Well Water Pressure Complaint Investigation

Purpose

"Low pressure" is the most mis-sold complaint on a private well, because the customer says pressure and usually means flow, and the pump is the most expensive thing in sight. A healthy pump delivers what its curve says, and the psi disappears somewhere between the tank tee and the fixture the customer is standing at, most often inside a filter housing nobody has opened in two years. This procedure measures the loss segment by segment and requires the segments to add up, so the answer is a named component rather than a hunch.

Scope

Covers investigation of a pressure or flow complaint on a private well system, residential and light commercial, from the pressure tank tee through the treatment train and distribution to the complaint fixture.

Does not cover the pump and tank verdict itself. Short cycling, a pump that will not reach cut-out and precharge diagnosis belong to the well pump service call SOP, which this procedure hands off to at step 3 rather than re-deriving. Does not cover well yield, which is a licensed well contractor's measurement, or water quality, which the water quality complaint visit SOP owns.

Roles and responsibilities

Procedure

1. Turn the complaint into measurable terms and reproduce it before touching anything. Name the fixtures, the time of day, whether it is hot only or both, and what else runs at the same time, then run that combination. Acceptance: fixtures named, hot or cold or both, the simultaneous-use condition stated, and the symptom either reproduced in front of you or recorded as not reproducible today. Wrong looks like starting at the pump because the customer said pressure. Stop rule: a symptom that will not reproduce gets a return visit at the stated time of day rather than a part fitted on description. Hazard: none created here, it is a conversation and a fixture opened, but run cold rather than hot so nobody is standing in a scald stream while watching a gauge.

2. Separate pressure from flow at the hose bib nearest the tank. Thread a gauge on the bib, read it static with everything closed, read it again with one fixture open, and time a container fill at that same bib. Acceptance: a static psi, a running psi and a timed fill converted to gallons per minute, all three written down. Wrong looks like calling a weak shower low pressure when the gauge holds 55 psi and the container says the system passes only a few gpm, which is a flow restriction and points somewhere else entirely. Stop rule: static pressure below the switch's cut-in with the pump idle means the pump or tank half has already failed, so go to step 3 and stop there. Hazard: the bib is under system pressure, so the gauge is threaded fully before the valve is cracked and the bib is opened slowly, because a partly engaged gauge launches off the threads.

3. Confirm the pump and tank are healthy before you walk downstream of them. Observe one full cycle hands off and record cut-in, cut-out and the run time in seconds. Acceptance: an observed cut-in and cut-out in psi and a timed run of at least about a minute per start under a steady draw. Wrong looks like reading 40/60 off the switch label when a previous tech may have turned the range nut. Stop rule: short cycling, a pump that never reaches cut-out, or a run under about a minute hands the job to the well pump service call SOP now, since no downstream measurement means anything on an unstable supply. Hazard: nothing is opened here, so no cover comes off the switch and no reading is taken standing in water.

4. Measure the drop across each treatment device at one stated flow, not at rest. Open a fixture to a fixed flow, then read pressure immediately upstream and downstream of the sediment filter, the softener, the carbon unit and the ultraviolet unit in turn. Acceptance: a psi drop for each device at the same stated flow, each set against the manufacturer's published clean drop at that flow. Wrong looks like taking the readings with nothing running, which shows zero drop across a completely blocked filter because a static column has no loss. Stop rule: a device measuring more than roughly twice its published clean drop is the finding, recorded before anything is changed, since a cartridge swapped before it was measured cannot be proven the cause. Hazard: filter and softener housings hold system pressure, so isolate and relieve through the housing's relief button before a housing is unthreaded, and never look at an energized ultraviolet lamp - it is an eye injury, so the unit is de-energized and left to cool before the sleeve is handled.

5. Measure at the complaint fixture and convert its elevation into psi before you blame anything. Read pressure at the complaint fixture with the same flow running, and measure its height above the gauge point used in step 2. Acceptance: a running psi at the fixture, the height in feet, and that height converted at 0.433 psi per foot for cold water at ordinary temperature. Wrong looks like blaming the pump for an upstairs complaint when two storeys of lift account for most of the difference. Stop rule: a fixture whose reading is explained by elevation and normal pipe friction alone is an expectations conversation, not a repair. Hazard: gauging at a shower or tub means working off a wet edge, so use a stable stand rather than the tub rim, and keep the test on cold so nobody is reaching through hot water.

6. Gauge the intermediate points before you accept any total. Take readings across the tank tee and its nipples, across every isolation valve between tank and fixture, and at the transition into any older piping. Acceptance: a psi drop at the stated flow across each valve and transition, or a written note that a segment could not be gauged and why. Wrong looks like assuming a full-port valve is open because its handle looks straight. Stop rule: a segment you cannot gauge is written down as ungauged rather than assumed at zero, because that assumption is exactly what makes the sum in step 7 close falsely. Hazard: an old galvanized nipple at the tank tee shears rather than turns, so the system is isolated and relieved to zero before a wrench goes on one, and the tank is supported so it cannot drop when the nipple lets go.

