Matthew Kotchen, Yale University
Abstract:
The phasedown of coal for electricity generation is considered vital to meeting global climate targets. Many countries have pledged to stop using coal, with some as early as 2030. While the United States has no target currently in place, several states do. In this paper, we examine the feasibility of phasing down U.S. coal-generated electricity given the existing fleet of power plants. In particular, we take consumption as given and evaluate how prioritizing natural gas generation over that of coal would change emissions and operating costs. To do this, we develop a replacement algorithm based on transmission regions and marginal cost comparisons. Using our preferred scenarios, we find that between 66 and 94 percent of coal generation could be replaced immediately, reducing electricity sector carbon dioxide (CO₂) emissions between 18 and 29 percent – equivalent to between 5 and 8 percent of total U.S. energy related emissions. The cost range is between $49 and $92 per ton of CO₂, where benefit-cost ratios are favorable in some scenarios considering local pollutant co-benefits alone. Despite the command-and-control nature of prioritizing natural gas generation, we find it relatively cost effective even in comparison to a Pigouvian tax. We examine sensitivity of the results to transmission regions, replacement cost conditions, natural gas pipeline capacity, and alternative fuel prices.
Bio:
Matthew Kotchen is a professor of economics at Yale University, with a primary appointment in the Yale School of the Environment and secondary appointments in the Yale School of Management and the Department of Economics. He is a research associate at the National Bureau of Economic Research (NBER) and a fellow of the CESifo Network. Professor Kotchen's research interests lie at the intersection of environmental and public economics and policy. Kotchen joined the Yale faculty in 2009 and has held previous and visiting positions at Williams College, University of California (Santa Barbara and Berkeley), Stanford University, and Resources for the Future. Professor Kotchen has also served as the Associate Dean of Academic Affairs, the Deputy Assistant Secretary for Environment and Energy at the U.S. Department of the Treasury in Washington, DC, the visiting chief economist at the Environmental Defense Fund (EDF), and as a member of the Environmental Economics Advisory Committee of the U.S. Environmental Protection Agency.
A Research Summary published in the NBER Reporter is available here. Kotchen also serves as the founding organizer and editor of Environmental and Energy Policy and the Economy and a co-editor of the Journal of the Association of Environmental and Resource Economists.
Summary:
Coal burning is still a major source of GHG and particulate pollution on the planet
Many countries, as well as US States have pledged to phase out the use of coal for power
2022: 20% of US power use (majority is natural gas)
Does this mean that the drawdown of coal is unavoidable or does it need policy incentives?
Probably not
Electricity use is rising (e.g. EVs, data centers)
Need more power to cover baseload
Solar, wind are too intermittent until grid-scale batteries are available
Research questions:
How feasible is to for US to phase out coal in the short term (using existing generation sources)?
Costs/benefits?
How would a phaseout compare to a carbon tax?
Challenge
This is a non-marginal change, question is about a hard push to phase out coal
Important to capture long-term impact (e.g. less coal power may make EVs less attractive)
Long-run / Grid-dispatch models capture the power grid, generator entry and exit
NEMS, ReEDs E4ST, PLEXOS, ISO model
Not good for discontinuous change
Instead use revealed preferences based on “availability”
Reduced form approach targeted to this research question
Focus on net load in PJM region
Net load: subtracts renewables and hard-to-dispatch technology (nuclear)
Different fossil-based generation technologies: natural gas, coal, combined-cycle
Make coal more expensive than natural gas
Model out power generation mix over time at hourly resolution to understand how much coal would actually need to be used using the current PJM grid
I.e. natural gas always gets priority over coal when allocating power
Observation: lots of spare gas capacity across US power generating regions
In Northeast you never have a shortfall
In both other regions there is just a little shortfall in the middle of the summer (e.g. due to AC use)
Supply-demand are connected geographically at the granularity of regions (coal generator in region X can be replaced by a gas generator in region Y), using different spatial region resolutions
Policies:
Assign a cost to different generators based on how they are used in real dispatch (this is an incomplete estimate of their real costs)
Baseline: Gas has priority over coal (not a ban)
Has artifacts because sometimes the cost of constrained dispatch is lower than unconstrained
Thus, too optimistic
Corrections:
RP Own: can only use gas to replace coal if it has higher cost
RP Max Coal: can only use a gas to replace the most expensive coal
RP Max Any: same but more expensive power generation mechanism of any kind (thus, coal may be used)
Solve power dispatch optimization problem
Data: Year 2022
Hourly power generation from EPA
Emission rates
Net capacity (CEMS)
Marginal cost estimates, across different power generator types
AP3 model to value local pollution and its health costs
Look at
Remaining coal use
Power generation costs
Emissions
Overall net benefits (environmental benefits - increased generation costs)
Result:
Was able to reduce power generation in US by 93%
Results in 29% reduction in CO2 emissions
$49/Ton CO2
Benefits from local pollution reductions is higher than costs
US emissions drop by ~7%
What if we use carbon tax to direct power generation instead of directly controlling power dispatch?
Optimize using revealed preferences choices
Priority dispatch archives 91% of the benefits of the carbon tax, which is a more efficient mechanism
Including natural gas pipeline capacity constraints doesn’t change the results much
From 385 coal power plants 110 shutdown completely, others curtail severely (>75%)
At 2019 prices the cost was $9/Ton (different gas vs coal prices)