Comments on Final Proposal

1. Please provide summary of your organization’s comments on the final proposal:

Even though CDWR is not directly impacted by the proposed changes in ESDER Phase 4, CDWR continues to support CAISO’s efforts in accommodating energy storage resources into the wholesale energy market. CDWR continues to monitor and analyze proposed ESDER policy changes that may facilitate real-time load bidding for participating resources.

2. Please provide your organization’s feedback on the state of charge parameters proposal:

None.

3. Please provide your organization’s feedback on the proposed Non-Generator Resource market participation agreements:

None.

4. Please provide your organization’s feedback on the proposed market power mitigation model for storage resources:

None.

5. Please provide your organization’s feedback on the proposal to reflect demand response operational characteristics:

None.

6. Please provide your organization’s feedback on the valuation and operational processes for variable-output demand response:

None.

7. Provide any additional comments on the final proposal:

None.

1. Please provide summary of your organization’s comments on the final proposal:

The California Efficiency + Demand Management Council (Council) appreciates the CAISO's efforts to create more flexibility for demand response (DR) resources to more accurately reflect their availability and capability through the Demand Response Operational Characteristics, and the Market Participation and Must-Offer Obligations for Variable-output DR components of this initiative.  However, the Council believes the use of an Effective Load Carrying Capability (ELCC) methodology to determine the capacity value of DR resources is highly problematic for the reasons described below.  In light of these strong concerns, the Council opposes this initiative with the caveat that we would support with caveats if the ELCC component were removed and addressed separately.

2. Please provide your organization’s feedback on the state of charge parameters proposal:

The Council reserves comment on this issue.

3. Please provide your organization’s feedback on the proposed Non-Generator Resource market participation agreements:

The Council reserves comment on this issue.

4. Please provide your organization’s feedback on the proposed market power mitigation model for storage resources:

The Council reserves comment on this issue.

5. Please provide your organization’s feedback on the proposal to reflect demand response operational characteristics:

The Council supports this component of the ESDER 4 final proposal in general but continues to object to the 1 MW minimum resource size to be eligible for this feature.  As the Council has already stated, establishing a 1 MW threshold may have no effect on the number of DR resources and could actually result in less or more expensive DR capacity if DR providers are prevented from grouping their customers in an optimal way.  There are several business- and operations-related reasons for sub-MW DR resources.  These include:

• DRP has insufficient customers within a subLAP. Under some conditions, a DRP may have not recruited enough customers within a subLAP to form a larger resource.  This can happen when the DRP is relatively new to the California market or perhaps their recruiting efforts are unsuccessful. Another related factor could be if a DRP is building a portfolio in a new (for them) or less populated subLAP.  It should be noted that there are 29 DRPs registered at the CAISO.  One result of so many competitors in the market is that some DRPs will have some smaller resources as they seek market share; in addition, aggressive recruiting efforts by DRPs can cause some customers to move from one DRP to another which can cause the size of some PDRs to fluctuate as DRPs are forced to rebalance their resources.
• Ease of settlement. A DRP with a customer having multiple locations (e.g. a grocery store chain) will often prefer to use a single, dedicated PDR for the customer’s locations within a subLAP.  This simplifies DRP settlement with the customer based on energy market revenues, whereas multiple customers’ locations within a single PDR can complicate the settlement process.  Similarly, scheduling coordinators will sometimes group a DRP’s customers in a subLAP into a dedicated PDR rather than mix customers from multiple DRPs into a single PDR.
• DRPs will group customers by common operational factors. When a DRP builds a PDR, it will seek to group locations by common operational factors such as desired frequency of dispatch, opportunity cost, dispatch duration, and response time.  This is necessary so that when a PDR is scheduled in the CAISO market, all of the locations composing the resource will be available to provide the committed energy when scheduled.  For instance, grouping locations that are capable of dispatching daily with locations that can only meet the minimum Resource Adequacy dispatch requirements would inevitably leave some customers dissatisfied because the PDR would be dispatched too frequently for some or not frequently enough for others.  The other option for the DRP would be to bid only a portion of the PDR’s capacity on a daily basis and all of the PDR’s capacity on a less frequent basis.  This would create problems because the PDR would not always be bidding the full number of MWs in its supply plan. 

The same conundrum exists when grouping locations with different opportunity costs and dispatch duration.  It would be impractical to place locations who have a preference to dispatch at $250/MWh in the same PDR as locations who prefer to be dispatched at $500/MWh because it would unnecessarily limit the dispatches of the $250/MWh locations.  Conversely, any bids of such a resource at $250/MWh would only result in the lower opportunity cost locations dropping load.  Similarly, some locations may be willing to reduce load for longer than four consecutive hours (e.g. a facility that will shut down for the day for a DR event) but others are only able to meet the minimum requirements to qualify for Resource Adequacy.

Finally, some locations may be capable of responding quickly enough to participate in the real-time market while others require more notification which limits them to the day-ahead market only.  Adding more granularity, locations that can participate in the real-time market can be further grouped by 60-minute, 15-minute, and 5-minute dispatch capability.

• Baseline variability. For larger PDRs, the baseline can sometimes be more variable if there are many locations or if there is too much diversity among the locations.  Greater baseline variability can sometimes “zero out” the actual load reduction of a market dispatch, especially for smaller dispatches, so dispatches of smaller PDRs are sometimes more accurately measured.  For example, if a 10 MW PDR receives a 100 kW market award, the curtailment could be lost in any changes to the baseline.

