Eskom s Draft Application to NERSA for Price. Increase in 2008/9

Transcription

1 Eskom s Draft Application to NERSA for Price Increase in 2008/9 Submitted to NERSA on 17 March 2008

2 Table of Contents Executive Summary Background and Introduction Primary Energy Demand Side Management Financial Analysis Implementation of the proposed price increase Analysis of South Africa s historic electricity prices Conclusion

3 List of Figures Figure 1: European coal prices have risen dramatically in Figure 2: Capacity reserve margin Figure 3: Diesel price index between 2001 and 2008 (Source: Stats SA 15 March 2008) Figure 4: Coal component of PPI index between September 2004 and January Figure 5: Coal Volumes per PS Contractual vs Budget for 2008/ Figure 6: Volume contribution of coal contracts Figure 7: Percentage Value contribution of Coal Volumes Figure 8: Scenario 1 Financial Analysis Figure 9: Scenario 4 Financial Analysis Figure 10: Scenario 5 Financial Analysis Figure 11: Debt: Equity scenario analysis on revised price increases Figure 12: Pre-tax Interest Cover scenario analysis on revised price increases Figure 13: Real Price Increase Scenario analysis Figure 14: Average price of electricity in c/kwh scenario analysis Figure 15: Annual Funding Requirements scenario analysis Figure 16: Comparison of Homepower and Homelight 2007/ Figure 17: Comparison of Businessrate 4 and Businessrate / Figure 18: Comparison of CPIX and Eskom s real price increase between 1975 and Figure 19: Eskom s cash reserves had Electricity prices increased with inflation List of Tables Table 1: Summary of adjustments to revenue and prices in 2008/9 and 2009/ Table 2: Range of indicative differential increases based on Scenario 5 (in nominal terms)... 9 Table 3: Range of indicative differential increases based on Scenario 4 (in nominal terms).. 10 Table 4: Comparison of PE, Opex and Capex Actuals vs MYPD allowance Table 5: Five-year Real Capital Expenditure of Eskom regulated entities Table 6: Nominal-year Real Capital Expenditure of Eskom regulated entities Table 7: Variable revenue allowances MYPD Table 8: Fuel cost adjustment - Summary of case studies examined Table 9: US Primary fuel pass-through summary compared to SA Table 10: DSM plan until 2010/ Table 11: Breakdown of DSM costs and MW savings Table 12: Breakdown of accelerated DSM Table 13: DSM Targets and cost comparison of actuals vs MYPD projections Table 14: Range of indicative differential increases based on Scenario 5 (in nominal terms) 54 Table 15: Range of indicative differential increases based Scenario 4 (in nominal terms) Table 16: R/month impact using the 60,87% scenario and options

4 Executive Summary Eskom is in the second year of the MYPD and is facing significant financial challenges for 2008/9, assuming the price increase of 14,2%. The expansion programme will cost R343bn over the next five years. There are further pressures of increased primary energy costs, and the need to reduce consumption through demand side management and power conservation projects. The impact on Eskom s financial sustainability is reinforced by Standard and Poor s (credit rating agency) placing the utility on credit watch. In order to address this crisis, Eskom recommends a revision of the price for 2008/9 from 14,2% to a 53% real increase or a 60% nominal increase. If the status quo remains, the price increase for 2009/10 will be almost 100%. Our analysis reflects an urgent need to double prices over the next two years and thereafter, once they are at acceptable economic levels, prices can escalate marginally above inflation. The financial analysis already assumes a R60bn government loan and aims to maintain Eskom s credit rating to enable the company to fund the build programme. Eskom fully recognises that the proposed increases could have a negative impact on the indigent customers with undesirable social impacts. In order to address this concern Eskom has proposed options to limit the impact of the proposed price increases on indigent customers. Other than Eskom s financial sustainability, the reliability of the electricity system is critical and the Power Conservation Programme (PCP) will help alleviate the tightness of the system. Hence a significant price increase complimented with stringent action e.g. electricity cut-off will stimulate energy efficient consumption behaviour which is fundamental to the country over the next several years. The consequence of a failed conservation programme is unthinkable with rolling interruptions and inconvenience to all. The Multi-Year Price Determination (MYPD) is based on a revenue cap mechanism and does not provide for price and mix variances of primary energy, while marginally catering for volume changes. Eskom had raised this 4

5 as a fundamental concern during the consultation process and public hearings of the first MYPD in February This concern was repeated during the rule change application process in During both processes the Regulator (NERSA) disagreed with Eskom to allow the recovery of prudent primary energy expenditures to the extent that it varied with original forecasts, other than that the regulator confirmed that the pass-through mechanism applicable to the DME OCGT IPP project is applicable to Eskom s own OCGT power stations as well. All costs incurred above that which was allowed by NERSA will impact negatively on Eskom s bottom-line. Due to the quantum involved, Eskom believes that it is necessary to change the current principle as the coal environment in particular is extremely volatile with prices on marginal coal purchases increasing above average coal cost. Prior to embarking on this process, Eskom also obtained a legal opinion on its options, to address the challenge. The opinion obtained states that unless NERSA permits Eskom to increase tariffs quite substantially, Eskom will most likely be unable to meet some of its repayment obligations. According to the Pricing Principles set out in the final determination of the MYPD issued in February 2006, NERSA states: Prices should enable licensees to be viable at an efficient level of costs. To the extent that new investment and capacity is needed that should be factored into prices as and when it is needed in order to keep the industry sustainable. Any smoothing of prices is linked to the need for Eskom to finance its investment programme to meet demand, given the possibility that certain financing ratios might otherwise come under pressure by the end of the control period or shortly thereafter. In addition NERSA states that the price of electricity should not compromise Eskom s ability to deliver on new capacity and reliability programmes. In order to meet increasing demand Eskom has operated more expensive plant during 2006/7 and 2007/8. In addition the price of fuel (mainly coal and liquid fuel) increased significantly compared to original forecasts. This culminated in additional costs of R7,5bn on primary energy compared to MYPD over the first two years. Subsequent to the recent load shedding 5

6 (January 2008), Government and Eskom have agreed to implement a Power Conservation Programme (PCP) which will reduce demand, reduce forced load shedding and lower the load factors of the high-cost-to-operate gas plant and swing coal stations (Majuba and Tutuka). The higher primary energy cost that Eskom had incurred in 2006/7, 2007/8 and projections for 2008/9 will result in the organisation barely breaking even in the current 2007/8 financial year and for 2008/9 could incur a significant loss. The proposed adjustments to 2008/9 will be premised on the following principles 1. Pass-through of prudent primary energy costs 2. Recovery of accelerated DSM costs in 2008/9 The level of adjustments will vary depending on the scenario selected. Clearly any further adjustment above the 14,2% in 2008/9 will contribute towards restoring financial sustainability, lower future price spikes, stimulate power conservation and encourage consumers to become more energy efficient. The tariff structure will also reduce the impact on low income households. In running the five scenarios below, the following common assumptions were made for all the scenarios: A maximum Eskom borrowing of R30bn annually over the next five years Government loan of R60bn, with an assumed phase-in of R6bn (2008/9), R12bn (2009/10), R20bn (2010//11) and R22bn (2011/12). Reduction in electricity demand due to PCP The following additional assumptions were made for each scenario: 1. Base assumes price increase of 9% real (14,2% nominal) in 2008/9, R60bn and primary energy costs pass-through of R2,4bn. Approved by Board on 27 February Primary energy 2008/9 assumes prices of 28% real in 2008/9, and primary energy cost pass-through of R5,3bn in 2008/9. The first two 6

7 years under-recovery of actual costs compared to the MYPD assumption is absorbed by Eskom. 3. Primary energy full 2006/7 to 2008/9 assumes price increase of 47% real in 2008/9, and primary energy cost pass-through of R13bn for the MYPD period. 4. Primary energy 2008/9 + DSM assumes price increase of 33% real in 2008/9, and primary energy pass-through of R5,3bn in 2008/9. The first two years under-recovery of actual costs compared to MYPD assumption is absorbed by Eskom. Accelerated DSM will incur additional costs of R2bn in 2008/9, the recovery of which would result in a further price increase of 5% to the consumer. 5. Primary energy full + DSM full: Assumes price increase of 53% real in 2008/9 and primary energy cost pass-through of R13bn for the MYPD period. Accelerated DSM will incur additional costs of R2,5bn over the MYPD period, the recovery of which would result in a further price increase of 6% to the consumer. 7

