New York City Energy Storage Peaking Turbine Replacement

The biggest air quality issue in New York State is compliance with the National Ambient Air Quality Standard for Ozone. In order to meet that limit the New York State of Department of Environmental Conservation (DEC) proposed regulations earlier this year to lower allowable nitrogen oxide (NOx) emissions from simple cycle and regenerative combustion turbines during the ozone season. The problem is that these turbines are needed to keep the lights on during periods when needed most so replacement is not very simple as I explained in an earlier post.

This post describes the State’s evaluation of the politically correct alternative, energy storage, to provide the power generated by these turbines. This post addresses the report findings for turbines that could be directly replaced by energy storage. I want to emphasize that the following represents my opinion and not the opinion of any of my previous employers or any other company with which I have been associated. I have been following the operational implications of these turbines and their effect on ozone for over 20 years.


The evaluation of using energy storage to replace these peaking units is part of the New York State Energy Storage Roadmap announced by Governor Cuomo in June 2018. As part of that effort the Department of Public Service (DPS) established an Energy Storage Deployment Program. On July 1, 2019, Energy Storage Deployment Program Report – Unit by Unit Peaker Study was submitted to the docket for Case 18‐E‐0130 – In the Matter of Energy Storage Deployment Program. DPS staff, working with New York State Energy Research and Development Authority (NYSERDA), Long Island Power Authority (LIPA), New York Independent System Operator (NYISO), NY Department of Environmental Conservation (DEC), Con Edison, and consulting firm Energy and Environmental Economics, Inc. (E3) prepared the report. The DPS December 12, 2018 Order Establishing Energy Storage Goal and Deployment Policy directed them to develop a unit‐by‐unit operational and emission profile study and methodology to determine which downstate peaking power plant generating units are potential candidates for repowering or replacement. My previous post describes these peaking turbines and more detail on the rationale for replacement so I will not repeat that material here.

The December 2018 DPS Energy Storage Goal and Deployment Policy specified what was to be included in the analysis. It was to “include a series of reliability and operational assessment studies looking at the equivalent level of ‘clean resources’ that could provide the same level of reliability as the existing peaker units. Hybridization and repowering with energy storage, as well as replacement with stand-alone energy storage, should be explicitly examined, according to the Roadmap.”

According to the description in Energy Storage Deployment Program Report – Unit by Unit Peaker Study:

The analysis relies on historical 2013 hourly operational and emissions data for the approximately 4,500 MW of affected peaking units across the state (almost entirely concentrated in New York City, Long Island, and the Lower Hudson Valley) to examine the technical feasibility of energy storage or energy storage paired with solar providing equivalent historical generation of the peaking units. Peaker operational and emissions data from 2013 was chosen because this reflects the peak NYISO demand year, and the correspondingly high levels of peaker operation which occurred in July 2013. This served as a proxy for representing peak‐level system operations, although theoretical peak system operations may impose incremental needs beyond those of 2013. The study did not consider system changes after 2013 that may impact how conventional peaking units and energy storage resources operate in the future, such as retirements of existing units, changes in the overall levels and patterns of demand, new transmission solutions, and/or the addition of more intermittent, renewable energy.


I am not a fan of the approach used in this analysis because I think it gives some mis-leading unit specific information. In the first place they considered all turbines as candidates not understanding that the primary purpose of some turbines is not to provide power during high load demand periods. They wasted effort considering the Jamestown Public Utilities turbine in Western New York that runs on the order of half the time. Peaking turbines are defined as units with an “average annual capacity factor of 10.0 percent or less over the past three years”. In addition there are turbines at steam boiler facilities that are necessary for “black start” situations when there is a blackout and the power necessary to start up the boiler is unavailable from the grid. Because that is a very rare instance the units are also run to provide power for peak power periods. In my opinion it would not be cost effective to dedicate energy storage for this application. You could not use it for peak loads because you never know when the grid power won’t be available. In conclusion the report considered units that should not have been included.

According to Table A-1 in the report, there are 3,780 MW of peaking turbines in New York. The report concludes that “Overall, at least 275 MW of peaking units, or around six percent of the total rated capacity of the fleet, are found to be potential candidates for replacement with 6‐hour energy storage sized to the maximum 2013 output of each peaking unit.” That means that a 6-hr energy storage system would be able to replace 7% of the existing peaking turbine capacity. The report goes on to say that “This number increases to over 500 MW when using 8‐hour duration storage”, but that only increases the replacement of existing capacity to 13%.

