RASC: Accreditation Reform (20230418-19)

The Environmental Sector appreciates the opportunity to provide the below comments responsive to MISO’s feedback request, RASC: Reliability Based Demand Curve (20230418-19), due May 5, 2023. We offer a few comments and questions below, organized by topic and slide.

MISO’s Clarifications and Rationale on Key Design Elements

Slide 9

We appreciate MISO’s openness to the future consideration of changes to the reference technology. As stated in our previous comments, we believe that the reference technology used to calculate CONE should reflect the least expensive technology that market participants are most likely to build.

Slide 10

While we agree that fewer PRA participants will necessarily lead to a steeper curve, we do not believe that this is necessarily a bad thing but rather is just a reflection of market reality. As stated in the middle box of this slide, it would still be better than the existing status quo. We only caution that MISO be careful to ensure any designs of alternatives for those market participants that opt out of the PRA are not eroded in such a way as to implicitly make participation in the PRA more attractive than it otherwise would be. In other words, we ask the MISO design the best options possible for market participants that opt-out, irrespective of any impact it may have on the RBDC and the PRA.

We support MISO’s proposed use of Expected Unserved Energy (EUE).

Slide 11

We reiterate our concerns that an annual participation model may result in greater-than-anticipated uplift costs, which ultimately would result in more emissions and higher prices for consumers. While it is true that the true-up is resource agnostic by design,^[1] in practice the true-up will still be used by the most expensive resources, which more often than not are non-renewable resources, and regardless, true-ups like this should be viewed as a fall-back to be avoided because of the excess costs they place on end-users. That said, we are pleased that MISO is already considering the expansion of the annual participation model to allow for resources to participate in blocks less than one year long (but longer than a season), which will help alleviate this problem to some degree.

Slide 12

Would you explain how retirements and suspensions were categorized into “Environmental Related” and “Economical”? More specifically, where is there overlap between these two categories and how is such overlap considered when deciding how to portray the cause of retirements and suspensions of resources? We understand that the data used on this slide came from the Independent Market Monitor and would thus appreciate any feedback that you could obtain from the IMM in this regard.

Impact of RBDC: Prior PRA

Slide 18

We are eager to learn more about MISO’s planned direction vis-à-vis the FRAP and AFRAP, and we look forward to learning more details at the next RASC meeting. In the meantime, are there any changes to the AFRAP design since MISO last presented details of it in the January 2023 RASC meeting that are now incorporated in the spread of potential impacts presented on slide 18?

Would it be possible for MISO to show the results of the analyses shown on slides 15-17 and 19-21 using new RBDC’s as impacted by one or two of the AFRAP participation scales shown on slide 18?

Next Steps

Slide 23

It is our understanding that MISO needed to have filed its RBDC proposal with FERC by June 1, 2023 in order for it to apply for the 2024/25 PRA. Slide 23 indicates that MISO is now planning to file its RBDC design with FERC by the end of this year, with any new design implemented for the 2025/26 PRA. If necessary in order to ensure the best RBDC design possible, when is the absolute last date by which MISO is able to file an RBDC design with FERC in order for it to apply to the 2025/26 PRA? While we do not wish to drag on the design phase any longer than necessary, we still want to understand where the most firm due date exists.

The NDPSC appreciates the opportunity to comment on the accreditation reform and the 3-year transition period presented during the April 2023, RASC.  We generally support the D-LOL approach and the use of Tier 1 and Tier 2 hours for nonthermal accreditation during the transition period. This would allow stakeholders time to make needed adjustments to their resource plans while balancing the pace with which the fleet is changing.

Minnesota Power appreciates the opportunity to provide stakeholder feedback to the RASC. 

Non-thermal Accreditation

1)     MISO Defined Key Takeaways from Slide 2 and Minnesota Power responses below each Item

a)     Under the proposed Direct-LOL approach, a reduction in accredited capacity would likely also reduce the Planning Reserve Margin Requirement (PRMR)

i)       The LOLE model is used to determine the PRMR.

ii)      The accredited values based on UCAP, ISAC, SAC are all independent of the LOLE modeling approach

iii)    Slide 8 is confusing calling out UCAP as an input to the LOLE model

(1)    It is our understanding that the GADS data is used to create a basis for the statistical draw of the LOLE model to have an outage draw on the system, and that the UCAP values aren’t used directly in the model. 

iv)    The conclusion stated of having a lower PRMR using Direct LOL is not in itself “good news” because if the Direct LOL accreditation shows a proportionately lower level of accreditation, there would not be the benefit of a lower PRMR. 

v)      At this point, we don’t know anything on the expected accreditation values, the impact of applying Direct LOL MW / ISAC ratios on a MISO wide-basis, so the net position due to an expected lower PRMR is not substantiated.

vi)    It is crucial to evaluate a full range of portfolios to show the impact of a transformed resource portfolio and how the accredited values are established both now and in the future.  This is crucial in order to assess the expected trend of having lower PRMR using the Direct LOL method for accreditation.  Without this full thorough evaluation of accredited MW values, we simply have no idea on the expected trend of the PRMR trends.[EP(1] 

vii)   In addition to the analysis, a key element that has not been discussed is how the seasonal distribution of the 1 day in 10 day Loss of load objective is established for all four seasons.  From the first seasonal LOLE model run, we have seen a very wide range on the PRM across the seasons.  The methodology on how the seasonal target values must also be thoroughly scrutinized and evaluated to make sure that the selection of the seasonal targets doesn’t result in unintended results of the seasonal Direct LOL accredited values by generation class. 

b)     MISO is proposing a 3-year transition with step-changes in accreditation with the goal of implementing Direct-LOL after 3 years

i)       The three year transition is crucial to evaluate the unknown issues that we won’t know at this point.  Based on the much more simple ISAC/SAC methodology, we found a number of issues that needed to be more thoroughly understood and updated. 

ii)      Another purpose of the 3 year transition approach should be to create a final approach on how the Direct LOL method will be implemented.  There are many many details that will need to be worked out, and the most important element of this evaluation phase is to determine if there is a proposed methodology that holds together. 

iii)    The resource attributes work must be moving in a much more substantive pace to make sure that the inclusion of the study findings are included in the final method. 

(1)    One example of how the attribute study findings could be included would be to use the LOLE model to show how the varying levels of different classes of generation will impact the system reliability, and give a leading edge of what type of generation will be given a more favorable resource accreditation. 