8 Based on the various assumptions the following adjustments would be required for 2008/9: Table 1: Summary of adjustments to revenue and prices in 2008/9 and 2009/10 Scenarios 2008/9 Real Price increase 2008/9 Revenues Additional revenues above MYPD Eskom under recovery of PE over MYPD Eskom under recovery of DSM over MYPD 2009/10 Real price increase (Estimate) MYPD 1% R44.5bn R0bn R15.4bn R2.5bn 1. Rule Change 9% R47.9bn R0bn R12.8bn R2.5bn 100% 2. Primary energy 2008/9 27% R53bn R5.3bn R7.5bn R2.5bn 69% 3. Primary energy full MYPD 47% R61bn R12.8bn R0bn R2.5bn 48% 4. Primary energy 2008/9 + DSM 5. Primary energy full + DSM full 33% R55bn R7.3bn R7.5bn R0.5bn 64% 53% R63bn R15.3bn R0bn R0bn 43% Affordability The high electricity price increase will have a definite impact on the economy especially on the indigent. There are a number of options that can be considered to address the impact of these high price increases on the poor. A simple and easily to implement method is applying a lower price increase to tariffs that should receive protection while increasing the other tariffs. A number of different options were modelled to show the impact of applying such a methodology. In all cases the Homelight tariff was given the lowest price increase as this is the tariff targeted at the indigent. With the 60,87% nominal price increase (53% real Scenario 5) The following range of price increases brought the same revenue as requested under scenario 5: 8

9 Table 2: Range of indicative differential increases based on Scenario 5 (in nominal terms) Option Homelight Businessrate and Homepower All other tariffs 1 35% 89% 61,9% 2 45% 70% 61,9% 3 30% 70% 63% 4 10% 100% 63,5% 5 26% 63,5% 63,5% Due to the small volumes on Homepower and Businessrate (1,6% of Eskom sales volumes)., the price increase to be applied to these tariffs has to be significantly higher than the 60.87%. Option 1 places a greater burden on Homepower and Businessrate. This Option also results in a 1% higher increase to other tariffs. Option 2 gives a slightly higher increase to Homelight than Option 1 resulting in a lower increase to Homepower and Businessrate. Option 3 explores a higher price increase to other tariffs Option 4 gives the lowest price increase to Homelight, but results in a very high increase to Homepower and Businessrate. Option 5 gives the same higher increase to Homepower, Businessrate and all other tariffs and a lower increase to Homelight. At an increase of 60,87%, a Homelight 20A customer using 100 kwh per month will pay R27,56 more per month and at an increase of 30%, R13,58 more per month (excluding VAT). For Homepower customers, in the worst case scenario ie the 100% increase, they will pay using 1000 kwh per month R more per month and at 70% R257,34 more per month. All options modelled increased the breakeven between Homelight and Homepower and may encourage customers to convert from Homepower to Homelight. Any decision by NERSA would have to allow for mitigation of this risk. 9

10 With the 39,61% nominal price increase (33% real Scenario 4) The following range of price increases brought the same revenue as requested under Scenario 4: Table 3: Range of indicative differential increases based on Scenario 4 (in nominal terms) Homelight Businessrate All other tariffs and Homepower 1 22% 70% 40% 2 28% 55% 40% 3 15% 55% 41% 4 10% 55% 41,4% 5 12% 41,7% 41,7% Eskom was also requested to look at the inclining block rate structure to achieve the above objective. With an inclining block rate the tariffs become significantly higher above a certain consumption level. The logic usually used for such a tariff is that inclining block rates would provide an economic incentive for customers to reduce their energy use and help motivate customers to reduce their bills, reduce demand growth and defer the need for new generation. There are a number of challenges associated with an inclining block rate to be used address the issue of affordability. These are: It does not reflect costs and gives no signal for peak usage and usage patterns no demand side management (efficiency) signal. To try use it to address poverty relief (through increasing subsidies) and energy efficiency (will reduce contribution to subsidies) are conflicting objectives. An inclining block rate tariff cannot be implemented currently to Eskom s prepayment customers as the meters do not allow for it. The level of sales to the Homepower and Businessrate tariffs are too low to provide sufficient revenue to provide a cross-subsidy to Homelight using an inclining block rate. 10

11 Based on international research there is considerable uncertainty over the likely reaction of customers to these tariffs will they respond at all and will the response be as intended and therefore, will the level of cross-subsidy be sustained as intended. There needs to be an accurate determination of the knee-point where the rate increases. MFMA submission to Parliament In terms of Section 42 of the Municipal Finance Management Act the 15,02% increase which has already been approved by NERSA was going to be implemented on 1 July 2008 for local authorities if tabled in Parliament on/or before the 15 March The Minister of Finance has since granted the extension to 15 April 2008 and Eskom proposes that the revised 2008/9 increase be tabled by 15 April Eskom proposes that the non-municipal increase be effective from 1 April 2008, and the revised municipal tariff be implemented on 1 July, taking into account that the increase is spread over 9 months compared to 12 months for the rest of the customer base. The NERSA decision for 2008/9 gave municipal tariffs a 15,02% increase i.e. a 1,18% higher than non-municipal tariffs. Table 4: Impact of a later price increase on Eskom s revenue The lower price increase using the NERSA approach results in a R2,2bn shortfall on the revenue requirement using the 60,87% scenario and R1,4bn using the 39,61% scenario. 11

12 Recommendation To be consistent with the principle that the tariffs should be reflective of prudent costs it is essential that prices be adjusted to reflect the cost of supply. Should this not be the case, then the under-recovery will need to be borne by the taxpayer, thereby further distorting the price levels. Eskom recommends the implementation of scenario 5, with a 53% and 43% real price increase in 2008/9 and 2009/10 respectively. This is based on full pass-through of prudent primary energy and DSM cost variances as compared to the revised MYPD for all three years of the MYPD. 1 Background and Introduction Multi-Year Price Determination Eskom is currently completing the second year of the first MYPD announced in February The MYPD spans over a three year period from 1 April 2006 to 31 March The decision of the MYPD by NERSA resulted in price increases of CPIX+1% being awarded to Eskom for each year of the MYPD. According to the Pricing Principles set out in the final determination of the MYPD issued in February 2006, NERSA states: Prices should enable licensees to be viable at an efficient level of costs. To the extent that new investment and capacity is needed that should be factored into prices as and when it is needed in order to keep the industry sustainable. Any smoothing of prices is linked to the need for Eskom to finance its investment programme to meet demand given the possibility that certain financing ratios might otherwise come under pressure by the end of the control period or shortly thereafter. In addition NERSA states that the price of electricity should not compromise Eskom s ability to deliver on new capacity and reliability programmes. 12

13 A fundamental principle in MYPD was that the primary energy costs were not eligible for pass-through of variances compared to original forecasts. However, Eskom disagreed with the then NER because of the volatility and unpredictability of that category of costs, with many components being outside the control of the organisations (for example, increases in steel prices, diesel costs and coal prices escalating beyond original estimates as a result of market dynamics). MYPD Rule Changes Application Eskom incurred primary energy costs that exceeded the amount allowed by NERSA by R1,8bn in the 2006/7 financial year. Eskom subsequently approached NERSA in September 2006, to discuss the escalations in primary energy and the impact of higher costs on the utility s financial sustainability. These discussions resulted in a formal application in April 2007, in which Eskom applied to NERSA for a rule change to MYPD which comprised of: 1. Primary Energy pass-through of prudent costs 2. Accelerated capital expenditure allowance for cost changes caused by both purchase price movements and project phasing changes during the build programme 3. Trigger for reopening of determination - based on a net Earnings band rather than purely on variance in revenue The result of the rule changes application would have equated to an 18.7% nominal price increase for 2008/9, based on the latest forecast of costs for 2008/9 that was available at 30 April At the time Eskom applied for the rule changes, it was highlighted that the forecasts for 2008/9 would likely change as the organisation was in the process of finalising their 2008/9 to 2012/13 five year financial planning process. This new five year plan will be used as the basis for the MYPD 2 submission in Primary energy cost is the main driver on Eskom s application due to changes in PE cost estimates from MYPD1 planning time to current plans. These costs have (in Eskom s estimates) increased by far more than was predicted with the MYPD1 application. The combined change is above R12 bn for the 3 years of MYPD1. 13

