Climate change, nuclear power, and the adaptation–mitigation dilemma more

Climate Change, Nuclear Power, and the Adaptation-Mitigation Dilemma *Natalie Kopytkoa, John Perkinsb a b The Evergreen State College, 296 Porter Way #12, Milton, WA, 98354 The Evergreen State College, 1806 24th Avenue NW, Olympia, WA 98502 *Corresponding author. 1-253-389-2613 fax: 1-204-835-2425 E-mail addresses: n_kopytko@hotmail.com, kopnat@evergreen.edu Abstract Many policy-makers view nuclear power as a mitigation for climate change. Efforts to mitigate and adapt to climate change, however, interact with existing and new nuclear power plants, and these installations must contend with dilemmas between adaptation and mitigation. This paper develops five criteria to assess the adaptationmitigation dilemma on two major points: 1) the ability of nuclear power to adapt to climate change, and 2) the potential for nuclear power operation to hinder climate change adaptation. Sea level rise models for nine coastal sites in the United States, a review of US Nuclear Regulatory Commission documents, and reports from France’s nuclear regulatory agency provided insights into issues that have arisen from sea level rise, shoreline erosion, coastal storms, floods, and heat waves. Applying the criteria to inland and coastal nuclear power plants reveals several weaknesses. Safety stands out as the primary concern at coastal locations, while inland locations encounter greater problems with interrupted operation. Adapting nuclear power to climate change entails either increased expenses for construction and operation or incurs significant costs to the environment and public health and welfare. Mere absence of greenhouse gas emissions is not sufficient to assess nuclear power as a mitigation for climate change. Keywords: Climate Change Adaptation Climate Change Mitigation Nuclear Power 1. Introduction Numerous analysts from industry, commerce, government, academia, and non-profits have promoted nuclear power as an appropriate mitigation for climate change. In essentially all cases the logic of the proposal is simple and appealing: • climate change results primarily from burning fossil fuels, which releases carbon dioxide to the atmosphere; • nuclear power yields no carbon emissions as electricity is generated; 1 • therefore nuclear power is an appropriate, indeed perhaps ideal, mitigation for climate change. Appealing as this logic model appears, it unfortunately ignores a wide range of other issues, each of which impinges upon the quest for reduced carbon emissions. Thus it is too simplistic and seriously misleads. The argument leads to easy conclusions about the suitability of nuclear power to temper climate change when in fact a more robust analysis suggests the opposite conclusion. Perhaps the single most important factor undermining the simple logic model stems from the fact that nuclear reactors require enormous amounts of water to cool or condense the coolant which transfers heat from the core to the turbines and cools the reactor core. This is why nuclear power plants are located near substantial amounts of water: the ocean, large lakes, and big rivers. If climate change affects the temperature, quality, or quantity of water, then existing nuclear power plants may be adversely affected. This paper examines several ways in which climate change has already affected water in ways that create problems for existing nuclear power plants. Specifically it examines the effects of sea level rise on nine existing coastal sites in the USA and the consequences of changes in water for inland reactors in France. Geographic Information Systems (GIS) models of sea level rise and a review of existing reports and published literature suggest that numerous existing plants have been or may be adversely affected by climate change. We call the set of interactions among climate change, water, and nuclear power the “adaptation-mitigation dilemma.” This term signals that existing and projected climate change threatens the operations and safety of existing plants and 2 poses other challenges to efforts to adapt to climate change. Thus existing nuclear power plants may not represent a good technology for mitigation of climate change. A separate question concerns the potential of new nuclear power plants to avoid the problems with water we identify in this paper. Maybe it’s possible to build new plants that don’t suffer the syndrome of problems in the adaptation-mitigation dilemma. For reasons explained in the conclusion of this paper, however, we believe that it may be quite difficult to fully avoid the dilemmas identified here. At the very least, avoiding these challenges will add costs and possibly increase the risks of nuclear power, both of which are already severe handicaps for this technology. This paper acknowledges that sharply differing opinions abound on what, if any, role is appropriate for nuclear power in the debates about climate change. It seeks, however, to shift the analysis and debates about nuclear power away from “Is it a good, safe, cost-effective way to reduce carbon emissions?” to “What can we learn about current nuclear power plants and how they have been or probably will be affected by the climate change that has already occurred?” With this shift comes the potential for analysis that is less fought with ideological baggage that hinders a clear understanding of nuclear power. 2. Climate Change, Nuclear Power, and Adaptation-Mitigation Dilemmas According to the Intergovernmental Panel on Climate Change (IPCC) (2007b), warming of the climate is unequivocal, as is now evident from observed increases in global average air and ocean temperatures. Human influences have: • Likely increased temperatures of extreme hot nights, cold nights and cold days. • Likely contributed to changes in wind patterns, affecting extra-tropical storm tracks and temperature patterns. 3 • More likely than not increased risk of heat waves, areas affected by drought since the 1970s, and frequency of heavy precipitation events. • Very likely contributed to sea level rise during the latter half of the 20th century. Impacts of climate change to safe and continued operation of nuclear power plants have been recognized by the International Atomic Energy Agency (IAEA); as a result, they are currently creating guide lines on adapting nuclear power plant design and operation to climate change. According to the IAEA (2003b), the major hazards to nuclear power plants are changes in the following: • • • • • Temperatures of the air and the sea The patterns, frequency and strength of winds The characteristics of precipitation, such as higher peak levels The flow rates of rivers Rises and anomalies in sea levels The operation of nuclear power plants must adapt to climate change. Beyond adapting the plants to climate change, however, planners and policy makers must also consider the implications of mitigation strategies in general (including nuclear power) for impacts on adapting to a wide range of consequences of climate change. The interactions of mitigation and adaptation have been recognized but remain unexplored (Klein et al., 2007). Given that other mitigation measures face similar problems, a specific set of criteria are needed to judge the severity of various adaptation-mitigation dilemmas. We propose criteria for evaluating the adaptationmitigation dilemmas specifically connected to nuclear power. 4 The adaptation-mitigation dilemma for nuclear power focuses on two broad problems that arise in mitigating for climate change. First, climate change challenges nuclear power as a mitigation technology because climate change affects reactor operations. Second, operation of nuclear power can impair the ability of natural and human systems to adapt to climate change or cause other environmental concerns. In order to address these two problems, we propose using the following five criteria to assess nuclear power as a mitigation measure for climate change: • Interrupted Operation: Can climate change thwart the future operation of nuclear power? • • Financial Costs: Does climate change increase the costs of nuclear power? Adaptation Impairment - Human Systems: Does operation of nuclear power potentially reduce the ability of human systems to adapt? • Adaptation Impairment - Natural Systems: Does operation of nuclear power have the potential to reduce the ability of natural systems to adapt? • Other Environmental Problems: Can climate impacts cause nuclear power to create other health and/or environmental problems? These criteria indicate serious consequences if met. • If nuclear power plants curtail production due to the impacts of climate change, then carbon-emitting sources of energy might be used as a replacement thereby voiding any benefit. • If climate change itself increases the cost of operations of nuclear power, then the benefits of nuclear power might no longer outweigh the costs and other mitigation options could become more financially attractive. 