RASC: Accreditation Reform Questions (RASC-2020-4,2019-2) (20231107-08)

The MISO FERC Filing is Q1 2024,  does this timeframe give stakeholders enough time to review the design details?

Can MISO define sub-classes by historic availability making Capacity values higher for above average performing resources?

An improved process should create incentives for Capacity resources to improve and join a higher sub-class.

The Environmental Sector appreciates the opportunity to submit the following feedback in response to MISO’s request at the November 7-8 RASC meeting. We answer the two specific questions in that request below. 

Question 1: What should MISO take into consideration to determine or define Resource Class?

1. There can be significant regional differences in the availability of wind and solar resources. MISO should do some analysis to determine whether solar and wind generation categories should be broken up based on a more granular level, such as one based on MISO’s planning regions (West, Central, East, and South). While we appreciate MISO’s acknowledgment of different regional accreditations between North and South, we believe that for weather-dependent resources such as wind and solar, a more granular breakdown may be appropriate. MISO North covers a very broad west to east footprint, which includes sufficiently diverse climatological conditions to warrant differing treatment of weather dependent resources based on their geographic locations. For example, by grouping wind resources in the wind-rich western region with wind resources in the east, the class as a whole will see lower accreditation, thereby dragging down the accreditation of wind in the West. While we appreciate that the DLOL methodology (including use of Schedule 53) is designed to shift accreditation within a class to those resources that are more productive at times of system risk, we believe that the geographic differences here are significant enough to warrant differing treatment. Similarly, while MISO’s North/South differentiation of resources works well for solar, which has real insolation differences north to south, regional (west to east) meteorological differences for wind are not similarly addressed by MISO’s proposal. Additionally, a more granular approach to accrediting wind will also help provide appropriate signals for future resource investment. To effectively serve load in a high renewable future, we will need to have a grid, including both generation and transmission, that is bigger than the weather systems. Thus we need accreditation to evolve with changing system needs, and provide the incentive for investment in wind and solar resources relative to the extent they can provide capacity to serve load as weather systems move across the country.
   

2. Likewise, there can be significant technological differences that create a substantial and real deviation in performance among individual generating resources within the same class. While MISO has attempted to address this fact by permitting those resources within a class that perform better than their peers to receive more capacity credit under MISO’s current proposed approach, we are still concerned that this differentiation is not large enough to incentivize investment in upgrades to improve capacity credit. Examples of beneficial investments that might be disincentivized include the addition of cold weather packages that enable certain resources to run (or be more robust and reliable) in very cold weather conditions, when their operation is most needed to ensure reliability.
   

3. For similar reasons, MISO should also determine accreditation for batteries based on their maximum output duration.  There will be a real value difference between 4 hour batteries and long-duration batteries that can provide capacity over a day or longer.  Keeping the batteries as a single class regardless of capability is likely to dissuade investment in longer-duration batteries.  

Question 2: Do you have suggestions for LOLE improvement ideas other than what MISO is already considering and shared in the November 8 RASC meeting? 

1. Consistent with the stakeholder motion for enhanced transparency from the March 1, 2023 RASC meeting (which passed with 56 in favor, 0 opposition, and 1 abstention), we ask MISO, with every RASC meeting and new analysis, to release all the relevant and permissible data it can undergirding its presentations, regardless of whether those data are specific to the point(s) MISO wants to make. The infrequent and limited sharing of data undermines the ability of the stakeholder community to provide effective feedback and help further improve the model and it creates uncertainty about MISO’s efforts. While this may seem time consuming at the outset, it may result in net saved time by creating more clarity alongside each presentation, requiring less time for MISO to respond to stakeholder comments, and requiring less time for stakeholders to fill the gaps that are left in such presentations. Together, this saved time should result in a better end product.
   

2. Improved modeling of storage: 

1. Storage dispatch: As was highlighted by the Invenergy/NextEra presentation at the August 23 RASC meeting, the accreditation value of storage is highly dependent on when and how it is dispatched. We understand that dispatching storage during LOL events in the model may result in a high accreditation value for storage today, but ultimately, consistent dispatch of storage during these events may shift the LOL event to hours after storage capability has been depleted. And as storage dispatch assumptions can shift the LOL events in the model, these assumptions will also impact the accreditation of other generation resources. We urge MISO to do more analysis and further consider with stakeholders what storage dispatch assumptions are most appropriate and how to value the capacity of these very flexible but energy-limited resources.

