UNITED STATES OF AMERICA BEFORE THE FEDERAL ENERGY REGULATORY COMMISSION

UNITED STATES OF AMERICA BEFORE THE FEDERAL ENERGY REGULATORY COMMISSION Price Formation in Energy and ) Docket No. AD14-14-000 Ancillary Services Markets Operated ) By Regional Transmission Organizations ) And Independent System Operators ) COMMENTS OF THE MIDCONTINENT INDEPENDENT SYSTEM OPERATOR, INC. Pursuant to the Notice issued on January 16, 2015, 1 by the Federal Energy Regulatory Commission ( Commission ), the Midcontinent Independent System Operator, Inc. ( MISO ) hereby submits comments on the Commission s questions related to the series of technical workshops to evaluate issues regarding price formation in the energy and ancillary services markets operated by Regional Transmission Organizations ( RTOs ) and Independent System Operators ( ISO ). I. PRICE FORMATION AND THE MISO MARKET VISION PROGRAM The development of appropriate price signals is a critical element fostering economically efficient electricity markets and helps ensure that MISO is able to deliver power to customers with a least-cost dispatch of available resources. MISO s markets and prices are designed to align the incentives of market participants with the reliable operation of the grid. Further, MISO continues to strive toward an energy market that provides timely, accurate and transparent 1 Notice Inviting Post-Technical Workshop Comments, Price Formation in Energy and Ancillary Services Markets Operated by Regional Transmission Organizations and Independent System Operators, Docket No. AD14-14-000 (January 16, 2015). 1

signals sufficient to recover as much of the costs to supply electricity as possible through the energy market itself. Where economically efficient alternatives in energy market design can be identified, they will be pursued in a manner that reduces uplift payments. To continue improving its market and provide stakeholders a roadmap for future market development, MISO has instituted a Market Vision Program that strongly aligns with the Commission s goals identified as part of its Price Formation docket. This program forms the basis of many of MISO s efforts to improve market price formation. 2 MISO s Market Vision Program s Guiding Principles, developed in collaboration with MISO s stakeholders and Board of Directors, were established to: 1. Support an economically efficient wholesale market system that minimizes cost to serve load 2. Facilitate nondiscriminatory market participation regardless of resource type, business model, sector or regional location 3. Develop transparent market prices reflective of marginal system cost and cost allocation reflective of cost-causation and service beneficiaries 4. Support market participants in making efficient operational and investment decisions 5. Maximize alignment of market requirements with reliability requirements of the system MISO focuses on the continued identification, prioritization and evaluation of initiatives and the tracking of progress in such efforts. From project review to implementation, the Market Vision Program ensures adequate consideration and effective evaluation of all market improvement opportunities, selecting those that best align with MISO s Market Vision. 2 MISO s Market Vision Program is located at: https://www.misoenergy.org/library/repository/communication%20material/market%20enhancements/2014%2 0Market%20Vision.pdf 2

To identify and address areas of improvement, the Market Vision Program looks comprehensively at MISO s market operations. Specifically, improving price formation has been a key focus in MISO s recent market enhancement efforts. Initiatives are undertaken with the objectives of improving efficiency of short-term prices under all operating conditions and providing long-term economic signals that support efficient development of resources consistent with long-term reliability and/or public policy objectives. In fact, MISO takes a holistic approach to ensure efficient market processes underlying price formation, and is committed to enhance unit commitment and economic dispatch processes, maximize economic utilization of existing and planned transmission infrastructure, and facilitate efficient transactions across seams with neighboring regions. Finally, MISO is a strong believer in market transparency and strives to improve availability of non-confidential and non-competitive market information. MISO has several high priority market enhancement efforts already underway that are designed to improve price formation within the MISO markets. These include: $1,000 Offer Cap Evaluation Emergency Demand Response Resource Enhancement (implemented on 3/1/15) Extended Locational Marginal Pricing, Phase I (implemented on 3/1/15) Emergency Energy and Demand Response Pricing MISO-PJM Interchange Modeling and Pricing Enhancements Ramp Capability Product MISO-PJM Interchange Optimization MISO-SPP Market-to-Market Coordination (implemented on 3/1/15) Additional price formation initiatives are in various stages of evaluation, including: Expanded Capacity for Contingency Reserves 3

