pumptest

Pumping test analysis in Python. Transmissivity, storativity, leakage, specific yield, skin and well losses from drawdown data, by the classic analytical solutions, with automatic fitting across any number of wells.

What it does

Solutions	Theis, Hantush-Jacob (leaky), Papadopulos-Cooper (large-diameter wells, with optional skin and fitted casing radius), Neuman (unconfined, delayed yield)
Straight-line methods	Cooper-Jacob, Theis recovery, step-drawdown well losses
Test types	constant rate, recovery, and any variable-rate schedule, by superposition
Fitting	least squares over every observation at once, with standard errors, correlations, RMSE and AICc for choosing between models
Diagnostics	the Bourdet log-derivative
Example data	seven published tests bundled in pumptest.datasets, with the published estimates for each
Dependencies	numpy and scipy only

How far to trust it

On three tests from the AQTESOLV manual and four open pumping tests analysed with welltestpy, the fitted parameters match the published ones, and on the same data the fit is at least as good. The well functions are checked against independent evaluations of their published formulas. Every number on the validation page is regenerated from the code, and a test fails if the page and the code disagree.

Thirty seconds

from pumptest import Theis, datasets, fit

test = datasets.load("sioux_flats")  # three observation wells, from the AQTESOLV manual
result = fit(test, Theis())
print(result.summary())

Next: Getting started.

Licence

MIT. Copyright 2026 Piezoline Technologies Private Limited. The bundled example data come from the TTim documentation (MIT) and from the Lauswiesen research site (CC BY-SA 4.0); see Citing.