Since late 1997, investor-owned utilities have been divesting power generation assets in record numbers. The process of selling large power plants is complicated, and the outcome of the sale is important to electricity customers (i.e. rate payers) and utility owners. This appendix presents three case studies describing the process of divesting power plants.

Case 1: Central Maine Power

Maine's restructuring law (LD 1804) requires divestiture of all generation by utilities. Exceptions are allowed for certain power purchase contracts, nuclear power plants, sites outside of the United States, and plants deemed by the Maine Public Utility Commission to be necessary for reliable performance of the utility's obligations. To respond to this law, Central Maine Power (CMP) placed its entire 2,110 MW asset portfolio up for auction. A total of 1,121 MW were sold in the initial auction. (See box for more details on CMP's asset divestitures.) CMP is still seeking buyers for the remaining assets. However, of the remaining 989 MW, only 127 MW must be divested.

The sale opened in May, 1997 with CMP's entire 2,110 MW portfolio of generation assets on the market, packaged by fuel type: fossil, hydro, biomass, nuclear, and power contracts. This included 862 MW of nuclear and power contract generation assets which were exempt from the mandated divestiture. Final bids were submitted in early December, 1997 and one month later CMP announced that FPL Energy had been selected to buy the fossil, hydro, and biomass packages. No buyers were selected for either the nuclear or power contract assets as CMP deemed none of the offers to be adequate. Approval by the Maine PUC and by FERC came in November, 1998. The sale closed in April, 1999.

This sale was highly controversial because of an appended Letter of Agreement between CMP and FPL in which CMP agreed to use its vote within the New England Power Pool (NEPOOL) to lobby for FPL's interests until FERC approved new guidelines for transmission access in the deregulated market. FPL was trying to maintain the priority of access to transmission lines that CMP had enjoyed under regulation. Some intervenors feared that this agreement, if allowed, would effectively put FPL in NEPOOL, giving it an advantage over other generators and violating the spirit, if not the letter, of Maine's restructuring law. CMP, however, saw the agreement as strictly limited in time and scope, and the PUC approved the sale, including the letter, on that basis.

In October, 1998, FERC did issue a ruling on NEPOOL's transmission access rules, ordering NEPOOL to revise the rules to lessen the burden on new generators connecting to the system. FPL felt that the ruling revoked the priority access that the CMP plants had previously enjoyed and considered this to be sufficiently harmful to the value of the plants that it filed suit in Federal court seeking a declaratory judgement voiding the purchase contract. The Court ruled in favor of CMP in April, 1999. FPL chose not to pursue the matter and closed the sale later that month.

The Auction Process

Public announcements and personal contacts with potentially interested bidders were used to generate interest in the sale. The assets were grouped by generation type to hold down the transaction costs of the sale. In phase I, a memo and reference manual for the auction were sent to all qualified bidders in June, 1997. Also, a document center was set up for bidders to review more detailed information on the plants. Tours of selected plants were conducted as part of the process. Non-binding bids were due by September 10, 1997. CMP and its financial advisor, Dillon Read, then reviewed these bids and selected final round bidders based on: 1) price offered, 2) financial ability of the bidder, 3) degree of deviation from the terms and conditions of the offering memorandum, 4) continued opportunities for current CMP employees, 5) flexibility to negotiate savings in power contracts, 6) assumption of CMP's collective bargaining agreement, and 7) ability of bidder to operate assets reliably in a competitive environment. Selected phase II bidders were sent an information packet with detailed financial information and a purchase/sale agreement form with terms/conditions that should be considered in submitting the final, binding bid. Phase II bids were due by December 10, 1997. CMP indicated that it would consider bids for partial packages, but clear preference would be given to bids made for complete packages.

The two-stage process was chosen to improve the chances of attracting serious bids. The first stage eliminates those unlikely to prevail, improving the odds for the remainder and increasing the resources they are willing to devote to a serious bid. However, the number of bidders must not be so low that their resources are devoted not to evaluating the assets but to forecasting their competitor's bid. CMP feared that this would generally lower the level of the bids.

Bundling assets was a method used to reduce administrative costs and improve chances for selling all assets. (In this method, low-value assets that will attract few, if any, bids are bundled with high-value ones.) Bundling may harm the total value of the assets if there are multiple buyers with different valuations for each plant, and all plants are valued by some bidders. (For example, Cape Station may have had more value as a pure real estate deal than as part of a power plant package.) CMP attempted to reduce this drawback by encouraging those wishing to bid on partial packages to form coalitions to bid on the entire package. This had the added benefit of reducing the number of bids to be considered.

