The Shower Conversion That Flooded the Hallway

The signal

Three weeks after the shop finished a tub-to-curbless-shower conversion, the customer's daughter called: the hall carpet outside the bathroom was soaked after every shower, out to about 3 ft from the door. The customer, a woman in her eighties, showers seated on a shower chair with the daughter helping and using the hand-held. The job had passed its flood test, the drain had been run on the last day, and nobody had seen a drop outside the shower.

Before diagnosis: water near power, and a wet floor under the user

Two hazards came before any theory.

Water and electricity. The hall had a receptacle at the baseboard about 2 ft from the bathroom doorway, inside the wet area of the carpet. The lead turned off that circuit at the panel and confirmed it dead at the receptacle with a tester proved on a known live source before and after, before kneeling on the wet carpet or pulling it back. The bathroom receptacle was on a GFCI that tested correctly. If water had reached the panel, the box or anything in the crawlspace near wiring, the next call would have been to an electrician, not further diagnosis.

The user. A wet hall and a wet bathroom floor on the way out of the shower is a fall waiting for the person least able to recover from one. Until the cause was found, the family agreed to lay an absorbent runner and towels across the doorway during and after showers, and the daughter walked her mother out each time.

The wet materials. Carpet pad that stays wet for more than 24 to 48 hours is where mold starts, per the EPA's guidance on moisture and mold. The shop pulled back the carpet, removed the wet section of pad, and set a fan and a dehumidifier in the hall the same day, rather than waiting for the diagnosis to finish.

The job as built

The bathroom is 5 ft by 8 ft. The tub was removed from the 60 in end wall, and the shower zone, 60 in wide and 36 in deep, was recessed into the framing so it sits flush with the rest of the bathroom floor at its entry edge. A linear drain runs along the back wall. The shower zone floor slopes to the drain; the remaining 60 in by 60 in of bathroom floor, from the shower edge to the door, was tiled over the existing plywood. The sheet waterproofing membrane covered the shower zone and its walls and ran 6 in past the zone's entry edge onto the bathroom floor. The old door threshold was removed so the floor runs flush into the hall, which is what a curbless conversion is for.

Hypothesis 1: a leak in the plumbing

The leak could be the drain body's connection, the drain line, or the valve and its supply in the wall. All three would wet the hall from underneath or behind, not across the top.

Test. With the shower off, the lead poured 5 gallons of water slowly into the linear drain channel with a hose at the grate, no splashing onto the floor, and then waited 30 minutes. A helper watched the crawlspace below the bathroom and hall with a light. Separately, the shower ran for 10 minutes with the hand-held hung in its holder, spray aimed at the back wall, nobody in it.

Result. Crawlspace dry under the drain, the trap and the hall. Hall carpet edge and the exposed subfloor at the doorway unchanged on a moisture meter, which read the same as a dry reference spot in a bedroom. A valve or supply leak would show between showers too, and the subfloor readings dropped steadily between showers once the fans ran. Eliminated.

Hypothesis 2: the drain could not keep up

Linear drains are sized by flow, and an undersized one ponds water at the grate that can sheet out across a curbless entry.

Test. The hand-held's flow, measured with a bucket and a timer at full open, came to 1.8 gallons per minute, below the 2.5 gpm federal maximum for showerheads at 10 CFR 430.32 and at the lower limit some states set (California among them). During the 10-minute run, the lead watched the grate: no standing water, the flow disappearing within the first few inches of the channel. The drain outlet and trap were the 2 in size plumbing codes commonly require for a shower, as adopted, which carries far more than 1.8 gpm.

Result. Eliminated. The drain was nowhere near capacity.

Hypothesis 3: the slope was wrong

Test. A 4 ft level and a tape across the shower zone and then the bathroom floor.

Result, inside the zone: the floor fell to the drain at about 1/4 in per foot across its 36 in depth, about 3/4 in of fall, which is the slope plumbing codes and membrane makers commonly call for. Fine.

Result, outside the zone: the bathroom floor was not level. Over the 48 in nearest the door it fell 1/8 in TOWARD the hall. The old house's floor had a slight sag toward the center of the house that the original tub and threshold had hidden. A floor that falls away from a drain will carry any water on it away from the drain. This hypothesis was not eliminated; it was half the answer, and it is useless on its own, because nothing in the plan had expected water on that part of the floor.

