Narrowing Estimates to a More Reliable Median
Heath, Technology Systems and Sustainability Analysis Group Manager Margaret Mann, and their colleagues at NREL’s Strategic Energy Analysis Center discovered that there is no shortage of reports and studies estimating the greenhouse-gas emissions of various energy technologies — in fact, there are hundreds for each — but they vary greatly in quality, consistency, and validity.
Narrowing these ranges of estimates is like hooking together a 1,000-piece jigsaw puzzle, or solving a devilish crossword.
Heath and his colleagues narrowed down the various studies to those meeting minimum thresholds for quality, relevance, and transparency. They looked upstream, making sure to include the materials that go into building a solar, wind, or coal plant. They examined in detail the emissions that occur when the plant is in operation. And they looked downstream: when a plant ceases to operate, what are the environmental costs of decommissioning it?
After harmonizing 397 LCA studies of greenhouse-gas emissions from photovoltaic (PV) projects, the study found a median of 45 grams of carbon dioxide emitted per kilowatt-hour of electricity generated. Just as important, the middle 50 percent of the range (75th minus 25th percentile) of differences narrowed — from 44–73 grams before the harmonization to 39–49 grams after the harmonization.
The 45-gram median for PV contrasts with a 1,001-gram median for coal.
Nuclear energy was found to have a large range even after harmonization reduced the published range by 50 percent, to 4–110 grams. The median harmonized estimate for nuclear is 12 grams, very similar to that of many renewable technologies.
Concentrated solar power’s medians were 26 grams for parabolic trough plants and 38 grams for power towers. Thanks to harmonization, the variability in estimates for CSP was reduced by 91 percent.
Thin-film solar cell technology demonstrated very low greenhouse-gas emissions: a median of 20 grams for amorphous silicon; 14 for cadmium-telluride; and 26 for copper indium gallium diselenide, with overall ranges reduced from 19–95 grams to 18–52 grams equivalent carbon dioxide per kilowatt-hour.
For wind, the harmonization study was able to narrow down the wide range of published estimates — 1.7–81 grams — to a harmonized range of 3–45 grams (a 47 percent reduction) and a more reliable median of 11 grams of carbon dioxide per kilowatt hour.
The Fine Art of Comparing Apples to Apples
“We did a systematic review of the literature — not just peer-reviewed journals, but government reports, theses, reports from non-government organizations and private companies,” Heath said. “We had to verify them all.”
The team brought in experts from different disciplines, including technology experts, literature search experts, and experts in LCA.
“We screened for quality and for transparency, and we screened out the studies that were of older technologies that probably won’t be going forward.”
If a study didn’t use sound methods, it was out. But if a good study lacked a particular variable — such as not considering decommissioning — the team used formulas to add in greenhouse-gas emissions from those activities so the otherwise-satisfactory study could be included.
“After our quantitative adjustments, each study included in the final analysis was consistent in the broad sense, if not in every detail,” Heath said.
Harmonization studies help guide policy locally, statewide, and nationally. Lobbyists, lawmakers, and the cost-analysis community can all use the harmonization estimates as building blocks to make their own estimates for specific projects or to guide policy.
“They can get numbers to plug into their own models; they can use our literature to assess the climate impacts of transitioning to new forms of energy,” Heath said.
Findings Appear in Special Edition of Journal of Industrial Ecology
The NREL method is being heralded as an important step forward in LCA studies that paints a clearer picture of the environmental penalties and benefits of different technologies.
NREL’s findings appear in six articles and an editorial in the April special supplemental issue of the Journal of Industrial Ecology on Meta-Analysis of Life Cycle Assessments.
Also contributing to the findings were subcontractors and researchers from DOE’s Brookhaven National Laboratory.
“This is great work,” said Reid Lifset, editor of the Journal of Industrial Ecology. “Garvin and his crew have done something very important.”
A Crucial Tool for Future Decision Making
“As a society, we need to better understand what the effects of our energy choices are,” Heath said. “Greenhouse gases and climate change are a part of the discussion. As we try to envision what our future energy system will look like, we need an accurate picture of what that transition will mean.”
“The scope of these syntheses is staggering and provides real insight into what can be done with the burgeoning research literature in industrial ecology,” Lifset said.
Peter Crane, Yale University’s dean of Forestry and Environmental Studies, agreed: “The application of meta-analysis to life cycle assessment is an important advance. Decision-makers seeking to make greener choices need a way to make sense of the information that is coming at them so quickly now. This can help.”
Heath added: “If we care about stabilizing our climate, we need to look at greenhouse gases. This study assembled the best available and most robust evidence. It helps us see the state of our planet, and it can help us make better decisions going forward.”