NYISO Wind Generation Record Publicity

The New York Independent System Operator (NYISO) issued a press release on December 17, 2019 announcing a new wind generation record for the state.  I disagree with the tenor of the press release and this post explains why I think it is inappropriate.  It was a lost opportunity to educate the public about the magnitude of the effort needed to meet the State’s ambitious clean energy goals.

Press Release

The press release stated:

Strong winds across New York State over the weekend pushed electricity generated by wind power to a new record.

The new record output of 1,675 megawatts (MW) was set during the 11:00 p.m. hour on Saturday, December 14, eclipsing the previous record of 1,651 MW which was set during the 8:00 p.m. hour on April 26, 2019.

When overall wind production peaked at 1,675 MW on Saturday night, it provided 11% of all energy generation in New York. The record output represents 84% of the 1,985 MW of installed wind capacity in New York State.

One megawatt is approximately the amount of electricity required to supply 800 to 1,000 homes. Interested parties can track the NYISO’s real-time fuel mix on our website, www.nyiso.com.

In my opinion the press release is thinly veiled propaganda support for the Climate Leadership and Community Protection Act (CLCPA) which includes a requirement to eliminate the use of fossil-fueled electric generation by 2040.  Announcing this record without qualifying the wind resources of New York gives the impression that the State is on track to meet that target.

Right off the bat note that the “record output represents 84% of the 1,985 MW of installed wind capacity in New York State.  Think about that.  The State has invested in wind energy and the best they have done is 84% of the total installed.  Unfortunately, when you look at the big picture serious problems show up.

Approach

I used two sources of data from NYISO to evaluate the existing New York wind energy resource.  For an overview I used the annual report that presents load and capacity data including historical and forecast seasonal peak demand, energy usage, and existing and proposed generation and transmission facilities.  The Load and Capacity Date Report or Gold Book is a featured report in the NYISO document library.  This post and a summary I posted in April 2019 use data in Table III-2 Existing Generating Facilities from those reports to describe the annual wind energy resources available.  Note that in 2018 all wind energy came from on-shore facilities.

The NYISO Real-Time Dashboard includes a window for the real-time fuel mix that includes the amount of wind generation being generated in the state.  The window also includes a link to historical data.  I downloaded data for all of 2018, sorted out the wind production numbers, and then calculated hourly averages to compare with the annual numbers from the Gold Book.  I use Statgraphics Centurion software from StatPoint Technologies, Inc. to do my statistical analyses and in this case I loaded the hourly data and calculated frequency distribution statistics.

Results

The NY 2018 Wind Facilities in the NYISO 2019 Gold Book table lists all the New York wind energy facilities.  The NYISO table provides the name plate ratings and 2018 net energy produced.  I used that information to calculate the annual capacity factor for each facility.  Note that there is a wide variation of capacity factors, that the highest is only 35.7%, and the state-wide capacity factor is only 24.5%.  In other words, New York wind facilities only provide a quarter of their name plate capacity.  So in the best hour wind energy has reached 84% of the nameplate capacity and over the year wind energy only produces only 24.5% of the possible power that could be produced.  But wait, there’s more.

Another wind-resource issue is the distribution of the hourly output.  The 2018 Hourly Wind Generation (MW) Frequency Distribution document lists frequency distribution data for all of 2018.  The histogram of wind output categories shows a skewed distribution such that low output is more frequent than high output.  The frequency tabulation for wind table shows that there were 10 hours when none of the 24 wind facilities in the state produced any power and that 32% of the time less than 200 MW per hour was produced.  The percentiles indicate that half the time hourly wind output is less than 346 MW and that for 876 hours (the tenth percentile) wind energy provides less than 49 MW of energy.

If New York has to rely on renewable energy in the future it is important to know the frequency distribution of wind at night when solar output is unavailable.  I used the New York City sunrise and sunset times and calculated when it was dark.  The 2018 Hourly Wind Generation (MW) Frequency Distribution at Night document lists the same statistics as before but only for night time hours.  While there was only one hour with no wind output and the frequency of hours with output less than 200 MW was down to 28% there still is a significant number of hours when there is no appreciable renewable energy being generated. The percentiles indicate that half the time at night hourly wind output is less than 367 MW and that for 876 hours (the tenth percentile) wind energy provides less than 65 MW of energy.  That means that energy storage is going to be absolutely necessary.

Conclusions

I think that the independent system operator has an obligation to the consumers of the state to tell it like it is.  This politically expedient press release did not mention any of the issues associated with wind energy relative to the CLCPA target.  Based on my results I am sure they could have easily found a day when the wind resources were weak.

I have no doubt that NYISO knows about these issues.  They know that the annual capacity factors are low.  I have never seen them publish the distribution of hourly wind output but I have to assume that they have looked at the resource in a similar manner.  I did not think it would be as bad as it is, and these results have important implications with respect to energy storage.  In an earlier post I estimated how much energy storage would be needed for one example period and the costs are startling.

Advocates for renewable power maintain that it is possible to address the problem of calm winds at one location by simply adding facilities in other locations where the wind is blowing.  If that were the case using New York resources the hourly distribution would not show that 5% of the time the total wind energy production for the entire state was less than 24 MW.  Furthermore, I suspect that even expanding the location of wind facilities to off-shore New York and adjoining jurisdictions is not going to significantly reduce the number of hours when wind resources are going to have to be supported by energy storage.  The fact that night time wind generation also shows significant hours with low levels exacerbates the need for energy storage because we cannot use solar to shave the amount needed.

I am very disappointed that NYISO ignored the opportunity to educate the public about the limitations of New York’s wind energy resource.  These results reinforce my position that New York State has to do a comprehensive analysis of the availability of renewable resources to determine a strategy for meeting demand with an all-renewable system.  Until that is complete, we are only guessing whether this can be done, much less how much this is all going to cost.  The NYISO should be pressing for this analysis and this was an opportunity to explain why it is necessary.

Generic Carbon Pricing Issues

The New York Independent System Operator (NYISO) is currently campaigning for its Carbon Pricing Initiative as the preferred approach to meet the requirements of New York’s Climate Leadership and Community Protection Act (CLCPA).  For example, they sponsored a blurb in the Politico New York Energy daily newsletter.  I have written extensively on my issues with the NYISO initiative and this post explains my concerns with carbon pricing schemes in general.

 The NYISO sponsored the following message in Politico New York Energy:

An increasing number of organizations recognize this unique, market-based solution as a viable, scalable option for helping to reduce carbon emissions. The World Economic Forum recently published an article by New York ISO, CEO Rich Dewey, Putting a Price on Carbon Will Help New York State Achieve a Clean Energy Future.

The World Economic Forum, an organization for public-private cooperation, engages the foremost political, business, cultural and other leaders of society to shape global, regional and industry agendas. New York, the 11th largest economy in the world, recently enacted the United States’ most aggressive climate change legislation. The New York ISO’s proposal for carbon pricing would embed a cost per ton of CO2 emissions in the sale of wholesale electricity, creating a price signal for investment in new clean energy resources. Read article.

Carbon pricing theory says that when the price of energy is raised by adding a cost for carbon, the increased costs at the higher CO2 emitting sources of energy will provide incentives to transition to lower or zero CO2 energy sources.  This is supposed to lead to the most cost-effective reductions.  I think that there are a number of practical reasons that carbon pricing will generally not work as theorized: leakage, revenues over time, theory vs. reality, market signal inefficiency, and implementation logistics.  Based on those concerns the NYISO plan is not going to solve anything in NYS.

Leakage

Leakage refers to the situation when a pollution reduction policy simply moves the pollution around rather than actually reducing it.  Ideally you want the carbon price to apply to all sectors across the globe so that cannot happen.  I don’t think a global carbon pricing scheme is ever going to happen because of the tradeoff between the benefits which are all long term versus the costs which are mostly short term.  I don’t see how anyone could ever come up with a pricing scheme that equitably addresses the gulf between the energy abundant “haves” and those who don’t have access to reliable energy such that “have nots” will be willing to pay more to catch up with those who have abundant energy.

