Panel Discussion on Managing Operational Risks with Dow, Marathon Petroleum, Plasco and ExxonMobil

My name is Renee BL, I work at Cisco. In the case of the Marathon or ALS project, I would like to know out of curiosity how many alarms do you get, and of those, how many false alarms are there? So how accurate is the system?

Hi, I'm Sid Snitkin from ARC. This is for John Sampson. John, I hope I say the correct name. I was fascinated with your talk and this issue when you talk about resiliency. I'm just wondering how, if that's in a sense kind of like business continuity planning. I know it's not exactly the same, or you may have a different definition for that. But that's my question: how the two would be compared.

Thank you. Can I ask a follow-up? Not a follow-up, but another one. I just have several questions. Patricia, if you want to comment or somebody else, that's up to you. But my second question was the idea we talk a lot about in this conference. We've had three different sessions on risk management, and we also kind of see this idea of all of these programs have to roll up in the end to an Enterprise Risk Management program. And I'm just wondering, it's easy for me to say that, it's conceptually easy to say they should all roll up, but what does that actually mean? When you get up to the corporate level, you talk about enterprise risk, is there somehow that some measures of what the risk, for example, of a plant having a disruption, someone's taking risk down at a facility that's going to affect your ability to deliver products and things like that?

And triggers would be something like the likelihood of an event has risen to a higher level of likelihood or something like that?

And could I just one more? I'm sorry, I don't want to... well, you and I can go outside, we can always come back to you.

Oh, that's fine. If there's somebody now, go ahead, we've got time.

No, I was just going to ask Patricia, you know, from Exxon, because I know that's not your specific job like John's, but I'm sure you're aware of what you do. I'm just wondering how you do that kind of stuff at Exxon.

Well, I think similar to what was just said, that depending on the type of risk and the type of business, there is a hierarchical, if you will, structure in place. So if you're talking about safety risk or financial risk associated with a project, our project management system would require risk assessments following a certain procedure, and depending on the risk, it takes different levels of management sign-off. Then when you start going more on the product side, you get into product safety risk, you get into financial risk, and again, there would be a combination of business considerations, sheet considerations, supply chain considerations, and those would be considered individually and then in aggregate. And again, depending on the significance, would determine the level of review that would be required. You can imagine when you get on our upstream side where we're exploring, we're producing way down under, way down yonder, tremendous considerations. And so there you've got the safety, you've got the operational, you've got the financial, you've got the social, you've got the political. And Rex Tillerson's a busy man. So, you know, not to take it lightly, we try to empower as low as appropriate so that we don't have all the risk management decisions tied at an inappropriately high level because then it becomes a bottleneck. But because the program and the review process is systematic, we think it's a fit for purpose. But you will find consistency across the countries, across the businesses, across the globe.

My question was for Don and Marathon. Putting that system in the way you did, it makes me wonder that there might have been some privacy issues. Some people that were against it, you know, they wanted not to tell you where they were going to be all day long. And I gotta believe in some countries it's actually a major privacy issue.

Yeah, we did not experience that. Now, there were some things that since that has the ability to track the worker, and we do keep that data, and we keep it for a very short period, a week. And the only reason we keep that data is if we did have an incident, we could go back and when we do the investigation, we could understand what steps led up to that. But I've, you know, we had to make a promise to the workforce that we'd never use that, and we never have. And I think it really comes down, you know, we're a non-union facility in Robinson, Illinois, and some would argue, well, it's easy to implement in a non-union. And to me, it's no different than union. I've worked in union plants and I've worked in non-union facilities, and I would say it's really the trust that the workforce has with management and the type of culture you generate. So if you have a righteous culture and you don't use it against people, then it doesn't to me matter if it's a union plant or non-union. And we just will not use it to track people's whereabouts. And I will tell you that really the acceptance has been very good because they understand that it's protecting them. And so that's been very positive. And in fact, it is on our business LAN. You know, we have very strict policies of separating the control LAN and business LAN. This is on the business LAN. And what's interesting, it's gone down and instantly, you know, they're calling up and wanting to know how come we can't get this fixed and back online.

Question to John. With respect to, my name is Ibrahim Hammed from Dolphin Energy. With respect to the custody of the ERM or the business unit who's interested to follow the risk register and the definition of the risk, the risk tolerance, usually in chemical companies or oil and gas, the upper hand come to the health, safety, and environment. With regards to Dow, how do you manage the risk? ERM, is it under the HSE or under the business internal audit or under the CEO office?

Another question for Don. Did you integrate the system with the safety system or the shutdown system in your solution?

No, the system is not integrated in any shutdown system. It's on our business LAN and it's not on the control LAN. And so it's, you know, obviously it's notification and response, but we don't tie it to any shutdown. The operators would have to elect to follow our protocols if there was a release and shut down a process unit or take the appropriate actions.

Or the safety system just for the operator view?

Yeah, just for the console operator at the central control room receives all the gas alarms and motion and panic.

Is there a plan to integrate it further with the plant, or you just keep it as an island?

Just keep it like that.

Okay, other questions? Steve Banker.

