An extremely small malfunction ... and then something terrible happened

Sifting the wreckage: Three Mile Island and the aftermath

AT 4:00 a.m. on the 28th of March of this year, an extremely small and unimportant malfunction happened at a nuclear plant in the state of Pennsylvania. Within a very short period of time, it would turn into an event of momentous impact, with repercussions felt over most of the world.

Two weeks after that accident, the President of the United States appointed a 12-member commission, and when we first met, in late April, I looked around the room and found one of the most diverse groups I have ever worked with - diverse in every possible way. The opinions on nuclear power ranged from the strongly pro-nuclear to the strongly anti-nuclear. The members had a wide variety of backgrounds: There were representatives of several academic specialties, a governor, a well-known environmentalist, a Washington lawyer with significant government experience, and Anne Trunk, the one member of the commission who lived through the accident. And, finally, there was the chairman. I did not fit any of the categories for membership on the commission. I am not quite sure whether I was there as a generalist, which is possible, or whether I was there because they felt that with so diverse a commission, my having had ten years' experience chairing faculty meetings would be an ideal background.

Let me say what I knew at the time I approached this assignment. I have a fairly strong background in science, but I know practically nothing about engineering, and in April I knew absolutely nothing about how a nuclear power plant worked. I was typical among members of the commission, each one of whom had some degree of expertise; yet few of us were experts - I should say none of us was an expert - on all the subjects that we had to investigate. Indeed, I had one additional distinction. On April 11, on my way down to the White House to be named chairman of the commission, my wife and I were madly trying to figure out what that agency with the initials N.R.C. was. The only such agency I knew was the National Research Council. I had never heard of the Nuclear Regulatory Commission. And now, six months later, the Nuclear Regulatory Commission wishes that they never heard of me.

Ol course, we quickly needed to find out how a nuclear power plant works, a process that can be described here in fairly simple terms. In the right-hand side of the diagram, the feedwater pumps drive water into the steam generator where the water is heated up and turned to steam, and that steam drives a turbine (or several turbines) to generate electricity. Then the water is cooled down again, condensed, pumped back in, and re-heated, and it keeps generating electricity. Incidentally, this part of the plant, I am told, is no different from any other kind of electric generating plant. It could be driven by coal or oil or whatever the fuel might be.

I want to call particular attention to the cooling towers. Because of the many ironies connected with this particular accident, the commission would spend a lot of time talking about the cooling towers, which became the symbol of Three Mile Island. They are tremendously impressive. They are frightening. They are probably the least important part of the entire plant. All they do is provide a third water system to cool off the steam, and they are used purely for cooling that water which has no radiation in it. All you see coming out of the towers is harmless steam. This is not where the danger is. They look huge, and they are totally unimportant.

But what is special about the plant? What is special is that it is not coal or oil or wood that heats this water but a nuclear reactor. That means there are uranium pellets in a metal casing, in something called zirconium . . . lots of them sitting around . . . and a very carefully controlled chain reaction, a reaction in which the very basic energy of the atom is used. In this particular case it is used for nothing more exciting than heating up water in the primary loop, which in turn heats up the water in the secondary loop (center section), which then drives a turbine. Enormous amounts of heat are created, and therefore a great deal of electricity can be generated.

Why are there two separate water systems? That is because the primary system comes in contact with nuclear materials and can pick up radioactive materials, while the secondary system never comes in direct contact with nuclear materials. It becomes heated by running through neighboring tubes, so it is free of radioactivity.

Now, the little things at the top of the reactor are absolutely crucial. They are called control rods. If anything goes wrong, they drop into the middle of all the nuclear material, and they are able to capture enough of the neutrons that are running around so that the chain reaction stops. That is the single most important safety feature of a nuclear power plant.

The whole primary system is enclosed in an enormous reinforced structure called the containment building. The theory is to contain within that area any radiation that might escape, and it can be isolated so that no radiation can escape to the outside world.

The part on the left of the diagram is normally the least important part. It is called the auxiliary building, for miscellaneous equipment. It would turn out to be terribly important in this particular accident because all the radioactive material that escaped went out through the chimney of the auxiliary building.

