As part of MISO’s continued effort to be transparent and provide stakeholders with information, MISO has released a draft white paper on Resource Accreditation. The draft white paper covers the design for the proposed resource accreditation methodology and is intended to provide clarity on the design/process.
To assist, stakeholders are requested to provide feedback on any section in the draft white paper that contains content that is unclear or that lacks information that may be helpful to develop an appropriate understanding of the proposed methodology.
Note: Any feedback that would be in debate of aspects of the methodology has been requested through separate feedback requests at the Resource Adequacy Subcommittee (RASC) and will not be considered through this solicitation for feedback.
Feedback on the draft white paper is due by May 31, 2023.
The PSC Wisconsin Office of Regional Markets staff appreciates the opportunity by MISO to provide feedback on its most recent whitepaper. Our staff considers the accreditation effort to be a key issue area that should be researched more to provide the fairest treatment to all resources. Recognizing that no accreditation method is perfect, and that all methods have their ‘pros’ and ‘cons’.
In Section 2.3, ‘Schedule 53 Methodology for Allocation of Class Level’, our staff would like for MISO to expand this example to those resource classes for which preliminary results will be shown in the July RASC meeting. These resources being: coal, gas, hydro, nuclear, pumped storage, run-of-river, solar, storage, and wind. We understand that only Class-level preliminary results will be shown in the July RASC meeting. However, it is imperative for future examples within this white paper to show the unit level allocation of accredited capacity for these resources. This is important as the educational point source for our state resource planners to base their accreditation plans going forward.
In Table 2-1, MISO has shown that the sum of individual ‘Final SAC’ will equal the Class-level accredited capacity determined by the Direct-LOL method. Our staff would like for MISO to further expand on this, and if possible, to use different examples or numbers that avoid confusion (such as making the two example generators have ICAP of 25 MW and 15 MW each, instead of both 20 MW).
It would be beneficial if the transition plan of Section 2.4 is accompanied by preliminary results akin to those shown in Section 2.3 Table 2-1. Even if it is an arbitrary ‘Unit 1’ and ‘Unit 2’ it will help resource planners to visualize the implication of the Class-level and Unit-level methods used in the 3-year transition period. In particular, the preliminary results we request are the coupling of Average-ELCC with Tier-1 and Tier-2 Hours for Wind, Peak Load with Tier-1 and Tier-2 Hours for Solar, and UCAP with Tier-1 and Tier-2 hours for Thermal resources.
In Section 4.6 (p. 19), the whitepaper makes mention of the ‘shift’ in risk hours. Our staff would like for MISO to show in the whitepaper the trend in risk hours ‘shift’ for the 20-year forward looking period. This look forward could be based on Future 2A, for example. MISO could show this trend in the form of a heat map, as shown in Figure 1, or in a linear graph based on the hour of highest % of total EUE (MWh) per year and season. We consider part of the statement behind using the Direct-LOL method is its ability to always value capacity during risk hours. Hence, showing its adaptability to the shift in risk hours over time is imperative for the greater understanding of the stakeholder community. It would also be helpful to show how the risk hours have shifted from the past 10-20 years if that is possible. These looks would give us a sense of where we are now in relation to where we have been, and where we expect to go.
Through the white paper, MISO makes use of the term ‘1-in-10’ as part of the implementation of the loss of load reliability standard; we consider important for MISO to include a better explanation in that ‘1-in-10’ refers to ‘1 event day’ over a period of 10 years, as shown in Figure 2. The LOLE standard is loosely used within the stakeholder community, and since it is at the core of the Direct-LOL method, we recommend MISO expand into a section of its own. Source here: Clarifying the Interpretation and Use of the LOLE Resource Adequacy Metric (techrxiv.org).
Furthermore, the collateral impact of the Direct-LOL on other resource adequacy planning metrics is of vital importance. We think that the white paper should also include a section that expands on the potential impact that the Direct-LOL would have on the Planning Reserve Margin Requirement (PRMR). The direct implementation shown in Slide 8 and 9 of the April RASC Meeting deserves greater explanation in writing, including a class-level implementation of the Direct-LOL to those resources previously mentioned. We would also recommend that MISO provide many more examples of how the PRMR is expected to change based on additions of wind and solar resources. MISO provided one example in the April RASC Meeting, but perhaps 3-5 examples would be more appropriate as a starting point.
One additional specific question from the whitepaper: on Page 15, bottom paragraph, it states, “average ELCC… aligns availability with need but not actual risk.” What does this mean? How does “need” vary from “risk” in this sentence?
