Field Experiences with V -Cone Technology II

Philip A. Lawrence McCrometer Inc

O&G Office Houston

TX. Abstract : The V-Cone differential pressure meter was introduced in the early 80’s. Initial customer review of the technology at that time was skeptical due to the radical change in emphasis of the flow regime from a central portion of a closed conduit (orifice plate) to fluid velocity profile around a centrally mounted cone at that time unheard of. Today much of the skepticism has passed due to the acceptance of large number of in field units > 40,000 This paper describes the principles and field use of the V-cone D.P. flow meter used in the role of wellhead metering, injection measurement allocation metering / custody transfer both topside and sub-sea in on-off-shore oil and gas production applications. Introduction : The V-Cone meter is a differential pressure device which, allows the fluid to be measured to pass around the exterior of a central dual cone the larger cone apex being upstream of a beta edge, and an acute angled smaller transition cone being on the downstream side of this edge. Flow velocities in the up-stream core of the pipe are then forced to mingle with pipe wall boundary layer velocities by radial expansion across the larger cone.

After passing this beta edge (formed between the two cones and the pipe wall ) the fluid then falls into a central region in which high frequency vortices form adjacent to the pipe wall and pipe center with a low amplitude. This parameter is particularly beneficial if trash laden fluids are to be measured which will be mentioned later in the paper. A significant flow conditioning/profiling affect is obtained by the geometry of the larger cone which cause an averaging of the fluid velocities directly up-stream of the cone beta edge. Independent testing has show this affect to occur with close coupled out of plane elbows, and single elbows at Reynolds numbers to 8,000-10,000 and above. Consideration must be given to any installation that employs an upstream disturbance producer other than an elbow , such as a gate valve or device which can produce jetting or high velocity fluid distortion past the cone. Practical experience, has shown that a 3-4 diameter upstream section will suffice to allow the sufficient fluid expansion needed to allow the cone to regenerate the velocity profile. (Fig 1)

Pressure Distribution

High Low Fig 1

Standard design concepts V-Cone with Welded Construction : This type of construction uses a support tube which acts as both a support and low pressure conduit connected to the large cone a horizontal low pressure tube passes through the center of the cone assembly and is attached to the support tube ( Fig 2). Machined Construction : A machined mono-block in stainless steel or other material is manufactured to allow a cone to be supported by a cross piece which can be removed allowing different cone diameters to be implemented. This allows flexibility of the metering system in respect to large turndowns. The meter is flangeless and sits between pipe flanges. see : fig 3

Installation Effects Pipe flow velocity profiles are rarely ideal, there are many installations where flow-meters exists in which the flow is not well developed. Trying to measure disturbed flow can create some problems in certain types of DP device. The V-Cone overcomes this problem by re shaping and distributing the up-stream velocity profile. ( Fig 4 ). This feature enables the meter to be placed in a small space with minimum straight run requirement. Thus saving real estate . Production and Test Separators Current design of multiphase separators allow overall uncertainty (on all three phases) of about 15% - 20% according to recent API /MMS comments. This can be due to operator control, time delay in stabilization of the vessel, incorrect design involving fluid levels, position of vessel in respect to the pressure head requirement (on liquid side) and also the main metering carry over of other product fractions due to undersizing . In particular where orifice plates are used it may be necessary to perform, plate changes to facilitate turndown (otherwise the performance of the measurement system could be compromised), and to use upstream

Fig 2

H . P. Port L.P. Port

High Pressure Low Pressure

Interchangeable Cone

Reshaping the Velocity Profile

Fig 4

( Fig 3 )