The final piece of equipment I want to describe is something called a pressurizer (center). It serves the following purpose: The water that cools the reactor and, in turn, heats the water in the pressurizer is at quite high temperatures, between 500 and 600 degrees. Since water boils at 212 degrees, you might ask how can there be water at that kind of temperature? The answer is by keeping it under high enough pressure, as in a pressure cooker, for example, it will not boil even at much higher temperatures. The purpose of the pressurizer, which contains part water and part steam, is to maintain an equilibrium to keep things at the right pressure. The operators are carefully trained to keep watching the water level in that pressurizer. If it is high enough, there is enough water. If it goes too low, it needs more water. That turned out to be one of the most important things in this particular accident.

LET me now turn to what happened. Just before 4:00 a.m., some of the crew were working on repairing an extremely unimportant part of the system - the polishers. (They are so unimportant, they are not even shown on the diagrams.) Water that runs through the turbine is circulated so that any "gunk" in it is removed by these polishers and therefore they have to be cleaned periodically. Something went wrong as they were being cleaned. We think, but are not absolutely sure, that a little bit of water from the polishers got into an air line, which in turn triggered something else that turned off a pump, and all of a sudden water stopped running into the steam generator. As a result, the turbine was shut off by the safety machinery, and there was no water circulating on the right-hand side.

As I pointed out, there is a dual function in the steam generator. The water in the primary loop heats up the water in the secondary loop to generate electricity. At virtually the same time, by passing its heat along to the secondary side, the water gets cooled off and can go back and keep the reactor cool. All of a sudden, with water not running into the steam generator, not very much heat was being removed from the reactor, and things began to heat up where it really counted. As the water got hotter, the pressure rose. But there is a provision for just such an event. Right above the pressurizer is a valve, the pressure-operated (or pilot-operated) relief valve, which acts like the lid of a steam kettle which, if the pressure gets too high, lifts up and lets out some steam. That valve lifted to release the pressure as it was supposed to do. When the pressure kept rising even higher, the next safety device was triggered: The control rods dropped in place in the reactor, just as they were supposed to, and the chain reaction stopped within a second. So, ten seconds into the accident, and given the original malfunction, everything had gone perfectly. Everything seemed to be completely under control. It was like hundreds of previous incidents that had happened with no bad effects.

One more thing about the reactors: When the operators turn off the chain reaction, not all the heat generation stops. About 94 per cent of it does, but not all of it, because the process of nuclear fission creates radioactive by-products - wastes - that continue to give off radiation, which in turn generates heat. After the fission is shut off, only six per cent of the original amount of heat remains. But the original amount was horrendously large and six per cent is still quite a large quantity, and it is extremely important to keep removing that heat.

The operators have two ways of doing that. One is through the steam generator. The other involves special emergency water systems that are designed to be much more than enough to remove all that heat. It is at this point in the Three Mile Island accident that things go wrong.

In this system practically everything exists in duplicate; when the water stops running, a whole second system is designed to start pouring water into the steam generator. That came on automatically as it was supposed to, but no water reached the steam generator. One operator madly scurried around trying to find out what was happening, and eight minutes later he discovered that some valves which were always supposed to be open under normal operations were left closed at some unknown time. We never did discover when they were left closed. The most likely time was during a maintenance operation two days earlier. And, although the people who were involved swear that they reopened them, we are not absolutely sure that their memory is right. At any rate, that is one of the minor mysteries; it is not of vast importance, because our analysis showed that the emergency water system coming on would not have made a great deal of difference in view of everything else that happened. But it was very important in terms of one operator's total confusion - scurrying around watching one side while horrible things were happening over on the other side.

The second thing that happened is that when the reactor shut off, the pressure started dropping. As the pressure dropped, the valve that opened under too high pressure was supposed to fall back into place. There was an electric signal to open it and an electric signal to close it, and there was the light in the control room that said the signal went out for the valve to close. What the operators did not know was that although the signal went out, the valve stuck open. Now, valves do stick open, and this particular valve had some history of sticking open. As a matter of fact, because of that, there was another valve behind it that they could close from the control room if this one ever stuck open. But they did not know that this valve had stuck open.

Things were getting to be quite serious because with the valve stuck open, precious water was running into a drain tank instead of staying inside the system to cool it. It would be well over two hours before the operators realized that that valve was open, and by that time tens of thousands of gallons of water had gone out through the valve and through one other route.

