Podcast: Revalidating Process Hazard Analysis: Getting Real Value

In this episode, Trish Kerin and Traci Purdum explore process hazard analysis revalidations and how to make them more effective. Kerin explains the difference between redoing a PHA and revalidating existing assessments, introducing the Delta HAZOP methodology that focuses on creeping change in facilities. She discusses triggers for revalidation, from legislative requirements to significant operational changes, and emphasizes the importance of selecting the right team and methodology. Kerin highlights how AI and machine learning can help gather data and identify trends, while stressing that human expertise remains essential. The key takeaway: approach revalidations with rigor and discipline, not as tick-box exercises, to truly identify hazards and manage risk effectively.

Transcript

Welcome to Process Safety with Trish and Traci, the podcast that shares insights from past incidents to help avoid future events. Please subscribe to this award-winning podcast on your favorite platform so you can keep learning with Trish and me.

I'm Traci Purdum, editor-in-chief of Chemical Processing, and joining me, as always, is Trish Kerin, director of Lead Like Kerin. Hey Trish, how are you?

Trish: I'm doing really well, Traci. How are you?

Traci: I'm doing all right. What's the most interesting thing you've learned lately?

Trish: Oh, there's been so many things. Sometimes my learnings are really just reminders of things I've forgotten. You work on a project, need to get people on board and engaged, and suddenly you remember: Oh yeah, I needed to do that consultation or ask that question.

Other times it's brand-new things — like some really fascinating renewable energy technologies being developed around the world. There are truly interesting ways of capturing solar energy using standard mechanical means and creative ways to store that energy for future use. It's a very exciting time.

Traci: It is, and it seems like every day something new is happening. It fascinates me how engineering minds work. I appreciate learning from everything I read and hear.

Today we're talking about process hazard analysis revalidations — specifically how to prepare effectively, assemble the right team and leverage technology to make the revalidation process smoother and more comprehensive. I'm going to throw some questions at you, and hopefully we can learn a lot today.

Trish: Yeah, go for it.

Why Do PHA Revalidations?

Traci: Are there triggers that necessitate a revalidation?

Trish: Yes. There are several reasons to revalidate PHAs performed in the past. In some jurisdictions, there's a requirement to revalidate at certain periods — usually about every five years. That's generally well-accepted practice, even if it's not legislatively required.

But there are also operational and change triggers. If you're making significant changes to your facility — I'm talking substantial changes, not the smaller ones managed through management of change — you should do another PHA to understand the impact.

And regardless, hazards can emerge over time that we didn't anticipate, and technologies to deal with those hazards change constantly. You really do want a reasonable cycle ensuring your PHAs still give you valid information and actually understand your risk.

Scoping Change

Traci: How do you effectively scope what has actually changed?

Trish: That's a really good question, and this is where we can get caught up. There are a couple of different ways to revisit PHAs.

The first is what we call a redo — where you basically start again with a blank PHA and do the whole process from scratch. Periodically that might be useful, especially if the facility has changed enormously over time.

The other is a revalidation, where you go through with your last PHA report and validate it against what you've got. You're not starting from scratch — you've got all this historic information.

One challenge with either approach is that after a period of time, you hit diminishing returns. You start to see less significant risk uncovered because you've previously dealt with it. In theory, you've identified it, put actions in place, done those actions and reduced the risk. So a lot of companies struggle to continue this cycle where they're putting enormous effort into work that's not getting new results.

That led us a few years ago, when I was leading the safety center, to develop a process we called Delta HAZOP. It was originally an Exxon Mobil process they very kindly gifted to the safety center. We brought together different companies with their experiences and created an enhanced version.

It's not actually a HAZOP — it doesn't use guide words. But it does look at different nodes in the process and prompts you with more preparatory information. The amount of data you capture before the workshop is higher than for a traditional HAZOP, but that makes the workshops shorter and faster. You put extra effort in at the start with one or two people gathering information, and you don't have three weeks of eight people sitting in a room together — maybe four days instead.

But more importantly, it can really uncover something you wouldn't necessarily find with a standard HAZOP redo or revalidation. It pulls out hazards that have crept in over time, as opposed to hazards in the original design, because it focuses on what has changed.

