Inherently Unsafe: The Hidden Issues Often Overlooked in Research Hazard Analysis

I suspect everyone knows the concept of inherent safety. Use a non-flammable instead of a flammable feed or make sure the system is designed so as to be unable to fail in a hazardous direction. Often, however, I encounter the reverse situation during a safety audit or a hazard analysis and risk assessment, an inherently unsafe mitigation that is ignored or unrecognized.

Some examples might be helpful in explaining this often unrecognized risk.

·????????A spring loaded relief valve set at 3 psig or less protecting glassware which is highly unlikely to relieve at the desired set pressure and so protecting the glass is problematical.

·????????A gas detector in an exhaust duct that would need an incredibly large release in the hood or ventilated enclosure to alarm increasing the potential for an ignition before detection.

·????????Assuming a glass tube handling flammable materials inside a tubular furnace with heating elements operating above the autoignition temperature will never fail.

·????????A manual operation that relies on the operator to set or hold a pressure, temperature, or flow below an unsafe level and fails to address the potential for human error.

·????????A reactor that must be disassembled and reassembled after each run and where a thermocouple providing high temperature protection must be removed and replaced properly each time increasing the potential that if could be left uninstalled and hence fail to protect the system.

I suggest that all these examples are inherently unsafe and highly likely to result in a much higher risk or an accident the longer they are allowed to exist. Yet many research hazard analysis and risk assessments will accept these “mitigative measures” as sufficient and move on to the next item with little if any discussion. At least until the accident investigation later causes them to recognize the issue.

What can be done to mitigate this type of risk? I think a few key guidelines could do a lot to address the problem.

·????????Always be aware of the potential for human error. Carefully review all manual operations and ask what if the yare not done properly, not done at all, or done at the wrong time or in the wrong sequence. Do not assume that anyone is so knowledgeable, so trained, or so experienced that they cannot make an error.

·????????Look for the numerous hidden manual operations rarely adequately addressed in most research hazard analysis and risk assessments. Cleaning, charging, setting up, taking down, loading, unloading, assembling, and disassembling are all key flags that there is a manual operation that needs to be reviewed. Many of these operations are poorly defined, if at all, during the hazard analysis and risk assessment. They are also areas most likely to be casually modified when they are found to be overly cumbersome or less effective.

·????????Recognize that manual operations are usually the highest risk activities deserving of the closest scrutiny as the operator is near the hazard and the potential for human error the highest.

·????????Personal protective equipment (PPE), such as safety glasses, face shields, goggles, aprons, fire retardant clothing, gloves, and steel toes shoes may all be prudent measures, but too often as soon as the box is checked that they will be worn the review moves forward without assessing if they are enough protection, all the time, for all the possible hazards.

·????????Always ask the tough what if questions about the reliability of equipment. Are we sure the relief device can work reliably under the operating conditions? (The number of times researchers have grudgingly admitted that clogging is “occasionally” a bad problem are legion.) Is there any way the alarm can be advertently or inadvertently bypassed, not installed, or installed improperly? Obviously, an operator who deliberately refuses to follow a safe procedure can always hurt themselves and others. This is probably not a credible scenario. A more credible scenario is can the operator forget to do something, would the operator be likely to want to take an unsafe but apparently efficient shortcut, could the operator be absent or distracted, how much focus is required, or how easy is it to do the wrong thing. These can all result in very credible potential hazards.

·????????Always question if an alarm or interlock will function as intended. Often new equipment with no operating history is assumed to work perfectly all the time to prevent a hazard. Asking how one can be sure of this often leads to easily overlooked issues.

·????????Ask how the operator knows things are set up properly. How do they know the knock out or trap is empty? How do they know the scrubbing solution is still viable? How do they know the tank has enough capacity to hold all the product? Often the answers to these questions are accepted pro forma with out any questioning. The knock out is drained after each run. What if that step is overlooked? How would the operator know? The scrubbing solution is checked periodically. How does the operator know what the period should be? How does the operator test or confirm the scrubbing solution is still effective enough? If there a potential for it to be used up quicker than expected? The answers to further probing often reveals a lot more gaps I the procedure and a much higher potential for incidents later.

For further discussion on some of these issues, you may want to consider some of these articles:

·????????Scared Safe: The Importance of Human Error when Evaluating Research Operations for Safety, https://www.dhirubhai.net/pulse/scared-safe-importance-human-error-when-evaluating-research-palluzi

·????????“My Laboratory is Very Safe.”: The Dangers of Myopic Looks at Laboratory Safety, https://www.dhirubhai.net/pulse/my-laboratory-very-safe-dangers-myopic-looks-safety-richard-palluzi

·????????Administrative Controls in Laboratories and Pilot Plants: When Might They Be an Acceptable Safety Measure?, https://www.dhirubhai.net/pulse/administrative-controls-laboratories-pilot-plants-when-palluzi

·????????“But It Is Only A Laboratory”: The Often Overlooked Hazards Inherent in Laboratory Operations, https://www.dhirubhai.net/pulse/only-laboratory-often-overlooked-hazards-inherent-richard-palluzi/

·????????The Ten Most Common Laboratory Safety Issues, https://www.dhirubhai.net/pulse/ten-most-common-laboratory-safety-issues-richard-palluzi/

·????????Ventilation Dilution: A Safe Way to Avoid A Fire or Explosion or a Placebo?, https://www.dhirubhai.net/pulse/ventilation-dilution-safe-way-avoid-fire-explosion-placebo-palluzi/

·????????“Do You Think We Should Consider This Risk”: The Need for Making Credible Scenarios in Risk Assessment “, https://www.dhirubhai.net/pulse/do-you-think-we-should-consider-risk-need-making-credible-palluzi/

·????????Ventilated Enclosures: Why Do They Often Fail to Work Properly, https://www.dhirubhai.net/pulse/ventilated-enclosures-why-do-often-fail-work-properly-richard-palluzi/