In the January 17 meeting of the Resource Adequacy Subcommittee (RASC), MISO presented accreditation reform draft tariff language and planned outage modeling methodology used in LOLE analysis. Stakeholders were invited to submit feedback.
Comments are due by February 7.
WEC Energy Group submits the following comments on the “Draft Tariff Redlines for Accreditation” that were posted with the January 17th RASC meeting materials.
Module A:
We do not understand why MISO wants to avoid the term Direct-LOL (DLOL) in its tariff (footnote on slide 24 of the Jan 17th RASC presentation). The current definition of Unforced Capacity (UCAP) should remain in the tariff to provide a distinction between UCAP and DLOL accreditation. Even in MISO‘s own presentations, UCAP is compared to DLOL accreditation. UCAP, as currently defined (including for wind and solar), is widely understood and used in many other processes. A distinction between UCAP and DLOL accreditation is necessary to provide transparency and avoid confusion.
Rather than redefining UCAP, the proposed tariff should redefine SAC. The definition of SAC only requires slight modification to account for the DLOL process without requiring any change to the definition of UCAP. Seasonal Accredited Capacity (SAC): The amount of Capacity, for a given Season, in MW, assigned to a Planning Resource after accounting for either its availability in the LOLE analysis or historic availability, as applicable. Schedule 53A only requires a change to replace UCAP with “SAC (resource class)”. SAC, rather than UCAP, is a much better description of the DLOL accreditation. DLOL accredition is not the unforced availability of a resource; rather, it is the availability of a resource during times of need (DLOL + expanded hours), regardless of why it is not available (forced outage, planned outage, or energy limited).
Schedule 53A:
WEC Energy Group shares the sentiment expressed by several stakeholders during the January 17th RASC meeting that a more formal definition of Resource Class is needed. At a minimum, expand the Resource Class definition to include resources with similar technologies, operating characteristics, fuel type, and vintage. Preferably, the tariff should include the same resource class definitions MISO has been using in recent presentations, until such time those definitions change through the stakeholder process.
Section II.A. should include a description of “expanded hours”, the hourly weights and the cap in the class-level calculations. At a minimum, the detail on these three items within the Jan 17th RASC presentation should be included in Schedule 53A.
Mississippi Public Service Commission (“MPSC”) and Louisiana Public Service Commission (“LPSC”) Response to RASC: Accreditation Reform (RASC-2020-4, 2019-2) (20240117)
At the January 17 meeting of the Resource Adequacy Subcommittee (RASC), MISO presented accreditation reform draft tariff language and planned outage modeling methodology used in LOLE analysis. Stakeholders were invited to submit feedback.
Comments are due by February 7.
Feedback
First, the MPSC and LPSC direct MISO to the MPSC’s earlier feedback – MPSC Response to RASC: Accreditation Reform (RASC-2020-4, 2019-2) (20230822-23). These concerns continue to be unaddressed:
1. MISO’s DLOL proposal, particularly as it applies to solar generation, continues to focus on limited and discreet hours of low margin, which undervalues solar generation’s contribution to resource adequacy.
2. MISO did not adopt Invenergy/NextEra’s proposed hour expansion. MISO’s limited hour expansion proposal is insufficient.
3. MISO has yet to address the IMM’s concerns regarding underestimation of storage needed to firm up forecasted renewable resources or over forecasting of renewables.
4. MISO has yet to explain storage dispatch assumptions. Will batteries be dispatched to meet operational reliability or based on owner economic signals? Should batteries be compensated based on cost of service (reliability function) or price signals (economic dispatch)? The method chosen will affect investment in storage.
Second, assuming MISO continues on its current course, the MPSC and LPSC agree with comments provided by WEC Energy group.
Third, considering implementation is planned for PY28/29, rather than rushing now to make making multiple (incremental) filings that acknowledge that accreditation is still a work-in-progress, the MPSC and LPSC ask that MISO complete the design, e.g.:
1. identify with specificity the resource classes,
2. identify weighting of hours,
3. address LSE allocation of PRMR proposal,
4. explain battery dispatch assumptions,
5. address the IMM’s concerns regarding storage assumptions and renewable over-forecasting, and
6. make repairs to the modeling.
MISO’s desire to complete these tasks must be weighed against the need of LSEs, who will be making billion-dollar investment decisions, to understand the entire proposal.
Fourth, MISO’s response to some Action Items were ambiguous. On Slide 18 – with respect to allocation of PRMR, “MISO acknowledges that any change to allocation of PRMR should involve careful consideration of several aspects, including but not limited to cost-of-service principles.”
It is not clear how allocation should vary based on cost-of-service. And it is not clear whether one of the considerations MISO will use addresses the need for “attributes” necessary to provide ancillary services in the future.
Before MISO creates a “proposal,” stakeholders need a forum to discuss these issues at length, as they will likely lead to significant financial investment and the potential for cost shifting among zones.
