In the July 11-12, 2023, meeting of the Resource Adequacy Subcommittee (RASC), MISO presented an updated accreditation proposal for stakeholder feedback. In your comments, please include feedback on the following:
Comments are due by July 28.
NG Renewables offers the following comments on the current proposal for D-LOL
Co-located/hybrid metering requirements: MISO has proposed to require additional component level monitoring for hybrid resources and co-located resources, regardless of whether they submit separate or combined offers. MISO should provide a definition of “component”. MISO should also provide some context around what value the additional metering will provide, as it is immediately clear what the value is for hybrid or co-located resources. For example, in a DIR-hybrid, will the component level metering be used to inform availability for purposes of accreditation, or will MISO continue to use the market-participant provided forecast/offer?
Takeaways from published class-level values: NG Renewables appreciates that MISO has provided estimated class level resource accredited values. These are extremely useful as a benchmark. MISO should also provide the values project out over time in addition to just the 2023-24 planning year, e.g., 2025, 2030, 2035, 2040. MISO should provide context around why solar resources appear to be less available during critical hours in the spring than the fall. It would also be helpful if MISO could provide an example of a solar+storage accreditation under D-LOL in both a co-located and DIR-hybrid context.
Hours used in D-LOL: NG Renewables appreciates that MISO has published initial D-LOL hours for solar and coal resources. However, NG Renewables also notes that the D-LOL hours significantly deviate from RA hours. For example, 2022 RA hours appear to capture far more hours during the winter season than the D-LOL approach. It would be helpful if MISO couldprovidesome granular context around why this deviation occurs, e.g., modeled behavior of a particular resource class, differences in modeled vs actual load, etc. NG Renewables understands that the two approaches will likely deviate but feel stakeholders would benefit from a detailed understanding of what has driven these deviations, even if this is only a sample case.
Transition timeline: MISO has verbalized that they intend to improve the modeling of storage resources. How does MISO plan to fit this modeling improvement into the transition timeline, if at all? What will stakeholder involvement look like?
NG Renewables appreciates MISO’s consideration and looks forward to further discussions.
MidAmerican appreciates the opportunity to provide feedback on the proposed accreditation reform (RASC-2020-4, 2019-2) (20230711-12).
MidAmerican continues to be greatly concerned about the proposed accreditation reform. The main issues can be broken into three issues.
MidAmerican has many questions about the model that needs to be included in the white paper. MidAmerican can easily demonstrate no correlation between wind speed and temperature, yet MISO seems to be correlating the two to create wind output assumptions. What power curves for the wind farms is MISO using? How is MISO forecasting load based on the various weather years? How does, for example, the 10,000 MWs of load added to the model in the fall, impact the overall results? Is the load added uniformly to every hour or is shaped with a peak of 10,000 MWs. When MISO makes statements like the PRMR is likely to increase as LOLE modeling enhancements are made to better reflect risk, it clearly indicates how modeling assumptions or even model capability will change over time causing volatility.
In addition, MidAmerican fully supports the comments already made by Xcel and Alliant. To repeat and reinforce the biggest concern, MidAmerican thinks MISO needs a process that is stable, and predicable to allow what should be the true goal of resource adequacy which it to make sure the generation portfolio on the system can reliably serve the load. Less focus on price signals and assuming that those prices can somehow be predicted so rational decisions will be made and more focus on informing those who are trying to build and retire generation, so the system is always able to reliably serve load.
Direct-Loss of Load results by resource class:
• North Dakota is concerned with the results for the thermal class showing a winter and spring drop off in accreditation as well as the high accreditation of storage. Please reevaluate the thermal assumptions and determine the drivers for the accreditation drop off as it doesn’t seem to reflect realistic reliability contribution. In particular, the effect of cold weather needs to be evaluated to make sure its impact does not incorrectly reduce thermal accreditation and create a disincentive to invest in this type of resource potentially causing future reliability issues. Thank you for recognizing the issues with how storage is being modeled and please pursue an approach that will recognize the correct valuation of the resource class.
Initial design proposal:
• We believe that MISO should begin exploring the potential locational divergences among resource classes, including solar, and further explore any divergences that may occur from the use of EUE. In cases where the modeling assumptions are difficult to capture, we encourage MISO to provide any of the insights into how accreditation may change in as granular way as possible uses future sensitivities and how the results align with the PRMR.
Co-located and single-offer hybrid proposals:
• We believe that the current proposals are sufficient but would be open to the possibility of modeling the most anticipated hybrid class, the solar-storage hybrid.
PRMR and LSE obligations:
• We strongly support using the hours of highest reliability risk and moving away from the peak demand when determining the PRMR. MISO must accredit resources based on their actual expected contribution to resource adequacy and use the clearest, most effective methods to determine those values. The requirement needs to be aligned with using accreditation determined by Loss of Load Hours. A forecast of the PRMR and accreditation will provide stability to the resource planning process for market participants.
Comments
of the
Association of Businesses Advocating Tariff Equity (ABATE),
Illinois Industrial Energy Consumers (IIEC),
Louisiana Energy Users Group (LEUG),
Texas Industrial Energy Consumers (TIEC),
Coalition of MISO Transmission Customers (CMTC),
Midwest Industrial Customers (MIC),
and
NIPSCO Large Customer Group (NLCG)[1]
Regarding
RASC: Accreditation Reform (RASC-2020-4, 2019-2) (20230711-12)
July 28, 2023
ABATE, IIEC, LEUG, TIEC, CMTC and MIC, as representatives of the End-Use Customers (EUC) Sector, and NLCG appreciate this opportunity to provide comments to MISO.