2. Long-duration storage: MISO only models storage as having a 4-hour capability. Multi-day storage is being piloted by MISO utilities, thus MISO should develop the modeling that will help determine multi-day storage’s value within MISO’s footprint now, rather than waiting until it is added to the fleet.
   

3. Modeling of solar-plus-storage hybrids: MISO should do more analysis to better understand the capacity contributions of solar-plus-storage hybrid facilities, and share that analysis with stakeholders. MISO is currently, and also within its accreditation reform proposal, treating the capacity of these resources with a sum-of-the-parts approach – that is, MISO is treating the solar component of a hybrid resource as falling into the solar accreditation category, and the storage component as falling into the storage accreditation category. We would like to understand how modeling their value as a unit may change their accreditation values. If modeling hybrids as a single generating unit is not possible, MISO should seek to improve both the way it models these resources and how it values their accreditation for the PRA. We would like any such modeling approaches to be carefully described, with examples provided so that stakeholders can better understand the implications for alternative modeling approaches.
   

4. It’s our understanding that MISO’s LOLE model assumes that all thermal resources operate at full load. Given how different this is from the reality of dispatch within the MISO system, it is important that MISO be clear about this assumption and give stakeholders an opportunity to discuss the tradeoffs of this methodology. We assume it is computationally more efficient to make this assumption, but we believe MISO should consider what implications this might have (compared to more evidence-based assumptions) on accreditation under the DLOL approach.
   

5. It’s also our understanding that MISO assumes a single thermal forced outage rate by fuel type per temperature observed. If true, this does not comport with the description of the cold weather adjustments contained in the MISO Seasonal Inputs for the 2022 LOLE Study. It also doesn’t comport with reality, in which multiple forced outage outcomes (common mode failures) are possible even at cold temperatures. We welcome clarification of this input and a conversation about its impact on the LOLE modeling.  
   

6. MISO’s communication of its LOLE modeling will be improved if it explains what option for scheduled maintenance MISO currently uses in modeling: 1) perfect foresight of a particular weather year and maintenance scheduled to minimize LOLE accordingly; 2) the worst January, worst February, etc. (regardless of weather year) and optimize a maintenance schedule around those outcomes; 3) the average January, average February, etc. and optimize maintenance to avoid LOLH; or 4) some other approach. MISO should also share with stakeholders the analyses it has performed to understand how well the selected methodology matches up with the reality of planned maintenance and adjust the methodology if a more accurate approach is possible.
   

7. MISO needs to do detailed validation of the assumptions about wind and solar shapes that are going into the LOLE model for periods identified as having LOL risk. Do the wind and solar shapes reflect the reality of the weather on the specific days identified as containing risk hours? Wind and solar generation, and demand, are not independent variables. This said, the relationships between high-risk days and the characteristics of the wind and solar resource are nuanced and require detailed study. Our understanding is that currently the method of estimating wind and solar shapes for periods where no data is available assumes that when the wind or solar profile is unknown, the best match for wind and solar profile will be a day with a known wind/solar profile and the closest match in MISO or LRZ wide average temperature. For example, if January 7, 1990 has an average temperature of 40°F, and January 9th, 2012 has the same average temperature in the region of interest, then the 1990 date will be assigned the same weather profile. MISO should check if this leads to low wind speed days being assigned to high risk periods, when in reality it was windy, or cloudy days being assigned to high risk periods when in fact it was sunny, which would yield lower capacity values for wind and solar via the DLOL methods, or indeed whether the converse is true, which would potentially create reliability issues. In any event, there may be identifiable biases in this methodology. Put another way, the wind and solar inputs going into the LOLE model may be more stochastic than they are in reality, and as a result the wind and solar capacity values emerging for risk periods may be biased.
   

8. We recommend that MISO conduct backcasts on multiple, recent weather years -- fixing the load and renewable profiles to match actual performance and then allowing the thermal outages to vary probabilistically to see whether an outcome similar to that weather year has a high or low chance of occurring. This would help identify areas that need improvement in the LOLE modeling, e.g. seasons that need adjustment to the inputs. It is critical that MISO ensures that the LOLE model results reasonably represent what it has seen historically.
   

