Loading summary
Transcript1 lines
- [00:03]
A
In the early 1970s, a small agency of about 81 people, backed by some of Washington State's local utilities, decided to build five nuclear power plants simultaneously. And it went as well as you might think with the country's biggest municipal bond bankruptcy up until then and four abandoned power stations. The story of Washington Public Power Supply System's attempt to pull off America's most complicated construction project is a fascinating mix of civil service bad estimates, nuclear immaturity and just plain mismanagement. In this video, a good old fashioned American nuclear energy debacle. This video is brought to you by the asianometry Patreon. Washington is a state in the Northwestern United States. They received statehood in 1889. Afterwards, its population boomed as settlers poured in from the east during the various gold rushes and and World War I. These larger populations needed water and power services and new utilities were founded to provide them. In major cities like Seattle and Spokane, those utilities were owned by the city and that helped keep their rates reasonably low. But the rural areas did not have that benefit. This left them vulnerable to private utilities that served them at extortionate prices. Their practices and monopoly status engendered anger and annoyance amongst the people. Urged on by a progressive organization called the Washington State Grange, the state authorities adopted Public utility districts or PUDs. PUDs are community owned non profit power utilities. Unlike regular private utilities subject to the State Utility Commission, puds are governed by commissioners elected by the local community. Over the next few years, these puds bought and resold power out of the federal hydropower system. Washington State has a rugged mountainous landscape. Winds blow moist clouds from the Pacific to these mountains causing them to drop their moisture in the form of rain. Lots of rain. These conditions are ideal for hydroelectric power. During the Great Depression, the US Federal government funded the construction of dams and hydroelectric power plants as part of its New Deal public works programs. Building these new dams led to questions about how the power would be distributed. So the government formed a bureau called the Bonneville Power Administration or the bpa. The BPA distributes and markets the power generated by the US Government's hydropower dams, reselling it to utility customers like the PUDS at cost. The PUDS in turn bring it to local residents while also handling customer service and other day to day things. These localpudz's heavy dependency on the power provided to them by the federal government led to concerns in the 1950s. That decade, the US federal government began to lose its taste for New Deal style involvement in the economy. In 1956, President Dwight Eisenhower pushed for changes in the funding model of the Tennessee Valley Authority, another federal agency similar to the bpa. Eisenhower, a fiscal conservative, saw the TVA and its ilk as socialism, and he halted several federal dam projects and even mulled selling the whole thing off. And if they can do that to the tva, then they can do it to the BPA too, which would leave the utilities high and dry. To head off Such a thing, 17 public utility districts formed a new agency a announced in mid-1956. The new agency was called the Washington Public Power Supply System. The name immediately received a fortunate acronym. WPPSS pronounced Whoops. I think it's important to make clear that wppss, let me just call them Washington Supply System was not a utility. Rather, their intent was to help shore up its member PUDS power generation sources by either purchasing existing dams from the federal government if available, or funding the construction of new ones. While the agency itself is authorized by state law, its projects are not funded by state taxes. Instead, the agency is authorized to issue non taxable state bonds to raise money. Washington Supply System started out as a small organization with little to do and and seemingly little reason to exist. Quickly, threepuds dropped their membership. The agency knew they had to find projects fast. As one manager colorfully put it, we better get pregnant or they'll dissolve us. An outside engineering firm hired as a consultant brought five potential jobs. One was a small hydroelectric project at the somewhat remote Packwood lake A about 25 miles outside of Mount Rainier, the namesake of those amazing cherries. In 1962, the agency issued about $13.7 million in bonds to help fund the Packwood Dam's construction. The dam project was completed two years later. Despite being delivered a few months late, the project was a big financial success. Now for this second part of the story, we're going to have to delve back a few years. During the Manhattan Project, the US Government set up a major plutonium production site along Washington state's largest river, the Columbia. The site was named Hanford, and its plutonium was used to manufacture two of the first atomic bombs. For a long time, hot water from Hanford's plutonium production reactors was simply returned to the river, which seemed wasteful. Why not use this waste heat to generate power for civilian