selection and treatment of stripper gas wells for production

Selection and Treatment of Stripper Gas Wells for Production Enhancement, Mocane-Laverne Field, Oklahoma Final Report October, 2000 – September 30, 2003 Scott Reeves Advanced Resources International 9801 Westheimer, Suite 805 Houston, Texas 77042 and Buckley Walsh Oneok Resources 100 West Fifth Street Tulsa, OK 74103-0871 August, 2003 U.S. Department of Energy DE-FG26-00NT40789

Advanced Resources International

SP082003-2107 i

Disclaimers

U.S. Department of Energy

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United Sates Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United Sates Government or any agency thereof.

Advanced Resources International The material in this Report is intended for general information only. Any use of this material in relation to any specific application should be based on independent examination and verification of its unrestricted applicability for such use and on a determination of suitability for the application by professionally qualified personnel. No license under any Advanced Resources International, Inc., patents or other proprietary interest is implied by the publication of this Report. Those making use of or relying upon the material assume all risks and liability arising from such use or reliance.


SP082003-2107 ii

Executive Summary

In mid-2000, the U.S. Department of Energy awarded Advanced Resources International an R&D contract to determine if methods to identify production enhancement potential based on engineering and virtual intelligence procedures, that was successfully employed in previous R&D could be applied to stripper gas wells. The overall procedure for the project was to select a field test site, apply the candidate recognition methodology, select wells for remediation, remediate them, and gauge project success based on the field results. This report summarizes the activities and results of that project. The site selected for the project was a Mocane-Laverne field in Oklahoma’s central Anadarko basin. The industry partner for the project was Oneok Resources, who operates over 100 wells in the field. Based on the results of the study, the following conclusions were drawn:

The Mocane-Laverne field appears to hold production enhancement potential. That potential appears to be greatest for stimulation (or restimulation), with artificial lift and compression being secondary opportunities.

A production enhancement candidate selection procedure, based on engineering type

curves, virtual intelligence and heuristic screening processes, was applied at the field, and opportunities were prioritized. The accuracy of the candidate selections were tested against the results from six actual workovers.

Two of the six workover wells were candidates based on the selection procedure, and

the other four were not. Workovers on the two selected candidates were “successful” based on criteria of a reserve cost of <$0.50/Mcfe, and the other four were “unsuccessful”. The two candidate wells provided almost a Bcf of added reserves as a result of the workovers at a cost of about $20,000, for an average reserve cost of <$0.02/Mcfe.

Even with a lack of reservoir data, and noisy production data, the overall

methodology (which combined data exploration, engineering type curves, artificial intelligence and heuristic screening) successfully identified remediation candidates. The procedure therefore appears to be effective at identifying production enhancement opportunities even when only limited data is available.


SP082003-2107 iii

Table of Contents

1.0 INTRODUCTION............................................................................................................................ 1

2.0 TEST SITE DESCRIPTION AND PRELIMINARY DATA EXPLORATION ............................. 2

3.0 REMEDIATION CANDIDATE SELECTION ............................................................................. 12

3.1 TYPE-CURVE ANALYSIS............................................................................................................. 12

3.2 ARTIFICIAL INTELLIGENCE ANALYSIS........................................................................................ 17

3.3 CANDIDATE SCREENING PROCESS.............................................................................................. 20

4.0 FIELD IMPLEMENTATION RESULTS...................................................................................... 23

5.0 CONCLUSIONS............................................................................................................................ 26

6.0 ACKNOWLEDGEMENTS ........................................................................................................... 27

7.0 REFERENCES............................................................................................................................... 28


SP082003-2107 iv

List of Tables

Table 1: Summary of Reservoir Properties..........................................................3 Table 2: Summary of Wells, Production Streams and Completions....................3 Table 3: Reservoir Pressures................................................................................8 Table 4: Parameters for Type Curve Matching..................................................13 Table 5: Input Parameters for ANN Model........................................................18 Table 6: Candidate Screening Criteria................................................................20 Table 7: Group “A” Candidate List....................................................................21 Table 8: Group “B” Candidate List....................................................................21 Table 9: Pressure Buildup Test Results..............................................................23 Table 10: Workover Results.................................................................................24

List of Figures

Figure 1: Location of Mocane-Laverne Field, Anadarko Basin ...........................2 Figure 2: Location of Study Wells........................................................................4 Figure 3: Well Completions Summary.................................................................4 Figure 4: Well Vintage Summary.........................................................................5 Figure 5: Breakdown by Stimulation Type...........................................................5 Figure 6: Stimulation Fluids.................................................................................6 Figure 7: Proppant Volumes.................................................................................6 Figure 8: Liquids Production................................................................................7 Figure 9: Liquid Producing Configurations..........................................................7 Figure 10: Distribution of Flowing Pressures.........................................................8 Figure 11: Distribution of Pressure Drawdowns.....................................................9 Figure 12: Utility of Compression..........................................................................9 Figure 13: Overall Completion Performance........................................................10 Figure 14: Current Completion Performance........................................................10 Figure 15: Reservoir Pressure Correlation............................................................13 Figure 16: Example Type Curve Matches.............................................................14 Figure 17: Summary of Match Quality - All Study Wells...................................15 Figure 18: Permeability – Thickness Results........................................................15 Figure 19: Thickness-Drainage Area Results.......................................................16 Figure 20: Fracture Penetration Results................................................................16 Figure 21: Well Performance Cross-Plot .............................................................17 Figure 22: Predicted versus Actual Well Performance, Stimulated Wells Model........................................................................................19 Figure 23: Skin Factor vs. Permeability...............................................................24