Additional feedback related to accreditation and the PRM

1. We repeat an earlier request that MISO has not yet sufficiently addressed. Specifically, we request further clarification regarding MISO’s proposed approach to allocating an LRZ’s share of the PRMR among LSEs within that zone. Slide 14 of MISO’s August 23, 2023 presentation described this step of the strawman solution as follows: “2. LSE allocation of LRZ share would be based on look back period consistent with accreditation.” On page 7 of our September 8, 2023 comments on the August 23 presentation, we noted our uncertainty regarding whether MISO meant “actual LSE demand in real life during modeled loss-of-load hours” and whether that would be “reflective of how the LSE would contribute to system need during a hypothetical loss-of-load event.” MISO’s October 27, 2023 response to this feedback does not resolve this ambiguity. On page 4, MISO states that “[t]he strawman proposes to allocate commensurate with contribution to risk within the LOLE model.” This response suggests that MISO will determine an LSE’s share based on modeled demand  (presumably, during only DLOL hours). It is unclear how this concept relates to the August 23 description of allocation based on a “look back period.” Moreover, on page 7, MISO’s October 27 response states: “Historical load information is used as an input into the LOLE model. Regarding the strawman proposal, actual historical load data would be used to allocate each LRZ’s PRMR to the LSE’s within that LRZ.” We seek clarification (and an explanation of the chosen approach) of whether “actual historical load data” means either (a) use of LSEs’ modeled demand within the LOLE model, which itself incorporates actual historical data; or (b) direct use of actual historical data, independent of the LOLE model.
   

2. Given the critical impact that the LOLE modeling assumptions have on the results of the DLOL accreditation results, it is important that all stakeholders have a thorough understanding of the key assumptions and algorithms that form the basis of these results.  Realizing that presenting more details to stakeholders adds to MISO staff (and stakeholder) time commitments, in the long-run it will help build trust and avoid other problems later if MISO can provide clear examples of the impacts potential alternative assumptions or approaches would have on the accreditation of different resource types. We understand the importance of implementing a new accreditation methodology in the 2028 timeframe MISO has identified.  But we urge MISO to take sufficient time to identify changes to the LOLE model and the impacts of those changes before filing the DLOL methodology at FERC.  MISO should not rush this filing in order to meet a Q1 2024 deadline when it is still possible to meet the 2028 implementation date with a later FERC filing.

Clean Grid Alliance Comments to the RASC on 
Accreditation Reform

December 8, 2023

 
At the November 7-8, 2023, meeting of the Resource Adequacy Subcommittee (RASC), MISO presented updates to its accreditation reform proposal. Stakeholders were invited to submit feedback on two targeted questions: 

1. What should MISO take into consideration to determine or define Resource Class? 
2. Do you have suggestions for LOLE improvement ideas other than what MISO is already considering and shared in the November 8 RASC meeting? 

Clean Grid Alliance appreciates MISO’s efforts and consideration of this feedback. We strongly encourage MISO to create a resource class for hybrid interconnections, particularly for solar + storage hybrids which constitute the vast majority of hybrid projects. We also strongly encourage MISO to create a resource class for long duration storage. Both of these resources, (hybrids in great prevalence) are expected to be online in the MISO system at the time when MISO’s accreditation reform proposal is proposed to be implemented. These resources are an important part of MISO’s changing fleet and consideration now will prevent a lot of inaccuracies and problems later. 

Clean Grid Alliance appreciates MISO’s consideration of this feedback. 

Xcel Energy appreciates the opportunity to provide feedback regarding accreditation reform.  We agree with the benefit of using a modeled DLOL approach with historical performance allocation.  However, we still have  concerns regarding the use of fuel type to distinguish the resource classes.  Two resources of different fuel types that have the same ISAC performance will receive different capacity accreditation values, dependent upon the class average of their respective group.    So poor performing units will have adverse impacts on the high performing units  in the same class since the class average will be reduced.  Taken to an extreme, a MP may decide to spend millions to upgrade the availability of their unit in the future, but if the other units in the class are poor performers and reduce the class average, the additional capacity accreditation expected from the capital investment may not materialize.