During the July 11-12, 2023, meeting of the MISO Resource Adequacy Subcommittee (RASC), MISO presented an update on its Direct Loss of Load (DLOL) capacity accreditation proposal for Capacity Resources. In addition, as part of that update, MISO for the first time proposed to potentially incorporate, into its planned November 2023 DLOL capacity accreditation filing at FERC, changes to the assignment of capacity obligations to Load Serving Entities (LSEs) under the MISO resource adequacy construct. Our comments respond to MISO’s new proposal in that area. Our silence with respect to any other aspect of MISO’s presentation on its DLOL capacity accreditation proposal should not be interpreted as tacit agreement with that aspect of the presentation. In addition, ABATE, IIEC, LEUG, NLCG, TIEC, CMTC and MIC each respectively reserve their right to further comment on MISO’s DLOL capacity accreditation proposal in the future.
On Slides 17 and 18 of MISO’s July 11-12, 2023 DLOL Capacity Accreditation Proposal Presentation, MISO proposed, potentially as part of its planned forthcoming November 2023 Direct LOL Capacity Accreditation filing with FERC, to align the hours that determine LSE capacity obligations (currently the MISO system peak hour for each MISO season) to the so called MISO Resource Adequacy hours. Those Resource Adequacy hours currently impact the capacity accreditation for Schedule 53 Capacity Resources and would impact the capacity accreditation for nearly all Capacity Resources under MISO’s DLOL Capacity Accreditation Proposal. Such a change in the assignment of LSE capacity obligations could greatly dilute the very strong price signal currently sent to LSEs to, where practicable, keep their load off of the MISO’s seasonal system peaks. It would do so by assigning capacity obligations based on average LSE demand over several hundred hours of the year rather than just LSE demand at the MISO seasonal system peak hours. It would also fail to consider that not all Resource Adequacy hours are equal. Some Resource Adequacy hours carry a much greater risk of a loss of load event occurring than others or the risk of a loss of load event that would involve significantly more firm load shedding than other such loss of load events. As a result, any proposed changes with respect to how capacity obligations are assigned to LSEs on a going forward basis would need to be made with great care; otherwise the changes could unintentionally act to adversely impact resource adequacy in MISO by eliminating the price suppression of load demand that is already present during the most critical hours of the year for resource adequacy -- currently the time of the MISO seasonal system peaks and particularly the time of the MISO summer system peak.
Given the high risk of unintended consequences and the complex interactions that would need to be considered to avoid those unintended consequences, ABATE, IIEC, LEUG, NLCG, TIEC, CMTC and MIC strongly oppose MISO attempting to address potential changes to the assignment of capacity obligations to LSEs as part of its planned November 2023 Direct LOL Capacity Accreditation filing with FERC. Any further discussion with respect to whether there is a need for going forward changes to the assignment of capacity obligations to LSEs within the MISO resource adequacy construct should be framed as a separate discussion within the MISO stakeholder process. That further consideration must be given sufficient time to provide for the analysis and discussion that would have to surround that issue to ensure that any proposed changes in this area are truly needed and the proposed changes, if any, would not have unintended adverse consequences with respect to resource adequacy.
Thank you for providing us an opportunity to provide the above comments. If it would be of help, we would be glad to discuss the above comments further with MISO and other stakeholders. Please do not hesitate to contact any of the following representatives:
Jim Dauphinais
Brubaker & Associates, Inc.
(Consultants to ABATE, IIEC, LEUG, NLCG and TIEC)
(636) 898-6725
Ali Al-Jabir
Brubaker & Associates, Inc.
(Consultants to ABATE, IIEC, LEUG, NLCG and TIEC)
(361) 994-1767
Ken Stark
McNees Wallace & Nurick LLC (for CMTC)
(717) 237-5378
Kavita Maini
KM Energy Consulting, LLC (Consultants to MIC)
(262) 646-3981
[1] ABATE, IIEC, LEUG, TIEC, CMTC and MIC are all MISO Members in the End-Use Customer Sector. NLCG is a non-MISO Member stakeholder whose members include large end-use customers within Indiana that are interruptible and/or have cogeneration facilities and that take service under NIPSCO Rate Schedule 831, which allows limited market purchases through Northern Indiana Public Service Company (NIPSCO).
WPPI appreciated the opportunity to provide feedback on MISO’s proposed changes to accreditation. To help us to better understand the results, WPPI requests MISO provide more detail on the inputs and assumptions used in the LOLE simulation that determined the DLOL results by resource class. In addition, it would be helpful to also see corresponding future-year accreditation estimates under existing accreditation methodologies as well as marginal ELCC accreditation estimates for wind and solar. Regarding PRM requirements and LSE obligations, WPPI requests MISO provide an example comparing an LSE’s requirement under current rules (i.e., based on peak demand) with an LSE’s requirement under MISO’s planned reforms, where the obligation is based load during the risky hours.
Lastly, WPPI believes MISO should revisit its capacity replacement requirements as part of accreditation reform, perhaps considering changes beyond those proposed by MISO’s IMM.
WPPI Looks forward to continued discussion on this topic.
The Public Service Commission of Wisconsin Office of Regional Markets (ORM) staff appreciates the opportunity by MISO to provide feedback on the updated accreditation proposal during the RASC July 11-12 meeting. This feedback should not represent a position in-favor or against the discussed accreditation methodology, but we offer this feedback as a continuation of the discussion since November 2022 when the Direct Loss of Load (D-LOL) method was introduced.