I don’t disagree with their conclusion that “Energy storage or a combination of energy storage and solar can contribute towards meeting NOx limits for a large number of units”. However there is a long way between “can contribute” and “will actually be an option used”. This is a preliminary scoping study. It notes that the “minimum size storage required to meet the NOx requirements can vary between units of the same facility” but does not recognize that the variation between sister units at a facility does not mean that one unit is more of a candidate than another.  The reality is that affected sources will adopt a facility‐wide strategy to meet the NOx limits and those strategies were not examined in this report.

There are other issues as noted in the Conclusion and Recommendations for Further Study. They note that “A more detailed analysis will be needed to understand the reliability impacts of specific unit replacements, especially as loads and resources change with greater electrification of transport and buildings and higher penetrations of renewables.” Many of these peaking units are in load pockets and changes in the load will drive whether energy storage is viable.


The report states that “A more detailed and thorough benefit‐cost analysis would need to be performed to understand the true economic viability of the replacement and/or hybridization options presented in this analysis.” Therein lies the biggest issue of energy storage – the cost. For those of us outside of Albany who care about costs a recently released report from the National Renewable Energy Lab (NREL): “2018 U.S. Utility-Scale Photovoltaics-Plus-Energy Storage System Cost Benchmark” provides information that can be used to estimate the costs of the energy storage option.

The NREL study lists costs for durations up to four-hours but the DPS report also includes six-hour and eight-hour durations. Table 3 in the NREL document, Detailed Cost Breakdown for a 60-MW U.S. Li-ion Standalone Storage System with Durations of 0.5–4 Hours, provides the information necessary to extend their projections to those different durations. I fit a linear regression model to describe the relationship between the specific costs and energy storage duration from the NREL table. 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. In this evaluation, in every instance, the reciprocal-X model (Y = a + b/X) statistic was the best choice and every regression had an R-squared coefficient great than 99.9% which indicates a strong relationship and suggests that these estimates are good enough for this analysis.

The NREL analysis includes all the costs for a greenfield energy storage project so I calculated values of retrofit potential costs that exclude the land acquisition costs. I estimate the installed cost for energy to be $343/kWh for an eight-hour battery system, $355/kWh for a six-hour battery system, and $380/kWh for a four-hour battery system.

The table NYC Energy Storage Peaker Replacement Summary lists data from the DPS study and calculated values. Table E1 in the DPS report lists the total nameplate capacity (MW) of peaking units that can potentially be fully replaced with storage to meet the 2025 NOx limits at 100% sizing to each unit’s 2013 peak generation. Note that I did not include the upstate turbine included in the DPS report in this analysis because it is not a peaking turbine. There are 36 MW of peaking unit capacity in New York City and Long Island that can be replaced with four hours of storage, 229 MW that can be replaced with six hours of storage, and 463 MW with eight hours of storage for a total of 728 MW. This is 18% of the 2013 peak load in New York City and on Long Island. Table A estimates the replacement cost estimate using the NREL report numbers and shows that replacing 18% of the load with Li-ion battery storage would cost $1.8 billion.

The cost per ton removed further demonstrates the staggering cost implications. I could not figure out which particular units were candidates for replacement because my analysis of Table B-1 did not result in the same number of units in each category. As a result I could not calculate the unit-specific cost per ton removed. Instead I just used the total emissions from all the sources the report’s Table 3: Peaking Units 2013 Operational Data. Table B shows the costs if all the emissions from all the peaking units came only from the 728 MW that can be replaced by energy storage. The cost to remove a ton of NOx is over $900,000 per ton and cost to remove a ton of CO2 is over $1,000 per ton. In order to put those numbers in perspective consider that the social cost of carbon (the alleged societal cost per ton of CO2 emitted) is currently around $50 by the Obama administration method and less than $5 by the Trump administration.


The report concludes “Overall, the findings suggest that there is an opportunity to consider replacing or hybridizing a substantial portion of the peaking units subject to DEC’s proposed NOx rule with a fleet of storage resources paired with solar. Such an outcome would potentially deliver significant environmental benefits, advance the state’s carbon reduction and clean energy goals, as well as benefit historically disadvantaged populations and communities such as environmental justice areas in line with the goals of the Climate Leadership and Community Protection Act.” However these results show that the cost of energy storage replacement is at least an order of magnitude greater than the cost of carbon’s impacts so this opportunity is not a cost-effective way to advance the state’s carbon reduction and clean energy goals.

Author: rogercaiazza

I am a meteorologist (BS and MS degrees), was certified as a consulting meteorologist and have worked in the air quality industry for over 40 years. I author two blogs. Environmental staff in any industry have to be pragmatic balancing risks and benefits and ( reflects that outlook. The second blog addresses the New York State Reforming the Energy Vision initiative ( Any of my comments on the web or posts on my blogs are my opinion only. In no way do they reflect the position of any of my past employers or any company I was associated with.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s