Note No.47 Competitive Contracting for Privately Generated Power What to do in the absence of competition in the market Robern Bacon The importance of competition the most efficient plant in the merit order and should indeed run on base load as specified in Independent power procducers (IPPs) are an the typical contract. But circumstances can increasingly inportant type of power genera- change, and after a few years. running the IPP tion project in developing countries. In manv as baseload might no longer be optimal. When of these countries. power sector reform involves a contract does not allow for such a contin- unbundling generation from transmission and gencv, the operational efficiencv of the svstem introducing private capital. Rather than pri- tends to decline. And contracts that provide vatizing existing generators. some governments for guaranteed sales to reduce private inves- prefer to create an enabling environment for tors' risk reduce the competitive pressure on IPPs. They have two objectives in doing so. them to operate their plants efficiently. First, by using truly competitive bidding to pro- cure new generation capacity, governments There is inevitably a tension between design- seek to minimize the costs of expanding power ing contracts to reduce uncertainty for the pri- supply. And second. by introducing the profit vate investor and running the power system as motive and competition in the power and en- efficiently as possible. The characteristics of ergy sector, they seek gains in operational effi- power sales agreements are pivotal in resolv- ciency that lead to lower prices for electricity ing this tension. This Note surveys a range of users than would otherwise have been pos- selling agreements-from the most rigid to the sible without subsidies. most flexible-and highlights their risks and benefits for operational efficiency. The second result is especiailly hard to achieve. and when achieved. it can be even harder to Some of the most common contractual arrange- sustain. Governments in most developing coun- ments for [PPs place no competitive pressure tries lack the proven track record of transpar- on existing suppliers. These arrangements will ent regulation of their power sectors that is not help improve the efficiencv of the sector. needed to attract private investment. Thus, They will only increase supply capacitv- long-term contracts (power purchase agree- though that can he valuable in a svstem facing ments) are required both to encourage entrv public financing constraints. When the entry hy potential investors and to safeguard their of IPPs fails to put competitive pressure on interests. Such contracts attempt to share the other suppliers. governments need to design risks between the parties in a predictable fash- other efficiency-enhancing reforms, such as per- ion. But heCause it is difficult to write legal formance contracts for the nonprivate genera- clauses that cover uncertainties about fLuture tors. And if the contract does not lead to marketing concditions. they cannot he fullv con- head-to-hiead competition between new IPPs tingent. so inefficiencies arise in system op- and the other suppliers, it should include per- eration. For example, at the time of signing a formance incentives to ensure that the IPP re- power purchase agreement, the IPP may be miains an efficient supplier. Industrv and Energy Department * Vice Presidencv for Finance and Private Sector Development Competitive Contracting for Privately Generated Power Methods of contracting for the sale of cases. penalties and bonuses are an important IPP power and energy part of the incentive scheme. There are three principal dimensions to power The price of energy usuallv is tied to an initial sales agreements: the selling prices for power cost estimate and a series of cost indexation fac- and energy, the amount of power and energy tors. An initial heat rate and initial fuel and oper- sold, and incentives to improve performance and ating and maintenance (O&M) costs are assessed disincentives to ensure that performance does for the plant, together with the appropriate in- not fall below a basic standard. Sales agreements dexes (for example, the consumer price index generally are based on a two-part pricing struc- for O&M costs and the average fuel price index ture-with separate pavments for capacity and of the sector for fuel costs). These determine the energy. But there are substantial differences in energy price. Often. the price is set to just cover the wav they deal with quantities-ranging from such costs. and as long as the indexes track the must-run" or take-or-pay' contracts on the actual costs exactly, there will be no change in plant's entire output to competitive dispatch. The the net revenue per unit of energy supplied. The stronger a sales agreement's guarantee of a mar- energy price is then designed. as in certain U.S. ket for the IPP's output. the more attractive the power pools, so that the IPP is indifferent to IPP becomes for the IPP sponsor and financiers. whether or not it is dispatched. The capacity price but the less pressure is created to generate that is collected because the plant was declared avail- output efficiently and the less competitive pres- able, not because it actuallv ran, and since it eams sure is applied to other generators. The choice no net revenue per unit of energy, there is no of contract structure must therefore take into gain from being dispatched. If strong enough, account the two-sometimes conflicting-aims the profit motive can encourage the firm to try to of attracting private finance and improving sec- "beat the index" in fuel purchases or O&M costs tor efficiency. so that the firm's actual costs rise less than its allowed costs. (The firmn is allowed to keep the The price of capacity in the power sales agree- difference-it is not passed on to the consumer.) ment usually is related to the capacity declared available, rather than to the actual capacity run. The contractual arrangements for determining it is likely to be set so that. at a given level of how much energy and power are sold can varv operation. the discounted revenue from capac- greatlv. The rest of the Note examines their ity payments will cover capital costs over the importance for the sector. life of che project. Contracts tend to set a target level for availability (sav, 80 percent) over the Must-run or take-or-pay contracts year. plus a bonus zone above this availability and a penaltv zone below it. Setting the target The least riskv form of contract for IPPs guar- well below the feasible availability under good antees the sale of a stipulated amount of power operating practices reduces the financing risk and energy for the life of the contract. W'hen of the IPP and thus the incentive for the op- this guarantee covers the entire projected out- erators to be efficient. If the IPP is one of the put of the plant. the 11'1' has an assLired market lowest-cost generators. it should be used as that it cannot lose without compensation. but much as possible. A honus pavment for avail- it also cannot increase its market share. Under abilitv a;bove the target can be used aIS an in- this must-run contract. there is no issue of eco- centive for higher production. Similarly, if the nomic dispatc h for the plant even when other price for capacitv allows an IPP to earn an eco- plants have lower costs. The subsequent entrv nomic return on capital at a capacity utiliza- of additional IPPs, each with a long-term con- tion below the target, penalties are needed to tract, can compound this problem. The pur- ensure that the IPP remains efficient. Recent chaser must pay for any contracted output that U.S. experience with IPPs shows that, in most it does not take from the [PP. This arrangement has three separate effects on a true merit order. andc svstemwicle generation the performance of the sector. First. there is no costs co