, 1982.2000 (in TCF)–Scenario 2B: Pessimistica

Total gas Nonassociated gas Assoc./dissolved

Year Production Product ion Reserve additions Proved reserve R/P gasb production

1982 . . . . . . . . . . . . . . . . . . . .1983 . . . . . . . . . . . . . . . . . . . .1984 . . . . . . . . . . . . . . . . . . . .

1985 . . . . . . . . . . . . . . . . . . . .1986 . . . . . . . . . . . . . . . . . . . .1987 . . . . . . . . . . . . . . . . . . . .1988 . . . . . . . . . . . . . . . . . . . .1989 . . . . . . . . . . . . . . . . . . . .1990 . . . . . . . . . . . . . . . . . . . .1991 . . . . . . . . . . . . . . . . . . .1992 . . . . . . . . . . . . . . . . . . . .1993 . . . . . . . . . . . . . . . . . . . .1994 . . . . . . . . . . . . . . . . . . . .1995 . . . . . . . . ........! . . .1996 . . . . . . . . . . . . . . . . . . . .1997 . . . . . . . . . . . . . . . . . . . .1998 . . . . . . . . . . . . . . . . . . . .1999 . . . . . . . . . . . . . . . . . . . .2000 . . . . . . . . . . . . . . . . . . . . 11.0 8.7 7.1 61.0 7.2

Cumulative production after 1982 = 254 TCF = 330/0 USGS remaining resource.

Note Rows and columns may not add exactly due to rounding.aAssumptions: Nonassociated gas new field discovery rate = 1500BCF/yr

Growth factor = 4.0Additional growth from price rises for old gas = 500 BCF/yr from 1985 to 1995.

bAssocla{ed/dissolved gas—gas found in the same reservoir with oil

SOURCE. Jensen Associates, Inc, contract submission to the Office of Technology Assessment, 1983

18.718.017.4

16.916.415.915.415.014.614.213.813.413.012.612.211.811.511.2

15.715.114.6

14.113.613.212.712.311.911.611.210.910.510.29.89.59.28.9

8.78.48.3

8.38.27.67.67.77.77.77,77.77.77.27.27.17.17.1

131.3124.6118.4

112.6107.2101.796.692.087.783.880.377.174.471.468.866.564.462.6

8.88.78.5

8.48.38.18.07.97.77.67.57.47.37.37.37.37.27.2

3.0

2.9

3.0

2.8 2.8

2.7

2.7

2.7

2.7

2.6

2.6

2.5

2.5

2.4

2.4

2.42.3

2.3

2.3

Table 25.–Lower 48 States Natural Gas Production and Reserves,1982.2000 (in TCF)–Scenario 2C: Very Pessimistica

Total gas Nonassociated gas Assoc./dissolvedYear production Production Reserve additions Proved reserve R/P gasb production

1982 . . . . . . . . . . . . . . . . . . . .1983 . . . . . . . . . . . . . . . . . . . .1984 . . . . . . . . . . . . . . . . . . . .1985 . . . . . . . . . . . . . . . . . . . .

1986 . . . . . . . . . . . . . . . . . . . .1987 . . . . . . . . . . . . . . . . . . . .1988 . . . . . . . . . . . . . . . . . . . .1989 . . . . . . . . . . . . . . . . . . . .1990 . . . . . . . . . . . . . . . . . . . .1991 . . . . . . . . . . . . . . . . . . . .1992 . . . . . . . . . . . . . . . . . . . .1993 . . . . . . . . . . . . . . . . . . . .1994 . . . . . . . . . . . . . . . . . . . .1995 . . . . . . . . . . . . . . . . . . . .1996 . . . . . . . . . . . . . . . . . . . .1997 . . . . . . . . . . . . . . . . . . . .1998 . . . . . . . . . . . . . . . . . . . .1999 . . . . . . . . . . . . . . . . . . . .2000 . . . . . . . . . . . . . . . . . . . .

Cumulative production after 1982 = 223.4 TCF = 29% USGS remaining resource.

18.517.716.916.1

15.314.513.813.112.512.011.511.110.710.410.19.8

9.39.1

15.514.814.013.3

12.511.811.110.49.99.48.98.68.27.97.77.47.27.06.8

8.78.07.56.4

6.15.45.45.45.55.55.55.55.65.65.55.55.55.55.5

131.3124.7118.1111.2

104.998.592.887.984.479.676.173.170.568.166.064.162.460.959.6

8.98.98.98.9

8.98.98.98.98.98.98.98.98.98.98.98.98.98.98.9

3.0

2.9

2.9

2.8

2.8 2.7

2.7

2.7

2.7

2.6

2.6

2.5

2.5

2.4

2.4

2.4

2.3

2.3

2.3

Note Rows and columns may not add exactly due to rounding

aAssumptions Nonassociated gas new field discovery rate = 1,500 BCF/yrGrowth factor = 30No additional growth from price rises for old gas

bAssoclated/dissolved gas—gas found in the same reservoir with 011

SOURCE Jensen Associates, Inc, contract submission to the Office of Technology Assessment, 1983



.


rea listic. Then, future ad d itions to p rove d reserve swere estimated, based on a subjective evaluationof the following factors:

q

q

q

q

Difficulty and expense of development.–Based on expected field sizes, depths,

known geology.

Announced leasing schedules.“ Ma turity” of p rovince.–The p ercent of to -tal expected resources that have alreadybeen developed.Rec ent d eve lopm ent history .-Espe c ia lly, therates of entry into proved reserves of the re-ma ining resourc es.

For ea c h region, it wa s ge nerally c onsidered

unlikely that a very high p ercenta ge of the re-maining undiscovered resource—say, 50 percentor greater—could be transferred into proved re-serves by 2000, and this situation acted as a strict

limit on production in some regions, for exam-p le, in the “ we st Texas and ea stern New Mex-ico’ region. * *

Tab les 26 and 27 present two sc enarios offuture gas production and reserve additions basedon the a bo ve ap proac h. Sc ena rio 3A projects thatone-qua rter of the ga s estima ted to b e a vailab lein undiscovered fields at the end of 1981 will bedisc ove red b y 2000. This c om pa res to 55 perc entof the und isc overed g as be ing d isc overed b e-tween 1945 and 1981, a period when larger pros-pects were available, but also when gas discoveryrates may have been hampered by low regulatedprices. In this scenario, gas production is pro-

jec ted to inc rea se in the Roc ky Mounta ins andGrea t Pla ins region, the Eastern Inte rior reg ion,

and the Appalachian region; in addition, produc-tion begins in Oregon-Washington and on theAtlantic c ontinenta l shelf. How ever, ma jor p ro-d uc tion de c rea ses a re p rojec ted for west Texasand eastern New Mexico, the midcontinent, andthe g ulf c oa st, all c ritica l ga s prod uc ers tod ay.

Scenario 3B assumes that exploration becomesmore efficient and that 35 percent of the re-sources in undiscovered fields can be discovered

* *To stabilize current gas production to the end of the century

in this region, 96 percent of the estimated undiscovered gas in the

region would have to be discovered by 2000. From 1970 to 1981,

23 percent of the inferred reserves plus undiscovered resources

were added to reserves.

by 2000. Even under this more optimistic sce-nario, however, gas production will decline to13.3 TCF by 2000.

Table 26.—Lower 48 States Natural Gas Productionand Reserves, 1982-2000 (in TCF)—Scenario 3A

Reserve ProvedYear Production additions reserves R/P

1981 . . . . . . . . . . . . .1982 . . . . . . . . . . . . .1983 .., . . . . . . . . . .1984 . . . . . . . . . . . . .1985 . . . . . . . . . . . . .1986 . . . . . . . . . . . . .1987 . . . . . . . . . . . . .1988 . . . . . . . . . . . . .1989 . . . . . . . . . . . . .1990 ......., . . . . .1991 , . . . . . . . . . . . .1992, . . . . . . . . . . . .

1993, . . . . . . . . . . . .1994 . . . . . . . . . . . . .1995 . . . . . . . . . . . . .

1996 . . . . . . . . . . . . .1997 . . . . . . . . . . . . .1998 . . . . . . . . . . . . .1999 . . . . . . . . . . . . .2000 . . . . . . . . . . . . .

18.518.617.616.816.115,715.315.114.714.514.314.214.013.713.5

13.413.313.012.812.6

21.610.910.911,211.211.211.211.211.211.211.211.211.210.110.1

10.310.310.310.310.3

168.6160.9154.2148.5143.5139.0135.0131.0127.4124.1121.0118.0115.2111.6108.2

105.1102.199.497.094,6

999999999988888

88888

Cumulative production after 1982 = 260.6 = 34°/0 USGSmaining resource.

SOURCE: J. P. Riva, Jr., A Projection of Converrtiona/lNatural Gas Production in the Lower 48 States to the Year 2000,Congressional Research Ser.vice/Library of Congress, June 10, 1983

Table 27.—Lower 48 States Natural Gas Productionand Reserves, 1982-2000 (in TCF)—Scenario 3B

.Reserve Proved

Year Production additions reserves R/P

1981 . . . . . . . . . . . . .1982, . . . . . . . . . . . .1983 . . . . . . . . . . . . .1984 . . . . . . . . . . . . .1985 . . . . . . . . . . . . .1986 . . . . . . . . . . . . .1987 . . . . . . . . . . . . .1988 ........, . . . .1989 . . . . . . . . . . . . .1990 . . . . . . . . . . . . .1991 . . . . . . . . . . . .

1992 . . . . . . . . . . . . .1993 . . . . . . . . . . . . .1994 . . . . . . . . . . . . .1995 . . . . . . . . . . . . .1996 . . . . . . . . . . . . .1997 . . . . . . . . . . . . .1998 . . . . . . . . . . . . .1999 . . . . . . . . . . . . .2000 . . . . . . . . . . . . .

18.518.718.017.316.816.215.815.415,015.115.615.715.214.814.514.213.913.713.513.3

21.612.112.112.112.112.112.112.112.112.112.112.112.112.112.112.112.112.112.112.1

168.6161.9156.0150.8146.1142.0138.3135.0132.1129.1125.5121.9118.8116.0113.6111.5109.7108.0106.6105.4

999999999

8.58888888888

Cumulative production after 1982 = 274 = 35% USGSremaining resource.

SOURCE: J P. Riva, Jr., A Projection of Conventional Natural Gas Production in the Lower 48 States to the Year 2000, Congressional Research Ser-vice/Library of Congress, June 10, 1983.




APPROACH NUMBER 4–GRAPHING THE COMPLETEPRODUCTION CYCLE

Projecting future gas production by graphingthe complete production cycle is based on theexpectation of M. King Hubbert that the completec ycle o f prod uction w ill some wha t resemb le abell-shaped curve and that knowing the areaunder the c urve—the tota l rec overab le resource —allows a reasonable facsimile of the entire curve

to be drawn, once a bout a third or more of theproduct ion cycle has been completed. Hubbert

used this approa c h in 195617 to show tha t then-

c urrent estimate s of the rem a ining oil resourcebase implied that oil production was on the vergeof peaking and then declining.

In this app lic a tion, gas prod uc tion va lues for

1900-82 were plotted, and three freeform curveswere extende d from the 1982 produc tion ratesuch that the area under the curves equal led therem a ining ga s resources estimated by, respe c -tively, Hub bert, USGS, and PGC (see ta b le 6).

These c urves a re shown in figure 24.

17M, K, Hubbert, “fNuclear Energy and Fossil Fuels, ” in Ameri-The curves show that Hubbert’s assessment im-

can Petroleum Institute, Dri// ing and Production Practice (1956),plies an extraordinarily sharp decline in produc-

clted in M. K. Hubbert, “Techniques of Prediction as Applied to tion, so tha t b y 2000 the to tal Low er 48 produc -

the Production of Oil and Gas, ’ O il and Gas Supply Modeling, S. tion rate wo uld b e ab out 3 TCF. Sinc e there is1, GaSS (cd.), National Bureau of Standards Special Report 631 May

1982.

little flexibility in drawing this curve, it appears

Figure 24.—Future Production Curves for Conventional Natural Gas in the Lower 48 States

22

i

I

21

20

19

7

6

5

4

3

21I

Year





unlikely that the range of uncertainty due to theselection of the curve’s shape is greater thanabout 2 to 5 TCF in 2000.

The curves representing the USGS and PGC gasresource assessme nts we re draw n so t ha t the de -clining portion of the curve resembles a mirror

image of the ascending portion. Both curves showproduction rates staying steady at least until 2000.A plausible physical interpretation of the curvesis tha t the y represent a resourc e b a se tha t still re-tains a substantial number of large fieldsamenable to rapid rates of production. Further-more, the shape of the curves is clearly alignedwith high demand for gas and prices that en-

courage substantial development drill ing as wellas vigorous exploratory efforts.

The USGS and PGC c urves ob viously can b eredrawn to reflect different conceptions of howthe production cycle might unfold. However, thenecessity of maintaining existing productiontrends in the early years and of tapering offgrad ua lly a s the resource is dep leted limits theop tion s. Figu re 25 sho ws the orig ina l USGS c urve

and a second curve that reflects a different con-ception, that of a production decline that com-menc es ear lier but proce eds at a more gradua lrate. This sec ond c urve might reflec t a futurewhere industrial demand for gas declines and ex-ploratory activity and development drill ing pro-c eed at a low er level. It might a lso reflec t a re-source base whose fields are smaller, in more

difficult to develop locations, and of lower aver-age permeabil i ty.

Figure 25.—Two Production Futures, One Resource Base: Alternative Representations of Future Productionof Conventional Natural Gas in the Lower 48 States, Based on the USGS (1981) Resource Assessment (mean estimate)

1900 1920 1940 1980 1980 2000 2020 2040 2080 2080 2100

Year





A RANGE FOR FUTURE GAS PRODUCTION

In c omp aring figures 24 and 25 to the produc -tion projections produced by the alternativeme thod s, som e interesting c onc lusions c an b edrawn. First, the higher end of the production

ranges, which shows essentially stable productionlevels out to 2000, appears to be quite compati-ble with the USGS and PGC curves, as drawn infigure 24. It should be remembered, however,that there are interpretations of the detailed phys-ica l na ture of the g as resource base tha t, while

compatible with the overall magnitude and eventhe regional estimates of USGS or PGC, could becompletely incompatible with the high yea r 2000production projection. The sec ond c urve in fig-ure 25 displa ys suc h a n a lterna tive interpretation,and there are more radical possibilities as well. *

A sec ond c onc lusion is that the lowe r end o fthe p rod uc tion rang e—ab out 9 TCF by 2000—isreally much too optimistic for a believer of theHubb ert o r RAND resource e stima te . This is be-cause the assumptions of the lower end of therange, while appearing to be pessimistic to a “re-

source optimist, ” may actually appear somewhatoptimistic to a “ resource p essimist. ” This end ofthe range assume s tha t the fairly low new fielddiscovery rates of the early 1970s are more real-istic as a long-term average than are the higherrates of the last few years, but it ignores the pos-sib ility that eve n the se low rate s might go downstill farther as resource depletion continues. Con-sequently, the true production implication of therange of resource base estimates cited in table6 is likely to b e a yea r 2000 range o f ab out 4 to

19 TCF rather than the range of 9 to 19 TCF ex-pressed by the first three projection approaches. *

As discussed in chapter 4, OTA b elieves tha t

the Hubbert and RAND estimates are overly pes-

simistic and that a more likely lower bound forthe remaining recoverable gas resources is about430 TCF rat he r tha n Hub b ert’ s 244 TCF orRAND’ s 283 TCF. This highe r va lue is c ompa ti-

b le with a 2000 produc tion rate of 9 TCF. Con-sequently, in our opinion, a reasonable range forLower 48 conventional natural gas production for

the year 2000 is 9 to 19 TCF. Simila rly, a reason-able range for 1990 is 14 to 20 TCF.

Finally, figure 24 illustrate s an imp ortant point

about the current “optimistic” assessments of therecoverable resource base: that these, too, im-

ply an inevitable decline in conventional gas pro-duction, although the date of decline is perhaps20 or 30 years later than that dictated by a pes-simistic (430 TCF) resource base assessment . Itmust be stressed, however, that the additional 20yea rs or so o f leewa y imp lied b y the mo re op ti-

mistic assessme nts see m likely to yield suffic ientc hang es in pric es and tec hnology to a llow theentry of nonconventional gas sources to the mar-ket and the movement of large amounts of con-ventional resources from “subeconomic” to‘ ‘economic. These p ot en tial sou rce s of g as p ro-duction are outside the boundaries of the re-

source base assessments and production forecastsd isc ussed in Pa rt I of t his report, but t hey will beextremely important to future U.S. gas production.

“It is important to remember that the kind of radical drop in pro-

duction dictated by the most pessimistic of the resource base esti-

mates will likely violate their baseline assumptions of maintenance

of existing cost/price relationships—except for Hubbert’s assessment

(Hubbert believes his methodology “captures” future changes in

price/cost relationships and technology). Although many presentgas customers can switch without extreme difficulty to oil prod-

*One such possibility would be a resource base that, while large, ucts or to electricity (assuming supplies of these are available), a

had most of its resources in hard-to-find, slow-to-produce fields. rapid drop in production would still tend to push gas prices sharplyThe future production “cycle” would then show a significant pro- upwards. This in turn would tend to Increase the resource base

duction drop in the next 20 to 30 years, followed by a very long by moving subeconomic resources into the economic, recoverable

period of low but stable production. category.

3 8 - 7 4 2 0 - 8 5 - 8




PUBLIC AND PRIVATE SECTOR FORECASTSOF FUTURE GAS PRODUCTION

Comparisons of alternative gas production fore-casts have many of the same problems as com-parisons of gas resource base estimates (see ch.

3, ta ble 4). The ec ono mic , reg ulatory, a nd o ther“scenario” conditions assumed for the forecastsare not always made clear. Because the range ofreasonable future values/assumptions for theseconditions are so broad, it is probably safe toa ssume that there are ma jor sc ena rio difference sbe twe en d ifferent fo rec asts. The resources mea s-ured ma y d iffer, with som e forec asts includingonly “conventional” gas and others including allmethane sources, especially gas from tight sands.The extent to w hich som e of the c om mo nly usedresource base estimates (which are importantvariables in some of the forecasts, directly deter-

mining finding rates or defining an upper limitfor cum ulative disc overies) c onta in unc onven-tional resources is not always clear, For exam-ple, the PGC acknowledges that as much as 20pe rc ent o f its estima ted “ po tential resource” isin tight sand s,18 but other estimates do not specifysuch a percentage. Consequently, even the fore-casters, themselves, do not always know howmuch tight gas is incorporated in their produc-tion forecasts. *

Table 28 presents the results of 21 pub lic andprivate sec tor produc tion forec asts of c onven-

tional Low er 48 ga s p rod uc tion.** All a re a fewyears old. Four of the forecasts explicitly includetight sands and/or Devonian shale; these are

noted on the table.

The e xtent of a greeme nt ab out future g as pro-duction displayed in table 28 is in sharp contrastto the v ery wide rang e p rojec ted b y OTA. For theyea r 2000, a rang e of 11 to 15 TCF/ yr–an ex-tremely narrow range, given different base

spotential Gas Agency, news release, Feb. 26, 1983.

*Further, there may not be agreement as to what constitutes “tightgas. ” For example, the Federal Energy Regulatory Commissionin-cludes a maximum permeability of 0.1 millidarcies in its definition,

while the National Petroleum Council used 1 millidarcy as the limitin its report on unconventional gas sources.

* “Including associated/dissolved gas (gas colocated with oil), on

a dry basis.

assumptions, forecasting methods, etc.—wouldenc om pass 13 of the 15 estimates ava ilab le fo rthat da te. In c ontrast, OTA b elieves that a n a pp ro-

priate range fo r year 2000 prod uc tion is 9 to 19TCF/ yr. Part o f this d ifferenc e ma y b e a ttrib utedto the fact that most of the values in the table rep-resent forec asts of “ mo st likely” ga s p rod uc tionrates, and there may be a tendency for such esti-mates to cluster together. In conjunction with thispossibility, a lack of documentation for many ofthe forecasts makes it unclear whether they area ll indep end ent, o rigina l estimate s. Som e ma ysimply be averages of other forecasts, reflectingthe “ c onventional wisdo m. ”

Of particular interest is a comparison of AGA’s

yea r 2000 estima te–1 2 to 14 TCF/ yr–and theproduction implications of the AGA-supportedPGC’ s gas resourc e a ssessment . PGC ’S assess-ment seem s most c omp atible with produc tionleve ls of 15 or 16 TCF/ yr, or highe r. If the AGA

produc tion forec ast is intende d to b e assoc iate dwith the PGC resourc e b ase, the n AG A is usinga most pessimistic interpretation of the resourcebase, at least from the standpoint of maintainingproduction rates at high levels during the nextfew decades.

A striking fea ture o f the ta ble is tha t a ll but o ne

of the forecasts project substantial declines in gasproduction, most within 10 years and all but theone “dissenter” by 1995. It is important to recog-nize that these forecasters include some prominent

ga s “ op timists.” Muc h of the current op timismabout gas’s future must stem from confidence insupplementary supplies from unconventionalsources, from Alaska, from Mexico and Canada,and from LNG imports. (Chapter 6 provides abrief discussion of the potential from all of thesesources except unconventional production,which is the subject of Part II of this report.) How-ever, it also seems likely that a resurvey of these

fo rec asts, using the latest 1984-85 results, wo uldshow a highe r range for the yea r 2000 p rod uc -tion than in the original 1982 and earlier forecasts.



Chapter 6

Gas Imports —An Overview



Chapter 6

Gas Imports—An Overview

Natural gas imports in 1983 totaled 918 billioncubic feet (BCF)1 and composed 5.5 percent ofthe to ta l U.S. dry ga s c onsump tion. The c urrentimpo rt sta tus and future imp ort p rojec tions a resummarized in table 29.

In evaluating potential import supplies of nat-ural ga s to the Low er 48 State s, the mo st o b vioussources are the border countries, Canada andMexico. Canada has been and probably will re-ma in our most imp ortant source of supp lemen-tal natural gas. In January 1983, the NationalEnergy Board recommended an additional 9.25trillion cubic feet (TCF) of reserves for export. Al-thoug h this de c ision nea rly doub les the e xport-ab le qua ntity av aila ble to the United Sta tes, a c -

tua l imp orts will dep end on U.S. dem and andcompetit ive pricing. An important uncertainty inthis reg ard is the e ffec t of rec ent Ca nad ian ini-tiatives to p ric e the ir gas mo re c om pe titively in

‘ U.S. Department of Energy, Energy Information Administration,“U.S. Imports and Exports of Natural Gas, 1981 ,“ june 1982.

the c urrent U.S. ma rket . In the long run, the in-crease in allowable exports will probably help en-courage frontier development.

Exports from Mexico will probably remain ator below 300 million cubic feet per day (MMCF/ day) in the near term, consistent with what theyhave been since the present contract was nego-

tiated in 1979. In fact, Mexico temporarily ceasedgas exports to the United States in the fall of 1984bec ause o f the p ressure fo r lower gas p ric es inthe c urrent w ea k U.S. ma rket. Althoug h Me xic anna tural ga s supplies a re b ount iful, the Me xic an

Government’s current export philosophy seemsto preclude significant increases in exports to theUnited Sta tes. Mexic a n c onsump tion is expe c tedto increase as the distribution infrastructuredevelops.

Alaska represents another large potential sup-ply; the Prudhoe Bay Field alone constitutes over10 percent o f the t ota l U.S. prove d reserves. At

Table 29.—Natural Gas Imports Summary Table

Natural gas supplied Allowable imports Proved reserve Range ofto Lower 48 States under recent estimates future export estimates

Source in 1983 licenses/contracts (Dec. 31, 1982) 1990 2000

Mexico 0.07 TCF 0.11 TCF 76 TCF 0.1-1.0 TCF 0-1.5 TCF(AGA/GER) (AGA/GER) (OGJ) (AGA, LA-Mexico)

Canada 0.7 TCF 1.75 TCF 97 TCF 1.0-2,5 TCF 1.0-3.0 T C F(AGA/GER) (AGA/GER) (OGJ) (AGA, LA-Canada)

A l a s k a o — 35 TCF ANGTS b

(EIA) 0.7-1.2 TCFPacific-Alaskan LNG

0.1-0.2 TCF(AGA)

LNG 0.13 TCF 0.9 TCFC 235 TCF (OGJ)d Variable—depends on future U.S.(AGA/GER) policy and pricing.

Total 0.9 TCF(EIA)

aThis range represents the highest and lowest estimates of the references cited.bAlaskan Natural Gas Transportation System.cThis value represents the total contract volumes for completed terminals.dReserve of SIX countries (other than U S.) currently exporting LNG

REFERENCES

AGA—American Gas Association, The Gas Energy Supply Outlook: 198.3-2000, October 1983.AGA/GER—American Gas Association, Gas Energy Review, various dates

EIA– Energy Information Administration, U.S. Crude Oil, Natural Gas, and Natural Gas Liquids, 1982 Annual ReportLA–Canada– Lewin & Associates–Canadian Natural Gas: A Future North American Energy Source? January 1980LA– Mexico– Lewin & Associates–Future Mexican Oil and Gas Production, July 1979.OGJ —Oil and Gas Journal, December 1982 and other issues.

SOURCE’ Office of Technology Assessment.

111




present, there is no natural gas production reach-ing the Low er 48 Sta tes, owing to t he lac k of ameans of transportation. Financing for a transpor-tation project is difficult to obtain because of cur-rent surplus supply and market prices belowlevels necessary for financial success. Despite a

wa iver pac kage to eliminate road bloc ks to p ri-va te fina nc ing, the Alaska Natural Ga s Transpo r-ta tion System p rojec t still ha s not a c hieved ad e-quate financ ing a rrange me nts. A riva l TransAlaskaGa s System wo uld ena ble North Slope ga s to b emarketed outside of the domestic market. Ametha nol c onversion alternative w ould a llow thegas to be marketed either domestically or inter-

na tionally. Neither of these a lterna tives appe arto have good prospects for the immediate future,

Throug hout the e a rly to mid -1970s, lique fiednatural gas (LNG) contracts were viewed as afavorable means of achieving long-term naturalgas supplies. Since that time, the supply scenario

has changed significantly, and LNG purchasersare now confronted with high-priced gas duringa time of gas surplus. In the near term, there islittle ince ntive to increa se LNG imp orts; how eve r,the availability of the long-term contracts and theop portunity to d iversify U.S. supp ly may proveto b e a ttrac tive in the future.

MEXICO

Mexic o ha d rep orted 75.4 TCF of p rove d re-

serves as of December 1981. Within the last 4years, large reserve additions have caused Mex-ico’s reserve-to-production ratio to double from30 to 60.

Most o f Mexic o’ s ga s prod uc tion is from we llsassociated with oil; nonassociated wells are typi-c a lly not p ut into prod uc tion. This p rac tice re-flec ts Mexic o’ s polic y of e xporting oil and usingnatural gas primarily to meet domestic energy de-

mands. Mexico exports only the surplus gas re-ma ining a fter dome stic de ma nd is met, whichc ould in the future b ec ome a limiting fa c tor toexport levels. Mexico’s current export maximumof 110 BCF/yr was established in 1979 by a con-tract with Border Gas, a U.S. pipeline company.2

This qua ntity is rec og nized as a c om promise

be twee n Mexic an p olic y ma kers, who b elieveenergy exports are necessary to bolster Mexico’sailing e c onom y, and those who be lieve the re-source should be saved for future domestic use.Because of the low gas demand and low marketpric e in the United Sta tes, ac tua l imp ort levelshad been reduced to the 60 percent minimumtake-or-pay level, causing 1983 imports to be

‘Border Gas is owned and controlled by six interstate pipeline

companies: Tennessee Gas Transmission Co., Texas Eastern Trans-

mission Corp., El Paso Natural Gas Co., Transcontinental Gas Pipe-

line Corp., Southern Natural Gas Co., and Florida Gas Transmis-

sion Co.

about 72 BCF.3 Mexican exports to the United

State s have now be en te mp orarily suspe nde d be -c ause o f the unfavo rab le m arket c ond itions.

Mexico has been successful in encouraging

conversions to natural gas, and, as a result, do-mestic gas demand has been growing at a rateof 13 percent per year.4 Because Mexico’s finan-cial condition has precluded investment in dis-tribution equipment, the primary constraint toincreased domestic consumption is a lack oftransmission and distribution capability. As thed istribution system deve lops and the proc ess of

converting end users to gas progresses, domes-

tic consumption will increase, which could fur-ther constrain the exportable surplus.

Ea rly in 1982, the M exica ns ta lked of inc reas-

ing exports to 500 MMCF/day and later to 1,000MMCF/day; however, these plans were not car-ried o ut, owing to prob lem s with ga s-ga thering

system s a nd bud ge t c utb ac ks. s The fa c tors thatwiII determine the actual value of future importsinclude the progress of Mexico’s economic growth

and attempted reliance on gas for domestic

‘j. L. Wingenroth and A. A. Bohn, “Mexican Gas Supplies to the

United States, ” Gd s Energy Review, American Gas Association,

August, 1984.

4Petro/eum Intelligenc e Week/y, Spec ial Supp lement, “Mexico’s

Expanding Role in World Oil Markets, ” June 28, 1982.

51 bid.




q

q

q

stimulate frontier exploration off the EastCoa st and in the Arctic;allow a 25 percent back-in interest for theCanadian Government on federal leases; andincrease the Canadian Government’s shareof petroleum revenues relative to those

received by industry and the producingprovinces.

The inc reased reg ulation of the NEP ha s ha d

a noticea ble neg ative impa c t on risk investment.Canadian operators and support companies haveleft Ca nada fo r mo re lucrative p rospec ts in the

United State s. Many p etroleum c om p an ies hav ecut expenditures and long-term projects and suf-fered severe losses. These effects, if not reversed,c ould lessen the qua ntity of g as prod uce d in theremainder of the century, thereby limiting theavailability of surplus for export to the United

States.Another factor affecting gas export is the level

of Canadian gas consumption. In an attempt toreduce the need for expensive foreign oil imports,

the Canadian Government is encouraging in-

creased use of natural gas and has provided sev-eral incentives for doing so, such as favorable gas

prices, grants, and loans. The NEB forecasts nat-ural gas demand to increase at 4 percent per yearduring the 1980s and 3 percent per year through-out the 1990s.12 Although the conversion proc-ess is p rog ressing slow ly, the q ua ntity o f g a s ava il-ab le to the United States could b e” c onstrained

if Canadian consumption increases substantiallyin the future.

Unde r current C ana dian export ag reem ents,natural gas exports will increase to about 1.6TCF/ yr by 1990 and then dec line t o a bout 0.15TCF/ yr by 2000.13 AG A estimates that between

0.8 and 1.8 TCF/yr will be exported by 1990 and1.0 to 2.4 TCF/yr by 2000.14 Lewin & Associatesbelieve that technological advances in the fron-tier areas and the development of unconventionalga s c ould a llow expo rts of 2.5 and 3.0 TCF/ yr in1990 and 2000, respectively.15

.‘National Energy Board, “Omnibus ’82 Backgrounder, ” Jan. 27,

1983.

1‘Ibid.I ~American Ga s Association, The Ga s Energ y SUpp lY Out /ook . .

1983 -2W0, October 1983,15Lewi n & Associates, Canadian Natural C,Is: A Future North

Ame rican Energy Sourc e, January 1980,

The ma ssive hydroc arbo n p ote ntial of Alaska

wa s rea lized with the d isc ove ry of t he PrudhoeBay Field in 1968, wh ich a dded 26 TCF to esti-mated U.S. proved gas reserves. Reserve esti-

ma tes for Alaska a ve rag e 35 TCF, and resou rcebase estima tes a re a s high a s 169 TCF.

16

Desp ite the sub sta ntia l qua ntity o f reserves in

Alaska, lac k of a transporta tion system ha s p re-c lude d ma rketing o f Alaskan ga s to the Lower 48States. The A laskan Natura l Ga s TransportationAct of 1976 directs the President, subject to con-gressional approval, to establish a means to trans-port Alaskan na tural ga s to th e Low er 48 Sta tes.To ensure domestic use of the resources, the Ex-port Ad ministration Ac t o f 1979 forbids the ex-port of North Slope hyd roc a rbons to non-U.S.

customers. Several transportation methods have

l b f JO t e n t i a { Ga S cornrnlt tee, Poten t ia / Supp / y Of Natura / Gas in

th e U. S., June 1983.

been proposed; not all of these have designated

the Low er 48 Sta tes as the final ma rket .In Sep temb er 1977, the Alaskan Na tural Ga s

Transportation System (ANGTS) was chosen overseveral alternatives. The 4,800-miIe pipeline wasto be routed from Prudhoe Bay across Alaska andCa nad a to Alberta, and split into a western legto California and an eastern leg to Illinois. De-spite a waiver submitted by President Reagan andapprove d b y Cong ress in mid-Dec em ber 1981,

to remove any legislative deterrents to private fi-nancing, the pipeline has not yet been financed.Investment capital has been difficult to attract be-cause the marketability of the gas is questionable.ANGTS is estima ted to c ost b etwe en $38.7 bil-lion and $47.6 billion 17 and de liver gas at p ric es

1‘America n Ga s Assoc iatio n, Ga s Energ y Revie w, vol. 10,

No . 1, january 1982.




cility (1 BCF/day) including production and li-que fac tion, transpo rtation, and rec eiving andvaporization facilities is around $3 billion to $7billion (1982$),25 depending on shipping distance.Although the cost of service is dependent on in-terest rates and shipping distance, a cost of

$3/MMBtu to cover liquefaction, transportation,and reconstitution of the gas would not beunusual .26

The future o f LNG d ep end s princ ip ally on pric -ing and policy. If the producing country demandsa high price at the wellhead—as an extreme ex-ample, a price approaching parity with oil on a$/Btu basis–then the d elivered p rice o f the g as- —

‘SD. Napoli, R. N., “Economics of LNG Projects, ” Oi/ a nd Gas

Journal, Feb. 20, 1984.zGBecause many of the capital costs are sunk, however, this cost

will not necessarily be added to the price of the gas.

ge nerally will be m uc h highe r than the p ric e o fc om pe ting fue ls. Currently, the p ric e of LNG is

higher than market-clearing levels in the UnitedSta tes, and c ould b e sold o nly by m ixing the g aswith lower cost gas and charging a price in l inewith the average cost. In fact, because the cur-

rent cost of liquefaction, transportation, and re-c onstitution ma y be almost eq ual to the ave rag ewellhead price of new gas,27 future imports ofLNG will probably require both a substantial in-c rea se in U.S. do me stic w ellhea d p ric es and ama rketing p olic y on the p art of the exporting na -tions that considers the transportation costs inpricing the gas at the well head.

—‘according to the February 1984, Natura/ Gas M (Energy infor-

mation Administration, DOE/EIA-0130 (84/OZ), the average price

for New Gas (NGPA sees. 102, 103, 108, and 109) was $3.59 per

MCF.



Part II

Unconventional Gas Supplies



Introduction and

Chapter 7

Summary:

Availability of Unconventional

Gas Supplies



Contents

Page

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ... ... ... ..I21The Definition Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ....122Relative Uncertainty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .........,122

Tight Gas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............................123

Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .........................123Resource Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .............125Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ................125

Resource Estimates . . . . . . . . . . . . . . . . . . . . . . . . . .......................126Production Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ................128

Gas From Devonian Shale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .........128Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......................,128Resource Characteristics . . . . . . . . . . . . . . . . . . . . . . ......................129Technology. ,. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..........130Resource Estimates . . . . . . . . . . . . . . . . . . . . . . . . . .......................131Production Estimates . . . . . . . . . . . . . . . . . . . . . . .........................133

Coalbed Methane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..................133Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..........................133Resource Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .........,.,.133Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...,...........134

Resource Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .................134Production Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....136

TABLES

Table No. Page

30.Ga s-in-Plac e Estima tes fo r Tigh t G as . . . . . . . . . . . . . . ...................12631. Ec onomic a lly Rec overab le Ga s a t Two Tec hno log y Levels . . . . . . . . . . .. ...12732. Devonian Sha le Rec overab le Resourc e Estima tes (TCF):

Ap pa lachian Basin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

33. Coalbed Methane Resource Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13534. Co m p a riso n o f Re c o ve rab le Resourc e Estima te s (TCF) . .................135

FIGURESFigure No.

26. Loc ation of Princ ipa l Tight Forma tion Basins .27. Prima ry Area of De vonian Sha le Ga s Potentia

Page

. . . . . . . . . . . . . . . . . . . . . . . . . 124. . . . . . . . . . . . . . . . . . . . . . . , 129



Chapter 7

Introduction and Summary: Availability of

Unconventional Gas Supplies

INTRODUCTION

The unc onventiona l na tural ga s resources in-clude low-permeability sandstone and limestoneformat ions (com mo nly know n a s tight sands or

tight gas formations), Devonian shales, methane-rich coal seams, and geopressurized aquifers.Methane hydrates—gas trapped with water in anic e-like state —and ab iog enic, or “ de ep ea rth”gas–gas supposedly originating from the ventingof m ethane from the Earth’s c ore—have rec entlybeen added to the unconventional roster. Dueto a combination of technical difficulties and highcosts of production, gas found in these geologic

c irc umsta nc es has not, for the most pa rt, bee nincluded in natural gas resource base estimates.However, higher gas prices and advances in tech-nology developments have recently made someof the unconventional resources more attractive.

During the past 10 years, several Governmentand private organizations initiated studies to de-termine the size of these additional resources andthe c ond itions nec essa ry to d eve lop the m. Thefirst comprehensive study of the unconventionalresources was completed by the Federal PowerComm ission in 1973 as part of the Na tiona l Ga s

Survey.

1

It was followed by studies by the Na-t iona l Ac ad emy of Sc ienc e,2 the Federal EnergyRegulatory Commission,3 Lewin & Associatesunder contract to the Department of Energy,4 a ndthe National Petroleum Council. sResults of these

‘ U.S. Federal Power Commlsslon, Task Force Report of the

Supply-Technical Ad\lsory Task Force–Natural Gas Technology,

In /VatIorra/ Cas SurIwy, Lot. 2, 1973.

] Natlona I Academy of Scwnces, Natural Gas From Uncon \ e n-

tmna / Ge olog ic Source s, 1976, Energy Research and Development

Admlnlstratlon Report FE-2271-1.

~Fecieral Energy Re8u Iatory Comm isslon, U.S. Department of

Energy, Nat/onal Gab Survey: Noncon~ent/onal Natural Gas Re-

sources, DOE/FERC-0010, June 1978,W. A. Kuuskraa, et al. ( Lewl n & Associates, Inc.), E n h a n c e d

Reco~ery of Uncon \ en t iona / Gas, Exec utl~e Summa ry, Vo/. 1, (X-

toher 1978, and 2 other vols,, U ,S. Department of Energy Publica-

tion tiCP/T270S-01, 02, 03.

5Natlonal Petroleum Courx II, Uncon~ entlorra/ Ga$ Sou rces, 5

\’o15,, 1980.

studies, and of other studies of individual re-sources, will be discussed in more detail in subse-quent chapters.

A general consensus emerges from these stud-ies that the total size of the unconventional nat-ural gas resource base is extremely large, that withhighe r pric es and more sop histic ate d tec hnol-og ies, significa nt q ua ntities of g as c ould be re-c ove red . Findings of the ea rly stud ies und oub t-ed ly p rovided the impe tus to include ga s fromtight formations, Devonian shales, coal seams,

and geopressurized brines in the high cost cate-gory (sec. 107) of the 1978 Natural Gas PolicyAc t (NGPA). The h ighe r allow ab le pric es for thiscategory were intended to promote near-term de-velop me nt of t hese resource s.

OTA’ s assessme nt d ea ls only with the gas re-source potential of the tight formations, Devo-nian shales, and coal seams. These resources arethe b est unde rstood of the unc onventional re-

sources and appear to have the most potentialfor contributing to supply within the next 20

years. Gas from tight formations and, to a lesserextent, from Devonian shales currently is beingproduced in quantities sufficient to cause substan-tive p rob lems with the de finition o f “ unco nven-tional,” as discussed below. Gas from coal seamsis also being produced, but in much smaller

quantities. In contrast, our present level of un-derstanding of the geopressurized aquifers sug-gests that they are less likely to be commerciallyviable gas producers within this century, althoughsome researchers vigorously disagree with thisview. Too little is known a bout t he m etha ne hy-drates and their production requirements to allow

an adequate assessment of their supply poten-tial. Finally, the p ote ntial of “ dee p e a rth ga s” is

only conjecture at this time because there is nogenerally accepted proof of its existence in com-mercial concentrations.

121

38-742 0 -85 - 9




The Definition Problem

“Unconventional” is not, perhaps, the best

term to characterize the gas resources under dis-cussion, although we will continue to use it inthis rep ort for the sake of simp lic ity a nd adher-

enc e to c ustoma ry usag e. Tight ga s and ga s fromDevonian shales, in particular, are not newly rec-ognized resources. Certain fields that fit in thesetwo ca tegories have b een p roduc ing g as for manyyea rs. These a nd othe r c urrently ec ono mic tightgas and Devonian shale formations are partly in-cluded in the conventional resource base esti-ma tes of the Potential Gas Committee, and a

small amount may be included in the estimatesof th e U.S. Geo log ica l Survey a nd ot hers. To e va l-uate its potential as an additional source ofsupp ly, the unc onve ntiona l resource should in-clude only those parts of the tight formations and

Devonian shales which have not been consideredeconomic to produce under existing economicconditions and technology. Coal seam methaneresources can be more easily categorized since

pa st p rod uc tion h as bee n low. They a re unlikelyto have been included in past estimates of con-ventional resources.

As noted in Part 1, conventional resources gen-

erally are categorized by defining boundary con-ditions in terms of “existing economic condi-tions,” “ current technology,” and other vagueterms. Unfortunately, there currently are nowidely accepted criteria defining these terms toallow a clear division between conventional andunconventional gas resources. Further, theboundary dividing conventional and unconven-tional resource s is c ontinuously c ha nging throughtime due to changing economic conditions, in-creased geologic understanding, and greatertec hnica l sop histica tion. The po orly d efinedboundary causes considerable confusion in deter-mining the size o f the unc onve ntiona l resource

and the am ount that it c an p otentially contrib-ute to to ta l U.S. na tural gas sup p ly.

The leve l of c onfusion is likely to c ontinue.

Most current estimates of the unconventional gasresource b ase a nd its supply po tentia l have a t-tempted to eliminate overlap with conventionalgas resource estimates by excluding areas withexisting production. But data sources for new pro-

duction from unconventional formations do notclearly distinguish between existing and new pro-ducing areas. Since passage of the NGPA, mostof the production data comes from Federal En-ergy Regulatory Commission (FERC) filings for sec-tion 107 (high cost gas) designation and from Pur-

chase Gas Adjustment (PGA) filings that recordgas purchases according to the NGPA categories.The FERC intended to exclude gas from existingproducing formations when it determined criteriafor d esigna ting forma tions eligible for sec tion 107classification. It was inevitable, however, that newwells drilled in a number of existing producing

formations would satisfy the FERC criteria and begranted section 107 prices. As a consequence,it is no longer possible to distinguish betweenarea s which p reviously we re e xcluded from as-sessments of t he unc onve ntiona l resource andthose w hich we re includ ed . Thus it is d iffic ult to

determine the extent to which resources classifiedas unc onventiona l in pa st assessme nts are nowbeing developed.

We will attempt to clarify and identify overlapbetween conventional and unconventional esti-mates of natural gas resource potential in thesub seq uent c hap ters. The rea de r should keep inmind that there are limited data available to makesuch distinctions and our conclusions are neces-sa rily tenta tive.

Relative Uncertainty

Estimates of gas-in-place, recoverable re-

source s, and future prod uc tion of conventionalnatural gas are characterized by a high level ofuncertainty, as described in the previous chap-te rs. Inevitab ly, simila r estima tes for the unc on-ventional natural gas resources will be more un-certain still. Many of the same categories ofunce rtainty, suc h a s lac k of g eologica l under-

standing, are magnified for the unconventionalresources. Further, whereas estimates for the con-ventional resource focus on existing and relativelywell understood technologies, most resource and

production estimates for the unconventional re-sources attempt to foresee new technological de-velopments, adding additional uncertainty. Final-ly, for the unconventional resources, there oftenis little of the production and discovery history



Ch. 7—introduc tion and Summ ary: Ava ilability of Unco nvention al Gas Sup p lies . 123

that serves as a guide to projections for conven-tional gas, and what history does exist applies

only to the small part of the overall resource thatwa s ac c essible to pa st disc overy and produc tiontechnology.

The gene ral level of unce rtainty assoc iated withestima tes of the unco nventiona l resource baseis somewhat different from that associated withprojections of future unconventional production.Although resource and production estimatesshare some uncertainties about the geology of the

resource, generally production estimates focus

on the most accessible, and best understood por-tion of the resource—at least for shorter term projections. On the other hand, projecting futurep rod uc tion req uires ma king assump tions ab outdrilling rates and development schedules, aboutthe pace of production research programs, about

pipeline accessibility, and about the future prog-ress of a number of institutional issues such as

controversies about the ownership of coal seamgas, leasing difficulties in Devonian shale devel-op me nt, etc . In OTA’s op inion, long-term p ro-

jections of future unconventional gas productionshould be viewed as at least as uncertain as, andprobably more uncertain than, estimates of gas-in-place and of recoverable resources at an

assumed price and level of technological devel-opment.

Finally, it cannot be overstressed that any esti-mates of future production and recoverable re-sources that would purport to be “most prob-able” estimates are explicitly relying on anassumption both of the level of effort that Gov-

TIGHT

Definition

Tight g as is na tural ga s tha t is found in forma-tions of sandstone, siltstone, silty shale, andlimestone that are characterized by their extreme-

ly low permeability–i.e., liquids and gases do notflow easily through them. Figure 26 shows themain tight gas-bearing basins in the Lower 48Sta tes. Tight g as reservoirs rep resent t he Iow-

permeability end of a continuum of gas-produc-

ernment and industry will put into the massiveresearch and development necessary to gain ac-cess to the greater part of the unconventional re-

source, and of the success of that R&D p rog ram .past disappointments in technological forecast-ing should serve a s a rem ind er tha t this type o f

estima te m ust a lwa ys be viewe d with a c ertaindegree of healthy skepticism. In addition, theseestimates are relying on assumptions of future gasprices and, in the case of production estimates,on assumptions of future gas demand. Both futurepric es and de ma nd m ust b e c onside red highly

unc erta in. For the se reasons, virtua lly a ll rec entestimates of recoverable resources and futureproduction rely on a scenario approach whereinthe effect of different price and technologyassumptions are examined.

In the following summaries, the term “gas-in-

place” denotes the total gas present in formationswhere some economic gas production is feasi-

ble; it therefore does not include every last mol-ec ule of g as present in the Ea rth. “Technicallyrecoverable resources” denotes gas expected tobe recoverable from these formations up to thelimits of know n tec hnolog y, with little rega rd toprice.6 “Remaining recoverable resources” de-notes gas that is expected to be recoverableunde r a set o f pric e a nd tec hnology a ssump tionsde fined by the estima tor.

eTh is category IS meant to I nc I ud e on Iy those rewu r( e$ that c a n

be extracted by technologies ordinarily used tor ~a~ production

For example, gas that theoretically could be obta[ned by ml nlng

and retorting shales wou Id be excl ucied.

GAS

ing reservoirs rathe r than a unique type. Over the

past seve ra l dec ad es, rising g as p ric es and im-provements in production technology have en-couraged gas producers to move to lower andlower permeability formations, and thus the

boundary between “convent ional” and “uncon-ventiona l, tight gas” has continually shifted. In

1978, in order to allow price incentives to en-courage production of high cost gas under theNGPA, FERC formally defined “tight gas” as hav-




Snake RiverCownyarp

\

Greater GreenRiver Basin

\

Figure 26.—Location of Principal Tight Formation Basins

Northern

Big Horn Great PlainsBasin Province

I

Ft. WorthBasin

AnadarkoBasin

SOURCE: Morgantown Energy Technology Center, (modified by OTA) Department of Energy

ing a specified range of permeabilit ies and pro-duction rates.7 The FERC d efinition is no t unive r-sally accepted by resource appraisers, though,and all estimates of resources and future produc-tion from tight format ions must be eva luated in

———7Under this definition, a tight gas reservoir is one having an aver-

age permeability of less than 0.1 md and a maximum production

rate prior to stimulation, dependent on depth, of 44 MCF/d at 1,000ft up to 2,557 MCF/d at 15,000 ft.

the context of their defined boundary conditions.For example, the current estimate of potential gasresource s published by the Pote ntial Gas Com-mittee—generally considered to be an estimateof conventional resources—contains over 150

TCF tha t PGC now c a tegorizes as tight gas re-sources. OTA estimates that at least 30 TCF of thegas counted as unconventional t ight gas by theNational Petroleum Council in its 1980 report areincluded as well i n the PGC resource estimate.



Ch . 7—introduction and Summary: Availability of Unconventional Gas Supplies 125

Resource Characteristics

Although gas from tight formations generallyis considered to be one resource, there are twodistinct types of tight formations with significantlydifferent p rod uc ing c ha rac teristics. Blanket fo r-

mations extend laterally over large areas, asbefits their name . They o c c ur either a s one thin(10 to 100 ft thick) gas-filled layer or as many eventhinner gas-filled layers alternating with clay-richlayers. LenticuIar formations consist of manysmall discrete reservoirs, often shaped like lenses,separated by shales and sometimes by coal seams.They o c c ur interspersed througho ut fo rma tions

hundreds of feet thick.

Three c ha rac teristics of the reservoir roc ks inthese formations are responsible for their lowpermeability. First, the g rains tha t form the roc k

are small, which causes the individual pores inthe rock and the connections between thesepores to be sma ll. This yields a very high ra tioof p ore surfac e a rea to p ore volume, allowinghigh ab sorption o f wa ter whic h c an physic allyobstruct gas flow; also, the small size of the porec onne c tions inhibits flow . Sec ond , the p roc ess

of dissolution and precipitation of minerals in theroc k, which ha s c ontinued ove r ge ologic time,has bloc ked o r imp aired some of the po re spa c eand connections between pores, further block-ing gas flow . Third , the c onsiderab le a mo unt o fclay often present in the formations can swell in

the p resenc e of wa ter and b loc k flow p aths, orc an b rea k ap art and plug op enings. The net p er-meability of the tight reservoirs also will be af-fected by any natural fracture systems in the rock,which p rovide a lternative p athw ays for gas flow .

Technology

Because of the poor flow characteristics of thereservoir roc k in tight forma tions, ec ono mic levelsof gas production generally can be achieved onlyby creating a manmade increase in permeability,by fracturing the reservoir rock surrounding the

we llbore. This is mo st c om mo nly ac hieved b y hy-draulic fracturing, which involves pumping a fluidu rider high p ressure into the we ll until sufficien tpressure is achieved to break down the rock. Be-cause the fractures would tend to close when thefluid is removed–especially in deep reservoirs

where the pressure of the rock is great—sand orothe r mate rials a re a dded to the fluid. These“ propp ants” settle out o f the fluids and are leftbehind in the fractures when the fluid is removed,

serving as wedges to prevent the fractures fromclosing.