Look-Ahead Multi-Interval Dispatch Look-Ahead Commitment, Phase II Combined Cycle Generation Offer Configurations Extended Locational Marginal Pricing, Phase II Elimination of Make-whole Payment for Contingency Reserve Deployment MISO appreciates the opportunity to participate in the Commission s series of workshops on price formation issues and looks forward to continuing to work with the Commission on these issues. The information collected in the docket complement and inform MISO s existing market enhancement efforts. MISO provides responses to the Commission s questions below. II. RESPONSES TO COMMISSION QUESTIONS 1. Offer Caps High natural gas prices during the winter of 2013-2014, as discussed at the price formation workshops, indicated that the current generic $1,000/MWh cap on energy offers ( offer cap ) might be insufficient to allow natural gas-fired generators to recover their costs when natural gas prices spike during constrained winter periods. a. Should the $1,000/MWh offer cap be modified? MISO Comment: Yes. The $1,000/MWh offer cap should be modified to ensure all resources are able to at least recover the incremental cost to produce energy supplied to the RTO markets. It is appropriate for any offer cap to ensure suppliers may reflect their actual operating costs into their bid. Offer structures that do not support cost recovery can increase the likelihood that a given and needed resource may not be available to the operators. In present terms, a higher 4

energy offer cap level would be an effective modification to account for recent experience with dramatically higher natural gas prices experienced in 2014. i. If the offer cap is modified, what form should the offer cap take? For instance, should a modified cap be set at a level greater than the current $1,000/MWh cap and apply even if a resource has costs greater than the new cap or should the offer cap be replaced with a structure that allows offers at the higher of marginal cost or the existing $1,000/MWh cap? Should it be a fixed cap or a floating cap that varies with the price of fuel (e.g., natural gas)? If a modified cap were set as a fixed offer cap, what should the new offer cap be? What should be the basis for determining the fixed offer cap? MISO Comment: At this time, a fixed energy offer cap is preferable to a dynamic price cap. A fixed energy offer cap provides both greater clarity of rule to market participants and also simplifies the process of implementing related markets that have an intrinsic tie to the maximum price level in the energy market. These may include transmission demand curves, emergency or scarcity pricing regimes and some ancillary services. Further, the current RTO market management software systems are designed with a fixed offer cap, and transition to a dynamic cap would likely require substantial system changes. A fixed cap will also provide a backstop check on the exercise of market power in fuel or other markets outside the RTO s scope that contribute to the cost of energy production and whose effects on energy clearing prices 5

would be difficult to remedy after the fact. MISO supports continuing attention to gas market transparency and liquidity to ensure that generation resources are able to avail themselves of a well-functioning fuel transportation and supply market. In the longer run, however, given growing dynamic nature of the cost inputs into energy prices, MISO plans to explore a transition toward a more dynamic price cap method that accounts for the increasing breadth of cost influences on energy price production. The quickly evolving set of cost influences may require a dynamic price cap method to keep pace with changes as they emerge in the future. ii. If the offer cap should not be modified or set such that marginal costs could be greater than $1000/MWh, how should the Commission ensure that suppliers with costs greater than the cap have the opportunity to recover those costs? MISO Comment: As noted, MISO believes the price cap should be modified to ensure marginal costs are reflected in the energy price. However, if an alternative to such changes is required, MISO s short-term approach during the winter of 2014/15 could be considered. On a short term basis, MISO implemented a method that allows incremental energy costs exceeding $1,000/MWh to be submitted in no-load cost offers and recovered through uplift payments. This short-term approach to the issue was adopted largely to address technical issues that remain uncertain in MISO s systems regarding the reliability of accepting energy offers above 6

$1000/MWh. As noted, however, this is not necessarily a preferable longterm solution as it mutes the price signal for the incremental and temporal cost of energy during such inherently scarce periods and reduces the effectiveness of energy hedges by Load Serving Entities by potentially exposing otherwise hedged customers to additional uplift costs. It is worth noting that marginal costs can potentially exceed an energy offer cap. Therefore, it is appropriate production cost guarantees to ensure cost recovery is available to resources with higher actual costs and to have a process to verify the appropriateness of those costs. iii. Do the real-time and day-ahead market clearing processes allow sufficient time to verify the cost-basis of the marginal resources that exceed the offer cap? Does the settlement process allow sufficient time to verify costs of resources that receive uplift associated with offers that exceed the offer cap? MISO Comment: MISO addressed these additional issues through the implementation of the short-term solution described above. Under its current approach, MISO s Day-Ahead and Real-Time market clearing processes provide sufficient time and procedures for cost verification as long as Market Participants: 1) follow the procedures to notify MISO that the gas price has caused their Energy Offer to exceed the $1000 Energy Offer Price Cap; and, 2) before submitting an Offer, have consultation calls with the Independent Market Monitor to initiate the adjustment of reference levels based on cost. During the settlement process, cost 7

based reference levels will be applied to settle the uplift for those resources. The make-whole payment may be adjusted after-the-fact for some cases involving Offers that do not exceed the applicable reference level. b. What are the advantages and disadvantages of having offer caps be set at the same level across all RTOs/ISOs? Would different offer caps across the RTOs/ISOs exacerbate interface pricing issues at RTO/ISO borders? If so, how? Would an offer cap that takes the form of the higher of marginal cost or $1,000/MWh create the same issues as setting different offer caps across RTOs/ISOs? MISO Comment: Having Offer Caps at the same level helps establish consistent shortage pricing levels between neighboring RTOs/ISOs. This, in turn, encourages the RTOs/ISOs to take appropriate actions to enhance reliability for the affected areas. Different offer caps across the RTOs/ISOs could exacerbate interface pricing issues when shortages occur. Issues might emerge, for example, if resources were committed day ahead in one RTO, but they decide to export power to a neighboring RTO with a higher cap price, only needing to buy through the original commitment at no more than a lower cap. Also, during tight operating conditions, differing offer caps that are not explicitly tied to different costs of production could also provide undue and counterproductive arbitrage opportunities for export trades between neighboring RTOs. c. What impact would adjusting the offer cap have on other aspects of RTO/ISO price formation (e.g., mitigation rules or shortage pricing rules)? Would other 8