6. Please provide your organization’s feedback on the valuation and operational processes for variable-output demand response:

The Council continues to have strong objections to using an Effective Load Carrying Capability (ELCC) methodology to calculate the Resource Adequacy (RA) value of DR programs and resources.  The CAISO has not demonstrated that using ELCC will yield a more accurate RA value than any other method.  The Council notes that during the recent heat storms in August and early September, DR resources played a significant role in avoiding more blackouts than otherwise would have occurred in the absence of these resources.  It would be unnecessarily risky for the CAISO to take steps at this point in time to devalue existing DR and disincentivize the entry of new DR at a time when new DR market participation is already on a decline, according to the Independent Evaluator report on the Demand Response Auction Mechanism 6 auction.  As the Market Surveillance Committee (MSC) stated in its draft Opinion of Energy Storage and Distributed Energy Resources Phase 4, “If too little credit is given to a particular resource type compared to its actual contribution to system adequacy, there will be too little investment in it, all other things being equal.”[1]

This is the time when California-based capacity resources should be encouraged, not discouraged because, as we have seen, high regional temperatures can sometimes limit imports from out-of-state resources.  The CAISO should remove the ELCC component of its proposal from the ESDER 4 initiative and take more time to address the many shortcomings in its study.  Additional development of the ELCC proposal is supported by potential areas of improvement that were highlighted by the MSC.[2]  Other shortcomings include:

1. The study examines IOU DR programs but no third-party (i.e. a non-IOU is the CAISO DRP) DR resources that are included in LSE supply plans. Third-party DR resources are subject to contractual performance obligations that are connected with their supply plan values, whereas IOU DR programs are not.
2. The study contemplates a single or perhaps a few representative ELCC factors for all DR resources which will undervalue the most effective DR resources and reduce the amount of capacity they can provide to the market.      
3. The CAISO has not indicated how frequently it would update its ELCC study. Because any updated factors will always lag the actual aggregate performance of DR resources, improvements in DR performance many not be compensated for many years.
4. The ELCC methodologies contemplated in the study lack transparency due to their complexity which is contrary to a well-functioning market.
5. The ELCC methodology would disincentivize new and high-performing DR resources as their performance would become linked to another lower-performing portfolio's DR resource.

The Council supports the Market Participation and Must-Offer Obligations for Variable-output DR component of the ESDER 4 initiative.  Allowing DR resources to reflect their capabilities through their bids is a reasonable and simple approach to assess their capability.

[1] http://www.caiso.com/Documents/Agenda-MarketSurveillanceCommitteeMeeting-Sep8-2020-EnergyStorage-DistributedResourcesPhase4.pdf, at p. 19.

[2] Ibid, at p. 25.

7. Provide any additional comments on the final proposal:

N/A

1. Please provide summary of your organization’s comments on the final proposal:

Support with caveats. 

2. Please provide your organization’s feedback on the state of charge parameters proposal:

CESA appreciates the ISO’s work on the development and future operationalization of an end-of-hour (EOH) state-of-charge (SOC) optional biddable parameter. As noted by the ISO, this parameter would offer stakeholders a means to manage their resources more effectively, ensuring they are able to fulfill their schedules while allowing them to retain the inherent flexibility of their resources. The implementation of optional parameters such as this one is aligned with the ISO’s commitment to integrate and manage storage resources using a market-based parameter to ensure that storage is able to deliver on their obligations, such as to maintain system reliability under Resource Adequacy (RA) contracts. CESA is specifically thankful for the ISO’s clarifications relative to ancillary services awards and their interaction with bid cost recovery (BCR) considerations – a concern CESA shared in previous iterations of this initiative. While this proposal represents a step in the right direction, CESA is still concerned with the proposed interactions between the EOH SOC parameter and the RA valuation of storage resources under the unforced capacity (UCAP) framework.

In Section 2.1.1 of the Draft Final Proposal, the ISO notes several stakeholders (including CESA) have pointed out the proposed interactions between the EOH SOC parameter and the UCAP framework would result in equipment operators foregoing the usage of the EOH SOC parameter. The ISO rejects these concerns, highlighting that it is developing tools that, along with the EOH SOC, will ensure storage RA resources meet their scheduled dispatches – specifically referring to the minimum charge requirement (MCR) proposal discussed within the RA Enhancements Initiative.[1] The ISO thus maintains its position to consider the use of EOH SOC as potentially detrimental to the UCAP evaluation process, arguing that the EOH SOC parameter was “originally intended to help storage resources manage their state-of-charge to meet other obligations (such as serving as a transmission asset, etc.)”[2] CESA disagrees with the ISO’s premises and conclusion.

First, the EOH SOC parameter should not be limited to being used to ensure meeting obligations other than RA. The ISO’s proposed parameter is robust and flexible enough to allow equipment operators to use it for a wide range of purposes. The flexibility provided by this parameter must not be limited to certain obligations; instead, it should be harnessed to mitigate the risks identified by the ISO. The provision of additional optimization and management tools is beneficial to the market as a whole, as it places the burden of effective operation and compliance on resources while providing increased certainty to ISO operators. As such, CESA considers the use of the EOH SOC parameter for RA purposes is both possible and desirable.

Second, it is not clear that the inclusion of additional proposals in separate initiatives addresses the issues related to the EOH SOC parameter within the proposed UCAP framework. CESA has highlighted several times in both ESDER and RA Enhancements initiatives that the potential UCAP impacts of the EOH SOC parameter will, in effect, make the existence of this parameter irrelevant for RA assets, especially as the vast majority of storage deployments will be contracted under RA contracts. Instituting a parameter that could hinder the RA value of an asset due to the chronology of UCAP evaluation renders the parameter useless. Given the time and effort stakeholders and the ISO have poured into the development of the EOH SOC parameter, this outcome would be inefficient and suboptimal. Prescribing the purpose of this parameter by establishing usage assumptions, as the ISO has done, goes against the spirit of this proposal, which seeks to bolster the management flexibility of non-generator resources (NGRs). Thus, it is unwarranted and detrimental to assume why an asset operator would employ this parameter. Within the Draft Final Proposal, the ISO states that under the UCAP framework it “may consider treating self-schedules and end-of-hour state-of-charge parameters that fall below the resources contracted value as a reduction in availability of the resource.”[3] CESA interprets this statement as a guarantee that the usage of EOH SOC up to an RA resource’s contracted value would not hinder its UCAP evaluation, and requests the ISO confirms this interpretation and clarifies it within the RA Enhancements Initiative.

