You are on page 1of 23

EVALUATION OF A ROXAR MPFM 2600 MULTIPHASE FLOW METER

A Report for

Roxar AS Bergen, Norway

Report No: 2009/265

Project No: ROX016

Date: August 2009

This report is issued as part of the contract under which the work has been carried out for the client.

NOTES
1 This report may be published in full by the client unless it includes information supplied in confidence by TUV NEL Ltd or any third party. Such information, if included within the report, shall be identified as confidential by TUV NEL Ltd. The prior written consent of TUV NEL Ltd shall be obtained by the client before publication by them of any extract from, or abridgement of, this report. The prior written consent of TUV NEL Ltd shall be obtained by the client before publication: • • • Where such publication is made in connection with any public enquiry, legal proceedings or arbitration. Where such publication is made in connection with any company prospectus or similar document. Where the client has notice that TUV NEL Ltd is seeking or intends to seek patent or like protection for any intellectual property produced in the course of rendering the services.

2a

2b

Norway Prepared by: Richard Harvey Date: 31 August 2009 Approved by: Gary Miller Date: 31 August 2009 Project No: ROX016 Report No: 2009/265 Page 1 of 21 August 2009 .COMMERCIAL TUV NEL TUV NEL East Kilbride G75 0QF UK Tel: +44 (0)1355 220222 Fax: +44 (0)1355 272999 EVALUATION OF A ROXAR MPFM 2600 MULTIPHASE FLOW METER A Report for Roxar AS Bergen.RESTRICTED .

.....................2 5 5........................2 2...............2 6 6.................3 6.................................................... 19 Project No: ROX016 Report No: 2009/265 Page 2 of 21 August 2009 ............................................3 7 INTRODUCTION .....................................................................................................................3 4 4............................................. 5 5 Cross-contamination Monitoring …………………………………………………...................RESTRICTED ............ 6 Pressure and Temperature ………………………………………………………… 6 REFERENCE MEASUREMENT QUALITY ....1 6..................................................................... 10 Liquid Flow Performance.. Traceability of Reference Measurements ………………………………………........3 3 3.......... 11 Gas Flow Performance.........................................................1 5........ 5 REFERENCE METERING SYSTEMS...............1 2........... 13 APPENDIX A: GRAPHICAL RESULTS ..............COMMERCIAL TUV NEL CONTENTS Page 1 2 2.............................. TEST SETUP AND PROCEDURES ............................................2 6.............................. 6 6 7 8 Meter Setup …………………………………………………………………………............ 3 4 Layout and Test Fluids ……………………………………………………………......................1 4.................................................................................................................... 12 Water Cut Performance...............................1 3.........................................................2 3....................................................... 9 TEST RESULTS ........................... TUV NEL MULTIPHASE FLOW FACILITY ........................................................................ 9 Test Procedure ………………………………………………………………………. Flowrate Measurements ........ 12 Oil Flow Performance... 5 Separation .. 12 CONCLUSIONS .......................................... Uncertainty Analysis ………………………………………………………………................................................................................................. 14 APPENDIX B: TABULATED RESULTS ................................................................ 4 Control...

The design is 80% lighter and half the length of the previous generation and incorporates a field replaceable Venturi insert. The TUV NEL facility logged all reference measurements and the test meter data were acquired via the meter’s data acquisition flow computer. Scotland from June to July 2009. Figure 1 Roxar MPFM 2600 The meter’s performance was initially characterised over a range of reference multiphase conditions by a Roxar representative. advanced signal processing and electrode geometry provides information not previously available. differential pressure and temperature.COMMERCIAL TUV NEL 1 INTRODUCTION This report presents the results from a flow test programme on a Roxar MPFM 2600 multiphase flowmeter. The Roxar MPFM 2600 meter (Figure 1) claims many design improvements over its predecessor the 1900 VI Multiphase Flowmeter. Project No: ROX016 Report No: 2009/265 Page 3 of 21 August 2009 . Roxar representatives were present for the meter commissioning and for the duration of the testing period. This report also describes the TUV NEL Multiphase facility and the test setup as well as data on the performance of the Roxar MPFM 2600 meter at the conditions tested. The aim of this evaluation was to assess the performance of the flowmeter with respect to the manufacturer’s stated specification against a matrix of multiphase flow conditions. The analysis of the data was conducted by comparing averages of the meter and facility parameters logged over the same test period. The test programme was conducted at the TUV NEL Multiphase Flow Test Facility in East Kilbride. including multiple flow velocity data and near wall measurements. A ‘blind’ test was then carried out over a defined matrix of conditions supplied by Roxar. The technology incorporates new compact sensor geometry which allows for measurements in individual sectors in addition to full cross-sectional area measurement. In addition.RESTRICTED . A Rosemount 3095 Multivariable Transmitter provides a compact. integrated measurement of pressure.