(2)    The use of a perfectly available CT to establish the 1 day in 10 target will in itself to come under greater scrutiny, because the divergence of the lack of peaking generation in the queue, compared to what I would expect to showing a larger amount of “perfect” capacity when determining the 1 day in 10 target.  In short, we will be asking for more PRMR based on the need for more “perfect” capacity, while not adding that type of capacity to the portfolio.  This would result in an overstated projection of system reliability, because the actual portfolio will not have as much dispatchable generation as what is being shown in the PRMR calculations. 

(3)    We don’t have a solid technical basis for using a simulated loss of load hour from a complicated statistical LOLE model to conclude that this is the approach of defining the hours that will be used to quantify the accreditation.  Clearly all hours are critical for maintaining the 1 day in 10 reliability target. 

(a)    Are the resources that are keeping the system from going into a state of unavailability for thousands of hours given proper accreditation?

(b)   Given the lower number of LOL hours for non-summer seasons, how will the LOLE model provide a defendable accreditation value for all generation classes?

(c)    Results should be shown across RRA and MISO Futures 2A cases for all resource types.

(d)   More LOLE modeling detail is needed to understand the distribution of output variables that are used to quantify resource accreditation. 

iv)    Due to the number of unknowns at this time, I don’t see the basis for making the assertion that the PRMR is expected to be lower using the Direct LOL method.

v)      MP recommends that MISO studies the Direct LOL method and alternative methods for a two year period.  During that time the current SAC methodology and auctions can be studied for lessons learned.  After two year study period start the three-year transition period.  In total – MP is recommending a five year time line.

c)     MISO is planning to share results from applying the Direct-LOL method to other resource types at the May RASC

i)       I am not understanding how we will only be seeing results on May 23-24, 2023, when the first time the method was proposed was November 30, 2022.  We have had no opportunity to provide any substantive input to the proposed method over these six months. 

(1)    The results for wind and solar across the RRA and Futures 2A have had a very wide range of results, and haven’t provided any level of understanding on the proposed accreditation methodology

(2)    The level of discussion on what the LOLE model is showing in terms of the accreditation by generation class will require a heavy lift and deep dive far beyond what we have seen to date. 

d)     Additional design discussions will continue through Q3 2023 with a targeted filing in Q4

i)       The filing should be limited to scoping the transition period, and not make any commitments of final implementation when we are in such a low position of understanding the method.

e)     Minnesota Power’s Core Objection to Using Direct LOL as a substitute for UCAP and ELCC for Wind and for all resources

i)       Current method discussion

(1)    Currently, the SAC methodology for thermal and hydro resources uses an overall measure of resource availability that reduces the ICAP MW by the forced outage impacts using GADS data as the standard reporting tool.  The wind is using average ELCC as a measure of showing the amount of load wind can serve on a system, taking into account the portfolio of other resources.

(2)    The ISAC/SAC method evaluates the resource availability over three years of hourly performance, taking into account the tightest margin hours for each season (Tier 2 ), and weighting the availability during those hours higher than the other hours of the year(Tier 1).

(3)    The LOLE model takes the GADS data and applies it to the ICAP of thermal resources, thereby providing a statistical representation of the unit availability when seeking to determine the total amount of capacity needed to meet the reliability target of 1 day in 10.

ii)      Direct LOL approach to replacing UCAP and ELCC

(1)    The Direct LOL accreditation method is using the simulated loss of load carrying capability hours as a primary means of assessing the availability of a resource.  

(a)    The hours of loss of load are deep within the modeling data details. 

(b)   The hours are clearly a subset of the hours used for Schedule 53 actual availability for 8760 hours a year. 

(c)    The narrow subset of hours does not provide an adequate assessment of resource availability for what is required 8760 hours a year, but makes the entire accreditation assessment based on a simulated occurrence of load not being served from resources. 

(2)    Determining the 1 day in 10 year standard

(a)    The utilization of the “Adjustment to UCAP” in the PRMR calculation introduces another perfectly dispatchable resource (or 100 percent load factor block of load) to arrive at the 1 day in 10 year standard, thereby creating more uncertainty as to the portfolio’s capability to reliably serve load, given the increasing departure of the resource types being added to the system compared to the perfectly predictable nature of the “Adjustment to UCAP” resource. 

iii)    Summary

(1)   The proposed Direct LOL accreditation methodology for all resources does not adequately provide a means of showing the availability of resources.  The narrow timeframe of loss of load hours does not provide an adequate window of evaluating the availability.  

Environmental Sector Comments on Accreditation Reform 

(RASC-2019-2, RASC-2020-4) (20230418-19)

1. The Environmental Sector Continues to Have Strong Misgivings About MISO’s Current Path

As an initial matter, we refer MISO back to our previous three rounds of feedback, submitted November 2 and December 21 of last year and March 24 of this year; most of the concerns and questions raised in those comments remain relevant to MISO’s most recent presentation on capacity accreditation. In particular, we note again below that the details of inputs and analysis of the LOLE model will determine whether applying the Direct LOL proposal results in fair and equitable capacity accreditation decisions; and as such, we urge MISO to release significantly more information on how the LOLE model operates and interfaces with accreditation and the PRMR decision as part of its engagement with stakeholders, consistent with the stakeholder transparency motion that passed at the RASC overwhelmingly earlier this year.

Furthermore, we believe it is appropriate to offer a response to slide 4 of the non-thermal accreditation deck from the April RASC,  in which MISO indicates three reasons or goals for accrediting resources.

1. To ensure seasonal Reserve Requirements are met

2. To inform long-term investment and retirement decisions by accurately representing

the capacity value of a resource in the prompt year

3. To reward resources for operating practices and attributes that serve the greatest

system need

We appreciate all three of these goals, but reiterate that we do not think it is necessary or optimal to try to meet them all with a single mechanism. The first of these is a threshold question and need not result in financial compensation. The second is important, but is in a time-frame far longer than the current PRA contemplates (while there is no certainty that one planning year’s auction clearing prices will closely carry over to future years). And the third is a worthy goal, but it can be achieved through compensation in the energy, capacity, and ancillary service markets, and need not simply be achieved through capacity accreditation.

