Private Well Supply and Pressure Tanks

The tank does not store water, it stores starts

A pressure tank on a private well is misunderstood the moment anyone calls it storage. Its water volume is almost irrelevant to supply; a household will drain the largest tank most contractors install in well under a minute of a running hose. What it actually holds is a delay — the interval between the pump shutting off and the pump starting again. Every gallon of drawdown buys a few more seconds of that delay, and the number of seconds is the whole design.

Pump motors fail on starts, not on hours. Inrush current at start is several times running current, and the heat from that inrush is dumped into the windings before the motor comes up to speed and flow past the motor can carry it away. A submersible motor sitting in a well is cooled by water moving up past the shell; during starting there is no flow yet. Stack enough starts close together and the winding insulation degrades, the thrust bearing takes repeated unloaded hits, and the control box or drive capacitors age at the same rate. A motor that runs continuously for hours is under far less stress than one that starts eighty times an hour.

So the working currency on a well job is cycles per hour and minimum run time per cycle. Manufacturers publish a maximum starts-per-hour figure for each motor size and a minimum run time, and those two numbers, not the tank's nameplate capacity, are what the sizing has to satisfy. Larger horsepower motors tolerate fewer starts than small ones — the trend runs the wrong way from what most people assume, because the bigger machine has more rotating mass and more inrush to dissipate. Confirm the figure for the specific motor going down the well rather than carrying a habit from the last job.

There is a second reason this framing matters on site. Almost every complaint a homeowner reports about a well — pulsing showers, a pump audibly kicking every few seconds, a pressure switch that chatters, a system that trips a breaker on hot afternoons — resolves to the same root: not enough drawdown between the cut-in and cut-out points. Diagnose in the language of cycles and the fault becomes obvious. Diagnose in the language of tank size and you end up selling a bigger tank that does not fix a waterlogged bladder or a pressure switch set two psi (0.14 bar) apart.

Drawdown is what the switch differential lets you have

Drawdown is the water that leaves the tank between cut-out pressure and cut-in pressure. Note what that definition excludes: the water still sitting in the tank at cut-in, which never participates, and the air charge, which does the pushing. A tank rated at a nominal volume delivers only a fraction of that as drawdown, and the fraction shrinks as system pressure rises. The same tank on a 40/60 setting (2.8/4.1 bar) gives noticeably more drawdown than on a 50/70 setting (3.4/4.8 bar), because the ratio between the absolute pressures at cut-in and cut-out governs how much of the air charge expands.

This catches people out on jobs with a long vertical run or a demanding fixture group, where the natural instinct is to raise both switch settings. Raising the pair without widening the gap costs drawdown and buys extra cycles. Widening the differential — moving from a 20 psi (1.4 bar) gap to a 30 psi (2.1 bar) gap — recovers it, at the cost of a more noticeable pressure swing at the fixtures. That trade sits at the centre of most commissioning arguments, and there is no universally right answer: a house with thermostatic shower valves tolerates a wide swing badly, while an irrigation-heavy property barely notices.

Pre-charge is the other half. A bladder or diaphragm tank wants its air charge set a couple of psi below cut-in, measured with the tank isolated and drained so the water side is at zero. Set it high and the tank runs out of water early — drawdown collapses and cycle rate climbs. Set it low and water sits permanently in the bottom of the tank doing nothing while the bladder never fully relaxes, which shortens its life. Charge with the pump off and the drain open, every time, and re-check pre-charge whenever the switch settings change. Adjusting a switch without re-charging the tank is one of the most common reasons a system cycles worse after a service visit than before it.

Old galvanised tanks without a bladder behave differently again. The air is in direct contact with the water and dissolves into it steadily, so the tank waterlogs and drawdown falls to almost nothing over months. That is what air volume controls and snifter valves existed to counter. If one of these is still in service, expect the cycle complaint to be seasonal and progressive rather than sudden — and expect the fix to be either restoring the air-charging arrangement or replacing with a bladder type.

What the well itself will give you

No tank arrangement can compensate for a pump that outruns the aquifer. Set a pump that delivers more than the well's sustainable yield and it draws the water level down to the pump intake, breaks suction, and either cycles on a low-pressure cut-off or runs dry and cooks the motor. The cycle count then goes vertical for a reason that has nothing to do with the tank, which is why yield gets checked before anything downstream is specified.

The number that matters is not the driller's flush-test figure taken over twenty minutes. It is the rate the well will sustain with the water level stabilised — static level, pumping level, and the drawdown in the borehole itself over a run long enough for the cone of depression to settle. A well that gives a healthy figure for half an hour and half that after four hours will be sized wrong if the short test is believed. Where the well is marginal, the answer is usually a lower-output pump plus atmospheric storage with a booster set, not a bigger pressure tank; the pressure tank cannot manufacture water the formation will not give.