3. The standard test section is constructed in 4-inch schedule 40 pipework. the nitrogen is exhausted to atmosphere from the separator. 4 and 6-inch meters.1 TUV NEL MULTIPHASE FLOW FACILITY Layout and Test Fluids The TUV NEL multiphase facility is based around a 3-phase separator which contains the working bulk fluids.COMMERCIAL TUV NEL 2 2. After passing through the test section. Perspex visualisation sections are available in 2. but many other sizes can also be accommodated. Piping and adaptors are available to allow testing of 2. topped to remove light ends and increase the flashpoint above 65ºC. For safety reasons nitrogen is used as the gas phase and can be delivered at up to 0.3 kg/s by evaporation of liquid nitrogen on demand. The delivery pressure of the nitrogen is up to 12 bar at the injection point. Currently test fluids in use are: • A mixture of Forties. Pressure control Three-phase separator Test meter Water Oil Liquid Nitrogen Pump Test meter Pump Gas Turbine reference meters Figure 2 Schematic of TUV NEL Multiphase Flow Facility Project No: ROX016 Report No: 2009/265 Page 4 of 21 August 2009 . The standard flange rating is ANSI class 150. 4 and 6-inch pipe sizes.7H2O). with kerosene substitute (Exxol D80) added to restore the original viscosity (Approximately 30° API gravity) • An aqueous solution of Magnesium Sulphate of concentration 83g/l (based on MgSO4. The oil and water are re-circulated around the test facility using two variable speed pumps. The facility is manufactured entirely from stainless steel and can thus utilise brine substitutes and dead crude oils as the working fluids in addition to deionised water and refined oils.RESTRICTED . Beryl and Oseberg crude oil. The test section can accommodate test setups of up to 60m horizontal and 12m vertical.

The PC is linked to the PLC via an EtherNet which permits a response time from command to action of less than one second.RESTRICTED .0 ltr/sec The nitrogen is metered through a choice of three gas turbine meters according to the flowrate required.2 Control An automated SCADA system allows a single operator to control the entire facility from a PC workstation.5 to 9. The PLC is linked to all the field instruments on the facility and can also be used to connect the outputs from meters under evaluation directly to the PC.53 to 19. and individual test points are also time-stamped. This acts as the storage vessel for the liquids currently under test.0 to 40. 2. in addition to separating the fluids for recirculation. Both metering circuits have a choice of two flowmeters.6 ltr/sec Project No: ROX016 Report No: 2009/265 Page 5 of 21 August 2009 .3 to 38. The calibrated ranges of these flowmeters are: ½-inch meter 1-inch meter 3-inch meter 0. 3 3.COMMERCIAL TUV NEL 2. with the following calibrated ranges: 1¼-inch oil meter 3-inch oil meter 0. During evaluations all reference instrument data are written to a text file on the PC workstation. The liquids can be re-circulated indefinitely across much of the operating envelope of the facility.19 ltr/sec 1.3 Separation At the centre of the facility is a large three-phase gravity separator. Various devices are employed inside the separator to speed up separation including baffles and parallel plate pack systems. Oil is metered using Faure Herman helical blade turbine meters. temperature is 40° C. according to the flowrate required. which contains approximately 25m3 of water and 15m3 of oil. It is additionally equipped with heat exchangers which allow the temperature of the oil and water to be maintained within ±1° over the range of approximately 10° to 40° T he normal facility operating C C.0 ltr/sec 5.5 ltr/sec 5.2 ltr/sec 6. The test files are date and time stamped.1 REFERENCE METERING SYSTEMS Flowrate Measurements The oil and water are separately drawn from the separator and pumped through the oil and water metering circuits respectively.0 to 41.38 to 2. The data are analysed offline together with the data from the meter under test.0 ltr/sec The water is metered using flat blade turbine meters with the following calibrated ranges: 1½-inch turbine meter 3-inch turbine meter 0. The level of fluid cross contamination is continually monitored and the reference liquid flowrates corrected.5 to 5. However at high liquid flowrates (especially oil-continuous flows) the degree of cross-contamination of the oil and water usually means that flow must be paused for a period to allow settling of the liquids. The separator is equipped with pumps and piping to allow transfer of settled liquids between the water and oil compartments.

