South Africa: Renewable Energy Market Transformation (REMT) Project

DRAFT REPORT South Africa: Renewable Energy Market Transformation (REMT) Project Economic and Financial Analysis Due Diligence 21 June 2004 ACKNO...
Author: Elvin Harper
6 downloads 1 Views 2MB Size
DRAFT REPORT

South Africa: Renewable Energy Market Transformation (REMT) Project

Economic and Financial Analysis Due Diligence

21 June 2004

ACKNOWLEDGEMENTS This report was prepared for the World Bank by Conningarth Economists, Pretoria, under contract [insert], and Peter Meier, under contracts PO 7607294 and PO 7622939. Arun Sanghvi was the World Bank Task Manager. Several technical experts made important contributions to this report: •

Small Hydro: Bo Barta, University of the Witwatersrand



Biomass Sugar Bagasse : Arnoud Wienese, Sugar Milling Research Institute



Biomass Pulp and Paper: Ian Davies, Campbell Davies Consulting



Landfill gas: Ray Lombard, Lombard de Mattos & Associates.



Wind Energy: Jason Schaffler, International Institute for Energy Conservation



Solar Water Heating: Dieter Holm [

].

[insert paragraph to thank World Bank peer reviewers, Bob Chronowski, . . . ] Notwithstanding these contributions, the responsibility for the opinions and recommendations of this report is solely that of the principal authors, and the views expressed do not necessarily correspond to those of the World Bank or the Government of South Africa.

ii

DRAFT

CONTENTS ABBREVIATIONS AND ACRONYMS .................................................................. vi 1. BACKGROUND ............................................................................................................1 The Renewable Energy White paper ..........................................................................1 REMT background .....................................................................................................2 The DME study ..........................................................................................................2 Objective.....................................................................................................................5 2. METHODOLOGY .......................................................................................................6 Financial analysis..........................................................................................................6 Interest rate and inflation assumptions .......................................................................7 Investment returns .......................................................................................................8 JSE Securities exchange .............................................................................................8 Views of a major commercial bank............................................................................9 Industrial Development Corporation(IDC).................................................................9 Development Bank of South Africa (DBSA) .............................................................9 Assumptions for equity returns.................................................................................10 The time profile of the FIRR ....................................................................................10 Economic Analysis......................................................................................................11 The discount rate ......................................................................................................11 Sources of concessionary finance. .............................................................................13 The Central Energy Fund .........................................................................................13 Development Bank of Southern Afrcia ....................................................................14 The Industrial Development Corporation.................................................................14 3. POOL PRICES AND LRMC ......................................................................................16 Institutional arrangements ........................................................................................16 The present setup ......................................................................................................16 Restructuring options................................................................................................17 LRMC estimates .........................................................................................................18 Pool price scenarios ....................................................................................................20 Prices in Scenario A: Regulated monopoly ..............................................................20 Prices in Scenario B: Wholesale competition...........................................................21 4. REPRESENTATIVE PROJECTS AND SUPPLY CURVES...................................24 iii

DRAFT

Small Hydro ................................................................................................................24 Background...............................................................................................................24 Small hydro rehabilitation ........................................................................................27 Hydro projects at water transfer schemes.................................................................29 New hydro schemes..................................................................................................30 The small hydro supply curve ..................................................................................32 Employment .............................................................................................................34 Black economic empowerment aspects ....................................................................34 Institutional barriers to small hydro development ....................................................35 Conclusions on small hydro .....................................................................................35 Sugar Bagasse .............................................................................................................36 Spare capacity...........................................................................................................38 Modifications to reduce process steam usage...........................................................38 Full-scale cogeneration plant....................................................................................40 Conclusions on sugar................................................................................................43 WIND...........................................................................................................................45 Background...............................................................................................................45 Baseline economic analysis ......................................................................................47 Capacity penalties.....................................................................................................50 Risk factors...............................................................................................................52 Conclusions on windpower ......................................................................................54 Pulp and Paper Industry............................................................................................55 Sappi Ngodwana.......................................................................................................56 Sappi Tugela.............................................................................................................57 Conclusions on pulp and paper.................................................................................58 Solar water heating.....................................................................................................59 Background...............................................................................................................59 SWH market transformation.....................................................................................60 Contribution to the 1000 GWh target .......................................................................61 Landfill Gas.................................................................................................................62 Assessed Projects......................................................................................................64 LFG supply curve .....................................................................................................67 Conclusions on LFG.................................................................................................69 5. IMPLEMENTATION SCENARIOS ........................................................................70 iv

