2. Well Performance Analysis in a Multiwell Gas Condensate Reservoir— Arun Field, Indonesia T. Marhaendrajana, Texas A&M University N.J. Kaczorowski, Mobil E&P (U.S.) T.A. Blasingame, Texas A&M University Presented at SPE Advanced Technology Workshop Well Testing in Gas Condensate Reservoirs 1-2 April 2000, Calgary, Alberta, Canada

3. Summary —Well Test Analysis A representative summary of the analysis and interpretation of well test data taken from the Arun Gas Field (Sumatra, Indo- nesia) (Single-phase gas analysis is used). 2-zone radial composite reservoir model is effective for diagnosing the effects of con- densate banking at Arun Field. Application of a new solution for the analysis and interpretation of well test data that exhibit "well interference" effects.

4. Analysis and interpretation of production data using (single-phase) decline type curve analysis: Permeability-thickness product Skin factor Original and movable gas-in-place Comparison of results from well test and production data analyses vary—but these variations appear to be consistent. Summary —Production Data Analysis

5. N Arun Field Located in Northern part of Sumatra, Indonesia Retrograde gas reservoir One of the largest gas fields in the world Arun Field has 111 wells: 79 producers 11 injectors 4 observation wells 17 wells have been abandoned Field Description General Information—Arun Field (Indonesia)

6. Major Phenomena Observed at Arun Field Liquid accumulation near wellbore (conden- sate banking) Need to know the radial extent of the condensate bank for the purpose of well stimulation. Well interference effects (well test analysis) Well interference effects tend to obscure the "standard" flow regimes—in particular, the radial flow response. This behavior influences our analysis and inter- pretation efforts, and we must develop an alter- native analysis approach for well test data affected by multiwell interference effects.

7. Well Test Analysis Strategy Condensate banking phenomenon: Used a 2-zone radial composite reservoir model—the inner zone represents the "con- densate bank," and the outer zone represents the "dry gas reservoir." (reported kh-values are for the "outer zone") Well interference effects: Developed a new method for the analysis of well test data from a well in multiwell reservoir where we treat the "well interference" effect as a "Regional Pressure Decline." (This pheno- mena is observed in approximately 35 cases)

8. Well Test Analysis: Examples Well C-I-18 (A-096)—Test Date: 28 Sep. 1992 Radial composite effects. Multiwell interference effects. Well C-IV-01 (A-060) [Test Date: 25 Feb. 1993] Radial composite effects. Multiwell interference effects.

9. Condensate banking region. Higher mobility region. Closed boundary at 160 ft? (includes non-Darcy flow). Pseudopressure Functions, psi Effective shut-in pseudotime, hrs Infinite-acting Reservoir Model (Does not include non-Darcy flow) Improvement of pressure derivative. WT Example 1: Log-log Summary Plot Raw data Corrected

10. Condensate banking region. Higher mobility region. Well C-I-18 (A-096) [Test Date: 28 September 1992] Shut-in Pseudopressure, psia Horner pseudotime, hrs (t p = 1.56 hr) WT Example 1: Horner (Semilog) Plot Raw data Corrected

11. WT Example 1: Muskat Plot (single well p avg plot) p p,avg = 1148.6 psia "Transient flow" Data deviate from the "Muskat line" —indicating interference effects from surrounding wells. Onset of boundary dominated flow (single well analogy). Shut-in pseudopressure, psia dp pws /d  t a, psi/hr Well C-I-18 (A-096) [Test Date: 28 September 1992]

12. WT E x. 1: "Well Interference" Plot (radial flow only) (  p p ')  t ae, psi  t a 2 /  t ae Well C-I-18 (A-096) [Test Date: 28 September 1992] Intercept is used to calculate permeability. Slope is used in the pressure correction. (  p p ')  t ae <0, indicating multiwell interference effects.

