Getting the most from safety inspections

JAY SHELLOGG

I enjoyed reading Graeme Rodden’s article “Know the State of Your Emergency Showers” in the Jan/Feb issue of Paper360°. I read it with a since of urgency, and reached out to a colleague of mine, Carlo Odoardi, to help me write this response.

The first step in correcting any problem is recognizing that you have a problem. In the article, Rodden quotes a presentation made by Larry Kilian of Haws at the recent PPSA conference. The article lays out the problem perfectly: “on average, only 25 percent of emergency shower/eye wash stations work properly and can provide proper first aid – despite weekly or monthly checks.”

In his presentation, Kilian calls out many possible reasons for non-conformance of emergency shower/eye wash stations, including such things as obstructed access, too much or too little flow, or no dust covers. To be sure, there are many more causes of failure, but determining those are better left to the Subject Matter Experts (SMEs)—the users of the systems.

In this article, I instead want to address two points: 1) How to determine what we must do to insure our system does what we want it to do; 2) What tolerable availability we are willing to accept, and what actions get us there.

Tolerable availability
Let’s discuss the second point first. An emergency shower is a very special type of industrial device, because it is a protective device. What it protects is our most precious resource: people. The thing about protective devices is that their failure only matters if the function they’re protecting also fails. That is, the emergency shower only needs to work in the event that someone gets contaminated. This is no different than a smoke detector, which only needs to work in the event of a fire. This sets up a very unique situation: these devices must be periodically tested with a failure finding task to ensure they will work.

One way we can arrive at the testing frequency is by understanding what availability we require and what the failure rate is. Let us start with what I believe is a straightforward idea: The more often we check if something is working, the higher we drive its availability.

As an example, you cannot know if an E-Stop works unless you check it. If you check once a month, all you can be assured of is that it was working (or not) the last time you checked. The more often you check (say daily), and repair any failures, the less time it will be in a failed state. Said another way, if you check it monthly, the longest it can be failed is 30 days; if you check it daily, the longest it can be failed is 24 hours. The more checks, the more availability.
Now think back to Rodden’s article, which tells us that 25% of all showers are in a failed state, even with weekly or monthly checks. That means that, checking every week or month, we find 3 out of 4 showers that are NOT working. With a little bit of applied statistics we can determine that, based on a failure rate of 75% with weekly or monthly inspection, we will achieve a maximum availability of only 63%. This means that only 63% of the time our emergency shower will work.

The question is, “IS THIS GOOD ENOUGH?” Maybe, but not likely. We know that the more often we check something, the higher we drive its availability, but what can we do if increasing the rate of inspection is not practical? We change the failure rate, as the article suggests, with better inspections. Not necessarily more, but better. There is something else we can do—decrease the demand rate on the shower. This means that we train our employees and maintain our process to diminish the need for the shower.

Getting the system to do what we want
This brings us back to point 1): How to insure our system does what we want it to do. This is how we improve failure rates.
On any given site, most emergency showers are of a similar age and construction; as the plant may have expanded and new showers were added, typically these new showers are added in batches. So while the new showers may be different from the old ones, they are similar to their contemporaries.

We tend to think that similar equipment can be managed the same way; however, causes of shower failure in the woodyard, bleach plant, recaust, or paper machines are all different. Have you ever seen a shower in the woodyard caked with lime dust, or a shower in the basement of a paper machine buried in fines, or a shower in the bleach plant obstructed with core buggies? Even though showers are very similar, when they are operated in different parts of the mill the maintenance strategy must be tailor-fitted for each area. In my experience, mills inspect all of the site’s identical showers with the same frequency, using the same parameters. We need to be prescriptive to each shower. This will improve our availability.

This sounds complicated and tedious, but it’s really not. Here is how we do it:

1. We gather a team of the potential users of the shower and anyone else who has something legitimate to say about its management. These are the SMEs.
2. The team builds an understanding of the operating context where the shower is located.
3. The team defines the functions the shower must perform.
4. The team defines the point at which the shower does not meet its required functions (functional failure).
5. The team identifies what causes the shower to fail (failure modes).
6. The team describes the effect of the failure (failure effect).
7. The team determines an appropriate proactive maintenance strategy, or
8. The team determines if a default maintenance strategy is appropriate.

Let’s look back at the reasons cited in the article—no dust covers—and run it through the steps above. One possible solution might be as follows:

• Function: The primary function of an eye wash is to provide wash water of a flow rate greater than 0.4 gal/min and less than 0.5 gal/min.
• Functional Failure: This shower does not provide wash water at all.
• Failure Mode: Dust covers left off following flow testing.
• Failure Effect: Following periodic flow testing, dust covers are not reinstalled. Airborne dust settles and accumulates in nozzles, eventually plugging the discharge. This only matters if an employee needs their eyes washed. Time to repair, 1.25 hours; cost of repair, $87.50.
• Proactive Maintenance: No scheduled maintenance.
• Default Action: Add “dust cover reinstallation” to the flow check test sheet.

It is important to recognize that the solution above is for example only and can in no way be taken as prescriptive. With each step there are several possible solutions depending on the operating context of the eye wash, and this article cannot possibly predict operating context. For example, if the failure mode had been “dust covers missing due to degradation of the plastic”, under “failure effect” we might have noted that, over time the plastic degrades, cracks propagate through the cover, and the cover falls off. In this case, with the help of the team, we have a proactive maintenance task: we look for the cracks beginning to develop. We would set our inspection frequency for half the time it takes between recognizing the cracks forming and the cover dropping off.

Finally, only the subject matter experts—the users/maintainers/owners of the showers—can determine this strategy. Consultants, contractors, and engineers can help—but only the mill’s SMEs are the true experts.

Conclusion
This process at first does seem overwhelming, but a team learns how to process the steps quickly. Ultimately, the output is the best maintenance strategy money can buy.

Thanks to Graeme Rodden and Larry Kilian for the interesting article and presentation that prompted this reply. I also want to thank Carlo Odoardi of COCO NET Inc. (coconetinc.wordpress.com) for helping me with this article, performing a peer review, and editing. If you have any questions or would like any details around this process please feel free to contact either Carlo (carlo.odoardi@cogeco.ca) or Jay (jayshellogg@strategicmaint.com).

About the author:
Jay Shellogg spent the last 16 years of his career working at a large pulp and paper mill, primarily as a senior environmental engineer and maintenance/reliability superintendent. During that time he encountered many challenges; in his own words, “Some I overcame, and some I didn’t.”

For a modest investment of $174, receive more than US$ 1000 in benefits in return. Visit www.tappi.org/join for more details.