Results
Data & QC
Diagnostic
Semilog
Model matching
Deliverability
Gauge record & detected structure
QC never edits your data — spikes, frozen intervals
and gaps are flagged and excluded from analysis, and the clock
offset between gauge and reported schedule is estimated from the
pressure breakpoints themselves.
Log-log diagnostic — Δp / Δm(p) and Bourdet derivative
Shaded bands are the detected flow regimes; dotted
guide lines show the reference slopes (+1 storage, +½ linear,
+¼ bilinear, 0 radial, −½ spherical) anchored on
the identified segments, so the focus region of the analysis is
visible at a glance.
Superposition semilog
Auto window maximises r² past the wellbore-storage
guard (1.5 log cycles) — the objective criterion validated against
published worked examples. Manual override: type a window and re-fit.
Candidate models
Ranking — statistical support, not truth
ΔAICc < 2 between models means this dataset
cannot discriminate them — geology, seismic and offset wells must
arbitrate. Parameter intervals are ±1σ (linearised);
correlated parameters carry honestly wide bars.
Inflow (LIT) from PTA results
AOF is an extrapolation of the fitted inflow
relation far beyond the tested rate window — quote it with the
uncertainty of the match, never as a standalone certainty.
VLP curve CSV
(optional) — q_mscfd, pwf_psia → operating point
Volumetrics
Any undetected boundary side must be entered
as a RANGE (re-run with low/high), never a single number. p/Z EUR
is volumetric depletion only — invalid under water drive.
Vertical lift performance
Two independent sources. Measured comes from the
test itself — each stabilised step gives (q, pwh, pwf),
so the tubing drop is observed, not modelled; it needs
p_wh in the rate schedule. Correlation works at any
THP but only as well as its assumptions. Match the correlation against
the measured points before trusting it away from the tested rates.