The DLOL class average and the total ISAC of a class needs to be similar for each class for this method to be unbiased.  When these statistics are not aligned, unreasonable results can occur across classes. MISO needs to demonstrate that using resource classes defined by fuel type or any other means is the most appropriate method to use and drives the correct incentives for availability.    The solution may be to calculate the DLOL class average for ALL thermal resources and then allocate by Schedule 53 performance.  This would ensure that all resource types are treated equally and are compared to all thermal resources for capacity accreditation, similar to what is done today for Schedule 53.  Can MISO calculate capacity accreditation by resource type (and preferably unit) under a DLOL approach using all thermal resources as one class?  We would expect that this approach would stabilize the accreditation year over year, which is critical for planning purposes.

We are very concerned that MISO appears to be moving towards a filing in the first quarter of 2024, with design elements (PRMR allocation, expanded risk hours, etc.) not fully vetted for impact.  In addition, there are many LOLE improvements that need to be evaluated now, so stakeholders have time to understand and determine the impact. Expecting stakeholders to support a filing that is not complete is not a reasonable approach.  With this in mind, we propose that MISO delay their filing for one year and reduce the transition time from three years to two.  This would keep the same implementation date but allow more time to fully develop the design elements, work on the highest priority of LOLE modeling improvements and ensure that the methodology appropriately values capacity and provides the correct signal for future investment. 

The NDPSC believes there is value in breaking out gas between CTs and CCs and thinks MISO should explore breaking down CTs based on size. We also think there is value in categorizing the wind and solar classes based on location.

The NDPSC urges MISO to remove the cold weather adder and provide the DLOL results to stakeholders. This would give stakeholders the means to develop additional improvement ideas.

December 8, 2023

 Re:  Orsted Feedback on November 7-8, 2033 RASC Resource Accreditation Questions

 At the November 7-8, 2023, meeting of the Resource Adequacy Subcommittee (RASC), MISO presented updates to its accreditation reform proposal.  Stakeholders were invited to submit feedback on two targeted questions: 

1. What should MISO take into consideration to determine or define Resource Class? 
2. Do you have suggestions for Loss of Load Expectation (LOLE) improvement ideas other than what MISO is already considering and shared in the November 8 RASC meeting? 

Orsted is a leader in renewable energy and, either directly or through its affiliates, develops, constructs, owns, and operates offshore and onshore wind resources, solar farms, battery storage and offshore transmission facilities.  In MISO, Orsted owns and operates 424 MW of wind generation across two wind farms and is developing additional wind, solar, and storage resources.  Given this commitment to participating in MISO, Orsted is directly impacted by resource accreditation policies.  Generally, we continue to be concerned, as noted in comments sent to MISO in May 2023, that the resource accreditation methodology does not adequately represent the contribution of all resources towards reliability targets.    

 Orsted offers responses to the two questions posed during the November RASC below and appreciates MISO’s consideration

 What should MISO take into consideration to determine or define Resource Class? 

With the proposed Direct Loss of Load (DLOL) methodology, resource accreditation is based on the incremental (marginal) contribution to reliability of each Resource Class.  The incremental contribution to reliability of a resource depends on the correlation between the resource’s available hours and the high-risk hours in the model.  Therefore, resources with highly correlated availability will have similar reliability contributions and can go in the same Resource Class. 

 Orsted would also note that the availability of different weather-dependent resources varies by geography and technology, even within the MISO footprint.  As such, an effort should be made to recognize the differences between the generation profiles of these resources when determining the appropriate Resource Class.  Similarly, the duration and capacity of storage resources can also affect their availability during high-risk hours.  In addition, improving how storage resource are evaluated for LOLE calculations, storage resources should be grouped into Resource Classes that capture their contribution to reliability.

 Do you have suggestions for LOLE improvement ideas other than what MISO is already considering and shared in the November 8 RASC meeting? 

MISO should work with stakeholders to ensure that storage and renewable resources are sufficiently modeled with granular data.  For storage resources, MISO should work with stakeholders to develop models for LOLE calculations.  For all resources, MISO should ensure the modeling of resource availability is complete and sufficiently granular.  Appropriate wind and solar generation profiles should be used that are based on geographical location of the resource and the technology used.  The weather correlation of fuel availability of thermal resources should be modeled.

