Sei sulla pagina 1di 13

August 2009

Field Application Comparison


Conditioning Type Orifice Plates (4-holed) and Generic Cone Meters

Mark W. Davis, Senior Staff Engineer


Shell Exploration & Production Co. Inc.
Houston, Texas USA

Philip A. Lawrence, Director of Business Development


Cameron Measurement Systems, Inc.
Houston, Texas USA

Abstract

Allocation measurement of produced hydrocarbon gas in the upstream area require different
measurement techniques and strategies than those used for standard pipeline quality gas measurement
due to the nature of the fluids in the system.

Wet gas, hydrate formation, liquid slugging, well clean-up debris and short meter runs all add to the
difficult task of collecting meaningful data from High Pressure Full Well Stream (HPFWS) allocation
metering points.

Various meter types have been employed for these flow regimes in order to meet the necessary accuracy
requirements and the arduous duty that normally occurs in these applications.

This paper details the real world experience of such a allocation metering system operating in Wyoming,
USA that has wide ranging environmental patterns from below - 30 deg F (-34 deg C) in the winter to + 90
deg F (+32 deg C) in the summer all which can have an impact on the measurement.

The operational experience and differences between the two metering technologies, conditioning orifice
plate versus cone will be detailed.

Meter selection criteria will be discussed based on facility design constraints in accordance with
regulatory and environmental requirements, allocation accuracy needs, along with sustainable and cost
effective approaches for implementation.

Data from the field operator will be shown together with performance, calibration data and computational
fluid dynamic imagery (CFD) for some meter types that were and are now being fitted to meet the
operator’s measurement philosophy.

Field Location

The location of the subject field is in West Central Wyoming consisting of remote well pad locations where
access can pose problems during the winter months due to various environmental factors such as
weather and animal migrations. From a sustainability standpoint, the remote measurement systems must
operate effectively with minimum maintenance during these months with limited access. - Figure 1
Figure 1.0

The measurement systems installed at the locations shown use full well-stream metering and are
adjusted by using a well test separator to calculate the volume correction factor for the wet gas metering.

The environmental issue is one of keeping the production facilities to the smallest footprint possible, due
to the impact on the local wildlife particularly the mule deer which use the Pinedale Anticline as a
migratory route. Figure 2 shows the typical landscape that is dealt with here. The local regulatory
authority are very stringent in managing and keeping the delicate balance needed at these locations.

Figure 2.0
Measurement Philosophy

Metering is required to fulfill a number of functions driven by regulatory, industry and internal
requirements. The metering function relates to process control, sales quantity and quality, and
hydrocarbon accounting.

Consistent with a need to simplify facilities, metering equipment is minimized to that deemed necessary to
run the business. The measurement systems are designed “Fit For Purpose” in order to reduce capital
and operating expense, and simplify volume reconciliation procedures.

Metering devices, flow conditioning equipment and ancillaries will be designed to meet the necessary
level of accuracy only. Such accuracy requirements must be in accordance with BLM and Wyoming Oil
and Gas Conservation Commission regulations, API standards, legal, accounting, Sarbanes–Oxley (Sox)
compliance, and reservoir management requirements.

Continuous readings are taken during the well testing process. The test separator gas volumes are
compared to the (HPFWS) Differential Pressure Meters (e.g. Cone meter, Conditioning Orifice Plate, or
similar devices) readings and a gas volume correction factors are calculated.

Once the gas volume correction factor (GVCF) is obtained, the gas only volumes from the HPFWS Meter
will be used to calculate liquid volumes. Liquid yields will be calculated based on the gas volume
produced during the test. The well test is used to calculate a daily well theoretical contribution percentage
for oil, water, and gas. This percentage is calculated using algorithms involving the daily readings from
the HPFWS Meter and the well test factors. Once the theoretical daily volumes are calculated, they are
compared with daily readings from the applicable sales meter for accuracy.

The daily sales are then allocated to each well based on each wells calculated contribution percentage.
The fiscal allocation occurs when the daily sales readings are replaced with final sales readings and the
same daily theoretical contribution percentage is applied.

