California Independent System Operator DRAFT – Work in Progress

 

California Independent System Operator

Market Design 2002 Project
Revised Draft Comprehensive Design Proposal and Project Time Line

For stakeholder review prior to presentation to the CAISO Board of Governors at
their February 7, 2002 meeting

IMPORTANT NOTE: To facilitate stakeholder comments on this draft, the deadline for submitting written comments has been extended to 9 A. M. Thursday, January 31, 2002. Please submit all written comments to Byron Woertz via e-mail ( bwoertz@caiso.com ).

About this Document

This document presents the CAISO’s proposed revisions to the MD02 design process and time line and to the major design elements as originally described in the January 8 Preliminary Draft Comprehensive Design Proposal. The revisions reflect the CAISO’s responses to the extensive constructive input received from stakeholders during the January 14- 17 focus groups and in written comments, as well as the outcomes of the MD02 design team’s continuing design efforts since the January 8 draft was released. The material in this document will be presented to the CAISO Board of Governors at their February 7, 2002 meeting, along with material from the January 8 draft that is unaltered by the revisions discussed below. To keep this document brief, repetition of material in the January 8 draft is kept to a minimum. The reader should therefore refer to that document for additional background and details on the MD02 project and the principal design elements.

The next section discusses the CAISO’s proposed changes to the original MD02 process and time line as stated in the January 8 draft. The subsequent section indicates areas where the CAISO is proposing revisions to or additional design details of specific design elements.

Changes to Original Process and Time Line

Readers should note the following proposed changes to the MD02 process and time line as originally stated in the Preliminary Draft of January 8, 2002:

1. CAISO Management will not be asking the Board of Governors to approve the Comprehensive Design Proposal at the February 7 meeting. Instead, Management will present a detailed status report on the MD02 project for the Board’s discussion and guidance, including 1) the scope and objectives of the project, 2) a description of the principal design elements, 3) a summary of the input received from stakeholders, 4) some of the high- level design issues remaining to be addressed, and 5) the proposed process and time line for completing the major deliverables of the project.
    
2. The CAISO still intends to make two FERC filings on MD02 by the end of March and by May 1, as originally stated, and the content of these will be presented to the Board for approval at the March 14 and April 25 meetings, respectively. However, the CAISO now proposes to modify the content of the two filings somewhat. Specifically, the CAISO does not intend to include the proposed Available Capacity Obligation, which has proven to be the most difficult element to design, in the first FERC filing. Instead the CAISO will take additional time as needed to design an optimally effective solution to the problem the Available Capacity Obligation is intended to address, namely, the problem of ensuring adequate supply capacity to meet California’s electricity needs at reasonable cost.
    
3. Since the Available Capacity Obligation will not be part of the first FERC filing, at this point it is not clear whether an ACAP Obligation can be implemented by September 30, 2002. The MD02 team will continue to work to develop the design and implementation details, and will assess the feasibility of September 30 implementation at a later date. At the same time, the team will develop alternative approaches that could be implemented in the absence of an ACAP Obligation to ensure supply adequacy for reliable grid operation and to prevent the exercise of market power when the current FERC mitigation measures expire.
    
4. In light of these changes in the MD02 process and time line, the CAISO will shortly develop and release a new schedule of activities and proposed milestones of the MD02 project, including opportunities for continuing stakeholder input to the process.

Revisions and Additional Details of Major Design Elements

This section is structured in two parts. Part 1 indicates proposed revisions and additional details for all elements of the proposed Comprehensive Market Design. Part 2 focuses specifically on elements needed to be in place by September 30, 2002, when the FERC mitigation measures expire. Given the decision not to ask the CAISO Board for a decision on the MD02 proposal at the February 7 meeting, the MD02 team has not tried to address all comments in the present document. Comments we have received on the January 8 draft and which are not addressed in the present document will be addressed in accordance with a revised process and time line to be released shortly.