market rule changes be necessary if offer cap levels were adjusted? Do other challenges associated with modifying offer cap rules exist? If so, what are they? If offer cap rules are adjusted, how quickly could RTOs/ISOs incorporate adjusted offer cap rules into their software and the market clearing process? MISO Comment: Adjusting the offer cap would likely require changes to shortage pricing rules and mitigation processes, as well as other products in the MISO markets, including operating reserves and the associated demand curves in order to maintain consistency of the simultaneous co-optimization process approved by the Commission. If a dynamic cap is implemented, such changes would likely require more complex changes that have not been previously examined and are uncertain in nature. As such, detailed assessment of all of these tertiary affects is needed before adjusting the offer cap. d. Should the same offer cap that applies to generation also apply to load bids? What are the advantages and disadvantages of applying an offer cap to load bids? MISO Comment: To the extent that load bids are able to set the marginal price, they should also be subject to the same Offer Cap rules as the generation offers. 2. Transparency At the Uplift and Operator Actions Workshops, some panelists addressed issues concerning insufficient transparency of uplift and operator actions. Improved transparency could inform resource entry and exit and market rule discussions; improved transparency could also improve market understanding, predictability, and confidence. a. What should RTOs/ISOs do to improve transparency of uplift credits and charges, unit commitment, and other operator actions? Please comment on the type of 9

information that would be useful, why it is necessary, whether it should be shared with specific resources or available to all, the timing of its release, and whether it is feasible to release the information in real-time. MISO Comment: MISO currently takes several proactive steps to publish information in order improve transparency. The information MISO currently makes available to market participants include real time commitments for congestion and capacity on a five minute basis, binding transmission constraints and a number of daily and monthly reports of uplifts, including Revenue Neutrality Uplift and Revenue Sufficiency Guarantee, providing summary and detail level information for market participants. Market participants are able to utilize this information to develop a good understanding of the system conditions and adjust their bidding patterns. MISO is continuously exploring opportunities to make additional information available that may be beneficial for participants, specifically related to reliability constraints that cannot be represented in the economic commitment and dispatch model. Additionally, MISO s Data Transparency Working Group solicits feedback and suggestions from stakeholders on data needs that can improve the transparency of MISO s market operation. b. What types of information should not be shared publicly? Why? What are the concerns with commercially sensitive information? MISO Comment: In general, market participant-specific and facility-specific information is confidential. This would include, in particular, offer/bid prices. Specific 10

information relating to an individual market participant or a particular unit must be protected as it may include confidential, commercially or competitively sensitive information, and/or Critical Energy Infrastructure Information ( CEII ). A detailed discussion would need to take place in the appropriate stakeholder forum to balance information needs against the unintended impact of providing information that is confidential for commercial or other reasons. The reasons for confidentiality include commercial or competitive sensitivity, business proprietary interests, gaming/manipulation/collusion risks, and classification as CEII ; and the no-conduit rule as between transmission and merchant/market-side personnel. Information is deemed commercially sensitive if it is proprietary (such as trade secrets), or if its disclosure is likely to prejudice the commercial interests of its owner (e.g., data on production costs, bidding strategies, or revenues). MISO already discloses to Market Participants substantial information regarding uplifts and operator actions. MISO believes the types of information it discloses, as mentioned above, sufficiently meets the need for transparency, while balancing confidentiality needs. c. Commission Staff s August 2014 report on uplift noted several issues with the consistency and granularity of uplift data provided as part of the Electric Quarterly Reports. What steps could be taken to improve the quality of uplift data required to be reported as part of the Electric Quarterly Reports? MISO Comment: The data provided in the Electric Quarterly Reports (EQRs) and reported to comply with FERC Order No. 760 needs to be analyzed to 11