Ultimately, I think that leakage will be a problem for any limited area carbon pricing policy.  Trying to force fit this global theory into the New York electricity market is an even more difficult problem.  As proposed, it will likely result in locational leakage where energy and emissions are not reduced but simply shift emission location within the inter-connected electric grid.   Additionally note that a carbon price on just the electric sector may result in leakage if more consumers generate their own power using unpriced fossil fuel.

Revenues Over Time

A fundamental problem with all carbon pricing schemes is that funds decrease over time as carbon emissions decrease unless the carbon price is adjusted significantly upwards over time.  This problem is exacerbated because over time reducing CO2 emissions becomes more difficult.  It has been observed that roughly 80% of the effects come from 20% of the causes and everyone knows the meaning of low hanging fruit.  This has been observed with regard to New York’s observed CO2 emission reductions to date.  New York electric sector emissions dropped 56% between 1990 and 2016 mostly by retiring old units and fuel switching to lower emitting fuels.  It can be argued that those reductions would have happened anyway because retirements and fuel switching were lower cost options without even considering CO2 emissions.   Furthermore, I believe that air pollution control costs increase exponentially as efficiency increases which makes this issue even more problematic.

This difficulty should be even more of a concern with CO2 emission reductions because at some point replacing existing fossil-fired generation not only has to consider the direct power output conversion costs but must also address dispatchability and grid support costs.  When those costs are included there will be a sharp increase in total costs per CO2 reduced.  Like many others, the NYISO Carbon Pricing Initiative proposes to use the social cost of carbon (SCC) as the carbon price.  The SCC cost increases over time but the costs over time do not increase enough in my opinion to keep pace with the necessarily more expensive total costs to maintain reliable electricity to consumers.

Theory vs. Reality

Another problem with carbon pricing theory is that in practice affected sources may not act rationally or as theory expects.  The Regional Greenhouse Gas Initiative (RGGI) is a market-based carbon pricing program and I have written extensively on it.  The academic theory for RGGI market behavior is that affected sources will treat allowances as a storable commodity and act in their own best interest on that basis.  If that were true affected sources would be purchasing allowances for long-term needs and “playing” the market to maximize earnings.  In practice RGGI affected sources plan and operate on much shorter time frames and have shown no signs of making compliance obligations a profit center.

Carbon pricing theory claims that when the cost of using higher emitting energy increases that will provide incentives to develop alternatives and discourage continued use of existing resources.  However, these incentives are indirect and again assume rational behavior in the market.  While theory says that a company that currently operates a fossil-fired plant will change its business plan and develop a renewable energy facility to stay in business, there are a whole host of reasons why the company may not go that route and instead treat the carbon price as a tax and continue to operate with that constraint.  In my opinion RGGI did not induce any NYS companies to change their business plans.

Market Signal Inefficiency

I am also concerned because the carbon price signal is an indirect inducement for emission reductions. CO2 emission reduction efficiency is an issue based on New York’s experience in RGGI.  The New York State Energy Research and Development Authority (NYSERDA) report New York’s RGGI-Funded Programs Status Report – Semiannual Report through December 31, 2018 (“Status Report”) describes how New York invested the proceeds from the RGGI auctions.  That report lists the many programs that are funded using RGGI proceeds as shown in Table 2 Summary of Expected Cumulative Annualized Program Benefits through December 31, 2018. There are six program categories: Green Jobs – Green New York, Energy Efficiency, Renewable Energy, Community Clean Energy, Innovative GHG Abatement Strategies, and Clean Energy Fund.

I combined the data for the six program categories in the Consolidated Summary of Expected Cumulative Annualized Program Benefits through 31 December 2018 table.  It summarizes the emission reduction benefits and costs for those categories.  The cost per ton reduced ratio ranges from $167 to $3,437.  At the high end the GHG Abatement Strategies category emphasizes long-term research and development.  Because this research could lead to a cost breakthrough this funding can be justified.  Looking at the other categories it appears that the more investments are focused on direct reductions rather than indirect investments the better the cost benefit ratio.  For example, the best ratio ($167 per ton removed) is in Community Clean Energy and that category includes direct support for renewable energy projects.   The Energy Efficiency category is an example of indirect support because investments in this category do not directly reduce emissions.  Instead the investments reduce energy use which reduces the need for energy production and indirectly reduces emissions.  However, the cost per ton removed, $425, is markedly higher than the best category.

Theory says that the carbon price alone can incentivize lower emitting energy production and that the market choices will be more efficient than government mandated choices. However, as a result of these observations, I do not think that carbon pricing schemes, like the NYISO initiative, that raise the cost of energy and do not include specific funding aspects will work as efficiently in the short term and in limited markets like New York as theory suggests.  There are risks involved so who is going to make the investments and when will they make investments?

Implementation Logistics

Finally, I believe that there are significant logistical issues associated with carbon pricing that the NYISO process has simply ignored.  In order to set a carbon price, you have to know what the carbon emissions are for every source providing energy to the market.  For a global all-sector pricing scheme, you could set the price as the fuel is produced so that everyone pays the cost all the way through its end use.  On the other hand, the NYISO has to set the price as electric energy is sold on a real-time basis.  That is a non-trivial problem.  In New York, NYISO knows which generator is running and has a pretty good idea of their emission rate.  However, the final emission numbers are not available real-time because the emission values reported to prove compliance are not finalized until quality assurance post processing is complete and that can be months after the fact. The more significant problem is that NYISO has no way to calculate imported electricity carbon emissions on a real-time basis so cannot assign a carbon price value that accurately reflects how imported electricity is being generated.  These issues have been glossed over to date.

Conclusion

The NYISO claims that “An increasing number of organizations recognize this unique, market-based solution as a viable, scalable option for helping to reduce carbon emissions market-based solution”.  I frankly don’t think those organizations have had actual experience with a carbon pricing initiative logistics and have not evaluated whether the carbon prices proposed will provide the market signals necessary to spur the necessary renewable development needed to meet any CO2 emission reduction goals as a viable, scalable option for helping to reduce carbon emissions for the CLCPA.

The success of any carbon pricing scheme boils down to the question whether the carbon price set will provide enough of an incentive for projects that produce emission reductions that displace today’s generators and eventually covers the costs to provide the dispatchability and grid support functions provided by today’s generation mix. There are no estimates that this will be the case for the NYISO initiative.

In my opinion, NYISO carbon price initiative support is based on parochial interests.  In the case of NYISO they appear to believe it will simplify the cost accounting for New York’s renewable implementation efforts.  I think they have under-estimated the difficulty implementing the infrastructure necessary to accurately track the price of carbon and have ignored the potential that the complex scheme needed to reduce leakage will lead to unintended consequences.  Other support appears to be based on the potential to make money and it is not clear that is in the best interest of the State’s desire to reduce CO2 emissions as cost-effectively as possible.

The more I study the practical implementation of carbon pricing schemes the more skeptical I become.  I think that there are a number of practical reasons that carbon pricing will not work as theorized.  Because a global program is impractical, leakage is always going to be a problem.  The carbon price has to be set such that revenues over time increase significantly.  The economists who support this theory seem to be blissfully unaware of the reality of the energy market. Based on observed results I think that indirect market signals are going to lead to less cost-effective reductions in the time frame necessary for the aggressive reduction rules.  Finally, no supporters seem to understand the very real problems of implementation logistics.

Update December 30, 2019:  Please check out the companion post describing additional problems with carbon pricing raised by Paul Homewood at Not a Lot of People Know That blog.

Transportation Climate Initiative Polling Results

I have previously posted on the Transportation Climate Initiative and its disconnect with reality.  To this point its promoters and stakeholders have labored in obscurity all the while believing that they have a mandate from the public to change the transportation system significantly.  Now there is a poll out claiming that the majority of the public supports the plan.

Survey

A new survey from MassINC Polling Group purportedly shows broad support for the Transportation Climate Initiative.  According to the press release:

A new set of polls of registered voters across the Northeast and Mid-Atlantic finds broad support for a multi-state policy to cap carbon pollution from transportation and invest in transportation improvements. That policy is currently being developed by the Transportation and Climate Initiative (TCI), a collaboration between 12 states and the District of Columbia.

