Cloud Over Comanche Peak

JUANITA ELLIS RESTS her elbows on her kitchen table, leans forward, and speaks softly. “I never even suspected it would come this far,” she says. “When I think about it, I’m just amazed. Nobody ever could have paid me to do what I’ve nursery and garden center in Oak Cliff. “I didn’t know grass from trees,” she says. But she was determined not to be ignorant about her husband’s business. “So I started reading the labels on the things we were selling,” she says, “and I started reading about pesticides and their effects.” In the newspaper and magazine articles she nursery and garden center in Oak Cliff. “I didn’t know grass from trees,” she says. But she was determined not to be ignorant about her husband’s business. “So I started reading the labels on the things we were selling,” she says, “and I started reading about pesticides and their effects.” In the newspaper and magazine articles she saw were mentions of nuclear waste and radioactivity, and how devastating an accident at a nuclear power plant could be to the environment and the people living around it.

Then, in December 1973, Ellis noticed in the paper that the Dallas City Council had scheduled a public hearing to consider whether Dallas Power & Light (DP&L, a division of TU) should become involved in a nuclear power project near Glen Rose. “I called everybody I knew,” Ellis says, “and everybody started calling [the city council] to get on the agenda to speak.” So many people called that the council postponed the hearing for two weeks. In the meantime Ellis and five of her friends formed CASE.

The members of CASE showed up at the hearing, but they lost this first round of the fight. The city council allowed DP&L to join the Comanche Peak project, despite CASE’S objection that DP&L didn’t need the electricity. But Ellis and her friends had not prepared a solid argument, so they watched helplessly as plans were laid for what has become the nation’s most troubled nuclear plant. And in the process, Ellis learned the value of going to hearings with her arguments well documented.

A NUCLEAR POWER PLANT WAS AN AMBITIOUS PROJECT FOR TU. The nuclear industry was in its heyday in the early Seventies-monster mill ion-kilowatt plants were being built all over the country, uranium was a domestically available cheap fuel, oil prices were climbing with the whims of OPEC, and the economy of Dallas/Fort Worth was beginning to boom along with the oil business. TU thought a dual-reactor plant would be the best and cheapest way to provide energy to a growing market.

In 1973 TU hired Brown & Root Inc. to construct the plant, and Gibbs & Hill Inc. of New York to design it. TU estimated they could build the plant in six years, and in December 1974, work began on the excavation of the nuclear plant’s foundation. From the start there were problems. The subcontractor hired to dig and level the foundation used too much dynamite and blasted the hole too big. And while the problem wasn’t especially significant-Brown & Root filled the cracks and holes with “dental” concrete to level the foundation-it did portend a trouble-filled future.

In that future, TU would make many mistakes in designing and building the plant, and the Nuclear Regulatory Commission (NRC), which made its own share of mistakes, would fail to assert its power and force them to do it right. Comanche Peak had three strikes against it before it even got started. First, Brown & Root was an inexperienced nuclear contractor. The company had built only one other nuclear plant, in Brunswick, North Carolina, and it was having big problems. Second, TU had no experience with nuclear plants, although they had built many conventional electric plants. And third, the NRC, created from the old Atomic Energy Commission in 1974, was almost as lax in their inspection and regulatory activities as was the A EC before them.

TU has always thought of Comanche Peak as they do their other coal- and natural gas-burning plants. Dick Ramsey, TU’s official spokesman, told the Fort Worth Star-Telegram in January 1987 that Comanche Peak “is a steam electric station. It’s a thermal plant. Only difference is, it’s got a different heat source.” Technically, Ramsey is right. A nuclear plant heats water to make steam that drives turbines that turn generators to make electricity. But the “only” difference is a tremendous difference, and coal or natural gas combustion is not even remotely similar to a nuclear reaction. “It’s a different ball game building nuclear than it is building fossil,” says Bob Warnick, the NRC’s chief inspector at Comanche Peak. “Different regulatory requirements, different quality requirements, different public expectations.”