reconcile differing reporting requirements between these two data sources. Any inconsistencies in the inclusion of certain costs or revenue should be reconciled so that a consistent set of information is presented in both of these reports. MISO provides EQR data to market participants for augmentation before it is submitted to FERC. This analysis and reconciliation needs to be performed at an individual ISO/RTO basis to account for the differences in the categorization of uplift costs or revenues. 3. Pricing Fast-Start Resources Commission Staff s December 2014 paper about operator-initiated commitments discussed how RTOs/ISOs relax the minimum operating level of resources to make certain block-loaded fast-start resources appear dispatchable to the pricing software, and thus eligible to set the market clearing price as the marginal resource. The paper also discussed how some RTOs/ISOs have modified the locational marginal price (LMP) framework to include start-up and no-load costs of certain fast start resources (e.g., New York Independent System Operator, Inc. s (NYISO s) Hybrid Pricing). a. During the Operator Actions Workshop, panelists explained that relaxing resource minimum operating limits can lead to incentive and operational issues such as over-generation. What tradeoffs are involved with relaxing the minimum operating limits of block-loaded resources to zero for purposes of price setting? Should relaxing the minimum operating level be limited to block-loaded fast-start resources, or should relaxation be available to a larger set of resources? MISO Comment: Allowing block-loaded fast-start resources to set prices by relaxing the minimum operating limits can obtain efficient prices that more 12

accurately reflect the true marginal cost of meeting load requirements. When block-loaded fast-start resources are committed to meet load, the real-time prices can be set by the lower-cost flexible units that are dispatched down to accommodate the full output of the block-load resources. Such price depression will result in uplift payments, muted incentives for scheduling of more costefficient energy in the day-ahead market, and inefficient export and import scheduling. MISO s ELMP addresses these issues by using partial commitment to relax the minimum operating limits in price setting. An issue with relaxing the minimum operating limits is that the dispatch produced in the pricing engine can be slightly different from that in the Unit Dispatch System ( UDS ). As ELMP is performed in an Ex-Post process for pricing only and UDS output is still used for physical dispatch, it will not cause operational issues. Compatibility of incentives can be affected if the prices are associated with a dispatch that deviates from the physical schedule. ELMP is carefully designed to achieve sufficient consistency, as verified by MISO s ELMP and LMP parallel operations tests. MISO believes the relaxation should be applied for a larger set of resources besides block-loaded ones. Accordingly, ELMP applies the relaxation to all faststart resources. 3 ELMP does this because price depression has been observed not only from block-loaded units, but also from other fast-start resources that are dispatched at the minimum operating limits when they are committed to meet load. There is no reason to differentiate the two sets of units if they involve the same issues and the ELMP method is available to address both. Moreover, if only block-loaded units are eligible to set price, there could be risks that fast-start 3 See the definition of Fast Start Resources in Section 1F of MISO s Tariff. 13

resources could sacrifice their dispatch ranges (whose flexibility can be very valuable) and offer themselves as block-loaded in order to set prices. Another set of resources that MISO is including in ELMP is Emergency Demand Response (EDR) resources. This inclusion is strongly supported by MISO s stakeholders and the IMM. As MISO is developing a price enhancement under emergency conditions, resources such as Load Modifying Resources (LMRs) would also be eligible to set prices under ELMP. b. What are the merits of expanding the set of costs included in the energy component of LMP (i.e., start-up and no-load costs)? What factors should be considered when expanding the set of costs included in the energy component of LMP? If the start-up and no-load costs of block-loaded fast-start resources are included in the LMP, how should they be included? For example, should start-up costs only be included during intervals when the resource starts up? MISO Comment: Incorporation of commitment-related start-up and no-load costs promotes price transparency to reflect the actual cost of dispatch and the underlying commitment to serve load. The commitment-related start-up and noload costs of fast start resources are essential parts of the marginal cost to serve load, and their appropriate reflection in prices can provide compatible incentives for market participants to follow commitment and dispatch instructions. For example, the system s operation will consider start-up and no-load costs and may not choose a unit that is able to fast start even though its incremental energy cost is relatively low. Therefore, the price should reflect the effect of fast start resources commitment costs on the marginal cost to serve load. In addition, 14

prices not capturing the full cost can cause revenue adequacy. In such instances, uplift payments have to be used to recover the cost. As pointed out in the FERC staff paper, the problematic uplift will undermine market efficiency and mute investment signals. 4 Nevertheless, the incorporation should be mainly focused on the commitment costs of fast start resources. Slow start units are usually online for a long time and their commitment costs can be covered by prices set by more expensive units. When expanding the set of costs to be recovered, the primary challenge is how to include these costs in an appropriate way. The full ELMP 5 or convex hull pricing provides the mathematic and economic foundation of how to incorporate commitment costs. MISO s ELMP is designed to capture such cost elements while considering the complexity of its implementation and its practical limitations. Another important issue is the comprehensive evaluation of the resulting prices. Expanding the set of costs in the energy component of LMP may appear to increase prices. However, ELMP/LMP parallel operations showed that ELMP equals the LMP most of the time and increases moderately when commitment costs are reflected. More importantly, the increase is offset by reduced uplift payments and the long term benefits of more compatible incentives and efficient 4 Staff Analysis of Uplift in RTO and ISO Markets, August 2014, http://www.ferc.gov/legal/staff-reports/2014/08-13-14-uplift.pdf. 5 P. Gribik, W. W. Hogan, and S. L. Pope, Market-Clearing Electricity Prices and Energy Uplift, Dec. 31, 2007, http://www.hks.harvard.edu/fs/whogan/gribik_hogan_pope_price_uplift_123107.pdf. 15