SP082003-2107 1

1.0 Introduction In 1996, Advanced Resources International (ARI) began performing R&D targeted at enhancing production and reserves from natural gas fields. The impetus for the effort was a series of field R&D projects in the early-to-mid 1990’s, in eastern coalbed methane and gas shales plays, where well remediation and production enhancement had been successfully demonstrated1,2,3. As a first step in the R&D effort, an assessment was made of the potential for restimulation to provide meaningful reserve additions to the U.S. gas resource base, and what technologies were needed to do so4. That work concluded that:

o A significant resource base did exist via restimulation (multiples of Tcf). o The greatest opportunities existed in non-conventional plays where completion practices

were (relatively) complex and technology advancement was rapid.

o Accurate candidate selection is the greatest single factor that contributes to a successful restimulation program.

With these findings, a field-oriented program targeted at tight sand formations was initiated to develop and demonstrate successful candidate recognition technology. In that program, which concluded in 2001, nine wells were restimulated in the Green River, Piceance and East Texas basins, which in total added 2.9 Bcf of reserves at an average cost of $0.26/Mcf5. In addition, it was found that in complex and heterogeneous reservoirs (such as tight sand formations), candidate selection procedures should involve a combination of fundamental engineering and advanced pattern recognition approaches, and that simple statistical methods for identifying candidate wells are not effective6. In mid-2000, the U.S. Department of Energy (DOE) awarded ARI an R&D contract to determine if the methods employed in that project could also be applied to stripper gas wells. In addition, the ability of those approaches to identify more general production enhancement opportunities (beyond only restimulation), such as via artificial lift and compression, was also sought. A key challenge in this effort was that, whereas the earlier work suggested that better (producing) wells tended to make better restimulation candidates, stripper wells are by definition low-volume producers (either due to low pressure, low permeability, or both). Nevertheless, the potential application of this technology was believed to hold promise for enhancing production for the thousands of stripper gas wells that exist in the U.S. today. The overall procedure for the project was to select a field test site, apply the candidate recognition methodology to select wells for remediation, remediate them, and gauge project success based on the field results. This report summarizes the activities and results of that project.


SP082003-2107 2

2.0 Test Site Description and Preliminary Data Exploration The site selected for the project was the Mocane-Laverne field in Oklahoma’s central Anadarko basin (Figure 1)7. It is one of the largest gas fields in the Anadarko basin and hence represents an excellent opportunity for reserve enhancement. The field produces from four main horizons, the Hoover and Tonkawa (Upper Pennsylvanian), the Morrow (Lower Pennsylvanian), and the Chester (Upper Mississippian) (in order of increasing depth). The uppermost three horizons are sandstones, and the lowermost (Chester) is a limestone. A summary of reservoir properties are presented in Table 1.

Figure 1: Location of Mocane-Laverne Field, Anadarko Basin

(from Atlas of Major Mid-Continent Gas Reservoirs7)


SP082003-2107 3

Table 1: Summary of Reservoir Properties

Hoover Tonkawa Morrow Chester Depth 4283 ft 5558 ft 7000 ft 7050 ft Pressure 1525 psi 1865 psi 2200 psi 2192 psi Temperature 112º F 130◦F 150º F 154º F Pay NA 15 ft 20 ft 18 ft Porosity 18 % 18 % 12 % 8 % Water Saturation

NA

NA

38 %

30 %

Permeability 60 md 40 md 25 md 1 md Gas Gravity 0.67 0.73 0.75 0.64

NA = not available The industry partner for the project was Oneok Resources, who operates over 100 wells in the field. A specific breakdown of the number of wells, completions and production streams available for study is provided in Table 2. An illustration of the locations of the wells included in the study (limited to the four horizons mentioned above) is provided in Figure 2. A breakdown of those wells in terms of completions and well vintages are provided in Figures 3 and 4. About 75% of the wells are completed in either the Morrow and Chester, and the wells can date back as far as the 1950’s to as recently as the 1990’s, representing a broad cross-section well vintage and completion practices.

Table 2: Summary of Wells, Production Streams and Completions Category Number Remarks

Wells 104

Production Streams 106 Two dual-tubing string wells.

Completions 131 82 single (incl. dual-tubing), 23 dual, 1 triple.