Fifth, the MPSC and LPSC are concerned that nuclear power’s DLOL accreditation is lower than it should be. Since MISO South is nuclear-heavy, this mis-accreditation must be addressed prior to MISO’s FERC filing. MISO is not addressing this issue because they claim that they do not have time to address it before the FERC filing. Failing to address this before filing is unreasonable as it will drastically affect MISO South’s ability to satisfy resource adequacy requirements.
The Arkansas Public Service Commission (APSC) provides its Comments to the Resource Adequacy Subcommittee regarding the proposed Direct Loss of Load methodology and corresponding changes to Resource Accreditation:
The APSC recommends that MISO file its proposal with the Federal Energy Regulatory Commission in the later quarters of 2024, while maintaining the proposed implementation date. The APSC recognizes that there are significant concerns with the LOLE model, and how MISO is proposing to address accreditation of storage resources. APSC therefore asserts that additional conversations are warranted to address these concerns.
The APSC also is concerned that MISO has not sufficiently listed or explained what the specific classes are for the class accreditation, or why MISO believes those classes are sufficiently distinct to justify being in separate classes, or why MISO believes generators within those classes are sufficiently similar enough to justify being in the same class. It is imperative that the specific classes are identified and substantiated so that Load Serving Entities may make informed decisions for which generators to acquire for purposes of meeting their capacity requirements. It is also imperative that generators are properly classified so as to recognize each generator’s specific traits which may differentiate it from other generators and allow for proper recognition of their capacity value.
Given the complexity of the issues and the numerous concerns of other stakeholders, the APSC believes that it is in the best interest of all stakeholders for proposal to be filed after additional conversation and refinement. Additional time would allow greater digestion and understanding of the models and methodology, and would likely result in a more reasonable proposal. Further, additional conversation would not necessitate delay in MISO’s implementation date, and therefore would not harm MISO’s efforts to ensure resource adequacy.
Xcel Energy appreciates the opportunity to provide feedback regarding Schedule 53A Tariff redlines:
We appreciate the inclusion of the requirement to provide indicative results until Schedule 53A is effective. However, one of the main reasons for using the DLOL methodology is to provide resource investment and retirement signals for MP's planning purposes, so we would like to see another paragraph added to require MISO to provide forecasted capacity accreditation and load obligation requirements for a specified time, such as 5/10/20 years out.
MISO needs to define metrics in the Tariff that will demonstrate the resources classes have been defined appropriately and do not bias a particular resource type.
Xcel Energy provides the following feedback regarding the MISO proposals at the January 2024 RASC.
Expanded Hours
We agree that expanding the critical hours used in the LOLE analysis to tight margin hours is appropriate and will likely increase the year over year stability of the results. We also agree that LOL hours should receive the highest weights when compared to the tight margin hours, but since these are modeled results it may be more appropriate, and transparent, to apply one weight to all LOL hours and a lesser weight to the tight margin hours. We are also not convinced that only the estimated margin should be used to weight each hour as it may also be important to incorporate the duration as well as the magnitude of the hours into the calculation, based on an analysis of historical critical events.
LRR Reliability Requirements Calculated with DLOL
MISO proposes to calculate LRR by applying the system-wide DLOL percentages to the zonal capacity. This does not account for the resource mix differences within each zone. We believe the approach should be refined to incorporate the zonal capacity by resource class.
Filing Timeline
We still believe that MISO should address the LOLE planned outage modeling improvements, define the metrics used to assess the appropriateness of the resource classes and provide at least a high-level concept of the LOLE modeling of storage resources prior to the filing These issues will have a significant impact on the future accreditation values and load obligations so stakeholders should have an opportunity to discuss and evaluate them prior to the filing, without impacting the implementation date.
DTE appreciates the opportunity to provide feedback on MISO’s accreditation reform proposal. DTE supports MISO’s move towards the DLOL accreditation methodology and DTE believes that the change will lead to more equitable accreditation across all resource classes.
However, there are some outstanding items we would like clarity on prior to the filing to FERC. The updated language related to resource class determination in the tariff redline does not provide enough detail pertaining to how classes will be determined, nor does it provide criteria to determine when resource classes may need to be updated in the future. DTE recommends this language be updated to include clearer requirements. DTE also requests a section be added to the tariff that describes how MISO will provide yearly forward looking DLOL class level accreditation forecasts based on RRA inputs for each year included in the RRA. These forward looking DLOL class level forecasts will be necessary to aid resource planners in accounting for MISO’s accreditation methodology. Without MISO’s forecasts in this area, it will be incredibly difficult for resource planners to understand future accreditation values of units and it is not practical to expect resource planners to be able to model MISO’s accreditation methodology for future years. Additionally, DTE still has concerns around the DLOL Class % methodology when dealing with classes that have a small sample size. DTE recommends that MISO come to the next RASC with a proposal on how to ensure that classes with a small sample size are not disadvantaged due to poor simulation results from the LOLE model.
