Administrative Issues
SCPPA and GPI submit that a deliverer approach would involve a larger number of regulated entities, and that this would complicate administration of the program. SDG&E/SoCalGas and Environmental Defense state that, while there would be more points of regulation for imports, the number would not be overly burdensome. As a potential benefit, Calpine suggests that having more actors in the market may help to increase liquidity and reduce the risk of market power.
SCPPA contends that no GHG emissions tracking device is available to permit identification of GHG emissions associated with imported electricity. SDG&E/SoCalGas submit that the same type of contract information would be used to assign emissions to imported energy under either a retail provider-based approach or a deliverer-based approach, and that there is nothing that makes this undertaking more challenging under the deliverer approach, as long as the required parties report the information.
Other parties (SCE, Calpine, and Morgan Stanley) assert that a deliverer approach would be less complex administratively than a retail-provider approach because only imports would have to be tracked under a deliverer approach while under the retail provider-based model all wholesale power transactions must be tracked in order to assign emissions to retail providers.
SDG&E/SoCalGas view emissions tracking and verification associated with the deliverer approach as being relatively transparent, because most of the participants in such a program would have close ties to the generation that they are selling. It states that the use of generator data would be a significant advantage for the deliverer approach compared to the retail provider approach, which would use default emissions values for all purchases of unspecified power, including power generated in California. However, GPI points to the dependence on default factors for many imports.
3.3.1.4. In-State Generators, Retail Providers for Imports as Point of Regulation
The only party in the proceeding that advocated for this model is EPUC/CAC, which later changed its position to support the deliverer approach. EPUC/CAC cited several possible advantages to a hybrid approach. According to EPUC/CAC, a hybrid approach: 1) "best aligns the incentives to reduce emissions with the source of those emissions," 2) "allows for greater accuracy in the tracking of emissions," 3) facilitates expandability, 4) "offers administrative simplicity," and 5) "can overcome legal challenge." While EPUC/CAC acknowledged that a hybrid approach would treat out-of-state sources differently, it asserted that the program could be designed to not disadvantage them and thus mitigate susceptibility to Commerce Clause challenge. It contends that, since the hybrid approach also would not directly regulate wholesale transactions, it should also overcome Federal Power Act (FPA) challenges.
EPUC/CAC asserted that, with an in-state generator/retail provider for imports hybrid, roughly 75-80% of California's load would be captured at the source. It argued that using a retail provider approach for imports could give California greater leverage in dealing with imported emissions, and that discovery of out-of-state sources could be incentivized by attributing a high default GHG emission rate to unspecified purchases.
Several parties contend a hybrid design would have significant disadvantages. SDG&E/SoCalGas, SCE, WPTF, and Calpine submit that a major problem with the hybrid approach would be its impacts on the CAISO markets. Calpine contends that such an approach would bestow a competitive advantage on out-of-state sources since they would not have to include a carbon price in their bids into the CAISO markets. SCE argues that carbon costs would not be imposed on imports bidding into the CAISO markets, and thus that importers would receive higher prices from the CAISO market with no emissions obligation. EPUC/CAC cited such concerns in reply comments and abandoned its support of a hybrid approach in favor of a deliverer approach.
Several parties contend that a hybrid approach would be at least as administratively complex as a deliverer approach. They submit that all load would need to be tracked to sources for the system to work. SMUD, Constellation, and PG&E are also concerned that a hybrid system would require extensive accounting to avoid double counting. DRA similarly states that such a system would require all of the reporting and tracking protocols associated with a retail provider-based system to account for imports, and would require the regulatory enforcement and compliance standards for generators associated with a source-based approach.
SDG&E/SoCalGas and SCE express concern that this option is vulnerable to challenges under the FPA and the Commerce Clause.
