Roundtable Meeting — June 2-3, 1999

Tape 6: 6-2-99

Side 1: John Noronha’s paper.

Sam Mannan: I’ll put myself in even a more jeopardizing position by making this question. Let’s assume everyone here is a layman or completely ignorant of what is benchmarking. Given that assumption, if you will lead us through what . . .how we would perform benchmarking on either chemical safety or anything else. Like for example the health example that we got into earlier in the morning. How that would work and why that could be used as a measurement system. I think that’s the key issue that this group needs to understand and then discuss whether or not that will work for chemical safety.

Irv: Sam, let me interrupt you just this sense. Could you define what you mean by benchmarking? Since I’ve used the word in a different sense.

John Noronha: I’m not sure if I can do it comprehensively, but let me attempt. Basically, it’s good to get raw data as each individual sees it, because everybody looks at it in a different light. And then try to coalesce it into some area of commonality of language.

Irv: My use of benchmarking was to establish best practices in a particular area as a benchmark, as a performance standard rather than . . .

John Noronha: That’s a very good example. That’s a very good example. That’s how many . . .

Irv Rosenthal: The best practices, not average or not . . .their consensual. That’s why I use it differently.

John Noronha: Well I think they’re the most formidable practice of benchmarking and that’s how CCPS and FDA and ……. They get the best practices of use by different companies and associations and attempt to get a consensus. And that’s good that’s a very good example. But as you apply [that] to an accident or release rates and safety goals it gets a little tricky. And what we’ve done, for example, in DIERS, I’ve been in that group for 4 years and . . . is to get some basic information on engineering practices, calculations, how much time do you use, how much cost do you use and then get from various companies. And then to develop another questionnaire, you’re gonna send through AIChE and perhaps through this center to about a 1000 potential users. And then try to develop some statistically relevant parameters and the metrics where we can derive where this information and that Eboni was trying to show and a couple of examples that was like . . . but yes it is elusive. You know that’s why we want to highlight, we’re not sure exactly how, that’s some very good information presented and we need to do something. We should have some commonalties. And if there are differences, then we need to explain what the differences are. So, having done that, from that you could perhaps evolve national safety goals.

John Susil: I would suggest and I’m like Irv. We’ve used benchmarking, think of it as best practices. What I heard you describe is more an assessment. And I would suggest you consider not using the term benchmarking because I have some concern that when this is presented in October to the audience at the safety conference that people are going to be thinking benchmarking in the traditional sense. And I know I was lost for awhile here in your presentation because I’m thinking traditional benchmarking, and trying to figure out where you were going. So, that may have been a big part of my problem.

John Noronha: Okay. Sam do you have any comment on that, or . . .?

Sam Mannan: Let me just follow up on that. I think also what this group needs to address is whether or not we go with the definition John is presenting, which is to do assessments. Or we go with the definition John Susil and Irv Rosenthal are saying that they are more used to, which is to establish real industry standards or real benchmarks . . .consensus standards. If I want to relate it to the health example, I mean, we can send out surveys to two million people and they come back with their blood pressure raised, and the average might tend to be something that’s unacceptable. Do we say that’s the benchmark? Well, that’s an assessment of what the average is. Do we live with that? I think that that would not be a good measurement system. I think we have to somehow translate this methodology you have to a real benchmark system.

Irv Rosenthal: Sam, this is not Alice in Wonderland where I’m the red queen and words mean what I mean it to mean. What I’m saying is the word has a predetermined meaning. And if you use it you’re gonna create confusion because the word has existed beforehand and its definition is fairly widely disseminated. If you’re looking for goals to find if you were to use a word benchmark you could say that our goal ought to be at the benchmark. That is the already determined best practice and our goal outta be aimed at best practice. But that’s not a consensus thing, that’s searching to find out that Xerox does this best and DOW does that best. And if you’re gonna do management of change, you gotta go to Celanese. And these establish what has been done by a firm in the real world and this best performance is what we are going to take as the standard towards which we aim. It is not doing an average. It’s . . .the word, I’m just saying, has empirically been defined, it’s been preempted and attempting to use it in a different fashion will just lead to difficulties in communication.