it is necessary to correct for expansion of the gas in the test section. 3. so that the exact gas volume fraction at the multiphase meter under test can be calculated. and is fully automated in the data processing program. 3. The pressure transmitters and platinum resistance thermometers are calibrated against standard equipment held in the multiphase laboratory. water and gas reference turbine meters are calibrated against the UK primary national standard facilities at TUV NEL.COMMERCIAL TUV NEL The gas flowrate is measured at the gas supply pressure (typically 9 . Project No: ROX016 Report No: 2009/265 Page 6 of 21 August 2009 . and it is important that these instruments are calibrated against accurate standards. Most reference instruments are calibrated annually. The conductivity of the water phase can be determined using a standard conductivity meter. 4 4.2 Cross-contamination Monitoring An additional bypass stream flows through monitors to measure the crosscontamination of the liquid phases. The calculation uses an equation of state for the density of nitrogen. The oil-in-water content of the water flow stream is derived from a density measurement using a Solartron densitometer. Density of the oil and water phases is also determined prior to a customer’s evaluation or test. This sample loop is taken from the main pump outlet.3 Pressure and Temperature For accurate volumetric metering of the gas phase. by operating the test section at reduced pressures it is possible to cover the full range of gas volume fractions. The pressure and temperature of the gas and of the multiphase mixture are therefore measured at a number of locations around the facility: • • • • At the reference gas meters At the inlet to the multiphase test section At intervals along the multiphase test section At the multiphase meter test location The reference gas flowrate is corrected to the multiphase meter test conditions using the pressures and temperatures at the two positions. with a traceable record of the calibrations being maintained.12 bar). The water-in-oil content of the oil flow stream is determined from a density measurement using a Parr Scientific densitometer. with gas superficial velocities up to 20 m/s in the 4inch test section. The fluid densities are determined using an Anton Paar DMA5000 laboratory densitometer and are used in conjunction with the contamination measurement densitometers to correct the reference liquid flowrates.1 REFERENCE MEASUREMENT QUALITY Traceability of Reference Measurements The instrumentation provided for the facility is of the highest accuracy practicable. However. passed through a densitometer and returned to the separator.RESTRICTED . The oil.

However. The biggest contribution to the gas flowrate uncertainty is the test section pressure. Over the majority of the operating range of the TUV NEL multiphase flow facility the combined uncertainties are: • Gas flow < ± 1. due to elimination of errors.0% • Water flow < ± 1.RESTRICTED . oil and water flowrates and the ratio between them as well as pressure. These uncertainties are calculated and displayed in Table 3 of Appendix B.0% One of the most significant contributing factors to the oil and water flowrate uncertainties is the uncertainty in the cross-contamination monitoring. All uncertainties quoted are expanded uncertainties based on a standard uncertainty multiplied by a coverage factor k=2. the uncertainties in GVF and in total liquid flowrate are much smaller than in the individual reference flowrates.COMMERCIAL TUV NEL 4. The resulting error in gas flowrate is greatest at low test section pressure (which usually occurs when testing at high GVF). temperature and liquid cross contamination levels. Project No: ROX016 Report No: 2009/265 Page 7 of 21 August 2009 . taking into account the uncertainties of the calibrated instruments. This will lead to the greatest error in oil flowrate at high water cut and the greatest error in water flowrate at low water cut. This provides a level of confidence of approximately 95%. The exact uncertainties of a particular reference flowrate will depend on the values of the individual gas. observed fluctuations in flowrates during tests. and combination of the readings of a number of instruments to give the final values.5% • Oil flow < ± 1.2 Uncertainty Analysis It is also important to be aware of the uncertainties which are present in the reference flowrates.

RESTRICTED .COMMERCIAL TUV NEL 5 TEST SET-UP AND PROCEDURES FLOW Figure 3 View Showing Roxar MPFM 2600 Meter Installation in Facility Test Section Project No: ROX016 Report No: 2009/265 Page 8 of 21 August 2009 .

