Static bottomhole pressure at multiple time points. Surface pressures don't work — wellbore hydrostatics corrupt the signal. Drop suspect points rather than include them.
| Time (yr) | P (psi) | Z | P/Z | Bg | Gp (Bcf) | Wp (MMbbl) |
|---|
| Z @ Pab | — |
| Pab/Zab | — psia |
| Bg @ Pab | — rcf/scf |
When P/Z curves upward, the Cole plot quantifies the aquifer. We model We with Fetkovich finite-aquifer formulation. Iterate J and Wei until F/Eg vs We/Eg is linear with slope ≈ 1.
For abnormally pressured reservoirs (initial gradient > 0.6 psi/ft) rock and connate-water expansion contribute materially to reservoir voidage. The naive P/Z plot overstates OGIP. Roach correction (P/Z)·[1 − ce·ΔP] with ce = (cf + cw·Swi)/(1 − Swi) restores linearity.
Havlena-Odeh linearization: F = G·Eg + We. Plot F (= Gp·Bg + Wp·Bw) versus Eg (= Bg − Bgi). Volumetric: line through origin with slope = G. Curvature → water influx or compartmentalization.
Compute OGIP from each individual pressure point (one-point material balance), plot vs time. Stable horizontal → correctly characterized reservoir. Upward drift → unmodeled aquifer / connected reservoir. Downward drift → missing production / compartmentalization.