Successful fracturing in tight formations is com-plex and fac es substa ntial obsta c les. Althoughfrac turing dates from the 1800s, and hydraulicfrac turing d ates from 1947 and has the b ene fitof the experience gained by thousands of sepa-rate fracturing treatments, the process is not fullyunderstood and extrapolation to new geologicsituations is difficult. Aside from the difficulty offorecasting wha t a frac ture will do, it is ha rd to

tell in a ny d eta il what a frac ture has done evenafter it has been completed and the well is pro-d uc ing (o r ha s prove d to b e unp rod uc tive). This

is a prima ry rea son why our extensive expe rienc ein fracturing has not been as much benefit in pro-

jecting future performance as might have beenexpec ted .

Despite the difficulties, fracturing has realizedconsiderable success in tight formations, at leastfor the blanket formations. Achievement of longfractures has become fairly consistent, and frac-tures ove r 2,000 ft long ha ve b ee n repo rted. Sub -stantial problems do remain, however. Opera-tors must reduc e the e xtent to which frac turesgrow vertically beyond the gas-bearing layers,lowering the overall eff iciency of the treatmentand losing reserves through water intrusion ordegradation of the reservoir “cap.” Problemsassociated with transporting proppants deep intothe fracture, to prevent fracture closure, and withd rilling fluid d a ma ge to fo rmations from the swe ll-ing or dislodging of water-sensitive clays must beoverco me. The d eg rad ation of pe rmea bility overtime, associated with gradual fracture closure orwith the blockage of po res and frac tures by a cc u-mulated clay or sand, must be prevented. Also,the level of success achieved in the blanket for-mations has not been transferred to the Ienticular

formations, where large-scale fracturing treat-ments have apparently been unsuccessful in con-

necting remote gas pockets, called lenses, to thewell bore—a necessary prelude to fully develop-ing the Ienticular resource. Developers of lenticu-Ia r forma tions have te nded to return to shorter,




less exp ensive frac ture treatments which may im-

ply lower gas recovery.

Resource Estimates

Gas-in-place estimates for t ight gas have been

made by the Federal Power Commission (1972),the Federal Energy Regulatory Commission (1978),Lewin & Associates for the Department of Energy(1978-79), and the Nationa l Petroleum Counc il

(1980). These a re shown in ta b le 30. All bu t the

NPC study limited their estimates to basins wherede tailed ap praisals c ould b e m ad e, prima rily ba -sins in the West and Southw est. The NPC useda method of extrapolation to incorporate basinswhere less data were available, and thus it is theon ly estima te for the to ta l U.S. resource .

In general, the later estimates build on the

ea rlier one s, are more sop histica ted , and havehad a c c ess to b ette r da ta . The Lew in and NPC

estimates should be considered the most credi-ble e stima tes to d ate of the in-plac e resource.However, there are substantial remaining uncer-tainties in even these estimates. Especially impor-tant are uncertainties in porosity and water satura-tion (which affect both the amount of gas presentand its flow prop erties), in the a rea l extent andthickness of the gas-producing portions of thetight formations, and, in some basins, in the

ge ologic history as it a ffec ted ga s forma tion a ndpreserva tion. These unc erta inties are imp orta nt

in the appraised basins and are critical in theNPC’s extrapolated basins.

s porosity and water saturation limit the

amount of gas that may physically be pres-ent in the reservo ir roc k. These p a rameters

are both extremely diff icult to measure ac-c urate ly and m ay vary over a w ide rang ewithin a small area. This implies that the useof a l imited number of data points to char-

Table 30.—Gas-in-Place Estimates for Tight Gas

Study Gas-in-place, TCFFPC , . . . . . . . . . . . . . . . . . . . . . . 600FERC . . . . . . . . . . . . . . . . . . . . . 793Lewin . . . . . . . . . . . . . . . . . . . . . 423NPC (appraised basins) . . . . . . 444NPC total . . . . . . . . . . . . . . . . . . 924

SOURCE’ Office of Technology Assessment.

q

q

acterize a basin—a characteristic of all theestimates—leaves considerable room forerror.Areal extent and thickness of the gas-bearing (pay) zones are direct determinants

of g as vo lume. These are d iffic uIt to meas-

ure in seve ral c irc umstanc es, for exam ple,pay thickness for blanket formations con-taining multiple thin gas-bearing layers(stringers), or a real extent for Ienticular sandswhen surface outcrops are not present.

Geologic history affects the volume of gaspresent because’ it determines the presenceof source materials, temperature and pres-sure histories critical to the formation andpreservation o f ga s, and the a va ilab ility ofa t rapp ing mec hanism. Substantial unce r-tainty occurs in any areas that have not beentested by d rilling , and ma y a lso exist for c er-

tain potentially productive layers in exploredterritory. Areas affected by this uncertaintyinclude the Northern Great Plains, the Pice-ance Basin, the northern part of the DenverBasin, and most of the extrapolated basins.

Although arguments have been made favoringboth higher and lower estimates of gas-in-place

than those found in the NPC estimate, with an

imp ortant exce ption, the a rgum ents for neither

view see m p rep ond erant. The except ion is the

argument, based on geologic theory and on thec urrent low level of d evelopm ent a c tivity, that

the NPC’ s estima te of 150 TCF for the NorthernGreat Plains’ gas-in-place is considerably toohigh. In OTA’ s op inion, th is is a d istinc t possi-bility. Otherwise, the NPC estimate of gas-in-place for the remaining 11 appraised basins–

444 TCF minus 148 TCF for the NGP, or 296TCF–should serve a s a reasona b le “ mo st likely”estimate for those basins. A considerable errorb and -pe rhap s +/ - 100 TCF–must be a ssigned tothe latte r value, how eve r. The NPC’s gas-in-p lac eestimate for the entire United States—924 TCF—isconsiderably less reliable because of extreme

unc erta inty in the 480 TCF assoc iate d with the

101 basins whose resource s we re estima ted byextrapolation rather than direct appraisal.

Estimates of the recoverable tight gas resourceshave be en ma de by Lewin & Assoc iate s and theNational Petroleum Council in conjunction with



Ch . 7–introduction and Summary: Availability of Unconventional Gas Supplies 127

their gas-in-place resource estimates, and by theGa s Resea rch Institute. These e stima tes a re ex-tremely sensitive to assumptions made aboutprice, level of technology, and gas-in-place.

Lewin and NPC first estimated technicallyrecoverable gas. Lewin computed a 50 percentrecovery of the gas-in-place while NPC computeda 66 percent recovery; this reflects NPC’s more

optimistic technology assumptions, as discussedlate r. Thus, Lew in’s estima te fo r rec ov erab le ga sfrom its appraised basins is 212 TCF, whereasNPC’ s is 292 TCF for a c om parab le set of a p -praised b asins, NPC estima ted a tota l U.S.rec overab le resource of 607 TCF.

Estima te s of economically recoverable gas re-sources vary over a wide range, from a conserv-ative 30 TCF (base technology, $3/MCF in 1979$,Ga s Resea rch Institute ) to 575 TCF (adva nc ed

technology, $9/MCF in 1979$, total United States,NPC), as shown in table 31.

At one extreme, GRI’s relatively low estimatesappear to reflect its desire to be conservative andto include only those tight resources that have

a very high probability of occurrence and recov-erab ility. At the othe r extreme, the NPC’ s c on-siderably higher estimates reflect the extensionof its ana lysis to the en tire United Sta tes, its fa -vorable assessment of gas resources in the North-ern Great Plains, and its confidence in the effec-tiveness of tight gas production technologies. As

note d in the d isc ussion of g as-in-plac e, the ex-trapolated portion of the resource base and theNorthern Great Plains resource appear to behighly uncertain. In addition, three key technol-

ogy assumptions made by NPC appear to be quiteop timistic, espec ially if taken to gether. Theseassumptions are:

1.

2.

3.

In

Fractures in Ienticular sands will contactlenses distant from the wellbore. Without

suc h c ontac t, gas rec overy in the Ienticularbasins will be d rastica lly red uc ed . At pres-ent, the ability to contact remote lenses hasnot been demonstrated.Present fracturing technology allows 1,000-ftfractures to be consistently achieved, and

advanced technology will achieve 4,000-ft fractures. The 1,000-ft frac tures do no tap pe ar to be the c urrent state o f the art i nshallow (e.g., Northern Great Plains) or len-tic ular formations, a nd 4,000-ft frac tures a p-pear optimistic for advanced technology inthese same geologic situations.

The longer fracture lengths can beachieved while reducing fracture heights,and thus reducing fracturing costs perfoot. This is op posite t o c urren t exp erienc e.On the other hand, there are approachesto a chieving these c onditions that d o a pp earplausible.

OTA’ s op inion , the op timism of this set ofassumptions implies that the NPC estimates ofrecoverable resources shouId themselves be con-sidered op timistic, tha t is, highe r than a “ mo stlikely” estimate.

In OTA’ s view , a ll ava ila b le estima tes of rec ov -erable tight gas are highly uncertain because ofpoorly defined reservoir characteristics and

Table 31 .–Economically Recoverable Gas at Two Technology Levels (TCF)

Price per MCF Base Advanced$ (study date) $ (1983) technology technology

Lewin (1977) . . . . . . . . . . 1.75 2.75 70 1493.00 4.70 100 1824.50 7.00 108 188

GRI (1979) . . . . . . . . . . . . 3.12 4.20 30 1004.50 6.00 45 120

6.00 8.00 60 150Total Appraised Total AppraisedNPC (1979) ., ... , ., . . . 2.50 3.35 192 97 331 142

5.00 6.70 365 165 503 2315.00 6.70 365 165 503 2319.00 12.00 404 189 575 271





technologic uncertainties. However, despite ourc ritic ism o f c erta in aspec ts of the NPC a na lysis,it seems basically sound to us, and we have littledo ubt tha t large qua ntities–at lea st a few hun-d red TCF–of tight ga s will be rec ove rab le pro-vided gas prices reach at least moderately high

levels in the future (e.g., $5 to $7/MCF in 1984$).

Production Estimates

Estimates of future production of tight gas have

been m ad e b y Lew in, NPC, and GRI, as we ll asby the Am eric an Ga s Assoc iation (AGA).

GRI’ s prod uc tion estimate s for the yea r 2000rang e from a bo ut 2 to 6 TCF/ yr de pe nding o npric e a nd tec hnolog y. Bec ause its estimates ofrecoverable resources are unexplained, thevalidity of this estimate is impossible to judge.

Both Lewin and NPC project high year 2000prod uc tion: 4.0 to 6.8 TCF/ yr for Lew in (at $3/ MC F, 1977$), 4.1 to 15.5 TCF/ yr fo r NPC (a t$5/McF, 1979$) depending on the phasing in of

advanced technology and the dril l ing schedule.The NPC e stima te is deliberate ly structured torepresent a goal attainable by a concerted effortat developing the necessary technology and ac-celerat ing the pace of development.

AGA used the NPC ana lysis as a sta rting point

and superimp osed more c onservative a ssump -tions about drilling rates, implementation of new

technologies, and initial production rates perwe ll. The result is a p rojec ted yea r 2000 prod uc -tion rate of 4.3 TCF/ yr, or 3 TCF/ yr if d efinitiona loverlap be twee n tight a nd c onventional reser-voirs is eliminated and a less optimistic outlookfor the potential of the advanced technologies isassumed.

Several factors imply that the more conserva-tive estimates should be considered more likely.

The slow rate of tec hnology d evelopm ent in theIenticular formations coupled with the impor-tance of the lenticular Rocky Mountain Basins inthe op timistic e stima tes is a c ritica l fac to r. Simi-larly, the Northern Great Plains would normallybe expec ted to p lay a ma jor role in future d evel-

op ment b ec ause muc h of the resource is projec ted to b e reco verable at relatively low c ost;however, here, too, there is controversy aboutthe ma gnitude of g as ava ilab le. The a bsenc e o fpipelines in many potential tight gas productionregions also implies a lower rate of developmentunless the market for new gas supplies improvesdramatically in the near future.

Aside from being sensitive to geologic (accessi-bility of Ienticular resource, magnitude of North-ern Great Pla ins ga s) and tec hnolog ic assump -tions, future production is also extremely sensitive

to gas prices and to the availability of competing,a nd less c ostly, conventional gas prospects. OTAis extremely skeptical of the possibility of reliablyforecasting either gas prices or conventional gasavailability in the time frame in question. Con-sequently, the range of plausible scenarios forfuture incremental t ight gas production encom-passes a year 2000 produc tion rate o f only 1 or

at mo st 2 TCF/ yr if c onve ntional ga s prod uc tionremains at high levels or if gas markets do notrebound from their current slump, or a rate of3 to 4 TCF/ yr, or pe rha ps even som ew ha t higher,if there a re ma jor tec hnology ad vanc es and a

c omb ination of strong ma rkets and high pric esfor unconventional gas, the latter in response todisappointing prospects for conventional gas sup-ply or a surge in gas demand.

——80ver and above product mn from tight format Ions now bw ng

exploited. Current production is about 1 TCF/yr.

GAS FROM DEVONIAN SHALES

Definition from the accumulation of organic-rich sedimentsin a shallow sea covering the eastern half of what

Devonian shale gas is gas produced from shales now c onstitutes the c ontinenta l United Sta tes.formed approximately 350 million years ago– The first Devonian sha le gas we ll was d rilled induring the Devonian period of geologic t ime– 1821, near Fredonia, NY, and mod erate levels



Ch. 7—introduction and Summary: Availability of Unconventional Gas Supplies q 129

of gas production (recently somewhat less than Resource Characteristics0.1 TCF/ yr) ha ve c ontinued to th e p resent. How-ever, despite its long history, the Devonian shale The Devo nian sha les oc c ur prima rily in the Ap -resource is still considered “unconventional” be- palachian, Illinois, and Michigan basins, showncause of its highly complex geology and because in figure 27. Past production has been primarilynew technology and higher prices will be re- in a sma ll portion of the Appa lac hian Basin, in

quired to exploit the major share of its potentially the Big Sandy Field in Kentuc ky and ad jac entrecoverable gas. West Virginia. The sha les are highly va riab le in

Figure 27.—Primary Area of Devonian Shale Gas Potential

\ \ Basin/’

)

SOURCE Johnston & Associates, OTA contractor




their makeup; they can be grouped according toc olor, with b lac k and brown shales having higherorganic content and gas content than the graysha les. The sha les a re a ric h sou rce roc k for na t-ural gas, but their porosity and permeability arevery low c omp ared to c onventional ga s reser-

voirs.9 Conseq uently, gas c ontent and flow ratealso are low by conventional standards. Furthercomplicating exploitation of the gas resource, theshales are sensitive to “formation damage’’—aninduced decrease in permeability—because theycontain water-sensitive clays that can be dis-lodged by fracturing fluids and block pores andfractures.

Portions of the sha le c onta in ne two rks of na t-ural frac tures, whic h tend to be predo minantlyi n a vertic a l pa ttern. These frac tures also tend tobe somewhat l ined up rather than random in di-

rec tion, a c ha rac teristic c alled “ a nisotrop y.” Thefracture systems provide potential flow paths forshale gas.

The sha le ga s oc c urs as free ga s in the frac tures

and pores of the shale and also as gas bound tothe physical structure of the shale (adsorbed gas).The a mo unts a nd p roduc tion m ec ha nisms of thedifferent modes of occurrence of gas are not fullyunderstood, and this lack of understanding com-plicates estimation of the recoverable resource.A prima ry uncertainty is the c ontribution of ad -sorbe d ga s to to ta l produc tion. Most ea rly esti-mates of shale gas resources are based on the no-

tion that the primary source of producible gas isthe free gas in the shale’s fracture network.Recently, many in the research community haveshifted to the view that gas adsorbed on the shalema kes the m ajor contribution to ga s produc tion.

Technology

As with the tight sands, production of Devoni-an shale gas depends on well stimulation to over-

come formation damage and the natural ly lowpermeability of the reservoir and open up path-

9Porosities generally are 1 or 2 percent compared to 8 to 30 per-

cent in conventional reservoirs; permeabilities range from 0.001

to 1.0 md compared to 1 to 2,000 md in conventional reservoirs.

The permeability difference means that, all else being equal, gasWI II f]~w 1 m i Illon to 2 m I I lion times faster i n a conventional reser-

voir than in the Devonian shales.

wa ys for the g as to flow to the we ll bo re. Unlikethe tight sands, howe ver, prod uc tion using c ur-rent technology generally cannot succeed unlessthe well intersects a natural fracture network, ei-ther directly or through an induced fracture. Animportant uncertainty is the extent to which new

technological development will allow productionfrom portions of the shale that do not contain awell-developed natural fracture network. Alsounlike the tight sands, producers generally haveused small fractures in the shales, a few hundredfee t o r less, no t th e ma ssive 1,000-to 2,000-ft frac -tures becoming more popular in the Westerntight sands.

Because of their extreme sensitivity to forma-t ion damage, the Devonian shales have been aprimary target for the development of new frac-turing tec hniques that a void such da ma ge. Stimu-

la tion by the use o f explosives ha s bee n p reva -lent in the sha les’ prod uc tion history, and moresophisticated explosive techniques may be prom-ising for future development. Also, the shaleshave be en a testing g round for new frac turing

fluids, including gas-in-water emulsions, nitrogen,liquid c arbo n d ioxide, a nd othe rs. The ga s-in-water emulsions, or foams, have dominated frac-turing in the Devo nian sha les in rec ent yea rs, butnitrog en ha s a lso g rown in use fo r sha llow we llsbecause it does not cause formation damage. Ni-trogen ha s limited ability to c arry p rop pan ts, soit is less useful at d ep ths whe re the ind uc ed fra c -

tures would close under the overburden pressureof the rock.

Fracturing has been extremely successful formany Devonian shale wells, but its overall rec-ord is very erratic. problems include extremevariation in the natural fracture systems from siteto site, lack of a systematic scientific method inapplying and evaluating fracture treatments, anddifficulties in accurately locating the gas-bearingzones. An unfortunate result of the trial-and-errorapproach is that no scientific basis for selectionof appropriate well stimulation techniques hasbeen developed.

Aside from problems encountered in fractur-ing, development of the Devonian shale resourcealso is hindered by problems in exploration andwell location. in general, sophisticated explora-



Ch . 7—Introduction and Summary: Availability of Unconventional Gas Supplies q 131

tion technology, such as seismic reflection, is notused in shale development. Besides the techni-c al p rob lems, which include ad verse terrain a ndlack of effectiveness of existing technology, theusual incentive for expensive geologic surveys—the a b ility to exc lusively d eve lop t he surveyed

area—is hampered by diverse land ownershipan d inad eq ua te State regulato ry system s tha t donot fully protec t d isc ove ry righ ts.

Resource Estimates

Several estimates have been made of the Devo-nian shale gas-in-place and recoverable re-sources. For example, recent estimates of the gas-in-place by the National Petroleum Council(1980), U.S. Geological Survey (1982), and Mon-santo’s Mound Facility (1982) encompass a rangeof 225 to 2,579 TCF for the Ap pa lac hian Basin,

the mo st sign ific ant o f the three sha le basins byfar. Differences in the e stimate s a re c aused p ri-

marily by the following factors:

. the use o f different b ound ary co nd itions forinclusion in the estimated gas-in-plac e re-

source;s including or excluding the less productive

gray shales;q substantial differences in the shale thickness

calculations because the measurement tech-niques were d ifferent;

varying levels of geochemical analysis under-

taken (this analysis can identify areas wheretem perature and p ressure c ond itions we repo or for ga s formation a nd preservation); anddifferent views ab out the am ount o f ga s in

eac h “ mod e” of oc currence within the shale

(in frac tures, in m ic rop ores, or bo und to theshale), and the extent to which it is properlymeasured in available studies of gas content. ‘O

Although no consensus about all of these fac-to rs c urrently e xists, OTA c onside rs it likely tha t

the Devonian shale gas-in-place is large, at least500 TCF and more likely well ove r 1,000 TCF.The size o f the in-plac e resou rce is no t rea lly themajor issue, however, Because, in general, theshale is a low-quality gas resource and economicreturns will be low, the major issue is the size of

the economically recoverable resource.

Tab le 32 shows seven d ifferen t estima tes of De-vonian sha le recove rable resource s. The estima tes

are not easily comparable because of differencesin tec hnology a nd ec onom ic assump tions, butthey appear to display a fairly broad range of ex-pectations about future shale gas development.For exam ple, the 1977 OTA stud y estima tes a

~~Recent studies have ind Icated that most gas content measu r~-

ments must be adjusted to account for gas that has escaped from

the shale samples prmr to measurement.

Table 32.—Devonian Shale Recoverable Resource Estimates (TCF): Appalachian Basin

Organization Year Estimate Conditions

Office of Technology Assessment. . . . . 1977 15-2523-38

Lewin & Associates . . . . . . . . . . . . . . . . . 1978-79 2-1o4-25

Traditional Advanced

National Petroleum Council . . . . . . . . . . 1980 3.3 - 38.9

15.3 - 49.9Pulle and Seskus (SAI). . . . . . . . . . . . . . . 1981 17-23Zielinski and Mclver (Mound) . . . . . . . . . 1982 30-50

Lewin & Associates . . . . . . . . . . . . . . . . . 1983 6.2-22.5

Lewin & Associates . . . . . . . . . . . . . . . . . 1984 19-44

After 15 to 20 yearsAfter 30 to 50 yearsAt $2-$3/MCF (1976$), current technology

(borehole shooting or hydrofracturing),150-acre spacing

Base caseAdvanced case for prices between $1.75-$4.50

For price levels between $2.50-$9, 160-acrespacing

Technically producible“Shot” wells, 160-acre spacingFor States of West Virginia, Ohio, and Kentucky

only, ’’shot” wells, 160-acre spacing

Technically recoverable, for most promising for-mations in Ohio. Maximum represents 80-acrespacing, advanced technology

Technically recoverable, for most promising for-mations in West Virginia. Preliminary values

SOURCE” Off Ice of Technology Assessment




large recoverab le resou rce–up to 40 TCF–a tmoderate prices and using conventional technol-og y. The 1977 Lew in stud y a nd the NPC stud yare considerably more pessimistic for similar tech-nology/price conditions, with estimates center-ing on about 10 TCF or lowe r. ” Both Lew in and

NPC e xpec t sha rp increases in rec ove rab le re-sources with higher gas prices and improvedtechnology, however. Lewin, for example, pro-

jects a rough doubling of recoverable resourcesthrough advanced technology that allows a sharpreduction in dry holes, completion of multiplezones through each well, and more effective frac-tures. And NPC projects a similar doubling of re-sources as prices move from the $2.50 to $5.00/ MCF range to the $5.00 to $9.00/MCF range. Fi-nally, recent studies by Lewin of currently pro-ducing portions of the shale in Ohio and WestVirginia, using a new reservoir simulation model,

indicate that a substantial increase in gas recoverycan be obtained with improved fractures, re-duced spacing of wells, and more efficient wellplacement that take account of the shale’s lowpermeability and anisotropy.

Aside from differences in assumptions about fu-

ture gas prices and other economic conditions,d ifferenc es in estimate s of Devo nian sha le ga srecoverable resources arise from several techni-cal uncertainties. One type is the set of geologi-cal uncertainties that underlie differences in thegas-in-place estimates, as described previously.Othe r unc ertainties a re a ssoc iated with the:

qability of new stimulation technologies to im-mediately increase the flow rate and main-tain an economic rate over the long term;

q ability of new exploration techniques toovercome the problems of finding areas withwell-developed natural fracture networks;

qability of advanced well-logging techniquesto accurately identify gas-bearing zones andallow greater stimulation success;

qdevelopment of production techniques thatwill allow economic production from shaleforma tions that d o not ha ve a well-devel-

oped fracture network; and

q the extent to which m ethane b ound to theshale matrix plays a major role in produc-t ion.12

[n OTA’s view, all of the existing studies that

estimate recoverable shale gas resources for spec-ified gas prices and technologies have significant

methodological and/or data shortcomings. For ex-ample, because of data limitations, the 1977 OTAstudy d id not und ertake a qua ntitative ana lyis ofthe geology of the Appalachian Basin; instead,it wa s forc ed to a ssume tha t 10 pe rc ent o f thebasin area would allow gas production at levelssimila r to the sma ll area now under produc tion.The e arly Lew in stud y eva lua ted rec ove ra ble re-sources using the assumption that most of therecoverable gas was fracture gas, an assumption

now being challenged. And the NPC study usesan emp iric ally d erived eq uation for ca lc ulatingthe recoverable gas that does not include several

variables—e.g., fracture density and thermal ma-turity of the shale—that ap pe ar to b e c ritic al tothe existence of recoverable gas. However, therecent Lewin analyses do combine a detailed res-

ervoir simulation approach with the latest avail-ab le d ata, a nd p roba bly should be co nsidered themost credible analyses to date. Based on our in-terpretation of the Lewin work, OTA considersit plausible that moderate increases in gas pricesc oup led with a vigorous resea rc h p rog ram to im-p rove we ll stimulation, we ll d iag nostics (e.g., log -ging), and exploration techniques and to a dvanc ethe state of knowledge of shale geology and pro-

duction characteristics could yield substantialqua ntit ies of reco verable g as from the Devo niansha les in the A ppa lac hian Basin, Although thelevel of uncertainty associated with any estimateis high, a figure o f 20 to 50 TCF for the recover-able resources in the fractured portions of the

basin seems reasonable, assuming prices some-what higher than today’s (perhaps $5/ M CF), op-timization of fracturing technology currently in

development, and easing of institutional barriersto development (including rationalization of wellspacing rules). A combination of still higherp rices–in the rang e o f $7 to $10/ MCF–and ad -vanced technology might boost the recoverable

I However, the Lewin estimate is predicated on a higher discount

rate because of its perception of higher risk.

12A major role in production for adsorbed gas implies a substan-

tially increased gas resource.



Ch. 7—introduction and Summary: Availability of Unconventional Gas Supplies q 133

resources to the 80 to 100 TCF level or higher.

Development of methods to produce gas eco-

nomically from shales that do not contain well-

developed natural fracture systems could substan-

tially increase the recoverable gas resource still

further; however, it is important to recognize that

the problems associated with developing theseunfractured shales may be insurmountable.


As with the tight gas analyses, those studies thatprojected future production from Devonian

shales did so by relying on “educated guesses”about available rigs and well drilling rates. All ofthe Devonian shale studies concluded that pro-duc tion within the next few de c ad es would b elimited to relatively moderate leve ls. For exam -ple, the 1977 OTA stud y conc luded tha t 1.0

TCF/ yr c ould b e a c hieved 20 yea rs a fter co m-me nc ing a n intensive d rilling p rogram. The firstLewin study p rojecte d a ma ximum produc tionra te o f 0.9 TCF/ yr in 1990 with ad vanc ed tec h-

nology and $4.50/MCF gas in 1977$ ($7.00/MCFin 1983$), but this was predicated on starting thedeve lopm ent effort in the late 1970s. Also, theLewin study p rojecte d a ma ximum rate of o nly0.3 TCF/ yr with c urrently a va ilab le tec hnolog y.

Finally, the NPC study projected a high of about1.4 TCF/ yr in 2000 with advanc ed tec hnolog y andvery high gas prices ($9/MMBtu in 1979$). At pricesmore in line with today’s, however, production

would have been only a fraction of this.

Development of the Devonian shales will be

c rit ic ally d ep end ent o n ma rket c ond itions, whic htoday are distinctly unfavorable to rapid advancesin production and certainly have delayed the de-velopment schedules projected in the early stud-ies. Furthermore, a rap id b uildup of p rod uc tionwould be hindered by institutional problems,divided land ownership, and difficult terrain. Onthe other hand, the most recent Lewin studiesof Ohio and West Virginia conclude that ad-vanced extraction technology and improved wellp lac em ent c ould substa ntially inc rea se individ -ual well produc tivity and tota l rec overab le re-sources. This imp lies a potent ial for a rap id

buildup of production under the right price andtechnology conditions. if market conditions im-prove very soon and exploration and productiontec hnology ad vanc es are ac hieved, OTA c onsid-

ers a p rod uc tion ra te of 1.0 to 1.5 TCF/ yr fromthe Devonian shales by the year 2000 or soonthereafter to be plausible, although optimistic.

COALBED METHANE

DefinitionCoalbed methane is natural gas formed as a by-

product of the coal formation process and trapped

thereafter in the coal seams. Unlike gas from tightsands and Devonian sha les, pa st p rod uc tion ofcoalbed methane has been very low. ’ 3 However,some important commercial recovery operationsha ve b eg un in New Mexico ’ s Sa n Juan Ba sin, inAlabama’s Warrior Basin, and elsewhere. Also,in the United States roughly 80 billion cubic feet(BCF) of c oa l gas is deliberate ly vented to t heatmosphere each year from working coal mines,

to remove the danger of explosion created by the

1 JHowever, gas formed i n coal seams and trapped i n adjacent

formations has been produced In quantity.

buildup of methane concentrations in the mineshafts.

Resource Characteristics

Methane is found in a ll c oa l sea ms, althoughits amount per unit volume or weight of coaltends to be proportional to the rank (carbon con-tent) o f the c oa l: higher rank co a ls suc h as an-thracite and bituminous coals may have from 200to 500 cubic feet of me thane p er ton of c oal,whe rea s the low est rank lign ite m ay c onta in 30

to 100 cubic feet per ton (CF/t). Gas content alsoincreases significantly with depth; Kuuskraa andMeyer, in their analysis of coalbed methane gas-in-place, assign an average gas content to bitu-minous coal of 150 CF/t for 1,000 to 3,000 ft




depths, and 400 CF/t for depths greater than3,000 ft.14

The m etha ne is found either ad sorbe d to thecoal surfaces-by far the most abundant source—or trapped in the coal’s natural fracture system,or “cleat.’ The frac ture system te nd s to b e

aligned so that, in an idealized form, the fracturesresemble a series of vertical, parallel slices madein a block of c oal—the “ fac e c leats’ ’ —with another,

less we ll-de ve loped series of vertic a l slic es, the

“ butt cleats, ” perpendicular to the face c leats.Because most coal beds are aquifers, water is alsopresent in the fracture network, and its hydro-static pressure plays a key role in keeping themetha ne from d esorbing from the c oa l.

Technology

Because coal is essentially impermeable, meth-

ane production depends on intersecting the nat-ural frac ture netwo rk to provide pa thwa ys for thega s to flow to the w ell. A second condit ion nec-essary for economic levels of production is to pro-mote the resorption of the gas from the coal intothe frac ture system by reduc ing the pressure inthe fractures. This usually involves dewatering thecoal to reduce hydrostatic pressure. Because therate of resorption is not a linear function ofp ressure–as reservoir p ressure d rop s, resorpt ionmay remain low until a critical pressure isreached, and then accelerate rapidly as the pres-sure drops further-effective gas recovery may

require drilling wells on relatively close spacingand pum ping wa ter from them rap idly and simul-ta neo usly in orde r to m aximize the pressure d rop.This prod uc tion me thod will also help to outrun

wa ter infiltra tion into the c oa l sea m. This p rac -tice of c lose spac ing is in sha rp c ontrast to thewide spacing used in conventional gasfields, be-cause the close well spacing tends to reducerecovery per well in conventional fields.

A variety of metho ds c an b e used to ena blewe lls to intersec t the vertic a lly oriente d na turalfracture network. Horizontal wells may be drilled

I ~V. A. Kuuskraa and R. F. Meyer, ‘‘Review of World Resources

of Unconventional Gas, ” IIASA Conference on Conventional and

Unconventional World Natural Gas Resources, Luxenburg, Austria,

june 30-july 4, 1980.

from within a w orking m ine or a spec ially d rilledsha ft. The la tte r me tho d is extremely exp ensive,however. Vertically drilled wells may be slantedtowards the horizontal, ideally so as to run par-allel to a nd within the c oa l sea m. Keep ing thewell within the seam is difficult, however, andthere are substantial operating problems leadingto increased c osts. Hydraulic frac tures a lso c anbe used to connect the well bore to the fracturesystem . How eve r, ind uc ed frac tures in the coa lseams tend to be short and tend to parallel ratherthan intersect the planes of the natural fractures.in minable seams, the tendency of the fracturesto propagate vertically may result in damage tothe rock above the seam, a potential hazard tofuture mining. Finally, a variety of problems asso-ciated with we ll dewa tering, formation da ma ge,etc., still face future efforts to recover coal seammetha ne. Although for the m ost p art these p rob -

lems ap pe ar to be a ma tter of refining a nd up -grading existing methods and technology ratherthan accomplishing major innovations, consid-erable basic research is required to understandthe controlling gas production mechanisms, de-velop an exploration rationale for identifying at-trac tive d rilling sites, and de velop ad vanc ed we llstimulation technology.

Resource Estimates

Gas-in-place estimates for coal seam methanehave been made by a variety of analysts and orga-

nizations, including most recently the Gas Re-search Institute (1980), National Petroleum Coun-c il (1980), Kuuskraa and Meyer (KM) o f Lew in &Assoc ia tes (1980), and the Dep a rtme nt o f Energy(1984). These and others a re shown in tab le 33.

The three 1980 stud ies a ll use b asica lly the

same me tho d—to m ultip ly USGS-de rived esti-mates of coal tonnage, subdivided according torank, by estimates of gas content for each rank.The na rrow sp read of estima tes–398 TCF (N PC)

to 550 TCF (KM) —reflec ts the m ethodolog ica lsimilarity. These estimates should be consideredas quite c rude, bec ause the a vailab le d ata on ga s

c onte nt are limited a nd va riab le, and the USGSestimates of deep coal resources below 3,000 ft—pa rtic ularly important b ec ause g as content in-

creases with depth—are uncertain.



Ch . 7—introduction and Summary: Availability of Unconventional Gas Supplies q 135

Table 33.—Coalbed Methane Resource Estimates

Resource in placeStudv TCF

Department of Energy (1984) . . . . . . .Kuuskraa and Meyer (1980) . . . . . . . .National Petroleum Council (1980) . .

Gas Research Institute. . . . . . . . . . . .Federal Energy RegulatoryCommission (1978) . . . . . . . . . . . . .

Deul and Kim (1978) . . . . . . . . . . . . . .Wise and Skillern (1978) . . . . . . . . . . .TRW (1977 . . . . . . . . . . . . . . . . . . . . . . .National Academy of Sciences

(1976) . . . . . . . . . . . . . . . . . . . . . . . . .

68-395

550

398

500

300-850318-766300-800

72-860

300

Table 34.—Comparison of Recoverable ResourceEstimates (TCF)

Technically or economically recoverable gas:KM 40-60

a

NPC . : : : $2.50 $5.00 $9.00

5 25 45 10% RORb

2.5 20 38 15% ROR2.0 17 33 20% ROR

GRI . . . . $3.00 $4.50 $9.00

10-30 15-40 30-60

aTechnically recoverable resource.bRate of return


SOURCES Adapted from AGA Gas Energy Review, September 1982, and C WByrer, T H Mroz, and G L. Covatch, “Production Potential for Coalbed

Methane in U S. Basins,” SPE/DOE/GRl Unconventional Gas RecoverySymposium, 12832 1984

The rec ent DOE stud y, a p ortion o f its Me th-

ane Recovery from Coalbeds Project, is a basin-by-basin analysis targeting only the most likely

ga s-bearing sea ms in eac h basin. Although allbasins are not included and each individual basinestimate does not include all potential gas-bearingareas i n that basin, the focus on the most prom-ising coal seams implies that the estimates mayrepresent a good starting point for evaluatingrecoverable resources. However, the range ofuncertainty in the multi-basin estimate, 68 to 396TCF, was exag ge rate d b y the me thod used to c al-culate the extremes of the range.

Estimates of the recoverable resource base have

been made by the NPC (1 980), KM (1980), and

GRI (1981). A summary of results appears in table34. The GRI estima te is the result o f a poll of ex-perts. The NPC estimates are derived by first dif-ferentiating the ga s-in-plac e resource ac c ordingto estimated production levels (million cubic feetper day per well), and then comparing per wellrevenues at any given price to the estimated costsof an ‘ ‘average’ well in order to determinewhether the gas is economically recoverable attha t p ric e. Uncerta inties in t he NPC results stemfrom a series of very broa d assump tions ab outgas content, recovery efficiency, the relationshipbetween coal seam thickness and gas production,

the expected long-term production behavior ofgas wells in coal beds, and several other factors.An important criticism of the NPC analysis is thatit relies for its data base on isolated, previouslyd rilled we lls tha t a re c onsiderab ly less p rod uc -

t ive tha n new we lls drilled ac c ording to the mo d-ern practice of close pattern drilling, and thus istoo p essimistic . OTA c onc urs w ith th is c ritic ismbut feels that the other areas of uncertainty, somewith less pred ic ta b le effec ts, are a t lea st a s im-portant.

The Kuuskraa and Me yer ana lysis d iffers sub -stantially from the others in that it uses an analyticmo de l of gas prod uc tion from c oa l sea ms, trea t-ing p rod uc tion a s a simp le d iffusion p roc ess. Theresults are extremely sensitive to assumptionsabout the spacing of the vertical fractures in thecoal seam and the magnitude of the diffusion con-stant. Also, while simple diffusion may be thecontrolling factor in some coal beds, it is likely inmost cases that the actual physical process is con-siderably more complicated and the simple mod-

el used in the analysis will yield only very approx-ima te results.

In OTA’ s op inion , none of the e xisting a na ly-ses provide an adequate basis for reliably estimat-ing the size of the coal bed methane recoverableresource. It is possible that recent basin analy-ses sponsored by DOE might provide enoughnew ba sic da ta to form a ba sis for a m ore credi-ble estimate of recoverable gas. However, it isnot clear that we have sufficient understandingof the production mechanisms to provide a trulynew, credible estimate as yet. Because of this lack

of und erstand ing, a n estima te o f rec overab le re-sources that attempted to encompass the credi-ble resource possibilities would have to span awide range , prob ab ly on the o rde r of 20 to 200TCF or so.



136 q U.S. Natural Gas Availability.’ Gas Supply Through the Year 2000


Both NPC and GRI calculated annual produc-tion estimates based on assumed drillingschedules. The NPC estimate assumes a con-tinuously rising gas price from the present to2000, with prices reaching as high as $9/MMBtu(in 1979$) by 2000. production is projected to

peak at more than 2 TCF/yr in the late 1990s. The

GRI estimates are both price and technology de-pendent. At $3/MCF (1979$) and using existing

tec hnolog y, only 0.3 TCF/ yr of p rod uc tion is projected for the year 2000. At $6/MCF and ad-vanc ed tec hnology, 1.4 TCF are p rojecte d,

The level of unc ertainty a ssoc iated with the seestima tes is ve ry high. The p hysica l c ha rac ter ofthe resource base is highly variable, so that pastexperience, which is limited anyway, cannotserve w ell as a guide t o future p rod uc tion, The

physical mechanisms controlling gas productionfrom coal seams are not well understood. Fur-thermore, there are important uncertainties con-

cerning legal ownership of the gas, environmentalconstraints associated with water disposal, unre-solved mine safety issues, and other factors thatmay serve to constrain future gas development;the effects of these factors is difficult or impossi-ble to predict .

On the other hand, successful development ef-fo rts suc h a s U.S. Steel’s effo rt in the Blac k War-

rior Basin and others provide encouragement thatcoal seam methane could prove to be an impor-tant future gas source. Estimates projecting produc-tion o f 2 TCF/ yr by 2000 may see m e xce ssive fromour present vantage point but conceivably could

bec ome more credible with advanced technol-ogy and strong demand. Key targets for technol-

ogy d evelopment a nd research include cha rac ter-ization of the deep coal resource, improvementof fracturing technology and deviated drilling

technology, and improved understanding of thegeologic characteristics affecting gas recovery,lea ding to a reliab le e stima te o f the reco verableresource.



Chapter 8

Tight Gas



Contents

Page

Introd uc tion . . . . . . . . . . . . . . . . . . . . .. ...139

Cha rac teristics of the Tight

Ga s Re s o u rc e . . . . . . . . . . . . . . . . . . . . . . 1 3 9

1) l--low M uc h Gas is Present?. . .......1402) How Well Can the Gas Flow? .. ....1413) How Will Tec hno log y Perform .. ....142

Gas-In-Place Estimates . ................142Methodologies . . . . . . . . . . . . . . . . . . . . .142Estimate Comparison and Discussionof Uncertainties . . . . . . . . . . . . . . . . . . . .145

Te c hn o log y. . . . . . . . . .................149H yd rau l i c Frac tu res . .. . . . .. . . . .. . . . .149Reservoir Cha rac teriza tionTec hnolog ies . . . . . . . . . . . . . . . . . . . . .. .154

Explo rat ion Te c hno log ies . . .. . . .. . . .. .156New Tec hno log ies and the Ind ustry .. ..156

The Rec ove rable Resource Base a ndProd uc tion Potentia l . . . . . . . . . . . . . . . . . . 157Me tho dolog ies and Results . . . . . . . . . . .. .157

Comparison of Estimates andDiscussion of Uncertainties . ..........161Annual Production Estimates . .........168

TABLES

Table No. Page

35.36.

37.

38.

39.

Tight Ga s-in-Plac e Estima tes (in TCF) ..144National Petroleum Council’s Gas-in-P lace Est imates . . . . . . . . . . . . . . . . . . . .145

Conventional and Tight Gas WellsCompared . . . . . . . . . . . . . . . . .......150Ma ximum Rec ove rab le Tight Ga sRe s o u r c e s, TCF . . . . . . . . . . . . . . . . . . .157Economically Recoverable Gas atTw o Tec hno logy Leve ls (TCFs) . ......158

Table No. Page

40.

41.

42.

Lewin & Associates Base v.Advanced Technology. . ............158NPC Base v. Advanced Technology

and Regulation. . . . . . . . . . . . . . . . . . . .160

Projections of Annual Tight GasP r o d u c t i o n , TCF . . . . . .: . ............169

FIGURES

Figure No.

28. Primary Tight Gas Basins29. Problems With Reservoir

Page

. . . . . . . . . . . 1 4 3

Mapping in

30

31

32

33

34

Lenticular Formations . .............148Conceptual Fractures Created byMassive Hydraulic Fractures for

Blanket and Lenticular Formations. .. .151

Comparison of RecoverableEstima te s. . . . . . . . . . . . . . . . . . . . . .. ..162Comparison of NPC Study ResultsWith GRI Present Case–Tight GasSands Recoverable Gas. . ...........164Annual Tight Gas productionEstimates. . . . . . . . . . . . . . . . . . . . . .. ..170

Incremental Tight Gas SandsProduction Rates As a Function

of R&D Advances . . . . . . . . . . . . . . . .. 17235. Annual Production by Basin–NPC

$5.00/MCF (1979$), 150/0 DCF RORStandard Scenario . . . . . . . . . . . . . . . . .173

36. Incremental Tight Sands Gas

37

38

Production Rates With and Without

the Northern Great Plains Resource .. 173Incremental Tight Sands Gasproduction Rates As a Functionof Rate . . . . . . . . . . . . . . . . . . . . . . . . . . .174Incremental Tight Sands GasProduction Rates As a Functionof Discount Rate . . . . . . . . . . . . . . . . . .174



Chapter 8

Tight Gas

INTRODUCTION

“ Tight g as” is na tural ga s tha t is found in roc kformations of extrem ely low permea b ility. Suc hlow-permeability formations are found in all gas-producing basins in the United States. Depending

on the volume of gas-in-place and the economic

viab ility o f its extra c tion, tigh t g a s forma tions ma yconstitute a large potential gas resource.

Bec ause of t he low permea bilities, fluid flow(both gas and liquid) through tight gas formationsis highly restric ted . Com merc ial volumes of gascan only be recovered by artificially fracturing therock to increase the area of the reservoir in con-tact with the wellbore. Considerable difficulties

in measuring key reservoir parameters in tight for-mations and establishing the productive areas and

zones leads to high uncertainty about the com-merc ial viabiIity of the w ell until afte r an expen-sive frac turing trea tment is c omp leted and thewell has produc ed for some time.

The tight ga s resource , of a ll the unc onve n-tional na tural ga s resource s, is thoug ht to havethe most potential for contributing to U.S. supplyin the next 20 years. Low levels of g as have b ee nproduced from tight gas formations for manyyears; significant development began in the early

1970s with the successful use of massive hydrau-lic fracturing techniques in the Wattenberg Fieldin Colorado. Incentive prices established in 1979by the Natural Gas Policy Act (NGPA) increasedthe relative attractiveness of the resource and pro-mo ted deve lop ment in the e arly 1980s. Since

then, however, interest in gas production from

this difficult resource setting has declined as pricelevels for tight gas dropped in response to the cur-rent supply surplus.

CHARACTERISTICS OF THE TIGHT GAS RESOURCE

Tight g as forma tions are defined in this reportas low-permeability sandstone, siltstone, silty

shale, and limestone’ formations deposited incontinental, shoreline, or marine environments.Ea rlier stud ies ha ve used the te rms “ tight sand s”or simply “tight” formations in addition to tight

gas to refer to th is resource . Som e o f the ea rly

studies included Devonian shales as tight gas for-ma tions, but later stud ies dea lt with the sha les

separately because of their different productionmechanisms and characteristics. This chapter alsoexc ludes Devon ian sha les from the “ tight g as”category, treating them in chapter 9 as a sepa-rate unconventional resource.

I Except for the Edwards LI me format Ion I n the Southwest,

limestone formations have not been Included In tight ~as resource

aswssments. ]oh n S h a rer, Ga s Rewarc h I n st It ute, persona I com

munl[ atlon, 1984,

Tight gas reservo irs represent the low perme -ability end of a continuum of gas-producing reser-

voirs rather than a unique typ e. In the p ast, thecutoff between a convent ional and a t ight gasreservoir has been somewhat arbitrary, based pri-marily on the economics of production or re-quirements for special production techniques.

The p ermea bility leve ls used to d efine the up pe rbo unda ry for tight g as have rang ed from 0.01millidarcy (md)2 to 1 md ; most forma tions iden-

tified as tight have permeabilities less than 0.1 md.(For comparison, the permeability of cement ison th e o rd er of 0.001 red .) The g ran ting of a sp e-ciaI price incentive to tight gas under the NGPAnecessitated a more precise definition. In 1978,the Federal Energy Regulatory Commission de-

—..—2 A millidarcy, abbreviated a~ ‘‘red, ” is a standard unit of

permeability, which measures the ease with which flulds (llquld\

and gases) can flow through porous rock.

139



140 q US. Natural Gas Availability: Gas Supply Through the Year 2000

fined a tight gas reservoir as one having an aver-age permeability of 0.1 md or less at subsurface(in situ) conditions of confining pressure and

water saturation, and a maximum production rate

prior to stimulation, de pe nde nt on de pth, of 44MCF/D at 1,000 ft up to 2,557 MCF/D at 15,000 ft.

The b ounda ry betw een c onventional gas andunconventional tight gas is still changing and willco ntinue to change w ith changing eco nomicsand the further development of production tech-nology for low-permeability reservoirs. Conse-qu ently, all estimate s of resourc es and future pro-duction from tight formations must be evaluatedin the context of their defined boundary condi-tions. Currently, the American Gas Associationis reexamining its definition of the convention-al/unconventional boundary. It appears likely thata result of this process will be to transfer resourcesfrom the unconventional to the conventional cat-

eg ory. The subseque nt de c line in the projec tedtight gas potential actually would reflect therecognition that current technology can allow ac-

cess to a substantial portion of this resource.

In evaluating the potential gas recovery froma well drilled into a tight gas formation, threeque stions need to be answered: 1) How muc hgas is present? 2) What conditions exist that con-trol the flow of gas? and 3) Can extraction tech-nology be successfully applied in this setting?

1) How Much Gas is Present?

The o c c urrenc e o f ga s in a g iven sec tion of aformation is dep endent on w hether ad equa te

source rocks and appropriate temperature con-ditions have existed which allowed gas to form.The a bility of the ga s to m igrate and the p resenc eof a trapping and a sealing mechanism are fur-ther requirements.

In the case of most tight gas reservoirs, the pres-ence of interbedded organic shales or coal seamsensures an ad eq ua te source for gas forma tion,By the sam e to ken, migra tion is not a p rob lemsince the g a s d oe s not hav e to travel far. The low

permeability of the formations themselves, to-gether with the overlying shales or other sealingroc ks, c onfine the gas within the sandstone s. In

many areas, gas is being produced from nearby

conventional reservoirs, and the required tem-perature conditions may sometimes be inferred.However, any estimation of potential tight gas re-sources in unexplored areas should considerthese temperature conditions as additional uncer-

tainties.

Som e o f the tight ga s forma tions dep ositedund er sha llow ma rine c ond itions, suc h as tho sein the Northern Great Plains, may represent anunusual type of natural gas occurrence. In theseareas it has been suggested that the methane hasformed biogenically (and at low temperatures andpressures) through decomposition of organic ma-

terial by micro-organisms. s Biogenic gas forma-tion allows gas to be present in areas that other-wise m ight be assumed not to be ga s-bea ring,

because they have never been exposed to thehigher temperatures and pressures generally asso-ciated with gas formation (see ch. 3).

Given that gas is present, the amount presentin any p ortion o f a tight gas format ion is prima r-

ily a func tion of the p orosity, tem perature, andpressure. These p a rameters define the spac e

ava ilab le to be oc cup ied by g as molecules andthe amount of gas that can be found in each unitvolume of available space. Water saturation isalso a n important c riterion a s wa ter co mp eteswith gas for the available space.

Porosity is the frac tion of t he roc k tha t is void

space, i.e., the spa c e remaining b etwe en a ndwithin mineral grains, after the grains are packed

together. Dissolution and precipitation of mate-rial by fluids percolating through the rock mayalter the original porosity. Porosity of tight gas for-ma tions typ ic ally range s from 3 to 12 perce nt.4

Co nventiona l reservo ir porosities rang e from 14to 25 percent or more.

Pore size is an important determinant of the

water saturation of the rock, and thus of gas vol-ume. Very fine grained rocks such as siltstones

and chalks, may have high porosities but the in-d ividua l po res a re very sma ll, resulting in a h ighpo re surfac e a rea to p ore volume ratio. Wate r

3D. D. Rice and E. C. Claypool, “Generation, Accumulation, and

Resource Potential of Biogenic Gas, ” AAPG Bu//et in, vol. 65, No.

1, january 1981.4That is, 3 to 12 percent of the rock volume is void space.



Ch . 8—Tigh t Gas q 1 4 1

molecules may be adsorbed on pore surfaces, re-duc ing the volume ava ilab le to b e filled by ga s

and increasing the water saturation. Because ofsma ll pore sizes, tight g as forma tions are gener-ally characterized by high to very high watersat u rations.

2) How Well Can the Gas Flow?

The m ost importa nt reservo ir c ha rac teristic toconsider in terms of the recovery of gas from tightgas forma tions is permeability because this ulti-mately controls how fast the gas can be produced.

Permeability is a measure of the ease with whichfluids can move through interconnected pores ofthe reservo ir roc k in response to a pressure g ra-dient. Thus, a volume o f roc k ca n b e b oth p o-rous—have a high percentage of void space—and have ve ry low permea bility if the individua l

void spaces are not interconnected or the chan-nels are too narrow to a llow ga s to m ove freely.Permeabilities of conventional gas reservoirsrange from 1.0 md to several darcies (thousandsof millidarcies). In contrast, the permeability ofa tight gas reservoir can be as low as 0.00001 md

(although at present, recovery generally is limitedto reservoirs with permeability greater than about0.007 red).