DRAFT

The energy supply curve of identified candidate projects ......................................70 REMT Phase I: the first 1000 GWh..........................................................................71 Pulp and paper ..........................................................................................................71 Landfill gas ..................................................................................................................72 Small hydro ..............................................................................................................72 Phase II: 4,000 GWh ..................................................................................................73 Pulp and paper ..........................................................................................................73 Sugar industry...........................................................................................................73 Solar water heating ...................................................................................................74 Summary...................................................................................................................75 Phase III: the 10,000 GWh Target ............................................................................76 6. CARBON FINANCE..................................................................................................77 South Africa’s Carbon Finance Perspective ............................................................77 Pulp and Paper ...........................................................................................................78 Landfill gas..................................................................................................................79 Total REMT-I carbon finance requirements ...........................................................80 Conclusions on carbon finance..................................................................................81 REFERENCES ...........................................................................................................82

v

DRAFT

ABBREVIATIONS AND ACRONYMS BEE CDM CER CPI CSIR CSWH DBSA DME ERR Eskom FBC FIRR GEF GoSA HH IDC IPECAC IPP JIBOR JSE LFG LH LHWP LRMC NER PBMR PCF PPA RE REMT RoE RoR RSA SAM SWH TA tc VAT

Black Economic Empowerment Clean Development Mechanism Certified Emissions Reduction consumer price index Council for Scientific and Industrial Research (South Africa) commercial solar water heating Development Bank of South Africa Department of Minerals and Energy (South Africa) economic rate of return Electricity Supply Commission (South Africa) fluidized bed combustion financial internal rate of return Global Environment Facility Government of South Africa high head (hydro project) Industrial Development Corporation International Panel for Climate Change independent power producer Johannesburg Interbank Offer Rate Johannesburg Stock Exchange landfill gas low head (hydro project) Lesotho Highlands Water Projects long-run marginal costs National Electricity Regulator (South Africa) pebble bed modular reactor Prototype Carbon Fund power purchase agreement renewable energy Renewable Energy Market Transformation (World Bank Project) return on equity run of river Republic of South Africa social accounting matrix solar water heating technical assistance ton of cane value added tax

vi

DRAFT

1. BACKGROUND

1. South Africa has very high levels of greenhouse gas (GHG) emissions, and ranks amongst the top ten producers of GHG emissions on a per capita basis in the world. The primary reason is that South Africa relies heavily on coal to meet its energy needs. In particular, South Africa has a large-scale coal-based power generation system. 2. But while coal is likely to remain a financially attractive energy source for South Africa, the Government is committed to reducing the country's GHG emissions, and increased renewable energy generation constitutes an important option to achieve this objective. The Renewable Energy White paper 3. In August 2003, the Government published its White Paper on Renewable Energy,1 which was approved by the Cabinet in November 2003. This argues for “an equitable level of national resources to be invested in renewable technologies, given their potential and compared to investments in other energy supply options.” The White Paper sets a target of 10,000 GWh of renewable energy contribution to final energy consumption by 2013, which represents 4% of the projected electricity demand for 2013. This 10,000 GWh target is in addition to the estimated 2000 renewable energy contribution of some 115,000 GWh/annum, generated from fuelwood and waste used by mostly rural households for cooking and heating. A 10-year renewable energy implementation plan will start in June 2004 by the Department of Minerals and Energy (DME). 4. The renewable energy contribution to this target is expected to include biomass, wind, solar and small-scale hydro, and includes grid connected power generation, and non-electric technologies such as solar water heating and bio-fuels that displace electricity. 5. The Government is committed to the introduction of greater levels of competition in electricity markets in South Africa, and has indicated that it will create an enabling environment to facilitate the introduction of independent power producers that generate electricity from renewable energy sources. To complement these reforms, Government would like to see a greater investment by the private sector in renewable energy power producers, and in the commercialisation and local manufacturing of renewable energy technologies.