The research surveyed registered voters in the seven largest states at the TCI table: Connecticut, Maryland, Massachusetts, New Jersey, New York, Pennsylvania, and Virginia. Overall, 66% of voters said they would support the policy, while 23% were opposed. Support ranged from a high of 71% in New York to a low of 60% in Virginia.

“This is a complex policy, and so we took the time to explain the basics of how it would work and how states might use the funds generated by it,” said Steve Koczela, president of The MassINC Polling Group, which conducted the polls with support from the Barr Foundation. “Support was broad, stretching across demographic and party lines and throughout the region.”

Across nearly every demographic group, more voters support than oppose their state joining in the program. Support was highest about younger voters (78%), non-white voters (75%), and lower income voters. Women (69%) were slightly more supportive than men (63%). The partisan gap was smaller than is seen on many issues in the current polarized climate, with Democratic support outpacing Republican support.”

 Voters were also asked to rate several potential uses for the funds generated by the proposed program. While majorities supported each item, two rose to the top overall and across the seven states polled: improving public transit and protecting transportation infrastructure from the effects of climate change. Generally, investing the proceeds of the program in existing infrastructure was favored over seeding newer technologies like electric vehicles, targeting communities most effected by pollution, or insulating drivers from higher costs at the pump.

The polls also asked voters about their views of climate change:

                • 73% of voters across the region think climate change is probably happening;
                • Among those, 85% think it is due at least in part to human activity;
                • 66% think climate change will be a serious problem for their state if left unchecked;
                • 57% think the federal government is doing too little to address climate change; and
                • 44% think the same of their own state government.

Voters also understand that transportation is contributing to climate change. When asked to rank six sectors of their state’s economy based on their greenhouse gas emissions, 65% ranked transportation first or second.

Despite all this, only 48% cited addressing climate change as a “major priority” for state government in their state. More (64%) considered improving roads, highways and bridges a major priority, just below health care costs (73%) and jobs and the economy (71%).

The disconnect between voters’ professed concern about climate change and the priority they place on addressing it is consistent with other polling MPG has conducted on climate change. “We’ve seen in other polling on this issue that voters believe in climate change and support policies that would address it, though not always because of the link,” said Koczela. “They may not always rank climate as a top immediate concern, but they do support actions to address it.”

About the Poll

These results are based on a survey of 6,395 registered voters across seven Northeast and Mid-Atlantic states: Connecticut, Maryland, Massachusetts, New Jersey, New York, Pennsylvania, and Virginia. Responses were collected via online survey interviewing November 12-19, 2019. Final survey data was weighted to known and estimated population parameters for each state’s registered voters by age, gender, race, education, geography, and party. Each state was then weighted to its relative proportion of registered voters to create an “Overall” regional average. This project was sponsored by The Barr Foundation.

My Thoughts

Steve Koczela, president of The MassINC Polling Group said “This is a complex policy, and so we took the time to explain the basics of how it would work and how states might use the funds generated by it”.  It is mind boggling to me that he never thought to include a question about costs.  Obviously if costs are low why wouldn’t everyone be in favor of “improving public transit and protecting transportation infrastructure from the effects of climate change”.

The published plan says that in December 2019 there will be a release of a regional policy proposal in the form of a draft Memorandum of Understanding (MOU), accompanied by modeling results that estimate the energy and emissions implications of different cap levels and investment scenarios, as well as potential costs and benefits of different program design options.  When it is released then there can be a poll that determines how much appetite the public has when they know potential costs.  Until then any poll is suspect.

I have one other problem with the poll itself.  I was unable to find details about the geographic distribution but because the only geographical categories listed were by state, I doubt that there was consideration of urban vs. rural.  The fact is that rural respondents are not going to rate improvements to public transit very high because it is of no value to them.  On the other hand, the gasoline tax increase needed to fund the TCI will disproportionally disadvantage anyone who has no choice but to use their personal vehicle.  For example, the National Household Travel Survey estimated vehicle miles traveled in 2009.  In the TCI jurisdictions urban vehicles mile traveled totaled 386.55 million miles, suburban miles traveled were 666.36 million miles, and rural miles traveled 703.38 million miles.  Failure to account for that geographic difference makes the poll suspect.

Finally, I think there is a corollary to the concept that you can prove anything with statistics.  In particular, you can prove anything with a survey poll.  The Barr Foundation sponsored this poll.  Their website states:

“Climate change is real. It is happening. It is accelerating. All over the world, people are experiencing its effects. And these are becoming more devastating every year—a trend that can be checked only through dramatic, global effort.”

I believe they got what they paid for.

NYISO Winter Peak Analysis Implications to CLCPA

A new report prepared at the request of the New York Independent System Operator (NYISO) addresses issues associated with an electric system reliant on renewable energy sources during the winter.  This post compares these results with my previous work related to New York’s Climate Leadership and Community Protection Act (CLCPA) and discusses the implications on that law.

The Citizens Budget Commission developed an overview of the CLCPA targets in Green in Perspective: 6 Facts to Help New Yorkers Understand the Climate Leadership and Community Protection Act.  The goals of the law are truly aspirational:

Reduce greenhouse gas (GHG) emissions:

      • Reduce GHG emissions to 60 percent of 1990 emissions levels in 2030;
      • Generate zero GHG emissions from electricity production by 2040; and
      • Ensure GHG emissions are less than 15 percent of 1990 emissions levels in 2050, with offsets to reduce net emissions to zero.
      • GHG offsets means that for every ton emitted into the air one ton is removed via GHG capture of some sort. For example, a company or individual can pay a landowner to leave trees standing that would otherwise be removed or plant additional trees to offset GHG emissions.

Increase renewable electricity:

      • Increase renewable sources to 70 percent by 2030; and

Develop or support:

      • 9 gigawatts (GW) of offshore wind electric generation by 2035;
      • 6 GW of distributed photovoltaic solar generation by 2025; and
      • 3 GW of energy storage capacity by 2030.
      • Conserve 185 trillion British thermal units (TBTUs) of annual end-use energy use by 2025, of which at least 20 percent should be from energy efficiency improvements in disadvantaged communities.
      • The CLCPA also requires between 35 percent and 40 percent of spending on clean energy or efficiency programs be in disadvantaged communities and mandates an air monitoring program in at least four such communities.

I have evaluated winter peak impacts in previous posts on New York Resource Adequacy Proceeding Comments, Solar Issues in Upstate New York , CLCPA Solar and Wind Capacity Requirements and CLCPA Energy Storage Requirements.  My primary concern is the requirement to generate zero GHG emissions from electricity sector production by 2040 coupled with the increased load needed to electrify the heating and transportation sectors enough to meet the 85% reduction by 2050 target.

Fuel and Energy Security In New York State Report

NYISO  had the Analysis Group do a forward-looking assessment of the fuel and energy security of the New York electric grid during winter operations.  The November 2019 final report was titled: Fuel and Energy Security In New York State: An Assessment of Winter Operational Risks for a Power System in Transition.  The objective was to assess winter fuel and energy security risks and identify key factors that would affect risks. Specifically, the study targeted potential reliability risks and impacts under severe winter conditions and adverse circumstances regarding system resources, physical disruptions, and fuel availability.  Importantly it is a snapshot of the winter of 2023-2024 before the CLCPA renewable energy and electrification of other sectors implementation really kicks in

Previously I have analyzed the effect of winter peaks and I chose 12/29/17 to 1/12/2018, a period that is included in their analysis.  The Analysis Group defined extreme weather events including the largest increase above average daily load over a long period as 14 days from 12/25/2017 to 1/8/2018 and more extreme shorter periods where they found in the last 25 years the fourth lowest 3-day cold snap was 1/4/2018 to 1/7/2018.  Then they evaluated different scenarios that included different combinations of “(a) timeframe for the development of new renewable resources; (b) capacity imports from neighboring regions; (c) potential retirement of units affected by the peaker rule; and (d) availability of natural gas for power generation”. The evaluation determined where these scenarios might cause problems.

