Integration Of Time-Lapse Geochemistry to Enhance Subsurface Characterization at Hydraulic Fracture Test Site II

Abstract

Understanding hydraulic fracture propagation remains a challenge for oil and gas operators when exploiting resource plays. Hydraulic Fracturing Test Site-2 (HFTS-2) is a well-studied industry and government consortium project area designed to further explore this challenge. The site is in the Wolfcamp Formation of the Delaware Basin. Several techniques were deployed to understand flow units at HFTS-2, including geochemical analysis of produced hydrocarbons, which provides a direct measurement of subsurface fluid flow. This information is integral to understanding the initial stimulated rock volume (SRV) and also the evolution of the drained rock volume (DRV) over time. Production allocation is the statistical analysis of geochemical parameters allowing for deconvolution of multiple reservoir contributions to hydraulic fracture networks. Samples are captured and analyzed throughout the wells’ lifecycles and tracking this temporal evolution is termed time lapse geochemistry (TLG). Through TLG, direct evidence of vertical drainage columns can complement fracture models, evidencing annealment of fractures and effects of nearby parent or child wells. At HFTS-2, four lateral wells drilled in 2019 were analyzed five times over the initial 18 months of production. Wells were landed in Upper Wolfcamp formations. Oil samples were sent to GeoMark Research for analysis. Utilizing measurements of lipid biomarkers, carbon isotopes, and common bulk oil properties, statistical approaches were employed to allocate contributions of oil families to the wells. Unique oil families were identified in Wolfcamp X/Y and Wolfcamp A formations. Temporal evolution of the contribution of these families to the produced fluids of the four wells show that initially, all the wells had contribution from vertically surrounding units. Subsequent time samples showed that vertical contributions were reduced, and significantly so in the lower wells, and by the time of the last sample, stratigraphic contributions were limited to rock units where the wells were landed. The allocations are integrated with logs, formation evaluation, hydraulic stimulation models, and microseismic to provide a wholistic view of field development. Microseismic and modelling suggest large SRV overlap (~60%) between Wolfcamp A and Wolfcamp X/Y wells and is supported by the early round of geochemical data. Significantly, the subsequent TLG allocation model suggests moderate (~5%) to very little overlap in contribution after 18 months. These conclusions illustrate that the modeled and measured SRV quickly change (over weeks to months) to geochemically measured DRV. The rapid change allows for the reconsideration of completions style and landing patterns in order to optimize resource extraction.