New Study: Clean Energy Least Costly to Power America's Electricity Needs
A new study published in the Journal of Environmental Studies and Sciences (JESS) authored by myself, Starla Yeh (NRDC), and Chris Hope (Judge Business School, University of Cambridge), shows that once estimates of climate change costs and other health impacts are taken into account, it would be cheaper to build new power plants from wind turbines or solar panels than from coal. It would also be cheaper to replace some of our dirtiest coal plants with these cleaner sources.
In other words, transitioning to cleaner energy won’t just help protect us and our children and grand-children from climate change, it’s also good economics.
The study results show that our electricity system, which generates fully 40 percent of the nation’s carbon dioxide pollution, is costly. Already, climate change is contributing to record heat waves, floods, drought, wildfires and severe storms. Such extreme weather caused more than $140 billion in damages in 2012. American taxpayers picked up nearly $100 billion of those costs, according to an NRDC report released in May, 2013.
This blog will start with the conclusions/policy implications of the paper, followed by a summary of the technical details for how I and my coauthors arrive at them.
In a nutshell…
Using widely available models of climate change pollution costs and other health damages from burning fossil fuels (i.e. from sulfur dioxide pollution), the study calculates the real cost we bear for electricity—not just what it costs to generate, but also the climate and health costs caused by power plant pollution. Our results support the need for strong and protective pollution standards for new and existing plants.
Taking all these costs into account, for new generation, we find that wind power costs about as much, and perhaps less, than conventional natural gas plants, and that any of the cleaner sources are less costly than conventional coal. This suggests that new coal and natural gas plants should not be allowed wherever wind is feasible, and that any of the cleaner sources should be chosen over conventional coal, including coal with carbon capture and storage (CCS).
For existing generation, we find that after taking into account all their costs, including climate change and health impacts, it would be less costly to replace many existing coal plants with new cleaner generation than to keep them operating. Replacement power from wind, natural gas, and natural gas with carbon capture and storage (CCS) would be justified based on the climate and health damages avoided by switching to these cleaner options. Under some scenarios, so would solar photovoltaic and coal with CCS.
These results mean that very strong pollution standards should be set for existing power plants. Compliance costs could be kept low by using a system-wide emissions-averaging approach. Emissions-averaging sets a limit over a fleet of power plants, rather than each one individually, allowing generators to meet the standard at low cost.
So…a story with a happy ending! How was it weaved?
Technical summary and the data
We arrive at our results in large part based upon a measure developed by economists called the “social cost of carbon,” or SCC. The SCC estimates how much damage one ton of CO2 emitted today causes now and into the future. Our study also includes economic damages caused by sulfur dioxide (SO2), another pollutant released by coal-fired power plants. Every year, SO2 causes thousands of premature deaths, respiratory ailments (e.g. asthma and bronchitis), heart disease and a host of ecosystem damages.
The figure below presents results for electricity costs for two SCC estimates in the article, both based upon a 12 member interagency analysis (click here for an interesting discussion of Republican efforts to stop the Environmental Protection Agency from considering climate damages when setting clean air standards). The first SCC estimate is from Johnson and Hope (JH, 2012), who redid the task force’s analysis using lower discount rates (more on discount rates later). The second is from the task force’s most recent analysis, which relied upon updated versions of the models it used in its earlier estimates not available when JH did their analysis..
I show costs just for the task force’s and JH’s central discount rates, 3 and 1.5, respectively. JH’s rate of 1.5 percent, half the task force’s, is roughly equal to the one used by the United Kingdom in the Stern Review (1.4 percent).
For each bar, total electricity costs for different generation technologies are given on a cents per kilowatt hour (kWh) basis, broken out at the bottom of each bar by production expenses (bottom number, for generation built in 2018 as estimated by the Energy Information Administration), SO2 impacts (middle number), and climate costs (top number); the total cost per kWh is shown at the top of the bar. (You can click on image to enlarge).
The task force’s damages are $33 per ton of CO2, and JH’s $122 per ton of CO2, for CO2 emitted in 2010 (2007$).