flow conditioners or velocity profile devices which add cost to a system. Long-term vulnerability using orifice plates in production separators can be demonstrated by examining public documents in the measurement field *. (*Examples cited in this document from Phillips Petroleum Embla–Platform NSFMW Gas Measurement for the Real World 1994) Orifice Technology issues: Beta Edge Degradation, Stagnation Area, and Deposition. During the use of a velocity profile sensitive device (as an orifice) it is necessary to confirm that the beta edge of the device is both clean and has not been compromised due to trash & debris. Current API paper standards address a clean dry and non-polluted gas, which does not take into account generally an offshore usage condition i.e.:-”wet gas.” This can result in high maintenance and intervention cost due to frequent plate changes. Where paraffin’s or sulphate deposition occurs in the pipe in almost all cases shift in the Cd is caused due to a re-circulatory effect at the rear of the orifice plate. There also can be a stagnation area up-stream of the device where heavy end hydrocarbons can collect. (see enclosed Photo examples from Marathon Oil site Appendix A and longevity issues later in the paper ) Wet Gas: Wet Gas Metering is a topic that has become more important to industry in recent years. However, the term "Wet

Gas" does not have an accepted definition in the oil and gas industry. The loose definition that is commonly agreed upon is that "Wet Gas" states the existence of relatively small volumes of liquid in a gas flow. Dry custody Gas contracts are sometimes applied to wet gas applications because of convenience, the subsequent impact in neglecting this condition can be severe during the life of the separator facility. In the natural gas production industry two separate definitions have appeared due to two separate metering conditions. These are: Undersized Separator Gas fields that start operational life producing natural gas with a set amount of entrained liquids degrade over time and the water and condensate quantity in the pipeline increases. Traditionally production natural gas flows pass through a separator and the dry gas line exiting the separator has a meter. However, with the degradation of these flows the separator is often not capable of 100% efficiency and a small quantity of liquid is entrained in the gas flow outlet (see Figure 5). Some engineers call this two-phase flow “Wet Gas”. This condition seldom has a Gas Volume Fraction or “GVF” (i.e. the ratio of the gas volume to total volume flowing) less than 99.5%.

Fig 5.0 Undersized Separator

Separator

Well 2 Wet Gas Flow

Dry Gas Sales Meter

Well 1 Wet Gas Flow

NGL/Retrograde

H2O

Well 2 Well 1

Traditional Conventional Gas Well Head

Upstream Unprocessed Flows In many conventional gas well head locations existing infrastructure is being utilised so that two or three wells may go to a common liquid separation point. This means that two-phase flows from different fields are going to the same separator. Due to increasing pressure from industry to provide more accountability of accurate flows per each well head, it may now require more attention to metering upstream of the communal pipe to the separator. Typically, the GVF of flows called “Wet Gas” in this condition can be down to an approximate limit of 90%. For example see Fig 6.0.

Fig 6.0 Upstream Two-Phase Flow Metering Situation. So when discussing “Wet Gas Metering” it is crucial that all parties agree that they are talking about the same type of flow. The V-Cone is currently capable of operating in both types of wet gas flow.

Common Wet Gas Metering Problems There are several technical problems common in typical wet gas flow metering applications. 2.1) In Gas Well Head Metering it is common for the available piping length allocated for metering to be short. It is also common for the proposed meter position to be close to pipe bends and valves and there is always the possibility of vibration and background noise. A meter must be able to operate in these adverse conditions. 2.2) In wet gas flows there may be a problem of liquid gathering in the lift system and slugging the pipe with fluid as it travels to the separator. These slugs may have large impact forces that can be applied to pipe work components. Therefore all components including flow meters must be able to survive slugging flow. 2.3) Wet Gas flows have the two-phases dispersed in the pipe in different ways depending on the flow conditions. These different dispersions are called flow patterns. Flow patterns are directly relevant to wet gas meter performances. Typical flow patterns are shown in Fig 7.0

