Archive for April, 2017

The number one goal of the Presidential EPA Transition Team was to restore the integrity of the scientific process.  Over the last two decades the EPA has moved away from using science to determine safe air and water quality standards to using manipulated science to justify regulations that met political goals.  Peer review became pal review by scientists receiving EPA grants, and often by peers affiliated with the researchers they are reviewing.  Key experimental data has been hidden.  President Trump’s recent Executive Order starts the process of restoring integrity by restoring unbiased peer review, and requiring public access for science studies and their underlying data.

The most egregious example of using junk science is the last administrations “War on Coal”, established as a campaign promise.  Unable to pass a national tax on carbon dioxide emissions with his own party’s legislative super majority, President Obama wrote an executive order to the EPA to create a regulation that would force states to adopt a carbon tax.  The EPA found they couldn’t do enough at the power plant level to meet their goal.  So, they used an obscure regulation to take over managing the entire electric grid with the Clean Power Plan.  They also ignored federal rules and calculated $30 billion in benefits when following the calculation rules yielded only $0.5 billion in benefits.  Fortunately the Supreme Court put a stay on implementation, and President Trump has ordered a review of the rule, and it will likely be overturned.  The President also requires cost/benefit calculation rules be followed.

Unable to close coal-fired power plants with existing regulations the EPA set out to create new regulations to do so.  However, not even the EPA can write a regulation without at least a fig leaf of cover for scientific justification, and a positive cost/benefit analysis.  A rule to limit mercury emissions that wound up closing hundreds of older, smaller coal-fired power plants showed only a $6 million dollar benefit compared to $9 billion in implementation cost.  No problem.  The expensive filtration systems needed to remove mercury would also remove fine particle pollution as a by-product.  The EPA merely included the exaggerated health benefits from lowered fine particle pollution to justify the mercury rule.

Those exaggerated health benefits are a good example of bad science.  The Harvard Six Cities Study on mortality from exposure to fine particulate matter is used to support the health benefits.  We have a rigorously determined air quality standard that says you can be exposed to up to 12 micrograms per cubic meter of fine particles all the time without injury, and certainly not death.  The Harvard study claims premature deaths occur from any exposure down to zero ignoring the standard.  No one gets to see the data or review its accuracy.  The Scientific Advisory Board tasked with reviewing the study is packed with EPA grantees and even reviewers from Harvard, and also never saw the supporting data.  Meanwhile human exposure studies, and real world exposure at levels way above the standard don’t confirm the Harvard results.

The EPA has used $380 billion in claimed health benefits from reducing fine particle exposure below the air quality standards to justify almost every major regulation of the last decade.  These include the new ozone standards, the cross state air pollution standards, and the carbon dioxide endangerment finding.  Part of the reason the EPA got away with using the Harvard study is the issue is complex and hard to understand without a lot of study.  If you want a detailed explanation of the problems with this study read the just published book “Scare Pollution” by Steve Milloy.

There are those who are claiming President Trump has started a war on science.  He is waging war, but it is against the use of junk science!

David T. Stevenson, CRI Director 

Member Trump Administration EPA Transition Team

E-Mail: DavidStevenson@CaesarRodney.org

Read Full Post »

Executive Summary

The most likely result of the EPA carbon reduction plan is; no actual emission reduction, a $50 billion a year national carbon tax, and reduced electric grid reliability, possibly leading to much higher electricity prices and blackouts.  This report summarizes fifteen false assumptions in the proposed regulation.  Meanwhile, while global emissions have continued to rise since 2005, US emissions dropped 750 million tons, or 12.6% without regulation, by far the most progress of any country.  Furthermore, electric demand has not been growing, and if that trend continues, we will meet the EPA 2030 CO2 emission reduction goal of 30% by 2018 if we do nothing!

With a regulation of this magnitude, the EPA must show monetized benefits exceed cost.  EPA press releases focus on a $93 billion a year benefit compared to $9 billion in compliance cost.  Changing the assumptions results in essentially zero benefits and $52 billion in compliance cost.

The basic plan is to drastically reduce the amount of coal used to produce electricity at existing power plants.  By 2030 the new Carbon Pollution Guideline will shut down about 11%, or 500 gigawatt-hours, of real electric generation.  The EPA hopes government subsidized energy efficiency projects will lower electric demand by that amount.  Experience has shown such subsidy programs are overwhelmed with “free-riders”, those who would have done a project without a subsidy, and by the “rebound effect”, where lower electric bills lead to higher electric use elsewhere.  Without the efficiency savings, the compliance cost estimate rises to $52 billion a year.

The lost base load power generation from coal fired power plants will lead to power shortages, and more use of expensive peak power, and possibly to rolling blackouts.  There simply won’t be enough generating capacity to meet demand.  We know about the $52 billion in compliance cost.  We don’t know how much peak power or blackouts will cost.

Climate change cost reduction accounts for $31 billion of the $93 billion savings estimate.  The estimate is based on the global cost of climate change, and a 3% discount rate to determine present value.   Meanwhile, the cost of compliance is based on only domestic expenses.  The US Office of Management & Budget, in no uncertain terms, requires cost/benefits analysis to use only domestic cost, and a 7% discount rate.  Applying these two basic guidelines reduces the EPA $31 billion savings estimate to about $0.5 billion.

Promoted as a climate change regulation, most of the estimated benefits come from lower health impacts of a coincident reduction in air pollution ($62 billion).  The EPA relies on flawed mortality rate studies multiplied by an inflated value of life estimate to come up with the $62 billion.  Any one of three assumption changes drops the $62 billion to essentially zero.  We can use better mortality studies, or assume exposures below the scientifically established National Air Quality Standards will not impact health, or use more realistic value of life estimates.

The EPA has concluded they cannot force existing coal fired plants to become more efficient to meet the carbon dioxide reduction goals using existing, reasonably priced technical solutions.  In fact they concede even small changes in efficiency will lower electric cost triggering higher sales from the plants thus offsetting the emission improvements.  A novel regulatory approach is being tried to cause states to do the regulation with a suite of strategies that offset emissions outside the actual power plants.  This approach opens the door to legal challenges.

The EPA claims to be giving the states flexible strategies for coming up with plans to meet the carbon dioxide reduction goals.  The problem is each strategy only works if coal fired electric generation costs more than other options.  The only way to do that is with a national carbon tax.  States will be forced to adopt a carbon tax to have their plans accepted by the EPA.