14 The 2008 financial plan was approved by the Eskom board in February 2008 and the primary energy costs were higher than the forecast of the 2007 plan. The result of the higher primary energy costs for 2008/9 with 14,2% increase would be a significant loss before tax. On the 14 December 2007, Eskom had engaged with NERSA to explain the consequences of the 14,2% price increase. The 2008 plan illustrates that even if Eskom was granted the 18,7% price increase, the organisation would still show a loss in the 2008/9 financial year. Eskom s application was based on the financial plan of 2006, which had the R150bn that was approved by the Board in February As stated during the rule change application we highlighted the risk of costs increasing as Eskom was conducting their Financial Plan 2007 process. The Financial Plan 2008 comprises R300bn (real rands) capex or R343bn (nominal rands). Another fundamental assumption is that it assumes 3000MWs reduction for PCP. Hence even if NERSA granted Eskom 18% price increase, the organisation will still incur a significant loss in 2008/9 as a result of the greater primary energy and DSM costs. This clearly demonstrates the need to move away from a cap on primary energy and cater for flexibility in terms of the rules to allow for prudent pass-through of these costs. In the April 2007 application, Eskom had requested three changes for the MYPD rules, however this application is only for the rule change to primary energy. Eskom accepts that NERSA would review the other two rule changes applied for in the second MYPD application. During 2007 NERSA completed their process of stakeholder engagements which culminated in their decision on 20 December 2007 comprising the following: o Decision on rules requested by Eskom was postponed to the MYPD 2 process o NERSA would provide Eskom with an ad-hoc adjustment of 9% real (14,2% nominal) for 2008/9. 14

15 In its rule change application Eskom indicated that it was willing to absorb the projected R6bn of higher primary energy costs for 2006/7 and 2007/8 projections, but requested recovery of all PE costs in 2008/9. The latest estimates show a total exposure of R12,8bn over the MYPD period as follows: R7,5bn for the first two years and R5,3bn for 2008/9. During the rule change process, Eskom highlighted to NERSA the costs of the accelerated 3000MWs DSM programme which would result in additional costs in excess of R2,5bn being incurred over the MYPD period. The costs for DSM in 2008/9 at the time of the rule change application was forecast at R1,8bn, however the 2008 plan indicates a cost of R2bn in 2008/9 due to the accelerated DSM programme. According to MYPD, the additional DSM costs will be recovered in MYPD 2. The latest estimate of the additional cost impact of the (further) accelerated DSM programme over the MYPD period is R2,5bn. Due to cash flow constraints it is necessary that Eskom recovers these additional DSM costs in 2008/9. Current Financial Plan 2007/8 During the current financial plan process, Eskom s forecasts have reflected further upward movements in primary energy costs which will have a direct impact on the bottom-line. Besides the escalating higher coal costs, Eskom decided to improve the level of stock days to its target which places additional price pressures on primary energy. 15

16 Coal (Quarter ahead excluding freight) Rising coal prices will have a significant effect upon generators, power consumers and the fuel mix given movements in the EU-ETS price in Q if this trend continues. Coal price ($/Metric tonne) ` Q4 has seen a massive increase in the price of coal, feeding through to generation costs 0 Mar-06 Apr-06 May-06 Jun-06 Jul-06 Aug-06 Sep-06 Oct-06 Nov-06 Dec-06 Jan-07 Feb-07 Mar-07 Apr-07 May-07 Jun-07 Jul-07 Aug-07 Sep-07 Oct-07 Nov-07 Month-Year Figure 1: European coal prices have risen dramatically in 2007 Source: Datamonitor, Spectron Figure 1 highlights the rise in European coal prices in Due to the steep change in primary energy and the capital expansion programme from R150bn to R301bn (R343bn nominal), the financial health of Eskom has come under tremendous pressures which resulted in Standard and Poor s (credit rating agency) placing Eskom on credit watch. The funding of the build programme in the short term would require a combination of Government support and significant price increases. In the 2008 Budget speech by Minister Manuel, National Treasury had announced a loan limited to a maximum of R60bn over the next five years. This leaves electricity prices to fill the gap as the only remaining material funding source. MYPD Performance Table 5 reflects Eskom s performance against the MYPD decision. Table 5: Comparison of PE, Opex and Capex Actuals vs MYPD allowance 2006/7 2007/8 2008/9 Rule MYPD Actual Var. MYPD Proj. Var. Budget Change Var. Cumulative Variance Primary Energy (R m) 11,117 13,039 1,92212,694 18,847 6,153 17,450 23,052 5,602 13,677 16

17 Opex (incl. depreciation) (R m) 21,974 19,785 23,875 23, ,394 27,546 2, ,189 Capex (R m) 14,992 17,862 2,87015,512 24,685 9,173 35,348 45,996 10,648 22,691 Note: Opex is limited to MYPD decision. The primary energy costs reflect a cumulative over expenditure of R13,7bn. Eskom is within the Opex allowance for MYPD, on the assumption of DSM being limited to R440m in 2008/9. Over the first two years the organisation has already spent R625m and R700m respectively - already more than the R1,3bn as originally assumed in total for DSM over the entire three year period. Capital Expenditure 2009~2013 (2007/8 Constant Rand Millions) In order for Eskom to meet the increasing demand and improve its capacity reserve margin, a number of supply options have to be undertaken. Due to international demand and supply constraints on new power plant construction, these plants costs have escalated substantially and have filtered through to Eskom s financial plans. Table 6: Five-year Real Capital Expenditure of Eskom regulated entities 2008/9 2009/ / / /13 Total Generation 32,002 56,007 48,875 42,024 40, ,436 Transmission 4,832 9,614 10,717 8,150 4,942 38,255 Distribution 5,811 6,054 6,550 6,715 7,169 32,299 Corporate 1,329 2,985 4,393 2, ,320 Net regulated total 43,973 74,660 70,535 59,331 52, ,310 Note: Electrification capital expenditure is funded by DME and is excluded from Eskom Holdings capex plan 17

18 Capital Expenditure 2009~2013 (Nominal Rand Millions) Table 7: Nominal-year Real Capital Expenditure of Eskom regulated entities 2008/9 2009/ / / /13 Total Generation 33,474 61,220 55,561 49,684 49, ,770 Transmission 5,054 10,509 12,183 9,636 6,077 43,459 Distribution 6,078 6,617 7,446 7,939 8,815 36,895 Corporate 1,390 3,263 4,994 2, ,744 Net regulated total 45,996 81,609 80,184 70,145 64, ,868 Note: Electrification capital expenditure is funded by DME and is excluded from Eskom Holdings capex plan Eskom was granted a real return on assets of 7,3% during MYPD. The results reflect an average historic return of 5% in 2006/7, and Eskom is currently projecting a 1% average historic return for the current year 2007/8, and a forecasted return of -3% for 2008/9. Eskom made a profit of R4bn after interest before tax in 2006/7, and the current year forecast shows a breakeven for 2007/8 and a projected significant loss in 2008/9, if the status quo of 9% real (14,2% nominal) remains without any further pass-through of primary energy costs. In order to improve and restore financial sustainability for Eskom, an allowance for pass-through of prudent on primary energy cost and allowing for a higher return on assets will result in positive signals being given to private investors to enter the industry in the near future. Power Conservation Programme In order to address the supply and demand challenges facing the industry a power conservation programme was announced, aimed at reducing demand by 3000MWs which could result in less scheduled load shedding and lower primary energy costs. PCP will also allow for more opportunities to do planned maintenance at power stations and improvement in reserve margin. Other initiatives to reduce demand revolve around the accelerated DSM plans to 18

19 add a further 815MWs in 2008/9. Significant tariff increases will definitely trigger a voluntary change in consumption patterns. Figure 2 shows that the all supply projects are required in the short term to increase the reserve margin to 10%. It also illustrates the benefit of having PCP which will push up the reserve margin to 20% assuming same build programme. Finally the graph only includes projects which have cash flow impacts over next 5 years. 30% 20% 10% 0% -10% '2006/07 '2007/08 '2008/09 '2009/10 '2010/11 '2011/12 '2012/13 '2013/14 '2014/15 '2015/16 '2016/17 '2017/18 '2018/19 '2019/20 '2020/21 '2021/22 '2022/23 '2023/24 '2024/25 '2025/26-20% -30% -40% -50% Reserve margin (Zero build) Reserve margin incl RTS Reserve margin incl IPP Reserve margin incl M edupi Reserve margin incl Bravo Reserve margin incl Ingula Reserve margin incl Steelport Reserve margin incl Nuclear 1 Reserve margin incl DSM Reserve margin incl PS 1 Reserve margin incl GAS 1 Reserve margin incl Wind 1&2 Reserve margin incl Co-generation (1100MW) Reserve margin incl OCGTs to CCGTs (400M W) Target PCP Impact Figure 2: Capacity reserve margin 19