We look forward to the continued support from MISO on this topic. The whitepaper is a key opportunity to put in writing the explanation of this method and its related impact. Thank you again for the opportunity to provide feedback.
--
[1] MISO, (May 11, 2023). Market Redefinition: Accreditation Reform.
[2] Stephen et. al, (2022). Clarifying the Interpretation and Use of the LOEL Resource Adequacy Metric.
Comments in MISO Direct LOL Whitepaper
Minnesota Power
May 31, 2023
Introduction
i. How is the intersection of LOLH and resource availability defined as the resource accreditation?
ii. How is the LOLE model expected to quantify resource accreditation by generation class?
iii. How will the methodology provide the breakdown on the accreditation by generation class from the total PRMR MW?
i. LOLH are a subset of all hours of operation and do not occur in the real world of system operations. The dependency on a simulation to show an hour where there aren’t enough resources to meet load is a technical challenge when seeking to quantify the accredited value of generation class.
ii. Using the LOLE model to quantify the PRMR is an established process. It is complicated and challenging to understand in detail, but we have this in place as a FERC approved methodology. This approach defines the PRMR using the LOLE model and post processes the “Adjustment to UCAP {1d in 10 y} (MW)” to come up with the PRMR MW.
i. Date/hour
ii. Outage draw #
iii. Weather year
iv. Load MW
v. Gen hourly MW total production for each gen class
vi. Forced outage MW total for each gen glass
vii. Planned outage MW total for each gen class
viii. LOLH indicator
ix. Any other hourly metrics required to show the accredited value
i. Season description
ii. Generation class MW accredited value
iii. Number of LOLH
iv. Unserved energy MWh
v. Any other seasonal metric required to show the accredited value
WPPI Energy appreciates MISO’s posting of an early draft of the Resource Accreditation white paper. Joint development of this document between MISO and stakeholders appears to provide the best path toward establishment of a concrete record of all factors relevant to an ultimate decision on accreditation changes.
To that end we provide the following initial comments:
Guiding Principles
The report cites the MISO Market Design Guiding Principles shared in Todd Ramey’s presentation to the February RASC meeting. While these principles may be useful for designing elements of the resource adequacy construct, we view them as having primary relevance to the energy and ancillary services markets. Accordingly, we would advise against uncritical adoption of these principles for resource adequacy purposes. We note that MISO has already developed resource adequacy guiding principles that MISO has reported, for example, in previous MTEP reports:
Average vs. Marginal Accreditation
This is an important topic that merits careful discussion. A few elements of the draft discussion raise questions. MISO says that the (average) wind ELCC calculation “measures the amount of perfect load needed to move from a system with one type of risk to another system with a different type of risk rather than the contribution of the resource class in aggregate to the current system like the method is intended to do.” This requires more explanation. It is not clear to us that the distinction MISO articulates here is a meaningful one, or that ELCC fails to measure the aggregate resource-adequacy contribution of the studied class.
Similarly, MISO faults average accreditation for the “inability to add the accreditation values of separate resource types together to measure the total accredited capacity of an entire fleet.” Do marginal accreditations add in this way? If not, then this would not appear to be a meaningful disadvantage of average accreditation.
MISO says that, using average-accreditation methods, “it is very difficult to accurately measure the interaction between resource types.” We suspect what MISO means here is that there is not a unique solution to this problem. But, if so, MISO should explain why a unique solution is necessary, and why it would not be adequate for MISO to simply define the methodology it would use.
Finally, MISO describes marginal accreditation methods as better linking resource planning to operations. While marginal accreditation likely provides a more accurate price signal for investment, as MISO discusses, it is not clear to us what this has to do with operations. If MISO can make this case with fuller explanation it should do so, and should otherwise stick to a smaller number of more substantial arguments.
Development of Direct LOL proposal
The paper provides very little description about how MISO arrived at its current proposal, with Section 4.1 providing virtually no discussion of Direct LOL method, and the paper presenting a stark step-change between the evaluation discussion of 4.5 and the conclusion discussion of 4.6. While this does appear to give a reasonable representation of the sudden adoption of Direct LOL in the stakeholder process, this paper would benefit from an effort to explain how the Direct LOL approach emerged and came to be MISO’s preferred choice.
We don’t find Figure 4-5 particularly useful, since we accept that MISO must move beyond existing non-thermal accreditation methodologies. The real question is not whether Direct LOL is superior to these existing methodologies, but rather whether it is superior to the other alternatives that were part of MISO’s October recommendation.