Seasonal variation deserves the same scepticism. Static levels in shallow wells move with rainfall and with neighbouring abstraction, and a system commissioned in a wet spring can start short-cycling on low water in late summer. Record static and pumping levels at commissioning and leave them with the homeowner, because the comparison a year later is what separates a failing pump from a falling water table.

Sanitary construction sits alongside yield as a non-negotiable. Grout seal, casing height above finished grade, a vermin-proof and watertight well cap, and separation distances from septic fields and other contamination sources are governed by state or provincial well-construction codes and local health authorities, and the required distances differ substantially between jurisdictions — verify against the authority having jurisdiction rather than a remembered figure. NSF/ANSI/CAN 61, Drinking Water System Components – Health Effects, is the standard normally invoked for wetted materials; where a pitless adaptor is used, its listing should be confirmed rather than assumed.

Matching pump output to drawdown so run time survives

Two systems with identical tanks can cycle very differently, because run time per cycle is drawdown divided by the surplus flow — the pump's output minus whatever the house is drawing at that moment. The nastiest case is not a large draw; it is a small one. A single toilet fill valve, a slow-filling humidifier, an underground irrigation leak, or a dripping outside tap draws a trickle that empties the drawdown volume slowly, then the pump starts and refills it in seconds because nearly all its output is surplus. Run time collapses even though total water use is trivial.

Design so that the minimum run time is met at the flow that produces the shortest run — which for a fixed-speed pump falls near half the pump's rated output, where surplus flow peaks against the available drawdown. Check that condition explicitly rather than checking a full-flow case that will always pass. On oversized pumps this single check is what exposes the problem, and the correction is either more drawdown, a wider switch differential, or a smaller pump.

Constant-pressure variable-speed systems change the arithmetic rather than removing it. A drive that throttles the motor down to match demand can hold a run going for long periods instead of cycling, and its tank requirement is much smaller — the tank exists mainly to absorb the small pressure excursions the drive cannot chase quickly enough and to prevent hunting at very low draw. Still fit a tank, still set pre-charge per the drive manufacturer's instruction, which is generally referenced to the drive's set pressure rather than to a mechanical switch cut-in. Below a certain minimum demand these drives shut down and restart, so the low-flow condition remains the one worth testing.

Pump protection belongs in the same conversation because it exists to catch the failure mode cycling causes. Low-yield cut-off devices, motor overload protection, and dry-run sensing all interrupt the start sequence when conditions are wrong; verify each functions during commissioning rather than trusting that it is wired. Electrical installation is governed by the wiring code in force locally — NFPA 70, National Electrical Code, in the United States, and equivalent national codes elsewhere — and submersible motor circuits, bonding and grounding of metallic well components carry specific requirements there.

Setting the system up on the day

Sequence matters more than any single adjustment. Prove the well first: static level, a sustained draw at the intended rate, pumping level stable. Then set the pump and check amp draw against the motor's service factor rating on all legs, because a motor pulling unevenly at commissioning will not survive high cycle counts. Only after that does the system get pressurised and work on the tank begin.

Charge the tank with the system drained and the water side open to atmosphere, using an accurate low-range gauge rather than a tyre gauge that reads coarsely at these pressures. Bring the pre-charge to its target, close up, and let the pump fill. Read cut-in and cut-out at the tank tee, not at a fixture and not off the pressure switch's own scale, and adjust the switch nuts in the order the switch manufacturer specifies — on the common differential-type switch the two adjusters interact, and turning the wrong one first means chasing both settings around in circles.

Then measure drawdown for real. Close the pump's supply, open a tap, and catch the water that comes out between cut-out and cut-in in a graduated container. That measured figure is the one to record. It exposes a partially waterlogged tank, a failed bladder, a pre-charge error, and a switch differential narrower than the dial suggests, all in a single test that takes a couple of minutes. Compare it against the design drawdown, and if it falls short, fix the cause before hunting elsewhere.

Finish with a cycle count under a controlled small draw — the low-flow case described earlier — and time the run. Write down cut-in, cut-out, pre-charge, measured drawdown, run time, amp draw per leg, static and pumping level, and the date. That record is worth more to the next person on the property than any equipment list, and it turns a future callout from guesswork into a comparison.

Disinfection and water quality close out the visit. Any well opened during the work gets shock chlorinated and then flushed until the residual clears, with a bacteriological sample taken after the residual is gone rather than during it; sampling procedure and acceptance criteria come from the state or provincial health authority. Where treatment equipment sits downstream, remember it adds pressure loss that rises as media loads, which quietly eats into the pressure available at fixtures and can push a homeowner into raising switch settings — straight back into the drawdown trade already described.