What We Found After Shower Demolition: Six Denver-Area Showers, Opened Up

We write a project page for most of the leak repairs we do, with photos taken as each layer comes off. This article pulls the demolition findings out of seven of those pages (six repairs and one inspection) and lines them up. We have left out customer names, prices and schedules. City names stay, because the age and style of construction in each area is part of the story. Where a project page is vague about a detail, we stay vague here rather than fill it in.

If you want the general theory first, start with What Causes a Shower Pan to Fail and Tile and Grout Are Not Waterproofing. This piece is the evidence behind both. For what demolition turns up in the rest of the bathroom, there is a companion article: Hidden Problems Found During Bathroom Demolition.

Denver: a one-year-old shower with screws through the liner

Second-floor bathroom. The shower was about a year old and had leaked from the start: after three or four uses, water came through the first-floor ceiling and the drywall sagged and bubbled. The wall tile was new and in good condition, so we did not want to tear it out to find the leak. We cut an access opening in the ceiling below, which showed the water was coming from the pan area rather than the walls or the plumbing, and let us look at the joists. Then we removed only the bottom four rows of wall tile, the floor tile and the mortar bed.

Under the bed was a PVC liner with several screws driven straight through it in one corner of the shower floor, below the flood level. The old liner had also been fastened to the framing improperly, and nothing suggested it had ever been flood tested. The drain connection itself was adequate, but the liner's seal at it was compromised. The framing had minor rot where water had been sitting, caught before the joists took real damage; the ceiling drywall below was saturated.

The fix: pull the liner and photograph the punctures for the homeowner, treat the wet wood, install a new 40-mil PVC liner with every fastener in its upper portion, re-seal the drain with silicone at the clamping ring and protect the weep holes with pea gravel, then flood test with the ceiling opening still open so we could check from underneath. After that, a dry-pack mortar bed sloped to the drain, a hexagon mosaic floor, the four rows of wall tile reset, and the ceiling repaired. The full record is on our leaking shower and pan repair page.

DTC: screw damage and missing dam corners under one row of tile

The homeowner noticed leaks showing up on the ceiling below the shower. We removed one row of wall tile and the existing pan. The old base was leaking, some areas had been damaged by screws, and the rubber corner pieces that waterproof the curb-to-wall junction (dam corners) were missing. The framing at the threshold showed mold and water damage.

Because the walls above that row were sound, this stayed a targeted repair. We dried the area, treated the exposed surfaces with KILZ, formed a new pre-slope, installed a new liner, floated the upper mortar layer and finished the floor in mosaic tile. The bench and the wall transitions in the repaired section were waterproofed with RedGard liquid membrane with reinforcing tape in the corners, and the wall tile was reset.

Commerce City: a 1990s liner with no pre-slope, and the floor under it

Second-floor shower. The first sign was small water stains on the living-room ceiling directly below. Our moisture inspection and the demolition that followed found that the original 1990s-era installation had used an inadequate vinyl liner with an improper drain connection and no pre-slope mortar bed, so water pooled against the liner seams until the material deteriorated. By then the plywood subfloor was saturated and falling apart, the floor joists showed early rot, and the ceiling drywall below was saturated, sagging and growing mold. The recessed lighting in that ceiling had to be checked by an electrician.

There was no partial version of this repair. Everything came out to the studs and joists. The structure was dried with dehumidifiers and air movers and checked with moisture meters, the subfloor was replaced with 3/4-inch plywood, the joists were sistered, and the pan was rebuilt as a traditional mortar-bed system: pre-slope, PVC liner run up the walls with folded corners, flood test, then the top bed. The walls and curb got a liquid-applied membrane, and the shower was finished in natural stone with a pebble floor and movement joints at the transitions.

Elizabeth: a liquid membrane on bare plywood, built in 2007

Mold appeared on the ceiling below the second-floor bathroom and spread to adjacent walls. The shower was built in 2007. When we opened the floor, we found a liquid-applied waterproofing membrane that had been put directly on the plywood subfloor and had failed outright. The mortar bed crumbled as it came out, the plywood underneath was soft and spongy and in places gone, and the joists showed early rot.

We cut out and replaced the damaged plywood with new OSB, inspected and treated the joists, then built the pan the traditional way: a pre-slope at 1/4 inch per foot, a 40-mil PVC liner with the corners folded rather than cut and run well above the curb, silicone at the drain with the weep holes kept open, a flood test and inspection, and a final mortar bed with pea gravel around the drain before a hexagon mosaic floor that works with the existing wall tile.

Arvada: a corner joint that leaked for years

An older tiled shower with a persistent leak, handled as an insurance-related reconstruction. On the first visit the signs were already on the surface: tiles loose, sections failing, and water getting through a corner joint in the tile assembly. Demolition showed significant water intrusion behind the tile, rotted wood in the subfloor area and damage to sections of the 2x4 framing. The project page does not pin the original pan failure to one detail, and neither will we; the age of the shower and the leaking corner are what is documented.

We replaced the compromised plywood, cut out and rebuilt the damaged framing, and built a custom liner-based pan with a pre-slope shaped to the footprint. The shower was finished with 12-by-24 wall tile, a mosaic floor and a recessed niche.