Part 1. Elements of the proposed Comprehensive Market Design

The proposed Comprehensive Market Design is comprised of the following elements, each of which is discussed below:

1. Available Capacity Obligation on Load Serving Entities
2. Forward Congestion Management
3. Forward Spot Energy Market
4. Changes to Ancillary Services Markets
5. Residual Day Ahead Unit Commitment
6. Firm Transmission Rights
7. Changes to Structure and Timing of Hour Ahead Market
8. Real- time Economic Dispatch Using Full Network Model
9. Bid Mitigation for Local Reliability Needs
10. Damage Control Price Cap on CAISO Markets.

Available Capacity (ACAP) Obligation on Load Serving Entities

This section contains new material.

The original motivation behind the proposed ACAP obligation has not changed. The main purpose of the ACAP obligation is to ensure that adequate capacity is committed on a daily basis to meet system load and reserve requirements and is available to respond to CAISO dispatch instructions to meet system imbalances and local reliability needs while minimizing the potential for suppliers to exercise market power. Under the original design of the California restructuring there was no entity with explicit responsibility to ensure adequate capacity. As a result the spot markets of the Power Exchange and the CAISO were vulnerable to market power exercise, and the CAISO frequently faced supply shortages right up to the operating hour. To remedy this problem the proposed ACAP obligation would apply to all Load Serving Entities 1 ( LSEs), thus placing the responsibility on them to procure adequate capacity to meet their expected peak monthly loads plus reserve requirements.

Following substantial discussion of the design of the ACAP obligation, both within the CAISO and with stakeholders at the focus groups, the team now recommends that the proposed ACAP Obligation be removed from the Phase 1 filing time line late March target) to allow additional
time to develop an optimal design that addresses the underlying problem. Therefore the present document does not provide additional detail on the ACAP Obligation. Rather, readers should consider the design details presented in the January 8 draft as now open to further discussion
and revision.

Forward Congestion Management (CM)

This section contains new material.

The present document does not modify the design of forward congestion management as proposed in the January 8 draft. To summarize, the MD02 team is proposing that forward CM should adjust generation, load, import and export schedules to clear congestion using a DC optimal power flow algorithm (DCOPF) and a Full Network Model (FNM) that includes all busses and intertie points within the CAISO- controlled grid, plus an external network representing the rest of the WSCC system to capture external loop flows. The proposed CM approach ensures that final schedules are feasible with respect to all transmission constraints as well as generator ramping and other performance constraints, 2 and renders the current distinction between interzonal and intra- zonal congestion irrelevant.
 

1 The term "load- serving entity" or LSE refers to any entity that provides electric energy to end-use consumers. While there are some non-utility electric service providers (ESPs) that serve end-use consumers under the direct access provisions of the California restructuring, the largest LSEs in California are the three Utility Distribution Companies (UDCs). The UDCs are the distribution system operators in their respective service territories as well as the default electric service providers for consumers who have not chosen a non- UDC direct access provider. The proposed ACAP obligation would of course apply to all LSEs. Moreover, since the CAISO transacts directly only with Scheduling Coordinators (SCs), the ACAP obligation would be applied through the SCs who schedule for LSEs.

2 The proposed concept of feasibility does not, however, require that final schedules reflect actual levels of load and generation expected in real time. Any shortfall between final schedules and the CAISO’s load forecast is addressed by the CAISO’s day- ahead residual unit commitment, discussed below.

Under the proposed CM approach SCs will submit Energy/ Adjustment Bids on their preferred schedules. Due to the complexity of the network model used for CM, the CAISO will need to use a mix of bids from across various SC schedules to clear congestion. As a result, balanced schedules for each individual SC become an option rather than a requirement. SCs who want to preserve physical bilateral contracts can submit high adjustment bids within the limits of any applicable bid cap) or no bids at all, thus becoming price takers for congestion charges, and hedge their congestion risks with FTRs. In addition, under this approach SCs will be able to submit demand bids unmatched by supply, or supply bids unmatched by loads, and the simultaneous energy/ CM algorithm will execute all economic trades and clear the market in a manner that respects transmission and generator performance constraints. Thus bid- based CM using a complex network model ultimately undermines the rationale for a balanced schedule requirement, while still accommodating physical bilateral scheduling for those market participants who wish to schedule in this manner.

