In directing the Federal Energy Regulatory Commission (FERC) to consider a proposed rule that would subsidize Eastern merchant coal and nuclear generators, the Secretary of Energy points to a study by IHS Markit as justification for DOE's proposal. But that study, which claims to put a price tag on “just how important our current balance of coal, gas, nuclear and renewable energy really is,”¹ has so many factual, logical and methodological flaws that it is wholly unsuitable to inform serious public policy.

The IHS Markit (IHSM) study, “Ensuring Resilient and Efficient Electricity Generation,” funded by the U.S. Chamber of Commerce Global Energy Institute (GEI), the Nuclear Energy Institute and the Edison Electric Institute, was released in September 2017.²  It was referenced in the Department of Energy’s September 28, 2017 Notice of Proposed Rulemaking (NOPR) and letter submittal to FERC.³ The IHSM study warns that the nation’s current coal and nuclear generation fleet must be kept in operation or annual energy costs will skyrocket, leading to loss of jobs and GDP and reducing disposable income for every American household. But those conclusions are based on bad logic, false facts and sloppy analysis, as outlined below.

The IHSM study correctly states that, “Current U.S. consumers benefit from a reliable, resilient, and cost-effective electric supply portfolio that employs a diverse set of generating technologies and fuel sources.” It notes the importance of diversity to reduce the magnitude and duration of harm to the grid from disruptive events. The study then asserts that accumulated subsidies, mandates and poor market pricing have distorted wholesale electric market prices, reduced generator revenues, and caused generation shares for “specific technologies” to fall below the shares associated with a “reliable, resilient and efficient electric supply portfolio.” Despite IHSM’s opposition to distorting subsidies and mandates, DOE uses this study to justify extremely generous subsidies for merchant coal and nuclear plants in the Eastern U.S. competitive power markets.

Resiliency

The IHSM report opens with the assumption that the current mix of U.S. generation capacity is already resilient and efficient, but it does not define either resiliency or efficiency.⁴ IHSM defines resilient generation as having a secure primary energy input supply chain for electric production. It asserts that because coal and nuclear plants — apparently even old, slow-ramping, non-cycling, high heat rate, high maintenance cost, low availability plants — have fuel on-site, they possess critical grid resilience virtues. However, IHSM misses the distinction between the “resilience” or availability of a single power plant, a resilient fleet of generators, and a resilient power system.  

On-site fuel availability protects a coal or nuclear plant against the narrow set of challenges associated with fuel shortages, such as coal train or barge delivery interruptions and gas pipeline disruption. But on-site fuel doesn’t protect a generator from extreme weather problems like the loss of transmission from the plant to customers, frozen conveyor belts, flooded or iced-over coal piles, too-warm cooling water, or the inability to get operators in to keep the plant on-line. During the Polar Vortex, 42% of the forced outages within PJM were the result of equipment issues at coal and natural gas units, not lack of fuel.⁵ Florida Power & Light’s Turkey Point nuclear plant suffered $90 million in damage from Hurricane Andrew in 1992; FP&L shut down its two nuclear units before Hurricane Irma hit Florida this year.⁶

The Rhodium Group used utility-reported data to estimate that the average U.S. customer experienced from 245 to 429 minutes of lost electric service per year between 2012 and 2016, finding 96% of those lost service hours due to weather events and only 0.00007% of that outage time due to fuel supply problems. Rhodium further found that over the same period, utilities with more coal and nuclear energy had slightly higher outage frequencies and durations.⁷ Since most customer outage hours arise from failures on the transmission and distribution system, rather than generation failures, it is probable that investments in transmission, distribution, distributed generation and energy efficiency will better improve power system resilience and protect customers than pursuing generator fuel security for its own sake.

Diversity

IHSM decries the loss of fuel diversity and argues that it is necessary to keep coal and nuclear plants on-line for the sake of diversity, without explaining how diversity improves reliability or resiliency or system economics.⁸ In fact, new, low cost, high-efficiency natural gas and renewable generation have been replacing coal, oil and nuclear plants, producing a grid that is more diverse today than ever before — as IHSM’s own graph (and EIA data) show. (Figure 1)

Figure 1 — U.S. Power Generation: Shares by Fuel, 1950-2016

Fleet fuel and technology diversity are higher today than ever before within PJM (which is moving toward an even split among coal, natural-gas-fired and nuclear generation) and most other regions; EIA projects that coal and nuclear power combined will contribute 43% of the nation’s electric fuel mix in 2030.⁹ This diversity makes the generation fleet better able to operate through problems at individual plants and more widespread disruptions from common modes of failure such as pipeline or coal train delivery failures, region-wide freezes or droughts.