investments that result from the economically efficient price signals rather than by relying on out-of-market uplift. 6 Analysis of full ELMP shows that commitment costs tend to be included in the intervals when the fast start resources are most needed. MISO calculates ELMP by allocating start-up costs of on-line fast-start resources over the minimum run time since they are usually turned on when most needed and stay online for at least the minimum run time. Allocation for offline fast-start resources are detailed below. c. Should off-line resources be eligible to set the LMP? If so, should start-up and noload costs be included in the price, or just incremental energy costs? MISO Comment: MISO allows off-line fast start resources to participate in ELMP pricing if they are feasible and economic to alleviate the shortage or transmission violation. Under some operational circumstances, off-line fast start resources may be available to resolve some shortages that may otherwise occur. In these situations, there is no real scarcity, in light of the availability of off-line fast start resources. By enabling the participation of such resources, ELMP can effectively reduce the inefficient transient price spikes. On the other hand, when MISO faces scarcity, ELMP reveals the shortage prices. By reducing the price spikes that inaccurately reflect scarcity, and allowing only price spikes that truly reflect scarcity, ELMP sends efficient price signals to the market to attract the supply and investment to intervals when they are needed. 6 S. L. Pope, Price Formation in ISOs and RTOs, Principles and Improvements, page 19-20, Oct. 2014, https://www.epsa.org/forms/uploadfiles/2cc210000016f.filename.epsa_price_formation_oct_29_2014_fina L.pdf. 16

When allowing off-line fast start resources to set price, ELMP fully considers start-up and no-load costs and also minimum generation cost. 7 ELMP parallel operation test runs show that if commitment costs are not included or are understated, off-line fast start resources can over-participate in pricing and depress prices. By working closely with the IMM and stakeholders, MISO designed ELMP to allocate commitment costs for off-line fast start resources to four real-time dispatch intervals. Based on MISO s study, this approach achieves the best balance in allowing economic off-line fast start resources to set prices. MISO will continue monitoring ELMP, and will adopt improvements as needed. 4. Settlement Intervals Panelists at the Shortage Pricing/Mitigation and Operator Actions Workshops generally supported sub-hourly, rather than hourly, settlement intervals as providing better incentives for resources to perform during shortage events and to make investments to enhance resource flexibility. a. What are the advantages and disadvantages of moving to sub-hourly settlements for the real-time market as they relate to price signals, market efficiency, and operations? MISO Comment: Moving to a sub-hourly settlement, with the interval aligned with the dispatch signal, enhances incentives for generation resources to follow dispatch signals. In MISO, Price Volatility Make Whole Payments provide a mechanism to address the disparity in dispatch and settlement intervals. However, these uplifts only partially address the effects of such disparity. MISO 7 MISO Tariff Schedule 29A, section II.B; Affidavit of David B. Patton dated December 15, 2014, page 15, filed in FERC Docket No. ER15-684-000. 17

expects that sub-hourly settlements would reduce the need for, and the amount of, these make whole payments. In this regard, MISO notes that starting June 30, 2015, interchange transactions will be settled in MISO at a sub-hourly interval to comply with FERC Order No. 764. At this time, MISO has not identified any disadvantage of sub-hourly settlement in the area of price signals, market efficiency or operations. b. What metering and RTO/ISO software changes would be needed to change settlement intervals from hourly to sub-hourly for the real-time market, and how long would these changes take to implement? Are there significant costs to RTOs/ISOs, and to market participants, of such changes? Are there any other impediments to adjusting settlement intervals? MISO Comment: Meter information would need to be supplied at a sub-hourly basis. As part of the implementation of the Ancillary Services Market in January 2009, MISO adopted load profiling algorithms that allow hourly meter values to be submitted, and that enable the derivation of a sub-hourly interval value from Supervisory Control and Data Acquisition (SCADA) values. MISO intends to use this same approach for market participants that have the capability to record and submit sub-hourly data. Other RTOs/ISOs have implemented five-minute settlement systems, and their market participants are apparently able to submit meter information at that level. One logistical impediment is determining the necessary changes to the settlement software to address the sub-hourly performance of settlements calculations. MISO will hold stakeholder discussions 18

