Choke Flow Modellingfor a Gas Condensate Reservoir 1- Department of Mineral Resources and Petroleum Engineering, Chair of Petroleum and Geothermal Energy Recovery, Montan University of Leoben, Austria 2- Faculty of Gas and Petrochemical, Persian Gulf University of Bushehr 3- Department of Mineral Resources and Petroleum Engineering, Chair of Drilling Engineering, Montan University of Leoben, Austria) 1- Mohammad Khosraviboushehri 2- Reza Azin 3- Abbas Roohi Abstract—wellhead choke is the only device used to regulate the production rate of a reservoir and thus can control formation draw down. A choke is basically designed with small opening which can prevent erosion problems and a leaking master valve which may killing the well to replace the valve. The main idea of using a choke is to maintain a sufficient back pressure for a reservoir to prevent formation damages such as gas or water coning and sand entry. The filed being studied in this project, located in Persian Gulf coastline, is a gas condensate reservoir. Since, no flow meter is installed on the wells the individual flow rate of each well is not available. But total flow rate of two platforms which is cumulative production of twenty wells’ flow rate is known instead. And, since it is necessary to determine each well’s production rate in order to analyze production data and predict the future of production, Pipesim simulator was used to simulate choke valve and calculate the gas and condensate flow rate of each well. Having opening percentage instead of bean size in adjustable chokes is a challenging issue in calculating production rate. Therefore, 77 data points of test separators were collected to simulate the choke and to obtain an applicable relation between the choke opening percentage and the bean size. At the end, total calculated enriched gas flow rate was compared to total measured enriched gas flow rate and consequently mean percentage error of 3.4 was obtained. Keywords— wellhead choke, bean size, choke opening percentage, enriched gas, gas condensate reservoir, Pipesim I. water coning and sand entry [2]. Due to high sensitivity of oil and gas production to choke size, production engineers should select the optimum choke size by accurate modeling of choke performance which plays an increasingly important role in the reservoir management. In other words, finding the optimum choke size can maximize economic recovery of oil and gas from a reservoir [2]. Choke flows can be classified into two main categories: critical flow and sub-critical flow. Critical flow occurs when fluid’s velocity will reach sonic speed and the flow rate is independent of downstream pressure. In this case, any fluctuation in downstream of the choke will not influence on upstream conditions. On the contrary, in sub-critical case, flow rate depends on both upstream and downstream pressures [3]. The dependency of flow rate to pressure ratio (downstream/upstream) is represented in the following figure. INTRODUCTION Gas condensate reservoirs have a different flow behavior from other gas reservoirs. These reservoirs are initially gas at the reservoir condition. However, with the beginning of production, pressure decreases below dew point once heavy hydrocarbons in gas phase start to form condensates near the wellbore and consequently may limit gas flow path in the wellbore which causes gas production decrease. Thus, Understanding the production rate of gas and condensate as well as PVT analysis and flow regimes would be essential to prevent a well from liquid hold up and from decreasing productivity of a well. The wellhead choke is the only device used to regulate the production rate of a reservoir and thus can control formation draw down. A choke is basically designed with small opening which can prevent erosion problems and a leaking master valve which may killing the well to replace the valve [1]. The main idea of using a choke is to make reservoirs to produce at the optimum rate while maintaining a sufficient back pressure for a reservoir to prevent formation damages such as gas or Figure 1 Flow rate versus downstream/upstream pressure [4] Theoretical and empirical methods are available in the literature to predict the choke flow performance. Tangeren et al. [5] did the first theoretical study on two-phase flow through restrictions. He assumed polytropic