List of Specific Recommendations

Cost benefits of CO2 emission reduction

1)      Use the UN IPPCC 2013 report release 5 instead of 2007 release 4.  The latest report downplays the connection between global warming and extreme weather events, and has a lower range of temperature and sea level rise.

2)      Use US OMB guidelines which require a 7% discount rate, and to compare domestic cost to domestic benefits (drops benefits from $31 billion a year to $0.2 to $0.7 billion a year).

3)      Do not establish a CO2 emission reduction goal until international negotiations are complete

Cost benefits of health improvements from air pollution reduction

4)      Count only exposure above the scientifically established National Air Quality Standard in calculating reduced premature deaths (drops the health benefits from $62 billion to $3 billion).

5)      Replace the flawed Harvard Six Cities study with the Cao et al study of Xi’an, China for the Chinese Academy of Science.  The Xi’an study sample size is 1000 times larger, with an exposure range 25 times larger, and the study finds a threshold level for health impacts (drops the health benefit from $62 billion to $1.25 billion).

6)       Use US OMB guideline which require a 7% discount rate which (drops health benefits from $62 billion to $57 billion).

7)       Eliminate fine particles of ammonium sulfate from health benefit calculations as they have been shown to be non-toxic.  Three quarters of the health benefit calculation is based on ammonium sulfate reduction (drops health benefits from $62 billion drops to $15.5 billion).

8)      Use lost earnings potential instead of willing to pay surveys to determine the value of one year of added life.  The willingness to pay survey is 100 times the rate of the lost earnings calculation (drops health benefits from $62 billion becomes $0.6 billion).

9)      The cumulative impact of these changes drops the $62 billion benefit to zero.

Compliance Cost

10)   Use US OMB guidelines which require a 7% discount rate (raises the 2030 compliance cost from $43 billion to $52 billion, and the net cost from $9 billion to $18 billion).

11)   Add free-rider and rebound effects to the $34 billion energy efficiency cost savings (drops cost savings from $34 billion to zero, and raises the net compliance cost to $52 billion).

12)   Assume total demand for coal will not drop because of export demand so electricity prices will not fall as the EPA predicts (raises compliance cost$1 billion, and a net cost to $53 billion)

Carbon reduction goals

13)   Use 2012 as the base year for all calculations

14)   Develop a specific CO2 emission goal for each state instead of a heat rate goal

15)  Include more new generation from nuclear power plants

16)  Use the US EIA “Low Electric Demand Case” for forecasting future electric demand


There are two main EPA reports covering the proposed guideline; “Carbon Pollution Emission Guidelines for Existing Stationary Sources: Electric Generating Units” (CPG), and the accompanying “Regulatory Impact Analysis” (RIA).  Page numbers will be referenced for both.  There are several options available to meet the emission reduction goal.  In press releases, the EPA quotes monetized benefits and costs of their plan.  The published numbers relate to Option 1 for individual states so our analysis will be based on the same Option 1 unless otherwise stated.  In a press release1 the EPA claims the new guidelines will reduce CO2 emissions by 30% by 2030 along with a 25% reduction in other air pollutants, and will result in $55 to $93 billion a year in benefits compared to only $9 billion a year in compliance cost.  The regulation was initiated by a Presidential Memorandum2 which was issued June 25, 2013, by executive order.

Key EPA assumptions are reviewed here to check for consistency within the Guideline proposal, for compliance with practices for conducting a cost/benefit analysis published by the United States Office of Management & Budget Circular A 94, and for sound science compared to alternative studies.  Information sources are listed in footnotes at the bottom of each page.

There are six sections:

1)      Review of key assumptions for calculation of cost benefits of reduced CO2 emissions, page 6

2)      Review of key assumptions used to calculate cost benefits of health impacts, page 9

3)      Review of key assumptions used to calculate compliance cost, page 14

4)      Review of key assumptions used to calculate carbon reduction goals, page 17

5)      Review of the Regional Greenhouse Gas Initiative, page 22

6)      Review of US EIA “Low Electric Demand” Case, page 26


1)       http://www2.epa.gov/sites/production/files/2014-05/documents/20140602fs-overview.pdf

2)       Presidential Memorandum – Power Sector Carbon Pollution Standards, June 25, 2013. http://www.whitehouse.gov/the-press-office/2013/06/25/presidential-memorandum-power-sector-carbon-pollution-standards  

Section 1: Review of key assumptions for calculation of cost benefits of reduced CO2 emissions

The EPA claims benefits of $31 billion3 a year by 2030 from CO2 emission reduction goals.  When proper assumptions are used the benefits disappear.  First, understand these are guidelines, not strict standards as is usually the case with EPA rules, and there are no penalties established for missing the goals4.  States are given flexibility in meeting the “goals”, and comment is welcomed by the EPA on changing the goals over the next year.  In fact, the EPA admits the benefits and costs cannot be calculated until the state plans are in effect sometime after 2017, and that the estimates they supply are merely “illustrative” of potential savings5.

The states will eventually calculate an actual carbon reduction goal in metric tons.  The EPA uses a range of Social Cost per ton of Carbon estimates calculated elsewhere6.  The cost/benefit estimates multiply the emission reduction tonnage by the cost/ton estimate to calculate the net cost savings.  The EPA lists many potential costs of climate change7:

  • More heat waves from increased average temperatures leading to more deaths
  • Increased ozone induced illness and death (ozone is produced on hot summer days)
  • More extreme weather from higher average temperatures adding to storm power
  • Increased coastal storms and storm surges due to rising sea levels as higher temperatures expand the ocean, and from increased glacial melt

There are numerous problems with the calculations of the Social Cost of Carbon.  The basis for the calculations is the United Nations Climate Change report.  The Interagency Working Group on the Social Cost of Carbon (IWGSCC) calculated the cost based on the 2007 version.  The 2013 version of the UN Climate Change report reduces the expected temperature rise caused by manmade greenhouse gases, such as CO2.  In turn that reduces all the potential impacts, particularly sea level rise.  The latest UN report also downplays the link between rising temperatures and extreme weather events and should be the basis of the proposed guideline.


3)       CPG page 57 (RIA Table 4-5, page 4-12)

4)       CPG pages 417-18

5)       RIA page ES-4

6)       RIA page 4-7 Interagency Working Group on the Social Cost of Carbon.