D-LOL results by resources class
It is helpful to see preliminary results of the D-LOL results implemented per resource class and its comparison to the UCAP method results. It would also be helpful for MISO to further clarify how are the (1) fuel shortage, (2) expected planned outage schedule, and (3) modeled unplanned schedules accounted for in the D-LOL implementation for coal and gas. Our expectation is for all the results shown in Slide 5 of the RASC July 11-12 meeting to be shown in a forward looking period for a 5 to 10-year period. While we understand that these results will be based on the SERVM results by Astrape, we ask MISO to use an energy growth curve that closely simulates that of Future 2A. This would more appropriately simulate the resource planning of existing resources based on an approach already familiar to stakeholders.
Initial design proposal
Based on the current status of information availability and stakeholder understanding, we recommend MISO strongly consider changing their initial design proposal and instead pivoting to a two-phase approach:
- First, adjusting the accreditation of solar resources to utilize the Average-ELCC as it is implemented for wind resources; and
- Second, delaying a filing regarding D-LOL for all resources until more information can be provided and vetted.
This would accomplish MISO’s initial goal of this entire non-thermal accreditation process (now going on 15+ months) of transitioning solar accreditation away from the current approach which utilizes afternoon hours of peak load, which MISO has demonstrated may not align with tight hours, especially as solar penetration increases.
Utilizing average ELCC could be an appropriate step towards getting solar to be accredited based on the LOLE model, instead of peak load. In addition, it has the benefit of largely being familiar to stakeholders because of the long-standing use of ELCC for wind resources.
This approach does not restrict MISO from continuing to explore D-LOL, but it provides a method to adjust solar accreditation now, without putting all of the accreditation eggs into the same basket as a wholesale overhaul of how all the accreditation beans are counted for every single resource in MISO.
The already-proposed 3-year transition plan would be a worthy ‘experimental’ period to further investigate how certain methods we are already familiar with could be implemented to other resource classes. This would also lead to more homogeneity by having a similar or the same accreditation scheme amongst non-thermal resources. Xcel Energy has also suggested this in the past. It would be beneficial to have preliminary data on this implementation for the next RASC meeting.
Additional detailed considerations of the D-LOL initial design proposal include:
Our staff views the class-level probabilistic approach as all the modeled possible scenarios of loss of load hours (LOLH), or instances, and the unit-level retrospective approach as the actual scenarios of tight margin hours that occurred. Tight margin hours, as described in Schedule 53, are the top 3% of tight margin hours within a 25% reserve margin or less per season. On the other hand, LOLH are instances whenever load (demand) is greater than generation. These two constraints are very different, and since the class-level step is the expanded probable scenarios of the unit-level step, we consider the constraints should be equal. We recommend for MISO to expand the class-level constraint of “load > generation” to the top 3% of tight margin hours within 25% reserve margin or less per season. With this change, both the class-level and unit-level accreditation steps will be evaluated under the same criteria and allow for a fair allocation of accreditation through the steps.
Our staff considers that the Expected Unserved Energy metric needs to be further studied before receiving any further consideration in anything Resource Adequacy related. This reliability metric has not been significantly explored before in the MISO footprint and greater understanding is necessary before implementation.
We appreciate MISO’s consideration in breaking down resource classes even further than those listed in Slide 5, and potentially considering their location. On one side, thermal resource, and other resources, would benefit by receiving unit specific (i.e., combined cycle, combustion turbine) accreditation consideration at the class-level. This could become particularly important if the proposed EPA power plant rules move forward and addresses emissions standards on a per unit basis of retiring, existing, and new units. This is because the new power plant rules utilize the unit type (i.e., combined cycle, combustion turbine) and the unit size (300 MW is threshold under review) as parameters to establish how each rule would apply. In addition, weather dependent resources (i.e., solar and wind) will have varied performance based on their location. Segregating resources by region (i.e., North/Central and South) might not be granular enough to fairly reflect the true availability of these resources. For example, the wind resources potential available in Wisconsin may be greatly different than Iowa. Solar resources will also have some location-based variability, particularly based on latitude (how far north or south) and season (there are more sunlight hours in the summer in MISO North, while but more sunlight hours in winter in MISO South).
At the unit-level, our staff suggests MISO to change the current approach of filling the seasonal 65-hour deficiency with the Annual Average Offered Capacity (AAOC). Our staff considers that by utilizing the AAOC it “socializes” the seasonal variation of each 65-hour data set. For this reason, we support the proposed change to fill deficient hours with seasonal class D-LOL %, however, this will only work if the LOLH are expanded to include those within the same margin threshold as written in Schedule 53. Otherwise, resources will have 65 “risk hours” that are not evaluated under the same criteria or metric.
Co-located and single offer hybrid proposal
Our staff agrees with MISO in using component level metering for both co-located resources and hybrid resources as part of the accreditation reform. We consider this as a fair evaluation to provide greater visibility on each resource’s capability. However, our staff considers that greater investigation is needed on ‘capping’ the aggregated accredited value at the shared point of interconnection. This limit might be restrictive depending on the season, and the variability of accredited capacity within the shared point of interconnection needs to be further evaluated.
Planning reserve margin requirement and LSE obligations
While we consider it appropriate to consider D-LOL availability into the PRMR calculation, we are also aware of the potential increase due to seasonal risk reflected in the LOLE study. For this reason, we request for MISO host a LOLE workshop that shows step-by-step how the study is completed on an annual basis and what has specifically changed with the incorporation of a seasonal construct.
Other
If MISO moves forward with the D-LOL proposal as initially proposed, the ORM staff suggests that MISO delay the tariff filing goal date of end-November until stakeholders can be assured that sufficient demand (load) data for Fall and Spring is available. This demand data is imperative to build proper, forward-looking trends for the planning reserve margin requirement in the respective seasons. As mentioned above, the implementation of the D-LOL will affect directly the PRMR calculation and with the lack of local or regional forecasted peak demand to calculate the specific PRMR to each shoulder season, it would just reduce veracity of the metric being calculated.