The first few years of Comanche Peak filled Juanita Ellis with a slow, creeping terror. Reports of problems at the construction site came to her from former employees: work crew chiefs who didn’t want to hear about problems the workers found; quality control inspectors who had been drafted from the ranks of trade workers and had to inspect their own work; harassment and intimidation of quality control inspectors who had been given unrealistic amounts of work to do in a given time period, resulting in hasty or incomplete inspections; painters who noticed pipe welds that didn’t look right and reported them, only to be told that their job was to paint, not to inspect welds. Ellis took affidavits from every former worker who would give one to her and sent them to the Nuclear Regulatory Commission, and later to the Atomic Safety and Licensing Board (ASLB), the “judicial” arm of the NRC, which began hearing complaints on Comanche Peak in 1979.

The NRC, although it was unrelenting in its support of TU, did conduct the federally required inspections and reports, and found even more problems that disturbed Ellis. The NRC said that the quality control system at Comanche Peak was poor, exactly the sort of thing Ellis had been hearing from the former employees, but the NRC said that it could be corrected and that it had caused no major problems. Sixty percent of the welds on pipes that would carry radioactive water were found to be defective, but the NRC said the system to detect errors was obviously working because the bad welds had been found. Then came a major mistake: TU built the concrete and steel supports for the reactor in Unit 2 out of alignment by 45 degrees. TU jack-hammered out the bad supports and rebuilt them in the right place. The NRC said the event did not indicate any major problems at the plant. Ellis was livid. It showed, she says, just how much attention to detail TU was paying as it built the new nuclear plant, and she contended that the same type thing might be happening all over the plant. But with the NRC supporting TU, the ASLB would not take Ellis’s concerns seriously, despite her growing pile of documentation.

Meanwhile, Ellis’s husband, Jerry, had sold his nursery business and retired, and Ellis began working as a secretary for a local insurance agent. The couple had no children and she spent her evenings compiling information on Comanche Peak to send to the ASLB. She could feel her other interests being crowded out by the nuclear plant, but she kept thinking that if she could just convince the ASLB that the plant had problems-and that those problems might be the tip of a very large iceberg-her job would be finished and she could rest. “Boy, was I wrong,” she says. “If I’d known then what I know now, I don’t know if I would have kept going.”



WHILE ELLIS WAS DOCUMENTING THE PROBLEMS of Comanche Peak, the whole nuclear industry was being shaken by two unprecedented accidents that terrified the public and pressured the NRC to crack down on every plant in the nation. Comanche Peak was no exception. In 1975, at the Brown’s Ferry nuclear reactor in Alabama, an electrician using a candle flame to check for air leaks caught some electrical insulation on fire. The fire threatened the control of the reactor, and a meltdown was only narrowly avoided. The NRC responded by changing the design and construction regulations for all nuclear plants being built. One of the changes was to require two widely separated sets of electrical wires leading to the reactor control room, where before there had only been one, and to use better, less flammable insulation material. TU had to go back and redesign major portions of Comanche Peak.

Then came the worst nuclear disaster in the history of the country, and the nuclear industry was changed forever. On March 28, 1979, Pennsylvania’s Three Mile Island Unit 2 went out of control. A pressure relief valve on the reactor’s cooling system stuck open and most of the coolant leaked out. The control room indicator light for that valve showed that it was closed, which confused the operators. While they made a series of wrong decisions, such as turning off the flow of coolant, the reactor temperature soared to more than 5,000 degrees, and the core partially melted and fell to the bottom of the reactor vessel. The danger of radiation leaking into the environment lasted for weeks before the situation was brought under control. A small amount of gaseous radiation was released, and some 120,000 gallons of radioactive water escaped into the Susquehanna River. Nuclear engineers, perpetually optimistic, point to the Three Mile Island incident as proof of how well their hardware works. “Three Mile Island,” says John Beck, vice president for nuclear engineering with TU, “is a dramatic example of how that multiple layer of protection has actually worked.”

After Three Mile Island, the NRC, the public, and activists like Juanita Ellis began scrutinizing the nuclear industry as never before. Many of the rules of the nuclear game were changed and enforced more strictly. Control rooms had to be redesigned with expensive new computers to help the operators understand better what was happening to the reactor. Operators had to be trained better, longer, using new methods, and exact duplicates of the control rooms had to be built on-site for training. Even rules unrelated to the Three Mile Island accident changed, such as stipulating strict enforcement of regulations on nuclear safety in the event of an earthquake. Comanche Peak, in one of the most seismically stable parts of the continent, had to abide by those rules, too.

