Uncertainty Study Aids Development-Plan Optimization in Deepwater Gulf of Mexico

Abstract

_ This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper OTC 32622, “Development-Plan Optimization and Uncertainty Study in a Major Deepwater Field in the US Gulf of Mexico,” by Mohsen Rezaveisi, SPE, Jennifer L. Campbell, and Paula L. Wigley, Woodside Energy Group, et al. The paper has not been peer reviewed. Copyright 2023 Offshore Technology Conference. _ This paper describes development-plan optimization and a probabilistic uncertainty study using Latin hypercube experimental design constrained to production performance in Lower Miocene (LM) reservoirs of the deepwater Gulf of Mexico Shenzi Field. The purpose of the development-plan optimization was to identify, rank, and characterize future development opportunities (i.e., infills and injectors) in the LM reservoirs to arrest field decline. The study uses history-matched dynamic simulation models. Introduction The Shenzi discovery is approximately 122 miles off the coast of Louisiana in a water column approximately 4,400 ft deep (Fig. 1). This part of the field is a partially filled three-way closure against a salt-cored anticline. The discovery includes four main sand units (A, B, C, and D) that exhibit various degrees of hydraulic connectivity based on pressure data and production history. The area of interest in the current study is the Eastern part of the field. Static Model Build The geomodel covering the area of interest was built from two seismically interpreted horizons and approximately 180 faults. The model is subdivided into 13 zones using well-based true stratigraphic thickness maps. The average cell height is 2–3.5 ft thick with heterolithics driving the lower end of the range. Sequential indicator simulation was used to populate facies using well-based net sand trend maps by zone. In zones with evidence of channelization, a second nested facies model was used to build channels into the model. Shenzi East reservoir properties are good because of the high-quality sand deposited in the field. Lower reservoirs (C and D) have slightly more depositional complexity because of the varying paleotopography and a lower average net-to-gross (NTG) of approximately 53%. Upper reservoirs (A and B) had a larger sediment supply and were more continuously deposited, which resulted in a larger average NTG of approximately 67%. Dynamic Model History Matching The Shenzi East reservoir has been on production since March 2009, providing abundant production data for history matching. Of seven producers, six are still active, and the reservoir also features three water injectors. Producers and injectors are identified with letters P and I, respectively, followed by well numbers. Well P7 was drilled after 6 months of production from other producers; formation pressure data showing reservoir depletion are available for history matching. The water injectors were drilled in 2012–2013, but only two have formation pressure testing data available. Well P1 was completed in all sands; Wells P2, P3, and P6 were completed in three sands (D, C, and B); and the rest of the producers were completed in two sands (D and C). Production is commingled in all producers, resulting in allocation uncertainties between sands. Water production has been observed in five wells, with the current water cut in the range of 35–65%. The two most updip wells, P1 and P7, are still producing at zero water cut.