John Noronha: I can see that.

Harry West: To me, benchmark is looked at as the best practices in the industry. Those companies that have the best practices. And I don’t see anything wrong with using the benchmarking concept. I think one proposal would be to write a type of program where there would be, maybe, participation from across the industry by 5 or 6 companies that we feel are the best performers, and then generate benchmarking standards for following the process safety management programs.

Fred Millar: I think what may be common to the discussion is that some kind of search for a standard. And maybe . . .my main exposure to the word benchmark was in the European context where the benchmark study that they did was in regard to quantitative risk assessment was done by the European community. And they . . .what they basically did was have 12 different risk assessment teams look at the same plant, you know, look at one common situation. I guess that’s what they called a benchmark, you look at one common situation, namely an ammonia plant in Greece. And everybody used their different techniques to analyze this client and then the study was to show how much variance there was in the estimates of risk based on the different methodologies that were used on common plant. I don’t know whether that lends itself to the tasks that this group is setting for itself or not. Although I think that, I mean, the common theme seems to be "a search for a standard and how do you go about that?" I don’t think that they came up with a standard in Europe because there was no agreement among the different nations about whether quantitative risk assessment was a good idea given how divergent . . .I mean they varied by a factor of 10,000 in their estimates of the risk of this facility.

John Noronha: I think we tried to describe that concept in out paper, you know in the handout.

Jon Averback: The process you described of collecting broad-based survey of actual practices is not an unusual technique to be . . .when you’re developing a regulation. I think the term benchmark, as Irv has been using it, would be analogous to the talk in a talk down regulatory approach. The talk isn’t always the standard setting the record. There . . .if you’re familiar with the water program there is best achievable technology, best professional judgement, etc . . .in setting standards. And, you know, some standards are applicable when you’re doing something from scratch versus other standards are applicable when you’re applying it to the existing technology. But, the baseline data on potential different goals should be a useful exercise in identifying targets. So I guess I’m, in a long-winded way, endorsing some of your goals.

Pam Kaster: Maybe I’ve missed something, but I’ve gotten confused. Are you talking about a developing a document that policy would be based on our best management practices? Or have we moved away from what Eboni was talking about, databases that the public uses to measure safety?

John Noronha: A little bit of everything. Again this is a on the slide we call it "graph 4" and we make the point that it is very elusive goals but we as John and Bill had suggested. One way to do it… recognizing the word benchmarking to get to develop consensus standards as a more commonly used phrase.

Pam Kaster: I guess a follow-up comment. If I’m understanding what you’re talking about, it’s not a document that’s gonna be as informative to a person like me as what Eboni was talking about. So I guess I’m a little confused on is this the product of the group or will this be one aspect of it?

John Noronha: Excuse me. Benchmarking will be, in particular, part of the overall program it helps different phases of the program. The full phases are critical aspects planning of the design report. Benchmarking, my suggestion would be kind of a . . .probably a general survey, in some cases require a statistical analysis. And it is just a tool to assess the four cases, it’s not a product itself.

Sam Mannan: Pam, if I may make a comment here. I don’t profess to be an expert in benchmarking or even understand the term benchmarking. But, first of all I’ll say this though that benchmarking for something that you have an idea what it should be it’s much more easier. For example, we know that blood pressure 120/90 is what is normally acceptable, normally less risky, okay. So, you can set a benchmark there . . .set a standard there, it doesn’t have to be consensus. But with due respect to my colleague, Dr. Rosenthal, that in a lot of the chemical process safety issues, you will not have these definitive determinations that so and so is the accurate number, okay. That’s why what you have to do is take industry sector by industry sector. Let’s say propane industry and let’s say we pick the top 10 performers from the databases Eboni gives us, okay, that these are the top 10 performers according to that database Eboni has given us. And then we send them a survey of what are they doing in these different areas. And even though we don’t have a definitive standard to compare with, because we know they are top performers, we take those surveys, average them out and then we call them consensus standards or benchmark that we compare the rest of that industry sector with. But the hard thing is that if you take propane and try to apply it to refineries and take refineries and try to apply it to petrochemicals and take transportation and apply it to stationary source you’ll get into trouble. Because the parameters are different, operating practices are different, all the issues are different.