As described in Section 3.2.2, we recommend that ARB adopt a cap-and-trade program that includes the electricity sector in California provided that ARB finds that the tests outlined in Part 4 and Part 5 of AB 32 are met. An integral component of a cap-and-trade program is the point of regulation. That is: which entities should have the compliance obligation within a cap-and-trade system for delivering GHG emissions reductions within the electricity sector?
To answer this question, we focus on what we believe are the five most important criteria. Those criteria are:
1. Environmental integrity. Here we focus on how each option deals with the problems of unspecified system purchases and imports in order to minimize the potential for leakage and/or contract shuffling, leading to real GHG emissions reductions from the electricity sector.
2. Compatibility with/expandability to potential regional and/or national GHG emissions cap-and-trade markets.
3. Accuracy and ease of reporting, tracking, and verifying GHG emissions reductions. Without accurate tracking, we cannot ensure that reductions are real, quantifiable, verifiable, and valid.
4. Compatibility with ongoing reforms of wholesale and retail energy markets. We focus, in particular, on potential interactions with the CAISO's new market design and the MRTU, while keeping in mind that some California entities are less involved with CAISO markets.
5. Legal issues.
We assume that, as a threshold matter, all options would have to be consistent with other federal, State, and local environmental requirements, such as those pertaining to criteria pollutants and toxic waste.
Below, we address each of the first four criteria in turn and discuss how each option for point of regulation does or does not meet the criteria. We stress at the outset, however, that none of the options meets all criteria fully. With any one-state cap-and-trade design in the electricity sector, there are inherent pros, cons, and tradeoffs. Our job is to weigh the pros and cons against our most important criteria. We note that there are other criteria that could be applied to this choice, as discussed in several ALJ rulings that have helped us reach this decision point and upon which parties have commented extensively. However, in this decision, we focus on those criteria that have led us to our recommendation for point of regulation in the electricity sector. We also discuss some other secondary criteria as they relate to the options under consideration.
As explained below, we conclude that the deliverer point of regulation best meets these four criteria. We then address some details regarding formulation and application of the deliverer point of regulation and consider legal issues (the fifth criterion listed above) related to the deliverer approach.
3.3.2.1. Environmental Integrity and Real GHG Emissions Reductions
In assessing the viability of the four options for point of regulation, obtaining real GHG emission reductions is our most important consideration. With any design, we must ensure that the system will deliver real reductions in GHG emissions into the atmosphere as required by AB 32. The two chief concerns here for California's electricity sector are that, while California imports approximately 20% of its electricity from neighboring states, those imports represent more than 50% of the GHG emissions from the sector and that, within California, unspecified system purchases are a substantial portion of purchases. Thus, to be effective, any system we design must address imported power and unspecified system purchases in some way. Since in Section 3.2.2 we recommend design of a cap-and-trade system for California that includes the electricity sector, we now examine how well options for cap-and-trade design address both in-state generation and imported power.
First we consider the option where in-state generators would be the point of regulation, without imports included in the cap-and-trade system. By not covering imports directly in the system, it is likely that there would be incentives for the electricity sector in California to reduce its GHG emissions by importing more power from out-of-state, without necessarily reducing emissions into the atmosphere at all. This is certainly true in the long term and likely true in the short term as well. As environmental costs begin to make in-state generation more expensive, the economic incentive to begin importing more power from uncapped out-of-state power plants would be strong. Therefore, this option appears to be the least desirable from the standpoint of environmental integrity.
The other three options (retail providers; deliverers; and a hybrid in which the point of regulation includes in-state generators and, for imports, retail providers) address imported power and unspecified in-state purchases in different ways.
With retail providers as the point of regulation, integrity of the system would be addressed by holding retail providers responsible for all of the power they deliver to consumers. In the hybrid option, retail providers would be responsible only for imported power. In order to make either of these retail provider-based systems function more accurately, it is likely that a tracking system and/or an emission attribute certificate system would need to operate parallel to the cap-and-trade system to ensure that contract shuffling is minimized under the model. It is impossible to track accurately all generation from each power plant to the retail provider that delivers it to a consumer. One option to deal with this problem is the development of default factors, but those factors are inherently inaccurate and create unintended and negative incentives for market participants. To reduce inaccuracies in retail provider-based systems, development of a mandatory emissions attribute tracking system which included imports likely would be needed.