2. MISO Needs to Provide Much More Detail on how PRMRs Would Be Impacted by the Proposed Direct-LOL Accreditation Approach

MISO provided information on only two slides discussing its intended treatment of PRMRs under its proposed Direct-LOL methodology, and gave only one example illustrating how it might change, on slide 9 of the Non-Thermal Accreditation Presentation on April 19 of this year.  According to the example provided, the PRMR would be adjusted based on the difference between the UCAP of the generating resources on its system, and the final accreditation those resources receive using the Direct-LOL method; thus, in its illustrative example MISO suggests that the PRMR would be reduced by 977 MW to account for the fact that the accreditation of wind and solar resources is 977 MW lower using the Direct-LOL method than simply by looking at their UCAP.  

While we appreciate the commitment to lowering the PRMR as nonthermal accreditation is reduced, and support MISO’s efforts in that regard, we do not believe the method MISO is proposing to employ here has been explained sufficiently.  Fundamentally, the question MISO should be answering is what the likely system needs (i.e. load) are during the high risk hours it has committed to focusing on in terms of accreditation; that requires looking at anticipated load during the LOL hours, which should be possible using the LOLE model.  The illustrative example employs a block “adjustment” to achieve the 1-in-10-year standard that is apparently similar to prior practice, but it gives very little explanation of how that “adjustment” will be calculated going forward.  As an initial matter, MISO needs to provide more detail on that calculation–and how it will change (or not) as MISO shifts to a seasonal capacity construct.  More examples, including some that are not “illustrative,” are needed.  But more fundamentally, MISO has already radically changed its resource adequacy construct by switching to a seasonal construct, and it is now proposing (even before observing the auction results for the first planning year under that new construct) to make significant additional changes to capacity accreditation reflecting a fundamental shift in accreditation priorities.  Given those changes, and the fact that MISO has made clear it is focused on resources’ performance specifically during high risk hours and across all seasons, it should explain why its proposal to continue using a flat LOLE-determined adjustment to UCAP to determine the system-wide PRMR is reasonable.

Finally, if MISO is going to move forward with a capacity accreditation construct that approximates marginal valuations, and reduces the PRMR accordingly, it needs to take steps to ensure that the PRMR reductions are allocated equitably to different regions or utilities based on their contribution to the shift of risk hours that resulted in a lower PRMR.  Applying any PRMR reduction identically to each region is neither fair nor efficient, because such a solution would socialize the reliability benefits created by regions with more significant renewable energy investments, without socializing the costs of those investments.  Thus, regions who have provided more of the renewable investments reducing the system-wide PRMR should also receive additional reductions in their PRMR.  And as part of its policy work formulating a PRMR reduction plan, MISO should open a stakeholder process to discuss and assess the available policy solutions to this problem. 

3. MISO’s Proposed Transition Plan is Reasonable, Assuming It Proposes a Reasonable Plan

While still noting our lack of support for the Direct LOL based accreditation approach, we do want to acknowledge aspects of MISO’s proposed transition that we support. We appreciate that MISO is proposing a three-year transition period to give load serving entities and generators time to adjust to and plan for the potential changes that will result from a new accreditation approach. We also appreciate that at the end of the three-year transition MISO intends to evaluate the accreditation of all resource types using the same methodology. But as we note below, the details of inputs and analysis of the LOLE model will determine whether applying the Direct-LOL approach to all resources will actually result in comparable treatment.  More transparency and discussion of these details are needed.

4. MISO Needs to Provide More Information on the LOLE Model to Ensure Comparable Treatment of Resources

MISO’s slide 8 indicates that under its proposed accreditation process, all resources would use the Direct-LOL output from the LOLE model. While this suggests comparable treatment of all resource types, we are not convinced that the availability inputs to the LOLE model are comparable, and thus these outputs may still result in discriminatory treatment for wind and solar resources. Correlated outage risk for wind and solar resources are captured by the inputs and analysis of the LOLE model. However, MISO has not yet proven that thermal resource availability inputs to the LOLE model (for instance during modeled “cold weather outage adder” events) will adequately capture correlated outage risk.

MISO indicated in slide 3 of its July 7, 2022 presentation to the Loss of Load Expectation Working Group (LOLEWG) that it will be incorporating changes in its LOLE model to represent increased outages correlated with extreme cold temperatures. The details about how this new analysis will be conducted are in an Astrape report also posted with the LOLEWG materials forJuly 2022. We appreciate the efforts of MISO and Astrape to improve the LOLE modeling to better approximate correlated outage risk for coal, gas, and other resources. But we request MISO and Astrape provide a presentation and stakeholder discussion at a future RASC meeting to share these details and whether the LOLE modeling results after incorporating this new methodology contemplate extreme winter events with significant correlated forced outages for thermal resources that are similar to the real experiences during winter storms Uri and Elliot.

We have asked for more detailed information on the LOLE model before; but in its April responses to stakeholder feedback, rather than provide a back cast of recent extreme weather events (as we requested), MISO pointed to the July 2022 LOLEWG presentation, which says, “MISO shared with stakeholders the details of the correlated cold weather outage adder, an improvement to the LOLE study that better aligns actual extreme weather events with more reasonable volumes of generation that would be forced offline due to weather.”  The Environmental Sector assumes that this refers to the improvements made for the 2022 LOLE model.  Certainly, we would expect this change to increase the number of LOL hours in the winter time.  However, the data last published to stakeholders showed only three LOL hours in the winter across four planning years.  

It is not enough to say the LOLE model will improve; MISO needs to test how significant the improvements will be that its change brings about. A backcast of several extreme weather events is an obvious way to do this and would be likely to help isolate what additional improvements may be needed or even whether the cold weather outage adder has overcorrected the problem.  For example, MISO stated in its January 18, 2023 RASC presentation that in modeling of 1,100 samples, 1,137 LOL hours were recorded in the winter time, but MISO has never indicated whether those hours are unique or repeated across multiple samples - and MISO has not yet provided those hours, despite committing to do so in its April response - so it remains unclear how many unique hours would count under the Direct-LOL proposal and when those hours would occur.  We appreciate MISO’s statement that “MISO uses actual historical unit-level outage events reported to GADS to develop the 5-year seasonal average outage factors for units in the LOLE model and is open to feedback into how to better align actual outage risks within the model.”  This is precisely the point of providing this response to MISO’s comments; we need more data from MISO to give more meaningful feedback.