The low p ermea b ilities of tight g as format ionsresult from a combination of factors that closeoff connections between pores, including smallgrain size, high clay content, and the cementa-

tion of g ra ins resulting from the prec ip ita tion ofd issolved ma teria ls.

In very fine grained and recrystallized rocks,where connections between pores are very small,the level of water saturation is a key factor indete rmining relative ga s permea bility. Seve ra lstudies have demonstrated that water saturationson the order of 60 to 70 percent can effectivelyred uce the p ermeab ility to g as to zero.

Clay content and composition in a formationis another important factor in determining its

permeability. Many clays are formed after depo-sit ion a nd tend to g row in such a w ay a s to b loc kpores and pore throats. Certain types of clays areexpa nda ble on c onta c t with freshwa ter. The largevolumes of wa ter introduc ed during d riIIing into

a formation by drilling muds and fracturing fluidsc an c ause suc h c lays to swe ll, further red uc ingpermeability of the formation. Gas and water flow

can also reduce permeability by displacing andplugg ing o pe nings with loosely ag grega ted or

platey c lays.

All the factors discussed above affect the ma-trix permeability of the reservoir rock itself,

d isreg arding any fa ults or frac tures. The bulk

permeability of the entire reservoir, however, maybe greater if there is a well-developed natural frac-ture system. Natural fractures occur as a result

of unequally distributed stresses at any time dur-ing or afte r forma tion of the roc k. They a re ub i-

quitous in all rock formations and occur at allsc a les. Large frac tures ma y extend for long d is-tanc es, while m ic roc rac ks are found at the sc aleof the smallest grain. At greater depths, underhigh overburden pressures, most small cracks and

som e larger frac tures are c losed . Depo sition ofc rysta lline ma teria ls by migra ting fluids a lso c anseal off fractures. s Nevertheless, the existence ofna tural fra c tures is extrem ely imp ortant to the ne tpermea b ility of tight g as format ions.

The flow o f gas in tigh t forma tions is a lso a func -tion of the shape of the reservoirs and their lateralc ontinuity—whe ther or not they e xist a s c ontin-

uous bod ies sp rea d ove r la rge area s or instea doccur as multiple smaller, discontinuous units.

Lateral continuity and geometry of gas-bearingunits control how much gas can find a flow path

to the wellbore. Most studies have responded tothe differences in these characteristics among theva rious tigh t fo rma tions by subd ividing the for-

mations into two categories commonly referredto as blanket formations and Ienticular for-mations.

Blanket formations consist of continuous gas-bearing deposits that extend laterally over a largearea. Blanket reservoir units, 10 to 100 ft thick,may be composed of sandstone, siltstone and siltyshale, or chalk or limestone interbedded withvery low pe rme ability sha les or non ma rine de-

‘However, If fractures form after hydrocarbons ha~e d)splared

the water layer, the fractures usually remain open. The mineral

saturated water is no longer present to precipitate solid crystals (Ovid

Baker, Mobile Research & Development Corp., personal commu -

nlcatlon, 1984).




posits, including coal seams. Alternatively theyc an oc c ur as millimete r to c entimeter thic k sand -rich layers, or “stringers,” alternating with clay-rich laye rs. The bu lk of c urrent tight g as prod uc -tion is from reservoirs in blanket formations.

Lenticular formations c onsist o f m a ny relatively

small, laterally discontinuous sandstone andsiltsto ne units, or “ lenses,” intermingled withshales and sometimes coal seams. These units aresimilar in mineral composition to the blanketsandstones except that they tend to have higherc la y c onte nts. They occur stacked vertically one

over the other, in formations hundreds of feetthick. The size, o rientation, a nd ge om etry of theIenticular units are variable. For example, some,suc h a s units formed by the filling of strea m c ha n-nels, are long and narrow and may have a pre-ferred orientation. Other deposits formed at thebends of stream meanders tend to be shorter and

wider and more randomly oriented.

The d isc ontinuous na ture of Ient icula r forma -

tions and the va riab le ge om etry of the lenses

ma kes them a more d iffic ult resource to qua n-tify, in terms of b oth g as-in-plac e a nd p ote ntial

for rec ove ry, For these reasons, despite their ex-

tensive occurrence, they have only recently beentargeted as a potential resource.

3) How Will Technology Perform?

The third c ritica l va ria ble for eco nom ic rec ov-ery of tight gas is the viability of massivescale wellstimulation and other extraction technologies inthe geological settings containing tight gas. Herethe issues encompass a large number of specifictechnology issues ranging from the ability to iden-

tify, using well logs, the attractive zones for gasrecovery, to developing fluid systems that canhelp c onta in a large vertic al frac ture w ithin ag iven roc k interva l. While wo rk is p rogressing in

these a rea s, much ne w resea rc h and de velop-ment is required before efficient technologies will

exist for tight g as extrac tion. The key tec hnolog iesare discussed later in this chapter.

GAS-IN-PLACE ESTIMATES6

A number of estimates have been made of the

resource base and production potential of t ightgas formations, beginning in 1972 with the Fed-eral Power Commission’s (FPC) National Gas Sur-

vey report. In 1978, the Fed era l Energy Reg ula-tory Commission’s (FERC) report updated andexpa nde d o n the FPC estima te of the tight g asresource . In 1978 and 1979, Lew in & Assoc iate s(Lewin), under contract to the Department ofEnergy, made a detailed appraisal of 13 tight gas-bea ring basins. In 1980, the Na tiona l Pet roleum

Co unc il (N PC) p ub lished its extensive stud y onthe to ta l U.S. resource in tight fo rmations, co m-

bining resource estimates for 12 basins that hadundergone an extensive appraisal, and 101 addi-tional basins whose estimates were based on ex-

61 n this section, a ncf the section on recoverable resources later

in the chapter, OTA has chosen to emphasize the 1980 NationalPetroleum Council Study on Tight Gas In its discussions of previ-

ous resource estimates. This emphasis is based on the excellent

reputation of the NPC report, which is widely cited in discussions

of the tight gas resource and future production potential, the re-port’s extensive documentation of assumptions and methodology,

and the fine level of detai I i n the basin and subbasi n analyses.

trapolation from the appraised basins. Finally, in1984, the G as Resea rc h Institute c ond uc ted a

series of sensitivity analyses of the tight gas re-source base, based on the NPC methodology.

Early estimates of the tight gas-in-place by theFPC and FERC only determined the resource forspecific basins. Subsequent estimates expandedthe number of basins under consideration andundertook more detailed appraisals within the in-dividual basins as more data became available.

The basins included in the various studies areshown in figure 28. A chronologic representationof the c hang ing resource estima tes for the ap-praised basins is g iven in tab le 35. Note that in-creasing the area under consideration has not

necessarily increased the estimated size of the re-

source.

Methodologies

The FPC and FERC studies defined the amountof gas within a unit reservoir volume by estimat-



Ch. 8—Tight Gas q 143

Figure 28.—Primary Tight Gas Basins

Basins included in resource assessmentsFPC 3, 5, 6 \J

FERC 1, 2, 3, 5, 6, 8

Lewin 1-8, 10, 12, B, D, HNPC 1-12 excluding Cotton Valley “sour”

1. Western Region1. Northern Great Plains

2. Williston

3. Greater Green River

4. Wind River

5. Uinta6. Piceance

7. DenverA. Snake River

B. Big Horn

C. Wasatach

D. Douglas Creek

E. Western Shallow Cretaceus Trend

NOTE Coastal regions include offshore areas

SOURCE National Petroleum Council

Il. Greater Southwest Region Ill. Eastern Region

8. San Juan L. Appalachian

9. Val Verde-Ozona Trend M. Black Warrior

10. Val Verde-Sonora Trend

11. Edwards Lime Trend

12. East Texas/North Louisiana Basin—Cotton Valley Trend

F, RatonG. AnadarkoH. Ouachita1. Arkoma

J. Fort WorthK. Western Gulf Coast

ing average gas-filled porosity7 and average reser-voir temperature and pressure conditions for

each of the formations or sections of formationsthat they evaluated. Using these figures togetherwith the net “pay” (i. e., gas-bearing) thickness

and the total productive area of the formation,each study team calculated the total gas-i n-place

in the reservoir.

—.~otal porosity times the fraction of void space filled with gas.

The FPC, in its 1972 study, set minimum cri-teria defining its interpretation of gas that couldconceivably be considered as recoverable. It re-stricted its evaluation to reservoirs with a mini-mum net pay thickness of 100 ft, less than 65 per-

cent water saturation, 5 to 15 percent porosity,and permeabilities of 0.05 to 0.001 md. It in-

cluded only formations at depths between 5,000and 15,000 ft and defined a minimum reservoirsize of 12 square miles. It also only considered




reservoirs in remote areas, and, among these,only reservoirs not interbedded with high-perme-ability aquifers.8 Its study concluded that the threeRocky Mountain Basins alone contained some600 TCF of tight gas (table 35).

The FERC study used the same methodology

as the FPC to calculate the gas-in-place butrevised some of the criteria and expanded thearea evaluated to include the San Juan Basin andthe Northern Great Plains. They included forma-tions with thinner pays (to 20 ft thick) and depthsas shallow as 1,500 ft, and allowed lower gas-filled porosities. Despite the expanded area andmodified formation parameters, the size of the

tight gas resource only increased from 600 to 793TCF (table 35).

Both the FPC and the FERC recognized thattight gas resources probably existed in other

gas-bearing basins. Since little data existed inthese areas, however, they felt the size of the ad-ditional resource could not be properly evaluated.

The Lewin and NPC studies used considerablymore elaborate methodologies to calculate thegas-in-place. Their methods were designed to

8These last two criteria were included because nuclear explosives

were being considered as a stimulation mechanism. The FPC ex-

cluded large portions of known tight gas resources which occurred

in more populated areas. The FERC study removed these exclu-

sions because by 1978 it seemed evident that nuclear explosives

would not be used.

take into account the variability of physical prop-er-ties, such as permeability, porosity, and net pay(total thickness of gas-bearing zones), within a for-mat ion.

The Lewin study used available well data to

divide each formation into “subareas” based onhomogeneous physical characteristics. The po-tentially productive portion of each subarea wasdetermined by multiplying by the wildcat successrate (fraction of wildcat wells that are commer-cial successes) for that region, The volume of gas-in-place calculated for each subarea was furtheradjusted to reflect a log-normal distribution of

“pay quality.”9 The Wattenburg Field in theDenver Basin, a relatively well-developed tight

sands gasfield, was used as a model to determinea characteristic distribution of pay qualities.

The NPC methodology for calculating the gas-

in-place in its appraised basins was similar to theLewin approach. Instead of using the same payquality distribution across all fields, however, theNPC determined a distribution of pay quality and

other reservoir properties for each individual

basin or subbasin based on existing well data. Inthis way it could calculate the gas-in-place for upto six permeability levels. The NPC study alsoevaluated a slightly larger area (13,000 more

qln the ~ewin report, pay quali ty is a funCtiOn of porosity,

permeability, and thickness of the gas-bearing zone.

Table 35.-Tight Gas-in-Place Estimates (in TCF)

FPC 1973 FERC 1978 Lewin1978 NPC 1980Appraised basins (gas-in-place) (gas-in-place) (gas-in-place) (gas-in-place)

Northern Great Plains/ Williston . . . . . . . . . . . . . . . . . — 130 74 148

Greater Green River . . . . . . . . . 240 240 91 136Uinta . . . . . . . . . . . . . . . . . . . . . . 210 210 50 20Piceance. . . . . . . . . . . . . . . . . . . 150 150 36 49Wind River . . . . . . . . . . . . . . . . . — — 3 34Big Horn. . . . . . . . . . . . . . . . . . . — — 24Douglas Creek. . . . . . . . . . . . . . —

— — 3

Denver . . . . . . . . . . . . . . . . . . . . — —

— 19 13San Juan . . . . . . . . . . . . . . . . . . — 63 15 3Ozona . . . . . . . . . . . . . . . . . . . . . — .— — 1

Sonora . . . . . . . . . . . . . . . . . . . . — .— 24 4

Edwards Lime . . . . . . . . . . . . . . — .— — 14Cotton Valley “sweet” . . . . . . . — .— 67 22Cotton Valley “sour” . . . . . . . . — .— 14Ouachita. . . . . . . . . . . . . . . . . . . —

—. - 5 —

Total . . . . . . . . . . . . . . . . . . . . 600 793 423 444



Ch. 8—Tight Gas q 1 4 5

square miles) of potentially productive land thanLewin, although the two sets of appraised basinswere similar (Lewin appraised 10 of the 12 NPCbasins). In addition, the Lewin study did notevaluate gas-in-place for permeabilities less than

0.001 md, whereas the NPC study included a

number of areas with permeabilities as low as0.0001 md and some areas with permeabilitiesas low as 0.00001 md. Despite these differences,however, the Lewin and NPC reports do not dif-fer substantially in their total estimates of gas-in-place in the appraised basins (423 v. 444 TCF,respectively). On the other hand, the estimates

for individual basins do differ considerably.

In addition to the 12 basin appraisals, the NPCstudy made the first attempt to estimate the totaltight gas resource occurring in all gas-producingbasins in the Lower 48 States. To do this it ex-

trapolated the results of its detailed basin ap-praisals to 101 remaining potential gas-bearingbasins. Extrapolated basins were classified accord-ing to their similarity to appraised basins; certainformations in the appraised basins were chosenas analogs to formations in the extrapolatedbasins. The NPC estimated an additional 480 TCFin place in the extrapolated basins as shown intable 36. Its total gas-in-place resource estimatefor the U.S. Lower 48 States is 924 TCF.

Estimate Comparison andDiscussion of Uncertainties

Appraised Areas

The sequence of gas-in-place estimates for tight

gas represents a continuing refinement of the esti-mation process. Each estimate builds on the

former–adding new data and evaluating newareas. In general, the more detailed analyses havetended to produce lower gas-in-place estimatesfor a particular area. For example, the early FPCand FERC estimates of gas-in-place for the Uinta

Table 36.—National Petroleum Council’sGas-in-Place Estimates

Appraised Extrapolated Total

(12 basins) (101 basins) (113 basins)

Potential productive

area (square miles) 53,000 68,500 121,500Gas-in-place (TCF). 444 480 924

SOURCE NPC Reporl vol V part I table 1

Basin are 150 TCF. The later Lewin and NPC esti-mates for the Uinta are only 50 and 20 TCF, re-spectively. These reductions can be in part ac-

counted for by an increasing realization of highwater saturation in the tight gas reservoirs andits negative effect on both gas content and gas

permeability.10 As discussed above, high watersaturations tend to reduce the total volume of gasin the reservoir rock as well as restrict the gasflow. The NPC study also reduced the assumed

thickness of total pay intervals in the Uinta froma 500- to 1,000-ft range to less than 500 ft, re-ducing the overall volume of the gas-producingzone and thus the total gas-in-place.

The NPC estimate represents the most compre-hensive estimate of the gas-in-place resource basethat exists to date. Because of its level of detailand its attempt to include gas in all gas-bearing

basins, it probably represents the best availablequantitative assessment of the size of the gas-in-place resource for tight gas. It shares, however,a drawback common to all the estimates: in

OTA’s opinion, none of the estimates adequatelyquantifies the extent of the uncertainty associatedwith the gas-in-place calculation.11 This may re-sult in the impression that the gas-in-place hasbeen very narrowly defined, whereas the actualrange of uncertainty may be quite large.

In most cases the estimators were fully aware

of factors contributing to uncertainty. To the ex-tent that their estimates are used by producersand others familiar with the industry, the level

of uncertainty may be understood. When the esti-mates are to be used by those without such a

common background, the inherent uncertaintiesneed to be explained in some detail. The follow-ing discussion describes the important factors

IoStrictly speaking, permeability does not affect the magn Itude

of gas-in-place, whereas gas content most certainly does. l-iowe~er,

most estimates of gas-in-place do not attempt to Include every mol-

ecule of gas, and may exclude gas from formations that have no

recoverable gas. Because permeability does atfect recot era blltty,

an altered estimate of permeability may cause a change In est)mated

gas-in-place. Nevertheless, gas content is the more closely related

of the two variables to gas-in-place.

I IThis statement IS not meant to imply that the NPC report dldnot discuss uncertainty. To the contrary, the technical reports pre-

sented substantive d iscusslons of the u ncertai ntles I n key parame-

ters, and used probability distributions rather than point estimatesI n several key calcu Iatlons. However, these d Iscusslons and calcu-lations were not translated Into error bands around the report’s pro-

jections of resources and future production.




contributing to uncertainty and their implicationsfor gas-in-place estimates.

Differences and uncertainties in volumetric esti-mates of the tight gas resource are often a func-tion of the level of understanding of the geologichistory of the basins and the physical character-istics of the formations. Critical parameters in-

clude the porosity and water saturation, the areal

extent and thickness of the gas-producing por-tion of a formation, and the actual presence of

gas and its geologic origins.

As discussed earlier, porosity and water satura-tion constrain the amount of gas present in adiscrete volume of rock. However, these param-eters are extremely difficult to measure accuratelyand may vary over a wide range within a smallarea, Available data consist of extrapolations fromexisting wells and are often insufficient for a valid

statistical analysis of a potential producing for-mation. Furthermore, conventional techniquesfor measuring and interpreting reservoir param-eters, where they have been used in the tight gasintervals, are often not applicable to these types

of formations. For example, conventional inter-pretation of well log data frequently results inhigher measured porosities than those actually

observed by laboratory analysis of core samples.Errors in porosity measurements can lead to even

larger errors in computed levels of water satura-tion, compounding errors in the estimates of gas-in-place. In the last few years, interpretative tech-niques have been modified to apply specificallyto low-permeability intervals, but these tech-niques are not yet routinely used.

The extent to which measurement problems of

this sort have affected tight gas resource estimatesis difficult to determine, although these problems

are likely to have been acute for the earlier stud-ies. I n the NPC analysis, well performance cal-culations for future wells were compared withactual well performance in producing wells tocheck the procedures for estimating porosity andpermeability .12 However, despite recent advances

in logging systems and interpretation techniques,there still were severe measurement problems at

the time the NPC estimates were developed. In

llovld Baker, Mobil Research & Development cOrP., Pf?rsOnal

communication, 1984.

fact, the discussion of measurement problems

that appears in the NPC report itself, reproducedin box H, implies that the NPC analysts could nothave been overly confident about the precisionof their resource calculations.

The areal extent of some tight formations, par-ticularly blanket formations, may be fairly well

documented in developed basins from data col-lected from exploration and development wells

drilled for conventional gas resources. Pay thick-ness may be fairly easy to determine for someblanket formations in which sandstone beds are

clearly defined. I n other blanket formations, thingas-bearing sand stringers occur finely inter-bedded with shales, making determination of thenet productive thickness extremely difficult.

The lateral continuity, thus areal extent, of the

gas-bearing portions of Ienticular formations may

be more difficult to determine than for blanketformations since the geometry of the individuallenses cannot be readily understood just by drill-ing and logging multiple wells within a formation.

For example, sand-rich intervals occurring at ap-proximately the same stratigraphic level in twoor more wells may or may not represent the same

lens (see fig. 29). Extrapolation from mapped sur-

face features of these formations is the best way,

at present, to determine dimensions, distribution,and orientations of the sand lenses. However, vis-ible surface manifestations of Ienticular formationsare not always conveniently present. Another wayto approximate the volume of Ienticular tight gasformations is to determine an average sand-to-shale ratio based on well log data for a portion

of the formation and extrapolate this over the areacovered by the formation. The accuracy of thismethod depends on the number of wells drilled

through the formation.

The presence of appropriate source materialsand temperatures and/or biologic agents for hy-drocarbon generation is a function of the geologichistory of a basin. Many tight gas resource esti-mates consider a large portion of the tight gas in-

terval in the Rocky Mountain Basins to be gas-bearing. There is, however, a dissenting opinion:that temperature, pressure, and source rock com-position throughout many Ienticular basins, suchas the Piceance, were not appropriate for the gen-



Ch. 8—Tight Gas q147

Box H.—What the NPC Report Had to Say About MeasuringKey Tight Gas Resource Parameters

“Experience to date indicates a broad variation of the critical parameters even within the same for-mation. Permeability routinely varies by orders of magnitude; gas-filled porosities can vary by factors

of 2 to 5; net pay can vary by factors of 2 to 50. Few data are available on the distribution of lens sizeand geometry, but extremely wide variations can be expected. Experience indicates that expensive,time-consuming measurements and analyses made for one well may not be sufficiently applicable toanother well in the same basin to design a fracture treatment . . . The significance of this is that it sub-stantially reduces the ability to extrapolate from one well to another. Thus, extrapolation over a widearea is difficult.

“In very low-permeability gas reservoirs, existing methods and tools for measuring critical parame-ters have been found to be inadequate for accurately characterizing the producing formations. Extremelylow permeability renders data from drill stem tests nearly useless. Well logging has failed to adequatelydistinguish gas-productive from water-productive zones. Net pay thickness and gas-filled porosity aremeasured with very low reliability. Current coring techniques tend to alter the rock properties priorto laboratory testing. Conventional laboratory tests of permeability have been shown to produce resultsthat vastly overstate actual permeability under reservoir conditions of water saturation and pressure.The distortion is greater at lower permeabilities, where recovery is more sensitive to permeability. Lab-oratory measurements and tests of rock strength and hardness are only now being developed and stand-ardized. No downhole permeability measurement technique has proved reliable in tight formations.Pressure testing of wells in tight formations requires vastly longer time periods than for conventionalformations (e.g., weeks v. hours) for comparable precision. Measurements of lens geometry have beenlimited to studies of outcrops of the formations. The nature and distribution of the subsurface gas-bearinglenses have not been characterized. Reservoir modeling also needs improvement. Competenttwo-dimensional flow models are available, but three-dimensional models have only begun to be de-veloped and are exceedingly costly to operate.

“Net pay is one of the most important parameters required for reservoir performance evalua-tion . . . Current methods for determining net pay have a very high probability for error.

“Accurate knowledge of permeability is essential for , . . predicting potential well perform-ance . . . Current methods for obtaining permeability . . . exhibit questionable reliability.”

SOURCE: National Petroleum Council, Unconventional Gas Sources: Tight Gas Reservoirs, Part 1, December 1980,

eration of large quantities of gas.13 This opinionholds that only a small proportion of the thou-sands of feet of vertical section of tight sandstonesis actually gas-bearing. If confirmed, this would

substantially reduce the estimated gas-in-place,and make basins like the Piceance much less at-

tractive for potential production. Similarly, theNPC estimate of the gas-in-place in the Denver

Basin included a caveat that as much as 30 per-

cent of the estimated gas-in-place in that basinmay not actually exist. Temperature and pressureconditions in the northern part of the basin maynever have been high enough to generate gas.

1‘j.W. ~ratton, ‘ ‘Fracturing Technologies for Gas Recovery F rom

Tight Sands, t’ contractor report to OTA, September 1983,

The volumetric estimate of gas-in-place in the

Denver Basin is based on gas contents extrapo-lated from the known gas-bearing portions of theformations occurring in the southern part of thebasin.

The most controversial of the appraised areasis the Northern Great Plains, estimated by theNPC to have nearly 150 TCF of gas-in-place in

shallow formations. Two-thirds of this gas is esti-

mated to be technically recoverable. The con-troversy centers around the origin of the gas,whether it is early thermogenic or biogenic, 14 and

.—14That Is, ~ormed by physical processes associated with high tem-

peratures (thermogenlc) or else by biological processes, usuall y

anaerobic d}gestlon (blogenlc).




Figure 29.—Problems With Reservoir Mapping inLenticular Formations -- -

1 2 3

Physicalreality

I Lenticular

sand/shalesequence

Blanketsandstone

some argue that there is likely to have been sig-nificant gas loss since the time of formation,unless sealing mechanisms have been unusuallyeffective.

Given the large discrepancies in estimates ofgas-in-place, and the important implications for

the natural gas resource base, the questions asso-ciated with the origin of the Northern Great Plains

gas and its preservation need to be resolved.

As might be expected, dissent from the NPCestimates is not restricted to those who are morepessimistic or are simply skeptical of the accuracyof the estimates. For example, several panelistsat OTA’s Workshop on Unconventional GasSources felt that the NPC’s resource estimatesreflected an overly conservative approach which

tended to discount or dismiss resource potentialunless there existed definitive evidence of its ex-istence. It was claimed that, in some basins whereextensive drilling records were available, NPCgeologists assigned either zero or a heavily dis-counted value of gas-in-place to sections wherethere had been no “gas shows, ” even when the“no shows” resulted simply from a lack of drill-ing. A case in point is the Denver Basin, men-tioned previously; in this basin, the NPC assumedthat only 1,600 of a total of 45,000 sections, or

3.6 percent, are gas-bearing. In the words of oneof these critics, this approach “places an unjusti-fied premium on the exploration technology andwisdom of our predecessors in the petroleumbusiness.” 18

In the opinion of these same critics, the con-

servatism displayed in the NPC study’s estimatesof productive area is duplicated in its estimates

of the thickness of the gas-bearing rock in the pro-ductive formations. For example, in the NPC anal-ysis of the Piceance Basin, one of the Ienticularbasins where high gas-in-place estimates havebeen questioned, the four gas-bearing formationshave a total thickness averaging about 6,400 ftbut are assigned estimated “net pay ranges, ” orgas-filled thicknesses, of:

Formation Net pay range (ft)

1. Ft. Union. . . . . . . . . ... . . . . 12-802. Corcoran Cozette . . . . 16-703. Mesaverde. . . . . . . . . . . . . . . . 20-2004. Lower Cretaceous-Jurassic . 7-35

. —18Ovid Baker, Mobil Research & Development Corp., letter of

Aug 3, 1984, to OTA.



Ch. 8—Tight Gas 149

The ranges are not additive, but at no point doesthe NPC assume that the gas-filled portion of theformations occupy more than 6 percent of thetotal thickness of sediment.

It is unlikely at this time that the existing argu-

ments about the relative “optimism” or “conserv-atism” of the NPC’s estimate of gas-in-place canbe resolved. In OTA’s opinion, the NPC estimate

of gas-in-place for the appraised basins, excluding

the Northern Great Plains, probably should stillserve as a reasonable “most likely” estimate, eventhough a considerable error band—at least +/- 100TCF—must be assigned to the estimate. As for the

Northern Great Plains, there appears to be a sub-stantial possibility that the NPC estimate is toohigh, but the evidence is by no means conclusive.

Extrapolated Areas

In the NPC analysis, the uncertainties that apply

to the appraised basins are magnified in the ex-trapolated basins. Not only is there a lack of ex-ploration and production data for these basins,but also the gas-in-place estimates were not gen-erated by geologists experienced in those basins,as was the case for the appraised basins. Extrap-

olations were made by assigning gas-in-placevalues to the estimated productive area in each

extrapolated basin using formations in appraisedbasins as analogs. In some cases this approachhas a high potential for error, since different partsof the country have undergone significantly dif-ferent geological histories which may have af-fected the amount of gas formed and preserved,

but which could not be properly accounted forin the estimation process.

For example, the Eastern region (primarily the

Appalachian and Black Warrior basins) was esti-mated to contain 49 TCF i n shallow formationsbased on analogies with formations in the North-ern Great Plains. However, the Northern Greatplains estimates assume a biogenic origin for the

gas, leading to the large volumes of gas-in-place.Even if biogenic gas formed in the Appalachianand Black Warrior basins, given their age (300million to 400 million years relative to 75 million

to 100 million years) and complex geologic his-tory, it is unlikely that much biogenic gas wouldhave been preserved to the present day. There-fore, the gas content per unit rock appears un-likely to be the same in the two regions.

There was widespread agreement among the

NPC study participants interviewed by OTA thatthe gas-in-place estimates for the extrapolatedbasins were highly uncertain.

TECHNOLOGYThe characteristics of the tight gas resource

base present gas producers with a variety of im-portant problems in locating and exploiting this

gas. Many of these problems result from the verylow permeability of the tight gas reservoirs andthe consequent need to use fracturing to achievecommercial levels of gas production. For exam-ple, full exploitation of the Ienticular reservoirsrequires the ability to contact lenses remote fromthe wellbore, yet our current ability to predict orcontrol fracture length and direction is poor. Inaddition, the presence of water-sensitive clays

raises the potential for formation damage fromthe fracturing fluids. Difficulties in accuratelymeasuring reservoir characteristics greatly com-plicate fracture placement and add to the eco-

nomic risk of stimulation. Also, the depth and

structure of some of the tight formations lead toextremely demanding requirements for fracturefluids and proppants. In this section, these prob-lems and the technologies available to overcome

them are briefly discussed. A longer discussionof fracturing technologies is presented in apdix B.

Hydraulic Fractures

Gas flows from a reservoir towards a well

pen-

borebecause of a pressure difference between thereservoir and the well bore. The rate of flow is de-

pendent on the difference in pressure and thepermeability of the formation. Fracturing is de-signed to increase flow rates by cracking thereservoir rock, exposing more of the reservoir sur-




face to the lower wellbore pressure. Hydraulic

fracturing is accomplished by pumping large vol-

umes of fluid down the wellbore, increasing thepressure on the rock formation until it breaks

down and fractures. Because the fractures wouldtend to close when the fluid is removed—espe-

cially in deep reservoirs where the pressure ofthe rock is great—sand or other materials areadded to the fluid. Left behind when the fluid isremoved, these “proppants” are wedged into thefractures and prevent them from closing,

Hydraulic fractures tend to be unidirectional,

generally extending out in opposite directionsfrom the wellbore. By convention, their lengthis measured along one wing (see fig. 30). Theirdirection and orientation (vertical, horizontal, orinclined) are controlled by the regional stressregime and the depth of the target formation. In-

duced fractures at depths greater than 2,000 ftare oriented in the vertical plane. At shallower

depths, such as might be found in the biogenic

gas reservoirs of the Northern Great Plains, frac-ture orientation may be horizontal.

Table 37 illustrates the need for fractures in tightformations and the substantial benefits that maybe gained from larger fractures. As shown in thetable, a well in a highly permeable (10 md) reser-voir can drain all the gas in two sections, or 1,280acres, of a field over a 15-year production life.An unfractured

19 well in a tight (0.001 md) blan-ket sand can drain only about 20 acres of the field

during a 30-year production life, and during that

I gActually, the process of ‘‘completing” the We l I involves an eX-

plosive perforation of the well lining that creates a small tracture,

here assumed to be 100 ft in length.

time would average less than 40 MCF/D of gasproduction, a very low rate. A massive hydrau-

lic fracture creating a 1,000-ft fracture would in-

crease the average production rate by nearly five-fold and allow complete drainage of the field with

six wells per section rather than nearly 30 with

unfractured wells. An advanced technology pro-ducing a 4,000-ft fracture would increase theaverage production rate by a factor of 15 andallow the tight field to be drained with a 320-acre(two wells per section) spacing, a fairly commonspacing in conventional gasfields.

Successful fracturing in tight formations is com-plex and faces substantial obstacles. A few criti-

cal points should be understood. First, the greatmajority of tight gas recovery heretofore has beenrestricted to areas “characterized by thick, fairlyuniform, blanket-type formations , . . (where) . . .

only a limited knowledge of the formation char-acteristics is necessary to stimulate economic pro-duction rates.”20 The majority of the resource,however, is more complex, and greater under-

standing of the geology and production mechan-ics is critical .21 Second, although fracturing datesfrom the 1800s, and hydraulic fracturing datesfrom 1947 and has the benefit of the experiencegained by thousands of separate fracturing treat-ments, the type of massive hydrauIic fractureneeded to begin to fully exploit the tight resource

has only been developed in the last 10 years orso. The process is not fully understood and ex-

trapolation to new geologic situations is difficult.

————Zocas Research Institute, Status Report.” GRI’s Unconventional

Natural Gas Subprogram, December 1983,

Z’ Ibid.

Table 37.—Conventional and Tight Gas Wells Compared

Wells per Reservoir areaFracture Ultimate Production section to Exposed by

Perm. length production life produce fractureType well (red.) (ft) (MMCF) (years) (all gas) (sq. ft)

Conventional . . . . . . . . . . . . . . . 10 100 31,700 15 0.5 34,000Tight—unfractured . . . . . . . . . .

0.001 100 390 30 29 34,000Tight—present technology ., . 0.001 1,000 1,838 30 6 340,000 Tight—advanced technology . 0.001 4,000 6,015 30 2 1,360,000NOTES Net pay = 85 ft.

Initial pressure = 5,500 psiMaximum recovery, gas/section =16 BCF for conventional,

11 BCF for tight gas wells.

SOURCE: R. W. Veatch, Jr and O Baker, “How Technology and Price Affect U S Tight Gas Potential, ” Petroleum Engineer International, January 1983.




Figure 30.—Conceptual FracturesBlanket and

Created by Massive Hydraulic Fracturing inLenticular Formations

rBlanket formations ~Well

bore

Casing

v

1 \ \ ”

Sand-f i l led

region offracture

Fractur

SOURCE: National Petroleum Council, Unconventional Gas Sources Tight Gas Reservoirs, Part /, December 1980




Third, aside from the difficulty of forecasting whata fracture will do, it is hard to tell in any detailwhat a fracture has done even after it has been

completed and the well is producing (or hasproved to be unproductive). In addition, as with

any new and rapidly developing technology,

state-of-the-art techniques are not necessarilystandard practice in commercial ventures. Forthese reasons, our extensive experience in frac-turing has not been as much benefit in project-ing future performance as might have been ex-pected.

Despite these difficulties, however, well serv-

ice companies and producers have been re-sponding vigorously to the various problems

encountered in tight formations and have devel-oped a number of variations to the standard frac-ture treatments. Of particular importance are thedevelopment of new fracture fluids and prop-pants which can increase fracturing efficiency andfracture conductivity. In addition, improved tech-niques for fracture containment (i. e., keeping the

fracture within the gas-bearing layer), fractureprediction, and onsite monitoring will result inmore accurate production estimates, greatly re-

ducing the risk of developing these high-cost re-sources.

Proppants.–Development of new proppants is

an important area of innovation by the servicecompanies in response to a need for material thatwill not crush at high pressures and is light

enough for the fluid to transport to the end ofthe fracture. Recent developments include ce-ramic beads and resin-coated sands that havelower densities than bauxite–the material cur-rently used in high pressure situations—and thuscan be more efficiently transported by the fluid.They appear to have sufficient strength for mostfracture applications.

Fracturing Fluids.–The need for a fracture fluid

that simultaneously can avoid formation damage

(clay swelling, etc.) and maintain a high capacity

to carry proppants in suspension into the frac-

ture has led to the development of very sophisti-cated fluids. A particularly significant develop-ment is “cross-l inking,’ which temporarilyincreases the viscosity of the fluid—and thus itsfracturing and proppant-carrying capability–by

linking together polymer chains. After emplace-ment of the proppant, this viscous fluid alters toa low-viscosity fluid so that it can flow back to

the wellbore, minimizing formation damage.

Fracture Containment.–A common problem inthin blanket sands is the difficulty of keeping thefracture within the blanket, or pay zone, so thatit does not waste its energy fracturing non-productive rock or actually cause reserves to belost by fracturing into a water zone or fracturingthe reservoir “cap.” In general, readily available

techniques have not been successful at contain-ing fractures within the pay zone. New tech-

niques have been proposed (including careful

placement of the fracture initiation point, verycareful control of the fracturing fluid viscosity andpressure, and use of floating proppants to sealoff upper non producing portions of the fracture)

and are being tested, as discussed in app. B.Fracture Prediction. -Understanding the mech-

anisms of fracturing and thus being able to pre-

dict fracture behavior is critical to reducing theeconomic risk of tight sands development to man-

ageable levels. The state of the art of fractureprediction, however, comes from sophisticatedmathematical models and laboratory experimentswith very little field verification.

Current practice in the field is to use relativelysimple analytical models against which to com-pare fracture behavior, proppant placement, trac-

ture length, and well performance. These simplemodels will often be inadequate in complex tight

gas situations, but the more sophisticated modelscurrently available are expensive, time-consum-ing, and dependent on input data that often arenot readily obtained. And, although laboratoryexperiments allow interesting possibilities fortightly controlled conditions and parametric anal-ysis, it is difficult to be confident that field-scalefractures will behave in the same manner as thesmall-scale laboratory fractures.

Field-scale testing, using experimental wells or

excavating a created fracture, can overcomesome of these problems but is extremely expen-sive. Only a few field-scale tests have been com-pleted and ongoing projects have been curtaileddue to Federal funding cuts. Nevertheless, they




are providing valuable information on actual frac-ture configuration.

The difficulties of fracture prediction are par-ticularly acute in the early stages of field devel-

opment, when theoretical models and analogywith other fields provide the only forecastingguides, Continued development of the field pro-

vides data for performance matching of prospec-tive sites with producing wells, and the percentageof successfuI well stimuIations should increasewith experience, As discussed below, however,this learning process may be disrupted by prob-lems associated with monitoring fracturing suc-cess and interpreting well performance.

Monitoring Fracture Behavior.–The ability to

monitor fracture behavior is critical to tight sands

development for two reasons. First, it is the meansby which the existing experience in fracturing can

be translated into an ability to predict fracturebehavior for new wells. The inability to measurewhat has actually occurred underground in pastfracture treatment is responsible for our poorpredictive capability. Second, it is critical to fielddevelopment, because knowing fracture location

is necessary in planning additional wells so as tominimize interference between wells. Most of the

technologies under development to monitor frac-tures in the field are adaptations of existing geo-physical and well logging techniques, They arediscussed in greater detail in appendix B. Somesuccess has been realized in determining fracture

height adjacent to the borehole, and in somecases, total fracture length and propagation direc-tion. Problems still exist in determining proppedlength of the fracture and vertical growth 22 at a

distance from the wellbore.

Most of these technologies are still experi-mental and costly, and are difficult to use. How-ever, fracture diagnostics has received major at-tention from service companies in the past fewyears, and innovation has been rapid. improve-ment and widespread commercial use of these

monitoring techniques would have a major ef-

fect in lowering the economic risk of tight sandsdevelopment.

~~vertical ~rO1~th is a n I m portant parameter because of the

cie~lrabl I Ity of keeping the fracture with In the pay lnter~ al. See Frac-

tur~ Ccmt<?/nmcmt abotfe.

Current Fracturing Success.—In tight gas forma-tions, improved fracturing technologies, devel-oped in blanket sands, have for the most partrealized considerable success. Increased fracturelengths in blanket formations often appear to beattainable simply by increasing the volume of

fluid and proppant pumped. Fractures over 2,000ft long have been reported. 23 Where designlengths have not been achieved, failure can gen-erally be attributed to fracturing out of the payinterval, inadequate proppant transport, or ex-tensive formation damage.

An additional but less obvious major problemmay still exist i n blanket formations. Becausemassive hydraulic fracturing in low-permeabilityformations is still a relatively new technology,

there are no data on whether the permeabilityof the fractures can be maintained through time.

There is concern that the fractures may close orbecome plugged before the 30-year productionhistories are complete. Counteractive measures,such as periodic cleanup treatments or multiplesmall fracture treatments over the life of the wells,have not been evaluated in terms of their costs,risks, and effect on well performance.

The overall success rate of massive hydraulicfracturing in tight formations is likely to improveto the extent that new technologies are devel-oped to counteract problems such as formationdamage and inadequate proppant transport. Acertain number of reservoirs, however, may

never be amenable to production using massivehydraulic fracturing. For example, adequate frac-ture containment in some reservoirs may neverbe possible due to the intrinsic characteristics ofthe rock. If reservoir boundary layers are substan-tially weaker than the reservoir rock, the fracturewill grow vertically at the expense of horizontalgrowth. In effect, much of the fluid and proppantis being used to create a fracture i n a nonproduc-tive interval. Recovery from such reservoirs can-not be based on the creation of 1,000-ft fractures.More work needs to be done in identifying suchformations and determining the optimal fracture

treatment to maximize recovery at minimum

cost .

Z3B . A , ) v la t t h e w s r W. K. MI I Ier, and B. W. Schlottman, ‘‘Record

Massive Hydraulic Fracturing Treatment Pumped In East Texas Cot-

to n Valley Sand5, ” [III <~nd G,IS }ourn,~l, Oct. 4, 1982, pp. 94-98.

3 8 - 7 4 2 0 - 8 5 - 1 1




Lenticular reservoirs represent a situationwhere new technologies have not been effective.The NPC-estimated base technology (1,000-ft

fractures) is not yet the acknowledged state of the

art, nor is there substantive evidence that frac-tures will penetrate lenses not actually intersected

by the well bore. Where large-scale fracture treat-ments have been attempted in Ienticular forma-tions, results have been disappointing. Commer-

cial flow rates were not attained. Insufficient datawere collected both before and after treatment

to evaluate why fracture treatments failed,24

Currently, most producers have returned to lessexpensive shorter fractures. In some cases theyare drilling wells in areas with thick vertical se-quences of Ienticular sands and stimulating multi-ple pay intervals by fracturing each interval withfractures on the order of 100 to 500 ft long,

Although per well stimulation costs are lowerin this case, cumulative production per well maybe less, since a smaller gas-bearing area is in con-

tact with the well bore. This technique may limitthe potential recovery from Ienticular formations,particularly in the short term. However, there are

some data indicating that, over the long term, gasflow from wells in Ienticular reservoirs, with rela-tively short fractures, may not exhibit the ex-pected production decline rates of a limited reser-

voir.25 Because the nature of natural fracture

systems and gas flow behavior in Ienticular for-mations is not well understood, it may be pre-

mature to draw conclusions about the long-termgas recovery of alternative fracturing strategies inthese formations.

Reservoir Characterization Technologies

Techniques and instrumentation to quantify

more accurately the physical properties of the

ZqAn alternative explanation of the production failure (instead of

failure of the fractures to reach design lengths) is that the fractures

may have achieved design lengths but that other problems, such

as water infiltration, and production problems caused by backflow

of the sand proppant, were primarily responsible for the disappoint-

ing gas flows. The originator of this explanation concurs, however,

that the “wells have not been tested and studied sufficiently to

decide if the fractures are successful or not.” Ovid Baker, Mobil

Research & Development Corp., personal communication, 1984.2 5 D . H . S t r i g h t , j r . , a n d J. 1. Gordon, “Decline Curve Analysis

in Fractured Low Permeability Gas Wells in the Piceance Basin, ”

SPE/LWESymposium on Low Permeability Gas Reservoirs, SPE/DOE

11640, 1983, pp. 351-356.

reservoirs and to help determine their relation-ship to one another are crucial to designing more

efficient stimulations and making accurate esti-mates of reserves and recovery rates for eco-nomic analysis, Accurate reservoir characteriza-tion reduces both the costs and risks of field

development.Conventional methods for determining reser-

voir parameters include: geologic mapping ofexposed areas of the reservoir formation, labora-

tory analysis of core samples and detailed sub-surface correlation of reservoirs, well logging, andwell tests. Considerable data have been collectedusing each of these techniques. Problems withthe techniques themselves have been identified.Only limited progress has been made to date inmodifying existing techniques and developingnew techniques to overcome their problems.

Laboratory analysis of core samples recoveredfrom a well provides the best estimate of reser-

voir properties at that particular site. The data canbe correlated with information from other wellsto derive a broader regional picture of the sub-surface conditions. Date from laboratory analy-ses can also aid in the interpretation of well logs.

Most samples for laboratory analysis are recov-ered using conventional coring techniques, Thedepth of recovery is known, and temperature andpressure information often is available. The re-

covered samples are subjected to a range of lab-

oratory tests to determine porosity, pore volume,permeability, water saturation, mechanical rockproperties, and mineralogy.26 Electrical and gamma-ray logs are also run on the samples for correla-

tion with well logs.

Because small changes in reservoir properties

can substantially alter estimates of the recover-able gas, the laboratory measurements need togive an accurate picture of actual reservoir con-ditions. Alteration of the physical properties ofthe rock during its collection represents an im-portant problem for tight formations, contribut-ing to inaccuracies in measurements of permea-

bility, water saturation and other parameters.27

26A. R. Sattler, “The Multiwell Experiment Core Program, ”

SPE/DOESymposium on Low Permeability Gas Reservoirs, SPEIDOE

11763, 1983, pp. 437-442.ZTNational petroleum Council, Unconventional/Gas sources: Exec-

utibre Summary, December 1980.



Ch. 8—Tight Gas 1 5 5

Techniques for unaltered core recovery exist

(e.g., the use of pressurized core barrels), but atpresent are too expensive and too unreliable forroutine use. The available techniques are mostpractical for research purposes or possibly for usein the early stages of developing a large field, to

assist in planning a development strategy,Well logs are among the more commonly used

techniques to determine reservoir properties.They are comparatively inexpensive to use, andresults generally can be obtained quickly. Unfor-tunately, many of the logging techniques usedin conventional reservoirs have given inaccurateresults in the unconventional formations,28 partlybecause some conventional methods of log anal-

ysis are not appropriate for low-permeability for-mations and partly because some analysts are notproperly accounting for the low-permeability

conditions .29

The complex mineralogic content, including

clays and cement, and poorly defined interfacesbetween producing and nonproducing zones intight formations distort log responses. Resultingporosity and water saturation measurements maybe too low or too high. Often the precision of

the tools is not sufficient for these reservoirs; e.g.,

a small error in porosity measurement can resultin a large error in inferred fluid saturation. Fail-

ure to distinguish gas-productive from water-productive zones, a common problem, 30 dras-tically complicates the selection of fracture

locations.

A prototype logging tool has recently been de-veloped using nuclear magnetic resonance (NMR)to determine porosity, fluid saturation, pore size

distribution, and bulk permeability of a forma-tion. Laboratory tests of the tool have been suc-cessful; however, the time required to obtain ac-curate measurements (on the order of severaldays) is likely to severely restrict its commercialapplication. 31 NMR logs may end up being most

z8G, c. Kukal, et al., “critical Problems Hindering Accurate Log

Interpretation of Tight Gas Sand Reservoirs, ” SPE/DOE Symposium

on Low Permeability Gas Reservoirs, SPE/DOE 11620, 1983.z9S. A. Ho]d; tch, Texas A&M University, PfYSOnd] COfYIm UniCd-

tion, 1984.JoNational petroleum Council, Ur?corrvent;of?a/ ~a5 SOUIC6’5: ~~-

ecutive Summary, December 1980.31 K. H. Frohne, Morgantown Energy Research Center, PerSOfIal

communication, 1984.

useful on experimental wells where they wouldbe used to gather baseline data for extrapolation.In the long term, it may be a less expensive tech-nology for obtaining reservoir data than coring.

Another approach to improved logging is to useavailable equipment with specialized interpretive

techniques to account for the log distortions asso-ciated with the tight formations. An example ofnew interpretive techniques is a system called

“TITEGAS,” which is based on equations which

define the response of conventional loggingtools.32 The system deals primarily with data fromdensity, neutron, and resistivity logs and makesa number of claims to accurately measure, amongother parameters, porosity, gas saturation, claycontent, and the presence of natural fractures incomplex, tight formations.

Improved interpretation of conventional logs

clearly is an important element of better reser-voir characterization for tight formations. Theneed for new interpretive techniques is not clear,however. OTA found some in the research com-munity who felt that, with available techniques,it was now possible to accurately measure reser-voir characteristics in most tight gas situations;

they blamed current problems on the failure ofmost practitioners to assimilate the latest ad-

vances. 33 There are many others who concludethat logging of tight formations is still error-proneand basically unreliable.

Well tests are also frequently used to em-pirically determine breakdown pressures (e.g.,the pressure required to initially fracture a for-mation) and gas flow rates. From these data, suchreservoir properties as in-situ stress and gas

permeability can be inferred. Creation of small-scale fractures is particularly useful in determin-ing the difference in the stress characteristics ofthe rock in the producing and non producing in-tervals, which in turn allows prediction of frac-ture containment.

3ZC, c , K u k a ] , “ A S y s t em a t i c Approach for the Effective Log AJM-

ysis of Tight Gas Sands, ” 1984 SPE/DOE/GRl Unconventional Gas

Recovery Symposium, SPE/DOE/GRl 12851, Pittsburgh, PA, May

13-15, 1984.JJS. A. Ho[ditch, Texas A&M University, persOnal communica-

tion, 1984.




Pressure-transient testing is the most common

technique used to get a rough estimate of gaspermeability prior to stimulation. These tests con-sist of producing or shutting in the well for a speci-fied period of time and measuring the pressure

drawdown or buildup. In low-permeability reser-voirs, however, this technique does not alwaysproduce useful results, especially during the short

time periods feasible for shutting in the well.

Exploration Technologies

No new technologies have been developed

specifically for exploration for unconventional re-sources, and they are not perceived to be a highpriority need because considerable resourceshave been identified by past drilling efforts. Someexisting state-of-the-art technologies used in con-ventional hydrocarbon exploration are applica-ble to the unconventional resources. These in-clude aerial and satellite imagery and geophysicalsurveying.

In tight sands, mapping of surface features from

aerial and satellite imagery has proved useful in

determining dominant structural trends.34 Suchdata may be useful in locating wells so that theymake the best use of existing fracture systems.

In tight sands, there may be some potential for

using seismic data to delineate the character ofbeds in a formation.35 36 If clusters of sand-richlenses and their orientation could be identified,

these data in conjunction with data on regionalstress patterns would determine the best locationfor a well. Three-dimensional, vertical, and cross

34J. A. Clark, “The Prediction of Hydraulic Fracture AzimuthThrough Geological Core and Analytical Stud ies, ” SPE/DOE Symp-

osium on Low Permeability Gas Reservoirs, SPEIDOE 11611 pp.

101-111.35T. L. DO&cki, “AppllcatiOrl of Areal Seismics to Mapping Sand-

stone Channels, ” Low Permeability Gas Reservoir Symposium,

SPE/DOE 9847, 1981, pp. 205-209.3 6 C A , S e a r I s , M. W. L e e , j . j , Miller, J. W. Al bright, J. Fried, and

J. K. Applegate, “A Coordinated Seismic Study of the Multiwell Ex-

periment Site,” SPE/DOE Symposium on Low Permeability Gas

Reservoirs, SPE/DOE 11613, pp. 115-117

borehole seismic surveys are being tested at theDOE multi-well site to delineate Ienticular bodiesand fracture zones in the Ienticular formation. 37

Initial results have been disappointing; the data

are not of sufficiently fine a scale to map the

spatial relationship of the lenses. Detailed geo-physical surveys are not likely to be used com-mercially in the near future due to high costs andcomplex time-consuming data reduction.

New Technologies and the Industry

It appears that both tight gas producers andwell service companies are willing to experiment

with new stimulation techniques. This may havecontributed to the rapid development of sophis-ticated fracturing fluids and proppants.