For the deliverer point of regulation, the entity that first delivers the power to the electricity grid in California would be held responsible for its emissions. This would capture emissions from electricity generated within California and electricity imported into California from out-of-state. The problem of in-state unspecified purchases would disappear, and this would be a major advantage. The carbon attributes of imports would be determined by the entity most likely to know what has been purchased and in the best position to provide verification documentation. Because of the increased accuracies of the deliverer approach in identifying the generating source of electricity, reported GHG emission reductions would also be more accurate and reliable. As a result, we conclude that the deliverer approach is the preferable alternative regarding the ability to ensure that reported GHG emission reductions are real.
3.3.2.2. Compatibility With/Expandability to Potential Regional and/or National Cap-and-Trade Markets
We want to design a system that is likely to be compatible with any regional and/or federal cap-and-trade system that may be established within the next few years. Negotiations are underway to design a Western region cap-and-trade system through the Western Climate Initiative, and a number of proposals are currently pending in the United States Congress. Thus, it appears likely that a regional and/or federal cap-and-trade system could be established within the next few years. It also appears likely that initiation of the compliance period for a regional and/or federal system could follow California's 2012 compliance initiation by at least a few years. Thus, at some point in the near future, a California cap-and-trade system will likely need to be linked to, or adapted to be compatible with a regional or national system.
Some parties have argued that it is not necessary to worry about compatibility of the design of the cap-and-trade system with a regional or federal system, because a regional or national program would render the California system obsolete. Others have argued that certain designs of a cap-and-trade system for California could co-exist with or link to a parallel federal or regional program. Both of these things could be true; it likely depends upon the ultimate design of each system. In the face of this uncertainty, we think it would be beneficial to design a system that is most likely to be similar to a federal or regional system.
As most parties have noted, all cap-and-trade systems operating to date have been source-based systems. These include not only the European Union Emissions Trading System, but also a number of cap-and-trade systems for controlling criteria pollutants within the United States. Therefore, this is the type of cap-and-trade system with which entities in the electricity sector in California, and the rest of the country, are familiar.
In addition, if the geographic scope and coverage of the cap-and-trade system is large enough, we need not worry so much about the potential for leakage and/or contract shuffling with entities outside of the capped area. If all, or at least most, of the emissions are covered under the cap-and-trade system, accounting for imports becomes fundamentally less of a concern. Thus, the point of regulation should be designed such that tracking of imports can be reduced or eliminated as the necessity of accounting for imports diminishes.
Under this criterion, we conclude that the retail provider point of regulation would perform least well in terms of compatibility with a national or regional system. This is because, as discussed above, most existing and proposed cap-and-trade systems are source-based in nature. While it may be possible, as some parties argue, that a retail provider system for California's electricity sector could be compatible with a national or regional source-based system, it is also likely that the retail provider point of regulation would produce a need for certain contractual and operational changes to the way electricity transactions currently take place. We prefer not to introduce this kind of shift into the electricity sector for what may be a period of only a few years (if a national or regional system supersedes our efforts here). We believe that it would be less complex and more effective to be able to integrate the cap-and-trade method chosen for the California electricity sector into an eventual regional or national system, rather than having it structured such that it could operate only in parallel to, rather than as an integrated component of, a broader system. Therefore, the retail provider point of regulation is the least preferred under this criterion.
All of the other three options (in-state generators with no imports under cap-and-trade, deliverer, and in-state generators/retail providers for imports) share a common component where, for most electricity sold in the state, the point of regulation would be at the generator level. Therefore, any of these three options could be integrated more easily into a regional or national system.