That's not just looking at your MOCs — you put MOC information into any HAZOP anyway. It's actually gathering data on the creeping change that has occurred in your facility. Creeping change happens every day. Equipment wears. The pump is no longer operating on its curve from when it was installed. The response to that alarm is now slightly different. The shutdown you built in, anticipating it would operate once every six months, now operates twice a week.

There are many different tools and techniques for PHA. You really need to find the right method for you — whether that's a redo, a revalidation or a Delta HAZOP. It's about picking the right methodology and applying it with rigor and discipline.

Assembling a PHA Revalidation Team

Traci: Does it make sense to have the same team do the revalidation?

Trish: I'm going to say yes and no. When you've got people who really understand the process well and were there at the start and all along, they bring unique depth of knowledge to the assessment. But if they were part of the original design or even the last HAZOP or PHA, and they think they got it right, they might get defensive if you challenge that.

So there are advantages and disadvantages, and you need to weigh that up and look at the organizational culture. But fundamentally you have to have people who understand the process you're looking at.

It may or may not be the original team. If you're likely to have defensive posturing, that's not going to give you a good safety outcome. You need people with open minds who are willing to explore and challenge.

Sometimes it might be useful to have people who were not in the original HAZOP. But they still have to understand the process and be the subject matter expert for their field.

If you have the opportunity to expand the experience and get different people involved, that's also a great development opportunity for your employees. It's good to spread it around. It's development for them. It lets them understand a new focus in their organization. It gives them new skills and helps them understand hazards more deeply, which can result in safety improvements.

So no straightforward answer. Mix it up a little bit. Manage the culture at the same time.

Traci: You mentioned picking the right methodology. How do you do that?

Trish: You really need to understand what the objective is. Are you trying to just check whether you've still got a safe system, or are you trying to uncover something deeper? Or are you just doing it because it's the scheduled time? You still go through the process seriously and try to uncover anything that may pop up.

Make sure you've got the right objective set at the start, and that will help you choose the right methodology. Then rely on your process safety and risk assessment experts — the people who do these assessments regularly. Take their advice. If they say you should be doing this methodology and here's why, and that's reasonable and they're your subject matter expert, we should listen.

They're not infallible — they can still be challenged and make mistakes. But on the whole, we want to take their advice, not just say from an operational perspective: No, you can't have any operators, you can't have maintenance, you don't have any time, there's no budget, just make it happen. That's a problem in an organization if we're not putting the right resources into doing the assessments. Listen to what your subject matter experts are saying. They're telling you things for a reason.

PHA Technology

Traci: We've done podcasts on utilizing AI in these types of situations — in the hazard analysis. Can we talk about how technology has changed revalidation preparation? Are there valuable aids we can look at?

Trish: I think the emerging technology of large language models, machine learning — what people commonly call AI — has great benefits in helping you prepare for assessments, sorting through enormous amounts of data and pulling together clear threads and trends we can now see. This is where this technology has amazing ability to help us.

What we need to be careful about is not assuming the technology can do our PHA for us. It can't. We need the human intervention, rationale and approach to doing these activities. But we can save a lot of time and effort by using AI technologies to collect and do some initial checking of data, even mapping it out in certain ways so that we as humans can use our ability in sense-making to see what it's telling us.

It can also be really good at highlighting important pieces of information we need to know — things like incident history in our facility, our company, our industry — so we can get better information going into the assessments to start with.

You'll recall on one of our podcasts a while ago, we had David Jamison from Salus Technologies, and he likened some of their AI methodology for HAZOP or PHA to that wise engineer on the shoulder saying, "Have you thought of this? Have you thought of this?" That's always stuck with me as a brilliant way to explain what we want to use this technology for. It's not to do the work — it's to help us do it better and more efficiently.

Traci: I think it would come in handy too with that creeping change, right? It can see that creeping change more readily than we can sometimes.

Trish: Yes, because it can see aspects of trending that we would not necessarily notice. It can look at enormous amounts of data in a relatively short period that would take a human a lot longer to map out and analyze, and it can present it to us in ways that we can look at and make sense of.

