Rim readings on a live sewer manhole
A three-person crew is preparing to enter a 4.2 m deep, 1.2 m diameter sewer manhole on a live municipal network. The gas tester holds a four-gas detector fitted with a 3 m sample line and a hand aspirator pump. Standing at the rim he draws a sample from just below the cover and records oxygen 20.8%, LEL 0%, hydrogen sulphide 4 ppm and carbon monoxide 6 ppm. All sit inside the site entry limits of 19.5–23.5% oxygen, 5 ppm H2S, 25 ppm CO and 5% LEL. A tanker is waiting, and the supervisor tells him to sign the entry permit so the crew can get down and clear a silt blockage. The manhole connects to a 600 mm interceptor sewer carrying domestic and trade effluent; the last rainfall was 30 hours ago. The day is warm at 29 °C with no wind. A tripod and harness are set at the rim, but the attendant has never operated the winch.
1. The rim readings are all within limits. Explain why they do not authorise entry, and set out the sampling programme you would insist on before signing the permit.
Read the worked answer
No. Four numbers taken just below the cover describe only the top of a 4.2 m column, and a sewer is not a well-mixed box.
- Density stratification. Hydrogen sulphide (relative vapour density about 1.19) and carbon dioxide (about 1.52) are heavier than air and collect at the invert. Methane (about 0.55) rises and gathers under the cover, where the open air above can dilute it below detection. The one place tested is the least likely place to find H2S and one of the more likely places to lose methane.
- The sewer is the source, not the manhole. A 600 mm interceptor with 30 hours of dry-weather flow holds sludge, slime and sediment. The silt blockage the crew is being sent to clear will release a plug of gas the moment it is disturbed. Warm, still air (29 °C, no wind) suppresses the convective mixing that would otherwise help.
- Sampling programme: test at the top, at mid-depth and at the invert; test the upstream and downstream sewer ends if they can be reached safely; test before entry, after any disturbance (rodding, jetting, breaking the blockage), at each break period specified by the permit, and whenever an entrant reports anything unusual. Only readings taken at the level of the entrant's head, at the work location, speak to the entry decision.
Sign nothing on the strength of the rim numbers.
2. How would you physically obtain a sample from the base of the manhole, and what could make that sample unrepresentative?
Read the worked answer
Purge the line, then the void.
- Work out the line volume. Three metres of 4 mm bore tubing is roughly 38 mL. With a hand aspirator delivering around 100 mL per stroke, you need well over the minimum: purge at least three line volumes plus the detector's internal sample path, and keep going until the readings stabilise. Follow the strokes-per-metre figure printed by the manufacturer rather than guessing.
- Do not sample through a long line while descending. Lower a rigid probe or a weighted line to the invert, keep the open end clear of silt and standing water, then draw the sample up.
- Water. A sewer is wet. Use a water trap and a hydrophobic filter, and stop the pump before the line takes liquid. A wet filter blocks flow and produces a false low reading while the pump still sounds normal.
- Adsorption. H2S sticks to silicone tubing and to dirty or damp lines. Use PTFE or the manufacturer's recommended material, keep lines short and replace them when contaminated. A slow-responding H2S channel is a classic sign of a dirty line.
- Response time and drift. A drawn sample is a spot value. Note the time of every reading, and repeat after the blockage is disturbed.
- Confirm what matters. Where a value decides the entry, confirm it with a second instrument or a colorimetric tube.
A sample that has travelled through 3 m of tubing and a wet filter is not the same as the atmosphere at the entrant's face.
3. At the base you obtain oxygen 19.9%, LEL 2%, H2S 14 ppm and CO 9 ppm. The supervisor proposes the crew wear the escape sets kept in the van and work fast. Decide and justify.
Read the worked answer
Refuse the entry. Require forced ventilation and, if it cannot bring the reading down, full working breathing apparatus with a BA-trained rescue team.
- H2S at 14 ppm is the deciding number. It exceeds the site's 5 ppm entry limit and exceeds any short-term exposure limit that can be defended (TLV-TWA 1 ppm with a 5 ppm STEL; the UK WEL is 5 ppm over 8 hours and 10 ppm over 10 minutes). It is in the range that irritates eyes and throat, and it is only three to six times below the concentration that deadens the sense of smell and then produces olfactory fatigue in the exposed person.
- 14 ppm is a minimum, not a maximum. It was taken before the blockage was broken. Disturbing silt releases H2S in bursts, sometimes within seconds, and the number can rise by an order of magnitude.
- Escape sets are not working breathing apparatus. They are short-duration, escape-only devices, and they give no protection to someone who has not donned them before exposure. Work fast is not an engineering control, and it is exactly the phrase used in the incidents that follow.
- What to do: hold the entry, leave the cover fully open, set up forced ventilation with the intake at the invert, run it until readings are stable, then re-test at all levels. If ventilation of a live sewer cannot get H2S below the limit, entry needs full BA, a BA rescue team standing by, and a permit that says so, not a four-gas meter and an untried winch.