20 2 Primary Energy Regulatory Treatment of Fuel Cost Variance In considering regulatory treatment of fuel cost variance the key areas of uncertainty are often set out in terms of: o Uncertain fuel price as measured on a unit basis, such as Rand per delivered ton of coal or rand per liter of kerosene). o Uncertain fuel mix taking into consideration that at the margin, various types of generating units are dispatched, each having different fuel costs (i.e. coal, gas, diesel, and kerosene.) o Uncertain energy demand leading to an incremental change in overall use of fuel for power generation As discussed below, the MYPD currently addresses uncertainty in energy demand, but does not account for uncertainty in fuel price or fuel mix. This means that variance in fuel price and fuel mix over the three year MYPD could lead to a windfall loss or gain to Eskom or its customers, and is not a desirable methodology. This is based on our experience since the MYPD was initiated. To further highlight the symmetry of this problem, note that (hypothetically) if Eskom had forecast 20% increases in fuel costs into its revenue application and energy prices than unexpectedly dropped during the following three years customers would have been equally disadvantaged under the current MYPD rules. Accounting for fuel cost variance under the current regulatory approach Under the MYPD as it currently stands, only a small proportion of overall fuel cost variance is accounted for via the correction factor. The correction factor provides for a revenue adjustment for each incremental kwh generated above 205 GWh per year 1 - to be added to Generation Division s revenue allowance, with a small additional factor for Distribution Division. 1 More specifically, energy generated is to be measured at the Transmission boundary under the rules set out under the MYPD. 20

21 The unit increases for each incremental kwh generated as set out under the MYPD are shown below 2. Table 8: Variable revenue allowances MYPD Cents per additional kwh (05/06 prices) Generation variable revenue Distribution variable revenue The simple correction mechanism (as defined in the MYPD) applied to Eskom s fuel costs is deficient in a number of aspects. First, the unit allowances (cents per kwh) are fixed through to 2009, even though it has become apparent that these rates no longer reflect Eskom s actual incremental cost of generation. Furthermore, the correction factor only addresses volume variance, but as shown previously in this document, volume variance accounts for only a small proportion of total variance in Primary Energy costs with price variance being the most significant factor. In regard to price variance, we do wish to note that the MYPD revenue allowance (thus Primary Energy cost allowance) is set in real terms, to be updated for actual inflation as defined by the CPIX. This would tend to provide some limited additional recovery of actual costs - to the degree that CPIX adequately tracks fuel cost inflation. However, CPIX is currently running at roughly 5%, whereas the PPI is increasing at 10% and SA coal mining costs at 18,2%. This CPIX adjustment factor for 2006/07 (as applied to the Primary Energy component of Eskom s revenue allowance, and depending on final CPIX inflation for the financial year) will likely amount to somewhere between R30 and R100 million 3. 2 Sourced from NER Multi-Year Price Determination of Eskom 1April 2006 to 31 March 2009 Draft Determination: Conditions, Methodology and Mechanism, December Eskom s 2006/07 initial revenue allowance and tariffs for 2006/07 were based n an assumed CPIX of 4,1% for the year. The MYPD provides that actual (ex post) CPIX will be used to reconcile any variance and that amount would be part of the correction balance to be brought forward to the next year s revenue allowance. The R30 million figure is based on 4,4% CPIX inflation, whereas the R100 million is roughly in line with 5,1% CPIX for the year. 21

22 Alternative approaches to fuel cost adjustment The treatment of fuel cost recovery within a regulated market environment is a crucial component of tariff setting. The key issue here is in dealing with ex post (actual) fuel costs within a cost of service regulatory regime where tariffs are typically based on ex ante forecasts of expenditures. There are a number of ways in which this matter can be dealt with one extreme being where there is no fuel cost adjustment provided for within (or even across) a regulatory period; to full pass-through of actual expenditures on primary energy where shown to be purchased in a prudent manner and where dispatch of various types of generating units is demonstrated to be done in an economically efficient manner. To identify the various practical options at hand, Eskom engaged sleconomics to examine how this problem has been dealt with in other countries in terms of: o The broad objectives considered in design of fuel cost adjustment mechanisms. o The frameworks used in regulatory review of unanticipated costs. o The administrative requirements as applied to a multi-year price determination. The sleconomics review 4 found that there appears to be broad consistency among the regulators canvassed in that the way in which fuel costs are recovered should: o Ensure that customers pay no more than the reasonable cost of supply. o Signal the true cost of supply via prices to consumers so as to promote efficient demand response. o Efficiently allocate risks between the utility and customers. o Provide a fair return so that future investment in generating capacity is facilitated. Furthermore, the review found that fuel costs are not generally seen as well suited to incentive based (fixed) cost allowances such as currently applied in the MYPD where variance in fuel costs are largely driven by external factors, and 4 sleconomics, Fuel Cost Adjustment International Experience, Final Report. December

23 difficult to predict one or more years in advance. In this case, a pre-set cost allowance does little to incentivise efficient fuel purchase by the utility, and simply lead to windfall gains or losses to the utility and customers. The broad trend has been to apply robust fuel cost adjustment mechanisms allowing for recovery of reasonable costs of supply. This is in recognition that, at the margin, fuel costs are to a large degree driven by external factors, and often diverge from forecasts made at the on-set of a tariff control. This is particularly important in a multi-year price control where there would not otherwise be the opportunity to re-set expenditure allowances. There is, however, the clear recognition that the utility must be accountable in regard to maintaining efficient cost of supply. In light of this, periodic reasonableness reviews (for example, on a yearly basis) are undertaken to establish the allowed adjustment in fuel costs to be reconciled against the next year s revenue allowance. These reasonableness reviews tend to focus on efficient fuel procurement and efficient dispatch of generating units, with comprehensive reporting requirements placed on the utility. A summary of the way in which fuel cost variation is treated in the case studies examined are provided below. 23

24 Table 9: Fuel cost adjustment - Summary of case studies examined Jurisdiction State of Andhra Pradesh (India) Treatment of variation in fuel costs Allows for recovery of actual fuel costs through a dynamic adjustment formula accounting for variation in the price and mix of fuels purchased. Adjustments are made on a quarterly basis. State of Gujarat (India) Allows for recovery of actual fuel costs through a dynamic adjustment formula accounting for variation in the price and mix of fuels purchased. Adjustments are made on an annual basis. State of Maharashtra (India) Fuel cost adjustment formula based on monthly change in fuel prices. Fuel mix is based on normative values determined by the regulator (we do not know if these normative values are fixed or vary with actuals). China Provides a mechanism linking coal and power prices whereby if the price of coal increases by more than 5 per cent in a six-month period, electricity prices can be adjusted. Under this mechanism, 70 per cent of coal price increases are transferred to end-users. Power generation firms bear the remaining 30 per cent. Philippines Provides for a Generation Rate Adjustment Mechanism which allows for the utility to petition for re-set of Generation Rates to reflect changes in actual fuel and purchased power costs where shown by the utility to be prudent. Turkey Does not provide for automatic fuel cost adjustment in end user tariffs. Wholesale tariffs are fixed and there appears to be no automatic or retrospective balancing in regard to variation (increases) in fuel costs. 24

25 Jurisdiction Ireland Treatment of variation in fuel costs Provides for a dynamic fuel cost adjustment. o o Fuel prices are adjusted on a monthly basis as input to a fuel index. Fuel mix is adjusted on an annual basis by updating the weights in the fuel index. Reconciliation of accounts on an annual basis. State of California (U.S.) A periodic review is undertaken in which actual primary energy costs are recovered subject to a reasonableness review by the regulator. Fuel costs are recovered at rates calculated on a daily basis (price and fuel mix). State of Texas (U.S.) An ex post review of actual costs is carried out on a periodic basis allowing for full pass-through of variations where proven to be reasonable and necessary: o o Changing fuel mix is allowed for subject to the reasonableness of dispatch. Fuel prices are allowed where demonstrated to have been procured on an arm s length basis. The international case studies examined suggests that fuel cost adjustment mechanisms based on reasonable actual cost of supply have been implemented by regulators and have been found to provide a fair and appropriate balance of risk between the utility and its customers. This is particularly the case where generation is regulated in a multi-year basis 5. Further examples of pass-through in the USA are highlighted in Table 10 5 We also wish to point out that the more static incentive based regulatory frameworks (fixed revenue or price cap regimes) are generally better suited to unbundled network businesses where costs are relatively easier to forecast over periods of three to five years, and where the industry is relatively dormant in term of demand growth and capacity expansion. 25

