Abstract

Well spacing and hydraulic fracture design optimization are among the most important challenges confronting companies operating in unconventional reservoirs. Field trials are time consuming and expensive. Reservoir simulation and/or rate transient analysis can help guide development decisions, but these calculations can be affected by non-uniqueness. For example, it is not possible to resolve permeability and fracture geometry using only production and pressure data in rate transient analysis. This work demonstrates that tracers can be used to reduce non-uniqueness. We quantitatively apply tracer measurements as part of the calibration and history matching of a fully coupled 3D hydraulic fracturing, geomechanics, and reservoir simulator. With the use of calibrated models, forward modeling and sensitivity analysis can be used more accurately to guide better decisions about well spacing and hydraulic fracture design. Tracers are complementary to data sources such as microseismic and distributed acoustic sensing, which focus on hydraulic fracture creation but provide less constraint on the producing behavior of wells, which ultimately drives asset financial performance.

Introduction

Operators in unconventional oil and gas reservoirs utilize a variety of economic metrics to guide investment decisions, including: net present value, rate of return, and capital efficiency. Horizontal wells and their completion comprise a large portion of an operator’s capital spend. Thus, it is critical to optimize this spend.

Fig. 1 depicts a theoretical well spacing sensitivity. There are numerous choices an operator can make regarding the completion of a well, including: stage spacing, cluster spacing, perforation configuration, fluid intensity, and proppant loading. This quickly becomes a complex problem when trying to link the cost of changing these parameters to the design objective, as depicted in the plot of net present value (NPV) vs. estimated ultimate recovery (EUR).

To perform a sensitivity analysis like the one depicted in Fig. 1, operators often use numerical simulators. However, it can sometimes be challenging to validate their underlying inputs. Non-uniqueness is particularly problematic in ultra-low permeability reservoirs because the time between diagnostic flow regimes exceeds the short window needed for investment decisions. This is especially true if the history matched model is constrained solely by bottomhole pressures and oil/water/gas rates. Simulation models matched solely to production data may yield orders of magnitude differences in permeability and may be matched with different conceptual models such as stimulated rock volume (SRV). Resulting path-forward models may result in major differences regarding critical development recommendations around well spacing and completion design (Fowler et al. 2019).