The use of Differential Pressure Technology (D.P.) to measure the full well stream wet gas is favored
because these types of devices appear to be less susceptible to the effects caused by the multiphase
aspects of the fluid. Although there is an effect of liquid in gas for all D.P. devices, it is known from
research that the main issue here is one of repeatable results from well test to well test with minimum
impact caused by liquid load changes.

Since a correction factor / allocation factor is calculated for each well location the most important
requirements are that the full well stream meter is robust, has a predictable and repeatable operating
envelope, and is representative between well tests. The normal well test period is 24 hours once per
month per full well stream meter to determine a gas volume correction factor (GVCF) that is used for the
time till the next well test period.- Figure 3.0

Well Test Manifold System

A well test manifold is installed in the flow line downstream of the HPFWS meters. There are chokes
situated upstream of the HPFWS meters on most of the locations. Samples are taken at least semi-
annually from each gas and condensate stream. Shrinkage factors are applied to the gross volume
obtained by the well test separator condensate/oil turbine meters to determine the net stock tank
condensate volumes.
Figure - 3.0 - Typical Allocation System

Meter Correction Calculations.

During the well testing process, the volume of gas is measured at the outlet of the test vessel and
recorded as the Standard Gas Volume (SGV). The indicated Full Well Stream Volume FWSV is also
recorded during the same time frame. Then the SGV divided by the FWSV which yields a gas volume
correction factor (GVCF) that is used to ascertain the gas volumes delivered through the HPFWS meter.

Thus;

FWSMF = SGTV / GVCF

Where;

SGTV = Standard Gas Test Volume (MCF).


GVCF = Full Well Stream Standard Volume FWSMF = Full Well Stream Meter Factor.

Condensate Volume Calculations


Volumes of Condensate produced during the well test are also recorded. The indicated “gross” volumes
usually are corrected to standard conditions (net volumes) in accordance with API MPMS 20.1.latest
revision. The resulting net volume is used as well as the gas volumes from above in ascertaining the
barrels of condensate per million cubic feet of gas yield (BBL / MMCF).

Thus;

CBMCF = CTV / (SGTV / 1000)

Where;

CBMCF = Condensate Barrels per MMCF. CTV = Condensate Test Volume.


Water Volume Calculations.

Produced water volumes are ascertained in the same manner as the condensate volumes detailed
before, since water at ambient temperature is assumed to be non-compressible. The only difference in
the method is that there is no correction from gross to net volumes during the well test period required.

Full Well Stream Meter Types (A) – 4 Holed Differential Orifice Plate Devices.

Originally the full well stream meters where selected on a cost basis and 4 holed orifice devices were
used because of this mandate for the application regarding the system application. Around 250 devices
were originally used at the Pinedale locations and field data for these devices was collected over an 18
month period. It was discovered over this period that various operational issues occurred resulting in less
than satisfactory measurement results causing the operator to re-think operationally and from a
metrological standpoint using this meter design for the application.

The 4 holed orifice meter is reported by the manufacturer to be successfully operating at other worldwide
locations. The performance of these devices is based on test data supplied by the manufacturers testing
and also test data derived from some laboratory work performed to some of the API 22.2 differential test
protocol requirements.

4-Holed Orifice Plate Manufacture/Design.

Typically a circular plate is cut from stainless or other materials and 4 circular holes equally spaced on a
radius are machined with a typical equivalent area ratio of a standard orifice plate to satisfy the
normalized flow rates per diameter. These betas or area ratios are reported to be available in 0.2, 0.4 &
0.65 currently the writer has no data regarding other beta ratio types or the effect of coefficient of
discharge. The device looks like a paddle as seen from the photographs of two models Figure 4 & 5

Figure 4 Figure 5

Beta Ratio

The effective area ratio (β) for all differential pressure devices is a function of the restriction versus the
pipe area, the manufacturer’s published data sheet table for the four holed plate meter shows the beta
ratio calculation for the device. The area ratio for all types of differential producers is the square root of
the smallest free flow area (differential pressure element) / largest free flow area (pipe), for a 4 holed
orifice plate this equation is shown next.- Figure 6 - (Equation) & Figure 7 - ( Data Table).
 4 Holed Orifice Plate (β) Equation – Figure 6

o.p. 