Description of existing accreditation methodologies
In discussion of the existing Wind ELCC methodology, we would request that MISO make clear what the source of wind-output data is. We understand this to be actual historical output, in which case this is not a strictly simulation-based accreditation approach. This would be useful to state explicitly.
Details of Direct LOL proposal
Implementation of Direct LOL accreditation will involve numerous details, many of which are not described in the document. We think it would be appropriate to provide fuller description that could set the stage for coming design discussions. For the time being, however, we take Figure 2-3 as illustrative of the expected approach. This figure raises a few questions. First, it appears that not all LOL Hours are equivalent, as some may involve much more expected unserved energy (EUE) than others. It appears to us that using these EUE values as weighting factors in the accreditation calculation would better use information available from the LOLE analysis and would tend to provide more accurate results. If MISO nonetheless proposes to treat all LOL Hours equally, it would be helpful to explain this choice.
The Generator Availability table in this figure appears to show no change in what appears to represent hourly wind-generator output across samples within a weather year. This may make sense for this type of generation, for which forced (equipment-related) outages are not a significant factor. For dispatchable thermal generators, however, does MISO plan to use the same approach? And if so, will the availability be an expected availability considering planned and forced outage likelihood, or a single deterministic value from a particular sample? We would suggest adopting an approach that best represents the probability characteristics of the resource type at issue.
Finally, we note that the suitability of a Direct LOL approach, as a simulation-based accreditation method, will depend critically on the quality of the correlated input demand and generation availability files. We would request that MISO provide more information on how these will be developed and validated.
Resource Classes
The white paper does not describe what resource classes MISO intends to use in its Direct LOL accreditation. This should be added. We see a potential need to distinguish among different types of solar installations, gas plants with and without backup fuel, and perhaps other resource details.
Use of Tier 1 for Wind & Solar
MISO included both Tier 1 and Tier 2 hours in the Schedule 53 accreditation approach, a decision that was driven by the desire to reduce volatility associated with periodic planned thermal generation outages. We understand that MISO would like to adopt a single universally applicable accreditation methodology, but we would suggest that differences in resource characteristics may still dictate differences in accreditation approaches. In particular, it’s not clear to us that inclusion of Tier 1 hours is necessary or appropriate for wind and solar generation, for which unavailability is caused by fundamentally different drivers than for thermal generation. At a minimum, this issue merits discussion in the paper.
Transition Plan
WPPI supports a transition plan to facilitate a relatively smooth shift to new accreditation values that allows for a gradual phase-in of the implementation of DLOL at the class level for all resources. MISO proposes a three-year transition, but it is not clear that MISO’s transition proposal will allow us to avoid abrupt accreditation shock in Planning Year 2028.
Projected Accreditation Results
MISO shared a small set of sample class-level accreditation values at the April RASC meeting, for summer 2023. This is helpful but stakeholders need to know also how to expect these values to change in the future as the resource transition proceeds. We also need to know how this would be affected by varying levels of storage. Accordingly, we would ask MISO to compile in the white paper estimated accreditation values for DLOL and other accreditation methods for the near term, and for one or more representative future years, and for a variation on that future year in which the amount of storage on the system is significantly varied, e.g., doubled. Without this information it is difficult for stakeholders to understand the implications of MISO’s proposed accreditation approach and to meaningfully participate in the discussion.
AMES, AMP, IMPA, MEC, MPPA, and SMMPA support WPPI's feedback.
IMPA would add that MISO should expand the white paper to include enough numerical examples to show all the ways that class-level accreditation can affect individual unit accreditation. In MISO’s two unit system example, ISAC is forced to be scaled by modeled, class level performance. In their example, Unit 2 (which performed well in Tier 2 hours), loses about 1 MW in accreditation because the class level modeling suggested that under a LOLE study, that class of unit wasn’t expected to perform as well going forward. If MISO were to assume a more extreme situation, however, where, e.g., the class level accreditation amount is reduced all the way to 5 MW instead of 20 MW, then both units could suffer reduced accreditation despite performing well individually. For completeness and transparency, MISO should amend the white paper to identify this phenomenon so that stakeholders can better assess and address the fairness and efficiency implications of MISO's proposal.
