Update June 30, 2020: I wrote a layman’s summary on this issue here.
Physicians, Scientists, and Engineers (PSE) for Healthy Energy is a multidisciplinary, nonprofit research institute that studies the way energy production and use impact public health and the environment. One of their recent programs is the Energy Storage Peaker Plant Replacement Project. That work formed much of the technical basis for the PEAK Coalition report entitled: “Dirty Energy, Big Money”. I have prepared two posts on that document (here and here). This post addresses the PSE report Opportunities for Replacing Peaker Plants with Energy Storage in New York State.
I am a retired electric utility meteorologist with nearly 40 years-experience analyzing the effects of meteorology on electric operations. I have been involved with New York peaking power plants in particular for over 20 years from a compliance reporting and operations standpoint and also evaluated impacts and options for this kind of source. This background served me well preparing this post. 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.
Energy Storage Peaker Plant Replacement Project
PSE defines the alleged problem in the introduction to this project as follows:
The United States relies on more than 1,000 natural gas- and oil-fired peaker power plants across the country to meet infrequent peaks in electricity demand. These peaker plants tend to be more expensive and inefficient to run for every megawatt-hour generated than baseload natural gas plants and emit higher rates of carbon dioxide and health-harming criteria air pollutants. Peaker plants are also typically disproportionately located in disadvantaged communities, where vulnerable populations already experience high levels of health and environmental burdens.
The text for the New York specific report describes the problem similarly:
Across New York, 49 oil- and gas-fired peaker power plants and peaking units at larger plants help meet statewide peak electric demand. These include both combustion turbines designed to ramp quickly to meet peak demand, and aging steam turbines now used infrequently to meet peak needs. More than a third of New York’s peaker plants burn primarily oil, and three-quarters are over 30 years old resulting in numerous inefficient plants with high rates of greenhouse gas and criteria pollutant emissions for every unit of electricity generated. Some of these plants are in very urban areas: ten plants have more than a million people living within three miles. One-third of the plants are located in areas the state considers to be environmental justice communities, where vulnerable populations typically already experience high levels of health and environmental burdens. New York has set energy storage targets and recently designed peaker plant emission reduction targets, providing an opportunity to replace inefficient, high-emitting peaker plants in vulnerable communities throughout the state with energy storage and solar.
Their proposed solution:
Renewable energy and energy storage systems are beginning to emerge as competitive replacements for this fossil fuel infrastructure. Simultaneously, numerous states across the country are designing incentives and targets to support energy storage deployment. Together, these developments provide a unique opportunity to use energy storage to strategically displace some of the most polluting peaker power plants on the grid.
In the Energy Storage Peaker Plant Replacement Project PSE did a screening analysis across nine states that identified peaker power plants that “may be prime candidates for replacement based on operational and grid characteristics, and whose replacement may yield the greatest health, environment and equity co-benefits”. They claim that their approach “aligns state efforts to adopt energy storage with environmental and societal goals”.
The following section is the summary of the report. Based on my review of the New York State-specific results I believe further study is needed to actually determine if all the peaker units identified can actually be considered candidates for replacement with energy storage and solar. I also worry that the PSE analysis is mis-leading inasmuch as it does not address the fact that peaker plants fulfill niche operational backup roles that vary widely across the country. I am familiar with New York State peaker plants and will show why that is important in New York.
“The majority of New York’s peaker plants are located in densely urban areas in New York City and Manhattan, a region that is in non-attainment for federal ozone standards. These include old, inefficient and oil-burning units near populations that experience high cumulative environmental health and socioeconomic burdens. The state’s new emission reduction standards for nitrogen oxides, along with its energy storage deployment goals, provide a clear opportunity to target inefficient and polluting facilities for replacement with cleaner alternatives, particularly in urban areas. In the attached table, we provide operational, environmental and demographic data for New York peakers and nearby populations. Indicators such as nearby population, emission rates, heat rate (a measure of efifciency), operation on poor air quality days, capacity factor, typical run hours, and location in an environmental justice community or in an import-constrained load zones downstate can help inform whether a given plant might be a good target for replacement with storage, solar+storage, demand response, or other clean energy alternatives. These data should be accompanied by engagement with accompanied by engagement with affected communities to determine replacement priorities and strategies.”
The New York report has four sections: New York State Policy and Regulatory Environment, New York State Peaker Plants, Nearby Populations, and Emissions and the Environment. I will address those sections in the following.
New York State Policy and Regulatory Environment
There isn’t much to comment on in this section. The PSE report only describes New York’s climate initiatives. Although the summary notes that New York has new emission reduction standards for nitrogen oxides, it does not highlight the fact that the regulation was specifically intended to address emissions from the old, inefficient simple cycle combustion turbines in New York City. I described New York’s specific initiatives in my background post on the PEAK Coalition Dirty Energy, Big Money report.