7)       CPG pages 62-63

Cost and benefit calculations go out many decades.  For comparison purposes, cost projections of this nature always are stated in terms of the net present value of future costs or benefits.  The biggest assumption is what discount rate to use.  For example, the EPA chose a 3% discount rate to come up with the $31 billion a year in savings from carbon reduction.  Using a 5% discount rate the $31 billion drops to $9.3 billion8.  The US Office of Management & Budget recommends a 7% discount rate9.  The EPA also uses a 7% discount rate for establishing a range of health benefits in this standard10.  Using a rough estimate a 7% discount rate would drop the carbon reduction benefit to about $3 billion, or about one tenth the benefit reported by the EPA!

The EPA also chose to report the global benefit of carbon reduction but the cost of compliance is based on domestic cost only.  Again, we go to guidance from the US Office of Management & Budget11 and the recommendation is to use only the domestic cost.   The 2013 IWGSCC report ignored the OMB, but the 2010 IWGSCC report estimated the domestic cost would be 7% to 23% of the global cost12.  Combining the 7% discount rate, and the domestic cost, lowers the 2030 benefits of carbon reduction to between $0.2 and $0.7 billion.

The IWGSCC report also admits it ignores “carbon fertilization” in calculating the Social Cost of Carbon.  Carbon dioxide is plant food, plants absorb it to grow.  The Heartland Institute summarized thousands of controlled experiments that consistently show the positive relationship between higher CO2 levels and increased plant growth13.  It is reasonable to assume the combined positive effects of carbon fertilization, a 7% discount rate, and using just the domestic impact of climate change, would eliminate the EPA’s $31 billion in reported benefits from carbon reduction.


8)       RIA Table 4-5, page 4-12

9)       US Office of Management & Budget Circular A-94 Revised  http://www.whitehouse.gov/omb/circulars_a094/ Base-Case Analysis. Constant-dollar benefit-cost analyses of proposed investments and regulations should report net present value and other outcomes determined using a real discount rate of 7 percent. This rate approximates the marginal pretax rate of return on an average investment in the private sector in recent years. Significant changes in this rate will be reflected in future updates of this Circular.Notes:

10)   RIA page 4-22

11)   International Effects. Analyses should focus on benefits and costs accruing to the citizens of the United States in determining net present value. Where programs or projects have effects outside the United States, these effects should be reported separately.US Office of Management & Budget Circular A-94 Revised http://www.whitehouse.gov/omb/circulars_a094/

12)   Interagency Working Group on the Social Cost of Carbon 2010, page 11, http://www.epa.gov/oms/climate/regulations/scc-tsd.pdf “with a 2.5 or 3 percent discount rate, the U.S. benefit is about 7-10 percent of the global benefit, on average, across the scenarios analyzed. Alternatively, if the fraction of GDP lost due to climate change is assumed to be similar across countries, the domestic benefit would be proportional to the U.S. share of global GDP, which is currently about 23 percent”

13)   Heartland Institute, “Climate Change Reconsidered II: Biological Impact”, http://heartland.org/media-library/pdfs/CCR-IIb/Full-Report.pdf

            Using the EPA’s selected 2005 base year, the United States is reducing CO2 emissions at a much better rate than most other major emission countries (Table 1), and is number one in both tons of reduction and percent reduction.  We are involved in CO2 reduction negotiations globally, and it is a questionable practice to take unilateral action which takes away our negotiating leverage.  No target should be selected until the negotiations are .

Table 1: GHG Emission by Country by Year in billions of tons

EPA comments table 1

Section 2: Review of key assumptions used to calculate cost benefits of health impacts

The argument has been made air pollution from burning fossil fuels to make electricity results in large increases to health care costs and lost work hours that are not reflected in electric prices.  The EPA claims benefits of $62 billion14 a year by 2030 from the health impacts of emission reduction goals.  The scientifically established National Air Quality Standards are set to a level that will not cause harm in even sensitive groups, such as asthmatics.  $59 billion of the $62 billion, or 95%, in estimated annualized health cost savings is based on the misguided assumption exposure to fine particle pollution (PM2.5) levels below the Air Quality Standard is harmful15.  Exposure levels dropped 33% from 2000 to 2012, and are solidly below Air Quality Standards (Chart 1)16, and that is true for every region of the country.

Chart 1

EPA comments chart 1

Fine particles below 2.5 micrograms (µg) are a wide range of particles including soot, and dust from forest fires, volcanoes, wind-blown dust, mold, and fossil fuel combustion.  The National Air Quality Standard limits exposure to 12 micrograms per cubic meter (µg/m3).  To put that in perspective, this is equivalent to putting soot particles weighing about 5000 times less than a grain of salt in a refrigerator.


14)   RIA figure 4-15, page 4-33

15)   RIA Figure 4-4, page 4-45

16)   US EPA Particulate levels, http://www.epa.gov/airtrends/pm.html

The EPA uses epidemiologic population studies comparing death rates over a period of time to determine the health impacts of PM2.5 especially mortality rates (98% of monetized cost).  The studies used17 are the 2012 Harvard Six City Study by Lepeule et al. and the 2009 American Cancer Society Study by Krewski et al.  The studies concluded a 14% (Six City), and 7% (ACS) increase in death rates would occur linearly for each 10 µg/m3 of added PM2.5 exposure18.  The EPA ignores anomalies in both studies that show a relationship between pollution and pre-mature deaths for people with an education of high school or less but not for people with education above high school19.  The Six Cities study finds a 65% higher relationship for men then for women20.  Clearly, there are other confounding factors at work that need explanation.

The health impacts from these studies are the basis of the EPA’s calculations of the benefits of lower pollution levels.  Note the EPA assumes there is no lower threshold for health impacts.  This is not a common assumption in toxicology where toxicity is always related to dose.  For example, aspirin and digitalis are key heart medications but a large enough dose of either is fatal.

A 2012 study21 of Xi’an, China, for the Chinese National Academy of Science, by Cao et al., found different results (Chart 1); a 0.27% increase in death rates for each 10 µg/m3, and a lower threshold for pre-mature deaths of 25 µg/m3.  The Xi’an study used the same methodology as the Six Cities and ACS studies.