SP082003-2107 4

Figure 2: Location of Study Wells

Figure 3: Well Completions Summary

3321.51

868.37

348.78

546.16

829.62

252.85

2066.77

6553.49

1839.57

2208.51

959.621467.85

1330.58

98.68

9233.18

8343.92

3951.63

6478.82

3819.18

1839.14

746.88

17066.74

9993.40

22232.01

22295.57

9743.42

27842.76

29436.26

2609.13

3377.87

2071.44

6584.80

4853.58

1635.17

4340.24

5861.99

503.00

732.17

1244.37

1445.51

779.37

357.27

1089.40

1317.25

2566.03

517.51

10.16

1277.58

2111.24

4446.67

6085.01

1733.52

1746.74

1655.35

5455.93

1545.93

137.49

1418.37

505.93

343.07652.28

922.69

5229.75

3054.42

1022.88

1057.63

14495.10

1559.98

1003.49

116.41

534.20

2598.45

3057.54

6466.25

2532.24

1591.42

7190.77

8000.57

4702.402828.44

1167.61

974.98

540.26

339.37

2120.61

516.89

937.28

8941.02

13666.04

415.29

12090.36

206.71410.28

2392.05

3567.78

3730.655730.87

867.42

6582.95

4260.461933.45

5018.97

3831.97

6490.64

660.87

236.07

90.80

341.49

28N/26W

24N/23W

5N/28E

1N/25E

BARBY FRY UNIT # 1-19

BUIS A # 1-8

LYNCH # 1-1

BEARD B # 1-9BEARD A # 1-8

DORMAN # 1-1

JUDY B UNIT # 1-5

KAMAS A # 1-11

KAMAS A # 2-11

DIXON # 2-29BARBY UNIT # 1-30

GIRK B # 7-35

GIRK B # 1-35GIRK B # 6-35

MILES H # 2-36

MILES N # 1-4

MILES J # 1-5MILES G # 3-31

MILES G # 1-31MILES H # 1-36

FICKEL B # 1-17

FERGUSON D # 1-18

FERGUSON E # 1-7FERGUSON F # 2-1

FICKEL D # 1-2 FERGUSON F # 1-1

BASS D # 2-14

FICKEL C # 1-12FICKEL C # 2-12

SHADDEN B # 1-25

SHADDEN B # 2-25 SHADDEN C # 1-30

BASS C # 1-12

BASS C # 2-12 ROYER A # 1-18

SPRANGLER A # 1-4

CONNER A # 1-3

BAGGERLY A # 2-32BAGGERLY A # 1-32

BAGGERLY C # 1-3BAGGERLY C # 2-3

SHADDEN E # 2-35SHADDEN E # 1-35

WOODBURY # 1-2

WOODBURY B # 1-36

SHADDEN D # 1-31

MILLER # 1-9HEADLEE # 1-8

HAMLIN, CHARLOTTE # 1-14

SHUMAN # 1-26

HIERONYMUS, J L # 1-18

RECTOR # 2-15

HARRISON # 1-20

HARRISON # 1-19ELLIOTT # 1-24

SCOTT # 1-30

TAFT # 2-10

ROBERT # 2-11

ROBERTSON # 1-16

TAFT # 3-10

NINE # 4-15

DUNAWAY # 2-32

CRAWFORD C # 2-27 CRAWFORD B # 1-26

BARKER # 2-20BARKER # 1-17

MCCLUNG B # 2-14MCCLUNG D # 3-15MCCLUNG D # 2-15

MCCLUNG C # 3-10

WHISENANT # 1-21

SHARP A # 1-15

BLAKEMORE # 1-11

STAPP # 1-16MILES F # 1-15

RIDGEWAY A # 1-7

EXLINE, HAZEL # 1-12

MOBERLY GAS UNIT # 1-18

JONES # 1-26

DUNAWAY B # 1A-9

MULBERRY A # 1-11

MCCLUNG D # 1-15

WHITE, MANION # 1-29 HELFENBEIN # 1-29

CRIGLER # 1-32

LAVERTY, ALBERT # 1-19

CATER # 1-4

MILLER # 1-4

GONSER, E.G. # 1-27

CRAWFORD C # 1-27

CRAWFORD B # 2-26

CRAWFORD B # 4-26

GIBSON # 3-25

CARVER # 1-31 CARVER # 2-31

SCHONLAU # 2-32

DUNAWAY # 1-32

LENZ, RUTH J # 1-7

BARKER # 1-20

SHUMAN # 1-21

MCCLUNG B # 1-14

MCCLUNG A # 1-11MCCLUNG A # 2-11

MCCLUNG C # 1-10MCCLUNG C # 2-10

Harper

Beaver

EllisAdvanced Resources International

Location of Study WellsDOE Stripper Well Project

Oneok Operated, Mocane-Laverne Field

Author:

Anne Taillefert

Date:

December 5, 2001Scale:

1’’ = 111,000’

M orrowC hesterH ooverT onkaw aM ultiple

P roducing Com pletions

M orrowC hesterH ooverT onkaw aM ultiple

P roducing Com pletions

Gas EUR (MMcf)

0-500

500-1000

1000-2000

2000-5000

> 5000

9%

16%

42%

33%

HooverTonkawaMorrowChester

32%

9%

17%

42%


SP082003-2107 5

Figure 4: Well Vintage Summary

Further insights into that cross-section of well types are provided in the following series of figures. Figure 5 shows that most stimulation treatments performed on the wells were hydraulic fracture treatments. Those that were matrix (acid) stimulation treatments tended to be in the Chester (carbonate) horizon. Some completions never did receive a stimulation treatment, and represent potential restimulation opportunities.

Figure 5: Breakdown by Stimulation Type

0

2

4

6

8

10

12

14

16

18

1950-1960 1960-1970 1970-1980 1980-1990 1990-2000Date

Freq

uenc

y


72%

12%

16%

FracMatrixNone

Stimulation Opportunity?

21 Completions

(19 wells)

72%

12%

16%

FracMatrixNone


21 Completions

(19 wells)


21 Completions

(19 wells)


SP082003-2107 6

In general, the hydraulic fracturing fluid was either acid, diesel/oil, or water, which was used in various forms – untreated, foamed or gelled (Figure 6). The proppant volumes used were generally quite small – less than 25,000 lbs per treatment (Figure 7). These cases too may represent restimulation opportunities.