purposes? In 1956, Washington's influential Senator Henry Scoop Jackson proposed a law to have the US government spend $65 million to build a dual purpose convertible reactor at Hanford. The bill faced substantial opposition from the Eisenhower administration and members of Congress who did not think that the Atomic Energy Commission, the US Federal agency then in charge of nuclear energy development, should be involved in generating power. So in 1956 it didn't pass. But by 1958, things had changed, in part thanks to the launch of Sputnik. The locals were also excited by the possibility of receiving energy from Hanford. So the proposal made it into a larger bill and Eisenhower reluctantly signed it, thus allocating money for the construction of a new plutonium reactor called the N reactor or new production reactor. Unlike Hanford's other reactors, this produced both plutonium for nuclear weapons and high pressure steam that can be used to generate electricity. Nobody in Congress opposed the plutonium part, lest they look weak on national security. But coal and private utility interests killed congressional funding for building the steam generating part of the facility. The Washington state government looked into picking up the slack. But the past as well. In early 1959, Washington supply system is in the midst of locking down the project for Packwood Lake. Flush off that success, Managing director Owen Hurd, with authorization from the board, starts investigating nuclear power. They eventually put themselves forward as a partner with the AEC to fund and build the 800 to 900 megawatt power plant. To get congressional opponents on board, Washington Supply System lobbies and offers to sell half the power generated to private utilities. With popular support rising, this revised Hanford generating project passes in 1962 and President Kennedy signs it into law. If you really think about it, the whole thing is a bit crazy. Washington's Supply System was just a few years old, was run by a thin staff of local utility operators and did not even have their own engineering staff. Yet it works out. Despite some labor issues, including some unofficial strikes and walk offs, Washington Supply System closely oversees the Hanford steam plant and shepherds it to completion. It begins operation in April 1966. At startup, the plant was the country's largest civilian nuclear power plant. By a large margin, it was a massive installation with eight pipes bringing hot steam to what were then the two largest turbine generators in the world. In 1967, Hanford single handedly accounted for 43% of the country's nuclear power. Washington sold $122 million of bonds at 3.26% interest rate to fund the construction. The project was finished below budget and the extra funds used to retire outstanding bonds. Despite the labor disputes, the whole thing can be called a pretty good success. Being associated with such a massive plant brought untold prestige. And soon enough, Washington's Supply System starts calling this generator hanford 1. A key assumption that Washington Supply System and the rest of the utility industry operated under in the late 1960s was that energy demand would grow. Per analyses and forecasts made by the aec, electrical loads in the region had grown at an average of 5 to 7% each year. And there seemed to be no reason that the trend wouldn't continue. The number of residential customers had grown 2.5 times between 1937 and 1967, and kilowatt hours those customers used grew from 1,200 to 12,200. Commercial and industrial customers loads were growing just as fast or even faster. Industrial customers used 90 times more power in 1967 than they were in 1937. Nobody questioned the assumptions behind these trends or the validity of the projections. Utilities instead worried about where to get the power to meet those needs. The Pacific Northwest region lacked reserves of good fossil fuels, and the number of ideal sites for new hydroelectric power plants was dwindling. So the only scalable alternative seemed to be uranium. In the late 1960s, the BPA's administrator brought together all the region's utilities and and discussed the need for nuclear energy to meet all that demand. After several years of wrangling, the group unveiled the Hydrothermal Power program, a sprawling 20 year plan covering the years from 1970 to 1990. Phase one called for seven large plants to be built from 1970 to 1980, a mix of coal and nuclear plants. With the BPA, by law, unable to build them. Washington's supply system quickly found itself responsible for building three nuclear plants. First WNP2 committed to in 1971, and then WNP3 and WNP1 in 1973. Looking back, it feels like an overreach, but it is important to note that Washington Supply System felt a civic duty to go where others cannot. The Hanford One job, which both the US Federal and Washington state governments turned down, is a prime example. Moreover, it was the start of the golden age of nuclear energy. After an initial hesitancy, utilities were adopting it wholesale. With 31 plants announced in 1967 alone. The TVA announced 17 by themselves. And even if Washington overbuilt a little bit, what's the big deal? The thinking of the time was that nuclear would make power too cheap to meter. Any extra capacity can be