During the last OMS-MISO Survey presentation, MISO used a Summer-like load trend for Fall and Winter-like load trend for Spring. In our judgment, it is not appropriate to base seasonal requirements on trends not specifically corresponding to the respective season, for the reasons already explained. By consequence, this could then send the incorrect signal to resource planners. Otherwise, MISO must justify to stakeholders that this approach would be acceptable. If MISO staff adjust the proposal based on feedback proposed in these comments, we envision the tariff filing date would depend on the scope of an updated proposal and ability for stakeholders to provide sufficient input to that proposal. Familiarity with an average ELCC approach for solar could streamline this.
The OMS Resources Work Group (OMS RWG) appreciates the opportunity to provide feedback to MISO on its most recent update on the Direct-Loss of Load (Direct-LOL) accreditation proposal. This feedback is from an OMS work group and does not represent a position of the OMS Board of Directors.
Direct-LOL Results by Resource Class
MISO’s posted Direct-LOL results based on the current Planning Year data provide a benchmark to show what the impact of the proposed change would be on the existing resource fleet and mix. However, to fully understand the implications of this new accreditation methodology, we need additional data and results with various changes in the resource mix. MISO must provide Direct-LOL results for the resource mix during several years in Future 2A, as this is the future used for transmission planning. We understand that Astrape will use its own forward-looking energy growth scenario based on the existing resources of Planning Year 2023-24; however, it is important for this scenario to closely align with the one that stakeholders are most familiar. It is critical to state resource planning to understand how MISO’s proposed Direct-LOL resource accreditation will change capacity values over time.
Initial Design Proposal
On the Design Elements listed on Slide 7 (Hour Selection, Direct-LOL calculation, and Resource Classes), MISO should consider the other potential options MISO has listed. For example, there is resource adequacy value in availability during hours with an extremely tight reserve margin (such as those that would prompt MISO going into Energy Emergency Alert conditions). Resource class granularity, such as more location-based (i.e., local resource zones) or with technology differentiation within a fuel category (ex. Gas CC vs CT and solar fixed panel vs solar tracking panel) would provide meaningful value. The OMS RWG requests that, at a minimum, MISO justify why its initial proposal is preferred to other potential options.
Co-located and single offer hybrid proposal
MISO should provide a model and resource class for the most prevalent hybrid combination in the queue – solar with storage. The OMS RWG understands that co-located and hybrid resources are being evaluated based on their point of interconnection (POI), but this needs further discussion and clarification. Co-located or hybrid resources could be broken into more classes based on the ratio of solar MWs to storage MWs, once more performance data from these resources is available.
Planning Reserve Margin Requirement and LSE obligations
Going forward, MISO should include UCAP accreditation and PRMR values when listing out Direct-LOL resource class accreditation and PRMR (Slide 16 of the presentation). Providing the UCAP and Direct-LOL results for the resource classes and the PRMR changes, side by side, will help determine the impact of the proposed change and the relative impact to each resource class.
DTE appreciates the opportunity to provide feedback on the updated DLOL accreditation proposal.
Seasonal Direct-LOL Results
DTE appreciates the summary of DLOL results for the 2023-24 Planning Year, and requests that MISO provide similar data for other cases (as further described below). In addition to the various questions on specific results raised by stakeholders in the RASC, DTE would appreciate further clarification on the Winter DLOL results for the Pumped Storage resource class. During the RASC, MISO seemed to indicate the low results compared to UCAP were potentially due to the small sample size of Pumper Storage units within the MISO footprint. If this is the case, DTE encourages MISO to consider improvements that limit the volatility of resource class percentages due to class size. Negatively impacting unit accreditation due to a resource’s class size seems counter to the goal of the DLOL accreditation approach. Further, DTE encourages MISO to work with stakeholders to release similar results for other requested scenarios (e.g. Higher renewable penetration, with/without cold weather adder on thermal units, expansion of hours beyond LOL, weighing DLOL based on EUE, etc.).
Modeling Improvements
DTE agrees with the MISO observation (as well as numerous stakeholders during the meeting), that significant modeling improvements are necessary in order to calculate an accurate DLOL percentage for storage resources. With a significant increase in storage deployment expected in the near future, DTE urges MISO to focus on improving storage modeling as soon as possible. More broadly, DTE supports MISO’s stated focus on various modeling improvements and believes they would lead to a more robust D-LOL accreditation method. Accordingly, DTE requests that MISO provide a list of improvements currently being considered by MISO and the potential timing for implementation. As individual or isolated improvements could have a greater impact on certain resource classes (i.e., implementing ramping constraints could reduce the availability of coal in a LOL hour and reduce accreditation), DTE encourages MISO to implement modeling improvements simultaneously as opposed to in series (to the extent possible). Stated another way, MISO should avoid disadvantaging a particular resource type due to the order in which it pursues modeling improvements.
Design Elements
DTE supports further discussion on each of the design elements suggested by MISO, with particular focus on the granularity of resource classes. Specifically, the potentially significant difference between modeled operations (as well as performance) of combined-cycle and simple-cycle gas units should be captured by separating the gas resource class. Additionally, DTE would also like to understand how the DLOL percentages would change if the resource classes were also differentiated by vintage (i.e. a new combined cycle might have a much lower outage rate than one built over a decade ago).
DTE appreciates the work MISO has done in investigating the best way to accredit co-located hybrid resources and agrees with the current MISO proposal to accredit based on component level and cap the aggregate accredited value at the shared interconnection service limit. Relatedly, and as mentioned above, improvements to storage modeling are a prerequisite to adequately accrediting hybrid resources.