 In addition, one of the shortcomings of Direct LOL as a methodology to assess the reliability contribution of resources to meet adequacy standards relates to how resources are valued.  The methodology values resources during hours where resources are no longer needed and not their availability during simulation hours that are critical to meet the reliability criterion.    This issue is illustrated in the case of energy storage and energy-limited resources.  The problem is partially addressed by expanding the number of hours included in the accreditation calculation.  Even with extended hours, Direct LOL places considerable weight on the availability of resources during hours where they are not needed to meet the one hour in ten adequacy criterion.  Orsted supports the use of extended hours as an improvement over considering only modeled loss of load hours in the accreditation.

 Respectfully Submitted,

 Lopa Parikh,

December 8, 2023

Submitted via:  RASC: Accreditation Reform Questions (RASC-2020-4,2019-2) (20231107-08) (misoenergy.org)

Re:  Orsted Feedback on November 7-8, 2033 RASC Resource Accreditation Questions

At the November 7-8, 2023, meeting of the Resource Adequacy Subcommittee (RASC), MISO presented updates to its accreditation reform proposal.  Stakeholders were invited to submit feedback on two targeted questions: 

1. What should MISO take into consideration to determine or define Resource Class? 
2. Do you have suggestions for Loss of Load Expectation (LOLE) improvement ideas other than what MISO is already considering and shared in the November 8 RASC meeting? 

Orsted is a leader in renewable energy and, either directly or through its affiliates, develops, constructs, owns, and operates offshore and onshore wind resources, solar farms, battery storage and offshore transmission facilities.  In MISO, Orsted owns and operates 424 MW of wind generation across two wind farms and is developing additional wind, solar, and storage resources.  Given this commitment to participating in MISO, Orsted is directly impacted by resource accreditation policies.  Generally, we continue to be concerned, as noted in comments sent to MISO in May 2023, that the resource accreditation methodology does not adequately represent the contribution of all resources towards reliability targets.    

Orsted offers responses to the two questions posed during the November RASC below and appreciates MISO’s consideration

1. What should MISO take into consideration to determine or define Resource Class? 

With the proposed Direct Loss of Load (DLOL) methodology, resource accreditation is based on the incremental (marginal) contribution to reliability of each Resource Class.  The incremental contribution to reliability of a resource depends on the correlation between the resource’s available hours and the high-risk hours in the model.  Therefore, resources with highly correlated availability will have similar reliability contributions and can go in the same Resource Class. 

Orsted would also note that the availability of different weather-dependent resources varies by geography and technology, even within the MISO footprint.  As such, an effort should be made to recognize the differences between the generation profiles of these resources when determining the appropriate Resource Class.  Similarly, the duration and capacity of storage resources can also affect their availability during high-risk hours.  In addition, improving how storage resource are evaluated for LOLE calculations, storage resources should be grouped into Resource Classes that capture their contribution to reliability.

2. Do you have suggestions for LOLE improvement ideas other than what MISO is already considering and shared in the November 8 RASC meeting? 

MISO should work with stakeholders to ensure that storage and renewable resources are sufficiently modeled with granular data.  For storage resources, MISO should work with stakeholders to develop models for LOLE calculations.  For all resources, MISO should ensure the modeling of resource availability is complete and sufficiently granular.  Appropriate wind and solar generation profiles should be used that are based on geographical location of the resource and the technology used.  The weather correlation of fuel availability of thermal resources should be modeled.

In addition, one of the shortcomings of Direct LOL as a methodology to assess the reliability contribution of resources to meet adequacy standards relates to how resources are valued.  The methodology values resources during hours where resources are no longer needed and not their availability during simulation hours that are critical to meet the reliability criterion.    This issue is illustrated in the case of energy storage and energy-limited resources.  The problem is partially addressed by expanding the number of hours included in the accreditation calculation.  Even with extended hours, Direct LOL places considerable weight on the availability of resources during hours where they are not needed to meet the one hour in ten adequacy criterion.  Orsted supports the use of extended hours as an improvement over considering only modeled loss of load hours in the accreditation.

