- 7-A. Alternatives for the Federal Columbia River Power System
- 7-B. Consistency with the Principles, Characteristics and Limitations of Competitive Markets
- 7-C. Allocation of Benefits
- 7-D. Public Purposes
- 7-E. Conclusions
This chapter focuses on the role of the Bonneville Power Administration in an increasingly competitive electricity market. The reason for this focus is at least four-fold. First, as a wholesale utility, competition is already here for Bonneville, and it can probably be counted upon to become more intense. Second, Bonneville markets the output of a public resource. As a consequence, Bonneville's governance is more an issue of public policy than is the governance of other utilities. Third, Bonneville is a major and integral part of the region's power system. In an average year, it controls the marketing of almost 40 percent of the electricity sold in the region, most of which is relatively low-cost federal hydroelectric power, and it owns and operates the majority of the region's electricity transmission system. [ Depending on how regional transmission is defined, Bonneville owns between 50 and 80 percent of the region's transmission system.] Bonneville, or its successor, will continue to be a major factor in the region's electricity markets, its economy and its environment. Finally, the federal system has other purposes, public and private, besides power production. As a result, the issues surrounding Bonneville seem more complex.
The subject of the Comprehensive Review is the entire Northwest energy system, but a primary focus of the review is the role of the federal power generation and transmission assets in a competitive power marketplace. A number of alternatives are being discussed for Bonneville. They range from a somewhat scaled-back version of the current federal agency to privatization. The emerging competitive electricity market raises a number of issues for any alternative. Some of the questions and issues that may be addressed in the Comprehensive Review are discussed below.
7-A. Alternatives for the Federal Columbia River Power System
The advent of the competitive market and Bonneville's recent financial difficulties have caused many to ask whether Bonneville should continue to be a federal agency. They point out that federal agencies do not typically compete with the private sector. Advantages such as tax exempt status, greater regulatory autonomy and access to financing from the federal Treasury could be interpreted as giving Bonneville an unfair competitive advantage.
Counterbalancing these advantages, however, are a number of statutory requirements that could hamper Bonneville's competitive position. These include the mandate to serve the requirements of Bonneville's public agency customers, regional preference, prohibition on resale of federal power, cumbersome rate processes and several cost transfers such as the residential exchange, the low density discount and the Bureau of Reclamation's irrigation pumping rate. Several of these are requirements that Bonneville has either recently been successful in removing or modifying or that the agency's representatives have mentioned as in need of relaxation.
Supporters of a less-constrained Bonneville continuing in its present federal status argue that Bonneville must be competitive to meet its responsibility to repay the Treasury and fulfill its "social" responsibilities. Opponents argue that Bonneville may become so competitive that it will be in a position to exercise undue market power.
This chapter considers several alternatives for the Federal Columbia River Power System. These alternatives were chosen to illustrate some of the issues. This is not an exhaustive list. The alternatives are:
- A continuation of Bonneville as a federal agency, but with limitations. For example, it might be limited to marketing the output of the existing system;
- A continuation of Bonneville as a federal agency free to compete in the electricity market with as many constraints as possible removed;
- Sale of the rights to market the output of the Federal Columbia River Power System to a public regional entity;
- Sale of the rights to market the output to a private entity or entities; and
- Leasing the rights to market the output to public or private entities.
Sale of Assets vs. Sale or Lease of Marketing Rights
The list of alternatives is limited to the rights to market the output of the federal power system, not sale of the dams or other generating assets. This does not mean that sale or transfer of the physical assets might not be desirable under certain circumstances or that it cannot be accomplished. However, the multi-owner, multipurpose nature of the Columbia River system greatly increases the complexity associated with a sale of assets compared to a sale or lease of marketing rights.
For example, ownership requires responsiveness to the requirements for a number of public or quasi-public purposes (e.g., flood control, recreation, fish and wildlife, navigation) as well as commercial purposes (e.g., power and irrigation water). This is not an absolute obstacle to selling the assets, since a number of dams in the Northwest were constructed by non-federal utilities for power generation only, with the other requirements imposed as a license condition. However, because of their 50-year term, license conditions are not necessarily as flexible as public ownership in ensuring public purposes are met.