Outside of the tariff filing, DTE believes 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. MISO should heavily consider taking learnings from the Astrape presentation pertaining to energy limited resource dispatch and how modeling of batteries can have a large impact on the accreditation of all resource types. Related to this, MISO should also take time to reevaluate how planned outages are modeled in the LOLE study as the placement of those outages will have a significant impact on the final DLOL class level percentages. MISO should provide stakeholders with analysis focusing on how different levels of fixed vs. variable planned outage placement impacts DLOL class percentages.
DTE also has concerns with the proposed LRR calculation methodology. Using the MISO system-wide resource class DLOL% will unfairly calculate the required LRR for each resource zone that has a zonal class level DLOL % that is different from the overall MISO DLOL %. DTE recommends that MISO move toward using a LRZ specific resource class DLOL% in the LRR calculation or using the actual SAC values for the resources in each zone for the calculation.
The OMS Resources Work Group (RWG) provides this feedback to MISO on its Direct-Loss of Load accreditation proposal. This feedback is from an OMS work group and does not represent a position of the OMS Board of Directors.
LOLE Model Considerations
The RWG notes that there is a potential risk to resource planning whenever changes are made to the LOLE model. That said, we have been an active participant in prior LOLE modeling enhancements, always supporting continued improvements to ensure the model accurately reflects the risks that the system is facing. We note that historically these changes have been implemented without requiring a tariff filing, and we expect that same process to continue.
To help set expectations, we suggest MISO make a list of and explain all planned modeling enhancements, the issues MISO is intending to solve with these enhancements, and identify any initial solutions that are known at this time. MISO should also explain why these changes do not need to be memorialized in the tariff.
Weighting & Cap
For the “pre-determined maximum number of hours in the Season” of hours beyond those with LOL events, MISO should explain whether a cap on the number of hours per Season per model weather year would be preferable (rather than the proposed per season across all weather years).
In the proposed Tariff language, MISO plans to calculate the Resource Class accreditation based on a weighted average of the Resource Class performance during critical hours, with a weighting system that assigns higher weight to the most critical hours. The RWG has significant concerns with the approach MISO presented at the January RASC to accomplish this. These concerns are summarized below:
For the hours beyond those with LOL events included in the critical hours, MISO should explain why taking a high number of hours from a single model weather year would not bias the results to one particular type of risk, in a way that conflicts with the equal weighting that each weather year receives in calculating the LOL metric.
Finally, the RWG asks whether MISO sees a need to evaluate the correlation between the performance of generators during critical hours and RA hours within each class. Is it possible that if these values are not strongly correlated, the process would send mixed signals such that resources in a class that perform well during LOL hours increase the size of a class pie for all resources within it but may not be the same resources that perform well during RA hours? Our understanding is that the unweighted expanded hours would have lessened the potential for mixed signals.
The Entergy Operating Companies ("EOCs")[1] appreciate the opportunity to provide feedback on MISO’s accreditation reform proposal.
DLOL Process and Timeline
The EOCs believe that the near-term LOLE improvement ideas identified in the RASC at the end of 2023 are critical to the final results produced by the DLOL methodology, especially the modeling of storage resources and the treatment of planned outages in the SERVM model. Additionally, the EOC’s believe that MISO needs to identify the final resource class definitions before filing with FERC and that the two most reasonable options for class definitions are to continue with classes defined by technology type or to combine all thermal units into a single class. For these reasons, the EOCs request that MISO complete the work on these LOLE improvement areas and class definitions prior to filing the DLOL proposal with FERC, which will require MISO to delay the target filing date to later in 2024. The EOCs believe that even if the FERC filing target is delayed beyond Q1 of 2024 that the current 28/29 PY implementation date can be maintained, while still providing Market Participants reasonable lead time to adapt to the changes. The EOCs understand that the improvements to the LOLE model will likely not be specified in the tariff language, but it is still critical for Market Participants to conceptually understand how the LOLE model will be adapted to implement DLOL in order to evaluate the tariff changes that MISO will be filing with FERC.
Planned Outage Methodology
As stated above, the EOCs request that MISO conduct a stakeholder process to evaluate how to improve the planned outage methodology utilized by the SERVM LOLE model. The current planned outage methodology used by SERVM does not attempt to allocate planned outages for each resource class in a fair or even manner. The model only focuses on the total number of MWs in planned outage each day and does not attempt to allocate an even proportion of nuclear, gas, coal, etc units on outage each day. As a result, random scheduling decisions made by the SERVM model can greatly penalize one technology class while benefiting another. There are two potential options to resolve this issue; (1) adjust the planned outage schedule on the front end (input) so that there is a fair allocation of planned outages for each resource class on each day, or (2) continue to use the existing scheduling methodology but perform a post-processing adjustment to resource class accreditation ratings to account for the unequal planned outage schedule used by the model. The EOCs believe that MISO should conceptually identify how this problem will be resolved before filing with FERC, but that it isn’t necessary for the identified solution to be implemented by the vendor in SERVM before filing with FERC.