1

White Paper on Renewable Energy. Department of Minerals and Energy. August 2003 1

DRAFT

1. BACKGROUND

REMT background 6. The Renewable Energy Market Transformation (REMT) project is designed to support the Government in setting up the policy and institutional framework required to transform the market for renewable energy. Using Global Environment Facility (GEF) funds, the project will provide technical assistance and capacity building. In addition, catalytic and declining GEF grants will be provided to remove the barriers to the utilization of solar water heaters by commercial customers. It is expected that the commercial solar water heating (CSWH) sector can be quickly transformed into a financially viable industry. 7. It is expected that the GEF grant will be about US$6.0 million, of which about $5.0 million is for technical assistance (TA) and $1.0 million for grants for CSWH, with the remaining financing to be provided by Government (about $1.5 million for TA) and private sources (about $9.0 million for CSWH). Phase I of REMT, which is expected to run from 2005-2008, aims to achieve the first 1,000 GWh/year of renewable energy of the White Paper target, and a potential follow-up phase (REMT-II, 2009-2012) plans to bring this total to 4,000 GWh. 8. This project is expected to trigger significant private sector investments in renewable energy power generation. These investments will be financed outside this project by a combination of private equity and debt, with debt financing facilitated by an output-based revenue stream provided by external carbon funds, in the first instance by the Prototype Carbon Fund (PCF). The DME study2 9. In 2003, the DME commissioned a study to develop supply curves for renewable energy to assist them in selecting the optimal mix of technologies for fulfilling the 10,000GWh White Paper target. This examined the following grid-connected renewable energy technologies • • • • • •



Biomass - Bagasse in the sugar industry Biomass - Paper and Pulp industry Landfill Gas utilisation to produce electricity Small-scale hydro Solar Water Heating in residential houses with water and electricity Solar Water Heating in commercial buildings Wind Energy.

10. The main results of the DME study are shown in Table 1.1. As shown in the financial supply curve of Figure 1.1,3 the large increments of renewable energy are in wind (class 4,5 and 6), medium hydro (10-50MW), and Sugar-bagasse (conversion into

2

COWI A/S and Conningarth Economists, Economic and Financial Calculations and Modelling for the Renewable Energy Strategy Formulation , Report to the Department of Minerals and Energy, Pretoria, February 2004. This study is hereinafter cited simply as the “DME Study”.

3

The DME study distinguishes between a so-called “static” financial cost, and a “dynamic” financial cost: the former reflects current costs and conversion efficiencies, the latter reflects likely developments by 2010. The dynamic cost curves are significantly different only for wind power, where significant capital cost decreases and improvements in efficiency (from taller towers and larger machines) may be expected. In this study, whenever the DME financial cost results are quoted, it is to the “dynamic” costs. 2

DRAFT

1. BACKGROUND

full cogeneration facilities).4 But with the exception of sugar-bagasse, none of these large tranches fall into the first 10,000GWh. It may be concluded that the first 10,000GWh will consist of a mix of technologies, requiring a broad focus, rather than emphasising a single technology. Wind may be the largest resource, but requires high costs to harness it. Table 1.1: Results of the DME Study GWh Output Biomass Pulp & Paper: Mill 1 Hydro: Large - Refurbishment Landfill gas: Large Landfill gas: Medium Landfill gas: Small Solar Commercial: Office & Banking Space Sugar Bagasse: Including High Pressure Boilers Biomass Pulp & Paper: Mill 2 Sugar Bagasse: Reduced Process Steam Solar Commercial: Hostels - Education Solar Commercial: Hospitals Solar Commercial: Hostels - Security Services Sugar Bagasse: Including High Tops & Trash Landfill gas: Micro Hydro: Large - Inter-Basin Transfer Solar Residential: Low Income Households Hydro: Large - RoR - LH Wind Energy: Class 1 Hydro: Small - Unconventional Wind Energy: Class 2 Solar Residential: Medium Income Households Solar Residential: Cluster Housing Solar Residential: Traditional Housing Wind Energy: Class 3 Solar Commercial: Shopping Space Hydro: Large - Diversion Wind Energy: Class 4 Solar Residential: Hotels Hydro: Large - Inter-Basen Transfer Solar Residential: Hotels Hydro: Small - RoR - HH Solar Commercial: Industrial & Warehouse Space Hydro: Large - Storage Regular Wind Energy: Class 5 Hydro: Small - RoR - Unconventional Hydro: Small - RoR - Refurbishment Wind Energy: Class 6 Wind Energy: Class 7 Biomass Pulp & Paper: Mill 3