Figure 6: Stimulation Fluids

Figure 7: Proppant Volumes

35%

4%

4%6%1%

17%

8%

6%

19%

AcidDeisel/OilWaterFoamed AcidFoamed Diesel/OilFoamed WaterGelled AcidGelled Deisel/OilGelled Water

0

10

20

30

40

50

60

70

80

0-25 25-50 50-75 75-100 >100Proppant Volume, Mlbs

Freq

uenc

y

Restimulation Opportunity?

(68 completions, includes acid jobs)

0

10

20

30

40

50

60

70

80

0-25 25-50 50-75 75-100 >100Proppant Volume, Mlbs

Freq

uenc

y






SP082003-2107 7

69

22 20

38

15

35

211

0

10

20

30

40

50

60

70Fr

eque

ncy

<1 1 - 10 10 - 100 >100Liquid Production Rate, bbls/mo

OilWater

36%

21%

43%FlowingPlunger LiftPumping Unit

Artificial Lift Opportunity?

36%

21%

43%FlowingPlunger LiftPumping Unit

Artificial Lift Opportunity?

The wells tend to produce small volumes of either oil or water (or both), Figure 8, but not all wells were equipped with artificial lift systems (Figure 9). Again, these cases might represent attractive production enhancement opportunities.

Figure 8: Liquids Production

Figure 9: Liquid Producing Configurations


SP082003-2107 8

21

35 34

97

0

5

10

15

20

25

30

35

Freq

uenc

y

0-25 25-50 50-75 75-100 >100Current Flowing Pressure, psi

AverageRange

40 – 780

20 – 400

90 – 390

40 – 170

180Chester

140Morrow

190Tonkawa

90Hoover

Pressure* (psi)HorizonAverageRange

40 – 780

20 – 400

90 – 390

40 – 170

180Chester

140Morrow

190Tonkawa

90Hoover

Pressure* (psi)Horizon

The estimated current reservoir pressures for each horizon, based on annual 24-hour shut-in tests, are provided in Table 3. For comparison, the distribution of flowing pressures for the wells is presented in Figure 10. Combining the information presented in these two figures, Figure 11 presents the distribution of pressure drawdown being achieved in the wells; only about one-third of the wells are on some form of compression (Figure 12). Well’s with low drawdown (high flowing pressures), yet not on compression, may represent another production enhancement opportunity.

Table 3: Reservoir Pressures

Figure 10: Distribution of Flowing Pressures


SP082003-2107 9

Figure 11: Distribution of Pressure Drawdowns

Figure 12: Utility of Compression A review of well performances by completion interval is presented in Figure 13. Noteworthy is that the Morrow appears to be the best performing interval overall, and that the average completion yields an impressive estimated ultimate recovery (EUR) of 3.8 Bcf. Current well performances are, however, quite low (hence the stripper well status). Figure 14 demonstrates that the majority of all completions currently produce less than 2,000 Mcf/month (66 Mcf/d).

21

2932

19

5

0

5

10

15

20

25

30

35

Freq

uenc

y

0 - 50 50 - 100 100 - 150 150 - 200 >200Pressure Drawdown

64%

36%

Not On CompressionOn Compression

Compression Opportunity?

64%

36%

Not On CompressionOn Compression

Compression Opportunity?


SP082003-2107 10

Figure 13: Overall Completion Performance

Figure 14: Current Completion Performance

0

2

4

6

8

10

12

14

16

18

20

0-0.5 0.5-1.0 1.0-5.0 5.0-10.0 >10.0EUR, Bcf

Freq

uenc

y


0

5

10

15

20

25

30

0-2000 2000-4000 4000-6000 6000-8000 8000-10000 >10000

Gas Rate, Mcf/m

Freq

uenc

y



SP082003-2107 11

In summary, the study wells represent historically good-performing, yet now older wells that have become depleted. The wells now also suffer from liquid production, which can inhibit gas flow. But, small-sized or no stimulation in some cases, the absence of artificial lift, and limited compression installation suggests that production enhancement opportunities do exist. The key objective of the project is to identify and rank those opportunities with more clarity and precision.


SP082003-2107 12

3.0 Remediation Candidate Selection Building upon the findings from the earlier R&D, the candidate selection procedure involved a combination of engineering, pattern recognition, and heuristic approaches. The specific procedure was:

o Perform engineering (type-curve) analysis to identify potential candidates. o Perform artificial neural network (ANN) and genetic algorithm (GA) analyses to identify

potential candidates.

o Combine the results of the first two steps, plus the findings from the data exploration presented in Section 2.0, in a heuristic process to identify and prioritize production enhancement opportunities.