Fig 7.0 Typical Two-Phase Flow Patterns

Separator

Dry Gas Sales Meter

To Customer

NGL/Retrograde

H2O

Well 2 Wet Gas Flow Meter

Well 1 Well 2

Well 1 Wet Gas Meter

Traditional Conventional Gas Well Head with Wet Gas Metering

Flow regimes For wet natural gas production the Annular Flow, Slug Flow and Stratified Flow are the most common flow patterns. Stratified Flow occurs in horizontal pipe lay outs only. Here the flow has separated phases, that is, the liquid runs like a river in the pipe with the gas flowing above the liquid. This flow pattern occurs with relatively low flowrates. Annular Flow is a flow pattern that exists at a higher flowrate and it consists of a liquid film around the circumference of the pipe and a gas core traveling through the pipe center with entrained liquid droplets. This can occur in any orientation and it is the most common flow pattern in wet natural gas production. In wet gas applications slug flow is caused by liquid being periodically held up and forced in “lumps” down the line. It can periodically appear when a flow conditions steady state flow pattern is any other type. 2.4) Most of the wet natural gas flows upstream of the platforms or processing facilities contain particulates. A meter has therefore to be sturdy enough to survive impact of these particulates. 2.5) In a number of wet natural gas metering situations the meter may not be in a position that is easy to reach and therefore the meter will be difficult to maintain. Therefore wet gas meters must be sturdy. The Main Types Meters Used The main meter types being developed as wet natural gas meters are Ultrasonic and Differential Pressure Meters. These are dealt with in turn here:

Ultrasonic Meters Ultrasonic Meter manufacturers are currently researching the possibility of developing an Ultrasonic Meter into a wet natural gas device but so far the published research has shown this to be an extremely difficult technical challenge. Whereas for many dry gas applications the Ultrasonic Meter is an excellent meter it is found with wet gas a number of problems arise. These include the chords flooding and therefore failing, liquid bridging the gap between transducer face and pipe wall (causing loss of signal), the signal strength being reduced by absorption in the liquid phase, the signal being deflected away from the desired path by refraction through the liquid phase and the background noise of valves etc. drowning out the signal. There is also a question on the sturdiness of the transducer faces in unprocessed flows where slugs and pressure pulses randomly occur, the temperature is often high and the flow contains potentially damaging particulates. Most of the recent wet gas ultrasonic research done on stratified flow which is difficult to re-produce in live field conditions Differential Pressure Meters Orifice Plate Meter Traditionally the Orifice Plate Meter was used to meter wet gas flows. In the last few years this has virtually stopped as it is now known that the liquid is held up at the plate and the resulting flow is not steady. The liquid tends to travel through the orifice in slugs. The result is an unsteady DP reading. This can be seen from Orifice Plate Meter wet gas photographs taken at South West Research Institution in 1997 see fig 8.0

Fig 8.0 : An Orifice Meter in a Wet Gas Flow. Furthermore, the Orifice Plate is susceptible to bending if struck by a slug or pressure pulse and the plate tends to act as a trap that gathers particulates. Appendix B shows the accumulation of debris on an orifice plate. Accumulation of Particulates on the upstream face of an Orifice Plate. Accumulation of Particulates on the downstream face of an Orifice Plate. Venturi Meter The Venturi Meter is a popular wet gas meter. It does not suffer the problems of an Orifice Plate Meter as it allows slugs and pressure pulses to pass through unobstructed due to the inlet being angled. (This feature also allows the Venturi to be self cleaning. Current Wet Gas Metering research Joint Industry Projects all include this meter in their test programs and its performance is reasonably well documented. ) A main difference between the Wet Gas Venturi Meter and the Wet Gas V-Cone Meter is that the minimum flow area (i.e. the “throat”) of the Venturi is along the

center line and the V-Cone Meters minimum flow area is at the periphery of the pipe. This gives the V-Cone an advantage as for wet gas flows as in single-phase flows the V-Cone can condition the flow as it passes the cone. The result is a steadier DP. Not only does the Venturi not condition the flow as effectively as the V-Cone it tends to hold up liquid at the inlet and therefore small slugs created by the Venturi meters presence periodically flow through the meter causing pressure spikes to be read. Venturi Meter testing in industry has led to the publication of correlations to correct for the liquid induced error. The Venturi Meters general performance is similar to the V-Cone Meters and hence these correlations are in fact very similar to the V-Cone wet gas correlations. However, the Venturi data sets tend to be slightly less stable. V-Cone The V-Cone Meter is a self cleaning device. The acceleration of the gas over the cone tends to remove any liquid and particulates that come into contact with the meter. As a visible example of that the photograph that was taken of the Wafer-Cone Meter (a type of V-Cone) which replaced the Orifice Meter shown in Figure Two is given in Figure Three. The cone had been in the same flow as the orifice plate for three times as long. The edge is clearly unaffected by the dirt in the flow. See Appendix A. The Cone is clean after being exposed for three times as long as the Orifice Plate Meter in the same application. A V-Cone Meter was installed in a wet gas test rig at South West Research Institute and the results visually showed the conditioning of the two-phase flow by the cone and the resulting stable DP.

