skip to main content

→ Top Stories:
Safe Chemicals
Defending the Clean Air Act

Dave Hawkins’s Blog

Unpacking EPA's Carbon Pollution Proposal

Dave Hawkins

Posted June 5, 2014 in Curbing Pollution, Health and the Environment, Solving Global Warming, U.S. Law and Policy

, , ,
Share | | |

Let me start by saying I am not going to be talking in this blog about the most important issue presented by EPA’s proposal to cut carbon pollution from existing power plants.  That key issue is how much will the rule reduce total U.S. carbon pollution by 2020 and beyond.  After all, that is why EPA is proposing these limits in the first place.  The good news is that we actually can cut pollution more than EPA has proposed and do so with huge health and economic rewards.  More on that in a later blog.

Now to what everyone seems to be focused on: how did EPA set the proposed reduction targets for states?  The core of EPA’s proposal is a set of state-by-state targets.  There has been a lot of confusion in the press and the blogosphere about how they were developed, what their impacts are, and how they relate to the expected emission reductions by 2020 and 2030. 

EPA developed these targets by considering the emission reducing potential for a common set of measures in the context of each state’s starting point in 2012.  Because the states have very different starting points, they reach different end points in 2020 and 2030, when considering a common set of tools.

Here are the basics.

The state targets are not based on some national emission reduction target.

EPA has estimated that its proposal would reduce power sector carbon pollution by 30% from 2005 levels in 2030.  Yet a number of state targets produce greater or lesser changes than 30% when measured from a 2012 baseline.  Commentators have puzzled over this.

The answer is quite simple.  EPA did not develop the state targets to achieve any preconceived national reduction target.  EPA developed the state targets by taking each state’s power system as it operated in 2012 and then used a common toolbox of pollution-reduction measures to assess how much of a cleanup each state could achieve by 2020, with additional cleanup from 2020 to 2030.

The national estimates of 26% reduction in 2020, and 30% reduction in 2030 (both measured from 2005 levels) are not separate goals that drove the state targets.  They are simply the cumulative result of each state meeting its customized targets.

Understanding the differences in state targets

EPA posted a huge amount of supporting material for the proposal and it is understandable that many have focused on a few tables – specifically those that list each state’s target and identify the change in the state’s carbon intensity compared to 2012.  

“Carbon intensity” is jargon for the average amount of carbon dioxide (CO2) in pounds that are emitted for each megawatt-hour (MWh) of electricity produced.  Think of it as the state’s body fat measurement—lower is healthier.  States are like people in one respect: some have much higher electricity “body fat” than others.  Coal is the butterfat in states’ electricity diet; they more they consume, the higher their carbon intensity.

But comparing different states’ 2030 targets to one another, or comparing a state’s target to its 2012 starting point, can be misleading.  What most people are interested in is “how hard will it be for my state to reach its target?”  Looking at the percentage change in a state’s carbon intensity alone will not tell you that it will be a breeze for some states to make large percentage changes, while it will require some more effort for some other states to achieve a smaller percent change. (Although, EPA’s targets are based on such modest improvements from today that no state will have to struggle to meet its target.)

Understanding the impact of using a 2012 start year

Some have argued that using a 2012 start year will “penalize” states that cut their carbon pollution before 2012.  That is not correct.  EPA’s approach is based on estimating what additional improvements are feasible for a state, given what the state has already done by 2012.  States that have done a lot already are ahead of the pack, but they don’t have as much left on the shelf as other states that have not yet started.  EPA’s approach accounts for that reality and does not require the same future effort from a state that has acted early as from a state that’s just starting.  In fact, the early actor states start EPA’s race closer to the finish line.  As I will discuss in a specific example below, this makes it easier for those states to hit their targets.

EPA’s toolbox (or “building block”) approach

States have great differences in their energy mixes.  EPA’s toolbox approach to setting targets seeks to deal with these differences by assessing how a set of four potential policies (“building blocks”) could lower each state’s carbon intensity from its 2012 starting point.

 The four building blocks

1. Make coal plants more efficient: Hardware and software tweaks to get more electricity out of a ton of coal.  (Not wasting fuel is a no-brainer.)