5 • If nuclear power compromises the ability of natural or human systems to adapt to climate change, then it can no longer be considered a solution to climate change. E.g., nuclear power adopted by one group could interfere with adaptation in another sector of the economy or in a neighboring state or nation. • Adaptation to climate change can suffer in the region that adopts nuclear power, or other environmental problems can arise. The benefit of mitigation is global but problems with adaptation are local or regional. Reductions in the ability to adapt may place an inequitable burden on regions that adopt nuclear power. Indicators, specific to nuclear power at inland and coastal sites, enable measurement of the criteria as shown in Table 1 and Table 2 respectively. While this analysis looks at impacts to coastal and inland reactors separately, they are not sharply distinguished. For instance heat waves could impact coastal locations, and hurricanes impact sites located many miles inland. Table 1. Criteria and indicators for inland sites. Criteria Interrupted Operation Financial Costs Indicator Unplanned shutdowns, power reductions Intake adjustments Alteration of cooling systems Flood protection Legal Battles (pertaining to water) Brownouts/Blackouts (heat waves) Thermal pollution Safety problems that include: Loss of off-site power Communication failure Restriction of evacuation routes Equipment malfunction Unplanned shutdowns Adaptation Impairment - Human Systems Adaptation Impairment - Natural Systems Other Environmental Problems 6 Table 2. Criteria and indicators for coastal locations. Criteria Interrupted Operation Financial Costs Adaptation Impairment - Human Systems Adaptation Impairment - Natural Systems Other Environmental Problems Indicator Unplanned shutdowns, power reductions Flood protection, revenue loss Loss of adjacent lands Loss of coastal habitat (coastal squeeze) Safety problems that include: Loss of off-site power Communication failure Restriction of evacuation routes Equipment malfunction Unplanned shutdowns The inland portion of the study focused on reactors in France, while the coastal portion studied reactors in the United States. France has the highest dependence on nuclear power of any country in the world, generating over 75% of its electricity from nuclear in addition to being the world’s largest net exporter of electricity (World Nuclear Association, 2008b). France has 15 inland nuclear power plant sites with 44 operating reactors. The inland portion looks at the impact of heat waves and then the impact of recent floods. The 104 operating reactors in the U.S. constitute the largest fleet of nuclear reactors in the world (World Nuclear Association, 2008c). Fifteen reactors at 9 sites are located within two miles (3.2 km) of either coast, while an additional 51 reactors at 31 sites are located in states along the east and gulf coasts. In the coastal portion of the analysis, the first section evaluates the vulnerability of nuclear power plants in coastal states to climate change by reviewing problems encountered during past hurricanes and determining areas that continue to be vulnerable, while the second section focuses on problems stemming from future sea level rise. 3. Methods 3.1. Climate change issues for inland and coastal reactors 7 Reports generated by the International Atomic Energy Agency (IAEA), the Autorité de S reté Nucléaire (ASN) in France, and the Nuclear Regulatory Commission in the United States (USNRC) provide the information on length of reactor shutdown and safety issues arising from heat waves, flooding and hurricanes. Utility reports and industry journals provide information concerning: the financial costs of adapting to climate, revenue losses from shut-downs, and changes to operating procedures such as temporary suspension of environmental regulations. This study synthesizes these widely scattered data to reach novel conclusions about climate change and nuclear power. 3.2. Sea level rise models and shoreline vulnerability analysis for coastal reactors Climate change affects coastal reactors in several ways. First, the rise in sea level can inundate reactor sites. Second, more intense storms, combined with sea level rise, can produce more severe episodic flooding and wind damage, even if the site is not permanently inundated. Third, rising sea levels can increase erosion and instability of shorelines, thus altering their shapes. The nine reactors sites selected for analysis in this study comprise all sites that lie within two miles (3.2 km) of either the Pacific or Atlantic coastlines of the United States (Figure 1). These nine reactor sites contain 15 reactors. 8 Figure 1. Location of all coastal reactor sites lying within two miles (3.2 km) of the Atlantic or Pacific coasts which were analyzed for vulnerability to sea level rise. Inundation modeling and assessing the impacts of storms faces substantial uncertainties in (a) amount and timing of sea level rise and (b) the intensities of storms, which are predicted to increase with climate change. Scenarios provided a way to see a range of potential impacts based on corresponding ranges of sea level rise, its timing, and a range of increased storm intensities. These scenarios drew upon the work of the International Atomic Energy Agency (IAEA) and the Intergovernmental Panel on Climate Change (IPCC). The IAEA’s report, Flood Hazard for Nuclear Power Plants on Coastal and River Sites recommends utilizing the results of investigations by the IPCC and using the upper bound of the 95% confidence interval to account for uncertainty. IAEA also suggests that the lifetime of a nuclear power plant, including decommissioning time, is 100 years (IAEA, 2003b). Over the next 100 years, the IAEA estimates an increase in 9 mean sea level of 35-85 cm. IAEA also advises that land subsidence should be considered along with climatic changes. Four different time ranges (base year 2008) underlie the scenarios for evaluating sea level rise at each of the nine coastal sites: 1) the end of reactor operations, 2) the end of reactor lifetimes, 3) 100 years, and 4) 150 years. The years remaining in operation for each reactor were determined by the license expiration date. Reactor lifetime was determined by subtracting the years in operation from 100 as recommended by the IAEA. Construction time was not included in calculating the reactors lifetime because of extended construction periods at several of the reactors included in the study. Sea level rises over the next 100 years and 150 years provided insights on the appropriateness of these nine sites for new reactor construction. Table 3 summarizes the four time-frames used to generate sea level rise scenarios. For end of operation and for life of reactor, the rate of sea level rise was based on current rates at 3 mm/year in California and Florida and 4.3 mm/year in the Northeastern United States. The higher rate for the Northeast includes land subsidence. The IPCC links all sea level rise scenarios to emission scenarios. For the 100 year time frame, the low and mid sea level rise scenarios in this report used the upper limit of the low and high IPCC scenarios. The 100 year high scenario used the IAEA estimate of 85 cm. For 150 years, the low and high IPCC scenarios were used. These IPCC estimates project an increased rate of sea level rise as the climate warms. For purposes of standardizing sea level rise in all nine sites, we also included the effects of a 1 meter rise, which some models indicate is possible by 10 2100 (Rahmstorf, 2007). For further explanation on sea level rise scenarios refer to Electronic Annex 1 in the on-line version of this article. Table 3. Scenario description and corresponding quantity of sea level rise for California and Florida, and the Northeast region. Scenario End of Operation Life of Reactor 100 year low 100 year mid 100 year high 1m 150 year low 150 year high Description/Rationale Global average of sea-level rise since 1993 Global average of sea-level rise since 1993 Upper limit of low IPCC scenario Upper limit of high IPCC scenario Estimate suggested in IAEA report Possible by end of century Low 100 + 50 x (3.9 mm/yr) or (4.3 mm/yr) Mid 100 + 50 x (9.7 mm/yr) CA/FL 3 mm/yr 3 mm/yr 0.39 m 0.59 m 0.85 m 1m 0.59 m 1.21 m Northeast 4.3 mm/yr 4.3 mm/yr 0.51 m 0.72 m 0.85 m 1m 0.72 m 1.21 m The greatest impact of sea level change occurs during extreme levels, i.e. storm surges, rather than as a direct consequence of inundation through changes in mean sea level (Bindoff et al., 2007). Therefore, storm scenarios must also be included in the analysis. As a baseline, the extent of flooding currently experienced during storms was modeled. To account for the possibility of increased storm intensity, the category of hurricane was increased by one and added to the projected rise in sea level. For a detailed description of storm models, please refer to Electronic Annex 2 in the on-line version of this article. Once sea level rise scenarios were established, sea level rise and storm conditions were modeled using Digital Elevation models in ArcGIS version 9.2. For further explanation of the GIS method, please see Electronic Annex 3 in the on-line version of this article. The scenarios generated here are also compared to the Design Basis Flood levels for each of the sites that are available from USNRC reports. This method does have limitations and errors, detailed in Electronic Annex 4 in the on-line version of this article. For instance, the elevation models for some of the sites have poor resolution and true mean sea level has diverged since the time of 11 measurement. In addition, elevation alone can not determine the location of a future shoreline. Coastal processes such as erosion remain a concern; therefore, a measure of the vulnerability of the coastline supplements this model. Review of data on shoreline vulnerability from USGS provides information on the vulnerability of shorelines throughout the U.S. (Thieler and Hammar-Klose, 1999a, b, 2000). The methods used in the data collection used data from local, state, and federal agencies and from academic institutions to provide an index of the relative vulnerability of different shoreline segments to sea level rise based on coastal geomorphology, rate of sea level rise, past shoreline evolution, and coastal slope. These variables identify those portions of the U.S. coastal regions the most at risk and the nature of that risk. Each coastal segment receives an overall ranking of risk: low, moderate, high, or very high. Please refer to Electronic Annex 5 in the online version of this article for a detailed discussion. 4. Results This section examines the impacts of extreme events on inland reactors in France and coastal reactors in the United States. The events of recent heat waves and floods on reactors in France and hurricanes on reactors in the United States were evaluated using the criteria outlined in the methods. Reviewing the impact of past climate variability provides one method to ascertain vulnerability to climate change. Modeling future climate offers a complementary method. The final subsection models sea level rise using ArcGIS and examines the vulnerability of the coastline to processes such as erosion. 