on the appropriate nature and degree of changes that need to be made to the settlement software. c. What are the advantages and disadvantages of changing from hourly to subhourly settlements in the day-ahead market? MISO Comment: The anticipated advantages of changing from hourly to subhourly settlements in the day-ahead market are the increased consistency between day-ahead and real-time markets and the lower production costs by better managing ramping issues. However, this is only true in the perfect-information world. Due to the imperfect knowledge of real-time net load or unexpected outages one day ahead, scheduling at the sub-hourly level may actually worsen the consistency and cost-efficiency than scheduling at the hourly level with some robustness. Moreover, headroom constraints are currently used at MISO to effectively obtain the ramping flexibility to meet intra-hour changes in system demands. The planned ramp capability product will further enhance the capability to manage ramping issues in real time and provide explicit price signals to incent flexibility as to be elaborated next. Besides the advantages and disadvantages, the associated computation complexity can prevent the change from being realistic given the current software capability. Changing from hourly to sub-hourly (1/N hour) settlements in the day-ahead market will increase (N times) the sizes of the unit commitment and economic dispatch problems. Especially for the mixed-integer unit commitment problem, the complexity and thus the computation time will increase drastically. This will pose major difficulty to clear the day-ahead market in time, and make it more challenging to 19

coordinate gas/electric markets. MISO do not believe that changing to sub-hourly day-ahead settlements is a high value effort for customers compared to other price formation issues discussed in the workshops. 5. New Products to Incent Flexibility Flexible resources that are capable of ramping up and down and/or starting up quickly provide value to the electric system. Panelists at the Operator Actions Workshop said that market designs which reward flexibility may stimulate investment in flexible capacity and provide resources more incentive to submit flexible offers. One panelist at the Operator Actions Workshop commented that existing market rules can create disincentives for resources to submit supply offers that reflect the full flexibility (for example, ramp rate, minimum run time, minimum operating level, maximum operating level, minimum down time) of their resources. In addition, panelists at the workshops discussed the need for locational reserve products to better reflect local needs for flexibility. a. How do RTOs/ISOs currently ensure that they will have sufficient flexibility during real-time? Specifically, to what extent are residual unit commitments used to acquire anticipated needed flexibility? MISO Comment: To reliably operate the system, flexibility is built into MISO s day-ahead unit commitment and real-time economic dispatch processes. For example, headroom constraints have been implemented in the Day-Ahead Market and the Reliability Assessment Commitment processes to obtain a set of commitments of units with sufficient ramping capacity. The headroom and reliability margin are closely monitored throughout the operating day to ensure 20

sufficient capacity is available. The current UDS is utilized to anticipate potential ramping needs by adjusting the anticipated load value. Residual unit commitments are used to maintain the needed flexibility as knowledge of system conditions improves throughout the day. Quick start combustion turbines are committed if there is an expected generator outage, transmission line congestion or deviation from the forecasted load. In addition, residual unit commitments are made as needed to address any un-modeled voltage constraints and local reliability requirements. 8 b. How are flexible resources compensated for the value that they provide to the system? Does that compensation reflect the value? Why or why not? If compensation to flexible resources does not reflect their value, how should RTOs/ISOs compensate flexible resources for the service they provide? MISO Comment: The compensation of flexible resources depends on the nature of their flexibility. For example, operating reserves are explicitly compensated by the marginal clearing prices to satisfy reserve requirements. Headroom constraints or residual commitment can affect pricing when the committed resources set prices, while make-whole payments are needed if a unit commitment incurs additional costs that cannot be covered by market revenues. MISO is also developing a ramp capability product to further recognize the value of resources ramping capability. The ramp capability product involves a cooptimization process with energy and ancillary service products. The clearing price of the product is established as the opportunity cost of being dispatched out 8 Such un-modeled requirements can be significant drivers of residual commitments. FERC staff paper, Dec. 2014 page 15. 21

of merit order in the current dispatch interval by withholding ramping flexibility to provide energy or ancillary services in future intervals. 9 This compensation provides economic incentives for resources to provide the required flexibility. A ramp capability demand curve further establishes the value between exercising the ramp capability to meet the current load and withholding the flexibility for net load in future intervals. MISO is working on the detailed design and construction of the ramp capability product, which is expected to be effective in 2016. c. What are the tradeoffs between sending a price signal through a short-duration shortage event versus establishing a ramping product that is priced separately? MISO Comment: MISO is dedicated to providing transparent pricing and incentives for the supply of resource flexibility. A separately priced ramping product and a short-duration shortage price signal both contribute to this effort and are synergistically used at MISO. The ramp capability product will utilize the existing resource flexibility in a reliable and cost-efficient way by better positioning the system to reduce the short-term scarcity and avoid committing expensive fast start resources. The explicit pricing and compensation provide an incentive to supply resource flexibility. The ramp capability product does not attempt to eliminate all scarcity events. Scarcity that cannot be economically managed by existing resource flexibility is revealed through price signals to attract more supply of flexible resources. d. What are the tradeoffs among procuring flexibility through unit commitments (e.g., headroom requirements) rather than through the ten-minute reserve products or through ramp products? 9 Prepared Direct Testimony of Joseph Gardner on Ramp Capability Product at pages 13-14. 22