Another significant impact of the proposed CM approach is its implications for forward energy trading. A separate day- ahead energy market along the lines of the former Power Exchange ( PX) may be a viable option in its own right regardless of the CAISO’s CM approach. At the same time, the proposed CM approach will effectively create a day- ahead energy market that runs simultaneously with CM. The MD02 team therefore does not see a need at this time for the CAISO to create a separate PX- type day- ahead energy market.

Using the FNM for CM does not mean that we must use 3000 busses for all scheduling and settlement purposes. Rather, it should be possible to aggregate busses to create "trading hubs" to facilitate energy trading, and "demand zones" to simplify load scheduling and settlement. At the time this document is being drafted, the MD02 team has not yet developed a more specific proposal for how to define trading hubs or demand zones, so we can not provide additional detail at this time. We recognize, however, that there are numerous questions about how trading hubs and demand zones in the CAISO system would work, particularly with regard to scheduling and settlement, and therefore this topic will be a high priority for additional development in coming weeks.

The proposed forward CM approach also raises a need for bid mitigation for locational needs, to prevent the exercise of local market power in areas of the grid or across transmission pathways where there is not a competitive supply of bids.

Finally, several parties commenting on the January 8 draft urged the CAISO to consider a fully integrated day- ahead market that includes ancillary service procurement and unit commitment service in addition to congestion management and energy trades. Software that performs this type of simultaneous optimization is well established and in use today by other ISOs. The issue of integrated A/ S is discussed below. Regarding unit commitment, the team believes it would be beneficial to offer a voluntary unit commitment service, which would optimally commit resources for those SCs who wish to participate while allowing others who wish to make their own unit commitment decisions to opt out of the CAISO service. This type of unit commitment would not eliminate the need for the Residual Unit Commitment discussed below.

Forward Spot Energy Market

This section does not contain new material.

As discussed in the section on Forward Congestion Management, the proposed approach necessitates a certain level of energy trading between SCs through the need to clear congestion using submitted bids. It also enables day- ahead energy trading by allowing market participants to submit unbalanced demand and supply bids, which the CAISO’s CM procedure would then clear in the most efficient manner subject to transmission constraints. Thus a forward spot energy market is implicit in the redesign of CM. At this time the team does not see that there would be a need for the CAISO to create an additional day- ahead energy market outside of CM.

Changes to the Ancillary Services Markets

This section contains new material.

The most significant design question is whether the CAISO should maintain today’s sequential A/ S markets, which run after the day- ahead congestion management procedure concludes, or should integrate A/ S procurement into the day- ahead CM procedure. The integrated approach is advantageous from the viewpoint of system efficiency, since it optimally allocates all available generating capacity to provide energy or unloaded reserve capacity. The sequential approach also offers some advantages, however, the principal ones being (1) a natural way to allocate transmission capacity to A/ S by associating A/ S capacity imports with intertie transmission that was not scheduled in day ahead CM, and (2) potential cost reduction by performing actual cost minimization as is done today through the Rational Buyer algorithm, rather than the bid cost minimization used in simultaneous optimization which ignores the fact that all bidders are paid the MCP rather than their bids. The team does not yet have a proposal on this question.

Residual Day- ahead Unit Commitment

This section does not contain new material.

The Unit Commitment envisioned in the MD02 effort is called "residual" because its intent is to supplement the final day- ahead scheduling process in which resources were committed either by market participants or in the CAISO’s integrated procedure. After day- ahead schedules are finalized, the CAISO would assess the adequacy of committed resources and scheduled firm imports to meet the CAISO’s forecast of next day’s system load and local reliability needs. The CAISO would then commit such additional resources whose start- up times require day- ahead
notice to ensure that enough capacity is on- line to meet the next day’s forecasted needs at both the system and local level. Resources so committed by the CAISO would be guaranteed recovery of their start- up and minimum load costs.