In contrast, state-of-the-art diagnostic tech-niques both for reservoir characterization and for

predicting and monitoring fracture behavior arenot commonly in use among producers. The pri-mary problem appears to be that these tech-

niques are costly and time-consuming. For smalland midsize producers in particular, with limitedacreage, there is little cost benefit to developinginformation leading to improved stimulation ifthey will only be drilling and fracturing a fewwells. Larger producers are more likely to makethis investment.38

The diagnostic techniques may ultimately bethe most important factor in making tight gas an

economically viable resource. However, becauseof the limited interest, the rate of developmentof these technologies is likely to be slow, particu-larly in the transition from proven concept tocommercial product. For these reasons, the in-dustry may continue to be dependent for sometime on external research efforts, such as thosesupported by the Department of Energy and the

Gas Research Institute, to develop the tight gasresource.

~71bid.

38J. W. Crafton, “Fracturing Technologies for Gas Recovery From

Tight Sands, ” contractor report to OTA, September 1983.



Ch. 8—Tight Gas q 1 5 7

THE RECOVERABLE RESOURCE BASE ANDPRODUCTION POTENTIAL

Estimates of economically recoverable gas fromtight gas formations are extremely sensitive to

assumptions made about price, level of technol-ogy, and the volume of gas-in-place. Existing esti-mates range from a conservative 30 TCF to almost600 TCF.

Some recoverable gas estimates include anassessment of the technically recoverable gas(also called maximum recoverable gas) in addi-tion to the economically recoverable gas. Tech-nically recoverable gas represents that gas recov-erable from tight formations under optimisticassumptions of technological development, as-suming price is relatively unconstrained. The eco-

nomically recoverable gas, in contrast, is subjectto both economic and technological constraints.It is usually calculated for several price and tech-nology levels and reflects the sensitivity of recov-erability to these parameters.

Detailed estimates of both technically and eco-nomically recoverable tight gas were made byLewin Associates and by the National PetroleumCouncil in conjunction with their gas-in-place re-source estimates. The Gas Research Institute (G RI)has also estimated the economically recoverable

resource in tight formations. The NPC study is themost comprehensive of these analyses and is dis-

cussed here in the most detail. The other studiesare included for comparison, particuIarly wheredifferent methodologies or assumptions are shownto substantially affect estimates. All these esti-mates assume two levels of technology. The

“base case” uses technologies currently available

or thought to be available in the near future. The

“advanced case” represents technologies thatmight be developed given a concerted researcheffort.

Methodologies and Results

Lewin EstimateThe Lewin report calculated the technically

recoverable gas using its “advanced case” tech-nology criteria, which assume a maximum well

spacing of six wells per section (107-acre spac-ing) and a maximum fracture length of 1,500 ft.They determined that 211 TCF, close to half ofthe total gas-in-place resource, was technicallyrecoverable. The percent of gas-in-place esti-

mated to be recoverable for the individual basinsranges from 10 to 80 percent, as shown in table 38.

Table 38.—Maximum Recoverable Tight Gas Resources, TCF

Lewin NPC

Maximum Percent Maximum PercentAppraised basins GIP

arecovery recovery GIP

a recovery recovery

Northern Great Plains/Williston . . . . . . . . . . . . 74 35 47% 148 100 68 ”/0Greater Green River . . . . . . . . . . . . . . . . . . . . . . 91 36 40 136 87 61Uinta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 18 36 20 15 75Piceance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 12 33 49 33 67Wind River . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 33 34 23 68Big Horn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 8 33Douglas Creek . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0.3 10Denver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 13 68 13 8 62San Juan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 12 80 3 2 67Ozona. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 0.6 60Sonora . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 16 67 4 2 50Edwards Lime. . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 9 64

Cotton Valley “sweet” . . . . . . . . . . . . . . . . . . . . 67 50 75 22 13 59Cotton Valley “sour”. . . . . . . . . . . . . . . . . . . , . . 14 10 71

Ouachita . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1 20

Total . . . . . . . . . . . . . . . . . . . . ., . . . . . . . . . . . . 423 212 500/0 444 292 66%Extrapolated basins. . . . . . . . . . . . . . . . . . . . . . . 480 315 66aGIP = Gas-in-place.




Lewin determined that 70 to 108 TCF could be could be produced under advanced case tech-economically produced under base case technol- nology conditions and the same price range, as

ogy conditions for well head prices ranging from shown in table 39. The different assumptions used$1.75 to $4.50/MCF (1977$) and $2.75 to $7/MCF for the base and advanced case estimates are(1983$), They determined that 150 to 188 TCF given in table 40.

Table 39.- Economically Recoverable Gas at Two Technology Levels (TCF)

Price per MCF

Study date 1983 Base Advanceddollars dollars technology technology

Lewin (1977) . . . . . 1.753.004.50

GRI (1979). . . . . . . 3.124.506.00

NPC (1979) . . . . . . 2.505.009.00

(2.75)(4.70)(7.00)(4.20)(6.00)(8.00)

(3.35)(6.70)

(12.00)

70100108304560

Total Appraised

192 97365 165404 189

149182188100120150

Total Appraised

331 142503 231575 271

Table 40.—Lewin & Associates Base v. Advanced Technology

Parameter Base case Advanced case

Fracture height , . . . . . . . . . . . .

Fracture length (one way)Shallow gas sands . . . . . . . .Near-tight sands . . . . . . . . . .Tight gas sands. . . . . . . . . . .

Fracture conductivity . . . . . . . .

Field developmentLenticular . . . . . . . . . . . . . . . .

Blanket . . . . . . . . . . . . . . . . . .

Net pay contactedLenticular sands

320 acres drainage . . . . . .107 acres drainage . . . . . .

Blanket sands . . . . . . . . . . . .

Dry hole rate:Lenticular ., . . . . . . . . . . . . . .Blanket . . . . . . . . . . . . . . . . . .

Discount rate. . . . . . . . . . . . . . .

4 times net pay(200’ minimum, 600

maximum fractureheight)

200 ‘500’

1,000 ‘

Decreases with

using currentand methods

depth

proppants

320 acres per well(2 wells per section)


17 % —

100%

300/0 20% 26°/0 (20°/0 real)

3 times net pay(150’ minimum, 400

maximum fractureheight)

500 ‘500 ‘

1 ,500‘

(Using improved proppants

and methods to maintainadequate conductivity)



—

800/0100%

200/010%16°/0 (10°/0 real )

SOURCE: V. A. Kuuskraa, et al., Enhanced Recovery of Unconventional Gas: The Program- Volume II, US, Department of Energyreport HCP/T2705-02, October 1978.




The Lewin estimates were developed usingreservoir simulation-based production curves forblanket and Ienticular reservoirs to determine

cumulative reserves from a well over a 30-yearperiod. Input parameters included reservoir char-acteristics for the individual basins and pay quality

intervals, fracture lengths, fracture conductivity,and drainage area of a well.

Production per year per well for the assumedwells, multiplied by the assumed gas price, givesthe positive cash flow for the economic model.

Offsetting costs include the initial investment inexploration, drilling, and stimulation, plus oper-ating and maintenance costs, royalties, and taxes.A required rate of return of 20 percent (real) was

used for the base case estimates, dropping to 10percent in the advanced case to reflect lower risks

associated with the greater predictive capabilities

and higher efficiencies associated with advancedtechnology,

National Petroleum Council

In contrast to the Lewin method of calculatingthe technically recoverable gas using its advancedtechnology criteria, the NPC determined the max-imum recoverable gas directly from the gas-in-place estimates. The maximum recoverable gasis defined as the total amount of gas that can beproduced from a reservoir before the gas pres-sure in the reservoir reaches the “well bore draw-down pressure.”39 This quantity is further modi-fied by a “recovery adjustment factor” meant totake into account the existence of very low per-meability areas within a field that are unlikely tobe productive, Recovery adjustment factors rangefrom 50 percent for blanket reservoirs with aver-

age permeabilities of 0.00001 md, to 95 percentfor blanket reservoirs with average permeabilities

of 0.3 md.40 Lenticular reservoirs use more con-servative adjustment factors. The NPC study de-

~qThe production of ga s depends on a pressure gradient eXiSting

between the wellbore and the rest of the reservoir. As the pres-

sure in a formation IS reduced through production, the reservoir

pressure and the wellbore pressure approach equilibrium. Even-tually no further gas WIII flow naturally, The INPC assumed that the

reservoir pressure .at this point would be approximately 10 percent

of the original formation pressure. Any gas remaining in the reser-

voir cannot be produced except by expending energy to induce

an artificial pressure gradient.40NPC Report, vol. V, part 1, table 15.

termined the technically recoverable gas for a setof appraised and extrapolated basins to be 292and 315 TCF respectively. Its results are also

shown in table 38.

The NPC estimates of economically recover-

able gas range from 192 TCF at $2.50/MCF (1979$)with base case technology to 575 TCF at $9/MCFwith advanced technology, as shown in table 39.The base case is represented as “current fractur-ing technology and well spacing regulations, ” butapparently allows for some evolution and im-

provement of the technologies over time as op-erators gain experience. The advanced case in-corporates new technologies developed by aconcerted R&D program. The methodology usedby the NPC to determine the economically re-coverable gas is similar to the Lewin methodol-ogy. Cumulative production per well was deter-

mined from type curves41

for base and advancedtechnologies. This gives the number of wells per

section required to produce all of the recoverable

gas in a given permeability level in a formation.The base and advanced technology criteria areshown in table 41. In the economic model, costsof drilling and fracturing and operating costs aresubtracted from income from production at vari-ous prices, and a marginal rate of return is cal-culated. All formations with a rate of return lessthan a prescribed value (10, 15, or 20 percent)are considered unprofitable.

It has been suggested that the NPC method ofusing type curves designed for 640-acre (one wellper section) spacing to determine cumulative pro-duction levels from blanket formations at 160-acre (four wells per section) and smaller spacingscould result in a substantial overestimate of therecoverable resource. The expected cause of thisoverestimate is the inability of the (640-acre) type

curve to account for the greater interference be-tween wells that would occur at 160-acre spac-

ing (interference reduces the production per

—-. . . . .4’Type curves are normalized representations of cumulative pro-

duction over time generated from reservoir simulation models. Be-

cause the data are generated in terms of dimensionless values, theyare applicable over a wide range in the reservoir parameters

(permeabilities, poroslties, net pay, pressure, temperature, gas com-

position, and fracture conductivity). Different type curves, how-

ever, are required for different well spacing rules. They are an effi-

cient technique for obtaining estimates of cumulative recovery and

producing rates for a large number of reservoirs.




Table 41 .-NPC Base v. Advanced Technology

Blanket formations Lenticular formations

1. Fracture height . . . . . . . . . . . 4 times net pay 3 times net pay 6 times net pay 4 times net payRange (ft) . . . . . . . . . . . . . . . . 200-600 150-400 300-1,000 200-600

2. Fracture conductivity(md-ft) . . . . . . . . . . . . . . . . . . . 500 1,000 500 1,000

3. Fracture length, wellbore totip (ft), permeability>=0.1 mdEffectively achieved . . . . . . . 1 ,000‘ 2,000 ‘ 1 ,000‘ 4,000 ‘Hydraulic design required . . 1 ,700‘ 2,500 ‘ 1 ,700‘ 5,000 ‘

4. Field development,wells per section(maximum) . . . . . . . . . . . . . . . 4 12 4 12Acres per well (minimum) . . 160 53 160 53

5. Lenses remote from thewellbore may be contactedby fractures? . . . . . . . . . . . . . NA NA Yes Yes

aProduct of fracture permeability and fracture width.

SOURCE: National Petroleum Council, Unconventional Gas Sources, vol. V, part 1, December 1980.

well).42 The overestimate of production per well

would be greatest at the higher permeabilitylevels where more of the economically produc-ible gas is found. This problem is not as seriousas it might appear, however, because of the waythe type curve data were used in the NPC report.If the NPC had computed total production from

wells drilled to determine gas recovery, assum-ing a standard “four wells per productive sec-

tion, ” then they would have overestimated themaximum recoverable resource. Instead, as notedpreviously, NPC determined the maximum recov-

erable gas independently of the type curves, ap-plying recovery factors to the gas-in-place. The

cumulative 30-year production of a well, derivedfrom the type curves, was used only to determine

the number of wells per section required to pro-

duce the maximum recoverable gas.

At the higher permeability levels, all of the

recoverable gas can be produced by less than thefour wells per section base case constraint. Thus,although the NPC may have urderestimated the

number of wells required to produce thesepermeability levels, unless the actual number ofwells required exceeds the four wells per section

constra int , it has not overestimated the amount

of gas that can be produced.

qlGruy petroleum Technology Inc. report to EIA, “Correlations

for Projecting Production From Tight Gas Reservoirs. ”

At lower permeabilities (0.3 md and lower),

where more than four wells per section are re-quired to produce all the gas in the formation,the NPC technique may overestimate the eco-nomically recoverable gas for the base case.Nevertheless, the amount of gas involved is notlarge. For example, a sensitivity analysis of the

NPC estimates shows that a shift to 160-acre type

curves would not have changed the estimate ofrecoverable reserves for the blanket formationsby more than 10 percent.43 A similar problemdoes not exist for the advanced case because allof the gas at all permeability levels can be pro-duced from less than the allowed 12 wells per

section.

Gas Research Institute

The most recent estimate of the gas recoverable

at different price and technology levels has beenmade by the Gas Research Institute and is alsoshown in table 39. These estimates are consider-ably lower than both the Lewin and the NPC esti-mates, ranging from 30 TCF at $3/MCF (1 979$)and base technology, to 150 TCF at $6/MCF andadvanced technology. GRI derived its estimateby making judgmental adjustments to existing

estimates. It concentrated on those basins where

43J P. Bra;ear, et al., “Tight Gas Resource and Technology Ap-

praisal: Sensitivity Analysis of the National Petroleum Council Esti-

m a t e s , 1984 SPE/DOE/GRl Unconventional Gas Recovery Sym-

posium, SPE/DOE/GRl 12862, Pittsburgh, PA, May 13-15, 1984.



Ch. 8—Tight Gas . 161

the tight sands resource was best understood and,in addition, attempted to eliminate tight gas that

might conceivably be considered conventional.Its conservative approach was specifically de-signed to provide an estimate of that portion ofthe tight gas resource which had a high probabil-

ity of occurrence.

Comparison of Estimates andDiscussion of Uncertainties

One of the major differences between therecoverable resource estimates is the differencein the total areas assessed, since the amount ofgas estimated to be recoverable is intimately tiedto the proportion of the gas-in-place actually in-cluded in the study and evaluated for recoverabil-ity. The NPC, Lewin, and GRI recoverable re-

source estimates each are based, implicitly orexplicitly, on a different gas-in-place resource

base. The NPC, by including both the appraisedand extrapolated basins, begins with the largestgas-in-place resource base. Lewin restricts its esti-mate to those basins for which considerable in-formation is available. GRI also restricts its esti-mate on this basis, but apparently further limitsthe resource base it evaluates by excluding all po-tential areas of overlap with conventional re-source estimates. Where the areas assessed byNPC, Lewin, and GRI overlap, however, thereare still substantial differences in the estimates ofrecoverable gas. These differences result from theunderlying assumptions used in the estimatingprocess.

In the sections that follow, we first compare

recovery estimates for a comparable area—theappraised resource—to show how different as-sumptions affect the estimates of technical y andeconomically recoverable gas. Next, we brieflydiscuss a portion of the resource base evaluatedby broad geologic analogy rather than directappraisal—the extrapolated resource. Finally, wediscuss briefly the geological/geographical bound-aries used in the resource estimates. Included in

each section is a discussion of the major uncer-tainties underlying the estimates and how theseuncertainties affect the amount of gas recover-

able. it is important to remember that many of

the uncertainties in calculating the gas-in-place

are propagated through to the calculations of thetechnically and economically recoverable re-source.

The Appraised Basins

The appraised basins of the Lewin and NPCstudies have comparable total volumes of gas-in-place, as discussed in the previous chapter. De-spite this agreement, their estimates for poten-

tial recovery of tight gas are significantly different(the technically recoverable gas is two-thirds ofthe gas-in-place in the NPC study compared tohalf of the gas-in-place in the Lewin study). Dif-ferences in percent of gas-in-place recoverableare particularly apparent on an individual basinlevel, as seen in table 38, and can be attributed

to the different assumptions used by the esti-mators to calculate both technically and eco-

nomically recoverable gas.

The NPC approach to estimating maximum re-coverable gas is independent of any technologyassumptions and effectively assumes that virtuallyall gas in a formation will ultimately be in con-

tact with a well bore. In contrast, the Lewin esti-mate is constrained by a set of advanced tech-nology criteria that do not allow the entiregas-bearing portion of a formation to be con-tacted, even under the most favorable conditions.

The Lewin estimate is more conservative than theNPC estimate in its assumed fracture lengths and,for Ienticular resources, in its assumptions aboutthe probability of producing from lenses not indirect contact with the wellbore.

Differences between the Lewin and NPC esti-mates of economically recoverable gas in the ap-praised basins are caused primarily by differencesin their assumptions about base and advanced

case technology conditions (tables 40 and 41).For the base technology case, the Lewin estimatesof economically recoverable gas are only two-thirds of the NPC estimates (see fig. 31). The ef-fects of shorter assumed fracture lengths in someareas, especially the Northern Great Plains, a con-

straint of two wells per section in Ienticular for-mations, and the constraint that only lenses indirect contact with the wellbore could be pro-

duced probably account for the lower estimatesof recoverable gas in the Lewin estimate.




100

0

Figure 31.— Comparison of Recoverable Estimates (for gas price in 1983$)

Comparison: Lewin Assoc.NPCGRI

NPC (total) advanced technology

NPC (total) base technology

rNPC (12 basins) advanced technology .

~-- - - - - - - - - - - -- -

/4 Lewin Advanced TechnologyF

NPC (12 basins) basetechn~y+ — . _

—* — “

“ - * ~ “ — “ — “- -GRI a d v a n c e d

--”” ‘--~”=” — — Lewin base technologyI

._ GR I b a s e. — . — “ —“—

I I I 1 I I0 1.00

Under advanced

2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.(

Wellhead price, $/MCF

technology criteria, there islittle difference between the NPC and Lewin esti-mates at the lower prices despite the NPC tech-nical potential for much higher levels of produc-tion using 2,000- to 4,000-ft fractures and up to12 wells per section. The increased “per section”production potential in the NPC analysis is prob-

ably offset by increased costs of longer fractures,which would make many of the lower permeabil-

ity prospects unprofitable at lower prices. The

lower marginal rate of return in the Lewin ad-vanced case allows it to bring a number of its pre-viously unprofitable wells on line, increasing thepercent of the technically recoverable gas thatcan be economically produced. In this case,tradeoffs between costs, rates of return, andrecovery per well allow two entirely different setsof assumptions to result in approximately the

same amount of gas produced. At higher prices,however, the higher fracturing costs of the NPC

estimate become less of a factor in determiningprofitability of a well and the NPC estimate ofrecoverable gas is again higher than the Lewin

estimate.

The above comparison of the Lewin and NPC

appraised basin estimates points out how differentassumptions of price, rate of return, and state oftech no logic development affect the estimates oftechnically and economically recoverable gas.The inherent uncertainty in some of these as-sumptions is the major factor in whether the esti-

mates can be considered an accurate represen-tation of the recoverable resource.

The primary uncertainty lies in the technology

assumptions for the different tight gas formations.The NPC assumptions are considerably more op-

timistic than those used by Lewin. Assumptionsmade by the NPC that are most likely to bechallenged include:

1.

2.

fractures in Ienticular formations will contact

(and allow production from) lenses distantfrom the wellbore;

currently available technology will allowpropped 1,000-ft fractures in all producing

situations, advanced technology will allow

4,000-ft fractures with fracture heights less




3.

than currently achieved with 1,000-ft frac-tures; and

every stimulation will be successful inachieving its design criteria (length, orienta-tion, and direction).

OTA considers these assumptions to be optimis-tic, especially for the base case, but also for theadvanced case as well. As noted in the technol-ogy discussion, not all of the base case conditionshave been met at the present time. Although

evolution and improvement of existing technol-ogy during normal operations44 will occur, the

base case criteria still appear optimistic. Some ofthe base case and advanced case criteria do ap-pear likely to be met over the long term undera strong R&D program; others may deserve to besubstantially modified.

1. Contacting Remote Lenses. A major assump-

tion of the NPC study is that massive hydraulicfractures in Ienticular formations can contact

lenses distant from the wellbore, in both the baseand advanced technology cases. This assumptionunderlies the inclusion of vast areas of Ienticularformations into the tight gas recoverable resourcebase. The NPC estimates that 126 TCF is recover-able from Ienticular formations in 4 of the 12 ap-praised basins—43 percent of the total recover-

able resource in all appraised basins. To date, noevidence exists that remote lenses can be con-tacted, despite a number of fractures completedin Ienticular formations. In recent analyses of tightgas resources, GRI has assumed zero contact ofremote lenses for its “present technology” case.45

The DOE Multiwell Experiment is designed toanswer this remote lens question; if the resultsof this experiment prove disappointing, estimatesof the amount of gas that can be recovered fromtight Ienticular formations may need to be sub-

stantially reduced.

A recent sensitivity analysis of the ability to con-

tact remote lenses has been performed using acomputer simulation of the NPC study. 46 Thisanalysis determined that the technically recov-

dqAs noted previously, such evolution and improvement is con-

sidered to be included in the base case technology conditions.45Briefingdocument fo r the Tight Gas Analysis System, Lewin &

Associates, Inc., 1984.qbThe Slmu[ation is called the TGAS Simulator, described in

Brashear, et al., op. cit.

erable gas from appraised Ienticular formationswould be reduced by about half if remote lenses

could not be contacted. For example, for anallowable 12 wells per section, recovery per wellwould drop from 100 to 48 BCF. This effect is

magnified for the economically recoverable gas,

especially at moderate prices, because some len-ticular formations cannot be developed at allunless remote lenses can be produced. For ex-ample, at gas prices of $2.50/MCF in 1979$ ($3.35in 1983$) and a rate of return of 15 percent,assuming advanced case technology, the NPC-estimated economically recoverable gas from theappraised Ienticular basins is 37 TCF if remotelenses can be produced and only 10 TCF if they

cannot be. Without production from remotelenses, the Ienticular tight formations may notplay a significant role in future U.S. gas pro-duction.

2. Achievable fracture characteristics. As notedearlier, there is considerable disagreement about

the extent to which the NPC-defined fracture

characteristics realistically reflect both currenttechnology and achievable future technology.

For the NPC “base” or current technology,OTA is particularly skeptical of the assumptionof 1,000-ft achievable fractures for the Ienticularresource (see the Technology section). RecentGas Research Institute analyses have used proppedfracture lengths of 600 ft and fracture conduc-tivities47 of 400 md-ft (v. NPC’s 500 md-ft) to rep-

resent “present” technology.48 49 As with changesin assumptions about contacting remote lenses,the effect of these changes in fracture character-istics is to substantially reduce the recoverablegas at all price levels. Figure 32 shows the com-bined effect on recoverable resources of the “nocontact of remote lenses” assumption and thechanges in fracture characteristics. For example,recoverable gas at $9/MCF (1982$) is reduced by

about 25 percent, from roughly 400 to 300 TCF.

qTFracture cond uc ti vit y : the product of fracture permeability and

width, a measure of how well gas can flow along the fracture.q86riefing document for the Tight Gas Analysis System, oP . cit.

49 GR115 “present tectlno[ogy” and NPC’S base case may not de-

scribe precisely the same technology level, however, because NPC’S

base case includes the evolutionary effects of future operations.

On the other hand, the GR I criteria were defined several years after

the NPC’S were.




Figure 32.–Comparison of NPC Study Results WithGRI Present Case—Tight Gas Sands Recoverable Gas

10.00 r . . . ,

8.00

6.00

5.00

4.00

3.00

. . . .1.00

0 ‘\ 1 I +

o 100 200 300 400

Economically recoverable gas (TCF)

GRI has also chosen to use more modest frac-ture characteristics than NPC for its “R&D Suc-cess” case, which includes improved geologicaland geophysical techniques, improved fracturedesign and diagnostics, and real-time fracturediagnosis and control:

50fracture length and con-

ductivity are assumed to be 2,000 ft and 800 md-ft, respectively, v. NPC’s 4,000 ft and 1,000 md-

ft. Although OTA agrees that these NPC “ad-vanced case” fracture characteristics appear tobe quite optimistic, our major concern lies withthe NPC assumptions for fracture height. GRI has

accepted the NPC’s fracture height criteria forboth the current and advanced cases.51 Accord-ing to these criteria, the 4,000-ft fractures will

have shorter fracture heights than the 1,000-ftfractures (e.g., for blanket sands, fracture height

is six times net pay for the 1,000-ft fracture andfour times net pay for the 4,000-ft fracture).

Because of the shorter fracture height, thelonger fractures will be cheaper per unit length

than the base case fractures, which in turn addssubstantially to the estimated recoverable re-source in the “advanced technology” case. The

fracture height assumption reflects NPC’s belief

that careful control of fracture placement and

—.—50Th~ latter improvement allows problems encountered in frac-

turing to be diagnosed and fixed in “real time, ” i.e., as the frac-

ture is being made.51 Briefing document, Op. cit.

fracturing fluid pressure will serve to maintainsmall fracture heights. This is the opposite of cur-rent experience, where longer fractures gener-ally are associated with increased fracture heightand higher (per unit length) costs. On the otherhand, current fracturing technology does not in-

clude reservoir and fracture-propagation model-ing capabilities that provide “real time” (during

the actual fracturing process) control of well treat-merits.52 GRI is sponsoring research to develop

this capability and clearly hopes that successfuldevelopment will allow longer fractures to be pro-duced without increasing fracture height. Futureimprovement in fracture control clearly is a cer-tainty. Nevertheless, assuming a reduction in frac-

ture height in conjunction with a fourfold in-crease in fracture length appears to be optimistic.

In addition, the NPC’s specification of fracture

height as a constant multiple of net pay thicknessserves to automatically favor thin, higher perme-ability intervals over thicker, low-permeability in-tervals, This is because fracturing costs area func-tion of the volume of fracture fluid required,which in turn is a function of fracture height. Con-

sequently, using NPC’s assumptions, thin, rich,

high-permeability pay intervals will tend to ap-pear very attractive to develop even though in

actual experience it may be difficult to achievevery long fractures in such intervals without ex-

ceeding the “four or six times net pay” fractureheight constraint. As a result, the NPC’s fracture

height assumptions may yield estimates of moregas at lower prices than maybe realistically pro-ducible. However, this effect may be counter bal-anced somewhat by the empirical relationship

between net pay thickness and permeability as-sumed by NPC. Because the NPC believed thatmost of the tight gas resource lies in the lowerpermeability rocks, they assumed that the higherpermeability gas-bearing zones would be thin andthe lower permeability zones would be thick. Be-cause the initial productivity, and thus the eco-nomic viability, of a well depends on the prod-

uct of thickness and permeability, the assumption

that higher permeability zones would essentiallyalways be thin is pessimistic. An alternative

sZGas Research Institute, Status Report: GRI’s (Jnconventiona/

Natural Gas Subprogram, December 1983,




assumption, albeit an optimistic one, that all payzones are the same thickness would yield a 46-

percent increase in the recoverable resource at$2.50 gas in 1979$ ($3.35 in 1983$) with the base

technology. The relative effect decreases at highergas prices and improved recovery technology.53

Finally, whether or not the NPC’s fracturing cri-

teria are generally realistic, OTA is concerned thatthey were universally applied to all formationsregardless of geologic character, except for theseparation of blanket and Ienticular formations.For example, using these fracturing criteria, theNorthern Great Plains becomes one of the dom-inant sources of recoverable gas from blanket for-

mations. Apart from the question of how muchgas exists in the Northern Great Plains, geologiccharacteristics of these shallow, low-pressurereservoirs can also restrict the amount of recov-

erable gas. For example, the presence of water-sensitive clays makes these reservoirs sensitive toformation damage.

54 Also, experience has shownthat fractures in shallow formations may propa-

gate horizontally rather than vertically, reducingthe net pay drained by each well, and significantlyreducing the recovery efficiency, Thus, the appro-priateness of assuming 4,000- or even 1,000-ftfractures in assessing recovery from these reser-voirs, and thus counting on up to 70 to 90 per-cent recovery of the gas-in-place, is questionable.The Lewin study assumed that short fractures (200to 500 ft) would be used in these reservoirs, and

assumed that fractures in the shallower forma-tions would propagate horizontally.55 Its ultimaterecovery efficiency in the Northern Great Plainsis only 47 percent of the gas-in-place (table 38).More accurate assessment of the recoverabilityof gas from these formations, using more site-specific technologies, is needed. For example, theuse of air-drilled open hole wells at close spac-ings may be a preferred method of productionconsidering the low drilling and completion costsfor these depths and the avoidance of formationdamage. 56 Ultimate recovery from this region ap-

pears more likely to be on the order of so per-

‘~J. P. Brashear, et al., op. cit.54 R. L. Gautier and D. D. Rice, “Conventional and Low

Permeability Reservoirs of Shallow Gas in the Northern Great

Plains, ” journal of Petroleum Technology, july 1982.ssLewin, vol. 11, pp. 3-57.

sGGaut;er and Rice, Op. cit.

cent (or less) of the gas-in-place rather than theNPC’s estimated 70 to 90 percent.

57

3. Success Rate of Fracture Treatments. Further,the NPC study assumes that each fracture treat-ment is successful in achieving design criteria.

Fractures are assumed to never grow verticallyout of the pay zone at the expense of fracturelength. They are never offset against existing nat-

ural fractures or intersect one another at closewell spacings. Formation damage is minimal. The

NPC approach assumes that technologic devel-opment will be able to overcome all of theseproblems. In fact, however, some of these prob-lems appear to be inherent to the formations.With better reservoir characterization, producers

may be able to better predict where fractures willgrow out of the pay zone or what the probabil-ity is that they may intersect, but they may never

be able to control many of these occurrences;other occurrences may be controllable but onlyat added cost. OTA feels that 100 percent frac-ture success is an overly optimistic assumption.

[n addition to uncertainties associated with thefracturing technology criteria, an important un-

certainty in the appraised recoverable resourceis associated with the geology of the NorthernGreat Plains:

The Northern ‘Great Plains

If gas does not exist in large quantities in the

Northern Great Plains, the estimated recoverableresource will have to be substantially reduced

regardless of the success of advanced technol-ogies. The NPC estimates that 100 TCF, one-thirdof the technically recoverable gas in the appraisedbasin estimate, can be produced from this region.

According to the NPC criteria, much of this gasought to be relatively inexpensive to produce.Approximately 54 TCF is estimated to be recov-erable at $2.50/MCF (1 979$) using base technol-ogies. Even the more conservative Lewin esti-mates predict that 21 TCF will be recoverable at

$1.75 (1977$) under base case conditions. De-spite this apparent incentive, however, there ap-

pears to be little current interest in developing

SzDudley Rice, U.S. Geological Survey, Denver, CO, personal

communication.




tight gas in the Northern Great Plains. No FERCfilings for tight gas designations in the Northern

Great Plains have been made to date. Reasonsfor the lack of interest may include external con-

straints, such as an absence of pipelines and mar-ket problems due to the existing surplus. How-

ever, there are other tight gas basins with similartransportation/market problems—e.g., the Uintaand Piceance basins—that have sustained at leastmoderate development efforts, so these problems

do not fully explain the lack of activity in theNorthern Great Plains.

On the other hand, a recent GRI analysis 58

claims that the current lack of activity and highestimated recoverable resource can be reconciledby comparing the profitability of individual pros-pects in the Northern Great Plains with prospectsin other tight gas basins. The Northern Great

Plains is described as “having a very geologicallydiffuse resource, which would require large num-

bers of shallow, low-pressure wells spread overgreat distances.”59 While the Northern GreatPlains contains more total economically recov-

erable gas than competing basins at any pricelevel, several basins in the Rockies and Southwestwere found to contain a number of individual for-mations offering greater profitability per thousandcubic feet of gas extracted. These prospects arecharacterized as offering larger field sizes and

greater well productivity and recoverable gas per

section than prospects in the Northern Great

Plains. Because developers make investmentdecisions on the basis of individual prospects andnot entire basins, development of the NorthernGreat Plains might not be expected to occur un-til the more attractive prospects elsewhere wereexploited. GRI also speculated that the inter-mingling of tight formations with established gas-

fields in the Rockies and Texas would also pro-vide an important incentive for developing these

basins first, because developers tend to want todrill in areas where they have had previous suc-cessful experience.

The questions concerning gas-in-place, produc-

ibility, and current activity levels call for an in-

58J. 1. Rosenberg, “The Economics of Tight Sands Gas Extraction

as Affected by R& D,” Gas Research Insights series, Gas Research

Institute, August 1983.

Jglbid.

dependent assessment of the Northern GreatPlains production potential. Such an assessment

should include a reevaluation of the gas-in-place,the engineering characteristics of the reservoirs,and the costs and technologic requirements ofproducing from these reservoirs. An analysis of

the external constraints and their temporal sig-nificance is also needed.

The Extrapolated Resource

The NPC approximately doubles the volumeof its gas-in-place estimate by including 101 ad-ditional basins in the resource base. Because itassumes that the formations in the extrapolatedbasins have similar producing characteristics toanalog formations in the appraised basins, it alsoapproximately doubles the estimated technicallyrecoverable resource as well as the economically

recoverable resource.It is generally agreed that the extrapolated

basins represent the most speculative part of theNPC resource estimate. In addition to the uncer-tainty as to the quantity of gas existing in thesebasins, there are further uncertainties in the ex-trapolation of engineering characteristics for gas-producing formations and the consequent esti-mates of production. There is sufficient informa-

tion in these areas to say that some gas is there,and that some of it may be economically pro-duceable, so the extrapolated resource certainlycannot be entirely disregarded. If FERC filings can

be considered indicative of productive areas (ifnot of actual volumes of gas), then the large num-

ber of filings for the Eastern and SouthwesternUnited States may bean encouraging sign for theproduction potential from extrapolated basins.

Recent studies have been undertaken to bet-ter characterize the extrapolated basins, 60 but

data sufficient to revise the NPC estimates are notyet available. Because the potential contributionof the extrapolated basins to the recoverable re-source has been estimated to be so large, moredetailed assessments remain a high-priority area

for future research.

mR. j. Finley and P.A. O’Shea, “Geologic and Engineering Anal-

ysis of Blanket-Geometry Tight Gas Sandstones, ” 198.3 S/?E/DOE

joint Symposium on Low Permeability Gas Reservoirs, SPEIDOE

11607.



Ch. 8—Tight Gas . 167

Many who use the NPC figures in their analy-

sis of gas supply consider only the recoverablegas in the appraised basins because of thespeculative nature of the extrapolated resource.OTA suggests that it is overly conservative todisregard the extrapolated resource altogether,

especially since FERC filings indicate that devel-opment is going forward in some of the extrapo-lated basins; however, until better estimates are

available, it may be more useful to consider theextrapolated recoverable resource estimates ina category separate from the more precise esti-mates of the appraised basins.

Effect of Boundary Conditions

Double Counting. –The separate estimation ofthe conventional and unconventional natural gasresource base by different groups, coupled with

poorly stated “boundary conditions” for eachestimate, can lead to some resources being

counted in both conventional and unconven-tional estimates (e.g., “double counting”). Thiswould lead to an overestimate of the total natu-ral gas resource.

Tight gas has the highest level of current pro-duction of all the unconventional resources. Be-

cause of the existing production, there is somecontroversy over how much of the tight gas re-source has already been included in conventionalresource estimates, and thus in estimates of future

conventional gas production. To the extent thatit is so included, it cannot be considered as a po-tential supplemental source of supply.

Most conventional resource assessments arenot well documented; and comparison of con-

ventional with unconventional estimates is notstraightforward. Recently, the Potential Gas Com-mittee (PGC) has attempted to determine whatpercent of its total conventional undiscoveredrecoverable resource 61 occurs in tight formations.It concludes that 172 TCF, or 20 percent of itstotal potential recoverable resource, is tight gas.62

61 I ncludi ng re50urCes made avai [able by the growth of already-discovered fields because of enhanced recovery, discovery of new

reservoirs, etc.GZThe pGC tight ga s estimate included both tight sands and DevO-

nian shales. However, conversations with the people responsible

for determining tight gas in the eastern region have indicated that

the Devonian shale contribution to the tight gas estimate is very

sma[l.

A considerable portion of this gas does not over-lap with the NPC tight gas resource, however,because it is below 15,000 ft (the NPC depthlimit), in areas already undergoing development(and thus not unconventional according to theNPC definition), or simply in locations not con-

sidered by the NPC study to be prospective.

There is no unequivocal data set that will allow

an accurate estimate of the overlap between con-ventional and unconventional natural gas re-sources. Estimates range up to 100 TCF,63 and thePGC has recently suggested that as much as halfof the tight gas in its estimate may represent anoverlap with the NPC resource64 An OTA anal-ysis of the overlap, presented in appendix C, in-dicates that the total overlap between the PGC-estimated 172 TCF of tight gas and the N PC-esti-mated 606 TCF of recoverable tight gas may be

as low as 30 TCF. Even the low side of the over-lap, however, though relatively small comparedto the total size and large uncertainties in the tightgas resource estimates, is still important in assess-ing the unconventional tight gas contribution tonear- and mid-term production because the 30TCF are in the most accessible, least expensiveportion of the tight gas resource. The main areas

of overlap occur in the blanket formations of theRocky Mountain Basins and the Cotton Valley

Trend in Texas and Louisiana. These areas cur-rently are the major producers of gas from tightformations, and the NPC report has predicted

these areas to be the main contributors to “un-conventional” supply in the next 20 years.

Deep Tight Gas.–Another problem arising frominadequately defined boundary conditions is thatcertain potential resources may be overlooked

entirely, resulting in an underestimate of the totalresource base. For example, the PGC estimated

that its conventional gas resource base includessome 89 TCF of gas in tight formations at depthsgreater than 15,000 ft. The NPC report acknowl-edges that large volumes of gas probably occurat depths greater than 15,000 ft, but they assumed

this gas to be too speculative a resource to in-

63vel10 KUUSkraa, Lewlrl & Associates, Inc., personal communi-

cation, 1984.MH. C. Kent, Director, Potential Gas Agency, testimony before

the Subcommittee on Energy Regulation, Senate Committee on

Energy and Natural Resources, Apr. 26, 1984.




elude in its tight gas assessment.65 Other expertshave concurred that large volumes of gas existin deep tight formations, in the Rocky MountainBasins as well as the Anadarko and Arkoma

Basins in Oklahoma and Arkansas. Many of the

deeper Rocky Mountain formations are Ienticular.

Apart from the PGC estimate, which is widelyviewed as an estimate of the conventional re-source, there are no quantitative estimates ofrecoverable resources in deep tight formations.Deep tight formations, other than those alreadyincluded in the PGC estimates, probably shouldbe included in future estimates of the unconven-

tional tight sands resource at least for the esti-mates of gas-in-place. They represent no more

speculative a source of gas supply than the un-fractured portions of the Devonian shales or the

deep coal seams, both of which generally are in-

cluded in gas-in-place estimates of unconven-tional resources.

Conclusions About Recoverable Gas

To summarize, OTA considers the NPC esti-mates of recoverable tight gas resources to be op-

timistic. The major reasons for this appraisal are:

1.

2.

3.

The Northern Great Plains. OTA considersit quite plausible that the recoverable re-

sources in the Northern Great Plains are con-siderably lower than the 100 TCF maximumrecoverable gas projected by the NPC.

Gas-in-Place. Aside from the Northern GreatPlains, OTA believes the NPC’s gas-in-place

estimates to be reasonable, in the contextof “most likely” values. The margin for er-ror is large, however, especially for the ex-trapolated resource; the estimate for the ex-trapolated resource should be considered asspeculative.Technology. OTA considers some important

aspects of NPC’s appraisal of current tech-nology and projection of advanced technol-

ogy to be over-optimistic. Of particular con-cern is the assumption that fractures will be

able to penetrate and drain lenses remote

.—bSAccording to Ovid Baker, Chairman of the NPC Tight Gas Task

Group, deep tight gas below 15,000 ft would cost more than

$12/MCF to produce, and the NPC upper limit on price was $9/MCF

(Ovid Baker, personal communication, 1984).

from the wellbore. Another important con-cern is that very long fractures (up to 4,000ft) can be achieved with a net decrease infracture height.

OTA also believes that all available estimates

of recoverable tight gas are highly uncertain be-cause of poorly defined reservoir characteristicsand technologic uncertainties. However, thereseems little doubt that large quantities—at leasta few hundred TCF—of tight gas will be recover-able even under relatively pessimistic technologiccircumstances, provided gas prices reach mod-erately high levels in the future (e.g., $5 to $7/ MCF in 1984$).

Annual Production Estimates

Any forecasting of actual production from thetight gas resource requires making a number of

assumptions over and above those made to esti-mate the recoverable resource, and similarly

uncertain. Projections must be made of the priceof gas and the state of technology development

over the time interval of the forecast. These arethe critical uncertainties. The amount of “highpotential” land immediately available for drillingmust be estimated. Finally, an estimate must bemade of the number of wells that can be drilledwithin a single year and the rate of increase infollowing years. The comparison of different fore-casts requires an understanding of these under-

lying assumptions.The Lewin study used two different develop-

ment schedules, described in table 42, to deter-mine potential production under base and ad-vanced technology conditions. The schedulesrepresent a rapid early development of the re-source. For the base case, tight sands productionis 3 TCF in 1990 and rises to nearly 4 TCF in 2000,at $3/MCF (1 977$). The advanced case contrib-utes 4 TCF by 1985 and nearly 8 TCF in 1990 but

declines to less than 7 TCF by the year 2000. Incontrast, the NPC standard scenario at $5/MCF(1979$) will contribute only 2 TCF by 1990 but

up to 9 TCF in the year 2000.

The NPC approach was to assume that thereare few external constraints on most of the fac-tors affecting production. Rather than a forecastof the “most probable” production, its supply




Table 42.-Projections of Annual Tight Gas Production, TCF

Assumptions 1990 2000

Lewin: All drilling begins in 1978 and IS completed by

2003Base case: drill probable acreage immediately, lag

drilling of possible acreage 9 years; 20 percent

DCF RORAdvanced case.’ lag drilling of probable acreage 3

years, lag drilling of possible acreage 9 years

but complete in 15 years; 10 percent DCF ROR

NPC: Standard scenario, phase in advanced technology,

5 percent by 1983, 100 percent by 1989 inblanket sands; 2 years lag in Ienticular sands;40 percent of most profitable prospects ( > 50%DCF ROR) drilled in first 20 years; 4 years from

first drilling to initial production.

2 times standard: faster technology developmentand drilling schedule,

1/2 standard.’ slower technology development anddrilling schedule

GRI:

Assume average production decline curves,increase in numbers of wells drilled by 200 per

year.

Base case.’ begin with 100 wells in 1984 to 800 in1988 and 10 percent to 15 percent growthbeyond 1988

Advanced case. begin with 200 wells in 1984;increase to 800 in 1987 with 10 percent to 15percent growth beyond 1987.

AGA:Price and rate of return not defined; average

production decline curve similar to GRI; lowerdrilling rates than NPC; slower implementation ofadvanced technology

Conservative case: lower initial average productionrates to account for areas where massivehydraulic fracturing cannot be used

Consensus case.’ modified to account for slowtechnologic development and overlappingconventional and tight definitions.

Pricea

($/MCF) Base Advanced Base Advanced

3.00 (4.70) 3.2 7,7 4.0 6.8

Price($/MCF)15% ROR 1/2 std Standard 2 x s t d 1/2std Standard 2 X std

2,50 (3.35) 1.1 5.8

3.50 (4.65) 1.3 7.05,00 (6.70) 0.9 1.8 2,2 4.1 8,3 15.59.00 (12.00) 2.5 10,5

Price

a

($/MCF) Base Advanced Base Advanced

3.12 (4.20) 0.48 0.93 1.99 3.444.50 (6,00) 0.53 1.02 2.51 5,216.00 (8.00) 0.57 1,12 3.31 6.04

Conservative High Conservative High

Calculated 1.1 1.2 4.3 6.2C o n s e n s u s — 1.2-3 b

aPrice in dollars at the time of the study(1983$)

bRange includes Devonian shale

scenarios are quite deliberately structured to rep-resent a maximum—the amount of gas that couldbe produced given a concerted effort to developthe required technology and no market restric-tions 66 on field development.

The NPC study combines the resource base

and economic model to develop several scenar-

ios for production of the tight gas resource. The

66Tha( is, producers can sell all the gas they can produce and

transport to potential markets, at the assumed wellhead price plustransportation costs.

scenarios incorporate a development schedulein which a certain percentage of the more prof-itable prospects in all basins are drilled first. Cer-tain constraints were incorporated in the devel-opment schedule to reflect the fact that not allthe leases on the most profitable prospects willbe immediately available for drilling. Develop-ment in some areas, such as the Northern Great

Plains, is delayed due to lack of pipelines. Ad-vanced technology is phased in according to spe-cific schedules. Annual production rates and cu-muIative additions to reserves for each scenario

3 8 - 7 4 2 0 - 8 5 - 1 2




were calculated for various prices and rates ofreturn. Results of the different scenarios areshown in table 42.

Figure 33 compares several of the NPC scenar-ios with the Lewin base and advanced case. The

primary difference between the two sets of esti-mates is the rapidity with which profitable pros-pects are developed in the Lewin study, and alsoin the period of time between discovery and ac-tual production from a prospect. In the NPC

study, it takes 4 years to place a well under pro-duction after drilling; in the Lewin study, produc-

tion begins the year after discovery.

GRI took a slightly different approach to fore-casting annual production from tight sands in the

earlier (1980) of its two analyses.67 It determined

a production decline curve representative of an

“average” well with a production life of 30 yearsand specified a rate of increase in the numberof wells drilled per year. Initial production ratesand rate of drilling varied for the base and ad-

vanced case scenarios. Details are given in table42, Its production estimates for base and ad-vanced technology cases range from 0.48 to 1.12TCF in 1990 and 1.99 to 6 TCF in the year 2000.

A more recent GRI production analysis68

i sbased on the TGAS model of the NPC study

67. C. Sharer and j. j. Rasmussen, “Position Paper: Unconven-

tional Natural Gas,” Gas Research Institute, Mav 1981.‘Sj. 1. Rosenberg, F. Morra, Jr., and M. M’archl’ik, “Future Gas

Contributions From Tight Sand Reservoirs, ” 7984 Irrterrratiomi CmResearch Conference, Sept. 10-13, 1984, Washington, DC.

Figure 33.-Annual Tight Gas Production Estimates(wellhead gas price = $4.70/MCF, 1983$)

1980 1985 1990 1995 2000

Year

“No gas pricespecified.



ChI. 8—Tight Gas 1 7 1

methodology, but adopts a series of advancedtechnology assumptions that are more modestthan the NPC’s. The production projections arebased on a price path rather than a constant realprice, with gas prices assumed to reach $6.20/MCFin 2000 and $7.80/MCF in 2010 (all prices in

1982$), and assuming a 15 percent discount rate.The model is deliberately calibrated to producea year 2000 base production level at around theaverage of current projections by AGA and DOE,

so the value of the analysis lies in the sensitivityanalyses around the base production level (de-scribed later), and not the level itself.

The production estimates for the base and ad-vanced technology cases range from 1,9 to 2.7TCF in 2000 and 2.9 to 5.6 TCF in 2010. An ad-ditional set of estimates was made for a casewhere no limits were set on gas demand, capital

and drill rig availability, or level of investment.As might be expected, the production estimatesfor this case are very high, 8 to 11 TCF in 2000and 8 to 14 TCF in 2010. The investment anddrilling implications of this latter case appearunrealistic, but the estimates give some ideaabout what might be possible with an emergencydevelopment program.

An additional estimate of future production

rates is given by the American Gas Association(AGA),69 AGA used the NPC’s analysis as a start-ing point and superimposed its own assumptions.Its initial estimate projected annual production

of 1.2 TCF in 1990 and 6.2 TCF in the year 2000.AGA attributes these lower estimates (relative tothe NPC forecasts) to lower drilling rates and a

slower implementation of advanced technology.A more conservative estimate, based on lowerinitial production rates per well, projects 1.1 TCF

in 1990 and 4.3 TCF in 2000. No price levels areindicated. AGA suggests that even lower annualproduction (from 1.2 to 3 TCF from both tightsands and Devonian shales) is likely due to defini-tional overlap between conventional and tightreservoirs (i. e., some of the production formerlyconsidered to be unconventional more properly

belongs in the conventional category) and a lessoptimistic outlook for the potential of advanced

————bgArrlerlCan GaS As s oc i a t i on , Tbe ~as ~r)f?rgy ~U~/J/Y @~~oo~:

1983-2000, October 1983.

technologies. This lower production range prob-

ably is not directly comparable to the NPC’s,however, because the AGA definition of uncon-ventional gas appears to exclude gas resourcesnot yet under development that are economicallyrecoverable at today’s price and technology—

resources that are included in the NPC’s tight gasresource base. Such resources, which are in a“grey area” between clearly conventional andunconventional resources, probably make up atleast 100 TCF of the NPC’s recoverable resource

and play a major role in NPC’s projected produc-

tion during the first few decades of tight gas de-velopment.

The time frame and extent of technological de-velopment is a major factor contributing to theuncertainty of actual production from the tightgas reservoirs. The production of most of the re-

source contained in the lower permeability for-mations and the Ienticular formations is stronglydependent on advanced technology develop-ment. A slower rate of technology developmentmay severely limit the potential contribution ofthe tight sands resource to U.S. supply in the next20 years. For example, NPC made its scenarioprojections based on immediate utilization of

base technology, and implementation of ad-vanced technology in all blanket formations by

the year 1989 and in all Ienticular formations by

1991. Events of the last few years have indicatedthat this rate of development is too optimistic.

As discussed previously, the base case technol-ogy criteria have not yet been met, especially notin Ienticular reservoirs.

Aside from the rate at which new technologywill be developed, the advanced technology

characteristics assumed by the NPC appear to bequite optimistic, and GRI has used more modestassumptions. However, in reality the specifica-tion of future technologies has always been a riskyundertaking, and past efforts in other technologi-cal areas have not tended to be wildly successful.As shown by figure 34 and by an examinationof the “base” and “advanced technology” pro-

duction projections in each of the studies, theadequacy of exploration and production technol-ogy is critical to the economics of tight gas de-velopment, and errors in projecting the future

state of technology may be translated into sub-




Figure 34.—incremental Tight Gas Sands ProductionRates As a Function of R&D Advances

8

6

4

2

01985 1990 1995 2000 2005 2010

Years

SOURCE: Lewin & Associates, Gas Research Institute.

stantial errors in projecting future productionlevels at a particular gas price.

Figure 34 shows GRI’s projected productionlevels at current technology, advanced technol-

ogy, and a modified advanced technology casewhere the ability to produce Ienticular reservoirsnot in direct contact with the well bore has neverbeen achieved. Because of the time delay in-volved in developing and implementing new

technologies, the three cases do not diverge un-til the mid-1990s, but by 2000 the advanced tech-nology case produces about 50 percent morethan the current technology case. By 2010, theadvanced case produces 100 percent more—5.6TCF/yr v. 2.9 TCF/yr About half of the additionalproduction would be lost, however, if remotelenses cannot be produced.

