California is a national leader in the transition to clean electricity, but integrating all that renewable energy onto the grid is no easy feat. Critics of renewables have long claimed that adding variable renewables (such as wind and solar) to the grid would force fossil-fueled power plants to operate less efficiently, leading to an increase in air pollution emissions. And as states start to integrate more renewables, those claims have come back.
But is this argument true? Do renewables really lead to increased air pollution?
(Spoiler: I crunched the numbers for California, and the answer is no.)
What’s the worry?
The argument that renewables will increase air pollution is alive and well, popping up most recently in North Carolina when Duke Energy sought changes to the air permits for a few of their natural gas power plants. Even UCS has been concerned about this issue, and one of the conclusions from our Turning Down the Gas in California study was that, as California adds more and more renewables to the grid, gas plant operations will continue to change in a way that could increase air pollution from those plants.
Renewable energy generation in California has drastically increased over the past decade.
Note: The data in this graph is from the California Energy Commission, and it does not include rooftop solar generation.
How exactly would more renewables increase air pollution?
The theory is that variable renewable generation would force fossil-fueled power plants to ramp down or turn off entirely during certain times of day to accommodate renewable generation. Then, as the renewable generation peters out, the power plants would need to ramp back up or turn back on to ensure enough electricity is on the grid. These changes in power plant operations (ramping, starting, and stopping) are collectively referred to as “cycling.”
The kicker here is that more power plant cycling, especially starting a plant, could lead to more air pollution emissions from that plant. Most natural gas, coal, and other fossil-fueled power plants have emissions control equipment designed to reduce the amount of air pollution that power plants emit. Emissions control equipment needs to be at a certain temperature to function properly, and the equipment isn’t particularly effective when a plant is starting because the equipment hasn’t been warmed up enough yet. So, if more renewables lead to more power plant cycling, then air pollution emissions could potentially go up.
Interesting theory, but does this really happen?
Again, this concern has been around for a while. Back in 2013, the National Renewable Energy Laboratory studied this exact question. They found that adding renewables significantly decreased air pollution emissions overall because, even though power plants were cycling more, they were generating less electricity and burning a lot less fuel. This decrease in power plant generation drove down air pollution emissions, overshadowing the effects of cycling.
But now that it’s 2019 and California has added a tremendous amount of renewable generation (there’s been roughly a 150% increase in solar and wind generation since 2013!), I wanted to find out how the operations of California’s power plants have changed over time and if those changes have impacted air pollution emissions.
California analysis overview
To answer these questions, I analyzed the hourly data for California gas plants collected by continuous emissions monitoring systems. (Most of California’s gas plants are required to use this monitoring equipment and report results to the EPA.) I included data from the last decade, and I broke up the analysis to examine three different types of natural gas power plants (combined cycle, steam turbine, and gas turbine). In general, combined cycle power plants are highly efficient but not very flexible, gas turbine power plants are very flexible but less efficient, and steam turbine power plants are old, inflexible, and inefficient, and they are mostly being phased out. Finally, I focused on nitrogen oxides (or NOx) pollution in particular, since NOx is the main air pollutant from gas plants. While NOx has negative health effects on its own, it is also a precursor to ozone and particulate matter, which are even more harmful to human health.
Are gas plant emissions changing?
Overall, I found that total NOx emissions from California’s power plants have been trending down over the last decade. It’s been a bumpy ride, but the trend is certainly downwards (see graph below).
The reason NOx emissions bounce around so much is because total NOx emissions depend, at least in part, on the overall level of natural gas electricity generation. For instance, NOx emissions plummeted in 2011, when record amounts of precipitation both in the Pacific Northwest and California led to a glut of hydroelectric power, significantly reducing gas generation. The very next year, 2012, saw very low levels of hydroelectric generation in California and the sudden closure of the San Onofre Nuclear Generating Station, causing gas generation (and NOx emissions) to spike back up to fill in the gaps.
There have also been changes in which types of plants are responsible for the most NOx emissions. While emissions from steam turbine units have tapered out over the last decade, emissions from gas turbines have increased. Given that power plants, particularly gas turbines, are disproportionately located in California’s most disadvantaged communities, this uptick in gas turbine emissions means that some of California’s communities might not be seeing the full benefits of this overall decrease in air pollution emissions.
While total NOx emissions from California gas plants have generally decreased over the past decade, a number of external factors (such as electricity generation from hydro and nuclear) have a pronounced influence on the overall NOx emission levels.
Note: Steam turbine NOx emissions start so high in 2009 and 2010 because an old, highly-polluting power plant continued to operate until it retired in 2010.
What’s causing the overall decrease in emissions?