26 Table 10: US Primary fuel pass-through summary compared to SA US Primary Fuel Pass-through Summary Fuel Mix State Name Type of Regulation Primary Fuel Pass Through Mechanism Primary Fuel South Africa Cost of Service No Coal 92% 0% 6% 0% 0% 0% 0% 0% 2% West Virginia Cost of Service Yes Coal 98% 0% 0% 2% 0% 0% 0% 0% 0% Indiana Cost of Service Yes Coal 95% 2% 0% 0% 0% 0% 2% 0% 0% Wyoming Cost of Service Yes Coal 94% 1% 0% 2% 2% 0% 1% 0% 0% North Dakota Cost of Service Yes Coal 94% 0% 0% 5% 1% 0% 0% 0% 0% Kentucky Cost of Service Yes Coal 92% 1% 0% 3% 0% 0% 0% 3% 0% Utah Cost of Service No Coal 89% 8% 0% 2% 0% 0% 0% 0% 0% Missouri Cost of Service Yes Coal 84% 4% 11% 0% 0% 0% 0% 0% 0% Iowa Cost of Service Yes Coal 76% 5% 11% 2% 5% 0% 0% 0% 0% Kansas Cost of Service Yes Coal 73% 4% 21% 0% 2% 0% 0% 0% 0% Colorado Cost of Service Yes Coal 72% 23% 0% 4% 2% 0% 0% 0% 0% Nebraska Cost of Service Yes Coal 65% 2% 28% 3% 1% 0% 0% 0% 0% Wisconsin Cost of Service Yes Coal 65% 9% 20% 3% 2% 0% 0% 1% 0% Tennessee Cost of Service Yes Coal 65% 1% 26% 8% 1% 0% 0% 0% - 1% Georgia Cost of Service Yes Coal 63% 9% 23% 2% 2% 0% 0% 1% 0% Minnesota Cost of Service Yes Coal 62% 5% 25% 1% 6% 1% 0% 1% 0% North Carolina Cost of Service Yes Coal 60% 3% 32% 3% 1% 0% 0% 0% 0% Alabama Cost of Service Yes Coal 55% 14% 23% 5% 3% 0% 0% 0% 0% Oklahoma Cost of Service Yes Coal 50% 47% 0% 1% 3% 0% 0% 0% 0% Mississippi Cost of Service Yes Coal 39% 34% 23% 0% 3% 0% 0% 1% 0% Florida Cost of Service Yes Gas 29% 43% 14% 0% 2% 1% 0% 10 % 0% Louisiana Cost of Service Yes Gas 27% 45% 18% 1% 3% 2% 3% 2% 0% Alaska Cost of Service Yes Gas 9% 61% 0% 18 % 0% 0% 0% 12 % 0% 48 South Dakota Cost of Service Yes Hydro 46% 4% 0% % 2% 0% 0% 0% 0% Washington Cost of Service Yes Hydro 6% 7% 9% 76 % 2% 0% 0% 0% 0% 84 Idaho Cost of Service Yes Hydro 1% 10% 0% % 5% 1% 0% 0% 0% South Carolina Cost of Service Yes Nuclear 40% 6% 51% 2% 2% 0% 0% 0% -1% Vermont Cost of Service No Nuclear 0% 0% 72% 21 % 6% 0% 0% 0% 0% Hawaii Cost of Service Yes Oil 13% 0% 0% 1% 5% 2% 0% 78 % 0% CRA International March 15, 2008 Coal Natural Gas Nuclear Hydro Other Renewables Other Other Gases Oil PS 26

27 With the above in mind, Eskom is of the view that international experience provides a sound basis in which to design and administer a more robust fuel adjustment mechanism within the broad framework of the MYPD. A broad proposal is provided in the following section. Proposed rule change The proposed rule change for Primary Energy Cost Adjustment is as follows: 1. Prior to each MYPD, Eskom will submit a forecast of Primary Energy costs as part of its MYPD revenue application (as is currently done). 2. Prior to the end of each financial year during the MYPD (e.g. after ten months of a given financial year), Eskom will provide NERSA with a report of actual Year-to-Date Primary Energy costs and with forecasted costs for that full year. Forecasted variances against original forecast costs for that year will be rolled in to the revenue allowance (with corresponding adjustments to WEPS) for the following year (positive or negative amounts). 3. At the same time, prior to the end of each financial year, the Primary Energy cost forecast for the following year will be updated as well. This updated forecast will be reviewed by NERSA and once approved will form the base for the following year s revenue allowance (with the variance component from the previous year (as per par 2 above) added or subtracted from this updated base amount). 4. Noting that Eskom must provide NERSA with adjusted WEPS tariffs for the following year prior to obtaining full year actuals for the preceding year, the variance against forecast for that preceding year will be based on 10 months actuals (to February of that year) and two months projections to year end Any error stemming from the last two month projections against full year end actuals will be reconciled as part of the next year s adjustment. To ensure that only efficient costs of supply are recovered by Eskom under the adjustment mechanism: 5. Eskom will provide NERSA with a final (independently audited) reconciliation report once full year end accounts are available, including an explanation for the adjustment in terms of price, volume and mix factors being different from forecast. 6. NERSA will have the discretion to review the audited reports and disallow cost adjustments if they are shown to be unreasonable. Disallowed cost 27

28 adjustments would be subtracted from the following year s revenue allowance NERSA will review existing and new fuel purchase costs to signal to Eskom if/where there are any aspects of the contracts that would lead to a disallowance of price increases NERSA will review any other aspects of variance in Primary Energy costs it deems relevant in review of the reasonableness of expenditure. The annual audit of Primary Energy costs would ensure that only efficient costs of supply are passed through to customers. Furthermore, the mechanism is symmetric, so that if fuel costs decrease in the future, this would also be reflected in the adjustment mechanism to the benefit of customers. Under this approach, there would not be windfall gains or losses to Eskom or its customers simply due to the unpredictability of fuel costs. This modification to the MYPD could follow existing administrative processes as Eskom in any case must provide an annual revenue and WEPS application to the NERSA taking into account existing correction factors. However, we do note that this rule change will lead to some additional administrative costs but we believe that the financial significance of this matter clearly warrants a more rigorous approach than is currently applied. Increase in Primary Energy Costs The Increase in Coal Costs is attributed to the following: Sharp increase in price of short term contract coal Increasing volumes of short term contract coal purchases Labour cost increases are significantly higher than inflation due to skills shortage Sharp increase in logistic costs Decrease in coal qualities has resulted in greater volumes of coal being burnt in order to generate the same quantity of electricity Improvement of future coal qualities has increased costs in the following manner Beneficiation costs would increase the costs on short term contracts 28

29 Yield loss of approximately 20% of production on the short term contracts Yield loss replaced with short term contract coal Reduced investment in new coal capacity with the change in the mining legislation to the Minerals, Resources and Petroleum Development Act (MRPDA) has caused a coal supply constraint. Diesel Basic Price From the September 2005 (MYPD 1 submission) to January 2008 the basic diesel price has moved from 349c to 517c, an increase of approximately 50%. This is an important factor as our imported coal attracts transport costs that literally double the price of coal delivered to the power stations ` 100 Jan-04 Mar-04 May-04 Jul-04 Sep-04 Nov-04 Jan-05 Mar-05 May-05 Jul-05 Sep-05 Nov-05 Jan-06 Mar-06 May-06 Jul-06 Sep-06 Nov-06 Jan-07 Mar-07 May-07 Jul-07 Sep-07 Nov-07 Jan-08 Diesel Oil CPIX PPI Figure 3: Diesel price index between 2001 and 2008 (Source: Stats SA 15 March 2008) 29

30 PPI Coal From the September 2005 (MYPD 1 submission) to January 2008 the index moved up from 194 to 259, or 29% ` Jan-04 Mar-04 May-04 Jul-04 Sep-04 Nov-04 Jan-05 Mar-05 May-05 Jul-05 Sep-05 Nov-05 Jan-06 Mar-06 May-06 Jul-06 Sep-06 Nov-06 Jan-07 Mar-07 May-07 Jul-07 Sep-07 Nov-07 Jan-08 Coal - PPI CPIX PPI Figure 4: Coal component of PPI index between September 2004 and January 2008 (Source SA 15 March 2008) 30