“All the documentation and prescriptive statements of how thou shalt proceed,” says John Beck, “evolved after the plant had a [construction] license.” TU had to go back and have major parts of the plant redesigned and rebuilt. The price tag for Comanche Peak soared to $2.2 billion in 1980.

And Juanita Ellis, always watchful, felt the tide of public sentiment changing. She renewed her efforts to bring contentions before the ASLB, learning the legal tricks that the TU lawyers used and throwing them right back at them. TU lawyers filed interrogatories asking, for example, for every document supporting Ellis’s claims of unsafe construction, or the name of every person who had made a safety-related mistake at the plant since it began. Such interrogatories are one of the techniques lawyers often use to discourage pesky “interveners” such as Ellis and CASE. But Ellis surprised the TU lawyers. She not only answered every interrogatory, personally typing hundreds of pages herself, but she hit TU with a barrage of her own interrogatories that they had to answer. TU was stunned. Juanita Ellis, “that woman” as they referred to her, could play hardball.

But while Ellis was learning the rules of the game, and the external events in the nuclear industry were swirling around outside of the Comanche Peak plant site, the problems continued to build up as construction continued. In designing and building Comanche Peak, TU used an “iterative” process. Literally, the term implies doing something over and over-and that’s exactly bow Comanche Peak has been built. The plant’s designer. Gibbs & Hill, drew the blueprints of how the plant was to be built and Brown & Root used the blueprints to do the construction. After a particular part of the plant was built, a draftsman would make a drawing of the “as-built” construction and send it back to the design engineers, who would check to be sure everything conformed to the original blueprints and government specifications. If it didn’t, the)’ would make new blueprint drawings Of how to fix the problem, give it back to Brown & Root to rebuild, make another as-built drawing, then send it back again to the designers. The plans would go back and forth until everything matched. In theory, iterative design works great. In practice, as layers of blueprints and as-built drawings build up. things can become confusing: the latest design changes can become lost or confused with earlier changes; engineers can end up making design modifications two or three steps behind the construction crew that is building that part of the plant; and construction crews can become confused about which design to use and when to use it.

As construction proceeded at the plant, confusion mounted. A construction crew was dispatched to install a door in a wall that wasn’t there-but was shown on the current blueprints to have already been built. Or a crew would be ready to put up a pipe support-and find an electrical cable tray already occupying the spot. The construction crews would do the best they could, improvising on the spot to make things fit, constantly assured by management that this procedure was normal and that any problems would be caught by the iterative process. Another as-built would be drawn of what the crew had designed and built on the spot, and sent back to the engineers. At any one time it was difficult or impossible to know which part, if any, of the plant conformed to federal safety regulalions.

Most of the construction problems at Comanche Peak, taken singly, were relatively minor, but the sheer numbers were daunting. More than 85,000 reports of potential safety violations (a large portion of them, by the way. filed after TU said the plant was ready to start up in 1982) have been filed by TU and the NRC to date at the plant-so the overall cumulative problem is enormous. In places the plant doesn’t remotely resemble the original design. It has become difficult to determine whether all the pieces of the final plant will work together correctly and whether the plant will be safe. Incredibly, that confusion took about eight years to show, Then, in 1982, Juanita Ellis’s efforts finally paid off.

IN THE SUMMER OF 1982 ELLIS GOT A call that changed everything. Mark Walsh, an engineer for TU. had quit his job in exasperation a few weeks earlier. He worked in the pipe-support group, which designed and checked the safety of brackets that held up the pipes that would carry coolant to the core. Walsh was convinced that the supports were unsafe, but none of his bosses seemed to care about his concerns. A couple of weeks after he left TU. he heard about Juanita Ellis.