IMPA also would add that the white paper should be amended to discuss the implications of the proposed accreditation method with respect to the timing of the LOLE Study and accreditation relative to the PRA. The white paper should explain the bounds of the additional accreditation volatility MISO's proposal could create. From there, MISO and stakeholders would better understand how the new accreditation method might warrant changes in the PRA administrative timeline that would give LSEs more time to adjust to accreditation changes.
I'd be happy to discuss.
David Sapper
dsapper@ces-ltd.com
Lansing Board of Water & Light (LBWL) supports WPPI’s feedback. - Natisha Wilkerson, Market Operation Analyst, LBWL
Xcel Energy provides the following feedback regarding the Resource Accreditation Draft White Paper:
The Environmental Sector appreciates the opportunity to weigh in on Version 1.0 of MISO's Resource Accreditation White Paper, released earlier this month. In keeping with MISO's request to limit feedback to "any section in the draft white paper that contains content that is unclear or that lacks information that may be helpful," we are not restating here previously shared concerns about the overall direction of MISO's proposed Direct-LOL accreditation methodology.
Starting in Section 2, we believe MISO's summation of its proposal lacks information that may be helpful to stakeholders or other parties attempting to assess the strengths of the proposal. As an initial matter, based on stakeholder questions, a clear comparison of the ELCC and Direct LOL approaches would also be helpful, as both are based on similar models and the same risk periods. But also, there are two key pieces of information missing from Section 2: first, MISO does not offer any detailed information about the underlying methodology it will use to "extract LOL hours and generator availability" (step 1 of the Direct-LOL method as described on page 6 of the white paper) using the probabilistic LOLE model. While the White Paper does a good job of describing how it uses the outputs of the model, MISO needs to offer a detailed workshop in which it runs through operation of the model, including all input assumptions and how they work, in order for stakeholders to be able to meaningfully evaluate how the direct-LOL method would work in practice. This request is consistent with the stakeholder motion that passed unanimously on March 8 (upon closure of the voting period), with 56 votes and a single abstention. That motion asked MISO to provide stakeholders better visibility and access into the models and data underlying the results it typically shares in RASC meetings.
We did hear the perspective of MISO staff in the meeting presenting this White Paper, indicating a preference to decide now on the accreditation methodology without digging into the details of the probabilistic LOLE model, and to instead work over the next several years to improve the model so that any accreditation method relying on the model can also be improved incrementally. And we do support MISO committing itself to a continuous process of improving its probabilistic LOLE model. However, it is not reasonable or realistic to expect stakeholders to sign on enthusiastically to MISO's Direct-LOL methodology without providing a full understanding of how the current iteration of the LOLE model will impact actual resource accreditation. Investment decisions are being made now that will dictate MISO's resource mix for the next decade or longer, and understanding the current iteration of the model is a fundamental need even if it may be improved down the road.
This critique is particularly salient based on the White Paper's mention of MISO's responsibilities as including "to measure current accreditation values, and forecast future values to inform investment and retirement decisions." MISO clearly expects projections of capacity accreditation under a direct-LOL methodology, based on the current iteration of the LOLE model, to be driving resource decisions. Indeed, we heard at the OMS Resource Adequacy Summit that the accreditation value signal from the PRA is not timely for utility resource planners, so a projection of future accreditation values will be critical for utilities to ensure they can meet their resource adequacy requirements in future PRAs. The White Paper did not, however, make explicit the form MISO's forecast of future values is expected to take, and how MISO is planning to provide those values on a timeline and in a form that utility resource planners can use as they plan their resource retirements and acquisitions down the road. We would encourage MISO to add that information to the next draft. We appreciate that MISO intends to provide at least some more of these details at the July RASC for various resource classes: if the presentation provides sufficient detail (as we hope it will), we would suggest that MISO delay finalizing its White Paper to be able to incorporate those details.
Second, Section 2 of the White Paper lacks a narrative explanation of how MISO is proposing to reduce the PRM under the Direct-LOL approach, to account for reduced accreditation of resources based on their performance during the highest risk hours. While we appreciate that MISO may still be evaluating how best to do this, it would be helpful for stakeholders to get a narrative explanation of the most recent MISO proposal. We hope it is clear that the PRM reduction method must be an integral part of any new accreditation regime, as well as questions about how to allocate any reduction in the PRM. So we were disappointed to see this topic almost entirely excluded from the White Paper.
