This paper was prepared for presentation at the SPE International Hydraulic Fracturing Technology Conference & Exhibition, Muscat, Oman, 11-13 January 2022.
In URTeC-123-2019, a group of operators and service companies presented a step-by-step procedure for interpretation of diagnostic fracture injection tests (DFITs). The procedure has now been applied on a wide variety of data across North and South America. This paper statistically summarizes results from 62 of these DFITs, contributed by ten operators spanning nine different shale plays. URTeC-123-2019 made several novel claims, which are tested and validated in this paper. We find that: (1) a ‘compliance method’ closure signature is apparent in the significant majority of DFITs; (2) in horizontal wells, early time pressure drop due to near-wellbore/midfield tortuosity is substantial and varies greatly, from 500 to 6000+ psi; (3) in vertical wells, early-time pressure drop is far weaker; this supports the interpretation that early- time pressure drop in horizontal wells is caused by near-wellbore/midfield tortuosity from transverse fracture propagation; (4) the (not recommended) tangent method of estimating closure yields Shmin estimates that are 100-1000+ psi lower than the estimate from the (recommended) compliance method; the implied net pressure values are 2.5x higher on average and up to 5-6x higher; (5) as predicted by theory, the difference between the tangent and compliance stress and net pressure estimates increases in formations with greater difference between Shmin and pore pressure; (6) the h-function and G-function methods allow permeability to be estimated from truncated data that never reaches late-time impulse flow; comparison shows that they give results that are close to the permeability estimates from impulse linear flow; (7) false radial flow signatures occur in the significant majority of gas shale DFITs, and are rare in oil shale DFITs; (8) if false radial signatures are used to estimate permeability, they tend to overestimate permeability, often by 100x or more; (9) the holistic-method permeability correlation overestimates permeability by 10-1000x; (10) in tests that do not reach late-time impulse transients, it is reasonable to make an approximate pore pressure estimate by extrapolating the pressure from the peak in t*dP/dt using a scaling of t^(-1/2) in oil shales and t^(3/4) in gas shales. The findings have direct practical implications for operators. Accurate permeability estimates are needed for calculating effective fracture length and for optimizing well spacing and frac design. Accurate stress estimation is fundamental to hydraulic fracture design and other geomechanics applications.