Tuesday, March 15, 2011

Long-Term Energy Generation Costs: Calculating Risk

The multiple-reactor accident at Japan’s Fukushima Daiichi Nuclear Power Station following its catastrophic earthquake and tsunami on March 12, which may turn out to be one of the most damaging nuclear plant accidents in history once again tragically highlights the hidden (that is, uncalculated) costs of differing energy technologies. The difficulty of calculating certain ancillary costs of power production, such as potential risk and long-term environmental cleanup and recovery leaves both government and industry partially in the dark when making important decisions regarding power generation investment and policy.

The simplified formulas used to calculate total lifetime cost of power plants, because they generally leave out longer-term and risk-based costs, dis proportionally effect perception of solar and wind energy solutions. These technologies are generally more expensive than coal, oil, and natural gas to install initially, but contain far fewer long-term environmental costs. However, because these costs are precisely the ones omitted from projections, the result is a skewed picture of solar costs via-a-vis other technologies.

Coal plant costs are perhaps the easiest to calculate because their design is relatively stable and standardized, and coal supply is domestic in the U.S. and also relatively stable; however, such calculations rarely (if ever) include the potentially massive costs of future environmental cleanup or mitigation. Though we know that these costs (coal is a significant contributor to global climate change) are enormous, they are notoriously difficult to quantify, and thus official projections of long term cost rarely contain a numerical indication of costs beyond construction, materials, and regulatory fees. Oil, natural gas, and nuclear power plants all pose similar potential environmental costs that aren’t included in cost projections—due not only to the direct difficulty of quantification, but also to the difficulty of projecting risk probabilities of accidents. The 2010 Gulf Oil spill will cost at least $30 billion in the long run, but none of these costs were figured into the development of BP’s fleet of wells and oil rigs in the Gulf Coast. Neither were the costs of containment and environmental cleanup factored into the construction of the Chernobyl, Three Mile Island, or Fukushima Daiichi power plants. Such accidents are extremely rare, but their costs can exceed the lifetime construction and operating costs of a plant by many orders of magnitude, and clearly should be factored in to overall projections of lifetime costs.

Researchers are the Argonne National Laboratory, run by the U.S. Department of Energy, have begun to address this problem with the formulation of a new model for calculating solar’s Levelized Cost of Energy (LCOE). Their formula is essentially an extension of one developed by industry that utilizes a Monte Carlo simulation to statistically select from probability distributions to account for uncertainty in a number of variables. These are precisely the variables that would otherwise drop out of the equation or be replaced by arbitrary constants. Thus this work provides a concrete methodology to extend LCOE calculations to complex, uncertain features of long-term energy cost. While this is enormously helpful for anyone calculating costs of solar energy generation, similar models need to be developed for other industries for an accurate comparison of costs to be possible. Still, a publicly-funded (and thus not serving any particular segment of industry) effort of this sort in an enormous step in the right direction: by highlighting the problem with previous formulas and blazing a trail toward more accurate models that account for uncertainty, the Argonne researchers are helping to institutionalize a new, more accurate and complete form of energy cost calculation.

Their full paper was published in the journal Energy and Environmental Science, here.