PRMR
DTE encourages MISO to host a more in-depth conversation on how MISO plans on determining the Planning Reserve Margin and distributing the PRM to the LSE’s. It is not abundantly clear how this will occur after this most recent RASC. Is MISO proposing allocating the PRM to LSEs pro rata based on their load during LOL hours, or differentiating based on each LSEs contribution to risk (i.e. fleet makeup and load shape)?
The Entergy Operating Companies ("EOCs")[1] appreciate the opportunity to provide feedback on MISO’s accreditation reform proposal. The EOCs continue to believe that Average ELCC is the correct methodology to be used for accreditation in MISO’s prompt year PRA because it more accurately measures the reliability contribution for existing resources than the DLOL methodology.
Direct-LOL Resource Class Results
The EOCs have the following questions/comments on the Direct-LOL Resource Class Results:
Direct-LOL Design Element
Storage Dispatch Assumptions
The storage dispatch assumptions used in the LOLE Model will have a significant impact on the class average accreditation results for storage resources and for wind/solar resources. This is because the decisions about when storage will be dispatched will greatly influence the number of loss-of-load hours and what time of day loss-of-load hours occur. As an example, if prior to a battery being dispatched, there are 8 loss of load hours, and the battery is fully discharged and able to eliminate 4 of the loss of load hours, that would result in MISO measuring accreditation during the 4 remaining loss of load hours where the battery is not being dispatched, and the battery would receive a zero accreditation value. Given that the battery reduced Expected Unserved Energy by 50%, it is unreasonable that the battery should receive a zero accreditation value. To alter this example, if the battery were to be discharged at 50% of its capability across all 8 hours of load shed so that none of the load shed hours were fully eliminated, then the battery would receive a 50% accreditation value under the DLOL methodology, which is a more reasonable outcome and better represents the true reliability contribution of batteries. Given the importance of storage dispatch assumptions, the EOCs request that MISO and stakeholders discuss this issue during a future RASC meeting.
PRMR Reduction
Moving from the current UCAP accreditation methodology to DLOL will result in declining resource accreditation values, which will ultimately result in lower PRMR values. Given that the amount of accreditation lost will be different for each LRZ, the EOCs believe that MISO should explore options for reducing PRMRs on an LRZ basis as opposed to uniformly lowering all LSE PRMRs by the same ratio.
LRZ Local Clearing Requirements
The Local Clearing Requirement should no longer be established on a UCAP basis. To determine each LRZ’s local clearing requirement, MISO should calibrate the zonal LOLE model to the 0.1 LOLE target and then sum up the LRZ’s DLOL capacity and the 0.1 LOLE MW adjustment. This will appropriately result in a LCR requirement on a DLOL basis.
The current MISO process for determining local clearing requirements is flawed because LCRs are established on a UCAP basis even though capacity is accredited on a SAC basis. For MISO wide requirements, UCAP and SAC are equal in magnitude so there are no problems with substituting one methodology for the other, but this is not true on an LRZ basis. Some LRZs’ total UCAP is greater than SAC while other LRZs’ total UCAP is less than SAC. The result of this LCR design flaw is that in some cases the MISO LRZ LOLE models will identify a set of resources that is able to meet the 0.1 LOLE target, but the sum of that group of resources SAC values minus the ZIA will be less than the LRZ’s UCAP LCR. This result defies the purpose of using an LOLE model to establish an LCR. If the LOLE model determines an LRZ is reliable with a given set of resources, then MISO’s resource adequacy construct should reach the same conclusion. The EOCs request that MISO fix this problem for the 24/25 PY.
[1] The Entergy Operating Companies are Entergy Arkansas, LLC, Entergy Louisiana, LLC, Entergy Mississippi, LLC, Entergy New Orleans, LLC, and Entergy Texas, Inc.
Illinois Commerce Commission (ICC) staff is interested in the Direct Loss of Load (DLOL) accreditation approach. ICC staff agrees that accreditation should reflect what resources are available at risky times. This is the most accurate measure of a resource’s value and should encourage developers to build the proper mix of resources that are needed for system reliability. MISO should provide more information location based DLOL accreditation. It seems plausible that resources in the North, Central, and South regions of MISO could perform differently. More granular accreditation could enhance system reliability and is consistent with the aim of the DLOL/marginal accreditation approaches.
ICC staff would appreciate more information, including examples, on the hybrid accreditation approach. The component-level approach seems workable, but more details would be helpful. Hybrid and storage resources will be important for future reliability, and their proper accreditation is crucial. ICC staff notes that accreditation alone may not be sufficient to maximize the reliability potential of storage and hybrid resources. MISO should consider new market products that incent the desired behavior from storage in order to ensure system reliability.
AMP, MPPA, and SMMPA submit the following feedback:
Please provide more information about the analytical methodologies underlying the D-LOL approach for calculating class average accreditation values. Providing unit-level projections without sufficient background information will not allow stakeholders to understand the implications of MISO’s proposal. Accreditation uncertainty won’t necessarily motivate any forward resource planning and procurement, and is less likely to promote efficient planning and procurement.
I am happy to discuss.
David Sapper
dsapper@ces-ltd.com
NextEra thanks MISO staff for seeking input on MISO’s proposals.