2. Use gas plants more effectively: Running gas power plants ahead of coal plants gets reductions from capital investments already made in the state.  (It also means more carbon pollution than efficiency and renewables, which is why it will make sense for EPA and the states to place much greater reliance on the next two building blocks.)

3. Increased wind and solar:  Grow renewable generation at a rate already being met by other states in the region.  (This is almost a lowest-common-denominator approach in the proposal. We can do better.)

4. Energy efficiency: Apply energy efficiency policies already adopted by many states.  (These are not bleeding edge requirements; much more middle of the pack levels of effort.)

These building blocks are not required actions: they are simply the types of actions EPA considers in setting the state’s target.  States are free to use all, any, or none of these approaches in creating their own plans and to use them in different proportions than EPA assumed in doing its study.


How state targets are set—Step by step

EPA starts with the state’s fossil power plant carbon intensity in 2012 and then adjusts that rate to include the effects of 2012 levels of non-hydro renewables generation and credit for continued operation of nuclear units at risk of retirement (6% of state’s nuclear capacity). In most states, that broader power system rate is lower than the fossil only rate.

Applying the building blocks:

Block 1: EPA calculates reductions from improving coal plant efficiency by 6% (a value documented as achievable in an exhaustive study).

Block 2: EPA identifies unused capacity in existing Natural Gas Combined Cycle plants in the state and calculates the reductions from increasing their capacity factor to 70%, displacing coal with the increase in gas generation.

Block 3: EPA calculates the reduction that would occur if the state grows renewable generation in an amount equal to the average required by current policies of other states in the same region.

Block 4: EPA calculates the reduction from implementing a policy to save 1.5% per year in electricity production through demand-side energy efficiency programs.  1.5% is a level adopted by a number of states already.  States that have no energy efficiency programs now have targets based on a slow ramp up to the 1.5% level.


Let’s take a look at an example State.

Michigan is a useful example to look at in detail.  It is a Midwest industrial state that burns a lot of coal and a lot of gas to make power: about three times as much coal as gas in 2012.  But it also started ramping up renewable energy and energy efficiency programs several years ago. 


How EPA’s approach operates in Michigan.

Michigan’s 2012 fossil fleet average emission rate was 1814 lbs CO2/MWh.

After including 2012 renewables and 6% of nuclear in the rate, MI’s starting electric system-wide rate drops to 1690 lbs.

Applying building block 1 (coal plant efficiency) brings the target for 2030 to 1603 lbs.

Applying building block 2 (increased operation of existing gas plants) brings the 2030 target to 1408 lbs.  This based on a shift of 22% of MI’s 2012 coal generation to existing MI gas plants.

Applying building block 3 (increased renewables generation) brings the 2030 target to 1339 lbs.   EPA’s guideline is based on MI simply achieving the current regional average renewables policy level and would result in a total MI renewable generation of 4.8 million MWh in 2020, increasing over the decade to 8.1 million MWh in 2029.  Thanks to renewables policies MI enacted in 2008, MI is already doing better than the EPA guidelines in this building block.   MI generated or acquired certificates for about 7 million MWh of renewables in 2013 and is projected to be at about 9 million MWh by 2015.  This is 15 years ahead of the target date for hitting the 8 million MWh level found in the EPA guideline.

Applying building block 4 (energy efficiency) brings the 2030 target to 1161 lbs.  This is based on MI starting to improve energy efficiency by 0.99%/year in 2017 and ramping up to a 1.5% level by 2020.  Thanks to energy efficiency programs now in operation in MI, the state is already ahead of schedule and on track to meet the 2030 target if it simply continues to achieve its current level of annual energy savings in the future.

EPA’s guideline assumes MI achieves a cumulative savings of 4.6% by 2020, continuing to a cumulative saving of 11.8% by 2029.   Simply by continuing its current rate of annual savings, MI will achieve a 5.2% cumulative savings by 2016—four years ahead of EPA’s guideline schedule—and will achieve cumulative savings of 11.7% by 2021.  Any additional efficiency savings resulting from building codes or appliance standards will put Michigan further ahead.  Because EPA’s guideline allows for average compliance over the decade, this head start by MI will provide it with additional flexibility.

EPA’s proposal allows for gradual implementation

As mentioned above, the state targets set by EPA are required to be met by 2030. Compliance with an interim target will begin in 2020 and that target can be met on an average basis over the decade.  This provides great flexibility in the timing of actions MI takes to reduce its power sector carbon intensity over time.  But it also presents a risk that the reductions we need in 2020 may get delayed.  EPA’s final rule needs safeguards to prevent this.

