Every crew that's worked a lift station or a manhole knows the smell. You crack the cover, and it hits you — rotten eggs, unmistakable. A few minutes into the descent, it seems to fade. By the time you're standing at the bottom, you can barely smell it at all.
Most people read that as good news. The smell went away, so the air must be clearing up.
It's the opposite. Your nose just stopped working, and it stopped working at almost exactly the concentration where hydrogen sulphide starts to kill.
Where the gas comes from
Hydrogen sulphide (H2S) in a sewer isn't a contaminant that leaked in from somewhere else — it's manufactured on site, continuously, by the wastewater itself. When sewage sits somewhere with little oxygen and little flow, anaerobic bacteria take over. Sulfate-reducing bacteria break down organic matter in the wastewater using sulfate instead of oxygen, and H2S is the by-product.
That process needs exactly the conditions you find in the confined spaces crews enter: warm, slow-moving or stagnant wastewater, low or no oxygen, and time. Wet wells, lift stations, force mains, and dead-end or low-flow lines are close to ideal breeding conditions for it. The gas builds fastest in the same low-velocity, low-oxygen spots that also happen to be entry points.
The smell disappears right where it starts to kill you
Here's the part that catches people off guard. H2S is one of the most detectable gases there is — the human nose can pick it up at extraordinarily low concentrations, down around 0.0005 to 0.3 parts per million. That's a tiny fraction of a single part per million. At that level, it's not dangerous. It's just unmistakable.
But smell and hazard don't move together as concentration rises. Above roughly 100 ppm, olfactory fatigue sets in: the nerve endings responsible for detecting the odor get overwhelmed and stop signaling. Continuous exposure in the 100–150 ppm range can knock out your sense of smell within two to fifteen minutes, and outright paralysis of the olfactory nerve has been documented at 150 ppm.
| Concentration | What's happening |
|---|---|
| 0.0005–0.3 ppm | Odor threshold — the smell is detectable, and the air is not dangerous |
| 10 ppm | BC's ceiling limit — must never be exceeded, at any moment (OHSR s.5.48) |
| 100 ppm | IDLH — immediately dangerous to life or health. Also roughly where olfactory fatigue begins. |
| 100–150 ppm | Sense of smell can be lost within 2–15 minutes of continuous exposure |
| 150 ppm | Paralysis of the olfactory nerve reported |
Read that table again. The concentration where the smell fades and the concentration that's immediately dangerous to life are the same number. The gas didn't get weaker on the way down that ladder. Your ability to detect it got disabled right as it became lethal.
Your nose is a low-concentration instrument. It has no way to tell you the difference between "gone" and "now at a lethal dose." Once you're past about 100 ppm, silence from your sense of smell means nothing at all — which is exactly why BC requires an instrument down there, not a nose.
It's also heavier than air, and it settles exactly where your crew stands
H2S has a molecular weight of about 34, against roughly 29 for air — meaning it's noticeably heavier and tends to pool rather than disperse. In a confined space, that means it collects at the bottom of a manhole, in the low point of a wet well, in a sump, in exactly the elevation where an entrant is standing or kneeling while a monitor clipped to their chest reads a level a foot or two higher. A reading taken at the wrong height can miss the worst of it entirely.
What "clean respirable air" actually requires
BC's OHS Regulation defines clean respirable air (s.9.1) as air containing no contaminant above 10% of its applicable exposure limit. For H2S, with a 10 ppm ceiling limit, that works out to 1 ppm — a tenth of the ceiling, and still well inside the range your nose can pick up if it hasn't already been overwhelmed. The regulation's bar for "clean" sits far below the point where a person would even describe the air as bad, let alone dangerous.
Why one reading at the top of the ladder isn't enough
Sewer H2S isn't a steady background level. It spikes — when sludge gets disturbed, when an upstream lift station kicks on and pushes a slug of septic wastewater through, when a valve opens and stagnant flow suddenly moves. A space that tested clean sixty seconds before a worker climbed in can be well past IDLH by the time they're at the bottom.
That's exactly why BC's testing requirements don't stop at the pre-entry check. Under s.9.25: testing has to happen within 20 minutes of entry (ss.1–2), it has to be repeated if the space is vacated for more than 20 minutes (s.3), and if a worker is inside a moderate or high hazard atmosphere, additional testing must be conducted as necessary to ensure the worker's continuing safety (s.4). Continuous monitoring is required "whenever practicable" (s.5). A single reading at the surface, taken once, tells you almost nothing about what the atmosphere will do once someone is standing in the low point twenty minutes later.
This is the same logic behind the oxygen reading most people over-trust — a good number at one moment, at one height, doesn't mean a good atmosphere throughout the job. It just means it was good at that moment, at that height. With H2S in a sewer environment, that gap is often measured in minutes, not hours.
The smell is a warning system with a hard ceiling. Past that ceiling, it goes silent right when you need it most — which is exactly why the testing has to be instrument-based, done at the right height, and repeated through the job, not trusted to a nose that's already been fooled once.