Finally, we identified one piece of information that we believe is unclear: in the last paragraph of Section 4.3, comparing marginal and average accreditation, MISO appears to conflate its proposed Direct-LOL approach to the implementation of a marginal clearing price in its existing energy markets. We don't believe this is a helpful comparison: energy markets (and of course capacity markets such as the PRA) do clear based on the bid offered by the “marginal” source of power for that given time period, but we believe this use of a "marginal" resource to set the clearing price in an auction is fundamentally different from the use of a "marginal" resource to establish resources’ accreditation levels. In particular, while using a marginal resource makes sense as a market tool to create an incentive for all resources to bid at their minimum price (by eliminating the fear of losing out on profit potential), accrediting resource classes based on the contribution of marginal additions creates no corresponding market incentive to the impacted (mostly existing) generators. But we are open to better understanding this comparison to energy markets and marginal pricing: if MISO believes this analogy is apt, it should explain it in more detail in a future draft.
Sections 2.1 & 2.2:
Does MISO feel confident that the LOLE model used in the DLOL sufficiently captures the effect of weather and the changing resource mix in determining when there are loss-of-load hours? In answering the question, we urge MISO to show how different futures across seasons change accreditation every five years through time so stakeholders can align their IRPs with future need.
Has MISO determined any potential drawbacks or challenges to the DLOL method utilizing the LOLE model?
Section 4.6:
MISO listed the evaluation criteria as impact, flexibility, feasibility, and stability and rated the DLOL method as all green. Are any criteria missing from this evaluation, and which criteria is the most problematic for it now and across the various futures?
AES Indiana (AESI) appreciates the opportunity to comment on the Non-Thermal Accreditation Reform presentation white paper published on May 11, 2023. These comments specifically cover the topics of Solar & Storage Accreditation and Planning Resource Auction (PRA) Impact.
Solar & Storage Accreditation
AES Indiana requests MISO present a plan for how storage resources will be handled under this new accreditation methodology. AESI would like to again restate our request to MISO, which was made in our March 7th and April 28th comments, that Solar+Storage resource accreditation and Storage resource accreditation by season be included in Resource Accreditation materials. Utilities in MISO are investing in Storage. For example, the 2022 AESI Integrated Resource Plan (IRP) calls for 200 MW of energy storage in the near term.
Also, AESI continues to remain concerned about the projection for the value of solar resources going forward under Non-Thermal accreditation proposed changes.
Planning Resource Auction (PRA) Impact
AESI found Table 2-1: Example Calculation of DLOL and the Schedule 53 Method for a Two-Resource System on page 8 a helpful explanation walking through the calculation example to the DLOL approach.
Section 3 of the white paper lays out accreditation calculations by resource class level and unit level. As AESI prepares for each PRA, some certainty is required for decision making. Since when one generator’s accreditation is incorrect and then corrected, that change impacts the other generators of the same resource type, certainty is difficult. Therefore, AESI poses to MISO the question, given this calculation impact, how can MISO increase the certainty of resource accreditation values under this methodology?
Mississippi Public Service Commission (MPSC) Feedback Regarding
MISO Draft Resource Accreditation White Paper (20230511)
MISO recently released its Draft White Paper on Resource Accreditation. The white paper covers the design for the proposed resource accreditation methodology and is intended to provide clarity on the design process.
On May 11, 2023, MISO requested stakeholder feedback on any content of the draft white paper that is unclear or that lacks information that may be helpful to develop an appropriate understanding of the proposed methodology. MISO asked that, in their feedback, stakeholders abstain from debating aspects of the methodology covered through separate feedback requests at the Resource Adequacy Subcommittee (RASC). Accordingly, the feedback provided herein is mostly didactic and editorial in nature, but it does not constitute endorsement of the methodology by MISO for accreditation of solar and wind resources which is still being debated in the RASC.
Feedback
Apex Clean Energy believes that Section 2.3 and the table therein (Table 2-1) is unclear.
Please explain in more detail how “Final SAC” is determined and the meaning of the red circle and arrow. Additional units with more varied ICAP values may help make this calculation and methodology clearer. Using two units, both with 20MW ICAP values, is not very illustrative. Please see the attached .pdf file for further information.
David Davis, Apex Clean Energy
Please add to the white paper:
This is needed to understand the impacts of the methodology changes and for forward looking resource planning purposes.
Alliant Energy continues to have concerns that MISO intends to accredit entire resource stacks on a marginal basis. In contrast, there are significant embedded accredited capacity contributions of base-level resources. Capacity accreditation can be calculated as a curve moving from embedded resources to incremental resources – and that incremental resources added today become embedded resources in the future.