Fig 9.0 below shows the same flow conditions as Fig 8.0

Fig 9.0 The V-Cone Meter with Wet Gas Flow.

Since these tests we have been involved in two Joint Industry Projects on Wet Gas Flow Metering. The resulting wet gas data sets have confirmed the good stability of the DP read by the V-Cone and allowed the creation of V-Cone wet gas correlations. As with previous DP Meter wet gas correlations McCrometer has produced a correlation that predicts the percentage over-reading of the gas flow to a non-dimensional number (often called the Lockhart-Martinelli Parameter) denoted by “X”. This non-dimensional number is in fact the ratio of the superficial liquid inertia to the superficial gas inertia and is calculated as shown in equation 1 below:

l

g

.

g

.

l

m

mX

where X = Lockhart-Martinelli Parameter Current AGA/API paper standards do not address wet gas installation

conditions,hence the use of an AGA/API standard to design a possible wet gas metering system can also end in higher intervention costs not generally assumed at the start up case. Greater liquid injection over time to obtain more products on declining wells can result in an increase in wetness at the gas outlet. Examples of this can range from under-reading of the meter or over-reading depending on Reynolds No, Beta ratio, and liquid mass fraction seen at the meter. Research in this field by Chevron and others indicates a Cd movement of over 2%-3% outside of the predicted API requirements due to wet gas with liquid loads of only 0.33bbl/MMscf as indicated in research. Reference: Dr V.C. Ting Chevron Corp :- Effect of Liquid Entrainment on Orifice Meters. Wet Gas Lab Test CEESI Nunn C.O. : Recent Wafer V-Cone testing at CEESI indicated a low susceptibility to Cd change with liquid load, base line values where plotted against numerous test loop instruments in a dry condition. Flow rates from 7 –70 feet / second in a 4 inch line size where used. The liquid rate was added to a maximum of 1 and 2 Bbl per MMSCF.The liquid hydrocarbon was a Decane derivative acceptable for use in closed surroundings. The results where plotted and the effects noted , further work is underway to see the effect of low D.P. ranges on repeatability and accuracy and Y factor changes ( currently the welded version has a Y factor eqution available to correct for density changes due to pressure ) ( See Appendix C fig 8 ) - (CEESI = Colorado Experimental Engineering Station Courtesy of Chevron Inc USA ).

B.P Amoco Field Test During March 2001 BP tested a V-Cone in an upstream field condition . A 4 inch wafer cone was tested with a 4 inch orifice plate 36 inches up-stream and a separator downstream . The results were very impressive. (See Appendix C fig 7 ) Longevity, Contamination, and Beta Edge Damage Lab Test During late 98 early 99 Marathon Oil installed test meters at there on-shore hydrocarbon facility in central Wyoming. The site was producing dirty wet gas with H2s and asphaltene contaminants. The result on the existing measurement system was not very pleasing to the client nor the local BLM (government) office whom collect royalty from these gas systems . The use of the V-cone was to see if the contamination would affect the meter the assumption that it would work was a driving force to implement the installation. 3 inch meters where fitted and the most severe well used as a test site. (See appendix A & “A-1:Orifice trash deposits” ) On inspection the orifice plate units showed trash build up after only three months usage with Asphaltene / Paraffin deposition at the up-stream inlet to the meter and contaminants after the plate in the low pressure region. On inspection of the V-Cone the unit (Appendix A) did not show the same severity, probably due to accelerated flow around the cone element. This seemed to keep the cone and sensing ports clear of deposition , thus maintaining a consistent D.P. across the meter. Entrained condensate liquid moved into slug flow condition periodically, which caused liquid to enter the orifice

