SPE 84287 Society of Petroleum Engineers Decline Curve Analysis Using Type Curves — Fractured Wells H. Pratikno, ConocoPhillips (Indonesia), J. A. Rushing, Anadarko Petroleum Corp., and T.A. Blasingame, Texas A&M U. “otain Geert aoe sos ty ee Cr tow ‘Sis he mary Tm raat Psa Sen trey SFE sign ar nye 10 pbs tee y Edn Seen of he osey of ‘ovouin igre, Ere peso Stn nw oat pee [Srumipsran of or ay whch foo nas eer orwan SPE PO. Sere bade akesoos USK havea Abstract This work provides the development, validation, and appli- cation of new decline type curves for a well with a finite ‘conductivity vertical fracture centered in a bounded, circular reservoir. This work fills a significant void in the modern inventory of decline type curves. In particular, this work is rectly applicable to production data analysis for cases taken from low permeability gas reservoir. Using an appropriate analytical solution for this cas, we pre- pared “decline” type curves for Fy values from 0.1 0 1000 — individual type curves are generated foreach Fup value using a range of rap values ffom 2 to 1000. The following "type curves" are provided © "Fetkovich” format rate-time decline type curves (con- ‘stant pressure case): gry VErSUS fry © "Fetkovich-MeCray" format rate-time decline type urves (equivalent constant rat eae): dru Vers Zp @ "Fetkovich-McCray" format rate-cumulative decline type curves: goa versus Nps \We provide an example demonstration ofthe methodology for decline type curve analysis using afield ease of continuously measured production rate and surface pressure data obtained from alow permeability gas reservoir. These solutions‘type curves provide an analysis/interpretation mechanism that has not previously been available in the petroleum literature, Compared to field data, we find that the traditional type curve solutions for an infinite conductivity vertical fracture are typically inadequate — and, the new solu tions for a well with a finite conductivity vertical fracture clearly show much more representative behavior. This vali- dation suggests that the proposed type curves will have broad utility in the petroleum literature — particularly for appli- cations in low permeability gas reservoirs Objectives ‘The following objectives are proposed for this work: © To develop and validate a series of decline type curves for a well with a finite conductivity vertical fracture centered in a bounded, circular reservoir © To provide a methodology for using decline type curves to analyze and interpret production or injection well performance for a well with a finite conductivity verti= cal fracture. © To demonstrate these new type curves using continu ‘ously measured production data (rates and pressures). In considering these objectives we note that we are strongly ‘motivated to provide these tools in light of the current high level of activity in the analysis and interpretation of reservoir performance data acquired from low permeability gas reser- voirs. We recognize that current methods based on the case of a vertical well with an infinite conductivity vertical fracture are overly-ideal for low permeability reservoirs — and we ‘must reconcile the need for a new decline type curve for a {finite conductivity vertical fracture. This rationale is the moti- vation for this work. Prior Work In this section we address the prior work performed on topics relevant to the analysis/interpretation of rate and pressure be- havior for a well with a finite conductivity vertical fracture. ‘The appropriate references are cited as follows: Topic Refs, Historical Methods — Proton Data Analysis 2 Production Data Analysis for Fractured Wells 3 Destine Type Curve Analysis —Fetkovieh McCray Format 6-10 ‘Well Solutions —Finte Conductivity Vertical Fractwe 11-15 Miscellaneous Wels Historical Methods — Production Data Analysis: We cite these references to orient the reader towards historical ‘methods for the analysis of production data. The Arps work (ret. 1) is a compilation of empirical results and field practices, up to about 1940, ‘The Fetkovich work (ref. 2) addresses most production data analysis technologies from 1940-1980, and should be helpful as a general reference.2 H. Pratikno, J. A, Rushing, and TA. Blasingame ‘SPE 84287 Produetion Data Analysis for Fractured Wells: The first reference in this section (Gringarten’) addresses the analytical {and very tedious) solution for a well with an infinite conduc- tivity or uniform flux vertical fracture in a bounded rectan- ‘gular reservoir. This was presented (and is still used) as an analysis mechanism for fractured wells — this solution is for- mulated for a well produced at « constant rate. ‘The work presented by Carter’ creates a "Fetkovich'-style production decline type curve for gas well. It is relevant to this discussion because Carter addresses the question of fractured well ana- Iysis (albeit in the form ofa negative skin factor), ‘The final reference (Fraim, et al.