MISO further stated in its April response to comments:

“The goal of the LOLE study is not to exactly predict the time periods in which the system will experience issues. By its very nature, a probabilistic assessment seeks to find “potential” risks and focuses on those that are most severe. Within the operational context the system will only experience a very small number of the risks predicted by the probabilistic assessment since each year represents only one within the tens of thousands of scenarios looked at in the LOLE study. The operational risks can be thought of as the realized risks. The main linkage between the potential and realized risks are the drivers of those risks rather than the highly specific condition that materialized. During MISO’s evaluation, it showed that the drivers of risk were well aligned between the probabilistic assessment and operational reality, but additional improvement is always warranted. As an example, the times of day in which the risks occurred were highly aligned, even if the specific days in which the events occurred were not.”

We understand that because the LOLE study is accounting for 30 weather-years, the results of the LOLE modeling will not perfectly predict the loss of load hours in any historical weather year.  However, given the Direct-LOL approach’s reliance on LOL hours for accreditation purposes, the question of whether the conditions under which those hours would occur is consistent with operational experience is very important.  That the times of day are consistent between the operational and modeled hours is not as relevant as whether similar conditions – weather, forced outage, renewable production, and load – are present in the model vis-a-vis actual experience.  It’s not clear how MISO would know that to be the case without doing the backcast that we have previously requested.

5. MISO Should Provide Direct-LOL Applicability Information Pertaining to Hybrid and Storage Resources at the May RASC Meeting; or Soon Thereafter

MISO indicated in Slide 16 of its April RASC presentation that it will present the results of its efforts to evaluate applicability of the Direct-LOL method to “other resource types.”  This list should include both storage resources and hybrid resources.  As we stated previously, both storage and hybrid generation facilities (especially solar-storage paired projects) are being built at a rapid rate, and represent a rapidly growing share of the available resources in the MISO region.  Their capacity accreditation may well be complicated due to the specific attributes of each resource type, and MISO and stakeholders will need time to work through how any accreditation regime might apply, in order to ensure that it results in comparable and equitable treatment to all resource types.

If MISO does not include storage and/or hybrid resources in the results it shares in advance of the May RASC Meeting, it should work expeditiously to provide those results as soon after that meeting is reasonably practicable.

The Entergy Operating Companies ("EOCs")^[1] appreciate the opportunity to provide feedback on MISO’s proposed solar/wind accreditation reforms.  MISO’s determination regarding this issue may have a significant impact on member utilities as they pursue renewable resource development that is key to achieving their and their customers’ goals; the EOCs therefore encourage MISO to give careful consideration to the feedback of its stakeholders in making that determination.

MISO’s Proposed Direct LOL-Methodology

The EOCs do not support MISO’s Direct-LOL methodology proposal. This Direct-LOL methodology will produce a very similar result to a Marginal ELCC proposal, and for this reason, the EOCs have the same concerns with MISO’s Direct-LOL methodology as we do with the Marginal ELCC methodology. Please see our previously submitted feedback to better understand our concerns with Direct-LOL/Marginal ELCC. The EOCs continue to believe that an Average ELCC approach is the correct methodology to use for determining the fleetwide accreditation of wind/solar resources in the prompt year PRA.

Entergy Proposal

In the April RASC Non-Thermal Accreditation on slide 15, MISO compared five various accreditation methods. The EOCs are disappointed that MISO did not analyze or consider the compromise proposal that the EOCs have described in prior feedback submissions which is to adopt a 50/50 blended Average ELCC and Marginal ELCC approach. The EOCs ask that MISO include this option in the analysis for future discussions on this topic. This blended approach  seeks to balance the competing interests of (1) sending a proper signal regarding what types of new capacity are needed within the market (Marginal ELCC) and (2) providing credit to existing resources that is commensurate with their true contribution to system reliability (Average ELCC). Including an Average ELCC component in the solar/wind accreditation would appropriately reward solar resources for the daytime reliability value provided by solar that causes system risk hours to shift to non-daylight periods.

Additionally, when performing future analysis of the various accreditation methods, the EOCs ask that MISO use a resource mix with higher solar penetration, rather then the 23/24 PY resource mix currently assumed on slide 15. Differences between the various methods will be more apparent when applied to a resource mix with higher solar levels.

Direct LOL PRMR Reduction

Under the Direct LOL method, MISO has proposed that PRMRs be reduced to account for the amount of wind/solar capacity that is lost due to using the Direct LOL methodology. The EOCs’ primary concern with this approach is that it would socialize the capacity benefit of wind/solar across the entire MISO system rather than providing the capacity benefit solely to the generation owners that are incurring the costs of owning/operating the resources. If the Direct LOL method is implemented by MISO, the EOCs believe that the PRMR reduction should be performed on a LRZ basis in order to reduce the amount of socialization that occurs. To explain further, the EOCs are proposing that the LRZs with the highest amount of solar/wind accreditation being lost due to the Direct LOL methodology should receive the largest reduction in PRMRs. This proposal does not eliminate the EOCs concerns related to socialization but the EOCs believe it would be a better approach then reducing PRMRs uniformly across all LSEs in MISO.

North/South RDT Constraint

The EOCs believe that the North/South RDT Constraint could have a significant impact on the true reliability value of solar/wind in future years and that by not accounting for this constraint MISO’s accreditation methodology could be sending inaccurate signals on where in MISO additional solar/wind resources would provide the most reliability value. Further, if the North/South RDT Constraint is not considered, it will likely result in an unfair distribution of capacity credit across resources in MISO North vs resources in MISO South. The EOCs request that MISO analyze this issue and include it as an item for discussion in future RASC meetings.

Transition

The EOCs support MISO’s implementation timeline of having the new accreditation changes taking full effect for the 28/29 PY.

Questions

The EOCs have the following questions for MISO:

• Under the new accreditation changes, how will MISO be accounting for resource deliverability considerations? Will unit level real-time injections be capped at the deliverability level as is currently done in the MISO wind ELCC calculation? Or will MISO only be applying the deliverability constraint in the final step of the accreditation calculation?
• Is MISO currently planning on accounting for the impacts of the North/South RDT constraint within the LOLE accreditation model or in some other post processing step?
• What is MISO’s plan/schedule for reforming the accreditation methodology for battery/storage resources and LMRs?