CMP's plan was to file for approval of the sale within 45 days of choosing the buyer and get PUC approval within 7 months of filing. The PUC found this timeline feasible providing the filing contained sufficiently complete and detailed information, including the complete purchase/sale agreement, an analysis showing that the sale maximizes asset value obtained, an analysis of replacement power for the interim between closing the sale and the opening of competition, and an analysis of the sale's impact on market power.

The selling price of the assets was substantially above their book value. Book value of the assets was approximately $231 million, and the selling price was $846 million. In part, this is due to the hydro assets which have a very low book value but are still in excellent operating condition. Maine's requirement that all power providers include at least 30 percent renewable power in their supply portfolio would also have pushed up the price. Third, FPL Energy's belief that existing generation assets would have priority access to the transmission grid increased the price they bid. CMP will use the proceeds of the sale to retire debt and perhaps finance a rate reduction.

FPL's plans for the assets include upgrading or replacing some of the older units and building 1,500 MW of new generating capacity on the sites.

Case 2: Pacific Gas & Electric Company

California's restructuring law (AB 1890) does not explicitly require divestiture. However, it does call for separation of transmission and generation, and it does require that no generator in the restructured market be able to exercise significant market power. Because of Pacific Gas & Electric's (PG&E's) size (the total nameplate capacity of its generation assets was over 14,000 MW) the California Public Utility Commission directed PG&E to voluntarily divest at least 50 percent of its fossil generation to mitigate its market power. PG&E chose to divest virtually all of its fossil generation, keeping only the 105 MW Humboldt Bay gas plant. (See box for more details on PG&E's asset divestitures.) (Because it is located on the site of a decommissioned nuclear plant, its sale would involve an excessive amount of regulatory red tape.) The sale was conducted in two auctions, splitting the plants among three buyers. The final stage in PG&E's generation restructuring is the auction of its hydroelectric generating assets. PG&E is keeping the 2,200 MW El Diablo nuclear plant.

PG&E's initial auction, proposed in October, 1996, offered four fossil plants for sale: Moss Landing, Morro Bay, Oakland, and Hunter's Point. In June, 1997, Hunter's Point was withdrawn from the initial auction and added to a proposed second auction which offered four more plants for sale: Potrero, Pittsburg, and Contra

Costa (all fossil plants), and the Geysers geothermal plants. The first auction began in September, 1997 and concluded with the November announcement that Duke Energy had been selected as the buyer. The sale generated little controversy and closed in July, 1998. The second auction began in April, 1998 and concluded in November, 1998 with Southern Energy selected to buy the fossil plants, and FPL Energy the geothermal plants. Subsequently, Calpine, owner of the geothermal steam fields that supply the Geysers plants, exercised its right of first refusal and supplanted FPL as the buyer of Geysers. The Southern Energy sale closed in April, 1999 and the Calpine sale in May, 1999.

The controversy in these auctions revolved around the Hunter's Point and Potrero plants. Both are old and inefficient, located in minority neighborhoods in San Francisco, and the subjects of repeated complaints that they pose a health hazard to the residents. They are also both "must run" plants, required for the reliable supply of power to the San Francisco area. (A transmission bottleneck limits the amount of power that can be delivered from outside.) San Francisco was afraid that the new owner would increase generation at the plants to maximize its revenue at the expense of the health of the residents. The city sought to buy the plants itself, but was late submitting a bid, and the PUC would not give it special status. After the city threatened to exercise its right of eminent domain to break the impasse, PG&E agreed to withdraw Hunter's Point from sale and close it down as soon as its "must run" status could be removed.

The Auction Process

On the advice of its financial advisor for the divestiture, Morgan Stanley, PG&E proposed a two-stage open auction for both auctions. The basic format of both auctions was the same. In stage 1, PG&E publicized the sale to potential bidders, providing basic information on the assets to be sold and the terms and conditions of the sales agreement. Interested bidders provided PG&E with evidence of their financial and operational qualifications, and a nonbinding bid. In the first auction, bids could be placed on any combination of plants; in the second, Pittsburg and Contra Costa were bundled as a single unit and separate bids were required for the Lake County and Sonoma County units of the Geysers geothermal plant. PG&E chose 5-10 final round bidders for each plant. In the second stage, PG&E provided detailed information in support of the due diligence being conducted by the bidders. At this time, the bidders were allowed to propose changes in the sales agreement--PG&E issued the final form of the agreement two weeks prior to the final bid due date. Each plant was sold to the highest bidder, assuming PG&E's reservation price was met and no unacceptable conditions were subsequently imposed by the reviewing agencies.

