ONTARIO NEWS

By AJ Goulding, London Economics International

 

1. Introduction

          1.1 Summary of findings

The Ontario Energy Association (OEA) issued a Discussion Paper on September 13^th, 2018 (hereafter, the “Discussion Paper”) entitled “Policy Case: Recommendations for an Ontario Load-Serving Entity Model.” While a load-serving entity (LSE) model can be effective, it may not work as intended if adopted as proposed by OEA. The OEA proposal could undermine the Independent Electricity System Operator’s  Market Renewal Process, and create further policy uncertainty in Ontario. Among the challenges with the OEA’s proposal are:

• its voluntary nature would result in a patchwork of arrangements across Ontario;

• it unnecessarily limits those who could serve as LSEs;

• it increases the burdens on the OEB and stakeholders;

• it undermines the proposed IESO capacity mechanism;

• it may not in fact reallocate risk; and

• it could be a distraction to local distribution company (“LDC”) management during a period of significant electricity market evolution.

          At a time when the focus of policymakers is increasingly on cost containment, it is unclear how the proposed LSE structure would benefit the sector or consumers.

 

          1.2 History of LSE discussion in Ontario

There has been little action around the development of an LSE model over the past decade, and the LSE model has not been a part of the foundation for the current IESO market renewal initiative.

AJ Goulding, London Economics International The LSE topic has been explored repeatedly by provincial government entities, the Ontario Energy Board (“OEB”), the IESO, and the former Ontario Power Authority (“OPA”) since before market deregulation in 2002. Conceptually, an LSE model would have resulted in a more decentralized decision-making process, where procurements were made at the individual LSE level rather than by a central planner. This was notably discussed in the Electricity Conservation and Supply Task Force report from 2004, where a central procurement agency was envisioned to meet procurement needs in the short term, but with LSEs playing an increasing role in procurement of new supply over time, leaving the central planner as the “default supplier of last resort”.[1] The OPA had also explored the potential for local distribution companies (“LDCs”) to serve as LSEs in stakeholder consultations from 2004 and through a pilot project in 2007.[2] Some key events surrounding these exploratory ventures are summarized in Figure 1.

 

          1.3 What OEA proposes

The OEA defines an LSE as “an entity that is responsible for securing electricity resources to meet the supply needs of the customers it serves.”[3] The OEA identifies supply needs as being inclusive of resource adequacy, and notes that the LSE could meet these needs through a variety of ways, including generation acquisition, contracting through power purchase agreements (“PPAs”), pursuing conservation and demand management (“CDM”), or acquiring or contracting with distributed energy resources (“DERs”). The OEA envisions this regime being based on a new OEB LSE license, which would obligate the licensee to follow a new LSE code.

          The LSE regime proposed by OEA would be voluntary, but limited to Ontario electricity LDCs. LDCs choosing to become LSEs would be required to submit integrated resource plans (“IRPs”) focused on incremental needs for their service territories – presumably in this case “incremental” to the resources already procured province-wide that are incorporated into the Global Adjustment. The LSEs will be required to create procurement plans that flow from their IRPs.

          Implicitly, the OEA is imagining that LDCs will become “active” LSEs, engaging in direct long term planning and direct contracting for resources. This is in contrast to the “passive” LSEs that OEA cites as examples; entities such as ConEd in New York, Eversource in New England, and PSEG in New Jersey engage in periodic auctions for future needs on full requirements basis (i.e., suppliers take volume risk) and pass through costs to consumers.[4]

 

2. Flaws in the proposal

As currently described, we do not believe that the OEA’s proposal is implementable. For the plan to be feasible, significant capacity building would be required at even the largest of Ontario LDCs. The plan sits uneasily within the current (albeit relatively dormant) competitive market arrangements in the province. Furthermore, it does not address how the concept of incremental needs would be defined in a manner that assured that the costs of existing generation obligations would continue to be appropriately shared across the province.

          Below, we highlight a set of additional critiques of the proposal that would also need to be addressed.

 

          2.1 Voluntary nature

The voluntary nature of the proposal is troubling. Presumably, if an LDC chooses not to become an LSE, than the responsibility to procure for that service territory would remain with the IESO. While the proposal opens the possibility of smaller LDCs banding together to create a joint LSE, it does not require it. There is potential for moral hazard here – those LDCs who think that self-procurement will benefit them will pursue it, while those that do not, won’t, but the impact may be to effectively shift shared costs from LSE customers to non-LSE customers while reducing the economies of scale that arise from provincial-wide procurement.

          In addition, if it is voluntary for an LDC to become an LSE, it is presumably voluntary for that LDC to remain an LSE. This means that over time LDCs could elect to cease being an LSE, then become one, and then stop again, creating volatility in the overall market framework. While the LSE code would presumably require there to be a significant amount of advance notice for an LDC to cease being an LSE, the voluntary nature of the responsibilities will make IESO planning more difficult.