4.1. Heat waves affect the operations of French nuclear power plants 12 According to the IPCC 2007 report, it is very likely that heat waves will be more intense, more frequent and longer lasting in a future warmer climate (Meehl et al., 2007). During July and August 2003, significantly above-average temperatures were observed throughout Europe, Scandinavia, and western Russia, with monthly mean temperatures exceeding the 90th percentile in each region (World Health Organization Europe, 2003). During the 2006 heat wave, temperatures in central and western Europe were 7˚C above historical maximums (MacLachlan et al., 2006). In recent decades, the most deadly weather events in industrialized countries have been extreme heat. The 2003 European heat wave claimed an estimated 15,000 lives in France alone (Lagadec, 2004; Poumadere et al., 2005). In France, warmer than average summers from 2003 to 2006 required extensive operational changes to maintain a steady power supply from nuclear power plants. In 2003, Électricité de France (EDF) attempted to balance production and comply with regulations on thermal pollution by stopping units at some plants (Autorité de sûreté nucléaire, 2003; Hibbs, 2003). The total power reduction during the summer of 2003 was 5.3 TWh: equivalent to a loss of more than 200 reactor days (Parey and Aelbrecht, 2005). Black-outs were avoided in France by exercising several options including: the purchase of energy on the wholesale power market (2800 MW), citizenship conservation (300 MW), negotiating lower loads from industry consumers (1700 MW) and reducing exportation to Italy (Parey and Aelbrecht, 2005). EDF’s contract with Italy was the only contract with a clause allowing interruption in the event of an emergency, and EDF was able to cut power exports by more than half (Hibbs, 2003; Poumadere et al., 2005). 13 Italy relies on France for much of its power supply. In 2003, Italy purchased 35.2 per cent of it’s energy imports from France (Power Engineering International, 2005). Consequently, many Italian cities experienced blackouts lasting several hours (BBC News, 2003). In addition, Italy, with a death toll of nearly 20,000, had the highest mortality rate in Europe during the heat wave (The Riviera Times, 2005). In 2006, similar measures were implemented such as citizen conservation and interruption of supply in the contracts that allowed for emergency measures. In addition, lessons were learned from the 2003 experience: Maintenance outages were postponed for coastal nuclear power plants and at the start of the heat wave EDF purchased 2,000 MW for an unspecified amount of time on the wholesale market (MacLachlan et al., 2006). Despite these precautions, however, EDF and the French government still made contentious decisions during the heat waves. In July 2003, the operators of Golfech, Tricastin, and Bugey were granted thermal release waivers 1°C higher than previously allowed. By August, an order was issued stating that the climatic conditions faced by France and Europe during the summer of 2003 were exceptional circumstances threatening the safety of property and persons, the continuity of public services, and the economic activity of the country. Thermal electricity production facilities (fossil fuel and nuclear) discharging cooling water into the Garonne, Rhone, Seine and Moselle river basins could continue to do so until the difference between the water temperature measured upstream and downstream after mixing at each of these installations was equivalent to 1 °C for installations fully equipped with cooling towers, 1.5 °C for those located on the Seine and Moselle rivers, and 3 °C for the other plants. 14 The French government limited the use of these measures in 2003 to the electricity production necessary for national consumption needs and for complying with international agreements. The electricity producers were required to monitor the environmental impact on human health and river fauna, especially fish, and inform the authorities daily of the temperatures recorded after mixing downstream at each of the plants concerned. The order ended on September 30th 2003 (Autorité de sûreté nucléaire, 2003). Five power plants out of 13 exempted from thermal discharge limits used the exemptions. Of those five power plants, four were nuclear: Bugey, Cattenom, Golfech, and Tricastin (Table 4). Tricastin consistently discharged water into the Rhone River at 2˚C or more above upstream temperatures between August 14th and August 27th. Bugey used it twice while Golfech used it 12 times (MacLachlan, 2004). Golfech heated the Garonne by 0.4˚C to 0.8˚C, potentially significant since the Garonne is already one of France's warmest rivers (MacLachlan, 2003b). Table 4. France’s inland reactors and sites issued thermal release waivers. Site Name (# units) Le Blayais (4) Bugey (4) Belleville (2) Cattenom (4) Chinon (4) Chooz (2) Civaux (2) Cruas-Meysse (4) Dampierre-en-Burly (4) Fessenheim (2) Golfech (2) Nogent-sur-Seine (2) Saint-Alban (2) Saint-Laurent-des-Eaux (2) Tricastin (4) Discharge River Garonne Rhône Loire Moselle Loire Meuse Vienne Rhône Loire Rhine Garonne Seine Rhône Loire Rhône Waiver 2003 3 ˚C 3 ˚C NA 1.5˚C NA NA NA 1˚C NA NA 1˚C 1.5˚C 3 ˚C NA 3 ˚C Waiver Used? No Yes NA Yes NA NA NA No NA NA Yes No No NA Yes Waiver 2006 3 ˚C 3 ˚C NA 1.5˚C NA 1.5˚C NA NA NA NA 0.3˚C 1.5˚C 3 ˚C NA 3 ˚C Waiver Used? No No No No No No No No No No No No No No No 15 Environmentalists criticized these measures, and many questioned whether waivers for thermal discharges would become a regular summer feature if the JulyAugust heat wave was indicative of climate change (MacLachlan, 2003a). Indeed, in 2004 the Nuclear Installation Safety Directorate (DGSNR) issued a permanent order allowing EDF to modify thermal discharge limits by 1˚C to 2˚C at the Bugey, Golfech and Nogent reactor sites and by up to 3˚C at Tricastin between July 1st and September 30th of each year. The order helped during the summer of 2005, although production at Tricastin had to be temporarily lowered due to high temperatures in late June (MacLachlan, 2005). Despite the increases, these measures did not appear to be sufficient during the 2006 heat wave. On July 22, 2006, the French government published an executive order allowing EDF to raise river water temperatures downstream of all its river-cooled nuclear plants--if necessary to preserve the stability of the grid and maintain power supply--until September 30th (MacLachlan et al., 2006). The order concerned plants located on the Garonne, Rhone, Seine, Meuse and Moselle Rivers, at sites that host 28 of EDF's 58 power reactors. The order didn't specify absolute temperatures for each river downstream of power plants, as the 2004 order did but instead referred only to the acceptable change between the intake and discharge stations. The order allowed a change in temperature of 0.5-1.5°C for plants equipped with cooling towers and 3°C for those without towers. The nuclear plants on the Garonne and the Rhone were already operating under the limits set in 2004 (MacLachlan et al., 2006). Fortunately, the authorization was unnecessary as none of the reactors utilized the 2006 waiver 16 (Autorité de sûreté nucléaire, 2006). Regardless many actions taken in 2006 were repeats of 2003 despite the benefit of prior experience. 4.2. Inland floods affect the operations of French nuclear power plants France’s inland reactors faced floods the winter following the 2003 heat wave; however, this was not the first time flooding challenged operation of reactors in France. The 1999 flood of Le Blayais revealed that flood protection had to be investigated and improved at many sites. Le Blayais nuclear power plant site consists of four 900 MW(e) pressurized water reactors (PWR) located 30 km southeast of the Atlantic ocean, on the banks of the Gironde estuary. The Design Basis Flood (DBF) calculated as 5.2 m French national datum level (NGF) was determined from the level of water resulting from the maximum astronomical tide and the 1000 year storm surge. The site is surrounded by a dyke made of earth and protected on the River Gironde side by a pile of stone blocks. Alongside the River Gironde, its height is 5.2 m NGF, and its height is 4.75 m at the other sides (IAEA, 2003a). The flooding event on December 27th,1999 resulted from a high tide, wind speeds of 100 km/h that generated waves estimated to reach a height of 2 m, combined with a 2.01 m storm surge. The maximum storm surge measured prior to December 27th, 1999 was 1.20 m for a 40 year historical series of data. Investigations carried out on the site after the storm, however, showed that the water had reached 5 to 5.30 m (Gorbatchev et al., 2000; IAEA, 2003a). The three units operating when the storm arrived shut-down due to loss of off-site power. Meanwhile, strong waves submerged the plant platform after water entered mainly on the northwest side of the dyke. The waves moved the rocks 17 causing part of the dyke to wash away alongside the River Gironde. The water reached a depth of around 30 cm in the northwest corner of the site, while an estimated 90,000 m3 of water entered the facilities (Gorbatchev et al., 2000; IAEA, 2003a). Units 1 and 2 were severely affected by incoming water: utility galleries, rooms containing outgoing electrical feeders, and the bottom of the fuel buildings were flooded; one of the essential service water pumps was lost due to immersion of the motors, and electrical switchboards were made unavailable (Gorbatchev et al., 2000). The Nuclear Safety Authority (ASN) asked EDF to repair all the flooded equipment and to upgrade flood protection. EDF raised the dyke around the nuclear site by 1 meter and equipped it with an alert system based on forecasts from MétéoFrance. EDF also developed an operational procedure to bring the site reactors to a safe state and protect the premises felt to be most important. Under these conditions, ASN authorized the two reactors most severely affected by the flood to restart in May 2000; however, ASN also asked EDF to take additional steps