Respectfully Submitted,

Lopa Parikh, 

Head of Electricity Policy

 lpari@orsted.com

(857) 291- 6592

The Entergy Operating Companies ("EOCs")^[1] appreciate the opportunity to provide feedback on MISO’s accreditation reform proposal.

 DLOL Methodology

Concerning the DLOL Methodology, the EOCs believe that:

• MISO should expand the selection of hours for the step 1 class average calculation to include a larger set of hours. This will improve accreditation stability and should better align with the tight-margin hours that are seen in MISO on an operational basis.
• MISO should have a separate class for CCGT units as shown in the November RASC presentation.
• Given the size of the nuclear class, MISO should verify through analysis that enough iterations are being run in the SERVM model so that the nuclear DLOL result has converged to a consistent value.

 DLOL Process and Timeline

The EOCs believe that the near-term LOLE improvement ideas identified in the November RASC presentation 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. The EOCs believe that MISO should complete work on these LOLE improvement areas prior to filing the DLOL proposal with FERC, which will likely require MISO to delay the target filing date to later in 2024. The EOCs believe that the current 28/29 PY implementation date can be maintained, while still providing Market Participants reasonable lead time to adapt to the changes, even if the FERC filing target is delayed somewhat beyond Q1 of 2024.

 PRMR Allocation

In prior MISO RASC presentations MISO has proposed a methodology for re-allocating PRMRs based on LSEs’ customer load during high-risk hours. The EOCs believe that this proposal should be considered separately from the DLOL methodology and if stakeholders and MISO agree to move forward with this proposal that it should be filed and implemented later than the DLOL accreditation changes. The DLOL accreditation changes on their own are very significant and complex, and the EOCs believe that MISO should allow LSEs time to adapt to the new DLOL accreditation methodology before also implementing new PRMR allocation changes.

 LRZ Local Clearing Requirements

The Local Clearing Requirement should no longer be established on a UCAP basis. To determine each LRZ’s local clearing requirement, MISO should calibrate the zonal LOLE model to the 0.1 LOLE target and then sum up the LRZ’s DLOL capacity and the 0.1 LOLE MW adjustment. This will appropriately result in an LCR that is on a DLOL basis.

 The current MISO process for determining local clearing requirements is flawed because LCRs are established on a UCAP basis even though capacity is accredited on a SAC basis. For MISO-wide requirements, UCAP and SAC are equal in magnitude, so there are no problems with substituting one methodology for the other; but this is not true on an LRZ basis. Some LRZs’ total UCAP is greater than SAC, while other LRZs’ total UCAP is less than SAC.  The result of this LCR design flaw is that in some cases the MISO LRZ LOLE models will identify a set of resources that is able to meet the 0.1 LOLE target, but the sum of that group of resources’ SAC values minus the ZIA will be less than the LRZ’s UCAP LCR. This result defies the purpose of using an LOLE model to establish an LCR. If the LOLE model determines an LRZ is reliable with a given set of resources, then MISO’s resource adequacy construct should reach the same conclusion.

AMES, AMP, MEAN, and MPPA support WPPI Energy's feedback.

AMP adds that MISO should define resource classes so that outliers won't significantly affect the class average regardless of class size.  Doing so is important for avoiding improper loss (or taking, as the case may be) of capacity values as well as improper over-confidence in resource class capabilities.  If MISO continues to use such broad class definitions by fuel and geography, then MISO should demonstrate empirically and statistically that the chosen class definitions do not contain outliers that significantly affect the class average.

Capital Power thanks MISO staff for providing an additional opportunity for stakeholders to comment on the Direct Loss of Load (DLOL) class definitions and design elements. Capital Power also appreciates that MISO aims to submit relevant Tariff revisions to the Federal Energy Regulatory Commission in Q1 2024. Recognizing that any proposed changes at this stage of the DLOL accreditation reforms must be workable within the corners of the proposal and recognize the reliability benefits of resource contributions, Capital Power recommends that MISO adopt an exemption mechanism to the class-average multiplier for capacity resources that have demonstrated high levels of performance and availability.

One issue with the current DLOL construct is that high-performing resources are penalized in accreditation because other class members perform poorly. This sends the wrong incentive to the region’s best performers.