Logging of the reference and test meter data was initiated simultaneously by the facility operator and the Roxar technician and both systems’ loggings were terminated automatically after the prescribed test period. covering the full operating range of the Roxar MPFM 2600 meter were conducted to verify the meter performance. Data logging was carried out for a period of 5 or 10 minutes per flow condition as requested by Roxar.2 Test Procedure Test points from the agreed test matrix were selected by the TUV NEL Facility operator in the order which best suits the facility operational efficiency. The reference flow conditions were then set and left to stabilise. This provided approximately 30 diameters of matching pipework prior to the meter entry. A test point was logged once the facility operator and the Roxar representative present were both satisfied that the required flow condition had been achieved and was stable.2. Project No: ROX016 Report No: 2009/265 Page 9 of 21 August 2009 . Initial in line static calibration checks were carried out on the various sensor elements. first on a series of well mixed flowing water oil ratios from 60 to 100% containing no gas.1 Meter Setup The Roxar MPFM 2600 multiphase flow meter was supplied with Grayloc 4GR40 standard connections and accompanying 3” schedule 160 pipework and adaptors to connect to the TUV NEL Multiphase Facility standard 4” class 150. The TUV NEL facility control computer and the Roxar MPFM data acquisition flow computer clocks were synchronised prior to testing. The conductance sensor was then calibrated. An extensive series of multiphase test points. Reference data was withheld from Roxar until the test meter log files were received and checked by TUV NEL. The gas phase was injected via a 4” tee piece approximately 4m prior to the 3” blind tee. Once Roxar personnel were satisfied with the accuracy of the meters output a ‘blind’ test was carried out over a range of multiphase flow conditions supplied by Roxar (Table 1). The protocol for this type of blind test is outlined in Section 5.COMMERCIAL TUV NEL 5. The gamma densitometer calibration was carried out on an empty pipe. then on these same water cuts through a range of gas volume fractions from 20 to 93%. The capacitance and conductance sensors were then checked on the facility oil with background water content and on the facility water solution respectively. schedule 40 pipework as shown in Figure 3. 5. In general 5 minute log times were used <65% GVF and 10 minute log times >65% GVF. A similar process was repeated for the capacitance sensor using water cuts from a ~1% background water content to 45% and at gas volume fractions up to 96% to derive the permittivity calibration data.RESTRICTED .

00 24.80 8.7 9.75 6.COMMERCIAL TUV NEL 6 TEST RESULTS The agreed test matrix is given in Table 1.00 31.4 9.6 3.7 WLR (w/c) (abs %) 4.00 27.1 5.40 19.00 0.80 1.30 13.20 m^3/hr Q Gas 55 70 130 20 55 15 130 15 40 70 100 30 130 130 30 55 15 20 55 80 180 bar oC Pressure Temperature 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 40 40 40 40 40 40 40 40 40 40 40 40 40 40 40 40 40 40 40 40 40 NOTE: Test points were not necessarily conducted in the order given above.20 16.9 3.50 19. The manufacturer’s published uncertainty for the Roxar 1900 VI is shown in Table 2.9 4.50 16.00 6.00 3.50 10.00 3.50 22.60 10.8 4.00 27.50 8. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 m3/hr Q Liquid 30 25 20 35 25 20 18 30 30 25 18 30 16 20 30 30 30 30 30 18 18 TEST MATRIX FOR ROXAR MPFM 2600 METER % GVF 65 74 87 36 69 43 88 33 57 74 85 50 89 87 50 65 33 40 65 82 91 % W/cut 0 0 0 10 10 20 20 35 35 35 35 45 45 60 80 80 90 90 90 90 90 m^3/hr Q oil 30.00 3.00 27.00 25. TABLE 2 (A) PUBLISHED UNCERTAINTY FOR A ROXAR 1900 VI METER TEST AT NEL GVF (%) 0-25 25-85 85-96 96-100 Sum combined uncertainty WLR (o/c) Gas Liquid (rel %) (rel %) (abs %) 10.2 4.20 12.50 8.00 24.00 3. Table 1 Test No. This specification is used for this test as currently no specific uncertainty specification has been established for the 2600 flowmeter.00 16.25 11.50 4.80 m^3/hr Q water 0. but were instead carried out in the most suitable order for facility operational efficiency.00 0.1 4.8 Project No: ROX016 Report No: 2009/265 Page 10 of 21 August 2009 .00 20.70 16.2 3.00 1.50 7.50 2.50 16.00 3.00 14.RESTRICTED .00 6.

