Troubled throughout its entire history, the Columbia Generating Station nuclear power plant at Hanford, Wash., will turn 30 years old in 2014. The plant’s opening in 1984 was marred by the failure to complete four of its sister plants in what at the time was the biggest municipal bond default in the history of US capital markets, not to mention serious flaws in plant design, construction and plant management. Plant managers say these early defects were eventually corrected. But after surviving its initial problems, further mishaps propelled the plant into a lifetime riddled with maddening human errors, mechanical breakdowns and substandard (or worse) operating performances.
By Paul Koberstein
and Robin Klein
A review of the plant’s 30-year record by Cascadia Times reveals that Columbia Generating Station today is one of the most dangerous nuclear power plants in America, if not the most dangerous, based on its poor safety and environmental record compared with the rest of the industry, and the deteriorating condition of some key components. It has settled into a mode of mediocrity and complacency. This series of reports will present evidence for this conclusion. ’
When approached from a nearby highway, the Columbia Generating Station appears to be missing the iconic concrete dome structure that’s exposed at most nuclear power plants. But the Columbia nuclear plant does have such a structure, known as the containment dome. It’s just hidden behind the nondescript sheet metal shell of a tall, cubical building (pictured above) that is surrounded by several smaller buildings, a small army of heavily armed security guards and a series of tall, barbed-wire fences.
The reactor and containment dome were designed by General Electric, the world’s second largest corporation which has become one of the two leading industrial players in the nuclear power industry, the other being Westinghouse. As at all nuclear plants, the reactor is the workhorse, creating heat from controlled nuclear reactions. The heat drives turbines, which generate electricity.
The reactor is housed within the containment dome, a cylindrical structure that is about three stories tall with 3-foot-thick reinforced concrete walls, and is shaped like an upside-down incandescent light bulb. The containment dome’s walls must be strong enough to withstand the build-up of extreme gas pressures on the inside, as well superhot temperatures and dangerously radioactive particles and gases that come out of the reactor. They must also be stiff enough to withstand any hurricane force winds, tornadoes, earthquakes and airplane crashes that could strike from the outside.
Finally the dome must shield the public and the surrounding environment from the radioactivity, although by design it does allow a tiny amount of these poisons to slip through.
The thick-walled dome is surrounded by a thinner-walled secondary containment building. Together the primary and secondary structures serve as the plant’s last line of defense in protecting the outside environment and the public in an accident.
On March 11, 2011, similarly-designed containment domes at the Fukushima Daiichi nuclear power plant in Japan failed to contain their toxic airborne contents. Hydrogen gas accumulated in significant quantities inside the primary domes at units 1, 3 and 4, reaching pressure levels that exceeded the structures’ ability to contain it, and leaked into the secondary containment buildings. At each unit, the gas caught a spark and exploded, blowing the roofs off of all three secondary containment buildings. The explosion at Unit 4 exposed spent fuel rods stored there to the atmosphere, creating an additional radioactive hazard.
Meanwhile, the nuclear fuel rods in reactors in Units 1, 2 and 3 melted when the power was cut off to the pumps that supply water to the reactors to keep them cool. Without a flow of coolant, the reactors became too hot and their nuclear fuel melted. The news media has blamed this catastrophic accident, one of the two worst nuclear accidents in world history, on a 9.0 earthquake and a 45-foot tsunami that hit the eastern coast of Honshu.
There’s no doubt that those natural disasters triggered the horrifying catastrophe that ensued, especially when they wiped out all power to the plant. But design errors in GE’s proprietary pressure containment system may have also played a significant role in exacerbating the damage.
The Fukushima accident released unfathomable amounts of radiation in the direction of a large human population and into the air, groundwater and Pacific Ocean.
When GE first introduced its nuclear power plant technology to the world in the 1950s and 60s, GE’s own engineers and management knew the containment design was deeply flawed and in need of testing and upgrades, according to memos introduced in lawsuits against GE that were filed by several American utilities that bought the reactors.
Dale Bridenbaugh, a GE engineer who testified in one of the lawsuits, said in a recent interview that the designers had failed to take into account all the dynamic forces that can boost gaseous pressures inside the containment dome during an accident. In 1977 and 1980, the NRC issued orders requiring American plants that had installed GE’s defective equipment to make certain retrofits.