When Walsh called, Ellis knew immediately what she had. The next ASLB hearing was two weeks away and Ellis worked feverishly with Walsh compiling a list of his complaints. When she submitted Walsh’s statement, the ASLB conceded that the allegations might have some merit. Ellis asked to subpoena a witness to bolster the allegations. She chose Jack Doyle, an engineer who had worked with Walsh and who had also resigned in disgust a few weeks earlier. Doyle had something that Walsh did not: documents. When Doyle resigned, his superiors marched him into his office and told him to put everything on his desk into a box and get out. What his bosses didn’t realize was that on his desk were several copies of pipe supports that Doyle believed had major problems. He hadn’t even thought about what he had in the box-until he got the subpoena from Ellis. Then he realized the significance of what he had, decided to cooperate, and took everything to her. “When I saw those documents he had,” Ellis says, “I couldn’t believe it.”

TU was blindsided by the Walsh/Doyle allegations. The woman who had been a thorn in their side for eight years suddenly became a sword. TU tried to counter the allegations by having a series of Ph.D.-holding engineers explain why they thought the pipe supports were safe. But the testimony of Walsh and Doyle was so compelling that the ASLB was shaken. The board ordered TU to take action on every allegation: either fix the problems or explain why they didn’t need to be fixed.

While working for TU, Walsh and Doyle had been charged with checking the stresses that pipe supports could withstand during an accident. They found that while the pipes were restrained by the supports from shaking sideways or up and down, they weren’t restrained from sliding back and forth or rotating. One of the main safety concerns in a nuclear plant has to do with the pipelines that carry coolant to the core. The supports that hold up those pipes are crucial, and must withstand the force of any shaking, whether from some internal cause or an earthquake. Walsh and Doyle believed that if a pipe dislodged somewhere and twisted and slid down the supports with several million pounds of force, it could break the pipe, potentially triggering a meltdown.

In December 1983, after listening for more than a year to the testimony of all the parties in the Walsh/Doyle allegations, the ASLB issued a Memorandum and Order on the hearings. They came down hard on TU.

“The record before us casts doubt on the design quality of the Comanche Peak Steam Electric Station.” the ASLB wrote. The ASLB wasn’t convinced by TU’s argument that Comanche Peak’s design was good because they had hired engineers with good judgment. The ASLB knew that behind every engineering and design decision should lie a traceable trail of sound mathematics. TU had shown no proof of such a trail. Walsh and Doyle, with Ellis’s help, had. And it led in a very different direction than TU would have liked.

The ASLB also indicted TU’s iterative design style of building Comanche Peak. On this count, TU had a powerful defender- the NRC. which stated in early 1983 that TU’s “iterative design review process has the capability to identify and correct pipe support design deficiencies.” But the board dismissed that argument out of hand: “The reason we reject these arguments is that we do not consider it proper for [TU| to wait until the end of its design process to attempt to locate and correct design errors.”

TU was frantic. “Every other word by the utility’s attorney in those days was ’Close the record! Close the record!’ ” Ellis says. TU didn’t want Ellis, and Walsh and Doyle, to have any chance of entering more allegations in the record. But the ASLB didn’t allow the record to be closed, so TU’s lawyers used another trick of the trade to get rid of intervenons. They filed motions for summary disposition for all of the Walsh/Doyle allegations. That meant that unless Ellis (with the help, of course, of Walsh and Doyle) could give a complete answer as to why the ASLB should not dismiss the Walsh/Doyle allegations, they would be thrown out. “It really puts the burden of proof on the intervener,” Ellis says. “It’s one of the time-proven memods of getting rid of intervenons. Except that we answered every one of those sons-of-guns!”

Besides dealing with the delaying tactics of TU’s lawyers, Ellis had to prepare a summary for the board of the important parts of the Walsh/Doyle allegations. Ellis quit her job with the insurance agent and worked long hours with Walsh and Doyle, typing everything herself. “Four hundred and forty-seven pages,” she says, shaking her head and sighing. “It was known in some circles of the NRC as the ’yellow bomb’ ” because it was bound in thick yellow paper. The bomb scored a direct hit.

Ellis had shaken the entire system, She had completely changed the direction the ASLB hearings were taking, and TU was devastated. ’T don’t think they [TU] took us seriously at first,” Ellis says. “I don’t think they took us seriously until that December 1983 order.” And while TU learned to take Juanita Ellis seriously, the cost of Comanche Peak climbed past $4 billion.