Moreover, political opposition could be increased by a sale of the assets, as opposed to sale of the marketing rights, at least if a proposed sale is to a private buyer. The Federal Columbia River Power System is built on an important natural resource for four states, two nations and many Indian tribes. Giving up public ownership of the dams is not an action that will be undertaken lightly. Apparently the non-power beneficiaries of the Southeastern Power Administration joined the power beneficiaries in opposing the recently proposed sale of that agency, in large part because it was a sale of the dams as well as of the power output. The proposal was killed in Congress.
A sale of marketing rights does involve a number of complexities. An important issue that would need to be resolved with a transfer of marketing rights is the degree of control afforded over the output of the dams. This is an issue for two reasons. The first reason is that the non-power constraints still allow flexibility, although not as much as in the past, in the decision to generate electricity or store water for later generation or other purposes. This flexibility is economically valuable. Bonneville currently uses this flexibility to maximize power value, within the constraints of the Coordination Agreement, dam operations requirements under the Endangered Species Act, and the Canadian Treaty. Any transfer of marketing rights will require a mechanism that can balance the ability to operate the system to maximize the value of power, versus operation of the system for non-power obligations.
A second reason degree of control is important is because of the "upstream/downstream" question. Storage releases from upstream dams usually constitute the bulk of the water flowing past downstream dams. The federal projects dominate the upstream storage capability. Coordination with downstream non-federal parties, primarily the mid-Columbia utilities, is essential both to optimizing the power output of the total system and to retaining the current rights of the downstream parties. The Coordination Agreement was developed, in large part, to resolve this potential for conflict. The issue is one of constraining the purchaser of the federal assets, especially if the purchase were of the dams or other assets themselves, but also if it were of marketing rights that include the flexibility to store water or generate electricity.
The complexities related to a transfer of marketing rights, however, appear much more manageable than those associated with a sale of assets. As a consequence, further discussion of non-federal alternatives will be limited to the sale or lease of marketing rights rather than the sale of generating assets.
Issues in Considering Alternatives
Whatever alternative is considered, it will be necessary to confront a number of issues. Some of the issues derive from the principles, characteristics and limitations of competitive markets discussed in Chapter 3. If Bonneville, or its successor, is to operate in a competitive electricity market, the principles, characteristics and limitations of that market will either apply, or the result will be a less effective market. Among the issues that should be considered in Bonneville's case are:
- The degree of separation of generation and transmission required to ensure that Bonneville, or its successor, cannot restrict competitors' access to the market;
- The degree of market power Bonneville, or its successor, might exercise as a result of its control of generation; and
- The ability to absorb competitive risks and rewards and the degree of congruence between those who take risks and those who reap rewards.
Related to the question of risk and reward are the terms of any sale or lease of the marketing rights and contractual constraints on any transfer of liability for the debt on the Washington Public Power Supply System nuclear projects.
Other considerations derive from Bonneville's historic role in the region, the public and quasi-public purposes Bonneville has fulfilled, and whether these purposes can be fulfilled in the future. These considerations include:
- Allocation of the benefits of the Federal Columbia River Power System through public and regional preference. The benefits are in the form of power sold at cost.
- Other public purposes, such as irrigation subsidies, mitigation of higher costs to serve low-density rural customers, access to the benefits of federal power for the residential customers of investor-owned utilities, and so on.
These issues are discussed in the following sections.
7-B. Consistency with the Principles, Characteristics and Limitations of Competitive Markets
Should Bonneville's Transmission and Generation Assets Be Separated?
Does the ownership by Bonneville, or its successor, of a very large percentage of the high voltage transmission in the region, combined with the rights to market the output of the Federal Columbia River Power System, give Bonneville market power inconsistent with a fair and effective competitive market? Functional separation of generation and transmission has been proposed as a requirement by the Federal Energy Regulatory Commission in its open access notice of proposed rulemaking for utilities under its jurisdiction. In its current form, Bonneville is not under FERC jurisdiction. Nonetheless, Bonneville is undertaking functional separation of transmission and generation within its existing organization. However, many fear that with pressure to repay the Treasury and at the same time keep power prices low, Bonneville will be tempted to exercise monopoly power over the federal transmission system to maximize the value of its power sales. If Bonneville's transmission system were sold along with the marketing rights to the output of the federal power system, it is likely there would still be similar concerns.
Setting up an independent, FERC-regulated grid operator for the region could insulate Bonneville, or its successor, and other transmission owners from monopolistic temptations. However, as long as Bonneville retains responsibility for marketing federal power, it cannot function as grid operator without facing the temptation to exercise undue market power. The conflict of interest between Bonneville as marketer of federal power and Bonneville as grid operator is unfortunate, given Bonneville's obvious strengths in the area of transmission.
