New York’s energy planning process continues its efforts to meet the aggressive goals of a reformed energy system that relies on renewable energy. The latest boondoggle in that effort is a plan to price carbon in the wholesale electric market. I have been following the process and submitting comments as an unaffiliated public party to the process. This post describes a vivid example of the difficulties of implementing economic theory related to carbon dioxide reduction programs.
I am motivated to submit comments in this process so that there is at least one voice of the unaffiliated public whose primary interest is an evidence-based balance between environmental goals and costs to ratepayers. There are significant hurdles to implementing carbon pricing in general and as proposed in the straw proposal that should be considered by the Integrating Public Policy Task Force (IPPTF). The questions in these comments are related to the total costs of the program.
This post is based on comments submitted as a private retired citizen. They 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. The majority of New York State (NYS) ratepayers are unaware of the ramifications of this proceeding or have any idea of the ramifications of incorporating the cost of carbon emissions into New York State (NYS) wholesale electricity markets. The basis of this initiative is a Brattle Group analysis that outlined a scheme to incorporate a state policy defined cost of carbon in the wholesale market to improve the overall efficiency of the New York Independent System Operator (NYISO) energy and capacity markets.
As of the date of this post one thing that has been conspicuously absent from the discussions has been the total expected cost. In my latest submitted comments I argued that it would be beneficial for all stakeholders to have the NYISO provide an analysis of historical data that shows what would have happened to the markets if the carbon price were in effect. I illustrated the problem estimating this cost by considering one historical hour. It appears that there is a significant overlooked component to this initiative. One feature of a carbon price scheme is usually revenue neutrality where the carbon costs are returned to the consumers to make it less regressive. However, in the New York State wholesale electric market case it looks like in addition to the carbon price itself there will be a general increase in market clearing prices. There is no mechanism to make that component revenue neutral.
In order to evaluate the emissions data I obtained data from the United States Environmental Protection Agency Clean Air Markets Division (CAMD) website for July 19, 2017 at hour 17. The CAMD website has hourly data for all emissions sources affected by national emissions trading programs. There are significant limitations to the data for this application but they should be indicative of the situation. I manually added the NYISO electric load control zones, made some guesses about whether some small units should be included or not, estimated CO2 emissions at some sources that are not required to provide that data, and revised some inconsistent numbers. There also is an inherent flaw in this approach because the EPA data set includes gross load whereas the NYISO loads are net loads. Also note that I excluded combined heat and power units and steam units from these calculations. My data are available upon request and the submitted comments describe the methodology in more detail.
The NYISO manages the state’s power grid and de-regulated energy market. In order to understand the implications of carbon pricing on New York electricity market costs some background information is necessary. There is an overview of the price setting approach used in the NYISO document NYISO Markets:
The energy market provides a mechanism for Market Participants to buy and sell energy at prices established through a competitive auction process designed to meet energy demands, or “loads,” with the least-cost resources available; or, through contractual, bilateral transactions where quantities and prices are arranged directly between wholesale suppliers and “load-serving entities” (LSEs) such as utilities. For energy purchases arranged through the NYISO’s auctions, the NYISO administers day-ahead and real-time auctions, resulting in a two-settlement process that sets the price of energy based on market and grid conditions at specific times. Further, the NYISO’s auctions reflect geographic conditions , establishing “Locational Based Marginal Prices” (LBMP) for energy that reflect local demand and supply conditions as well as any constraints that may exist when moving energy across the grid to meet demand. The first settlement is based upon the day-ahead bids and the corresponding schedule and prices, or day-ahead commitment. The second settlement is based upon the real-time bids and the corresponding real-time commitment (RTC) and real-time dispatch (RTD). Market Participants may participate in the DAM and/or the real-time market. Roughly 94% of energy is scheduled in the day-ahead market, while the remaining 6% is accounted for in the real-time market. About 40% of the energy settled in the day-ahead market is scheduled through bilateral contracts.
The NYISO Zone CO2 Cost July 19 2017 at hour 17 table lists the gross load, heat input, CO2 mass, and CO2 rate in lbs per mmBtu and tons per MWhr for the entire state, by LBMP zone and aggregating Downstate and Upstate zones. The source data show that the hourly CO2 emissions range from 681 tons per hour at the remaining coal plant to 1.2 tons for a partial operating hour at a natural-fired turbine. More importantly the CO2 emission rate (lbs/mmBtu) data only lists three general emission rates corresponding to natural gas, oil, and coal fuels. If the results for this hour are generally consistent throughout the year then the efficacy of this program to lower CO2 emissions is questionable. There are slight differences within these rate categories but there are relatively minor. The New York Department of Environmental Conservation recently announced a new regulation that will for all intents and purposes ban the future use of coal so this program cannot be expected to shut down the use of coal. The oil generating units do not burn oil for economic reasons so this program cannot be expected to change the use of those units relative to natural gas units. The difference in CO2 emission rate for the natural gas units is so small that this program cannot be expected to lead to the use of lower emitting units. Therefore, this program will not likely cause fuel switching due to the price of carbon.
According to the NYSERDA Patterns and Trends report, in 2014 the electric sector CO2 emission rate was 39,406,671 tons per year. If the carbon price is $50 per ton then we can expect this program to generate a minimum of over $1.5 billion dollars per year. The hourly carbon price based only on emissions ($50 per ton times the total tons in the previous table) gives a state-wide cost of $440,373 with $173,995allocated Upstate and $266,978 allocated Downstate.
The Brattle report proposed that the only impact to consumers would be directly related to the carbon price. However, the NYISO has not done an analysis of the potential impact of the carbon price on the wholesale electric market to determine if there could be a general increase in market clearing prices. If that is the case then the consumers will be paying a whole lot more than just the carbon price and there will be no way to even to try to make any extra costs revenue neutral.
I used the hourly data to estimate LBMP zone costs in theNYISO Zone CO2 Cost July 19 2017 at hour 17 table. I assumed that the zone cost equals the total load times the maximum CO2 rate (tons per MWhr) times the Social Cost of Carbon (Tab “LBMP”). Because of the magnitude of the carbon price I also assumed that the price of carbon sets the price of the most expensive unit in the zone. If that presumption is correct then the results are far different than the example estimate simply multiplying emissions by the cost of carbon. The total statewide cost is $773,644 and the Upstate portion is $209,394 and Downstate is $564,251. Note that most of the additional cost is due to a $306,048 increase Downstate because the Upstate cost only increases $35,999.
Based on this example I believe it is necessary and appropriate for the NYISO to provide estimates of the expected historic market response to the carbon price for an entire year based on hourly LBMP values. The NYISO knows the marginal economic unit and can use the USEPA data to show the marginal and maximum emission rates, CO2 mass/MWH and CO2 mass/mmBtu. At the proposed price of carbon that analysis could determine what would happen to the LBMP prices.
In addition to the financial impacts we can estimate what kind of impacts the carbon price will have on generation patterns. Based on the CO2 rates in the example hour it appears that we will find very small shifts in the marginal economic unit. Only when we have annual results can we verify whether this proposed program will have any effect on carbon emissions.
Finally, the analysis I recommend will not only estimate how the carbon price will affect LBMP prices but also provide information about where those revenues end up. If my assumption that the LBMP prices are based on the maximum emission rate but the residual that goes back to the consumers is based on the actual rates for each generator then the only facility that fully pays its residual is the maximum emission rate unit. All the other units contribute less to the consumer. The NYISO should provide the analysis so that we can determine what portions of the LBMP price increases remain with which generator sectors and what residuals could be returned to the LSEs. Finally, we can estimate the portion of LBMPs that could be credited to new renewables.