New York State Peaker Plants
This analysis and report were intended to provide background information to support “clean energy alternatives” for peaker plants. A primary component of that information is identification of peaker plants. The technical documentation describes peaker power plants and the selection criteria used in their screening analysis. PSE states “The phrase peaker plant commonly refers to fossil fuel-burning power generation used to meet peak demand on the electric grid, but the term itself does not have a precise definition”. Actually, for EPA reporting purposes there is an exact, regulatory definition. 40 CFR Part 75 §72.2, states that a combustion unit is a peaking unit if it has an average annual capacity factor of 10.0 percent or less over the past three years and an annual capacity factor of 20.0 percent or less in each of those three years.
PSE chose to select peaking power plants based on the following criteria: fuel type: oil & natural gas; Capacity: ≥ 5 MW; capacity factor: ≤15% (3-yr. avg.); unit technology type: simple cycle combustion turbine, steam turbine & internal combustion; application: entire peaker plants & peaking units at larger plants; and status: existing and proposed units. Relative to the peaking power plants subject to EPA reporting requirements, the biggest difference is that the PSE criteria selects small units between 5 and 15 MW that are so small that their emissions and impacts are generally considered insignificant. Those facilities do not report continuous emissions monitoring data that the units >15 MW do.
Briefly, PSE collected data from EPA and EIA then screened it with their criteria to identify peaking units. They calculated operational and emissions data. Then they compared operational data with ambient monitoring data and found periods when the peaking units operated during periods of high ambient levels. This is a straight-forward number crunching exercise and I have no comments on the methodology.
The technical and policy documentation for the Energy Storage peaker plant replacement project includes a section titled “Grid requirements: transmission constraints and capacity needs” that includes a discussion of New York. For the most part PSE relied on the New York Independent System Operator analyses of the peaker plants. They note that the impacts of removing capacity is highly location dependent quoting NYISO reports: “lower amounts of capacity removal are likely to result in reliability issues at specific transmission locations” and that NYISO did not “attempt to assess a comprehensive set of potential scenarios that might arise from specific unit retirements”. Despite the fact that NYISO cannot make specific recommendations PSE goes ahead and makes recommendations for five plants in New York City and five plants on Long Island that are “replacement opportunities” in PSE Peaker Documentation Table 5.3.
While I am certainly no expert on New York City reliability requirements I believe that there are ramifications not considered by PSE. The NYISO Gold Book Data for Table 5.3 Replacement Opportunities table provides additional data for the PSE opportunities. First note that PSE did not identify peaking units that operate at facilities with other units. There is a combustion turbine at Northport and Arthur Kill that operates with the capacity factor listed. PSE apparently does not understand that the primary purpose of those units is for black starts, that is to say when they provide power necessary to start the steam turbine units when there is no off-site power available. In theory battery storage could be used for that but because of reliability considerations the battery would have to always be kept with enough energy to start the plant for the very rare occasion when there is a blackout. There is no way that could be cost-effective. My table also lists the fuel burned and it is instructive that all but one of the units listed can burn kerosene or number 2 fuel oil. There are specific requirements for minimum oil burning when there is a possibility that the gas supply could be cut off. Because this is not the standard peaking power plant replacement scenario, I am not sure whether battery storage would be cost-effective for this requirement.
Advocates for “clean energy alternatives” point out that New York has a law that requires that no electricity will be generated by fossil fuels in 2040. Until such time that the State has a plan to meet that goal that explains how reliability and affordability can be maintained, then I will continue to believe that meeting that aspirational goal is more than simply a matter of political will. For example, the Gowanus power plant has a nameplate capacity of around 540 MW. For all the Article Ten solar energy applications currently in the queue 5.4 acres per MW was the lowest spatial requirement. That means that solar panels totaling at least 4.6 square miles will be needed to replace this source of in-city generation. While that may be possible, there are a host of logistical issues starting with the need to provide the power where it is needed when it is needed. New York City is a load pocket relative to the rest of the grid but there are numerous smaller load pockets within the city.
The report notes that “Ten of the New York peaker plants each have more than a million people living within a three-mile radius. The most urban plants tend to also be in relatively low-income, minority communities, due to both the location of some facilities in low-income, environmentally overburdened communities of color.” In my background post on the PEAK Coalition Dirty Energy, Big Money report I described the environmental justice concept of dis-proportionate impacts. I do not know how to deal with dis-proportionate impacts when the location of some facilities impact rich communities at the same time they impact low-income communities.
PSE developed a “cumulative vulnerability index that integrates data on health burdens (asthma, heart attacks, premature birth rates); environmental burdens (ozone, particulate matter, toxics, traffic proximity, lead paint, and hazardous facilities); and demographic indicators (low-income, minority, linguistically isolated, and non-high school-educated populations)”. It is vital to determine the effect of the peaker power plants relative to all the other impacts on the admittedly over-burdened environmental justice communities.