Setting up a separate federal transmission agency with control over Bonneville's transmission assets is another option. Establishing FERC jurisdiction over this entity's transmission tariffs identical to its jurisdiction over investor-owned utility transmission tariffs could go far toward limiting Bonneville's market power. This separate federal entity might be able to play the role of independent grid operator, as well.
The idea of privatizing the transmission grid has also been raised. The resulting private transmission company would be regulated by FERC and would be allowed to earn a rate of return on its investment. It is not clear what benefits would be associated with private ownership compared to continued federal ownership. For example, could private ownership and operation of the system result in sufficient efficiencies compared to federal ownership to offset the higher return on investment a private owner would receive? If not, the result would be a net increase in the cost of transmission in the region.
Some have also suggested that it would be possible to sell the transmission system for more than its remaining debt, with the "profit" used to buy down the debt on Bonneville's high-cost generation. Since many of the users of the transmission system are not firm power customers of Bonneville, such a sale might be a mechanism for spreading the cost of Bonneville generation more broadly. This would undoubtedly raise issues of fairness and might not pass FERC scrutiny on the ground it would result in recovering generating costs in transmission charges.
Market Power
A fundamental question that will have to be resolved, whether Bonneville continues as a federal agency or whether its right to market federal power are sold to another entity, is that of market power. Does the entity have market power in any important electrical product as a result of its control over a large portion of the hydropower system? If so, what remedies are appropriate?
Bonneville, for example, clearly has the ability to influence spot market prices, at least at some times of the year. There may be other power products -- storage and load following, for example -- that the hydropower system is particularly able to provide. Competitive markets for these products may not exist. If not, some degree of regulation of their prices may be necessary.
If Bonneville, or its successor, is to be a full participant in the competitive power market, it may be necessary to sell the marketing rights to more than one party. This raises issues of how the output of the system would be allocated. These issues are probably manageable. Limiting Bonneville's role to an allocation of power to its customers, with limited ability for Bonneville to market any residual power, probably accomplishes the same end. If the structure of the wholesale electricity market in the region evolves toward a mandatory pool, it may be possible to mitigate the market power associated with Bonneville power marketing. In such a pool, prices are set by the marginal bid price in any period. Experience in the United Kingdom, however, indicates it is possible to exercise market power through a mandatory pool if there is sufficient concentration of ownership or the rules for operation of the pool are poorly set. The interaction between market structure and market power should be investigated.
Markets, Risks and Rewards
Competitive markets imply the risk of business failure and loss. Conversely, they also imply the possibility of success and profits. Whatever form Bonneville, or its successor, takes in the future, it will have to be able to accommodate the possibility of either profit or loss.
Risk, Reward and Federal Ownership
The risk of long-term loss poses a problem for Bonneville as a federal agency in the transition to competitive electricity markets. Bonneville has long been subject to the risk of year-to-year fluctuations in hydropower output, risks of fish and wildlife restoration costs and risks associated with treaty obligations to the region's Indian tribes. In the past, Bonneville has been able to absorb these risks because its costs have been consistently below market. The advent of competition poses the possibility and, in recent months, the reality that market competitors may undercut Bonneville's prices for extended periods of time.
As a federal agency, Bonneville has no stockholders to absorb the business losses that are bound to happen, to a greater or lesser extent, in a competitive environment. Instead, the federal Treasury ultimately bears the burden of losses in excess of what can be covered by Bonneville's financial reserves. Bonneville has not yet incurred any long-term losses, and past missed Treasury payments were subsequently brought up to date, with interest. However, one of the mechanisms by which Bonneville has lowered its proposed 1996-2001 rates is a reduction of the probability of full, on-time repayment of its Treasury obligations. On the other hand, recent agreements to limit fish recovery costs have raised the probability of meeting the Treasury payment.
As electricity generation evolves toward a fully competitive industry, the possibility of long-term loss needs to be addressed. Stranded costs due to the competitive transition represent one form of loss, but dealing with the current level of stranded costs, difficult as it may be, will only require a one-time solution to a one-time problem. It should not be assumed that losses could not recur due to changes in technology, customer choices, and so forth. Bonneville's financial problems are generally considered short term (over the next three to five years, for instance), but that is not guaranteed. In a competitive market, prices are independent of a company's own costs. Generally, customers cannot be expected to bear any of the burden of either short-term or long-term losses, since they will simply find a different supplier if the current supplier tries to raise prices above market levels.