          A more conventional approach would be to require all LDCs to become LSEs, even if some (presumable smaller) LDCs band together to do so. Each would be required to procure for its needs, and would be required to participate in the capacity mechanism by either demonstrating that their forward needs have been fully met through bilateral contracts, or purchasing needs through the capacity auction. This approach would more clearly delineate roles and responsibilities in the Ontario market, though it would continue to raise issues of transparency and liquidity in the capacity mechanism.

 

          2.2 Inability for large customers or retailers to serve as LSEs

The OEA appears to rule out the possibility of retailers and individual customers becoming LSEs.[5] The paper argues that retailers should not be LSEs because their relative share of the market has been shrinking since 2002, and that large Ontario customers are not as “sophisticated” as those in places like California. The first assertion is not relevant, and the second, untrue (and contradicted by discussion elsewhere[6] regarding large customer capabilities). The market share of retailers in general has no bearing on whether a specific retailer would be able to take on the role of an LSE. Indeed, giving retailers the option to become LSEs could result in a rejuvenation of the retail sector, provide competition for LDCs in the LSE role, and enable greater choice for customers regarding supply portfolios. On the customer side, the paper presents no evidence of lack of sophistication for large customers – indeed, both retailers and large customers in Ontario today have a greater understanding of the wholesale market than LDCs do. Regardless, a lack of sophistication today is not evidence that such skills and knowledge cannot be gained when needed; if LDCs can acquire these capabilities, so can retailers and large customers.

          By preventing retailers and large customers from becoming LSEs, the OEA proposal creates a protected market for LSEs. This is likely to reduce creativity and increase costs should the LSE model be adapted. If the LSE proposal is truly about enhancing the “buy side” of the market[7], then having more entities that are eligible to be LSEs can only help achieve this objective. Instead of limiting the entities that can become LSEs, the suggested LSE Code should establish minimum standards for the role, and allow any entity that meets these standards to become an LSE. In addition to retailers and customers, this could include gas LDCs, large generators such as Ontario Power Generation (“OPG”), or any other entity that meets the requirements.

 

          2.3 Additional layer of planning and procurement processes

Ontario currently has a centralized planning regime focused on long term plans at the wholesale level. The Discussion Paper does not envision this disappearing, and indeed the IESO, in order to implement a functioning capacity mechanism, would need to continue creating long term outlooks. However, the Discussion Paper imagines LSEs preparing IRPs that in turn lead to procurement processes that are scheduled so as to take place prior to IESO auctions for capacity. There are a number of issues with this. First, given that the capacity auctions are forward auctions for a period several years into the future, LSEs are going to need to conform their planning horizon to match that of the IESO capacity mechanism. Secondly, the IESO in its long term outlook may now not know how many LDCs will have elected to become LSEs, leading to uncertainty around the IESO’s own planning processes. Third, the new LSEs and the IESO will need to develop a common planning and forecasting framework, so that as LSE load is subtracted from the load for which IESO is procuring capacity, there is consistency in methodologies.[8]

          While the OEA suggests customers will benefit from “efficiencies with respect to planning and resource procurements,”[9] it is hard to see how having more entities engaged in creating more plans and hosting more procurements results in efficiencies.

 

          2.4 Greater burden on OEB and stakeholders

The OEA proposal would place a significant additional burden on the OEB and also on stakeholders. The OEB would need to develop the LSE code and LSE license, and host processes to involve stakeholders in doing so. This alone is likely an 18 to 24 month initiative. Then, the OEB would need to review applications to become LSEs. Even if we assume that only 10 out of the more than 50 electric LDCs decide to become LSEs, that means that the OEB and stakeholders will need to participate in periodic reviews of 10 separate procurement plans, on top of what will presumably be ongoing regulatory proceedings regarding development and implementation of the new capacity mechanism. It is unclear that the net additional benefit, if any, of having such an LSE framework would be relative to the costs, not only of implementing it, but of subsequently maintaining it.

 

          2.5 Undermines capacity auctions

Utility self-supply has created challenges in the Independent System Operators (“ISOs”) where it is allowed. Effectively, the OEA is proposing something similar. By allowing entities serving a large proportion of Ontario’s load to effectively opt-out of the capacity mechanism, the proposal would reduce the number of participants in the capacity auctions, and relevance of the resulting price signal. By removing volume from the capacity auctions, the LSE proposal would result in smaller plant sizes for new build, potentially increasing costs. There would also be uncertainty from developers regarding how much load would be available to serve through the capacity market, making it more difficult to anticipate the types of projects that would be most useful.

          This uncertainty would spill into the design parameters for the capacity mechanism. For example, as the IESO considers the values for the Cost of New Entry (“CONE”), the unit sizes it uses to calculate CONE may no longer be relevant because the residual need met by the capacity mechanism would be smaller, meaning CONE could be understated for the remaining load. Results from the capacity auctions in general would be less meaningful for planning purposes as they would not reflect the value of capacity as a whole. Market participants would need to aggregate cost reporting from the individual LSEs along with the prices from the capacity mechanism auction in order to understand the total cost of capacity in Ontario.