expansion of gas that is dispersed uniformly in the mixture having liquid as the continuous phase in critical conditions. Fortunati [6] developed the first correlations for both critical and subcritical flow and the boundary between these regimes. Ashford [7] presented a model for two-phase critical choke flow based on the work of Ros [8]. Gould [9] plotted the Ashford boundary, showing the dependency of boundaries to the polytropic exponent. Ashford and Pierce [10] derived an equation to predict the critical pressure ratio. Pilehvari [11] 1|P age These studies include the following methods: Fortunati presented an experimental method for both critical and subcritical cases. and Ros [8] modified his method and introduced some new factors which could fit the method to a variety of fluids and flowing conditions. The Gas equivalent of produced condensate In gas condensate reservoirs. The filed being studied in this project. n. He also utilized the same assumption as Ashford and Pierce and suggested an equation for critical pressure ratio. Guo et al.7 psi. Several different relations have been developed for gas and liquid flow rate calculations so far. gas specific gravity. condensate API. Table 2 Mean GLR for ID1 & ID2 from 2005 to 2010 22000 Mean GLR for ID1 (SCF/STB) 21000 2005-2008 48000 21000 2009 30000 22000 2010 Mean GLR for ID2 (SCF/STB) year C. no flow meter is installed on the wells the individual flow rate of each well is not available. In table 1 fluid properties of reservoir are presented based on the analysis of PVT samples Table 1 Fluid properties for ID1 and ID2. Other studies suggested some new methods for prediction of critical and sub-critical flow that describe the behavior of multi-phase fluids flowing through a choke. since it is necessary to determine each well’s production rate in order to analyze production data and predict the future of production. Some of these equations resulted from experimental measurements and some other derived based on physical theories [9]. Sachdeva et al. [13] [1] evaluated the accuracy of the Sachdeva's model using field data from oil and gas condensate wells in Southwest Louisiana.also studied choke performance under sub-critical conditions. The Perkins' models [13] presented an approach to estimate the critical pressure ratio and the mass flow rate in a way nearly same to Sachdeva et al. Later. nominal choke size and the gas/liquid production ratio. the gas equivalent of one stock tank barrel at standard condition (14. • Choke upstream pressure and temperature • Bean size • Correlation determination of choke pressure drop • Choke downstream pressure B. But total flow rate of two platforms. assuming isentropic multiphase homogeneous mixture. In black oil model. other researches such as Baxendell [14]. [12]. Assumptions used in choke simulation for ID1 and ID2 1) Black oil model is used for choke simulation. Perkin used the mass continuity equations and developed a relation for total mass flow rate. And. Pilehvari [11]. Since. METHODOLOGY A. [1] minimized models error using different values of choke discharge coefficient (CD). (GE). Again a try and error method is need in using this equation as in Sachdeva and Ashford equations. 60℉) is defined as [2]: 2|P age . Comparisons of the results indicated that Sachdeva's model generally underestimates gas and condensate flow rates. He used the homogeneous compound assumption and specified the critical and sub-critical boundary with the aid of experimental data [6]. In compositional model. Calculation of the sound velocity or the separating boundary of the critical and sub-critical flow is necessary for evaluating the behavior of compressible fluid in chokes [7]. II. and water cut percentage should be available. The first empirical relation for critical flow was developed by Gilbert. 3) As water cut percentage has a slight influence on choke pressure drop calculation.672 53. Inputs for Pipesim software Platform Gas specific gravity API condensate ID1 ID2 0. Pipesim simulator was used to simulate choke valve and calculate the gas and condensate flow rate of each well in different dates. information about fluid components and their physical properties are required. cumulative gas production includes total gas produced and stock tank liquid production.689 0. which is cumulative production of twenty wells’ flow rate is known instead. [12] presented a model to predict mass flow rate through chokes for both regimes. Achong. the water cut value is assumed to be zero in this reservoir. He introduced new factors relating upstream pressure to gas/condensate ratio. which has to be converted into its gas equivalent. 