Another important area of uncertainty is the po-

tential for errors in resource assessment. A num-ber of the resource estimates for individual basinshave been questioned. Figure 35 breaks out theNPC standard development scenario into thecontributions by each basin. In this way, the ef-fect of delaying or otherwise constraining devel-opment of certain basins can be more easilyvisualized. Significant errors in the resourceassessments for these basins might be reflectedin errors in the production estimates. However,because operators will shift from one target to

another if the first does not meet their expecta-tions, the effect on national production of anoverestimate of resources in a single basin mayinclude both an overestimate of production from

that area and a partially offsetting underestimateof production from adjoining areas.

A significant contributor to annual productionis the group of Eastern basins. These fall in theextrapolated portion of the resource assessment;recoverable resource estimates for these basins

are subject to considerably greater uncertaintiesthan those for the appraised basins.

Another area that deserves particular attentionis the Northern Great Plains, which has been thesubject of considerable argument concerning themagnitude of its recoverable resource. As illus-trated by figure 35, the Northern Great Plains(NGP) plays only a moderate supply role in the

NPC standard scenario. One cause of this moder-ate role may be that the NPC assumed that lackof pipeline availability would significantly delaydevelopment in this region. However, the rela-tive development costs of the NGP’s shallow gas

resource are projected to be quite low, and asupply analysis that assumed fewer supply restric-tions would project a more extensive role forNGP gas. Figure 36 illustrates GRI’s TGAS pro-

jection of tight gas production with and without

the NGP, with current and advanced technology.The projection assumes a gradually rising gas

price rather than the NPC’s constant real price.For both the present and advanced technologycases, removal of the NGP drastically curtails totaltight gas production, especially in the short term.For the present technology case, production only“recovers” to about half the reference case by

the year 2010. The effect of “losing” the NGPin the advanced technology case is less severe,

with production in 2010 about two-thirds of thereference production.

Another potential supply constraint, the total

size of the resource, probably will not significantlyaffect production rates over the next 20 years.in fact, if development of the tight gas resource

continues to be as slow as in the last few years,

this boundary constraint is not likely to be feltuntil well into the next century. it may, however,have an impact on the long-term contribution to

supply. Comparison of the NPC and Lewin sce-narios demonstrates the effect of a smaller totalresource on production. Supply from the Lewinappraised resource will begin to decline near the




Figure 35.—Annual Production by Basin— NPC, $5.00/MCF (1979$), 15°/0 DCF ROR, Standard Scenario

.

D

I

Midcontinent

Rocky Mountains I

Northern Great Plains

1985 1990 1995

SOURCE National Petroleum Council.

Figure 36.—Incremental Tight Sands Gas ProductionRates With and Without the Northern Great Plains

Resource8

6

4

2

01985 1990 1995 2000 2005 2010

Year

SOURCE Lewln & Associates, Gas Research Institute.

2000 2005 2010 2015

Year

year 2000, apparently due to the boundary con-straints imposed by the size of the resource. Incontrast, production from the total U.S. resourceappraised by the NPC continues to increase wellinto the next century and remains as a significantsource of supply until at least 2040 (NPC report,fig. 27).

The absence of pipelines in many potential tight

sands production regions further constrains near-and mid-term development of the tight sands re-source. Development of new regions has always

been something of a problem for gas develop-ment: pipeline companies cannot get approvalto build new lines without evidence of proved

reserves, whereas producers are reluctant to drill

.




and prove reserves without the presence of anearby line. For tight gas fields, pipelines may re-

quire evidence of larger reserves than are pres-ently required for conventional fields, becausethe tighter reservoirs produce more slowly.

In the short term, pipeline constraints will af-

fect the pattern of development. Fields near ex-isting pipelines will be developed first. Alreadymuch of the new production from tight sands

fields is coming from the Southwest, which has

excess pipeline capacity. In other areas withoutpipelines, despite potentially large volumes of gas

and low estimated costs of extraction, develop-ment is likely to be delayed for several years dueto lack of pipeline capacity.

Aside from technical considerations, futuretight gas production will be dependent on futuregas prices and investment discount rates. For ex-

ample, figure 37 illustrates how production willchange if prices are 20 percent higher or lower

than the projected values in GRI’s analysis. Theprice variation produces a production variationin the present technology case of about 1 to 3TCF in the year 2000, a major difference in pro-

duction for a relatively modest range of gas prices.Similarly, the range of year 2000 production pro-

jections for the advanced case will be about 2to 4 TCF. Because future gas prices are likely to

be closely tied to unpredictable oil prices, thechance of estimating year 2000 gas prices to with-in better than +/- 20 percent—and thus, estimat-

ing production to better than about +/- 1 TCF—seems remote.

Figure 38 shows the effect of changes in dis-count rate on production. Discount rate is essen-tially a proxy for the perceived risk associated

with an investment. The range of discount rates—10 to 20 percent–displayed in the figure seemsa reasonable measure of uncertainty for the pres-ent technology case because the profile of thepotential developers, the development climate,and the physical uncertainty associated with de-velopment expenditures are not well understood

at this time.As illustrated by the figure, discount rate is a

crucial variable over the entire course of devel-

opment and for both base and advanced tech-nology. For example, in the year 2000, the 10

Figure 37.—incremental Tight Sands Gas ProductionRates as a Function of Gas Price

8

01985 1990 1995 2000 2005 2010

Year


Figure 38.—incremental Tight Sands Gas ProductionRates As a Function of Discount Rate

I

Present I

1985 1990 1995 2000 2005 2010Year


point spread in discount rate introduces an uncer-

tainty of about +/- 50 percent in the expected pro-duction for the present technology case and+ 100 percent, –5o percent for the advancedtechnology case.

In the NPC and Lewin reports, and possibly in

the GRI and AGA reports as well, little or no con-sideration was given to the possibility that mar-ket problems might constrain the future produc-tion of tight gas. in the past few years, however,declines in gas usage and the widely perceivedoptimistic prospects for conventional gas supplyand price stability have altered even the enthusi-asts’ perception of the near-term future of tightgas and other forms of unconventional gas, It isgenerally considered improbable that massive fi-

nancial resources will be channeled into devel-opment of very low permeability formations ifample prospects of conventional gas are avail-



.


able. Consequently, many supporters of the NPCstudy, while remaining convinced of the accuracyof the gas-in-place and recoverable resource esti-mates, no longer consider the high production

projections to be very likely. To a certain extent,

more pessimistic projections of future tight gas

production can be self-fulfilling because theiracceptance is likely to discourage the researchnecessary for a major expansion of tight gas pro-duction. on the other hand, at least a moderaterate of improvement in tight gas exploration andproduction expertise and technology will con-

tinue regardless of markets, because of continuedresearch on oil production from tight formations

and the momentum of existing research programsand tight gas development.

OTA believes that future tight gas developmentwill be closely linked to the availability of con-

ventional gas resources. As discussed in Part I ofthis report, we do not believe that the year 2000

supply of conventional gas can be projected with-out a large error band. Consequently, we are

skeptical of our ability to reliably project tight gasproduction to the year 2000 except on a “whatif . . .“ basis.

On an optimistic basis, we do believe that theyear 2000 incremental production of tight gas 70

zOThat is, over and above the 1 TCF/yr or so of tight gas produced

today that is generally included in “conventional” productionfigures.

can reach 3 or 4 TCF/yr, or perhaps even some-what higher, if the present gas bubble ends withina year or two, markets for gas remain firm andreal prices increase steadily for the remainder ofthe century, and the industry is confident of thelong-term marketability of tight gas and thus is

willing to make the necessary investments inR&D. Such a future is consistent with the pessi-mistic end of OTA’s projected range of 9 to 19TCF/yr for year 2000 production of conventional

gas.

On the other hand, there are plausible circum-stances that could stifle future tight gas produc-tion, including high conventional gas production

and stable or declining real prices, low future de-mand for gas, or the loss of industry confidencein future gas marketability. A pessimistic scenariomight involve year 2000 incremental production

of 1 to 2 TCF/yr. Beyond 2000, the size of therecoverable resource, as affected particularly byproduction technology and the availability of gasin the Ienticular basins and the Northern GreatPlains, will play a critical role in determining the

magnitude of production. In addition, the pro-duction rate will always be extremely sensitiveto the introduction of any new technological ad-vances affecting production costs and recoveryefficiency.



Chapter 9

Gas From Devonian Shales



Contents

Page

introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... ... ... ..179

Characteristics of the Devonian Shale Resource . . . . . . . . . . . . . . . .. ... ... ...179

Gas-In-Place Resource Base . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..182Methodologies and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...182

Estimate Comparison and Uncertainties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .186

Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .........................187Fracturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...........................187Deviated and Directional Drilling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ....189Exploration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..........................189

Recoverable Resources and Production Potential . ........................190Methodologies and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .191Estimates Comparison and Uncertainties. , . . . . . . . . . . . . . . . . . . . . . . . . . . . . .199

Annual Production Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... ... ....201

TABLES

Table No. Page

43.44.45.

46.47.

4849.50.

51.

52.

53.

54.55.

Devonian Shale Resource Base Estimates (TCF) . .......................182Equation Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........184Estimates of ln-Place Natural Gas Resources in the Devonian

Shale of the Appalachian Basin . . . . . . . . . . . . . . .......................184

Devonian Shale Recoverable Resource Estimates (TCF) Appalachian Basin .191

Summary of Major Differences Between Lewin Base and AdvancedCases in Devonian Shale Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193Lewin &Associates: Results of Economic Analysis, Summary Table . ......194Summary of Producible Gas Estimates, Appalachian Basin . . . . . . . . . . .. ...195Effect of Spacing on the Devonian Shale Recoverable Resource . . . . . . . ...196

Results of Lewin Assessment of Technically Recoverable Gas in Ohio,By Stimulation Method After 40 Years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .198Results of Lewin Assessment of Technically Recoverable Gas inWest Virginia, by Stimulation Method After 40 Years... . ...............198

Comparison of Estimated Appalachian Devonian Shale ResourcesRecoverable With Borehole Shooting, Well Spacing of 150 to 160 Acres ..200Lewin &Associates Results for Annual production Estimates . ............202Potential Incremental Supply of Devonian Shale Gas in theAppalachian Basin: NPC -High Growth Drilling Schedule. . ..............200

FIGURES

Figure No. Page

39. Primary Area of Devonian Shale Gas Potential . .......................18040. Averaged Production Decline Curves for 50 Devonian Shale Gas Wells ...18141. Shale Gas Potential in the Appalachian Basin . ........................18542. Distribution of Devonian Shale Gas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...187



Chapter 9

Gas From Devonian Shales

INTRODUCTION

Although it is often presented as a gas source

of the future, Devonian shale gas actually has along production history: the first Devonian shale

gas well was drilled in 1821, near Fredonia, NY,and “modest” production from Devonian shalewells began around the 1920s and has continuedto the present.1 Cumulative production from theshales during all the years of production has beenless than 3 trillion cubic feet (TCF), most from the

‘Potential Gas Agency, “Potential Gas Resources From Non-

conventional Sources, ” Potential Supply of Natural Gas in the

United States, May 1981.

Big Sandy Field in Kentucky and adjacent West

Virginia, and current production is only about 0.1TCF/yr.

Because of its history, Devonian shale gas maybe thought of as a “conventional” gas resource.It is also an unconventional gas resource, how-

ever, because of its complex geology and becauseadvanced exploration and extraction technol-ogies and higher prices may be able to transformit into an important component of U.S. gas supplyfrom its current status as a very limited, if stilllocally important, gas source.

CHARACTERISTICS OF THE DEVONIAN SHALE RESOURCE

Devonian shale gas is defined as natural gasproduced from the fractures, pore spaces, andphysical matrix2 of shales deposited during theDevonian period of geologic time. As illustratedin figure 39, Devonian shales occur predominant-

ly in the Appalachian, Illinois, and Michigan

basins. The shales formed approximately 350 mil-lion years ago in a shallow sea that covered theeastern half of what now constitutes the continen-

tal United States. organic-rich muds and siltswere deposited in the sea and subsequentlyburied by younger sediments. The high pressuresand temperatures that accompanied burial of thesediments resulted in the formation of natural gasfrom the organic material.

The gas content of the shales is proportionalto the amount of organic material, and more pre-

cisely the organic carbon, present in the rock.The organic material occurs as microscopicallythin layers, alternating with mineral layers. Theactual physical color of the shales is indicativeof their organic content: black and brown shalesgenerally have higher organic contents and there-fore more gas than gray shales.

‘That is, a portion of the gas is adsorbed, or bound, to the actualphysical structure of the shale.

Other determinants of the gas content are theorigin of the organic material, that is, the type oforganisms (algae, pollen, woody plants, etc.) thatformed the sedimentary layers which became theshale, and the physical conditions, especially thetemperature, to which the organic material wasexposed. For example, blooms of algae appearto be the major source of Devonion shale gas,whereas terrestrial organisms are less promising

sources of gas.3

And temperature conditions be-tween 60° C (140° F) and 150° C (3020 F) areoptimal for the formation of petroleum (oil andgas).4

Unlike accumulations of natural gas that areconsidered conventional, and unlike tight sandsgas, Devonian shale gas did not migrate fromsource rocks to reservoir rocks and accumulatein a trap. Instead, the low permeability of theDevonian shale prohibited most of the gas fromescaping. As such, the shale is effectively thesource, reservoir rock and trap for the gas. How-ever, some gas originally present in the shale may

3R. A. Struble, Evaluation ot’ the Devonian Shale Prospects in the

Eastern United States, U.S. Department of Energy Report

DOE/MC/19143-1305, undated.

4j. M. Hunt, Petroleum Geochemistry and Geology (New York:

Freeman, 1979), p. 617. Cited in R. A. Struble, op. cit.

179




Figure 39.—Primary Area of Devonian Shale Gas Potential

SOURCE: Johnston & Associates, OTA Contractor,

have migrated out of the formation, escaping to

the atmosphere or forming conventional gas ac-cumulations in nearby sandstones.

The reservoir characteristics of Devonian shalediffer substantially from those of conventionalreservoirs. Porosities range from 8 to 30 percentin conventional reservoirs; Devonian shale ma-

trix porosities are generally 1 to 2 percent. The

permeability of the shales is also significantlylower. Conventional reservoirs have permeabil-

ities in the range of 1 to 2,000 millidarcies (red),whereas Devonian shale matrix permeabilities

generally range from 10 -5 to 10 -6 md. Eventhough portions of the shale contain natural frac-tures, fracture permeabilities tend to be low, rang-



Ch. 9—Gas From Devonian Shale . 181

ing from about 0.001 to 1 md; in most cases, per-meabilities are less than about 0.1 md. s Assuggested by these statistics, gas flows much lessreadily from most Devonian shale reservoirs thanfrom conventional sandstone reservoirs.

The natural fractures in the shale, which most

often occur in a vertical pattern, are critical tosuccessful production. Such production gener-ally requires intersecting these fractures to utilizethe increase in overall permeability that they pro-mote. Because the shale fracture systems in thegreat bulk of the Appalachian Basin are still quite

tight, however, achieving high recovery efficien-cies generally also requires inducing new, proppedfractures in addition to connecting with the nat-ural system. Also, aside from their “tightness,”the shale fracture systems tend to be somewhatlined up rather than being random in direction–a

property called “anisotropy.” Optimum fracture

design and well spacing are affected by this prop-erty, e.g., a rectangular well spacing patternaligned with the direction of anisotropy will in-

crease gas recovery over the usual squarepattern. 6

Typically, production from Devonian shalewells is at first relatively high, followed by a steadydecline to a base level which can remain con-stant for over 50 years. Four production curvesrepresenting the averaged production of multi-ple wells are included in figure 40. The shape ofthe production curves probably is a result of the

multiple ways in which the gas occurs in the rock:in pore spaces, in the fracture system and ad-sorbed to the shale matrix. The initial productionis composed primarily of the free gas containedin the fracture network immediately connectedto the wellbore and that pore gas which readilymigrates to the well bore. The base level then rep-resents the rate at which the gas diffuses throughand desorbs from the shale matrix. However, the

relative contribution of each of the three distinct“sources” of gas in the shale is not completely

‘V. A. Kuuskraa and D. E. Wicks, “Devonian Shale Gas Produc-

tion Mechanisms, ” 1984 International Gas Research Conference;

V. A. Kuuskraa, et al., Technically Recoverable Devonian Shale Gas

in Ohio, Lewin & Associates Report for Morgantown Energy Tech-

nology Center , July 1982.

6V. A. Kuuskraa and D, E. Wicks, “Devonian Shale Gas Produc-

t i on Mec han i s ms , ” 1984 In ternat ional Gas Research Conference.

Figure 40.—Averaged Production Decline Curves for

understood, and there are alternative interpreta-tions of the precise composition of the base pro-duction level. One interpretation is that the baselevel is primarily adsorbed gas that is being re-leased by the shale matrix as the pressure drops.An alternative explanation is that the base level

gas is primarily gas from other gas-bearing inter-vals that are connected to the primary (stimu-lated) interval by the vertical fracture network inthe shale.

Deciphering the relative roles of these twomechanisms is critical to estimating the recover-able resource. At one extreme, if the base levelof production is mostly adsorbed gas and thereis little communication between gas-bearing in-

tervals, then the intervals that are not currentlybeing stimulated may be available for productionin the future. At the other extreme, if there is lit-

tle resorption of gas and the base level is dueto vertical communication between intervals,

then the number of targets for economic produc-tion is drastically reduced and the recoverableresource will be far less. In the latter case, the




major part of the gas-in-place (Kuuskraa and The current available data appear to support theWicks estimate that adsorbed gas represents over “resorption, little vertical communication” inter-80 percent of the total7) will not be available for pretation, but these data are limited to a veryproduction with currently foreseen technology. small geographical area.8

BCharles Komar, Morgantown Energy Technology Center, per-7Kuuskraa and Wicks, op. cit. sonal communication, 1984.

GAS= IN-PLACE RESOURCE BASE

The Devonian shale resource is becoming in-creasingly well characterized as a result of recentefforts to better understand the geological char-acteristics of the resource and its size. Work per-formed as part of the Department of Energy’s(DOE) Eastern Gas Shales Project (EGSP) has pro-vided substantive geological and geochemicaldata.9


As illustrated in table 43, several organizationshave estimated the size of the Devonian shale re-source base. The three most recent estimates ofthe in-place resource were made by the National

gThe results of the project are summarized in a report by R. A.

Struble, Evaluation of the Devonian Shale Prospects in the Eastern

United States, U.S. Department of Energy Report DOE/MC/19143-

1305, undated. This report provides a comprehensive review of

the “state-of-the-art” of Devonian shale resource analysis up to

about 1982.

petroleum Council (NPC) in June 1980 and the

U.S. Geological Survey (USGS) and the MoundFacility (operated by Monsanto Research Corp.)in 1982. The NPC study evaluated the Devonianshale resource in the three shale basins, whereasthe other two restricted their estimates to the Ap-palachian Basin.

National Petroleum Council10

The NPC estimated the gas-in-place resourcefor each of the three major basins. The primaryvariables in the in-place resource calculationwere shale thickness, areal extent and gas con-tent, with gas content assumed to be uniform

throughout each basin. These parameters wereestablished differently for each basin, dependingon the type and quantity of information available.

locational Petroleum council, “Unconventional Ndtural Gas-Devonian Shales,” June 1980.

Table 43.-Devonian Shale Resource Base Estimates (TCF)

Organization Year Basin evaluated Estimate

National Petroleum Councila. . . . . . . . . . . . . . . 1980 Appalachian . . . . . . . . . . . . . . . . 225 to 1,861 (125 to 1040b)

Michigan . . . . . . . . . . . . . . . . . . 76Illinois. . . . . . . . . . . . . . . . . . . . . 86

U.S. Geological Surveyc . . . . . . . . . . . . . . . . . . . 1982 Appalachian . . . . . . . . . . . . . . . . 577 to 1,131Mound Facilityd . . . . . . . . . . . . . . . . . . . . . . . . . . 1982 Appalachian . . . . . . . . . . . . . . . . 2,579 (l,440

b)

Lewin & Associatese. . . . . . . . . . . . . . . . . . . . . . 1980 Appalachian . . . . . . . . . . . . . . . . 400 to 2,000

Federal Energy Regulatory Commissionf. . . . 1978 Appalachian . . . . . . . . . . . . . . . . 285

Smithg. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1978 Appalachian 206 to 903

aNational Petroleum Council, Unconvenyional GasSources, Tight Gas Reservoirs Part 1, December 1980.bConsidering “drillable” area only.C R.R. Charpent ier et al. , Es t imates of Unconvent ional Natural Gas Resources of the Devonian Shale Of the Appalachian Basin, USGS Open-File Report 82-474, 1982.dR, E, Zielinski and R D, Mclver, Resource and Exploration Assessment of the oiI and Gas Potential in the Devonian shales of the Appalachian Basin, DOE/DP/0053-1 125,

undated.eV. A. Kuuskraa and R. F. Meyer, “Review of World Resources of Unconventional Gas,” in IIASA Conference on Conventional and Unconventional World Natural Gas

Resources, Laxenburg, Austria, June 30-July 4, 1980.fU.S., Department of Energy, Nonconventional Natural Gas Resources, Report DOE/FERC-0010, 1978.gE. C. Smith, “A Practical Approach to Evaluating Shale Hydrocarbon Potential,” in Second Eastern Shales Symposium, Vol. II, U.S. Department of Energy, Report

METC/SP-7816, 1978, pp. 73-87.




Ch. 9—Gas From Devonian Shale q 183

In each basin, the black and gray shale thick-nesses were multiplied by their respective gas

contents and their areal extents to arrive at thegas-in-place estimate for the basin.

In the Appalachian Basin, the gas-bearing zoneincludes the gray and black shale units that

overlie the Onandaga limestone and underlie theBerea sandstone. The thickness of the blackshales was determined in two ways: first, bygamma-ray well log data which detects the highradioactivity content characteristic of organic-rich

shales; second, by thicknesses determined visual-ly from core sample color. The two black shale

thickness calculations yield substantially differentresults, which in turn yield very different gas-in-place estimates. The gas content of the shales wasdetermined through off-gassing data from rockcore samples.11 Values of 0.6 and 0.1 cubic feetof gas per cubic feet of Appalachian shale were

obtained for black and gray shales, respectively.The areal extent of the Appalachian shale was

determined to be 111,100 square miles.

The gas-bearing unit in the Michigan Basin isthe Antrim shale, which extends over 35,400square miles and contains both black and greyshales. The thickness of the Antrim was deter-mined strictly by well logs and is poorly definedwhere the Antrim grades into the barren Ellsworth

shale in the western portion of the basin. Also,there were no core samples available for off-gassing experiments to determine the quantity of

gas present in the unit. In the absence of thesedata, the gas content of the Michigan Basin shaleswere assumed to be the same as those of the Ap-palachian Basin on the basis of similarities in thewell production data for the two basins.

The New Albany shale group, covering 28,150

square miles, is the gas-bearing unit in the IllinoisBasin. Neither gamma-ray well log data nor core

sample data were available to determine thethickness of the units, and therefore the black andgray shale thicknesses could not be differentiated.The thickness of the entire sequence was deter-

mined from USGS maps and used in a simplevolumetric resource calculation. off-gassing datafrom cores were available to determine the gas

.-.——1 Ioff.gassing measures the amount of gas that desorbs from a

known volume of core over a specific period of time.

content. The thickness of the sampled units inproportion to the total group thickness was usedto establish a weighted average of 0.62 cubic feetof gas per cubic foot of shale for the entire NewAlbany group.

The results of the NPC study suggest that esti-

mates of the quantity of gas present in the Ap-palachian Basin are sensitive to the assumedthickness of the black shale and to the inclusionor exclusion of the lower quality gray shales.Based on thicknesses determined by gamma-raylogs, and including only the black shales, the gas-

in-place is estimated to be 225 TCF. Based onblack shale thickness determined visually fromUSGS samples and including the gray shales,1,861 TCF is estimated to be present. (The rangefor the black shales only is 225 to 1,102 TCF.) Ifareas that are not drillable12 are excluded, the gas-in-place estimate is reduced to 125 and 1,040 TCF

for the log and sample thicknesses, respectively,The gas-in-place estimates for the Michigan andIllinois basins are 76 and 86 TCF, respectively,

but are more uncertain because of the lack ofdata.

U.S. Geological Survey13

The USGS estimate of Devonian shale gas inthe Appalachian Basin recognizes three catego-ries of shale gas: macrofracture gas, microporegas, and that gas which is adsorbed, or attached,to the clay matrix. Unlike the early Lewin & Asso-

ciates study (1 978-79), the USGS attributes onlya small amount of the total volume of gas-in-placeto macrofractures; it assumes that most of the gasis in micropores or bound to the organic matterin the shale matrix.

The Appalachian Plateau province and a smallsegment of the Valley and Ridge province were

divided into 19 areas, termed plays. The charac-teristics of each play were described in terms ofphysical location, unit names, thickness, organiccontent, maturation level and type of hydrocar-bon present, tectonic or structural attributes, and,

subsequently, a brief description of the produc-I zBeCaUSe of difficult terrain, presence of buildings and other de-

velopment, designation of the land as protected parkland, etc.1 JR. R. Charpentier, et al,, Estimates of Unconventional Natural

Gas Resources of the Devonian Shale of the Appalachian Basin, ”

U.S. Geological Survey Open-File Report 82-474 (preliminary), 1982,




tion potential of the play. The volume of gas foreach of the 19 different areas was calculated usingEquation 1, below:

The parameters used in the equation are ex-plained in table 44. The volume of gas for each

area was summed to obtain a total basin estimate.

The analysis explicitly recognizes two severedata problems:

q limited quantity of data for most of theassessed area, and

q large sampling errors and differences in inter-pretation.

Because of the limited quantity and quality ofdata, a range rather than a point estimate of the

gas present was developed. The Monte Carlomethod of estimation was employed to acquire

the range.14

———lqln using a Monte Carlo method, appropriate variables in the

equation are specified by a probability distribution rather than by

a point estimate. Then, the equation is “solved” for the depen-

dent variable–in this case, gas-in-place–a large number of times

by randomly sampling the probability distributions. In this way, aprobability distribution is obtained for the dependent variable (gas-

in-place). The “solution” to the equation can either be expressed

by the probability distribution itself, by its mean or median, or per-

haps by a range defined by some probability that the actual value

is within its borders (e. g., “there is an 80 percent probability that

the correct value is in the range X to Y“).

Table 44.—Equation Parameters

Symbol Meaning

G . gas-in-place0 rnacro “ : : . average macrofracture porosity as a fraction

of total volumeTH s . . . . . . . average thickness of organic-rich shalesPr . . . . . . . . . average reservoir pressure (psi)P s . . . . . . . . standard pressure (14.73 psi)Tr . . . . . . . . . average reservoir temperature (oR)T s . . . . . . . . standard temperature (5200R)z . gas deviation factor (0.9)Area” ; ; ; ; . area (square miles)0 micro . . . . average content of microporosity gas at

standard temperature and pressure as a

fraction of rock volumeSOR . . . . . . average volume ratio of adsorbed gas toinorganic content

ORG . . . . . . average organic content as fraction of rockvolume

SOURCE U S. Geological Survey, “Estimates of the Unconventional Natural GasResources of the Devon Ian Shale of the Appalachian Basin, ” 1982

The data used were acquired from a variety ofsources. The configuration of the gas shales wastaken from geologic cross sections, isopachs(thickness maps), and other geological maps com-piled by the USGS. Maps were also used to esti-mate thickness, organic content and average

depths, which when combined with temperatureand pressure gradients yielded average reservoirpressures and temperatures. Micropore gas esti-mates were achieved by plotting gas content—acquired from off-gassing data from canned coresamples—against the amount of organic matterin the sample. The slope of the resultant curverepresents the ratio of adsorbed gas to organicmatter. The intercept (gas content at the pointwhere organic matter is zero) represents themicropore gas content.

The results of the USGS study are compiled intable 45. The 95th fractile (F 95is a low estimate

and signifies that there is a 95 percent chance thatthere is more than 577.1 TCF present. The 5th

Table 45.—Estimates of In-Place Natural GasResources in the Devonial Shale of

the Appalachian Basin

Natural gas resources(trillions of cubic feet)

Low HighPlay F 95

a F 5

a Mean

1. North-Central Ohio . . . . . . . 17.9 34.22. Western Lake Erie . . . . . . . 21.7 31.33. Eastern Lake Erie . . . . . . . . 2.1 3.34. Plateau Ohio , . . . . . . . . . . . 44.4 76.25. Eastern Ohio . . . . . . . . . . . . 35.2 55.16. Western Penn-York. . . . . . . 20.4 28.27. Southern Ohio Valley. . . . . 19.7 36.28. Western Rome Trough. . . . 38.0 74.09. Tug Fork . . . . . . . . . . . . . . . 13.7 25.9

10, Pine Mountain . . . . . . . . . . . 10.7 18.711. Plateau Virginia . . . . . . . . . 3.9 10.212, Pittsburgh Basin . . . . . . . . . 76.8 129.913. Eastern Rome Trough . . . . 70.7 132.514. New River . . . . . . . . . . . . . . 38.5 91.715. Portage Escarpment . . . . . 8.5 21.316. Cattaraugus Valley . . . . . . . 10.4 23.217. Penn-York Plateau . . . . . . . 98.1 195.218. Western Susquehanna. . . . 24.1 67.719. Catskill . . . . . . . . . . . . . . . . . 22.1 75.8

25,926.5

2.7

59.944.724.327.756.019.714.6

7.1102.1100.363.114.616.6

146.044.947.6

Entire basin . . . . . . . . . . . 577.1 1,130.9844.2NOTE: All tabulated values were rounded from original numbers. Therefore, totals

may not be precisely additive.aF

95denotes the 95th fractile; the probability of more than the amount F

95is 95

percent. F5is defined similarly. Because of dependency between plays, these

fractiles (unlike those in many other studies) are additive.

SOURCE U.S. Geological Survey, “Estimates of the Unconventional Natural GasResources of the Devonian Shale of the Appalachian Basin, ” 1982




fractile (F5 is a high estimate and indicates that

there is only a 5 percent chance of there beingmore than 1,130.8 TCF present.15 The mean esti-mate is 844.2 TCF. Although the USGS did not

estimate recoverability, it compiled a map illus-trating shale gas potential (fig. 41) which qualita-

tively indicates the potential for recovery in eacharea based on gas-in-place and the presence ofnatural fracture systems.

The Mound Facility16

The Mound study incorporates extensive geo-chemical data into its volumetric analysis of the

1 Sconseq Uently, there is a 90 percent probabi Iity that the gas-in-place is within the range 577.1 to 1,130.8 TCF.IfJR, E. Zielinski and R. D. Mclver, “Resource and Exploration

Assessment of Oil and Gas Potential in the Devonian Gas Shalesof the Appalachian Basin, ” 1982.

gas-in-place in the Appalachian Basin. Organicgeochemical analyses were performed on over2,000 individual core samples and on an addi-tional several hundred well cuttings to evaluatethe quality of the shale units as sources of natu-ral gas and other hydrocarbons. In particular, the

analyses focused on three primary determinantsof gas content: the quantity of organic carbonpresent, its origin (i.e., the nature of the organicmaterial that provided the carbon, e.g., spores,pollen, herbaceous plants, algae, etc.), and itsthermal maturity .17

1 Thermal maturity is the extent to which the organic matter In

the rocks has been “cracked” by heat. Cracking is the process by

which long hydrocarbon chains are broken to form simpler

molecules such as those comprising methane gas.

Figure 41 .—Shale Gas Potential in the Appalachian Basin

SOURCE USGS, “Estimates of Unconventional Natural Gas Resources of the Devonian Shale of the Appalachian Basin, ” 1982

3 8 - 7 4 2 0 - 8 5 - 1 3




Several general conclusions were formulated

about the resource potential of the AppalachianBasin. The most important conclusion is that theDevonian shales are exceptionally rich source

rocks; were it not for their low porosity andpermeability, the shales would represent “oneof the greatest oil- and gas-producing provinces

of the world.”18B Other conclusions point to thenonuniformity of the resource and establishwhere the generally rich nature of the sourcerocks may not apply. For example, the quantity

of organic material in the basin decreased to theeast. The organic-rich rocks in the extreme north-western and western portion of the basin hadhigh potential for hydrocarbon development, butthey were only slightly thermally altered andnever reached their full gas-bearing potential. Inthe deeper portions of the basin, too much heatwas generated, thereby lessening the potentialfor finding hydrocarbons.

The gas-in-place analysis differed from the other

estimates in the way in which the gas content wasdetermined. Mound felt that a large portion ofthe gas escaped during the process of obtainingthe core and before the core could be sealed inthe gas-tight canister, To improve the accuracyof the measurement, Mound developed a con-trolled off-gassing experiment where the rate of

gas release from the core is measured. Data fromthese experiments were used in combination withother geologic and geochemical characteristicsof the formations to develop equations which

determined the “indigenous,” or original, gascontents of the rock. The methodology wasverified by the use of a pressurized core barrel,which extracts the core under in-situ pressure,thereby limiting premature gas release. Moundthus concluded that the revised values more ac-curately reflected the quantity of gas originallypresent in the rock.

The gas-in-place estimates were determined foreach of 17 separate stratigraphic intervals, or

lall. E. i!ielirlski and R. D. Mclver, ‘ ‘Syn thes is o f o rgan ic

Geochemical Data From the Eastern Gas Shales,” SPE/DOE 10793,

in Proceedings of the Unconventional Gas Recovery Symposium,May 1618, 1982, Pittsburgh, PA.

units, in the Appalachian Basin by first combin-ing the indigenous gas contents (MCF/acre-foot)with the thicknesses of the gas-bearing shale toprovide the areal distribution of the total gas ineach unit. The data were contoured, as illustratedin figure 42. Next, the acreage contained withineach contour area was integrated, multiplied by

the appropriate gas areal density (MCF/acre), andsummed to yield the total gas-in-place for theunit. This methodology resulted in a gas-in-placeestimate for the 17 units composing the Appa-lachian Basin of 2,579 TCF.

Estimate Comparison and Uncertainties

The methodologies used by NPC, USGS, andMound to determine the gas-in-place are allvolumetric estimates based on multiplying gascontent, shale thickness, and areal extent, butthey differ substantially in their computation

methods and input data. The major difference ap-pears to be in the computation of gas content.NPC used a basin-wide average based on off-gassing data available at the time. Both theMound and USGS analyses use a more sophisti-cated, disaggregate approach, with USGS cal-culating separately the macrofracture, micro-porosity, and adsorbed (bound) gas for 19 areasin the Appalachian Basin, and Mound determin-ing gas contents for 17 stratigraphic units in thebasin using equations based on geochemicalanalysis, and contouring and integrating theresults across the basin. Both the USGS and

Mound estimates had access to new off-gassingdata developed by the Eastern Gas Shales Proj-ect. The Mound approach is the most optimisticof the three because it incorporates a calculationof gas lost in obtaining and measuring the coresamples, a factor not considered by the NPC andapparently not considered significant by theUSGS.

In OTA’s judgment, the physical evidence citedand calculations made in the Mound report ap-pear plausible; the Mound estimate of 2,579 TCFgas-in-place (1,440 TCF in “drillable” areas)

seems a reasonable estimate given the availabledata.




Figure 42.— Distribution of Devonian Shale Gas

[

MCF Gas/Acre

100 )

I .“

q TENN-9o

-

0

Stratigraphic interval: early marcellus time

SOURCE: Zielinskl and Mclver, 1982,

TECHNOLOGY

Fracturing

Because of the extremely low permeability ofthe shales, production of Devonian shale gasdepends on exploiting the natural fracture net-

work in the rock and on enhancing gas flow byartificially stimulating the well. Over 90 percent

of all Devonian wells require stimulation in or-der to yield gas in commercial quantities, andeven stimulated wells will not be successful unless

there is already a well-developed natural fracturenetwork.

For most of the Devonian shale’s productionhistory, wells were stimulated by filling large por-

tions of the wellbore with explosives and allow-ing the detonation to shatter the rock surround-ing the well. This basic method was first used in1865 and is still in use. It generally is consideredless of a “fracturing” method than simply as a




means to overcome the formation damage causedby drilling in the shales. (Its primary structural ef-fect is to fragment the rock immediately surround-ing the well bore.)

As in the tight sands, hydraulic fracturing is be-coming increasingly used today in the shales. The

fractures being created in the shale formationsare small, however, not the massive 1,000-ft frac-tures becoming more popular in the Westerntight sands. They also generally carry less prop-pant than in the sands. The shales are extremelysensitive to formation damage during fracturing,especially because of the presence of water-sensi-tive clays in the shale that can be dislodged bythe fracturing fluids and plug pores and fractures.Although formation damage is a problem withtight sands and coal seams also, Devonian shalesmay be the most sensitive. Devonian shales have,

as a consequence, served as a testing ground for

a number of new fracturing fluids.

Many of the fluids developed to minimize for-mation damage are foamed, using a gas phase

to reduce the amount of water required. Foamedfluids are gas-in-water emulsions, where the sur-face tension of the bubbles holds the proppant–

particles that become wedged in the fractures andhold them open–in suspension. Nitrogen (N 2

is the most common gas used. Properties thatmake foam suitable for the Devonian shale in-clude low volumes of water, high efficiency increating fractures, high proppant-carrying capac-

ity, low friction during pumping, and sufficientenergy within the gas phase to allow recovery ofmost of the fluid without pumping. By 1980-81,foam technology had advanced so rapidly thatit dominated fracturing in the Devonian shales.

Pure nitrogen has also been used as a fractur-ing fluid. It is not an efficient fracture fluid and

can only be used to fracture small depth inter-vals of 10 ft or s0.19 In addition, it is not an effec-tive carrier of proppants, and consequently is ef-fective only at shallow depths where the fracturesare less likely to close. However, nitrogen does

not adversely affect the formation and it has prov-en very effective in increasing gas flow. As a re-

sult, nitrogen fracturing quickly became the pre-ferred method for many production situations.Because of the newness of this type of treatmentand its relative lack of propping effectiveness, the

ability of nitrogen fracturing to maintain produc-tion levels over the long-term is uncertain.

Current trends in fracturing technology are cen-tered on the developing of fluids that are bothnondamaging, like nitrogen, and can carry prop-pants more effectively. possibilities include:

q

q

q

q

Stabilized foam.–Although similar to theoriginal foams, water has been reduced from25 to 10 percent, and proppant-carrying ca-pability is enhanced by a gelling agent thatstiffens the foam. This is a high cost methodthat has been used only sparingly.Liquid carbon dioxide.–This method has theability to transport proppants. As the liquid

C02 warms, it reverts to the gas phase andeasily flows back out of the hole with mini-mal damage to the formation. Liquid CO2

fracturing is a relatively expensive process

and somewhat more dangerous to use thanfoamed fluids. In addition, the casing andpumping materials must be capable ofwithstanding very low temperatures.Shale oil.—This method combines a nitro-gen-driven fracture with the subsequent injection of shale oil—obtained from previousdrilling–as a proppant-carrying agent to pre-vent fracture closure.

Water-based nonreactive solvents.–Thesesolvents can be used either after a fractureto clean the formation or as the fluid baseof a stabilized foam fracturing treatment. Thissystem is still experimental.

Finally, producers have attempted the use ofradically improved versions of the original ex-plosive fracturing used since the 1800s. In tailoredpulse loading, a propellant charge is ignited topressurize the wellbore at a much slower ratethan is achieved with conventional explosives.

20

The loading rate, or rate at which the energy

stored in the propellant is released, can be con-trolled to create different types of fractures. For

lqFor example, a nitrogen fracture might affect only the reser-

voir rock between 900 and 910 ft in depth.

20J. W. Crafton, “Fracturing Technologies for Gas Recovery From

Tight Sands, ” OTA contractor report, September 1983.




example, at intermediate loading rates, multiplefractures form radially around the well bore.21 Atslow rates, fractures form in an analogous man-ner to hydraulic fractures, directionally controlledby the regional stress field.

This technique has only been used on a small

scale for prefracturing tests.22 However, it isthought to have significant potential for use ona larger scale, especially because it causes littleformation damage. Commercial application in theDevonian shales may occur in the near future.

The record of success of well stimulations inDevonian shales is mixed. Although new stimula-tion technologies have increased gas productionfrom a number of wells, many have not bene-fited from stimulation and the specific reasons for

their lack of success are not well understood.

One problem is the local variability and unpre-dictability of fractured zones. Wells offset shortdistances from producing wells may not intersecta productive fracture system.23 Improved tech-nologies or exploration strategies to locate andcharacterize fracture systems are critical to eco-nomic development of Devonian shales.

Another problem is the difficulty in extrapolat-ing successful stimulation techniques from onesite to another. Although some producers havebeen quick to try the newest in technologies, noone has yet established criteria for choosing tai-lored pulse loading over nitrogen or perhaps liq-

uid C02 injection. Most stimulations appear tobe conducted on a trial-and-error basis and in-adequate records are kept to determine the rea-sons for success or failure of a particulartechnique.

A third problem in successful Devonian shalegas production is accurately determining the payinterval. Because many wells do not have signif-icant gas shows prior to stimulation, it is difficult.———zlThe radial fractures possible with tailored pulse loading may

have special potential in areas where the natural fracture systemsin the shale display strong directional tendencies. Radial fractures

may outperform long hydraulic fractures when these tendencies(permeability anlsotropies) are substantial (Kuuskraa and Wicks,

1984, op. cit.).Zzjohnston & Associates, inc., “The Status and Future of Produc-

tion Technologies for Gas Recovery From Devonian Shales, ” OTA

contractor report, 1983.231 bid,

to determine the interval within the shale se-quence which is most likely to contain recover-able gas. Consequently, fractures may not beproperly located to optimize production.

Finally, no technology now exists or is beingconsidered to produce gas from those portions

of the Devonian shales where the natural frac-ture system is not well developed. This severelylimits the overall production potential of the re-

source.

Deviated and Directional Drilling

The only other technology that currently hasany potential for increasing production in un-conventional reservoirs is one that allows drillingwells that either intersect more of the reservoir

rock or intersect more of the natural fracture sys-tem. Thus, if reservoirs lie in an essentially

horizontal plane and natural fracture systems ina more or less vertical plane, a well drilled atsome angle from the vertical would intersectmore gas-productive natural fractures.

Directional drilling has frequently been sug-gested as a technology applicable to Devonianshale gas production .24 Production requires in-tersection of natural fractures and most of thesefracture systems are vertical. The major drawbackis the problem of formation damage. A drill bitdrilling at an angle from the vertical encountersincreased frictional resistance and, if drilling in

a fractured formation, runs a greater risk of hav-ing the wellbore collapse. Drilling muds areneeded to reduce friction and hold the holeopen; however, drilling muds may cause consid-erable formation damage. One experimental de-viated well has been drilled in the Devonian

shale, in Meigs County, Ohio, but its intent wasmore to determine the natural fracture spacingthan to test a new production technique.

25

Exploration

Unlike the sophisticated exploration techniques

used in frontier areas such as the Western Over-

.2qOffice of Tech rlology Assessment, ‘‘Status Report of the Gas

Potential From Devonian Shales of the Appalachian Basin, ” 1977.Z> Charles Komar, Morgantown Energy Technology center, per-

sonal communlcatlon, 1984.



190. U.S. Natural Gas Availability: Gas Supply Through the Year 2000

thrust Belt and offshore, the “exploration tech-niques” used for locating Devonian shale wellsare often little more than near-random selectionbased on the availability of land. The failure touse sophisticated exploration technology reflects

a number of factors. First, it is difficult to build

an exploration block of any size in the Appalach-ian shale basin because of the diversity of landownership. Second, leasing problems—e.g., thelack of well spacing requirements in some States

in the shale area—aggravate the problem becausewells on adjacent properties can get as close to

a successful well as the property line allows.There is little incentive to invest in expensive

seismic surveys if the costs cannot be recapturedby exclusive development of the surveyed area.Third, existing technology is not fully effective inlocating the subsurface features whose under-standing is critical to drilling success, and the

steep terrain drives up the cost of techniques suchas reflection seismology.

Aerial and satellite imagery may prove usefulin Devonian shale exploration because they canidentify Iineaments—characteristic topographicfeatures–which may be related to fault and frac-ture zones and may contain information on re-gional stress patterns. Concentrations of surfacefractures may indicate the presence of subsurfacefracture networks that could serve as potentialreservoirs. 26 Skeptics feel that surface expression————

Z6R. D. MCIVer, j. R. Kyle, and R. E. Zielinski, “Location Of Drilling

Sites in the Devonian Shales by Aerial Photography, ” SPE/DOE10794, SPE/DOE Unconventional Gas Recovery Symposium, pp.

51-53.

of fractures does not accurately reflect subsur-face conditions (i.e., fractures may curve at depth

or may not extend to the gas-bearing rocks). Tosupport their position, they cite failures of wellsoffset from producing wells along a surface linea-

ment.27

Subsurface mapping of variations in the rockformations is also important in the shale region.

A key mapping tool for the shales is well loggingusing a combination of gamma-ray and neutron-density logs. Induction logs, spectral logs, ther-mal decay logs, noise logs, and temperature logsare often run in combination with the gamma-ray and neutron-density logs. Some operators alsorely on a combination of gamma-ray, bulk-den-

sity, and resistivity logs. (For a brief descriptionof the various types of well logs, see box B-1 inapp. B.) Unfortunately, current well logging tech-niques are not adequate to detect open fracturesthat do not actually intersect the boreholes, sothey are only of limited use in mapping fracturepatterns. Also, some States in the shale basins donot require full disclosure of well log data, andthis further limits the ability to construct usefulsubsurface maps from the existing data.

~johnston & Associates, inc., oP. cit.

RECOVERABLE RESOURCES AND PRODUCTION POTENTIAL

The Devonian shale gas-in-place resource in

the Appalachian Basin has been estimated at be-tween 225 TCF (the NPC low estimate) and 2,579TCF (the Mound estimate). As discussed above,the higher end of the range appears most credi-

ble on the basis of existing data. However, muchof the gas-in-place is unlikely to contribute tofuture gas supply, for the most part because geo-logical conditions make the gas extremely diffi-

cult to produce. The quantity of gas likely to beproduced is a function of the price of the gas,

the available technology, the associated costs ofproduction, and a variety of other factors, suchas institutional barriers, that will influence deci-

sionmaking on production. Several organizationshave attempted to estimate the size of the recov-erable resource. Reports have been issued whichdescribe the resource and designate the most




favorable areas for production.28 Production sce-narios have also been established by further

assuming drilling and development schedules. Ingeneral, the estimates of recoverable resourcesand future production are based on extrapola-tion of past production, for which there is a siz-

able amount of data due to the long productionhistory of the shales. However, the productiondata are limited in important ways. Available pro-duction histories are for the most part limited towells using traditional production technology,

that is, “shooting” with explosives at wide wellspacing. Also, these histories are affected by a va-riety of factors aside from the nature of the gasresource. These factors include differences inmarket conditions, well operating practices, pro-duction techniques, the use of workover treat-ments, and pipeline pressures.29 Extrapolation ofproduction data therefore should account for

these variables, yet the lack of data and the com-plexity of the necessary analysis makes such anaccounting quite difficult. None of the existing

ZBFavorable areas are designated by State in DOE/METC reports

118-124 and on a play basis in the USGS report entitled, “Estimates

of Unconventional Natural Gas Resources of Devonian Shale of

the Appalachian Basin, ” 1982.Z9R. E. Z i e l l n s k i a n d R. D. Mclver, Resource an d hph?t ion

Assessment of the Oil and Gas Potential in the Devonian Gas Shales

of the Appalachian Basin, U.S. Department of Energy, Morgantown

Energy Technology Center Report DOE/DP/0053-l 125.

studies of recoverable resources have attemptedsuch an accounting. Also, it is not clear to what

extent the drilling represents a true unbiased sam-ple of what might occur on undrilled acreage ifthe same production techniques were used. Theseproblems with the available data are discussed

later.


As illustrated in table 46, several estimates ofrecoverable resources have been made. Theseinclude early estimates by the Office of Technol-ogy Assessment (OTA), Lewin & Associates, andthe National Petroleum Council (NPC). Each esti-mator assumed economic and technologic pa-rameters to establish estimates of recoverable re-sources in the Appalachian Basin. Both Lewin &Associates and the NPC extended their analysis

to include annual production estimates orscenarios.

More recently, Pulle and Seskus of Science Ap-plications, Inc. (SAI), Zielinski and Mclver ofMound, and Lewin & Associates also estimatedthe recoverable resource in the AppalachianBasin. Pulle and Seskus

30used past production

30C$ v. Pul[e and A. P. Sesktjs, “Quantitative Analysis of the Eco-

nomically Recoverable Resource, ” U.S. DOE-METC, 1981.

Table 46.—Devonian Shale Recoverable Resource Estimates (TCF): Appalachian Basin

Organization Year Estimate ConditionsOffice of Technology Assessment. . . . . 1977 15-25

23-38

Lewin & Associates . . . . . . . . . . . . . . . . . 1978-79 2-1o4-25

After 15 to 20 yearsAfter 30 to 50 yearsAt $2 to $3/MCF (1976), current technology

(borehole shooting or hydrofracturing),150-acre spacing

Base caseAdvanced case for prices between $1.75 to

$4.50traditional advanced

National Petroleum Council . . . . . . . . . . 1980 3.3 - 38.9

15.3 - 49.9Pulle and Seskus (SAI). . . . . . . . . . . . . . . 1981 17-23Zielinski and Mclver (Mound) . . . . . . . . . 1982 30-50

Lewin & Associates . . . . . . . . . . . . . . . . . 1983 6.2-22.5

Lewin & Associates . . . . . . . . . . . . . . . . . 1984 19-44

For price levels between $2.50 to $9, 160-acrespacing

Technically producible“Shot” wells, 160-acre spacingFor States of West Virginia, Ohio, and Kentucky

only, “shot” wells, 160-acre spacing

Technically recoverable, for most promisingformations in Ohio. Maximum represents80-acre spacing, advanced technology

Technically recoverable, for most promisingformations in West Virginia. Preliminaryvalues

SOURCE: ???




data and Delphi estimation to compute a meanvalue of 20.2 TCF for the recoverable Ap-palachian resource using explosive fracturing at160-acre spacing. Zielinski and Mclver31 utilizedSAI data to estimate the recoverable resource,also based on explosive fracturing. They felt that

sufficient data were not available to make areliable estimate, but produced a preliminary esti-mate of 30 to 50 TCF for the minimum recover-able gas in West Virginia, Ohio, and Kentucky.The most recent estimation effort was performedby Lewin & Associates under contract to DOE’sMorgantown Energy Technology Center (METC)and was an estimate of the technically recover-able reserves in the most favorable Devonian for-mations in Ohio and West Virginia (the estimatesfor West Virginia were published only in draftform at the time of this report).32 To OTA’s knowl-edge, this estimate is the only one currently

available to use reservoir simulation. Using thissimulation capability, the analysis explores theramifications of alternative fracture technologies,well spacing, and well patterns on the size of the

recoverable resource.

Office of Technology Assessment33

The OTA report was published in 1977 and wasthe first study that attempted to evaluate therecoverable Devonian shale gas resource in theAppalachian Basin.

