Flushometer Rebuild and Adjustment

Purpose

A flushometer rebuild is not finished when the valve stops leaking. It is finished when the valve delivers the volume the fixture was designed around and clears the bowl in one flush, both measured. The failure this prevents is the rebuild that ends at "no drip": a valve throttled down at the control stop until it stops running will also stop clearing the bowl, and the shop gets called back in a month for a plugged closet that is really a starved flush. The other failure it prevents is the opposite one, a valve quietly passing water past close, which is the largest continuous water loss in most small commercial buildings and is inaudible in an occupied restroom.

Scope

Covers manual and sensor-operated diaphragm and piston flushometers on water closets and urinals in offices, small retail, small restaurants and apartment common areas, including kit replacement, seat inspection, control stop adjustment and delivered-volume verification.

Does not cover control stop or valve body replacement, which takes the branch down and routes to the shared riser shutdown and tenant notification SOP. Does not cover fixture or carrier work, owned by the wall-hung fixture and carrier SOP. Kit part numbers and model families for one manufacturer are covered in that manufacturer's own rebuild article; this procedure is the shop standard that applies whatever make is on the wall.

Roles and responsibilities

Procedure

1. Watch the valve through a full cycle before you close anything, and classify the failure. Flush it and stay with it for three minutes. Acceptance: the fault named as one of runs continuously, short flush that leaves the bowl uncleared, will not start, or holds a trickle after the refill ends. Wrong looks like starting a rebuild on the customer's description, which uses one word for four different faults. Stop rule: a valve that behaves correctly through three cycles gets no kit; find the fault or leave it alone and write down what you observed. Hazard: standing over a cycling closet puts your face above an aerosol from the bowl, so stand back from the rim and flush with the lid down where there is one.

2. Read the gpf stamped on the valve, the diaphragm and the fixture, and record all three. Acceptance: three numbers written down, and any mismatch called out. Wrong looks like assuming the fixture's rating tells you what the valve delivers; a 1.6 gpf kit sitting in a valve on a 1.28 gpf bowl over-flushes forever, and a low-volume kit in a valve on an older bowl leaves the trapway uncleared. Stop rule: where the three disagree, the fixture's rating governs the kit you fit, and the mismatch goes on the record so the next tech does not undo it. Hazard: none at this step, it is reading castings and labels, but getting it wrong is how a shop rebuilds the same valve twice.

3. Rule the fixture out before you rule the valve in. Look into the trapway with a mirror or a light and check for an obstruction, and check the bowl rim holes for scale. Acceptance: a clear trapway and open rim holes confirmed by eye. Wrong looks like a rebuilt valve on a bowl with a mop head in the trap, which reproduces every symptom of a weak flush. Stop rule: an obstructed fixture is cleared or replaced first, and the valve is not opened until the bowl is proven clear. Hazard: never reach into a trapway by hand and never mix an acid descaler with a chlorine-bearing bowl cleaner, because that combination liberates chlorine gas in an enclosed stall; ventilate the room and use one product at a time per its own label.

4. Measure the as-found delivered volume at the meter before the valve is opened. With every other fixture in the building closed, read the meter, run ten complete flushes, read it again and divide by ten. Acceptance: an as-found gallons per flush figure recorded beside the fixture's rated gpf from step 2. Wrong looks like skipping the baseline because the fault is already obvious, which leaves nobody able to say afterwards what the rebuild actually changed. Stop rule: a valve that never stops running cannot be measured in ten flushes, so record the baseline as not obtainable with that reason and carry on to step 5 rather than inventing a figure. Hazard: this holds the building's other fixtures closed, so tell the tenant before you do it and never close a valve serving a kitchen or a boiler make-up without asking who depends on it.

5. Close the control stop, prove it holds, then relieve the valve. Close the stop with the correct key or driver, then operate the handle once to drop residual upper-chamber pressure. Acceptance: the stop closed, the flush producing nothing, and no flow at the spud for a full minute. Wrong looks like a stop that closes to a dribble because its own seal is gone. Stop rule: a stop that will not hold means the isolation you need is upstream, and that is a branch shutdown, not a field workaround; route it and reschedule rather than opening the cover against live pressure. Hazard: the cover of a live flushometer carries full line pressure, so the stop is proven and the valve is relieved before a wrench touches the cover, and you stand to the side of it either way.

6. Remove the cover and inspect the seat before you reach for the kit. Unthread the outer and inner covers with a smooth-jawed wrench, catch the residual water, lift the old assembly out. Acceptance: a seat that is smooth and unpitted to a fingernail drawn across it. Wrong looks like a pitted or scaled seat, which no new diaphragm will seal against and which sends the valve straight back to running. Stop rule: a pitted seat takes a seat repair kit where the make offers one, and a corroded or eroded body is a valve replacement quoted through the lead, not a second diaphragm. Hazard: this water is fixture-side and on a urinal it carries uric scale, so wear gloves and eye protection, and never scrape or wire-brush dry scale in a closed stall; wet it first and bag the debris.

