SPE 109625 Estimating Reserves in Tight Gas Sands at HP/HT Reservoir Conditions: Use and Misuse of an Arps Decline Curve Methodology

This paper presents the results of a simulation study designed to evaluate the applicability of an Arps decline curve methodology for assessing reserves in hydraulically-fractured wells completed in tight gas sands at high-pressure/high-temperature (HP/HT) reservoir conditions. We simulated various reservoir and hydraulic-fracture properties to determine their impact on the production decline behavior as quantified by the Arps decline curve exponent, b. We then evaluated the simulated production with Arps' rate-time equations at specific time periods during the well's productive life and compared estimated reserves to the true value. To satisfy requirements for using Arps' models, all simulations were conducted using a specified constant bottomhole flowing pressure condition in the wellbore. Our study indicates that the largest error source is incorrect application of Arps' decline curves during either transient flow or the transitional period between the end of transient and onset of boundary-dominated flow. During both of these periods (principally the transient period), we observed bexponents greater than one and corresponding reserve estimate errors exceeding 100 percent. The b-exponents generally approached values between 0.5 and 1.0 as flow conditions approached true boundary-dominated flow. Agreement between Arps' suggested b-exponent range and our results using simulated performance data also indicates that, if applied under the correct conditions, the Arps rate-time models are appropriate for assessing reserves in tight gas sands at HP/HT reservoir conditions. Introduction Tight gas sands constitute a significant percentage of the domestic natural gas resource base and offer tremendous potential for future reserve and production growth. According to a recent study by the Gas Technology Institute (GTI), tight gas sands in the US comprise 69 percent of gas production from all unconventional natural gas resources and account for 19 percent of total gas production from both conventional and unconventional sources. The same study estimates total domestic producible tight gas sand resources exceed 600 Tcf, while economically recoverable gas reserves are 185 Tcf. Most of the resources assessed in the 2001 GTI study were at depths less than 15,000 ft, yet the natural gas industry continues to extend exploration and development activities to much greater depths. In some geologic basins, those depths are approaching 20,000 to 25,000 ft. Many of these deep natural gas resources are not only characterized by lowpermeability, low-porosity reservoir properties, but these reservoirs also exhibit abnormally high initial pore pressure and temperature gradients — i.e. high-pressure/hightemperature (HP/HT) reservoir conditions. Similar to conventional natural gas resources, tight gas sand reserves are routinely assessed with Arps’ decline curve techniques. The original Arps paper suggested the decline curve exponent, b, should fall between 0 and 1.0 on a semilog plot. However, we often observe values much greater than 1.0, particularly in tight gas sands at HP/HT reservoir conditions. Deviations in observed b-exponents from the expected range suggest Arps' rate-time relationships may not be valid for modeling the decline behavior of tight gas sands at HP/HT conditions. More importantly, inappropriate use of the Arps models may cause significant reserve estimate errors in these unconventional natural gas resources. Since these depths and extreme reservoir conditions require wells that are very expensive to drill, complete and operate; it is imperative that we understand both the well productivity and production decline behavior. We also need to determine the applicability of the Arps rate-time equations for assessing reserves. To address these concerns, we have conducted a series of single-well simulation studies to develop a better understanding of both the shortand long-term production decline behavior and to identify those parameters affecting the production decline. In this study we simulated a range of reservoir and hydraulic fracture properties, including: Vertical heterogeneity from layering, permeability contrast among layers, horizontal permeability anisotropy, and stressdependent reservoir properties; 2 J.A. Rushing, A.D. Perego, R.B. Sullivan, and T.A. Blasingame SPE 109625 Variable effective fracture conductivities and lengths, unequal fracture wing lengths, two-phase and non-Darcy flow, and stress-dependent fracture properties; and Reservoir temperatures ranging from 300 to 400F and initial pore pressure gradients ranging from 0.60 to 0.90 psi/ft. We evaluated the simulated production with the Arps ratetime equations. Reserve estimates were obtained at various time periods during the well’s productive life by extrapolating the best-fit Arps model through the simulated production. Our assumed economic conditions for estimating reserves were either a rate of 50 Mscf/d or a producing time period of 50 years, whichever came first. Reserve estimate errors were computed by comparing those estimated reserves to the “true” value. For this paper, we define the “true” estimated ultimate recovery (EUR) to be the 50-year cumulative production volume. For reference, we also summarize the Arps rate-time equations in Table 1, given below: Table 1 — Summary of the Arps' rate-time relations (Ref. 1)