OTA established production estimates based

on 15 to 20 years of production data from 490wells in three productive areas of the AppalachianBasin. Wells from Cottageville and Clendenin,WV, and Perry County, KY, were grouped ac-cording to the quantity (high, medium, or low)of gas produced. Average production rates werecalculated for both shot and fractured wells ineach group and used to calculate the recoverableresource assuming a productive area of 16,300

..—3’R. E. Zielinski and R. D. Mclver, “Resource and Exploration

Assessment of the Oil and Gas Potential in the Devonian Gas Shales

of the Appalachian Basin, ” 1982.

32 v . A . l ( u u s k r a a , e t a l . , Technically Recoverable Devonian shaleGas in Ohio, Lewin & Associates Report for Morgantown Energy

Technology Center, July 1983; and V, A. Kuuskraa, et al., Tech-

nica//y Recoverable Devonian Shale Gas in West Virginia, Summary,

1984 (draft).‘~office of Technology Assessment, “Status Report of the Gas

Potential From Devonian Shales of the Appalachian Basin, ” 1977.

square miles— 10 percent of the entire basin areaof 163,000 square miles. A well spacing of 150acres was assumed, yielding approximately69,000 wells. The economics were determinedusing an after tax net present value (ATNPV)model, a discount rate of 10 percent, and a well-

head price for gas in the $2 to $3/MCF range(1976$).

The findings as reported by OTA are summa-rized below:

q

q

q

The Devonian shale resource could be pro-duced without developing new productionequipment and techniques.The Brown shaless34 (as they were called by

OTA) could yield between 15 and 20 TCFduring the first Is to 20 years of production.After 30 to 50 years, cumulative productioncould reach 23 to 38 TCF.

Because of the tremendous drilling effort andthe time required to develop the-necessarypipeline infrastructure, as many as 20 yearsmay be required before annual production

reached 1 TCF.

A critical factor in OTA’s analysis is the assump-

tion that only 10 percent of the Appalachian Basinwill prove to contain gas recoverable at the as-sumed price using conventional technology. Thisassumption is based on the general argument thatpast drilling has not been random, but insteadhas been skewed to the high-quality areas–a uni-

versal tendency in resource development—andalso upon observations of the clustered natureof existing development, the considerable depthsand/or thinness of the shales in some undevel-oped portions of the basin, the poorly developedfracture systems in other undeveloped areas, andthe lack of success of drilling in some of theseareas. This assumption that much of the unde-veloped acreage in the basin will not prove tobe productive is undoubtedly correct qualitative-

ly, but there appears little quantitative basis forthe choice of 10 percent as the productive frac-tion; it is essentially an educated guess.

‘QBrOWn Shales are generally younger than black shales and have

more hydrocarbons in the organic material. The organics in the

black shales are closer to elemental carbon. (V. Kuuskraa, 1982,

“Unconventional Natural Gas,” in Advances in Energy Systems and

Technology, vol. 3.) Brown and black shales are commonly referred

to jointly as black shales.




Lewin & Associates I35

Lewin & Associates addressed the gas poten-tial of Devonian shales in a series of 1978-79reports entitled “Enhanced Recovery of Uncon-ventional Gas. ” The purpose of the Devonian

shale portion of the study was to estimate the eco-nomic potential of the resource, based on em-pirical data such as geology, reservoir perform-ance, and costs.

The Lewin & Associates study evaluated gas re-

covery potential with respect to price for baseand advanced cases. The parameters assumed foreach case are listed in table 47. Economic anal-yses were performed to determine the economi-cally recoverable resource at $1.75, $3.00, and$4.50/MCF (1977$) for both the base and ad-

vanced cases. The base case represents the de-velopment of areas with producing characteris-

tics similar to areas under production today, usingavailable small-scale hydraulic fracturing tech-niques with design fracture lengths of 100 to 200ft. The advanced case added a mix of strategies

to the base case to increase production andenlarge the size of the recoverable resource, in-cluding:

jJLeWin & Associates, Inc., Enhanced Recovery of UnCOn VentiOna\

Gas, U.S. Department of Energy reports HCP/T 2705-01, 02, and

03, 1978-79.

q

q

q

extension drilling into deep shales, with im-proved stimulation (in eastern West Virginiaand Pennsylvania);dual completion, i.e., stimulating two sepa-rate gas-bearing intervals from one well, withimproved stimulation (in Ohio), to allow eco-

nomical production of marginal prospects;andimproved recovery through advanced stim-ulation technologies and closer well spacing(in eastern Kentucky and western West Vir-ginia, the center of current production).

The evaluation was confined to the Appalach-

ian Basin. Of the entire basin area of 210,000square miles, only 62,000 square miles were con-sidered as potential shale gas-bearing areas. The62,000 square mile area was divided into 12 ana-lytical areas based on similar geologic character-

istics, drilling or production histories. Approx-imately 5,000 square miles of that area includedalready proven areas or sites of previous produc-tion, leaving 57,000 square miles as probable andpossible gas-bearing areas.

Actual production data from several gas com-

panies and 250 individual wells were collectedand cumulative production decline curves estab-lished for each area. The production curves wereadjusted for “play out, ” to compensate for thefact that fields tend to produce less as drilling

Table 47.—Summary of Major Differences Between Lewin Base and Advanced Cases in Devonian Shale Analysis

Strategy item Base case Advanced case

Source characterization: Eligible areas . . . . . . . . . . . . . . . . . . . . . . . Probable areas Probable and possible areasDry hole rates . . . . . . , . . . . . . . . . . . . . . . . 20% 10%

Technology: Completions . . . . . . . . . . . . . . . . . . . . . . . . Single Dual where a low producer is underlain by

other productive payRecovery efficiency per unit area . . . . . . Current levels Improved by 20 percent in higher

producing areasEconomics:

Risk—reflected in discount ratesaof . . . 21 ‘/0 160/0

Development: Start year for drilling

Probable area . . . . . . . . . . . . . . . . . . . . . 1978 1981 (R&D effect begins)Possible area . . . . . . . . . . . . . . . . . . . . . 1987 1987

Development paceProbable area . . . . . . . . . . . . . . . . . . . . . 17 years to completion 13 years to completionPossible area . . . . . . . . . . . . . . . . . . . . . 17 years to completion 15 years to completion

aDiscount rates include a constant ROR based on 10 to 150/0 and an inflation adjustment of 6°/0

SOURCE: Lewin & Associates.




moves into extension areas, and for stimulationtechnology improvement, to compensate for thedifferences between the old explosive fracturingand hydraulic fracturing. The adjusted curveswere then used to estimate 30-year cumulativerecovery per well for each area. These estimates

were then used in the analysis of the economicpotential.

The economic analysis used a discounted cashflow model. Net cash flow was calculated by sub-

tracting investment costs, operating costs, and allother allocated costs from the cash flow acquiredfrom production revenues. The net cash flow forthe 30 years of production for each well consid-ered in the study was discounted to arrive at thenet present value. The areas with a positive netpresent value were assumed to be developed inaccordance with the timing schedule designated

for each case.

The results of the economic evaluation aresummarized in table 48. The base case estimatesare quite pessimistic—at $4.50/MCF (1 977$), or

$7/MCF (1 983$), a very high price in today’s mar-ket, total recoverable resources are only 10.5TCF. The somewhat more optimistic advanced

case, which reaches 18 to 25 TCF at the sameprice, obtains most of its added recovery from

the deep drilling and dual completions, with im-proved recovery in existing producing areasyielding only 2.1 TCF at this price.

An important consideration in this analysis isthat Lewin considered there to be little differencein per well recovery efficiency between the baseand advanced case, despite the more effectivefracturing attainable in the advanced case. Themajor difference between the two cases is the

Table 48.- Lewin & Associates: Results of EconomicAnalysis, Summary Table

Economically recoverable

Price, 1977$ (1983$) Base case Advanced case

$1.75/MCF ($2.75/MCF) . . . . 2 TCF 5 TCF$3.00/McF ($4.70/MCF) . . . . 8 TCF 16 TCF

$4.50/MCF ($7.00/MCF) . . . . 10.5 TCF 18 to 25 TCFb

aCurrent proved reserves are 1 TCF and the following estimates represent addi-tions to reserves.

bThe range reflects the geologic uncertainty with regard to natural fracture in-

tensity.

SOURCE: Lewin & Associates.

more rapid drainage attainable with the improvedstimulation technology, which greatly improvesthe economics of recovery and moves marginalareas into the “economically recoverable” range.The source of this interpretation is the belief atthe time that the primary source of producible

gas is the fracture porosity.36 This was thoughtto imply that the recovery efficiency of even bore-hole shooting would be quite high, in the neigh-borhood of so percent, with little improvementobtainable from more effective fractures. It cur-rently is believed, however, that much of thelong-term well production is from the resorptionof gas bound to the shale matrix, and that theactual recovery efficiency of borehole shootingis only a few percent. Lewin’s new work, de-scribed later in this section, folds this new under-standing of the source of recoverable Devonianshale gas into its analysis (see the discussion of“Lewin & Associates 11”). An important implica-tion of this understanding is that improved frac-turing should increase ultimate recovery, not justaccelerate production.

National Petroleum Counci137

The NPC also estimated the quantity of pro-

ducible gas in the Appalachian Basin for differentlevels of technology and price. Three levels oftechnology were considered in the analysis: tradi-tional (borehole shooting), conventional (conven-tional hydraulic fracturing), and advanced (unique

fracturing techniques and deviated drilling). Ad-vanced technology was assumed to double theproduction increase achievable from the use of

conventional technology.38 Conventional tech-nology was assumed to increase production over

traditional borehole shooting by O to 57 percentdepending on the open flow rates of the wells. 39

MI n other words, it was thought that most of the recoverable gas

was free gas stored in the natural fracture systems.jzNational petroleum COU nci 1, Unconventional Gas SOUrCt?S—

Devonian Shales, 1980.JaThis assumption was based on experiments performed i n

Kanawha County, WV. Three advanced technology wells had pro-

duction increases of 230 percent over wells stimulated by tradi-

tional shooting. Conventionally fractured wells showed 80 percentincreases in production over traditionally shot wells.

39 T he a d v a n t a g e s d fracturing over borehole shooting decline

as the unstimulated flow rate increases; with flow rates above 300MCF/D, fracturing was assumed to be no better than shooting.




The quantities of potential reserves for price levelsbetween $2.50 and $9.00 (1979$) were estimatedby performing discounted cash flow analysis for10, 15, and 20 percent after tax rates of return.

NPC’s initial objective was to utilize existingproduction data to predictor extrapolate produc-

tion in undeveloped areas. They intended tomodel the average well production decline ofeach county to the hyperbolic equation below:

where C l, C2, and C3, were empirically derived

constants for that county. When the existing pro-duction data were fit to the equation, they dis-covered that all the decline curves, regardless ofcounty, could be represented adequately with theuse of 3 and 2.5 for C2 and C3, respectively.Therefore, apparently Cl can serve as an indexto characterize average production decline for

each county. The relation of C l to cumulativeproduction can be determined by integrating thehyperbolic equation over the appropriate time

period. For example, 30-year cumulative produc-tion is equal to 4.43 Cl.

Several parameters, such as the various thick-ness estimates and depth, were evaluated for cor-relation with the Cl values. The thickness of theblack shale as determined by gamma-ray logs wasthe only parameter that correlated to C l; it didso with a linear coefficient of 0.213, that is, theaverage county black shale thickness as deter-

mined by logging can be multiplied by 0.213 toobtain the Cl value for each county. This valuewas used as the basis for the “traditional tech-nology” case.

The results of the economic analysis are tabu-lated in table 49. The potential reserve is that por-tion of recoverable gas that is economically pro-ducible at a given price. The total producible gasis the total cumulative amount of in-place gas that

Table 49.—Summary of Producible Gas Estimates

can be produced over the wells’ 30-year lifetimes,under the specified technological conditions, ir-respective of price.

The major findings presented in the NPC re-

port are listed below:

q

q

q

Average well production can be modelledto a hyperbolic decline curve as representedby the equation below:

C l is an index to characterize average pro-duction decline and is linearly related (lin-ear coefficient of 0,213) to black shale thick-

ness as determined by log data.The total producible gas using conventionaltechnology is 37.4 TCF, which is approx-imately 30 percent of the 125 TCF estimated

gas-in-place in drillable formations assumingthe lower value of shale thickness based on

gamma-ray well log data (see the discussionof the NPC estimates of gas-in-place earlierin this chapter), and about 4 percent of the

1,040 TCF gas-in-place using the visuallydetermined thickness.The average price requirement for produc-tion of 37.4 TCF is $6.75/MMBtu (1979$)at10 percent rate of return. Approximately 15TCF can be produced at prices up to $3.50/ MMBtu.

Pulle and Seskus (SAl)40

This analysis essentially extrapolates productiondata from 1,534 Devonian shale wells (with 10years or more of production history) to full de-velopment of the Appalachian Basin’s shales. Thebasin is divided into 10 subregions based on thethickness of the radioactive (black and brown)

‘C . V. Pu Ile and A. P. Seskus, Science Applications, Inc., Quan-

titative Analysis of the Economically Recoverable Resource,

DOE/MC/08216-l 57, May 1981.

Appalachian Basin (constant 1979 dollars and 10°/0 ROR)

Cumulative potential reserves (TCF) v. price ($/MMBtu) Total producible2.50 3.50 5.00 7.00 9.00 gas (TCF)

Traditional technology . . . . . . . . . . . . 3.3 8.5 11.4 14.9 16.6 25.3Conventional technology . . . . . . . . . . 7.3 14.5 19.5 23.5 27.0 37.4Advanced technology . . . . . . . . . . . . . 11.8 20.1 27.2 32.9 38.9 49.9

SOURCE: National Petroleum Council.




shales, the drilling depth of past production, anda measure of the thermal maturity of the shalecores that have been obtained. To obtain an esti-mate of the recoverable resource, wells are as-sumed to be drilled on 160-acre spacing, usingexplosive shooting for stimulation. Past produc-

tion histories are used to estimate 30-year cumu-lative production for wells in half of the subre-gions; for the other subregions, production isestimated by combining the opinions of four ex-perts in a Delphi procedure. The percentages ofdry holes are estimated using the same Delphiprocedure.

The estimated mean recoverable gas under the160-acre spacing is 20.2 TCF, with an estimated95 percent probability that the total lies between17.06 and 23.34 TCF. However, the “95 percentprobability” is a statistical value based on the

assumption that the distribution implied by the

historical data is a perfect reflection of future pro-duction from new wells. This seemingly high levelof probability should not be treated too seriously.It does not account for errors introduced by theDelphi procedure, by changes over time in well“shooting” techniques, or by the possibility thatpast well locations were not random but insteadrepresent some selection on the basis of relative

prospects for success.

The analysis does not include an evaluation ofthe effect of price on well spacing, so it is not clearwhat gas price corresponds to the estimated 20,2

TCF of recoverable resources. On the other hand,the authors show how total recovery is likely tovary with well spacing; table 50 shows the varia-

tion of recoverable resources with assumed wellspacing. This estimate is based on a theoreticalmodel applied to only four wells, so the resultsshould be treated as very tentative. Also, the esti-mated recovery—and revenue—per well for 10-acre spacing is only one-fifth of the per well

recovery and revenue for 160-acre spacing. Thisimplies that the gas price needed for economicrecovery of the 67,9 TCF resource for 10-acre

spacing will be five times the price needed for

recovery at 160-acre spacing, all other things be-ing equal. A countervailing factor, however, isthat gathering costs are quite high in the Appa-lachian region, and closer well spacing and the

Table 50.—Effect of Spacing on theDevonian Shale Recoverable Resource

—Relative Ratio of Total

Spacing number increase in recoverable(acres) of wells resource resource (TCF)

160 . . . . . . . 1 1.00 20.280 . . . . . . . 2 1.68 34.0

40 . . . . . . . 4 2.44 49.320 . . . . . . . 8 3.10 63.010 . . . . . . . 16 3.36 67.9

SOURCE: C. V. Pulle and A. P. Seskus, Quantitative Analysis of the Economical.

ly Recoverable Resource, US. Department of Energy Report DOE/MC/ 08216-157, May 1981.

resulting higher production levels per unit areawould lower these costs.

Mound Facility (Zielinski and Mclver)41

Zielinski and Mclver of the Monsanto Corp.’sMound Facility have reviewed the NPC and SAIestimates of recoverable resources in the Appa-lachian Basin and, using the SAI data, derived analternative, admittedly preliminary estimate of therecoverable resource in West Virginia, Ohio, andKentucky.

Zielinski and Mclver’s review of the NPC esti-mates noted the following:

1.

2.

3.

——

There are important numerical discrepancies

between well production values reported byNPC as derived from their equations, andvalues actually calculated using theseequations.

The NPC analysis derived a relationship forthe initial production rate of a well by search-ing for correlations only with variables whichhave little to do with production, and pickeda variable (gamma-ray log response) for therelationship only by default. Neither gamma-ray log response nor any of the other varia-bles examined bear any relationship to theorganic matter type or thermal maturation,both critical factors in determining gas po-tential.The NPC found that a single equation couldrepresent well production for the entire Ap-

41 R. E. Zielinskiand R. J.Mclver, Resource and Exploration Assess-

ment of the Oi/ and Gas Potential in the Devonian Gas Shales of

the Appalachian Basin, Mound Facility Report to U.S. Department

of Energy, DOE/DP/0053-l 125, undated.



—


palachian Basin. This may imply that, for theshot well technology used in most of thewells in the production histories, the fairlyuniform porosity and permeability values ofthe shale dominate average gas production.This, in turn, implies to the reviewers that

closer well spacing should significantly im-prove recovery if shooting is the method ofstimuIation.

The review of the SAI estimate noted the fol-lowing:

1. SAI divided the basin into 10 subregions byevaluating four variables—radioactive shalethickness, drilling depth, stress ratio, and ameasure of thermal alteration. Only the lat-ter variable has any documented relation-ship to gas production, and this relationshipis a limited one. Consequently, the extrap-

olation of production data in a subregion toundeveloped portions of that region may notbe valid.

2. SAI assumed that well production in eachsubregion would be distributed Iognormally,but the actual production data in the five

subregions where production data was avail-able did not tend to indicate a Iognormal dis-tribution.

Zielinski and Mclver also were concerned thatneither the NPC nor the SAI estimates accountedfor the effects of external influences–market

forces, operating policy, production practices– on production, but instead implicitly assumed

that past production was dependent only on thephysical nature of the resource.

Based on the above concerns, Zielinski andMclver’s conclusion was that “the two . . . studies

do not form a sound foundation for the esti-. . .mation of recoverable resource. ”

Zielinski and Mclver have developed a prelimi-nary estimate of recoverable resources in Ohio,Kentucky, and West Virginia based on the obser-vation, from SAI production data, that the pro-

duction per unit area of the developed portionsof these States follow a clear pattern, i.e., Ken-tucky’s production tends towards 2.3 X 10 -3

TCF/mi 2, Ohio’s towards 0.7 X 1 0-3 TCF/mi 2,and West Virginia’s towards 1.5 X 10-3 TCF/mi 2.

Extrapolating these values to prospective butundeveloped acreage yields a total recoverableresource of 30 to 50 TCF for the three States, forthe same technology (“shooting”) and well spac-ing (160 acres). This estimate is based on theassumption that different market conditions, pro-

duction practices, etc., did not affect the Stateproduction averages, and also that past well sitingwas random. The authors consider the 30 to 50TCF value to be a minimum for recoverableresources because improved stimulation technol-ogy, closer well spacing, or use of remote sens-ing techniques to improve well siting can indi-vidually or in combination increase total recovery

as well as production rates. For example, Zielinskiand Mclver predict that halving well spacing to80 acres will essentially double the recoverable

resource.

Lewin & Associates II42

A recent estimate by Lewin & Associates makesuse of the extensive data collection and analysiseffort of DOE’s Eastern Gas Shales Project, e.g.,the geochemical analyses of Mound, and com-bines this with reservoir simulation to estimatethe technically recoverable gas resource of theLower and Middle Huron Intervals of the OhioDevonian shale.

The Lower and Middle Huron Intervals repre-sent only a portion of the resource base that po-tentially can be exploited; they contain about 50

TCF of gas-in-place, compared to an estimatedgas-in-place of 390 TCF for Ohio and 2,579 TCFfor the Appalachian Basin .43 The Huron Intervalshave been the traditional targets for past drillingin Ohio, and the great majority of available drill-ing data applies only to these intervals. The re-covery potential of the remaining 340 TCF inOhio is unknown. However, nearly half of thegas-in-place in the entire basin is considered byMound to be undrillable because of surface con-straints such as roads and towns, and thus therecoverable resource for the remainder of Ohio

42 V E K u u 5 k r a a , e t a l , , I -e w ; rl & A s s o c i a t e s , Inc., T e c h n i c a l l y,.

Recoverable Devonian Shale Gas in Ohio, prepared for the Morgan-

town Energy Technology Center, U.S. Department of Energy, July

1983.4JFrom the Mound Study.



%$q U.S. Natural Gas Availability: Gas Supply Through the Year 2000

is unlikely to be as large, in relationship to its gas-in-place, as in the Huron Interval. Nevertheless,

the Lewin estimates should be recognized as rep-resenting only a limited portion of Ohio’s Devo-nian shale gas potential, although the most pro-spective portion.

Selected results of the Lewin analysis are shownin table 51. The results should be interpretedcarefully because they refer to the expected phys-ical results of a specified quantity of drilling andstimulation without regard to economic feasibil-ity. In other words, they are comparable to the“technically recoverable” or “maximum produc-ible” resources of other estimates. The results areparticularly interesting, however, because Lewin’suse of reservoir simulation provides for a morecredible estimate of the effects of improved stimu-lation technologies and smaller well spacing. Asshown in the table, both methods of improving

gas recovery could be extremely successful in theAppalachian Basin. According to the report, 80-acre spacing has already started to supplant themore traditional 160-acre spacing in new drillingin the basin, Accordingly, the 8.7 to 10.5 TCF pro-

jected as the result of improved but relatively con-ventional technology at 160-acre spacing prob-ably represents a pessimistic estimate of actualrecoverable resources assuming high gas prices.This result has interesting implications for thefuture potential of Devonian shale gas in view of

the limited portion of the total Appalachian re-

source represented by this analysis.

A more recent Lewin study, available in draft

at the publication close of this report, estimatesthe technically recoverable Devonian shale re-sources in West Virginia. Table 52 summarizes

the results, which apply to the Huron, Rhine-street, and Marcellus shale intervals. These inter-vals represent the most promising shale prospectsin the State, although only 70 TCF out of a total

of 125 TCF of gas-in-place for the intervals wasactually appraised. Insufficient reservoir datawere available for the nonappraised portions ofthese intervals. In addition, hundreds of TCF ex-ist in lower quality shale formations that may bedevelopable at some point, but not with simpleextensions of today’s technology.

The results of the West Virginia assessment are

even more optimistic than the Ohio results, giventhe estimated 25.4 to 32.7 TCF technically recov-erable resource based on improved but readilyattainable technology and 160-acre spacing.Coupled with the probability that Kentucky willprove to have recoverable resources somewherein-between those of Ohio and West Virginia,44

the Lewin results imply that the Devonian shalerecoverable resource is considerably greater thanimagined by all of the previous estimates re-

viewed herein.

Table 51 .—Results of Lewin Assessment ofTechnically Recoverable Gas in Ohio, by

Stimulation Method After 40 Years

dqvel 10 I (u u 5 k r a a , Lewin & Associates, Inc., personal Comm u n i-

cation, 1984.

1. Present technology, 160-acre spacingqBorehole shooting . . . . . . . . . . . . . . . . . . . . . . .6.2 TCF

Il. Improved but readily attainable technology,160-acre spacing:q Small radial stimulation (30 ft radius). .. ....8.7 TCFq Small vertical fracture (150 ft wings) . .....10.5 TCF

Ill. Advanced technology, speculative,160-acre spacing:q Large radial stimulation (60 ft radius). .. ...10.2 TCF. Large vertical fracture (600 ft wings) . .....15.2 TCF. Large vertical fracture, 80-acre spacing

(600 ft wings) . . . . . . . . . . . . . . . . . . . . . .....21.0 TCFIV. Changed well patterns (3 to 1 rectangle,

taking account of permeability anisotropy)with improved or advanced technology:yields added recovery of 5 to 10 percent/well.

SOURCE: V. A. Kuuskraa, et al., Technically Recoverable Devonian Shale Gas in

Ohio, Lewin & Associates Report for Morgantown Energy TechnologyCenter, July 1983.

Table 52.-Results of Lewin Assessment ofTechnically Recoverable Gas in West Virginia, by

Stimulation Method After 40 Years

1. Present technology, 160-acre spacing:qBorehole shooting . . . . . . . . . . . . . . . . . . . . . .19.1 TCF

Il. Improved but readily attainable technology,160-acre spacing:q Small radial stimulation (30 ft radius). .. ...25.4 TCFq Small vertical fracture (150 ft wings) . .....32.7 TCF

Ill. Advanced technology, speculative:q Large vertical fracture (600 ft wings) . .....88.4 TCF

SOURCE: V. A. Kuuskraa, et al., Technically Recoverable Devonian Shale Gas in

West Virginia, Summary, 1984 (draft).




Estimates Comparison and Uncertainties

OTA’s examination of the several estimates ofDevonian shale recoverable resources estab-lished some important concerns:

First, with the exception of the latest Lewin &Associates reports on Ohio and West Virginia, theattributes of advanced recovery technologieswere described vaguely, and a reservoir simula-

tion model that could predict the effect of longerfractures or other attributes of advanced technol-ogies was not available. Consequently, estimatesof the effects of advanced recovery technologieson recoverable resources should be consideredat best either “educated guesses” or extrapola-tions based on quite limited evidence.

Second, several of the estimates extrapolateavailable data on existing production to the en-

tire Appalachian Basin by using methods that relyon guesswork, on arguable assumptions, or onapparent relationships with variables that do notseem likely to be strongly related to resourcerecovery potential. For example, the early OTAstudy did not formally relate well production tomeasurable physical attributes of the areas underdevelopment, presumably because there were in-adequate data. Instead, the study assumed that10 percent of the basin will be productive in the

same manner as the area now under production,and that the remainder of the basin will be un-productive; the choice of 10 percent was not ex-plained, but is essentially an “educated guess.”Both the Pulle and Seskus and the NPC estimatesrelied on predictive variables—in NPC’s case,shale thickness as measured by gamma-ray logs—which seem likely to be only limited predictorsof gas recovery. The Zielinski and Mclver estimatefor West Virginia, Ohio, and Kentucky, admit-tedly a preliminary, crude attempt, is based onthe assumption that the existing wells, used forextrapolation, were randomly sited, so that theirproduction would be representative of whatwould occur in the untested portions of the shalearea. This type of assumption is more tenable inthe Appalachian Basin than it would be else-where because sophisticated exploration tech-niques were not used to site the existing wells,

the haphazard availability of land for drilling mayhave interfered with pattern drilling, and other

factors. In our view, however, the information

gained by earlier drilling is certain to have di-rected subsequent drilling to better-than-averageprospects, and the direct extrapolation used inthis study will tend to lead to an overestimate ofthe resource recoverable by borehole shooting.

Third, only a very small part of the Devonianshale has been tested by drilling, and conse-quently the available studies focus on the re-sources recoverable from the producing shale in-tervals, or those closely resembling them, i.e.,those with well-developed natural fracture sys-tems. Much of the gas-in-place exists in intervalswhich are not currently productive. For the mostpart, the large portion of this gas probably couldnot be produced with current technologies andprices. However, it may be possible to economi-

cally recover much of this gas with new technol-

ogy, especially at higher prices. The existingstudies cannot account for this possibility, andprobably there is no credible way at present todetermine the true potential of these intervals.Nevertheless, it must be made clear that the ex-isting estimates of the Devonian shale resourcepotential do not include this speculative portionof the resource, and that there is at least the pos-sibility that, with further technology develop-ment, the gas ultimately recoverable from theDevonian shales may be considerably larger than

currently estimated.

Fourth, all of the estimates may suffer from theproblem that past production has been influ-enced by the market and by other external in-fluences on production. The studies all implicitlyassume either that the resource characteristics aredominant in determining production character-istics, or else that any external influences on pro-

duction will remain essentially unchanged dur-ing the period of development of the resource.Zielinski and Mclver have noted that these pro-duction influences might have affected the ulti-mate recovery of past wells, and should be takenaccount of when extrapolating to future pro-

duction.

Comparison of the resource estimates of thevarious studies is made difficult by the first prob-lem, the lack of clearly defined criteria for “ad-vanced technologies, ” and also by differences in




assumed gas prices and study areas. However,relatively clear comparisons can be made of theresources recoverable by traditional boreholeshooting. Table 53 compares the resource esti-mates of six studies for borehole shooting at 150-

to 160-acre spacing and, for three of the studies,

at prices moderately higher than current marketprices for new gas. The estimates of Pulle andSeskus, Zielinski and Mclver, and the recentLewin study do not specify prices and are morein the nature of “technically recoverable gas. ”They are best compared to the NPC “total pro-ducible” resource of 25.3 TCF and the 1977Lewin estimate of 10.5 TCF for gas prices of$4.50/MCF ($7.00/MCF [1983$]); the latter should

be close to a “technically recoverable” limit be-cause of the rapidly diminishing returns for fur-ther price increases apparent in the Lewin analysis.

Because there has been extensive experiencewith borehole shooting, the estimates for recov-erable resources using this technology should bethe most reliable. There are serious differencesamong the estimates in table 53, however, andwe believe these differences reflect some of ourconcerns with the individual studies. The Zielin-ski and Mclver estimate, which should be con-sidered optimistic because it assumed that sitingof past wells was random and therefore that theirproduction experience would be representative

of undrilled acreage, is in fact the highest of theestimates of technically recoverable gas. The 1977

OTA estimate is not based on a geological evalua-tion of the prospects of the undeveloped portionof the basin, and probably should be down-played; it is, in fact, at considerable variance with

the NPC and early Lewin estimates, which aremore pessimistic. On the other hand, the earlyLewin estimates also appear to be considerablymore pessimistic than the recent Lewin estimatesfor Ohio and West Virginia; it seems likely thata new Lewin estimate of the Appalachian gas

recoverable at about $4.70/MCF (1983$) wouldbe considerably higher than the 8 TCF predictedby the early study. The difference between theearly and more recent Lewin studies is empha-sized by the recognition that the early study in-cludes some hydraulic fracturing in its recover-able resource estimate; presumably, its estimatefor the resource recoverable with borehole shoot-ing only would have been even lower than 8 TCF.

In conclusion, OTA’s best resolution of existingresource studies is that the Devonian resourceultimately recoverable with borehole shooting at160-acre spacing and gas priced at $4 to $5/MCF(1983$) is at least 10 TCF and, based on the im-plications of the recent Lewin work, most prob-ably is somewhat higher. Ultimately, if the Lewinanalyses prove to be substantially correct, about

30 to 50 TCF may be recovered with this tech-nology and spacing at very high prices. These areextremely conservative estimates of the actual gaspotential of the Appalachian Basin, however, be-cause neither the technology assumption nor thespacing assumption are realistic. Many producersin the basin have begun to use hydraulic fractur-ing, which increases ultimate recovery per well,and well spacing in the less permeable areas hasbegun to be decreased to 80 acres, which im-proves recovery per section. With boreholeshooting as the “baseline” technology, halving

Table 53.—Comparison of Estimated Appalachian Devonian Shale Resources Recoverable With Borehole Shooting,Well Spacing of 150 to 160 Acres

Gas price ($/MCF) Recoverable resource Total producibleStudy (data) study date (1983)

a(TCF) (TCF)

OTA (1976) . . . . . . . . . . . . . . . . . . . . .... ... ... .. .$2 to $3 ($3.03 to $4.95) 23 NANPC (1979) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $3.50 ($4.70) 8.5 25.3Lewin I (1977). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $3000 ($4.68) 8

b10.5+

Pulle and Seskus (1981) . . . . . . . . . . . . . . . . . . . . . . None 17-23

Zielinski and Mclver (1982)(West Virginia, Ohio, Kentucky) . . . . . . . . . . . . . None 30-50Lewin II (1983) (portions of

Ohio and West Virginia only) . . . . . . . . . . . . . . . None 25.3aThe 1983 gas price is obtained by applying the GNP Price Deflators published by the Bureau of Economic Analysis, Department of Commerce.blncludes hydraulic fracturing as well as borehole shooting.





the spacing seems likely to yield at least a 70-percent increase in recoverable resources, pri-marily because the area of influence of each wellis smallest with this technology and thus the ad-ditional wells will interfere least with the adjacentwells. The incremental benefit of reduced spac-

ing will decrease for recovery technologies thatcontact more of the formation because of inter-ference between adjoining wells; however, Lewin

estimated that halving the spacing will still add40 percent to the recoverable resource in Ohio

even when the baseline technology achieves 600-

ft fractures, which should maximize interference

between wells.

The use of conventional hydraulic fracturingand more advanced stimulation also can greatlyincrease the recoverable resource. The N PC esti-

mates that conventional fracturing will yield a 50-to 70-percent increase in recovery over boreholeshooting. They project an additional 40-percentincrease over the conventional fracturing with themore advanced stimulation techniques, but thisestimate is based on very limited experience. Therecent Lewin study estimates that a stimulationachieving 600-ft fractures can more than double

the recoverable resource over that obtainablewith borehole shooting, and that a combinationof this technology with 80-acre spacing can morethan triple the resource.

These estimates imply that the recoverable re-source in the Appalachian Devonian shales may

prove to be quite large, perhaps 80 or 100 TCFor even higher with high gas prices and substan-tial improvements in recovery technology. How-ever, the current limited capability in reservoirsimulation and our limited geologic understand-ing imply that estimates of recoverable resourcesusing advanced technology should be viewed ashaving quite high uncertainty.

Finally, as noted previously, none of the esti-mates consider the possibiIity of producing fromshale intervals that do not contain well-developednatural fracture networks. This portion of the

shale is a highly speculative resource, and its gas-in-place may never become recoverable. Never-theless, it does present some potential for futurerecovery.


Annual production estimates may be calculatedby estimating the number of wells to be drilledand postulating a drilling schedule. The numberof wells drilled is determined by the area con-sidered to be drillable and the well spacing

assumed. The larger the well spacing, the fewerthe number of wells that may be drilled in a given

area. The USGS argues that drainage patterns ofDevonian shale wells are too variable to assumea constant spacing.

Lewin & Associates determined 1990 produc-tion estimates based on an available acreage of57,000 square miles, a well spacing of 150 acresper well and the drilling and development sched-ules outlined for the base and advanced technol-ogy cases. The resulting estimates are shown intable 54.

Annual production and additions to reserveswere also calcuIated in the NPC study. The num-ber of wells drilled were constrained by an avail-able acreage of 62,000 square miles, a well spac-ing of 160 acres per well and “low” and “high”drilling schedules. The low scenario assumedthere would be initially 12 rigs drilling in Devo-nian shale in 1980, and a 12-percent increaseeach year thereafter. The high scenario assumed15 rigs were active in 1980, with 15 rigs added

each year through 2000. All rigs were assumedto drill 35 productive wells per year. The results

of this analysis are included in table 55. As shownin the table, the high scenario depends on ex-tremely high gas prices into the 1990s. Even withadvanced technology, the year 2000 production

rate of 1.35 TCF/yr requires gas prices above$10.00/MCF (1983$).

It is not clear whether or not the annual pro-duction estimates projected by the NPC andLewin accounted for some important factors thatcan influence the rate at which the resource isdeveloped. The effects of these factors are notreadily quantifiable, but they can be used to qual-ify the production potential estimates.

One factor that definitely will affect the pro-

duction rate is the availability of adequate leases

for exploration and drilling. In fact, in most ac-

38-74~ O - 8 5 - 1 4




Table 54.—Lewin & Associates Results for Annual Production Estimates

Price, 1977$ (1983$) Production rates (TCF/yr)

($/MCF) Base case Advanced case

1.75 (2.75) . . . . . . . . . . . . . . . Peak 0.1 in 1990 Peak 0.3 in 1990declines thereafter declines thereafter

3.00 (4.70) . . . . . . . . . . . . . . . Peak 0.3 in 1990 0.6 in 1990, hold to 1995

gradual decline thereafter declines thereafter4.50 (7.00) . . . . . . . . . . . . . . . Remains constant at 0.3 Increases to 0.7 to 0.9 in 1990

through 2000

SOURCE: Lewin &Associates, Inc., Enhanced Recovery of Unconventional Gas, U.S. Department of Energy Report HCP/T2705-01,02, and 03, 1978-79.

Table 55.—Potential Incremental Supply of Devonian Shale Gas In the Appalachian Basin:NPC High Growth Drilling Schedule (production and reserve volumes [BCF] and price [$/MMBtu]) (constant 1979 dollars)

1980 1985 1990 1995 2000Annual productive wells drilled. . . . . . . . . . . . .Cumulative wells . . . . . . . . . . . . . . . . . . . . . . . . .Traditional technology:

Annual production rate . . . . . . . . . . . . . . . . . .Annual reserve additions . . . . . . . . . . . . . . . .Cumulative additions . . . . . . . . . . . . . . . . . . .Incremental price at 10°/0 ROR . . . . . . . . . . .

Conventional technology:Annual production rate . . . . . . . . . . . . . . . . . .Annual reserve additions . . . . . . . . . . . . . . . .Cumulative additions . . . . . . . . . . . . . . . . . . .Incremental price at 10°/0 ROR . . . . . . . . . . .

Advanced technology:Annual production rate. . . . . . . . . . . . . . . . . .Annual reserve additions . . . . . . . . . . . . . . . .Cumulative additions . . . . . . . . . . . . . . . . . . .Incremental price at 10°/0 ROR . . . . . . . . . . .

770770

15200200

<2.50

17240240

<2.50

21290290

<2.50

3,40012,500

190

8903,300<2.50

2201,0403,800

<2.50

2701,2904,800

< 2 . 5 0

6,00037,300

4301,2508,800

<5.00

5501,660

11,000<3.50

7002,030

14,000<3.50

8,65075,300

6201,110

14,300<7.00

8651,690

19,600<7.00

1,1102,170

25,100<5.00

11,300126,400

690720

18,400<12.00

1,0051,140

26,100<9.00

1,3551,600

34,500<9.00


tive areas of Devonian shale drilling, perhaps 95percent of all wells drilled are located on the basis

of the availability of Iand. Most of the Devonianshale in the Appalachian Basin occurs in theolder, more populated areas, which over theyears have been divided into small tracts (by oilindustry standards). Building an exploration blockof any size is difficult and expensive with manytitle problems. Also, a trend towards short-termleases of 1 year or less has made establishing an

orderly exploration program difficult. Some Statesalso have no well spacing requirements, makingany tract capable of holding a potential drillingrig a drill site. As a result, a good well may be

jeopardized by other wells that are placed too

close to it.

The type of operators in the Devonian shalewill also influence the development of the re-source. The major oil companies have not in-vested in Devonian shale wells, due principally

to land problems and poor well performance.The majority of operators are small companies

financed by drilling funds or direct investmentgroups. Although these small operators can move

swiftly into “hot” acreage, their cash flows aregenerally not sufficient to allow much explora-tion. As a result, many wells are drilled with lit-tle regard to geological conditions. A lack of fundscould also inhibit the use of costly yet more ef-fective evaluation and stimulation techniques,

thereby reducing the quantity of gas ultimatelyrecovered,

Aside from these uncertainties, and the obviousuncertainty introduced by our inability to project

future economic and market conditions such asgas prices, demand, and availability of capital,production projections share most of the uncer-tainties associated with the estimates of recov-erable resources discussed previously in thischapter. Of particular interest are uncertainties




in the effects of improved stimulation technol-ogies, since increases in fracture areas will affectproduction rates as well as total cumulative pro-

duction. Consequently, production projectionsassuming the use of advanced technologiesshould be considered substantially more uncer-tain than projections assuming explosive fractur-ing. In the latter case, extrapolating from historicproduction data should be an acceptable proce-dure for projecting future production, althoughour lack of data on production practices andother factors that might have influenced past pro-duction rates requires that a substantial errorband be placed around the results.

Production levels of about 1.0 TCF by the year2000 would seem to be readily supported by theavailable studies: the 1977 OTA study concludedthat 1.0 TCF/yr could be achieved 20 years aftercommencing an intensive drilling program, as-

suming relatively moderate prices; the first Lewinstudy projected a maximum production rate of0,9 TCF/yr in 1990, with advanced technologyand $4.50/MCF gas (1 977$) ($7.00/MCF [1983$]);and the NPC study projected a 1.0 TCF/yr pro-duction rate in 2000 using available technology,although admittedly at a very high price ($9.00/ MCF [1979$]). However, high production levels

of Devonian shale gas in the Appalachian Basinusing currently available or moderately improvedtechnology implies a massive expansion of drillingin an area where such an expansion is institu-tionally and physically difficult. Also, the produc-tion levels in the various studies were derived byassuming arbitrary drilling levels and extrapolat-ing production data from quite limited areas tothe basin as a whole. On the other hand, the re-cent Lewin work in Ohio and West Virginia im-plies that gas recovery (and production rates) per

section can be increased substantially with re-duced spacing, tailoring of drilling patterns topermeability anisotropy, and improved fractur-ing. Increased recovery per section could allowa more rapid expansion of production by easingproblems of pipeline construction and land as-semblage. Consequently, if gas market conditionsin the Northeast improve very soon, institutional

barriers to production are reduced or overcome,and exploration and production technology ad-vances are achieved, OTA considers a produc-

tion rate of 1.0 to 1.5 TCF/yr from the Appa-lachian Devonian shales by the year 2000 to bequite plausible. Achievement of the prerequisiteconditions in the short time span involved is,however, still somewhat optimistic.



Chapter 10

Coalbed Methane



Contents

Page

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........................207

Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .................207

Gas-in-Place . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...............209Methodologies and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...209Uncertainties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...................211

Production Methods and Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....212Production Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..............212

Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ......................215

Recoverable Resources and production Potential . ........................217Methodologies and Results . . . . . . . . . . . . . . ............................217Estimate Comparison and Uncertainties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..219Annual Production Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..........221

TABLES

Table No. Page

56. Coalbed Methane Resource Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21057. Coal Resource and Gas Content Assumptions . .................,......21058. DOEGas-in-Place Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..21159. Comparison of Recoverable Resource Estimates . ......................21760. GRl Coalbed Methane Annual Production Estimates (TCF). . . . . . . . . . . ....223

FIGURES

Figure No. Page

43. U.S. Coal Regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .........20844. Variation of Gas Content by Rank, Temperature. . . . . ..................20845. Variation of Adsorbed Gas by Rank . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..20946. Methane Gas Adsorbed on Coal as a Function of Pressure . .............21347. Pressure Drawdown Curves for Three Wells in a Line . .................21448. 20-Year Production Prediction for Gas and Water Production From a Well

Pattern Designed to Allow Rapid Water Removal . .....................21449. Deviated Drilling into a Coal Seam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...21650. Estimated Distribution of Bituminous Coal by Seam Thickness . ..........21851. Well Production Histories in San Juan and Other Basins . ...............22052. Annual U.S. Production Rates of Coalbed Methane as a Function of

Time–Vertical Wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........22253. Annual U.S. Production Rates of Coalbed Methane as a Function of

Time–Horizontal Wells . . . . . . . . . . . . . ..............................222



Chapter 10

Coalbed Methane

INTRODUCTION

Methane in coal seams traditionally has beenviewed as a hazardous waste product of the min-ing operation, rather than an energy resource inits own right. In fact, gas in a coal seam only con-tains 1 to 2 percent of the energy capacity of thecoal itself. As a consequence, an estimated 217thousand cubic feet per day (MCF/D) or 80 bil-lion cubic feet per year (BCF/yr) of methane isvented to the atmosphere from U.S. mines, ’ with-out thought of recovery, to increase mine safety.The search for additional natural gas resourcesin the 1970s fostered an interest in economicallyrecovering this “wasted” gas. By removing this

gas before mining begins and either using it onsite or selling it to the natural gas market, energyconservation could be combined with increasedmine safety. In addition, it was realized that a po-tentially large gas resource lies trapped in seamsof coal that will likely never be mined becauseof their depth or physical characteristics.

Although it is widely acknowledged that thecoal bed methane resource is large, early eco-nomic assessments suggested that it had little po-tential for economic recovery barring very highgas prices. More recent evidence from wells pro- .ducing gas from coal seams at current prices sug-gests a more optimistic outlook is justified. It ap-

pears that in some areas with highly favorablegeology, commercial volumes of gas are recover-able at current prices using existing technologies.

Current production efforts include nearly 100producing wells drilled by various operators in

Alabama’s Black Warrior Basin, early efforts byCarnegie Natural Gas Co. and Equitable Gas Co.in Pennsylvania and West Virginia, a variety ofwells in the San Juan Basin of New Mexico, and

others. z

IV. A. Kuuskraa and R. F. Meyer “Review of World Resources

of Unconventional Gas, ” IIASA Conference on Conventional and

Unconventional World Natural Gas Resources, Luxenburg, Austria,

]une 30-july 4, 1980.

2). L. Wingenroth, “Recent Developments in the Recovery of

Methane From Coal Seams, ” Gas Energy Review, vol. 10, No. 9,

September 1982, American Gas Association.

CHARACTERISTICS OF THE COALBED METHANE RESOURCE BASE

Coalbed methane is defined as natural gastrapped in coal seams. The location of the Na-tion’s coal resources and associated methane ac-cumulations are depicted in figure 43. Approx-imately two-thirds of the resource is located inthe West and Midwest and the remainder is lo-cated in the Appalachian Basin.

Methane forms as a byproduct of the coalifica-

tion process. s With increasing temperatures, the

rank (carbon content) of the coal increases, andlarger volumes of methane and other volatile con-stituents are produced (fig. 44). As volatiles are

~Coalification: the formation of coal from organic-rich sediments,

under intense heat and pressure.

driven off by the increasing temperatures, thecoal shrinks, giving rise to a pervasive natural frac-ture system called the “cleat.” Although muchof the generated gas migrates out of the forma-tion, some remains in the coal seam adsorbed

to the coal pore surfaces, and some is trappedin the pore spaces and fracture system by thereservoir pressure. In sharp contrast to conven-tional gas reservoirs, where essentially all of thegas is trapped in the pores and fractures, the ad-sorbed gas is the dominant source of coal bed

methane and plays the major role in production.The volume of adsorbed gas appears to be a func-tion of depth (pressure) and coal rank, as shownin figure 45. Nevertheless, given the vagaries of

207




Figure 43.—U.S. Coal Regions

SOURCE: Department of Energy.

Figure 44.-Variation of Gas Content by Rank, Temperature

20 Coal rank

170(338)

220 (428)

270

50 100 150

Yield, liter/kg

200

SOURCE: J. M. Hunt, Petroleum Geochemistry and Geology (San Francisco: W. H. Freeman & Co., 1978) ( fig.5-7, p. 165).



I

Ch. 10—Coalbed Methane q 209

I 4700

600

200

100

Depth, meters

SOURCES: A G Kim, “Estimating Methane Content of Bituminous CoalbedsFrom Adsorption Data,” BuMines RI 8245, 1977 c.f ; G. E. Eddy, C. T.Rightmure, and C. W. Bryer, “Relationship of Methane Content of CoalRank and Depth, Theoretical vs Observed,” SPE/DOE Unconventional Gas Recovery Symposium, 10800, 1982

the geologic process, methane content in coalis highly variable from seam to seam and evenwithin the same seam.

The quality of the gas present in coal seams isalso somewhat variable, but generally is quitegood. The heat of combustion ranges from 950to 1,050 Btu per cubic foot. The gas has few im-purities; carbon dioxide and water vapor are theprimary undesirable components. Sulfur dioxide

and hydrogen sulfide gases are absent even inthe more sulfur rich coals.

Coal in itself is essentially impermeable. Bulkpermeability of a coal seam depends on howwell-developed the cleat is. Generally, there isa dominant system of vertical fractures, the so-called “face cleat, ” and a less developed systemof vertical fractures perpendicular to the facecleat, the “butt cleat, ” the nature of the face cleatis critical to the coal’s production characteristics.The importance of the natural fracture system,

together with the critical production role played

by adsorbed gas, establishes a close parallel be-tween coal seam methane and the Devonianshale gas resource.

Many coal seams contain water and thus thereservoir pressure is partialIy a hydrostatic pres-

sure caused by groundwater. Although in some

cases the water is the original product of the coalformation process, often the water infiltrates thecoal from the surface or from overlying aquifers.The presence of this water has profound effectson gas production from the coal seams.

GAS-IN-PLACE

Existing gas-in-place estimates are summarized Methodologiesin table 56. These range from a low of 68 TCFto a high of about 850 TCF. Most estimates arebased on the U.S. coal resource; differences arisefrom varying assumptions of gas content (volumeof gas per cubic foot or ton) of the coal and cri-

teria for defining coal seams as good targets forrecoverable gas.4 Several of the most recent esti-mates are discussed in more detail below.

4These criteria are important because gas-in-place estimates gen-

erally consider only gas found in formations that contain poten-

tially recoverable gas.

and Results

National Petroleum Council, Gas ResearchInstitute, and Kuuskraa and Meyer

Among the more recent studies, the estimatesby the National Petroleum Council (NPC), the

Gas Research Institute (GRI), and Kuuskraa andMeyer (KM) are very similar in methodology andin results. The estimated resource in place rangesfrom 398 TCF (NPC) to 550 TCF (KM).

All of these studies use the 1974 U.S. GeologicSurvey’s (USGS) coal resource data—an estimate




Table 56.—Coalbed Methane Resource Estimates

Resource in placeStudy (TCF)

Department of Energy (1984) . . . . . . . 68-395Kuuskraa and Meyer (1980) . . . . . . . . 550National Petroleum Council (1980) . . 398Gas Research Institute (1980) . . . . . . 500

Federal Energy RegulatoryCommission (1978) . . . . . . . . . . . . . 300-850

Deul and Kim (1978) . . . . . . . . . . . . . . 318-766Wise and Skillern (1978) ., . . . . . . . . . 300-800TRW . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72-860National Academy of Sciences

(1976) . . . . . . . . . . . . . . . . . . . . . . . . . 300SOURCE: Adapted from AGA Gas Energy Review, September 1982; and C. W.

Byrer, T. H. Mroz, and G, L. Covatch, “Production Potential for Coal-bed Methane in U.S. Basins,” SPE/DOE/GRl Unconventional GasRecovery Symposium, 12832, 1984.

of “minable” coal resources—as the basis for

their estimates. The USGS assessment is broken

down into identified and hypothetical resourcesat depths less than 3,000 ft and hypothetical re-

sources at depths greater than 3,000 ft. The iden-tified resources are further broken down by coalrank—anthracite, bituminous, subbituminous,and lignite. For the methane estimates, Kuuskraaand Meyer have subdivided the hypothetical re-sources by rank in approximately the same pro-portion as they occur in the identified resources.

The NPC, GRI, and KM analyses then multiply

the coal resource by an assumed gas content todetermine the gas resource in place. All assume

that gas content varies with rank and depth.

Assumptions are compared in table 57. Although

the KM estimates disaggregate the gas content ofcoal to a greater extent than the GRI or NPC esti-mates, their assumed gas contents, averaged, areessentially the same as the GRI and NPC values.