HE WALSH/DOYLE ALLEGATIONS spurred a series of reports, reviews, and studies by both the NRC and TU that resulted in many changes to Comanche Peak: 200 seals on the holes for electrical wiring through the containment building’s walls, which protect against radioactive material leaking out during an accident, had to be replaced because they were found to be weak; workmen were found to have ruined a $40 million condenser by using sledgehammers to force some of its tubes into place; 45 percent of the pipe supports (5,342 of them) had to be redesigned and rebuilt.

While the hardware of the plant was being repaired, evidence surfaced of problems of a much deeper sort, raising questions about the ethics and competence of TU management. A 1978 consultant’s report highly critical of the quality control process at Comanche Peak was hidden from the public and the ASLB by then-vice president of TU Lou Fikar until 1985. And on at least two occasions, TU has been found by the ASLB to have made “material false statements” to the NRC-about an improper concrete pour on the dome of Unit 1, and about the incorrect use of U-bolts in pipe supports.

Ellis says that she thinks that TU managers were so awed by the multiple safety measures in the design of the plant that they thought the plant would be safe no matter what they did. “I think they convinced themselves and a lot of the people that worked there that because of that redundancy, they could cut comers here and there, and that it wouldn’t matter,” she says.

Following these revelations, the ASLB was outraged (in a judicially restrained way, of course). On June 1, 1987, they released a statement saying that TU still had not adequately responded to the 1982 Walsh/Doyle allegations, that proof of the plant’s safety was still deficient, and that the board thought CASE had handled the complex issues at stake better than TU had.

Meanwhile, the Region IV office of the NRC in Arlington, which was the overseer of Comanche Peak, found itself in trouble, too. “When we got into this,” Ellis says, “it had to have been an insult to the NRC, because basically what we were saying to the whole world is we don’t think the NRC is doing its job. If they had been, we wouldn’t have felt it necessary to get in there and bring out the facts.” The special investigation office of the NRC at the Bethesda. Maryland, headquarters reviewed Region IV’s Comanche Peak operations, and found three major problems: there had been harassment and intimidation of at least one inspector, resulting in the downgrading or deletion of TU safety violations; the Region IV office had failed to carry out required inspections of TU’s quality assurance program; and safety inspectors had failed to file all the necessary documents to prove that they had inspected what they said they inspected.

In March of ’87, the NRC headquarters yanked Region IV from the Comanche Peak project and created a special projects office in Bethesda to oversee the troubled Texas plant (along with the deeply troubled plants of the Tennessee Valley Authority). A highly experienced crew of eight inspectors was gathered from around the country and sent to Comanche Peak to oversee TU’s most recent “get-well” plan. Most observers believe that the new Comanche Peak watchdogs will be tough because the whole nuclear and environmental community is watching. “We try hard to get out there,” says Bob Warnick, the new head NRC man on site, “to look and see that that plant is built right. We have to see that they set up a system to ensure that they identify and solve their own problems. Then we just have to monitor that system and make some independent checks. But when they’ve got 8,000 [employees], and we’ve got eight, we’re spitting in the wind. It’s tough.”

To make matters worse for TU, a fight has broken out between TU and the minority owners: Texas Municipal Power Agency (TMPA), with 6.2 percent; Brazos Electric Power Cooperative, with 3.8 percent; and Tex-La Electric Cooperative, with 2.17 percent. The minority owners became disgruntled over paying millions more than they had originally agreed to. With the end nowhere in sight, Brazos and Tex-La decided to stop making their payments to TU. TMPA continued, but threatened a lawsuit. So, in May 1986, TU sued the minority owners for breach of contract. Three weeks later the minority owners sued TU for failing to complete the plant according to the original cost estimate and time frame and for managing the construction in a “grossly negligent” manner. Some 30 million pages of legal documents have been filed and the trial date has been set for August 1988.

Then, just when TU officials might have been thinking that nothing else could happen to them, came the nuns, Into the fray marched two Catholic nuns’ groups, the Sisters of St. Mary of Namur in Fort Worth and the Medical Mission Sisters in Boston. The two groups own 450 of the 143 mill ion outstanding shares of TU stock. At the annual stockholders’ meeting in May 1987, the sisters called for a shareholder vote to force TU to consider canceling the plant or converting it to coal. TU calmly predicted that the request wouldn’t get even the 3 percent necessary to force its consideration again at next year’s meeting. The ladies of the cloth garnered 11 percent, revealing a serious breach of stockholder confidence in TU management. The sisters’ demand for coal-conversion might seem outlandish, but the Zimmer nuclear plant in Cincinnati, Ohio, was derailed in just this fashion. The plant was 97 percent complete in 1984 when it was converted to a coal-fired plant because its safety and quality could not be proven.