13. Condensate banking region. Higher mobility region. Closed boundary at 330 ft? (includes non-Darcy flow). Improvement of pressure derivative. Infinite-acting Reservoir Model (Does not include non-Darcy flow) WT Example 2: Log-log Summary Plot Pseudopressure Functions, psi Effective shut-in pseudotime, hrs Well C-IV-01 (A-060) [Test Date: 25 February 1993] Raw data Corrected

14. Shut-in Pseudopressure, psia Horner pseudotime, hrs (t p = 2.61 hr) WT Example 2: Horner (Semilog) Plot Well C-IV-01 (A-060) [Test Date: 25 February 1993] Raw data Corrected Condensate banking region. Higher mobility region.

15. WT Example 2: Muskat Plot (single well p avg plot) p p,bar = 1573.5 psia Onset of boundary dominated flow. "Transient flow" Shut-in pseudopressure, psia dp pws /d  t a, psi/hr Well C-IV-01 (A-060) [Test Date: 25 February 1993]

16. WT Ex. 2: "Well Interference" Plot (radial flow only) (  p p ')  t ae, psi  t a 2 /  t ae Well C-IV-01 (A-060) [Test Date: 25 February 1993] Intercept is used to calculate permeability. Slope is used in the pressure correction. (  p p ')  t ae <0, indicating multiwell interference effects.

17. Correlation of Well Test Results—Arun Field Maps: kh (outer-zone (gas) permeability). skin factor. non-Darcy flow coefficient. Radius of condensate bank. Correlation of non-Darcy flow coefficient and the permeability-thickness product (kh).

18. kh distribution ap- pears reasonable. 3 major "bubbles" of kh noted, pro- bably erroneous. kh shown is for the "outer" zone (when the radial compo- site model is used). kh Map

19. Skin factor distri- bution appears very consistent. Areas of "high skin" indicate need for individual well stimulations. Skin factors are calculated based on the "inner" zone of the radial compo- site model (when rc- model is used). Skin Factor Map

20. This map indicates a uniform distribution. "high" and "low" regions appear to be focused near a single well. Relatively small data set (30 points). D (Non-Darcy) Map

21. Good distribution of values—"high" spots probably indicate need for individual well stimulations. Relatively small data set (32 points). Condensate Radius Map

22. D-kh crossplot indi- cates an "order of magnitude" correla- tion. Verifies that non- Darcy flow effects are systematic. D (Non-Darcy)—kh Crossplot Slope = 2

23. Production Data Analysis: Arun Field Well C-I-18 (A-096) Limited history (no EUR MB analysis possible). Erratic performance. Reasonable match on decline type curve. Well C-IV-01 (A-060) Good history (well was down for almost two years in 1993-1995). Sparse p/z data for EUR MB analysis. Early data match on decline type curve is questionable. Late performance data deviate from material balance trend on decline type curve, indicat- ing "well interference" effects.

24. WPA Example 1: Well History Plot

25. WPA Example 1: Decline Type Curve Plot Fetkovich-McCray Decline Type Curve (No Well Interference Effects)

26. WPA Example 2: Well History Plot

27. WPA Example 2: EUR MB Plot

28. WPA Example 2: Decline Type Curve Plot Fetkovich-McCray Decline Type Curve (No Well Interference Effects)

29. WPA Example 2: Decline Type Curve Plot Fetkovich-McCray Decline Type Curve (Includes Well Interference Effects)

30. Correlation of Production Analysis Results—Arun Field Maps: kh skin factor Crossplots: kh WT —kh WPA s WT —s WPA Production data analyzed using decline type curve analysis—single-phase (gas) p p and t a used. Flow Properties:Volumetric Properties: Plots: G vs. time EUR MB vs. time kh vs. time Crossplots: G—EUR MB EUR PI —EUR MB

31. kh shown is com- puted using decline type curve analysis on early production (p p and t a used). kh distribution ap- pears reasonable (albeit lower than WT estimates). A few "bubbles" of kh noted, these are probably erroneous. kh Map

32. Skin factors com- puted using decline type curve analysis on early production. Skin factors from WPA are lower than WT estimates. Skin Factor Map

