Dry drain trap
A Socratic walk-through of the dry drain trap — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why does a bathroom nobody uses start to smell of drains?
A cloakroom nobody has used since spring. Nothing has gone down that basin, nothing has rotted in it, nothing has been spilt. And it is the one room in the house that smells of sewer. The bathroom in daily use — same pipework, same waste, vastly more of everything unpleasant passing through — smells of nothing.
That inversion is the whole clue, and it points the wrong way for the obvious explanation. If dirt caused the smell, the busy room would be the offender. So the question is not what the idle drain has accumulated. It is what the busy drain keeps doing that the idle one has stopped.
Reasoning it through
REASONING #Begin with what a drain is connected to. Every waste pipe runs, eventually, into the same sewer, and a sewer is an anaerobic digester with no lid, generating hydrogen sulphide, methane, ammonia and a long tail of volatile organics. That gas is under no particular pressure, but it moves wherever the pressure is lower — and a warm house, with air rising and leaking out at the top, sits reliably below the ground beneath it.
So the sewer is always trying to breathe into your house. What stops it? Not a valve, in most fittings. Bend the pipe below the fixture into a U and a small quantity of water sits in the bottom of the bend and cannot run out; gas arriving from the sewer side meets that plug and stops. The seal is nothing but water held by gravity in a bend — which is why it is called a trap, and why the depth of that water matters. Codes specify a minimum seal depth, commonly around 50 mm, though that varies by jurisdiction and fitting and I am recalling it rather than deriving it.
Now the answer arrives on its own. Water evaporates. In an unused room the trap is never refilled, the seal shrinks, and one day it is a damp bend rather than a plug. The gas walks straight through. The busy bathroom is not cleaner; it is simply topped up several times a day by ordinary use.
Is that the only way to lose a seal? No, and the alternatives matter because they are the ones that come back. A fixture discharging fast can siphon its own trap dry behind the departing water — self-siphonage. A larger flow passing in the stack alongside can drag the pressure down and pull a neighbouring trap out with it — induced siphonage, which is exactly what vent pipes exist to prevent. A blockage further down can push the other way, blowing a seal out as bubbles. And a hair or a thread of lint draped over the weir of the bend will wick water out of the trap by capillary action for weeks, quietly.
Each of those has a different signature, which gives us a test. Pour a litre of water into every trap in the room — basin, bath, shower gully, the standpipe behind a washing machine, and the floor gully people forget. The refuting observation: if the smell is unchanged an hour later, the problem is not a lost seal at all, and you should be looking for a trap that was never fitted, an unsealed washing-machine standpipe, or a cracked pipe in the void. If the smell goes and stays away a fortnight, it was evaporation and the fix is nothing more than running the tap monthly. If it goes and returns within a day or two, refilling is not the answer, because something is removing the water faster than you can put it back — and if the return coincides with another fixture draining, the venting is at fault, not the trap.
The analogy
THE ANALOGY #Think of a doorway with no door, kept shut by a bucket of water balanced across the threshold. Nothing about the bucket is clever; it simply occupies the gap. It works perfectly and invisibly for as long as somebody keeps topping it up, and the moment the household stops walking through that doorway the topping-up stops with them. The room is not breached by an intruder. It is breached by an absence.
A bucket empties only if you let it, whereas a trap can also be actively pulled dry by pressure changes elsewhere in the system — and the doorway image suggests the barrier is doing a job, when the water is doing nothing but sitting there being incompressible.
Clarifying the model
THE MODEL #The folk explanation here is worth naming precisely, because the advice attached to it is correct and the reason given for it is not. "Pour water down the drain to wash away the smelly gunk" gets people to do exactly the right thing for exactly the wrong reason. Nothing is being washed away. Nothing was there. You are refilling a seal, and that is why a cupful is enough where a bottle of bleach is not: bleach kills the biofilm it lands on and then evaporates alongside the water, leaving the trap just as empty a month later.
That distinction earns its keep when the seal is intact and the smell is real anyway. A washing machine that smells is a different animal — a lipid biofilm on the machine's own surfaces, which does respond to heat and oxidiser. Sort them by where the smell is strongest: at the plughole, look at the seal; in the drum or on the door gasket, look at the film.
Two honest limits. "Sewer gas" is a mixed and variable thing, and the case against it at household concentrations is largely about hydrogen sulphide and about nuisance rather than a well-quantified dose — I would not put a number on it. And self-sealing waste valves, sold precisely for rarely-used fixtures, are a genuine fix but a mechanical one: they have a failure mode, whereas a water seal has only an empty state that any tap can undo.
A picture of it
THE PICTURE #How to readStart at the rounded terminal and follow the parallelogram, which is the single action the whole diagnosis rests on: refilling the traps. Each diamond is a question with both answers labelled, and the branches sort four different faults apart by timing rather than by smell. Red shapes need a plumber; green is the harmless case. The back-edge from "Run the tap monthly" is the loop a used bathroom performs by itself several times a day.
What became clearer
WHAT CLEARED #The barrier between a house and its sewer is a few centimetres of water sitting in a bend, maintained by nothing but the habit of using the room. A drain does not smell because it is dirty; it smells because it is idle, and the fix is not cleaning but replacement of something that quietly went missing. Once that is clear, the diagnosis stops being about odour and becomes about time: how long the smell takes to return after a litre of water tells you which of four quite different faults you have.
Where to go next
ONWARD #- Why drainage systems need vent pipes at all, and what a stack's pressure does when a bath empties into it.
- How a washing machine's own biofilm produces a similar smell by an unrelated mechanism.
Key terms
TERMS #| Term | What it means |
|---|---|
| Trap | a bend in a waste pipe holding a plug of water that blocks gas from the sewer side. |
| Seal depth | the vertical height of water held in that bend, and the margin available before pressure or evaporation empties it. |
| Induced siphonage | loss of one fixture's seal caused by a different fixture discharging past it in a shared stack. |
| Self-sealing waste valve | a membrane device fitted in place of a water trap, intended for fixtures used too rarely to keep a seal. |
Every term the collection defines is gathered in the glossary.