Chart 2 PM 2.5 and Mortality in Xi’an, China

EPA comments chart 2


17)   RIA page 4-19

18)   Health Effects Institute, “Reanalysis of the Harvard Six Cities Study and the American Cancer Society Study of Particulate Air Pollution and Mortality”, page 1, http://pubs.healtheffects.org/getfile.php?u=273

19)   IBID, page 19

20)   IBID, page 16

21)   Environmental Health Perspectives, “Fine Particulate Matter Constituents and Cardiopulmonary Mortality in a Heavily Polluted Chinese City” http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3295342/

The Xi’an mortality rate is just 2% the Six Cities study used in the EPA $62 billion health impact claim!  The key differences are the Xi’an study covered 1000 times more people, and is the second most polluted city on earth where exposure to PM2.5 ranged up to 25 times higher than the Six Cities study (Table 1).  Note the pattern, the higher the sample size, and exposure range, the lower the mortality rate.  Health benefits would drop to $27 billion using the ACS study, and $1 billion using the Xi’an study.

Table 2:  Mortality Rates, Sample Size, and Exposure Range for Three Studies

EPA comments table 2

Fine particles are made up of many different kinds of particles from many different sources such as forest fires, volcanoes, wind-blown dust, mold, and fossil fuel combustion.  The EPA “assumes all particles, regardless of their chemical composition, are equally potent in causing premature mortality”22.  The EPA then warns, “the scientific evidence is not yet sufficient to allow differentiation of effect estimates by particle type”.  Most of the fine particulate matter from coal fired power plants comes indirectly from sulfur dioxide emissions forming particulate ammonium sulfate.  Studies show no link between ammonium sulfate and health issues.  In fact, the compound has been used as an inert control material in studies of health impacts of aerosols23.   Three quarters of the $62 billion EPA health benefits calculation comes from planned reductions in sulfur dioxide24, so the benefits drop to $15.5 billion.

The EPA does not only use population studies.  They have also conducted laboratory testing on individuals by exposing them to high levels of PM2.5, at over five times the EPA standard for 24 hour exposure.  The sole basis for an EPA claim of the danger of short term exposure is a case study25 based on a single individual who went into atrial fibrillation 23 minutes into exposure.  The subject was back to normal 2 hours later.  The 58 year old volunteer had a history of heart problems, had a body mass index of 34.9 with a 45” waste, and her father died of heart disease at age 57.


22)   RIA page 4-41

23)   Joel M. Schwartz and Seven E. Hayward, Air Quality in America: A Dose of Reality in Air Pollution Levels, Trends, and Health Risks (Washington, D.C.: American Enterprise Press), 2007, p. 150-151

24)   RIA Table 4-15, page 4-33

25)   Case report: “Supraventricular Arrhythmia Following Exposure to Concentrated Ambient Air Pollution Particles ‘, Evironmental Health Perspectives http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3279446/

The EPA failed to disclose the lack of reaction of three hour exposure involving forty other subjects.  Based on this single case study, EPA Administrator Lisa P. Jackson testified about PM2.5 before Congress in September, 2011: “Particulate matter causes premature death. It doesn’t make you sick. It is directly causal to you dying sooner than you should.” Mrs. Jackson also testified that PM2.5 kills about 570,000 Americans annually, about 25 percent of all U.S. deaths (Washington Times 1/22/13).  The highest estimate of added premature deaths from air pollution is 75,000/year26 from the Six Cities study, and could be as low as 1,500 based on the Xi’an study.

The model output suggests a 29%27reduction in PM2.5 emissions will save up to 6200 premature deaths in 203028.  Chronic Lower Respiratory disease is the third largest killer in the US29.  Chronic bronchitis, and emphysema (together known as Chronic Obstructive Pulmonary Disease or COPD) account for 93% of the Chronic Respiratory disease deaths.

It should be noted the primary cause of COPD (90%)30 is long term damage from smoking.  According to the COPD Foundation an estimated 30% of diagnosed patients continue to smoke.  So it is reasonable to say most of the premature deaths were a direct result of smoking.  Would differential rates in smoking after diagnoses explain the different outcomes based on education, and sex in the Six Cities and ASC studies?

The estimated number of expected premature deaths is multiplied by $10.1 million/premature death31.  This is based on “Willingness to Pay” surveys.  These surveys ask people what amount of money they would be willing to spend for one additional year of life.  Clearly, this is an exaggerated measure.  For example the American Lung Association estimates the total cost of all COPD premature deaths at $12.4 billion32 but the cost was spread over 134,000 deaths for an average cost of about $93,000/premature death.  Compares this to the $62 billion EPA estimate for 6200 premature deaths. The cost covers lost potential earnings with a 3% discount factor, and is reasonable as most of the premature COPD deaths (85%) occurred in people over 65 years old.  Benefits drop from $62 billion to $0.6 billion.


26)   Harvard School of Public Health, “Harvard Six Cities Study Follow up”, page 1, http://archive.sph.harvard.edu/press-releases/2006-releases/press03152006.html

27)   RIA Table 3-7, page 3-21

28)   RIA Table 4-18, page 4-36

29)   Center for Disease Control, Vital Statistics

30)   Center for Disease Control Fact Sheet, http://www.cdc.gov/tobacco/basic_information/health_effects/respiratory/index.htm

31)   RIA page 4-22

32)   “Trends in COPD”, http://www.lung.org/finding-cures/our-research/trend-reports/copd-trend-report.pdf

            The health benefits calculations suffer from the same discount factor problem as the climate benefits.  The $62 billion in health savings is based on a 3% discount rate.  As stated earlier, the US Office of Management & Budget recommends using a 7% discount rate on cost/ benefit studies when the time horizon is more than six years.   The health benefits drop to $57 billion33 using the 7% discount rate for the Six Cities study, to $24 billion for the ACS study, and to $1 billion for the Xi’an study.

It is clear the health benefits estimates suffer from a number of poor assumptions.  The Six Cities study needs to be dropped.  It has the smallest sample by far, has the smallest range of exposure to fine particle pollution, and has several internal anomalies suggesting the study has missed confounding issues.   The assumption there is no lower limit of toxicity for fine particle pollution flies in the face of the science of toxicology, the Xi’an study results, and common sense.   Ammonium sulfate clearly poses no danger and needs to be subtracted from the exposure tonnage estimates.  The EPA should follow the OMB guidelines for cost/benefit analysis and should use a 7% discount rate.  Finally, power plant emissions declined 63%34 between 2005 and 2012.  All calculations should be based on a 2012 base year.