The orders, however, did not apply to foreign GE-designed plants. Worldwide there are 87 GE-designed reactors like the ones that exploded at Fukushima, including 35 in the US, 25 in Japan, 7 in Sweden, 6 in Germany, 4 in Taiwan and 2 each in Mexico, Finland, Switzerland and Spain, according to a database maintained by the International Atomic Energy Agency.
It is not clear whether the six Fukushima reactors or any of the other foreign plants were ever modified to correct the containment defects, despite a statement that GE issued March 16, 2011, four days after the Fukushima accident.
In this statement, GE only went only so far as to assert that it understood that modifications were indeed implemented at Fukushima. “We understand that all (GE-designed) containment units at Fukushima Daiichi … addressed these issues and implemented modifications in accordance with Japanese regulatory requirements,” the GE statement said, which is published on its web site.
But judging from the tepid wording of the statement, “We understand…” GE did not say with certainty that the retrofits were in fact made. It is less clear who, if anyone, made those retrofits if they were made, and whether there were any follow-up inspections to ensure that they were done properly. It seems reasonable to assume that If GE had been able to speak with any certainty, or could assure the public that it had inspected the work, it would have done so. GE did not respond to an email seeking clarifications of this issue.
These containment weaknesses had been detected in General Electric-designed plants that were built before 1980. The three GE-designed Fukushima plants that exploded were built from 1967 to 1978. GE waited until 1974 before publicly disclosing it knew about the weaknesses in its containment domes, stating that “in certain accident and non-accident situations,” the domes could be subjected to pressures “that could structurally damage the steel containment and the equipment inside it.”
But the company refused to pull their reactors off the market and the NRC did not order any recalls. Instead, the NRC ordered the dozens of nuclear plants that had insufficiently strong containment domes to install billion-dollar retrofits.
In the 1980s, several US utilities that bought the defective reactors sued GE. The long list of plaintiffs included Long Island Light, owner of the now-defunct Shoreham plant on Long Island Sound, and the Washington Public Power Supply System, owner of Columbia Generating Station.
Federal District Judge Alan A. McDonald ruled in the WPPSS lawsuit that GE “didn’t fully understand” phenomena “that it had been observing in tests since 1958.”
By selling reactors that it knew were defective, McDonald said that GE had been playing “a sophisticated form of Russian Roulette.”
Russian Roulette is a potentially lethal game of chance in which a player places a single round in a revolver, spins the cylinder, places the muzzle against his head, and pulls the trigger. In his analogy, the gun was the nuclear plant and the bullet was the vast amounts of radiation they can release in an accident.
Moreover, McDonald said that GE “hid serious doubts” about the safety of the reactors it sold to WPPSS. “The concealment constituted bad faith and nullified (a) contract provision limiting GE’s liability,” the judge said, as was reported in the trade publication Nucleonics Week.
These defective, inadequately tested machines were installed in the backyards of communities from rural Washington state to upstate New York without the public’s knowledge of their flaws. The people living near these plants became human guinea pigs in GE’s appalling product development research experiment, raising ethical questions about the way the company’s brand of commercial nuclear power plants were introduced to the world.
Documents filed in McDonald’s courtroom show that from 1958 to 1974, as GE engineers were still trying to work out bugs in GE’s proprietary “boiling water” reactor design at the same time that the company’s sales force was out selling boiling water reactors to unsuspecting electric utilities around the globe.
One of these engineers was the aforementioned Bridenbaugh, an ex-manager in the company’s nuclear division who, along with two other engineers, abruptly quit their jobs in 1976 because they were worried that the plants were not safe. Together the engineers were known in media reports at the time as the “GE Three.”
Later, Bridenbaugh testified before Judge McDonald as a witness in the WPPSS lawsuit. In a recent interview, Bridenbaugh referred to federal Nuclear Regulatory Commission regulations that require containment systems at nuclear plants to constrain any radioactive material that might be released, with no significant leakage.