5 0 .1 Liquid Flow Performance Figure 4 shows the liquid flow rate plotted against the reference gas volume fraction for the Roxar MPFM 2600. The maximum error in the measured liquid flow rate in oil continuous flow was -2.2 The numerical results from the test programme are given in Table 4 at the end of this report.2 and an allowance for installation and process flow effects Table 2 (B) combined in quadrature. in water continuous flow it was -3. The error bands shown on these graphs do not constitute any performance specification claimed by Roxar. The standard deviation is defined as: Project No: ROX016 Report No: 2009/265 Page 11 of 21 August 2009 .5%. Instrument Uncertainty 1. shown in Table 2 (A) for this type of meter. The liquid flow errors show a total spread of 6. Figure 7 shows the same errors with respect to liquid flow. 6.1% with a mean offset of -0.96 % Combined Uncertainty 4.3 0 .2 2. Volumetric flow rates measured by the reference gas meters have been converted to the pressure and temperature conditions as reported by the test meter to permit a direct comparison. Figures 4 to 6 illustrate the performance of the test meter during the test programme. Some deterioration in performance may be interpreted at the higher GVF test points but is not significant up to the maximum GVF tested of 91. The uncertainty bands displayed represent the actual Roxar published uncertainty.65%. consisting of the Roxar 1900 meter uncertainty Table 2 (A). Measurement accuracy appears consistent across water and oil continuous regimes.RESTRICTED . the singlephase reference meter uncertainties detailed in Section 4.59% and a standard deviation of 1. liquid and water cut measurement capabilities with respect to GVF.96 % NEL 2 2 2 NEL 2. It can be seen from Figure 4 that all points lie within specification on liquid flow. The graphs show the gas.2 2. The data are compared to a composite error band.COMMERCIAL TUV NEL TABLE 2 (B) ROXAR PROVISION FOR FLOW LOOP CONTRIBUTION TO COMBINED UNCERTAINTY GVF NEL Qgas Qliq 1 WLR_o 1 WLR_w 1 Ref.81%.77%.96 % Process Uncertainty 4 0 . No significant trend is apparent though data is spread over a limited liquid flow range. Also included are Figures 7 to 11 presented for guidance only.

75%. The ±10% bands shown on this graph are for guidance only. The gas flow errors show a total spread of 15.4 Oil Flow Performance Figures 10 and 11 are included for information only as Roxar do not make any specific claim on oil flow rate accuracy.7%. The performance appears to deteriorate slightly in the water continuous regime and although still well within specification. No deterioration of performance with respect to GVF is detectable from this data set.9% of the total flow is oil (1.3 Water Cut Performance The water cut performance of the Roxar MPFM 2600 meter is shown in Figure 6 with respect to GVF where.9 m3/hr). Project No: ROX016 Report No: 2009/265 Page 12 of 21 August 2009 .41% absolute. The meter’s gas flow performance was unaffected by the continuous liquid phase and displayed a maximum error of -8. This shows that there was only 1 test point recorded between 45 and 80% water cut. 6. The uncertainty bands displayed represent the actual Roxar published uncertainty. Figure 9 shows the same data plotted against reference water cut. Under these conditions the MPFM measured the oil rate to within 0.89%. shown in Table 2 (A) for this type of meter. 6. The water cut errors show a total spread of 3. data is only available at two water cuts of 80% and 90%. These water cut set points were chosen to avoid operating in the oil/water continuous transition region.45% and a standard deviation of 4.2 Gas Flow Performance Figure 5 illustrates the Roxar MPFM 2600 meter gas flow performance.37% and represents a flow point in which less than 0. Both these test points represent conditions where the oil flow is at the lowest measurable by the facility (also with the highest reference uncertainty) and where the oil fraction in the multiphase stream is less than 2%. The water cut level at which the continuous phase changes can vary in the TUV NEL Multiphase facility from 55% to 70%.87% with a mean offset of 0. It can be seen from these graphs that only 2 of 21 points fall outside a band of ±10%.55 m3/hr absolute deviation from the reference. all 21 test points were within the water cut performance specification. In oil continuous flow the maximum deviation from the reference water cut was 1. all of the test points are within the meters gas flow accuracy specification. In water continuous flow the maximum deviation was 2. All oil flow rate measurements are well within the theoretical oil uncertainty that can be calculated from the stated uncertainties in table 2(A). or when the same errors are plotted against reference gas flow (Figure 8). once again. Test matrix point 21 in particular displays the maximum deviation from the reference of -28.77% with a mean offset of -2.COMMERCIAL TUV NEL 6. In line with the liquid flow results.RESTRICTED .78%. The ‘flipping’ point largely depends on the cleanliness of the single phase liquids at injection but also on the degree of mixing through the test section which in turn depends on GVF and velocity.42% and a standard deviation of 0.