31 The graph below details the performance of the current long term (cost plus and fixed price agreements) against the contractual agreements. Except for Tutuka and Arnot power stations the volumes extracted from the mines will exceed their contracted volumes in 2008/9. Tons (000's) 20,000 15,000 10,000 5,000 - Arnot Kriel Lethabo Hendrina Tutuka Matla Duvha Kendal Matimba Contractual tonnage Budget Figure 5: Coal Volumes per PS Contractual vs Budget for 2008/9 Figure 5 depicts the amount of coal acquired from the different coal contracts. The short term contracts increase and will average approximately 30% over the next 5 years. Due to the increasing demand in electricity, incremental volumes are delivered by the more expensive stations which are linked to short term contracts. These would include RTS stations and swing stations Tutuka and Majuba. Tonnage Contribution 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Cost Plus Fixed Price Short Term Contracts Figure 6: Volume contribution of coal contracts 31

32 Value Contribution 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Cost Plus Fixed Price Short Term Contracts Figure 7: Percentage Value contribution of Coal Volumes Figure 7 details the cost impact to Eskom as a result of increasing the coal purchases on short term contracts. The graph also depicts the increased cost differential for purchasing coal on the short term contracts when comparing the different contracts to Figure 6. With the assistance of Power Conservation Plan (PCP), primary energy cost increases in 2012/13 can be reduced by at by as much as 25%~40% subject to the rules and the requirements of the PCP. 32

33 3 Demand Side Management Accelerated DSM strategy Since the beginning of 2008, South Africa experienced an unprecedented level of electricity supply shortages. At the time Eskom had already committed itself to an aggressive DSM target. Eskom s 2007 financial plan for five years included a DSM target of 3000MW. Looking into the future of electricity demand, DSM is a critical tool to ensure that Eskom has sufficient capacity to supply the required demand. In order to achieve the required DSM target, Eskom has developed an implementation plan that is split into two phases. The first phase is from January 2008 until December The second phase is from December 2008 into the future. The first phase is the backbone of the achievable target. The second phase is supplementary to the first phase. The implementation plan will accelerate the existing programme to ensure delivery of approximately 347MW by the end of December The existing initiatives will deliver a further 58MW. 33

34 Table 11: DSM plan until 2010/11 DEMAND SIDE MANAGEMENT OPERATING EXPENDITURE AND 3 YEAR PLAN PROGRAMME AREA 2006/07 Actual 2007/08 Projection 2008/09 Budget Proposal PLAN Previous Year Current Year TOTAL 12 Months 2009/ /11 Rm MW Rm MW Rm MW Rm MW Rm MW TOTAL INCOME TOTAL OPERATING EXPENSES RESIDENTIAL PROGRAMMES Efficient Lighting (CFLs) Solar Water Heating Shower Aerators Smart Metering Geyser,Pipe Insulation Household Cooking Conversion to Gas ESCO PROJECTS Commercial Industrial Redistribution OTHER DSM INITIATIVE EXPENSES Awareness and communication (including extensive Residential programme and education programmes) street lights and Gov buildings 15 Power Alert PROGRAMME OVERHEADS Department overhead Monitor & Evaluation Costs NET OPERATING EXPENSES R R R 2, The main drive to reduce the load will be to install CFLs throughout South Africa. This should result in a saving of 0.03kW per bulb exchanged and installed. The intention is to purchase 19.8m CFLs for delivery from the end of May 2008 to ensure that the bulbs can be installed in houses at a rate of 3m per month until the end of December A number of initiatives will be run in parallel. The roll-out of CFLs will occur with a house-to-house roll-out, corporate and government employees, exchange with major retailers and the Eskom employee roll-out. The implementation plan includes other initiatives that will contribute towards a significant and sustained demand reduction. Eskom intends to encourage its customers to install solar water heaters (SWH), shower roses, geyser blankets and convert to gas stoves. House-to-house CFL roll-out is targeted at the low income households as access to these households is easier than access to high LSM households. The target is to deliver 4m CFLs per month from April 2008 to August

35 However, a realistic target will be to deliver a maximum of 2m from April in the current year. Thus dedicated resources have been commissioned to ensure that this can be achieved. A Project Management company with experience in this field will be commissioned to assist with the roll-out. The benefit of the house to house roll out is that the incandescent bulbs are collected and destroyed. This programme is intended to realise 100MW by the end of August 2008 and an additional 247 MW by the end of December Corporate and Government employees It is estimated that government buildings and corporate buildings will exchange (5MW) and (21MW) inefficient bulbs respectively for an equal amount of CFLs. Exchange with major retailers Eskom has targeted Pick n Pay and Massmart for this programme. There is limited ability to collect inefficient bulbs and have them destroyed. It is estimated that this initiative will deliver or 15MW. Massmart has committed bulbs and the balance will be contributed by Pick n Pay. The Eskom employee roll-out The CFL exchange with Eskom employees is intended to realise bulbs being exchanged. This will result in 6MW saving. Solar water heaters The installation of a standard SWH unit will reduce the demand by an average of 0.6kW per unit. Thus if these can be installed it has a large savings potential although the cost is prohibitive at present. The challenge that should be resolved in order to accelerate this process is that there are seven registered suppliers in the local market, while over a hundred (100) suppliers can supply a product that is not SABS approved. 35

36 The next challenge is that the subsidy for the installation of SWH is inadequate at 20% of the cost. The level of the subsidy does not encourage customers to install SWH in large numbers. It is recommended that the subsidy be increased to 50% of the cost of installation. There must also be a drive to reduce the overall cost of the units as the cost/mw is six times that of CFL s even with a 50% subsidy. Due to the lack of progress with the existing SWH programme, Eskom proposes that it purchase units by September 2008 to ensure that it can install in Eskom villages as well as mining villages for major mining houses. This will realise a 7MW saving. Shower roses The aerated shower roses save 0.26kW per installation. It also saves water. An estimate of units that will be installed by end December is This equates to 5MW. There is potential for up to 25MW if specifications, contracts, procedures are in place before the end of March Geyser blankets Geyser blankets are expensive at a unit installed cost of R500 and only provide a benefit of 0.05kW. This equates to less than 2 CFLs. Due to the high cost, large storage space required and difficulty in getting access to houses, it was decided to link this initiative to the shower rose and SWH rollout. Thus estimated installation is from to units. This will only result in a reduction in load of 1MW. However, the energy saving relating to geysers is notable being around 100kWh per annum per geyser. Thus it is still advisable to attach this initiative onto others. Conversion to gas stoves The intention in this strategy is to develop a programme similar to that of SWH whereby a subsidy is offered to purchase the units. Thus the procurement and sale of the units will be via the retailers. The amount of stoves that could be sold in this way is estimated to be in the order to in 12 months. The amount of power saved per stove is 0.6kW for a 4 plate stove and 0.8kW 36

37 for a 4 plate stove and gas oven. The amount estimated to be saved for the year is 36MW. ( units) Table 12: Breakdown of DSM costs and MW savings TOTAL 12 Months 0 to 5 Months 5-9 Months 10 to 12 Months PROGRAMME AREA MW MW MW MW CFL Programmes CFL Mass Rollout Eskom Employees Government Buildings & Employees Retail Corporate Gas Stoves Solar Water Heating Shower Heads Geyser Blankets Smart Metering TOTAL MASS ROLLOUT Without compromising the initiatives implemented for the last ten months of the year, a more robust and pervasive implementation plan should be rolled out in order to secure the benefits enjoyed. The implementation plan beyond the ten months ending in December 2008 has two aspects to it: o Streamline the project approval process for DSM initiatives. o Ensure technology is in place for load reduction by means of tariffs and load limiting. The aim of the streamlining is to ensure that a project proposal from ESCO or any other person is processed within two weeks so that the project can be executed and benefits realised promptly. At present the DSM process can take in excess of two years for projects to be approved. There are numerous devices that are being developed that can perform various tasks relating to possible load reduction. It is necessary to categorise these based on function and determine the implementation of these into the future. 37

38 Load limiters and warning devices These devices enable customers to shed their load prior to being cut off. These devices should be remote controlled and enable the utility to warn the customers that load reduction is necessary, failing which, the supply will be disconnected. This can be accompanied with a reduction of breaker size or merely a disconnect if the current is above a certain limit. The cost of these devices needs to be below R600 and should be able to be retrofitted to prepayment and billed customers. Load limiting devices shift load and do not reduce load in the long term. It is thus an alternative to load-shedding where the customer can still retain supply for lights and entertainment. Smart metering The Smart meter normally combines the abilities of the load limiter with that of the ability to introduce time of use tariffs. A pilot needs to be initiated in 2008 with Eskom customers to enable roll-out to the industry from Use of smart meters for theft reduction It is estimated that smart metering can reduce electricity theft by 1.5kW per installation in low LSM areas. In the Gauteng area alone this should result in a load drop of MW. It is also important to note that DSM will not necessarily reduce the load of those that bypass meters. 38