The requested changes are all reasonable, and defensible.  If adopted, they would produce much more realistic health benefit estimates.  Unfortunately, for the EPA’s case, the health benefits disappear.


33)   RIA Table 4-15, page 4-33

34)   US Energy Information Agency, “Power Plant Emissions Continue to Decline”, http://www.eia.gov/todayinenergy/detail.cfm?id=10151

Section 3: Review of key assumptions used to calculate compliance cost

The EPA calculated the compliance cost for meeting the guidelines to be $8.8 billion a year by 203035.  This adds the change in coal plant capital expenditures for new generation and heat rate improvements, the ongoing cost of operating pollution control equipment, and the cost of energy efficiency improvements.  The calculation subtracts the lower coal prices and operating costs from expected lower electricity demand.  From a practical perspective, the EPA is claiming the only cost as the cost for energy efficiency programs ($42.7 billion36 in 2030, at 3% discount rate), and subtracts a projected lower cost for operating generating units ($34 billion37) to arrive at a net cost of $8.8 billion.  We note again, a 7% discount rate should be used which raises the cost for the energy efficiency programs to $51.8 billion, and net cost to $17.8 billion.  However, as described below, the assumption electricity demand will fall and lower operating cost is probably incorrect, and the total compliance cost estimate should total about $52 billion.

The EPA uses a levelized cost of saved energy of $90/megawatt-hour38 times the net expected megawatt-hours saved from energy efficiency investment to calculate expected savings.   The saved energy is adjusted for several factors discussed below.

The EPA discusses a concern about the “rebound effect”39 in regards to forcing the improved efficiency of coal fired generators.  Increased efficiency results in lower operating costs, and lower prices.  Lower prices increase sales potentially wiping out the CO2 reduction gains from the efficiency projects.  The same thing happens when a customer sees a lower electric bill from putting in a more efficient heating unit and raises the thermostat to be warmer.  The EPA also discusses the need to adjust for “free-ridership”40 which occurs when someone applies an energy efficiency grant, or tax credit, against a project they would have done anyway.


35)   RIA Table 3-8, page 3-22, Option 1, State

36)   RIA Table 3-4, page 3-19

37)   RIA Table 3-9, page 3-23, Option 1, State

38)   CPG page 226

39)   CPG page 157

40)   CPG page 226

The assumptions of how much energy efficiency savings are lost to rebound and free-ridership are not provided.  So how often do these adverse effects occur?  A 2013 University of Maryland study41 on energy efficiency grants of up to $3000 for heating and air conditioning units showed free-ridership at up to 89%, and no energy was saved because of the rebound effect.  Meanwhile, a control group of homeowners who improved their heating and air conditioning units without grants saved 16% on energy used.

Similarly, popular subsidies for energy efficient lighting, and refrigerators would have mostly free-riders as federal regulations have pretty much wiped out inefficient alternatives.  Also, for example, the rebound effect occurs when an old refrigerator is moved to the garage.

The EPA also discusses the lack of success of the nine state carbon tax, Regional Greenhouse Gas Initiative (RGGI).  RGGI runs a quarterly auction where permits are sold to owners of electric generating units to allow CO2 emissions.  The revenues are supposed to be used to fund energy efficiency projects to reduce CO2 emissions.  The cost of buying the permits is passed on in higher electric bills.

The nine states did see a 40% reduction in CO2 emissions, but the EPA notes, “RGGI was not the primary driver for these reductions”42.   Data for electric generation, by fuel, by state, is available from the US Energy Information Agency.  A detailed analysis by CRI showed 68% of the drop in emissions was from fuel switching to low carbon natural gas and no carbon nuclear, hydro, and wind, and 32% was from closing coal fired power plants.  The fuel switching was mainly driven by market forces, and the coal plant shut downs were driven by EPA air pollution regulations.  The lower regional generation was made up by imports from non RGGI states.  RGGI had no impact on the reduction.  There was no discernible impact from energy efficiency after six years, and $1.8 billion of RGGI revenue.

The US Energy Information Agency tracks CO2 emissions over time divided by inflation adjusted Gross Domestic Product to measure Carbon Intensity.  Since the early 1970’s Carbon Intensity has improved about 2% each year, and is expected to do so through at least 204043.  Individuals, and businesses, continually make cost saving investments in energy efficiency improvements in their own best interest.  Regulations such as fuel efficiency standards for motor vehicles, and appliances are also captured in this trend.


41)   “Free Riding, Up sizing, and Energy Efficiency Incentives in Maryland Homes” , Anna Alberini et al, University of Maryland, College Park, August 11, 2013, http://www.feem.it/userfiles/attach/20139301611194NDL2013-082.pdf

42)   CPG page 98

43)   US EIA, Annual Energy Outlook 2014, Figure MT-7 on page MT-4, http://www.eia.gov/forecasts/aeo/pdf/0383(2014).pdf

Over the last decade a large number of government subsidized energy efficiency programs have been tried to accelerate investment in energy efficiency44.  However, the slope of the energy intensity curve has not changed suggesting the subsidies are having little impact.  The EPA is suggesting new massive investment in subsidized energy efficiency programs with the intent to add another 1.5% a year45 efficiency gain on top of the historical 2% a year gain.

All the real world evidence suggests energy efficiency subsidy programs do not result in significantly reduced energy usage, and so will not decrease CO2 emissions.  We can reasonably conclude the $34 billion in savings projected by the EPA will not be realized.  Using the program cost basis with a 7% discount rate, the real cost of compliance with the Carbon Pollution Guideline will be $52 billion, not $9 billion as the EPA has advertised.  A key assumption limiting the compliance cost is the EPA estimate (based on US Energy Information Agency estimates that did not consider exports) the cost of electricity will fall about 3% in 2030 compared to the reference case46, and that is mostly based on an 18% drop in coal prices47 caused by lower demand.  With coal demand growing worldwide it is unlikely coal prices will drop.  If not, the difference in price of 0.4 cents/kilowatt-hour times 4,051 gigawatt-hours of expected demand48 will add $1.6 billion a year to the compliance cost.


