Preliminary results from an ISO-NE study show that the quantity of reserves needed in an increasingly renewable future will be a function of how well semi-dispatchable resources can be curtailed, the Planning Advisory Committee heard on Wednesday.

Patrick Boughan, ISO-NE senior engineer for resource studies and assessments, presented that finding during a discussion of the ancillary services and marginal unit emissions components of the New England States Committee on Electricity’s requested 2019 Economic Study. (See ISO-NE Planning Advisory Committee Briefs: April 23, 2020.)

The RTO employed Dartmouth College’s Electric Power Enterprise Control System (EPECS) simulation tool for the study, with modifications and improvements made to the software program since it was previously reviewed with the PAC in December 2017, he said.

The analysis reviews both the use of select reserves currently required by ISO-NE, such as a 10-minute spinning reserves, as well as other types of reserves that are not required but have physical qualities that can be tracked and analyzed, such as load-following reserves.

NESCOE, Anbaric and RENEW Northeast last year each requested separate studies from ISO-NE. (See “Modeling More Offshore Wind, Slowly,” ISO-NE Planning Advisory Committee: March 18, 2020.)

Regarding ancillary services, NESCOE in its initial study request said, “As the market needs change, new grid opportunities may be identified to address challenges, including load following, regulation, operating reserves and operation during low-load periods.”

NESCOE said its study request conformed with the Tariff, “as it considers the potential economic benefits of relieving transmission constraints and shows the benefits of interconnecting increasing amounts of offshore wind in alternative locations.”

EPECS assumes semi-dispatchable resources are infinitely curtailable to maximum amounts, but if this assumption is revised, more reserves will be needed, the report said.

Estimated Marginal Emissions

Boughan said the RTO developed two ways to calculate marginal unit emissions for the NESCOE study and developed two complementary analyses using outputs from GridView, a software program that simulates the economic operation of power systems in hourly intervals for periods ranging from one day to many years.

This varies from how the marginal resource is determined for the annual marginal emissions analyses (MEA) conducted by the RTO; however, using the GridView model allows analysis of future scenarios that are not available via the MEA method, Boughan said.

One approach to determining marginal unit emissions compares results of two scenarios with different amounts of wind, calculating the change in annual emissions per additional megawatt-hour of energy produced by offshore wind.

The second approach is to find the most expensive generator able to respond to dispatch signals — the “implied” marginal unit, which sets marginal emissions, the report said.

By analyzing only cases without transmission constraints, the RTO was able to cleanly quantify the change in CO2 emissions because of offshore wind production additions. There is only a small change in annual emissions and megawatt-hours because of transmission constraints caused by offshore wind, and internal interfaces create difficulties for determining the most expensive generator, Boughan said.

Drawing from both methods, the NESCOE study finds that 30 to 40% of the emissions reductions come from reduced dispatch of high CO2-emitting municipal solid waste and coal generators, even though they are marginal less than 5% of the time.

The study also observes that municipal solid waste resources may not, in reality, be marginal generators because of the other services they provide.

The second approach, implying marginal emissions in GridView simulations, provides a slightly lower estimated marginal CO2 emission rate than observed from the first approach, comparing two simulations with different amounts of offshore wind.

The study concluded that results from GridView simulations do not exhibit the range of marginal units that a historical locational marginal unit analysis would contain.

Not only does GridView not apply bidding strategies to resources, but results from the simulations do not exhibit the range of marginal units that a historical locational marginal unit analysis would contain. Resources such as energy storage charging and discharging are price takers and are never identified as marginal. Only natural gas, coal, municipal solid waste and wood-fired units are seen to be marginal.

ISO-NE will continue studying ancillary services in the 2020 Economic Study to further determine the needs of the system and will work with the PAC to confirm the assumptions needed, Boughan said.

The RTO’s next steps are to publish the final NESCOE study by July 1, the final Anbaric study in June or July and the final RENEW report by July.

Anbaric 2019 Economic Study Follow-up

The addition of 8,000 to 12,000 MW of offshore wind plus assumed resource retirements resulted in Southeast Massachusetts/Rhode Island (SEMA/RI) export interface constraints, the PAC heard during a follow-up to the March presentation on the Anbaric 2019 Economic Study.

Haizhen Wang, the RTO’s lead engineer for resource studies and assessments, presented preliminary results that show natural gas-fired resources were required to partially replace retired nuclear generation in all Anbaric scenarios.

Because of its intermittent nature, offshore wind does not follow loads, and the study illustrates intervals when demand is high but offshore wind is low, especially during summer, Wang said.

2020 Economic Study Scope, Assumptions

ISO-NE Manager of Resource Studies and Assessments Peter Wong presented the first of two presentations planned on the 2020 Economic Study draft scope of work and high-level assumptions for production simulations.

National Grid requested the study to model year 2035 to provide insight into wholesale energy market impacts, unit economics, utilization of resources, and the role of bidirectional transmission capability and battery storage in meeting the needs of a system with a high proportion of variable resources.

Several stakeholders asked Wong about the assumed potential bidirectionality of the existing external ties, including the Highgate, HQ Phase II and New Brunswick interconnections. The study proposes to “treat them as bidirectional, if physically capable, with a focus on Hydro-Québec, given the likelihood of coupled supply and load in New York.”

“If the 1,200 MW from the New England Clean Energy Connect [NECEC] is essentially base-loaded, if you had export over Phase II to do the type of spillage absorption … doesn’t that just mean that we’re backing down on the amount of NECEC exports, since on a net-interchange basis the control area is just reducing the amount of its imports?” asked Tom Kaslow of FirstLight Power Resources.

NECEC is a $950 million project to deliver 1,200 MW of Canadian hydropower to the New England grid in Lewiston, Maine, along a 145-mile transmission line controlled by Avangrid subsidiary Central Maine Power.

The study will neither include an assessment of FCM outcomes nor ancillary service prices, Wong said.

“This study requires that numerous resource types of different sizes and locations be added to the system, making it nearly impossible to develop any meaningful results without lots of effort and also many assumptions,” Wong said.

For example, the large quantities of solar generation are dispersed projects for which the interconnection studies are being done through the distribution utilities interconnection process, and “we don’t have any good handle on what is needed,” he said.

“And some of these storage facilities, well-sized and well-located standalone storage proposals should not be triggering the need for any substantial upgrades,” Wong said.

The preliminary schedule for the 2020 Economic Study is to finalize production simulation assumptions for the three scenarios at the June/July PAC meeting; present draft production simulations results and identify sensitivity scenarios and assumptions in Q3; present sensitivity scenarios simulation results and draft ancillary services (EPECS) results in Q4; and present draft and final reports in the first quarter 2021.