Finally, CESA recommends that the ISO reconsider the MCR proposal in the RA Enhancements Initiative, which would be discriminatory to storage resources and erode market-optimal behavior of storage charging and discharging in the market. Given the EOH SOC parameter being developed in this initiative, the ISO’s concerns about storage being able to deliver on their RA obligations can be addressed through this market-based tool. CESA will raise our specific comments and potential alternative solutions to address the ISO’s concerns in the RA Enhancements Initiative, but we thought it was important to highlight the relevance of the EOH SOC parameter in addressing the issues noted in the RA Enhancements Initiative.

[1] Draft Final Proposal at 3.

[2] Draft Final Proposal at 4.

[3] Draft Final Proposal at 9.

3. Please provide your organization’s feedback on the proposed Non-Generator Resource market participation agreements:

CESA offers no comment at this time. 

4. Please provide your organization’s feedback on the proposed market power mitigation model for storage resources:

CESA appreciates the work the ISO has done to ensure energy storage is governed by a set of reasonable and prudent market mitigation measures. While currently limited, CESA understands the ISO’s intention to establish these rules in light of the substantial upcoming additions of energy storage resources to the CAISO footprint. In general, CESA agrees with the ISO’s general notion to establish DEBs that represent a general upper bound of costs for storage, as this approach will mitigate the methodological inaccuracies related to the difficulty of perfectly modeling all the factors that determine the marginal cost of energy storage. CESA is generally supportive of the ISO’s proposed approach, offering specific comments on a component of the default energy bids (DEBs) below.

With regards to energy costs, CESA reiterates its concerns with the assumption that all storage resources will charge at the times with the lowest prices. This assumption may not be generalizable across all storage resources, which might be optimizing their charging behavior with consideration of products and services beyond energy (e.g., ancillary services). In the Draft Final Proposal, the ISO notes that CESA’s concern might be misplaced as energy revenues make up the bulk of revenue for these resources and the ISO expects this trend will become more pronounced as the penetration of storage increases in the CAISO footprint.[1] CESA disagrees with the conclusion of the ISO as it is based on assumed performance of all future resources rather than the fact that a substantial set of storage resources are currently dedicated to provide regulation and ancillary services rather than daily energy arbitrage, and some future storage resources may continue to operate in this manner. As such, this assumption could in fact unduly lower the DEB of storage assets, resulting in unfair reductions in compensation due to a lack of understanding of the charging costs these assets incurred.

Furthermore, CESA recommends that the ISO consider additional energy cost categories for distribution-connected storage resources that interconnect under the Wholesale Distribution Access Tariff (WDAT) to participate in the ISO’s wholesale market. For example, while transmission access charges (TACs) are not assessed against wholesale storage charging for later resale, per Order No. 841, this policy may not be broadly applicable to all distribution-connected storage systems, where Order No. 841 left it up to distribution utilities to establish potential storage charging rates on a case-by-case basis. Southern California Edison Company (SCE) currently has WDAT contract demand rates that apply to most distribution-connected storage resources going forward, where such costs should be reflected in the DEB calculation. These rates for SCE-territory projects may evolve, and other distribution utilities may consider something similar. The ISO should be aware of these additional cost categories and adapt the DEB calculation methodology to reflect them.

[1] Draft Final Proposal at 21.

5. Please provide your organization’s feedback on the proposal to reflect demand response operational characteristics:

CESA offers no comment at this time. 

6. Please provide your organization’s feedback on the valuation and operational processes for variable-output demand response:

CESA offers no comment at this time. 

7. Provide any additional comments on the final proposal:

CESA offers no comment at this time. 

1. Please provide summary of your organization’s comments on the final proposal:

Please see DMM's comments here:

http://www.caiso.com/Documents/DMMComments-EnergyStorageandDistributedEnergyResourcesPhase4-FinalProposal-Sep162020.pdf

2. Please provide your organization’s feedback on the state of charge parameters proposal:

Please see comments in Item 1.

3. Please provide your organization’s feedback on the proposed Non-Generator Resource market participation agreements:

Please see comments in Item 1.

4. Please provide your organization’s feedback on the proposed market power mitigation model for storage resources:

Please see comments in Item 1.

5. Please provide your organization’s feedback on the proposal to reflect demand response operational characteristics:

Please see comments in Item 1.

6. Please provide your organization’s feedback on the valuation and operational processes for variable-output demand response:

Please see comments in Item 1.

7. Provide any additional comments on the final proposal:

Please see comments in Item 1.

1. Please provide summary of your organization’s comments on the final proposal:

CLECA supports the implementation of a max run time parameter for Demand Response.

CLECA continues to oppose the use of DR ELCC as it has not been shown to be superior to the current Load Impact Protocols.

2. Please provide your organization’s feedback on the state of charge parameters proposal:

No comments at this this issue.

3. Please provide your organization’s feedback on the proposed Non-Generator Resource market participation agreements:

No comments at this this issue.

4. Please provide your organization’s feedback on the proposed market power mitigation model for storage resources:

No comments at this this issue.

5. Please provide your organization’s feedback on the proposal to reflect demand response operational characteristics:

It is CLECA’s understanding that the implementation of a maximum run time parameter would improve the ability to accurately reflect demand response operational constraints in the model. 

Given the experience with the recent heatwave and the increased use of demand response, CAISO should expedite this implementation prior to next summer.

6. Please provide your organization’s feedback on the valuation and operational processes for variable-output demand response:

CLECA appreciates that CAISO will continue work with the CPUC on the qualifying capacity proposals for demand response. 

CLECA appreciates the effort by the CAISO to exempt demand response (DR) from RAAIM penalties as it does for other weather-sensitive resource adequacy resources, and allow them to bid  a variable amount based upon availability as opposed to a fixed qualifying capacity value.[1]  This would resolve one barrier to including DR on supply plans.[2]  Allowing variable bids without penalty would avoid having to discount DR’s true capability on supply plans to meet the 1 in 2 peak in order to avoid RAAIM penalties for weather-sensitive DR programs.  When DR’s true capability to meet the RA peak is discounted, then the shortfall is replaced with other less preferred resources.  If DR’s value is artificially devalued, then the CPUC might reduce incentives, which would reduce DR participation.  Under the current rules, the impact of placing DR on supply plans and being subject to RAAIM would create additional unnecessary replacement cost.  The result is contradictory to California policy to procure energy efficiency and demand response prior to other resources, including renewable resources. 