sensing lines and also be retained after the plate. The cone meter did not show this problem due to the straight through design. The regular “ blowing “ off of the plate was deemed a severe problem in man hours and traveling to the site, plus the effect on accuracy this caused. With the lack of liquid retention using the V-cone the system now runs within the BLM (us gov) guidelines. Damage Test Damage testing of the V-cone wafer meter was recently performed, this involved determining a base line on a calibration rig over several flow rates, after which intentional damage to the cone beta edge was performed in a somewhat severe manner. The photographs and data * are shown below Figs 10 and 11 overleaf Fig 11 Test Results (damage testing) The deviation from the test shows the Cd shifted by app. + 0.3% which is within the uncertainty of the McCrometer calibration station.

This initial test is currently being superceded by further tests with multiple damage regimes to view the effect per incident. This work is a pre-courser to the use of the meter in a sub sea “non-intervention” environment. (See appendix D ) Weight Penalty ( platform design): Current design of orifice systems require large up and downstream pipe lengths, this is not good if you area user since cost of real estate is premium offshore due to the large weight related cost in installing :- large pipe runs, orifice carriers, and supporting structures. Current weight penalty costs in the Gulf of Mexico can be from app. 12$ per pound to 25$ per pound. Installing a system including all piping requirements and a cast carrier can amount to high dollar amounts in the platform support requirements. This can be compounded on deep water platforms. Therefore using a device, which has low weight and reduced up-downstream piping needs, can reduce overall client installation costs Installation cost relating to weight penalty : The V-Cone can perform its own velocity profile conditioning as part of the meter design . The installation envelope can be reduced significantly and direct coupling with elbows can be straight to the meter flange face without Cd performance

degradation due to swirl or profile skewing. Please note that the V-Cone weight is approximately only 1444 lbs for a 16inch #900 meter, 1161 lbs for 14 inch meter and only 945lbs for a 12 inch which is significantly lighter than both orifice carriers and respective up-downstream piping for these size systems ,which can weigh more than 2.5 tons, per stream. V-Cone Technology Technical Attributes: Real World accuracy using laboratory calibration is possible to facilitate an accurate performance over 10-1 turndown better thanthat of conventional differential producers. Performance of +/- 0.5% at 10-1 turndown, with trash resistance, low beta edge degradation ( Beta edge is after the flow on V-Cone) and no stagnation area make the unit ideal for high cost intervention areas and long-term performance. The low weight of the device will ensure economy of installation without having to purchase long piping lengths and high dollar flow conditioners/profilers (certain 12 inch diameter flow profile generator units can cost over $8-10,000 dollars).

SSuubb SSeeaa IImmpplleemmeennttaattiioonn aanndd DDeessiiggnn Currently 45 precision tube units are in service in a sub-sea wellhead marine environment, in Norway, Angola, and South China Seas . The main usage has been water injection metering , however, allocation gas metering has been a recently accepted philosophy with the device. Implementation to >12000 feet is acceptable with a new configuration and special sensor housing. (See Appendix D )

Appendix A

Wet Gas Photographs Wellhead Metering

Wafer Body of V-Cone 3 inch Diameter ( 9 month inspection )

Appendix A-1

Wet Gas Photographs Wellhead

Three months inspection for wet gas line (H2s + paraffin’s)

Appendix B

GasProduction Separator

( Typical Arrangement )

( BP Champlain 457 Separator Site Wy.USA. )