The bottom line for MI and other states

EPA’s proposal establishes customized targets for each state, based on very modest rates of improvement in four areas that are applied in a non-discriminatory manner to all states.  Of the four building blocks used by EPA to set state targets, MI is already doing better on two of them—energy efficiency and renewables. 

The remaining two—coal plant efficiency improvements and increased operation of existing natural gas plants—are straightforward, low cost measures and the guideline provides ample time to carry out these activities, if MI chooses to use those strategies.

But EPA’s guidelines do not require MI or any state to use any particular measures.  For example, states can skip the coal plant patch up and gas plant ramp up measures and prioritize the cleanest energy resources: renewables and energy efficiency.  EPA’s guidelines provide states with complete freedom to choose the approaches that best meet each state’s needs, requiring only that the state achieve the overall target in 2020 and make further progress to 2030.

Share | | |


Michael NobleJun 5 2014 07:38 AM

Best explainer so far. Experience matters, David.

Craig OrenJun 5 2014 10:31 AM

great explanation, Dave. This is the best thing I've read so far.

Michael BerndtsonJun 5 2014 12:52 PM

I'm a bit confused about carbon intensity, lbs CO2 per MWh.

Emissions are typically determined on a mass per time basis. Like pounds of something per daily/monthly average and an annual allowable total. This is a mass of pollutant permitted for discharge into the atmosphere or body of water.

Carbon intensity seems like a concentration. The volume analogy would be mass of pollutant per volume of media. For air and water discharge this would be mass of pollutant in the vapor stream discharging into the atmosphere of mass of pollutant in an aqueous stream being discharged into a body of water, respectively.

So let's assume a carbon intensity goal of 1000 lbs of CO2 per MWh. A utility or electricity provider could add controls and optimize its plant performance to achieve that intensity on the one hand. And pay for those upgrades by ramping up electricity generation (i.e. burn more coal) as long as the intensity is at or below 1000 lbs CO2. This is sort of called fiddling with the denominator to assuage the numerator.

Am I missing something? Is there a mass per time based emissions permitting I'm forgetting about? The existing operating permits still apply?

Lisa WozniakJun 5 2014 02:18 PM

Excellent explanation. So useful. Many thanks (and love that Michigan is the focal point)

Alan RamoJun 5 2014 03:29 PM

If EPA is using an intensity factor, how does this account for growth in demand and capacity? After all, China is reducing its intensity, and their emissions are dramatically increasing.

Michael BerndtsonJun 5 2014 08:49 PM

I'm less confused after skimming the EPA proposal, rereading the post, and pondering a while. My above comment may not be all that pertinent. Brilliant, but wrong.

I guess with MWh in the denominator there isn't much a state can do with coal. Additional electricity would have to come from nuke and renewables - and to some degree natural gas.

Theoretically, a ton of coal generates about 1.8 MWh per ton of coal. A ton of coal theoretically emits 5720 pounds of CO2 upon combustion. (EIA's rules of thumb)

The carbon intensity just in theory would be 3,105 (say 3,000) pounds of CO2 per MWh.

States intensities goals range from about 600 to 2,000 pounds CO2 per net MWh

Let's assume the US average for fuel mix in 2011: coal, 42%; nuke, 19%; natural gas, 25%, renewables, 13%.

Let's assume the following carbon intensity (lbs CO2/Mwh): coal, 3000; nuke, 0; natural gas 1500; renewables, 0. This is a guess for the last three.

Average carbon intensity for the US would be:
3000*0.42 + 0*0.19 + 1500*0.25 + 0*0.13 = 1635 pounds CO2 per MWh

I guess maybe it makes sense. For instance California is required to achieve an intensity of around 600 pounds CO2 per MWh. Cali must have mostly renewables and nukes, no?

It would be easy for a state with 100% coal to achieve the 6 percent cut, as proposed by EPA. Maybe that's why Illinois (ComEd) decreased it nuke generation from about 80 percent to around 40 percent to jack up intensity, before having to reduce it per EPA.