7. Close the arithmetic: the measured drops have to sum to the measured loss. Subtract every measured device drop, valve drop and the elevation conversion from the step 2 running pressure, and compare the result with the pressure measured at the fixture. Acceptance: predicted and measured agreeing within 5 psi, with the largest single loss named. Wrong looks like changing the most convenient part when the sum was never run, which is how a shop replaces a softener costing single-digit psi. Stop rule: a gap wider than 5 psi means a segment was never measured, so return to step 6 and gauge until it closes; nothing is corrected on an open sum. Hazard: none physical, this is arithmetic done with the system running normally, and the cost of getting it wrong is a part sold that changes nothing.

8. Correct the largest measured item, re-measure at the same points at the same flow, and put every disturbed device back in service. Change or repair the item step 7 named, then repeat steps 4 through 7 unchanged. Acceptance: the corrected device within its published clean drop, fixture pressure improved by close to the psi removed, the sum closing again within 5 psi, and every housing, bypass and lamp back in normal service. Wrong looks like a softener left in bypass or an ultraviolet unit left unpowered after the pressure test, which trades a pressure complaint for an untreated water complaint. Stop rule: a fixture that does not gain close to the removed loss means a second restriction, so the sum runs again before the ticket closes. Hazard: this is the step that puts pressure back, so a filter housing is refilled slowly with the relief button open until water appears, hands and face are clear of the housing while it comes up to pressure, and the ultraviolet unit is confirmed lamp-on and in alarm-free service, because that is the protective function this procedure took offline.

The record this produces

Ten fields filed against the property: the reproduced complaint and its simultaneous-use condition; static and running psi at the tank tee; the timed flow there in gpm; observed cut-in, cut-out and run time; a drop per treatment device at a stated flow with its published clean drop beside it; drops across valves and transitions; fixture elevation in feet and its psi conversion; measured pressure at the complaint fixture; the sum check and its gap; and the post-repair repeat of the same set.

The next tech reads the per-device drops against their own, which turns a filter change from a calendar guess into a measured interval for that household's water. The office reads the sum check when a customer disputes the repair, because a documented predicted-versus-measured agreement beats a technician's recollection. The lead reads a distribution loss growing across visits as early evidence for a repipe conversation rather than another cartridge.

One worked pass, including a failure

Two storey house on a private well, softener, carbon filter, ultraviolet unit and a sediment filter ahead of them. Complaint: the upstairs shower goes weak while the washer runs.

The combination reproduces on the first try. At the hose bib nearest the tank, static reads 58 psi and running with the test fixture open reads 52 psi, and a container fill gives 5 gallons in 60 seconds, so 5.0 gpm is the flow every later reading is taken at. A cycle observed hands off gives cut-in 40, cut-out 60 and a run of 95 seconds, over the roughly one minute floor, so the pump and tank half is left alone.

At 5.0 gpm the drops come in as follows: sediment filter 22 psi against a published clean drop of 3, softener 8 psi against a published 8, carbon 4 psi against a published 5. The upstairs shower stands 9 feet above the gauge point, and 9 feet at 0.433 psi per foot is 3.9 psi.

The sum fails on the first pass. Total measured device drop is 22 plus 8 plus 4, or 34 psi. From the 52 psi running at the tank tee, 52 minus 34 leaves 18, and 18 minus the 3.9 psi of elevation predicts 14.1 psi at the shower. The shower actually reads 6 psi running. The gap is 14.1 minus 6, or 8.1 psi, well outside the 5 psi closure gate, so the stop rule fires: nothing is changed and step 6 is repeated with more gauge points.

The missing segment is a globe valve on the branch feeding the upstairs bath, left roughly three quarters closed after an earlier repair, measuring 8 psi on its own at 5.0 gpm. Opened fully, the shower reads 13 psi, and against the 14.1 psi prediction that is a gap of 1.1 psi, inside the gate, so the sum now closes. The sediment filter is still the largest single item at 22 psi against a published 3, so the cartridge is replaced, which removes 22 minus 3, or 19 psi. Predicted at the shower is 13 plus 19, or 32 psi, and the re-measurement reads 32. The housing is refilled with the relief button open until water shows, the softener is confirmed out of bypass and the ultraviolet lamp confirmed lit before the ticket closes.

References

• Filter, softener, carbon and ultraviolet manufacturer literature for the published clean pressure drop at a stated flow, which is the only benchmark that makes a measured drop meaningful in step 4.
• Pump and tank manufacturer literature for delivery in gpm at the operating pressure, used to sanity check the step 2 flow reading.
• See related: the Plumbing well pump service call SOP, which owns the pump and tank verdict this procedure hands off to at step 3, the well pressure tank replacement and setup SOP, and the water quality complaint visit SOP.