It is important to note that the SERVM planned outage methodology only needs to be re-evaluated if MISO breaks up the thermal resources into separate classes. If there is a single class for all thermal resources, then this is no longer an issue because the impact of planned outages is being contained to only one class and we don’t need to worry about fairly allocating the impact across multiple classes.
Lastly, the EOCs believe that the SERVM planned outage scheduling logic needs to be consistent across all resource classes with regard to using a fixed (user pre-determined) outage schedule or providing planned outage rates to the model and allowing the model to determine the outage schedule. In prior LOLE studies, MISO has used a fixed outage schedule for nuclear outages while using a planned outage rate input for other technology types. Going forward, the EOCs request that MISO use a planned outage rate input for all technology types.
LRZ Local Reliability Requirements
The EOCs do not support the LRR calculation methodology presented by MISO in the January RASC. This methodology would determine LRRs based on assigning MISO-wide class average accreditation values to all units which would create a disconnect or misalignment between the LRZ LRR values and the accreditation values in each LRZ and could falsely cause an LCR constraint to bind in the PRA, triggering CONE pricing. Instead, the EOCs believe that MISO should calculate the LRZ LRR values using unique individual accreditation ratings for each unit in an LRZ.
DLOL Methodology
The EOCs believe that MISO should use the non-weighted expanded hours approach presented at the end of 2023. MISO’s latest proposal places too much emphasis on the loss of load hours relative to the tight margin hours. Additionally, the EOCs are concerned that caping the total number of hours at 1,950 has the potential of constraining the LOL hours to a small number of modeled weather years which may create high volatility from one planning year to the next as older weather years drop off and newer ones are added. The EOCs request that MISO perform a volatility analysis on the DLOL results examining the impact of the 1,950 cap vs not using a cap.
[1] The Entergy Operating Companies are Entergy Arkansas, LLC, Entergy Louisiana, LLC, Entergy Mississippi, LLC, Entergy New Orleans, LLC, and Entergy Texas, Inc.
AES Indiana (AESI) appreciates the opportunity to comment on the updated MISO accreditation reform proposal. As shared in our May and September 2023 dated comments, AESI still has concern over solar accreditation value degradation and understanding storage accreditation into the future. In addition, we’d like to comment on how the term “Resource Class” is noted in the Tariff redlines. Specifically, the granularity of each Resource Class needs better definition. For example, within the Resource Class “Storage”, AESI requests that MISO differentiate between a 4-hour storage asset and a 6-hour storage asset. For example, the 6-hour storage asset would have greater ability than the 4-hour storage asset to provide during a peak event, and therefore should receive capacity accreditation accordingly. Another example, for thermal resources, “Resource Class” differences between a single fuel and dual fuel combustion turbine (CT) units also have different capacity value and should be accounted for.
Overall, AES Indiana continues to need to understand the impact and devaluing of non-thermal resources with this new DLOL approach to accreditation. We are constantly working through an RFP process for new generation and IRP modeling that requires correct forecasting accreditation by resource type for our planning and RA needs. AESI urges MISO to provide a better accreditation outlook for Storage as a Resource Class. MISO has not shared in public stakeholder materials results of the model with improved battery modeling inputs which would lower accreditation compared to what is being presented now. For example, in the November 2023 RASC Presentation “Market Redefinition: Accreditation Reform” on Slide 21, Storage shows 85% for 2027 and 89% for 2032 for the Summer Season. Is this a final and accurate depiction of Storage accreditation value? Do these values, which are near the class average under the current methodology, simply highlight future resource changes but not the changes needed from complete and final DLOL modeling? To further highlight the need for more clarity on Storage, the EEI MISO CEO Deck titled “Executing through the Energy Transition: MISO Update for EEI”, dated January 11, 2024, on Slide 8, Storage is not even listed. Also, does MISO plan to use RA hours for Storage, and if so, does that disproportionately impact storage accreditation? Until these open questions are answered, AESI would ask MISO to delay filing these Accreditation changes at FERC.
See attached comments from MidAmerican Energy Company.
AMES, AMP and IMPA generally support WPPI's feedback.
I'm happy to discuss.
David Sapper
dsapper@ces-ltd.com
Invenergy appreciates the opportunity to provide comments on the Accreditation Reform (RASC-2020-4 and 2019-2) presented at the January 17 RASC.
Tariff Redline Feedback:
Module A DLOL Definition:
The Tariff definition for Direct Loss of Load (DLOL) is missing from the redlines. Rather, MISO has opted to modify the Unforced Capacity definition. A closer look also reveals the use of the acronym UCAP also needs to be included in the Unforced Capacity definition, much like Module A defines ties the ICAP acronym to the Installed Capacity “(ICAP)” definition. We find the mingling of Unforced Capacity (UCAP), which the industry understands as a calculated value using forced outage rates, and Direct Loss of Load (DLOL) will confuse matters. We encourage MISO to delineate the two terms and avoid comingling the two. MISO was able to do that with the introduction of Seasonal Accredited Capacity (SAC) in earlier reforms.