Where d = Hole size of each of the holes

Where Pid = Pipe Internal Diameter

4-Hole dc (from Total Effective Beta


Nominal Pipe Pipe ID Orifice Plate Manufacturer Flow Per Equation
D Schedule (Pid) (o.p.) d - 0.4 Literature) for Area (Figure-6)
Beta a 0.4 Beta
3" 40 3.068 0.614 1.228 1.1844 0.400
80 2.9 0.614 1.228 1.1844 0.423
160 2.624 0.614 1.228 1.1844 0.468
Area Ratio Table Figure 7- (Four Holed O.P. Manufacturer Data Table)

Baseline and Disturbance Testing 4 Holed Orifice Plate

The effective area ratio changes with pipe schedule as can be seen from the table 5 b above below is test
data from a recent test of a 4 inch 4 holed device tested with disturbance effect testing (1/2 moon plate)at
2 , 5 and 7, Pipe Diameters (D’s) – data - Hollister & Dyer- 2008 - Figure 8

Figure 8
Full Well Stream Meter Types (B) - Generic Differential Pressure (D.P.) Cone Meter

The generic cone meter uses a beta ratio equation similar to an annular orifice, which incorporates the
flow around the outside of the cone annulus. The equation for this type of device is shown in Figure 11.

All Cone Meters (non-wall tap design) consist of a conically shaped differential pressure producer fixed
concentrically in the center of a pressure retaining pipe using a cantilever member, by which a differential
pressure can be obtained across the interface of two cone frustums via an internal port-way system.

This allows the downstream pressure P2 to be measured in the center of the closed conduit via a port-
way. The fluid is linearized through the meter throat (annulus) within a region defined by the differential
producer cone surface area and the interior surface of the closed conduit, whilst also flattening the
velocity profile in the throat region of the device – Figure 9. The upstream pressure P1, being measured
at the pipe wall.- Figure 9

Computational Fluid Dynamics Image (50% ‐ Section) 
P1 P2
Meter Tube Wall

Meter Throat

FLOW
P2
Separation
                   Figure 9 (CFD) 
Field Acceptance

The concept of using the center of the cone above to monitor the downstream pressure has shown
certain benefits from historical data compared to conventional differential pressure meters such as the
following :-

a) Flow Conditioning Effect. (shorter runs)


b) High Turndown Ratio. ( up to 10-1 )
c) Some Static Mixing.
d) Wet Gas usage at Lockhart & Martinelli Wet Gas Test Values LM = up to 0.35 & LM = 0.5.
(One test only at 0.5 by Stevens - NSFMW 2003)
e) Low Maintenance application : design is robust.

This was one of the reasons that the operator made a decision to try this type of device at the site in
Pinedale Wyoming. The low maintenance aspect was a deciding factor !

Y Factor (Expansibility Coefficient)

The Y factor equation (expansibility coefficient for gasses) for this type of device is per the Reader-Harris-
Peters, (NEL-TUV in U.K.) generated equation thus - (eq.10)

Y = 1‐ (0.649+0.696 β^4)  ……………………………………………………………..  (10)

This equation has been accepted in Industry and is shown to work for cone type meters using the
geometry shown in Figure 13 from 2 inch diameter meter sizes and larger.
Beta Ratio Cone Meter

Cone meter area ratio’s  (β’cone) are varied to accommodate the measurement of different flow rates by
changing the cone length and thus the cone diameter. This changes the effective diameter of the cone in
relation to the pipe diameter and thus the beta or effective area ratio and ultimately the velocity across the
Beta edge boundary, the cone meter equation is shown in Figure 11 below.

Generic Cone Meter (β) Equations - Figure 11 Honed Meter Body- Figure 12

D2 − d 2 & d = D2 − (D∗ β)2


βcone =
D
Generic Cone Meter β cone
Beta D d
D 2.9 2.9
d 2.42198 2.42198
Beta 0.55000 0.55

Generic Cone Meter Geometry


ΔP
    Taps:   Static                Low Pressure 

Beta edge boundary 

Figure 13

Commercially made cone meters operate generally in  Beta ratios from 0.45 - 0.85. As the beta ratio
becomes larger (approaching beta 0.8- a smaller cone diameter) the meter performance changes and the
measurement uncertainty may become greater with disturbed velocity profiles.- Figure-13

This “performance” effect is caused by the reduced interaction between the cone area and the fluid, i.e. A
smaller cone has a smaller dynamic effect on the fluid so the flow linearization aspect of the meter can be
reduced.