Emissions and the Environment – Air Quality Impacts
In order to determine the relative impact of peak power plants we have to consider their air quality impacts. In order to be permitted to operate, all power plants have to evaluate the potential impacts of their emissions relative to the National Ambient Air Quality Standards (NAAQS). There are two types of standards. Primary standards provide public health protection, including protecting the health of “sensitive” populations such as asthmatics, children, and the elderly. Secondary standards provide public welfare protection, including protection against decreased visibility and damage to animals, crops, vegetation, and buildings. Air quality models combine information about the emissions, operating characteristics, and meteorological conditions to estimate the ambient concentrations from the power plants and those estimates are compared to the NAAQS. If the contribution from the facility directly causes an exceedance of any NAAQS limit then the plant cannot operate until changes are made to reduce the impact. If nothing can be done to reduce the impacts lower than the limits then it cannot be permitted to operate.
The air quality modeling used to permit a power plant to operate considers pollutants like sulfur dioxide and nitrogen dioxide that are directly emitted by the plant. Power plants also emit pollutants that are precursors to other pollutants that form in secondary reactions. Modeling secondary pollutants is more complicated and ascribing the impacts of particular facilities on air quality is more difficult. Permit conditions for secondary pollutants such as ozone and inhalable particulate matter can also limit emissions of the precursor pollutants.
In my opinion the most difficult air quality issue today is ozone attainment because the emission source characteristics and meteorological conditions are not only complex and difficult to understand but also because making the reductions necessary are costly and impactful. Ground-level ozone is created by chemical reactions between oxides of nitrogen (NOx) and volatile organic compounds (VOC). Those pollutants are emitted by cars, power plants, industrial boilers, refineries, chemical plants, natural sources and other man-made sources and when they chemically react in the presence of sunlight, they create ozone. Ozone is most likely to reach unhealthy levels on hot sunny days in urban environments but because NOx and VOC as well as ozone can be transported long distances by wind, rural areas are affected and urban areas are affected by sources far upwind.
It has been observed that when widespread transportation restrictions are implemented (e.g. during the Atlanta Olympics) that there is a marked improvement in ozone levels. However, the fact is that there is little societal desire to maintain the draconian restrictions of automobile use that produce those improvements. For peaking power plants, the problem is that the conditions most conducive to create ozone are also the hot and muggy conditions that increase electricity demand for cooling, so the peak load of electric generation produces the most emissions at the worst time. However, in order to provide the power necessary to keep the lights on when people really want and need it, the existing power grid has peaking facilities. In my second post on the PEAK coalition report I described the process used to determine if these units are needed and why I think they have not been replaced yet.
Recall that PSE developed a cumulative vulnerability index that integrates data on environmental burdens including ozone and particulate matter. The point of this entire discussion is that ozone is a secondary air pollutant and the vast majority of ambient PM2.5 from power plants is also a secondary pollutant. In other words, there is a lag between the time of relevant emissions and creation of either ozone or PM2.5. As a result, the accused peaking power plants do not create the air quality impact problems alleged to occur to the environmental justice communities near the peaking power plants. In fact, because NOx scavenges ozone the peaker plants reduce local ozone if they have any effect at all.
The PSE report notes that “These data should be accompanied by engagement with accompanied by engagement with affected communities to determine replacement priorities and strategies.” I do not want anyone to misunderstand that I am not arguing that something should not be done about New York City’s simple cycle combustion turbine peaking power plants. They are old, inefficient and relatively dirty. However, in order to do the right thing, we need to understand all the background information. The PSE analyses and the PEAK Coalition vilification of fossil-fired power plants only tells one side of the story and, inasmuch as most of the alleged environmental impacts are based on ozone and PM2.5 impacts, they misleadingly imply much more of an environmental benefit to the affected communities than will actually occur if the existing power plants are replaced by the latest generation of natural-gas fired power plants.
As noted in my post on the feasibility of the “clean energy alternative”, I have reservations about that proposed solution. Even though the cost for developing renewable energy resources is allegedly cheaper than the cost of equivalent fossil-fired energy resources, the cost to ensure that electricity is available when and where it is needed for the two resources are not even close. Because renewable energy is intermittent energy storage is required and my feasibility post demonstrated those costs are immense and would have to drop by an order of magnitude to make the solar+storage option comparable in cost.
The PSE report Opportunities for Replacing Peaker Plants with Energy Storage in New York State includes a table that lists all the power plants in New York State that meet their screening criteria defining a peak plant. The title of the report suggests that this list contains facilities that could be replaced by energy storage. However, it includes steam turbine units that burn residual oil. Because those units burn an expensive fuel, they don’t run much but because their operating costs are relatively low, they can be kept available for the rare occasions when they are needed. I was working at one of the named Upstate plants, Oswego, when the 2003 Northeast blackout occurred. When the transmission system lost power, three nuclear units nine miles east of the plant had to shut down. In order to bring the system back on-line, both of Oswego’s 850 MW units were turned on, ran for a combined 231 hours and generated 71,684 MWh. I cannot think of any scenario where it would be in the best interest of New York to build enough energy storage to replace the Oswego power plant for this type of incident.