Consequently, the increased detail in their esti-mate does not contribute to a substantial differ-ence in the calculated gas-in-place.

The NPC estimate excludes all coal resources

at depths less than 300 ft, assuming these coalseams contain essentially no recoverable gas. Theexclusion of shallow coals appears reasonable be-cause the lower pressures may have allowed anygas originally contained in shallow seams to haveescaped to the surface. However, the NPC alsoassumed that a full third of the identified andhypothetical coal resource between O and 3,000ft occurs above 300 ft; also, the NPC apparently

assumed that the bulk of this shallow coal is bitu-minous, with high gas content. Thus, the NPCanalysis excludes from consideration a large per-centage of the higher-gas-content coal. This con-clusion appears to be the primary reason that theNPC estimates are 100 to 150 TCF lower than the

GRI and KM estimates.5 The exclusion appearsoverly pessimistic because it is the lower rank

5,41though G RI also appears to exclude the coal resource at less

than 300 ft from their gas-in-place calculations, in fact their total

coal resource base is equal to the USGS coal resource base (3,968

X 109 tons) from O to 6,000 ft.

Table 57.-Coal Resource and Gas Content Assumptions

Coal rank

Zero to 3,000 ft Greater than 3,000 ftKuuskraa & Meyer Kuuskraa &

<1,000 1,000-3,000 >3,000 NPC GRI Meyer NPC GRI

A) Coal resource assumptions (billion short tons): Anthracite (A) . . . . . . . . . 845

739 1,58446 60

Bituminous (B) . . . . . . . . 1,001 1,300A + B . . . . . . . . . . . . . 845 739 1,584 1,047 1,360

Subbituminous . . . . . . . . 538 470 1,008 1,137 1,520Lignite . . . . . . . . . . . . . . . 538 470 1,008 504 700

Total . . . . . . . . . . . . . . . 3,600 2,688 3,580Total (all depths) . . . . 4,000 3,076 3,968

B) Gas content assumptions-cubic ft/ton: Anthracite and

bituminous . . . . . . . . . 150 250 197a200 200

Subbituminous . . . . . . . . 60 100 79a

80 80Lignite . . . . . . . . . . . . . . . 30 50 39 a 40 40

176112112

400 388 388

500 200 200200 200 200100 200 200

aWeighted average.




Ch. 10—Coalbed Methane . 211

(thus lower gas content) coals that tend to occur

at shallower depths.

Department of Energy (DOE) —MethaneRecovery From Coalbeds Project

The DOE Coalbed Methane Basin analysis isthe first attempt to estimate the gas-in-place ona basin-by-basin level. The DOE approach targetsthe most likely gas-producing coal seams in eachcoal-bearing basin. Wherever possible, they haveestablished a range of gas contents for the tar-geted coals in each basin and calculated a gas-in-place. Data were obtained from a variety ofproducing wells and test wells. They have com-pleted studies of 14 basins with an estimated totalgas-in-place of 68 to 396 TCF. Results for the 14basin analyses are summarized in table 58. The

high end of this range is essentially compatible

with earlier estimates; the low end is very con-servative, being the product of lower estimatesof both target area and gas content.

DOE appears to have made a number of subjective judgments in delimiting its target areas.For example, DOE selected for inclusion in theresource base only those coal seams with highreported gas contents, high rank, and thick cumu-lative sections, without setting any quantitative

criteria for the selection. In addition, assessments

of several basins have not been completed. Thus,its estimate is conservative in terms of total gas-in-place. Because it focused on formations thatare the most likely to contain recoverable gas,however, the gas-in-place estimates may repre-sent a valid basis for an estimate of the technically

recoverable resource.

Uncertainties

The wide range of gas content in coal seams,

seen clearly in table 58, is the primary factor con-tributing to uncertainty in gas-in-place estimates.The range in the DOE gas-in-place estimates—

over a factor of 5—may not be an unreasonablereflection of the true uncertainty at this time. Thelevel of uncertainty will only be reduced as more

data are obtained on gas content of specific coal

seams. However, the impetus to obtain moredata may only come as producers move to de-velop these resources.

The other major factor contributing to uncer-tainty is the lack of data on coal resources atdepths greater than 3,000 ft. The USGS coal re-source estimate is limited to potentially minableseams, and may substantially underestimate thegas-bearing resource. Very little information is

available on the rank, reservoir characteristics,

Table 58.— DOE Gas-In-Place Estimates

Gas contents Estimated total gas-in-place (TCF)

Basin (CF/ton) Minimum Maximum

Eastern: Northern Appalachian , . . . . . . . . . . . . . . . . . . . . . . 30-420 61.0Central Appalachian . . . . . . . . . . . . . . . . . . . . ... , . 125-400 10.0 48.0Illinois , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30-150 5.2 21.1Warrior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-600 11.0Arkoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70-700 1.6 3.6Richmond . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ND 0.7 1.4

Western: Piceance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 -410+ 30.0 110.0Powder River 1.45 5.9 39.4Greater Green River . . . . . . . . . . . . . . . . . . . . . . . . . 13-539 0.2 30.9San Juan. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-135 + 1.8 25.0Western Washington . . . . . . . . . . . . . . . . . . . . . . . . 32-86 3.6 24.0Raton Mesa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-492 8.0 18.4

Wind River . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . a 0.5 2.2Uinta , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-443 0.2 0.8

Total . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . ..—

67.7 395.8ND—no dataaAssumes deep coals will contain some gas

SOURCE. C. W Byrer, T H. Mroz, and G L Covatch, “ProductIon Potential for Coal bed Methane in U.S Basin s,” SPE/DOE/GRI Unconventional Gas Recovery Sym-posium, 12832, 1984




and gas content of the deep and unminable coals.Because deep coals are likely to be of higher rankand have higher gas content than shallower coals, bgas-in-place estimates that assign to the deepcoals the same coal rank distribution found in theshallow coals may be too conservative. However,

very low permeabilities, particularly in an-thracites, may exclude some of these coals assources of economically recoverable gas re-sources, absent significant advances in wellstimulation technology.

‘Deeper coals are more likely to have been exposed to high

temperatures, which in turn influence rank and gas content. See

fig. 44.

To the extent that the deep coal seams are notconsidered to be viable targets for mining, manyof the legal and institutional constraints to pro-

ducing methane from minable coal seams will notbe applicable to these deep seams. This may in-crease their attractiveness to gas producers. Refin-

ing the estimates of the deep gas resource, alongwith incorporating improved gas content datafrom the newly drilled basins, are the most im-portant tasks remaining in establishing a morecredible estimate of the coal bed methane gas-in-place.

PRODUCTION METHODS AND TECHNOLOGY

Production MethodsProducing coal seam methane is considerably

different from producing natural gas in conven-tional reservoirs. Production rates in conventional

reservoirs are primarily a function of permeability,whereas in coal seams, methane production isalso dependent on the rate at which the adsorbedmethane diffuses into the fracture network, or“cleat.” If the permeability of the coal’s fracturenetwork is very low, then permeability will be thefactor controlling production rates. However,when the fracture network is relatively permeable

and is connected to the well bore, or when thefracture network is not well-developed (and thus

the surface area for diffusion to take place islimited) production is more likely to be limitedby the rate of diffusion of the adsorbed methaneinto the fracture network.

These different limiting factors have importantimplications for the probable effects of fractur-

ing. If permeability is controlling, fracturing shouldincrease production by enhancing the flow path

from the fracture network to the wellbore. If dif-fusion is controlling, however, fracturing is un-likely to greatly affect production because it can-not add greatly to the surface area available forresorption, and any increased permeability it cre-ates will not add to production. 7

7Lewin & Associates, Inc., Enhanced Recovery of Unconventional

Gas, Vo/urne ///: The Methodology, U.S. Department of Energy re-port HCP/T2705-03, February 1979.

It is necessary to reduce the pressure in the frac-ture systems in order for gas to desorb from the

coal and be available for production. Figure 46shows how reducing the pressure will reduce thevolume of gas adsorbed. The pressure/gas vol-

ume curve, which is typical of coal seams, isstrongly nonlinear: a unit pressure drop has farless effect on resorption at high pressures thanit does at low pressures. As a result of this non-linearity, there may be little or no gas produc-tion until the pressure in the formation is reducedto the level where the rate of resorption per unitpressure drop begins to accelerate.

Because the reservoir pressure generally is a

hydrostatic head associated with the groundwaterin the coal seam, reducing the reservoir pressuremeans dewatering, i.e., pumping the water outof the seam. Water removal also increases therelative permeability of gas in the fracture net-work, allowing more gas to flow to the well bore.This also tends to reduce the pressure in the for-mation, further increasing the ability of gas to

desorb from the coal,

As pumping the water from a well commences,

the reservoir pressure is first reduced in the im-mediate vicinity of the wellbore, with the areaof the pressure drop spreading overtime. The rateof gas production generally will increase withtime as more and more area achieves the largepressure drop necessary to cause rapid resorp-

tion. This production increase with time is in




Figure 46.— Methane Gas Adsorbed on Coal as aFunction of Pressure

22

Maw Lee Coal

18

t A

16LB

14 -

12 -

c

10 -

8 -

6 -

4

SOURCE: S. C Way, et al , “Role of Hydrology in the Production of Methane FromCoal Seams,” Quarterly Review of Methane From Coal Seams Tech.nology, vol 1, No. 2, August 1983, Gas Research Institute,

sharp contrast to the more normal production de-cline experienced in conventional gas wells. s

This model of gas production from coal seamsmay not apply to single, isolated wells. In coalswith highly permeable, interconnected fracture

systems, the effect of pumping over time willdraw the pressure down in small increments overa wide and expanding area with little change inthe pressure distribution near the wells.9 Because—————

as. C. Way, et al,, “Role of Hydrology in the Production of Meth-

ane From coal Seams,” Quarter/y Review of Methane From Coa/Seams Techfto/ogy, vol. 1, No, 2, August 1983, Gas Research in-stitute.9“New Advances in Coalbed Methane, ” Intercomp Resource De-

velopment & Engineering, Inc. (appears as app. C in K, L, Ancell,

Coa/ Degasification, An Unconventional Resource, Dowdle Fair-

child & Co., Inc., Houston, TX).

resorption is strongly nonlinear, favoring a largepressure drop, the areal extension of a moderatepressure drawdown is likely to yield little addi-tional gas. The solution to this problem is to

somehow bound the drainage area of the wellsso that larger pressure drops occur over time.

One method is to drill a closely spaced patternof wells and pump them simultaneously, delib-erately creating interference between adjacentwells. Such interference will effectively halt orbound the areal spread of pressure drop from asingle well. This practice is in sharp contrast tonormal practice in conventional gas fields, where

close spacing and well interference are avoidedbecause they reduce average recovery per well.

Figure 47 shows pressure drawdown curves for

a group of three wells. The broken lines repre-sent the pressure curves associated with each wellin isolation; the solid lines are the actual pressure

curves that result from the three well system, re-flecting the interference effects of the wells on

each other. Close spacing of wells allows more

of the formation to achieve the sharply reducedpressures necessary for maximum resorption andproduction of gas. The advantage of closer spac-ing is particularly apparent when water can in-filtrate the formation. Pumping from isolated wellsmay simply be unable to remove the water fasterthan it can infiltrate, and thus such pumping willnot successfully dewater the formation and pro-duce the gas; simultaneous pumping from agroup of wells generally can “outrun” the infiltra-tion. In a formation where water infiltration is aproblem, the rate of pumping also becomes crit-ically important, since a higher pumping rate maybe necessary to outrun the infiltration and suc-cessfully draw down the pressure.

One well spacing pattern uses exterior wells toproduce water and provide interference, whileinterior wells produce most of the gas.10 Simu-lated production from this configuration predictsconsistent high flow rates over a 20-year term,as shown in figure 48.

Iolbld ,




Figure 47.— Pressure Drawdown Curves for Three Wells in a Line

Q, Q2 Q3

Impermeable

SOURCE: S. C, Way, et al., “Role of Hydrology in the Production of Methane From Coal Seams,” Quarter/y Review of Methane From Coal Seams Technology, vol. 1, No. 2, August 1983, Gas Research Institute.

Figure 48.— 20-Year Production Prediction for Gas and Water Production From a Well PatternDesigned to Allow Rapid Water Removal

History I 20-year prediction Time, days

SOURCE: INTERCOMP Resource Development and Engineering, Inc.



Ch. 10—Coalbed Methane 215

Technologies

Increasing production, at least from shallowwells, does not depend primarily on the devel-opment of new technologies. The primary targetfor improved technologies are the deep coal

seams. There we need better reservoir characteri-

zation techniques. New drilling and completiontechniques also are required in deep wells wheredrilling fluids or cements used for completionsare likely to cause extensive formation damage.Stimulation technologies also have not beenhighly successful in the lower permeability deep

coal seams.

Drilling

Most current production of methane from bothminable and unminable coalbeds uses verticalwells drilled through the coal seam. Major prob-

lems involve extensive formation damage in-duced by drilling fluids and by cements used tocomplete the holes. In shallower wells, air orwater drilling and open hole completions11 can

counteract these problems. Improved productionfrom multiple completions12 and from deep coalseams where open hole completions may not bepractical will require new technology develop-ments. Current research programs sponsored bythe Gas Research Institute are addressing theseproblems.

Another problem inherent in vertical drilling

is the difficulty of intersecting the vertical frac-ture network, or face cleat. Well stimulation maybe required to connect the wellbore with the nat-ural fracture system and thus provide a pathwayfor gas to flow to the well. Another remedy is to

slant or deviate the wells from the vertical to in-tersect the face cleats. Ideally, the well can bedrilled parallel to the seam, as sketched in fig-ure 49. This technology requires considerable im-provement and cost reductions to be considered

a realistic option for coal seam methane produc-tion. Keeping the wellbore in the coal seam isquite difficult, and drilling costs are significantly

1‘That is, completing the well by perforating the gas-bearing rock

formation without first casing and cementing the wellbore In the

vicinity of the formation.1.2That Is, pr~ucing from multiple seams with a single well.

higher than for vertical wells. Dewatering devi-ated wells may also be a problem.

In minable coal seams, horizontal wells may

be used. These wells usually are drilled fromwithin the mine workings, perpendicular to theface cleat, and generally have high rates of gas

drainage. The gas recovered from the boreholesis pumped through a separation unit to removeassociated water before the gas is piped out ofthe mine.13

The main difficulty in drilling horizontal holesis keeping within the coal seam. Consolidated

Coal Co. (Consol) has developed a mobile hori-zontal drilling system with special features for

methane production. Three- to four-inch diame-ter holes may be drilled to lengths greater than2,000 ft using a guidance system to keep the bitwithin the seam and methane is piped out through

closed-loop plastic pipes.Horizontal wells may partially escape depen-

dence on the mining operation with a system thatuses horizontal holes drilled radially from the bot-tom of a central vertical shaft. However, the ex-

pense of the shafts may dictate that the wholeoperation can succeed financially only if theshafts can be re-used later on for the mining oper-ation; thus, it is not clear that this drilling systemactually will sever the tie between mine and gasrecovery operation. This method has not yet beentried in the United States.

Stimulation

Where low permeabilities area problem, stimu-lation is used to increase the flow of gas to the

well by increasing the area of the natural frac-ture system in contact with the wellbore. Hydrau-lic fracturing is the most common stimulationtechnique used. As with such treatments in the

——-.13A major (but no nt e ch n i c a l) roadblock tO s u b s t a n t i a l prod uc-

tion of methane from these types of horizontal boreholes is that pro-

duction may be dependent on the mine operation, If the mine were

to close, the methane recovery operation might also be forced to

cease. For example, at Kerr McGee’s Choctau Mine i n the Arkoma

Basin, the coal seam methane production operation was completelyset up, equipment installed, and approvals acqulreci when the mine

was closed for lack of a coal market, Financing for a methane

recovery project that is dependent on mine operation will tend to

be difficult to obtain.




SOURCE: Gas Research Institute.

Devonian shales, the most widely used fractur-ing fluids are nitrogen foam and water-based gel.

A sized proppant may be included to hold openthe newly formed fractures. The amount of eachingredient used depends on the fracturing fluidpressure required, the coal seam thickness, thefracture length desired, and the cost. When thefluid injection is completed, the induced pressureis released and the well prepared for production.

The effects of hydraulic fracturing may be en-

tirely different in coals than in sandstone reser-

voirs. An induced fracture in sandstone will typ-

ically extend outward at substantial length fromthe well bore. Coal formation fractures are gen-erally shorter and wider than sandstone fractures.

The difference is attributed to the plasticity of thecoal and dissipation of the compressional energyinto the cleat system. ’4

Several problems may be encountered in stim-ulating coalbed wells. One is the tendency forproppant material to flow back into the well boreand create pump malfunctions during dewater-ing. Another problem is orienting the fracture tointersect rather than parallel the planes of the ver-tical fractures, or face cleat, in order to intersectas many fractures as possible. This may be diffi-

cult because the original stress field in the coal——

14M. G. Doherty, “Methane From Coal Seams, ” InternationalConference on Small Energy Sources, Los Angeles, CA, Sept. 9-18,1981.




clearly favored fracture directions parallel to theface cleat. A third problem, pertaining to min-able seams, lies in containing the fracture withinthe seam. Some mine operators feel that fractur-ing can cause structural damage to the roof rock,increasing the potential for mine collapse. Ongo-

ing work by the U.S. Bureau of Mines is attempt-ing to evaluate the extent to which this concernis valid. 15

15 M A T r e v i t s , M . E. H a n s o n , and V. L. Ward, ‘‘Methane Drain-. ,

age: Identification and Evaluation of the Parameters Controlling in-

duced Fracture Geometry, ” SPE/DOE Unconventional Gas Recov-

ery Symposium, May 16-18, 1982.

RECOVERABLE RESOURCES

The coalbed methane resource base is large

but, like the other unconventional resources, therecoverable portion is significantly less than thegas-in-place. Economic and technological condi-tions are the primary factors governing theamount of the resource that will contribute tofuture supply. Other factors, such as legal andenvironmental issues, also are likely to influencecoal bed methane production, particularly fromminable coal beds. Even without the hard-to-predict effects of these other issues, however, theuncertainty associated strictly with technicalissues is high. The NPC, in describing its estimatesfor the recoverable gas resource, calls them “a

qualified and educated guess, ” and “nothingmore than an order-of-magnitude projection basedon current information."16 Although the scien-tific understanding of coal bed methane produc-

lbNdtiO~dl Petroleu M COU ncil, Unccmventiona/ Gd s $XJfCe5: COA/

Seams, June 1980.

Pumps

Water removal can in general be accomplishedwith existing pumping technologies, but the largeamount of water that is produced during dewater-ing is a strain on pumping equipment and fre-quent maintenance is often required. pumps arealso apt to become clogged with the coal finesremaining in the well after drilling. Dewatering

deeper wells may require the development oflarger capacity pumps. Nevertheless, solutions to

these problems are more a matter of refinementof existing technology than radical innovation.

AND PRODUCTION POTENTIAL

tion has improved in the last 4 years, no estimates

of the economically recoverable resource havebeen made since the 1981 GRI estimate.


Estimates of the recoverable resource basehave been made by the NPC (1 980), Kuuskraa

and Meyer (1 980), and GRI (1 981). A variety ofdifferent economic and technology assumptionswere used to obtain the estimates shown in table

59.

National Petroleum Council17

[n the NPC study, recoverable resources werecalculated by identifying that portion of the re-source that would yield sufficient production perwell to cover the costs of an “average” well (with

—.—.—171 bid.

Table 59.—Comparison of Recoverable Resource Estimates

Technically or economicallyrecoverable gas (TCF) Assumptions

KM . . . . . . . . . . . . . 40-60 30-45°/0 recovery of target resource of 135 TCF, no price constraints, butrecoverable resource limited to bituminous seams >3.5 ft thick,subbituminous seams >10 ft thic k

NPC . . . . . . . . . . . . $2.50 $5.00 $9.00 Production costs define minimum production levels,

5.0 25 45 10%\o ROR which in turn define minimum economic thickness at 32.5 20 38 150/0 ROR MCFIDlft (bituminous), 1,2 MCF/D/ft and 0.6 MCF/D/ft2,0 17 33 200/0 ROR (subbituminous and lignite, respectively) Gas is used

onsite,GRI . . . . . . . . . . . . $3.00 $4.50 $9.00 Expert judgement

10-30 15-40 30-60 Existing-advanced technologiesSOURCE Off Ice of Technology Assessment

3 8 - 7 4 2 0 - 8 5 - 1 5



218 qU.S. Natural Ga s Availability: Gas Supply Through the Year 2000

a depth of 3,000 ft, 12-year well life, assumed 10percent production decline rate, and 90 percentsuccess rate). This was done by the followingmethod:

1.

2.

3.

Using county-by-county coal resource esti-mates, a distribution of the coal-in-place re-

source according to seam thickness was con-structed for each grade of coal. The distribu-tion for bituminous coal is shown in figure50 as a plot of cumulative coal-in-place v.

minimum total seam thickness.By examining data on production rates perfoot of seam thickness for existing wells, val-

ues of 3 MCF/D/ft (bituminous), 1.2 MCF/D/ft

(subbituminous), and 0.6 MCF/D/ft (lignite)were estimated for the production rates perfoot from the coal resource in place. Then,for each grade of coal, minimum total seam

thickness was converted to minimum “per

well’ ) production rate. This rate can, in turn,

be converted into minimum gas price nec-

essary to pay for the well.

It is assumed that gas recovery will be 50 per-

cent of the total gas-in-place in coal seamssatisfying the minimum thickness criteria,and that a random 10 percent of the coal-in-place will not be available for drilling.18

——.—16Neither of these values are further substantiated i n the report.

Figure 50.— Estimated Distribution of BituminousCoal by Seam Thickness

0 10 20 30 40 50 60

Minimum total seam thickness (feet)


4.

Using these values and the assigned valuesof gas content (200 cubic feet per ton forbituminous, 80 CF/t for subbituminous, and40 CF/t for lignite), cumulative coal-in-placecan then be converted to recoverable gas re-source.

The final result is a relationship between therecoverable resource and gas price. The esti-mated recoverable gas resources for threegas prices, assuming the gas to be used on-

site without compression, are shown in table59.

The report does not give results for the casewhere the gas is scrubbed, compressed, andgathered for delivery to a pipeline. However,comparison of plots of gas price v. necessary pro-

duction rates for onsite use and pipeline sales(figs. 4 and 5 in the NPC report) imply that pipe-line delivery will add approximately $1.00/MCF

to production costs. The actual effect on pro-ducer incentives is not clear, however. On the

one hand, it is not uncommon for pipelines topay for gathering and compression costs, whichreduces the gas price required by producers tomake a profit. On the other hand, for existingcoal bed methane projects, initial compressionand gathering cost generally have fallen on theproducers. 19

Kuuskraa and Meyer20

A large portion of Kuuskraa and Meyer’s gas-

in-place estimate of 550 TCF was recognized asbeing within coal seams that were too thin orwhose gas content per unit volume was too lowto exploit. Assuming the favorable resource to oc-cur only in bituminous coal seams greater than3.5 ft thick and sub-bituminous seams greaterthan 10 ft thick, Kuuskraa and Meyer estimated

that about 135 TCF of methane is present in themost favorable coal seams. The technically re-

coverable resource was determined to be 30 to45 percent of the favorable resource, or 40 to 60TCF, based on calculations of the amount of gas

Iqvello KUUSkraa, LeWin & Associates, Inc., persona! cornrnuni -

cation, 1984.

ZOV. A. Kuuskraa and l?. F. Meyer, ‘ ‘Review of world Res ou rc es

of Unconvent ional Gas , ” IIASA Conference on Conventional and

Unconventional world Natural Gas Resources, Luxenburg, Austria,

June 30-)uly 4, 1980.



—


Figure 51 .—Well Production Histories in San Juan and Other Basins




bituminous coals range from 45 to 20 ft, cor-responding to production levels of 135 to 60MCF/D. The NPC data on actual wells, however,indicate that seam thicknesses in all cases exam-ined were less than 25 ft and most were less than10 ft, with production rates in all cases less than

70 MCF/D. Although the sales price of the pro-duced gas and the profitability of the wells is notknown, presumably some of these wells are prof-itable, and it does not seem likely that the prices

paid for this gas could be much above the givenrange. This implies that the cost of these wellsmust have been lower than the NPC’s calculatedaverage well costs.

On balance, the examination of uncertain as-sumptions in the NPC study appears to indicatethat their analysis may have been overly pessi-mistic.

The Kuuskraa and Meyer analysis differs sub-stantially from the NPC analysis, especially be-cause it calculates recovery efficiency from an

analysis of diffusion from the fracture network

rather than assuming a recovery efficiency. This

exposes the KM analysis to some different kindsof uncertainties than those encountered by NPC.In particular, as noted in the earlier Lewin re-por t , 28 the resuIts are extremely sensitive to

assumptions about the fracture intensity in theseams and the diffusion constant. For example,for Western coals, a change in the spacing be-tween vertical fractures, from 1- to 5-ft intervals,

reduces the 10-year recovery efficiency from 30to 2 percent, essentially eliminating the economicrecovery potential from these coals. Fracture in-tensity is not well-documented, especially for

deeper coals.

Another potential problem with the KM anal-ysis is that it is uncertain whether or not its sim-ple diffusion model adequately represents the ac-tual physical production mechanism in the coalseam. For example, the model and associated as-sumptions imply uniform production behavioracross the seam, whereas in reality production

behavior in existing coal seam methane projects(i.e., the Black Warrior development) has fluc-

tuated widely from well to well .29 This implies that

we do not yet fully understand the gas produc-tion mechanism.

Because of the substantial remaining uncertain-ties and the lack of recent economic analyses thatcould take into account the latest understandingof the nature of the coal seam methane resource,OTA is reluctant to project a new estimate of therecoverable resource, However, in our opinion

the NPC estimates for moderate prices–e.g., 2.5TCF (at 15 percent rate of return) for gas pricesof $2.50/MCF in 1979$ ($3.35/MCF in 1983$)—are overly pessimistic, and are based on past ex-

perience that does not reflect recent productioncapabilities associated with improved operatingpractices such as closer well spacing. What is crit-ically needed is a reevaluation of the economicsof recovering this resource given our better un-

derstanding of the resource and improved pro-duction methods. Information that would helpsuch an estimate is a disaggregation of the re-source base based on gas content as well as seamthickness. The data collected for the DOE basinanalysis may be sufficient to provide the basis fora new analysis along these lines.


Both NPC and GRI calculated annual produc-

tion estimates. NPC estimated annual productionthrough the year 2000 for both production from

vertical wells and from shafts with horizontalholes. The vertical well development scenario ex-tends over a drilling period of 18 years withrecovery of the resource for 28 years. The devel-opment schedule requires beginning with 80 rigsthe first year and adding 80 rigs in each of thenext 7 years, with each rig drilling 45 producingwells per year. The resulting annual productionestimates are included in figure 52.

For production from shafts with horizontalholes, the NPC assumed that 50 shafts would bedrilled in the first year, with a 20-percent increasein the rate of adding new shafts every year, At

this drilling rate, a 22-year program would be re-quired to recover the total projected gas resource

2 9 v e l [ o K u U s k r a a , l-e~~,l n & Associates, Inc., personal com m u n i-

cation, 1984.

3 8 - 7 4 2 0 - 8 5 - 1 6




Figure 52. —Annual U.S. Production Rates ofCoalbed Methane as a Function of Time—

Vertical Wells

Incremental price ($/MMBtu) at 10°/0 ROR

o 3.50 5.00 7.00 9,00

1985 1988 1992

Year

SOURCE National Petroleum Council

in a period of 35 years. The annual

rates are depicted in figure 53.

996 2000

production

GRI used the recoverable resource estimatesacquired from the poll and further assumed pro-duction and drilling rates. Production rates wereestimated at 30 and 100 MCF/D for existing andadvanced wells, respectively. The number ofwells drilled per year was assumed to be 200 from1983 to 1986 with a 10 to 15 percent increase

per year thereafter. Utilizing the resource base

as a limit, annual production estimates were cal-

culated. The 1990 and 2000 production estimatesare included in table 60.

The NPC study assumes that the gas price willincrease steadily to $9/MCF (1 979$), Given cur-

rent and expected future market conditions, thisassumption appears unrealistic. On the otherhand, if, as it appears, a majority of the recover-

Figure 53.- Annual U.S. Production Rates ofCoalbed Methane as a Function of Time—

Horizontal WellsIncremental price ($/MMBtu) at 10°/0 ROR

o 2.50 3.50 5.00 7.00

3.0

1.0

0.5

01985 1988 1992 1996 2000

Year

SOURCE: NPC.

able resource can be recovered at prices on the

order of $5/MCF (1979$), and if little is requiredin the way of technologic advances, increasing

levels of production might still be expected. Whatis required is an increased level of producerinterest-interest which at present is constrainedby questions of ownership, mine safety, environ-mental concerns, and other institutional consid-erations. Some of these issues are discussedbelow.

Legal Constraints

Court decisions to date have attempted to re-solve several legal questions associated with coalseam methane production, but without muchsuccess. Previous litigation has centered on theissue of resource ownership and whether themethane is a resource in its own right or an in-

trinsic part of the coal. The U.S. Steel v. Hoge




Table 60.–GRI Coalbed Methane Annual Production Estimates (TCF)

1990 2000

Market price Existing Advanced Existing Advanced

(1979$/MCF) technology technology technology technology

$3.00 . . . . . . . . . . . . . . . . . 0.06 0.22$4.50 . . . . . . . . . . . . . . . . .

0.29 0.950.07 0,23 0.35 1.2

$6.00 . . . . . . . . . . . . . . . . . 0.07 0.24 0.42 1.4

SOURCE GRI

case of March 1980 set a precedent on both theseissues. Although lower court decisions deter-

mined that methane is a natural gas occurring in

coal, and that the land owner owns the meth-ane until the gas rights are released, in Decem-ber 1983 the Pennsylvania Supreme Court re-versed that decision, remanding ownership to thecoal owner.

The recovery of coalbed methane on Federal

coal lands also is burdened with unanswered

legal questions. The current position of the So-licitor’s Office of the Department of Interior is thatownership of a coal lease does not include rightsto the coal bed gas, but that a reservation of gasdoes,30 and that coalbed gas is leasable under theoil and gas leasing provisions of the Mineral Leas-ing Act. This position has not been tested in thecourts, however. Drilling permits for coal seammethane recovery have been issued, although

administrative delays are a problem.

Environmental Constraints

The primary environmental issue associatedwith coal seam methane production is disposalof the water produced with the gas in those coalseams characterized by high water contents.Since dewatering is a primary production require-ment, large volumes of water must be pumpedfrom the subsurface. The quality of the water

varies from slightly acidic to slightly alkalinedepending on the site location. The environ-mental regulations of the State determine wheth-er the water must be treated, and such decisionswill influence the economic viability of the re-covery project.

3 f 3A d e t a i l e d s u m m a r y o f t h e l e g a l situation IS p r e s e n te d In j. H.

Kemp, “Coalbed Gas: Recent Developments in the Ownershlp andRight to Extract Coalbed Gas, ” The LJdmafl, November 1982.

Institutional Barriers

There are other factors that will influence the

contribution of coal bed methane to futuresupply. Institutional barriers characteristic of thecoal industry will deter or possibly preclude pro-duction in some instances. A primary institutionalbarrier is the lack of interest exhibited by the coalcompanies. According to industry analysts, sincethe companies’ primary interest is coal mining,they tend not to want to become involved in the

more long-term nature of the methane produc-tion industry, particularly when the economicsare marginal. Investment incentives may be re-quired to create interest in producing the meth-ane rather than venting. Alternatively, if new anal-yses demonstrate a real economic advantage toproducing gas prior to mining, the coal industrymay become more interested in overcomingproblems created by the mining schedule. Oneproblem with degassing prior to mining is theshort time period between the beginning of gasproduction and the mine opening that has oftenbeen allowed. Numerous wells are required, in-

curring high capital costs which cannot be re-couped without a longer production period.

Another institutional barrier to production is thestrong concern with worker safety associated withcoal mining. The issue of whether stimulationcauses unacceptable damage to the mine ceil-ing has not been resolved. The Bureau of Minesinitiated a program at four sites to determine theeffects of stimulation. Due to various problems,however, work was completed at only one site.The site evaluation indicated that there were noadverse effects on the mining operation, but the

limited extent of the test precludes extrapolationof the results to other sites. No firm evidence ex-

ists to link stimulation effects to mine collapse at

numerous other operating facilities, but until the




technique is proven not to cause damage, many able resources, imply that the current productioncompanies will be reluctant to invest. projections are not adequate, and effort should

be focused on establishing a new, more scien-In OTA’s opinion, the above uncertainties, tifically based estimate of the potential contribu-

coupled with the technical and economic uncer- tion of coal seam methane to future U.S. gastainties mentioned in the discussion of recover- Supply.



Appendixes



Appendix A

Calculation of Additional Reserves FromIncreased Gas Recovery in Old Gasfields

OTA calculated the effects of higher gas prices on

gas recovery in “old” gasfields by modifying a previ-ous analysis of those effects conducted by the Shell

Oil Co.1 The OTA analysis is discussed in detail in arecent OTA staff memorandum and is summarizedhere.

The analysis focuses on the expansion of “old gas”reserves, which are defined here as all reserves thatdo not qualify for “new natural gas” status under the

Natural Gas Policy Act (NGPA). In general, old gasis gas in reservoirs that were discovered (and reservesreported) prior to about mid-1977; however, the pre-

cise boundaries are more complicated than this.

Shell’s Analysis

The Shell study assumed that all old gas would re-main at low prices under the NGPA and would riseto $3.50/MMBtu, the assumed free market price,

under a price decontrol policy. Shell calculated theeffect of a $3.50 gas price on recovery in the Nation’s

old gasfields by the following method:1. Calculate the Nation’s “responsive reserves,” that

is, the old gas reserves that might grow if their pricesgo up. Some reserves, such as Alaskan North Slopegas, gas dissolved in oil, and, to a lesser extent, gasin water-drive reservoirs will not respond much to agas price increase and were not included in the anal-ysis of reserve growth. For example, gas dissolved in

oil is responsive primarily to oil prices, because thevalue of the oil in the reservoir far outweighs the valueof the gas. If oil prices go up, more oil will be pro-

duced and thus more gas will be co-produced with it.Shell’s estimate of responsive reserves in 1981: 115

TCF.2. Calculate reserve growth in sample fields where

adequate data are available. Shell evaluated the ef-fects of a price increase to $3.50 on lower abandon-ment pressures and well reworkings, infill drilling, and

well stimulation for 14 large sample fields. The lowerabandonment pressure calculation involves comput-ing the gas flow that will produce revenues equal to

operating costs3 for the new and old gas prices. The

difference in reservoir pressures corresponding to the

‘ C S Matthews, Inc rease /n Untted States ‘‘old” Gas Reser~e5 Due to

Oeregu /a t i on , Shell 011 Co , April 1983

‘Office of Technology Assessment, Staff ’ Mem orandum on th e Etiects of

Decontro l on ( Id Ga s Reco~ery, February 1984

‘Th IS IS approxl mately the abandonment poI nt tor the well, 51 nce profits

are zero at this poI nt

“new” and “old” flows, and the additional reserves

corresponding to this pressure difference can then becalculated by using the physical gas laws. The infilldrilling calculations were made using reservoir simu-lation and extrapolation from previous infilling experi-ence. The well stimulation calculations are based onan engineering judgment that an additional 1.5 per-

cent can be added to ultimate recovery by this means:

Reserves remaining in sample fields . . . . 41.3 TCF

Reserve growth, lower abandonment pressures

a n d w e l l r e w o r k i n g . 9.6 TCF

Reserve growth, infill drilling ... ., . . . . . . 7.6 TCF

Reserve growth, well stimulations = 1.5 percent of ultlmaterecovery 4

3. Scale up the sample results to the Nation, assum-ing that, except for well stimulations, the results will

scale by the ratio of the remaining reserves:

Scaling factor = 11 5/41.3 = 2.8

National reserve growth for lower abandonment pressures and

well reworkings = 9.6 X 2.8 = 27 TCF

National reserve growth for infill drilling = 7.6 X 2.8 = 21 TCF

By examining available production records and esti-

mates of remaining reserves, Shell estimated that theultimate recovery represented by the 115 TCF of re-sponsive reserves is 475 TCF, thus:

National reserve growth for well stimulation =

0.015 x 475 = 7 T C F

Assuming that 3 TCF of the infill drilling would occuranyway at presently available prices (an incentiveprice of $2.75/MMBtu in mid-1983),

Total national reserve growth due to higher prices =

27 + 21 + 7 – 3 = 52 TCF

OTA’s Modifications to Shell’s Analysis

OTA has made a number of modifications to Shell’soriginal calculations based on a detailed review ofShell’s methodology, an evaluation of alternative datasources, a review of literature on infill drilling andother topics related to gas recovery, and a numberof telephone interviews with geologists and petroleum

engineers. The most important of the modificationsare:

1. Scaling to the Nation. OTA determined that an

appropriate scaling factor should be related as closelyas possible to the original volume of gas in the fields.The basis of Shell’s scaling factor, remaining reserves,is tied closely to the production history of the fields.

‘Ultlmate recovery = cum u Iatll e production plus remal n I ng rwerl es

227




Ultimate recovery, on the other hand, is more directlyrelated to the original gas volume and is a more appro-priate basis for the scaling factor. Using Shell’s owncalculations, the use of ultimate recovery as the basisfor the scaling factor yields an increase in the expectednational reserve growth of 36 percent (all else being

equal).2. Responsive reserves. As noted above, Shell as-

sumed that all old gas would remain at low pricesunder the NGPA, so that, for the calculation of respon-sive reserves, Shell estimated the total old gas reservesand subtracted only those reserves that would be

physically unresponsive to higher prices. However,the NGPA provides for the decontrol or price escala-tion of most old intrastate reserves by 1985, and thedecontrol by 1985 or 1987 (depending on depth) ofall gas from infill wells in old intrastate fields. Conse-quently, these reserves will receive a high decon-trolled price whether or not any additional decontrolmeasure is passed, and thus are not “responsive” to

such a measure . . . . they should be subtracted fromShell’s calculated responsive reserves. Shell also madesome minor errors in its original calculation of totalold gas reserves; it treated all “extensions” added toreserves since 1977 as old gas, whereas some of these

reserves qualify as “new” NGPA Section 102 gas and

should have been excluded from the calculated totalof old gas reserves.

Data on the amount of reserves in each NGPA cat-

egory are not available. OTA used data on reservevolumes in interstate and intrastate commerce, in-

terstate pipeline purchases by NGPA category, andlimited production data by NGPA category to estimatethe volume of old gas reserves in each category, and

the volume of responsive reserves. Our estimate ofresponsive reserves was 63.4 to 71.4 TCF for lowerabandonment pressures and well stimulations, and 59to 66 TCF for infill drilling, as compared to Shell’s 115TCF estimate for each category. Consequently, all else

being equal, Shell’s results are overstated by the ratio

of “incorrect” to “correct” reserves, or by a factorof about 1.6 to 2.0.

3. Abandonment pressures. Shell’s estimates of thecurrent abandonment pressures in its sample fieldsgenerally are considerably higher than the estimatesof alternative analysts, for example, the American GasAssociation’s Committee on Natural Gas Reserves. Ahigher current abandonment pressure implies a larger

growth potential, so applying the alternative, lowerpressures would yield a lower estimate of the addi-

tional reserves available from the growth of olderfields. Specifically, applying the alternative pressure

estimates in those fields where such estimates areavailable more than halves the estimates of growth po-

tential, from 8.1 TCF to 3.0 TCF. The uncertainty asso-ciated with these alternative abandonment pressureestimates was factored into OTA’s estimates of fieldgrowth potential.

4. Infill drilling. A key point of contention withShell’s analysis of infill drilling is the extent to whichthe potential reserves may be available at today’s

prices without any legislative changes. Shell’s predic-

tion that only 3 TCF of a 21 TCF potential would beforthcoming at today’s prices is based partly on itsassumption that the low level of infill drilling activityof the past few years must reflect a lack of economicprospects. However, a variety of factors other than aninadequate price may have played a role in the cur-rent inactivity. These factors include the current gassurplus, opposition by pipelines or consumers, s op-position by other producers in the same field,6 and

State prorationing rules that prevent producers fromincreasing production rates. OTA’s discussions withproducers have lead us to believe that more than 3

TCF of the total infill potential would eventually bedrilled at current prices. The range of infill potential

in Scenario 1 reflects the possibility that as much asone-third of Shell’s “after decontrol” infill potentialcould occur eventually without any further legislativechange.

5Be( a use the current Inflll Incentive price of about $2 .85/MMBtu appltes

to all gas from the Inflll well, Including gas that could have been produced

from adjacent wells at a lower price.

6Because the potential for drainage across the field means that the other

producers would have to lnhll also or face the 1 0 s 5 of some of their gas.



Appendix B

Fracturing Technologies

Introduction

A characteristic of “unconventional resources” is

that, while increased prices generally are an impor-

tant condition for full commercialization, new tech-nology developments are also required. In the longrun, technology may have more impact than price onthe amount of gas recovered. Studies of unconven-tional resources that OTA has reviewed have con-cluded that the amount of gas that could be producedwith existing technology at prices considerably higherthan today’s is less than the amount of gas that could

be produced with advanced technology at currentprices for new gas. For example, the National Petro-leum Council’s tight gas study estimates that more gas

can be produced for $3.00 per thousand cubic feet(MCF) with advanced technologies than can be pro-

duced for prices up to $9.00/MCF using base case

technology. In response to this perception, a consid-erable amount of Government and industry researcheffort has gone into developing more advanced tech-nologies for unconventional gas recovery,

In the past 5 years the state of technological devel-

opment has advanced. With and sometimes without

additional Government financing, producers havebeen willing to try innovative approaches. Neverthe-less, a high failure rate still exists in probing certain

types of unconventional gas formations. The follow-ing discussion describes successful new developmentsin fracturing technologies and delineates areas wherework remains to be done. This appendix will serve togive the reader more insight into the validity of the

assumptions used in the various estimates of recover-able resources and production potential.

New Technology Developmentsin Fracturing

The objective of fracturing a low-permeability reser-voir is to increase the surface area of the formationthat is in direct contact with the well bore. The pres-sure gradient between the lower permeability forma-tion and the higher permeability fractures is the driv-ing mechanism for the gas flow. Thus, the greater the

area over which such a gradient can be established,the larger the volume of gas flowing at a given point

in time.Technologies for fracturing gas reservoirs are not

new. Explosives have been used in Devonian shales

since the late 1800s. Detonation of explosives shat-ters the rock immediately around the well bore, effec-

tively increasing the well bore diameter. A large-scalevariation on this theme was tried in the late 1960s intight sandstone formations using nuclear explosives.The generally unsatisfactory results (possibly due to

melting of the reservoir rock from the heat of explo-sion or permeability damage due to compaction offine particles) and the lack of public enthusiasm for

potentially radioactive gas put a quick end to thisprogram.

Hydraulic Fracturing Technologies

Hydraulic fracturing is the most commonly usedfracturing technology today. The concept was first de-

veloped in the 1940s for use in conventional oil and

gas reservoirs. In the early 1970s producers began toincrease the size of the fracture treatments to gener-

ate longer fractures, on the order of 1,000 ft in low-

permeability sandstones.1

This technique, known asmassive hydraulic fracturing (MHF), is now a commonmeans of well completion in the tight sand formations,although it is certainly not applicable in all cases.Devonian shale and coal seam reservoirs use more

conventional size hydraulic fracture treatments to

create fractures ranging from 100 to 500 ft. Fracturesin these types of reservoirs are designed to intersectnatural fractures, which serve as the primary pathwaysfor gas flow.

Hydraulic fractures are created by pumping largevolumes of fluid down the well bore. The fluid exerts

pressure on the rock formation, eventually creatinga fracture. Fluids generally carry proppant materials,

such as clean coarse sand, which are left in the frac-tures and hold them open when the fracturing fluid

is removed. The induced fracture has a considerablyhigher permeability than the surrounding formation.

Hydraulic fractures tend to be unidirectional, gener-ally extending out as wings in opposite directions fromthe wellbore. By convention, their length is measuredalong one wing. Their direction and orientation (ver-tical, horizontal, or inclined) are controlled by the re-gional tectonic forces and the depth of the target for-mation. Most fractures at depths greater than 2,000ft are oriented in the vertical plane. Figure B-1 sche-

matically represents a hydraulic fracture.

Major research efforts in fracturing technology have

focused both on increasing the fracture length and

‘C. R. Fast, G. B, Holman, and R. j. Corlln, ‘‘The Application ot Massive

Hydraullc Fracturing to the Tight Muddy j FormatIon, VVattenberg Field, Col-

orado, ” )ourna/ of Petro/e urn Tec hnology, january 1977, pp. 10-16

229




Figure B-1 .—Conceptual Fractures Created by Massive Hydraulic Fracturing in-Blanket and Lenticular Formations

FI

t

rBlanket formations

Fractureheight

1

Wings

Wellbore



App. B—Fracturing Technologies 231

solving some of the problems that reduce the effec-tiveness of the fracture in increasing rates of gas flow.

Maximizing the effective length of a fracture is not

simply a function of increasing the volumes of fluid

pumped into the well bore. One must make sure thatthe proppant is transported to the end of the fracture,

and effectively holds the fracture open once the frac-turing fluid is removed. Other problems that must beovercome include minimizing damage to the forma-tion caused by the fracture fluids and containing thefracture within the “pay interval,” the layer where gasis present. If the fracture intersects a permeable zone

that allows the fracturing fluid to “leak off” at a highrate, further penetration of the fracture may becomeimpossible because the fluid loss prevents further pres-

sure from being built up.Each of the objectives of fracture research will be

discussed in turn:Maximizing High Conductivity Fracture Length.–

Accomplishing this objective involves choosing appro-priate fracture fluids and proppants. The proppants

should be strong enough not to crush as the fracturecloses, and of sufficient diameter to overcome anytendency to become embedded in the formation.Also, they must be light enough to be carried by the

fracturing fluid to the design fracture length withoutsettling.

Conventional practice is to use clean rounded sandas the proppant material. It is the least expensive prop-pant and, at shallow and intermediate depths, has suf-ficient strength to hold the fracture open withoutcrushing. At greater depths where closure pressuresare higher and proppant crushing prevalent, a strongermaterial is needed.2 Producers most commonly use

sintered bauxite under these conditions. However,bauxite has two drawbacks–high cost and high den-sity. Because of the latter, it is difficult to transport thebauxite particles to the end of the fracture. Servicecompanies are rapidly developing alternate interme-diate and high strength proppant materials. A num-ber of these materials, including ceramic beads andresin-coated sands, have lower densities than baux-ite and appear to have sufficient strength for most frac-ture applications. More work needs to be done to de-velop materials with densities lower than sand and

adequate strength to maintain high fracture conduc-tivities.

The need for a fracture fluid with a high capacity

to carry proppants in suspension has led to the de-velopment of very sophisticated fluids. These include

‘R. A. Cutler, D. O Ennlss, A. H Jones, and H. B. Carroll, ‘‘Compa risen

of the Fracture Concfuctlvity of Commercially Available and Experimental

Proppants at Intermediate and High Closure Stresses, ” SPE/DOE Syrnp osIurn

on Low Perrneab l/lty Gas Reservoirs, SPE/DOE 11634, 1983.

water- and hydrocarbon-based polymer liquids andgas-charged emulsions and foams.

The water-based fluids use organic polymers for fric-

tion reduction, fluid loss control, and viscosity en-hancement. The polymers are long chains of organicmolecules which bond loosely with the water, form-ing gels. The resultant fluid is thicker than water and

has a higher surface tension. It flows with less tur-bulence, can suspend greater volumes of proppants

and does not leak off into the formation as rapidly aspure water.

Probably the most significant technical developmentin fracturing fluids is the process of cross-linking.Cross-linking is a chemical reaction which bonds

polymer chains together, effectively increasing the

viscosity of the fluids as much as an order of magni-tude. The reaction is timed so that cross-linking oc-curs just as the fluid arrives at the fracture entrance.The increased pressures required to pump the thickerfluid will widen the fracture and the enhanced viscos-ity can carry the proppant greater distances. At theend of the treatment the fracture fluid warms up tothe higher reservoir temperatures and the cross-linkedpolymers break down. Now significantly lower in vis-cosity, the fluid can leak off into the formation or flow

back out of the wellbore, leaving the proppant in place.Hydrocarbon-based fracture fluids behave similarly

to water-based fluids and can also be cross-linked.They are used in instances where water-based fluidsare likely to cause significant formation damage—asin the presence of water-sensitive clays. However, ingas-bearing reservoirs the introduction of a third phase(oil, in addition to gas and water) may further impedethe flow of gas in the formation.

Minimizing Formation Damage.—In addition tofluids designed to improve proppant transport, moreexotic fracture fluids have been designed to addressthe problem of formation damage. Fracture fluids havebeen a major factor in causing formation damage.Fluid leak-off into the formation can block pores, espe-

cially if the gels are not completely broken down. in-troduced fluids can also cause clays to swell, or dis-lodge fine particles to block pores.

All three types of unconventional gas reservoirs are

susceptible to formation damage. Devonian shalesmay be the most affected because they have naturallylow water content and high clay content. Devonian

shales have, as a consequence, served as a testingground for a number of the new fracturing fluids.

Many of the fluids developed to minimize forma-tion damage use a gas phase to reduce the amount

of water required. Foamed fluids are gas-in-wateremulsions, where the surface tension of the bubblesholds the proppant in suspension. Such fluids cannot




transport large volumes of proppant long distance—they are generally used for shorter fractures. Nitrogen(N,) is the most common gas used in foamed fluids,although CO2 can also be used. Producers are experi-menting with increasing the percentage of gas from75 to 90 percent of the total fluid volume.

Pure N2 has also been used as a fracturing fluid. Itis not an efficient fracturing fluid as it requires very

high injection pressures. However, nitrogen fractur-

ing has proved very effective in increasing gas flowbecause it does not adversely affect the formation. Ni-

trogen gas cannot carry proppants, therefore it is onlyeffective at shallow depths where the fractures are lesslikely to close. Whether wells fractured with nitrogenwill maintain higher production levels over the longterm is still unknown.

Some wells have been fractured using liquid C02,which has the ability to transport proppants. As theliquid C02 warms, it reverts to the gas phase and easilyflows back out of the hole with minimal damage to

the formation. Liquid CO2 fracturing is a relatively ex-

pensive process and somewhat more dangerous to usethan foamed fluids, In addition, the casing and pump-ing materials must be capable of withstanding very lowtemperatures.

The tradeoffs of minimizing formation damage,transporting proppants, and containing costs all en-ter into the decision of which fracture fluid is used.Generally, an important element in the decisionshould be laboratory compatibility tests between for-mation cores and the proposed fracturing fluids, which

can identify potential damage problems.Containing the Fracture within the Pay lnterval.–

As a fracture propagates outward from the wellboreit may also grow vertically up or down, Vertical growth

occurs at the expense of lateral growth, thus reducesthe effective length of the fracture for the same vol-ume of fluid pumped. Those portions of the fracturewhich extend above and below the gas-producing in-

terval are essentially wasted; also, they may extend

into water-bearing strata which will adversely affectgas flow.