While CAISO appears to offer a proposal to resolve the supply plan and RAAIM issue, it makes it conditional on the CPUC’s adoptions of its DR Effective Load Carrying Capability (ELCC) proposal.[3]  Based upon CAISO’s DR ELCC study, which contains flaws (discussed in more detail below), many DR programs would be subject to reductions to their qualifying capacity value despite being able to provide higher load shed during a 1 in 2 peak event such as occurred in August and September.  This would result in reduced DR program incentives and therefore DR participation, which would be replaced by less preferred resources. 

CAISO’s offer of a requirement of placing DR on a supply plan under the current must offer rules versus the CPUC adopting DR ELCC and being exempt from RAAIM poses an unnecessary dilemma.  Based upon CAISO’s flawed ELCC results to determine qualifying capacity, either option will lead to a non-optimal outcome which is the loss of participation in demand response programs. Both result in reduced demand response which is contrary to California policy.

The Supply Slide Working Group report discussed developing a DR forecast, either derived from the Load Impact Protocols (LIP) or other approaches, and using that forecast as the measurement in a must offer requirement.[4]  This would appear to resolve CAISO’s concern about DR bidding in what is actually available and having a must offer requirement.[5] 

 It appears CAISO is proposing ELCC because it is used for wind and solar.  However, the requirement of using ELCC for wind and solar in the CPUC’s RA program is due to state law and not because ELCC was determined to be superior to other counting methodologies for Resource Adequacy. 

The Load Impact Protocols (LIP) should not be replaced with another approach until it has been properly vetted and found to be superior to LIP and is an appropriate fit in an RA accounting program.  Using ELCC for DR has not been shown to be superior or a good fit for the RA program. Indeed, several issues with using ELCC for DR have been raised, notably by CLECA.

            The DR ELCC results presented by the CAISO in its May proposal suffer from several fundamental flaws.  First, the study used the RA qualifying capacity values which are grossed up for the planning reserve margin and losses against bids that do not contain any gross up.  This is comparing apples to oranges and produces misleading results about DR performance and the LIP.  Second, SCE identified that the use of the CAISO’s own bidding rules prevented resources from bidding their full capability.  Again, this error also results in misleading results about the DR performance and the LIP.

            In the Market Surveillance Committee’s (MSC) opinion on the DR ELCC proposal, it commented that modeling ELCC has become very complex due to the correlations between weather, load, solar, wind, and battery operations.[6]  The MSC members mentioned that the effort is a good start, that but more work is required.  CLECA agrees.  If the modelling is not performed correctly, it produces scenarios that may be unrealistic and, if improperly weighted, gives misleading results.  The impact of this error could result in mis-valuing the ability of resources to avoid loss of load.

            The August and September heatwaves indicate that the time of system stress was during the peak during the month.  The times of the peak resulting in Stage 2 Emergencies were mostly between 4 pm – 8 pm.  If the ELCC is not showing those times as periods of shortfall, then there is a problem with the modeling.  However, the LIP for qualifying capacity averages the values from 4-9 pm, which covers the period during most of the Stage 2 Emergencies. There is no reason to conclude that the hours of most concern are outside of a peak event during the hours of 4 – 9 pm.

[1] CAISO ESDER 4 proposal at 44.

[2] There may still be problem for the CPUC’s Resource Adequacy Procedures if a DR program is included on a supply plan. This is because the benefit of utility demand response programs is provided to other load serving entities as they are still responsible for paying for demand response related costs.

[3] CAISO ESDER4 final proposal at 44.

[4] See Supply Side Working Group Report (June 28,2019) at 19-30

[5] CAISO ESDER4 final proposal at 45.

[6] Market Surveillance Committee, Sep 3, 2020, Opinion on Energy Storage and Distributed Energy Resources Phase 4, at 25.

7. Provide any additional comments on the final proposal:

No comments at this this issue.

1. Please provide summary of your organization’s comments on the final proposal:

CPUC staff requests further discussion on the proposed market power mitigation model for storage resources.  CPUC staff has remaining questions regarding how the proposed model will address lithium-ion batteries’ unique costs which include the costs associated with cycling beyond design specifications and/or deep discharge cycles.

2. Please provide your organization’s feedback on the state of charge parameters proposal:

CPUC staff supports the ESDER end-of-hour state-of-charge bid parameter proposal

CPUC staff continues to support CAISO’s end-of-hour state-of-charge (SOC) bid parameter.  This support was previously stated in CPUC staff comments submitted to the CAISO on May 17, 2019, July 11, 2019, and November 15, 2019.  CPUC staff supports the proposal because it allows scheduling coordinators to achieve optimal use of an energy storage resource throughout the day through desired end-of-hour SOC bid parameters.  The SOC bid parameter proposal thus assists with achieving the Multiple Use-case Application (MUA) framework developed jointly by CAISO and CPUC and assists with ensuring that energy storage systems meet any reliability obligations.

3. Please provide your organization’s feedback on the proposed Non-Generator Resource market participation agreements:

CPUC staff has no comment on this proposal at this time.

4. Please provide your organization’s feedback on the proposed market power mitigation model for storage resources:

CPUC staff supports a dynamic ESDER default energy bid (DEB) formula to account for lithium-ion batteries’ variable costs associated with cycling beyond design specifications and deep discharge depths.

CPUC staff understands and appreciates that the ESDER DEB proposal will be further discussed through a separate stakeholder initiative.  The CAISO and MSC have proposed enhancements to the DEB formula to account for energy storage cycling costs, however, additional discussion is still needed to arrive at equitable solutions that address  lithium-ion batteries’ unique costs which include the costs associated with cycling beyond design specifications and/or deep discharge cycles.