AMP suggests that MISO more completely explain the connection between accreditation and the LSE capacity requirements (PRMR and LCR).  Specifically, MISO needs to explain the extent to which MISO would (or could) ensure that resulting seasonal PRMR and LCR values would be lower than current levels if MISO moves to marginal accreditation (all other factors held constant). 

I am happy to discuss.

David Sapper

dsapper@ces-ltd.com

Ameren, holding company for Ameren Illinois Company (d/b/a Ameren Illinois), Union Electric Company (d/b/a Ameren Missouri) and Ameren Transmission Company of Illinois appreciates the opportunity to provide feedback on MISO's accreditation reform.  Ameren Missouri supports an accreditation methodology that is consistent between Schedule 53 and Non-thermal (renewable) resources.  Direct LOL is a step in the correct direction to achieve comparable treatment.  Ameren Missouri would appreciate additional information and discussion around RA hour applicability and other related details.  Due to the initial application of Direct LOL to non-thermal resources and its effects, Ameren Missouri would like to have additional discussion specifically, where RA hours are applicable to renewables.  While we agree with a "phased-in" approach, the schedule may be slightly aggressive since further discussion has been requested by stakeholders.

Comments from the MISO Cities and Communities Coalition Participants on the Resource Adequacy Subcommittee—RASC: Accreditation Reform

May 5, 2023

Submitted on behalf of the MISO Cities & Communities Coalition

To: The Resource Adequacy Subcommittee

Re. Direct Loss of Load Methodology

The MISO Cities and Communities Coalition (MISOCCC) appreciates the opportunity to submit this feedback to MISO on its Direct Loss of Load (LOL) methodology for calculating wind and solar accreditation and subsequent implications for the Planning Reserve Margin Requirement (PRMR). MISOCCC is a coalition of communities across the MISO footprint that coordinates collective action to further the individual clean energy, economic development, decarbonization, affordability, and grid reliability goals of participating communities through engagement with MISO leadership, its staff, and its stakeholders.[1]

MISO’s proposed Direct LOL methodology will directly impact the ability of MISOCCC members to achieve their respective energy goals. MISO’s resource accreditation method sends fundamental investment signals to the region’s utilities that serve MISOCCC members. Notably, these investment signals are often fed into long-term planning processes, not near-term investments. As such, we strongly urge MISO to adhere to the core principle of compensating resources based on their reliability contributions derived from an average loss calculation. By valuing resources on the margin only, MISO’s proposal fails to value resource reliability contributions in a way that is compatible with the resource investment decision-making processes used throughout the region.

Fair Credit to All Resource Types Through an Average Calculation Approach

MISOCCC urges MISO to ensure its capacity accreditation process is fair for all resources including large-scale renewables, storage, hybrid systems, and distributed energy resources. The probabilistic modeling approach favored by MISO’s RASC is prone to errors and limitations due to a rapidly changing resource portfolio as well as inaccurate weather forecasting. While the marginal accreditation methodology is consistent with how energy markets value locational marginal pricing, the average accreditation methodology values a portfolio of resources’ reliability contributions, which is the purpose of resource adequacy (RA).

MISO’s current proposal undervalues renewable and other clean energy resources based on their reliability contributions and undercuts the broader market for these technologies. MISOCCC members are actively seeking clean energy deployment in support of their energy affordability, reliability, and decarbonization goals. As written, MISO’s proposal is in direct conflict with public policy goals and our desire to maximize the economic value of clean energy and distributed energy resources because the proposed accreditation would favor building out a new generation of gas plants rather than clean energy technologies including renewables and storage. MISO has not demonstrated that these technologies cannot provide comparable reliability value as thermal resources.

It is critical to link RA accreditation with real-time operational performance during periods of reliability risk to ensure the RA construct efficiently meets the regional reliability needs.

In addition, we urge MISO to provide additional examples of how the PRMR may be affected by its selected accreditation methodology. A single example showing directional alignment between the PRMR and the proposed marginal Direct-LOL accreditation methodology is not nearly sufficient to alleviate our concerns over potential cost increases resulting from a manufactured gap between the PRMR and accredited capacity in the MISO region.

Finally, if MISO plans to consider LOL methodologies for other resource types (i.e., fossil-fueled resources) as it has indicated previously, MISOCCC requests that MISO provide a timeline for that process. MISO should consider concurrently implementing LOL methodology for all resource types.

Thank you for considering our feedback. We look forward to continuing to work with MISO and stakeholders to ensure MISO’s wind and solar accreditation reforms are aligned with the policy and reliability needs of MISOCCC members and the region more broadly.

Sincerely,

The MISO Cities and Communities Coalition

Jeremy Caron

Sustainability Program Manager

City of Des Moines, Iowa

Jason Ludwigson

Sustainability Coordinator

City of La Crescent, Minnesota

Eric Shambarger

Environmental Sustainability Director

City of Milwaukee, Wisconsin

Kim Havey

Director, Sustainability

City of Minneapolis, Minnesota

Greg Nichols

Deputy Chief Resilience Officer, Office of Resilience & Sustainability

City of New Orleans, Louisiana

DTE appreciates the ability to provide feedback on the accreditation reform proposal. DTE Energy believes the fundamental purpose of the MISO Resource Adequacy construct is to maintain reliability in a given planning horizon by ensuring adherence to federal Loss of Load Expectation standards. Given shifting risk profiles resulting from the generation transformation, DTE agrees that greater focus on high-risk hours is warranted, and therefore supports further discussion and potential adoption of the Direct LOL methodology. While further discussion is necessary on items such as PRMR impacts and potentially concerning the socialization of benefits due to reduction of PRMR, the Direct LOL methodology will adequately accredit resources based on their contribution during high-risk periods and capture broader risk profiles than what would be captured in Schedule 53 alone.

DTE further strongly supports an accreditation construct that uses a common calculation methodology across the various resource types (as is being proposed by MISO), as well as a transition methodology to prevent drastic changes in requirements and expectations.

The OMS Resources Work Group (OMS RWG) appreciates the opportunity to provide feedback to MISO on its proposed accreditation reforms including impacts on PRMR calculation and a three-year transition period. This feedback is from an OMS work group and does not represent a position of the OMS Board of Directors.