Schedule 53A Indicative Results:
Invenergy highly supports MISO publishing indicative results during a transition period. This will result in a higher probability for the reforms to be successfully executed as both staff, processes and technology improvements benefit from iterations of implementation before a cutover to financially binding production results. We cannot emphasize this enough. This strategy by MISO is a win-win for both MISO and Market Participants.
Schedule 53A Resource Classes:
The Resource Classes section ought to be more robust and defined. Detail that we recommend MISO include:
By doing so, this would be similar to how the Tariff manages Local Resource Zones. For example:
In a similar fashion, resource class should also be memorialized in the Tariff.
General Reform Feedback:
DLOL Methodology:
We agree with other stakeholder feedback that storage dispatch heuristics are newer concepts that have significant impact to the accreditation of the resource and pursuant to FERC’s rule of reason must be included in the Tariff.
If MISO is successful in getting FERC acceptance of the DLOL capacity accreditation reforms, Invenergy supports MISO further exploring an Energy Equity dispatch methodology over a chronological method used in the SERVM LOLE software when implementing. This will measure more accurately and assign capacity value to resources based on the contribution to system reliability during periods of highest risk. A chronological dispatch of storage in the software likely does not comport with the marginal value of resources. Rather, in simulations on days with load shed driven by the exhaustion of energy limited resources, the dispatch of these resources can be prioritized to shave the net load peak, resulting in constant hourly Expected Unserved Energy (EUE) across all hours with any energy limited dispatch. We believe this Energy Equity method aligns with MISO’s guiding principles and we encourage MISO to spend more time on this, working out those details with Stakeholders and memorializing the method into the Tariff.
Planning Reserve Margin Requirement (PRMR)
We agree with several stakeholders that the PRMR detail should not be overlooked and should be paired with the DLOL reforms filing either together in the same filing or in a separate but related joint filing, similar to what MISO did with recent generator interconnection reforms. Failing to do so does not allow for a full evaluation of MISO’s contemplated reforms.
Invenergy thanks MISO staff for their time and consideration.
AES Clean Energy commends MISO's effort on Resource Accreditation reform and welcomes the opportunity to provide feedback. While we support the transition to the DLOL method, we believe that at present MISO’s proposal for accreditation reform filing at FERC is premature. We highlight our concerns and propose solutions to address them. Finally, we seek answers to some questions we have about the overall resource accreditation process.
Concerns and proposed solutions:
Lack of clarity on key elements in the Tariff:
The Timing and Impact of proposed enhancements to LOLE model on the downstream process of accreditation:
While we appreciate MISO's emphasis on future accreditation values for long-term investment, we express concern about the misalignment between planned LOLE model enhancements and premature filing of accreditation reforms. MISO highlights in the Attributes Roadmap a “As the model improves, results of downstream processes (such as accreditation or requirement setting) will be impacted. Some of these recommendations may have significant implications in those processes." If with the proposed enhancements, accreditation values change frequently and significantly, this will send confusing and noisy signals to all market participants and can erode the credibility of MISO published accreditation value. Furthermore, publishing forward accreditation values with DLOL methodology where the LOLE model itself is changing will further add to the confusion regarding accreditation values.
At present stakeholders who oversee effectively communicating to their respective institutions, investors, state authorities who make PRMR procurement and investment decisions, are already grappling with the LOLE model, DLOL methodology and its implications for resource accreditation. Furthermore, the added layer of a premature resource accreditation reform filling at FERC that lacks clarity on key design elements like resource class definition, calculation of critical hours, treatment of storage resources etc. will further complicate communication and accentuate uncertainty already existing in a transitioning market. We urge MISO to consider the unintended consequences a premature FERC filing on Resource Accreditation reforms can have. Economist Nicholas Bloom in his Econometrica paper finds, “higher uncertainty causes firms to temporarily pause their investment and hiring.” (Source: Bloom, Nicholas. “The impact of uncertainty shocks." Econometrica 77.3 (2009): 623-685.” (Highlighted for emphasis).
Project development and online implementation require substantial time (several years) and financial commitment (millions of dollars). Unfortunately, uncertainty regarding resource accreditation and classification, critical hours calculations etc. adds to this burden. Essentially, generators are expected to invest heavily without knowing the potential value their resource will hold. Demanding investment decisions under such opaque conditions is unreasonable and unfair to generators. The lack of robust capacity forecasts poses a significant hurdle for generators making investment decisions and can significantly hurt resource adequacy within MISO footprint.
