What a Floor Drain Needs That a Fixture Drain Does Not

Why this matters

Every trap in a building is a small volume of water standing between the occupants and the sewer, and almost all of them are refilled every time somebody uses the fixture above them. A floor drain has no fixture above it and no user. Its seal is maintained only by whatever replenishment somebody deliberately arranged, and if nobody did, the seal is on a countdown from the day the building opened. That is the one structural difference, and six separate requirements fall out of it. This card is the specification, with the fields filled in for a real location and the measurement that sets the one number nobody ever measures.

First actions on finding a dry one

• Pour water in it before anything else. A dry floor drain is an open pipe to the sewer, and it is open in the direction that matters: gas comes in, and it does so continuously.
• If the space smells strongly of sewer gas, ventilate it to outside air before working in it. Sewer gas is an inhalation hazard and the methane in it is flammable, and a mechanical room is full of ignition sources.
• Probing a trap means handling sewer water. Nitrile gloves, hands away from the face, wash before eating or before the next call. If the line has surcharged and there is any chance of a pressurized release, add splash goggles.
• Standing water in a mechanical room around energized equipment is an electrical problem first. Open the disconnecting means, lock it, and verify dead at the point of work under 29 CFR 1910.333(b)(2) in general industry or 29 CFR 1926.417 on construction work before anyone wades in to find the drain.

The one structural difference

A lavatory trap is refilled several times a day as a by-product of the lavatory being a lavatory. Nobody has to think about it, nobody schedules it, and it does not appear in any maintenance program because it does not need to.

A floor drain is refilled when a floor is washed, a relief valve discharges, a condensate line runs, a primer fires, or a person walks over with a bucket. In a great many rooms none of those five things happens, ever. So the floor drain is the only drain in the building whose seal has a budget rather than a supply, and everything below is either about extending that budget or about the other jobs the drain has picked up by virtue of sitting at the low point of the room.

The six fields of the spec

Field 1: seal depth. A conventional fixture trap seal is on the order of 2 inches, and the minimum is set by the plumbing code as adopted and amended by the local jurisdiction. A floor drain in a space with no regular flow gets a deep-seal trap instead, commonly 3 to 5 inches. The reason is arithmetic rather than tradition: evaporation removes depth at a rate set by the room's air, so doubling the depth roughly doubles the time before the seal is gone. It buys time. It does not remove the requirement for replenishment, and specifying a deep seal and then walking away is the most common way this gets half-done.

Field 2: replenishment source, named explicitly. One of: regular washdown flow, an indirect waste that discharges to the drain often enough to matter (a condensate line, an equipment drain, a relief discharge), an automatic trap primer, or a scheduled manual pour written into somebody's route. Pick one and write it on the drawing or on the maintenance schedule. Which one is right, and specifically when a primer is the wrong answer, belongs to a sibling card; the point here is that "none" is a specification error rather than an omission.

Field 3: strainer and sediment capture. A floor drain collects whatever gets swept toward it, and the trap is a poor place to store it. A removable strainer plus a sediment bucket in the drain body catches solids where somebody can lift them out. Without one, the solids go into the trap and then into the branch, and the drain that never gets used becomes the drain that is always partly blocked.

Field 4: trap arm and vent. The floor drain's trap needs venting for the same reason every trap does, and it is the trap most likely to be treated as an afterthought during a slab pour. Two things go wrong here: the vent is omitted because the drain "does not get used", and the trap arm runs long because the drain sits wherever the floor slopes rather than wherever the stack is. Slope and trap arm length route to the adopted code and to sibling cards; the field note is that a floor drain's geometry is set by the floor, not by the plumbing, so it is the one that gets stretched.

Field 5: backwater protection. The floor drain is at the low point, so when the sanitary line surcharges it is the first opening water comes out of. A backwater valve protects fixtures below the elevation surcharge can reach; it belongs on the drains that need it rather than on the whole building, and it needs service access, because a valve nobody can reach is a valve nobody maintains.

Field 6: flashing clamp and weep path. On a waterproofed floor - a shower pan, a kitchen, a mechanical room with a membrane - the drain body has a clamping ring that grips the membrane and weep openings that let water travelling on top of the membrane get into the drain. Fill those weeps with mortar during a tile set and water reaching the membrane has nowhere to go; it stays under the finish and finds a wall.

Worked example: the filled-in spec for one location

A boiler and pump room in a small commercial building. One 2 inch floor drain in the corner. The complaint is an intermittent sewer smell in the corridor outside, worst in winter.

Field 1, seal depth as built: 4 inch deep seal, confirmed by probing. Field 2, replenishment source as built: none. There is no washdown routine, no condensate line routed to it, no primer, and no entry on any maintenance schedule. Field 3: strainer present, sediment bucket present and about a third full. Field 4: vented, trap arm short, no issue found on camera. Field 5: no backwater valve; the room is above the surcharge elevation for this building so none is required here. Field 6: no membrane in this room, so not applicable.

Field 2 is the finding. Now measure what that costs.

The measurement. Probe the seal with a dipstick, record, come back at a known interval, probe again.