against the risk of flooding before the first quarter of 2001 (Autorité de sûreté nucléaire, 2000). In September 2000, EDF proposed an anti-surge device consisting of riprap and a wall was placed on top of the dyke that had already been raised in March. After conducing surge tests, the ASN requested the height of the wall be raised to 8.50 meters (Autorité de sûreté nucléaire, 2000). The flooding which occurred at Le Blayais Nuclear Power Plant revealed a potential mode by which the safety of all the units of a single site could be jeopardized (Gorbatchev et al., 2000). Operating teams were unprepared to deal with an incident that affected all the reactors on a site; it was difficult to asses the 18 situation from the data available in the control room; and the site had no suitable control procedures for managing a situation involving loss of outside electrical power sources combined with flooding (Autorité de sûreté nucléaire, 2000). The ASN wanted to take full advantage of the lessons learned to improve protection of all the reactors in France against flooding. In March 2000, the ASN asked EDF to review the existing measures for dealing flooding at all of EDF’s nuclear sites (Autorité de sûreté nucléaire, 2000). However, the nuclear installation safety directorate (DSIN) told EDF it must not wait to act until the completion of revised calculations of maximum flood risk (MacLachlan, 2001). Action was particularly urgent at the Belleville PWR site, where EDF studies showed the ''safe'' flood level equivalent to the level of the maximum historical flood with a 15% safety margin was up to 1.4 meters higher than the level in the plant's design (MacLachlan, 2001). DSIN asked EDF to advise what measures would adapt the site's protections to the new data. A subsequent study by safety experts at Institute of Radiological Protection and Nuclear Safety (IPSN), focused on external risks that might not have been sufficiently considered in the original design, showed that many EDF sites were under protected. EDF proposed a 100-million-franc (approximately 13.5 million USD) program to build a levee around the Belleville site that would raise flood protection by 1 meter. In the meantime, EDF planned to reinforce the site's system of mobile flood protection walls (MacLachlan, 2001). On the morning of December 2nd 2003, the ASN activated the emergency center due to storm conditions. The Cruas nuclear power plant takes its cooling water from the Rhone River. During the night of December 1st and 2nd 2003, intake of 19 water containing large amounts of mud and plant debris impaired the cooling systems. The rapid deterioration of the exchange capacity of the cooling systems led the Cruas plant operator to trigger the on-site emergency plan as a preventive measure and shutdown the reactor. The Tricastin plant, located further downstream than Cruas, takes its cooling water from the Rhone bypass canal at Donzère. Pumping of water containing large amounts of plant matter and mud triggered shutdown of the cooling water pumping system and consequently automatic shutdown of the 3rd and 4th reactors on the 2nd and 3rd of December 2003 respectively (Autorité de sûreté nucléaire, 2003). Fearing a deterioration of the situation, the Tricastin plant triggered its own emergency plan on the night of December 2nd 2003. Late afternoon on December 3rd 2003, the status of the nuclear installations, the weather forecast and the flow rate of the Rhone river were considered to be satisfactory enabling a number of alerts to be lifted and the gradual restart of reactors (Autorité de sûreté nucléaire, 2003). Lessons learned from 1999 flood were applied during the 2003 events. EDF headquarters maintained a supervisory team to monitor the changing situation in the Rhone valley and in the Loire (Belleville, Dampierre, Saint-Laurent and Chinon) and Garonne (Golfech) valleys days before the maximum flood levels were to be reached on the rivers (Autorité de sûreté nucléaire, 2003). In accordance with their procedures, the operators of the nuclear power plants on the Loire took preventive protection measures, especially for site access problems due to submersion of access roads. Furthermore, the build-up of detritus in the hydroelectric dam upstream of the Golfech plant was released in collaboration with the Golfech plant operator (Autorité de sûreté nucléaire, 2003). 20 The 2003 floods revealed another problem: the safety of certain installations during flooding events depends to a large extent on the behavior of off-site structures not belonging to EDF, particularly the Cruas and Tricastin nuclear power plants. Evaluating the robustness and the surveillance and upkeep of these structures entails a complex decision-making process between the stakeholders, the authorities, and EDF. Therefore, ASN asked EDF to continue the exchanges initiated between the licensees of these structures and to keep it informed of any difficulties (Autorité de sûreté nucléaire, 2007). 4.3. Hurricanes affect the operations of US nuclear power plants The predictability of a hurricane's path allows time for preparation. The USNRC has an established hurricane response program that is implemented each year during hurricane season. Integrated emergency plans are developed in a coordinated manner between the facility licensee and State and local authorities, with oversight by the USNRC and Department of Homeland Security/Federal Emergency Management Agency (DHS/FEMA)(Leach et al., 2006). Yet, review of posthurricane reports shows that carelessness in licensee actions poses a serious problem. Moreover, formal procedures require that each nuclear power plant take actions under weather conditions specific to each site, including shutdown of the reactor in anticipation of hurricane force winds (Leach et al., 2006). While shutdown of the reactor is vital to safe operation, the restart of the reactor requires approval from both the USNRC and FEMA that can take days to weeks. In addition, several safety issues repeatedly arise during storm events including: the loss of offsite power, failure of communication systems and alarm systems, and obstruction of site access. 21 Safe operation of nuclear power plants and accident recovery depend on the availability of AC power, which comes to the plant from the external grid. An assessment conducted in 1998 of loss of offsite power events at U.S. nuclear power reactors found that sixteen of the 22 events resulting from severe weather occurred at only 5 sites. The five sites were Pilgrim Station in Massachusetts, Crystal River in Florida, Brunswick in North Carolina, Millstone in Connecticut, and Turkey Point in Florida. The units at these sites have diverse designs with little similarity in the design or redundancy of electrical power supply; therefore, the proximity of the sites to the east coast was a major factor in the frequency of loss of power (Atwood et al., 1998). Hurricane Andrew, a Category 5 storm in 1992, was the first time a hurricane significantly affected a commercial nuclear plant (U.S. Nuclear Regulatory Commission, 1994). The analysis for wind indicated a need to modify the flood wall and improve the emergency procedure for Category 5 hurricanes. Hurricane Andrew caused damage to a number of non-safety structures and equipment at Turkey Point including: collapse of all steel-framed turbine canopies, damage to one of the chimneys belonging to the fossil fuel units, movement of the base anchors for the vent stack on the Unit 4 containment, failure of the ductwork from the radioactive waste building, and the collapse of the non-safety high-water tank onto the fire protection pumps and pipes thereby rendering one of the fire protection systems inoperable. This event demonstrated the need to either design non-safety structures and equipment to withstand the postulated events, or assure that the consequences of their failure would not disable the safety functions of safety-related structures, systems and components (U.S. Nuclear Regulatory Commission, 1994). 22 Prior to the storm, on August 23, 1992, the licensee shut down both reactors and placed them in the “hot standby” condition as required by the plant emergency procedures. The plant lost all offsite power during the storm and for over five days after the storm (Leach et al., 2006). Many false alarms in the spent fuel containment created concerns because it was not accessible during the storm (IAEA, 2003a). In addition, the security system sustained extensive damage specifically to equipment including: lighting, cameras, intrusion detection equipment, protective area fencing, and the entrance building (Leach et al., 2006). Furthermore, wind damage caused the loss of all communication at Turkey Point Nuclear Generating Station. As a result of this experience, the USNRC arranged for portable satellite communication equipment to be available at sites as required. However, during Hurricane Katrina in 2005 satellite phones did not work as well as anticipated. Operators had to go outside to use the phones because cloud cover interrupted the reception (Leach et al., 2006). Indeed, hurricanes during the 2003-2005 season continued to cause the same problems encountered in 1992, specifically communication issues, loss of off-site power, loss of emergency sirens, and obstruction of site access as summarized in Table 5. 