We understand that model improvement is a continuous process. Aligning with MISO’s insights and findings from Attributes Roadmap, and keeping in mind the three-year transition period MISO is considering, we propose:
By efficiently using the transition period, MISO can ensure that the solutions MISO puts in place are robust, better understood, hence guide investment decisions as intended.
For the reasons discussed, we strongly believe that a FERC filing by Q1 2024 is premature and will only add to the pre-existing uncertainty in the MISO market that can have sever unintended consequences for investment decisions and resource adequacy within MISO footprint. We are committed to working together with MISO during the transition period and are happy to continue this conversation to ensure a smooth market transition in MISO.
General questions on resource accreditation using LOLE model:
The Environmental Sector appreciates the opportunity to provide feedback on MISO’s proposed Draft Tariff Redlines for Accreditation (hereinafter, “Draft Tariff”)[1] and its accompanying January 17, 2024 RASC Presentation (hereinafter, “Presentation”).[2]
The Environmental Sector believes that MISO is issuing the Draft Tariff prematurely. At the January 17, 2024 RASC meeting, MISO introduced a new weighting concept that will have a material effect on how MISO’s accreditation program will operate, and MISO should consider official stakeholder feedback before incorporating this concept into their accreditation proposal. MISO should also wait to file the new accreditation proposal until it has completed its LOLE model enhancements initiative.
Proposed Weighting of LOL and Tight Margin Hours
The Environmental Sector has four major concerns with the weighting proposal that MISO first introduced in the Presentation.
First, the weighting proposal is arbitrary. In the example of the hourly weights calculation shown on slide 9 of the Presentation, MISO chose to weight each hour based on its deviation from the largest margin achieved among all LOL Hours and Expanded hours in a given season. This resulted in “Hour 2” being given a weight that is 13 times larger than “Hour 3,” and infinitely larger than “Hour 1” (which set the standard). MISO has not offered a policy justification for so powerfully (and linearly) determining the weight of different hours–in particular, why this weighting approach is more appropriate than other methods. As an example, a stakeholder who is a member of our sector suggested alternative weighting proposals to MISO staff in the course of the Presentation (e.g. weighting “Hour 1,” “Hour 2,” and “Hour 3” at 1.0, 1.13, and 1.01 respectively; or at 1.0, 2.3, and 1.1 respectively), and MISO staff responded that they had not considered these or other calculation methods. MISO’s chosen weighting methodology may be mathematically coherent, but that by itself cannot justify one weighting option chosen among many possible (and equally mathematically coherent) alternatives.
Second, MISO’s weighting proposal undermines MISO’s previous commitment from the November RASC meeting to appreciably expand the hours considered under its resource accreditation methodology as a means to consider LOLE modeling errors. While MISO’s proposed DLOL accreditation methodology attempts to be more accurate than its predecessor, much of that inaccuracy remains because of its reliance on the LOLE model, which MISO has acknowledged needs to be updated to account for the realities of a modern energy mix. The addition of expanded hours to the DLOL analysis was meant to counteract the inherent inaccuracy found in all models, including the LOLE model, and also to smooth the volatility of annual accreditation values. But MISO’s proposed weighting methodology significantly reduces the benefit of adding Expanded Hours to its accreditation calculation in the first place, by severely curtailing their influence on unit accreditation and in some cases eliminating them. The obvious result of a weighting system that gives vastly more weight to the LOL hours than to the tight margin hours will be that the resulting accreditation values are predominately based on the LOL values. This is a step backwards.
Third, MISO’s weighting proposal is unnecessary because, as MISO discussed on slide 10 of the Presentation, MISO’s proposal already subjects expanded hours to a cap of 1950 hours minus the number of LOL hours (LOL hours themselves are not subject to a cap). As a result, not only is there no guarantee that all expanded hours will be included in a given season, there is no guarantee that any expanded hours will be included, i.e. if there are 1,950 or more LOL hours in a given season. The addition of the weighting methodology proposed by MISO acts to further reduce the value that expanded hours provide to MISO’s proposed resource accreditation methodology; but their utility was already limited to seasons where there are fewer LOL hours.
Finally, and perhaps most importantly, irrespective of the mechanics of such weighting methodology, MISO has not provided any meaningful opportunity for stakeholders to review the necessity of this proposal and its impact on accreditation values. This is particularly problematic, because this concept is already in the Draft Tariff,[3] meaning that MISO is proposing to file in its Tariff a concept that MISO has not yet sought stakeholder feedback on. This undermines the entire purpose of stakeholder participation and is inconsistent with MISO’s obligation to operate an open and inclusive stakeholder process. The concept of weighting–both the mechanics and the presence of the concept itself–deserves greater consideration and additional stakeholder discussion before MISO files the Draft Tariff with FERC.