Without looking at the details, one of the most important aspects of the analysis is immediately apparent: carbon costs, and therefore total costs, vary dramatically with the discount rate. This is because economists “discount” damages that occur over many years into a (lower) “present value” (click here for a discussion of discount rates). While production expenses and SO2 damages are immediate, climate change harms our children, their grandchildren, and many future generations to come, so discounting matters a lot.
A quick digression is needed here, to make this issue concrete. The best way is by example. When discounted at 1.5 percent, $100 of future damages imposed on your kid 50 years from now counts as $48; the same damages discounted at 3 percent count as $23. One hundred years from now, that same $100 in damages are counted as $23 and $5, respectively.**
Okay, back to the results.
At the task force’s 3 percent rate, wind, conventional natural gas, and natural gas with CCS are all cheaper than either a new or typical coal plant in operation.
At JH’s 1.5 percent discount rate, all of the cleaner technologies are cheaper. A particularly striking result is that climate damages are more than four times production expenses (13.3 cents versus 3 cents/kWh) for a typical coal plant currently in operation.
Two other results merit mentioning.
First, the total costs of wind and natural gas are approximately equal, 8 versus 7.5 cents even for the task force’s higher discount rate. In fact, though not in the graph, by 2018, they are 8 versus 7.8 cents, respectively. The task force’s SCC for 2018 emissions is $41/ton of CO2 (JH estimated damages only for 2010 emissions, while the task force has a schedule many years into the future, with damages increasing over time as climate change accelerates). Further, as the authors note, climate damages from methane leaks from gas drilling are not counted; if they were, wind would likely be significantly less costly than natural gas.
Second, while new solar and coal with CCS cost about the same as new conventional coal at the task force’s discount rate, for emissions in 2018 conventional coal is more expensive than these lower-emitting choices: 13.3 and 13 cents for solar and coal with CCS, respectively, versus 13.8 cents for new coal (this is also not depicted in the graph).
Three factors that make the results conservative
There are three key factors that make fossil fuel generation is even more costly than our analysis indicates.
First, most of the damages from climate change are not included in the SCC; the damages are therefore greatly underestimated. To name a “just” a few (there are many more): forest fires, drought, smog (and associated asthma and other respiratory illnesses), interaction effects from higher sea levels and storm surge (which is what made Superstorm Sandy so devastating), increasing food prices from drought, dried up canals where goods can’t be shipped, water shortages, forest dieback from pest infestations, coral reef disintegration, ocean acidification, and socio-economic conflict from shrinking water availability and other resources. These are left out because they are much harder to measure than the damages included in the calculations, such as higher air conditioning costs, property lost to higher seas, and flooding from coastal storms.
Second, JH relied upon the older version of the task force’s model. Holding the discount rate constant, JH’s SCCs would be much higher with the task force’s updated model. JH also did not estimate SCC values for emissions in the future years; the SCC rises over time, reflecting increasing marginal costs of CO2 emissions.
Finally, generation costs are likely to be overestimated for the cleanest technologies, to the extent that innovation is likely to bring down their costs. If one looks at past projections for renewable energy technologies, one consistently finds cost overestimates. Further, the generation cost estimates exclude upstream externalities associated with fossil fuel extraction, such as methane emissions from natural gas wells and spills from pipeline transmission, and land disturbances from coal mining.
Some concluding remarks
Our results make a good case for strong protective standards on carbon pollution from our nation’s power plants. We can help slow climate change while at the same time reducing the real cost (i.e. costs inclusive of pollution damages) of our electricity. Replacing our dirtiest coal-fired generation with cleaner options like wind, solar, and natural gas, and building new generation from these same sources instead of coal, is good economics.
Climate change is a growing threat to our children and grandchildren. Even though power plants are the nation’s single largest source of carbon pollution, there are no federal limits on how much they can emit. We limit the amount of mercury, arsenic, soot, cyanide, and other harmful pollution from these plants. It's time to cut this carbon pollution. We owe it to ourselves, and to future generations.
** The SCC $/ton values represent the sum of damages in all future years from one ton of CO2 emitted in a given year; the example above is just a hypothetical amount of damages equal to $100 in one year not associated with any particular amount of CO2 emitted.