23 Table 5. Summary of reactors impacted by recent hurricanes. Storm (Year) Isabel (2003) Isabel (2003) Isabel (2003) Isabel (2003) Isabel (2003) Isabel (2003) Isabel (2003) Isabel (2003) Charley (2004) Frances (2004) Frances (2004) Jeanne (2004) Katrina (2005) Katrina (2005) Katrina (2005) Reactor (State) Calvert Cliffs (MD) Hope Creek (NJ) Salem (NJ) Harris-1 (NC) Limerick(PA) Peach Bottom (PA) Three-Mile Island (PA) Surry-1 (VA) Brunswick (NC) St. Lucie -1&2 (FL) Crystal River-1&2 (FL) St. Lucie (FL) Grand Gulf (MS) River Bend (LA) Waterford-3 (LA) Safety Issues LOES LOES LOES LOES LOES LOES LOES OSA LOOP, LOES a Reason for Shutdown (approximate length) Not Shutdown Electrical faults from salt water deposits (days) Electrical faults from salt water deposits (days) Not Shutdown Not Shutdown Not Shutdown Not Shutdown Loss of power to intake pumps (days) LOOP (days) b CI Preventative Measure (days) c CI, LOOP Preventative Measure (days) LOOP (days) Reduced power to restore stability to grid. Reduced power to restore stability to grid. Preventative Measure (weeks) LOOP(partial) d LOES NA e CI, LOOP LOES: Loss Of Emergency Sirens, OSA: Obstruction of Site Access, LOOP: Loss Of Off-site Power, CI: Communication Issues. a Emergency Notification System direct connection with the NRC was lost, as was the Emergency Response Data Acquisition Display System link between Unit 1 and the NRC. b Manual trip of unit-2 occurred as a result of steam generator level oscillations while reducing power. c Loss of telephone lines and difficulty was encountered with satellite phones. d Emergency Response Data System link to NRC was lost. e Winds knocked out 17 of 43 emergency sirens placing the number of sirens less than the operability rate of 75%. While shutdowns during storms tend to be only days, the change to normal operating procedures that occur when sirens, communication, off-site power and site 24 access are lost or restricted, alone or in combination, becomes problematic. In addition, equipment failures often arise during stormy conditions. Brunswick, St.Lucie, and Crystal River experienced unexpected equipment malfunctions or failures during the 2004 hurricane season. Brunswick had failures of the B-train standby gas treatment. St. Lucie experienced problems with a feed-water regulating valve and a breaker for an intake cooling water pump. Crystal River had an overloaded alarm system and failure of an emergency lube oil pump for a main feedwater pump turbine. However, negligent actions on the part of licensees are of even greater concern. After Hurricane Francis passed the St. Lucie site, the reactor auxiliary building’s missile shield doors were found open thus risking exposure of safetyrelated equipment to tornado-induced missiles. The licensee stated that the doors could have been open for several years (Kauffman, 2005). In addition, all three sites had breaker faults or failures related to salt contamination or moisture intrusion. Moreover, the licensee at one site stated that preventive and corrective maintenance activities had not identified moisture buildup as a condition requiring corrective action (Kauffman, 2005). 4.4. Sea level rise potentially affects the operations of US nuclear power plants While the previous section describes the problems encountered during past storms, this next section uses models built in ArcGIS to determine future flooding from a combination of sea level rise and storms. These models do not include coastal processes such as erosion; therefore, the coastal vulnerability index, determined by experts with the USGS, complements inundation models. 25 Results are summarized in tables for each reactor site. The numbers in each cell represent the amount of sea level rise in meters. The top row is sea level rise alone and each subsequent row contains increasingly intense storm conditions, while the columns are successive time-scenarios. For sites with multiple reactors in operation, time-scenarios were based on the newest reactor. “Potential for Flooding” is the sea-level rise at which the models predict that the site will start flooding, but the flood waters have not reached structures on the site, or covered the roads. “Considerable Flooding” describes conditions that cause flood waters to reach structures on the site or block roads to the site. “Site Inundated” indicates that the entire site is covered in flood waters. While models show that some sites will flood, design and operational procedures are supposed to ensure safe operation. 4.4.1. Turkey Point The two Turkey Point units are located on the west shore of Biscayne Bay in Miami-Dade County, Florida. Turkey Point Reactor-4 began operation in 1973 and the license expires in 2033. From 2008, Turkey Point had 25 years remaining in operation and the total life of the reactor was determined to be 65 years. The plant grade level is 5.49 m (18 feet) above mean low water (MLW), and the site has been flood protected to an elevation of 6.1 m (20 feet) above MLW. Components vital to safety, with the exception of the intake cooling water pumps, are protected against flood tides and wave runup to 6.7 m (22 feet) above MLW on the east side of the units by a continuous barrier consisting of building exterior walls and stop logs for the door openings. Additional protection against flooding is provided by placing safety equipment on pedestals or providing curbs, use of closed doors with water26 tight sills, floor drainage systems with sumps and sump pumps, and water level alarms (Haney, 2006) The model of sea-level rise (without storms) predicts that Turkey Point will not have a problem with flooding between now and the end of operations (2033) and the life of the reactor (2073). If a new reactor were constructed today, then in 100 years (2108) the site has a potential for flooding at the low level of predicted sea level rise. Considerable flooding occurs (without storms) at (a) 100 years (2108) at the mid- and high-levels of predicted sea level rise, (b) at 150 years, and (c) if sea level rise reaches 1 meter (Table 6). When storms (northeasterns and hurricanes) are factored in, Turkey Point is already subject to flooding. Northeasterns and hurricanes of Categories I and II cause considerable flooding. As time progresses and sea level rises, the flooding situation becomes even more precarious. The model predicts that a Category I High storm will inundate the site by 150 years (2158), and Category II storms will inundate the site by the end of the reactor's life (2073). Storms of Category III or higher already will inundate the site, and the inundation becomes progressively deeper in the future (Table 6). According to these scenarios, a Category V storm would cause flooding conditions that exceed the probable maximum surge (5.6 m) for the site and approach the design basis flood level (6.1-6.7 m) within the lifetime of the reactor. In terms of site access, roads could potentially flood at 0.7 m, and are completely covered at 0.9 m as shown in Figure 2 and Figure 3 respectively. As revealed in Figure 4, the site is almost completely flooded at 2.5 m. 27 Table 6. Sea level rise and storms for Turkey Point, by scenario. Table Key: No Flooding Potential For Flooding Current Storms End of Operation 0.1 0.6 1.2 1.5 2.4 3.7 4 5.5 6.1 0.7 1.3 1.6 2.5 3.8 4.1 5.6 6.2 Considerable Flooding Site Inundated 100 Yrs High 0.9 1.5 2.1 2.4 3.3 4.6 4.9 6.4 7.0 Sea Level Rise 1 m 1.0 1.6 2.2 2.5 3.4 4.7 5.0 6.5 7.1 150 Yrs High 1.1 1.7 2.3 2.6 3.5 4.8 5.1 6.6 7.2 Scenarios Conditions Sea Level Rise Alone Nor’eastern Cat I Low Cat I High Cat II Cat III Cat IV Low Cat IV High Cat V Life of Reactor 0.2 0.8 1.4 1.7 2.6 3.9 4.2 5.7 6.3 100 Yrs Low 0.4 1.0 1.6 1.9 2.8 4.1 4.4 5.9 6.5 100 Yrs Mid /150 Yrs Low 0.6 1.2 1.8 2.1 3.0 4.3 4.6 6.1 6.7 Values for sea level rise are in meters above mean sea level for inundation modeling, while values used by the U.S. NRC in design basis calculations reference mean low water. Storm strength increases along each subsequent row, while columns represent time scenarios and corresponding sea level rise. Figure 2. Turkey Point with a sea level rise of 0.4 m and 0.5 m. 28 Figure 3. Turkey Point with a sea level rise of 0.9 m. Figure 4. Turkey Point with a sea level rise of 2.5 m. Sea level rise, inundation, and storm surges all threaten the Turkey Point site-now and in the future--but additional threats also arise from erosion with consequent changes in the shape of the shoreline. Coastal vulnerability analysis provides a 29 means to move beyond looking at just inundation. In addition, the presence of a development along an already vulnerable coastline increases the vulnerability of adjacent coastal segments to erosion. Turkey Point receives an overall coastal vulnerability index (CVI) ranking of high due to very high rankings received in geomorphology, slope, and tides (Table 7). Table 7. Coastal vulnerability of the shoreline at Turkey Point. Reactor Turkey Point Tide Very High Waves Moderate Erosion Moderate Sea Level Rise Low Geomorphology Very High Slope Very High CVI High The USGS ranked coastal features according to vulnerability to sea level rise. The average of each of these features determined the coastal vulnerability index (CVI) (Thieler and Hammar-Klose, 1999a). 4.4.2. Calvert Cliffs The Calvert Cliffs nuclear power plants are in Calvert County, Maryland, on the west bank of Chesapeake Bay, approximately halfway between the mouth of the bay and its headwaters at the Susquehanna River. The first of the two power plants on the Calvert Cliffs site began operation in 1977 (the date on which these scenarios are based) and the license expires in 2036. From 2008, Calvert Cliffs had 28 years remaining in operation and the total life of the reactor was determined to be 69 years. UniStar is currently applying for construction of a new reactor at this site. The current reactor is approximately 152.4 m (500 feet) from the shore, while the new proposed reactor will be a 304.8 m (1000 feet) from the shoreline (MACTEC Engineering and Consulting Inc., 2008). The flooding conditions considered in design include: the probable maximum flood (PMF) on streams and rivers, potential dam failures, probable maximum surge and seiche flooding, probable maximum tsunami and ice effect flooding. The Nuclear Island of the new site is at an elevation of 24.8 m (81.5 ft) with respect to the reference level. Safety-related structures of Nuclear Island have a minimum 30 grade slab or entrance at elevation 25.8 m (84.6 feet). The maximum flood level at the intake location is an elevation of 12 m (39.4 ft) as a result of the surge, wave heights, and wave run-up associated with probable maximum hurricane (UniStar Nuclear Development, 2008). The site is at a high enough elevation that flooding from storm surges does not pose a big problem. Flooding could potentially happen