A well that used to make 1,000 gpm is making 900. Somebody has to decide what to do about it, and the decision is expensive whichever way it goes.
There are three usual reasons a well makes less water: the aquifer has dropped, the well itself is plugging, or the pump is wearing out. Each one has a different fix and a different price. Rehabilitate a well whose real problem is a worn pump and you spend a six-figure sum and get your flow back for about a week — until the pump is still worn. Put a new pump in a well that is plugging and the new pump runs off its curve from the day it is installed. Do either on a well where the water table has fallen and you have spent the money on the one thing that was not broken.
The good news is that the three leave different fingerprints, and you can read them from four numbers most districts already collect.
The four readings
For any test you want to compare against the well’s baseline, you need:
- Static water level — the level with the pump off and the well recovered.
- Pumping water level, with the flow the well was actually making at the time, from a meter.
- Discharge pressure at the wellhead.
- The pump’s published curve, so you can see what head it should make at the flow it is making.
From the first two you get drawdown (pumping level minus static) and specific capacity (flow divided by drawdown). From the first three you get total dynamic head — pumping level, plus discharge pressure in feet (psi × 2.31), plus the friction loss in the column. The fourth is on the pump submittal, and how to read a pump curve covers it.
Three wells that each lost 10%
Here are three identical wells. Same construction, same pump, same 60 psi discharge pressure. At baseline each made 1,000 gpm with 50 feet of drawdown. Today each makes about 900 gpm. Only one number in each column tells you why.
| Reading | Baseline | Well A | Well B | Well C |
|---|---|---|---|---|
| Flow | 1,000 gpm | 896 gpm | 905 gpm | 901 gpm |
| Static water level | 180 ft | 215 ft | 182 ft | 181 ft |
| Pumping water level | 230 ft | 260 ft | 257 ft | 226 ft |
| Drawdown | 50 ft | 45 ft | 75 ft | 45 ft |
| Specific capacity | 20.0 gpm/ft | 20.0 | 12.0 | 20.0 |
| Total head, measured | 379 ft | 406 ft | 404 ft | 373 ft |
| Head the curve promises at that flow | 379 ft | 406 ft | 404 ft | 405 ft |
| What changed | — | The aquifer | The well | The pump |
Well A — the aquifer. The static level has fallen 35 feet. Specific capacity has not moved: the well gives up exactly as much water per foot of drawdown as it did when it was new. The pump is fine too — it is making exactly the head its curve promises. It is simply lifting water from 35 feet deeper, and a centrifugal pump that has to lift more makes less. Nothing about the well or the pump is broken.
Well B — the well. The static level is where it always was. But the well now needs 75 feet of drawdown to make what 50 feet used to — specific capacity is down 40%. That is the screen, the gravel pack or the formation right around the borehole plugging up, and the pump is doing its honest best against the extra lift the plugging forces on it. The pump is still on its curve.
Well C — the pump. Static level normal. Specific capacity normal — the well is as clean as the day it was drilled. But at 901 gpm the pump’s curve says it should make 405 feet of head, and it is making 373. It is 8% short of its own curve. That is worn impellers and bowls, opened-up wear rings, or lineshaft and bearing trouble. A pump efficiency test on Well C would also show wire-to-water efficiency well down from baseline.
We have kept drawdown proportional to flow to keep the arithmetic readable. A real well’s specific capacity rises a little at lower flow, because the part of drawdown lost at the well face grows with the square of the rate. That makes a falling specific capacity at a lower rate — Well B — even more telling, not less.
The order to check them in
Work through the readings in this order. Each step either explains the loss or passes it on to the next.
- Check the system first. Has the discharge pressure changed? A higher zone pressure, a valve left part-closed or a new connection downstream raises the head the pump works against and costs flow with nothing wrong in the ground at all. On a drive, check the speed too — head moves with the square of speed, so a pump tested at 57 Hz and compared against a 60 Hz curve looks 10% worn when it is not. Correct for it with the affinity laws before comparing anything.
- Look at the static level. Compare it with the same season in earlier years, and with neighboring wells and monitoring wells — DWR’s California’s Groundwater Live maps current groundwater levels across the state and how they have changed over 1, 3, 5, 10 and 20 years. If your well’s static is falling and so is everyone else’s, it is the aquifer.
- Compare specific capacity at the same rate and the same pumping time. A decline of 15% to 25% from the well’s baseline is the range most practice uses to schedule rehabilitation. Comparisons only mean something when the rate and duration match, so pull the test that matches best, not the most recent one.
- Put the test point on the pump curve. Measured head at the measured flow, corrected to the same speed, against the curve. A pump more than a few percent below its curve across more than one test is worn.
