Roundtable Meeting — June 2-3, 1999

Wayne Bissett (continued from Tape 3) One is a legislative initiative. The Canadian Environmental Protection Act received third reading from our parliament just on Monday night. It has a new section on environmental emergencies, which allow us to develop regulations, such as you have proposed as safety management or the RMP world. It still has to go to senate and given that we have a two tiered level of government, the senators want to justify their existence by taking some time to view the legislation. So, it will probably come into play in the fall. The second is a status report that ­ has produced. Unfortunately, although I have a copy here it won’t be released until June 15^th, so I can’t share it with anyone yet, maybe I won’t be let back in the country if I do share it. It has chapters on each sector and how they are doing, it also has chapters on each problem identifying sites that we have confirmed to date and the communities therein and where they stand in terms of process safety management and also emergency preparedness. Obviously, the goal is to join the two of them together. We think by producing this report and trying to do it again every two years it will encourage, promote, coerce some of the communities and some of the companies to move a little faster in light of the fact that there is no legislation. I am sorry for taking so long but I just wanted to say that there are different approaches but they do have their downfalls as well.

Irv Rosenthal: I would like to comment on a number of items. I for one believe that the regulation should be set at the floor rather than trying to define good practice in terms of the statute. I think the system is too complex to define in a rigid sense. I think there is a place for good practice in definitions of it and we have a parallel rule system and devising good practice that can be used as a standard in terms of negligence and other actions under common law would be an important contribution. But I don’t think it is too dynamic a thing to be tempted under regulation, which is a very rigid system. In regard to the point that Fred raised about the worst case consequences, it is not a measure of how safe. One of the advantages is that they take into account the likelihood of an event occurring so that they deal with the impacts of zoning. So you have to convince the authorities that either you meet probabilistic goals and take into account the likelihood and move towards reducing inventories. The problem with reduction in inventories as a means of worst case consequences you get the thing that Bob remarked on, you get firms that play games. They put in small quantities of just one tenth percent below the regulated amount or one pound below it and they get multiple deliveries, so we have to approach that carefully. A report on worst case consequences because one of the routes to safety is to reduce inventories to a level that is needed. Competition in our present system, rigidity that will tend to affect the economics of our processes which will affect our economy and in turn the health of our people.

Dr. Mannan- I am glad that my two-part question stimulated quite a bit of discussion. I am not going to discuss anymore on that issue but I wanted to clear up a procedural issue. If you look around the room there is representation from public interest groups, environmental groups, industry, government agencies, unions, professional organizations and so on so forth. But I am the first one to admit that it probably does not represent 100% of stakeholders, and in fact if we did want to represent 100% of the stakeholders, this U table would have to be set up in the football stadium at Texas A&M University. When you get into a larger group like that it is difficult to have interactive discussion. We spent quite a bit of time determining who should be at the roundtable and how can we make it represent a large portion of the stakeholder group. I dare say that it is probably not as low as 25% of the stakeholders, probably it is higher than 50%, even 75%. Finally, your handout includes two lists: the invitee list and the attendee list. If you peruse the invitee list you will see there are a lot of organizations there that would fit within the small business mode that are not here. I will tell you that the main problem that we face in dealing with small business is that it is very difficult to get them to spare the time to come here. Even if you volunteer to partially or completely fund their trip, it is very difficult to even convince them to spend two days here. They think it is probably a waste of time. Many did not accept our invitation. My challenge to the rest of the group is that tomorrow we will go into our breakout groups and one of the things that needs to be addressed is if we have the right representation of stakeholders. Should the people here today be here or someone else? The reason that I pose the question is that if we do not have the basics then how are we to move forward. So lets question everything and come up with solutions that will really move us in the right direction. I am happy in one respect that it has generated a lost of discussion, but in another respect as John was telling me it might take until 3:00 to get to lunch.

Pat McNulty-- Several people this morning mentioned the evolutionary process which we are now addressing. In listening to Jonathon talk about the genesis of section 112r, I have to conclude that it is also an evolutionary process. The only way to find out if section 112r can work more efficiently than it is designed to work now, is to enforce it and interpret the data that is generated. This morning as Irv Rosenthal was talking he discussed marginal operations and how they frequently fail. Some of the marginal operations are small firms who have limited resources, both financial and human. Bob Barrish mentioned the fact that these are very difficult to ensure compliance. Some of the challenge is how do we ensure compliance with the 67 or 68 thousand firms, many of which are small with limited resources, follow 112r and what can be learned from the data generated so we can modify 112r so that we can continue the evolutionary process. There are two aspects that fascinate me with 112r legislation. One is the process safety program that is a reflection of the safety process that is in place. Although, I can understand the concern about the program, if the safety process is not in place, we are not going to have the benefits of 112r. The two challenges here are how to enforce it and how to interpret the data that we generate.