Since the pressure was not diminishing to proper levels, two minutes after the accident started the main safety system came on by itself. It is called the high pressure injection system, and it pumps in, through two pumps, a thousand gallons a minute - vastly more than is normally needed to cool everything off. We have carefully checked the calculations; even with that valve stuck open, there was plenty of water to bring everything under control. But now the third thing went wrong. The operators turned off one of the pumps and cut the other one way back, so that instead of a thousand gallons a minute there were less than 100 gallons a minute, which created a situation where the amount spilling out of the system was significantly more than the amount being let in. That is the fundamental and direct cause of why a minor incident turned into a major accident.

The operators realized none of this. Their supervisors who came to help out did not recognize the problems, and about 100 minutes into the accident - a little over an hour and a half - they had a very serious problem. Water was boiling. There was not enough water to cover the core. The metal casings that held the uranium started bursting, and eventually enormous damage was done to the entire core. For hours they kept trying to deal with it. They tried one thing after another, and from analysis done a month later it is clear that some of their efforts made things better and some made things even worse. It took most of that day before they got the system even roughly under control. That is how the worst accident in the history of commercial nuclear power happened.

IT was quite natural, I suppose, that given those facts all the early reports said: "It is very simple. The equipment had only minor problems. It is all the fault of the operators. It's what they did wrong that caused the accident."

It certainly looked that way to us, too, and at our first public hearing where we could take testimony under oath, we called all the operators and their supervisors to tell us why in heaven's name they did what they did. And, to a man, they testified that they had never been prepared for anything like what they faced there. We pushed them very hard, and we could not shake them. At that hearing our investigation, which had concentrated on equipment, took a very sharp turn. From that point on - although we continued to look at the equipment and made a number of recom- mendations for improving it - we followed a path leading to our major conclusion: That the real problem was not the equipment but the people. Before we were through we would find "people problems" in more different places than we would ever have guessed at the beginning.

To carry on this changed investigation we hired a new, very tough legal staff, with a new chief counsel with experience in investigative legal work. The first thing he advised me to do was to sign what is called a broad-based document subpoena, something I had never seen before. For example, the subpoena would say to a utility: "Produce every single piece of paper you have in your files that deals with Three Mile Island-2." I said to our chief counsel, "Do you have any idea how many pieces of paper that would result in?" and he said, "No." I said, "Could you take a guess?" and he said, "Oh, 100,000 pages." I said, "Stan, you are totally out of your mind!" And then I signed the subpoenas.

Fortunately, our chief counsel was absolutely right, and I am glad that I did not use my better judgment. Out of that enormous haystack we would find some little bits that would make all the difference in understanding the real cause of the accident of Three Mile Island.

We would acquire documents that at the end filled 300 feet of shelf space in a library; we would talk to hundreds of individuals; we would take formal, sworn depositions from over 150 people; and we would hold numerous public hearings with testimony under oath. And out of that we pieced together the following picture.

THERE was one incident from which we probably learned more than anything else we found. We learned it because of one single document that our legal staff somehow found in an enormous mound of papers from one company.

Davis-Besse is a nuclear power plant in Ohio that had an incident in September of 1977 with certain strong similarities to what would happen a year-and-a-half later at Three Mile Island. For somewhat different reasons, the turbine shut off, so the problem was first on the secondary side. The pressure-operated relief valve opened as it was supposed to, the reactor worked correctly, and the chain reaction stopped - and the valve did not close. Furthermore, the emergency water system came on and the operators turned it off. Some of that probably sounds familiar.

As this document showed, however, there were some significant differences between Three Mile Island and Davis-Besse. The operators at Davis-Besse figured out that the valve was stuck open in 20 minutes instead of two hours and 20 minutes. Also, it was a newly fueled plant, and therefore there was much less residual heat to be removed than in an older plant. Finally, they happened to be operating at less than ten per cent power. That made a very great deal of difference.

Someone who investigated the earlier incident for the Babcock and Wilcox Company, the company that manufactures the steam supply system, said, "I'm terribly worried. I don't understand why the operators turned off the water." The answer he learned was that what the operators were looking at was not the pressure in the core but at the now-famous pressurizer. They said, "The level of water in the pressurizer kept rising, and therefore we knew we had to turn off the water. We had too much water." Yet all the recorded evidence shows that the water was disappearing from the system, and the investigator became very worried about it.