It can really start to show that creeping change — that we actually have a slight trend change on this particular parameter that corresponds with something else happening now. But correlation is not causation, and that is something we need to be very careful about. Just because two things occur at the same time does not mean they're related.

For example, the classic: I'm in Australia at the moment. We've tragically had a number of shark attacks around our beaches lately, and there is a correlation between shark attacks and ice cream sales. They're absolutely correlated. But there's no causation between shark attacks and ice cream sales.

Traci: Unless they're out of chocolate ice cream, I guess.

Trish: Right. We need to understand that just because two things appear to have a correlation does not mean they have a direct causal link.

Challenges in PHA Revalidation

Traci: We talked about one challenge — the potential for people to get their feathers ruffled. Are there other challenging aspects of addressing these previous PHAs?

Trish: Yes. One of the key other issues is the biases we have as we go into the assessment. We might look at the report, the previous information, the team that were there and make an assumption that it must be pretty accurate because they were pretty smart people. Or likewise, we might look at it and go: Oh, I don't trust this information because they're not the brightest ones we've got.

Either way, we're making a judgment based on no useful information. If we think they were pretty smart people, they must have got it right, we may not challenge it as vigorously as we should. We may assume things are correct when they're actually not.

That's one of the big challenges with revalidation — it has this information presented to us, and it can lead us to say: Yeah, that's still right. Yeah, I think that's still right. And not actually say: Do we actually still have the same context here? Has something changed that makes this result no longer valid to us?

That's one of the biggest challenges — we can either assume they got it all completely right, so there's really nothing for us to do, or assume they got it completely wrong and we need to redo the whole thing. We need to look at the information, take it in, and be willing to challenge and say: Well, is that still accurate? Where's my evidence that tells me it's still accurate? Not just going on: Oh yeah, I think it is. No, I know it is because — where's my evidence? And I think that's what we need to really focus on. If we're going to make determinations on something, we need to be able to back it up with evidence.

Traci: Sort of bake in the challenge points — the poke. Let's have a session where we try to poke holes and put everybody on that same even field that we're not poking holes at you, we're just trying to poke holes in everything we're looking at.

Trish: Absolutely. De Bono's Six Thinking Hats can sometimes help with these sorts of conversations and challenges, where you actually give people specific roles in the conversation. One of those thinking hats is the black hat, and the black hat is there to almost be the contrarian and challenge everything, just to make sure this counterpoint is put on the table so we can with evidence either back it up or dismiss it.

The other thing — you're very right — it is not about the person, it's about the process. One of the sayings we have in Australian rules football is: You need to play the ball, not the man. Go after the process, not the person. I've seen that used in risk assessments I've seen done in Australia. It's like: No, we're playing the man here. This is wrong. Bringing people back to play the ball, not the man, and focus on the process, not the person.

Traci: Trish, is there anything you want to add?

Trish: My concluding thought is: Focus on going into your risk assessments with discipline and rigor. I've said that before, and that's about making sure we don't just assume these are tick-box exercises we're doing because the law says I have to do it every five years.

What we want you to be doing is actually challenging that data and looking: Is there a better methodology? If you don't think you're getting value out of just doing the same process every five years, go find a better methodology. There's many different processes out there. Go take a look at one of them and test it in your facility. Test it with your regulator if that's what you need to do, and see how you can get to the point of getting value out of putting people in a room together, because it's very expensive.

You need to make sure you're getting value out of it. It's not just a tick-a-box activity — it's actually there to identify your hazards and consequences and to manage that risk, so hopefully you don't see those consequences come to fruition.

Traci: Well, Trish, you are here to help us drive home that we need to rely on our subject matter experts, that correlation isn't causation, and sharks and chocolate ice cream don't match up, and to talk about the process and not the person. I appreciate all of the wisdom you have bestowed upon us today.

Unfortunate events happen all over the world, and we will be here to discuss and learn from them. Subscribe to this free podcast so you can stay on top of best practices. You can also visit us at ChemicalProcessing.com for more tools and resources aimed at helping you run efficient and safe facilities.

On behalf of Trish, I'm Traci, and this is Process Safety with Trish and Traci. Thanks again, Trish.

Trish: Stay safe.