30 diameters

Wellhead

4 inch line

Orifice V-Cone

36 inches 12 inches

30 diameters Proved Orifice Meter With Mobile Gas Prover

Gas + Water + Hydrocarbons

Gas OUT

H2O

NGL / Retrograde

V-Cone vs. Orifice Plate & Separator

420

440

460

480

500

520

540

17:0

0:00

19:0

0:00

21:0

0:00

23:0

0:00

1:00

:00

3:00

:00

5:00

:00

7:00

:00

9:00

:00

11:0

0:00

13:0

0:00

15:0

0:00

Time

MC

FH

MCF-Vcone

MCF-Norwood

MCF-Orifice

Appendix C Data from BP/ Norwood Separator Site Wy

Wafer V-Cone CEESI Calibration Test--Wet Conditons, 0.5 Beta at 80 psig (Raw data)

0.86

0.88

0.9

0.92

0.94

0.96

0.98

0 200000 400000 600000 800000 1000000 1200000

Pipe Reynolds Number D=3.826"

Dis

char

ge

Co

effi

cien

t,

Cd

Cd-Dry

Cd-1bbl/MMSCF

Cd-2bbl/MMSCF

--TIME-- MCF-Cone

MCF-Norwood

MCF-Orifice

DP-Norwood

DP-Orifice -PSIA- -DEGF- --MMBTU--

17:00:00 504.053 504.41 489.526 153 51.5 373.5 119.9 479.708

18:00:00 509.756 496.5 495.586 148 52.5 375.6 119.9 485.135

19:00:00 519.024 526.25 505.496 165 54.7 374.9 119.6 493.955

20:00:00 496.692 504.04 484.337 152 50.6 371.7 119.5 472.702

21:00:00 495.705 484.87 482.686 141 50.3 371.9 119.4 471.762

22:00:00 500.691 493.45 488.556 145 51.3 373.2 119.4 476.507

23:00:00 502.829 506.04 491.551 152 51.9 373.8 119.4 478.543

0:00:00 497.401 503.08 485.979 152 51.2 369.8 119.3 473.376

1:00:00 491.94 496.79 480.75 150 50.2 369.4 119.3 468.18

2:00:00 487.84 490.04 476.625 145 49.1 371.2 119.3 464.277

3:00:00 495.004 491.04 484.341 144 50.6 371.5 119.2 471.096

4:00:00 500.487 504.5 490.445 152 51.8 372.3 119.2 476.314

5:00:00 488.422 486.87 477.947 142 49.4 370.7 119 464.831

6:00:00 484.779 484.45 474.372 142 48.9 369 119 461.364

7:00:00 483.022 479.67 472.455 138 48.6 368.4 119 459.692

8:00:00 479.931 483 469.689 142 48.3 366.5 118.9 456.75

9:00:00 477.279 479.58 467.129 140 47.6 367.8 118.9 454.226

10:00:00 476.231 476.83 466.232 138 46.9 370.9 118.8 453.228

11:00:00 486.98 478.54 477.817 139 50.7 361.3 118.6 463.46

12:00:00 481.302 481.29 472.392 145 49.4 362.8 118.6 458.055

13:00:00 470.997 463.2 462.513 128 45.4 376.7 118.7 448.248

14:00:00 470.623 471.2 461.351 131 0 373.2 118.7 447.891

15:00:00 470.861 474 461.711 134 45.7 373 118.7 448.119

16:00:00 499.061 477.33 489.357 136 54.2 355.3 118.6 474.957 Totals 11770.91 11736.9711508.843

Appendix D

Reference Documents / Data : Hayward A Basic Guide and Source

Book for Users 1973. Szabo / Winarski V-Cone Meter for Natural Hypnar Gas Flows 1992. Miller Flow Measurement (Latest Edition)

Handbook GRI Wet Gas Research 1997

V-Cone 4 inch Diameter

Lawrence V - Cone Technology 1998 (Old Wine in a new Bottle)

Lawrence D.P. Metering for the New 1999 Millennium Lawrence Wellhead metering using V-Cone Technology NSFMW (UK) 2000 Bright BP Amoco Wy Test Separator 2001

data

Steven Wet Gas Measurement 2002

Well 2 Wet Gas Flow Meter

H2O NGL/Retrograde

well 1 Traditional Conventional Gas Well Head ith Wet Gas Metering

To Customer Dry Gas Sales Meter Separator well 2