WB Note (The Enron Saga: Power Play in India between MSEB and DPC)

INDIA

DABHOL POWER PROJECT

Introduction

1. In July 1992, ENRON power Development Corporation, (ENRON, USA) and the Maharashtra State Electricity Board (MSEB) entered into a Memorandum of Understanding for Enron to set up a 2550 MW power Station to build, own and operate basis, at Dabhol, about 150 miles South of Bombay in the state of Maharashtra. This proposal was subsequently scaled down to 2015 MW and is referred to in this note as the project. The project would run in base load on natural gas, imported from Qatar in the form of Liquefied Natural Gas (LNG). The size of the project in base-load operation, matches output of one LNG production train. A two stage development is envisaged stage 1 involves construction of 690 MW, stage 2 an expansion to about 2015 MW, consisting of 3 X 635 MW combined cycle units and 110 MW additional gas turbine peaking capacity. Stage I is scheduled to be commissioned in the first quarter of 1996 (assuming financing is arranged by the end of 1993), running initially on distillate fuel oil. Stage 2 is scheduled to be commissioned in 1998, if the financing and the contract for the supply of LNG can be finalized by the end of 1993.

Power Purchase Agreement (PPA)

2. The power purchase agreement (PPA) provides for separate capacity and energy charges. MSEB would pay not only the capacity charge but also the energy charge if the generation (due to MSEB's despatch) does not reach the agreed minimum level (currently envisaged by ENRON to be 80-85%). ENRON has indicated that it is prepared to guarantee a 90% availability and its calculations assume dispatch at 90% plant load factor. Since the plant is designed to consume the output of one LNG train and in light of the PPA provisions, MSEB will have no option but to run the plant in base load with minimal variation.

Power Market

3. The addition of a 2015 MW station, in forced base-load operation, would place a significant constraint in MSEB's power system. System simulations show that while the project would provide additional generation to help most increasing loads in peak periods, it would displace lower cost coal-fired generation in the off-peak periods.

4. The Western Region system and India as a whole have surplus generation capacity in off-peak periods (8-10 hours during night time. Although the output of coal-fired stations is reduced during the off-peak periods within occupational constraints, system frequencies commonly exceed 50 Hz off peak, a concrete sign of surplus capacity. The imbalance between base load generation (mainly coal) and peak-to-intermediate load generation (mainly hydro supplemented by natural gas in combined cycle generation) has worsened steadily during the 1980s and is likely to worsen further in the 1990s and possible even beyond. Surplus coal-based power in the off-peak periods is likely to continue in the foreseeable future.

5. Since MSEB's system cannot fully absorb the output of the project in off-peak periods, MSEB have to sell surplus LNG based power to other states. At best, MSEB may be able to sell coal-based power in the Southern Region, at or close to its variable generating cost, in off-peak periods. As a result it would incur financial losses. LNG generation at a variable cost of about paise 150/kwh would displace coal-based power costing paise 30/kwh. This loss, adjusted to economic terms, would have to be included in the economic analysis as a project cost and thus reduce project benefits.

Market for LNG Power In India

6. Natural gas-fired combined cycle stations provide an attractive option for intermediate to base load generation. In the Indian system, given the current and the projected availability of lower cost coal fired base load generation during off-peak periods, combined cycle stations are economically attractive mainly in intermediate load service. The availability of natural gas constraints currently the use of gas for power generation. Imported LNG has a potential market niche supplementing domestic natural gas. Targeting the Dabhol Project to take advantage of this niche would require modifications in the project design and size as well as in the PPA. Given the relative high cost of LNG compared to local natural gas, a comparison between LNG based power and coal based power in intermediate load generation would still be necessary to ensure that this indeed meets the least-cost condition. Due to its higher capital cost, a coal fueled station is generally not competitive in intermediate load service, but the high cost of LNG increases coal's competitiveness.

Macro-economic considerations

7. The implementation of the project would place a significant long-term claim on India's foreign exchange resources. The estimated annual fuel cost is about US $ 500 million, subject to escalation as provided for in the PPA (a fuel cost pass through to MSEB is envisaged). Contractually, capacity payments in foreign currency would start at about US $ 175 million (70% foreign exchange portion only considered here) in 1996 and escalating at 5% per annum, reaching US $ 400 million in 2015.