It has been recognized in the last few years that the

main factor controlling containment of a fracture isthe difference in the stress characteristics of the rocksmaking up the producing and nonproducing zones.

The stress on the rocks, or “in-situ stress,” is a func-tion of the mechanical properties of the rock and theregional stresses acting on the rock. Thus, the same

type of rock at different locations or depths or in dif-ferent tectonic environments may have different in-

situ stress characteristics. A sand-shale interface in theCotton Valley Sands may effectively contain a fracturewithin the sand zone. A fracture in the Piceance Basin

may break through a similar sand-shale interface. Sim-ilarly, different rocks will have large differences in theirmechanical resistance to fracturing and thus requiresubstantially different applied pressures for fracturing.

No commercially available technologies exist today

that successfully deal with fracture containment. Somerecent research efforts have focused on developing

innovative techniques to control the growth of frac-tures out of the pay zone. GRI is testing three ap-proaches: 3

q

q

q

Fracture initiation placement–the well casing is

perforated above or below the pay zone allow-ing the fracture to grow vertically into the pro-

ducing interval. A field test of this technique wasperformed in July 1983.4 Preliminary results in-dicate increased flow but further cleanup is nec-essary before final results can be assessed.Controlled process zone–the fluid viscositiesand pumping rates are controlled to get prefer-

ential initial leak-off in the pay zone. This shouldresult in more penetrating rather than taller frac-

tures. This technique is still being tested in thelaboratory.Lightweight additives–impermeable floatingproppants are used to seal off upper, non-produc-ing portions of the fracture. Appropriate proppantmaterials are currently being tested.

Predicting and Monitoring Fracture Behavior.–

Another important research objective in improvingfracturing technology is to develop techniques to pre-

dict and monitor fractures. To know in advance howa fracture is likely to perform or to be able to tell inthe field whether a fracture is conforming to designparameters increases the chances that stimulation will

be successful. Furthermore, as fields become more de-veloped, it is important to know the direction andlength of a fracture, and thus the drainage area of awell in order to minimize interference from subse-

quent wells. Fracture diagnostics probably is one areawhere the most innovation has occurred in the last

few years.Predicting fracture behavior.–State-of-the-art pre-

diction of fracture behavior comes mostly from for-mulation of sophisticated mathematical models andcomparison of the model results with results of lab-oratory experiments. There has been little field verifi-cation because of the cost and technical difficulty in-volved in obtaining a detailed picture of the physicalresults of fracturing.

3“GRI’s Unconventional Natural Gas Subprogram, ” Status Report, mcem-

ber 1982.

4“New Fracturing Technique Undergoing Tests, ” oil and Gas )ourna/, Aug.

8, 1983.



App. B—Fracturing Technologies q 2 3 5

Box B-1. --Well Logs

Well logs are measurements of rock formationcharacteristics taken by devices, called sondes,that are lowered into the wellbore on an elec-tric wireline and transmit back information to a

surface recording device. There are a great va-riety of these devices. The most common are:

1. Eletrical logs measure the electrical char-acteristics of the rock surrounding the well-bore before the well is cased. Electrical logsmeasure either resistivity or spontaneouspotential.qResistivity logs pass an electric current

through the rock formation and measureits ability to conduct electricity. Theselogs help to determine the type of fluidcontained in formations and the relativesaturation of oil and water. There are a

variety of resistivity logs, including: –Induction logs, which measure forma-tion resistivity in wells drilled withfreshwater drilling fluids or with non-conductive fluids such as air or oil.

–Laterolog, which can identify thinnerformations than ordinary resistivitylogs. These are used with saltwaterdrilling fluids.

–Microlog, which is used to identify theporous and permeable zones by meas-uring the resistivity of the thin layeraround the wellbore that is invaded bydrilling fluids.

q

The spontaneous potential log measuresthe electrical potential created by the dif-ferences in salinities between the forma-tion water and drilling fluids. This loghelps to differentiate between rock types(e.g., sand and shale) and to define for-mation water salinity.

2. Radioactive logs measure either the natu-rally occurring radioactivity in the rock for-mation or the response of the formation tobombardment by neutrons or gamma rays.

These include:qGamma-ray logs record the naturally

occurring gamma rays in the rock forma-tion surrounding the well bore. They dif-

ferentiate between shales and other for-mations, or measure the amount of shalein the formation.

* Neutron logs bombard the formationwith neutrons and measure the inducedgamma rays: They delineate porous for-mations, and indicate the amount of fluid(and, in some ca ses, fluid type). They areuseful in locating gas zones and deter-mining rock types.

qFormation density logs measure the scat-tering of gamma rays bombarding the for-mation from a source on the logging tool.They are used to determine porosity, and

help in determining rock types in con-junction with sonic or neutron logs.

qGamma spectrometry logs measure boththe scattered neutrons and the gamma-ray spectrum from neutron bombard-ment. They help in measuring ’hydrocar-bon saturation, porosity, formation watersalinity, and rock types.

3. Acoustic veIocity, or sonic logs, measure the

velocity of an acoustic (sound) wave along

the wall of the borehole. They are used tomeasure porosity, to distinguish betweensalts and anhydrites, to detect shales withabnormal pressures, to determine rock

types, and to identify fractures.4. Nuclear magnetism logs measure the effects

of applying a large magnetic field to therock formation. They are used to measurepermeability, porosity, producibility, andwater saturation.

5. Temperature logs are used to identify zoneswhere drilling mud is being lost into the for-mation or, in air-drilled wells, the locationsof gas entry into the wellbore.

SOURCES: F. A. Giuliano (cd.), Irrtruducfion to 0# and Gas Technology, 2d ad., M#comp ~ De#qmMM an d Engineering, Inc., Houston, TX, 1981;Brit~shPetroleum Co. Ltd., Our Industry Petrokum. 1977; and Schhsmberger Well !kwiccs, @40k -’m Catalog, Ig8J.



Appendix C

Double Counting

A Comparison of the PGC andNPC Tight Gas Estimates

A major problem with estimating the natural gas re-

source from unconventional reservoirs is determininghow much of this unconventional gas has alreadybeen included in estimates of the conventional re-source. The primary areas of overlap would be the tightsands and Devonian shales. Much of the tight sands rep-resents the lower end of a continuum of gas-producingreservoirs. Except for its lower porosities and perme-abilities, the “blanket” portion of the tight sands re-source is quite similar in other respects to conven-tional formations and, in fact, gas is presently beingproduced from tight blanket formations and even, insome cases, from Ienticular formations. The Devonianshales have been producing gas since the early daysof petroleum development in this country. Obviously,

these categories cannot be considered entirely newadditions to the resource base.

There is no clear-cut boundary between gas that hasbeen included in conventional resource estimates andthat which has not. The cutoff point for the conven-tional resource varies from assessment to assessmentand tends to be loosely defined on the basis of ratherill-defined economic and technical constraints. For ex-ample, the Potential Gas Committee (PGC) defines itsresource estimate to include gas from “all wells whichwould be drilled in the future under assumed condi-tions of adequate economic incentives in terms ofprice/cost relationships and current or foreseeabletechnology.” 1

The PGC assessment, Potential Supply of Natural Gas in the United States, is one estimate of the con-ventional resource which overlaps with the “uncon-ventional. ” Given its broad definition of what consti-tutes the undiscovered recoverable resource, the PGCchose not to define a physical cutoff point, such asa permeability limit, to separate out tight gas from con-ventional gas. To do so would exclude from the re-source base gas that conceivably could be producedunder the assumptions of reasonable price and tech-nology. Thus, the PGC has consistently designatedsome “tight” gas as part of the conventional resource.

Recently, the PGC has made an attempt to deter-mine the percent of its total resource estimate that oc-

curs in tight formations. (It includes in the tight gascategory both tight sands and Devonian shales.) For

I Potential Gas Agency, Potential Supply of Natural Gas in the United States

(as of Dec. 31, 1980), 1981.

each of its reporting areas (fig. C-1 (a)), it estimates the

percentage of gas that occurs in tight formations,above and below 15,000 ft. Table C-1 gives the PGCestimates, in TCF, for each reporting area.2 The total

tight gas included in these estimates is 172 TCF, or20 percent of the total potential resource.

The following analysis compares the PGC tight gasbreakout with the National Petroleum Council’s (N PC)estimates of tight sands and Devonian shale re-sources. s It has been suggested that there may be aconsiderable amount of overlap between these twoestimates. Because estimates of the United States’ gasresource base and future supply often add conven-tional and unconventional gas contributions, elimina-tion of any overlap would decrease the projected totalresource base and supply. Additionally, since an over-lap is most likely to occur among the most attractivegas prospects, elimination of the overlap may affect

near- and mid-term supply forecasts disproportionately.In order to compare the PGC breakout with the NPC

estimates, the assumptions underlying the estimatesneed to be reviewed. Some of the assumptions are

documented, others have been confirmed throughpersonal communications.

The definition for tight gas used by the PGC is simi-lar to the FERC definition and includes all gas in for-mations with average permeabilities less than 0.1 mil-Iidarcy (red). The NPC report does include some gasin formations with average permeabilities greater than0.1 md, but the amount is small, less than 1 trillioncubic feet (TCF). Therefore, the permeability levels forthe two estimates are generally compatible.

Not all the gas in the PGC tight gas estimate willoverlap the NPC estimate. For example, PGC tight gasincludes gas from new pools and reservoirs i n forma-tions that are already being produced. By definition,this gas is mostly accounted for in the probable cate-gory.4 The N PC report specifically excludes areas

already producing tight gas from its evaluation, sinceits objective is to estimate “new potential reserve ad-ditions.” Thus, tight gas in the PGC probable category,amounting to some 56 TCF, cannot be part of anyoverlap between the two estimates.

The NPC report does not include any potential gasresources in formations at depths greater than 15,000ft, although it postulates that a significant additional

ZBased on the 1982 revised figures for the total resource: Potential Gas

Agency news release, February 1983.

3Natlonal Petroleum Council, Unconventional/ Gas Sources, 1980.4Harry Kent, Director, Potential Gas Agency, personal communication,

1984.

236



App. C—Double Counting q 237

NOTE: From Figure 1—Reporting areas and total potential gas supply. 1982 Potential Gas Committee

NPC Tight Gas Basins n

1. WesternGreat Plai1. North2. Wil l is

Rock MO

3. Great4. Wind River Other Known Southwest5. Uinta Basins studied by NPC F. Raton6. Piceance G. Anadarko7. Denver H. Ouachita

Il. Greater Southwest Region 1. ArkomaOther Known Western Southwest Appraised J. Fort Worth

A. Snake River 8. San Juan K. Western Gulf CoastB. Big Horn 9. Val Verde-Ozona TrendC. Wasatach 10. Val Verde-Sonora Trend Ill. Eastern RegionD. Douglas Creek 11. Edwards Lime Trend L. AppalachianE. Western Shallow Cretaceus Trend 12. Cotton Valley Trend M. Black Warrior

3 8 - 7 4 2 0 - 8 5 - 17




Table C.1.— PGC Estimate of Gas Occurring in Tight Formations Included in Its 1982 Estimate of TotalU.S. Undiscovered Recoverable Resources (in TCF)

Probable Possible Speculative

PGC area <15,000 >15,000 <15,000 >15,000 <15,000 >15,000 Total

A 23.49 — 0.4 — — 23.89B::::::::::::::::::::

—

0.32 0.18 0.24 0.3 0.3 0.48 1.82c . . . . . . . . . . . . . . . . . . . . — — — — — — —

D . . . . . . . . . . . . . . . . . . . . 0.8 — 3.0 — 3.0 0.8 7.6E . . . . . . . . . . . . . . . . . . . . — — —

G 0.28 -

—. . . . . . . . . . . . . . . . . . . . 0 3 0.54

-048 0.12

-016 1.88

H . . . . . . . . . . . . . . . . . . . . 21.75 5 14.43 14.56 7.02 29.2 91.96I 0.94 — 0.33 – 1.27Jn” : : : : : : : : : : : : : : : : : : : —

— —

1.98 14.06 3 20.06 40.2Js . . . . . . . . . . . . . . . . . . . 0.49 — 0.38 – — 0.87L . . . . . . . . . . . . . . . . . . . .

—

0.04 0.3 0.3 0.6 0.7 0.4 2.34

Total . . . . . . . . . . . . . . . 55.87 50.72 65.24 171.83SOURCE” Potential Gas Agency. Potential Supply of Natural Gas in the United States (as of Dec 31, 1982), Report of the Potential Gas Committee, Colorado School

of Mines, June 1983

resource could exist at these depths. However, overhalf of the PGC tight sands estimate is found at greater

than 15,000 ft depths—89 TCF total, 81 TCF in the pos-

sible and speculative categories. We assume that thereis no overlap between the PGC tight gas below 15,000ft and the NPC estimate.

The gas projected by the NPC to be recoverablefrom tight sands at $5.00/MCF,5 with a 15 percent dis-counted cash flow rate of return (DCF ROR) and usingbase technology, is assumed to represent a reason-able upper economic limit to gas that might be in-cluded in the PGC estimate6 (the N PC’s “maximumrecoverable” gas would be an extreme upper limit).In other words, tight gas considered produceable at

less than $5.00/MCF using present technology is likelyto be included in the PGC tight gas estimate.

In summary, the most potential for overlap exists

between the PGC tight gas in the possible and specu-lative categories at less than 15,000 ft and the NPCtight sands and Devonian shales gas recoverable at$5.00/MCF (1979$) using base technology. This isgraphically represented in figure C-2. It should be

noted that the overlap determined by a straightforward

comparison using the above assumptions may be toolarge. PGC used FERC criteria as a guide to definingthe tight formations and the FERC interpretation ofwhat constitutes tight gas has tended to be generousrelative to the NPC interpretation.

Other specific assumptions need to be made tocompare individual areas. These are discussed in moredetail below.

—‘For simplicity, the N PC prices (in 1979 dollars) are used In this analysis.

bThe “boundary conditions’ for the PGC resource estimate are imprecise,

and no Iimlt on price is specified other than what may be inferred from the

phrase “adequate economic Incentives In terms of price/cost relationships, ”

To be precise, the NPC definition of Its “base case” technology allows

evolutionary improvements (n presently available technology.

Areas of Overlap

The PGC reporting areas and the NPC appraised and

extrapolated basins are shown in figure C-1 (a) and(b), respectively. Table C-2 lists the comparable areasand notes where the PGC has specifically identifiedtight gas included in the potential resource. The totalPGC estimate of gas in tight formations is about 172out of a total of 870 TCF, or approximately 20 per-cent of the remaining undiscovered recoverable re-source. NPC estimates 607 TCF of recoverable gasfrom tight sands8 and an additional 25 TCF, at least,

from Devonian shales. Our comparison attempts to

determine how much of this 633 TCF of gas has al-ready been included in the PGC estimate of conven-

tional undiscovered resources and cannot be consid-ered as additions to the U.S. resource base.

Our analysis indicates that the greatest potential foroverlap occurs in the Rocky Mountain region coveredby the PGC reporting area H (fig. C-1 (a)). The com-parable NPC area is the Rocky Mountain Basins plusthe Northern Great Plains (fig. C-1 (b)). This area is

already the site of considerable production of gas from

tight formations (e.g., from the Wattenberg field of theDenver Basin). The extent of the duplication is sum-marized in table C-3.

For a first approximation, we assume that the PGC

tight gas does not include gas from the Northern GreatPlains.9 Basins that probably are included in bothRocky Mountain estimates are the Greater GreenRiver, Uinta, Piceance, Wind River, and Denver

basins, Within these basins, the most likely overlapoccurs between the blanket formations of the NPC re-

@This IS the “maximum recoverable” gas. Using present technology, NPC

estimates that 365 TCF would be recoverable at $5.00/MMBtu

‘Kent, op. cit.



App. C—Double Counting q 239

Figure C-2.— PGC and NPC Categories of Tight Gas

NPC tight gas(includes both tight sands and Devonian shales)

SOURCE Off Ice of Technology Assessment,

Table C.2.—Comparable Areas—PGC Report and NPC Report

PGC reporting areasArea A. . . . . . . . . . . . . . . . . . . . .

Area B. . . . . . . . . . . . . . . . . . . . .Area D. . . . . . . . . . . . . . . . . . . . .

Area G . . . . . . . . . . . . . . . . . . . .

Area H . . . . . . . . . . . . . . . . . . . .

Area I . . . . . . . . . . . . . . . . . . . . .

Area J north. . . . . . . . . . . . . . . .Area J south . . . . . . . . . . . . . . .

Comments NPC basinsIncludes some Devonian shales Appalachian Basin Devonian shales

recoverable by normal drilling, wellst imulation, and completion and Eastern extrapolated tight sandsanalogous in geologic setting toprevious production

— Eastern extrapolated tight sandsIncludes tight formations of Travis Peak Cotton Valley appraised, Southwest

and Cotton Valley extrapolated (east) — Edwards Lime Trend, Southwest

extrapolated (south)Includes tight formations of Greater Rocky Mountains appraised, Northern

Green River, Uinta, Piceance, and Wind Great Plains appraised, WesternRiver extrapolated

— San Juan appraised, Southwestextrapolated (west)

— Southwest extrapolated (east) — Val Verde Ozona-Sonora appraised,

Southwest extrapolated (central)


Table C-3.–Comparison of PGC and NPC Tight Gas Resource, Rocky Mountain Region (in TCF)

Recoverable Maximum PGC area HNPC Rocky Mountain Basins

a$5.00/McFb recoverable <15,000 ft Tight gas Percent overlap

Blanket . . . . . . . . . . . . . . . . . . . . 19.7 34.5 Probable 21.75 NonePossible 14.43Speculative 7.02 )

62-100

Lenticular. . . . . . . . . . . . . . . . . . 68.9 164.9 NoneCombined. . . . . . . . . . . . . . . . . . 12.6 15.5 NoneTotal , . . . . . . . . . . . . . . . . . . . . . 101.2 214.9 Total 43.20 10-21aIncludes. appraised Greater Green River, Uinta, Piceane, and Denver basins, and other Western extrapolated basins.

bGas recoverable at $500/MCF (1979$), 1570 DCF ROR, assuming base technology.





port and the PGC possible and speculative categoriesat depths less than 15,000 ft. We are assuming thatthe PGC estimate does not include any gas in lenticu-Iar formations or in combined blanket and Ienticularformations. This assumption may not be strictly cor-rect because individual lenses have been produced

in past drilling by directly intersecting the lens with

the wellbore. There is no existing technology, how-ever, for producing lenses remote from the well bore.

Much of the Rocky Mountain gas occurring in blan-ket formations appears to be included in both the NPC

and PGC estimates. The NPC estimated range of gasrecoverable in blanket formations, from gas availableat $5.00/MCF, 15 percent DCF ROR, and base tech-nology, to the maximum recoverable gas, is 20 to 34TCF (see vol. V, table 9 of the NPC report). This in-cludes 7 to 10 TCF in extrapolated blanket formationsin this region. The NPC estimate for gas in blanket for-mations is very close to the PGC estimate of 21 TCFof possible and speculative gas occurring in tight for-

mations at less than 15,000 ft, and probably represents

a duplication of the PGC estimate.Another significant area of overlap may occur in the

Cotten Valley Trend of east Texas and Louisiana. Thisarea is included in PGC area D and is one of the ap-praised basins of the NPC report. The NPC range ofgas in the Cotton Valley, from $5.00 to the maximumrecoverable, is 7 to 12 TCF, which probably overlaps

the 6 TCF of possible and speculative tight gas in for-mations less than 15,000 ft as estimated by the PGC.

In south Texas, the NPC estimate ranges from 44

(at $5.00) to 60 TCF (maximum recoverable). The esti-mate covers extrapolated formations as well as the ap-praised Edwards Lime Trend, with an estimated gaspotential between 6 TCF (at $5.00) and 9 TCF (max-

imum recoverable). it is also covered by PGC repor-ting area G. Here, PGC estimates 0.66 TCF of gas inthe possible and speculative categories above 15,000

ft. It is likely that the PGC estimate, even if it does notspecifically refer to the Edwards Lime Trend, is dupli-cated somewhere in the NPC appraised plus extrap-

olated formations.In the PGC’s area 1, including the San Juan Basin,

most of their estimated tight sands gas is derived frominfill drilling of the Dakota and Mesaverde formationsand is most likely included in its probable category.The 0.33 TCF remaining in the possible category may

be new gas occurring in these formations, and most

or all of it may overlap the NPC estimates of 1.49 to

2.31 TCF for the appraised San Juan Basin. Althoughthe NPC extrapolates an additional 11 to 16 TCF inthis region (which would include gas in the RatonBasin in northeastern New Mexico), the PGC estimateprobablv does not overlap with any extrapolated gas.

A large quantity of gas–40 TCF–is estimated by thePGC to occur in tight formations within its reportingarea Jn. Most of this gas is thought to be found in theDeep Anadarko and Springer sands. Thirty-six TCF arefound at depths greater than 15,000 ft. This leaves only4.1 TCF in the possible and speculative categoriesabove 15,000 ft to potentially overlap the NPC tight

gas. All the NPC basin estimates in this region are ex-trapolations with a total range from 16 to 24 TCF. Itis likely, but not conclusive, that the 4.1 TCF of PGC

gas does overlap NPC gas.The amount of overlap in tight formations in the

Eastern United States is more difficult to determine.

This area encompasses reporting areas A and B of thePGC report and the Eastern U.S. extrapolated tightsands and the Appalachian Basin Devonian shale ofthe NPC report. in area A, excluding the estimatedprobable gas, which is likely to be primarily gas fromproducing formations in Devonian shales and inter-

Iayered sandstones, leaves 0.4 TCF in the possible cat-

egory. In area B, there are 0.54 TCF in the possible

and speculative categories at depths less than 15,000 ft.Producible gas from Devonian shales as estimated

by the NPC falls in the range of 12 TCF at $5.00/MCFand 25 TCF maximum recoverable using traditionaltechnologies only. The extrapolated tight sands re-source for the Eastern United States ranges from 72to 101 TCF. In OTA’s opinion, the 0.94 TCF of gas inthe PGC areas A and B are likely to be included some-where in the total of the NPC extrapolated tight sands

and the Devonian shale resource,Table C-4 summarizes the total overlap between

NPC and PGC estimates, amounting to approximately30.5 TCF. The percent reduction due to duplicationfor the total NPC tight gas resource (including both

tight sands and Devonian shales) is 8 percent for thegas recoverable at $5.00 and 5 percent for the max-imum recoverable gas.

The amount of overlap, then, is not vitally impor-tant in terms of reducing the size of the total additionalresource from unconventional reservoirs. More impor-

tant are the specific areas of overlap, since these oc-

Table C-4.–Overlap of PGC and NPCResource Estimates

NPC total recoverable gas

PGC total

tight sands and Devonian shales Total

tight gas At $5.00a

Maximum overlap

171.83 TCF 376.6 TCF 633.5 TCF 30.5 TCF——aGas recoverable at $5.00/MCF (1979$), 15%. DCF ROR, assuming base tech-

nology,




App. C—Double Counting q 2 4 1

cur in areas which have been predicted to be maincontributors to supply in the next 20 years.

For example, the NPC estimates that the RockyMountain Basins will contribute over 14 TCF to pro-

duction over the next 20 years and 43 TCF to reserveadditions, according to its standard development sce-nario. It is likely that much of this production will befrom the blanket formations, as these are generally the

more profitable prospects. However, if these forma-tions are already partially counted in conventional re-source estimates, and these estimates are used in fore-casting supply, what the NPC is estimating cannot beconsidered as additions to existing projections offuture conventional supply.

Another primary contributor to the NPC reserve ad-dition and supply forecasts in the next 20 years is theGreater Southwest, including primarily the Cotton

Valley, the Val Verde Ozona-Sonora Trend and theEdwards Lime Trend. The Cotton Valley Trend poten-

tial, however, appears to be duplicated in the PGCreport; thus it, also, cannot contribute additional re-

serves or supply.This analysis deals only with the overlap between

the PGC and the NPC assessments of the natural gasresource. Similar duplication is Ii kely to exist in other

geologically based estimates such as the U.S. Geologic

Survey’s (USGS) estimate of undiscovered recoverablegas resources.10 However, because of varying ap-proaches to estimating the resource, no categoricalstatement of the amount of overlap between conven-tional and unconventional resource estimates can bemade.

A final comment needs to be made regarding the

PGC estimates of tight gas recoverable from forma-

tions at depths greater than 15,000 ft. This gas repre-sents a resource additional to the N PC estimated tightsands resource, In general, these resources would beconsidered even less economic to produce than theNPC gas because of the higher costs and greater tech-nical difficulty of drilling and fracturing at these depths.However, Potential Gas Committee members felt thatthe technology did exist to produce tight gas from

deep formations, and under certain conditions theremight be sufficient incentive to produce this gas.

Nevertheless, we feel that the 89 TCF of deep tightgas in the PGC estimate should be regarded with atleast as much, if not more, caution than the NPC esti-mates in terms of evaluating their potential for con-tributing to near- and mid-term supply.

10G L, Dolton et al., Es(jfna(es o f U n d ls c m w e d R e c o v e r a b l e COfJ\’en-

tlonal Resourc es of O Il an d (2s In the United States, U.S. Geological Survey

Circular 860, 1981.



Appendix D

Glossary

adsorbed gas: Natural gas that is physically bound tothe surfaces of the reservoir rock.

anaerobic: Conditions that exist only in the absence

of oxygen.anisotropy: A characteristic of certain rocks whereincertain properties, such as permeability, exhibit dif-ferent values when measured along axes in different

directions.

anticline: A fold, generally convex upward, whosecore contains stratigraphically older rocks.

associated dissolved gas: Natural gas that occurs to-gether with oil in a reservoir, either dissolved in theoil (dissolved gas) or as a gas cap above the oil

(associated gas).blanket formations: Thin gas-bearing formations that

take the form of one or several stacked layers ex-tending laterally over a wide area.

borehole shooting: A method of stimulating increasedgas flow by detonating explosives inside the bore-

hole of a well.cleat: The pervasive, vertically oriented natural frac-

ture system in coal seams.coal seam (coalbed) methane: Natural gas formed as

a byproduct of the coal formation process andtrapped in the coal seam.

combination trap: A trap for oil or gas that has bothstructural and stratigraphic elements.

deviated drilling: Drilling that has been deliberately

angled away from the vertical.

Devonian shale gas: Gas trapped in the shales of Devo-nian age located in the Eastern United States, pri-

marily in the Appalachian, Michigan, and Illinoisbasins.extension test: A well drilled to extend the areal limits

of a partially developed pool. May sometimes be-come a new pool discovery well. Also known asoutpost well.

fault: A sudden displacement of rock strata along a

fracture.field: Composed of a single pool or multiple pools that

are grouped on or related to a single structuraland/or stratigraphic feature.

formation: A rock mass composed of individual beds

or units with similar physical characteristics ororigin.

formation damage: A reduction in permeability causedby drilling, fracturing, or producing a well–e.g.,by the plugging of pores by water-sensitive clays

dislodged or caused to swell by water-based frac-ture fluids or drilling fluids.

formation water: Water present in a water-bearing for-mation under natural conditions, as opposed to in-troduced fluids, such as drilling mud.

infill drilling: Drilling at a smaller spacing than calledfor in the original development plan, designed tospeed up production and/or increase ultimaterecovery.

interference: A condition whereby adjacent wells in

a field are close enough together that their areasof (pressure) influence overlap, generally reducing“per well” gas recovery below the level that wouldbe obtained with an isolated well.

lens: An individual reservoir in a tight Ienticular for-mation (see below), often oval in cross-section.

Ienticular formation: A thick formation containinglarge numbers of small, separate, lens-like reser-

voirs interspersed with impermeable shales or coal.

Iineament: A linear feature of the Earth’s surface thatmay reveal a subsurface feature such as a fault.

log, well log: Measurements of the physical proper-ties of a reservoir, taken while drilling, generallyby lowering measurement devices down the well bore.

massive hydraulic fracturing (MHD): Creation of large,manmade fractures in reservoir rock by pumpingfluids into a well under high pressures. “Frac” jobsgenerally are considered “massive” when the vol-ume of fluid used is 100,000 gallons or more, butthere is no universally accepted criterion.

methane: The primary constituent of natural gas, the

gaseous hydrocarbon CH4.

natural fracture system: A series of fractures, often

aligned in some way, created by natural processes.new field wildcat: A well drilled in search of oil or gasin a geological structural feature or environmentthat has never before been proven productive.

new pool wildcat: Well drilled in search of pools above(shallower pool test), below (deeper pool test), or

outside the areal limits of already known pools infields that have already been proven productive.

May sometimes become an extension well.nonassociated gas: Natural gas that occurs in a reser-

voir without oil.outpost well: See extension test.pay: A rock stratum or zone that yields oil or gas.permeable: Having the property or capacity of a po-

rous rock, sediment, or soil for transmitting a fluid;it is a measure of the relative ease of fluid flow

under unequal pressure.petroleum: A general term for all naturally occurring

hydrocarbons, whether gaseous, liquid, or solid.

242



App. D—Glossary . 243

play: A rock formation or group of formations within

a sedimentary basin with geologic characteristicssimiIar to those that have been proven productive.A play serves as a planning unit around which anexploration program can be constructed. May alsorefer to the exploratory effort, often following a sig-nificant discovery, that uses a geologic idea to de-

termine where petroleum can be found.pool: A subsurface accumulation of oil and/or gas in

porous and permeable rock, having its own isolatedpressure system. Theoretically, a single well could

drain a pool. Also known as a reservoir.porosity: The percentage of the bulk volume of a rock

or soil that is occupied by interstices (gaps betweenthe particles that compose the rock), whether iso-lated or connected.

proppant: Small particles of a hard material (sand,

bauxite, etc.) that are suspended in fracturing fluid,to be left behind when the fluid is removed to pre-vent the created fractures from closing under thepressure exerted by the overlying rock.

prospect: An area that is a potential site of economi-cally recoverable petroleum accumulation basedon preliminary exploration.

province: A region in which a number of oil and gaspools and fields occur in a similar or related geo-logical environment.

reserves: The portion of the total gas resource basethat has been identified by drilling and estimateddirectly by engineering measurements, and that isrecoverable at current prices and technology.

reservoir: See pool.reservoir rock: Any porous and permeable rock that

yields oil or gas. Sandstone, limestone, and dolo-mite are the most common reservoir rocks, but gas

accumulation in the fractures of less permeablerocks also occurs.

resources: The total amount of oiI or gas that remainsto be produced in the future. Generally does notinclude oil or gas in such small deposits or under

such difficult conditions that it is not expected to

be produced at any foreseeable price/technologycombination.

secondary migration: The movement of fluids withinthe permeable reservoir rocks that eventually leadsto the segregation of oil and gas into accumulationsin certain parts of these rocks.

sedimentary basin: A low area in the Earth’s crust,

caused by Earth movements, in which sedimentshave accumulated.

sedimentation: The act or process of forming or accu-mulating sediment in layers, including such proc-

esses as the separation of rock particles from thematerial from which the sediment is derived, thetransportation of these particles to the site of depo-

sition, the actual deposition or settling of the par-ticles, the chemical or other changes occurring inthe sediment, and the ultimate consolidation of thesediment into solid rock.

source rock: Sedimentary rock in which organic ma-terial under pressure, heat, and time was trans-formed to liquid or gaseous hydrocarbons. Source

rock is usually shale or limestone.stimulation: Any process that mechanically or chem-

ically disturbs the reservoir rock in order to increasegas flow to the well.

stratigraphic trap: A trap for oil or gas, resulting fromchanges in rock type, porosity, or permeability, thatoccurs as a result of the sedimentation processrather than structural deformation.

structural trap: A trap for oil or gas resulting from fold-

ing, faulting, or other deformation of the Earth.

thermal maturity: The extent to which the organicmatter in sedimentary rocks has been “cracked”-broken into simpler molecules–by heat.

trap: Any barrier to the upward movement of oil or

gas that allows either or both to accumulate. A trapincludes a reservoir rock and an overlying imper-meable roof rock; the contact between these is con-cave, as viewed from below. See also stratigraphic,structural, and combination traps.



Index



Index

AAPG. See American Association of Petroleum

Geologistsabandonment pressures, 73-75, 93, 228

abiogenic gas. See “deep source gas”

acidizing, 74Alabama, 8, 59, 133, 207, 219

Alaska, 7, 17, 26, 43, 53, 66, 106, 114-115Alaskan Natural Gas Transportation System, 112,

114-115Alaskan Natural Gas Transportation Act of 1976,

114American Association of Petroleum Geologists, 62

American Gas Association, 9

addition to reserves and, 24, 54, 83

conventional/u nconventional boundary and, 140decline in RIP and, 95economic incentives to recovery and, 66-69extensions and new discoveries and, 88-91import estimate, 113, 115production estimate, 106-107

tight gas estimate, 169-171Anadarko Basin, 22, 45, 57, 65-66, 76, 88, 168

anaerobic process, 29, 242

ANGTS. See Alaskan Natural Gas TransportationSystem

anticline, 30, 242Antrim shale, 183Appalachian Basin, 57, 149

coal seam methane in, 207, 211Devonian shales in, 8, 129, 132, 179-181, 183,

185-186, 190-203

undiscovered resources in, 59

aquifers, geopressurized, 6, 26, 121Arkansas, 59, 88, 168

Arkoma Basin, 168, 211Atlantic shelf, 46Austin Chalk, 88

availability, natural gas. See supply, natural gas

Baltimore Canyon, 22, 53, 76Big Sandy Field, 129biogenic gas, 29-30, 147-149

biomass, 26

bituminous coal, 218-219Black Warrior Basin, 8, 136, 149, 207, 219, 221Blanco Basin, 73blanket formations, 125, 141-142, 151, 153, 230,

242

Border Gas, 112borehole shooting, 200, 242Bromberg/Hartigan, 19-20, 49, 61

“bubble” of gas, 3, 17, 25

California, 53, 57, 114Canada, 7, 17, 26, 106, 111, 113-115carbon dioxide, liquid, 188-189

Carnegie Natural Gas Co., 207

“Circular 725, ” 47, 54“Circular 860,” 47, 50, 54

“cleat,” 134, 207, 212, 215-216, 242coal seam methane

defined, 6, 133, 207estimate uncertainties, 7-9, 134-136, 211-212FERC estimate, 135, 210

in-place estimates, 134-135, 209-212NPC projections, 3, 209-210, 217-219, 221-222

production, 7, 207, 212-224

as recoverable resource, 3, 5, 7-10, 26, 29,121-122, 133-136, 210-212, 217-221

regions, 208Colorado, 139, 233

Colorado School of Mines, 51conservation, 3, 94, 112

Consolidated Coal Co., 215conventional gas. See also proved reserves;

recoverable resources; reserves; resources

defined, 5parameters of supply estimate, 5, 50production cycles, 20

resource estimates of, 5-9, 19unconventional boundary, 140year 2000 production, 6-7, 18

Corcoran Cozette, 148

data availability, 18

decontrol, iii, 9, 75“deep source gas, ” 6, 29, 43, 53, 60, 66, 75-76,

121deliveries, curtailment of, 3Delphi approach, 42-43, 48, 50, 192, 196Delta provinces, 71demand for gas

exploration and, 5, 17, 45production and, 5-6, 17, 94

Denver Basin, 71, 76, 126, 144, 147-148, 157

depletion, 39, 41, 69, 105resorption, 212-213

Deul and Kim, 135, 210development strategy, 5

Devonian shalesin Appalachian Basin, 8, 129, 132,

183, 185-186, 190-203areas found, 8, 129, 132, 179-187,defined, 6, 128, 179-182development problems, 122, 130

79-181,

90

247




FERC estimate, 182in-place resource, 182-186, 190NPC projections, 3, 131, 182-183, 186, 191,

194-196, 199-203

OTA assessment (1977), 192production from, 7, 181-182, 190-203

as recoverable resource, 5-6, 8-10, 26, 121-122,

128-133, 190-201resource estimates, 131-133, 182-183, 186,

190-201dewatering systems, 10, 212-214discovery, natural gas

demand and, 5, 17, 45estimate of prospects, 17-18, 39extrapolation from past trends, 21, 23-24, 43-45,

66, 88inadequate indicators for, 23new field, 24-25, 45-46, 61-64, 81-90, 98-99,

1 0 5

technical prospects for, 5, 45double counting, 167, 236-241

Drew, L. J., 71drilling, 5, 24. See also infill drilling

advances in deep, 22costs, 22, 66-67, 76, 94, 215deviated, 189, 242directional, 189exploratory, 33-34, 44, 46, 54, 73improved technology, 10, 83

low- v. high-risk, 62, 68NPGA and, 66

success rates, 68-69

East Anschutz Ranch, 66Eastern Gas Shales Project, 135, 182, 186, 197

Eastern Overthrust Belt, 7economic conditions, natural gas and, 20, 24, 41,

45, 66-67

Effects of Decontrol on Old Gas Recovery, iiiEIA. See Energy Information AdministrationElectric Power Research Institute, 3

Ellsworth shale, 183energy conservation. See conservationEnergy Information Administration, 10, 24, 34,

88-90, 95, 107“Enhanced Recovery of Unconventional Gas,”

193environmental constraints, 223Equitable Gas Co., 207

exploration, natural gas, 69basic techniques, 33-34evolution of technology, 17, 20, 82-83high-risk, 62, 67-68

history of, 57

unconventional gas, 130-131, 156, 189-190

Export Administration Act of 1979, 114“extensio ns,” 23-24, 34, 65, 75, 81, 86-90,

98-102, 242Exxon, 49, 60, 76, 107

Federal Energy Regulatory Commission, 121-122

coal seam methane estimate, 135, 210Devonian shale estimate, 182

tight gas estimate, 123-124, 126, 139-140,142-144, 166, 236

Federal Power Commission, 66, 121, 126, 142-144foamed fluids, 188

FERC. See Federal Energy Regulatory CommissionFletcher Field, 94Florida, 59Fort Union, 148fracturing, 74, 94

coal seam methane, 216-217Devonian shales, 130, 187-189

technologies, 229-235tight gas, 8-10, 125, 127, 139, 149-154, 163-165

Garrett, R. W., 49gasfields

field size, 42, 62-65, 69-73new discoveries, 24-25, 45-46, 61-64, 81-90,

98-99, 105

new gas from older, 7, 21-22, 24, 46, 66, 73-75,

227-228role of small, 7, 21, 53, 66, 69-73, 84

gas hydrates, 6, 121Gas Research Institute, 9

coal seam methane estimate, 135-136, 209-210,215, 217, 219, 222-223

mineback experiments, 233tight gas estimate, 127, 156-158, 160-164, 166,169-172, 174

Georges Bank, 7, 22

Gold, Thomas, 29“growth factor” projections, 98-102Gulf of Alaska, 46, 53, 76Gulf of Mexico, 7, 22, 53, 57, 59-60, 71, 76, 84,

95

history-based estimates, 43-46, 52-56, 60-61, 66,

99, 203Hubbert, M. King

methodology of, 54-56, 60-61, 81-82, 103-104

OTA assessment of estimate, 77, 105resource estimate, 19-20, 39, 41, 43, 47, 49, 65

Hugoton field, 62, 65, 73, 95Huron Intervals, 197-198

hydrates, gas, 6, 121hydrocarbons, liquid, 29-30



Index q 2 4 9

IGT estimate, 49Illinois, 114, 129, 179, 211imports, 7, 26, 39, 111-116infill drilling, 73-75, 93, 228, 242

Jensen Associates, Inc., 99

Kansas, 57, 59, 65, 73, 95Kentucky, 129, 192-193, 197, 199Kuuskraa and Meyer (KM), 133, 135, 209-210,

217-219, 221

leases, 46, 83, 123, 190, 201-202legal constraints, 3, 222-223lenses, 8, 10, 125, 242Ienticular formations

defined, 125, 142, 242gas recovery from, 127, 148, 151, 154, 156,

162, 172, 175, 230Lewin & Associates

coal seam methane estimates, 219

Devonian shale estimates, 131-134, 182-183,191-194, 197-203

Mexican gas and, 113-114,tight gas estimates, 126-127, 142, 144-145,

157-159, 161-162, 165, 168-170, 172-174

unconventional resource estimates, 121limestone formations, 6, 121, 123, 139, 141, 183liquefied natural gas, 7, 17, 26, 106, 112, 115-116

LNG. See liquefied natural gas

logging, 9Louisiana, 59, 65, 84, 88, 91, 95

Lower Cretaceous-Jurassic, 148low-permeability reservoirs, 6, 18, 25, 121,

123-124, 139

McKelvey Box, 40-41, 45Mesaverde, 148methane. See coal seam methaneMethane Recovery From Coalbeds Project, 135,

211-212, 221

Methanol, 115Mexico, 7, 17, 26, 106, 111-113Michigan, 129, 179, 183Mineral Leasing Act, 223mine safety, 8Mississippi, 59Mobil Corp., 49models, 81

Monsanto Corp., 131, 196Monte Carlo simulation, 47-48, 184Morgantown Energy Technology Center, 192Mound Facility, 131, 182, 185-186, 190-191,

196-197Multiwell Experiment, 163, 233

National Academy of Science, 121, 135National Energy Board, 111, 113

National Energy Plan (Canada), 113-114National Gas Survey, 121, 142National Petroleum Council

coal seam methane estimates, 3, 209-210,217-219, 221-222

Devonian shale estimates, 3, 131-136, 142,144-149, 182-183, 186, 191, 194-196,

199-203

tight gas estimates, 3, 124, 126-128, 142,144-149, 154, 157-175, 236-241

unconventional gas estimates, 3, 121natural gas basics

defined, 29how formed, 29-31nonassociated, 30where found, 30-31

Natural Gas Policy Act of 1978, 22, 66, 74-76,121-123, 139, 227-228

Nehring, Richard, 52-54, 58, 60, 71, 77

New Albany shale, 183New Mexico, 8, 65, 133, 207“new pool discoveries, ” 23-24, 34, 65, 81, 86-90,

98-102NGPA. See Natural Gas Policy Act of 1978nitrogen foam, 188, 216Northern Great Plains, 165-166

tight gas and, 8, 126-127, 140, 147-150, 157,

161, 165-166, 168-169, 172-175

North Slope gas, 112, 114-115NPC. See National Petroleum Council

offshore natural gas, 8, 25, 46, 50, 57, 60Ohio, 57, 132-133, 189, 192, 197-200, 203

Oklahoma, 59, 65, 73, 94, 168onshore deep gas, 7-8Outerbanks, California, 53Outer Continental Shelf, 66

Overthrust Belt, 53, 57, 65-66. See a/so EasternOverthrust Belt; Western Overthrust Belt

Pennsylvania, 207Pennsylvania Supreme Court, 223permeability, 6, 30, 94, 145Permian Basin, 22, 71, 73, 76, 88PGC, See Potential Gas CommitteePiceance Basin, 126, 144, 146-147, 157, 211, 233pipeline imports, 7, 17-18, 26, 39, 112

policy, energy, 9-10policy, natural gas, 9-10, 66Potential Gas Committee

Devonian shales and, 52

OTA assessment of estimate, 77, 82production estimate, 103-106




resource estimate, 3, 5, 17, 19, 39-40, 49-52,

58-60, 65, 100tight gas estimate, 236-241

unconventional gas estimate, 122, 124, 167-168Potential supply of Natural Gas in the U n i t e d

States, 236

Powder River Basin, 76

Powerplant and Industrial Fuel Use Act of 1978,iii, 3

pricescontrolled, 22, 67, 74, 76conventional gas, 5-6declining, 3, 22, 66

and economics of recoverability, 3, 5, 9, 17, 41,45, 66-67, 73-76, 83, 163, 193-194, 201

increased gas, 3, 9market level, 9, 75pressures to decontrol, iii, 9, 75

technology development and, S, 9tight gas and, 8unconventional gas and, 5, 7-8, 121, 139, 163,

193-194, 201wellhead, 5, 17, 26

probabilistic estimates, 46-48, 52

production, natural gas. See also coal seam

methane; Devonian shales; tight gasAGA estimate, 106-107

assessment of, 17-26, 39, 97, 100-106bases for OTA projections, 25basic mechanics of, 35

coal seam methane, 7, 207, 212-224demand and, 5-6, 17, 94enhancement, 221950-85, 4, 19pessimistic scenarios for, 5, 39, 101

potential, 81-107projection for year 2000, 6-7, 17-19, 81, 97,

100-102, 105-106tight gas forecasts, 7, 168-175uncertainties in estimates of, 5-7, 18-25, 105-106USGS estimate, 103-105

proppants, 125, 150, 152, 188, 216, 243proved reserves, 40. See also reserves; recoverable

resources

additions to, 4, 18, 23-24, 43, 61, 65, 81-97current level of, 3-4, 17, 23rate of additions to, iii, 18revisions to estimates, 90-93total U. S., 1977-1983, 34

uncertainties in calculations, 46-47withdrawals, 18

Prudhoe Bay Field, 111, 114

Pulle, C. V., and Seskus, A. P., 131, 191-192,195-196, 199-200

pumps, 213, 217Purchase Gas Adjustment, 122

RAND Corp.

OTA assessment of estimate, 77, 105

resource estimate, 19-20, 47, 49, 52-54, 61-62Reagan, Ronald, 114recession, effect of, 112

recoverable resources. See also coal seam

methane; Devonian shales; tight gasalternative estimates, 49, 55Hubbert estimate, 19-20, 39, 41, 43, 47, 49, 65PGC estimates of, 3, 5, 17, 19, 39-40, 49-52,

58-60, 65, 100price changes and, 3, 5, 9, 17, 41, 45, 66-67,

73-76, 83uncertainties in estimates, 5, 8, 17-18, 52, 58USGS estimates of, 3, 5, 39-40, 47-50, 59-60,

65, 82, 100, 104

regions, gas supply, 4, 31, 59, 100-102regulations, government, 45, 66, 75, 94reserves. See also proved reserves

additions to, 4, 18, 23-24, 43, 54, 61, 65, 81-97defined, 3, 40“inferred,” 491981-2000, 3, 97, 100, 102

OTA estimate of, 8, 39-50, 56-74, 97reserves-to-production ratio

decline in, 95implications, 23-25, 81, 93-96projected, 19, 96-97

for years 1945-80, 95reserves, proved. See proved reserves

resources. See also recoverable resources; supply,natural gas

credibility of estimates, 60-61, 77defined, 3estimate comparison problems, 41, 45-47, 60-61geology-based estimates, 42-43, 45, 48, 52-53,

60OTA assessment of base, 39-50, 77, 82PGC estimate, 3, 5, 17, 19, 39-40, 49-52, 58-60,

100

uncertainties in extrapolating discovery trends,23-24

uncertainties of magnitude and character, 7-9,18-24

undiscovered, 54, 59, 82USGS estimate, 3, 5, 17, 19, 39-40, 47-50,

59-60, 65, 82, 100, 104



I ndex 251

responsive reserves, 228“revision s,” 23-24, 34, 65, 75, 81, 98-102risk taking by industry, 83rocks, source, 30, 35, 46, 125, 243R/P ratio. See reserves-to-production ratio

SAI estimate, 196-197sandstone formations, 6, 121, 123, 141, 147, 183San Juan Basin, 8, 65, 73, 144, 157, 207, 211, 220

satellite imagery, 156

Scheunemeyer, J. H., 71sedimentary basins, 29-30shale oil, 188

Shell Corp., 19, 49, 76, 107, 227-228shortages, 5Soter, Steven, 29

Southeast Georgia Embayment, 22, 53, 76South Texas Lobo Trend, 88stratigraphic traps, 7, 22-23, 30, 53, 66, 76-77, 243

“stripper” wells, 22subduction, 71supply, natural gas

long-term trends, 3, 5, 17national perception of (1 978), 3, 17for next few decades, 5, 39

private sector forecast, 107shortage, 5short-term, 3, 17surplus, 3, 17, 25, 113warning of shift in, 9-10

synthetic gas, 18, 26

tailored pulse loading, 188, 234technology, new

and currently uneconomic production, 5, 9development, 9forecasting problems, 7, 17, 41fracturing, 229-235gas recovery and, 3-5, 9, 24, 45, 73, 83

Texas, 59, 65, 84, 88, 95

Texas Railroad Commission District, 73, 87, 91tight gas

AGA estimate, 169-171deep, 167-168defined, 6, 123-125, 139-140FERC estimates, 123-124, 126, 139-140, 142-144,

166, 236in-place estimates, 144-149, 168NPC projections, 3, 124, 126-128, 142, 144-149,

154, 157-175, 236-241

PGC estimate, 236-241price, 8production forecasts, 7, 168-175recoverable, 5-6, 8-9, 122-128, 139-149, 157-168

sands, 26, 52, 88, 121, 139, 156, 174uncertainties in estimates, 139, 145-149,

161-168, 236-241TransAlaska Gas System, 112traps, petroleum, 30, 32, 46, 243TRW, 135

unconventional gas, 26, 39. See a/so coal seammethane; Devonian shales; tight gas

defined, 5, 121-122

exploration, 130-131, 156, 189-190Federal role and, 121NPC supply projections, 3, 121parameters of supply estimate, 5-7, 50, 121-123PGC estimate, 122, 124, 167-168prices, 5, 7-8, 121, 139, 163, 193-194, 201

USGS estimate, 103-105, 122, 131, 134, 183-186U.S. Bureau of Mines, 223U.S. Congress

Alaska and, 114

perception of long-term supply, iiiU.S. Department of Energy

contractors, 121, 126, 142Eastern Gas Shales Project, 135, 182, 211-212,

221Methane Recovery From Coalbeds Project, 135,

211-212, 221mineback experiments, 233Morgantown Energy Technology Center, 192

policy, 10research, 156

resource estimate, 134, 210U.S. Department of the Interior, 223U.S. Geological Survey

coal resource data, 209-210Devonian shale estimates, 182-185

gasfield growth estimate, 99production estimate, 103-105resource estimate, 3, 5, 17, 19, 39-40, 47-50,

59-60, 65, 82, 100, 104unconventional gas estimates, 103-105, 122,

131, 134, 183-186USGS. See U.S. Geological SurveyU.S. House of Representatives

Committee on Energy and Commerce, iii, 4Subcommittee on Fossil and Synthetic Fuels, iii,

4U.S. Senate

Committee on Energy and Natural Resources, iii,4

Subcommittee on Energy Research and Develop-ment, iii, 4

U.S. Steel, 136, 219U.S. Steel v. Hoge, 2 2 2




Van Driest, 11, E. R., 49, 52Ventura Basin, 53

Warrior Basin, 133, 211

Wattenberg Field, 139, 144well. See also well stimulation

classification, 33-34

completions, 67conventional v. tight gas, 150expenditures for 67

ne w gas from old fields, 73-75safety factors, 223technology, 8-9

tight gas and, 153-154

Western Overthrust Belt, 4, 7, 22, 45-46, 51, 53,60, 76, 88-89, 189-190

West Virginia, 129, 132-133, 192-193, 197-200,203, 207

Whitney Canyon/Carter Creek, 66wiIdcats

u.s. natural gas availability: gas supply through the year 2000

Documents