Care must be taken when using smaller cone diameters where valves or other flow disturbance
generators are in line and upstream of differential pressure cone device. The cone meter installations at
the facilities had short upstream lengths at 4-5 D’s, depending on the site location.
The cone meter manufacturer should be able to advise of the minimum straight lengths per Beta ratio and
diameter versus ReD, regarding this effect. The cone meters used in this application were also honed to
improve surface roughness effects and roundness - Figure 12.

The mass flow rate equation for generic cone meters is exactly the same as per any standard differential
pressure device, (Orifice,Venturi.) with the exception of the C.d. implementation and beta ratio calculation
which is usually derived from empirical testing by the manufacturer or other independent laboratories and
not generally by mathematical iteration.

Flow Conditioning Effect (Cone Meters)

It is known from research that the use of a cone shape concentrically mounted in a closed conduit (pipe)
can facilitate a flow conditioning effect by “velocity profile re-distribution”. This effect seems to occur over
quite a wide Reynolds Number (ReD) range and appears to be enhanced farther away from the transition
region (ReD:8000-10,000).

90 Deg. Out Of Plane Testing of a 4 inch Generic Cone Meter

The graphical test results and photograph of the test are shown next for a 90 deg out of plane elbow test
regime conducted at South West Research Laboratory (SWRI) in San Antonio in 2005 at 165psig ambient
conditions with a 4 inch diameter 0.45 β meter − Figure 14

Figure 9

Figure 14

API Chapter 22.2 Testing of Generic Cone Meters

Recently various generic cone meters supplied by Cameron Measurement Systems were tested at the
Colorado Engineering Experimental Engineering Station (CEESI) facility in Nunn Colorado to meet the
API chapter 22.2 test protocol requirements. Test results for a 2 inch 0.45 beta meter which involved a full
suite of disturbance testing is shown next.

The test is extremely aggressive regarding the up & downstream condition tests with a ½ moon orifice
plate (50%valve test) at 3 diameters upstream and 0-diameters on the outlet of the meter, Out of Plane
Elbows at zero D,s, a swirl generator close coupled at three diameters, and finally a standard baseline
over 12 points for comparison against the disturbance tests.- Figure15.
Figure 15

Installation and Operating Criteria

The meters were installed in separate streams situated in a steel cabin. D.P. Temperature and Static
Pressure data was collected and transmitted to a locally mounted Supervisory Control Data Acquisition
System (SCADA) to allow computation of the flow rate. The process conditions varied over the well life so
a meter sizing was selected to manage the following Natural Gas metering needs:

Process Condition and Operating Criteria

a) Initial Condition : 10 MMcf/day @ 2000 psig density 0.0443 lbm//ft^3

b) Majority Condition : 6 MMcf/day @ 600 psig density 0.0443 lbm//ft^3

c) End Life (new) : 1.5 MMcf/day @ 600 psig density 0.0443 lbm//ft^3

d) Old End life data : 1.0 MMcf/day @ 600 psig density 0.0443 lbm//ft^3

Meter sizes were determined electronically from a Cameron supplied propriety software which gave a 3
inch - 0.55 Beta cone meter as the selection to fit the existing pipe-work and flow needs.

Pressure rating for the piping was ANSI Class # 2500. Meters were manufactured from carbon steel with
316 stainless steel cone assemblies.

The meters where installed locally by a skid manufacturer near to the field facility. The facility staff helped
to ensure that the meter factors and dimensional /calibration data was inputted correctly into the SCADA
system, original proof calibration of the meters was performed at FCC Ltd on water.

The wells shown in examples (A) 3.7MMcf/Day and (B) 1.5MMcf /Day relate to original predicted process
conditions for the facility with the four holed plate operating at 50% of the listed “Majority Condition” and
the generic cone meter operating at the listed “End Life (new)” condition, as shown in the operating
process criteria listing above. Graphical results are shown in - Figures 16 &17.
Field Allocation Results

Example (A) Four Holed Orifice Plate

Average Daily Rate = 3,757-MSCF/day.