Quite a few factors affect the total level of NOx emissions from California’s power plants. As I’ve already mentioned, the overall amount of natural gas generation is one of those factors. After gas generation spiked up in 2012, the overall level remained roughly the same until 2016, when gas generation started declining significantly (see figure below). This decline over the past few years certainly helped to push down the total level of NOx emissions.
After a sharp rise in 2012, electricity generation from natural gas plants remained at similar levels until it declined over the last few years. Overall gas generation fluctuates in response to external factors (such as electricity generation from hydro and nuclear).
Note: The “Total” generation data is from the California Energy Commission, and it represents generation from all gas plants in California. The EPA data analyzed here only includes data from a subset of the state’s gas plants, accounting for roughly 80% of electricity generated from natural gas in California.
Another factor that affects NOx emissions is the average NOx emissions rate, i.e. how much NOx a power plant emits in order to generate one unit of electricity. As older, dirtier gas plants were retired, and as newer gas plants with more effective emissions control technologies came online, the average NOx emissions rate has fallen significantly over the past decade (see weighted averages in figure below).
While the combined cycle NOx emissions rate has not changed very much over the past decade, the gas turbine rate and the overall rate have trended downwards over time. Since so much of California’s natural gas electricity generation comes from combined cycle plants, the overall weighted average NOx emissions rate tracks very closely with the combined cycle emissions rate.
Note: The overall NOx emissions rate includes all gas plants in the EPA data, including steam turbines. Furthermore, the NOx emissions rate for combined cycle plants is likely overestimated here because the EPA data does not include electricity generation from the steam turbine portion of each combined cycle power plant.
What about cycling?
As I mentioned earlier, cycling (ramping, starting, and stopping) of gas plants can lead to increases in NOx emissions because emissions control technologies are less effective when a plant is cycling. California’s gas plants, particularly gas turbines, have been starting much more frequently over the past decade (see figure below), with a very steady increase in the number of starts over time. While the massive increase in gas turbine starts has not dramatically increased the gas turbine NOx emissions rate, all these starts are undoubtedly keeping the overall NOx emissions rate higher than it otherwise would be.
California’s gas turbines have been starting more and more frequently over the past decade as more renewables have come onto the grid. While this increase in starts could theoretically drive up NOx emissions, other factors, such as an overall decrease in natural gas electricity generation, have kept NOx emissions trending downwards.
Note: Renewable generation data includes solar PV and wind generation only. It does not include rooftop solar data. Additionally, it’s important to note that not all starts produce the same amount of NOx emissions. NOx emissions from starts generally depend on the capacity (MW) of the plant, and since gas turbines usually have a smaller capacity than combined cycle plants, gas turbine starts may emit less NOx than combined cycle starts.
Are we in the clear?
With overall NOx emissions and NOx emissions rates both going down, recent trends have been encouraging. However, California’s gas plants have been cycling more and more frequently, and UCS will remain on the lookout for future changes in air pollution emissions.
California’s communities still face some of the worst air quality in the country, so action to improve air quality is urgently needed. UCS remains committed to ensuring that changes in power plant operations do not lead to an increase in air pollution in California’s communities.
California is far ahead of the rest of the country in adding large amounts of renewables to the grid, and the rest of the world is watching as the state marches into uncharted territory. Because no one has ever decarbonized their electric sector to the extent that California plans, we need to keep watch for a wide range of unanticipated impacts to ensure the transition to clean energy goes as smoothly as possible.
Analysis details:
- This analysis used EPA data for California gas plants collected by continuous emissions monitoring systems. Not all California gas plants are required to report emissions to the EPA, so this analysis focuses solely on the plants responsible for roughly 80% of California’s in-state electricity generation from natural gas.
- I conducted this analysis on a unit level, not a plant level. Some power plants in California have multiple unit types (combined cycle, gas turbine, steam turbine) installed at the same facility, and the results reported here are broken up and based on unit-specific operations. Importantly, this means that the results for power plant starts are for unit starts, not entire power plant starts.
- When analyzing NOx emissions and NOx emissions rates, I visually inspected the data and removed erroneously high values for specific units for specific years. For almost all years, the removed values accounted for less than 2% of generation for each type of gas plant. The highest removed values were for gas turbines in 2013 and 2014. In those two years, removed values accounted for 8-9% of all generation from gas turbines, which means the NOx emissions for gas turbines in 2013 and 2014 are likely being underreported by 8-9% in this analysis.
- I removed erroneously high values for NOx emissions from this analysis because, when emissions monitoring equipment is not working correctly, there is a conservative process for estimating emissions that can significantly overstate emissions during these periods. However, there is no way to distinguish between periods when an emissions monitor was not working vs. periods when emissions control equipment was not working, so it is possible that some real values were incorrectly removed from this analysis. To prevent such errors, I cross-checked all values with emissions reported by the California Air Resources Board before removing them. All values that were cross-checked and removed were substantially higher than the emissions reported to the California Air Resources Board.