“We weren’t certain that those plants would be able to survive the accidents that they were supposedly designed to contain,” he said, echoing the thoughts of other GE engineers who made similar statements as early as 1964.
As a project manager, Bridenbaugh said he was assigned the task of evaluating data on the dynamic forces that accidents could create inside the containment structures at the 23 GE plants then in operation in the United States.
“I went to my boss in January 1976 and said if these numbers that we are generating right now are correct, we’ve got to shut these plants down,” he said. “I worked diligently on this until February, 1976 when I became disillusioned and resigned.”
He said he became disillusioned because both GE and the federal Nuclear Regulatory Commission had decided to allow plants to remain open while they awaited repairs. “It was obvious they no longer met licensing criteria.”
Bridenbaugh wasn’t the first GE engineer to sound the alarm, according to several internal GE documents presented in McDonald’s courtroom. For several years, these documents were sealed by the courts, but were later unearthed by Daniel Pope, a professor at the University of Oregon who dug them up in a Lexis-Nexis search while doing research for his 2008 book, “Nuclear Implosions: The Rise and Fall of the Washington Public Power Supply System.” Pope shared the documents with Cascadia Times.
One engineer in 1968 documented the existence of unexplained vibrations inside the containment dome. Another 1970 disclosed the existence of “severe jumping and banging” within the dome during tests. In one demonstration, engineers saw the equipment inside the containment dome “literally jump off the floor,” according to a New York Times report in 2011.
On August 31, 1970, the manager of GE’s Advance Systems & Analysis Design unit issued a memo regarding “Outstanding Containment Questions.” The document discussed a number of issues in the containment design of the plant that GE sold in 1971 to WPPSS. The paper said GE was trying to “dump” the containment problem onto WPPSS, but a “favorable outcome” would have required a “GE rescue operation.”
The paper predicted that a solution “would not be terribly expensive and would not come out of GE’s pocket.” In 1972, another GE engineer stated that observed weaknesses in the containment dome “have not been considered in any containment design.” In other words, GE engineers who designed the containment domes had gotten the math wrong when they calculated the amount of internal pressure they would be expected to withstand. Unfortunately they erred on the low side.
WPPSS lawyers uncovered memos from GE’s engineers that specifically questioned whether the containment structures could remain stable during a “loss of coolant accident (LOCA)” scenario eerily identical to the one that unfolded a half-century later at Fukushima in which the plant loses all capability to keep its reactor cool and constrain its rising temperatures before a meltdown.
At Fukushima, a total loss of power killed the pumps that were designed to deliver coolant to the reactors. GE sold six models of boiling water reactors and three models of containment structures, the Mark I, II and III. WPPSS bought a Mark II containment system for Columbia Generating Station in 1971.
In 1975, when construction of the containment dome at Columbia Generating Station was more than 80 percent complete, the NRC began an inquiry into the defects in GE’s containment domes, and eventually required several utilities, including WPPSS, to rip their domes apart and rebuild then to meet new safety standards. WPPSS began its retrofit in March 1977, delaying plant start-up by 18 months and costing about $1.2 billion in construction and redesign expenses.
The defects and deceptions cited by McDonald in the WPPPSS case were typical of GE’s corporate behavior, judging from evidence presented in other, similar litigation involving other GE-designed plants. A news article in the New York Times in 1987 about a lawsuit filed by three Ohio utilities said the utilities claimed that General Electric sold them a reactor “knowing it was flawed.” The WPPSS lawsuit accused GE accused of fraud, misrepresentation and breach of contract for failing to provide it with a “safe” nuclear reactor.
GE had a simple motive to hide the flaws and deceive its customers: money. From 1972 to 1974, GE sold 40 nuclear plants to utility customers in the US. It was not about to ruin a bustling business by being truthful about defects in its products.
For example, in 1975 a GE engineer named Henry E. Stone wrote a memo about the variety of failures, technical problems and serious structural defects at GE reactors had not been resolved. The memo was labeled “strictly private” and “GE confidential: Subject to protective order, Zimmer litigation.”
But Stone’s supervisor, A.J. Bray, general manager of GE’s nuclear reactor division, was taken aback by what he described as the memo’s “negative tone.” “If any of our customers ever get a copy of this, we are in real trouble,” Bray wrote. “All of the comments may be true, but why does GE have to put it into print to ruin a business?”