The analysis included the following relevant conclusions (two key points underlined by my emphasis):

    • With the continued operation and availability of most of the assets currently expected to be in place in the winter of 2023/2024, the NY grid contains sufficient diversity and depth of fuel supply to support reliable winter operations. This result is consistent with the historical operating experience in recent past winters, including during severe weather conditions.
    • Meeting the state’s renewable and clean energy goals can provide valuable reliability support, and may be particularly true with respect to offshore wind. Delayed realization of renewable resource additions (as compared to the 2017 CARIS Phase 1, System Resource Shift case levels that are assumed under initial conditions) can lead to potential LOL events that would not otherwise occur when combined with other adverse system conditions. The potential magnitude and pace of change to the resource fleet stemming from requirements under the CLCPA may be of far greater importance for evaluation than the considerations, scenarios and physical disruptions evaluated in this fuel and energy security study with respect to winter operational risks.
    • The availability and contributions of adequate levels of natural gas-fired and oil-fired (or dual fuel) generating resources is necessary to maintain power system reliability in cold winter conditions in the near-term. This is particularly true for Long Island and New York City. Simply put, avoidance of potential loss of load events in these load centers, under plausible adverse winter conditions, requires operation of natural gas and oil-fired units. Reduction in the generation available from such resources – whether through capacity retirements, low initial oil inventories, reduction in natural gas availability for power generation, or interruptions in the ability to refuel oil tanks throughout the winter represents the most challenging circumstances for reliable winter system operations in New York over the coming years.

Implications for CLCPA

The analysis notes that the “potential magnitude and pace of change to the resource fleet stemming from requirements under the CLCPA may be of far greater importance for evaluation than the considerations, scenarios and physical disruptions evaluated in this fuel and energy security study with respect to winter operational risks”.  I agree because I believe that it is absolutely necessary for the State to prove that when the energy load increases when other sectors are electrified that fuel and energy security can be maintained without using fossil fuels.

The analysis also states that “Simply put, avoidance of potential loss of load events in these load centers, under plausible adverse winter conditions, requires operation of natural gas and oil-fired units”.  The CLCPA requirement that all electric energy must come from non-fossil fired sources in 2040 is an extraordinarily difficult goal to meet.  The political calculus to include this in legislation was not backed up by any analysis.  The state has to show how this can be done as soon as possible lest New York resources be squandered on an impossible quest.  As I show below, the actual renewable resource may not support this target because of logistical issues and even if it does there may be immense costs.

Need for Renewable Energy Resource Analysis

In my back of the envelope analysis of the summer peak energy storage requirements I used actual wind speed data to estimate the New York off-shore wind resource.  New York State awarded the first two contracts for off-shore wind projects in July 2019.  The Equinor 816 MW winning project press release said “The project is expected to be developed with 60-80 wind turbines, with an installed capacity of more than 10 MW each”.  Among the many details redacted in the public version of their proposal was specific information on the proposed wind turbines.  The public version included a diagram of the proposed wind turbine size as compared to the Chrysler building and showed that top tip of the blade at 250 m.  I estimated the hub height to be 173 m by scaling the drawing.  In this analysis I characterized wind energy output as a function of observed wind as follows.   I found a wind turbine power output variation curve that had a cut-in speed of 3.5 m/s and a cut-out wind speed of 25 m/s. Using that wind variation curve, I estimated that output of each 10.2 MW wind turbine will equal 0.971 times the wind speed minus 3.4.

For the input meteorological data, I found a National Oceanic and Atmospheric Administration buoy located 30 nautical miles south of Islip, NY (40°15’3″ N 73°9’52” W) that I used to represent NY offshore wind resource availability. The observed wind speed at the hub height is proportional to the logarithm of the height above ground.  For that calculation I assumed a hub height of 173 m and a surface roughness of 0.0003 using the buoy anemometer height of 4.9 m. I downloaded hourly NDBC data for 2018 and 2017 and calculated the wind energy output for every hour in the period 12/25/2017 to 1/8/2018 using that relationship and the wind turbine output variation equation I derived.

The key finding is that there were two no wind energy output periods on 3-4 January 2018 during an intense cold snap when electric load is high as shown in the New York Off-Shore Wind Generation Estimate for 9000 MW CLCPA Off-Shore Target table.  I was surprised to see that the wind resource went to zero during a high load period not only when the winds were light on January 3 but also when a deep low pressure developed and the wind speeds exceeded 25 m/s on the very next day.  The wind generation estimate table lists the output from a single 10.2 MW wind turbine, 80 turbines in the Equinor proposed wind facility and for all 9,000 MW of Cuomo’s CLCPA target of 9,000 MW of off-shore wind.  It is important to note that adding even more wind turbines still does not preclude the need for substantial energy storage.  While all the New York off-shore wind resource may not go to zero simultaneously that resource is going to be highly correlated across the available area so they all will track closely.

Conclusion

Every time I look at the meteorological data relative to the winter peak I get a surprise.  I expected that the winter observed peak load would occur during very cold weather associated with a slowly moving high pressure system that originated in the cold northern plains large enough to cover the entire northeastern US.  The resulting multi-day period of clear skies, light winds, and inherent cold temperatures would result in very high energy demand for heating.  The early January 2018 high load period was very different.  Weather maps for this period show (January 2018 Weather Maps) a relatively small high pressure system in the central US on January 2 that moved east ahead of a storm system on January 3.  The high pressure was strong enough over the New York offshore wind region that winds were less than 3.5 m/s for five hours on January 3.  However, the storm system moved eastward and re-developed into a strong storm just off the coast on January 4 with an eleven-hour period of greater than 25 m/s wind speed 13 hours after the light wind period ended.  By January 5 the storm had raced northeast to the Canadian Maritimes but was pumping cold air back across New York State.

This period must be analyzed in more detail by New York State to determine whether the CLCPA requirements endanger fuel and energy security.  If the assumptions I used for no wind power due to light winds and strong winds are correct then there will be 16 hours of no wind power in a 29-hour period during the coldest extended duration cold weather event that the Analysis Group identified after analyzing 25 years of data.  Furthermore, it also overlaps fourth worst 3-day cold snap.  The State needs to estimate what the future load will be when the home heating and transportation sectors are electrified to meet the CLCPA emission reduction goal and then assess whether renewable resources will be adequate during the entire extended duration period using the proprietary energy output information in the renewable energy proposals submitted to the State not only in the NYSERDA off-shore wind program but also the Article Ten applications.  This analysis has to be done for the entire state and obviously will lead to an estimate of the amount of energy storage necessary in 2040 when no electric energy from fossil-fired facilities is allowed.  It is not clear to me if there is enough space available where it is needed to site all the renewable and energy storage necessary.  Even if there is enough space, this analysis will provide the information needed to estimate how much all this will cost.

Frankly, it is laughable that the New York State legislature and Governor Cuomo enacted a law mandating specific energy and emission reduction goals without doing such an analysis first.  I believe it is time for the energy professionals in the State to step up and demand such an analysis before the State squanders money on a system design that can only be implemented with massive wind, solar and energy storage development.  Even if this system could be developed it will surely be expensive.  Just how much is anyone’s guess until such a study is completed.

CLCPA Renewables Needed for Doubled Electric Load

The Brattle Group recently released a report entitled “Achieving New England’s Ambitious 2050 Greenhouse Gas Reduction Goals Will Require Keeping the Foot on the Clean Energy Deployment Accelerator”.  The primary emphasis of this blog is on New York environmental and energy policy issues so this post applies their renewables needed estimate to the New York’s Climate Leadership and Community Protection Act (CLCPA) requirements.

In particular, the Brattle report notes that in order to reduce greenhouse gas emissions from the whole economy it will be necessary to electrify transportation and heating which will increase electric load.  They estimate that in order to meet New England’s goal of an 80% reduction by 2050 that electricity demand could be “twice the current level by 2050”.   I have yet to see any New York agency estimate of future load for the CLCPA so I decided to try to estimate how much renewable energy would be needed if New York electric load doubles by 2040 at the same time that New York electric load is supposed to eliminate fossil-fired generation.  I used that estimate to estimate how much renewable energy capacity (MW) in the form of on-shore wind, off-shore wind and solar could be needed in New York in 2040.  I determined how much power (MWh) was used in 2018, doubled that, and then guessed how the necessary capacity could be distributed between on-shore wind, off-shore wind and utility solar.