In cases where significant environmental impact is a possibility, California's Environmental Quality Act requires an Environmental Impact Report to be completed by the PUC, detailing mitigation requirements. This was done for the second auction, in large part because of the controversy over Hunter's Point and Potrero. Remediation costs totaling nearly $90 million were imposed on PG&E, which it may recover through the Competitive Transition Charge.⁽⁴⁹⁾

The California PUC is also charged with ensuring that the deregulated electric power system will continue to run reliably and that no generator will be able to exercise market power. The distribution of PG&E's assets among three buyers satisfied the goal of mitigating market power. The reliability question is handled in part through the designation of some plants as "must run" status plants, which places obligations on the owner of the plant. California's restructuring law also contributes to the continuity and reliability of plant operation by requiring the new owner to contract with the old owner to operate the plant for two years from the closing of the sale. Lastly, the requirement of proof of operational expertise at stage 1 of the auction to be considered a qualified bidder helped satisfy the goal of continued reliability.

In November, 1998 PG&E began the final phase of its divestiture, submitting a plan to transfer its hydroelectric generation to its unregulated affiliate, PG&E Generating. PG&E chose to divest via transfer rather than auction for economic reasons. First, it was thought that the transfer could be accomplished in as little as 6 months, compared to over 2 years to complete the auction process. This would allow PG&E to end its stranded cost recovery, and thus its rate freeze, well before the March 31, 2002 deadline. Second, the transfer avoids the large Federal capital gains taxes that would be due if the plants were sold at auction. These savings would be applied to PG&E's stranded costs, benefitting California's ratepayers. The value of the transferred assets was to be assessed by outside experts, as required by California's restructuring law.

This plan was highly controversial and drew criticism from environmentalists, consumer groups, municipalities, State regulators and State legislators, all staking a claim to what was expected to be a very valuable asset. The Association of California Water Agencies (ACWA) assessed the value of the plants at between $3.14 billion and $4.34 billion. The ACWA saw no merit to market power criticisms of a transfer, but warned that the relicensing of the plants would likely reduce their value, either through increased environmental mitigation costs or through reduced generation capability. Several bills were introduced into the California Legislature championing various sides of the issue, including one by PG&E and its allies seeking approval for the transfer. The PG&E bill proposed setting the plant's value at $3.3 billion, about $2.5 billion above book value. However, the 1999 legislative session ended without any action having been taken. On September 30, 1999 PG&E filed an application with the PUC outlining an auction plan for the hydro plants, splitting them into 20 bundles. PG&E Generating would participate in this auction.

The El Dorado hydroelectric project has been separated from the rest of the hydroelectric system and sold. It had suffered severe damage from winter storms in recent years and PG&E decided it was not economically worthwhile to repair the damage. The "buyer," El Dorado Irrigation District, bought El Dorado to obtain the water delivery assets of the project and plans to dismantle the power plant.

With the exception of El Dorado and Geysers, all plants sold brought in considerably more than their book value. For example, the Potrero, Costa, and Pittsburg power plants sold for $801 million. Their book value was $256 million. The reason for El Dorado's low price was noted above. In the case of the Geysers, the likely reason is supply constraints on capacity utilization. Although rated at 1,224 MW, the current condition of the geothermal steam fields supplying the plants restrict their effective capacity to 665 MW. The net excess of price over book value plus transaction costs will be used to lower PG&E's stranded costs. Calpine, owner of the Geysers steam fields, purchased the power plants in order to unify steam field and power plant operations, reducing costs to California consumers and extending the life of the assets. Duke and Southern both plan on actively participating in the merchant power market in California. They are somewhat constrained by the "must run" status of most of their units and environmental restrictions on the operation of others (Potrero and Pittsburg). Several of the older units will probably be upgraded or replaced with new, larger units.

Case 3: Portland General Electric

In 1996, the Governor of Oregon issued a statement of principles as a guideline to restructuring. However, the Oregon legislature has not yet passed restructuring legislation. To adapt to the new environment, Portland General Electric (PGE) is voluntarily divesting all of its generation assets. It intends to become a regulated transmission and distribution company and thus is seeking to sell all of its generation and related assets.

PGE filed its divestiture plan with the Oregon PUC in September, 1997, choosing Merrill Lynch to serve as its financial advisor in the sale. By taking advantage of the current excess demand for generation assets, PGE, like General Public Utilities System and Montana Power, hopes to realize a premium on the sale of their assets before the increasing number of States with restructuring laws that require divestiture glut the market and bring prices back down. (See box for more details on PGE's asset divestitures.)

