Monday, May 21, 2018

Nuclear Power part 3: Economics

                                                                                                written 24 March, 2018
                                                                                                published 31 Mar 18

            This third article on nuclear power focuses on economic costs.
            The earthquake and tsunami at Fukushima in 2011 showed that, within hours, a nuclear facility could transform from a $60B asset to a $700B, multi-decade, liability.  But nuclear power has never made economic sense. 
            To justify the billions of dollars invested in developing the atom bomb during WW2, the US government created the Atoms for Peace program in 1954, promising atomic energy "too cheap to meter".  Because of potential liability issues, no bank was willing to invest in this new technology, so the Price Anderson Act was legislated to limit corporate liability to $450M per reactor accident.  Capitalizing profits, and socializing losses, is a mark of dualistic economics.
            As another corporate subsidy, the government took on the responsibility for the disposal of the high level radioactive waste, charging a modest fee to the utilities.  The disposal fund currently totals $30B, but the disposal problem remains unsolved, with more than $40B spent on the defunct Yucca Mountain facility. 
            There is an inherent conflict between safety and cost.  When the first reactors were being designed, nuclear power was not well understood, and the constant design changes were expensive.  A decision was made to include a known weakness in the design in order to expedite construction.  The Fukushima reactors are a version of that design, and their containment failure, after total loss of cooling water, was a risk known for decades.  There are 35 reactors of this design currently in operation, including 23 in the US.
            In the 70s, when nuclear power was all the rage, Washington Public Power System committed to build 5 reactors.  By 1983, with only one plant completed, cost overruns, construction delays, and declining demand forced a mid-construction halt on two plants, and complete cancelation of the last two planned plants.  This caused the largest default on municipal bonds in US history.  Similar delays and cost overruns on current projects in Georgia and South Carolina, caused Westinghouse to declared bankruptcy last year, and the parent company, Toshiba, was forced to sell lucrative chip divisions to stay solvent.  Customers have been billed for years for nuclear power they will never receive.
            In California, the San Onofre reactors were shut down due to unexpected expenses for steam generators failures, leading to discussion about who will pay remaining debt and decommissioning costs, the ratepayers or the stockholders.  At Diablo Canyon, the remaining reactors operating in the state, PG&E agreed to a shut down when the license expires in 2024, and invest in renewables instead.  Operating an existing nuclear plant is no longer cost-effective compared to other power sources, and six US reactors have been shut down in the last 5 years.  Last year alone, 10,000 MW of solar was installed in the US.
            A standard 1000 MW nuclear plant, operating at 90% for 40 years, will cost about $10B to build, another $1B to be refueled every 18 months, and at least $750M to decommission.  This gives a lifetime power cost of $0.038/KWhr.  A utility scale solar array can be installed for $1.34/W, operate for 20 years, averaging 5 hr/day full capacity.  This gives a lifetime power cost of $0.037/KWhr.
            Nuclear plant construction costs have consistently run over budget and time schedule, because they are complicated, unique designs, not modular.  The costs of decommissioning are only estimates, since none of the large reactors have ever been completely decommissioned to greenfield status.  The storage of high level wastes has yet to be solved anywhere on the planet.
            The primary design challenge with solar power is intermittent production, but the recent developments in storage battery technology are beginning to address this issue.  The modularity of both solar and storage are well suited to mass production, which makes costs drop over time.  The other advantage is scalability, which means that solar and storage can be designed to fit very specific needs and grow as the needs change.  Nuclear plants are large, centralized, and capital intensive.  This toxic, obsolete power source is at odds with a world that is moving to smaller decentralized systems of modest means.