44)   CPG page 107-108, 225

45)   CPG page 225

46)   RIA Table 3-23, page 3-42

47)   RIA Table 3-18, page 3-38

48)   RIA Table 3-11 page 3-27

Section 4: Review of key assumptions used to calculate carbon reduction goals

States must submit a plan to meet the CO2 reduction plan.   If a plan is not submitted, or approved, the EPA will supply one.  A state may submit a plan to just lower emissions at specific EGUs, or may chose to offset emissions with an energy efficiency plan, a carbon tax, or a renewable energy plan.  In reality, the most likely result will be no reduction in CO2 emissions!  We may see significantly higher energy prices, and lower electric grid reliability, possibly leading to black outs.  More nuclear power is not seriously considered by the EPA.

There is considerable confusion of what the goals are.  The 30% reduction goal is from a 2005 base year.  The emission rate was 2,434 million metric tons in 2005, and, in the most optimistic case, will be reduced to 1,701 million tons by 203049.  But, by 2012, EGU’s had already reduced emissions to 2,023 metric tons50, or 17% by switching from coal to natural gas for lower cost, and by closing older, smaller coal fired EGU’s that were not worth upgrading to meet other EPA air pollution reduction standards51.

The actual state specific goals are calculated from a 2012 base year compared to a base case52 of future emissions.  The base case assumes no carbon reduction goal is set and calculates CO2 emissions will grow to 2,256 metric tons by 2030.  So, compared to the base case, CO2 emissions will only fall 25% (from 2,256 tons to 1701 tons), according to the EPA.  The EPA should drop the reference to 30% reduction by 2030 from 2005, and use a 25% reduction by 2030 from 2012 to end this confusion.

Compounding the problem further is the EPA goals are actually going to be measured based on the “Adjusted Output Weighted Average Pounds of CO2 Per Net Megawatt-hour From All Affected Fossil Fuel Fired EGU’s in the State”, or Heat Rate, and is specific for each state53.  An actual goal will not be known until the state plans are accepted sometime around 2017, and will be calculated by dividing fossil fuel power plant emissions by the total power generation from all sources to come up with an average heat rate from all sources including the avoided generation resulting from energy efficiency projects54.  A simple CO2 emission target for each state would be a lot simpler.


49)   RIA table 3-6 page 3-20

50)   CPG Table 4 page 71

51)   Mercury & Air Toxics Standard requires coal fired EGU’s to meet lower levels of emissions for mercury with expensive filtration equipment

52)   RIA table 3-5 page 3-20

53)   CPG Table 8 page 346

54)   CPG page 355, Expressed as a formula, the equation for the annual rate computation is:

[(Coal gen. x Coal emission rate) + (OG gen. x OG emission rate) + (NGCC gen. x NGCC emission rate) + “Other” emissions] / [Coal gen. + OG gen. + NGCC gen. + “Other” gen. + Nuclear gen. + RE gen. + EE gen.]

The EPA recognizes it cannot meet the CO2 emission reduction target by simply relying on regulations targeted at coal fired power plants.  The EPA is required to do a cost/benefit analysis for any regulation with over $100 million in compliance cost.  In this guideline efficiency savings of about 6%55 were identified for existing power plants that could be achieved at a cost of $6 to $12/metric ton of CO2.  In contrast, conversion of coal fired steam generators to burn natural gas is estimated to cost56 $83 to $150/metric ton of CO2, and was considered too high in this guideline for consideration.  The EPA expects that the costs of integrating a retrofit Carbon Capture System into an existing facility would be substantial and will not require it.

As discussed above, the increased power plant efficiency is expected to lower cost, and price, which, in turn, would lead to higher volume.  The added sales would wipe out any CO2 reductions from the efficiency improvements.  How can production be switched from coal to other fuels?  The obvious answer is to increase the price of coal fired electricity relative to other fuels.  That is most easily accomplished by a carbon tax that impacts coal more than natural gas, or other low carbon options.  A state wanting its plan adopted by the EPA would find a carbon tax plan to be the easiest path to acceptance.

Since the EPA cannot meet emission goals “inside the fence”, they are offering states a range of options “outside the fence”.  These strategies include creating a carbon tax and trade auction with the auction revenue funding energy efficiency projects, requiring the use of more wind and solar power, maintaining the nation’s nuclear fleet, and shifting the dispatch priority to use more power fueled by natural gas.

Regional grid operators determine power plant dispatch priorities based on bid prices from each power plant.  The lowest priced plants produce the power that flows into the grid.  Typically coal has had the lower price compared to natural gas fired plants.  However, natural gas produces electricity with about 40% less CO2 emissions.   The EPA wants state plans to consider ways to shift dispatch away from coal steam to natural gas combined cycle (NGCC) power plants.  Capacity Factor is a measure of the percent of time a power plant runs compared to the time it is available to run.  The EPA goal is to reduce the Capacity Factor of coal plants to 73% by 2030 from a base case of 79%57.  NGCC power plants would see Capacity Factor grow from a base case of 42% to 51%.  Below is information on current Capacity Factors and trends (Chart 3).


55)   CPG page 143, 405

56)   CPG page 146, 247

57)   RIA Table 3-10, page 3-25

Chart 3: Capacity Factors for coal steam & natural gas combined cycle power plants 2008-2013

EPA comments chart 3

Source: US Energy Information Agency, Electric Power Monthly, May, 2014 Table 6.7.A

Chart 3 shows coal steam plants Capacity Factors have already dropped from 73% in 2008 to 60% in 2013, a drop of two percentage points a year.  The Capacity Factor could go up somewhat as some coal fired plants shutdown because of the EPA Mercury & Air Toxics Standard.  The Capacity Factor of NGCC plants increased from 40% to almost 50%, an increase of 1.7 percentage points a year, almost replacing lower generation from coal.  It would seem the 2030 goal has already been met.   But let’s assume the EPA case is correct.  How do we push the market to dispatch more natural gas?  Again, a carbon tax will raise the price of coal fired plants relative to natural gas, and is likely to be approved by the EPA.

Another consideration is availability of natural gas.  By law, if there is a supply problem, natural gas flows first for heating systems.  If there is a shortage NGCC power plants come second.  Also, to obtain the lowest prices, NGCC plants sign contracts for interruptible service for natural gas.  Because of this the winter months see the lowest Capacity Factors for NGCC units.  In fact, the Capacity Factor dropped to 38.5% (from the 50% average) in March 201458.  There is no reason to believe this situation will change by 2030 regardless of electric prices.  In fact, the EPA’s favored case shows generation capacity for NGCC plants will drop 4% compared to the base case59.   The Carbon Reduction Guideline doesn’t just affect coal.