He went to one of the top engineers of the nuclear division at Babcock and Wilcox - a man called Dunn - and Dunn, who had studied these problems, realized what must have happened and wrote a memorandum that turned out to be a historic document because he was absolutely correct. The document says that if you do the wrong things and water starts boiling, what will happen is that the steam will drive the water higher up in the pressurizer, and therefore the pressurizer can even be full, with water escaping from the top, without it telling anything at all about the water in the core itself. The core could be almost empty. That was something they had never considered before. Mr. Dunn became very seriously concerned, and he started writing memos to colleagues saying, in effect, "We've got to do something to warn the operators that under such-and-such circumstances, looking just at the pressurizer can be totally misleading. They have to be taught to keep that water pouring in, because if this accident had happened at full power it could have left the core seriously uncovered and resulted in major core damage." Of course, that is precisely what would happen at Three Mile Island.

That memorandum was written 13 months before the Three Mile Island accident. It was written by an employee who was not in any way criticizing his company's equipment, so that company had every selfish interest to get that memorandum out. We tracked that memorandum all through the company and found that it just got lost in a bureaucratic mess and never got out.

The memo went to the training division, and they asked the same questions the operators asked: "But we thought we were always supposed to watch the pressurizer. If we let too much water in, the pressurizer could get full and that can cause some equipment damage." (The damage would be quite minor, incidentally.) And Dunn answered, "But for heaven's sake, when you are running the risk of the core getting uncovered and having a horrendous accident, surely you are not going to worry about that!" They were not sure, so they sent the memo on to a special division that worries about the system overall. That division did not have time to deal with it because they had just gotten some new regulations from the Nuclear Regulatory Commission (a bit of irony) and were told to drop everything for three months.

Eventually, the chiefs of these two groups got together - in front of a coffee machine - and one of them asked, "Have you looked at that memorandum?" The second chief claims he said "Oh, yes, he is absolutely right," but the other man hears him saying "Oh, no, it is perfectly all right," meaning, "don't worry about it." As they both testified, that is how the whole thing "dropped between the cracks." Because of that, the instructions never got out until two weeks after the accident at Three Mile Island.

An exact parallel occurred within the Nuclear Regulatory Commission, where an inspector quite low in the hierarchy became worried about the same DavisBesse incident. He tried warning people. He kept going to his superiors, and he got absolutely nowhere. Finally, he took advantage of something called the "open door policy" and went to see two of the five commissioners of the Nuclear Regulatory Commission, telling them about the problem. They took him very seriously and asked for an investigation. Unfortunately, the date of that visit was March 21 of this year - seven days before Three Mile Island. Too late to make any difference.

There was an engineer within TVA (which has similar equipment), who, purely on theoretical grounds, got worried about the same sort of thing happening. He issued a number of warnings - one of which concerned this kind of problem that could lead to operator misunderstanding. Those warnings were sent to the utility, to Babcock and Wilcox, and to the Nuclear Regulatory Commission, and for all kinds of reasons none of them ever did anything about it.

It is very clear that if any of those groups had correctly reacted to the warnings, the Presidential Commission would never have been created.

WE then started looking at the training of operators. I will give just a slight idea of what we found. There seemed to be no minimum educational qualifications for somebody entering operator school. It is essentially an on-the-job training program; it has no precise curriculum. There is shared responsibility between the utility and the steam supplier, with no precisely developed rules as to exactly what is to be covered by each one. Neither side teaches the fundamentals of nuclear power. It is a sort of button-pushing school. Even when they do receive practical, hands-on experience, there are some serious deficiencies. For example, a key training device is something called the simulator, which is a computer that can reproduce what happens in the control room. NASA used simulators for its space flights, and they are used for training airplane pilots. This simulator was a perfectly good one for ordinary operations. But our investigation showed that when they tried to reproduce on it the conditions of the accident at Three Mile Island, they found it was not programmed to reproduce those conditions. Therefore, the operators had to be right in saying that they had never been trained for anything like that.

Speaking of the control room, I am afraid I made a slip of the tongue early in the investigation. When we toured the plant, I came out of the control room and said, "That equipment is 20 years out of date." I received some severe industry criticism for that remark. They are quite right; the remark was not correct, because we later found a document by a very highranking N.R.C. official, written ten years ago. As he pointed out, the equipment was then 20 years out of date.