1. To address this, Capital Power proposes that MISO apply the following three criteria to identify the region’s best performing resources, and exempt these resources from the DLOL class-level multiplier when being accredited for capacity (all three should be met): Average of >80% availability* in each last 12 seasons over the last 3 planning years
2. Average of >50% availability* in each the last 36 months over the last 3 planning years
3. No outages or derates due to Fuel Supply Availability** affecting more than 40% of Capacity over the last 3 years
* where availability is calculated hourly by dividing RT Offer EmerMax by GVTC for all hours (ie. without considering any exemptions for planned outages)
** as defined in BPM-008 Section 4.11

The first two criteria ensure that resources have demonstrated strong historical availability in all months and all seasons over the past three years even when accounting for planned outages. The third criteria ensures that resources have not subject to severe derates due to fuel supply issues.

When taken together, the three criteria are sufficiently restrictive to ensure that only the region’s best performing assets will meet the threshold, and essentially creates a new class of very high performing assets that would be exempted from their technology class average. These criteria will incentivize those lower performing class members to improve availability and firm up energy delivery.

The proposed exemption mechanism is expected to be simple to implement as it is designed to fit easily into MISO's existing DLOL construct by using the 3 year historical horizon and developing new criteria based on parameters already used in MISO's planning processes. We would like to reiterate our appreciation for the opportunity to comment and welcome a discussion around the proposal/suggestions for consideration.

LBWL supports WPPI's feedback.

MISO is proposing a new accreditation methodology less than a year after implementation of the most recent significant accreditation reform.  This makes it imperative, in WPPI’s view, for MISO to proceed carefully and deliberately in designing a new methodology, so as to reduce the need for follow-up fixes.  We see many issues with MISO’s proposal that appear unlikely to be resolved on MISO’s current filing timeline.  Accordingly, we urge MISO to focus on making changes as necessary to address stakeholder concerns rather than satisfying a pre-determined filing timeline far ahead of the proposed effective date.  We propose that MISO do so by retaining the implementation timeline but pushing the filing back to Q1 2025, shortening the transition period by 1 year.

WPPI remains open to the general approach MISO proposes.

In recent years MISO has consistently emphasized the priority they place on an accreditation method that provides incentives for resources to be available at the time of greatest need.  We have concerns that MISO’s proposed approach may fail to provide some incentives that could be valuable.  This particularly arises when low-frequency, high-impact events are not appropriately reflected in 3-year-historical unit-level accreditation.  This approach may tend to not adequately account for resource features such as:

• Wind generator ability to continue to produce (even if subject to a reduced cap) at very low temperatures that may be experienced in likely capacity-shortage conditions, such as -30C conditions in northern MISO as experienced during Winter Storm Uri;
• Availability of firm gas or alternate liquid fuel for natural-gas-fired generation;
• Energy storage exceeding 4 hours at rated output.

We share the concern, raised by Xcel, that nuclear generators with lengthy shoulder-season refueling outages will have those outages treated, for accreditation purposes, the same as short-lead-time planned outages for other thermal units, even though refueling outages are typically entered further in advance than nearly any other outage.  MISO and stakeholders should consider in discussion at RASC whether that is appropriate.

At the November RASC meeting, MISO showed wind accreditation under DLOL that varied significantly between the south and north-central subregions.  WPPI would ask that MISO show the extent to which that distinction, which we expect is appropriate, would be captured via unit-level accreditation under MISO’s proposed treatment of wind as a single resource class.  We note that, depending on the outcome of this review, it may be appropriate to distinguish between subregional locations for wind accreditation.  We would appreciate similar analyses for different types of gas-fired generation, including CT, CC and ST generation, and different equipment ages (which might be substantially capturable via CT MVA ratings).

In the absence of more information on such analyses, it is difficult for us to intelligently comment on needed changes to DLOL resource classes.

MISO acknowledges concerns or at least the need for further review of the “extended hours” that include additional hours where margins are less than 3% of demand, and the modifications to accreditation for energy-limited resources like batteries.[1] MISO has committed to additional improvements to its modeling to better align its forecasted accreditation method with the risks it is trying to capture. Further, there have been requests at several previous stakeholder meetings regarding the need to further delineate the broadly defined resource classes to ensure a reasonable allocation of the class average prior to the application of unit-specific performance.