LOLE Model Enhancements
While the Draft Tariff presents a superstructure for the mechanics of the DLOL resource accreditation methodology, the actual impact of the DLOL methodology remains largely unknown with respect to how it may impact accreditation values of particular resource classes, because it is based entirely on the LOLE model. This is particularly troubling, because MISO itself has acknowledged that its LOLE model needs updating, and it is actively working to “enhance” or modify the assumptions in its LOLE model. What that means is that, although MISO has previously provided indicative accreditation values for resource classes, these values cannot actually be relied on by stakeholders and most importantly system planners, because the LOLE model is expected to change those values (in some cases greatly). This is especially true for electric storage resources, for which MISO has made clear on multiple occasions it is planning a major modeling update with as-yet uncertain anticipated modeling outcomes. The modeling of planned and maintenance outages will also have a significant impact on the LOLE results, which MISO has not yet completed. Stakeholders need to not only understand the DLOL proposal, but also how that proposal interacts with the LOLE model to achieve specific accreditation values, in order to be able to fully evaluate the impact of MISO’s DLOL methodology.
The Environmental Sector supports improving the LOLE model and believes that such improvements, especially with respect to storage and planned and maintenance outages, should be completed prior to MISO’s filing of its upcoming accreditation tariff. These improvements should not take years but months to determine and will not impact MISO’s proposed implementation date for the DLOL proposal. Considering the fact that the DLOL accreditation methodology will not go into effect until the 2028/29 planning year, MISO can afford to–and indeed should–take the time to clarify its LOLE model changes, as well as the resource classes it plans to use (as we discuss further below), prior to filing so that stakeholders will have a much fuller picture with which to analyze the efficacy of MISO’s proposed resource accreditation methodology. MISO has indicated that it wanted to file its resource accreditation methodology in March in order to provide parties more certainty when conducting state-level integrated resource planning, but even FERC approval of MISO’s switch to DLOL accreditation will not grant the “certainty” MISO claims. Rather, such certainty will come only when MISO clarifies these key assumptions in its LOLE model, because the results of the methodology itself will be uncertain until LOLE model “enhancements” are complete.
Draft Tariff language
Putting the critical issues raised above aside, the Environmental Sector offers the following feedback on the specific tariff language MISO has offered for review. Please note that our feedback below is entirely focused on MISO’s phrasing decisions and should not be construed as direct support for or opposition to any underlying policy choice.
Resource Classes:
Section I.C. of Schedule 53A briefly describes how MISO would define the resource classes that will be used to identify accreditation values. This language is too vague and overly broad. The vagueness will make it difficult for stakeholders to predict which resource classes will be used. The broadness could permit large changes within MISO’s Business Practice Manuals that undermine the certainty the definition seeks to create. Thus, stakeholders will not be able to easily predict future classes and accreditation values for their generators, which will interfere with their ability to appropriately plan for future generation needs. We request that MISO provide more detail on how resource classes will be defined, or in the alternative include the resource classes in the tariff.
The term “Resource Class” as used in this Schedule 53A shall mean a group of Capacity Resources, except External Resources, with similar technologies and/or operating characteristics that are modeled together for the purpose of Resource Class-level UCAP calculations pursuant to the provisions set forth in this Schedule 53A.
DLOL Provisions:
The tariff provisions in Section II.A of Schedule 53A describe the methodology that MISO intends to use to determine the class-wide accreditation values for each resource class. This section does not adequately describe the DLOL methodology, because it contains several significant policy choices that could and should be translated to specific written descriptions, as contemplated by the Commission’s Rule of Reason and judicial precedents.[4] Where key parameters in a complex methodology are left unstated (e.g. “smallest differences” or “higher weight” in the language below), a lack of interpretive principles either in the tariff text itself or in industry practice will leave Market Participants and stakeholders at a loss to understand what is intended.
Resource Class-level UCAP in a Season will be determined by calculating the combined availability and performance of all resources within that class during critical hours in the LOLE analysis. Critical hours are defined as the set of hours with loss of load events and, for those Seasons with a small number of hours, the hours with smallest differences between available capacity and load in the hours (subject to a pre-determined maximum number of hours in the Season and a threshold for the magnitude between available capacity and load). Resource Class-level UCAP will be calculated as the weighted average of the Resource Class performance during critical hours, with a weighting system that assigns higher weight to the most critical hours.
“Critical hours'' should be defined in terms of actual hours per season, including the hours used for a cap and the percent margin threshold used to determine expanded hours. This is consistent with how RA hours are discussed later in Section 53A, where the target is 65 such hours and where calculation steps for various concepts are actually delineated.
Similarly, the weighting mechanism needs to be more clearly defined, just as the weight between Tier 1 and Tier 2 hours are defined in Section VI of Schedule 53A. Please see our comments in the first section above for further discussion with respect to the concept of weighing critical hours.
[1] MISO, Draft Tariff Redlines for Accreditation, published on January 23, 2024, available at https://cdn.misoenergy.org/20240117%20RASC%20Item%2007a%20Draft%20Tariff%20Redlines%20for%20Accreditation631388.docx, hereinafter “Draft Tariff.”