If Bonneville is to continue as a federal agency, there are at least two risk-related questions that must be answered. First, with the greater risk exposure associated with a competitive market, will the federal Treasury continue to fulfill the risk-bearing function? If not, what are the options for bearing that risk? Second, should the Treasury be exposed to additional risk as a result of new resource development by Bonneville? The Northwest Power Act obligated Bonneville to meet the requirements of its preference customers and authorized Bonneville to acquire resources to meet those requirements. Those customers, however, are under no obligation to purchase power beyond the periods established in their contracts.
Under the Power Act, Bonneville was granted the authority to acquire resources because, at the time, new resources were large, required long lead times and were very expensive. Small public utilities and even investor-owned utilities were not expected to be able to shoulder the risks of such huge investments without federal backing. This is much less the case in today's utility world. New combustion-turbine technologies, for example, are smaller in scale, less expensive and require far shorter lead times to develop. The risks to utilities from resource development are more manageable, and other entities can develop and market these resources. Consequently, if Bonneville is to continue as a federal agency, there may be reason to limit its role to marketing the output of the existing system. Bonneville no longer needs to take on the risks associated with new resource acquisitions because the utilities themselves are more financially able to manage those risks.
Just as a competitive market implies risks, it also implies the possibility of rewards or profits. It is possible to construct scenarios in which Bonneville's costs are once again below market prices. For example, when the debt on the Washington Public Power Supply System nuclear plants is retired beginning in 2011, it appears likely that Bonneville's costs would be well below market prices. When and if this occurs, will the federal government be willing to allow the region to retain the reward in the form of either profits or below-market prices, or will it want to appropriate some or all of the benefit for the Treasury?
Risk, Reward and Regional Public Ownership
One set of alternatives to continued federal ownership of the marketing rights of the federal power system involve some form of regional public lease or ownership. Regional public ownership is a mechanism for ensuring that potential benefits are retained by the region, at least to the extent that the terms of the sale or lease leave room for benefits.
Risk, however, is still an issue. For a general-purpose government entity (e.g., a state or municipal government), shortfalls are managed by shifting budget accounts or raising taxes. For a non-taxing public entity (e.g., a wholesale generation-only analog of a public utility district), probably neither is possible. If market prices fall below costs, and customers have access to the market, there is no entity to absorb the loss. The same holds true for a non-profit, non-governmental entity purchasing Bonneville's marketing or generation assets. One alternative is federal government guarantees of the debt of such an entity. However, as demonstrated by the savings and loan problem of the 1980s, such backing can substantially distort investment incentives and become a major problem for taxpayers, although it is often perceived to be without cost when it is proposed.
Risk, Reward and Private Ownership
In the private corporate economy, stockholders bear the business risks and incur whatever profits or losses result from taking those risks. Stockholders lose if there are stranded costs. They win if their firm is more efficient than its competitors. Privatization of Bonneville's power marketing function would resolve the allocation of risk and reward in a manner that is consistent with the private economy. This includes transferring any return that might be earned by the regional system to the participants in the sale -- the federal government and the private purchaser.
What is an Appropriate Price for the Rights to Market the Output of the Federal System?
Any sale or lease of the marketing rights for the federal power system involves determining a price. The process of determining a price is one of assessing potential risk and potential reward. Under most circumstances, no one should pay more than market value, and the Treasury should not accept less than embedded cost, unless it is greater than the expected market value. Of course, in this instance, the assessed market value is not certain. Market value depends on the relationship between future costs and the future market price of electricity, both of which can be estimated, but not known. Consequently, there is a great deal of room for negotiation of the price. Because of this uncertainty and the possible desire for immediate deficit reduction, the government could accept a price lower than embedded cost.
In trying to establish market value, it is important to specify the operative time horizon. With a short time horizon (the next five-year rate period, for example), there may be little difference between the market value of the output and embedded cost. Over a longer term (a permanent sale, for example), a buyer might expect there to be more market value in a system that is dominated by fixed costs and low variable costs when the environment is one of variable gas prices. On the other hand, the fixed-cost burden might be considered a liability in an environment in which generating costs and efficiencies are being improved, and fuel prices are stable or declining.