This however should not detract from the overall excellent performance of the test meter. In a comparison against the manufacturer’s performance specification. gas flow and liquid water cut. During the test programme a total of 21 test conditions were logged across an agreed matrix of multiphase mixtures on a Roxar MPFM 2600 Multiphase Flow Meter. Project No: ROX016 Report No: 2009/265 Page 13 of 21 August 2009 .RESTRICTED .COMMERCIAL TUV NEL 7 CONCLUSIONS The test programme results have been analysed and presented numerically and graphically showing the performance with respect to the specification published by Roxar (Table 2 (A)). The meter performed extremely well on liquid flow. The matrix was rather limited at 21 test conditions with low liquid flow turndown and limited water cut ranges. all of the meters individual measurements were within the current published uncertainty specification.

COMMERCIAL TUV NEL APPENDIX A GRAPHICAL RESULTS Project No: ROX016 Report No: 2009/265 Page 14 of 21 August 2009 .RESTRICTED .

RESTRICTED .COMMERCIAL TUV NEL Liquid Flow Error 2600 Meter 15 10 Error in Liquid Flow (%) 5 0 -5 O/C -10 W/C Combined Uncertainty -15 0 10 20 30 40 50 60 70 80 90 100 Reference GVF (%) Figure 4 Error in Meter-Indicated Liquid Flow Rate Versus Reference GVF for Roxar MPFM 2600 Meter Gas Flow Error For 2600 Meter 30 20 Error in Gas Flow (%) 10 0 -10 O/C -20 W/C Combined Uncertainty -30 0 10 20 30 40 50 60 70 80 90 100 Reference GVF (%) Figure 5 Error in Meter-Indicated Gas Flow Rate Versus Reference GVF for Roxar MPFM 2600 Meter Project No: ROX016 Report No: 2009/265 Page 15 of 21 August 2009 .

RESTRICTED .COMMERCIAL TUV NEL Water Cut Error For 2600 Meter 15 10 Error in Water Cut (% Abs.) 5 0 -5 O/C W/C -10 Combined Uncertainty (O/C) Combined Uncertainty (W/C) -15 0 10 20 30 40 50 60 70 80 90 100 Reference GVF (%) Figure 6 Error in Meter-Indicated Water Cut Versus Reference GVF for Roxar MPFM 2600 Meter Liquid Flow Error 2600 Meter 15 10 Error in Liquid Flow (%) 5 0 -5 O/C -10 W/C +/.5% -15 0 2 4 6 8 10 12 Reference Liquid Flow (l/s) Figure 7 Error in Meter-Indicated Liquid Flow Rate Versus Reference Liquid Flow for Roxar MPFM 2600 Meter Project No: ROX016 Report No: 2009/265 Page 16 of 21 August 2009 .

10% -30 0 10 20 30 40 50 60 Reference Gas Flow (l/s) Figure 8 Error in Meter-Indicated Gas Flow Rate Versus Reference Gas Flow for Roxar MPFM 2600 Meter Water Cut Error For 2600 Meter 15 10 Error in Water Cut (% Abs.) 5 0 -5 O/C -10 W/C +/.RESTRICTED .COMMERCIAL TUV NEL Gas Flow Error For 2600 Meter 30 20 Error in Gas Flow (%) 10 0 -10 O/C -20 W/C +/.3% -15 0 10 20 30 40 50 60 70 80 90 100 Reference Water Cut (%) Figure 9 Error in Meter-indicated Water Cut Versus Reference Water cut for Roxar MPFM 2600 Meter Project No: ROX016 Report No: 2009/265 Page 17 of 21 August 2009 .

10% -30 0 2 4 6 8 10 12 Reference Oil Flow (l/s) Figure 11 Error in Meter-indicated Oil Flow Rate Versus Reference Oil Flow Rate For Roxar MPFM 2600 Meter Project No: ROX016 Report No: 2009/265 Page 18 of 21 August 2009 .COMMERCIAL TUV NEL Oil Flow Error 2600 Meter 30 20 Error in Oil Flow (%) 10 0 -10 O/C -20 W/C +/.10% -30 0 10 20 30 40 50 60 70 80 90 100 Reference GVF (%) Figure 10 Error in Meter-indicated Oil Flow Rate Versus Reference GVF for Roxar MPFM 2600 Meter Oil Flow Error 2600 Meter 30 20 Error in Oil Flow (%) 10 0 -10 O/C -20 W/C +/.RESTRICTED .

RESTRICTED .COMMERCIAL TUV NEL APPENDIX B TABULATED RESULTS Project No: ROX016 Report No: 2009/265 Page 19 of 21 August 2009 .