DLOL results by class: NextEra does not support MISO’s Direct-LOL (DLOL) methodology proposal. NextEra reiterates concerns that the Direct-LOL methodology’s capacity accreditation results are similar result to a Marginal ELCC proposal. The DLOL methodology is well suited for measuring the reliability value that the next wind/solar resource will provide to the MISO system during the modeled loss of load events, but does not measure the cumulative reliability value of the existing wind/solar fleet in the MISO system. Directly responding to the results by class, NextEra identifies the need for more information surrounding the assumptions used in MISO’s LOLE and DLOL models to determine availability of resources, how hours are determined, and assumptions around current fleet performance/availability (i.e. installed solar is currently ~4,000MW and the model incorporates ~12,000 in some seasons). The limited information has provided no visibility into the drivers for DLOL values and the resultant accreditation impacts after the Schedule 53 process is integrated. Moreover, the results appear inconsistent with current technology-specific contributions to reliability. This is especially true for nuclear, which varies from 91% to 80% across all seasons despite having high availability outside of planned outages. The DLOL results for wind in the winter are low and inconsistent with existing methodologies for capacity valuation and past operating data. The same is true for solar accreditation in the summer.
Initial design proposal: Given the time constraints during the RASC, there continues to be uncertainty around the details of the DLOL design, and impacts to PRMR, obligations. For these reasons, we believe we need to continue attempt to clarify the issues throughout the stakeholder process and delay FERC filings until stakeholder-led improvements can be incorporated into the filing. NextEra requests that MISO schedule additional workshops or increase the allotted time at future RASCs in which DLOL will be discussed.
July 28, 2023
Comments from Environmental Sector on RASC: Accreditation Reform (RASC-2020-4, 2019-2) (20230711-12)
The Environmental Sector appreciates the opportunity to provide feedback to MISO regarding the Market Redefinition: Accreditation Reform, presented at the Resource Adequacy Subcommittee (RASC) on July 12. During the July RASC meeting, MISO requested feedback on: DLOL results by resource class, initial design proposal, co-located and single-offer hybrid proposal, and Planning Reserve Margin Requirement and LSE obligations.
In addition to commenting on these issues below, the Environmental Sector requests that MISO and Astrape hold a workshop dedicated to further discussion of the resource class accreditation results and the models and assumptions that were used. The workshop should give attendees the option to attend in person or virtually. There were many questions raised at the July 12th meeting that need further discussion. Even MISO indicated it was still working to fully understand the results presented. At least three hours, if not more, should be dedicated to that discussion. Below we offer a series of questions which we hope can be answered at that workshop to help better understand and reflect upon the proposal. We request that MISO provide data to stakeholders that support MISO’s answers for all data/result related questions.
DLOL results by resource class
Why is solar accreditation (both ELCC and DLOL) higher in the fall than in the spring, especially considering that days (that is, sunlight periods) in the spring are longer than in the fall?
Why are the wind DLOL results lower (compared to most other seasons) in the winter when the UCAP results are significantly higher in winter? Please provide relevant data.
When is MISO planning to update its storage modeling? Will that be in advance of its filing of the DLOL accreditation proposal at FERC? If not, when will the DLOL tariff be updated to reflect changes in storage modeling?
Please further explain why MISO expects the accreditation values of storage to decrease.
Storage accreditation is dependent on the way storage dispatch (charging and discharging) is modeled. In fact, all resource class accreditation may be impacted by storage modeling to the degree that it shifts the LOL hours. We appreciate that MISO intends to improve its modeling of storage. Please provide more details on how MISO intends to change the modeling of storage in the future. In addition, we request MISO engage stakeholders on the storage modeling changes and seek input on the assumptions and methodology.
We believe MISO is only evaluating accreditation for 4-hour energy storage resources. How will storage accreditation values change for 6- or 8-hour storage?
How does MISO intend to capture portfolio effects, consistent with the recommendations contained on page 36 of ESIG’s February 2023 report, “Ensuring Efficient Reliability; New Design Principles for Capacity Accreditation,” and as further discussed throughout that report (e.g. pages 28-30)?
Does the LOLE model account for congestion on the transmission network? If yes, please explain how.
How does the fact that certain seasons have significantly fewer LOL hours impact the accreditation of resources in those seasons? Please provide example data to help illustrate the answer.
Initial design proposal
MISO should consider the “Other Potential Options” listed on slide 7. Please provide spreadsheets and documentation to show the impact of these options:
Using an additional expanded set of hours, such as the 3% of tightest margin hours. (Using the 3% tightest margin hours could provide a more realistic value of the capacity of resource types and also result in less volatility of accreditation values year over year.) Please provide analysis to compare how adding additional hours (3% or 1% of tightest margin hours) differs from accrediting based on LOL hours alone.
The impact of weighting LOL hours based on the amount of Expected Unserved Energy in those hours.
The location of resources for resource class accreditation, either subregionally or based on MISO’s four planning regions (East, Central, West, and South).
Please explain why setting the PRM based on high-risk hours, which may be at a lower level than peak load, will ensure sufficient capacity resources to meet peak load, and provide an example illustrating it. We understand that all resources will be included in the LOLE model. But would setting a PRM lower than what would be needed to serve peak load potentially result in retirements or selling capacity in other neighboring markets that would be needed to serve peak loads in MISO? Has MISO considered a PRM mechanism that addresses both the peak load requirement and the capacity requirements at non-peak LOL hours?
A forecast of future accreditation values will aid utilities in resource planning. We appreciate that MISO is planning to provide such a forecast via the annual Regional Resource Assessment (RRA), and we support using the RRA as the space to provide that forecast. A forecast that is 5, 10, and more years in the future will provide more useful information than the accreditation values in the current PRA.