5) Although there is not a great deal of differences between GLR in the daily reports. located in Persian Gulf coastline. Perkins [13] included the three phase effects for the polytropic expansion exponent.5 55. some information including gas liquid ratio (GLR). ID1 and ID2.9 2) Knowing the gas condensate flow through chokes. Assuming ideal gas behavior. Guo et al. mechanistic model is considered to calculate total pressure drop for both critical and sub-critical correlations. and also found the mixture average velocity at the throat. The filed wells are of deviated and cased-hole types and completed in monobore. is a gas condensate reservoir. Required Pipesim data for simulation • Phase behavior model determination The phase behavior model of gas condensate reservoirs can be either compositional or black oil model. the average values of GLR are considered for platforms ID1 and ID2 in the period of 2005 to 2010 which are presented in table 2. 67 03/09/2008 04/09/2008 22 21 222.8 0.6 83. it is essential to have the correct value of bean size for simulating a choke by Pipesim simulator [15] in order to calculate the gas and condensate flow rate of individual well.6 120 120 20391 21079 1.4 0. In addition.6 D.3 21368 1. q (gtot).6 Bean Size (inch) E. following relation suggested as regression equation: 1.8 83.55 06/01/2011 25 207.44 06/01/2011 15 231.0 1.1 83.3 120.0 Bean Size (inch) 1.38 06/01/2011 21 214. simulation results of bean size for some of data points in ID1-01 well are listed.0 1.49 1.2 0. a relation between the bean size and choke opening percentage is suggested based on regression analysis of each well.8 0. Bean size results by choke modelling of test separator data As mentioned before. 4) Presents an exceptionally good trend for the high opening range (more than 30 %) and the low opening range (less than 10 %). Choke opening percentage and bean size relation After obtaining bean sizes.0 0 5 10 15 20 25 30 35 40 45 50 Choke Opening (% ) Figure 4 Regression analysis results ID1-13 well 3|P age .8 120. In Table 3.9 228.6 Bean Size (inch) choke opening Date 0. 77 data points of test separators were collected to simulate the choke and obtain an applicable relation between the choke opening percentage and the bean size.3 21301 1. the total reservoir gas production.0 0 Wellhead T. SPD3-01 ID1-01 2.2 0. is given by equation 2: Figure 2 to figure 4 illustrate the regression analysis results for the data points of simulated choke for some wells of ID1. 2) Preferably it is simple and includes few numbers of constants. On the other hand.86 06/01/2011 20 220.2 120.1 120 22010 1.6 82. Therefore.0 0 5 10 15 20 25 30 35 40 45 50 Choke Opening (% ) Figure 3 Regression analysis results ID1-02 well SPD3-13 ID1-13 2. the constants of relations obtained for each well and the corresponding coefficient of determinations are indicated in table 4.32 06/01/2011 24 209.4 Table 3 Simulated Bean size for some test separator data points for ID1-01 well head p.1 16096 0.1 83 120. 3) Has an increasing trend from 0 to 100 percent choke opening. 1.3 79.7 83 120.4 0. 0. Therefore. Downstream p. An applicable relation should meet the following condition: 1) Determination coefficient (R2) should approach to 1 which indicates a good correspondence with the data points.2 18152 1. GLR Simulated bean size % barg °C barg SCF/STB inch 02/09/2008 36 212.4 26271 1.2 1. hereafter called enriched gas.8 0.59 5 10 15 20 25 30 35 40 45 50 Choke Opening (% ) Figure 2 Regression analysis results ID1-01 well SPD3-02 ID1-02 2. the adjustable chokes used in these platforms provide the accessibility of choke opening percentage.Where ϒc is the specific gravity of condensate (water=1) and Μwc is the molecular weight of condensate. Considering above conditions.8 83. 7 ID1-13 1.44E-05 2.9784 ID1-07 100.93E-05 1. The steps of this procedure are presented as follows: 1) A specific date (e.15 4416 3.40E-05 0.12 5869 4.74 107 0.04 115.54E-05 0.94 113. 