PRMR Calculation

 Ørsted values the opportunity to comment on MISO’s latest presentation covering their proposal for capacity accreditation of wind and solar resources. 

We continue to believe that the capacity accreditation methodology proposed by MISO does not adequately represent the contribution of all resources towards reliability targets and does not treat different technologies in a non-discriminatory way. 

MISO staff and stakeholders have recognized that the Planning Resource Auction (PRA) has the purpose of compensating resources for their contribution to reliability over the upcoming planning year, yet through the proposed accreditation, MISO will fail to fully compensate some classes of resources for their contribution to maintaining resource adequacy.

MISO’s presentation[1] during the RASC meeting on March 1^st illustrates the unequal treatment of intermittent resources under the proposed Direct Loss of Load (LOL) accreditation.   Slide 5 shows the shift of “high risk hours” away from peak demand hours due to increased solar penetration.  This shift results in a reduction of the Planning Reserves Margin Requirement (PRMR), as shown in the example presented[2] during the April 19^th RASC meeting. 

The reduction in PRMR due to the shift of “high risk hours” is the result of solar resources meeting a portion of MISO’s capacity requirements.  But under MISO’s proposal, instead of directly compensating solar generators for that capacity, as they would for other types of generation, MISO will socialize the contribution of solar resources by reducing capacity margin requirements.   Solar resources have been so effective at reducing the risk of loss of load during peak demand hours, that the highest risk hours have shifted away from peak demand.  MISO interprets the lower demand during “high-risk hours” as a reduction of the PRMR, but in reality, a portion of the PRMR is met by solar resources that receive little to no capacity payments under Direct LOL accreditation.

We continue to encourage MISO to consider an accreditation methodology, such as average effective load carrying capability (ELCC) model that accounts for resources’ actual contribution to maintaining the desired Loss of Load Expectation (LOLE) metric.

During the April 19^th RASC meeting, participants asked MISO staff to expand the example shown in slide 9 of MISO’s presentation.  We support this request and ask MISO to show the change in the LOLE metric that would result from the capacity requirements that were met by wind and solar being met by thermal generation instead. We believe that comparable treatment of different types of resources should mean that a MW of accredited wind or solar capacity would be, from a reliability standpoint, equivalent to a MW of accredited capacity of any other resource.  We also reiterate our request for more detail regarding the development of the LOLE model used for reserves margin and accreditation calculations. 

With the evolution of the fuel mix for electric generation, it is essential that all resources are accurately accredited for their contribution to maintaining resource adequacy.  Since the proposed Direct LOL accreditation is based on the availability of resource classes during the simulated loss of load events in the LOLE model rather than availability during risk hours, ensuring that modeling assumptions are accurate is critical.

Rainbow Energy Center (“REC”) thanks MISO for the opportunity to provide feedback on MISO’s non-thermal resource accreditation proposal. REC supports MISO’s efforts to implement a Direct Loss of Load (D-LOL) accreditation methodology for all generating technologies. As an initial step, REC agrees with MISO’s analysis that a unit-level assessment for wind and solar resources based on performance during RA Hours and overlaying resource class performance during high loss of load risk hour will accurately measure the reliability value of capacity resources. The more accurate reliability measurement will allow MISO to commit resources that have a high likelihood of contributing to reliability during periods of reliability risk and system need.

REC emphasizes that the D-LOL methodology aligns with MISO’s System Attributes initiative. The Regional Resource Assessment survey results show the MISO region facing reliability risks as retirements outpace capacity additions. In accrediting capacity resources through the D-LOL, MISO will place higher value on available resources that provide long-term duration, fuel assurance, flexibility, and availability. In turn, reliability will improve.

MidAmerican appreciates the opportunity to provide feedback on accreditation reform, including Planning Reserve Margin Requirement (PRMR) calculation and a transition period.

MISO should focus more on the resource adequacy construct with the objective of making the process more transparent and predictable. Only MISO currently has the hourly data across weather years to assess tight margin hours. MISO should either provide the necessary data to the market or develop more simplistic methods of translating this information such that load serving entities can make informed decisions regarding long-range resource adequacy.

When providing out-year D-LOL accreditation percentages, please provide assumed renewable penetration of each resource for respective out-year. A published look at varying the penetration amounts and the subsequent accreditation under D-LOL would be beneficial for planning purposes. Subsequently, how would the PRM vary from year to year with these various scenarios? Hour-ending and month of the year that the accreditation is based on would also be necessary on a yearly basis. For planning, all resource classes would be essential with a more detailed assessment of various hybrids.

Submitted on behalf of East Texas Electric Cooperative, Inc. (ETEC):

ETEC appreciates the opportunity to submit the following questions and requests to MISO regarding accreditation reform:

ETEC appreciated the discussion between marginal versus average ELCC that was held last year including in the October 2022 RASC meeting. While ETEC understands that MISO’s thinking has evolved to the Direct-LOL approach, ETEC continues to think that MISO need to justify its preferred Direct-LOL approach in a marginal versus average discussion.

While ETEC appreciates MISO providing some information on PRMR impacts from Direct-LOL, much more information in this area is needed. Stakeholders need to understand the impact to PRMR across all seasons (not just summer) and based on the full scope of MISO’s proposed plan to extend Direct-LOL to all resources (not just wind/solar). Furthermore, if MISO is going to implement the Reliability-Based Demand Curve (RBDC), then PRMR impacts need to be shown in terms of the RBDC as well.

ETEC believes that accreditation changes that will reduce the aggregate amount of capacity / supply are only half of the picture that stakeholders need to understand. By shifting risk accounting to the supply side through accreditation, MISO needs to also present the impact of the reduced demand side requirement. If assessing and sharing this information necessitates a discussion / proposal on modifying load capacity requirement calculations, then MISO should include that.

MPSC Staff would like to thank MISO for its consideration of previous feedback on a transition time that will allow states to align changes with IRPs. We support a 3-year transition starting in PY2025/26 with implementation by PY2028/29. (or 5 years if other states need more time to align with IRPs) This will give states and LSE time to adjust to the changes and incorporate into state planning mechanisms

 Thinking about accreditation from a high level, we support the flowing principles:

• Support applying a unified accreditation methodology to both thermal and non-thermal resources
• Comparability and equity across resource types should be a key pillar of the reform effort (as highlighted by the ESIG report)
• The methodology should be robust enough to accommodate the future resource mix and computationally feasible as different resource types develop.
• Would prefer to talk about accreditation and resource attributes together instead of in a linear fashion
• It seems that the direction and strength of the change in accreditation (including the marginal vs. average discussion) will influence how much work will then need to be done on attributes.
• As the IMM previously highlighted, a more robust accreditation methodology (like a marginal approach) may mean less work will be needed to ‘fix’ resource attributes 

MISO has indicated its intent to move towards Direct LOL accreditation of non-thermal resources.  WPPI offers our comments on several aspects of this proposal.