Co-located and single-offer hybrid proposal
We are concerned that the “sum of the parts” concept does not adequately capture the combined benefits of a hybrid resources, such as storage-solar integrated resource. In particular, we are concerned that MISO’s storage modeling method simply added to its solar modeling method will impose imagined constraints that ignore the real, symbiotic interaction between storage-solar hybrid facilities. For example, in the summer, all storage facilities that are part of hybrid storage-solar facilities likely will be fully charged heading into the risk hours (which typically will be in the early evening when MISO reaches higher penetrations of solar). However, this benefit of a hybrid storage-solar resource may not be identified under MISO’s current approach. In fact, MISO identified the “storage fully charged heading into risk hours” assumption as a problem they have to address in updated storage modeling.
Please provide an explanation and evidence to show that the sum-of-the-parts accreditation methodology for hybrids does not reduce the full accreditation value these resources would receive if they were treated as a single resource.
While we understand that MISO may not be able to come up with a multitude of hybrid models to accredit different hybrid resource combinations, it is generally expected that the largest percentage of hybrids will be storage-solar hybrids. Could MISO at least develop a storage-solar hybrid model?
Planning Reserve Margin Requirement and LSE obligations
Why is PRMR expected to increase with LOLE modeling enhancements? What are those enhancements? When will they be implemented? Will MISO engage with stakeholders on the specifics (especially inputs and assumptions) of those modeling enhancements? Please provide examples of how the PRMR will change.
Please clarify what MISO means by “translation to PRA,” from “percentage based” to “MW based, TBD” (slide 18)? Including, under “MW based,” what is “TBD,” and when will that be determined?
Invenergy thanks MISO for the opportunity to provide comment on the updated accreditation proposal provided to stakeholders at the July 2023 RASC.
Invenergy appreciated the provision of Solar and Coal DLOL data as well as preliminary results by technology. Such materials are critical for good-faith efforts to evaluate the benefits and shortfalls of DLOL.
After preliminary analysis, Invenergy is concerned that the DLOL hours flagged by the model may be too narrow and underestimate system risk. It must be noted that Invenergy acknowledges that the SERVM software and LOLE model are best practice and industry standard. However, preliminary analysis shows substantial misalignment between DLOL hours (in the data provided) and MISO’s previously published RA Hours.
Misalignment between DLOL Hours and RA Hours suggests that the forward-looking model may not be capturing hours which have recently experienced risk:
Invenergy acknowledges that DLOL Hours and RA Hours have different inputs (predictive modeling vs historical data), scopes (loss-of-load hours vs 65 tightest hours per season), and applications (class-level vs unit performance).
However, comparable measures of risk should expect to see some convergence. This does not seem to be the case with DLOL Hours, which heavily weight risk during one part of the year during the same hours, whereas RA hours seem substantially more diverse not only across seasons, but also across hours.
Back-of-the-napkin analysis shows that RA Hours capture far more DLOL Hours than the inverse, which gives the impression that RA Hours capture more periods of risk.
This issue will not be addressed if RA Hours are merely applied to unit-level performance.
[Please see additional attachment for visualization.]
(1) Invenergy would request that MISO provide initial results for the potential options listed on slide 7 of agenda item 08ai, specifically expanding hours within a certain margin. Invenergy would also ask MISO whether there is any perceived reliability risk in the difference between DLOL Hours and RA Hours.
(2) Invenergy would recommend that MISO provide more detail on how data transparency will be enhanced, were the DLOL model to be implemented. With stakeholders able to focus on inputs to one model, the increased resources and feedback that can be allocated would result in more robust assumptions, if access to data is provided.
(3) Invenergy supports MISO’s preliminary exploration of hybrid participation models and asks that more time be allocated to this discussion. Invenergy would be interested to know whether purely BTMG storage added to solar behaves differently (and can therefore provide different reliability value) than co-located storage connected to both a solar project and the grid.
(4) Invenergy encourages further discussion of the change in PRMR attributable to DLOL. Without thoughtful discussion, there is a risk that the benefits of PRMR reduction will not contribute to the capacity price signal.
Invenergy thanks MISO staff for their consideration and looks forward to ongoing collaboration.
- .
Minnesota Power Direct LOL Stakeholder Feedback
Minnesota Power Appreciates the opportunity to provide feedback regarding the proposed changes to the resource accreditation methodology. There have been robust stakeholder interactions at the RASC meetings. Stakeholder feedback process is a crucial element in working through the process of refining the proposed methodology.
Overview Comments
It is critical to consider that this proposed methodology for the Direct Loss of Load method has not been utilized in any other Regional Transmission Organizations (RTOs), ISOs, any NERC Regional Entities, or any other international resource adequacy application.
Defined Approach
The LOLE modeling has been established to determine the planning reserve margin for the complete resource portfolio to quantify at a portfolio level the system planning reserves (requirement??). The LOLE modeling includes a wide range of sensitivities including unavailability of resources due to either planned or forced outage events. Taking the step of using the same model to determine the accredited capacity by generation class has an appealing attribute of using the same model and “currency” of accredited MW as what has been used to determine the planning reserve margin needs?, but the distinction of how the LOLE model comes up with the total system portfolio capacity requirements and how the DLOL can derive the accredited MW by generation class has not been shown. This is a key component of understanding.
We know that the adjustment factor is one example of how the LOLE model is able to arrive at the 1 day in 10 target, with either a perfect 100% load factor load or an ideal peaking plant being used if more resources are needed to reach the target. Suffice it to say, stakeholders need more information understanding how the accreditation by generation class derivation of the accreditation is a defendable approach, especially in light of an expected resource transformation on the system and the significant impact to accredited capacity value for certain resource types over the four seasons.