8. Bank projections indicate that (a) the current account and external reserve positions, and the external debt/expert ratio would improve (b) the debt service ratio would rise initially, but fall to 22% in FY 1998 (compared to 26% in FY 93) and (c) foreign reserves in FY 98 would be at over 10 $ billion or less annual foreign exchange claims are considered affordable by India in this scenario. For example, in year 2000, fuel import for the project would account for about 1.2% of India's merchandise imports, capital service would account for about 1.7 % of India's merchandise exports. There is down side risk of much lower than the projected 5.6% annual GDP growth; and a much less favorable balance of payments and reserve position could emerge.

9. Most critical for the project with a non-tradable output (and with no direct foreign exchange earnings) is the relative growth of India's exports and imports. Bank projections show export growth consistently exceeding the growth of imports, contributing to the relative base case set of indicators outlined above. Import demand depends on the pace of further liberalization, and on the performance of the energy sector, in particular the domestic production of oil and gas. Price of oil is an important variable; a Bank sensitivity analysis shows that a 33% increase could lead into foreign reserves declining to half the projected reserve. While such risks reinforce the need for prudent macro-economic policies and vigorous export promotion, the project would even in that scenario still be considered affordable for India.(1)

10. Assuming that the alternative would be to build additional coal-fired stations using local coal, the outflow would be limited to debt service for foreign borrowing, foreign exchange. Such foreign exchange outflow would be significantly below those of ENRON project (about 30-40% only of the Enron outflow) and would be even lower depending on the extent of domestic manufacturers' participation in the equipment supply. The foreign exchange outflow resulting from the development of coal-fired generation using imported coal would be much closer to the outflows of the ENRON project (the difference would depend on the source of equipment and the possibly different escalation of LNG and coal prices).

11. It has been argued that in the absence of the project, India would not implement the domestic or the imported coal option, due to financing constraints. True additionally of the project is limited to the foreign equity, even as concerns that component, Enron’s partners (GE and Bechtel) might well invest in coal based power projects in India in case the project does not proceed. ENRON is unlikely to participate in coal projects, but may be interested by another gas-based project. The major part of the project financing proposed to be provided by the Bank, US Exim Bank and domestic capital market, should be available to finance reasonable alternative projects, such as a series of 500 MW coal projects using imported coal. Not proceeding with the project may somewhat delay power development, but it is unlikely to cause a lasting setback to India's private power program.

______________________________________________________________________

1. India should according to the considerations, afford to commit to number of power projects of this size or smaller private power projects in the 1990s. The main question then becomes which projects it should pursue and whether LNG base-load power is among the high priority options.

Cost and value of LNG Power

12. ENRON indicated that the total cost of power, in base load generation at 90% plant load, is about US cents 7.0/kwh, of which about US cents 3.3/kwh is for capacity. Assuming industrial sales in areas close to the project location and on additional investment of 20% for transmission and distribution and assuming 10% for system losses, the average cost at the consumer level would at the minimum be US cents 7.4/kwh. This corresponds to about Rs. 2.4/kwh in 1993 prices. These figures do not include import duties and taxes on equipment and LNG. The former has been reduced to 20% in India's FY 94 budget, the later is understood to be at 16%. Using these figures, the resulting retail revenue from LNG power would have to be Rs. 2.4/kwh in 1993 prices, equivalent to Rs. 4.6/kwh in 1998 prices, with no provision for administrative and other MSEB overheads (or cross-subsidization of agricultural and residential consumers, para. 14).

13. The cost of LNG based power is comparable to that of diesel generation and lower than that of diesel pumping. The willingness to pay for electricity by industrial consumers has been estimated at about Rs. 2.2/kwh in the Western region and Rs. 2.4/kwh in Bombay in the second Maharashtra appraisal report, below the cost of LNG power. The Bank's standard project economic analysis concludes that the project is not viable. The project therefore would have to be justified as a special effort to meet the electricity demands of consumers expressing a willingness to pay higher tariffs for guaranteed and/or additional supply at system peak periods. However, it remains to be established that a demand for such power exists.

14. Electricity prices given in 1998 prices above demonstrate the need for substantial tariff adjustments for MSEB. Its current average revenue is about Rs. 1.4/kwh, the revenue from the most lucrative industrial consumers is in the Rs. 2-2.5/kwh range. Prices for industrial consumers would at the minimum have to be doubled in nominal terms to recover the cost of LNG power. This indicates annual increases of about 15-20% against the Bank's 5-6% inflation estimates. It should be noted that to the extent that the industrial customers actually take LNG power. However, the above tariff would not provide for cross-subsidization hence an across the board adjustment instead of a special LNG power tariff for specified consumers would be needed.