39 Summary of initiatives Table 13 shows the proposed acceleration of 3000MW from five years to three years including the current acceleration process. Table 13: Breakdown of accelerated DSM PROGRAMME AREA 2007/8 2008/9 2009/ /11 Total MW MW MW MW MW RESIDENTIAL PROGRAMMES ,301 Efficient Lighting (CFLs) Solar Water Heating Shower Aerators Smart Metering Geyser and Pipe Insulation Household Cooking Conversion to Gas ESCO PROJECTS Commercial Industrial Redistribution TOTAL This DSM programme is aggressive and it is a cornerstone to balancing supply and demand of electricity in the short-term. It is necessary to constantly monitor the system requirements to ensure that the load reduction and energy efficiency is in line with the other programmes and is complementary to them. Table 14: DSM Targets and cost comparison of actuals vs MYPD projections 2006/7 2007/8 2008/9 Total MYPD costs R400m R421m R440m R1,261m MYPD MWs 153 MW 153 MW 153 MW 459 MW Actual/Projected costs R625m Actual /Projected MW 169MW 400MW 815MW 1384 MW 39

40 4 Financial Analysis Eskom recommends implementation of scenario 5, with a 53% and 43% real price increase in 2008/9 and 2009/10 respectively. This is based on full passthrough of primary energy costs and DSM cost variances over the MYPD to be implemented in 2008/9. Five year financial plan assumptions Eskom Board has approved the next five year financial plan cognisant of the following assumptions: - Reduction in electricity demand due to PCP - Pass-through of prudent primary energy costs - Allowance of capital expenditure from a timing and cost perspective - Government loan of R60bn, with assumed phasing of R6bn (2008/9), R12bn (2009/10), R20bn (2010//11) and R22bn (2011/12). - Maximum Eskom borrowing of R30bn annually over the next five years - Capital expenditure of R343bn over the next five years - Return on assets of vary in short term due to funding constraints, but targets a return of 10% real over medium term Scenarios for revision of price increase in 2008/9 1 Base assumes price increase of 9% real (14,2% nominal) in 2008/9, R60bn and primary energy costs pass-through of R2,4bn. Approved by Board on 27 February Primary energy 2008/9 assumes prices of 28% real in 2008/9, and primary energy cost pass-through of R5,3bn in 2008/9. The first two years under-recovery of actual costs compared to MYPD assumption is absorbed by Eskom 3 Primary energy Full 2006/7 to 2008/9 assumes price increase of 47% real in 2008/9, and primary energy cost pass-through of R13bn for the MYPD period. 4 Primary energy 2008/9 + DSM assumes price increase of 33% real in 2008/9, and primary energy pass-through of R5,3bn in 2008/9. The first two 40

41 years under-recovery of actual costs compared to MYPD assumption is absorbed by Eskom. Accelerated DSM will incur additional costs of R2bn in 2008/9. 5 Primary energy full + DSM full: Assumes price increase of 53% real in 2008/9 and primary energy cost pass-through of R13bn for the MYPD period. Accelerated DSM will incur additional costs of R2,5bn over the MYPD period, the recovery of which would result in a further price increase of 6% to the consumer. Financial summary The following are extracts from financial information for the various scenarios. Pretax interest cover /7 2007/8 2008/9 2009/ / / / Debt Equity 2006/7 2007/8 2008/9 2009/ / / /13 25% Return on Assets 20% 15% 10% 5% 0% -5% 2006/7 2007/8 2008/9 Return on assets 2009/ / /12 Average ROA 2012/13 Figure 8: Scenario 1 Financial Analysis 41

42 If the status quo of 14,2% remains, Eskom will make a loss before tax in the range of R1bn to R10bn which will place additional pressures on funding. More importantly catch up will be required in 2009/10 to stablise the situation with real increases of approximately 100%. Pretax interest cover /7 2007/8 2008/9 2009/ / / / Debt Equity /7 2007/8 2008/9 2009/ / / /13 25% Return on Assets 20% 15% 10% 5% 0% 2006/7 2007/8 2008/9 Return on assets 2009/ / /12 Average ROA 2012/13 Figure 9: Scenario 4 Financial Analysis In scenario 4, Eskom will make a loss after tax of between R1bn and R10bn with a 33% real increase in 2008/9 followed by a further increase of 64% real in 2009/10. 42

43 Pretax interest cover 2006/7 2007/8 2008/9 2009/ / / / Debt Equity /7 2007/8 2008/9 2009/ / / /13 25% Return on Assets 20% 15% 10% 5% 0% 2006/7 2007/8 2008/9 Return on assets 2009/ /11 Average ROA 2011/ /13 Figure 10: Scenario 5 Financial Analysis Scenario 5 will provide a profit after tax of between R1bn and R5bn with price increases of 53% real in 2008/9 followed by 43% real in 2009/10. Financial sustainability The financial health is determined by its credit ratings awarded by international credit rating agencies. The key ratios targeted by these ratings agencies are interest cover and debt equity, with both ratios largely being a function of prices (awarded by the NERSA). 43

44 Debt Equity Debt Equity (incl prov) '2007/8 '2008/9 '2009/10 '2010/11 '2011/12 '2012/13 Base PE 2008/9 PE Full MYPD Threshold PE 2008/9 + DSM 2008/9 PE Full MYPD + DSM Full Figure 11: Debt: Equity scenario analysis on revised price increases The Debt:Equity ratio graph illustrated in Figure 11, shows that Eskom is below (better than) the threshold of 2:1 ratio, except for the base case which equals 2.04:1 ratio in year 2011/12. 44

45 Interest cover 6.0 Pre-tax Interest Cover '2007/8 '2008/9 '2009/10 '2010/11 '2011/12 '2012/13 Base PE 2008/9 PE Full MYPD BBB Rating PE 2008/9 + DSM 2008/9 PE Full MYPD + DSM Full Figure 12: Pre-tax Interest Cover scenario analysis on revised price increases The base case shows a negative pre-tax interest cover in 2008/9. Between the period 2008/9 and 2010/11 the ratio shows that the pre-tax interest cover is above (better than) the threshold of 3 times but reducing in later years. The base case shows a spike in 2009/10 of 98%. However, to solve the spike in 2009/10 the PE Full MYPD + DSM Full scenario increased the price in 2008/9 to 53%, thereafter 43% in 2009/10. Phasing of tariff adjustments over the future will only prolong the agony and result in the wave of spike increases. Similar to petrol prices there is no major impairment in increasing prices in line with global trends and once these global pressures are reduced future price reductions can be implemented. 45

46 Price increases of electricity 100% Real Price Increases (above CPIX) 80% 60% 40% 20% 0% '2007/8 '2008/9 '2009/10 '2010/11 '2011/12 '2012/13-20% Base PE 2008/9 PE Full MYPD PE 2008/9 + DSM 2008/9 PE Full MYPD + DSM Full Figure 13: Real Price Increase Scenario analysis The base case shows a spike in 2009/10 of 98%. However, to solve the spike in 2009/10 the PE Full MYPD + DSM Full scenario increased the price in 2008/9 to 53%, thereafter 43% in 2009/10. Phasing of tariff adjustments over the future will only prolong the agony and result in the wave of spike increases. Similar to petrol prices there is no major impairment in increasing prices in line with global trends and once these global pressures are reduced future price reductions can be implemented. 46

47 Average Price of Electricity 50 Price of electricity in c/kwh '2007/8 '2008/9 '2009/10 '2010/11 '2011/12 '2012/13 Base PE 2008/9 PE Full MYPD PE 2008/9 + DSM 2008/9 PE Full MYPD + DSM Full Figure 14: Average price of electricity in c/kwh scenario analysis The price of electricity in c/kwh is below the cost of supply in 2007/8 and 2008/9. 47

48 Funding requirements 60,000 Annual Funding Requirements 50,000 Rand millions 40,000 30,000 20,000 10,000 0 '2007/8 '2008/9 '2009/10 '2010/11 '2011/12 '2012/13 Base PE 2008/9 PE Full MYPD PE 2008/9 + DSM 2008/9 PF full MYPD + DSM Full Treasury limit Treasury limit + R60bn loan Figure 15: Annual Funding Requirements scenario analysis From 2008/9 Eskom will require more the than estimated annual limit that the capital markets will be able to provide. The additional loan of R60bn from National Treasury would enable Eskom to fund its capital expansion program. 48