After hearing the case for seven years, Judge McDonald, a Reagan appointee to the federal bench, concluded that GE did not disclose its doubts about its reactors to WPPSS because it “was concerned about its market position, profits and potential liability.”
General Electric did not respond to a request for comment on this story. However, Reginald Jones, CEO of GE in the 1970s, assured a group of security analysts in 1975 that he saw nuclear power as the future of energy and that GE would continue to invest in it. “And as long as we can make these investments, and contain our risks, then we’re going to continue with this strategy.”
The “Zimmer litigation” referenced by Stone’s memo was a lawsuit filed by Cincinnati Gas & Electric over defects at its Zimmer Nuclear Plant, located on the Ohio River east of Cincinnati.
GE never intended for the Bray memo to be released, but it was filed along with several other documents in open court by lawyers for the plaintiffs. GE said this release was a breach of a protective order, which GE had expected would ensure confidentiality.
But an alert reporter for the local newspaper took notice, and soon stories began appearing about an alleged “12-year cover-up” of a “secret report” which contained “undisclosed safety problems,” according to a two-volume, 1200-page history of GE’s nuclear plant problems. published by the company in 1987.
GE’s history accused the newspaper of publishing “misleading” articles which “raised concerns in communities where GE (reactors) are in operation.” The confidential documents also proved troublesome when utilities and public officials in other states where GE-designed plants were located, began demanding copies.
GE’s semi-exhaustive history book neglected to give credit to Bridenbaugh and other whistle-blowers who raised questions about the plants’ safety in the 1960s and 1970s, leading to the retrofits.
Meanwhile, the Zimmer plant never produced a single watt of nuclear power. Before the plant opened in 1983, it was converted to coal.
In 1990, the WPPSS lawsuit was declared a mistrial after a jury wasn’t able to reach a unanimous verdict. Judge McDonald ruled that WPPSS could change its complaint against GE to “negligent misrepresentation” rather than fraud and breach of contract.
A second trial was about to start in 1992 when a settlement was reached. As the Seattle Times reported at the time, GE settled the case for $134.9 million worth of goods and services, but paid no cash. However, GE agreed to increase the power output of the WPPSS reactor by 50 megawatts, an increase that could generate about $16.5 million worth of electricity in a year. In 1999, GE agreed to pay an additional $30 million in cash and provide certain other nuclear plant services as part of a renegotiated settlement agreement.
WPPSS’ lawsuit attracted little media attention at the time it was tried in the late 1980s. For one thing, the trial was conducted in Tucson, Arizona, far away from the glare of the Seattle news media and Seattle’s vocal anti-nuclear community.
For another, everyone was much more interested in the sexier story of WPPSS’ massive bond default. More than $2 billion had vanished and retirees lost their retirement savings in the collapse. WPPSS had tried to build five nuclear power plants at one time, but only one, the Columbia Generating Station, was ever finished. The skeletons of the other four continue to rust across the rural Washington landscape, stark reminders of the times.
By 2011, two decades after Judge McDonald issued his warning about General Electric’s game of Russian Roulette, and a half-century after GE’s engineers had initially expressed their concerns, the lawsuit had been long forgotten. Then the Fukushima plants exploded. The Russian Roulette bullet had finally gone off.
The Fukushima accident proved to be an enormous disaster for the 300,000 people living in Fukushima, a city on the east coast of Honshu, Japan’s largest island.
The Fukushima accident is contaminating the North Pacific Ocean with radiation. In June 2012, 56 percent of Japanese fish catches tested by the Japanese government were contaminated with cesium-137 and –134. Cesium levels near the Fukushima coast shot up to an astonishing 45 million times the pre-accident levels.
Pacific bluefin tuna with elevated levels of radiation were caught off the California coast.
The people living near nuclear accidents or releases are often called “downwinders” because the air they breathe has often been contaminated by pollution from a source located upwind. Residents of the Tri-Cities area near the Columbia Generating Station know well what it is like to be a downwinder. Since World War II, they have lived downwind from the highly polluted Hanford Nuclear Reservation and its now-closed nuclear reactors and bomb factories. Some have suffered a series of serious health problems as a result.