Seller: Portland (Oregon) General Electric (a subsidiary of Enron Corporation)

Asset: 3,030 MW of generation and supply contracts, split into 5 packages (which included Boardman [330 MW coal], Beaver, Bethel, and Coyote Springs [830 MW gas/waste], Pelton and Round Butte [408 MW, hydro], Clackamas, Bull Run and Sullivan [202 MW hydro], and 1,260 MW of generation contracts)

Asset: 323 MW share of Colstrip (coal)

Buyer: PP&L Global, Inc

Details: Sold in conjunction with shares of Montana Power and Puget Sound Energy in November, 1998; PGE's share of the price was $230.5 million (book value was $219 million)

Asset: 33.5 MW share of Centralia (coal)

Buyer: TransAlta

Details: Sold in conjunction with the other 7 owners of the plant in May 1999; PGE's share of the sale price was $13.85 million (book value was $4 million)

The Auction Process

PGE proposed a two-stage auction process for qualified bidders, with sealed bids, and selection made on the basis of price plus imputed value of other terms and conditions. They favor a two-stage auction because: (1) it is expensive to develop binding bids on generation assets and bidders are unlikely to commit the necessary resources until they have some indication that their chances of success are reasonable, and (2) conducting due diligence is expensive for the seller as well, as they must make company resources and senior officials available to all bidders. The use of nonbinding first-round bids to filter out weak bidders quickly reduces the cost of exploring a sale, provides the second round bidders with the signal they need that their chances are reasonable, and cuts administrative costs to the seller. Sealed bids help the company to maximize value received for the assets--in a public auction the winning bid will almost surely be only slightly larger than the second place bid, even if the winner was willing to go much higher to acquire the assets. The use of imputed value for the other terms and conditions of the sale, rather than price only, helps maximize the overall value of the sale and improves the chances of obtaining regulatory approval in cases where these conditions are important to the community.

PGE's plan was partially approved by the Oregon PUC in January, 1999. The divestiture of fossil assets and power contracts was not controversial and was approved. However, the proposed divestiture of hydroelectric generation was controversial.

The Oregon PUC agreed with the intervenors that the sale of PGE's hydroelectric assets was not in the best interest of the State. The issues they cited were:

• The sale would have an adverse impact on and would be adversely impacted by the relicensing of the hydroelectric projects. In particular, the PUC felt the sale is likely to delay the relicensing process, despite the FERC's assurances to the contrary. Further, the uncertainties of the relicensing process would likely lower the bids for hydroelectric plants, as would knowledge that the sale would receive close scrutiny by the PUC.


• Hydroelectric's low cost is a major reason that Oregon's electricity rates are among the lowest in the Nation. The PUC felt complete merchant status for all generation would almost surely raise average prices, mostly to residential customers. Retaining the hydroelectric plants would lower Oregon's dependence on market purchases and reduce price volatility.


• Properly evaluating and allocating the sale's benefits is difficult. The PUC felt mixed sales of hydroelectric and fossil plants would make it difficult to ensure that the hydroelectric assets were properly valued. Further, it argued that since the sale is not reversible, if the anticipated benefits did not appear, it would be too late to backtrack. Finally, PGE's plan was to amortize the benefits over 5 years; the PUC argued that, because of the long life of hydroelectric assets, this would deny the benefits of the sale to many future users of the power from those plants.

As an alternative to the sale of the plants to an outside company, the PUC offered a plan in which the hydroelectric assets would be spun off to an affiliated generating company of PGE.

At present, PGE is awaiting the action of the Oregon legislature before deciding on how to proceed with its planned divestiture. Because of the expense in bidding on generation assets, the support of the PUC is an important element in attracting good bids. If it is likely that the PUC will not approve the sale, or place expensive conditions on it, then the assets become less valuable to the bidder. Bids will be lowered in compensation for these expected additional costs, and fewer resources will be committed to generating a bid. The sales of PGE's shares of the Centralia and Colstrip plants were conducted separately from the proposed auction of PGE's other assets. Each was sold in conjunction with shares held by the other owners of the plants, in order to maximize the sale value. That is, selling a majority stake in a plant will likely attract better bids than the separate sale of several minority stakes.

49. This is a charge to the ratepayer to cover a utility's costs as a result of California's electricity industry restructuring program.

• Proceed to Appendix C
• Return to Appendix A
• Return to Front--Table of Contents