The procedures followed and results for each of these three steps are presented below. 3.1 Type-Curve (TC) Analysis The ideal approach for selecting production enhancement candidates is to understand the relative contribution of reservoir quality and completion/production practices on the performance of each individual well, and select the wells with high upside potential for remediation. An approach to accomplish this is through the use of production type-curves8. Type curves can provide estimates of reservoir permeability, skin and drainage area from relatively limited (production) data. There are several limitations with this technique, however. First, the typical models are for single-layer reservoirs, and the multi-layered nature of most tight sand plays render the results suspect. Second, the noise-level of the production data normally available, plus the inherent interdependencies of the output parameters, makes achieving a unique result difficult. Finally, because this method requires values of net pay, porosity, fluid saturation and other reservoir parameters for each well, some interpretive and potentially labor-intensive petrophysical evaluation is required, and the errors associated with such interpretations are introduced into the process (especially problematic in tight formations). Recognizing these limitations however, such approaches have been beneficial in a relative sense to identify production enhancement candidates. The first step in the type-curve matching process for this project was to assemble the input data required to analyze on each well. Such data included producing history (volumes and pressures), and reservoir properties (i.e., net pay thickness, initial pressure and temperature, porosity, water saturation, and fluid properties, among others). That data was obtained from both well files and public data sources, and is listed in Table 4. It would have been preferable if this data, specifically net pay, porosity and water saturation, would have been available on an individual well basis.


SP082003-2107 13

Oneok Gas Wells - Formation Pressure Gradient vs. Time

R2 = 0.9717R2 = 0.9752

R2 = 0.8058

R2 = 0.9247

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

Oct-54 Mar-60 Sep-65 Mar-71 Aug-76 Feb-82 Aug-87 Jan-93 Jul-98 Jan-04Time

Pres

sure

Gra

dien

t, ps

i/ft

Hoover psi/ftTonkawa psi/ftMorrow psi/ftChester psi/ftLinear (Hoover psi/ft)Linear (Tonkawa psi/ft)Linear (Morrow psi/ft)Linear (Chester psi/ft)

Table 4: Parameters for Type Curve Matching

Hoover Tonkawa Morrow Chester Initial Pressure Actual pressure if available, from correlation if not. Net Pay 18 ft. Porosity1 Water Saturation2 Temperature1

18%

65%

112°F

18%

30%

130°F

12%

34%

150°F

8%

37%

154°F Gas Gravity 0.68 Producing Pressure Last measured value 1 – Gas Atlas value (reference 7) 2 – Well data value (from well files) A particularly noteworthy parameter for each well was the initial reservoir pressure. Since the wells were drilled over a large time period, during which time active development occurred, the initial reservoir pressures were very different depending upon when a particular well was drilled. Therefore, a correlation of reservoir pressure versus time was prepared using the annual 24-hour pressure surveys. The results of that correlation, by formation, are presented in Figure 15. If an actual pressure survey was available for a particular well when it was completed, that was the value used for the type curve matching. If not, then the value from the correlation, at the appropriate date and for the appropriate formation, was used.

Figure 15: Reservoir Pressure Correlation


SP082003-2107 14

Using these data, type curve matches for each completion in the dataset were generated. As mentioned earlier, due to quality of production data in general (which can be very noisy), type curve matching involves a high degree of interpretation. As an example, Figure 16 presents what would be considered an “excellent-good” and a “fair-poor” match in terms of quality. One can easily envision that (many) other interpretations could also result from the same data. When all matches were complete, a summary of match quality was prepared (Figure 17). Clearly, the overall quality of the results was less than preferred.

Excellent-Good

Fair-Poor

Figure 16: Example Type Curve Matches


SP082003-2107 15

Figure 17: Summary of Match Quality – All Study Wells The results of the type curve matches are presented in Figures 18 through 20. Note that these results are on a production stream basis, not necessarily an individual completion or well basis, as many wells had commingled completions. Figure 18 provides the permeability-thickness results, and indicates an average value of 28 md-ft. Figure 19 provides the thickness-drainage area results (i.e., effective reservoir volume), and indicates an average value of 31,000 acre-ft per well. Finally, Figure 20 provides the effective fracture penetration results (assuming an infinite conductivity fracture), indicating an average ratio of drainage area to fracture penetration of about 8 (i.e., the average fracture penetrated 10-15% of the distance to the drainage boundary.

Figure 18: Permeability-Thickness Results

GoodFairPoor

20%

35% 45%

Note: No “excellent” matches.

GoodFairPoor

20%

35% 45%

GoodFairPoor

20%

35% 45%

Note: No “excellent” matches.

1

60

39

60

10

20

30

40

50

60

Freq

uenc

y

<1 1 - 10 10 - 100 >100Permeability-Thickness, md-ft

Average = 28 md-ft

1

60

39

60

10

20

30

40

50

60

Freq

uenc

y

<1 1 - 10 10 - 100 >100Permeability-Thickness, md-ft

Average = 28 md-ft


SP082003-2107 16

Figure 19: Thickness-Drainage Area Results

7

5544

05

101520253035404550

1

Xe/Xf

Figure 20: Fracture Penetration Results

1

45

54

6

0

10

20

30

40

50

60

Freq

uenc

y

<1000 1000-10,000 10,000-100,000 >100,000Drainage Area x Thickness, Acre-ft

Average = 31,000 acre-ft

1

45

54

6

0

10

20

30

40

50

60

Freq

uenc

y

<1000 1000-10,000 10,000-100,000 >100,000Drainage Area x Thickness, Acre-ft

Average = 31,000 acre-ft

<5 5-10 >10


SP082003-2107 17

To compute production enhancement potential, a well performance cross-plot was created (Figure 21). Here, a normalized well performance indicator – in this case the estimated ultimate recovery divided by the reservoir drawdown – was plotted against a reservoir volume & transmissibility function – specifically the product of effective reservoir volume and permeability. The result is a simple, yet reasonably representative model of well performance for the +/-100 wells in the dataset. Wells with performances that fell short of the model expectations (i.e., those that fall below the trend line) were identified as production enhancement opportunities. The ranking of those opportunities are based on the absolute magnitude of the deviation from the trend line (note that the well performance indicator is plotted on a logarithmic scale). This approach essentially “lumps” all three categories of production enhancement – restimulation, artificial lift, and compression – into one assessment, and does not distinguish which type of enhancement opportunity each well represents.

