Pumping Test Interpretation in a Layered Sand-and-Gravel Aquifer
A municipal well is completed in a confined sand-and-gravel aquifer 32 m thick, overlain by 12 m of clay. A 72-hour constant-rate pumping test was run at 1,800 m3/day with two observation wells at 25 m and 120 m. Drawdown in the near well at 25 m shows a steep early segment, a flattening between 10 and 900 minutes, and a renewed downward trend after 1,200 minutes. The far well at 120 m shows a smooth, slowly rising drawdown curve with no inflection. Barometric pressure fell steadily by 6 hPa over the test, and the pumped well penetrates only the upper 60 percent of the aquifer thickness. The regulator asks whether the test supports a defensible transmissivity for a permit renewal.
1. Interpret the three-segment shape of the near-well curve. What mechanisms plausibly explain the flattening and the late renewed decline, and how should each be handled in analysis?
Read the worked answer
The steep early segment reflects wellbore storage and possibly a low-permeability skin; early data should be excluded from a Theis or Cooper-Jacob fit until wellbore storage effects die out (roughly when the log-log derivative flattens).
The mid-test flattening between 10 and 900 minutes is consistent with a higher-transmissivity layer contributing early, then a lower-K layer dominating as leakage or vertical flow develops. A bounded or leaky response, or partial penetration effects, can produce similar flattening; the far-well curve, smooth and without inflection, argues against a simple no-flow boundary at 120 m.
The renewed decline after 1,200 minutes is most plausibly a boundary effect (recharge or a lower-K unit being drained) or a gradual decline in pumping efficiency. Barometric efficiency of a confined aquifer means the 6 hPa fall (~0.6 m of water) adds apparent drawdown; correct the record using barometric efficiency before interpreting the late trend.
Practical handling: fit the middle-time straight line on a semi-log plot for transmissivity, use the far well to confirm, treat the late segment as boundary-affected and do not use it for storativity, and flag partial penetration as a source of vertical-flow error requiring a correction or numerical model.
2. The pumped well penetrates only the upper 60 percent of the aquifer. How does partial penetration bias the near-well drawdown, and what does this mean for the permit transmissivity?
Read the worked answer
Partial penetration concentrates flow near the open interval, producing additional drawdown from vertical convergence that is not captured by a two-dimensional radial solution. The effect is strongest near the pumped well and diminishes with distance; this is consistent with the steep near-well curve and the smooth far-well response.
If the near-well data are fit without correction, the apparent transmissivity will be biased low because the extra convergence drawdown is attributed to a lower K. The far well, at 120 m and beyond most of the convergence zone, gives the more reliable estimate.
For a permit, report transmissivity from the far well (or from a partial-penetration-corrected near-well analysis), state the uncertainty range, and note that the near-well fit is a lower bound. A defensible permit value should use the far-well estimate with an explicit caveat about vertical anisotropy and partial penetration.