Brattle Key Findings

The Brattle Group prepared their report on behalf of the Coalition for Community Solar Access.  According to the news release there were several key findings:

      • Electricity will play a critical role in decarbonizing the New England economy. As a result, electricity demand will grow substantially and could well be twice the current level by 2050.
      • In supplying this growing demand for power, both solar photovoltaic (PV) and offshore wind will likely play a critical role.
      • Merely maintaining the current rate of clean energy resource deployment will cause the region to fall short of its targets. Currently planned clean energy resource generation for 2019–2030 in New England amounts to approximately 830 MW per year. This represents a significant increase from the historical generation of 280 MW per year from 2010–2018.
      • However, to achieve the 2050 targets, New England will need to accelerate clean energy resource additions to between 4 and 7 GW per year on average between 2021 and 2050.
      • To reach these levels, annual clean energy resource additions will need to continue to grow by approximately 9% per year through 2050.

Procedure

The first step is to determine how much generation was used in 2018.  I used the 2019 NYISO  Load & Capacity Report “Gold Book” Summary of Table III-3c 2018 Annual Net Energy Generation by Zone and Type to estimate 2018 energy generation by generator type and I have included a consolidated summary of data by fuel type.  Note that fossil fuel accounted for 41.9% of the generation in 2018.  Total load was 135,585 GWh so a 2040 doubled load will equal 271,170 GWh.

Estimating the feasibility for future development scenarios is difficult.  The Projected Net Energy Generation in 2040 table lists the assumptions I made to determine how New York generation could meet the doubled load.  There are three main components of the response: expand output from pumped storage and conventional hydro, maintain nuclear, and expand renewable combustion (landfill gas and municipal waste incineration); reduce load through energy efficiency, tighter codes & standards, behind-the meter (BTM) distributed solar and BTM non-solar distributed generation; and utility scale on-shore wind, off-shore wind and solar.

In the first component I estimated future generation as a function of fuel type.  The CLCPA law requires that all fossil-fired electric generation be replaced by 2040 so I zeroed out potential generation in those categories.  Although I have seen claims that additional hydro generation can be developed, I doubt that those resources could ever exceed a 10% increase over current levels.  In 2040 I estimated that only Nine Mile 2 nuclear would be operating because all the other nuclear generating stations would be over 60 years old.  The renewable combustion category includes internal combustion from landfill gas as well as wood and refuse fired steam generators.  I assumed that those resources could only increase 10% over current levels.  The result is that Component 1 can only cover 18.4% of the estimated load.

Reducing energy use through energy efficiency, improvements in codes and standards for energy use, behind the meter solar, and behind the meter non-solar distributed generation will reduce the amount of energy needed in Component 2.  Table I-b: Summary of NYCA Annual Energy Forecasts in the Gold Book lists GWh projections for these categories in 2040. The Gold book also includes tables that provide the net energy and the estimated capacity for each of these categories.  Since the publication of the 2019 Gold Book, the distributed solar deployment target was increased to 6,000 megawatts by 2025, up from 3,000 megawatts by 2023.  I adjusted my projection for that by assuming 6,000 MW in 2025, that observed energy per capacity would remain the same and that the annual increase after 2025 would be the same as the Gold Book projection.  Instead of 5,928 GWh from 4,525 MW of distributed BTM solar I estimate 9,847 GWh from 7,517 MW in 2040. The result is that Component 2 can only cover 13.6% of the estimated load.

The last component of the expected load is utility-scale wind and solar. On the basis of the assumptions for components 1 and 2, these resources will have to provide 68% of the load or 184,285 GWh.  In my opinion, this requirement will lead to a large increase in imported generation but if we assume that is not the case, this will require a massive buildout of large renewables.  In round numbers this buildout will require development of 9,214 GWh of generation output per year as shown in the Projected Renewable Energy Resources Needed to Meet Doubled Annual Electric Load in 2040 table.

In order to determine how much capacity will have to be developed to generate that much energy capacity factors have to be used.  In 2018, 1,739 MW of on-shore wind produced 3,985 GWh of net energy and 32 MW of utility solar produced 49 GWh of net energy.  Those observed values can be used to determine an on-shore wind capacity factor of 26% and a utility solar capacity factor of 17%.  I assumed those capacity factors for 2040.  Off-shore wind capacity will be developed by 2040 and I assumed a capacity factor of 50% based on the NYSERDA assumption for its recent awards. Using those capacity factors, I arbitrarily picked an annual development rate to meet the 9,214 GWh annual rate target.  I calculated that 1,050 MW of on-shore wind, 1,500 MW of utility solar and 1,050 MW of off-shore wind which will generate 9,293 GWh additional energy per year which exceeds the target development rate.

Sanity Check

One question is whether my arbitrary future renewable energy choices are reasonable.  I compare those numbers to an earlier wind analysis and NYS announced projects below.

The National Renewable Energy Laboratory sponsored the Eastern Wind Integration and Transmission Study (EWITS) and that report includes an estimate of future NY wind requirements.  The study was designed to examine the operational impact of up to 20% to 30% wind energy penetration on the bulk power system in the Eastern Interconnection of the United States. It included development of a database of wind resource and plant output data for the eastern United States.  One of the scenarios modeled 30% penetration of wind energy into the total energy consumed.  The scenario was labeled “Aggressive On- and Offshore” and required “a substantial amount of the higher quality wind resource in the NREL database” and noted that a “large amount of offshore generation is needed to reach the target energy level”.  The total wind capacity projected for the New York State Independent System Operator to meet a 30% penetration level was 23,167 MW of on-shore wind and 9,280 MW of off-shore wind.  The total on-shore wind capacity projected for the arbitrary choice is 20,000 MW and that compares well with the EWITS value, but the arbitrary choice for total off-shore wind is 23,000 MW more than double the EWITS value.

Another sanity check is to compare the annual generation development predicted with the queue of New York Projects.  In New York all utility scale projects over 25 MW are required to go through an Article 10 review process.  As part of that process all the projects in the review queue are tracked.  The New York State Department of Public Service Article Ten Project Queue table lists the MW capacity for projects summed by technology proposed categories and the year the permitting process began.  In the best year 1,478 MW if wind projects started permitting and in 2019 to date there have been 3,193 MW of solar and solar plus storage projects have started permitting.  Therefore, it appears that my guess that 1,050 MW of on-shore wind and 1,500 MW of utility solar development per year is possible.

Conclusion

The Brattle estimate for the additional energy needed to power all the electrification necessary to reduce CO2 emissions in New England by 80% is double the current load.  Thus, when New York State gets around to proposing how much energy might be necessary for the CLCPA, it is not unreasonable to expect that it will be at least two times the current amount.

At some point the intermittent and diffuse nature of renewables will have to be addressed.  Because renewables are intermittent, storage to cover light winds and low solar irradiance will be required.  Because they are diffuse transmission links to move the power as needed are also required.  As a result, the renewable generating facilities will not only have to replace existing fossil-fired dispatchable load but also provide support to the transmission system that is currently provided by those facilities.  I believe that will add to the amount of renewables needed.

When New York State gets around to proposing how much energy might be necessary for the CLCPA, it is not unreasonable to expect that it will be at least two times the current amount.  The wild guess estimate of the renewable capacity does not appear to exceed any feasibility threshold but there are some caveats.  For all components of the proposed response the problem of diminished returns on investment cannot be dismissed.  For example, it is not unreasonable to expect that the best on-shore wind development sites will be developed first and may in fact be already developed.  Consequently, later developments are going to have to go to sites with a lower potential wind resource so the capacity factor for future developments will be reduced.  The same problem should be expected for distributed solar, utility solar, and energy efficiency investments.