Figure 21: Well Performance Cross-Plot 3.2 Artificial Intelligence (AI) Analysis The second analytic technique utilizes virtual intelligence, specifically artificial neural networks (ANN) and genetic algorithms (GA), to identify production enhancement candidates. Artificial neural networks are utilized to recognize highly complex patterns in how various input parameters (e.g., geologic, drilling, completion, stimulation, and workover data) impact the output (i.e., production). The relative contribution of “uncontrollable” geologic/reservoir parameters can thus be separated from “controllable” drilling, completion, and stimulation

y = 23.322x 0.4767 R 2 = 0.7924

1.00

10.00

100.00

1000.00

10000.00

100000.00

10 100 1000 10000 100000 1000000 10000000

A*k*h, md-acre-ft

TC Incremental

y = 23.322 x 0.4767 R 2 = 0.7924

1.00

10.00

100.00

1000.00

10000.00

100000.00

10 100 1000 10000 100000 1000000 10000000

A*k*h, md-acre-ft

Type Curve Incremental

EUR

/(Pi –

Pw

f), M

cf/p

si


SP082003-2107 18

parameters. This is, in effect, the separation of reservoir and completion components (i.e., permeability and skin). Genetic algorithms are then used to “optimize” the “controllable” input parameters for each well, and those wells where the greatest discrepancies exist between actual well performance and optimized performance are identified as production enhancement opportunities. In this case, the actual input parameters selected for the ANN model and their units are listed in Table 5; the output (dependent) variable was the estimated ultimate recovery (EUR). Data is generally sought to fill four types of data: space & time, reservoir, completion/stimulation & production practices. Noticeably lacking from the dataset was reservoir quality information. At a minimum, pay zone thickness is desirable on an individual well basis, if not porosity, fluid saturations and permeabilities, if available. Two models were actually constructed for this analysis; one for wells that had been previously stimulated, and one for wells that had not been previously stimulated. Both models were fully-connected, feed-forward/back-propagation models with three layers (input, hidden, output). For the first model, 93 wells existed, of which 62 were used for training the model (66%) and 31 were used to test it (33%). For the second model, only 13 wells existed – 9 were used for model training (69%) and 4 for testing (31%).

Table 5: Input Parameters for ANN Model Space & Time Completion/Stimulation

○ X (Long) ○ Y (Lat) ○ Perf Depth for each Horizon (ft) ○ Date of First Production (date)

○ Treatment Interval (depth) ○ Treatment Type (fracture, acid) ○ Fluid Type (gelled, acid, water) ○ Fluid Volume (gals) ○ Proppant Volume (lbs) ○ No. Stages (#)

Reservoir Production Practices ○ Zone (name) ○ Net Perf Thickness (ft)

○ Producing Method (flowing, gas, lift, pumping)

○ Flowing Pressures (psi) ○ Last Rate (Mcfe/d)

Output (dependent) parameter – EUR (Mcfe).


SP082003-2107 19

The model predictions versus actual well performance for the first model (stimulated wells) are presented in Figure 22. The correlation coefficients indicate that a reasonably good match to actual results is achieved by the model.

Figure 22: Predicted versus Actual Well Performance, Stimulated Wells Model

The genetic algorithm analysis was then performed to determine restimulation potential for the wells. Specifically, for each well, the following stimulation parameters were optimized:

o Stimulated (Y/N) o Stimulation Fluid

o Fluid Volume

o Proppant Volume

The magnitude of any differences between actual well performance (EUR) and predicted performance with optimized stimulation parameters was the ranking criteria for restimulation opportunity. To estimate the production opportunity represented by artificial lift and compression, the performance of each well was first modeled assuming the installation of a pumping unit, and second by reducing the flowing pressure to 28 psi (the average value for wells on compression). The total incremental potential for all three opportunities – restimulation, artificial lift and compression-was then summed and ranked on that basis. Note that in this case, the total incremental could be allocated between the three production enhancement categories.

Actual vs. Predicted EURStimulated Wells Model

0

5,000,000

10,000,000

15,000,000

20,000,000

25,000,000

30,000,000

0 5,000,000 10,000,000 15,000,000 20,000,000 25,000,000 30,000,000

Actual EUR, Mcf

Pred

icte

d EU

R, M

cf

TestingTraining

Training R-squared = 0.99Testing R-squared = 0.87


SP082003-2107 20

3.3 Candidate Screening Process ARI’s earlier work has shown that no single analytic approach on its own is as effective at identifying production enhancement candidates as the integration of multiple techniques. Therefore a heuristic approach to the selection of candidate wells was applied. The criteria for selection included the results of the data exploration, the type-curve analysis, the artificial intelligence analysis, and also the most recent producing rate (prior R&D indicated that the better the most recent rate, the better the candidate a well is for production enhancement). These criteria are listed in Table 6.

Table 6: Candidate Screening Criteria Based on these criteria, two lists (“A” and “B”) were created. The “A” list represented wells that had at least four “hits” of the above criteria, and the “B” list were wells with at least three “hits” each. Those wells are presented in Tables 7 and 8.

