Stratified Hydrogen Sulfide in a Lift-Station Sump Entry
A contractor is preparing to enter a 3.2 m deep, 1.4 m diameter concrete lift-station sump to replace a corroded guide rail. The entry permit was issued this morning after a single set of readings taken at the top of the opening with a calibrated four-gas meter: oxygen 20.4 percent, LEL 0 percent, H2S 6 ppm, CO 0 ppm. A blower sits at the rim with a 1 m flexible duct.
The attendant lowers the meter on a line to 2 m as a check and reads H2S 22 ppm with oxygen at 20.1 percent. The entrant is standing by in a full-body harness on a tripod with a 15-minute escape respirator, but the planned working respiratory protection is a half-mask air-purifying respirator with H2S cartridges. The site supervisor says the sump has been entered a dozen times in the last two years without incident, that the 22 ppm reading is just a slow meter, and that the crew has a four-hour window before the line must be back in service. The general contractor's safety officer is off site until tomorrow.
1. What do you do in the next ten minutes?
Read the worked answer
Stop the entry. No one goes into that sump until the atmosphere at the work location has been characterised and the permit reissued. The permit is invalid because it was issued on a single top-of-opening sample, not on readings at the point where the person will breathe.
In practice: keep the area clear and post the opening, notify the entry supervisor and the general contractor, and have the meter re-checked and calibrated if there is any doubt. The 22 ppm reading is a real reading; a slow response explains a delayed number, not a false one, and it argues for more monitoring, not less.
Re-classify the space as permit-required with an atmospheric hazard and require a revised permit: continuous monitoring worn by the entrant at the work location, forced-air ventilation with the duct run down to the bottom rather than left at the rim, a verified pre-entry reading from top, middle and bottom, a rescue plan with trained rescue personnel, and supplied-air respiratory protection with an escape cylinder instead of the air-purifying respirator.
If the four-hour window cannot be met under those conditions, the window moves. Notify the client and the safety officer in writing the same day, and record the stop-work with the reading, time, meter identity and calibration date.
2. Explain the technical basis for rejecting the top-only sample and the half-mask air-purifying respirator.
Read the worked answer
Hydrogen sulfide is heavier than air, with a molecular weight of 34 against about 29 for air and a relative density near 1.19, so it accumulates in low points. In a quiescent, poorly mixed sump the concentration stratifies, and a reading taken at the rim says nothing about the bottom where the entrant's head will be. A blower at the rim on a 1 m duct can short-circuit across the opening and leave the lower volume essentially unpurged, so ventilation must be delivered to the bottom and then verified by measurement.
On the respirator: a half-mask air-purifying respirator has an assigned protection factor of 10 and relies on cartridge capacity and no breakthrough. It is not appropriate where the concentration is unknown, stratified or could rise into the immediately dangerous to life or health range of 100 ppm. Hydrogen sulfide also deadens the sense of smell at higher concentrations, so odour is no warning.
The 10 ppm instrument alarm is a workplace exposure threshold, not a safe-entry limit; many jurisdictions set a 15-minute short-term exposure limit of around 5 ppm, so even 22 ppm is already an overexposure. Continuous personal monitoring with a lower alarm, supplied air and a rescue capability are the defensible controls.
3. The supervisor's 'we have done it before' argument is raised again at the gate. How do you answer it, and what do you document?
Read the worked answer
Answer it directly: prior successful entries are evidence that nothing went wrong before, not evidence that the hazard is controlled now. The sump is a live sewer structure; biological activity generates hydrogen sulfide, and the concentration changes with flow, temperature and time of day. Relying on history is normalisation of deviance, and the fact that the previous entries were made under the same single-sample permit means the earlier risk was never actually measured either.
Then escalate rather than argue. State that you are exercising stop-work authority, that the entry will not proceed under the current permit, and refer the matter to the entry supervisor and the general contractor. If the supervisor persists, go up the client's line and record that you did.
Document: the stop-work notice with time and location, the readings at the rim and at 2 m with the meter serial number and calibration date, the permit number, who was notified and when, the reason the permit was withdrawn, and the conditions required for reissue. Retain the record in the project HSE file. The schedule pressure is real, but the consequence here is a fatality in a confined space, and the professional duty to protect health and safety takes precedence over a four-hour service window.