A Fukushima-like explosion at the Columbia Generating Station could bring similar consequences. It could lead to serious health effects throughout the Pacific Northwest, forcing massive evacuations of cities and towns, contaminate the Columbia River and its salmon runs, and render vast stretches of prime agricultural land worthless for centuries.
Large portions of the United States are potentially at risk as well. Most of GE’s nuclear reactors are located near population areas east of the Mississippi River. More than 58 million people live within 50 miles of a GE nuclear reactor in the US.
The nuclear power plants discussed here are known as boiling water reactors. Today, there are 35 boiling water reactors currently operating in the United States and five others that are defunct. Each was designed by General Electric. Many of the other 65 operating plants in the US are known as pressurized water reactors and have had serious problems themselves. Westinghouse, a major manufacturer of these competing designs, itself has had to fend off a series of lawsuits filed by its customers.
Meanwhile, another serious unresolved problem with GE containment systems has emerged: the discovery of gas bubbles trapped in the pipes of the reactor’s emergency core cooling system. These bubbles can disable or damage the system’s pumps when they are trying to cool the superheated reactor during an accident.
If the pumps ingest gas, they “may become inoperable,” according to a study of the problem by scientists at Purdue University. A pump that swallows air can overheat, causing its casing to thermally expand, exceeding tolerances, the NRC has found.
“Since these components typically have tight tolerances, a significant amount of thermal expansion will cause these tolerances to be exceeded,” said an NRC engineer, who asked not to be identified for fear of retaliation by his employer.
Despite more than 10 years of study, the NRC has not yet found a solution to the gas bubble problem. Nevertheless, it continues to allow the GE plants keep on running. The results could be catastrophic.
The Fukushima accident triggered a new wave of concern at the NRC, which on March 19, 2013, ordered further retrofits at the Columbia plant and the other similar plants in the US. The order required the plants to install vents that would relieve some of the pressure in the containment domes during a nuclear accident.
But instead of requiring the plants to also install filters to prevent radioactive particles from also escaping through the vents, as recommended by the NRC staff and anti-nuclear activists, utilities will be allowed to consider alternatives. The Commission overruled a staff recommendation and decided, in a 4–1 vote, not to require the filters because of opposition from the nuclear power industry.
Only Allison MacFarlane, chair of the commission, voted in favor of the filtered vents.
As Paul Gunter, of nuclear power opponent Beyond Nuclear, said, the non-filtered vents will pose a safety risk to to anyone living near a GE-designed plant. “Venting an accident without a filter,” he said, is like “fire-hosing downwind communities with massive amounts of radiation.”
Gunter also noted that GE’s plants “are aging and deteriorating with fundamentally flawed containment systems. They are inherently dangerous. These reactors should be immediately closed.”
David Lochbaum, a nuclear engineer with the watchdog group Union of Concerned Scientists, said “the filters would remove 99 percent of the contamination.”
Moreover, Charles K. Johnson, director of the Joint Task Force on Nuclear Power for Oregon and Washington Physicians for Social Responsibility, a group that is calling for the permanent closure of the Columbia Generating Station, stated, “this half-measure upgrade is unlikely to prevent a hydrogen explosion and a massive release of radiation in a worst-case accident scenario.”
These comments echo similar views expressed 40 years ago by top federal nuclear regulators. The New York Times reported in 1972 that Stephen H. Hanauer, then a safety official with the Atomic Energy Commission, a predecessor of the NRC, recommended that GE plants be discontinued because they presented “unacceptable safety risks.”
In 2011, The Times noted that among the concerns cited by Hanauer “was the smaller containment design, which was more susceptible to explosion and rupture from a build-up in hydrogen — a situation that may have unfolded at the Fukushima Daiichi plant.”
Also in 1972, Joseph Hendrie, who in 1977 became chairman of the Nuclear Regulatory Commission, said the idea of banning GE-type reactors “was attractive,” the Times reported. But Hendrie added that a ban on GE’s technology “could well be the end of nuclear power.”