The question of which energy technologies generate the most greenhouse-gas emissions — cradle to grave — now has a more precise answer, thanks to a meta-analysis of life cycle assessment (LCA) studies done by the U.S. Department of Energy’s (DOE) National Renewable Energy Laboratory (NREL).
The new, robust analysis weighed the emissions estimates per kilowatt-hour from raw materials, manufacturing, transportation, operation, and decommissioning to get the best apples-to-apples comparisons. Sure, during operation solar panels release virtually no emissions, versus the tons of greenhouse gas produced by a large coal plant. But what about the emissions generated from the manufacture of solar panels versus, say, the turbines required for coal- and wind-based energy?
NREL’s LCA Harmonization Project gives decision-makers and investors more exact estimates of greenhouse-gas emissions for renewable and conventional generation, clarifying inconsistent and conflicting estimates in the published literature and reducing uncertainty.
The analysis found that from cradle to grave, coal-fired energy releases about 20 times as much greenhouse gas into the atmosphere per kilowatt-hour as solar energy. Wind and nuclear energy are on relative par with solar energy. Natural gas generation wasn’t included in the final analysis but is generally assumed to emit about half as much greenhouse gas per kilowatt-hour as coal.
What’s more, the “study of studies” narrowed the huge ranges of estimates sometimes as much as 90 percent, presenting a more reliable look at the likely greenhouse-gas emissions from different technologies.
Decision-Makers Need Environmental Costs Before Giving Go-Ahead
Lifetime greenhouse-gas emissions are an increasing concern for lawmakers and investors who must weigh the merits of a new coal-fired plant versus, say, a wind farm, and need to know not just the relative dollar costs but also the potential harm or benefits to the environment.
Until recently, emissions estimates ranged wildly, sometimes because vested interests had a stake in demonstrating that a certain technology’s emissions were high or low. For instance, if decommissioning costs aren’t included in a total-emissions estimate for nuclear energy or natural gas, those studies give artificially low figures.
NREL was seeing surprisingly high emissions numbers for concentrating solar power (CSP) plants, but deeper digging found that many studies combined the numbers from both CSP and natural gas when a utility used combustion of natural gas to supplement solar-energy generation. When the harmonization process allowed CSP emissions to stand on their own, the numbers plunged.
NREL looked at more than 2,000 studies across many generation technologies, applied quality controls, and greatly narrowed the range of estimates to reach reliable medians for greenhouse-gas emissions.
“This methodology allows you to arrive at a better precision, so you can say with more certainty that this is the benefit you get from using this technology rather than that technology,” said NREL Senior Scientist Garvin Heath, who led the project. “Anyone who wants a true comparison of the greenhouse-gas costs should benefit from this.”
Heath noted that today’s decisions on new plants will still have ramifications decades from now. Owners and investors will need to know about greenhouse-gas emissions and their possible effect on the bottom line, while policy-makers need to know the long-term implications of greenhouse gas on climate.
Investors “need to be very forward looking,” Heath said. A power plant is long lived, and its attributes and shortcomings are locked in for decades. That’s why investors push for estimates of greenhouse-gas emissions before they invest.
President Obama’s clean-energy standards require these estimates for each technology as a way to assign credits or discounts for building new plants. Credits, discounts, and the possible future price of carbon all figure heavily into decisions on which technology to choose.
“Analysts and decision-makers want a more robust sense or a narrower range of uncertainty to make the best decisions,” Heath said. Until now, no one has tried to differentiate between low- and high-quality estimates in a comprehensive way.