Another report issued by the Chamber of Commerce directly contradicts the IHSM study on diversity. The Chamber GEI’s 2017 “Index of U.S. Energy Security Risk” — issued in the same week as the IHSM report, with a press release quoting the same GEI official — acknowledges that U.S. energy security risk continues to fall (i.e., improve). GEI finds the factors that contribute to lower energy security risk include lower electricity prices across much of the U.S. due to low fuel prices, generous production subsidies driving the growth of wind and solar capacity, the continuing growth of domestic natural gas production, and electric capacity margins above 20%. In particular, GEI anticipates relatively stable electric capacity diversity (based on the HHI market share concentration index) through 2019, with the capacity diversity index becoming less risky thereafter.¹⁰ (Figure 2)

Figure 2 — U.S. Electricity Capacity Diversity Index

Flawed logic

IHSM writes that the “improving relative cost of natural gas-fired generating technologies … has not produced what one expects in an economic textbook marketplace: the orderly economic replacement of unprofitable, obsolete generating technologies with new, profitable state-of-the-art natural gas-fired generating technologies.”  But that exact thing has in fact happened — over the past decade, unprofitable, old, inefficient coal and small nuclear plants have been steadily displaced by new natural gas-fired and renewable generation. 

IHSM claims that with coal and nuclear retirements, the grid has become less reliable and resilient. That claim is wrong – in fact, FERC, NERC and all of the regional reliability coordinators report continuing improvements in bulk power system reliability. These improvements arise from better grid management and planning tools, better coordination between electricity and gas supply, better winterization and emergency management practices, and lower plant forced outage rates as older, less reliable plants have been replaced by new, more flexible generation and demand response. IHSM does not specify the reliability and resilient services it claims coal and nuclear plants deliver, nor estimate their value.

IHSM warns that subsidies and mandates for renewables have distorted market prices and caused divergence from a cost-effective generation portfolio: “Increasingly, the U.S. electricity supply is being shaped by subsidies and mandates for favored technologies and fuel sources based on flawed cost assessments typically involving simple levelized cost analyses that ignore the power supply cost implications of balancing electricity demand and supply in real time.”¹¹ But IHSM fails to address whether the costs of maintaining nuclear and coal plants themselves constitute a distorting subsidy for favored technologies (while using simple levelized costs to argue some of its points).

IHSM repeats that grid resilience can be improved cost-effectively by keeping more nuclear and coal plants on-line. However, since most customer outages are due to severe weather-caused failures of electric transmission and distribution, keeping uncompetitive coal and nuclear plants on-line is unlikely to reduce customer outages.

Bad data

IHSM’s study uses bad data to bias its results. It analyzes a “less diverse portfolio” using erroneously high capital costs and per kWh prices for wind and solar generation relative to actual capital prices and contracts today. Both EIA and Lazard Frères estimate that unsubsidized wind is competitive with existing conventional generators, particularly since wind plant capacities have improved to 40-50% (which IHSM overlooks).

IHSM constructs a less diverse portfolio that discounts the value of renewable generation, saying it provides inferior services such as frequency response. IHSM then raises portfolio costs by assuming that wind generation must be backed up by gas-fired generation — even though ALL power plants are backed up by all the other supply and demand response on the system. Recent advances in control technology enable new renewable generation to provide many essential reliability services at a speed and cost that is equal to or better than slow-moving, low-flexibility coal and nuclear plants.

IHSM points to renewable policies as a major cause of low electric prices and coal and nuclear plant retirements.  This contradicts the DOE Staff Report to the Secretary on Electricity Markets and Reliability and many other analyses, which agree that low electric prices are due to the combination of low natural gas prices, flat electric demand and market competition, while most retirements are due to the plants’ age and inefficiency exacerbated by environmental and relicensing upgrade costs. 

IHSM overstates the importance of the 2014 Polar Vortex event as a justification for generation diversity (ignoring other events such as hurricanes, earthquakes, wildfires and droughts) and misrepresents plant failures during that event. IHSM ignores the facts that since 2014, generation owners have improved their winterization practices; several RTO markets have established capacity performance mechanisms to assure better generation availability in future disruptive events; all reliability coordinators use continually improving grid planning and operations methods (including coordinating electric and gas operations); and the entire industry is improving emergency event preparation and management.