The in-state generator option for point of regulation with no inclusion of imports under cap-and-trade is likely the most forward-compatible of the options. This approach would transition more easily into a larger geographic cap-and-trade system. But, as we mentioned above, it is the least favorable alternative for environmental integrity.
The option of in-state generators with retail providers as the point of regulation for imports is likely second-best, since the retail provider portion likely could be easily abandoned at such time as the states from which California imports become covered under a regional or national cap-and-trade system. However, the deliverer point of regulation could be modified to eliminate its coverage of imports, though the process may not be as simple, as some regulations or designs may need to be modified once imports are captured under a regional or national system.
3.3.2.3. Accuracy and Ease of Reporting, Tracking, and Verifying Emissions Reductions
We want to design a system where emissions can be accurately and reliably reported, tracked, and verified. Using this criterion, the option of in-state generators only, with imports not covered under cap-and-trade, is the simplest among the choices. But, as we have discussed, this option has serious flaws under our most important criterion of environmental integrity.
The retail provider point of regulation is a less preferable alternative under this criterion. In order to make the retail provider option function accurately, it is necessary to track all electricity generated to serve California customers from the point where it is generated to the point where it is delivered to a retail provider's customers. While this may be relatively easy in the case of in-state generation owned by a utility company that uses the power for its customers, it becomes most difficult in the case of purchases from unspecified power plants. The best way to make such a system work would be to undertake a West-wide tracking system for emissions attributes, perhaps with tradable aspects similar to RECs for renewable energy, where the attributes are tradable separately from the commodity electricity. While such an option would be theoretically workable, in our judgment the administrative complexity and time required to set up such a system render this among the less preferable alternatives.
Similarly, the in-state generator/retail provider for imports option is also administratively complex. In order to make such a system work and hold retail providers responsible only for their imported power, their entire electricity portfolio would need to be tracked, with the in-state generation portion netted out to determine the portion of the portfolio attributable to imports. Thus, all of the tracking or attribution necessary under the retail provider point of regulation would also be necessary under this alternative, with an added layer of complexity to conduct the proper accounting to subtract out in-state generation. Thus, we also find the in-state generator/retail provider for imports point of regulation option to be less preferred under this criterion.
We conclude that the deliverer point of regulation is the most workable. This is because each deliverer is responsible for reporting and tracking the GHG attributes of its power as it is delivered onto the California grid. For in-state generation, generators (or other entities that own the power when it is delivered to the grid) are tracked, similar to a system in which only in-state generation is capped. Similarly for imports, the party that owns the power as it is delivered to the California grid is held accountable. This removes the need for complete tracking from generation source to delivery to customers, as under the retail provider system, and also removes the need for complex netting of in-state generation from the retail provider portfolios, as under the in-state generator/retail provider for imports system. Making the deliverer the point of regulation moves the compliance obligation as close as possible to the generation source, which increases the accuracy of knowledge of GHG emissions attributes of the generation sources. Therefore, we find the deliverer point of regulation to be the preferred option for accurate reporting, tracking, and verification of emissions in the sector.
3.3.2.4. Compatibility with Wholesale and Retail Energy Market Reforms
In discussing this criterion, we begin by noting that our purpose in designing a cap-and-trade system for the electricity sector in California is fundamentally to reduce GHG emissions from the sector, including all electricity consumed in California as well as all electricity produced in the State regardless of where it is used. In doing so, we do not wish to interfere with the functioning of the wholesale market for electricity in the State. Instead, we aim to produce the environmental result required under AB 32 with the least impact possible on wholesale electricity markets. We recognize, however, that the cap-and-trade market is likely to cause the price of some electricity sold through the wholesale market to rise. To the extent that happens, our goal is to have that price effect be transparent to and consistent for all participants, without any distortionary impact.
In addition, in order to meet not only the 2020 goals under AB 32, but also the more aggressive 2050 goal of reducing GHG emissions 80% below 1990 levels, as established by Governor Schwarzenegger,16 we will need to focus much more on the kind of electricity infrastructure built to serve California consumers, and not simply the type of generation dispatched in the wholesale markets.