used to lure another aluminium smelter to town. Finally, the BPA threw in another safety net called net billing. Explaining this in full would be a bit tiring. The short version is that net billing means the BPA commits to buying nuclear power more expensive than its hydropower and spreads out the cost difference across all of its utility customers. Net billing is meant to help spread out the high cost of building new power generating sources across a wide population. And from a development standpoint, I get it. But it also commits every utility customer to the plant. And critically that holds, even if the plant is never built. Basically, as soon as things got rolling, changes had to be made. In October 1970, Washington Supply System had to relocate the 1,100 megawatt WNP2, initially estimated to cost $500 million, from its original planned location at Roosevelt Beach National Park. Staffers wanted a six year study to see whether the plant's hot water would affect the local clam population. So WNP2 was relocated to Hanford, where the environmental studies had already been done. A few Months later, in January 1971, Washington Supply System received an unpleasant call from the federal government. The Nixon administration was cutting spend and with plutonium needs declining thanks to arms agreements, they wanted to shut down Hanford One's N reactor and the 860 megawatt power plant attached to it. A furious lobbying effort followed. Various players in Washington State said that the shutdown would create a power crisis in the area. Ordinary people sent letters to the government asking to reverse the shutdown. A compromise was reached in mid-1972. The old N reactor can run for a few more years. But since its weapons capability cannot be practically removed, the whole thing had to go. So to prevent a permanent loss to the grid, Washington Supply System pledged to build a replacement WNP1 to replace Hanford One. That is how Washington's Supply System ended up committing to two nuclear projects in 1973. And why WNP2 and WNP1 both happen to be at Hanford, while WNP3 is at the town of Satsup, west of olympia. And why WNP2 was scheduled to start before WNP1. Washington supply system now had to manage a budget that had abruptly tripled from $500 million to $1.6 billion. For an organization of just 81 employees in 1971, this is a substantial ask. They rapidly hired and moved to a new office to keep up with the monumental task of of three nuclear power plant builds. At the same time. They had to adjust to the requirements of a new era. The 1970s marked the rise of the environmental movement, leading to new regulations issued by the AEC leading to constant design changes. The newly passed law, nepa, also necessitated extensive studies and environmental impact reports that cannot be ignored or downplayed, especially in the wake of the momentous Calvert Cliffs court decision. Finally, the net billing safety net broke. A 1972 tax ruling by the IRS regarding non taxable bonds, plus the rapidly rising costs of the hydrothermal power program. Hitting BPA customers meant that net billing can no longer be used for nuclear projects after WNP 1, 2 and 3. So tell me why, oh, why did Washington's supply system then decide to add another two nuclear reactors? It ties back to the energy crises. In the late 1960s, people talked of an impending energy crisis. Then in the 1970s, it seemed like those days had come. A light snowpack in 1972 meant less water for the region's hydropower dams and causing factories to shut down. Then in 1973 was the Yom Kippur War and the resulting oil embargoes. With the gas rationing, conservation programs, etc. You know the deal. People at both the BPA and utilities believed that 1973 would be just the beginning. The only thing to do was to push forward to phase two of the hydrothermal power program and build more nuclear plants. The hope was that twinning WNP4 and WNP5 with WNP1 and 3 at Hanford and Satsup can realize some $400 million of cost savings. And at the time, the feeling about progress at the three plants was that they were slow but on the right track, a feeling that later turned out to be incorrect. With net billing no longer available, utilities were asked to commit themselves to building the plants and then estimated to cost $2.25 billion with the BPA's aggressive help, including a potential threat of shutoff by 1983. 88 utilities signed an agreement by 1976. One major exception to the trend was the big utility Seattle Light, which owned its own hydropower stations and was not heavily dependent on the bpa. They asked the Seattle City Council for authorization to buy into the WMP4 and 5 projects, and the council voted no, saying that it cost too much. So that is how Washington's supply system committed to five nuclear power projects in five years. And as it turned out, many of the assumptions motivating that urgency were incorrect. First of all, the Washington State region was not hurtling towards a power crisis. As they like to say, past performance is not indicative of Future returns. The 5 to 7% annual growth forecasts had been prepared by a body called the Pacific Northwest Utility Conference Committee, which compiled historical data from utilities and made a crude projection. But some of the historical growth can