at 3.9 m and above, corresponding to an increase in storm intensity within the life of the existing reactor as shown in Table 8. Figures 5 and 6 illustrate the potential flooding at 3.9 m and at 4.6 m respectively. The flooding in the illustration does appear serious; however, it is ranked as “potential for flooding” because the structure that appears to be covered by flood waters belongs to the cooling water intake system. Table 8. Sea level rise scenarios and results for Calvert Cliffs. Table Key: No Flooding Potential For Flooding Scenarios Conditions Sea Level Rise Nor’eastern Cat I Low Cat I High Cat II Cat III 0.6 1.2 1.5 2.4 3.7 Current Storms End of Operation 0.12 0.72 1.32 1.62 2.52 3.82 Life of Reactor 0.30 0.90 1.50 1.80 2.70 4.00 100 Yrs Low 0.51 1.11 1.71 2.01 2.91 4.21 100 Yrs Mid/ 150 Yrs Low 0.72 1.32 1.92 2.22 3.12 4.42 100 Yrs High 0.85 1.45 1.92 2.22 3.12 4.42 Sea Level Rise 1 m 1.00 1.60 2.20 2.50 3.40 4.70 150 Yrs High 1.21 1.81 2.41 2.71 3.61 4.91 Values for sea level rise are in meters above mean sea level for inundation modeling, while the U.S. NRC documents do not specify the reference elevation used in design basis calculations. Storm strength increases along each subsequent row, while columns represent time scenarios and corresponding sea level rise. 31 Figure 5. Calvert Cliffs with a sea level rise of 3.9 m. Figure 6. Calvert Cliffs with a sea level rise of 4.6 m. While the Calvert Cliffs’ site has only a potential of flooding, the simplicity of the surge model used in this study could lead to overly conservative results. The configuration of the coastline and contours of water depth accentuate the level of sure at this location. The largest tidal flood that is likely to occur under the most 32 severe meteorological and hydrological conditions in Cheaspeake Bay is 4 m (13 feet) above the national geodetic vertical datum, while waves can reach an additional 1.5 m (5 feet) (Ward et al., 1999). Nonetheless, this level of surge is accounted for in the design of the new reactor and the existing reactors are at an elevation of at least 11 m (36 feet). While Calvert Cliffs appears resistant to flooding, the site ranked very high in erosion and geomorphology and received a very high ranking for the coastal vulnerability index (CVI) as shown in Table 9. In the U.S., the Chesapeake Bay region is ranked the third most vulnerable to sea level rise behind Louisiana and Southern Florida due to large losses of land from erosion. Maryland is currently losing approximately 58 acres of land per year to shore erosion (Maryland DNR, 2007; Mayrland DNR, 1999). Table 9. Coastal vulnerability of the shoreline at Calvert Cliffs. Reactor Calvert Cliffs Tide Very High Waves Moderate Erosion Very High Sea Level Rise High Geomorphology Very High - High Slope Low CVI Very High The USGS ranked coastal features according to vulnerability to sea level rise. The average of each of these features determined the coastal vulnerability index (CVI) (Thieler and Hammar-Klose, 1999a). 4.4.3. San Onofre San Onofre is located in north San Diego County, California and is fronted by a narrow beach along the Pacific Ocean. San Onofre Nuclear Generating Station (SONGS) Unit-3 began operation in 1983 and the license will expire in 2022. From 2008, the reactor had 14 years remaining in the operating license and the life of the reactor was determined to be 75 years. Tsunamis caused by active trench system are considered along with those generated by large scale tectonic movement. Structures designed to protect the site include the seawall. The plant grade is at an elevation of approximately 6.1 m (20 feet) MLLW (Southern California Edison Company et al., 2002). This elevation is well above the maximum seawater elevation of 4.8 m (15.8 feet) mean lower low 33 water that is predicted to occur in the event of a maximum tsunami coincident with storm surge and high tide (Haney, 2006). Flooding does not occur under any of the scenarios as shown in Table 10. The sea wall and the plant grade is at a high enough elevation to prevent flooding from coastal storms. Table 10. Sea level rise scenarios and results for San Onofre. Table Key: No Flooding Scenarios Conditions Sea Level Rise SLR and El Niño Storm Surge Storm Wave End of Operation 0.04 0.3 0.7 1.5 Life of Reactor 0.23 0.53 0.93 1.73 100 Yrs Low 0.38 0.68 1.08 1.88 100 Yrs Mid /150 Yrs Low 0.59 0.89 1.29 2.09 100 Yrs High 0.85 1.15 1.55 2.35 Sea Level Rise 1 m 1 1.3 1.7 2.5 150 Yrs High 1.1 1.4 1.8 2.6 Values for sea level rise are in meters above mean sea level for inundation modeling, while the U.S. NRC uses mean lower low water in design basis calculations. Storm strength increases along each subsequent row, while columns represent time scenarios and corresponding sea level rise. Similar to Calvert Cliffs, San Onofre does not flood, but receives a high rank for coastal vulnerability as shown in Table 11. The San Onofre power plant is located on the coastal terrace, which is underlain by Miocene marine rock capped by Pleistocene marine and nonmarine sediments (Kuhn, 1980). These Pleistocene sediments are essentially horizontal and are easily eroded from the bluff face and along the canyons. Approximately 80 percent of the cliffs between the power plant and Target Canyon six miles to the south, on Camp Pendleton, consist of landslides (Kuhn, 1980). Furthermore, where protective measures project or extend seaward beyond adjacent unprotected lots, there is immediate erosion and notching of the unprotected sites. As beach sand levels fall, storm waves tend to converge on projecting structures and the waves refract toward unprotected lots (Kuhn and Shepard, 1983). The San Onofre facility itself is not at risk from erosion or flooding 34 owing to massive double-seawall protection; however, adjacent beaches have narrowed since 1985 (Griggs et al., 2005). Table 11. Coastal vulnerability of the shoreline at San Onofre. Reactor San Onofre Tides Waves Erosion High Low Sea Level Rise Geomorphology Moderate Slope Low CVI High Moderate High The USGS ranked coastal features according to vulnerability to sea level rise. The average of each of these features determined the coastal vulnerability index (CVI) (Thieler and Hammar-Klose, 2000). 5. Discussion and Conclusions The major objective of this inquiry was to analyze (1) the effects of actual and predicted climate change on existing nuclear power plants and (2) the effects of using nuclear power on the ability of human society and ecosystems to adapt to climate change. Ultimately this objective sought insight on whether nuclear power plants currently in operation and proposed or under construction provide a useful technology for mitigating the existing and projected effects of climate change. In addition, the study sought insights on the future of nuclear power. Five criteria provided the framework to accomplish the needed analysis: (1) interrupted operations, (2) financial costs, (3) impairment of the adaptation of human systems, (4) impairment of the adaptation of natural systems, and (5) creation of other environmental problems. In this discussion we examine each of the criteria to understand the strengths and weaknesses of nuclear power as it is now. Water emerges as a key component of climate change interactions with nuclear power plants, and the effects of climate change on water are embedded in a larger set of debates about nuclear power: unresolved issues of safety, their costs if built, and the current global capacity to build enough plants to matter. In the final section, we weave together our assessment of climate-change induced water issues with these other matters and discuss the future of nuclear power in mitigating climate change. 5.1 Interrupted operations 35 While extreme climate events interrupt operation at both inland and coastal sites, the consequences at coastal locations are generally not as severe. Reactors typically resume power generation soon after storms have passed. In addition, during severe storms evacuations occur and therefore demand for energy is low. In contrast, heat waves threaten continued operation of inland reactors at a time of peak demand. A combination of low-flows due to drought and warmer temperatures in summer months raises the temperature of cooling waters prior to the onset of heat waves. Due to physical constraints related to the Carnot efficiency, nuclear reactors produce less energy with an increase in the temperature of cooling water. When heat waves hit, and nuclear power plants must reduce power to comply with thermal release regulations, the supply shortage becomes even greater. Consequently, power needs to be supplemented from other sources that potentially emit greenhouse gases. It can be argued that the interruptions of operation are intermittent and other sources of energy encounter similar interruptions. However, the timing of the supply interruptions and the actions needed to continue nuclear power operation remain a concern. 5.2 Financial costs Providing that lessons are learned quickly, climate change adaptation could be addressed with investment in defenses or changes to operation; however, the cost of these changes could be significant. Locating reactors farther from the coast have increased piping requirements that add to construction costs as do flood protection measures. Moreover, adjusting inland sites to drought1 and heat waves incurs both construction and operating costs. Cooling towers, particularly dry cooling towers, utilize less water, but involve greater construction and operation costs. For instance, 36 wet cooling towers reduce the overall efficiency of a power plant by 3-5% (Australian Uranium Association, 2007). While dry cooling towers uses less than 10% of the water required for wet cooling towers, these systems have much greater construction costs and require energy generated by the plant to run (Australian Uranium Association, 2007; World Nuclear Association, 2008a). 