The timeline for implementation of the new construct should be delayed at least one year to the 2029-30 planning year to provide sufficient time for utilities with an obligation to serve load to complete generation resource plans which take into account the new construct, identify generation resource locations, submit generation interconnection requests, have the requests processes, obtain permits from regulatory authorities and construct the resources. MidAmerican remains supportive of MISO’s Reliability Imperative and agrees that changes in the accreditation construct are needed to advance reliability. The concerns identified above point to the need to reasonably balance the pace of the tariff’s timeline for the transition to this new design with the amount of time needed for load serving entities to adjust to the new policies. As MISO transitions its resource adequacy policies, load-serving entities need time to react; the proposed transition period is too short because:

• Conducting long-term generation planning is becoming very difficult and time-consuming due to increased transparency requirements and the need to fully understand and model the new construct;
• New dispatchable generation takes at least 5 years to install due in part to the 2-3 years required to get through the MISO interconnection queue and time required to obtain environmental and local permits; and
• MISO’s pause of the 2023 interconnection queue will drive the need for additional time in the transition.

Regardless of the ultimate policy determined through the MISO stakeholder process, MISO needs to

begin now to share significantly more of its model inputs, and MISO needs to accelerate its development of “bookend” scenarios of its own that provide information to the market regarding the impacts of various resource mixes.

More specifically regarding model inputs, MidAmerican continues to encourage MISO to provide relevant information to support modeling needs of the region because the resource mix and related risks are regional, not load serving entity – specific.  Any information MISO needs to enable it to determine risk hours with the current resource mix or in future resources mix scenarios is information that should be available to all that are developing long term resource plans. As an example of information needed, historical performance MISO GADS by unit class and by Local Resource Zone to determine Direct LOL data. Similar tables to the forced outage tables provided in the LOLE study are needed for planned outage data assumptions used by MISO to determine Direct LOL. This information is needed by unit Class and by Local Resource Zone.

Alliant Energy is looking forward to material discussion on MISO’s verbal proposal to provide the modeling data from the Direct-LOL analysis. MISO should draft a document for stakeholders that explains what information will be provided, how it will be provided, and specifically how LSEs will be able to use this data for short and long-term resource planning purposes.

 As part of its resource adequacy changes, MISO should endeavor to consider LSE resource planning needs in addition to footprint reliability needs. It is critical that resource adequacy changes accommodate a methodology for LSEs to anticipate, predict, and adequately react to such changes. For example, Alliant Energy is looking for MISO to provide current and future resource accreditation rates, and PRMRs, at critical time slices of the day such as gross peak, net peak, as well as morning and evening hours around solar shoulder performance hours.

WEC Energy Group recommends that MISO consider resource class definitions that allow comparable generation resources to demonstrate their relative availability during Tier 2 and Tier 1 hours.  The DLOL resource class accreditation "pie" should represent resources that have comparable fuel types and dispatch characteristics so that superior performers are rewarded with a larger slice of the accreditation "pie".  While all capacity resources have a must offer requirement, not all resources are similarly situated in terms of the dispatch stack, which may have implications on their availability when called upon.  For example, a regularly maintained CCGT that is frequently committed and dispatched is more likely to be available than a CT that runs very little and may experience start-up issues.  Additional DLOL accreditation scenarios are necessary to ensure that top performers within a resource class are rewarded, and vice-versa.  Further subdividing the gas-fired resource class may be appropriate given the wide range of dispatch characteristics within that class (CCGT, re-fueled coal, CT, RICE).

As noted in our previous comments, MISO and stakeholders have made significant progress on the DLOL approach.  However, several issues require additional analysis and discussion, such as the resource class item within the previous paragraph.  WEC Energy Group recommends that we continue to forecast a 2028-2029 implementation of DLOL but take more time to resolve open issues (which provides stakeholders with more understanding and confidence that should minimize protests).  A Q1 2024 FERC filing is far too aggressive given the importance of this initiative.

Due to electrification and the flatness of the winter peak DLOL daily load curve, there is little opportunity to charge Battery Energy Storage Systems.  Thus, BESS capacity accreditation should be zero for the winter period.