49 5 Implementation of the proposed price increase Tariff structures to address affordability to the indigent The high electricity price increase will have a definite impact on the economy especially on the indigent without appropriate intervention. There are a number of options that can be considered to address the impact of these high price increases on the poor. It, however, is important before an option is chosen and implemented to understand firstly how Eskom s tariffs are structured and what the impact of the options is. Eskom offers its residential customers the option of 2 tariffs namely, Homelight and Homepower. Homelight is a single energy rate life-line tariff. This tariff has no fixed charges and is structured in such a way to provide a subsidy to the low consumption customers. The Homelight tariff, therefore, is usually the tariff where the majority of Free Basic Electricity (FBE) is provided, a low average monthly consumption is experienced and where the average Eskom rate is below the cost to supply, i.e. it is subsidised. The other residential tariff is the Homepower tariff. This tariff recovers its cost to supply and makes a small contribution to subsidies. This tariff is suitable for higher residential consumption supplies where there is a fixed charge differentiated on the size of supply, i.e. higher fixed charges for greater capacity supplies. The following graph shows the breakeven between Homelight and Homepower based on 2007/8 rates. 49

50 Comparison of Homepower and Homelight 2007/ Average consumption for Eskom on Homelight 300 R/month Homepower 4 Homelight 60A Homelight 20A Consumption per month Figure 16: Comparison of Homepower and Homelight 2007/8 The same principle is also true for Eskom s tariff to its small commercial customers - Businessrate. Businessrate 4, aimed at lower consumption smaller commercial customers, is a single energy rate tariff, while Businessrate 1, 2 and 3 have a high fixed charge which increases as the capacity increases. The following graph shows the break-even between these two tariffs. 50

51 Comparison of Businessrate 4 and Businessrate / R/month Businessrate 1 Businessrate Consumption per month Figure 17: Comparison of Businessrate 4 and Businessrate /8 Before any option is considered it is also important to understand the current situation around the shortage of energy and demand on the Eskom system and the response obtained from the different customer categories to reduce their electricity consumption. One can consider whether this additional price increase could be utilised to include a stronger pricing signal for efficient use of electricity to the appropriate customer categories. The Minister of Public Enterprises recently informed the public that large industrial customers have delivered demand reduction since February, but that significant savings from the commercial and residential sectors have not yet been achieved. With the higher price increase being requested, the big issue to be resolved is then how to protect the indigent (who already consume the minimum due to poverty) and to provide the correct pricing signals to higher consumption customers in the residential and commercial categories who are considered to be the least energy efficient. It is accepted that the indigent should receive some protection from high price increases and one way to do this is through differential price increases, i.e. to increase the current subsidies to the applicable tariffs. This is achieved by 51

52 implementing a lower price increase to the relevant tariffs while the increased subsidy is raised on other tariffs. A number of different options were modelled to show the impact of applying such a methodology. In all cases the Homelight tariff was given the lowest price increase as this is the tariff targeted at the indigent. Another option that has been publicly mentioned is the inclining block rate to be applied to residential and small business customers. This tariff structure, however, has some challenges. 5.1 The inclining block rate tariff structure The inclining block rate tariffs is touted as a structure that can achieve a demand side objective, but is also being mentioned that is can provide a cross subsidy to the indigent. With the inclining block rate, the energy rate tariffs becomes significantly higher above a certain consumption level (knee point). The logic usually used for such a tariff is that inclining block rates would provide an economic incentive for customers to reduce their energy use and help motivate customers to reduce their bills, reduce demand growth and defer the need for new generation. Eskom offers a choice of 2 tariff options, i.e. Homepower and Homelight, to its customers in this category with the indigent most likely already on the Homelight tariff. There are a number of challenges associated with an inclining block rate to be used address the issue of affordability. These are: It does not reflect costs and gives no signal for peak usage and usage patterns no demand side management (efficiency) signal. o Customers who consume only small on-peak quantities crosssubsidise customers using higher amounts in peak periods. 52

53 To try use it to address poverty relief (through increasing subsidies) and energy efficiency (will reduce contribution to subsidies) are conflicting objectives. An inclining block rate tariff cannot be implemented currently to Eskom s prepayment customers as the meters do not allow for it. The level of sales to the Homepower and Businessrate tariffs are too low to provide sufficient revenue to provide a cross-subsidy to Homelight using an inclining block rate. The above assumes that the Homepower customers will not reduce consumption due to the higher price and additional revenue will be obtained from them. This may or may not be true, but based on international research there is considerable uncertainty over the likely reaction of customers to these tariffs will they respond at all and will the response be as intended and therefore, will the level of crosssubsidy be sustained as intended. There needs to be an accurate determination of the knee-point where the rate increases. It should ideally be higher than the break-even between Homelight and Homepower, but low enough to earn enough additional revenue to cover the cross-subsidy and this is the biggest challenge. If the break-even is too low, customers using electricity in the lower block may be encouraged to use more and if it is too high, you create hardship for the medium level of consumers or encourage them convert to Homelight. It creates revenue uncertainty, is not proven to reduce costs and therefore cannot be considered as the optimal solution. For the above reasons it is not considered a feasible option to address the needs of a lower price increase to the indigent. 53

54 5.2 Differential price increase Another method of addressing the issue of affordability is to have a higher increase for the Businessrate and Homepower tariffs and to offset this with lower increase to Homelight. It is to be noted that Homepower and Businessrate together make up only 1,6% of Eskom sales volumes and Homelight 3,11%. A number of scenarios were modelled With the 60,87% nominal price increase (53% real Scenario 5) The following range of price increases brought the same revenue as requested under scenario 5: Table 15: Range of indicative differential increases based on Scenario 5 (in nominal terms) Option Homelight Businessrate and Homepower All other tariffs 1 35% 89% 61,9% 2 45% 70% 61,9% 3 30% 70% 63% 4 10% 100% 63,5% 5 26% 63,5% 63,5% Due to the small volumes on Homepower and Businessrate, the price increase to be applied to the tariffs has to be significantly higher than the 60.87%. Option 1 places the greatest burden of the reduced increase to Homelight on Homepower and Businessrate and increases the breakeven between the two tariffs to 1100 kwh per month. This may encourage customers to convert from Homepower to Homelight, which may result in a revenue loss. This risk would need to be mitigated. This Option also results in a 1% higher increase to other tariffs. Option 2 gives a slightly higher increase to Homelight than Option 1 resulting in a lower increase to Homepower and Businessrate. 54

55 Option 3 explores a slightly higher price increase to other tariffs as well. This Option shares the burden between all tariffs by applying a slightly higher increase to all other tariffs. For this option the break-even between the tariffs is 800 kwh per month. Option 4 gives the lowest price increase to Homelight, but results in a very high increase to Homepower and Businessrate. Option 5 gives the same higher increase to Homepower, Businessrate and all other tariffs and a lower increase to Homelight With the 39,61% nominal price increase (33% real Scenario 4) Using the same principles for Option in the 60,87% increase scenario, the 39,61 nominal % impact was modelled. The following range of price increases brought the same revenue as requested under the 39,61% average price increase scenario: Table 16: Range of indicative differential increases based Scenario 4 (in nominal terms) Option Homelight Businessrate and Homepower All other tariffs 1 22% 70% 40% 2 28% 55% 40% 3 15% 55% 41% 4 10% 55% 41,4% 5 12% 41,7% 41,7% As with the 60,87%, the above scenario increases the risk of tariff conversion between Homelight and Homepower. Rand impact per customer per month Even though the increase may seem high, in order to put them into perspective this needs to be shown on the impact per customers per month. The table below shows the impact using the indicative price increase options for the Homelight and Homepower tariffs at their average consumption levels for the 60,87% scenario. 55

56 Table 17: R/month impact using the 60,87% scenario and options At an increase of 60,87%, a Homelight 20A customer using 100 kwh per month will pay R27,56 more per month and at an increase of 30%, R13,58 more per month (excluding VAT). For Homepower customers, in the worst case scenario ie the 100% increase, they will pay using 1000 kwh per month R more per month and at 70% R257,34 more per month. 5.3 Equity considerations of Eskom applying differential price increases to its customers versus residential and commercial tariffs within municipalities One of the motivators for going for higher price increases for residential and commercial was stated to be a signal to encourage these customer categories to conserve electricity as they are currently not meeting the target to save 10% of their consumption. It, however, is very important to understand the customer base in South Africa. Eskom has only a small number of customers in the medium to higher consumption residential categories vs the number of customers for similar tariff categories in the municipal areas. Therefore, in order to be effective in the South African context NERSA would rely on Government mandating a similar approach to be implemented by municipalities to their customers in those categories. 56