5.3 Impairment of the adaptation of human systems The financial resources needed to protect nuclear power plants from coastal impacts could in itself impair the adaptation of human systems. The funds needed to protect nuclear facilities reduce the funds available to protect other coastal developments. In addition, the protection of one piece of shoreline can increase erosion further along the shore. For example, losses of beach area have been noted near the San Onofre nuclear power plant. In many instances, damages to the coast are not immediately apparent; however models demonstrate the potential for future impacts to the coast. While adaptation at coastal locations involves planning for the future, recent events demonstrate that operations of nuclear power plants during heat waves at inland locations have already impaired the ability of human systems to adapt. In terms of lives lost, heat waves are the most dangerous of extreme climate events; therefore, reliability of the energy grid is essential to ensure public safety. Blackouts during heat waves were avoided in France by reducing energy exports to Italy and by taking measures that impaired the adaptation of natural systems. 5.4 Impairment of the adaptation of natural systems According to the USGS, many coastal nuclear power plants operate in regions that are among the most vulnerable to sea level rise in the United States. Coastal habitats must adapt to a rise in sea level; however, development along the 37 coast leads to a loss of coastal habitats as the land types are not able to “move” inland to accommodate to rising seas. Restricting development along the coast remains the best strategy to reduce loss of coastal habitat. Sites for nuclear power plants must be suitable for 100 years, thereby removing planned retreat as an option at these locations. While occurring slowly, the cumulative losses of coastal habitat become substantial over time. Similarly, the impacts of thermal pollution during heat waves might not cause large immediate fish-kills. However, the combination of higher temperatures from climate change and the warm effluent from nuclear power can cause serious changes to ecosystems. In France, easing of environmental regulations has become a permanent feature due to the issuance of an order in 2004 allowing higher thermal releases during summer months. 5.5 Other environmental problems Nuclear power has the potential for catastrophic accidents and consequently widespread environmental damage, unlike any other form of energy. The potential costs of not adapting nuclear operations to climate change are exceptionally high. Safe operation during extreme climate events remains a challenge. For one, the uncertainty in predicting climate change poses a problem for safety. Historical flood levels can no longer serve as an adequate predictor of future floods. As seen in France, recent floods have exceeded design basis levels. Regardless of design parameters, storms at coastal locations continue to be a problem because they often lead to the failure of multiple systems, and despite previous experience, failures in alarm and communication systems continue to occur. In certain cases, licensees have shown a low awareness of potential problems caused by external events. Moisture buildup leads to equipment failure; nonetheless, 38 a licensee at one site did not recognize the problem as something requiring preventative and corrective measures. In addition, after a hurricane had passed a site in Florida, the missile shield doors that protected safety related equipment were found open and according to the licensee these doors could have been open for several years. These examples indicate that licensees do not always take proper action in dealing with external events; moreover, they are not prepared for the issues that will arise due to climate change. 5.6 A look at the past, present, and future of nuclear power: Is it a suitable choice for climate change mitigation? The adaptation-mitigation criteria should be considered in conjunction with the other issues surrounding nuclear power particularly the existing issues tied to safety, cost, and the ability to construct the necessary number of nuclear power plants. Even in the absence of climate change, safety remains an important factor dissuading full public acceptance. The world now has more than 50 years experience with this technology, but memories of Three Mile Island and Chernobyl feed public distrust despite improvements made in safety and efficiency since those accidents. In addition, consensus eludes engineers and policy makers about how to deal with spent fuel. Although not addressed in this article, climate change will factor into the selection of suitable sites for storage facilities. Similarly, the potential costs incurred to adapt to climate change should be considered in the context of the already rising price tag of new nuclear power plants. Historically, nuclear power plants have been subject to cost and time construction overruns. Recent construction projects have not been successful in dismissing this syndrome. For instance, construction of the Olkiliuto plant in Finland, by the French company Areva, is over budget by 1.7 billion euro and behind schedule by 3 years 39 (Schneider et al., 2009). The ability to construct the required number of reactors to matter could be severely hampered if construction costs continue to rise. The matter of scale needed for nuclear power to have a significant effect as a mitigation technology must be addressed. Pacala and Socolow (2004) suggest that with planning new sources of CO2 could be eliminated thereby keeping global CO2 emissions at the current level of about 7 Gt of carbon per year. They estimate that one way of achieving 1Gt of the required reductions would entail adding about 700 new reactors of 1000 MWe each. This level of investment would approximately triple the existing world nuclear generating capacity of about 370 GWe (Schneider, et al., 2009). In their extensive analysis of costs, the retirement of existing reactors, capacities for building new reactors, subsidies, and unresolved safety issues, Schneider, et al. (2009) suggest that it is unlikely that the world’s nuclear industry could achieve the task of building up to 700 new reactors. Climate change provides one more obstacle to future reactor construction by limiting the number of suitable sites. Nuclear power requires an all or nothing approach; the high investment will not payoff unless reactors can be deployed at many sites. Constructing nuclear power plants at existing sites is the quickest option, but existing nuclear power plants already have vulnerabilities to climate. For instance, water shortages currently inflict many regions of the world and the situation will worsen with climate change. The Hadley Centre Global Climate Model predicts that the proportion of the land surface in extreme drought will increase from 1% for the present day to 30% by the end of the twenty-first century (Burke et al., 2006). Seasonal changes in precipitation, such as reduced snow pack and drier summers, pose problems for operation as well. Therefore, appropriate inland sites 40 for nuclear power plants will be more limited in the future because of diminished water resources. The lack of suitable sites at inland locations can not be addressed simply by locating nuclear power plants along the coast due to restrictions on coastal development. Table 12 summarizes the current and future challenges with nuclear power operation using existing technologies. Table 12. The Adaptation-Mitigation Dilemma for existing and new nuclear power plants. Criterion Interrupted operations Existing nuclear power plants Extreme storms interrupt both coastal and inland reactors. Consequences are most serious for inland reactors during heat waves: loss of power when it’s needed for public health and safety. Retrofitting US and French reactors with dry cooling systems will have high costs and will reduce power output of plants. Protecting coastal and inland reactors against floods and inundation will be expensive. Funds to protect nuclear power plants from sea-level rise will compete with funds needed to protect other infrastructure. Protection against erosion can exacerbate erosion in nearby coastal areas. Blackouts during heat waves have already impinged on health and safety. Existing coastal plants are already subject to sea-level rise, storms, and erosion. Protecting these plants will hinder the ability of coastal ecosystems to “move” inland as sealevels rise. Valuable coastal ecosystems services will thereby be diminished. Thermal pollution from inland plants is already altering ecosystems for inland reactors. Safe operation during extreme events at coastal sites is already more difficult. Predictions of storm and flooding events at coastal and inland sites already is hindered by new yet still unknown frequencies of extreme events. New nuclear power plants Extreme storms and heat will make identification of proper sites more difficult, and potential for interrupted operations will therefore increase. Moving new coastal reactors further from shore to protect from sea-level rise and storm flooding will produce extra expenses for extended piping. New inland and coastal plants will require extra consideration and expenses, which will compete with other needs to develop new infrastructure. Financial costs Impairment of adaptation of human systems Impairment of adaptation of natural systems Other environmental problems Avoiding sites that hinder adaptation of ecosystems to rising sea levels will exacerbate shortage of good sites for plant construction. Lower amounts of water at inland sites will necessitate either more expensive dry cooling or acceptance of higher levels of thermal pollution. New coastal and inland sites will have same difficulties with extreme events already experienced by existing plants. 