Aurora: the subfloor around the drain

Water had been escaping this shower long enough to soak the subfloor around the drain and rot the framing under the adjoining tub deck, and none of it was visible from the finished side. We stripped the lower wall tile and the flooring to expose the subfloor and the drain line. Around the drain the plywood was saturated and delaminating. Our page lists the usual paths for a slow leak like this, a cracked grout joint, a failed pan connection, a gap where caulk should have been, without naming which one it was, so we leave it there. The rebuild went framing, subfloor, cement substrate, membrane, and only then a new white shower base, tile walls and a tiled bench.

The one we caught before the tile went on

Not a demolition, but it belongs here because it shows the same mistakes before anything hides them. A Denver-area homeowner asked us to look at a pan another contractor had just built. The liner's inside corners had been cut rather than folded, the adhesive work at the folds was missing or incomplete, and the clamping drain was not assembled correctly. We put a test plug in the drain and poured in roughly half a bucket of water. It came out from under the pan almost immediately. The project stopped there, before the mortar and tile, which is the cheapest point at which any of the showers above could have been caught.

Line the findings up and they stop looking like bad luck. Each failure below maps to a specific requirement. We distinguish the source: the model building code (the IRC, which Denver-metro jurisdictions adopt in different editions with amendments), the liner manufacturer's instructions (Oatey, whose PVC liner is what we and many installers use), and trade or professional guidance.

One pattern is not in the table because it is in every row: the first symptom was not in the shower. Five of the six homeowners found out from the ceiling below. That is why regrouting or recaulking never had a chance in any of these cases; see Why New Grout Won't Fix a Leaking Shower.

The hole in the Denver liner was a few screw heads wide. The leak in Elizabeth was a membrane that was too thin in places. Neither sounds like a reason to replace a subfloor. The reason is time. The EPA's guidance for any wet building material is that it must be dried within 24 to 48 hours to avoid mold growth. A shower pan leak wets the wood again every morning, in a closed cavity with no airflow, and nobody knows until the ceiling shows it. By then the question is not whether the wood can dry but how much of it is still wood.

That is also why the fix in four of the six showers included structural work, and why we cut back to sound, dry wood instead of drying in place. In Commerce City and Elizabeth the subfloor was replaced and the joists were sistered or treated; in Arvada and Aurora the plywood and sections of framing were rebuilt. Tim Carter's advice to homeowners is the same as ours: confirm the studs and subfloor are dry and sound before anything is closed back up. Denver is the exception that proves it. The leak was caught after a handful of showers, the rot was minor, and treating the wood was enough. Speed, not the size of the hole, decided the scope.

The other half of the answer is that water does not stay where it enters. It follows the slope of the subfloor, soaks along the bottom plates, and comes out at the lowest point: the ceiling below, or the framing under an adjoining tub deck in Aurora. Opening the shower is how we find the entry point. Opening the ceiling or the adjacent floor is how we find out where it went.

Nothing in the fixes above is exotic. They are the ordinary sequence, done in order, with the two or three details the original installers skipped. Our practice on every shower we build or rebuild is the same:

• A code-listed membrane. Either a PVC liner meeting ASTM D4551 with folded corners, dam corners at the curb and no fasteners near the floor, or a bonded membrane meeting ANSI A118.10 applied to the manufacturer's coverage, over a proper substrate rather than bare plywood.
• A sloped base under the membrane. The pre-slope that Commerce City never had, so water that reaches the liner moves to the weep holes instead of sitting in the bed.
• A flood test before tile. Plugged, filled, held, and checked from below when there is a below. On the Denver job we left the ceiling open for exactly that reason.
• Dry, sound structure first. Wet framing and subfloor come out or dry out before a new pan goes in. The EPA's 24-to-48-hour window has usually passed by the time we arrive, so "out" is more common than "dry."
• The smallest opening that answers the question. Denver and DTC kept most of their wall tile. Commerce City, Elizabeth and Arvada could not. The decision follows the findings, not the other way around; see Shower Repair vs. Complete Rebuild.

Does every leaking shower need this much demolition?

No. Two of the six, Denver and DTC, were repaired under the existing wall tile with one to four rows removed. What decided it was the condition of the walls above the cut and how far the rot had spread, not the age of the shower. The three full rebuilds all had rotted subfloor and framing damage that could not be reached any other way.

Can the wet wood just dry out once the shower is open?

Sometimes, if the leak is caught early, as in Denver. The EPA's rule for wet materials is drying within 24 to 48 hours. Wood that has been wet for months, which describes Commerce City, Elizabeth, Arvada and Aurora, has usually delaminated, softened or started to rot, and the sound answer is to cut back to dry wood and rebuild.

How can I tell which of these my shower is before anyone opens it?

You cannot fully, and neither can we. A flood test tells us whether the pan holds; spraying the walls, niche and curb separately narrows a wall leak; a look from below, through a basement ceiling or a small access opening, often shows the corner. The findings in this article all came after that diagnosis, not instead of it.

My shower is new and already leaking. Is that normal settling?

No. A new shower that leaks is a shower with an installation defect, like the Denver pan with screws through the liner or the inspected pan with cut corners. Both would have failed a 15-minute flood test before the tile went on. Stop using it, document the ceiling, and get the pan diagnosed before anyone proposes a surface fix.