33. kh WT estimates are clearly higher than kh WPA estimates. kh WT estimates are "current," kh WPA estimates are "ini- tial." Variation is system- atic—decline type curve analysis uses earliest production data for kh (and s) estimates. kh WT —kh WPA Crossplot

34. s WT estimates are "current," s WPA estimates are "ini- tial." Could argue that this plot shows the "evolu- tion" of the skin fac- tor. s WPA estimates should be higher, tied to kh estimation in decline type curve analysis. s WT —s WPA Crossplot

35. Gas Reserves from Decline Type Curve Analysis

36. Estimated Ultimate Recovery from Material Balance

37. Flow Capacity (kh) from Decline Type Curve Analysis

38. G-EUR MB crossplot indicates excellent agreement of com- puted results. G (Decline Type Curve)—EUR MB Crossplot

39. EUR PI —EUR MB cross- plot shows excellent correlation of results. Verifies that these analyses are consis- tent. EUR PI —EUR MB Crossplot

40. Well Performance Analysis in a Multiwell Gas Condensate Reservoir— Arun Field, Indonesia T. Marhaendrajana, Texas A&M University N.J. Kaczorowski, Mobil E&P (U.S.) T.A. Blasingame, Texas A&M University Presented at SPE Advanced Technology Workshop Well Testing in Gas Condensate Reservoirs 30 September- 1 October 1999, Houston, TX.

41. Well Test Analysis: Examples (extra) Well C-IV-11 (A-084)—Test Date: 5 Jan. 1992 Multiwell interference effects. Well C-IV-11 (A-084)—Test Date: 4 May 1992 Radial composite effects. Multiwell interference effects.

42. Closed boundary at 150 ft? (includes non-Darcy flow). Pseudopressure Functions, psi Effective shut-in pseudotime, hrs Infinite-acting Reservoir Model (Does not include non-Darcy flow) Improvement on pressure derivative. Example 3: Log-log Summary Plot Well C-IV-11 (A-084) [Test Date: 5 January 1992] Raw data Corrected

43. Shut-in Pseudopressure, psia Horner pseudotime, hrs (t p = 1.62 hr) Example 3: Horner (Semilog) Plot Well C-IV-11 (A-084) [Test Date: 5 January 1992] Raw data Corrected

44. Shut-in pseudopressure, psia dp pws /d  t a, psi/hr p p,bar = 1920 psia Onset of boundary dominated flow. "Transient flow" Example 3: Muskat Plot (single well p avg plot) Well C-IV-11 (A-084) [Test Date: 5 January 1992]

45. (  p p ')  t ae, psi  t a 2 /  t ae Example 3: "Well Interference" Plot (radial flow only) Well C-IV-11 (A-084) [Test Date: 5 January 1992] Intercept is used to calculate permeability. Slope is used in the pressure correction. Presence of multiwell interference effects is unclear

46. Condensate banking region. Higher mobility region. Closed boundary at 197 ft? (includes non-Darcy flow). Improvement on pressure derivative. Infinite-acting Reservoir Model (Does not include non-Darcy flow) Pseudopressure Functions, psi Effective shut-in pseudotime, hrs Example 4: Log-log Summary Plot Well C-IV-11 (A-084) [Test Date: 4 May 1992] Raw data Corrected

47. Condensate banking region. Higher mobility region. Shut-in Pseudopressure, psia Horner pseudotime, hrs (t p = 1.63 hr) Example 4: Horner (Semilog) Plot Well C-IV-11 (A-084) [Test Date: 4 May 1992] Raw data Corrected

48. p p,bar = 1882.8 psia Onset of boundary dominated flow. "Transient flow" Shut-in pseudopressure, psia dp pws /d  t a, psi/hr Example 4: Muskat Plot (single well p avg plot) Well C-IV-11 (A-084) [Test Date: 4 May 1992]

49. (  p p ')  t ae, psi  t a 2 /  t ae Example 4: "Well Interference" Plot (radial flow only) Well C-IV-11 (A-084) [Test Date: 4 May 1992] Intercept is used to calculate permeability. Slope is used in the pressure correction. (  p p ')  t ae >0, no clear indication of multiwell interference effects.