Jim Makris: You know, I’m listening to this discussion of benchmarking, metrics etc. First, after listening to Dave, we have to do, and this is part of the benchmarking process, is to establish what the baseline is. Nobody at this point knows what the baseline is because we don’t know, according to what I’ve heard today, the current state of chemical safety in the United States. A lot of people have some ideas. Some people will say it’s terrible, some people will say it’s great. All of us agree that it needs to be improved, we just don’t know where we’re starting. That’s the first part of it. The second part of it is to somehow envision where we would like to be. Many of us would like to say that the only acceptable ultimate goal is zero accidents. We’ll probably agree on something short of that as something that is attainable in the foreseeable future. Then we need to look at organizations, methods, companies that are above that, significantly above that baseline, or have approached, individually, that corporate goal that we established determine what they are doing to reach that. I think that’s . . .I think that we’re getting tied up. Isn’t that essentially what we need to do? And we need to agree, as part of that, on what things we need to measure. I would assert that measuring fatalities is a little ridiculous. We better be beyond that pretty quickly. And, so whatever it is we need to agree on that. And that’s the whole purpose. Let’s not get tied up on the semantics of whether it’s benchmarking or whatever it is.

Sam Mannan: Benchmarking is just one methodology, we don’t have to use it. Okay, that’s just one methodology out there to establish best practices and we don’t have to use it. We can come up with any other measurement system. But we wanted to make sure we gave this group an example of what’s been practiced out there. John supplied it to, for example, the DIERS relief valve project. Can it be applied to chemical safety? I don’t know.

Jim Makris: I think we need to be very careful when we look at the term best practices, because if we use that and get very prescriptive in what a particular organization has to do. We need to recognize that an organization’s safety culture varies from organization to organization and quote unquote what works for you as far as a particular practice may be such a radical change for another organization it has a detrimental affect.

John Noronha: Good point. I think the term "generally accepted practices" is better than "best practices" and there should be a methodology that is acceptable.

Jerry Poje: Not acceptable, even something stronger than that. But define as good practices. Sort of a catalog of things that a company can do based on what other companies have done that will help them reach the next plateau, whatever that happens to be.

John Noronha: Greg, what would you use at Rohm and Haas?

John Susil: I want to make the point again though, Sam, that . . .and I agree with Jim, we shouldn’t get all tied up in semantics. But none the less, I fear that if you try to use the term "benchmarking" for what Sam has described here of an assessment and see what the average is. You’re just gonna confuse people. The word has a generally accepted meaning or generally understood meaning and I really think that you’re gonna confuse people.

Sam Mannan: It thought I said what we outta do is use the term to establish a baseline, set some goals and then determine the practices that can be done in industry to get there. We can call that some other thing.

Dave Willette: Since nobody around the table is satisfied with where we are, and I would submit that the public isn’t either, that we’re involved with the journey here. And I think what we outta be looking at is auditing the process, measuring the process by which we’re going to get to some end point. The end point has to be determined based upon the particular industry or the particular business or whatever that you’re engaged in and maybe a single methodology to determine what that end point would look like. But I think the real issue is we outta be looking at management systems. I think the CMA deal where you have management systems verification. That that program has an element of what I’m talking about. I think we outta be thinking about how do we measure how well they are doing and how can they report on how well they are doing. And that whoever "they is" has determined the end point that’s becomes acceptable based upon that industry or endeavor and then to the way in which you measure progress.