Part of the shareholder dissent may reflect concern over who is going to pay for Comanche Peak. If the plant doesn’t go on line, the stockholders will pay. A decision by the NRC that Comanche Peak is not safe to operate would cause the value of TU stock to plummet, and the credit rating agencies would certainly lower TU’s rating from its current top slot of A-plus.

And if the plant goes on line, then the rate payers will pay. TU is predicting that the Texas Public Utility Commission (PUC) will grant a 10 percent rate increase per reactor. With Comanche Peak’s two reactors, that means a 20 percent increase in electric rates. But getting a rate increase through the PUC is by no means easy. In fact, the PUC hearings could be almost as involved as the NRC licensing hearings for Comanche Peak. The PUC will consider whether or not TU management acted prudently with the rate payers in mind. That means they can question TU’s decision to build Comanche Peak in the first place, and to continue building it when problems began cropping up. If the PUC cannot be convinced that TU acted responsibly, they can disallow part of the cost of the plant from being passed on to consumers, or not allow TU to charge any extra at all. But that’s all speculation now; TU can’t apply for a rate increase until the plant begins to make electricity, and without an NRC license, that will never happen.

JUANITA ELLIS SEES ANOTHER BATTLE coming down like a cold front over the plains from the north. TU, which wants to load nuclear fuel in the summer of 1988 and begin commercial operation in early 1989, has requested that the ASLB resume hearings for Comanche Peak’s operating license; they should be well under way by early next year. The main issues Ellis will address are the quality control of the design and construction of the plant and the ability of TU’s management to handle the construction and operation of a nuclear plant.

Ellis says she plans to pursue the issues vigorously, but she is realistic. “Let’s face it,” she says. “There’s never been a plant where a license has been just flat-out denied.” The N’RC has denied licenses before, but it’s sort of like TU’s iterative design process-the utility repairs some problems and then asks the NRC for a license again, over and over, until either the plant gets a license or the utility runs out of money. “I can’t really see TU [going broke],” Ellis says. “If anything 1 think this utility is so stubborn they’d never do that.”

The biggest remaining questions about Comanche Peak are: will it go into operation? Is it sate? TU would give a resounding yes to both, and Ellis concedes that the plant will probably get a license if for no other reason than TU’s stubbornness. But will Comanche Peak ever be safe? “I don’t think so,1’ Ellis says. “Not at this point in time. I think it’s too late.”

“1 think they have changed it.” Dr. George Crawford, a nuclear physicist at SMU, says, “from a completely unsafe plant, into one in which at least the safety is a question mark. Because of the lack of proper supervision on the part of TU and the federal government, that plant is always going to be suspect.”

But it’s not about hardware and safety procedures that Ellis is mostly worried. She is afraid that the attitude of TU and its management, which she believes created all of the design and construction problems in the first place, will spill over to the operation of the plant when there will be highly radioactive material on site. Sitting at the kitchen table in her house that bulges with Comanche Peak documents, Ellis clasps her hands together and speaks in a tired voice. “There’s nothing to indicate that the people who are going to operate the plant are going to do a better job than the people who constructed it and designed it. Why should we have any more confidence in those people?”

Ellis is also getting tired. Her life has been completely absorbed by fighting Comanche Peak. She says her couch disappeared under boxes of documents, and several years later, in 1985, they threw it out, along with the rest of the furniture in the living room and all but one of their bedrooms. She and her husband now live in one bedroom and a kitchen with documents spilling into it from the living room, where metal shelves line the walls packed with boxes of filed documents. An IBM copier rests on top of their old console television set, and tens of thousands of pages of documents sit in stacks on the floor. Does she ever think about giving it up? “Only about three times a day,” she says, laughing. But she won’t and she knows it. “I have these visions.” she says, of growing old and “walk ing with a cane down the aisle into another hearing on Comanche Peak.”