The control room is the biggest Tinker Toy you have ever seen in your life. There are hundreds of flashing lights all over the place. Each one suggests a certain kind of alarm. I have never quite gotten over an incident that happened while we were touring the plant. Alarms went off, and people did something about them, called somebody, pulled a lever and pushed a button. There are little transparent labels, and behind each label is a light that blinks to show which alarm has gone off.

After the third alarm during our tour, I looked to see where the alarm was. Nothing was blinking. So I finally raised my hand and asked the person who was giving us a fascinating lecture if he would mind telling what was going on. He asked the operator, who said, "Oh, it is very simple. One of the light bulbs must have burned out, and therefore we don't know which alarm went off." I then watched for five or ten minutes while an auxiliary operator removed those little labels, one at a time, until he found the burned-out light bulb, and screwed in a new light bulb. It started blinking, and then he knew which alarm went off. Then they did something about it. I am afraid I did go so far as to say publicly that I did not think that was modern technology at its most glorious.

FROM there we took a very intensive look at the Nuclear Regulatory Commission. In terms of operator training, I should tell you that section has a minute staff - I think there are eight or nine full-time employees in it, plus some part-time people. They really just did a token sampling and administering of quite routine exams, and even there we found, for example, that an operator could fail badly that part of the exam having to do with the safety systems, but get a good enough overall grade to pass the examination. We looked at their licensing and found all kinds of difficulties with the licensing process. Perhaps the most troublesome was that in the licensing of a given plant, if there was a serious problem that was applicable not just to that plant but to a lot of plants, they would label the problem "generic." Then the plant did not have to worry about that problem because it would go through a different machinery. Some of those "generic" problems have been sitting around for years and years.

We looked at their enforcement arm, because all their rules and regulations mean nothing unless somebody' enforces them. We found that this was probably the weakest part of the Nuclear Regulatory Commission. Here we had an enormous amount of help because there have been two or three other studies in the recent past that had scathing reviews of the performance of this particular branch of the N.R.C.

We found, as they had in previous studies, that a great deal of the so-called inspection of plants was a paper inspection. Not everything can be personally examined. But we found out, for example, that TMI-2 was checked thoroughly by an inspector a few months before the accident and he never looked at a single piece of equipment. He just looked at the paper records of the company. That did seem extreme to us. When inspectors found problems, again and again we had testimony that there would be great reluctance on the part of superiors to follow up on them if it required really tough action.

What troubled us most as we worked our way up the hierarchy of the Nuclear Regulatory Commission is that we never did find put who runs the agency. There are five commissioners on top, one of whom has the rank of chairman, but the law quite clearly says that the five commissioners are equal. They are picked for their diversity of views, which is an important safety feature in terms of getting many views represented. But it is not terribly good for running a huge agency on which the safety of nuclear power plants depends. Indeed, we reached the conclusion that in a very real sense, no one was running this particular agency.

We looked at how they responded to the emergency and found it was a state of total confusion. There was a lack of communication, and instead of making things better, at several places N.R.C.'s actions made things worse. The most outrageous example is the now-famous hydrogen bubble. It has been known for a long, long time, both in nuclear and non-nuclear plants, that if you uncover certain metals at sufficient temperature, the metal casing is going to interact with water and create hydrogen. Apparently, this had never been worked out by the Nuclear Regulatory Commission, and they had no idea whether that hydrogen would blow up or not. On the fourth day of the accident they were calling all over the country, trying to get an answer to that question. They came to the wrong conclusion - that is, at least some of them did, because they split 50-50 on it. They spread it to the press and scared the daylights out of people when, as a matter of fact, the scientific reality is that there was no possible way that hydrogen could have exploded.

As we studied the N.R.C. overall, our most important judgment was that it is an agency hypnotized by equipment. It had a firm belief that equipment can be made fail-safe, and as a result the N.R.C. totally ignored the human element in nuclear power.

This accident could clearly have been prevented in many ways. But the President's Commission on Three Mile Island has overwhelmingly concluded that given all the problems - the human problems that we found within the utility, its supplier, and particularly the Regulatory Commission - if this particular accident had been prevented, an accident like it was eventually inevitable.