58)  US Energy Information Agency, Electric Power Monthly, May, 2014 Table 6.7.A http://www.eia.gov/electricity/monthly/epm_table_grapher.cfm?t=epmt_6_07_a

59)  RIA Table 3-11, page 3-27

The EPA suggests completion of six new nuclear power plants in various stages of construction, and six nuclear plants scheduled for shutdown be kept open.  No nuclear power beyond that level is suggested.  This is surprising since nuclear power emits no CO2.  They are expected to be competitively priced by 2030, and are a reliable base load power source.   Also surprising is the EPA’s faith they can reverse the six planned closings.  One is in virulently anti-nuclear Vermont.  The EPA gave Vermont a waiver from submitting a CO2 reduction plan as there are no fossil fuel fired power plants in the state.  Vermont will have to import power to make up the loss, and some could come from coal.  So, the EPA has zero leverage to reverse the closure plan in Vermont, and most likely little leverage elsewhere.  Any realistic plan for carbon reduction needs to include more nuclear power.

The EPA also encourages more renewable power such as wind and solar.  However, the estimated generation growth from these sources is insignificant by 2030.

The core of the EPA plan is to replace about 500 gigawatts of real coal fired generation capacity with wishful thinking energy efficiency programs58.  Energy efficiency demand response reductions are often called “negawatts” as they are theorized to eliminate the need for real power generation.  The Federal Energy regulatory Commission issued Order 745 requiring grid operators to compensate electric customers to use less power during peak demand periods at the same rate as actual power generators.  Negawatt availability grew rapidly.  The Order was vacated by the US Court of Appeals, DC, as it violated the Federal Powers Act and intruded in retail market regulation reserved for the states60.  Basically, the court ruled demand response programs are not equal to real power generation with real operating expenses.

With much lower demand response payments how do we pay for the EPA’s ambitious goal?  Again the answer is a carbon tax.  The RGGI program has set cost caps for each permit sold at auction to release a ton of CO2.  Auction prices have been close to the cost cap price.  The cost cap price rises to $10/ton in 2017, and thereafter increases 2.5%/year.  Therefore, the price should be about $14/ton in 2030.  The EPA predicts emissions of 1,701 million tons in 203061, so a carbon tax should raise about $24 billion for energy efficiency projects by 2030, compared to $52 billion in forecasted cost of projects needed to reach the EPA goal.


60)   Energy Business Law, “Divided Court of Appeals Panel Vacates FERC Order 745 on Compensation of Demand Response”, http://www.energybusinesslaw.com/2014/05/articles/ferc/divided-court-of-appeals-panel-vacates-ferc-order-745-on-compensation-of-demand-response/

61)   RIA Table 3-5, page 3-20

In reality, the EPA plan only works if there is a national carbon tax, exactly the President’s goal in issuing his Executive Memorandum to begin this rule making process when he couldn’t get Congress to pass a carbon tax.

Discussed at length in Section 3 of this report, energy efficiency programs suffer from both free-rider, and rebound effects that can overwhelm the expected savings.  The EPA acknowledges it has not accounted for a “system wide rebound effect”62 in its analysis.  What happens if we get to 2030 and the coal plants are shut down but energy efficiency isn’t supplying the expected demand reduction?  The result will be unpredictably high prices, and possibly blackouts with catastrophic impacts on businesses, and individuals, especially the poor!


62)   RIA footnote 49, page 3-18

Section 5: Review of the Regional Greenhouse Gas Initiative63

Since the most likely path most states will take to comply with the EPA Carbon Reduction Guideline is to institute a carbon tax, this section describes, in detail, the RGGI program used in nine northeast states.  The basic things to know are it adds cost while not reducing CO2 emissions, and it will result in a minimum carbon tax of $14/ton by 2030 to possibly $28/ton.  That works out to $24 to $48 billion a year at 1,701 million tons in 2030, or roughly $6 to $12/megawatt-hour average across the United States.  A typical residential customer uses about a 10 to 12 megawatt-hours a year so would pay $60 to $144 more a year.  A large industrial customer might pay $1.7 to $3.4 million more a year.

The regional cap and trade program began in 2007 with ten states agreeing to reduce carbon dioxide emissions from power plants by 10% by 2018.  Power plants in these states would need to buy permits in quarterly auctions for each ton of emissions.  Each state had an allotment of permits roughly equal to their average emissions between 2002 and 2006.  The full permit allotment was to be auctioned through 2014 followed by a cut back of 2.5% a year through 2018.  The cost of the permits is passed on to electric distributors who pass the cost on in electric bills.  The auctions are run by RGGI, Inc. which receives 10% of the proceeds.

Auctions began in September 2008 and prices ranged from $3.00 to $3.50/ton for the first year.  Prices collapsed to the $1.86/ton reserve price starting in September, 2009 as emissions had dropped by 34% in 2009 and permit supply was abundant.  The recession played a minor role but cheap natural gas led to fuel switching from coal, and federal and state regulations requiring lower sulpher dioxide and mercury emissions led to the closing of older, smaller coal fired power plants.  New Jersey pulled out of RGGI in 2011 putting further downward pressure on prices as speculators left the market.  With the effects of the recession behind us, CO2 emissions fell another 4% by 2012.  We estimate the 40% reduction in emissions by 2012 was caused 68% by fuel switching and 32% by coal plant closings with essentially none of that impacted by RGGI itself.

We can compare changes in electric generation by fuel in the nine state RGGI group to the rest of the country (Table 2).  In both cases high carbon content coal and petroleum fired generation declined replaced by low carbon natural gas and wind power.  A major difference is the nine RGGI states lost 9% of its generation capacity while the rest of the country grew by 3%.


63)   Unless otherwise stated all information is available on the RGGI.org website

Table 3: Major Fuel Mix Changes 2004 to 2012 in million MWh

EPA comments table 3

The RGGI states now import 8% of their power at a price premium instead of being balanced with generation and use (Table 3).  Generators in the RGGI states must add the cost of carbon allowances when bidding into the electricity market and are at a competitive disadvantage making it less likely new fossil fuel power plants will be built in the RGGI states.  States without sufficient generation capacity to meet their needs plus a reserve capacity pay a premium price for power through annual Capacity Auctions, and through hourly and day forward bidding practices.