Ray Skinner: John, I think I agree with David. I think we have to look at three levels: what’s new in practice, what’s a good practice and what is the best. And you always lay out a goal to improve, but you gotta set that baseline. Now David said earlier that OSHA, picked up advice from that group, in a hotel down in Crystal City and if you talk about accident prevention, that’s really where you gotta start. So we have the OSHA standard modified by the RMP rule. That’s almost a new practice for us, that’s compliance, that is our baseline, then how do you decide that you want to move beyond that baseline.

Tim Gablehouse: I’m sort of sitting here trying to rack my brains, how I would report this conversation to citizen groups back home, the fire department, emergency response and entity and all them folks who are awfully interested about these issues, are they safe, are visitor community safe, what in the heck is, what isn’t risk? And while I very much respect and appreciate the conversation about, you know, top performers and specific methodologies to achieve those kinds of levels of performance and stuff like that. I mean the inherent question that I’m gonna get asked is "When am I okay? When am I safe? When are my kids safe? Is my school safe? " And it’s not an in-plant issue, it’s an offsite consequence issue. So, when Sam talks about top performers, I have in my mind, you know, top performers from the standpoint of making my community feel safe. Making my community feel like "here is an entity that respects my desire to have reduced risk in this community." Is this a community that is responsive and participates in my emergency response activity and my emergency planning functions? Is this a community that makes its workers feel safe? Do we have open access to information? All that kind of stuff. I mean, I don’t care if you call it benchmarking or widgets, but if what we’re talking about is trying to establish, you know, performers and figuring out what works to achieve high levels of performance. Let’s make sure we that we keep in mind that an awful lot of this is an inherently subjective thing. And who is a top performer in helping the community manage its overall risk and making the community feel safer and reducing the risk outside the plant fence is always going to be in the eyes of the beholder regardless of what the internal accident rates are. And that’s materially a different "beast" in a lot of ways. And what I would ask, you know, this group and to a certain degree, preempting what I was going to say later, but what I would ask the group to keep in mind is that to the degree that y’all focus on, you know, internal accident rates inside the fence boundary kinds of stuff, you’re not addressing a great big other audience out there that has to be addressed.

John Noronha: By the way, in the report we were trying to select the words . . .to use the words subjectively. We’d rather use the word qualitative rather than subjective. If something is purely subjective then you really don’t have a statistical inference versus objective.

Irv Rosenthal: I think we have to recognize, or have to try to come to grips. Are we talking about measures of the individual facility feeling safer or are we talking about public health issue? In a public health issue we’re talking about a statistical movement and may not offer comfort to any particular industry or individual in that population. It seems to me to be talking about setting a national goal you could set it as a minimum standard for any facility and hopefully that’s what’s been addressed in the regulations. And perhaps we need to define, because these regulations are not performance generated. They are process. In other words you need to go through a process hazard analysis. It can be a HAZOP, a What-If, a fault-tree, a checklist or something equivalent. It does not specify the quality of them and we could talk about specifying the quality of what is a reasonable program. And we could specify on the particulars. I think, however, in setting national goals we have to also consider this as a public health issue. The statistical occurrence of accidents is probably of insignificant factor to anyone individual. You know, they’re not that frequent that we look at deaths per se. They are not the most significant part of the world. They are interests of public health just as some other disease. And those cases we have to set the goal in terms of what we think is achievable with the resources we have to bring to bear on it. What we can offer for the individual, if we wish to set goals as what constitutes good practice or best practice. We can do that in a set of exercises, but it’s a different thing. My only reason for dealing with the semantic issue is that unless we’re clear on what we’re talking about, we can’t have reasonable discussions. And therefore, I ask for a definition of what we’re going to do and what the goal is. And I think there are different goals. If we’re going after improvements, and I can see setting a standard, setting a goal to improve. To improve process safety management systems by ten units as measured by the CCPS Process Index. There is a goal you can measure it now, you can do a statistical sample, you can do your advertising, you do your promotion education and go back and measure it. Any goal we set has to define the measure of instrument, the starting point, how we’re going to sample, how we’re going to determine, and if we can’t do, we’re out of luck. That is not a benchmarking exercise. If we want to assure a facility that they are at least above a good practice and then above best practice, we have to go through exercises of surveying firms and good practice is again a term of art. You have to define that in some way. It is not clear what good practice is as the lawyers know. Best practice has been defined by benchmarking. You can go benchmarking and find, by consensus, that practice in each industry for each facet the process safety management and risk management program and define the best in that industry as the benchmark for that segment of the program and measure people against it.