Accounting for Cycling Beyond Design Specifications

As stated in prior comments submitted on March 20, 2020 and June 15, 2020, CPUC staff requests that the CAISO consider a dynamic DEB for lithium-ion batteries to address the noted cost difference when a lithium-ion battery is cycled beyond its designed specifications.  The CAISO states that the costs associated with “storage resources operating beyond their design specification… may be between 2 to 3 times larger than those costs when operating within them.”[1]  For this reasons, the CAISO intends to account for the increased costs of operating outside a designed range by using the highest cycling cost value in the proposed DEB formula.  This approach is not sufficiently precise and may lead to inequitable outcomes for ratepayers. 

Accounting for Deep Discharge Cycles

CPUC staff would appreciate a more robust stakeholder discussion on methods to account for the costs difference between shallow and deep cycling for lithium-ion batteries to develop a DEB formula that achieves equitable outcomes for ratepayers and storage operators.

As the CAISO states, the revised straw ESDER Phase 4 proposal “included a dynamically calculated default energy bid that could change on an interval-by-interval basis directly with depth of discharge or specific dispatch for storage resources.”  However, this approach was abandoned for the aforementioned approach of using the highest cycling cost value in the DEB.  As stated, CAISO’s proposed approach could have inequitable outcomes for ratepayers.

The MSC also notes in its Opinion that “deep discharge cycles have been an underrecognized costs in electricity markets.”[2]  Nevertheless, the MSC proposes a static $/MWh value to represent storage cycling costs and suggests further refinements to this approach as “computational capabilities improve.”[3]  A static cycling cost value could have inequitable outcomes for storage operators.

[1] CAISO Energy Storage and Distributed Energy Resources Phase 4 Final Proposal, August 21, 2020, p. 25.

[2] CAISO MSC Opinion on Energy Storage and Distributed Energy Resources Phase 4, September 3, 2020, p. 10.

[3] CAISO MSC Opinion on Energy Storage and Distributed Energy Resources Phase 4, September 3, 2020, p. 11.

5. Please provide your organization’s feedback on the proposal to reflect demand response operational characteristics:

CPUC staff has no comment on this proposal at this time.

6. Please provide your organization’s feedback on the valuation and operational processes for variable-output demand response:

CPUC staff has no comment on this proposal at this time.

7. Provide any additional comments on the final proposal:

CPUC Staff Supports an Alternative to the Minimum Charge Requirement Proposal

CPUC staff requests an alternative to the Minimum Charge Requirement (MCR) proposal the CAISO introduced at the March 3, 2020 ESDER Phase 4 stakeholder meeting.  This proposal would impose an MCR on energy storage resources in the real-time market to ensure energy storage capacity is available to meet day-ahead market awards.  If implemented, this proposal would make energy storage resources less flexible in the real-time market and would likely reduce their value to the grid.

As an alternative to the MCR, the MSC, Western Power Trading Forum[1] and CPUC staff recommend improvements to the CAISO’s real-time market optimization.  Specifically, the MSC states in its Opinion on ESDER Phase 4 that

if the real-time horizon for operations in mid-day or later is extended to cover the evening peak; the real time optimization can determine when it is best to charge and discharge energy or to just hold it, as opposed to a highly conservative rule that is imposed because the present real-time market cannot see beyond its limited time horizon.  Therefore, although it may not be practical at this time to extend the time horizon of the RTPD and RTD markets, evaluating the potential for such extensions should be a priority.”[2]

Conclusion

CPUC staff appreciates the opportunity to comment on the ESDER Phase 4 stakeholder initiative and looks forward to continuing involvement in this stakeholder process.

[1] The Western Power Trading Forum (WPTF) comments on the Energy Storage and Distributed Energy Resources Phase 4 Second Revised Straw Proposal, March 16, 2020, p. 5. “WPTF asks that the CAISO evaluate what extended horizons could be feasible. It may be the case that extending the horizon to, for example 10 hours, may be long enough to address most of the issues raised herein as well as during other previous stakeholder initiatives.”

[2] CAISO MSC Opinion on ESDER Phase 4, September 3, 2020, p. 18.

1. Please provide summary of your organization’s comments on the final proposal:

OhmConnect opposes with caveats the California ISO (CAISO) Final Proposal for the Energy Storage and Distributed Energy Resources Phase 4 (“ESDER 4”). OhmConnect is concerned that the implementation of final proposal topic #4 (vetting qualification and operational processes for variable-output demand response resources) will preclude and limit demand response participation at precisely the time, given the events of the last thirty days, that demand response is needed the most. 

2. Please provide your organization’s feedback on the state of charge parameters proposal:

 OhmConnect has no feedback on this proposal.

3. Please provide your organization’s feedback on the proposed Non-Generator Resource market participation agreements:

 OhmConnect has no feedback on this proposal.

4. Please provide your organization’s feedback on the proposed market power mitigation model for storage resources:

 OhmConnect has no feedback on this proposal.

5. Please provide your organization’s feedback on the proposal to reflect demand response operational characteristics:

  OhmConnect supports the removal of the minimum 1 MW curtailment capability and registration of a Pmax value greater than 1 MW requirement.

6. Please provide your organization’s feedback on the valuation and operational processes for variable-output demand response:

OhmConnect understands that the CAISO is proposing an Effective Load Carrying Capability (ELCC) methodology that would still require further vetting with regulatory authorities, including the California Public Utilities Commission (CPUC), before it is fully adopted as a way to establish Resource Adequacy qualifying capacity. However, OhmConnect is extremely concerned that the proposal as written will limit the total availability of demand response precisely at a time when all resources, including demand response, are needed by the California grid. As written, the ELCC would haircut demand response resources unless they meet the description of a “Perfect Generator”. Furthermore, the haircut would be applied in addition to the load impact protocol analysis, which already does not measure the maximum capability of a demand response resource. As the Final Proposal states, the maximum capability would be determined “under 1 in 2 peak load weather conditions”,[1] which is generally not equivalent to the maximum capability of a resource. Put another way, a demand response resource is fully capable of outperforming its modeled performance under 1 in 2 peak load conditions, especially if the conditions are 1 in 10, or 1 in 100. The use of 1 in 2 peak load conditions to measure maximum capability is flawed for this analysis.