Table 4: Electric Generation vs. Use by state 2010 MWh x1000

EPA comments table 4

The US Energy Information Agency64 reported the northeast is losing fuel diversity by depending so much on natural gas (44% of electricity supply in 2012).  The northeast has also expanded the direct use of natural gas in homes and businesses for heat, hot water, and industrial use.  There is insufficient natural gas transmission pipeline infrastructure to serve the rapidly expanded use of gas and there were shortages during the winter of 2014.


64)   US EIA, “Northeast Grows Increasingly Reliant on Natural Gas for Power Production” Nov 12, 2013, http://www.eia.gov/todayinenergy/detail.cfm?id=13751

RGGI, Inc. announced a major change to the auction plan in February, 2013, to be effective January 1, 2014.  The 10% reduction goal would be increased to 45% with an additional 10% reduction by 2020.  There is a Cost Containment Reserve of allowances to be released if auction prices exceed a cap equaling $4 in 2014, $6 in 2015, $8 in 2016, $10 in 2017, and then rising 2.5% a year thereafter.  The primary reason for the rule change was to raise more revenue.  The change will increase revenue from $168 million in 2012 to $1,032 million by 2017.  Revenue changes by state are listed in Table 4 with 2013 showing increased revenue from the planned rule changes, and 2017 assuming all allowances go at the Cost Cap Reserve Price Cap.

Table 5: RGGI Revenue Forecast by State by Year $ Millions

EPA comments table 5

The first quarterly auction under the new rules was held March 5, 2014, with a settlement price at the CCR trigger price of $4/ton.  The entire annual CCR was released as demand was three and a half times higher than the offered allowances as speculators entered the market in hopes of buying at $4 and selling at a higher price later.  Electric generators were only able to buy 45% of the offered allowances so will need to buy additional allowances from speculators at a higher price on the secondary market.  Secondary market prices have tended to move in step with auction prices. So, for example, a speculator who purchased allowances for $4/ton in March might sell them for as much as $6/ton within a year.  In contrast, the last auction in 2012 saw only 53% of the offered allowances purchased with all of them going to electric generators at a price of $1.93/ton.

The latest auction was held on June 4, 2014, with all allowances sold, with 45% of the allowances going to speculators, and at a price of $5.02/ton.  Demand was 2.9 times the supply with no CCR allowances available as they were all used in the first auction.

A report65 from the Regional Greenhouse Gas Initiative manager, RGGI, Inc., summarizes all the wonderful things the nine participating states have done with the money raised at quarterly auctions of carbon emission permits between 2008 and 2012.  The Executive summary claims 92% of almost a billion dollars raised was “invested” in energy efficiency projects that will return $2 billion in energy bill savings, and save 8 billion short tons of carbon dioxide (CO2) emissions.  The report further claims over 3 million households and 12,000 businesses received these benefits, and the tax caused power plants to emit 40% less carbon dioxide.  Looking at the report critically reinforces why we don’t want a national program like this.

  • Power plant emissions of CO2 have dropped about 40%, but 68% of the reduction was from fuel switching from coal to natural gas to save money, and 32% was from closing down older, smaller coal fired plants that were not worth the investment to meet new federal pollution emission standards. RGGI had no impact on the reduction.
  • Over three quarters of the 3 million residential beneficiaries were low income people who received an average of $50 in electric bill assistance over a five year period, and they would have been better off if the RGGI program hadn’t increased their electric bills in the first place.
  • Before RGGI, the nine states generated 100% of the electricity they consumed. By 2012 so many generating plants closed the RGGI states were paying a premium price to import 8% of their electricity.  Generation grew 3% over the same time period in non-RGGI states.
  • Most of the energy efficiency projects claimed had no post project audits to confirm the energy savings were real, and didn’t consider “free riders”, recipients who would have done projects without RGGI grants. The report also did not consider the “rebound” effect where energy savings are used to increase energy use elsewhere.  For example, one study66 showed 89% of beneficiaries who received grants to buy energy efficient heat pumps were free riders, and the free riders saved no energy because of the rebound effect, while homeowners without rebates saved 16%!
  • One sample project cited for Delaware discussed the low income Weatherization Assistance Program for only 104 homes that was so poorly done (based on a federal audit) every home had to be checked for sloppy and incomplete work by contractors, some of whom were investigated for fraud67. The program was shut down for two years to fix the problems.
  • Without RGGI, data from the US Energy Information Agency shows the US economy has become 2% more energy efficient every year since the 1970’s simply from residents and business acting in their own best interest68. The 2% trend is expected to continue at least through 2040.  Even if the claimed CO2 reductions from RGGI are correct, they represent less than 1% of the savings that would occur anyway, and came at a cost of $1 billion.


65)   “Regional Investment of RGGI CO2 Allowance Proceeds, 2012”, http://www.RGGI, Inc.com

66)    “Free Riding, Up sizing, and Energy Efficiency Incentives in Maryland Homes” , Anna Alberini et al, University of Maryland, College Park, August 11, 2013

67)   “Weatherization Audit Points to Need for Change” www.delawareonline.com , may 28,2010

68)   US Energy Information Agency http://www.eia.gov/todayinenergy/detail.cfm?id=10191

Review of US EIA “Low Electric Demand Case”

The EPA uses the US Energy Information Agency Annual Energy Outlook 2014 to develop their reference case.  EIA also develops alternative cases and one case considers lower electric demand69.  It results in faster coal fired power plant shutdowns, and lower total electric generation.  The result is CO2 emissions drop to 1,774 million tons by 2018, a 29% reduction from 2005, almost meeting the 2030 goal.  The reference case used by the EPA to develop its forecast assumed electric demand would grow about 1% a year.  The low demand growth case assumes electric demand will remain flat through 2040.  The importance of this case is it reflects exactly what happened between 2008 and 2012 despite economic recovery.  The historic connection between the growth in Gross Domestic Product and electric demand has been broken.  It is likely the EIA will move to a no growth reference case in the 2015 Annual Energy Outlook.


69)   US EIA Annual Energy Outlook 2014, “Implications of low electricity demand growth”, 4/30/14, http://www.eia.gov/forecasts/aeo/elec_demand.cfm

David T. Stevenson, Director

September 8, 2014










Read Full Post »