Irene Jones: One thing I recognize from many of the comments that have been made throughout the day is there are certain themes and certain issues and certain concerns that have been stated and then repeated. And so, what I was wanting to know if, from the point-of-view of Haskell, is there a way to start gathering in a focused way, certain of these things which have been repeated issues that will then serve as a framework or a skeleton upon which we start hanging certain of the specific items that come out of this meeting. In order to move us forward to make some progress forward in the actual collection of many of the more poignant concerns, comments, and issues, which we’ve heard during the day. I am thinking about the meeting that was held in Washington, in December that was facilitated very professionally and very productively I might add by the Chemical Safety Board, in particular Dr. Jerry Poje, and many of you were present at that meeting. And one of the most impressive things that occurred to me at that meeting was that we had as many divergent positions, passionate compelling arguments on either side. But, by the time the afternoon had started, people were starting to shake their heads around the table in concurrence that certain things were being said again and again and again. And we started to consolidate some of the similarities of these issues and those served as the focal point for the report, that was then very professionally done, by the Mary Kay O’Connor Center that then served as the basis for the presentation that Jerry was able to make to the Senate. This entire process I see to be repeated here, this very important group of stakeholders. The collection of this information, in my notes that I’m trying to make as we go forward, are not coming together yet. And they’re not coming together yet because I don’t see us visually focusing on things, and forgive me for that’s the way I think. But, I keep seeing that there are categories of things that we keep on saying. And I would like us to start, if we could, putting those down somewhere. I mean, I don’t mean to know how to facilitate better than you, because please, I don’t. I don’t want to accept that responsibility. But, I would say that if we could try to do that I don’t think we’re gonna go over the same area again. What we will find is we will have more compliments of that item because we naturally will gravitate toward doing that. Does that sound like something that you folks would be interested in doing?

Jim Makris: I think you’re right on. You called it get more focused, right? Fred Millar said, you know in kind of a friendly way to all of us "you’re not very serious about this." You now, it seems to me that those are two sides of the same coin. It is that we are wondering in some ways and we are . . .but we’re all committed to where we’re trying to go. You know, I really do believe that maybe we needed to go through this "blood letting" or this, at least, open discussion and relationship building of today so that we can do very much the sort of thing that Irene just spoke of as we move through this tomorrow. You know, as I see tomorrow, the schedule is much more interactive, it is obligated to be interactive, it’s powerfully facilitated by Haskell. I see flipcharts all the around the room, I suspect tomorrow we’re gonna raise something, like one of these pieces of paper and I suspect that’s part of the plan. So I . . .my own guess is that you’re right on, Irene, Tim, I thank you too. And I’m hoping that when we go into tomorrow with all this new background we’ll be in a place to say let’s, you know, let’s get moving toward our objectives.

