Seeing a notable gap between EDFM-modeled SRV and fiber DAS on a March 2025 Delaware Basin Wolfcamp A zipper pad: the dual-perm compositional model (GEM + EDFM) matches early rate but overpredicts 12‑month oil by about 25% unless we relax cluster efficiency and add stronger near‑wellbore tortuosity. Curious if others are updating frac‑hit transmissibility multipliers or parent depletion maps to close this, and what that’s doing to your 2‑year EUR forecasts.
We saw the same on a 2024 Wolfcamp A zipper; “relax cluster efficiency” helped, but switching to a time‑decaying frac‑hit transmissibility and adding pressure‑dependent near‑wellbore conductivity trimmed the 12‑mo overcall without killing the early rate — otherwise the model was still drinking through a firehose. Small caveat: constrain EDFM cells to fiber‑implied half‑length/height or you’ll fake SRV growth; did you tie the decay to local dp/dt?
Overpredicting 12‑mo oil by about 25% on a March ’25 Wolfcamp A drove us nuts too; in GEM+EDFM we fixed it by tying fracture conductivity to effective stress (Cf(σ) with proppant compaction) and making EDFM transmissibility decay with Δσ_n, not just time — brought us within about 5% without nuking cluster efficiency. We also used an asymmetric frac‑hit multiplier toward the nearest parent off the depletion map, @ian_swift99; small caveat: you need a decent closure‑stress/DFIT trend or you’ll over‑penalize late time.
EDFM ran hot for us on a March ’25 Delaware Wolfcamp A too; the fix wasn’t more transmissibility tweaks but making the matrix–fracture exchange coefficient time‑dependent (“diffusion‑limited” shape factor) with cap‑pressure/kr hysteresis during cleanup. That kept the early GEM+EDFM rate match to fiber DAS but pulled 12‑mo oil down about 20–25% without touching efficiency or added tortuosity. If you want a simple knob, let transfer ramp with sqrt(t) instead of staying constant, @n_greenleaf78.