Gas detection for confined space entry: what to test and the order

Confined space atmospheric testing: what are the limits?

The standard defines a hazardous atmosphere as one that may expose employees to the risk of death, incapacitation, impairment of the ability to self-rescue, injury or acute illness from any of five causes. Oxygen below 19.5 percent or above 23.5 percent. Flammable gas, vapor or mist above 10 percent of its lower flammable limit. Combustible dust at or above its LFL, which in practice means dust so thick it obscures vision at 5 feet. Any substance with an OSHA exposure limit at a level that could expose someone above that limit. And any other condition that is immediately dangerous to life or health. A space that fails any one of these is not acceptable for entry until ventilation or another control brings it inside the limits and the test confirms it.

For the four-gas monitor that means: O2 reads 19.5 to 23.5 percent, LEL reads under 10 percent, H2S and CO read under their limits. The common factory alarms sit at those lines or below: O2 19.5 and 23.5, LEL 10 and 20 percent, H2S 10 and 15 ppm, CO 25 to 50 ppm low and 100 to 200 ppm high, depending on the maker and version. The written program may set them lower; it may not set them above the standard’s numbers for entry.

How do you test a confined space, step by step?

1. Bump test the monitor that day, and zero it in clean air outside the space. A monitor that was not shown to work that morning has not tested anything.
2. Open the space and let it breathe, but do not lean in. Test from outside. The entry supervisor stays at the opening; the probe goes in, not the head.
3. Sample at the top, the middle and the bottom, and at any point where the entrant will work. Gases stratify: methane and ammonia rise, H2S, CO2 and propane sink, CO mixes. A single reading at the hatch is not a test. Use a pumped (sample-draw) monitor with a hose or probe, and hold the test for at least the instrument’s minimum response time plus the time the manufacturer gives for the sample to travel the hose, before reading. Appendix B says to test about every 4 feet in the direction of travel and to slow down when a probe is used.
4. Read oxygen first. If it is outside 19.5 to 23.5, stop: ventilate, retest. Do not evaluate the LEL reading in a low-oxygen space.
5. Then combustibles. Under 10 percent LEL is the line. Remember the sensor was calibrated to one gas, usually methane or pentane, and reads other vapors differently; the manual gives correction factors.
6. Then toxics, H2S and CO on a four-gas unit, plus any other gas the space could hold, which may need a fifth sensor, a PID for solvent vapors, or a detector tube.
7. Record the readings on the permit, with the time and the tester’s initials, and show them to the entrants. If ventilation was used, keep it running and test again after it has run; the permit shows the before and after.
8. Monitor during entry. The entrant wears a monitor, the attendant watches readings, and in continuous or large spaces the monitoring never stops. Anyone whose monitor alarms leaves, and the space is re-evaluated before re-entry.

Which monitor does the testing: diffusion or pumped?

The pre-entry test from outside the space needs a pumped unit: an internal pump draws the sample up a hose from the bottom of the space to the sensors at the opening. A diffusion unit only reads the air it is sitting in, which at the hatch is outside air. Once inside, the entrant can wear a diffusion unit, which is smaller and runs longer on a battery, or the same pumped unit with the hose removed. Many crews own one pumped four-gas for the attendant and testing and diffusion units for the entrants. A pumped unit with the pump failed or the hose kinked reads outside air and looks fine; the pump alarm and a block test (cover the hose end and watch for the stall alarm) before each test are part of the routine.