Lee Feldstsein: I had a comment which, I guess it may . . .more have been said better before Jim summed up his. This is Lee Feldstein, National Safety Council. I was in a similar quandry as Tim and when I get back to the office I’m gonna to have to talk to my boss and maybe some National Safety Council members. And say "What did you learn here? What happened? What does this mean to us?" And, I guess I was starting to get lost in the benchmarking and I wasn’t very sure what it meant in terms of a product. And I have a sense of what some of our members would want and that is "Well, how do I compare against other people who do . . .who are good? And what do I need to do to improve?" There are all sorts of implications on that but never the less, I mean, that’s something concrete, something that they can see where they stand. And you can categorize it by industry, by geography, you know, a number of other important criteria. But something that folks can kinda see where they are and where they need to go.

Tim Gablehouse: I wanted to emphasize that , I would be stealing my thunder actually, I’m thinking about like the small business guy. I make fluorine but I don’t have the resources of John Susil, or Eastman Kodak. I thought benchmarking was basically a method by which the larger companies, the government and so forth would provide me with some information as to how I should do a certain thing. Let me give you an example. I don’t know how to . . .how often should I recalibrate my relief valves on my . . .on a tank? Or should I use third party people to do it? All these other kinds of details that are part of this, not exactly best practices, because John’s way of doing it may be one way and Eastman Kodak’s another. And I choose. Isn’t part of the goals that we’re trying to talk about is being able to take the benefits of the safety practices that have been learned by these larger companies and provide them in a simple way to smaller companies. That’s what I thought benchmarking would facilitate us doing, and I basically agree with the National Safety Council because that’s what our stakeholders are looking for.

Pam Kaster: I guess just a question. In order for this conversation to make sense to me I need a way to understand what the cumulative, or to borrow another word, what the synergistic effect or impact of all of these practices are on my community ‘cause I have so many fence-line neighbors. So that’s an important component.

Irv Rosenthal: I’d like to go back again to say we have a fundamental issue to face here. What I heard from a friend in terms of making Chlorine is we could attempt to provide technical service to individual people or to find what our good practices or best practices for the millions of operations that compose the chemical industry, the processes. And I think that’s an impossible task and not appropriate for this group. That’s appropriate for technical associations for the Chlorine associations for consultants for a variety of groups. What we might have a goal is to find the means of seeing that that type of assistance was more readily available. And if we did that we might hope that this would improve the process safety practice and contribute to an overall reduction in goals. It’s a public health measure. I think that we have to look at the issues here that we’re talking about national goals which . . .and a means of furthering those goals. And a national goal by its very sense has got to be approached in a public health way of thinking.

Jerry Poje: If I could just add to that comment. I think we’ve already stated pretty clearly the importance of baselining, of establishing good practices and then trying to move towards a benchmark of what is the highest. That clearly is something that could be applicable but in a different fashion to any sub-sector that we have under consideration. Clearly the need for all those sub-sectors to participate in such a process, I think is an important goal. Now how it gets driven by this small audience is . . .it’s not so easily observed to how what we could do to recommend in that arena. But I think establishing this conceptual basis that it is the most responsible thing in any sector to have an understanding of baseline, to have an understanding of what is good practice. And to have an understanding of the need to constantly re-invent that highest end of practice and to communicate that to all the stakeholders who would have concerns as Pam would is quite important. I think it’s worthy of us being challenged to come up with some practicalities on how we would mobilize ourselves in the next day and a half to completion.