• We note that the proposed Direct LOL approach, which is entirely simulation based, makes it especially important that MISO appropriately model the interrelationship between intermittent-resource output and load, including any correlations.  We encourage MISO to devote particular attention to this task, including evaluating MISO’s assumptions against historical data and sharing these evaluations with stakeholders.
• We agree that PRMR should decline consistent with accreditation of the resources modeled in the LOLE analysis (which are intended to reasonably represent expectations of future auction clearing).
• We are supportive of a transition period for significant changes in accreditation methodology.  We agree with AES Indiana, however, that MISO’s proposed transition for non-thermal resources looks more like a simple delay than a phase-in, as the transition from Average ELCC to Direct LOL goes into effect in the final year, which will likely result in a distinct accreditation shock in that year .  This is something stakeholders should understand.  We agree also with WEC Energy Group that MISO should post DLOL accreditation estimates for informational purposes even before these are used for actual accreditation.
• We are open to discussion of blended accreditation methods such as suggested in the table on slide 15.
• We are also open to differences between accreditation methods for different types of resources.  In particular, Tier 1 hours were included in thermal-resource accreditation because of the significance of long planned outages for these resources.  This is much less significant for wind and solar resources, and we have doubts that inclusion of Tier 1 hours would be appropriate for these resources.
• MISO described the table on slide 15 as showing largely similar values across accreditation methods.  We note that this may be, at least for solar resources, largely attributable to low current levels of solar penetration.  As solar penetration increases, we expect that significant differences are likely to emerge, and thus that the low-penetration numbers may be of limited significance.  Repeating this calculation with simulated higher levels of solar penetration might provide additional useful information, as would showing the impact of energy storage at high renewable-penetration levels.

AES Indiana (AESI) appreciates the opportunity to comment on the Non-Thermal Accreditation Reform presentation discussed at the April 18-19, 2023 Resource Adequacy Subcommittee (RASC).

PRMR Calculation

AES Indiana agrees with the March 17, 2023 Capacity Accreditation Workshop presentation explaining that reliability can be met with an average ELCC or peak hours approach. AESI views wind and solar as  being penalized under a D-LOL/marginal ELCC/RA hours accreditation approach since reliability can still be maintained using average ELCC/peak hours for accreditation as long as the reserve margin is also adjusted upward.  The Capacity Accreditation Workshop guest presenter indicated that a system can maintain reliability using average ELCC/peak hours for wind and solar accreditation; however, the system will require a higher planning reserve margins to ensure enough capacity is available.  Wind and solar have an important place in resource/system planning as a sustainable bridge resource until emerging & sustainable baseload technologies become available. 

Transition Period

The transition approach MISO presented on Slide 12 of the April RASC presentation is not much of a transition.  Unit Level accreditation, which is the bulk of wind and solar accreditation, still switches to Schedule 53 in 2025/2026 which will significantly impact Load Serving Entity (LSE) planning.  AESI urges MISO to share more information so Stakeholders can be informed on their position to the potential D-LOL MISO filing.

Need for Solar+Storage Resources and Storage Only Resources Information

AES Indiana would like to restate our request to MISO, which was made in our March comments, that Solar+Storage resource accreditation and Storage resource accreditation by season be included in the next presentation on Non-Thermal Accreditation. We expand that request to include by PRA year, which was also requested by other stakeholders in the April meeting. To date, MISO has not addressed this request or made any statements about when information may be available after MISO modeling and review. MISO’s response in the April RASC presentation that examples haven’t been provided because there is only a small volume of Storage on the system is insufficient. The queue includes a significant volume of Storage and these potential projects could be included in modeling for illustrative review and understanding.  LSEs, including AES Indiana, are planning to invest in Solar+Storage and Storage resources and therefore need this information.   

• Alliant Energy looks forward to the additional information MISO will provide in May. In particular we are looking for forecasted accreditation rates for various resources with an understanding of how those values will evolve over by time and vary by zone.
• The PRMR calculation slide did not provide helpful insight on underlying calculations in the D-LOL methodology. MISO should elaborate further with its May update.
• We appreciate MISO’s 3 year transition period to give LSEs time to react where possible.
• We continue to have concerns, as do other stakeholders as indicated by the recent rollcall vote, that a marginal approach (like D-LOL) does not fairly represent the contribution/value of all resources in the portfolio. Resources like solar need the recognition of offsetting the gross peak and core LOLE risk, in addition to offsetting the net peak and remaining marginal LOLE risk. Other thoughts as previously shared in March:
• MISO’s marginal-only approach does not reflect long-term capacity value of all resources since it ignores embedded benefits.
• MISO needs to reward both embedded capacity benefits and evolving incremental needs.
• Perhaps incremental-only benefits need to be addressed in different markets – or at least a different approach than what MISO proposes. Ongoing discussion with stakeholders is needed.
• Resource additions are likely to go through phases of initial marginal benefit, then become embedded benefits in the future as new marginal needs develop.
• This embedded vs incremental issue falls squarely in the realm of MISO’s proposed attributes initiative, so the issue should be tabled until that time.
• We support accelerating MISO’s reliability attributes conversation, which needs to be part of a holistic approach, rather than piecemeal on everything:  DSDC, non-thermal, etc.

Xcel Energy appreciates the opportunity to provide feedback regarding MISO's capacity accreditation proposal at the April 2023 RASC.   We appreciate the three-year transition before a probabilistic approach at the class level for all resources is implemented.  We surmise that three years may not be long enough to allow for LOLE data inputs and assumptions to be improved and evaluated.    We prefer a probabilistic approach for the class averages, but are not completely convinced at this point that Direct-LOL is the best methodology.  One of the benefits of using a probabilistic approach is the ability to forecast capacity accreditation and load obligations in the future, incorporating impacts from issues such as a changing resource mix, new generation technologies, and electrification.  One of the disadvantages of using a probabilistic approach is the difficulty for LSEs to forecast long-term impacts for their planning purposes.  Therefore, regardless of the probabilistic approach selected, MISO will absolutely need to provide long-term forecasts for capacity accreditation and load obligations for LSE's future resource planning purposes.