OhmConnect is also concerned that the Final Proposal states that “the current load impact protocols…[do] not necessarily align with when resources are needed to avoid a loss of load event when considering the availability of other resources on the system.”[2] There appears to be a subtext that demand response resources cannot support emergency grid conditions. However, this is directly contradicted by the CAISO’s own numerous press releases throughout August and September lauding the use of targeted conservation (i.e. demand response) to avoid widespread power outages.[3] Demand response is available and effective in mitigating grid emergencies and avoiding power outages. Taking steps to reduce the overall availability of demand response would be unfortunate and counterproductive.

For these reasons, we urge the CAISO to ensure that the ELCC will not preclude or unreasonably limit demand response participation in the CAISO before adopting the Final Proposal.

OhmConnect is supportive of a bidding option for demand response resources to provide variable bids. We would support adoption of this methodology independent of the ELCC methodology.

[1] Final Proposal, p. 43

[2] Final Proposal, p. 43.

[3] August 17, 2020 “No electricity outages expected today, says California ISO”

August 19, 2020 “No power outages expected today; consumers answered the call again”

September 5, 2020 “Conservation helps grid avoid outages today, says ISO”

September 6, 2020 “Grid avoids calling for outages with help from conservation, says ISO”

7. Provide any additional comments on the final proposal:

OhmConnect appreciates the opportunity to provide feedback on the CAISO Final Proposal, and the efforts of the CAISO and E3 to model the impact of an ELCC methodology. Nevertheless, we ask that the CAISO definitively show that adopting the ELCC methodology for demand response will not unreasonably reduce the amount of available demand response. In absence of this analysis, we are deeply concerned that the adoption of the ELCC methodology will exacerbate the challenges the grid faces.

1. Please provide summary of your organization’s comments on the final proposal:

 PG&E’s comments can be summarized as follows:

• More details are required on how the CAISO plans to automate the minimum state of charge constraint in order to protect day-ahead (DA) awards
• The opportunity cost formula in the market power mitigation section should be corrected/clarified in order to avoid confusion
• The CAISO should confirm the scenario that exists when bid cost recovery (BCR) payments are at risk during end-of-hour (EOH) state-of-charge (SOC) parameter use
• CAISO’s demand response (DR) proposal represents a fundamental change to the current program design of DR and PG&E supports the CAISO coordinating with the CPUC in the Resource Adequacy proceeding.

2. Please provide your organization’s feedback on the state of charge parameters proposal:

PG&E identified an issue with the EOH SOC parameter and RTD market horizons which has BCR implications. In discussions with the CAISO, it was confirmed that EOH SOC parameter on market outcomes will only begin mid-way through the hour, and CAISO settlement systems can only implement changes to settlements on an hourly basis. This has the potential risk of underpayment of BCR for those intervals which have already cleared (up to the first six RTD intervals). The CAISO is aware of this risk, but is more concerned by the possibility of overpaying resources BCR in the last six intervals due to uneconomic dispatch caused by the EOH SOC.

PG&E understands the CAISO to be making a trade-off between placing some BCR payments to Scheduling Coordinators at risk and the possibility of overpayments or abuse this optional parameter. If PG&E’s understanding is incorrect, the CAISO should provide more clarification.

Another important subject raised in stakeholder discussions is the CAISO’s intention to make the minimum charge constraint on RA battery resources automatic. PG&E understands this as the CAISO’s attempt to protect DA awards (i.e. DA must offer obligation) without forcing SCs to use the EOH parameter. This concept requires more language in the ESDER4 proposal and should therefore not be forced into the RA Enhancements initiative without first being addressed in ESDER4. It is imperative to understand how this construct would be implemented because the details would impact how the parameter will likely be used by scheduling coordinators. For example, would the minimum charge constraint be automatically adjusted all day or just a few hours ahead of the EOH target?

3. Please provide your organization’s feedback on the proposed Non-Generator Resource market participation agreements:

No comments. 

4. Please provide your organization’s feedback on the proposed market power mitigation model for storage resources:

PG&E appreciates the CAISO’s clarification of the language in the default energy bid formula as it pertains to “variable costs”. However, the formula for opportunity costs (page 25) is still written in a way which suggests the opportunity cost (OC) will always either increase or remain the same compared to the previous day’s OC. PG&E does not believe that this is CAISO’s intention, therefore recommends further clarification.

PG&E presents the following example to illustrate its point. Let’s assume the CAISO is calculating the opportunity cost component of Monday’s default energy bid, and the actual cleared DA prices for the previous day (Sunday) were $45, $35, $32, $30, $27, $31, $40. Since the fourth highest hourly price was $32, the equation becomes:

                                                    OC_MON = $32 * MAX ( (DAB_MON / DAB_SUN) , 1)

And if we assume that (DAB_MON / DAB_SUN) < 1 , we arrive at the following:

Then, Monday’s DA market clears with prices $10 less than Sunday’s (e.g. $35, $25, $22, $20, $17, $21, $30). So the following day, Tuesday, the OC component becomes:

The CAISO’s notation of OCT-1 suggests that instead of $22 in the equation above, we should rather use OCMON  which as we’ve demonstrated before is equal to $32. In other words, there needs to be way to differentiate the OC calculated value with the 4^th highest hourly price from the previous day’s cleared DA market.

5. Please provide your organization’s feedback on the proposal to reflect demand response operational characteristics:

 No comments

6. Please provide your organization’s feedback on the valuation and operational processes for variable-output demand response:

As the Final Proposal maintains the variable output demand response proposal represented in the draft final proposal, which recommends the ELCC as the evaluation methodology in determining DR qualifying capacity, PG&E reiterates its concerns with ELCC and adds one technical comment.