Not Compressed with <50 psi current drawdown (18

Data Exploration

Unstimulated (21 completions, 19 wells) Flowing with>100 bbls/mo (3 production streams)

production streams)

Type Curve (TC) Incremental (top 45 wells)

Artificial Intelligence (AI) Incremental (top 45 wells) Stimulation Artificial lift Compression

Current Producing Rate (top 45 wells)


SP082003-2107 21

Table 7: Group “A” Candidate List Data Exploration Opportunities Top 45 Rank

Well Stimulation Artificial Lift Compression TC Incremental

AI Incremental*

Last Rate

Girk B 1-35 N Y N 9 28 (95.0.5) 31

McClung A 1-11 Y N N 1 23(100.0.0) 1

Miles H 2-36 Y N N 5 41(97.0.3) 38

Schonlau 2-32 N N Y 4 31(87.0.13) 44

Taft 2-10 N Y N 30 16(83.7.10) 3

*Breakdown = % Stim. % pump.% compr.

Top Candidate

Table 8: Group “B” Candidate List

Data Exploration Opportunities Top 45 Rank

Well Stimulation Artificial Lift Compression TC Incremental

AI Incremental

Last Rate

Barker 1-17 N N N 43 43(100.0.0) 23 Cater 1-4 N N N 3 13(100.0.0) 32 Crawford B 1-26 Y N N N 30(100.0.0) 28 Crawford B 2-26 Y N N N 15(100.0.0) 35 Fickel B 1-17 Y N N N 5(92.0.8) 6 Fickel C 2-12 Y N N 37 36(83.0.17) N Hieronymus 1-18 N N Y 32 7(58.1.40) N Jones 1-26 N Y N 33 N 24 Judy B 1-5 Y N N N 3(81.6.13) 22 Kamas A 2-11 Y N N 7 45(57.43.0 N McClung B 1-14 Y N N N 40(100.0.0) 39 McClung C 1-10 Y N N N 8(100.0.0) 33 Miles G 1-31 Y N N N 42(96.0.4) 34 Moberly Unit 1-18 Y N N 11 25(100.0.0) N Rector 2-15 Y N N 22 14(89.0.11) N Scott 1-30 Y N N 45 1(88.0.12) N

Shuman 1-21 Y N N N 33(100.0.0) 2


SP082003-2107 22

Interestingly, from the breakdown of production enhancement opportunity by type, based on the artificial intelligence analysis, the greatest opportunity for production enhancement is via restimulation, and not with artificial lift or compression. With this understanding, within each candidate group the wells were further prioritized according to the following “rules”:

o Has an unstimulated horizon in the well. o Ranks in the top 45 based on both artificial intelligence and last rate.

o Is in the top 30 ranking for at least one of the above criteria.

Using these rules, certain wells were highlighted as candidates in each of the two groups, and are indicated by boxes. These represent the top priority candidates for production enhancement.


SP082003-2107 23

4.0 Field Implementation Results

Subsequent to the candidate prioritization analysis, Oneok implemented a workover program to evaluate the selections and enhance field production. In total, six workovers were performed. The selection of wells on which workover treatments were performed was based on both the results of this study, and Oneoks’ established procedures.

The workover results are presented in Table 9. For this set of wells, one was from candidate selection Group “A”, one from Group “B”, and the remaining four were not selected by the process. To evaluate the success of each workover, the pre- and post-workover production was analyzed by Oneok, to arrive at an incremental EUR as a result of the workover. Note that the EUR estimates presented are net to Oneok, not gross. In addition, the net cost to Oneok (not gross) was documented, and the cost of added reserves computed in units of $/Mcfe.

Table 9: Workover Results

Group Well Completion Date Type Incremental EUR, Net to

Oneok (MMcfe)

Cost Net to Oneok

($)

Reserve Cost

($/Mcfe)

A Taft 2-10 Morrow Jul-01 Paraffin Treatment

133 $8,149 $0.06

B Jones 1-26 Morrow Aug-01 Acidize 816 $11,117 $0.01 n/a Beard 1-8 Morrow &

Chester Jan-02 Gas Gun 104 $97,700 $0.94

n/a Beard 1-9 Morrow & Chester

Apr-02 Gas Gun 0 $38,858 infinite

n/a Gonser 1-27 Chester Mar-03 Gas Gun 22 $35,894 $1.63 n/a Miles G 3-31 Morrow &

Chester Feb-02 Gas Gun 2 $78,400 $39.20

Here, the two selected candidate wells had by far the lowest cost of added reserves, suggesting that the candidate selection procedure did, in fact, correctly select “good” candidates. In contrast, the wells not selected by the process had considerably higher costs per unit of added reserves. Workovers on the two selected candidates were “successful” based on criteria of a reserve cost of <$0.50/Mcfe, and the other four were “unsuccessful”. The two candidate wells provided almost a Bcf of added reserves at a cost of about $20,000, for an average reserve cost of $0.02/Mcf. This is a promising result given the noisy production data and lack of detailed reservoir information on an individual well basis.

An aspect of these results that must be considered is that each of the workovers performed on the “non-candidates” were Gas Gun stimulation treatments, which were not performed on any of the “candidates”. One might come to the alternate conclusion that this type of stimulation is simply not effective, and not necessarily that the candidate selection process was accurate.