Firm Transmission Rights (FTRs)

This section contains new material.

The present document does not propose any revisions to the FTR proposal contained in the January 8 draft. Readers should consult that draft for the design details. We note, in response to numerous comments, that the proposed new designs for congestion management and FTRs would be slated for Spring 2003, probably with a target date of April 1 so as not to impact the FTRs recently auctioned for the April 1, 2002 to March 31, 2003 period.

The MD02 team can offer its initial views on some of the questions posed in the previous draft.
Specifically:

• Whether to adopt a "use it or lose it" model for FTRs similar to the model being considered by RTO West. In that approach, FTRs would generate revenue only for parties who actually schedule on the system, and would not have value for an investor who wants to buy the FTRs to earn the associated stream of congestion revenues. The way this is achieved is by allowing payment of congestion revenues only to offset the congestion charges a party actually incurs from scheduling on the system. The CAISO’s design team does not at this time favor this approach. Rather, we would propose to retain a model whereby the financial costs and benefits of FTRs accrue to the FTR holder regardless of the holder’s scheduling behavior.
   
• Whether the new FTRs would be obligations, which impose a cost on the holder when congestion is in the opposite direction of the right, or options only which impose no cost when congestion is in the opposite direction. The team recognizes that the obligation model is unfamiliar in California, and that numerous parties have expressed preference for the option model. At this point, however, the team is concerned that (1) the methodology for running a simultaneous feasibility assessment of option- only FTRs has yet to be fully developed, much less tested and utilized by a functioning ISO, and (2) the obligation model allows larger quantities of FTRs to be released and offer more complete risk- management capabilities.

Finally, we note that an extremely important issue to be addressed in the coming weeks is how to allocate FTRs initially, i. e., whether to provide an initial set of FTRs to load- serving entities or to rely completely on an open auction for the initial allocation. The team does not yet have a proposal on this issue, but expects to initiate a full assessment of the alternatives shortly.

Changes to the Structure and Timing of the Hour- ahead Market

This section contains new material.

Numerous parties have expressed a need to more the hour- ahead market closer to real time. While many details remain to we worked out, the MD02 team is considering a simplified HA market design that would perform only congestion management and energy trading, and would close to market participant submissions perhaps as late as 60 minutes before the start of the operating hour. For this to be workable for the CAISO, however, it may be necessary to stop accepting Supplemental Energy bids at the same time. This would allow CAISO operators to run CM and energy, and then to issue pre- dispatch instructions as needed from among the energy bids that were not accepted in the CM/ energy procedure.

Allowing a window for energy trading as late as 60 minutes before the start of the operating hour will provide an opportunity for resources to be dispatched for hourly periods, with an hourly price commitment and timing that is near real time. (Resources with longer start- up times will have had the opportunity to be matched with loads in the day- ahead market.) The MD02 team believes that the trading opportunity created by the revised HA time line should satisfy the need that has been expressed by inflexible resources (i. e., those with operating characteristics that are not well- suited to 10- minute real- time dispatch) for a 60- minute dispatch market.

Real- time Economic Dispatch using Full Network Model

This section does not contain new material.

This element was fully discussed in the context of the CAISO’s CMR project and was described in the CMR proposal issued in January 2001. Consistent with the CMR recommendation the MD02 team proposes a security- constrained economic dispatch for the real time market, to fully
take into account all transmission constraints, local reliability needs, loop flows, generator operating constraints, and imbalance energy needs. This approach would produce nodal realtime energy prices, which would be paid to supply resources but could be aggregated for settling load deviations. This approach is a logical extension of the real- time market pricing changes being introduced in the CAISO’s filing of the permanent Target Price fix.

Bid Mitigation for Local Reliability Needs

This section contains new material.