Jim Overman: You know I was cautioned to stop talking about DOW, and I’m sorry I’m from DOW and that’s what I know. But if you look at section 17 that I turned in, we set very specific goals for our company by the year 2000. You look on the second page of that we said we would reduce the injuries and illnesses per 200,000 work hours by 90%, that’s OSHA recordable illnesses and injuries. Is that the way everybody should do it? I wouldn’t begin to claim that, that was an issue for us. If you’re . . .this is a worldwide goal. If you’re in a facility that had 10 of ‘em, when we set these goals, you can only have one. If you’re in a facility that had 100 of them, you can have 10, otherwise you make your goal. And the company doesn’t make its goal. We said we’d reduce the loss of primary containment instances, leaks breaks and spills by 90% corporately worldwide in a 134 different manufacturing sites, that’s a 136 per year and we defined what a loss of primary containment was. We did the same thing for transportation incidents per 10,000 shipments. Process safety incidents, that’s fires and explosions. Motor vehicle incidents. And we’re still looking at how to get our hands around incidents with DOW products at customer facilities. We don’t know how we do that. But we’re looking at it. My challenge is that each person in this room needs to do the same thing for his or her facility or his or her thing. And if we do that we can come up with a consensus of national goals. That or reduction of incidents in 10 years by 90%, if so, let’s find out where we are now and let’s take 10% of that number and say by 2010 that’s where we’ll be. And if you make it, if you contribute positively, if you make 9% or 10%, you’re good, if you make 15%, you’re hurting the average. It’s real simple. It’s not that complicated, if we’ll look at a goal and go get it and do it.

Jim Makris: I love it, he always says it in a way that we can all understand. On the other hand there is one sentence that he didn’t include. And that is: what is the number now for the 140 plants at DOW . . .what is the number now and then what will the number be after it is over and how can we all know that it occurred? It is not a problem only for DOW. I mean the whole issue of data today, the whole Eboni point, was how can we know where we are? So that we can agree to that kind of a goal and that we can say we can make it. And who’s gonna be the honest broker of the doubt? Who’s gonna be the person that knows that DOW now has 2700, and DOW’s willing to say it and that’s not a good number, and it’s only going to be 270 at some committed point in time. And how are we going to add your number to Monsanto’s and to Sam’s Refinishing Company, I mean that’s the issue and, you know, we’re real close to starting to define what it is we’re trying to do. And Jim, what you said is perfect if we can make it transparent.

Jim Overman: I would add to that that the challenge is public. The data is public. It’s: www.dow.com/environment/goal2005. Okay what we have to agree on is a few definitions: What is a primary loss of containment? How big is it going to be to be considered in the database? It’s something bigger than a drip and something less than a catastrophic release like Bhopal. Okay, we have to agree on are we gonna use OSHA rules to determine injuries and illnesses per 200,000 people. What are we gonna do? Those things are details that need to be worked out. Once we agree on that, again, I propose that companies that want to voluntarily get in the program. Two, make the results public via Internet, via some other process. And three, give the data to a common source, whether it’s somebody like Sam, CC . . .you know it doesn’t make a difference who it is. We give the data to a third person, follow the rules then there outta be a reward or recognition for that. And it’s real funny, we work for baseball hats and plaques, it doesn’t take a lot to get us going. The firefighter group will tell you that a firefighter will do anything for a baseball hat. Well a company will do almost anything given that the CEO can stand-up in front of TV and get a reward from some agency.

John Noronha: Jim that’s right. That’s what we consider, again, phase one, to collect and assess all available data and to analyze it. Analyze their techniques used, then attempt to have consensus.