 

          2.6 Enhanced load forecasting capabilities would be required at Ontario LDCs

For those LDCs who opt in to being the LSE under the OEA plan, significant human capacity building would be required. Ontario LDCs would need a standardized, rigorous approach to load forecasting using statistical methods and large customer analysis. If LDCs are going to be preparing IRPs, the IRPs need to be based on a sophisticated and granular understanding of load. While the OEA paper relies on the idea that the LDCs are closest to the customer, and thus best able to serve them, this needs to ultimately translate into forecasting expertise. While LDCs have some experience preparing CDM plans, the knowledge that is required to prepare credible IRPs may be lacking. For some LDCs it make take several years and significant outside support to develop these capabilities – capabilities that are currently more advanced at the IESO.

 

          2.7 Government backstop fails to reallocate risk

The report argues that “an effective LSE planning and procurement model should result in more appropriate risk allocation…”[10] It argues against socialization of generation costs; this ignores the fact that such socialization may potentially enable economies of scale, and that regional differentiations in prices to final consumers have been politically challenging in Ontario. Curiously, however, the report appears to believe that customers will bear less risk under the LSE model than they do currently. However, for that to be the case, either the LSEs themselves will need to bear more risk, or the suppliers will. LDCs are not going to (and should not) take on more risk.[11]

          The document suggests that there should be a government entity that will “financially backstop”[12] LSE contracts with third party providers in case they do not deliver new projects on time. The report does not, however, say who should pay for the costs of this backstop. If it is not the LSE’s customers themselves, then the concept fails the very principles of cost causation that the LSE proposal ostensibly is designed to meet. Indeed, placing IESO in the backstop role merely assigns to them the role that they are already playing. This means that the IESO would lose the ability to contract directly for needs (or meet them through the capacity mechanism) yet be responsible were an LSE to fail in its own contracting processes.

          At most, the OEA’s proposal reallocated risk among customers; it does not reallocate risk away from customers. While adopting a more passive role for the LSE might shift some risk to suppliers, this comes at a cost that suppliers will build into their bids in such processes.

 

          2.8 Unclear how LDCs benefit

One puzzling aspect of the paper is that it is not clear how the LDCs themselves would benefit. The Discussion Paper makes no mention of how the LDCs would be compensated for taking on this significant activity. Presumably the assumption is that the LDCs would be able to use their role as an LSE to push back on the threat from DERs within their own system and as a bridge towards owning in ratebase a more diverse range of resources subject to periodic market tests through procurements. Without either of these motivations, LDCs would be taking on a substantial effort with little to show for it. Becoming an LSE would require a large amount of senior management attention, potentially distracting LDC leaders from other pressing initiatives.

 

3. Concluding remarks

The LSE concept is neither new nor novel. As we have noted in the past, having multiple LSEs that collectively cover all load in a jurisdiction is a conventional means of creating a robust “buy side” for the market. However, the concept works poorly when it mixes regulated and non-regulated entities, incorporates entities with minimal risk management capabilities, or forces entities with capped rates to procure spot.

          For the LSE model to work, it needs a robust retail sector contracting with suppliers for varying terms potentially coupled with a default supply model. In many such default supply models, LDCs provide a pass-through service either of spot market costs or laddered multiyear contracts whereby a portion of non-switching load is contracted each year and the price to consumers is the weighted average of all of the outstanding contracts. This is not, however, what the OEA is proposing.

          The OEA appears to envision a world in which some LDCs become LSEs, the LSEs do not supply on a pass-through basis, and the LSEs ultimately become vertically integrated utilities themselves. Parts of the province not covered by LSEs would continue to be supplied by the IESO through residual capacity auctions.

          A properly implemented LSE model clearly delineates responsibility for serving load, avoids conflicts of interest among parties, minimizes the potential for cross-subsidies, and increases consumer choice. Established in this fashion, such a model can both promote competition and provide sufficient customer exposure to long term price volatility to encourage hedging through contracts of various lengths, in turn facilitating new build when needed. While this type of LSE model relies on decentralized planning and decision making (other than through the presence in many cases of a centralized capacity mechanism), it does redistribute risk among market participants, minimizes the potential for stranded costs, and allows for transparent price signals.

          On the other hand, poorly implemented LSE models exacerbate conflicts of interest, suppress wholesale prices, distort capacity mechanisms, and crowd out independent investment. The greater the number of players who are able to obtain revenues through ratebase or external subsidies, the harder it is for market participants without such access to remain in business. To varying degrees, these challenges can be seen in the California, MISO, and New York markets.

          While the proposal can be improved upon – by, for example, making it mandatory rather than voluntary, continuing to prohibit LDC ownership of generation, creating common analytic frameworks and building LDC capabilities, allowing entities other than LDCs to be LSEs, and increasing the centrality of the capacity mechanism – it ultimately will be difficult to implement. For suppliers, it would fragment the procurement landscape, potentially undervalue existing resources, and create planning uncertainty. Suppliers will benefit most from the IESO Market Renewal Process reaching a conclusion and certainty regarding nuclear refurbishment. The OEA proposal is a distraction that may draw attention of policymakers away from important ongoing initiatives.