There is a fundamental difference between how resource adequacy values DR and how the CAISO prefers DR to operate.  The CPUC should provide clear guidance. Specifically, the RA program design assesses a resource’s contribution at the monthly peak, while the ELCC methodology favors a resource that is available 24/7. The DR load impact evaluations have been evaluating DR capacity at the monthly peak, in keeping with the RA program design. What the CAISO is proposing represents a fundamental change to the purpose of DR, because DR has been designed to be primarily peak-shaving, although DR is also available to provide load reduction for non-peak hours. Currently, RA has not expected DR to provide as much impact on the average day as during the peak. If the CAISO requires DR to operate differently by applying different evaluation criteria, PG&E believes the CPUC should first clarify the purpose of DR, considering various factors holistically associated with the program redesign (including its cost-effectiveness). As such, PG&E supports the CAISO bringing up this matter with the CPUC in the RA proceeding for better alignment between the two agencies.

On a technical note, as CAISO was informed that the DR bids do not include the 15% PRM and T&D line losses, PG&E would appreciate the CAISO or its consultant updating the comparison between DR bids and the RA Qualifying Capacity. The previous results cast a negative light on DR performance, overstating the difference between DR bids and the QC. An update, which incorporates our comment on the PRM and line losses, would be in order.

7. Provide any additional comments on the final proposal:

no comments

1. Please provide summary of your organization’s comments on the final proposal:

SDG&E supports the final proposal with the caveats discussed below.

2. Please provide your organization’s feedback on the state of charge parameters proposal:

SDG&E generally supports the concept of the optional end of hour state of charge (EOH SOC) tool.  However, SDG&E is still concerned that the usage of the tool would impact the resource’s future ability to sell resource adequacy capacity under the CAISO’s Resource Adequacy Enhancements initiative.  The resource should not be penalized for providing its full capability in a single hour, one hour at a time, to the CAISO because the CAISO’s methodology divides EOH SOC value of one hour by a rolling four hour time frame.

Additionally, SDG&E requests that the CAISO provide more discussion of how the minimum charge requirement, also proposed in the RA Enhancements initiative, will work in tandem with the EOH SOC.  Which tool or constraint will impact or override the other tool or constraint in real time?

3. Please provide your organization’s feedback on the proposed Non-Generator Resource market participation agreements:

SDG&E generally supports simplifying the agreements, or consolidating such contracts, as the CAISO proposes.   

4. Please provide your organization’s feedback on the proposed market power mitigation model for storage resources:

SDG&E looks forward to additional discussions on market power mitigation and default energy bids for energy storage resources.

5. Please provide your organization’s feedback on the proposal to reflect demand response operational characteristics:

SDG&E supports the proposal to establish parameters that better define operational characteristics of demand response resources.  The CAISO notes that this optional parameter is limited to proxy demand resources that have a curtailment of 1 MW or greater.  SDG&E requests the CAISO to clarify whether resources that have a pmax of less 1 MW will continue to receive dispatch instructions at pmin of 0 MW because such resources would not be eligible to utilize this proposed parameter.  It may be helpful to limit the pmin of resources with less than 1 MW to be the pmax such that CAISO market optimization does not dispatch such resources at 0 MW.  The benefits of not dispatching demand response resources at 0 MW should be applicable to all such resources rather than just resources that has a pmax of greater than 1 MW.

6. Please provide your organization’s feedback on the valuation and operational processes for variable-output demand response:

SDG&E understands that the CAISO is not proposing to make any tariff modifications for the qualifying capacity methodology for variable output demand response resources.  The CAISO explained during the stakeholder call that it will be submitting this proposal into the CPUC’s RA proceeding.  As such, SDG&E looks forward to additional discussions in the CPUC’s RA proceeding to better understand the proposed expected load carrying capability (ELCC) and compare it to the current load impact protocol methodology. 

SDG&E generally supports further review of the CAISO’s ELCC proposal.  Drilling down, more granularly, into what DR is actually providing, how it is being bid and dispatched, and how it is being valued, are all issues worthy of scrutiny.  SDG&E appreciates the CAISO’s diligence in understanding DR and working toward assurances that DR is providing a resource/service that is performing to the best of its ability and being compensated fairly, particularly in order to protect ratepayers.   While there have always been optimistic views of DR, and what it can deliver, if it is found to not be able to deliver what the markets need with the reliability of other resources, it should not be compensated the same.  Demand response should and must compete well against other resources without special treatment in order to provide the most value to ratepayers for those costs.     

SDG&E agrees that it is important to develop appropriate and comparable qualifying capacity methodologies for utility, demand response auction mechanism, and third-party demand response resources reflective of their contribution to reliability.    

SDG&E requests the CAISO to consider developing processes to allow for demand response resources to properly participate in the CAISO market due to the variable nature of the resource.  Based on the discussion during the stakeholder call, the CAISO does not have a timeline or proposal to implement such changes if the CPUC were to adopt this methodology.  SDG&E believes this is an important step to ensuring that demand response resources can better participate in the CAISO markets and meet their must offer obligations.  

On all issues related to demand response, as the CAISO intends to work with the CPUC as required in how this proposal might be implemented, SDG&E looks forward to participating in working out those details as part of the stakeholder process.

7. Provide any additional comments on the final proposal:

SDG&E believes the CAISO should postpone the board of governors’ approval until the default energy bid and market power mitigation proposal is also complete as a whole package rather than splitting this initiative into multiple proposals requiring separate board of governors’ approval.

1. Please provide summary of your organization’s comments on the final proposal:

SCE appreciates the opportunity to offer comments on the CAISO's final proposal in the Energy Storage and Distributed Energy Resources Phase 4 initiative. Although SCE supports the CAISO's commitment to scheduling flexibility for resource scheduling and the ease of resource participation in its markets, SCE has some residual concerns about the following matters:

a) the end-of-hour state of charge parameter,

b) the Market Surveillance Committee's assumption of sunk cost for charged resources within the market power mitigation discussion of the Committee's opinion rendered on ESDER4; and

c) the need for further work on the ELCC methodology used in the determination of the capacity credit for demand response resources within the resource adequacy program. 