• Day 0: 4.0 inches
• Day 21: 2.6 inches
• Drop: 1.4 inches over 21 days, so 0.067 inches per day

Correction 1, confirm the drop is actually evaporation. A raw drop includes any seal lost to siphonage or back-pressure during the interval, and attributing those to evaporation overstates the rate and produces an interval that is too short. Checked before using the number: the branch serves nothing else, no fixture upstream discharged during the interval, and the vent is clear. On that evidence the 1.4 inches is evaporation and the rate stands.

Correction 2, what the error term is. Both readings are taken on the same dipstick in the same trap, so any fixed offset in where the stick registers the weir is a systematic error common to both and it cancels in the subtraction. What remains is the reading spread, about one sixteenth of an inch per reading, independent between the two readings. Independent spreads combine in quadrature: 0.0625 times the square root of two is about 0.09 inches on a 1.4 inch difference, roughly 6 percent. That does not change any conclusion below.

Correction 3, the functional floor is not zero. A trap with half an inch left in it has no reserve against the pressure events that a sibling card covers, so treat the practical floor as the conventional 2 inch fixture seal rather than as dry. The budget is therefore from 4.0 inches down to 2.0 inches, which is 2.0 inches.

• 2.0 inches divided by 0.067 inches per day equals 30 days

Correction 4, that 30 is a bound, not a schedule. Evaporation rate depends on the air temperature, the humidity and the air movement over the trap, and this is a boiler room in winter, which is the hottest and driest that trap will ever see. Written honestly with one inequality and no interval: the safe replenishment interval is less than 30 days. In summer, with the boiler off, it will be considerably longer, and that is exactly why the smell complaint is seasonal.

Setting the interval. Schedule at 50 percent of the measured budget so the seasonal swing is absorbed: 30 times 0.5 equals 15 days, rounded down to a 14 day cycle that lands on the same weekday. Re-measure once in each season and adjust; do not carry a winter interval through a summer or the other way round.

What to actually pour. Approximating the seal as the pipe's cross-sectional area times the depth, a 2 inch trap holding a 3 inch seal contains roughly 9 cubic inches, about 5 fluid ounces. That is the order of magnitude, not a specification, because a cast trap's real geometry differs. It is also the reason a small measured pour does not work: water poured in displaces the water already there and runs straight through, so restoring a seal is not a matter of adding the deficit. Pour a generous volume, at least a quart, slowly. Overfilling costs nothing and under-pouring leaves the trap where it was.

Where the facility and the local sewer-use ordinance permit it, a small quantity of a sealing liquid floated on top slows evaporation substantially. Treat it as a supplement that lengthens the interval, not as a replacement for the pour, and confirm it is permitted before introducing anything to a drain on a customer's behalf.

The failure mode this catches. The common program is "pour water in the floor drains monthly", applied uniformly to every drain in the building. Against a 30 day budget in the boiler room, monthly is already at the edge and it fails in a cold snap when the boiler runs harder. Against a walk-in cooler vestibule, where the air is cold and near saturation and evaporation is slow, monthly is servicing a drain that would hold its seal for most of a year. The uniform schedule is simultaneously too slow where it matters and wasted everywhere else, and no complaint ever surfaces from the drains where it is wasted.

What flips the specification. A drain receiving a condensate line, an air handler pan drain or an equipment relief discharge has field 2 satisfied by that flow and needs no scheduled pour while the equipment runs. Note in passing, since the two sit together in a search: the compressed-air and steam-side drain traps covered elsewhere in this library are a different device, solving condensate removal rather than gas exclusion. A drain in a food preparation area may not accept a sealing liquid at all. A drain below the surcharge elevation makes field 5 the leading requirement rather than a note, and the backwater valve's access becomes part of the spec. And a drain in a room washed weekly needs none of the replenishment machinery, which is why this problem is invisible in kitchens and obvious in mechanical rooms.

Reading the seasonal complaint

The most useful diagnostic fact about floor drain seal loss is that it is seasonal in a heated space and not seasonal in an unheated one. Warm dry moving air pulls the seal down fastest, so complaints cluster in the heating season, or year-round where dehumidified air is pushed through a space continuously. A sewer smell that arrives every winter and clears every spring, in a room nobody uses, is this mechanism until proven otherwise, and two dipstick readings three weeks apart prove it.

The mirror case keeps the cue from being over-applied: a smell that appears at a fixed time of day, or right after a particular piece of equipment runs, is not evaporation. Evaporation has no clock. That pattern points at a pressure event, and a sibling card covers which one.

References

• 29 CFR 1910.333(b)(2) for electrical lockout and verification in general industry; 29 CFR 1926.417 for the construction counterpart
• The plumbing code as adopted and amended by the local jurisdiction, which sets minimum trap seal depth, venting, backwater valve requirements and floor drain provisions
• The sewer-use ordinance as enacted by the local authority, which governs what may be introduced to a drain
• See related: What a Trap Primer Is Solving and When It Is the Wrong Answer; Drain Pipe Slope + Venting Quick Reference; Drain, Waste + Vent (DWV) Reference