be attributed to the fact that the Pacific Northwest had some of the lowest power costs in the country thanks to its plentiful hydropower dams. No sophisticated economic Analysis was done to see what might happen if the price of power went up, and it came to be that rising energy prices and good conservation programs blunted that demand curve. During the Seattle light drama, an outside consulting firm was brought in to do their own estimates of future electricity demand. And what they came up with was 1.5%, far lower than even Seattle's low 4 to 5% forecast. The second major miscalculation was believing that nuclear power plants would get cheaper as they got bigger. Back in the 1960s, GE told potential customers that net cost per kilowatt for a 1,000 megawatt plant would be half that of a 200 megawatt plant. There had been no data to back this assertion, and as they tried to build these new facilities, they found it was not true. In 1967, the AEC estimated that the 1100 megawatt plant would cost $134 million. Two years later, that was revised up 79% to $240 million. That price trend would only go up in the coming years. But those do not excuse the fact that the single biggest factor behind the debacle was Washington Supply System's struggle to manage the project. Within the restraints put upon it. Washington grew its staff to 500 and then 2,000 people. But even so, they were ill equipped to handle what was without question the biggest megaproject in the United States at the time. These five plants were rushed to construction with no attempt to achieve economies of scale. Consider the French nuclear industry. Many of the French plants built in the 1970s had the same 900 megawatt design with the same components and shared a single supply chain. Washington Supply Systems five plants, on the other hand, were of three different designs done by three different designers. It was noted that a state law nominally prevented consolidation of the jobs and and that no standard design then existed. Nevertheless, the savings would have been substantial. An EDF was a massive organization run by a tight knit group of educated elites, and they kept a tight grip on the project, embedding their own people at the engineering sites, and strictly refused to take on design changes even when they seemed good on the surface. Washington's supply System, perhaps overwhelmed, exercised lax control over its contractors. They elected three separate architect engineers, apparently to spread out the risk, who then struck different contracts with different supply and engineering firms. Contracts were awarded while the designs were still being finalized, and contractors were allowed to make changes on the fly, leading to careless modifications, component delays, lots of waiting around, shoddy work, and extensive rework. One concrete and steel job was initially estimated at $40 million. But the moving goalposts caused its price to balloon to well over $200 million. In another major incident in June 1980, the Nuclear Regulatory Commission, or NRC, fined Washington Supply System $61,000, where they discovered that the sacrificial shield wall, a protective barrier between the reactor vessel and the containment buildings, was done profoundly, terribly and would not survive an accident. The issue was that bad work was being done on top of bad work. Inspectors reported cracks in the wall's structural beams since 1977. Even so, two contractors built on top of its flawed structure. This is not on the workers, but their supervisors. Washington initially thought the fix would take three weeks and ended up taking 13 months of 24. 7 work by over 100 workers. And then there was extensive labor strife between the contractors and the labor unions doing the work. In 1976, a major strike by the Plumbers and Steamfitters union shut down work entirely for four months, costing up to $500,000. Each day, Washington's Supply System struggled to mediate between the contractors and its unions. The unionized workers felt that they were being blamed for what they thought was incompetent management. And they are, of course, right. But it was a two way street. And I should note that this was the 1970s, a time of volatility, and the workers were not doing their best work. They did whatever they can get away with. Walk offs and strikes continued throughout the next few years. Between April 1978 and July 1979, there were over 100 work stoppages or secondary boycotts. And at Satsop alone, there were 18 labor disputes in just 1979. And then there are the things that just make you sad. The hope had been that twinning the number four in five stations would save money. But the time between construction start of these two build phases were so close together that none of those benefits were ever realized. In August 1978, heavy rains in Sapsop washed away all the excavation work done for WNP 3 and 5 into a nearby creek, adding $51 million to the cost. The area gets an average of 90 inches of rain each year. This should have been accounted for. And then in early 1979, came a double hit. First, the movie thriller called China Syndrome, depicting a possible nuclear reactor meltdown. And 12 days after the movie was released came the nuclear incident at Three Mile island, an accident that shattered the