SPE 71514 A Semianalytical pz Technique for the

SPE 71514 A Semianalytical p/z Technique for the Analysis of Abnormally Pressured Gas Reservoirs Ronald Gunawan Gan, VICO Indonesia and T. A. Blasingame, Texas A&M University

Objective To present a new technique that can be used to : · Calculate gas-in-place for an abnormally pressured gas reservoir using only average reservoir pressure and cumulative production data. · Calculate pore volume compressibility as a function of reservoir pressure.

Presentation Outline · Introduction · Overview of Existing Methods · New Method · Field Examples · Conclusions

Introduction l p/z schematic for a normally-pressured volumetric gas reservoir p/z Gp G

Introduction p/z schematic for an abnormallypressured gas reservoir l p/z Gp G Gapp

Introduction · Reasons for the non-linear p/z behavior: n Rock and water compressibility effects — "rock collapse theory" (Hawkins, 1969) n Shale water influx (Bourgoyne, 1989)

Existing Methods · Methods based on presumed knowledge of system compressibility: n Hammerlindl (Constant Compressibility), 1971 n Ramagost & Farshad (Constant Comp. ), 1981 n Yale et al. (Variable Compressibility), 1993

· Methods based on presumed knowledge of system compressibility (continued) n Fetkovich, Reese, and Whitson - 1991 - Derived General Material Balance Eq. - Define cumulative effective compressibility, - ce represents the cumulative change in hydrocarbon PV caused by compressibility effects (and water influx).

· Methods which do not require a prior knowledge of system compressibility n Roach - 1981 - very sensitive to initial pressure. - method sometimes doesn’t exhibit a negative intercept (which is not possible). n Bernard - 1985 - using Least Squares approach. - very sensitive to data scatter. n Ambastha - 1991: Type Curve Approach - non-uniqueness problems.

New Method · Satisfies both "rock collapse" and "shale water influx" theories · Develops 2 new plotting functions: 1. 2. · Requires production data only (p and Gp)

New Method · Uses general material balance equation (proposed by Fetkovich, et al. ) · Rearranging, we obtain

New Method · Calculate the ce(pi-p) function for each p/z versus Gp trend ce(pi-p) = ? ? ? p/z ce(pi-p) = ? ? ? Gp G Gapp

New Method · For early time data (1 st straight line) : · For late time data (2 nd straight line) : where: A is the inflection point

New Method · Plot of log ce(pi-p) versus (p/z)/(pi/zi): inflection point log ce(pi-p) G/Gapp=0. 6 h G/Gapp=0. 7 G/Gapp=0. 8 (p/z)/(pi/zi)

New Method · Plot of log ce(pi-p) versus (p/z)/(pi/zi) : log ce(pi-p) inflection point h (p/z)/(pi/zi)

New Method · (p/z)/(pi/zi) 1 Infl. Point: Gp. A/G, (p/z)A /( pi /zi ) h 0 Gp/G 1

New Method · 1 (p/z)/(pi/zi) G/Gapp=0. 6 Inflection point h G/Gapp= 0. 8 G/Gapp=1 0 Gp/G 1

New Method · 1 (p/z)/(pi/zi) Inflection point h G/Gapp=0. 8 0 Gp/G 1

New Method · n Dynamic Type Curve Matching. Automatic Matching using SOLVER m(Excel function for non-linear regression). n

New Method · Data required for analysis: n Fluid property data n Initial Reservoir p and T n p and Gp data

New Method · Computer program: Visual Basic Application in MS Excel n Only requires MS Excel n Easy to use - especially for analysis n

Data Analysis Sheet

Example 1: G is too low

Example 1: G is too high

Example 1: Correct G

Example 2: Long transition period

Example 3: Early time data

Example 4: Synthetic Dry Gas Case

Example 4: Backcalculated cf · Procedure to calculate cf vs. p from production data: 1. Get from type curve matching 2. Use the following equation to calculate 3. Calculate cf (p): :

Example 4: Backcalculated cf

Conclusions · We have developed a straightforward approach for analyzing p/z versus Gp behavior for abnormally pressured gas reservoirs — the approach considers that two straight-lines must be observed on the p/z plot. · The proposed method determines gas-in-place without using system compressibility data. Only p, Gp, and fluid property data are required.

Conclusions (continued) · Our approach of using ce(pi-p) versus (p/z)/(pi /zi) and (p/z)/(pi /zi) versus Gp/G as dynamic type curve matching functions has been shown to work extremely well. · Using our new method, it is possible to calculate rock compressibility as a function of pressure from p and Gp data

Conclusions (continued) · The "dynamic type curve matching technique" used for calculating gasin-place from production data is more representative (and more stable) than the non-linear optimization method provided by SOLVER.

SPE 71514 A Semianalytical p/z Technique for the Analysis of Abnormally Pressured Gas Reservoirs Ronald Gunawan Gan, VICO Indonesia and T. A. Blasingame, Texas A&M University