There are other issues of course.  First, and foremost, is cost.  There are also logistical issues developing this much renewable energy over this time frame.  Finally, note that given that the expected lifetime of these renewable projects is on the order of 20 to 25 years, there will be an eventual de-commissioning and re-development steady-state of constant investments required forever in the future.

Hay Harvest Climate Trend?

At a recent meeting I ran into Lois New who, before she retired, was the Director of the New York Department of Environmental Conservation’s Office of Climate Change.  I have known her for years and we worked together during RGGI stakeholder meetings. During our conversation she mentioned that her neighboring farmers were seeing the effects of climate change because they were having more trouble getting hay in before winter.  I said I thought it was more likely weather, she disagreed, and that ended the conversation.  This post looks at data to see if there is, in fact, a climatic trend for worse weather for haying.

First, let’s define weather and climate.  According to the National Oceanic and Atmospheric Administration’s National Ocean Service “Weather reflects short-term conditions of the atmosphere while climate is the average daily weather for an extended period of time at a certain location.”  The referenced article goes on to explain “Climate is what you expect, weather is what you get.”

New York State policy is all in that there is an imminent and inevitable climate change catastrophe that can only be averted if we do something.  In this case New York’s version of doing something is the Climate Leadership and Community Protection Act (CLCPA) which was enacted last summer.  It is described as “the most ambitious and comprehensive climate and clean energy legislation in the country”.  When Governor Cuomo signed the bill he said:

“The environment and climate change are the most critically important policy priorities we face.  They literally will determine the future – or the lack thereof. Even in today’s chaos of political pandering and hyperbole there are still facts, data and evidence – and climate change is an undeniable scientific fact.”

In order to rationalize these statements Governor Cuomo has a long history of attributing any observed unusual or extreme weather to climate change’s effects being seen today.

Paul Homewood at the Not a Lot of People Know That web page authored a couple of recent articles slamming NY Governor Cuomo for a couple of examples. The first article points out that his statement that “we did not use to have hurricanes, we did not have super storms, we did not have tornadoes” is dead wrong.  The second article entitled “Cuomo’s Fake Claims About Extreme Rainfall” noted that Cuomo implied in an MSNBC interview (referenced in the blog post) that extreme rainfall was getting much worse in NY State.  However, Homewood showed that “there is absolutely no evidence of that at all” at the long running Ithaca station or a New York City station.

As a member of the Governor’s inner climate circle Ms. New must have been a part of the public relations campaign justifying the CLCPA.  Her claim that farmers cannot complete harvesting hay because of climate change is entirely consistent.  This post will look at the facts, data, and evidence that there is a climate effect on haying.   Homewood referenced a link to precipitation data at http://climod2.nrcc.cornell.edu/ which is what I needed to do this analysis.

Approach

According to Mother Earth News in a Guide to Growing, Harvesting and Baling Hay “there are three steps involved in turning a green crop into what can rightfully be called hay:

      1. Cutting (followed by partial drying.)
      2. Windrowing (followed by further drying.)
      3. Baling hay or stacking hay.”

For our purposes the key is that freshly cut hay has to be dried because if hay is tied into tight bales when it still contains moisture it will go through a curing process that creates heat which can lead to self-combustion.  Therefore, farmers do not want to bale their cut crop until the moisture is less than 20%.  To do that the cut grass is allowed to dry for up to several days, then raked into rows and allowed to dry out most of the moisture.  Once dry then the hay can be baled.

The Northeast Regional Climate Center data sets provide processed values for temperature and precipitation for many observing sites in New York.  I chose to look at two stations with long records: Ithaca (1894-2019) and Mohonk House (1896-2019 with 1899 missing).  I downloaded the daily maximum, minimum, and average temperatures, precipitation amount, snowfall amount, snow depth and growing degree days.  A growing degree is the difference between average temperature in deg F and 50.  For example, if the average daily temperature is 60 deg F then there are ten daily growing degree days for that date.

I assumed that in order to harvest hay that the farmer would need to have at least four days when there was no precipitation greater than 0.05”.  Whenever that threshold was reached or exceeded the Harvest-Day parameter was set to one.  The total number of days that met this criterion in each month was summed along with the number of growing degree days per month for the growing season that I set as May through October.

I did a simple analysis of the two data sets.  I calculated the growing degrees and number of harvest days for each month in the growing season that I defined as May 1 to October 31.  I summed these values for the whole growing season.  I also summed values for the fall hay harvest season that I defined as August and September.  I fit a linear regression model to describe the relationship between growing degree days and hay harvest days by year for the whole growing season and just August and September. I use Statgraphics Centurion software from StatPoint Technologies, Inc. to do my statistical analyses because it enables the user to choose the best relationship from 27 different linear regression equations.  I determine which linear regression model provides the best fit and then use that model to describe the data. If the calculated probability value (P-value) is less than 0.05, there is a statistically significant relationship at the 95.0% confidence level and I defined the test result as significant. In addition, I calculated simple statistics to describe the two data sets.

Results

The purpose of this analysis is to evaluate the claim that farmers are having more trouble getting their hay harvested because of climate change.  If that were in fact the case then we would primarily expect to see a trend in decreasing hay harvest days and to a lesser extent a decrease in the number of growing degree days.  The linear regression statistical results for two sites over the entire growing season and August and September were evaluated.  Overall eight tests were done with the following results:

Mohonk Growing Season hay-harvesting days

      • Insignificant reduction in the number of hay-harvesting days over the growing season

Mohonk Growing Season growing degree days

      • Significant increase in the number of growing degree days over the growing season

Mohonk August and September hay-harvesting days

      • Significant increase in the number of hay-harvesting days in August and September

Mohonk August and September growing degree days

      • Significant increase in the number of growing degree days in August and September

Ithaca Growing Season hay-harvest days

      • Significant reduction in number of hay-harvesting days over the growing season

Ithaca Growing Season growing degree days

      • Significant reduction in growing degree days over the growing season

Ithaca August and September hay-harvest days

      • Insignificant reduction in number of hay-harvesting days in August and September

Ithaca August and September growing degree days

      • Significant reduction in growing degree days in August and September

Of the eight tests: three were consistent with the hypothesis that getting hay harvested is getting more difficult because hay-harvesting days and growing degree days decreased, three were inconsistent with the hypothesis because hay-harvesting days and growing degree days increased, and two tests were statistically insignificant.  Note that if growing degree days are going down that is an indication that temperatures are cooling and not warming.

Another way to look at the climate change impact is to statistically evaluate the data.  The Ithaca Hay Season Climatological Normal and Last 30 Year Dataand the Mohonk House Hay Season Climatological Normal and Last 30 Year Data tables include climatological normal data, the last 30 years of data, and summary statistics.

I believe that one way to check climate change claim effects is to check the 30-year average from the beginning of the period of record with the 30-year average at the end of the period of record.  I also believe that there are multi-year weather cycles of differing lengths and if that is the case then arbitrarily picking these two periods may not be representative.  Therefore, consider this comparison with caution. At Mohonk House the growing season hay harvesting days decreased by two days but the August and September days increase by three days.  The Ithaca growing season hay harvesting days decrease by five days but the August and September days stayed the same.

Importantly, these data also indicate that there is a lot of inter-annual variation in hay harvesting days.  The standard deviation of the August and September data at Ithaca was 6, the minimum was 2 and the maximum was 28. At Mohonk House the standard deviation was 8, the minimum was 0 and the maximum was 43.  The difference between the first 30-years and the last 30-years is less than the standard deviation variation.  Therefore, I don’t think it is reasonable for anyone to claim that they can discern a climatic trend.

So why is there an impression that hay harvesting is getting worse?  A quick review of the last 30 years of data indicates to me that this impression is consistent with weather variations.  At Mohonk House in 2014 and 2015 there were a couple of years that were great for hay harvesting with half the days (31) in August and September suitable.  Jump forward to the last two years and there were only 11 and 12 days suitable.  In that short time frame of reference an alarming trend is evident.  However, also recall that the average number of hay harvesting days is 18 and with a standard deviation of 8.  So even though  2018 and 2019 data suggest there may be a problem, the data are within one standard deviation of the mean which means that they are well within natural variation observed since 1896.