credibility of the American nuclear industry and led to a massive revamp. In the NRC's overview, Washington supply System employees later blamed the NRC and its safety regulations for 50% of the project's overruns in its first four years. This estimate has been contested, with others instead blaming Washington's general mismanagement. The estimated price tag of $4 billion for all five plants had been too optimistic from the very beginning. Even as utilities were committing to plants 4 and 5, the strikes, delays and new changes were blowing the original estimates out of the water. By 1978, the plants were delayed by two years and costs went up to $8.5 billion. In 1979, that had gone up to $11.7 billion. By the end of 1980, the time delay had stretched to 13 years and cost to $17.3 billion. Those costs had to go somewhere. Rates rose very quickly from 0.351 cents per kilowatt hour in 1979 to to 2.2 cents in 1983. A significant portion of those rises can be attributed to the power plants. This was 1979. Per polls, the number one issue that people cared about that year was inflation and energy. The absolute worst thing to be doing was raising rates. At the same time, Washington Supply Systems interest rates rose from 5.93% in 1977 to 11.77% in 1981. Part of this was the inflationary environment. But there were also concerns that they were selling too many bonds. From 1978 to 1980, 20% of all the gas and electric bonds came from Washington alone. By 1980, they were selling $200 million every 45 days. Half of the company's costs were just the interest on its debt. Worse yet, the people buying those bonds were not given proper information about the state of the project. In late 1979, a mutual fund manager reportedly toured the site himself and then immediately sold every Washington Supply System bond they had. Yet Washington kept selling $200 million worth of bonds every time, almost right up until the end. The early 1980s saw the US economy hit by a sizable recession. Unemployment rates in Washington state doubled from 1979, lumber mills closed and electricity demand plunged. With the rough economic situation, there was no possible justification to finish all five power plants. In early 1981. The new Washington Supply System director, Robert Ferguson, orders a ground up cost estimate for the project. The new estimate puts the cost of completion at $23 billion, or about $2086 billion today. As one estimate put it, enough to build the Great Wall of China with union labor. Ferguson realized that the organization simply cannot do it and recommends that the board put a one year moratorium on construction for plants four and five. The board is shocked, feeling that they had not been told about all this, but agrees the fallout is immediate. Wall street analysts predict that plants 4 and 5 will not be completed. Thousands of workers in Hanford are laid off, and they protest. The ratepayers rose up in fury, suing their local utilities to keep them from paying for the plants. In January 1982, WNP 4 and 5 are canceled, but of course, the bondholders still want to get paid back for the bonds they bought for those projects. The utilities and their ratepayers were on the hook, in some cases up to $12,000 per customer. However, the Washington Supreme Court makes a somewhat dodgy ruling that the utilities and energy cooperatives did not legitimately enter the agreement for four and five, releasing them from their obligations to the project. So in August 1983, Washington's supply system defaults on the $2.2 billion of bonds issued for Stations 4 and 5. It was the largest municipal default in American history up until then. As with any lawsuit, the bondholders sued, alleging malfeasance, incompetence and fraud. The lawsuit dragged on for six years and was finally resolved in late December 1988, with the 75,000 bondholders getting 10 to 40 cents back for every dollar they put in. The lawyer certainly made out well, earning $33 million, though they asked for 100. The estimate to finish WNP 1, 2 and 3 was over $12 billion, a ridiculous number. They raised $1.6 billion in an attempt to continue before deciding to cancel 1 and 3 with both at over 60% completion rate. So they can just focus on WNP 2. WNP 2 finally opened in 1984 at the cost of nearly $3,000 per kilowatt. To compare, the French had an average weighted cost of 1,200 per kilowatt. It is now called Columbia Generating Station and has worked fine without much drama, which is exactly what you want of a nuclear power plant. In 1995, the remains of WNP 1 and 3 were demolished. The two plants at Satsup are still out there and are frequently the subject of YouTube exploration videos. Finishing the first nuclear plant somewhat redeemed Washington's supply system, but a lot of the Baggage remained. In 1998, they renamed themselves to Energy Northwest after paying $260,000 to deal with a trademark issue. Since then, they have kept their heads down successfully developing other renewable forms of energy, like the Nine Canyons wind turbine project and the Horn Rapids solar projects alongside Packwood and Columbia. A troubled start, but a reliable cornerstone since. Alright everyone, that's it for tonight. Thanks for watching. Subscribe to the channel. Sign up for the Patreon and I'll see you guys next time.