There are, therefore, two basic conceptual choices for the term of the transfer: permanent or limited term. A permanent transfer might be the simplest, but it has the greatest possibility of deviations from subsequently observed market values, primarily because of uncertainty-related discounting by the purchaser.
The alternative is a limited-term transfer, for example, an auction every five or ten years. The shorter the term, the closer the result will be to the observed market value of the system output.
The duration of the transferred rights will affect the perceived value of those rights. Uncertainty about future value might be reduced by shorter-term sales. A series of shorter-term auctions may produce higher prices for the sale of marketing rights than a one-time long-term sale if system value rises over time.
Purchasers of a longer-term right would take into account the potential net value above cost in the out years, but that net value would be discounted because of timing and (most likely) uncertainty below a simple sum of the forecast net values. A longer-term sale will produce more revenue up front than a series of shorter-term sales, depending on how purchasers perceive future risks. The relative values of a long-term sale compared to a series of short-term sales would have to be explored further using various parties? discount rates and expectations about net value of the system in future years.
Contract Constraints on Transfer of Nuclear Power Plant Assets and Liabilities
The Bonneville Power Administration assumed responsibility for paying the principal and interest on the bonds for the construction of the Washington Public Power Supply System's nuclear plants 1, 2 and 3. A transfer of the marketing rights of the federal system would have to address this responsibility. A preliminary examination of this question was conducted more than a decade ago. This analysis was focused on sale of the physical assets. The same issues would appear to be relevant to a lease or sale of the marketing rights.
The examination concluded that the various WPPSS-related contracts (bond resolutions, project agreements, net-billing agreements) appear to severely constrain the ability to assign the WPPSS marketing authority and financial liability away from Bonneville without, ultimately, the consent of the bondholders. The only alternative that offered a clear transfer path was to pay off the bonds at the time of transfer. Other approaches that did not require immediate payment were considered possible, but are affected by legal ambiguities that would need to be resolved or do not meet the test of completely transferring the assets away from Bonneville. Resolution of these questions would be a necessary condition for any sale of marketing rights.
7-C. Allocation of Benefits
How the possible benefits of the Federal Columbia River Power System are allocated is an important and difficult question for the region. Any change in the status quo has the potential to alter that allocation of benefits.
The Basis of the Benefits -- the Hydropower System
The essential "regional benefit" provided by Bonneville is financial -- the difference between a free-market price of electricity and the low historic costs of the hydropower system and its associated transmission system, largely constructed by the federal government. While much of that benefit has been diluted by past nuclear investments and by the general lowering of the market price level in recent years, that benefit was substantial at times in the past and could be substantial in the future, depending on changing electricity generating technologies and fuel markets.
Distribution of Regional Benefits
The benefits of the federal hydropower system were widely distributed in the region prior to 1973, when Bonneville's existing 20-year firm power contracts with investor-owned utilities were not renewed. Between 1974 and 1981, customers of investor-owned utilities had no access to firm power from the federal hydropower system. In 1981, as a result of the Power Act's residential exchange provisions, the financial benefits of federal hydropower were again made available to residential and small farm customers of investor-owned utilities. In 1985, Bonneville revised its average system cost methodology, and the residential exchange benefit to investor-owned utility customers was reduced. It is expected to be reduced even more after 1997, with the phase-out of the residential exchange. Although there have been changes over time, the primary beneficiaries of the hydropower system have historically been a wide spectrum of public and investor-owned utilities, and direct-service industrial customers.
The distribution of whatever future benefits can be produced by the system will be, at least in part, a function of the ownership of the rights to market the output of the system and the risk that goes with that ownership. One possible outcome might be continued federal ownership and continued willingness on the part of the federal government to be the ultimate bearer of risk, ensuring that the benefits of the power system go to some or all of Bonneville's traditional regional customers. Bonneville was created to achieve such public purposes as regional development, which go well beyond market risk and reward relationships. Whether the federal government will be willing to maintain the current allocation of risks and benefits in a world of competitive wholesale electricity transactions is a question the region must confront.
If some sale of the marketing rights is undertaken, the terms of the sale will, as discussed earlier, result in some distribution of risk and potential benefits between the Treasury and the buyer. Who the buyer is will determine who receives the buyer's share of potential benefits and risks. A private buyer will take the risks and return whatever future benefits can be produced to its investors. If the buyer is a consortium of Bonneville's current customers, then the risks and the potential benefits would be allocated to those customers. If the buyer is some entity created by the Northwest states, the benefits as well as the risks would go to the states to be further allocated as determined by the states, perhaps to taxpayers or to specific customers.