2. Please provide your organization’s feedback on the state of charge parameters proposal:

While SCE understands that multiple use and contractual arrangements drive interest among stakeholders in the end-of-hour state-of-charge (EOH SOC) parameter, caution should be exercised by the CAISO in the tradeoff between providing greater scheduling flexibility for resources and the reliability of the grid. Thoughtful representation of values for the EOH SOC parameter will be required of Scheduling Coordinators for compliance with the contractual terms and conditions that govern individual storage resources if the prevalence of infeasible schedules is to be avoided as outcomes from the market optimization software. SCE anticipates that with great variability among specified values of the EOH SOC parameter, longer solution times will be required for the arrival at solution outcomes from the market optimization software, while also facing a reduced feasible solution set. Simultaneously, this circumstance also does not lead to a positive contribution toward increasing supply flexibility. In addition, the awards made in the day-ahead market are unlikely to prevail in the real-time market for those resources with feasible or infeasible schedules in the DAM when the EOH SOC parameter is used given the granularity differences in the dispatch intervals between the two operating environments. 

The likelihood of infeasible schedules in the day-ahead and real-time markets is made manifest by the poor alignment between the last advisory interval for the FMM, RTPD and RTD software environment and the differences in the time horizon for the market optimization process in the day-ahead and real-time markets absent any consideration of inaccuracies in the load forecast used in both markets. 

SCE believes that SCs may be better served to manage market risk and the potential for triggering market mitigation by using the value of energy alone in the offer curves (with the EOH SOC parameter made implicit) for scheduling their resources during periods of significant net load ramp variability. The value of energy may be accompanied by the EOH SOC parameter in the shoulder periods when prices tend to be lower and foregone opportunities tend to be lower in value during those hours. 

SCE concurs with the CAISO’s commitment to ensure that storage resources with energy and ancillary services awards will be dispatched in a manner that reserves sufficient capacity for the ancillary services awarded whether that results in under-delivery of any energy award made to the resource. However, SCE is concerned that between the infeasible schedule associated with the attainment of an infeasible EOH SOC target requested for the resource, and the resource owner’s exposure to the under-delivery of energy awards when insufficient capacity exists to deliver the energy and ancillary services awards simultaneously, storage resources may incur penalties for uninstructed deviation from their day-ahead schedules and will very likely be ruled ineligible for bid-cost recovery when these infeasible situations arise. As such, while SCE understands and supports the CAISO’s position toward penalty application, the CAISO is urged to apply the design in a way that mitigates the chances of resources ending up in positions of infeasibility unrelated to resource performance. 

SCE remains optimistic that the learning period for minimizing such negative experiences among SCs in the market will be quite lengthy though SCE prefers that the outcome proves inexpensive for consumers. 

3. Please provide your organization’s feedback on the proposed Non-Generator Resource market participation agreements:

SCE supports the CAISO’s decision to allow stand-alone storage resources to be governed by the participating generator agreement under the non-generator resource model while allowing storage resources that are coupled with demand response resources to be governed by the participating load agreement consistent with all metering and interconnection requirements that govern those arrangements. 

4. Please provide your organization’s feedback on the proposed market power mitigation model for storage resources:

SCE prefers to defer further feedback on the current proposed market power mitigation model for storage resources contained within the draft Final Proposal in light of the opinion rendered by the Market Surveillance Committee on market power mitigation. In particular, the MSC’s assumption that a storage resource that is fully charged at the beginning of the time horizon for the market optimization software, but is partially discharged at the end of that horizon, the value of the residual energy in the storage device should be regarded as a sunk cost for future periods outside of that current time horizon. Questions that are arise from the MSC’s views are: 

a) Is the value of that residual energy really $0/MWh whether additional charging occurs? 

b) If zero with no additional charging, does that outcome leave the resource with self-scheduling as the only option? 

c) Will the default bid for such a resource be $0/MWh when cycling is absent and no further charging occurs? 

d) If additional charging occurs, will the resource be required to value the blocks of energy differently? 

?

Valuing the energy stored in the resource based on when charged, relative to the optimization time horizon, inside or outside, will add another layer of complexity to resource scheduling and market power mitigation since there will be many tranches of energy to track through the operating day for individual resources with more than one charging period. 

5. Please provide your organization’s feedback on the proposal to reflect demand response operational characteristics:

SCE reiterates the desire for expedited implementation of the maximum daily run time parameter. The CAISO’s ESDER4 proposed implementation timeline is October 2021, which is after the summer event season. SCE’s preference is to have the maximum daily run time parameter activated before the summer event season next year to systematically resolve the issue of DR resources being dispatched beyond program limitations. 

6. Please provide your organization’s feedback on the valuation and operational processes for variable-output demand response:

SCE appreciates the CAISO’s acknowledgement of the discrete dispatch limitation in SCE’s 2019 bid quantities used by E3 in the ELCC study. E3’s analysis suggested that DR does not bid into the CAISO market, in aggregate, at levels equal to its NQC value.1 SCE recreated the ELCC study’s comparison of monthly max bids to NQC values, using the modified bid quantities without the discrete dispatch limitation, to prove that SCE’s DR resources were in fact bidding near NQC levels. The CAISO’s final proposal also recognizes that the modified bid quantities should be considered in future studies. 

SCE looks forward to properly vetting the ELCC methodology in track 3 of the CPUC RA proceeding. 

Any analysis of the Demand Response contribution to reliability should look closely at the recent performance during critical grid conditions. Based on SCE system-level load data, SCE has observed an exceptionally strong performance of its DR portfolio – in excess of the MW levels credited to DR in the Resource Adequacy process. These observations are not surprising considering that the value used in the RA process is a 1-in-2 DR forecast from the Load Impact Protocols – meaning that in time of high loads, the load reduction potential also increases.  

7. Provide any additional comments on the final proposal:

N/A