The China Syndrome: Can It Happen Here?



THE THING THAT SETS NUCLEAR power plants apart from conventional power plants is the potential for disaster. That’s the reason nuclear plants are designed and built with multiple layers of protection; and why the government created the Nuclear Regulatory Commission to watch over the design and building of a plant every step of the way; and why watchdog groups contend that anything short of perfection in a nuclear plant is unacceptable. If there were no potential for disaster, Comanche Peak would have been built and in operation years ago.

A nuclear plant, like any electricity-generating plant, boils water to make steam that turns turbines that turn generators. It’s just that a nuclear plant uses the heat of a controlled nuclear reaction to boil the water. The reactor core, packed with rods of enriched uranium, sits inside a thick-walled steel vessel, which itself sits inside a massive steel-reinforced concrete containment building. Highly pressurized water circulates through the uranium rods, carrying away the heat to boil water and keeping the reaction cool-about 600 degrees Fahrenheit-and under control.

One of the worst things that could happen would be a loss of the cooling water, caused by a number of things: an earthquake breaking a major pipe; a welded joint that has become brittle with age and suddenly bursts; or a pressure relief valve that sticks open when the operator thought it had been closed, as happened at Three Mile Island. Without coolant the core begins to heat up rapidly, climbing thousands of degrees within seconds. Dr. George Crawford, a nuclear physicist at SMU, says the temperature “can go up into the millions. If it goes completely out of control, it could vaporize.” But before that happens the reactor computer should “scram” the core, dropping reactor control rods into the core and stopping the nuclear reaction. This is where there is potential for the second worst thing to happen. An operator could misread what is going on and manually stop the scram; the sensors in the reactor that feed into the computer system could fail, keeping the computer from scramming; or the control rods could hang up because the core is so hot it begins to melt and lose its shape.

Without control rods the core temperature soars into the millions of degrees, everything inside the core melts, and the reactor vessel explodes, which is what happened at Chernobyl. At Comanche Peak, as at all but a few American plants, there is a containment dome over the reactor that is designed to contain even a Chernobyl-type explosion. And here is the potential for the third nightmarish possibility: if the dome is weak because of poor construction or design, or if the seals over one of the many holes for wires and pipes passing through the containment dome do not hold, then the containment dome will burst and highly radioactive material will spew out into the atmosphere. Everything downwind will be coated with radioactive dust. But even that’s not the end of the problem at the plant.

The core, now melted, continues to react and make heat. If it gets hot enough, up to several million degrees, the molten core begins to eat through the bottom of the reactor and into the ground. That happened at Chernobyl, where workers quickly tunneled underneath the core and poured in massive amounts of cement in hopes of slowing or stopping it. Fortunately it stopped. Had the melted core reached the water table, a steam explosion would have resulted, with radioactive material being blown high into the atmosphere and the ground water contaminated. That has never happened.

The effects of such a disaster are difficult to predict and are constantly debated in the nuclear community. The only “experiment” of true disaster magnitude has been Chernobyl. Half of the ejected radioactive material fell within twenty-five miles of the plant. The other half rose three miles into the atmosphere in a plume of radioactive smoke and dust and was dispersed over half the earth’s surface. Thirty-one people died as a direct result of the accident, and the estimates for delayed cancer deaths range from 5,000 to 75,000.

In the event of a severe accident at Comanche Peak, Dallas will be threatened in several ways. If the wind is blowing from the southwest at the time (which it does thirty-three days of the year, according to the National Weather Service), then Dallas is more likely to receive large doses of radioactivity. Dr. Richard Webb, a nuclear industry consultant who helped solve the Three Mile Island accident, has estimated that up to 300,000 acres (the size of Texas and Louisiana combined) could be permanently ruined for agriculture in a worst-case scenario. Crawford of SMU simply says, “If something happens to that plant [Comanche Peak], it’s going to cover Fort Worth, Dallas, a tremendous area.”

John Beck, TU’s vice president for nu clear engineering at Comanche Peak, when asked if the worst-case scenario, the China Syndrome, could ever possibly happen at Comanche Peak, says, “That is not allowed by the regulations.” – J. P.