We recommend that MISO use an average ELCC approach for the solar class level to more appropriately value solar during the transition.  We are concerned that the step change from the current class level approach for solar based on peak load performance to the D-LOL approach (based on high risk hours) could be large; using an average ELCC approach during the transition would mitigate this and align with capacity accreditation methods for wind. 

We note that storage and hybrids were not included in the transition plan and believe that MISO stated that capacity accreditation for these resource types during the transition will not change. There is proposed legislation in Minnesota to require 3000 MW of energy storage systems by 2033, so we need a forward look at capacity accreditation for storage resources.   We are hopeful that MISO provides more insight on storage and hybrids in the upcoming white paper; this may indicate transitional revisions. 

The PSC Wisconsin ORM Staff appreciates the opportunity from MISO to provide feedback on these issues regarding accreditation. We understand there is an upcoming white paper related to the Direct-LOL method which may answer some of the questions posed in this feedback. We look forward to continued discussion on these topics and responses to this feedback, along with the comments we have posed in our March Feedback (https://www.misoenergy.org/stakeholder-engagement/stakeholder-feedback/2023/rasc-accreditation-reform-rasc-2019-2-rasc-2020-4-20230228-0301/)  through this current feedback window, the whitepaper, and/or future meetings with MISO.     

It is important for MISO to establish the clear difference, or similarity, between LOL Hours from the D-LOL method at the Class-level and Resource Adequacy (RA) Hours from Schedule 53 at the Unit-level. Please include examples, and how this information will be shared in the future in line with stakeholder requests for more transparency. In the March 1 RASC meeting, part of the benefit stated by MISO for the D-LOL method was its ‘easier’ replicability relative to the ELCC methods.  Our questions and comments include:

 1. Will the set of LOL hours be shared with stakeholders like with the RA Hours?

 2. MISO should address stakeholder concerns, raised many times, that the estimated RA Hours do not match the actual Max Gen hours.  Understanding why this mismatch is occurring will help us understand whether RA Hours are close to accurate or not, which is now a foundational part of accreditation.  If RA Hours is going to be used for non-thermal generation, we simply must have a better understanding.  We need to know that DLOL Hours and RA Hours are reasonably accurate.

 3. In addition, academic and industry efforts are underway nationally/globally to look at risk hours approaches and may identify new and better methods than the RA Hour approach or DLOL.  MISO and stakeholders need to monitor these developments and be prepared to pivot if a better approach is identified.

PRMR Calculation

By analyzing Slide 9 from the April 19 RASC meeting, our staff noted that MISO has provided an example of the impact of D-LOL to the calculation of PRMR. However, we need for this example and future examples to provide much more context and detail (how many MW of wind were included in that calculation, how many MW of solar, what were the changed contributions of wind and solar respectively and not in aggregate, and graphics showing how the load shape / generation shape changed in context with the changing risk hour). Under this detailed approach, we would like to also see: How the PRMR would change when the D-LOL is implemented to thermal resources, storage, and hydro?

3-year transition plan proposal

We also appreciate the transition plan that was presented in Slide 12 of the April 19 RASC meeting.  From an operational standpoint, we support a transition period.  However, we stress that a “transition period” is not a substitute for further discussion and evaluation of the foundational gaps we have pointed out time and again between forward-looking accreditation to signal what future resources need to be built, as compared to vertically integrated utilities which build resource plans based on meeting their near-term load obligations at least cost. Forward-looking accreditation, which devalues components of the present-day generation mix, does not seem to work together with the resource plans from vertical integrated utilities, and MISO must better explain and provide different alternatives for how to navigate these foundational differences.  As stated above, one aspect that continues to need real explanation is how the PRMR changes as accreditation changes.

As to the details of the transition period provided by MISO, we would like to see the 3 following questions answered:

1. What would be the preliminary capacity credit (%) results for the 3-year transition (PY25-26, PY26-27, and PY27-28) for wind resources when implementing Average ELCC (at the Class-level) with the Schedule 53 RA-hours (at the Unit-level)? Similar to those presented in Slide 21 of the January RASC meeting, but again, including the Unit-level accreditation.

2. Similar to wind resources, what will be the preliminary capacity credit (%) results for solar resources of implementing Peak Load Accreditation (at the Class-level) with the Schedule 53 RA-hours (at the Unit-level)? Similar to those presented in Slide 21 of the January RASC meeting, but again, including the Unit-level accreditation.

3. Similar to wind and solar resources, what will be the preliminary capacity credit (%) results for thermal resources of implementing the Unforced Capacity (UCAP) (at the Class-level) with the Schedule 53 RA-hours (at the Unit-level)? Similar to those presented in Slide 21 of the January RASC meeting, but again, including the Unit-level accreditation.

Our staff considers it imperative to have these preliminary results presented in order for us, and our states’ stakeholders, to have a proper view of this 3-year transition plan. With that being said, when presenting the results for thermal resources, it would be important to specify this by their specific class (i.e., combined cycle, combustion turbines, nuclear, etc.).

Assuming a Direct-LOL methodology is adopted to accredit capacity resources, WEC Energy Group supports a 3-year transition period as a necessary component of the proposal.  During the transition period, it is essential that the Direct-LOL methodology is applied to each seasonal PRA for informational, educational, stakeholder readiness and process improvement purposes.  LSEs require substantial time and outreach to educate their executives, resource planning teams, customers and regulators on a resource adequacy approach that shifts from the traditional gross peak load hour approach to a construct that identifies capacity need during simulated loss of load hours.

The calculation of the PRM within a Direct-LOL methodology should reflect load levels and required capacity during the LOL hours from the LOLE model, to align with the accreditation of resources during those same hours (rather than the traditional seasonal coincident peak).  WEC Energy Group believes additional stakeholder discussion is needed on the calculation of the PRM (and PRMR) within a Direct-LOL approach because it is a fundamental departure from the traditional calculation reflecting load and required capacity during seasonal peak load hours.