Marketing and Pricing
Historically, Bonneville has been constrained in its marketing of power from the Federal Columbia River Power System to giving preference first to its public agency customers and second to the region. The restrictions on out-of-region sales have recently been relaxed, but not eliminated. The term of out-of-region surplus sales is limited to seven years, and regional customers retain a right of first refusal on surplus sales (regional customers are given the opportunity to match the price offered by an out-of-region customer).
The marketing restrictions on Bonneville may not be as vital an issue for Bonneville customers today as they once were. The approximate convergence of increasing Bonneville costs and falling market prices have effectively eliminated the price advantage that Bonneville's power once carried. Some Bonneville customers have been willing, at least temporarily, to leave the federal system and the risks associated with that system in order to buy from the market. However, Bonneville's costs may be below market prices in the future. How should the power from the Federal Columbia River Power System be marketed now, when its costs are close to market prices, and in the future, when the costs may be below market prices? Should it be marketed on a preferential basis, with any surplus made available to the broader market, or should its marketing be unconstrained?
A corollary question has to do with the pricing of the power from the Federal Columbia River Power System. The below-market pricing of Bonneville power has historically been the mechanism by which benefits have been delivered to Bonneville's regional constituencies. The marketing of Bonneville power at cost has been a major reason why many Northwest utilities and their consumers have enjoyed rates well below the national average. This raises the question of how possible future benefits, to the extent they can be retained for the region, are to be returned -- in the form of prices that are again below market or in some other form, for example, cash dividends.
However, below-market pricing was one of the main inefficiencies that led to the dramatic over-investment in the region's nuclear plants in the late 1970s. This problem might or might not recur in the future. If the dividend is continued in the form of below-market prices, but Bonneville is no longer in the resource acquisition business (or is in it on the basis of specific acquisitions at market prices), then this distortion will be eliminated. If, however, Bonneville acquires new resources to meet its customers? load growth and sets its prices by averaging the costs of existing resources with those of the new acquisitions, it will be reinstating the price distortions of the 1970s. A better method of conveying the dividend to regional beneficiaries needs to be designed.
7-D. Public Purposes
The Federal Columbia River Power System has historically supported a number of "public purposes" beyond that of providing power to its customers at cost. These have included cost transfers that benefit different classes of customers, such as reduced rates for irrigation and low-density rural customers and the residential exchange that benefits the residential customers of investor-owned utilities. Some would also include centralized funding of conservation, activities to encourage renewable resource development, and other forms of research, development and demonstration. Fish and wildlife costs are a cost of producing power and are among the purposes for which the hydropower facilities are operated. Without attempting to sort out which costs truly cover public purposes, the question for the region is which of the public purposes should be maintained and how the region can best accomplish those purposes in the context of a competitive power market.
The principles of competitive markets suggest that subsidized rates are not the way to accomplish public purposes. Such rates are both inefficient and, because they cause the unsubsidized customers? rates to be higher, they create an opportunity for competitors to exploit. To the extent the power system can earn a profit, public purposes can be supported from those profits. To accomplish this, however, dividends for other purposes must be reduced. How to balance profits and public purposes is a legitimate policy decision that will have to be addressed in the course of the Comprehensive Review.
However, competitive markets don't guarantee profits. It may be that the power system cannot be counted upon to earn a profit or one that is sufficient to support both public purposes and the return requirements of the risk-bearing owners, whether public or private. If that is the case, other non-market mechanisms to support those public purposes may be required. These mechanisms could include a regulatory requirement applied to the monopoly elements of the business, a general tax or a charge for use of the transmission or distribution system, a tax on generation or fuel use, or development standards for new energy facilities.
7-E. Conclusions
Many argue that the Bonneville Power Administration, as currently configured, violates several of the principles for a competitive market. Bonneville combines generation and transmission in one entity. It has substantial market power. Market risk is ultimately borne by the Federal Treasury. And it carries out several public purposes that may be difficult to support in a competitive wholesale power market. At the same time, Bonneville is at the heart of the regional power system and embodies many of the values of the region. Deciding the future role of Bonneville is a key task of the Comprehensive Review of the Northwest Energy System. Successful resolution of the Bonneville question will set the stage for an efficient and competitive regional power system.