Figure 16

Field Allocation Results

Example (B) Generic Differential Pressure Cone Meter

Average Daily Rate = 1461.5-MSCF/day

Figure 17
Conclusions

In conclusion, both the 4-holed orifice plate and cone metering technologies provided wet gas
measurement accuracy for well level allocation when adjusted by the “GVCF” from the well test in this full
well stream wet gas application.

The cone metering technology was chosen by Shell E&P because of enhanced operational and in-field
documented allocation accuracy, with more robust wear capabilities in this particular flow regime.

Reduced meter inspection frequencies and less problem interventions were also observed with this type of
device.

References

(1) Hayward A. “A Source Guide for Users” Edition Published 1978


(2) Bagge, D.J., “Evaluation of Ketema, V-Cone Flowmeters” Test Report E 1705 S 92, SIREP,
1992.
(3) Ifft, S. A and Mikkelsen, E.D - “Pipe Elbow Effects on the V-Cone Flowmeter” North Sea Flow
Measurement Workshop, Peebles, Scotland, 1992
(4) B. K Lee, N.H. Cho and Y. D Choi, 1988 “ Analysis of periodically fully developed turbulent and
heat transfer by k- ε equation model in artificial roughened annulus”. Int J. Heat Mass Transfer,
Vol,31, pp 1797-1806
(5) B. H. Chang and A. F Mills, 1993, “Turbulent flow in a channel with transverse rib heat transfer
augmentation”, Int J. Heat Mass Transfer. Vol 36, No, 6, pp 1459-1469
(6) B. E Launder and D. B Spalding, 1974, “The numerical computation of turbulent flows”, Comput
Meth Appl Mech Engng, Vol 3,pp 269-289
(7) C. K. G Lam and K. A Bremhorst, 1981, “Modified form of the k-w model for predicting wall
turbulence”, Journal of Fluid Engineering, Vol 103, pp 456-460
(9) D.C Wilcox, 1993, Turbulence Modeling for CFD, DCW Industries, Inc
(10) J.Y.Yoon, 1993, Numerical Analysis of flows in channels with sand dunes and ice covers, Ph.D
Thesis,Department of Mechanical Engineering, The University of Iowa.
(11) D.C. Wilcox, 1988, “Reassessment of the Scale Determining Equation for Advanced Turbulence
Models” AIAA Journal, Vol 26, No 11, pp 1299-13
(12) M. C. Richmond and V. C Patel, 1991, “ Convex and Concave Surface Curvature Effects in Wall-
Bounded Turbulent Flows”, AIAA Journal, Vol 29, pp 895-902

(13) J.Tyndall, 1988, A Numerical Study of Flow over Wavy Walls, M. S. Thesis, Department of
Mechanical Eng. Univ. Iowa, Iowa City.
(14) H. C Chen and V. C. Patel, 1988, “ Near-Wall Turbulence Models for Complex Flows Including
Separation” AIAA Journal, Vol 26, pp 641-648

(15) Braid C Mr. (Barton Canada) first principle calculations for flow computers 1999.
(16)D. D. Knight, 1982, “ Application of Curvilinear Coordinate Generation Technique to the
computation of Internal Flows”, Numerical Grid generation, Elsevier Science Publishing
Company, pp 357-384
(17) Sveedmen SWRI Homogenous Model and OFU liquid effect( report)
(18) Ifft Wet Gas Testing at SWRI 1997 - V-Cone Meter
(19) Lawrence-Wellhead Metering by V- Cone Technology NSFMW 2000 Gleneagles, Scotland,UK
(20) Reader Harris-Peters. Y factor Expansibility Equation (NEL-TUV)
(21) Braid-Cameron (Canada) Cone Equations for Flow Computers a Technical Document 2006
(22) Lawrence-CBM Measurement by D. P. Cone Meter CII Conference India 2007
(23) Lawrence-ISHM Oklahoma Class 1320 Wet Gas May 2008
(24) Hollister - Dyer Comparison of Established Orifice, Venturi & Nozzle with Cone, Conditioning
Orifice, and Wedge Meters. NEL-TUV Production & Upstream Measurement Workshop 2008
(25) Davis.M.W - Shell Exploration and Production Allocation Methodology and Data Set Shell E&P
Pinedale 2009.

Potrebbero piacerti anche