7) Simulated rich gas flow rate is compared with the reported enriched gas flow rate.01E-05 1.8 ID2-01 3.8 ID2-14 105.9686 ID1-04 103.9842 ID1-02 97.61E-05 0.65E-05 0.65 104.94 3698 2.2010 can be observed in the following tables.68 133.2010) for simulation is picked out.62 3321 2. Therefore.11 42.06 3402 2.9761 ID2-07 2.54 101. It is also possible to perform the same simulation for the production days in which one of two platforms is shut down.82 4582 3.7 4441 3.56 99.Table 4 Constants of relations and the corresponding coefficient of determinations for all wells III.9445 ID2-14 1.11.52 100.18 ID1-12 97.9848 ID1-06 1.97E-05 ID1-12 recorded.79E-05 ID1-10 1. 4|P age .50E-05 8.36 96.9 111. Studies showed that the suggested model may be used successfully to calculate the gas and condensate flow rate of each platform separately. 3.53E-05 2.94E-05 3.83 Calculation results for 63 field data set are presented in the table below.3 ID2-04 70.9 ID2-10 99.9873 ID2-12 2.9782 ID1-07 1.90E-05 9.21 4691 3.87 72. upstream pressure and temperature. bean size is determined according to choke opening percentage for all 20 wells of ID1 and ID2 platforms.11.84E-05 4.6 E-05 2.9807 ID1-02 1.6 1.97E-06 0. Table 5 Production flow rates results for ID1 & ID2 simulated by Pipesim in 3.g.95 ID2-03 99. 2) As on that date.64E-06 0.5 ID2-02 2.8 3527 2.9791 ID2-10 1. our derived choke relation is subjected to a practical test when half of the production system is out of service.35 3978 3.9297 ID2-06 3.63 102 ID2-12 102.44 5065 4.5 ID2-07 96.72 4896 3. 3) Using equation 3.08E-05 1.9607 RESULTS A.36E-05 0.17 ID2-09 110.9923 ID1-09 111.8 E-05 0.51E-05 0. and the relative error percentage is Well number Table 6 Percent error for ID1 & ID2 in 3/11/2010 Percent error Total enriched gas flow rate (reported) Total enriched gas flow rate ( simulated) 1.46E-05 0.2010 gas flow rate condensate flow rate equivalent condensate rate enriched gas flow rate MMSCF/D STBD MMSCFD MMSCFD ID1-01 103.23E-05 0.64 102.9726 ID1-09 1. Applying this relation. Total enriched gas flow rate obtained from choke simulation of the field data set In the previous section.11. 6) Total enriched gas flow rate of all wells is defined as simulated enriched gas flow rate. 4) Simulating the choke in Pipesim to obtain gas and condensate flow rates of each well. a relation between the bean size and the choke opening was derived.63E-05 1. Results for the specific date of 3.43E-05 0.6 ID2-03 2.79E-05 0.9615 ID1-04 2.61 3220 2.05E-05 1.68 105.8 108.49E-05 2.95E-05 3.9852 ID2-04 2. flow rates of enriched gas for each well are calculated using equations 1 and 2.77 107.56E-05 1.75E-05 0.32 3311 2.52 4614 3.2 ID1-13 101.9 ID2-02 41. choke opening percentage and downstream pressure are extracted from the production daily reports.5 ID1-10 92.59E-05 0.5 ID2-06 119. 5) After the gas and condensate flow rates were determined.64 2354 1.9722 ID1-06 129.7 104. well a b ID1-01 1.93E-05 0.7 3. choke is simulated based on the field data set and furthermore the enriched gas flow rate of each well is calculated.78E-05 1.68 2063.3 2091.89 111.95 4725 3.2 ID2-01 109.9763 ID1-03 107.84E-05 0.9789 ID2-09 2.03E-05 1.16 122.64E-05 0.70E-05 1.9837 ID1-03 1.82 4219 3.94E-05 3.05 3635 2.85 1395 1.46E-05 0. The symbol “*” in the following table indicates the mentioned dates. ID2.9 05/05/2007 2316.1 06/01/2008 2309.1 14/1/2006 2108.1 01/12/2009 2294.63 2193.96 2.16 4.9 28/10/2009 2306.86 7.02 5.07 3.2 05/12/2005 1991.57 1741.8 24/11/2006 2289.19 0.16 0.4 1/10/2006* 1154.5 19/6/2007 2306.83 20/9/2010 2095.14 6.07 1724.6 14/11/2005 1823.51 2313.7 19/4/2007 2065.01 2273. ID1.Table 7 Final Choke simulated results for ID1 and ID2 Total enriched gas flow rate (reported) Total enriched gas flow rate (simulated) (MMSCFD) (MMSCFD) 03/11/2010 2091.8 19/8/2008 2223.3 28/10/2010 2133.88 2105.24 1985.41 5.33 7.4 01/09/2009 2348.1 2115.06 0.5 19/10/2006 1908.3 16/7/2006 2037.8 14/10/200* 1125.7 02/03/2010 1854.8 08/07/2009 2320.2 03/04/2006 2150.3 24/4/2007* 1138.2 27/2/2007 2294.13 8.68 10/07/2010 Total enriched gas flow rate (reported) Total enriched gas flow rate (simulated) (MMSCF) (MMSCFD) 01/07/2007 2306.7 4/10/2008* 1066.8 04/09/2008 2300.77 2123.83 8.98 1927.1 1.02 2216.7 09/01/2007 2209 2156.49 2049.47 1928.1 1925.51 0.12 2.05 6 Date percent error Date percent error 17/2/2009 2316.83 1.6 2157.63 2204.9 6.9 26/2/2006 2112. CONCLUSION Main objectives of this study was Calculating gas and condensate flow