[2] MISO, Market Redefinition: Accreditation Reform, presented on January 17, 2024, slides published on January 23, 2024, and available at https://cdn.misoenergy.org/20240117%20RASC%20Item%2007a%20Accreditation%20Presentation%20(RASC-2020-4%20and%202019-2631379.pdf, (hereinafter, “Presentation”)
[3] Draft Tariff, Schedule 53A, Section II.A.
[4] See, e.g., City of Cleveland, Ohio v. FERC, 773 F.2d 1368, 1376 (D.C. Cir. 1985).
This feedback is submitted to MISO on behalf of the Independent Power Producer and Power Marketer Sectors (the Joint Sectors). The Joint Sectors appreciate MISO considering stakeholder feedback on the DLOL proposal. The Joint Sectors also understand that resource accreditation must evolve with markets and changes in resource mix. Generally, we are concerned that MISO may file an incomplete proposal at FERC, when there is no time pressure. Implementation is set for 2028/2029. With the extended implementation timeline, the Joint Sectors feel all stakeholders will benefit from additional time to refine resource class definitions, Tariff language, the impact of the entire accreditation equation, and the redefinition of demand through the DLOL process.
The DLOL proposal assigns capacity value based on the assumptions MISO inserts in the Loss of Load Expectation (LOLE) model. All of the LOLE assumptions used in this LOLE model have not been published in advance of MISO submitting the DLOL proposal and have not been fully defined prior to finalizing the proposal. The assumptions that are essential to the DLOL calculation but not defined include: 1) critical hours or the formula to calculate critical hours, 2) weighting calculation, 3) resource classes, 4) calculation and application of DLOL, 5) PRMR and allocation to LRZs.
Furthermore, the LOLE model is based on assumptions such as planned outages, risk distribution amongst seasons (currently 0.1 in the Summer Season and 0.01 in the remaining seasons), modeling of energy limited resources (storage), and resource classes. Under the current proposal, there is no accounting for how MISO will utilize these assumptions in the LOLE model and this introduces the risk that these assumptions will change year-over-year. The variability in assumption development creates meaningful concern around future capacity accreditation values for all resource classes. This opaqueness frustrates investment decisions, asset deployment, integrated resource planning development, and project valuation for all market participants. The Joint Sectors request MISO include parameters on accounting of these assumptions in its DLOL calculation and LOLE model in the Tariff and present to stakeholders prior to the submission of the DLOL proposal to FERC.
Separately, the DLOL proposal disconnects resource availability from capacity accreditation. The resource class definitions fail to reward resources with higher-than-average availability and performance. The Joint Sectors encourage MISO to revisit the issue of resource class definitions that MISO raised in Q4 2023 and develop a methodology to incent improved availability and performance.
Coupled with adding resource class criteria to the Tariff, we encourage MISO include a timeline for redefining classes prior to each Planning Year, criteria which might trigger re-evaluation, and the actual result of the classes themselves. By doing so, class definition and redefinition would function similar to the Tariff’s handling of Local Resource Zones todays. For example:
In order to add certainty to the resource adequacy and accreditation process, resource class definitions and LOLE model assumptions should be memorialized into the Tariff.
Clean Grid Alliance, Solar Energy Association and American Clean Power Comments on MISO’s Accreditation Reform to the RASC
Feb 7, 2024
At the January 17, 2024 Resource Adequacy Subcommittee (RASC) MISO presented accreditation reform draft tariff language and the planned outage modeling methodology proposed for use in the LOLE analysis. Stakeholders were invited to submit feedback:
Clean Grid Alliance (CGA), the Solar Energy Industry Association (SEIA) and American Clean Power (ACP) appreciate the opportunity to provide feedback to MISO on its proposed accreditation reform filing presentation. Our primary concern with MISO’s proposal is that the proposed tariff language does not include core details on LOLE assumptions that have a significant impact on capacity accreditation. These are material details that should be reflected in the tariff, not a business practice manual. Furthermore, the absence of this detail makes it challenging for us and our members to offer either support or recommendations for adjustments.
Our specific concerns, include:
A model is as good as its inputs. MISO must finalize the details on the assumptions and design elements before filing. We request that MISO delay the filing by three months or whatever time is necessary to create a complete, transparent and “stakeholder reproducible” process for the filing that includes all details of the capacity accreditation methodology in the Tariff. This delay would not impact MISO’s proposed implementation date of planning year 28/29 and would allow for meaningful evaluation of the proposal by all impacted stakeholders.
[i] FERC Docket No. ER-22-495-000 p. 44.
WPPI reiterates our comments from the last round of feedback on this item (https://www.misoenergy.org/engage/stakeholder-feedback/2023/rasc-accreditation-reform-questions-rasc-2020-42019-2-20231107-08/).
In addition, we offer the following comments on the posted draft tariff changes: