 Okay, I wanted to show you an aerial photograph of the Texas City plant. This was taken from a satellite actually in about six months before the accident occurred. And so what you can see over here in the refinery is a raffinate splitter column. This is basically separating light weight alkanes from the refining process. And of course it's a complicated process, but the key other piece of equipment that I want to point out is the blowdown stack that would relieve any liquid from the vent of the raffinate splitter column. I also want to point out a building here which is a, basically it's a plain steel building that used to house the control room for the plant, for this raffinate splitter unit. And it was, at that point, had just been idled and replaced with a more robust building, a blast resistant building down in this part of the plant. And then this area here in the photo is empty, but over the next few months, British Petroleum, which was the owner of this plant, installed 13 temporary trailers. So these were structures that they brought in by truck and placed on the site and they were used to house a crew of workers that were working on a rebuild in the raffinate splitter area and also in this adjacent area down here which was the fluid cat cracker area. So the trailers were installed here. This is a ground level view of the raffinate splitter column, okay, and then this is the blowdown stack. The other thing I wanted to highlight in this picture is a number of trucks that are on the ground. And I don't know about refineries or chemical plants in India, but in the state of Texas it's very common for there to be pickup trucks that are driving all up and down the roads of the plant. And of course in this case, there are many trucks in this area. Now what happened was on the evening before, on the evening of March 22nd, the evening shift decided to begin the start-up of the raffinate splitter column which is part of the overall start-up of the plant. Now the adjacent cracker unit was still not started and probably wouldn't start for some time, but they had other parts of the plant that they had to operate and catch up. So the column was started by the evening shift operator beginning to feed the regular raw material feed into this column. And the operator did so, of course the first thing that he needs to do is to fill up the bottom of the column. But what he neglected to do was to open the bottom valve from the column. And normally what you really want to do is open the manual bottom valve and put the level of the bottom of the column on automatic control. He did not do that and began to feed liquid to the column. And so the liquid accumulated in the bottom and rose higher and higher in the column. And then the morning came and the next shift arrived and they said hello and goodbye. And the new shift came on, so they did not discuss at the change of shift what was actually happening in the plant other than they had just started up the column. So the new operator arrived and saw that feed was coming in and simply watched it rise in the column. And the liquid continued to rise higher and higher in the column to actually a fairly good height without the second operator knowing that it was actually accumulating. And of course the next step of the process is to start heating the bottom. So they fired up the re-boiler and started heating the bottom of this column. So this is the re-boiler area down here. So the bottom of the column started to get hotter and hotter and hotter. And of course as the liquid is now halfway up the column there's a great deal of pressure above the re-boiler. And so now it is possible to heat the contents in the re-boiler well above the normal boiling point for that liquid. So that continued for a while and then all of a sudden the day operator looked at the control panel and saw that the level was at 80 percent and it's supposed to be at 70 percent. And he said, oh I must have forgotten to open the bottom valve. And so at that point he opened the bottom valve and when that happened the liquid flowed out of the bottom of the column to the receiving tank through the feed bottom heat exchanger. And so now we have super heated bottoms heat exchanging with the cold feed and the feed then started to boil. And so the feed entered the middle of the column and now there's a liquid level above the feed. So now there's a gaseous feed entering at a very high rate into the middle of the column and that big mass of gas just propelled the liquid up out of the top of the column and down through the downcomer and out through the condenser and all the way down to the vent stack. And in a matter of a few minutes the base of the vent stack filled with liquid and then the vent stack continued to fill with liquid and then pretty soon liquid was spraying out at the top of the vent stack. Now you can imagine that liquid spraying out of the top of the vent stack is not a good thing and there was a new operator that looked over at the vent stack and saw that that liquid was coming out of the vent stack and came back to the control room and announced this. He said you must pull the emergency alarm and call for evacuation. But the head operator said no that can't possibly be happening. Meanwhile there are trucks that are driving up and down the street and one truck was seen to be, all of a sudden the engine began to race because it was getting now a much more rich fuel mixture into its carburetor and of course the engine got extremely hot and that caused a flash fire so the liquid flashed back to the vent stack and then exploded. This is some of the damage that occurred. This is actually the base now of the raffinate splitter column which as you saw from the aerial picture was really quite a distance away from the vent stack. So the damage traveled quite a distance and if it traveled that distance across the plant you can imagine here's another view. This is a pipe bridge that was destroyed, turned into a twisted metal. This is the inside of that former control room building, it hadn't been emptied yet. The walls didn't cave in but because of the pressure to generate from the explosion they flexed and when they did everything that was sitting against the wall was bumped into the middle of the room and this is a trailer that contained 12 workers from Jacobs Engineering which was doing the turnaround in the cat-cracker unit and these 12 workers that were in this trailer were killed instantly and as you can see this is reduced to ash and rubble. The principle of level control in this column is actually a pretty simple one. It's been commonly used for many, many years. It's used by differential pressure and the equation is very simple. It's the density times the height of the liquid, that's the density of the height of the gas but of course that's negligible and so you really measure when the liquid is between the sensors you're reading the height of the interface. However, when the liquid goes above the sensor you're no longer reading the height of the liquid, what you're reading instead is the density of the liquid and as you begin to heat, as the bottom of the column got hotter and hotter the density dropped lower and lower and so the level which was reading 100% now started to drop and had reached 80%. In fact, if the operator had waited another interval of time he might actually have seen that the level of the column looked normal and that could have been even worse. There are many things that can be learned from this accident and this is just one page. The official Incent Investigation Report is almost 1,000 pages including I think 200 pages of explanation and 800 pages of tables and graphs and figures. It's a really very good reading that I recommend to anybody who is studying chemical engineering. It may take a while but it's definitely worth reading. First of all, if the operator was following the basic written procedure, this accident would not have occurred because the very first thing that he would have done is open the bottom valve and put the level control on automatic and so the learning for you is that it's not enough to just assume that the operator knows what he's doing but also that a procedure is written and it's written well and that it's followed and that's very important. The second thing was that the first operator didn't really notify the second operator what the status was in the plant and typically what you should do, the two operators should do is sit down and say, okay I have done this and I have done this and I have not done this and I have not done this. Every next step is the following and just the act of having that discussion would probably have brought up the fact that that bottom valve wasn't open. Now the basic knowledge, this is something, the operation of a level indicator, it's something that's taught to operators as part of their basic training and in fact it's even taught to engineers when they go to a plant for the first time because it's not unusual to forget this, you learn this in school but you can forget it too and I know that my first time that I worked on a distillation column a very senior person came to me and sat me down and said make sure you know this. This is something that's very important. The, as it turns out the supervisor for the day shift was called on a family emergency and he assumed that everything was well in hand and he left the plant for an hour just to make sure that everything was well but the, I think it was with his family. Unfortunately he left during the hour that this accident occurred and if he was paying attention you know he should have known, he should have known better. And in fact all you really needed to do was to look at the trend on the DCS and had they done that they would have seen the level was dropping from 100% to 80% and would have asked why and it should have been then at that point become very obvious especially to a supervisor. Of course the emergency evacuation alarm was not sounded, there were trucks driving in the process area and in fact it's against the safety procedures of the plant and of almost every plant for vehicles to be driving especially during a start up because during the start up it's, that's where you discover where you know equipment may not be connected properly, there may be leaks, it's a very good reason to not drive a vehicle and so the truck should not have been on the road at that time. Now management of change, this is a very important thing to know as far as operating a chemical plant because it's our nature, it's as engineers we want to make things better and if we see that there's a need to do either accomplish a task or make something more efficient we want to make that change as fast as we can. Usually sometimes if we don't think very carefully about the nature of those changes something bad can happen and in this case there was a very simple management of change which was the installation of temporary trailers in an area where there was going to be a start up process unit and again temporary trailers are not designed to withstand explosions and so they made the decision to install these trailers without considering the possibility that there could be an explosion during start up and that was a very important thing to to to consider and but you know this is not really a process type of change but you know even if you were going to change let's say feed rates or location of the feed in the column you know the distillation parameters all these things are changes that have to be managed. Sighting guidance. This speaks to industry standards for the distance that you need to keep between operating units and different kinds of buildings, occupied buildings. So if you have a building that's occupied it has to be able to withstand an explosion and the building that was that I showed you that was located below the unit that was you know the that new reinforced control room indeed survived the blast. On the other hand the trailers which were not designed to be occupied were not able to withstand the blast and they should have been located a considerable distance away. Now if you had followed the guidelines that exist the American Petroleum Institute recommended practice 752 and the vapor cloud explosion guideline of of CCPS the those trailers were not sighted they were still sighted too close but had you followed these guidelines and sighted them they would have been able to withstand the flash fire but not the explosion and so in fact they're not conservative enough and we have to go back and make some revisions and actually extend those distances. Another very important thing that you need to do when with the start of any new piece of process equipment is a pre-start-up safety review which is to make sure that that everything that needs to be in place is in the proper place and one of the very first things that you would do at a pre-start-up safety view is ask the question is the bottom valve open okay and so you know that would be the kind of thing that that would need to be done but in addition it looks for you're inspecting to make sure that let's say lined line blinds line blanks were removed or that instrument connections were reconnected or you know basic basic piping what modifications were complete and all these things had to be done but they were not done okay now let's let's go to the management level okay these these are all these people all you know what we're in the business world and they got there through being a chemical engineer and they believed that BP had a very good safety program and in fact they're right BP had a very good safety program but they made they became confused what they what they were confused is they looked at the rate at which employees were getting injured the occupational injury rate and that was getting very very low and in fact was one of the lowest in industries in industry so they they came to the conclusion that that not only is the occupational safety program of BP good but also that the process safety program is good there's two different things one deals with people and the other deals with the engineering the safe engineering of equipment the hardware and the software and and in fact the process safety program especially at this that this plant was was not not as good as it could be even though this plant had an excellent occupational safety record and so the the leadership of BP did not come into the plant to do inspections they said the plan is obviously doing very good and so you know they they they ignored the all of these conditions which had they done a proper inspection would have have discovered they would have discovered that the culture for process safety was nowhere near as strong as the culture for for safety so people were very you know they would wear their seat belts when they drove the trucks but they even though they should not have been driving the trucks in the first place so that gives you an idea of the of the two different kinds of cultures and there are many there there are many more much more to learn this is how you can receive read the information that BP presents on on this accident it's a BP response calm and it can be downloaded and and and read at your at your convenience there's a additional information that will be coming out over the over the next a few months a very high-level panel called the Baker panel was convened last year and their report is due in fact it may have been released today it's it's sometime this week or possibly next week also the the US chemical safety board will issue a report in in March of 2007 this is an accident investigation board of the of the United States and they will have many recommendations for not just for BP but also for industry in general and so we look forward to seeing seeing their recommendations and finally we'll do a comprehensive review of of all the recommendations to industry by it by all the appropriate panels at the CCPS conference which is in April of next year