°) presents a comprehensive suite of decline type curves forthe case of a well with a finite conductivity vertical fracture. These type curves were derived ‘numerical simulation and the decline variables were established using regression to estimate a relation between the dimensionless decline variables and ther and Fy variables. ‘Our present work uses analytical solutions to develop the required type curves, and we also establish a relation for the decline variables — although we derive our relations from the analytical solution. While there are other modern techniques developed in the present work (e.g, the rate integral and rate integral-derivative functions, as well as the rate-cumulative type cure), it is important to note that these are essentially refinements — the Fraim, et af. work is robust and functional, and should be regarded as a major contribution in this area, Decline Type Curve Analysis — Fetkovich-MeCray Format: ‘The use of the "Fetkovich-MeCray" type curve format (i.e. the rate integral (qo4) and rate integralderivative (dod) functions, and later, the material balance time function tpg) ‘evolved as an effort to present data functions with better reso- lution (gow and gra), 88 well as to account for variable-ratl variable pressure drop conditions (pj). Refs. 610 provide the current technologies for the analysis of production data using decline type curve analysis. Well Solutions — Finite Conductivity Vertical Fracture: This section recounts the solutions we have used or adapted in order to model the case of a well with a finite conductivity vertical fracture produced at a constant rate in a bounded reservoir. Cinco, et al’ presented a semi-analytical, real domain solution for a well with a finite conductivity vertical fracture in an infinite-acting reservoir. This semi-analytical solution is tedious, but accurate Houze, er ai.'" presented a Laplace domain solution for a well with an infinite conductivity vertical fracture in an in-finite- acting reservoir, while Cinco and Meng"’ provide a Laplace domain solution for a well with a finite conductivity vertical fracture in an infinite-acting reservoir. ‘The use of a Laplace domain solution offers some convenience, and while such solutions are not necessary (¢.g., the work of Fraim, ef al.*), ‘we will use a Laplace domain solution in this work. Using the techniques presented by Ozkan and Raghavan,'*"* for the case of a well with an infinite conductivity vertical fracture, we develop a "desuperposition” solution for the case of a well with a finite conductivity vertical fracture in a finite (circular) reservoir — which is based on the transient flow solution given by Cinco and Meng."” Miscellaneous: Refs, 16 and 17 provide details of the "Gaver- Stehfest" algorithm, the most common toot used to numeri- cally invert Laplace transform solutions, Ref. 18 is the re- search thesis developed by the principal author on the subject of this work, Desuperposition Solution We have developed and validated a new "desuperposition” solution for the ease of a well produced at a constant rate with a finite conductivity vertical fracture in a finite-acting (cir- cular) reservoir. Schematically, this “desuperposition” solu- tion is expressed as: (Pound Sie codavy se (Poni Sie conus se (Potts cee vo (I) We readily acknowledge that other solutions may exist for a well with a finite conductivity vertical fracture in a bounded system — our goal in using desuperposition was to develop a relatively fast and accurate routine for this particular case ‘The proposed solution is validated using numerical simulation and should be considered appropriate for the analysis of well performance data (flowrates and pressures) taken from wells, which have under-gone hydraulic fracture stimulation treat- ‘ments, TThe desuperposition solution for the case of a well producing at a constant rate with a finite conductivity vertical fracture in 4 finte-acting (circular) reservoir" is given below: Poawllxp|sLyp Op Stap.s)= fl . Re q . “5 J, ape shea do Jo psy pm 9 ull xp) vl ee . A sonnei (2) Where #=sf(s) — ls) is the interporosity flow function for the dual porosity reservoir ease (f(s)=1 for this work). Validation of the Desuperposition Solution: In order to validate Eq, 2, we use results generated by numeri cal simulation for F,j=0.2x, = 2x, 10x The comparison of solutions is shown in Figs. 1-4 — we note excellent agreement between our desuperposition solution and the numerical simu- lation results. We consider this validation sufficient to pro- ceed with the development of new decline type curves using the desuperposition solution. Kyi nUdarep) Pseudosteady-State Performance Correlation: boys Recalling the general (dimensionless) identity’ for pscudo- steady-state performance flow, we have: @) Po,pss D4) = Sop ss +2 D4 ~