The DLOL methodology quantifies the generation class level availability of resources, using the LOLE model. The grouping of resource by class has been the (proposed??) approach to seek for a class level accredited value, expecting that a resource level assessment of accreditation using this approach would be “lumpy”. The most basic question of defending how the accredited value can be established by using the simulated LOL hours has not been answered. LOL hours are a very limited subset of identifying the availability of a resource and are by definition a modeled approach to simulating the hour where there aren’t enough resources to meet load. Resource adequacy is seeking to quantify the expected availability of a resource, and focusing on the simulated hours where there are not enough resources to meet load, is a small subset of the 8760 hours in a year.
The DLOL methodology is a marginal resource adequacy model, as has been explained with the expected trend of increased levels of solar generation on the system shifting the system peak to a later hour in the day when the solar generation isn’t able to provide power. The concern of properly accrediting resources to provide a stable value proposition for capital investments hasn’t been adequately described or defended. Furthermore, this methodology appears to not give credit for the value capacity resources bring to all hours of the year. It is focusing on developing accredited capacity value, and associated capacity revenue, to resources only available during a select few hours of the year. MP appreciates the focus on reliability during LOL hours, but it significantly devalues resources that provide resource adequacy during other times of the year. Consideration needs to be given how these other capacity resources are recognized and financially compensated for providing reliability during all hours of the year.
Another issue that hasn’t been adequately described is the viability of quantifying a generation class accreditation that is expected to shrink by 75%. In addition, the PRMR summary on slide 16 of the July RASC meeting grouped all generation into one line item, which was also not explained.
Presentation of Results and Recommendations
The RASC presentations have included high level objective descriptions regarding resource transformation, and now we have initial results showing ICAP/DLOL ratios by generation class compared to ICAP/UCAP ratios by generation class and a PRMR summary by generation class. Hourly reporting of LOL hours for coal and solar have also been provided to show the mechanics of how generically identified resources MW levels are summed up for each season, and the average approach to the DLOL creating an accreditation value for each season.
Risk of Load Issue
Linkage of DLOL to RBDC
More details on the hourly datasets provided and the LOLE model.
The following information is shown from the hourly datasets:
It was appreciated and useful to be able to see the hourly data for coal and solar to show the accredited calculations and to also understand the summary of LOL hours across the weather years to get a better understanding on what years have identified LOL hours. More data is needed to dig deeper into the understanding of the how the LOLE model can calculate the DLOL accredited values by generation class.
The following data is not known or understood from the hourly datasets
In summary, Minnesota Power has considerable concerns about the uncertainty on how the DLOL will allow for a stable and predictable methodology of quantifying resource accreditation, capability to equitably value all capacity resources contribution to resource adequacy during all hours, and insufficient time and information provided by MISO to thoroughly analyze given this is a first of its kind methodology. The small set of simulated hours where the loss of load is aligned with the modeled availability of resources, and limited level of evaluation of how the methodology works for a range of resource mix changes during unprecedented energy transformation all contribute to the concerns.
Xcel Energy appreciates the opportunity to provide feedback regarding accreditation reform.
DLOL Results by Resource Class
MISO needs to determine the potential reasons for the significant decrease to the thermal units winter and spring accreditation values. One reason may be due to the optimization of planned outages around the peak instead of the LOLH. Another reason may be the cold weather outage adder, so we agree with other stakeholders that comparing the accreditation values with and without this adder would be beneficial to isolate the impact.
Now is the time to start evaluating impacts from modeling assumptions and enhancements. MISO and stakeholders need to develop a complete set of improvements based on using the LOLE model for DLOL and PRMR; adding improvements in a piecemeal fashion could drive improper investment/retirement incentives in the near term.
We appreciate MISO's commitment to providing forecasted capacity accreditation values based on the Futures and ultimately the RRA. Ultimately, we would prefer a snapshot every 5 years, if possible, to use as an input to our Resource Plans.
PRMR
We strongly recommend that PRMR is calculated based on the LOLH to align the requirement with the accreditation. This marginal obligation calculation softens the societal benefit allocation issue with solar as it shifts the risk hours later in the day. As MISO stated during the RASC, the PRMR based on the LOLH for PY 23-24 needs to be included when the DLOL results for each resource are provided to the MPs.
We appreciate MISO's commitment to providing forecasted PRMR based on the Futures and ultimately the RRA. Ultimately, we would prefer a snapshot every 5 years, if possible, to use as an input to Resource Plans.
Co-located and Hybrid Resources
We agree with MISO that the sum of the output from the hybrid components capped at the interconnection service is appropriate, provides the proper signal for investment and provides the most flexibility for a hybrid to be comprised of any combination of resource types. There is no need to create a distinct hybrid capacity accreditation methodology for hybrid or co-located resources.
Initial Design Proposal - Options
We would like to see the year over year stability of using only LOLH vs using LOLH plus additional high risk hours. We are also interested in seeing analysis of EUE weighted LOLH. Are there significant differences? Regarding location-based resource classes we wonder if this would be reasonable only if PRMR is regionally calculated?
Please provide more information on why nuclear is not get more accreditation as a class in these models. The class Seasonal D-LOL results on page 5 of the presentation vary from 91% to 80% for nuclear. This seems incredibly low given the historical performance of nuclear (even in Winter Storm Elliot described in the other presentation on July 11th) and the fact that it has on site fuel that is not really subject to delivery issues like other thermal assets. Its seems that the current ISAC method is a much better indicator of the performance of nuclear generation.
Initial design proposal
MISO needs to push back its October-November FERC filing due to the issues identified below. MISO needs to provide additional information and resolve uncertainties so that stakeholders have at least 60 days to review before MISO submits a FERC filing. Specific issues and concerns include:
DLOL results by resource class
Planning Reserve Margin Requirement and LSE obligations