Dr. Angela Summers: I’m stuck somewhere between statistics, lies and more lies. How to tell the liars from the statisticians. And we’ve been arguing a lot about semantics. I mean assessment versus benchmarking, injuries and deaths as being goals versus production or property losses. I work every day with companies who are trying to make decisions on process control design and safety instrument and system design and consequence mitigation systems. And somehow they’re trying to track the risk of the process back to what they should be expending in terms of capital dollars on making improvements in the process. And there’s not . . .there’s something’s that there’s been a general consensus on. I think at a point we’ve decided what is an acceptable fatality rate per industry. And that was partially European driven and partially driven by normal statistics of what is your chance of having a fatality in your home and should it be about equivalent at the work place? But when we look at getting deeper into what’s acceptable in terms of environmental release or an impact to the public, that’s a lot harder thing to put our hands around because it’s a lot harder from a statistical basis to have good information on potential impacts. A lot of the scenarios that we’ve been talking about, in terms of worst case scenarios, are truly scenarios that a lot of us don’t believe will happen. And some of us . . .a lot of those scenarios don’t represent risks that we’ve actually seen. I don’t think worst case scenarios represent the Phillips explosion in Pasadena. So I’m a little confused about . . .and I know we’ve been talking about this a long time and I should . . .I thought I was intelligent enough I’da figured it out by now. What are we trying to benchmark? Are we just trying to come up with "what is a national goal that we should only have our predicted fatality rate of 1 in 10,000 years for industry. And that an acceptable loss time injury rate is 1 in 1,000 and that an acceptable band-aid cut is 1 in 100? Is that what we’re tying to set as a goal? I mean Harry was talking about some specific design issues and Instrument Society of America argued for 10 years to come up with a process by which to design safety systems, let alone, actually coming up with anything prescriptive that would tell people how to design them correctly. I don’t think that it’s anywhere within the bounds of this group, or really even within the bounds of the general engineering societies to come up with anything that’s prescriptive. But I’m confused on what we’re trying . . .what the benchmark study is supposed to do other than take a poll and say what’s an acceptable fatality rate for you. What’s an acceptable injury rate for you? What’s an acceptable property damage for you? Is that what you’re trying to come up with? Or are you trying to develop what are the programs that process procedures that each individual company is implementing and did they reach their target?

John Noronha: There you go, that’s what I was leading to. What we did, since about 1990 the DIERS group which consists over 120 companies, identical benchmark, to use the word, so the assessment of their programs. But rather than coming up with objectives and how many releases and how many fatalities, what we passed, each company, what we consider conventional better and extraordinary in the old company. And this is a heating, cooling systems, mixing, whatever, okay, …... And each of them listed what would be extraordinary in my company was conventional somewhere else. So what we did was list all the relative conventional better ……. They’re all good systems. And then we’ll let each company decide on their own the cap of ………. to be used in different designs. You’re right, rather that looking at a prescriptive thing, we should be looking at most generally acceptable, so to balance what I would consider conventional right thing for a given hazard, I would have to use an extraordinary heating system was a conventional cooling system. So that would balance the worst …….. And that’s a good point you brought up. We’re looking at programs, safety programs, and rather than being prescriptive, give them the ….. of what’s available and generally accepted.

Dr. Angela Summers: I think you’re setting a goal for yourself that is going to be very difficult. When I look at any particular company, we could name anybody that’s in the Fortune 500 if we wanted to, even within their own company individual plants and individual processes how they actually handle risk and mitigate risk is completely different.

John Noronha: That is correct.

Dr. Angela Summers (cont’d): Now there is going to be consistency within the corporate philosophy, how they approach risk and how they try to establish programs by which they mitigate risk. The actual details of how they implement in terms of, you know, we can’t even agree in this country whether or not safety systems should be de-energize to trip or energize to trip. And that outta be an easy one.

John Noronha: That is correct. I agree with you. By the way we, that’s a very good point, it took us practically years to develop a position statement to address this issue. I do have a copy of that, in terms of giving a flexible issue of various alternatives. And if you’d like to see it, the language was generally acceptable toward these 117 companies at a time, now it’s a 130.

Greg Keeports: I agree with a lot of what you had to say, I mean, in terms of how one has to be realistic in terms of the tasks you take on. John, you and I have talked for many years about the DIERS process. And I think it’s fair to say that one of the characteristics of the DIERS process is that industry discovered some very important problems by doing a really wonderful set of research over a 10 year period with a lot of cooperative work and discovered some very serious problems. But that the overall process is not very transparent and that is no accountability to public about whether small companies, for example a medium sized company are adopting any kind of DIERS type of methodology to correct their vulnerabilities in terms of Emergency relief systems. So it seems to me that the key question here has to do with do we want to encourage everybody to do their own thing or do we want to have some kind of public accountability. I don’t think we even need to talk about coming up with joint accident data and all that kind of stuff, unless we start to think about that we need to have some kind of public accountability. . .

End of Tape 6