Rate transient analysis ("RTA") is used to determine fracture and reservoir parameters and well geometries in unconventional reservoirs. It is applied to production forecasting, evaluating completion effectiveness, type well construction and well spacing optimization. Choke management and data quality during the initial production period (i.e., within the first two months of a well's life) has been recognized as crucial to optimizing well performance in unconventional reservoirs and in the early determination of completion effectiveness. This paper illustrates the additional significance of the initial production period by showing how model matches to initial production data are used to seed models for EUR estimation.
Linear flow analysis is one of the most common methods for analyzing well performance in unconventional reservoirs. One of the key insights presented in this paper is that accurate, hourly, automated measurements that preserve reservoir transients indicate that fracture interference may be occurring earlier than suggested in linear flow analysis. Fracture interference was identified with hourly data on Log-Log plots of Rate Normalized Pressure ("RNP") vs. Material Balance Time ("MBT") and Flowing Material Balance ("FMB") plots generated from numerical models and field data. Numerical models were used to test and validate well geometry assumptions indicating early interference for single and multifracture examples and then the workflow was applied to wells in the SPE Data Repository.
Fracture and reservoir parameters determined from analysis of initial production data adequately seeded models for an oil well and a gas well in the SPE data repository. The model match using the fracture half-length and permeability from the initial production period started to over predict cumulative production when a unit slope was observed on flow regime identification plots. The model matched the data when constrained by the contacted volume for each well, indicating that the analysis method can be accurate when boundaries are defined and can help identify the magnitude of lost performance if the fracture half-length and permeability degrade, or interference occurs. This workflow has now been applied to hundreds of wells across multiple unconventional basins in the US.
Early fracture interference has been proposed previously, but to the authors’ knowledge this is the first time it has been demonstrated using hourly initial production data. The reservoir transients that may indicate earlier facture interference are masked by lower resolution daily data and low-quality hourly data with measurement errors. This paper illustrates the additional significance of high frequency/high quality initial production data by showing that the facture half-length, permeability, and number of dominant fractures determined within the first 2 months of production are representative of performance until interference occurs or performance starts to degrade.