Under today’s zonal CM system, real- time bid mitigation becomes necessary when the merit order sequence of dispatching real- time energy to meet system imbalances fails to resolve intrazonal congestion or satisfy a local reliability need. In these cases the CAISO has to take bids out- of- sequence (OOS) to address the local problem. Suppliers who are aware of their strategic locations in such situations are said to have locational market power, and often exploit this by bidding extremely high to increment their output (INC), or extremely high in the negative direction to decrement their output (DEC), knowing that the CAISO must dispatch them. To limit exploitation of locational market power we expect to propose specific mitigation measures similar to those approved by FERC for other ISOs. Such measures are appropriate regardless of the granularity of forward congestion management or real- time locational pricing, i. e., under both full nodal pricing and a more aggregated approach. In addition, once the proposed new forward CM procedure is put in place, it will be necessary to implement bid mitigation in the forward market as well, since locational market power can be exercised by resources needed for local reliability to ensure the feasibility of forward schedules. The CAISO does not have a more detailed proposal on this element prepared at this time, but does intend to model this element on successful designs used by other ISOs.

Damage- control Price Cap on CAISO Markets

This section does not contain new material.

Without the price mitigation provided by the FERC market mitigation orders, the spot markets will be vulnerable to occasional extreme peak prices. All other ISOs have some level of damage- control price cap to limit the adverse cost impacts of an unusually severe price spike. The MD02 team proposes that such a cap be established for the CAISO markets, but has not yet developed a recommendation on the level of this cap.

Part 2. Elements needed by September 30, 2002, when the FERC mitigation measures expire

To date the FERC mitigation provisions have been helpful in limiting the exercise of market power; specifically, the must- offer obligation has targeted physical withholding while the price cap in non- emergency hours has targeted economic withholding. Thus the objective of the MD02 team is to identify those elements that would be most effective in ensuring adequate supply at competitive prices after September 30. These elements would then comprise the Phase 1 FERC filing mentioned earlier in this document.

To place these critical elements in context, the reader should keep in mind that the CAISO Board has already approved the filing of Real Time Market Pricing changes (i. e., the permanent " Target Price" fix) and Interim Intra- zonal Congestion Management (including real- time bid
mitigation for local reliability needs). We assume these elements will be in operation before September and will complement the additional elements identified below.

The January 8 draft indicated the following elements for implementation by September 30:

1. Interim CAISO Available Capacity (ACAP) procurement that will transition to fully effective ACAP obligation on load- serving entities (LSEs),
    
2. Capability for day- ahead energy trades to shape next- day supply needs (e. g., the dayahead energy market as directed in the December 19 FERC order),
    
3. CAISO day- ahead residual unit commitment, and
    
4. Damage control price cap on CAISO markets.

Of the elements above, items 3 and 4 are part of the Comprehensive Design discussed in the previous section. The remainder of this section discusses only items 1 and 2.

Interim CAISO ACAP Procurement

This section contains new material.

As noted at the beginning of this document, due to the complexity of the policy and design issues involved, the MD02 team is no longer confident that it is feasible to have the ACAP Obligation adequately specified in a time frame that would allow September implementation, even in an interim manner as proposed in the January 8 draft. The team will continue to work through the ACAP issues, and will reassess the feasibility of September implementation at a later date. At the same time the team will also consider the alternative design elements that would be needed to be in place upon expiration of the FERC mitigation measures currently in effect, should it not be possible to have a workable ACAP design by that time. The team does not have an alternative proposal developed at the time this document is being prepared.

Day Ahead Energy Market

This section contains new material.

FERC’s December 19, 2001 order directed the CAISO to file a proposal for a day- ahead energy market by May 1. At the focus groups we discussed the possibility of relaxing the market separation rule along with the other elements that would take effect on September 30, 2002, as a way of enabling day- ahead energy trades in the near term. Participants in the focus groups were largely supportive of this idea, and initial assessment internally indicates that it would be relatively straightforward to implement. Essentially, the adjustment bids submitted with dayahead schedules (which would not need to be balanced) would serve also as energy bids, and would be optimally cleared to eliminate congestion while creating an hourly day- ahead energy price in each of the CAISO’s three zones. The same provisions could be implemented in the hour- ahead market as well. Based on the initial positive response to this idea the MD02 team is continuing to assess the pros and cons while proceeding to develop additional design details.