In California, the main centralized wholesale electricity market is operated by the CAISO. The CAISO has undertaken a multi-year effort to redesign its electricity markets under the MRTU process. Its new market design is due to go into operation this year. This market redesign comes under the jurisdiction of the Federal Energy Regulatory Commission (FERC). The MRTU process will lead to both a day-ahead and a real-time energy market, and one goal of the market redesign is to encourage the scheduling of more power through these markets, which will enhance efficiency of dispatch, increase market liquidity, and provide more operational flexibility to the CAISO to balance the system.
With these facts in mind, we have evaluated the point of regulation options against their likely impacts not only on the CAISO's MRTU markets, but also on California's energy markets in general.
We first examine the in-state generator option with no imports in the cap-and-trade system. Under this system, we expect that in-state generators would reflect the increased cost of compliance with the AB 32 regulations in their wholesale bid prices. However, because out-of-state resources would not face any cap-and-trade compliance costs, their costs would not change. They could nevertheless raise their bid prices by an amount slightly less than the allowance price, capturing a rent from California consumers while still being dispatched ahead of in-state resources. The result would be a less efficient use of both generation and transmission infrastructure coupled with a wealth transfer from California consumers to out-of-state generators for no environmental gain.
This situation is a major disadvantage of this option for point of regulation, because it would cause the price of in-state electricity generation to rise in California, while imports on which California relies would see a smaller impact, if any. This is true regardless of emission allowance allocation policy, because the in-state generation price would reflect either the actual cost of the emissions allowances or their opportunity cost. Out-of-state generators, which would not face the compliance cost of the GHG regulation, would be able to sell their power at a relatively lower price than in-state generators. Thus, the system would produce leakage, violating the environmental integrity principle outlined above. For that reason, as we state above, we do not prefer this alternative.
The deliverer and the in-state generator/retail provider for imports point of regulation options are similar in terms of their impacts on wholesale markets. In either case, a compliance obligation would be placed on some entity for all power that has emissions associated with it. All power generated in California or consumed by California customers would reflect the cost of compliance with the cap-and-trade program.
There is still some risk of distortion in the MRTU markets with the in-state generator/retail provider for imports hybrid option, because some low-emissions imported power may face incentives to self-schedule in order to identify its low-emission characteristics to the entity responsible for the emissions for this imported power. This is because low-emissions power offers additional value to retail providers by reducing the number of allowances that need to be retired. This is value that low-emissions out-of-state generators can partially capture when their power is sold on a specified basis. Moreover, this approach may also induce leakage through the CAISO markets as out-of-state generators, with no compliance obligation, could bid into the markets at a lower price than in-state generators. California retail providers who purchase from the CAISO markets could only be held responsible for allowances for imports after bids have cleared.
The magnitude of potential MRTU market distortion may be relatively small under this option, however, because imports represent only about 20% of the power sold in California, and low-emissions imports represent an even smaller percentage. Under either of these point of regulation options, the approximately 80% of generation produced and sold in California through the markets would have no incentive to self-schedule to identify their emissions characteristics, because they would be responsible for their own compliance with the program. The deliverer approach avoids these perverse impacts on the CAISO markets since out-of-state entities delivering power to CAISO would also have to factor allowance costs into their bids.
Finally, we consider the potential for a retail provider point of regulation system to interfere with the functioning of the wholesale electricity markets in California. It is likely that the risk of interference with markets is the greatest with this option. This is because the incentive that would induce clean imports to sell to California on a specified basis under the in-state generator/retail provider hybrid approach would apply to in-state sources as well under a retail provider point of regulation. This would reduce the flexibility of the CAISO to schedule resources based on economic and/or operational considerations, instead forcing it to dispatch units that are self-scheduled due to relatively low GHG emissions characteristics. Thus, wholesale prices from low-emission generation would rise, and further costs would likely result from the higher transaction costs of negotiating specified purchases and the foregone efficiencies of the pooled CAISO markets.