Hypothesis 4: the waterproofing stopped too soon

The membrane ended 6 in past the shower's entry edge. Beyond it, tile and grout on bare plywood. Grout is not waterproof.

Test. This needed seeing, not measuring. With the customer's agreement, and without her in the room, the daughter showed the lead exactly how the shower routine goes: a chair set where her mother's shower chair sits, at the front of the zone near its entry edge, since that is where her mother can transfer to it from her walker; the daughter standing outside the zone, on the bathroom floor, using the hand-held to rinse her mother's feet, legs and back.

Result. Water landed on the bathroom floor out to about 30 in beyond the zone edge, which is 24 in past the end of the membrane. In 4 minutes the sheet of water on the bathroom floor had crept along the 1/8 in fall to the doorway and onto the hall subfloor. Moisture readings in the bathroom floor's plywood, through two grout joints the lead drilled at about 18 in and 36 in from the zone edge, were elevated; a third at the zone edge, over membrane, was dry.

The cause

The waterproofing and the slope were drawn around where the showerhead points. The water went where the user and her helper go.

A standing user in a curbed shower keeps water inside the curb. A seated user at the front of a curbless zone, with a caregiver standing outside it and a hand-held moving around her body, puts water on the whole bathroom floor every time. The shop had built the right shower for a person standing alone, and the walkthrough would have shown it the person it was actually for: she sits, she sits near the entry because that is where she can transfer, and someone helps her. See related: Curbless Showers and Where the Water Goes, which owns the design rule; this case is what skipping it looks like.

The floor's fall toward the door then took the water that should have been the bathroom's problem and delivered it to the hall.

The repair

The repair had to make the whole bathroom a wet area without adding a curb, which would have defeated the conversion.

1. Full-floor membrane. The tile and setting bed outside the shower zone came up. Cutting and grinding tile and thinset releases respirable crystalline silica, so the work used wet cutting and a HEPA-shrouded grinder with dust extraction and respirators under a written program, per 29 CFR 1926.1153, with the bathroom door closed and the dust barrier up. Sheet membrane was installed over the whole bathroom floor, bonded to the existing shower membrane with the maker's lap and seam method, and turned up the walls 4 in all round.
2. Re-slope the doorway. The floor's 1/8 in fall toward the door was corrected by feathering a leveling layer from nothing at the room's midpoint up to 1/4 in higher at the doorway, so the last stretch of floor falls back toward the room. That 1/4 in is the vertical change the 2010 ADA Standards allow without a bevel (303.2), a reference rather than a requirement in a private home, and one the customer's walker and a wheelchair cross without catching. The membrane runs to the doorway under it.
3. A collapsible water dam at the zone edge. A soft rubber dam set at the shower's entry edge, which folds flat under a wheel or a foot and stands up again, to slow the sheet of water leaving the zone.
4. Flood test the new membrane across the whole floor for 24 hours with the doorway temporarily dammed, before any tile went back.
5. Hall subfloor dried to readings matching the dry reference spot before new pad and carpet went down. No visible mold growth was found on the subfloor; had there been, the work would have gone to a mold remediation scope first.

A second linear drain at the zone's entry edge was considered and set aside for now. It would have caught the water at the source, but it meant opening the floor framing again, and the family wanted the bathroom back.

Restoring the hall circuit. With the carpet re-laid and the subfloor dry, the lead closed the hall circuit's breaker standing to the side of the panel, then tested the receptacle and confirmed the GFCI protecting the bathroom still tripped on its test button.

How the fix was confirmed

The daughter ran the same rinse demonstration on the finished floor. Water reached the same 30 in from the zone edge, stopped at the doorway crest, and ran back to the drain within about 2 minutes. Over the following week of normal showers, moisture readings at the hall doorway subfloor, taken through an access point under the threshold trim, stayed at the dry reference.

References

• 29 CFR 1926.1153, respirable crystalline silica in construction, for cutting and grinding tile and setting beds.
• 10 CFR 430.32, federal water-conservation standard for showerheads; several states, California among them, set lower limits.
• 2010 ADA Standards for Accessible Design, sections 303 (changes in level) and 608 (shower compartments): a reference for a private home, not a requirement there.
• U.S. EPA, guidance on mold and moisture in homes, for the 24-to-48-hour drying window.
• See related: Curbless Showers and Where the Water Goes; How to Convert a Tub to a Low-Threshold Shower.