Technical evidence

Why “Just Swap the Kit” Doesn't Deliver Water Savings

A valve-kit-only retrofit can look identical to a real water efficiency project on paper, and still fail on flush performance, on the M&V report, and over the life of the contract.

1. The approach: drop a lower-flow kit into the existing bowl

A common lower-cost retrofit approach treats water conservation as a parts swap: a 1.6-gpf toilet gets a 1.28-gpf diaphragm kit, a 3.5-gpf toilet gets a 1.6-gpf kit, and the fixture is logged as converted. It's fast, it's cheap, and on a spec sheet it looks identical to a genuine efficiency upgrade; the same gpf number appears in the project close-out documents either way. The problem is what that number doesn't capture: whether the bowl the kit was dropped into can actually turn that reduced water volume into a complete flush.

2. The bowl was never redesigned, only the valve was

The historical record on low-flow toilets is consistent on exactly this point: the water volume and the fixture geometry have to be engineered together, or performance collapses. When EPAct '92 first forced the market to 1.6 gpf, manufacturers that simply shrank the water volume without redesigning the bowl and trapway produced the weak, double-flushing fixtures that gave low-flow toilets their original bad reputation. The volume changed; the bowl didn't. The same lesson repeated when the industry pushed to 1.28 gpf: EPA WaterSense's own toilet specification documentation identifies a flushometer-valve toilet's flush valve seat diameter, flapper geometry, and overflow tube height as interdependent, meaning the parts and the bowl are calibrated to each other as a matched system, not interchangeable components.

Getting a 1.28-gpf fixture to perform reliably took manufacturers years of real engineering work, not a lower-flow valve dropped into an existing design. Kohler's AquaPiston canister valve, TOTO's Tornado cyclonic rim, and comparable siphonic-jet and fully-glazed-trapway designs were developed specifically to convert a reduced water volume into enough velocity and siphon action to clear waste in one pass, and the performance gap between a fixture with that engineering and one without it is large and measured. On the industry's own Maximum Performance (MaP) test, a poorly matched 1.28-gpf setup commonly scores only 350–500 grams, squarely in what the trade calls double-flush territory, while a properly engineered 1.28-gpf fixture reaches 800 to the test's 1,000-gram ceiling. Both fixtures can carry the identical "1.28 gpf" label. Only one of them clears a bowl in a single flush.

A retrofit that installs a lower-flow kit into a bowl that was designed and cast around the old, higher flush volume is, by definition, not the matched system WaterSense certification actually tests. It's a valve change wearing the label of a fixture change, and the fixture's real-world performance was never re-engineered to match.

3. The M&V blind spot: one flush in the test, two flushes in the building

A standard flush-volume measurement, the kind an M&V protocol typically records, tests a single flush and reports the water that flush used. If a kit-swapped fixture measures 1.28 gpf on that test, it goes into the savings calculation at 1.28 gpf. What that single-flush measurement cannot see is user behavior: if the bowl doesn't clear on the first flush, occupants flush a second time as a matter of course, and the M&V report has no way to record that unless someone is specifically watching for it.

This isn't a hypothetical gap; EPA has documented exactly this failure mode at the design level. WaterSense's dual-flush toilet specification originally assumed occupants would choose the reduced flush mode roughly twice as often as the full flush mode, which was the basis for its claimed 20% water savings. After the specification had been in the field, EPA received consistent feedback from its own program partners that real usage wasn't coming close to that assumed 2:1 ratio, forcing EPA to revise the entire specification because the lab-basis assumption didn't match how buildings actually behaved. If EPA's own efficiency program had to correct course because a reasonable-sounding usage assumption didn't survive contact with real buildings, a kit-swap retrofit's single-flush lab number is at least as vulnerable to the same gap, with the added problem that nobody is watching for it in the field.

The practical result: a fixture that measures 1.28 gpf on a bench test but requires two flushes per use in the building isn't delivering 1.28 gpf of real consumption. It's delivering something closer to 2.56 gpf, worse than the 1.6-gpf fixture it replaced, while the M&V paperwork still shows the improvement the project was sold on. The designed savings were real on the test bench and are never going to show up on the utility bill.

4. The 15-year problem: kits don't stay put

Even setting aside day-one flush performance, a kit-only retrofit has to survive the full term of a performance contract, commonly 7 to 15 years, for the guaranteed savings to actually be realized. The IPMVP itself names this directly as an M&V risk: it explicitly flags "the likelihood of degradation of originally achieved savings over time" as something a project has to plan for, not something that only happens if a project is poorly run. Standard practice under IPMVP also reduces the frequency of active savings measurement once initial performance is proven, relying on lighter monitoring in the years that follow. That means a kit that gets changed back in year 4 of a 15-year term may not be caught again until year 10, if it's caught at all.

The failure paths are the same ones documented across our other research on fixture drift. A diaphragm has only a 4–6-year service life before it needs replacing anyway, and when it does, facility staff fixing a double-flushing complaint (the exact complaint a poorly matched kit produces) have an easy fix sitting on the shelf: a higher-flow kit that's known to solve the complaint, installed in the identical valve body. A retrofit that was already borderline on performance at installation is the retrofit most likely to get "fixed" back to a higher flow rate the first time someone complains: quietly, with no record, well before the contract term is up.

5. Why this matters for how a project should be built

None of this means flush valve kits are the wrong tool; recommissioning a fixture back to its designed performance is often exactly the right fix, as covered in our companion research. The difference is whether the kit is matched to a bowl and trapway that were engineered to perform at that flush volume, verified with more than a single-flush lab number, and checked again later in the contract term rather than assumed to persist. A retrofit program built on that standard (fixture assessment before the kit is chosen, field verification that accounts for real double-flush behavior, and periodic re-checks across the full contract term) is a fundamentally different product than a valve swap that happens to carry the same gpf number on the close-out report.

← Back to all water insights