Conclusion

The points that I want readers to remember are that climate numerical analysis results are likely ambiguous, picking a climatic trend out of weather records is not simple, and, most importantly, any statistically significant trends are likely smaller than the observed inter-annual variation.   As a result, anecdotal claims of observed changes of weather parameters due to climate change are likely biased and unsubstantiated.

This data analysis shows ambiguous results.  It suggests that there is conflicting support for a climate-change induced problem with hay harvesting in August and September.  Mohonk House data indicate a statistically significant trend in more days suitable for harvesting hay whereas Ithaca data indicate a trend towards less days suitable for harvesting hay but the trend is insignificant.  At both stations there is a negative statistically significant trend in the number of growing degree days.  Depending upon your intent, statistics can “prove” an argument that there is a problem or there isn’t a problem.

In order to do a comprehensive analysis to settle the question would take a lot of work.  Before doing any more analysis work, the evaluation data used should be confirmed as appropriate.  In order to represent the New York region adequately, stations across New York and the region would all have to be analyzed similarly.  It might also be appropriate to look at each month when haying is done.

I think the comparison of possible trends against inter-annual variation is illuminating.  If there is a climate change signal the difference between the first thirty years of the records with the last thirty years should show changes.  What differences that do exist are smaller than the observed variations.  All changes are less than one standard deviation from the mean.  I believe that this is a consistent problem for lines of the so-called evidence for climate change impacts observed in New York.

Based on this analysis I believe that anecdotal claims of observed changes of problems with hay weather parameters due to climate change are likely biased and unsupported by the data.  The variations noted and ascribed to climate change are in fact due to weather.

I also think that similar analyses of other claims would provide similar results.  The Governor’s claim “Even in today’s chaos of political pandering and hyperbole there are still facts, data and evidence – and climate change is an undeniable scientific fact” and his tendency to blame any unusual weather on climate change are not supported by this analysis.  In my opinion, careful evaluation of data and evidence for most of his claims would find similarly ambiguous and less certain results.

RGGI Lessons to Date – November 2019 Edition

I recently had a simple version of this RGGI article published at Whats Up with That .  This article provides more details and considers other issues with the Regional Greenhouse Gas Initiative (RGGI).  The program is ten years old and has been touted as a successful example of a “cap and dividend” pollution control program and now it is being proposed as the model for a similar control program in the Transportation Control Initiative (TCI).  This post looks at the numbers to see if this praise is warranted and whether RGGI is a good model for the proposed TCI.  Ultimately the question is whether any cap and trade program for carbon dioxide (CO2) can be successful.

I have been involved in the RGGI program process since its inception.  I blog about the details of the RGGI program because very few seem to want to provide any criticisms of the program. I have extensive experience with air pollution control theory and implementation having worked every cap and trade program affecting electric generating facilities in New York including the Acid Rain Program, Regional Greenhouse Gas Initiative (RGGI) and several Nitrogen Oxide programs.  Note that my experience is exclusively on the industry side and the difference in perspective between affected sources trying to comply with the rules and economists opining about what they should be doing have important ramifications.  The opinions expressed in this post do not reflect the position of any of my previous employers or any other company I have been associated with, these comments are mine alone.

Background

RGGI is a market-based program to reduce greenhouse gas emissions. It is a cooperative effort among the states of Connecticut, Delaware, Maine, Maryland, Massachusetts, New Hampshire, New York, Rhode Island, and Vermont to cap and reduce CO2 emissions from the power sector.  According to a RGGI website: “The RGGI states issue CO2 allowances which are distributed almost entirely through regional auctions, resulting in proceeds for reinvestment in strategic energy and consumer programs. Programs funded with RGGI investments have spanned a wide range of consumers, providing benefits and improvements to private homes, local businesses, multi-family housing, industrial facilities, community buildings, retail customers, and more.”

The RGGI states developed a cap during a long stakeholder process that was based on historical operations and emissions.  The cap is the regional budget for CO2 emissions.  The Nine-State RGGI Region Emissions, Original RGGI Cap, and the Adjusted RGGI Cap table lists observed emissions and two caps.  The original cap was developed before the fracking revolution changed the cost of natural gas such that it became significantly cheaper than coal and residual oil.  After natural gas prices dropped so much the original projections for emissions were so out of tune to what was happening the RGGI states developed an adjusted cap to account for that development.

In order to determine if RGGI is successful and a program to emulate let’s define some metrics.  The primary goal of the program is to reduce greenhouse gas emissions (GHG) from the electric generation sector so quantifying the emissions change from before the program to the present is a key metric.  Another appropriate metric is cost efficiency per ton of CO2 reduced compared to the Social Cost of Carbon (SCC).  This parameter is an estimate of the economic damages from emitting a ton of CO2 and is widely used to justify GHG programs.  For a comparison metric I will ignore issues with this parameter even though I agree with the following by Paul Driessen and Roger Bezdek: “The SCC assumes fossil-fuel-driven carbon dioxide emissions are causing dangerous manmade climate change, and blames U.S. emissions for every conceivable climate-related cost worldwide. But it fails even to mention, much less analyze, the tremendous and obvious benefits of using oil, gas and coal to power modern civilization.”

In addition to these metrics we have to look at lessons learned and considerations that are not yet resolved to determine whether RGGI is a good model for future control programs.  I address the following: the theory and reality of historical trading programs, control options for affected sources, allowance management, allowance ownership and allowance costs.  Most of these issues were not discussed in the What’s Up with That post.

Theory and Practice

I agree with the argument that economic incentives or market trading programs reduced emissions more cost-effectively than a command and control program.  However, it is instructive to look at the reasons why emission reductions occurred because the theory does not necessarily drive the observed reductions.  It is necessary to review historical performance of RGGI to determine why CO2 emissions reductions have occurred so we can reasonably expect a similar result in other applications like the TCI..

Let me first point out that there is a fundamental difference between the way affected sources operate in emissions trading markets and the way economic theory predicts they should operate.  I believe that electric generating affected source allowance management is different than theory because the affected sources do not treat allowances as a storable commodity or a financial asset in the usual sense of the term.  Instead allowance management is overwhelmingly driven by regulatory requirements for the current compliance period. i.e., do I have enough allowances to cover expected emissions?  Financially it is simply another cost of operating and not a potential profit center.  The important difference is that the academic economic theory holds that affected sources are looking years into the future, but in reality, there is no such long-term time horizon for affected sources.  Their decision on the quantity of allowances to buy is driven by their expected operations in the period between auctions and, at most, the entire compliance period.  Also note that most companies include a small margin for operational variations and regulatory compliance considerations. Because of the differences between the way affected sources operate and the way economic theory says they should operate, I have little faith in the models that predict future allowance margins.

The Acid Rain Program (ARP) was undoubtedly a successful program because it lowered emissions more than expected at far lower costs than predicted.  This program was open and transparent so all emissions and allowance data are available.  In order to meet the initial emission cap target of a 50% reduction, affected sources were awarded half of their historical emissions.  Although it is common practice to vilify this program for giving away the allowances for free the rationale is still valid today. The concept for the acid rain program was that power plants would install SO2 control equipment and if they over-controlled their emissions, they could sell the excess allowances earned to other facilities that could not install the control equipment as cost-effectively.  This approach incentivizes over-control because affect sources can subsidize control equipment investments they made by selling excess allowances.  This cost reduction efficiency brings down overall costs.   It turned out that fuel switching and technological improvements were so effective that far greater than expected reductions occurred.  Fuel switching occurred because technology to burn lower sulfur coal was developed and railroad de-regulation opened the market to transporting coal cost-effectively over very long distances. Another subtle point is that the ARP allowance bank was earned, that is to say excess allowances in the bank represent over-controlling emissions lower than the cap limits.