Fred Millar ­ Particular legal question that is that on pages 13 and 14 in his discussion he mentioned that some people are quite unhappy with EPA’s interpretation of the RMP law. RMP is basically a very small summary of the information which is gathered. John says that some critics of this make the statutory argument that the clean air act actually defines the risk management plan as a giant set of data to be available to the public. In fact the EPA’s regulatory strategy has been, and I have been a critic that has chastised the EPA for progressively weakening this rule as it would a law, but the outcome is do you anticipate that in fact there is likely to be some legal challenges when companies keep information on their plant sites and don’t give it to the public, providing instead a very summary RMP document. Do you anticipate that there may be some lawsuits from communities saying that we have been denied what the law says we should have access to?

Jon Averback - I don’t think there will be a successful challenge to it if some folks wanted to say that more information be made available. The statutes to be implemented by the rule, we work thru the rulemaking process, there was an opportunity to challenge our rulemaking decisions. No environmental group challenged our decisions so I assume they concurred.

Dr. Haskell Monroe ­ Ladies and gentlemen, as an observer it seems to me that we are trying to operate in the real world somewhere between "not in my backyard" and "sell it to me cheap", which I think is the real world and what we ought to be doing. Therefore as we go to lunch I suggest that you sit with a variety of folks and continue this discussion. As you walk out please pick up one of the Bush Library brochures because right next to us is a very interesting facility. Tonight when you go to eat and tomorrow at noon please sit with a different group.

<LUNCH>

Dr. Haskell Monroe­ Our next speaker is one of the superb students of today who is here. I presume that you have reviewed the material on Eboni McCray. I was just chatting with her, she grew up in Greenville, MS, I asked her what she wanted to be doing in 20 years and she said retired. But then she chuckled and said she would like to be in the real world of trying to improve American industry and help it measure up to the goals that it should aspire to meet. She is going to talk to us about accident databases and after she finishes if any of you would like to share with the group as a part of the discussion of her paper any goals that your organization or company may have that relate to this. We do not want a speech on what a great corporation yours is, but we would be appreciative to hear you share any information on how your organization is moving forward.

Eboni McCray ­ First I am going to give you a brief overview of the presentation, a brief introduction as to why we are here today, and the objectives that we are trying to meet. Followed by a description and analysis of the databases that I included in the study, some conclusions which will lead to more questions and future recommendations and finally some acknowledgements. Just a quick note, I do believe that some of my tables were omitted from your binder. We have handed out the omitted pages. Introduction: why have the accident databases become so important? Obviously, over the past 25 years the occurrence of several catastrophic accidents have basically focused the attention on chemical safety from both the public genre as well as chemical industries and government. Because these accidents have involved millions of dollars in lawsuits as well as many lives being lost, we need to report that we have improved safety standards. As a result of the desire to improve safety standards, governments have placed mandates on chemical companies to help provide more explicit safety guidelines to prevent the occurrence of catastrophic accidents such as the Bhopal incident in 1984. Thus, in a rush to compile the best safety guidelines, numerous databases have emerged that investigate various aspects of workplace safety. Because these databases have emerged with little to go on with regard to how the data should be collected, what are the important aspects to looks at, we have to look at what is out there now to decide what it is we need for the future. First we need to look at our objectives as chemical industry representatives. First we need to analyze these accident databases and hopefully develop a safety timeline. Basically what we want to do is find out how far we have come and more importantly where are we going as far as safety is concerned. Hopefully this will give us a better definition or understanding of the chemical industry as a whole. Where are safety practices? Are we doing things that really have made no change? What are our needs as an industry to make sure the guidelines that we are following are really affecting change and preventing such accidents as the Bhopal incident from occurring but also everyday occurrences of injuries and fatalities that may be avoided because our safety practices aren’t what they should be.