001 All uncertainties quoted are expanded uncertainties based on a standard uncertainty multiplied by a coverage factor k=2.10 35.560 1.63 2.001 0.90 15.86 5.01 0.916 0.08 U % 1.633 0.95 5.RESTRICTED .90 0.002 0.005 0.35 0.003 0.96 1.65 0.003 0.06 8.39 5.65 0.756 1.720 2.73 0.01 0.91 2.625 0.445 5.88 0.162 0.190 1.002 0.02 1.004 0.001 0.998 1.657 0.004 0.56 15.570 1.667 0.807 39.04 5.641 0.61 38.60 0.144 1.004 0.004 0.90 15.06 0.33 0.01 0.729 0.73 3.41 0.006 1.70 1.35 2.89 0.43 0.004 0.606 0.94 0.87 0.23 4.004 0.56 U Water/Liq ratio 0.004 0.49 0.52 5.158 1.96 0.573 1.896 0.575 1.00 37.580 0.89 0.779 0.004 0.004 0.003 0.696 1.002 0.89 0.49 8.002 0.74 0.59 22.005 0.10 0.002 0.004 0.45 0.506 0.550 0.005 0.819 2.44 4.11 0.749 0.551 1.01 1.525 1.172 1.131 1.48 8.57 8.004 0.049 1.COMMERCIAL TUV NEL TABLE 3 REFERENCE UNCERTAINTIES FOR TESTS ON ROXAR MPFM 2600 METER Gas flow MUT l/s 4.005 0.004 0.29 U % 1.003 0.35 0.96 0.633 0.39 6.012 Oil flow MUT l/s 4.547 1.481 1.87 4.599 0.680 0.66 4.55 3.65 0.91 0.004 0.004 0.77 6.43 4.573 1.41 1.09 4.49 0.13 26.73 2.73 6.547 1.36 4.004 0.90 0.94 0.003 0.712 2.886 1.005 0.615 1.77 6.536 1.82 8.614 0.86 0.59 19.167 5.38 0.04 0.33 6.749 0.82 0.003 0.80 0.711 0.39 8.935 0.520 2.059 4.80 0.59 U Gas volume fraction (GVF) 0.32 5.91 0.004 0.775 4.27 0.84 39.34 0.004 0.40 8.182 Total Liquid flow MUT l/s 5.90 0.998 1.552 1.94 5.005 0.53 5.54 0.25 53.69 0. Project No: ROX016 Report No: 2009/265 Page 20 of 21 August 2009 .69 0.14 8.630 0.32 6.58 0.571 1.41 4.30 Water flow MUT l/s 1.43 6.61 9.82 2.93 5.71 8.01 8.676 0.22 9.35 0.51 7.005 0.561 0.96 5.09 0.94 0.91 5.11 11.84 0.899 0.03 3.35 0.815 0.05 8.003 0.762 0.532 1.002 0. This provides a level of confidence of approximately 95%.580 1.52 0.001 0.98 2.19 0.004 0.924 0.71 1.28 15.03 0.60 19.46 0.88 U Test Point 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 0.492 0.542 1.36 4.02 8.53 7.967 34.657 0.20 0.90 0.26 U % 0.442 47.73 7.