It is on that that we based our overall conclusion: "To prevent nuclear accidents as serious as Three Mile Island, fundamental changes will be necessary in the organization, procedures and practices - and, above all - in the attitudes of the Nuclear Regulatory Commission and, to the extent that the institutions we investigatedare typical, of the nuclear industry."

WHERE did we go from there? Our commission then entered its very difficult decision-making stage. We taped every single word of those discussions, and transcripts were released in late October so future historians can see how we struggled from a tremendously diverse set of opinions to an overwhelming consensus. This consensus would lead to 44 recommendations, of which 36 were unanimous and of which no other received more than one negative vote, though there were some abstentions. Frankly, given the composition of our commission, I consider that an almost miraculous outcome.

Of course, there were many things that we could not do. We did conclude that our evidence was not such as to recommend that nuclear power should be abolished. But we did not feel that we could say go full speed ahead with nuclear power, either.

We did not look at the whole industry. We did not look at the energy mix available to the United States. It was not part of our charge, and there was no way we could have done that in six months along with everything else that we did. Therefore, while we have some understanding of the present dangers of nuclear power and have made recommendations that we hope will reduce those dangers significantly, we did not have a basis to compare nuclear power to alternate sources of energy such as coal, which seems to be one of the most plentiful materials available in the short-run. However, some serious scientists are worried that if coal is burned in large enough quantities it will change the climate of the earth with totally unknown consequences for mankind. We did not look at that.

We concluded that we should speak only to what must be done to nuclear power. The decision of what is the best energy mix for the United States must be resolved by the political process through the President and Congress. What we did do is to recommend those changes in nuclear power that we thought were absolutely necessary.

We started with a restructuring of the Nuclear Regulatory Commission, because that agency must be changed. We recommended that it go to the form of a strong administrator, who would be chosen from outside the present agency. We feel that this is crucial. We want him or her to have enough freedom to bring a great deal of new blood into the agency. And since we recognize that the N.R.C. structure had one advantage - namely, that somebody was watching over what other people were doing - we want to replace that with something that would play the same role. We decided that it would best be put outside the agency, because a committee of five cannot run that commission. So we recommended a permanent oversight committee for both the N.R.C. and the nuclear industry, with a diverse composition of members to whom people could write with their concerns. It would be a watch-dog committee to report at least annually to the President and Congress.

We recommended sweeping changes for how this agency should function, which I hope will be accepted whether or not our single administrator is accepted. I'll provide just two examples: On licensing we have recommended vastly higher qualifications for a utility before it is allowed to operate a nuclear power plant - qualifications stringent enough, I suspect, that a number of present utilities would not be allowed to operate such plants. Regarding enforcement, we issued a long list of recommendations for strengthening that area, one of which is that every single licensee should be periodically reviewed in depth as to whether it is performing up to the standards of that license. We made a number of recommendations to the utilities on organization, on procedures, on standards of maintenance that were certainly not satisfactory in this particular plant. We recommended a completely new approach to operator training so that candidates would first have to graduate from a training institution where they will learn the fundamentals of nuclear power before they get on-the-job training. And we recommended very strongly that utilities pay a great deal more attention to the procedures given to operators, particularly to emergency procedures, because some of the ones we reviewed were at best confusing and possibly totally wrong. (For example, we had a half-hour argument in one of our public hearings, just to try to figure out which of two procedures the operators were supposed to follow during that particular accident.) We made a number of recommendations on equipment, the most important of which has to do with modernization of the control room.

THEN we turned to the mitigation of accidents once they happen. There is the question of emergency preparation and response. Fortunately, in the case of Three Mile Island, although that plant is now a total mess, the containment building worked well enough so that most of the radiation did not get out. Having retained the best experts in the country as consultants, we are convinced that the direct physiological health effects are going to be minimal or non-existent. However, there was major mental stress, and we concluded that was the greatest direct health hazard.

But if there had been significant releases of radiation, we felt that the preparation was atrocious and the response was worse. The Nuclear Regulatory Commission had come up with a magic number for establishing a safety zone. Through a crazy calculation that we tried to explain in our document, they somehow established a two-mile safety zone as a magic radius around this plant. Fortunately, Pennsylvania has tougher laws; that state requires five miles. And those zones were for much worse accidents than this one. In the middle of this accident, where people suffered very little direct harm, authorities were planning for an evacuation of up to 20 miles from the plant, and we noted that the way that planning was done had absolutely nothing to do with real life.