7. Fit the kit dry, with the bypass orifice oriented as the manufacturer shows. Wipe the seat with a non-abrasive pad, seat the new assembly, thread the inner then the outer cover. Acceptance: covers hand tight plus the fraction of a turn the manufacturer's instruction states, and no gasket pinched out of its groove where you can see it. Wrong looks like a diaphragm fitted backwards, which will not seat, or a cover cranked with a pipe wrench, which cracks it and marks a chrome finish the tenant will notice. Stop rule: if the cover will not run down by hand, back it off and re-seat rather than forcing it. Hazard: do not sand or file the seat to clean it, because the scratch pattern leaves a leak path that no kit will ever close.

8. Open the control stop in stages and set delivered volume at the stop, not at the handle. Crack the stop, let the upper chamber fill, then open it and cycle the valve. Acceptance: a full open cycle with no hiss at rest, then the stop backed toward closed only far enough that the bowl still clears, per the make's own adjustment instruction. Wrong looks like the stop used as a cure for a valve that is still passing water, which hides the fault behind a starved flush. Stop rule: if the valve hisses at rest after the chamber fills, close the stop and re-seat the diaphragm before adjusting anything. Hazard: this is the step that puts line pressure back with your hands on the valve, so open the stop in stages with your body clear of the cover, and keep the key on the stop until you have watched a full cycle.

9. Re-measure delivered volume the same way and set it against the step 4 baseline. Repeat the ten-flush meter measurement with the building's other fixtures closed. Acceptance: measured gallons per flush within a shop default of 10 percent of the fixture's rated gpf, tune the band to your own accounts. Wrong looks like calling it good on the sound of the flush, which cannot distinguish a 28 percent over-flush from a correct one. Stop rule: outside the band, the kit is the wrong rating or the stop is misadjusted; correct the cause rather than splitting the difference at the stop. Hazard: the other fixtures go closed a second time here, so re-confirm with the tenant before this run rather than assuming the first agreement still covers it.

10. Verify bowl clearance and a still meter, then restore and clean up. Run the paper test: a full sheet of tissue clears the trapway in one flush. Acceptance: the bowl clears on three consecutive flushes, and with the fixture at rest for three minutes the meter's low-flow indicator does not move. Wrong looks like one good flush signed off, when the fault was intermittent to begin with. Stop rule: a moving indicator at rest means the valve is still passing and the job is not done, whatever the volume test said. Hazard: leave the valve pressurized and watched for a few minutes after the last flush, since a cover gasket that will weep does it then, and dry the floor before you leave the stall open to the public.

The record this produces

Per valve: fixture location tag, valve make and model, the three gpf readings from step 2 and any mismatch, the seat condition, the kit fitted, measured gallons per flush before and after, the bowl clearance result, and the still-meter result. Where the stop was found holding or not holding, that goes down too.

These land on the fixture's line in the property record. The next tech reads the measured gpf to see whether this valve has been rebuilt into the wrong volume twice. The account owner reads the seat condition, because a building with three pitted seats is a body-replacement conversation, not three more kits. The property manager reads the before-and-after volume when they ask what the maintenance contract did about the water bill, and that is the only number that answers it.

One worked pass, including a failure

Small office, four-stall restroom, ticket says one closet flushes twice on its own. Step 1 finds the valve holding a trickle for the full three minutes after the refill ends, so the fault is passing past close.

Step 2 fails and takes its stop rule. The casting reads 1.6 gpf, the old diaphragm 1.6, and the bowl 1.28 gpf. The three disagree, so the fixture's rating governs the kit that gets fitted: this valve takes a 1.28 kit, and the mismatch goes on the record so the next tech does not undo it. The bowl is a later low-consumption fixture the valve was rebuilt past at some point.

Step 3 finds the trapway clear. Step 4 takes the baseline before anything is opened: with the other fixtures held closed, the meter reads 16.4 gal across ten flushes, so 16.4 divided by 10 is 1.64 gpf against a rated 1.28. The difference is 1.64 minus 1.28, or 0.36 gpf, and 0.36 divided by 1.28 is 28 percent over. That is the figure the previous rebuild's throttled stop was hiding.

Step 5 closes the stop, which holds cleanly. Step 6 finds the seat smooth. Step 7 fits the 1.28 kit. Step 8 opens the stop in stages and sets it.

Step 9 re-measures the same way: 13.1 gal across ten flushes, so 1.31 gpf. That is 0.03 over 1.28, and 0.03 divided by 1.28 is 2.3 percent, inside the 10 percent band, so step 9 passes on the post-rebuild number. Step 10 clears a full sheet of tissue on all three flushes, and the meter's low-flow indicator is still after three minutes at rest. The record carries 1.64 gpf before, 1.31 gpf after, the casting-versus-fixture mismatch, and a note that this valve must not be rebuilt with a 1.6 kit again.

References

• Flush valve manufacturer's rebuild and adjustment instruction for the make on the wall, which is the source of the kit rating, the cover tightening figure and the bypass orientation.
• Fixture manufacturer's data for the bowl's rated gallons per flush, which governs the kit when the valve casting disagrees.
• IPC Section 604.4 for maximum water closet and urinal flush volumes as adopted by your authority having jurisdiction, in the edition it enforces.
• See related: the Plumbing commercial restroom preventive visit SOP, which flags valves into this one, and the shared riser shutdown SOP when the control stop itself has to come out.