On the other hand, coal is not wrong to say there is a war on them. This is a shot across the bow. Assuming electricity consumption will only increase and coal is set pretty much as is - we're going to have to build a lot of nuke plants and drill a lot of natural gas wells - until wind and solar can pick up the slack.

Steve AplinJun 6 2014 02:22 PM

Amazing to me that we can talk about a "carbon intensity" target of 1161 pounds per MWh for Michigan without somebody pointing out how HIGH that is.

My home province, Ontario Canada, has a grid-level "carbon intensity" (I call it CO2 intensity per kilowatt-hour, or CIPK) of less than 53 grams per kWh.

That's about two ounces per kilowatt-hour -- or about one-tenth that of the TARGET for Michigan.

Shouldn't you be looking at how we do it? Ontario is an industrial jurisdiction like Michigan -- we have lots of auto manufacturing, plus other very electricity intensive industry like nickel mining/smelting and pulp and paper.

If we can do all this with a grid CIPK of around two ounces of CO2 per kilowatt-hour, maybe you should be implementing our strategy. I mean, you are trying to lead the way, right?

Steve LargentJun 6 2014 05:56 PM

Thanks for breaking this down step by step, very helpful!

The one part I'm still struggling with is the 1.5% annual energy efficiency savings. In the EPA's goal calculation, they provide the following % of avoided MWh sales for MI for each year, 2020-2029:

4.6%, 5.7%, 6.8%, 7.8%, 8.6%, 9.4%, 10.1%, 10.8%, 11.3%, 11.8%

Not only does growth slow at the end of the period (as opposed to ramping up), but this is a CAGR of 11% for the annual savings rate over the period. Am I missing something?

Thanks for any help you can provide and thanks again for thourough and logical explanation of the rule.

Dave HawkinsJun 6 2014 06:20 PM

Yours is a quick one to answer: This spreadsheet at EPA's web site provides the annual and cumulative savings figures EPA used for each state:

Dave HawkinsJun 6 2014 06:22 PM

Because EPA's method is a bottom-up analysis for each state based on where it was in 2012, it can result in some pretty unambitious targets. In the public comment period we and others will be submitting information that we think supports more ambitious targets. Thanks for the information about Ontario.

Dave HawkinsJun 6 2014 06:27 PM

Using an intensity factor does not prevent an increase in emissions over time if the state's demand continues to increase. That is why the energy efficiency component of EPA's analysis is so important. And implementing energy efficiency in the real world is much more important. NRDC's proposal for standards under section111(d), issued in 2012 and updated in March 2014, concludes that by applying available energy efficiency measures, US power sector emissions could be reduced by 35% from 2005 levels by 2030.

wdJun 6 2014 09:26 PM

So, it appears the EPA is determined to control CO2 in the atmosphere by setting standards of CO2 emission for US power generating plants. No?

(I'm trying to imagine the dialogue from the 1070's when the legislation was born. A legislator proposes an amendment to control CO2 as well. We hold our breath (appropriately). Does anyone think the legislation would have passed with CO2 so included?)

Anyway, if we model CO2 in the atmosphere for the globe, US generating plants will surely show up as one factor. But, the interest in CO2 is driven by the greenhouse effects. No?

So, where do we consider other gases? Where do we consider my neighbor who habitually leaves on the yard lights? Where do we consider me when I exhale (CO2 in part)? How about sources of CO2 outside of the US? What about manufacturing, commercial, farm, and residential burners of fuels that produce CO2? What about my car that burns gasoline? Perhaps a tree planting contest? Why fixate only on electricity generating plants?

I've already cut my utility usage by 50%. I did it without any standards or regulations. Also, I participate in a voluntary electricity rate program intended to reduce peak loads at the utility.

I don't want to spoil anyone's fun, but I'd have more patience with the EPA's efforts if there were a corresponding scheme on "who pays" where the advocates have a chance to pay more for desirable green energy while others have the option to select plain vanilla if they wish.

I've also lived in Ontario. Mr. Alpin makes some good points. (Plus, he is breathing some of what MI emits.)

Comments are closed for this post.


Switchboard is the staff blog of the Natural Resources Defense Council, the nation’s most effective environmental group. For more about our work, including in-depth policy documents, action alerts and ways you can contribute, visit

Feeds: Dave Hawkins’s blog

Feeds: Stay Plugged In