rate for 20 wells of 2 platforms.75 1123.3 26/4/2008* 1132.1 01/09/2007 2288.86 4 27/8/2007 2282.9 31/5/2006 2055.37 1114.9 0.13 2212.13 2145.7 06/12/2007 2289.3 24/6/2009 2321. Obtained results represent that Pipesim simulator is able to model choke performance for Gas condensate reservoir.8 Mean percentage error =3.26 2173.56 6.76 4.35 6.21 2.38 2213.7 24/5/2010 2231.98 2.77 2.42 0.13 2210.13 0.9 30/10/2005 1747.88 2173.85 0.5 1045.06 3.13 1076.3 08/12/2008 2311.5 14/6/2006 1775.74 0.26 2341.93 7.92 6.13 2275.7 29/9/2008* 1083.1 02/02/2008 2294.63 1046.42 2. after model the choke performance.7 20/6/2010 2082.01 2300.2 2079. the bean sizes 5|P age .44 3.3 08/07/2008 2266.51 2313.4 29/4/2010 2122.13 2119.09 5.89 6.6 29/11/2009 2293.97 0.9 22/4/2009* 1089.06 0.1 2141.86 5 11/03/2009 2318.1 24/6/2008 2300.36 2215.2 6.2 06/01/2010 2316.4 IV.63 2238.54 7.5 1142.13 2349.67 1.08 0.4 06/10/2007 2308.3 17/3/2008 2299.68 2063.5 02/08/2006 1725.83 1699.09 7. In the first step.01 2327.44 1.72 1.7 20/12/2006 2277.69 3.73 2.38 1157.96 2003.17 1.34 3.62 2214.26 1.99 1088.32 1849.83 2211.93 15/8/2010 2177.91 4.86 1979.89 2326.82 1709.89 1.31 3.23 1.83 0.64 2329.37 1.2 06/01/2009 2309.38 2269.35 0.51 2298.51 2184.73 1.81 2137.21 1791.3 5/11/2007* 1124.17 5.7 03/04/2009 2312.13 2227.76 2130.63 1119.74 3.88 2320.3 18/3/2006 1908.4 6/10/2008* 1034.7 04/03/2007 2288.98 2008.13 2198.51 2209.4 03/09/2006 1979. J.. P. P.4 which is not considered as a great difference. [3] Masoud Ahmadi Nia. the total enriched gas flow rate of ID1 and ID2 in different dates is calculated. 2007.E.. [15] Craft. [12] Sachdeva. J.H. V. 27. A. C. [10] Ashford.. This statistical error demonstrates a value of 3. “Two-phase flow choke performance in high rate gas condensate wells”. 1145–1152. T. Afterwards. A. A.. [4] Guo. Lyons.E.. “Two phase Flow Through Wellhead Chokes”. Brill. This choke relation is subjected to a practical test when half of the production system is out of service. [2] Nasriani. “Compressibility effects of two-phase flow”. B. Pet. Pet.. Phys. J.F.A. SPEDC 271. “Two-Phase Flow Through Chokes”.. T..850–863. 2011. [11] Pilehvari.. 2008. 374.. 1974.. H. F. W. “Discussion of paper: an evaluation of critical multiphase flow performance through wellhead chokes. “An Improvement on Gilbert Type Choke Performance Relationship for Iranian Gas Condensate Reservoirs”.. Afterwards.E. Revised by Terry. [14] Baxendell. 2002. N. “Improvement in Sachdeva’s Multiphase Choke Flow Model Using Field Data”. Blais. 2008. Al-Bemani. 20. “Aplicability of Sachdeva's choke flow model in southwest Louisiana gas condensate wells”.. 1975. Dodge.B. 1949. 637– 645. Pipesim ver.S.843. Res.E..C. [9] Gould. 1991..C. Studies show that the suggested model may be used successfully to calculate the gas and condensate flow rate of each platform separately.S. SPE 3742.M. R. Tech.P. R. IChEC. [8] Ros. R..R.. Ghalambor. Kalantariasl. second ed. B. Appl. Consequently. Sci.. total simulated enriched gas flow rate is compared with the reported enriched gas flow rate by the mean of percentage error.. [6] Fortunati.. H. 2. A. P. 1993.. Seifert. 1972. B.L. Ghalambor. 1957... Tech. SPE 75507. SPE Paper 145576.A. “Determining multiphase pressure drop and flow capabilities in down hole safety valves”.E. this model will help to determine the optimum choke to obtain the optimum production Finally.1986 [13] Perkins. Hamid Reza Nasriani. Pet.J. Schmidt.. “Bean performance-lake wells”. a relation between bean size and choke opening percentage is derived which is used to calculate the bean size for different opening percentage by which the specific flow rate of each well can be obtained. “An evaluation of critical multiphase flow performance through wellhead chokes”.H.. M.1976. “An analysis of critical simultaneous gas–liquid flow through a restriction and its application to flow metering”.Tech. J. [7] Ashford.. [16] Schlumberger. Petroleum Production Engineering Elsevier Science &Technology. “Critical and subcritical flow of multiphase mixtures through chokes”. ACKNOWLEDGMENT VI. 1980.. P. Applied Petroleum Reservoir Engineering. 1960.K. [5] Tangeren. Z. 6|P age . by Ashford”.1 User Manual. A. “Experimental study of subcritical two-phase flow through wellhead chokes”.C. SPE15657. Pierce. Ronald E. F.. REFERENCES [1] Guo. App.for test separator data are obtained accurately. 2011.