In a hybrid planar-DFN frac/flowback history match on a 2020 Wolfcamp A well in Reeves Co., flipping tNavigator from Carter leakoff to pressure-dependent leakoff bumped my 5-yr oil forecast by 6.8% via slower closure and slightly higher effective conductivity. Is that a reasonable sensitivity in your models, or am I masking completion variability with a leakoff knob?
‘6.8% in 5-yr oil’ is plausible, but only if PDL is DFIT-calibrated — match shut-in G-function slope and late-time derivative, then rerun. Freeze geometry and sweep the PDL exponent/Cw +/-25%; if EUR moves more than your pad P50 spread, you’re masking completion variability. Quick ref: pressure dependent leakoff | Page 1 | Search Results | OnePetro.
Seems reasonable, but I’d sanity-check by freezing geometry and residual width, then sweeping only the PDL coef/exponent and seeing if your early flowback pressure derivative still lines up. @kevin_m193’s DFIT point is solid; also compare against stage-by-stage shut-in mini-falloffs right after screenout — if the uplift disappears there, you were turning the leakoff knob into a EUR volume knob. Do you have stage-level WHP/CHP during controlled drawdown to run that quick slope check?
I’d gate that 6.8% by reconciling PDL to first-30‑day water recovery and the shut‑in on that 2020 Wolfcamp A — if modeled leakoff mass doesn’t match cleanup, PDL’s absorbing completion noise. Then do a swap test: keep PDL on but disable stress‑dependent conductivity; if the EUR bump dies, it’s really SDC/closure, not leakoff. @kevin_m193 have you seen the water-balance check flag this before?