With that in mind, I had the task of deciding which accident databases I would include in the study. Let me preface this by saying by no means am I saying that these 8 databases are the best that are out there, the ones that should be included, or the only ones that should be included. What I tried to do is choose databases that offer us a broad spectrum of what is out there. Some are very specific in the type of industry that it investigates and some are very general mandated by such government agencies as OSHA. So in looking at those that I have chosen, the first one being the OSHA Statistics of Workplace Injuries Annual Survey which compiled data from 1974 ­ 1991. Next we have Census of Fatal Occupational Injuries which actually is an improvement of the first one in order to provide more detailed information. Office of Pipeline Safety ­ the Hazardous Materials Incidents database.

Eboni continues her presentation….

Voice #1: Could you characterize why the reporting requirements have changed to capture more incidents/less incidents?

Eboni McCray: Certainly I feel that in reading about the databases that they have worked to capture more incidents. But maybe the organizations that they report to and how they report things have changed - maybe they wouldn’t actually report this to the OSHA statistics, maybe they would report it to the department of transportation, maybe to different agencies. But the point is that we have worked to report more incidents to make sure none of these can basically fall through the cracks. One point I would like to note too, if you were wondering if maybe this is just because the number of people in that industry has changed. If you will look in your binder you will see that I have provided you with some statistics about how many employees are employed to give you an idea of the number of employees working in that industry did not change drastically over that period of time. So we are looking at what I see pseudo real decrease in the number of fatalities occurring.

Delilah Barton: On the contact with objects and equipment vs. exposure to harmful substances ­ would going into confined spaces fall into contact with equipment or under environment?

Eboni McCray: That depends. First of all, going into a confined space does not necessarily mean you are going to be killed by that. Certainly going into a piece of equipment such as a reactor or column you would have to suffer some secondary mode in order to be killed by that. Certainly going into a confined space is not going to kill you, but if you go into a confined space and a worker above you drops a wrench on your head and you are killed, that is something to be considered. Maybe you want to reword your question, but going into a confined space I don’t think would warrant a fatality.

Dr. Mannan ­ I think that what she is talking about is when you go into a confined space and you are exposed to flammable or toxic…

Eboni McCray­ Those would go under actual exposure. I apologize, I didn’t understand your question. Any exposure to vapors or liquids go under the exposure to harmful substances and/or environment. Basically, in a hazardous or harmful environment you would have to have contact with some type of substance in order to be killed. Whereas contact with equipment, we are thinking of things that you are working with actually than exposure that is considered harmful.

Delilah Barton: Would it also include electrical contact?

Eboni McCray: Yes

Irv Rosenthal: Do you have any idea at all what percent of the facilities that fall under fires/explosions/fatalities fall under the OSHA PSM Rule?

Eboni McCray: No, I wouldn’t have any idea of the percentage because the data that I receive is strictly from these databases and unfortunately we get no detailed information about the exact number of people that are reporting each year. So therefore to differentiate between a particular company or industry as far as numbers of people reported these accidents, we cannot calculate an actual percentage.

Irv Rosenthal: Can you identify the facility in which it occurred? And does that facility fall under the OSHA PSM rule?

Eboni McCray: Not for this particular database. These two do not identify the particular facility where the accidents occur. It is a census, meaning that these companies voluntarily report this. So therefore a record of who reported it and what is going on is not provided in this database. Granted if you requested a report, they could provide that information, but in the database itself the names and specific facilities is not noted.

Irv Rosenthal: Thank you.

Eboni continues presentation - Office of Pipeline Safety Database

Lois Epstein: Actually, I have used the database on line and there is a different level that you go to that does provide specific incident information

Eboni McCray: Specifically for 1994?

Lois Epstein: Specific to each incident. I can talk to you about it at the end.

Eboni McCray: That is good information for all of us. Thank you, I was unaware of that.

Voice #1: Do you know whether or not the number of fatalities, is that for everybody, public or…?

Eboni McCray ­ Just for the company, employees, anybody working on site.

Continues w/presentation - Hazardous Materials Incidents Database

Voice #1: Are the fatalities and injuries due to impact or to exposure?

Eboni McCray­ The cause of death is not indicated in this database. Perhaps Lois could tell us if that data is available as well.

Lois Epstein - I can only speak for the pipeline database.

Eboni McCray ­ No, as far as that database is concerned, the actual cause of death is not documented. Therefore, we do not know if the conductor of a particular railway was killed when he hit another train or if maybe there was leak and he was exposed ­ that data is not available in this database.

Eboni continues w/presentation ­ Bureau of Transportation Statistics database.

Voice #2: Can I conclude then that data was available on 5 year cycles for 75, 80, 85…

Eboni McCray: That is the assumption that I made. If you will see in your binder that I have noted that the data begins in 1975, 1980, 1990 and then each consecutive year.