MISO Comment: Headroom constraints are enforced in the unit commitments to procure flexibility at MISO, and ramp products are further developed to manage short-term net load variations and uncertainties. The use of headroom constraints results in improved day-ahead real-time price convergence and cost savings. Nevertheless, the ramp capabilities of resources may not be fully reflected in prices. The ramp product will explicitly model ramp capabilities in a cooptimization process with energy and ancillary services throughout the commitment and dispatch. These products are expected to improve operational efficiency and reliability to reduce short-term scarcity, and provide transparent pricing and incentives for the supply of flexibility. e. Does allowing combined-cycle natural gas resources to submit different offers for different configurations facilitate more efficient price formation? What are the advantages and disadvantages to generators of bidding these configurations? MISO Comment: Combined-cycle units are currently offered in an aggregate model at MISO, either as a normal single generator or as separate Combustion Turbines ( CTs ). Allowing configuration-based offers can enable more accurate bidding and possibly greater flexibility to the system to facilitate more efficient price formation. Such types of offers would require major and complex changes to the unit commitment and economic dispatch software, particularly with the size of MISO s system. If system changes are not well addressed, the resulting commitment and dispatch decisions will likely be sub-optimal and can adversely affect pricing. MISO is evaluating the combined-cycle offer configurations and potential solutions. 23

With configuration-based offers, market participants do not need to do their own configuration changes after the Day-Ahead Market is cleared. However, configuration-based offers can also cause generators to lose their offer or operational flexibility in some ways. The advantages and disadvantages of configuration-based offers may vary depending on each market participant s situation. 6. Operating Reserve Zones A lack of sufficiently granular reserve zones could be muting efficient price signals. At the Shortage Pricing/Mitigation workshop, the NYISO panelist noted that NYISO is considering establishing a new reserve zone and the PJM Interconnection, L.L.C. (PJM) external market monitor indicated that he believed PJM s shortage pricing rules were not sufficiently locational. For instance, last year PJM experienced shortages in the American Transmission System, Inc. (ATSI) footprint that did not trigger shortage pricing because the ATSI zone is not a reserve zone. a. How does the establishment, elimination or reconfiguration of reserve zones affect price formation? What should the triggers be? From experience, do the RTOs/ISOs have the appropriate reserve zones defined? Are additional, fewer, or different reserve zones needed? MISO Comment: MISO s market design establishes clear processes for the establishment, elimination, and/or reconfiguration of operating reserve zones ( ORZ ). ORZs are established to ensure Regulating Reserve and Contingency Reserves are dispersed, in accordance with Good Utility Practice, in a manner that 24

prevents adverse operating conditions that affect the reliability of the Transmission System. b. Are processes in place for adding, removing, or changing reserve zones adequate for efficient price formation? MISO Comment: MISO performs studies to determine ORZ configuration on a quarterly basis concurrent with Network Model updates. ORZs may be reconfigured and/or eliminated prior to the next quarterly review if: 1) a condition or event occurs, including, but not limited to, an unplanned transmission facility outage, a Generator Forced Outage, or an event of Force Majeure; 2) such condition or event results in an adverse reliability condition that cannot be resolved through normal operating procedures; 3) such condition or event has a projected duration of two or more Operating Days; and 4) MISO determines such adjustment is necessary to ensure the reliability of the Transmission System. ORZ requirements studies are performed daily and published 48 hours prior to the start of the operating day and establish hourly operating reserve requirements. Operating reserve requirements are established on a market-wide and zonal basis. MISO may clear more resources on either a market-wide or zonal basis in order to enforce requirements. MISO uses market-wide and zonal demand curves, in both the day-ahead and real-time markets, to price operating reserves when insufficient reserves are cleared to meet requirements. 25

7. Uplift Allocation Uplift allocation rules might impact resource participation decisions in RTO/ISO markets. For example, uplift allocation rules might incent participation in day-ahead markets or drive decisions on how to use financial products. a. Do uplift allocation rules reflect cost causation or mute potential investment signals? If so, how? MISO Comment: MISO believes costs that are allocated via uplifts are closely aligned with cost causation. Over recent years a number of initiatives have been implemented to address allocation of costs to better align with cost causation. This methodology, coupled with the availability of more detailed reports on uplift costs, provides information for market participants to understand the allocation of these costs. b. What philosophy should govern uplift allocation? Do any of the RTOs/ISOs have a best practice? What is it and why is it a best practice? MISO Comments: Uplift payments should be aligned with cost causation as much as feasible. MISO s mechanisms for allocating and uplifting current costs are closely aligned with cost causation. c. Should uplift allocation categories reflect the reasons for committing a unit and incurring uplift? Would disclosing these reasons through publicly available data improve uplift transparency and provide information to facilitate modifications of the allocation of uplift costs? MISO Comment: MISO allocates unit commitment make whole costs based, as much as reasonably feasible, on the reasons for committing a unit (e.g., capacity, 26