ARC. Hi, this is for John Sampson again at Dow. What strikes me is I'm on the supply chain side. So, you know, five years ago, maybe we talked about supply chain risk management, we didn't use that vocabulary, we didn't talk about resiliency. And there's certainly a whole lot more attention now than there used to be. And there's councils. And I've got to say, I was very, very impressed with what you described. But are you sort of part of any other councils, and what are they? And who do you think are the best companies at risk management in the world based on participation in those kind of councils?

If I can jump in, and oh sorry, P, I was just going to say something that we've talked at a number of different sessions over the last today is about collaboration and partnering. And when you collaborate and partner, you often have more eggs in fewer baskets. And so from a supply chain standpoint, you might be getting your short-term, long-term price advantage, but when you do have disruption for whatever situation, actually the risk is higher. And we actually experienced a near miss with hurricanes in the Gulf a couple years ago. And so I will tell you that our Lubes business, I mean, we provide lubricants, we provide fuels, and those are the products that municipalities need to rebuild after a local situation, whether it's weather or something else. And so we have enhanced the risk management process associated with supply chain disruption to make sure we're taking the right balance of cost and OpEx versus disruption and your ability to recover after the fact. And I think a number of us, as we've shortened our list of suppliers and as we've increased our strategic partnering, those are good words, those are right words, but you got to balance short-term and long-term when you're putting those together.

The question I was thinking of, like I know we're talking about operational risk management through multiple lenses. And I was especially intrigued, and I guess this is really a question for Bill and Chris, because you looked at operational risks from the design and build stage, and having a repeatable process was extremely important to your operation. But I saw something on your screen that looked a little bit like the digital plant. And from what I understand, the digital plant has always been a Mecca, but you also desire to limit the amount of stick builds. So you're using module yards and various sources. Can you talk a little bit about how you're using technology to manage that handover and quality control to support your operation?

It's a matter of being able to have simultaneous engineering shops work with the same product in a coordinated manner. So we'll farm out very specific pieces that are meant for modularization, and those will be designed by that particular module shop, by that particular engineer. They'll do all of the drawings around that. They'll use the same system, and then we'll have an integration engineer that'll integrate all those pieces. So they'll take the product that's designed by the modular shop, make sure it fits into the larger scheme of things. So it's maybe it is a Mecca, but I think that's the only tool that we can use to at least know where we're trying to get to and work through that.

So the question was about compliance versus risk management and how the interactions relate.

I guess I'll take that one to start, and then I'm sure others will have opinions to add to it. There's no question that at all times, under all circumstances, we want to be compliant. And when I say we, I say we, our own employees, our own operators, as well as any contractors or suppliers that we deal with. And that's the beginning. And so yes, at ExxonMobil, we have our systems to plan, assess, and ensure that we are compliant. Any contractors and suppliers that we deal with need to attest that they have processes and systems in place to demonstrate similarly. I will tell you that in the last couple years, there have been situations where you went to a contractor who subbed, who subbed, and it ended up that some of our supplier quality processes did not anticipate fourth-generation subbing. And we found ourselves in some supplier situations where the quality was not what was expected. And so we've enhanced the systems because of that. But I would tell you that there's sometimes where yes, being compliant is appropriate. There's other times where you want to go more than that for various reasons, and our systems will be set at that higher threshold.

So I think we've got time for one last question.

One last question. Question for Don. You may not be willing to answer this, but what was the cost to deploy a system like that? You know, are we talking a million bucks, are we talking more or less?

Yeah, maybe, you know, I probably don't want to share that. But I will tell you that it was, I can tell you on the installation, I can give you a little bit of insight. There is, we optimized the cost of installation. And how we did that is as we designed the location of the APs, we worked with our embedded electrical contractor. So instead of going through a process that we normally would go through, is have engineering develop a construction package, and then from that construction package we issue it out in the field and we go build. But what we instead did is we had Accenture work with our embedded electrical contractor in the placement of the APs, and they really field ran it because the electrical contractor knew where the power was, they knew where the fiber was. And so they basically ran, and then we did as-built drawings afterwards. So I think one of the things that is really nice about that is we implemented it on a fairly quick basis. You know, you can go into a process unit and get it done in a few weeks. And then also we minimized the cost there. And then one of the things we did go back and look at, maybe a different way of looking at it, is when you look at the cost of the hardware in the field, the backend systems, and the monitors themselves, is we sort of tried to break it down into a cost, you know, how much does it cost per worker per hour, that type of thing. And I think came back with a fairly low cost there of, you know, sort of like a dollar an hour type thing to cover each worker.

So a dollar per hour per worker?

Yes.

Okay, all right, that kind of gives me an idea. What are your plans for your other facilities? You know, this is obviously a pilot facility, and at some point you would want to extend it, you know, if you felt the value proposition was there.

Yeah, I think right now is we got it fully implemented, fully functional at the Robinson Refinery. And what we wanted to do is just make sure that, you know, the system performed, was stable, and it has been. So we haven't had any problems there. And so our plans are to roll it out to our other large refineries. Not only first we're looking at our Garyville facility, Garyville, Louisiana, and Catlettsburg, Kentucky. But then I think in the mix of all this, as we got thrown in, we had a large acquisition with our new Galveston Bay Refinery. So I think that's sort of put things a little bit on hold.