514 5.65 0.553 39.02 0.15 0.318 2.60 88.14 36.93 4.82 33.82 4.45 65.02 5.56 58.433 8.380 37.58 40.23 34.39 Mean Pressure Bar g 4.19 68.87 32.93 0.43 16.68 44.12 0.53 36.45 34.86 2.771 0.61 79.13 2.75 5.41 0.571 0.50 87.93 4.97 87.53 38.94 48.280 5.02 8.76 40.99 39.83 4.98 40.55 6.86 79.46 49.864 7.12 59.413 6.66 4.40 88.66 33.87 19.65 6.58 64.69 8.95 34.409 2.97 5.64 24.COMMERCIAL TUV NEL TABLE 4 RESULTS SHOWING REFERENCE AND TEST METER FLOWRATES WITH ASSOCIATED DEVIATIONS FOR METER 1600-08 Test Point 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Test Date 8-Jul-09 8-Jul-09 8-Jul-09 8-Jul-09 8-Jul-09 8-Jul-09 9-Jul-09 9-Jul-09 9-Jul-09 9-Jul-09 9-Jul-09 9-Jul-09 9-Jul-09 9-Jul-09 9-Jul-09 9-Jul-09 9-Jul-09 9-Jul-09 9-Jul-09 9-Jul-09 9-Jul-09 Start Time 16:05:03 16:23:06 16:47:36 17:04:45 17:21:53 17:48:56 10:37:11 11:04:20 11:40:47 12:06:03 12:23:21 12:42:28 13:08:32 13:38:35 13:58:21 14:29:51 14:53:06 15:19:16 15:53:02 16:16:49 16:54:43 Duration min 0:05:00 0:10:00 0:10:00 0:05:00 0:10:00 0:10:00 0:05:00 0:10:00 0:10:00 0:05:00 0:05:00 0:05:00 0:10:00 0:10:00 0:10:00 0:10:00 0:05:00 0:10:00 0:05:00 0:10:00 0:10:00 No.90 58.69 1.54 73.18 54.139 4.34 21.12 -1.05 74.029 5.01 4.74 69.75 35.52 58.43 0.34 89.62 88.39 39.70 40.605 6.98 12.956 7.733 5.40 40.111 5.65 -1.365 8.276 1.00 4.72 27.486 5.631 3.900 8.83 -1.61 8.32 59.24 0.37 15.000 2.57 -6.00 41.25 0.00 4.17 5.05 40.24 -0.02 33.78 -3.81 5.27 1.03 -2.66 34.16 65.50 -5.46 91.66 34.12 11.92 65.02 -0.96 16.92 22.74 0.07 0.07 -0.68 63.12 8.55 36.085 0.48 6.270 1.029 4.671 5.69 47.708 6.92 39.220 4.04 -6.11 8.52 9.41 65.98 20.89 87.059 8.87 0.539 4.358 9.52 -8.595 4.95 4.70 0.726 5.13 84.82 5.78 9.963 19.26 0.52 73.42 45.61 87.547 0.99 -0.52 69.43 1.491 4.727 8.13 86.77 -2.90 Test Meter Measurements Water Cut Gas Flowrate Value Error Value Error l/s % % % water 4.84 5.81 5.956 7.48 2.55 89.70 -0.91 4.327 0.02 0.80 0.44 4.05 0.550 2.15 65.977 5.83 -3.30 39.46 -0.43 15.955 9.87 4.99 1.93 1.06 -2.28 8.43 4.19 -0.850 2.866 0.59 8.812 15.05 55.24 8.12 91.89 80.906 8.260 0.19 6.41 5.08 51.727 6.707 6.02 -0.68 -0.74 14.45 -2.617 5.43 19.08 40.10 4.69 6.08 0.27 -1.95 -0.98 4.22 34.84 0.931 22.96 34.57 -8.315 11.79 -7.10 41.28 5.RESTRICTED .24 73.59 -0.87 Test Start Time 16:04:58 16:23:08 16:47:32 17:04:41 17:21:57 17:48:52 10:37:08 11:04:10 11:40:39 12:06:03 12:23:19 12:42:22 13:08:34 13:38:31 13:58:21 14:29:41 14:53:11 15:19:08 15:53:02 16:16:49 16:54:39 Liquid Flowrate Value Error l/s % 5.42 34.26 4.37 -1.04 73.14 11.26 2.10 4.11 14.053 53.76 88.63 0.137 3.80 15.71 89.64 89.03 4.19 43.351 2.95 65.83 40.15 33.00 39.56 39.93 -0.50 8.93 6.04 89.39 5.63 87.08 82.06 81.003 26.39 0.56 -2.33 -2.64 2.86 5.596 0.433 1.87 91.07 1.75 9.904 19.19 -0.093 Water Cut % 19.389 4.020 35.519 4.766 8.600 38.82 -1.098 4.71 39.044 0.83 40.50 9.686 8.60 46.395 15.32 39.473 8.95 0.46 81. Records 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 Mean Temp degC 41.93 -6.08 4.086 5.23 87.82 10.85 89.85 Reference Measurements Oil Water Liquid Gas Flow Flow Flow Flow l/s l/s l/s l/s 4.530 8.86 41.39 9.44 92.573 0.440 4.45 1.15 Project No: ROX016 Report No: 2009/265 Page 21 of 21 August 2009 .37 Gas Fraction Value Error % % gas 44.22 33.71 -0.29 2.81 0.04 1.229 15.50 90.77 -0.34 88.386 15.860 0.24 19.43 1.051 1.38 0.56 Corrected Gas Gas Flow Fraction l/s % 4.16 0.25 52.06 4.54 82.63 8.18 -2.821 8.75 -8.32 84.56 49.54 20.20 -1.26 81.68 19.60 15.897 6.29 41.16 41.03 39.88 -3.273 3.45 -1.12 19.06 38.22 1.61 66.43 37.25 -1.91 40.964 8.26 -2.89 3.38 40.07 Gas Fraction % 42.043 6.45 91.27 7.82 41.09 2.71 0.88 -0.74 88.80 5.21 73.24 41.23 54.08 1.