While complaining that subsidized renewable resources give an unfair advantage to wind and solar generation, IHSM does not recognize subsidies such as nuclear power plant insurance or state zero emissions credits to nuclear power plants.

IHSM claims that customer electricity bills would show more monthly variability under the less diverse portfolio.  This ignores the fact that most customers use flat electric rates and state regulators actively shield customers from volatile rates and bills.

While decrying the price volatility associated with natural gas pipeline and delivery problems associated with the 2014 Polar Vortex and the 2011 Texas Eastern Transmission disruption in Pennsylvania, IHSM ignores the facts that most natural gas generators use hedging methods to stabilize their fuel prices, while retail electric providers use hedges to reduce vulnerability to real-time fuel-induced price volatility.

Flawed methodology

IHSM’s study contains several methodological flaws, starting with the comparison of a “less efficient diversity” generation portfolio to actual grid performance over the period 2014 through 2016. IHSM specifies this portfolio as having zero nuclear, coal and oil-fired generation and 20% less hydro generation, and replaces all of the missing generation with 200% more intermittent renewable resources and many new natural gas plants. Then IHSM runs its less diverse portfolio through grid and economic models and compares the estimated results for the less diverse portfolio to the actual results for the same period.  Not surprisingly, the less diverse portfolio — which was defined and structured to be more expensive — produces drastically worse economic results and costs than the actual outcomes.

This approach has several problems:

• The less diverse portfolio is improbable and unjustified. IHSM removes all coal- and oil-fired and nuclear generation plus 20% of hydropower capacity from the 2014 through 2016 generation mix, even though the Energy Information Administration (EIA) expects that most of that generation will remain on-line through 2030. For 2030, EIA’s 2017 Annual Energy Outlook projects 183 GW of coal generation, 52 GW of oil and natural gas steam plants, 97 GW of nuclear power, 304 GW of renewables (including hydro, which EIA does not expect to fall from current capacity levels), and 384 GW of natural gas combined cycle and combustion turbine capacity.¹²
• The less diverse portfolio replaces the missing generation (apparently overnight, with associated transmission) with new, full-cost wind, solar and natural gas-fired generation, even though there existed plenty of already-built natural gas capacity that could have been run at higher capacity factors to make up part of the sudden capacity loss.  
• IHSM inflates its alternate portfolio’s cost to electricity customers by removing all of the Investment Tax Credits and Production Tax Credits that offset wind and solar costs — even though those subsidies were actual payments that lowered customers’ payments over the 2014-2016 period and will continue for years ahead.
• IHSM concludes that, “subjecting the current U.S. economy to the less diverse U.S. power supply power price increase would trigger economic disruptions, some lasting over a multi-year period,”¹³ ignoring the fact that these hypothesized disruptions are created by the analysts’ own tortured assumptions rather than a realistic alternative scenario. 
• IHSM does not explain why it compares three years of past results (backcasting for 2014 through 2016) rather than forecasting multiple generation scenarios with a consistent methodology.
• A more reasonable analysis (forecast or backcast) might retain all of the uneconomic, now-retired coal and nuclear power plants and run them at forced generation levels to estimate the costs and reliability impacts of keeping this high-cost generation on-line relative to continued retirements with more gas and renewables.

IHSM claims that its less resilient portfolio erodes electric reliability, causing “more frequent power supply outages that add to the current U.S. average expected outage rate of 2.33 hours per year,” with a cost reported at another $75 billion per hour.¹⁴  But the study explains neither the higher outage level nor the additional outage costs.  The latter estimate is particularly curious because – for perspective – all of the electric outages due to Superstorm Sandy in 2012, in multiple states over multiple days, are estimated to have cost customers between $27 and $52 billion.¹⁵

Conclusion

IHSM’s study on resilient, reliable, efficient generation does not actually explain much useful about grid resiliency, reliability or efficiency. IHSM claims to favor efficient, transparent electric markets and oppose mandates and subsidies, but its analysis muddies rather than illuminates these issues. This study, which DOE cites to justify cost of service subsidies for Eastern merchant coal and nuclear plants, has so many logical, factual and analytical flaws that it should be ignored rather than cited.