The Regional Greenhouse Gas Initiative (RGG) is a cap-and-invest program that has been touted as a model for a TCI cap-and-invest trading program because of its “success”.  Although the RGGI states claim that the program is open and transparent the fact is that there is no allowance ownership information available.  There is no question that CO2 emissions have come down in the RGGI states since the inception of the program but it is important to determine why they have come down.  I will address that point later on.  There is a fundamental difference in the way that affected sources treated RGGI as opposed to ARP, namely ARP was considered a control program and RGGI was considered a tax.  Because there are no cost-effective add-on controls for CO2 at existing power plants there are limited options to meet the cap.  Because allowances all have to be purchased and the incremental cost was low plant control programs to reduce CO2 through efficiency were not implemented.  The allowance bank does not represent earned reductions below the cap limits.  Instead the bank is made up of allowances purchased at auctions and on the market.  The RGGI states in their program reviews were very concerned that the allowance bank was large and have taken steps to adjust the allowances sold at auction to force the bank smaller.  In the naïve belief that RGGI investments significantly reduced emissions the RGGI states have also reduced the cap going forward.  As a result. RGGI going forward is going to be significantly different that RGGI in the past and that has ramification on its value as a model for TCI or any other future emissions trading program.

I noted above the distinction between the ARP “earned” allowance bank and the RGGI allowance bank.  Because the ARP affected sources over-controlled emissions below their cap levels, they earned the allowance bank.  That means the bank represents surplus allowances that are not needed for compliance so it does not matter who owns them.  On the other hand, all RGGI allowances were purchased at one time or another by anyone who offered a high enough price at an auction or on the market.  Because allowance ownership is not transparent, we only know the number of allowances owned in the following three categories:

      • Compliance-oriented entities are compliance entities that appear to acquire and hold allowances primarily to satisfy their compliance obligations.
      • Investors with Compliance Obligations are firms that have compliance obligations but which hold a number of allowances that exceeds their estimated compliance obligations by a margin suggesting they also buy for re-sale or some other investment purpose. These firms often transfer significant quantities of allowances to unaffiliated firms.
      • Investors without Compliance Obligations are firms without any compliance obligations.

To this point in 2019 the affected sources with compliance obligations have been able to get the allowances needed to cover their emissions from auctions and the market.  However, at some point going forward this will change and it will make a difference.  I have addressed the status of RGGI emissions and allowances elsewhere but briefly because the allowance cap is being reduced so much, the affected sources are going to have to go to the investors without compliance obligations.  This is uncharted territory and, at a minimum, I expect that the allowance prices will spike upwards.  Note that this price spike provides no dividends for CO2 reduction investments because the dividends are earned at the initial sale.  But it could be even worse if the entities without compliance obligations withhold allowances and create a shortage such that affected sources do not have enough allowances to run.

RGGI supporters who claim it is successful point to emission reductions of 40 to 50%.  In order to evaluate the RGGI emissions reduction claims I used data from the Environmental Protection Agency Clean Air Markets Division air markets program website.  Emissions data from the electric generating unit (EGU) sector are available from before RGGI started to the present, so I downloaded all the EGU data for the nine states currently in RGGI from 2006 until 2018.  In order to establish a baseline, I calculated the average of three years before the program started.  As shown in the RGGI Nine-State EPA CAMD Annual CO2 Emissions table the total emissions have decreased from over 127 million tons prior to the program to just under 75 million tons in 2018, for over a 40% decrease.  Note that these numbers are slightly different than the previous table because different sets of sources are used.

However, when you evaluate emissions by the primary fuel type burned it is obvious that emissions reductions from coal and oil generating are the primary reason why the emissions decreased.  Note that both coal and oil emissions have dropped over 80% since the baseline.  Natural gas increased but not nearly as much.  I believe that the fuel switch from coal and oil to natural gas occurred because natural gas was the cheaper fuel and had very little to do with RGGI because the CO2 allowance cost adder to the plant’s operating costs was relatively small.   There is no evidence that any affected source in RGGI installed add-on controls to reduce their CO2 emissions.  The only other option at a power plant is to become more efficient and burn less fuel.  However, because fuel costs are the biggest driver for operational costs that means efficiency projects to reduce fuel use means have always been considered by these sources.   Because the cost adder of the RGGI carbon price was relatively small I do not believe that any affected source installed an efficiency project as part of its RGGI compliance strategy.

As a result, the only reductions from RGGI that can be traced to the program are the reductions that result from direct investments of the RGGI auction proceeds. Information necessary to evaluate the performance of the RGGI investments is provided in the RGGI annual Investments of Proceeds update.  In order to determine reduction efficiency, I had to sum the values in the previous reports because the most recent report only reported lifetime benefits.  In order to account for future emission reductions against historical levels the annual reduction parameter must be used.  The Accumulated Annual Regional Greenhouse Gas Initiative Benefits table lists the sum of the annual avoided CO2 emissions generated by the RGGI investments from three previous reports.  The total of the annual reductions is 2,818,775 tons while the difference between the baseline of 2006 to 2008 compared to 2017 emissions is 59,508,436 tons.  The RGGI investments are only directly responsible for less than 5% of the total observed reductions!

In order to argue that RGGI emission reduction programs are a good investment relative to the expected societal cost of CO2 emissions the Social Cost of Carbon (SCC) parameter can be used.  SCC values range widely depending on assumptions, but if you use a discount rate of 3% and consider global benefits like the Obama-era Environmental Protection Agency (EPA) did then the 2020 SCC value is $50.  The Accumulated Annual Regional Greenhouse Gas Initiative Benefits table lists the data needed to calculate the RGGI CO2 reduction cost per ton.  From the start of the program in 2009 through 2017 RGGI has invested $2,527,635,414 and reduced annual CO2 emissions 2,818,775 tons.  The result, $897 per ton reduced, is 18 times than the current EPA SCC value for United States benefits.

There is another key lesson from RGGI that applies to any CO2 emissions marketing control program. There is an important difference between cap and trade programs for SO2 and nitrogen oxides (NOx) emissions and cap and invest programs for GHG emissions.  There are add-on control options for SO2 an NOx whereas there isn’t any cost-effective option for CO2.  In the ARP the affected sources could directly control their compliance.  In RGGI there were limited direct options for the affected sources and, going forward especially, they are going to have to rely on indirect reductions, i.e., someone will build a zero-emitting plant that displaces enough output from a fossil plant that enough allowances are available to cover the affected source requirements.  The ultimate control strategy for a emissions marketing CO2 control program is to run less and hope power is available from somebody else.

Conclusion

I believe that RGGI is not the success that its adherents believe. Based on the numbers there are some important caveats to the simplistic comparison of before and after emissions.  Fuel switching was the most effective driver of emissions reductions since the inception of RGGI.  Emission reductions from direct RGGI investments were only responsible for 5% of the observed reductions.  RGGI investments in emission reductions were not efficient at $897 per ton of CO2 removed.  In my opinion those are not the hallmarks of a successful program.

I want to highlight a related point.  In order to determine emission reduction efficiency from the RGGI investment reports, I had to sum the values in the previous reports because the most recent report only reported lifetime benefits.  The RGGI website only lists the lifetime benefits of RGGI investments in 2017 but those parameters are useless for the most obvious application.  In order to account for future emission reductions against historical levels the annual reduction parameter must be used.  It is hard to not believe that excluding the accumulated annual reductions was deliberate because the numbers are so poor.

As a model for future programs, RGGI successfully proved that a regional entity could implement a cap and auction program.  However, the actual cause of observed reductions and ability of affected sources to make the reductions proposed should be considered before other programs adopt the RGGI model.  I considered the use of the RGGI model in Transportation Climate Initiative Draft Framework Cap and Invest Caiazza Comments  that were submitted as part of their stakeholder process.  I concluded that there are so many differences between a program for mobile sources and electric generating units that simply implementing a tax and investing the proceeds as proposed would be less likely to have serious problems with unintended consequences and unanticipated issues.

As a result of the issues raised in this post, I believe that it is fair to ask whether any cap and trade program for CO2 can be successful if the ultimate goal is a significant reduction in emissions.  Because CO2 from fossil fuels is such an integral part of our lifestyles a large reduction in emissions is going to have to require changes in lifestyles.  Therefore, the question becomes will people accept lifestyle changes such as giving up the gas automobile with all its current advantages over any alternative as a result of indirect CO2 pricing?