Integrated Petrophysical Evaluation of Unconventional Reservoirs in the Delaware Basin

Abstract

Abstract The Delaware Basin of west Texas and southeast New Mexico has seen a resurgence of drilling activity reflecting advancements in horizontal drilling, hydraulic fracturing technologies, and the industry's focus on oil assets. The targeted reservoirs include the Delaware Mountain Group sands, Bone Spring tight sands, and the Avalon and Wolfcamp shales. These formations represent a series of multiple-stacked transgressive sequences comprised of naturally-fractured, low-porosity interbedded carbonates, clastic sands, and shales. Formations are composed of varying amount of quartz, calcite, dolomite, kerogen, illite, albite, and pyrite. This mix of minerals leads to grain densities that vary from 2.5 g/cc to 2.7 g/cc and pose a major challenge when estimating porosity, water saturation, and net pay. A grain density uncertainty range of 0.2 g/cc can increase the error bars on porosity by 6 porosity units and dramatically impact resource estimation. Therefore, an uncalibrated petrophysical interpretation in this complex environment leads to large uncertainties in calculated values. Addressing the grain density issue requires a clear understanding of the mineralogy from core XRD and availability of geochemical logs. Additionally, a high quality logging suite including a triple-combo, NMR, spectral GR, dipole sonic, imaging logs, and geochemical logs are needed. These are analyzed by using calibrated mineral models, mudlogs, drilling parameters, comparison with core data, and production tracer data to deliver a reliable interpretation to be used for production forecasting. If the logging suite consists of only triple-combo and Spectral GR logs, analytical techniques in conjunction with mineral modeling can be used to estimate total organic carbon content (TOC) and porosity but such methods yield higher uncertainty in petrophysical parameters. This paper describes a case study on a complete logging and interpretation program. A workflow is presented based on mineral modeling of both pilot holes and laterals wells depending on the available data. One key lesson learned in this exercise is to understand the accuracy and precision of each measurement and plan ahead for redundancy as operational constraints can pose a challenge when relying on only one technique or technology for interpretation. Our results show that default uncertainty bounds for the logging suite may need to be changed and the error bars have to be widened to account for log repeatability. Comparing the rotary cored depths and the resistivity imaging tools indicate issues with depth control that can be attributed to tool string motion and cable tension. Imaging logs showed many drilling induced fractures in the target intervals but formation testing with straddle-packers in the pilot did not provide any successful pressures or formation fluid samples due to the low permeability and lack of a natural fracture network in the near-wellbore region. However, we were able to successfully induce fractures in multiple zones using a micro-frac tool and the results compare favorably with geomechanical logs. Logging while drilling (LWD) measurements in the lateral showed significant lithology variations is compared and validated with production tracers.