In addition to the workovers, pressure buildup tests were also run on six wells. The results of those tests are provided in Table 10. Two of the tests were performed on wells from candidate selection Group “A”, three were from Group “B”, and one well was not a candidate based on the analytic process described here. Generally speaking, for a well to be a “good” stimulation (or restimulation) candidate, the desired traits are high formation permeability and a high (positive)


SP082003-2107 24

skin factor. To visually assess the occurrence of these traits, by Group, skin factor was plotted as a function of formation permeability (Figure 23). In this plot, the best candidates would be located in the top-right of the graph, and the least-likely candidates in the bottom-left.

Table 10: Pressure Buildup Test Results

Group Well Completion Date Permeability (md)

Skin Factor Remarks

A Girk B 1-35 Chester Aug-02 3.6 2.3 A Miles H 2-36 Morrow &

Chester Aug-02 0.9 -3.8 Chester only

B Cater 1-4 Chester Jan-02 0.6 -4.7 B McClung B 1-14 Hoover Mar-03 0.1 -1.8 B Shuman 1-21 Hoover Jul-02 262 -0.5

n/a Kamas A 1-11 Morrow & Chester

Jul-02 1.4 6.8

n/a (Kamas A 1-11)

A Girk B 1-35

B Shuman 1-21

B McClung B 1-14

A Miles H 2-36

B Cater 1-4

-6

-4

-2

0

2

4

6

8

0.1 1 10 100 1000

Permeability, md

Skin

Fac

tor

Figure 23: Skin Factor vs. Permeability


SP082003-2107 25

Examination of the figure suggests that the candidate selection process yielded unclear results. On the one hand, it correctly selected two wells (Girk B 1-35 and Shuman 1-21), but apparently mis-diagnosed the remaining four (selecting three “non-candidates” and not selecting one “candidate”). Additional workovers on these wells, some of which are already planned, will help shed light on this issue.


SP082003-2107 26

5.0 Conclusions Based on the results of this work, the following conclusions have been drawn:

The Mocane-Laverne field appears to hold production enhancement potential. That potential appears to be greatest for stimulation (or restimulation), with artificial lift and compression being secondary opportunities.

A production enhancement candidate selection procedure, based on engineering type curves,

virtual intelligence and heuristic screening processes, was applied at the field, and opportunities were prioritized. The accuracy of the candidate selections were tested against the results from six actual workovers.

Two of the six workover wells were candidates based on the selection procedure, and the

other four were not. Workovers on the two selected candidates were “successful” based on criteria of a reserve cost of <$0.50/Mcfe, and the other four were “unsuccessful”. The two candidate wells provided almost a Bcf of added reserves as a result of the workovers at a cost of about $20,000, for an average reserve cost of <$0.02/Mcfe.

Even with a lack of reservoir data, and noisy production data, the overall methodology

(which combined data exploration, engineering type curves, artificial intelligence and heuristic screening) successfully identified remediation candidates. The procedure therefore appears to be effective at identifying production enhancement opportunities even when only limited data is available.


SP082003-2107 27

6.0 Acknowledgements This work was funded by the U.S. Department of Energy, National Energy Technology Laboratory, under grant number DE-FG26-00NT40789. The DOE technical project manager was Gary Covatch. The authors would also like to acknowledge and appreciate Oneok Resources, for agreeing to participate in this project and allowing the results to be published.


SP082003-2107 28

7.0 References 1. Reeves, S.R., Morrison, W.K., Hill, D.G.: "Pumps, Refracturing Hike Production from Tight Shale Wells", Oil and Gas Journal, February 1, 1993. 2. Kuuskraa, V.A., Reeves, S.R., Schraufnagel, R.A., Spafford, S.P.: "Economic and Technical Rationale for Remediating Inefficiently Producing Eastern Antrim Shale and Coalbed Methane Wells", SPE 26894, Proceedings of the SPE Eastern Regional Meeting, Pittsburgh, November 2-4, 1993. 3. Lambert, S.W., Saulsberry, J.L., Reeves, S.R.: "Coalbed Methane Productivity Improvement/Recompletion Project in the Black Warrior Basin", GRI Topical Report 95/0034, February 1996. 4. Reeves, S.R.: "Assessment of Technology Barriers and Potential Benefits of Restimulation R&D for Natural Gas Wells", GRI Final Report 96/0267, July 1996. 5. Reeves, S.R., “Natural Gas Production Enhancement via Restimulation,” GRI Final Report 01/0144, June, 2001. 6. Reeves, S.R., Bastian, P.A., Spivey, J.P., Flumerfelt, R.W., Mohaghegh, S., Koperna G.J.: "Benchmarking of Restimulation Candidate Selection Techniques in Layered, Tight Gas Sand Formulations Using Reservoir Simulation", SPE 63096, Proceedings of the SPE Annual Technical Conference and Exhibition, Dallas, October 1-4, 2000. 7. Texas Bureau of Economic Geology, “Atlas of Major Mid-Continent Gas Reservoirs,” report prepared for GRI, 1993. 8. Cox, D.O., Kuuskraa, V.A., Hansen, J. T.: “Advanced Type Curve Analysis for Low Permeability Gas Reservoirs,” SPE 35595, Proceedings of the Gas Technology Conference held in Calgary, Alberta, April 28 – May 1, 1996.

selection and treatment of stripper gas wells for production

Documents