Therefore, we conclude that the retail provider point of regulation has the most potential to interfere with the functioning of the wholesale markets.
In addition to the wholesale market reforms being undertaken by the CAISO, the Public Utilities Commission is currently exploring the possibility of restoring retail competition in R.07-05-025. In this decision, we take no position on whether, when, or how direct access should be reopened in California. We note, however, that reopening direct access could result in increased market share for retail providers that rely heavily on market purchases of energy to meet their customers' needs. Decreases in long-term contracts, which are chiefly entered into by vertically integrated utilities, would likely increase the amount of unspecified energy flowing through California's markets. The deliverer point of regulation can best accommodate such a development as it does not require source-to-sink tracking of all transactions.
3.3.2.5. Conclusions Regarding Compatibility with the First Four Criteria
In evaluating the point of regulation options against our key criteria above, we conclude that the deliverer point of regulation best meets the four criteria evaluated above. Each of the other options has serious shortcomings regarding one or more of our priorities.
A deliverer point of regulation would provide for the environmental integrity of the cap-and-trade system by covering imported power as well as in-state generation. The deliverer point of regulation shares a number of common characteristics with a source-based point of regulation, making it likely to be compatible with the eventual design of a cap-and-trade system that is broader in geographic scope (regional and/or national). The deliverer point of regulation also would improve the ability to report and track emissions in the sector and would minimize the impact of AB 32 GHG regulations on California's wholesale electricity markets.
Having determined that the deliverer point of regulation best meets the four criteria examined above, we turn to certain details regarding the manner in which compliance requirements should be determined in a cap-and-trade system with a deliverer point of regulation for the electricity sector.17
We conclude that the most useful formulation of the deliverer point of regulation approach is that the point of regulation would be the entity that owns electricity as it is delivered to the grid in California. In most situations, this would be the entity that owns the electricity on the portion of the physical path just before the point where it is delivered to the California transmission grid, which would be the busbar for in-state generation or the first Point of Delivery in California for imported power.18 Where electricity is first delivered to the California grid at the distribution level, the deliverer definition results in the following: (i) for generation facilities that are connected to a retail provider's distribution network, the deliverer would be the entity that owns the electricity as it is delivered to the distribution network, and (ii) for electricity delivered directly to California retail customers of a multi-jurisdictional utility from out-of-state sources, the deliverer would be the multi-jurisdictional utility.19 Recognizing that electricity is an instantaneous commodity, we call the entity that owns the electricity as it is delivered to the California grid the "deliverer" of the electricity for purposes of establishing GHG responsibility. We recommend that deliverers be required to surrender allowances associated with the electricity's GHG emissions.
Deliverers would include generators, operators, retail providers, marketers, and any other types of entities that own electricity as it is delivered to the California grid. While the deliverer often may be the owner or operator of the generating unit, it could also be any entity that purchases or otherwise has a contractual arrangement such that it owns the electricity as it is delivered to the California grid.
The proposed decision and parties' comments on the proposed decision addressed several possible exceptions to our determination of the manner in which deliverers should be identified for the purpose of GHG compliance obligations. We address these proposed exceptions in turn.
16 See Governor Schwarzenegger, Executive Order S-3-05, June 2005.
17 As explained in Section 3, electricity that is wheeled through California is not included in the electricity sector for purposes of establishing GHG regulations pursuant to AB 32. As explained in Section 4.2.2, we defer the issue of whether electricity generated by CHP facilities should be included in the electricity sector.
18 In this situation, the deliverer would be the owner that delivers the electricity to the first Point of Delivery in California, not an entity that accepts ownership of the electricity for the first time at that Point of Delivery.
19 We understand that the multi-jurisdictional utilities generate or purchase electricity out-of-state and that the electricity is delivered at the distribution level directly from out-of-state to their California retail customers.