Load Forecasts

15. System simulations referred to in the above discussion use CEA load projection for Maharashtra. The projections assume an annual load growth of over 7% in the 1990s and about 6% in the following decade. This projection is higher than the forecast used in the appraisal in the 2nd Maharashtra Power Project. The later was based on Maharashtra's estimated capacity in supply, which due to financial constraints was judged insufficient to meet CEA's forecast.

16. MSEB maintains that CEA's projections is too low, on the basis of pending applications from industries for additional supply not considered in the CEA projection. Such application amount to about 4100 MW. MSEB's assessment is that about 50% of these applications would materialize, thus increasing the peak load by 2100 MW by 1997 at a 77 % load factor. This would absorb all LNG power and eliminate the projected surpluses. Should this load scenario materialize, surpluses would indeed diminish. However, this would not invalidate the conclusion that on LNG based power plant operating in base load is not the least cost option for expanding power supply.

17. The suggested load increase is unproven and the proposed high forecast is not a suitable basis for evaluating the project. Three points warrant attention regarding the proposed high forecast. First, the full additionality of the loads in pending applications is questionable and the resulting growth rate for industrial electricity consumption is unproven. MSEB's industrial sales in FY 93 were about 16400 Gwh and are projected to increase to about 22300 Gwh in the CEA forecast, adding about 6000 Gwh MSEB's proposed additional load to be connected by that year corresponds to 12000 Gwh. Under this assumption industrial load would double, i.e. grow at an average rate of about 20% for 4 years, compared to CEA's already respectable 8% average annual growth.

______________________________________________________________________

2. Given the large size of the project to MSEB system, time allocation of the output to the most lucrative consumers of course somewhat questionable. The locational advantage is clear the project will generate power not far from the main load centre in Maharashtra, the Bombay metropolitan area.

Second, the willingness of the applicants on MSEB's 4100 MW list to pay the above mentioned price, involving more than doubling the current industrial tariffs, has not been tested. Applications may well reflect expectations of the relative low electricity prices continuing. Before a high forecast could be adopted as the basis of the analysis of the new state-wide forecast (involving the review of CEA forecasts and pending applications) would need to be undertaken, taking price explicitly into account in the forecast. Third, MSEB is unable to connect the new load gradually between 1993 and 1997 as it does not have additional generation capacity, The load is unlikely to wait until 1996-1998 and then come in over 1-2 year period; a more gradual increase is likely.

Conclusions :

18 Although the ongoing review of the methodology would require marginal adjustments in some of the assumptions, it is highly unlikely that they would change the results

of the analysis. The main conclusions of the analysis are summarized below :

(a) The project is not a least-cost choice for base load power generation compared to Indian coal and local gas. Even if domestic fuels are not available, imported coal would be the least-cost option for base-load generation for MSEB with current environmental standards. The additional requirement of fuel-gas desulphurization would narrow coal's advantage, but in all likelihood not eliminate it; an environmental premium would be required to close the gap. In addition to lower fuel cost, the coal alternative is more attractive due to its flexibility, capacity can be added in 500 MW units in steps to meet the growing demand instead of a 2,000 MW in 1998 in the project;

(b) The unique features and risks of this LNG-based project (large minimum consumption, dependence on one power generator and on one LNG supplier) need to be considered; they offset LNG's environmental benefits over coal. The LNG price escalation cost is still to be confirmed for the project; international practice is to be the price of LNG to oil or coal price;

(c) ENRON's current proposal would require MSEB to dispatch the plant as a base load unit at 80-85% minimum plant factor( referred to as plant load factor, PLF, in India). This would prevent the operational flexibility of a combined cycle plant,

(d) Stipulations of MSEB power operations indicate that the project would add more capacity than needed to meet the projected load growth in 1998 and would also result in uneconomic plant dispatch (lower variable cost coal power would be replaced by much higher variable cost LNG power). This would adversely affect the economic viability of the project and place a heavy financial burden on the MSEB; and

(e) Substantial adjustment in electricity tariffs would be required to recover cost of the project from the consumers and to safeguard MSEB's financial position. Given the large share of the project in MSEB's power supply in the initial years, across-the-board adjustment would be required. Adjustments limited to special industrial consumer categories would not be sufficient as their capability to continue to cross-subsidize, by paying more than the already high cost of LNG power, would be limited.

April 30,1993.