1. POWER GRIDS AND CDM METHODOLOGIES Workshop for CDM stakeholders The World Bank Buenos Aires December 8, 2004

2. Power Grids and CDM Methodologies u Characteristics of deregulated power markets in LAC l Dispatch and spot market l Contract market l Criteria used for deciding new power plants u Planning and dispatch: case of Chile u What does a CDM project substitutes? Case of Chile u Comments on the CDM baseline methodologies CONTENT

3. Power Grids and CDM Methodologies u Most LAC have developed deregulated Wholesale Electricity Markets -WWM- l Dominican Republic, Guatemala, Nicaragua, Salvador, Panama, Colombia, Ecuador, Peru, Bolivia, Chile, Argentina u Main features: l Competition in generation and supply. Deregulated prices l No central planning n Indicative Planning intended to provide information, only l Open access to the transmission system l Centralized dispatch based on merit order of variable operating costs (most countries) l Spot prices based on the system´s short run marginal costs (most countries) Characteristics of deregulated power markets in LAC

4. Power Grids and CDM Methodologies u Usual practices Characteristics of deregulated power markets in LAC Dispatch and spot prices based on offered prices Colombia, Salvador Dispatch based on merit order and spot prices based on variable costs All others Capacity price separated from energy price All, except Salvador Capacity price based on the annuity of capital of a Gas Turbine (least cost for peaking-reserve) All, except Salvador Supplies mostly based on bilateral contracts and agreed prices Chile, Peru, Colombia, Dominican R., Nicaragua, Guatemala, Panama Supplies mostly based on levelized (6 months) spot prices Argentina, Bolivia, Ecuador

5. Power Grids and CDM Methodologies u Usual practices for contract supplies l Contracts established for 2 to 15 years l Contract prices based on: n First years of the reform: average of expected spot prices n Due to risk of fluctuations, the current practice is to set prices close to a proxy of the system development costs (capital + operation costs of new units) n Equivalent to concept of opportunity costs (costs of competitors) l Indexed with US CPI and fuel prices n In accordance to cost structure Characteristics of deregulated power markets in LAC

6. Power Grids and CDM Methodologies u Usual practices for spot supplies l Gencos sell to “the system” l Discos purchase from the system l Price = average of expected spot prices for following 6 months l Adjustment “ex-post” in accordance with actual spot prices u Links between spot prices (including capacity component) and long term contract prices: l A permanent discrepancy in one direction cannot be sustained l Under equilibrium conditions, both converge l In theory, revenues from the contract market should be closed to those from the spot market l Contracts provide stability in revenues Characteristics of deregulated power markets in LAC

7. Power Grids and CDM Methodologies u A Genco decides to invest: l When the revenues through contracts and/or the spot market covers its capital + operating costs l Spot revenues assessed for a likely expansion sequence n Investments decided under competitive pressure l Likely expansion sequence = such as spot revenues (capacity + energy) provide a given a return on assets l Return depends on the country risk and power sector risk n LAC: 10-14% on assets (real term, after taxes) l The possibility to sign LT contracts sometimes “accelerates” the decision to invest Criteria used for deciding new power plants

8. Power Grids and CDM Methodologies u Practical approach: l 1- Determine a likely expansion sequence n Includes power plants already decided n New power plants added when spot prices get a level such that they provide a given return on assets n The expansion sequence includes the project (several sizes, start-up dates) n In hydro-thermal system, a least cost thermal “base-load” technology is usually selected l 2- Compute spot prices for that expansion sequence l 3- Computes revenues for the project selling P and E into the spot market l 4- Determine if and when revenues pay the project´s capital + operating costs Criteria used for deciding new power plants

9. Power Grids and CDM Methodologies u Spot and contract prices in the central Interconnected System WEM prices in Chile

10. Power Grids and CDM Methodologies u Since the gas interconnection with Argentina: price driver are 370 MW combined cycle units burning NG (base load) u Before: coal units Costs Planning and dispatch: case of Chile Investment192 MM US$ (520 US$/kW) Fixed O&M4 MMUS$/year Fixed gas transport cost17 MMUS$/year Variable non fuel + NG price14 US$/MWh Total average cost (11% return on assets, 85% plant factor) 36 US$/MWh Return from capacity payments11 US$/MWh Required return from energy payments 25 US$/MWh

11. Power Grids and CDM Methodologies l Since new CC units (most economic development units) require 25 US$/MWh, the MW planned must be “below” the existing coal units for dispatch l In order to avoid being the marginal unit Planning and dispatch: case of Chile Hydro reservoir Coal = 28 US$/MWh Gas = 14 US$/MWh Run of River =0 US$/MWh Coal is marginal SRMC= 28 Gas is marginal SRMC= 14

12. Power Grids and CDM Methodologies l Same would happen with new hydro power plants (instead of NG cc units): if too much capacity is added, the SRMC would drop l What does a CDM (base load) project substitute? Coal or NG? apparently coal (and some NG) What does a CDM project substitutes? Case of Chile Hydro reservoir Coal = 28 US$/MWh Gas = 14 US$/MWh Run of River =0 US$/MWh Coal substitution? SRMC= 28 Gas substitution? SRMC= 14

13. Power Grids and CDM Methodologies l In fact substitutes 100% NG since it will displace (or reduce the size) of a new cc unit l The same amount of coal is burned What does a CDM project substitutes? Case of Chile Hydro reservoir Coal = 28 US$/MWh Gas = 14 US$/MWh Run of River =0 US$/MWh Coal substitution SRMC= 28 Gas substitution SRMC= 14 The CDM project substitutes MW in the cc unit

14. Power Grids and CDM Methodologies l Apparently a small CDM project is marginal and might not affect investment decisions n Certainly, a large project close to start construction or under construction will not be postponed. n But “marginal” does not mean “nil” n The impact of even a small project is relevant in the spot market n Example.  A 50 MW CDM base load project (less than 1% of the 7000 MW installed capacity in the Chilean SIC) has an impact of -1.5 US$/MWh in the spot prices if nothing is postponed (4% on prices) n The incumbents perceive an impact of 1.5 x 35 TWh = 52 MMUS$/year if they do nothing n => Very shortly they will postpone projects in order to avoid prices below development costs What does a CDM project substitutes? Case of Chile

15. Power Grids and CDM Methodologies u The “Combined Margin methodology” averages the Operating Margin (OM) and the Build Margin (BM) u We believe that only for very small projects the OM is applicable l Under the condition that the total capacity of all projects remains small u In principle, the appropriate methodology would be to recalculate the expansion sequence with and without the project. u Given the complexity, lack of transparency, gaming, etc., we believe that the BM methodology should have more weight l For base load CDM projects, a good approach seems to estimate the BM through the reduction in the capacity that is likely to be installed as base load (hydro and cc units in the case of Chile) l This is the “proxy plant method” Comments on the CDM baseline methodologies