The original reference for periodic leak testing comes from 29 CFR 1910.1047 App A, and is entitled “Substance safety data sheet for ethylene oxide (non-mandatory).”
Section VII-2-i of the document states:
Leak detection. Sterilizer door gaskets, cylinder and vacuum piping, hoses, filters, and valves must be checked for leaks under full pressure with a Fluorocarbon leak detector (for 12/88 systems only) every two weeks by maintenance personnel. Also, the cylinder piping connections must be checked after changing cylinders. Particular attention in leak detection should be given to the automatic solenoid valves that control the flow of EtO to the sterilizer. Specifically, a check should be made at the EtO gasline entrance port to the sterilizer, while the sterilizer door is open and the solenoid valves are in a closed position.
Since the source document dates back to 1998, its focus is on external cylinder-based, rather than internal cartridge-based ethylene oxide sterilizers. Still, the reference to a Fluorocarbon leak detector—with no mention of an ethylene oxide detector, readily available at the time—is curious.
Moreover, instructing people to check for leaks, without further guidelines, is ill-advised. Such qualitative leak checks can rapidly deteriorate into a fool’s errand, especially if no corresponding increase in employee exposure is also observed.
Note that the details on leak detection have since been modified, and the following is what now appears in the later (and current) document “OSHA’s Small Business Guide for Ethylene Oxide.”
Section 2-A-ii-c of the document states:
Employers who are required to create a written compliance program because their employees’ exposures are over the permissible exposure limit must also produce a schedule for routine leak detection surveys.
Some businesses that use EtO find it helpful to test equipment such as sterilizers, pipes, tanks, and fittings at least every two weeks to confirm that there are no leaks.
One of the peculiar conceits of the Small Business Guide (from Section 2-A-ii introduction) is the notion that “[A]rea monitoring is complementary to personal monitoring, but can never be used instead of it.” This attitude may derive from decades old and now obsolete area monitors, that had such high ranges that they were only usable as catastrophic area leak detectors.
Technically, area monitoring is not required by OSHA, as all of their standards refer to employee exposure. However, it has long been recognized that achieving compliance solely with the use of personal monitors (or badges) is far from easy, and may not even be possible in many cases.
At best, badges can provide a history of exposure, but allow no instantaneous alarms. At worst, the history will record that the employee was overexposed to the toxic chemical as a fait accompli, and the employer is in violation. Thus, the only “protective” value of badges is as an after-the-fact warning that something is wrong.
As to personal monitors, even if they were available for all regulated toxic chemicals, and they were also outfitted with time-weighted average alarms—which most assuredly is not the case—the employer would still be faced with the matter of archiving all the exposure data.
Absent highly unlikely contrived circumstances such as the employee somehow opening the sterilizer while it is running—yet not setting off an area instantaneous alarm—if the concentration of EtO in the area is shown to be less than the allowable level for employee exposure, the employer is in compliance.
Surely, this is a clear example that area monitoring CAN be used instead of personal monitoring.
For these reasons, the vast majority of employers install area monitoring (often with data acquisition), when ethylene oxide (and virtually every other toxic chemical) is being used in the workplace.
Checking for leaks in the areas mentioned by OSHA is admirable, but such leaks will also manifest themselves as excursions in the area monitoring data, possibly even generating an instantaneous alarm. Typically, once these excursions occur, the source of the problem is determined, and it may be one of those OSHA-described leaks.
It is puzzling that OSHA did not mention another exposure scenario, which is at least as common as mechanical leaks: Poor ventilation design.
We are aware of far too many facilities whereby the exhaust air was inadvertently dumped back into the work area. In one particularly egregious case, the customer complained about the “wrong” high EtO readings he was getting, even after they were verified with two other analytical methods. The source of the problem was shown to be errant ventilation, and, sadly, the matter was not easily resolved.
In our judgment, continuous area monitoring is more than equivalent to checking for leaks every two weeks.