You will sometimes see a failing pump listed among the causes of falling specific capacity. At the same flow it is not one. A weaker pump makes less water and draws the well down less, and specific capacity stays where it was or edges up. That is exactly why specific capacity is the right number for judging the well: the pump cannot fake it.
When it is more than one
Real wells rarely fail one way at a time. A twenty-year-old well in a declining basin can have all three going on at once. The same four readings split the loss between them:
- the static change tells you how much of the lost flow is the aquifer,
- the specific capacity change tells you how much is the well,
- the gap to the pump curve tells you how much is the pump.
That split is what lets a district spend in the right order. If most of the loss is the well, rehabilitate it and consider pulling the pump in the same mobilization, since the rig is already there. If most of it is the pump, the well can wait. If most of it is the aquifer, neither fix brings the water back, and the conversation is about setting depth, bowl selection and how hard the basin can be pumped.
Three causes, three price tags
Order of magnitude, for a municipal well in Southern California, from our own scopes:
- The aquifer: no rehabilitation fixes it. The options are lowering the pump setting if there is room above the screen, re-bowling the pump for the new head, pumping less, or leaning on other wells. Some of these cost very little; the expensive mistake is rehabilitating a well that did not need it.
- The well: a full-scope rehabilitation — video survey, brushing, chemical treatment, swabbing and airlifting, disinfection, disposal, start-up and before-and-after testing — runs on the order of $165,000.
- The pump: a pump replacement runs on the order of $175,000. A rebuild can cost less when the column and shaft are sound.
Two of those three are big enough to need a board item. Getting the diagnosis wrong does not waste a little money. It wastes the whole job.
What confirms it
The readings indicate. A cause is confirmed by evidence that looks at the thing itself:
- Aquifer: the static trend matches the monitoring network and the neighbors.
- Well: a downhole video survey showing encrustation or biofouling on the screen, a step-drawdown test showing well loss up since the last one, or water chemistry and bacteria results pointing the same way. See when to rehabilitate a water well for what waiting costs.
- Pump: a pump efficiency test showing the head and efficiency shortfall, and in the end the teardown, which shows the wear.
What to do with this on Monday
- Find each well’s baseline — as-built or post-rehabilitation test: flow, static, pumping level, discharge pressure.
- Take the four readings on any well that has lost flow before anybody writes a scope for it.
- Check discharge pressure and drive speed first. The cheapest diagnosis is the one where nothing is wrong.
- Compare static with the neighbors, specific capacity with the baseline at the same rate, and the test point with the pump curve.
- Write down which fingerprint you found and what would confirm it, then order that test — not the fix.
Quick answers
Why is my well producing less water?
Usually one of three things: the water table has dropped, the well screen and gravel pack are plugging, or the pump is worn. A falling static level points to the aquifer; a falling specific capacity at the same rate points to the well; a test point below the pump curve points to the pump.
Does a worn pump lower specific capacity?
Not at the same flow. A worn pump makes less water and draws the well down less, so specific capacity holds steady or rises slightly. A falling specific capacity at the same rate is the well.
How much does specific capacity have to drop before a well needs rehabilitation?
Common practice schedules rehabilitation when specific capacity has fallen 15% to 25% from the well’s baseline, compared at the same pumping rate and duration.
See the aquifer, well and pump separated on one of your own wells. Request a walkthrough →
Sources
- Water Replenishment District of Southern California. Specific Capacity — A Measure of Well Performance, Well Problems, and Aquifer Transmissivity, Technical Bulletin TB02. Source for specific capacity declining as perforations or filter pack plug, the need to use the same time frame for every comparison, and the 25% decline trigger attributed to Driscoll (1986).
- Southeast Hydrogeology. Using Specific Capacity to Monitor Well Performance. Source for physical, chemical and biological plugging as the causes of well-side decline, and a 15% decline as a rehabilitation trigger.
- California Department of Water Resources. California’s Groundwater Live. Source for statewide groundwater level conditions and their 1- to 20-year changes.
- Driscoll, F. G. Groundwater and Wells, 2nd edition. Johnson Division, 1986. Source for well loss and aquifer loss as the two parts of drawdown, and for specific capacity as the measure of well condition.
- Hydraulic Institute. ANSI/HI 14.6, Rotodynamic Pumps for Hydraulic Performance Acceptance Tests. Source for comparing a measured head and flow against the pump’s rated curve at the same speed.
Every link above was retrieved and read on 2026-09-25; the California’s Groundwater Live link was replaced and read on 2026-09-30, after the Department of Water Resources moved the site. Paid standards are named by designation so you can check them against your own copy.