We have recommended that, where feasible, plants should be located as far away as possible from major population centers, and the states in which they are located should be required to have emergency plans. We also recommended significantly better public education.

THE last subject I want to deal with is the hardest one. It was most unfortunate for our commission that, breaking all the trust we had in each other, someone gave premature leaks to the press that were highly selective and that distorted the press coverage in the period before we went public.

Those of us who had promised not to speak until October 30 were caught in our commitment. A number of false things were said during that period on the subject of the moratorium, with the result that when the report came out it was terribly hard to get our most important recommendations across because everybody wanted to know about the moratorium. I think whoever did that leaking did it intentionally. For example, in the leaks it was suggested that somehow there were lastminute changes in the rules under which a moratorium could be adopted. The fact is that my first memo dealing with rules of adoption for recommendations was written on August 14, more than two months before the submission of the report. In that memo I suggested that in order to have enough consensus so that people would listen to us, perhaps eight votes out of twelve would be a reasonable requirement. That was changed at the final meeting from eight to seven, and the seven-vote rule was agreed to by a unanimous vote of the entire commission. I mention that because a very different story was printed in the media.

What did happen to us on the moratorium was this: There were three separate votes, and eventually eight different commissioners would vote for at least one of the moratorium proposals, i.e., for a temporary halt to new construction permits. We could never get seven of them to vote for the same one; we got six on the first, six on the second, and four on the third.

I'll be glad to tell you my own voting record. I strongly supported the first proposal, argued for it over two meetings, and voted for it. It was a moratorium until the President and Congress had an opportunity to consider our recommendations. I did not vote for the second proposal, because on that one the present Nuclear Regulatory Commission could have called off the moratorium, and I'm afraid I didn't have any confidence in them. The third one was a straight two-year moratorium. Only four of us voted for that, and mine was one of those favorable votes.

This issue has been blown totally out of context. Quite frankly, whether there would or would not be a moratorium would not be a decision of our commission. It will be up to the President or Congress whether one occurs or one does not occur. They know how we voted and how we feel.

The sad thing that happened was described by one congressman at the end of the joint hearing. He pointed out that the biggest story never got proper coverage. That story is the vote we unanimously adopted that says no license - neither a construction permit nor an operating license - should be issued until three separate requirements are satisfied: 1) that the new safety requirements that we and others recommended have been implemented; 2) that the applicant fulfill the much higher qualification for running a nuclear power plant, including having an acceptable operator training program; and 3) that those licenses should be made conditional on the approval of state and local emergency plans.

We have no way of enforcing these requirements, as we have no way of enforcing any other recommendation. But if this recommendation should be given teeth, let's say, by the President and Congress, it could be the most powerful recommendation that we made.

WHEN all of this began, I made two promises. One was that this commission, if I had anything to do with it, would tell the truth. And this commission did tell the truth precisely as we found it, irrespective of the views we held when we began the investigation. Secondly, I promised that unlike most such reports, we would produce one that you can read. The compliment 1 was most pleased with was the instance when two senators, one Republican and one Democrat, independently of each other, went on at length to say that this particular report is the first such report they have ever seen that is written in English and that people can read.

I have something to ask of you. You are going to read all kinds of statements about this report, although the press overall has been generous to us. But the media cannot give you complete coverage of this document. This is a strong, highly structured report, and we worked hard to produce a concise package that makes sense as a package and that gives an overall picture that should be viewed in its entirety. I believe it is a report of integrity. You cannot get that from the press reports, no matter how good they are, because they cannot print something so long. The commission's report will be available, and I very strongly urge you to do one thing: Please read it for yourself and then make your own individual judgment as to whether the President's Commission on the Accident at Three Mile Island rendered a verdict that was fair.

With the projected image of the power plant looming behind him, President Kemeny reports to the College on Three Mile Island.

"The control room is the biggest Tinker Toy you have ever seen ... I asked a person if he would mind telling what was going on."

"Breaking all the trust we had in each other, someone gave premature leaks to thepress . . . [and] a number of false things were said on the subject of the moratorium."

This article is adapted from the tapetranscript of a speech to the College communityon November 2, less than 24 hoursafter John Kemeny returned to Hanoverfrom Washington. As is customary withhim, the speech was made without aprepared text.