Voice #2: So it would appear that they were on a 5 year cycle.

Eboni McCray: Yes, it would appear so for some reason. Why I don’t know because that doesn’t make any sense to me if you are developing a database.

Eboni continues.

Dr. Mannan: There is a very important point that Eboni is trying to make here that goes back to what Lois also referred to and that is that these databases, we get them and take it as is, or take the upper cut and use that as analysis. But when you look deeper into it, you say oh, okay, now I understand what it means. When she first picked up this database and brought it into our meeting, the first question we asked is where is this number coming from. She brought it back next week and said that is the --- accident. It wasn’t even a chemical hazard but oxygen generator in commercial flight. The issue is that it was wrong usage of transportation…

<Inaudible discussion>

Eboni McCray: There is no differentiation. The people were killed by the crash as a result of the fire. I really don’t think they were killed by the oxygen generator itself. At first glance I would think there was some hazardous chemical that people were exposed to, not that they were on a plane that was carrying oxygen tanks.

Eboni continues with BTS database presentation.

Lois Epstein: Hazardous liquid is petroleum and hazardous substances is the toxics.

Eboni McCray:­ Okay, that was not apparent to me in reading the database. Thank you.

Lois Epstein: Is the data for intra or inter state transportation?

Eboni McCray: It is for interstate transportation

John Susil: Is your property damage normalized for inflation?

Eboni McCray: No, that is a good question. So you would expect it to increase over time. But the magnitude, as I understand inflation to be, the magnitude still suggests that the property damage has steadily increased over the inflation costs as the number of accidents has decreased. So that to me seems disproportionate. The magnitude still suggests a true upward trend and not just inflation.

Dr. Mannan ­ She is more convinced of that conclusion than I am. There is no way to determine if total volume of the industry increased or not. The information is available but it is not normalized.

Eboni McCray: Now looking at the same database as far fatality and injury data. We see that fatalities for hazardous liquids and natural gas show significant changes from year to year. There is no trend and there is nothing that we can really say as far as improvements or changes to the negative. But what we can say that a steady decline in hazardous fatalities is indicative from 1994 to 1996. But again I must say, is this really significant, that in only three years the numbers declined from 21 to 18 and 18 to 15. Is that a true decline? One other note in looking at magnitude, we see that natural gas causes more injuries than hazardous liquids. In one point I was discussing earlier, we assume that pipeline transportation is the safest mode, generally we are talking about safety to human beings. But when we look at the environmental aspect, how many more people will be injured if hazardous liquids contaminate ground water and things of that nature. We need to keep this in mind when we are defining safety and safe measures.

Finally, there has been a general decline in natural gas injuries since 1970. From 233 in 1970 to 72 in 1996. Certainly, from year to year during this time span we have seen fluctuations, but I do believe in general the injury rate has decreased and that is a true estimate.

Question: What happened in 1994?

Eboni McCray: Good question. That data is not offered to us. One thing I wanted to point out too about 1994 is that the incidents in 1994 with regard to hazardous liquids was fairly high. The property damage was almost $100 million. But look at 1994 number of incidents, they are minimal only about 222. There is only 1 fatality and 1858 injuries.

Dr. Mannan: My guess is there was one incident that killed one person but hurt a lot of people.

Irv Rosenthal: But if one incident harmed that many people it would be all over the place.

Voice #1: Isn’t that the case in New Jersey where an apartment complex was surrounding a pipeline and they had a leak and burned the whole place down?

Voice #2: Would an injury equate to evacuation?

Eboni McCray: To me an injury is anybody that was affected health-wise by a particular incident. That could be offsite as well as onsite. I would assume that with over 1800 people they would have spilled offsite.

Irv Rosenthal: The apartment complex that was totally damaged and may have been reported that number of people particularly some of the class action suits.

Eboni McCray: But is that included in the Bureau of Transportion?

Voices: Yes, I would think so as a pipeline

Eboni McCray: Let’s go on to the next database which is National Board of Boiler and Pressure Vessel Inspectors Incident Report. The reason that I chose this one is that it is a very specific database and I thought it would be helpful to see what other databases are out there that are incident specific. Certain boilers and pressure vessels are used throughout the industry so they would be relevant to many industries as a whole. Therefore, we are looking at the vessel types that are reported to this specific incident report. One of them that is accounted for is power boilers, water heating boilers, steam heating boilers, and what they term as unfired pressure vessels. Now, across the board what I have found…<End of tape>