41 In trying to predict the future of nuclear power, it is worthwhile to consider both potential and existing technology. Further into the future, new reactor designs might not have the same vulnerabilities as existing nuclear power plants. However, Generation IV reactors have not been built and tested and will not be deployed until 2030 at the earliest (OECD Nuclear Energy Agency, 2010). Climate change mitigation can not afford the time it takes for these reactors to get from the drawing board to operational. In the mean time, reactors now under construction, and planned for construction for the next two decades, require large volumes of water to operate. These reactors have a predicted operational life of fifty to sixty years. Adopting dry cooling systems remains the only way to avoid the problems with water shortages, but the additional expense could make nuclear power uneconomical. Currently no nuclear power plants in the US or France use dry cooling. Human ingenuity has few limits and nuclear power plants could be built to withstand greater storms, but financial limitations exist. For example, several sites in the United States have plans to construct AP10002 reactors because the original design had been certified by the USNRC. However, the manufacturer changed the design of the shield building to reduce costs and the USNRC found the revised design of the shield building to be vulnerable to severe weather. Striking the right balance between safety and cost continues to be a challenge. Achieving the desired level of safety, and minimizing the impact to climate change adaptation will likely be too expensive at many locations. Therefore, according to the criteria outlined here, nuclear power is not and will not be a suitable mitigation measure. REFERENCES Atwood, C.L., Kelly, D.I., Marshall, F.M., Prawdzik, D.A., Stetkar, J.W., 1998. Evaluation of Loss of Offsite Power Events at Nuclear Power Plants: 1980-1996. Idaho National Engineering and Environmental Laboratory, pp. 1-123. 42 Australian Uranium Association, 2007. Nuclear Power Reactors Briefing Paper #64, Melbourne. Autorité de sûreté nucléaire, 2000. Pressurized Water Reactors Chapter 11, 2000 Annual Report, Paris, France, pp. 193-261. Autorité de sûreté nucléaire, 2003. Electricity Generating Nuclear Power Plants Chapter 11, 2003 Annual Report, Paris, France, pp. 265-344. Autorité de sûreté nucléaire, 2006. EDF's Nuclear Power Plants Chapter 12, 2006 Annual Report, Paris, France, pp. 285-472. Autorité de sûreté nucléaire, 2007. EDF's Nuclear Power Plants Chapter 12, 2007 Annual Report, Paris, France, pp. 299-357. BBC News, 2003. Italy warns of power blackouts, BBC News. Bindoff, N.L., J. Willebrand, V. Artale, A. Cazenave, J. Gregory, S. Gulev, K. Hanawa, C. Le Quéré, S. Levitus, Y. Nojiri, C.K. Shum, L.D. Talley, A. Unnikrishnan, 2007. Observations: Oceanic Climate Change and Sea Level, in: Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor, H.L. Miller (Eds.), Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. Burke, E.J., Brown, S.J., Christidis, 2006. Modeling the Recent Evolution of Global Drought and Projections for the Twenty-First Century with the Hadley Centre Climate Model. Journal of Hydrometeorology 7, 1113-1124. Gorbatchev, A., Mattei, J.M., Rebour, V., Vial, E., 2000. Report on flooding of Le Blayais power plant on 27 December 1999. Institue for Protection of Nuclear Safety, Paris, France. Griggs, G.B., Patsch, K.B., Savoy, L., 2005. Living with The Changing California Coast. University of California Press, Berkeley and Los Angeles, California. Haney, C., 2006. Actions To Be Taken In Response To The Hurricane Lessons Learned Task Force Report (WITS NO. 200600198, TAC NO. MD1832), in: Leach, M.N. (Ed.). U.S. Nuclear Regulatory Commission, p. 30. Hibbs, M., 2003. EDF permitted to raise discharge temperatures at six nuclear sites. Nucleonics Week 44, 3. IAEA, 2003a. Extreme external events in the design and assessment of nuclear power plants. International Atomic Energy Agency (Engineering Safety Section), Vienna, Austria, p. 114. IAEA, 2003b. Flood Hazard for Nuclear Power Plants on Coastal and River Sites, IAEA Safety Standard Series. International Atomic Energy Agency, Vienna, Austria, p. 81. Kauffman, J.V., 2005. Impact of the 2004 Atlantic Hurricane Season on U.S. Nuclear Power Plants. U.S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research, p. 33. Klein, R.J.T., Huq, S., Denton, F., Downing, T.E., Richels, R.G., Robinson, J.B., Toth, F.L., 2007. Inter-relationships between adaptation and mitigation, in: Parry, M.L., O.F. Canziani, J.P. Palutikof, P.J. van der Linden, C.E. Hanson (Eds.), Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. Kuhn, G.G., 1980. Greatly Acclerated Man-Induced Coastal Erosion and New Sources of Beach Sand, San Onofre State Park and Camp Pendleton Northern Sand Diego County, California. Shore & beach, 9-13. Kuhn, G.G., Shepard, F.P., 1983. Beach Processes and Sea Cliff Erosion in San Diego County, California, in: Komar, P.D. (Ed.), CRC Handbook of Coastal Processes and Erosion. CRC Press, Inc., Boca Raton, Florida, pp. 267-299. Lagadec, P., 2004. Understanding the French 2003 Heat Wave Experience: Beyond the Heat, a MultiLayered Challenge. Journal of Contingencies and Crisis Management 12, 160-169. Leach, M.N., Billings, S.A., Kahler, R.E., Bolling, L.A., Hufham, J.W., Wert, L.D., Laura, R.A., Morell, G.K., Jackson, T.J., McLean, M.L., Rogers, A.H., 2006. Task Force Report 2005 Hurricane Season Lessons Learned Final Report. U.S. Nuclear Regulatory Commission, p. 90. MacLachlan, A., 2001. French Regulator Tells EDF To Protect Reactors From Flood Now. Inside NRC 23, 6. MacLachlan, A., 2003a. Environmentalists protest order on French river effluent temperature. Nucleonics Week 44, 8. MacLachlan, A., 2003b. Few power plants in France used discharge exemptions, says EDF. Nucleonics Week 44, 6. MacLachlan, A., 2004. EDF authorized to exceed discharge temperature for second year in row. Nucleonics Week 45, 7-8. 43 MacLachlan, A., 2005. Heat forces EDF to curb output at Tricastin, look at other units. Nucleonics Week 46, 7-8. MacLachlan, A., Marshall, P., Hibbs, M., 2006. Amid heat wave, French reactors get broader discharge exemption. Nucleonics Week 47, 1-5. MACTEC Engineering and Consulting Inc., 2008. Supplemental Environmental Resource Report Calvert Cliffs Nuclear Power Plant Unit 3 Calvert County, Maryland. Prepared for: Calvert Cliffs 3 Nuclear Project, LLC, Raleigh, North Carolina, p. 183. Maryland DNR, 2007. DNR Answers Questions about Sea Level Rise In Response to IPCC Report. Maryland Department of Natural Resources. Mayrland DNR, 1999. Shoreline Erosion in Depth, Maryland shorelinesonline. Maryland Department of Natural Resources, Coastal Zone Management Program, Annapolis, MD, p. 4. Meehl, G.A., Stocker, T.F., Collins, W.D., Friedlingstein, P., Gaye, A.T., Gregory, J.M., Kitoh, A., Knutti, R., Murphy, J.M., Noda, A., Raper, S.C.B., Watterson, I.G., Weaver, A.J., Zhao, Z.C., 2007. Global Climate Projections, in: Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor, H.L. Miller (Eds.), Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. OECD Nuclear Energy Agency, 2010. Generation IV International Forum FAQ. Organization For Economic Cooperation and Development Pacala, S., Socolow, R., 2004. Stabilization Wedges: Solving the Climate Problem for the Next 50 Years with Current Technologies. Science 305, 968-972. Parey, S., Aelbrecht, D., 2005. Short and Long Term Implications of the August 2003 Heat-wave on the EDF Generation system, Powerpoint presentation. EDF. Poumadere, M., Mays, C., Le Mer, S., Blong, R., 2005. The 2003 Heat Wave in France: Dangerous Climate Change Here and Now. Risk Analysis 25, 1483-1494. Power Engineering International, 2005. Italy: Italy's lack of natural resources and its fragile infrastructure are the cause of several probelms that have to be solved without an obvious source of investment, Power Engineering International. Rahmstorf, S., 2007. A Semi-Emprical Approach to Projecting Future Sea-Level Rise. Science 315, 368-370. Schneider, M., Thomas, S., Froggatt, A., Koplow, D., Hazemann, J., 2009. The world nuclear industry status report 2009, with particular emphasis on economic issues. German Federal Ministry of Environment, Nature Conservation and Reactor Safety. Southern California Edison Company et al., 2002. CDP Application No. E-00-014. California Coastal Commission, San Francisco, p. 39. The Riviera Times, 2005. Italy revises 2003 heatwave casualties, The Riviera Times. Thieler, E.R., Hammar-Klose, E.S., 1999a. National Assessment of Coastal Vulnerability to SeaLevel Rise: Preliminary Results for the U.S. Atlantic Coast. U.S. Geological Survey, Woods Hole, Massachusetts. Thieler, E.R., Hammar-Klose, E.S., 1999b. National Assessment of Coastal Vulnerability to SeaLevel Rise: Preliminary Results for the U.S. Gulf of Mexico Coast. U.S. Geological Survey, Woods Hole, Massachusetts. Thieler, E.R., Hammar-Klose, E.S., 2000. National Assessment of Coastal Vulnerability to Sea-Level Rise: Preliminary Results for the U.S. Pacific Coast. U.S. Geological Survey, Woods Hole, Massachusetts. U.S. Nuclear Regulatory Commission, 1994. Effect of Hurricane Andrew On Turkey Point Nuclear Generating Station, NRC Information Notice 93-53, Supplement 1:, p. 4. UniStar Nuclear Development, 2008. Probabilistic Risk Assessment And Severe Accident Evaluation, UniStar Nuclear Calvert Cliffs Unit 3 COLA Final Safety Analysis Report. Ward, L.G., Rosen, P.S., Neal, W.J., Pilkey Jr., O.H., Anderson, G., Howie, S.J., 1999. Living with the Chesapeake Bay and Virginia's ocean shores. Duke University Press, Durham and London. World Health Organization Europe, 2003. The health impacts of 2003 summer heat-waves Briefing note for the Delegations of the fifty-third session of the WHO Regional Committee for Europe, 12. World Nuclear Association, 2008a. Cooling Power Plants. World Nuclear Association, 2008b. Nuclear Power in France. World Nuclear Association. World Nuclear Association, 2008c. Nuclear Power in the United States. World Nuclear Association. 44 1 Potentially the largest climate impact to nuclear power operation, drought, has not been addressed in this article. Electronic annex 7 explores some issues that have arisen due to water shortages in the southeastern United States. 2 The Advanced Passive 1000 reactor is a pressurized water reactor with simplified design and passive safety features. 45