Abstract

This paper demonstrates how geochemical production allocations can be used to calibrate reservoir simulation models and improve the optimization of well spacing and hydraulic fracture design in unconventional assets. Geochemical analyses provide quantitative assessments of flow by layer over time. This allows numerical models to be fine-tuned to realistically capture the productive fracture height for wells landed in different stratigraphic layers. Model calibration that relies on production and pressure history alone often fails to uniquely resolve important differences in productivity and fracture geometry. Diagnostics such as distributed acoustic sensing, microseismic, and sealed wellbore pressure monitoring capture total hydraulic fracture extent but do not characterize the producing behavior. Thus, it is very valuable to utilize diagnostics that directly assess the producing length and height of fractures. Vertical flow allocation is particularly important in formations with multiple productive benches, such as in the Midland Basin.

This work reveals the connection between completions, geomechanical inputs (such as minimum horizontal stress and toughness), and the geochemical production allocations as demonstrated by the vertical distribution of proppant. This work also identifies the roles that well spacing and drawdown play in time-lapse geochemical production allocations.

Introduction

With attention throughout the industry on delivering shareholder returns, unconventional oil and gas operators are intensely focused on maximizing capital efficiency. Operators have multiple design levers at their disposal carrying productivity and cost implications, including well spacing vertically and laterally, lateral length, stage spacing, cluster spacing, proppant loading, and fluid intensity.

Fig. 1 demonstrates a hypothetical design sensitivity with the wide range of possible outcomes for both value and resource recovery. Development optimization is challenging because of the complex array of design permutations and the variability of formation properties across a basin.

Operators often turn to reservoir simulation to approach these challenges in a robust and cost-efficient manner. However, the utility of simulation models in unconventional assets can be limited by non-uniqueness in the history matching calibration phase. Multiple combinations of permeability, fracture geometry and connectivity, and resource distribution can lead to equally viable matches to production data. Utilizing models that match only production data for optimization sensitivities may lead to vastly different design implications for future development (Fowler et al. 2019).