constraints, voltage, reliability must run). This level of reason-based cost allocation provides transparency of the uplift allocations. 8. Market and Modeling Enhancements At the Uplift and Operator Actions Workshops, panelists highlighted various drivers of persistent, concentrated uplift and operator actions, including constraints that are not incorporated into market models. Panelists also noted that certain constraints are difficult to model accurately or to incorporate into both the day-ahead and real-time market models. These include local voltage constraints and reliability constraints such as N-1-1 contingency constraints. a. Assuming that RTOs/ISOs should improve their market models to better reflect the cost of honoring reliability constraints in energy and ancillary services market clearing prices, what types of constraints should RTOs/ISOs include in their market models, and what types of constraints should be handled by manual commitments? Of those reliability constraints that should be in the market models, which reliability constraints should RTOs/ISOs prioritize? MISO Comment: Transmission constraints that can be managed through commitment and dispatch should be included in the market models. Panelists were correct that certain constraints are difficult to model accurately in certain market models. MISO utilizes thermal proxies and offline studies to create operating guides. MISO is investigating methods to identify voltage and local reliability commitments in the day-ahead process. Incorporating N-1-1 constraints into the day-ahead market is likely too conservative, unless there is a known load pocket issue. The day-ahead market includes planned outages in each 27

hour. By adding the N-1 transmission constraints to the topology that includes the planned outages, the result is a subset of likely N-1-1 constraints as compared to an off-line study with the assumption that all of the transmission system is in service. Incorporating N-1-1 constraints on top of outages can result in seeing congestion that is unlikely to occur in real-time. In the real-time market, state estimation is used to provide the base topology, including equipment already out of service. The N-1 constraints are utilized in the dispatch of the generation. Similar to the day-ahead context, the net effect of utilizing N-1 constraints in the real-time dispatch with the open branches is very similar to an N-1-1 constraint in an off-line study that assumes all transmission is in service. The indirect effect identified in the staff paper on energy and ancillary service prices due to committing additional units may also occur by including N- 1-1 constraints in addition to outages. Additional generation commitments due to the N-1-1 constraints will result in additional supply and might reduce clearing prices. In the case of a load pocket, the transmission system is weakly interconnected to the pocket, and N-1-1 constraints need to be utilized to reliability commit enough generation to serve the load inside the pocket. A load pocket is a commitment issue until the transmission system can be expanded to allow enough energy to flow into the load pocket to enable the load to be served following the worst single contingency. Manual commitments are appropriate for reliability issues or to replenish reserves after deploying contingency reserves. Occasionally, multiple units may go offline in a short amount of time. Manual commitments may be needed to replenish 28

contingency reserves in the expected time identified in the reliability standards. Local reliability issues may also drive the need for a manual commitment. If a resource supporting local reliability trips, a replacement unit may need to be manually committed to replace the resource. Deploying contingency reserves would result in increasing or calling on generation that may not support the local reliability issue. b. In 2013, ISO New England Inc. (ISO-NE) increased its replacement reserve requirement to reduce the need to schedule additional resources above the load and reserve requirements in its Reserve Adequacy Analysis. PJM has a similar proposal to increase day-ahead and real-time reserve requirements when extreme weather is expected. In what circumstances can such practices improve efficiency of price formation? MISO Comment: To the extent that flexible capacity is committed through a unit commitment process that is not based on the procurement of a need that is being priced in the market there is an increased opportunity for uplift costs to support that commitment decision. As a general rule, allowing increased transparency and procuring the specific reliability service needed allows better aligns the market and reliability operations of the system. The improvement of this alignment is the basis of MISO s ramp product implementation noted above. Increasing the reserve requirements through headroom to address flexibility or operational risks allows the RTO to be better prepared for extreme weather. Procuring the headroom through additional reserves in the Day-Ahead allows the 29

RTO to better reflect, in pricing, the need for extra flexibility to be reliable throughout such weather as opposed to a unit commitment outside of that process c. Do transmission constraint relaxation penalty factors improve the efficiency of price formation? If so, should these penalty factors be allowed to set the energy price if a transmission constraint is relaxed? MISO Comment: Transmission constraint penalty factor relaxation can be helpful in certain situations. In constraints that are managed by Transmission Loading Relief ( TLR ), the market may not have effective resources to assist in controlling the constraint. Relaxing the penalty factor allows the market to help according to the capability of the resources. In situations involving other RTOs, market-to-market constraints relax and use the information along with the other RTO s shadow price to determine the most effective way to co-manage the constraint. MISO saw the most improvement by moving to a demand curve for transmission constraints. Allowing transmission constraint relaxation along with appropriate penalty factors improves the efficiency of both dispatch and pricing. It helps to avoid taking unnecessarily costly actions to resolve a minor transmission violation and the penalty factors should be allowed to set the prices to be consistent with the physical operation of the transmission system. MISO has implemented transmission constraint demand curves with stepped penalty factors to approximate the value of an increment of transmission for different severities of transmission violations. The demand curve provides greater transparency to the congestion being experienced by the constraint. Flexibility to utilize constraint 30