65 273 32 215 160 224 3,795 39 570 581 267 339 1,483 191 526 930 820 63 205 78 1,339 254 159 167 121 6,964 5,109 284 95 2,232 77 210 158 24,841 108 19 31,139 2,705 5

Cumulative Financial Cost GWh output R/kWh 65 339 371 586 746 969 4,764 4,804 5,374 5,955 6,223 6,562 8,045 8,236 8,761 9,691 10,511 10,574 10,779 10,857 12,197 12,451 12,610 12,777 12,898 19,862 24,971 25,255 25,349 27,581 27,658 27,868 28,026 52,867 52,975 52,994 84,133 86,838 86,843

0.10 0.11 0.17 0.18 0.19 0.23 0.22 0.23 0.24 0.30 0.30 0.30 0.29 0.30 0.30 0.35 0.34 0.38 0.34 0.40 0.42 0.42 0.42 0.45 0.45 0.43 0.51 0.49 0.47 0.50 0.48 0.54 0.51 0.58 0.56 0.58 0.70 0.82 0.98

4

The DME study also prepared economic supply curves, but these reveal few differences in the ranking of technologies. Moreover, the economic supply curves in the DME study already embed allowances for GHG credits, which makes it difficult to separate out which part is economic based just on production costs, and which part is economic only when GHG credits are taken into account. Moreover, since carbon finance also affects the financial costs, this study prefers to keep these separate. 3

DRAFT

1. BACKGROUND

Figure 1.1: The DME Study financial supply curve 1.2

1

R/kWh

0.8

WClass6 [31,100 GWh] 0.6

WClass5 [28,800 GWh] WClass4 [5,100 GWh] Hydro [6,964 GWh]

0.4

SugarBagasse[3,795GWh]

0.2

0 0

20

40

60

80

100

cumulative energy, 1000 GWh

11. Indeed, this is confirmed by closer examination of the supply curve for the first 10,000 GWh (Figure 1.2). This reveals that: • • •

wind is not part of the first 10,000 GWh all of the landfill gas potential is within the first 10,000GWh (and occupies a significant portion of the first 1,000 GWh) The balance of the 10,000 GWh target is made up of commercial solar water heating, small hydro, pulp&paper, and the sugar industry.

Figure 1.2: The supply curve: the first 10,000 GWh 0.4 Solar Residential: Low Income Households Hydro: Large - RoR - LH

R/kWh

0.3

Hydro: Large - Inter-Basin Transfer Sugar Bagasse: Including High Tops & Trash Sugar Bagasse: Reduced Process Steam Solar Commercial: Office & Banking Space , Pulp&PaperMill2 Sugar Bagasse: Including High Pressure Boilers

0.2

Landfill gas: Small Landfill gas: Medium Landfill gas: Large

Hydro: Large - Refurbishment Biomass Pulp & Paper: Mill 1

0.1

0 0

2

4

6

8

10

12

cumulative energy, 1000 GWh

12. It must be noted that the DME financial cost estimates are based on a general evaluation of costs without consideration of the capital structure of actual projects – i.e. to derive levelized costs per kWh, capital is priced at the assumed discount rate over the 4

DRAFT

1. BACKGROUND

project lifetime. Generalized “standard electricity generating plants” were created for this purpose, using broad averages for key technical parameters (such as load factors in hydro projects). Financial costs defined in this way may not necessarily reflect the financial attractiveness of projects for private sector implementation, and it is for this reason that the present study examines the financial characteristics of actual potential transactions in more detail. Objective 13. The objective in this economic and financial due diligence study for REMT is to examine a sample of potential projects for each of the sectors in the DME study. At least 50% of the contribution of each technology to the 1,000 GWh target for REMT should be based on assessments of specific projects.5 That is not to say that the projects evaluated in this report are necessarily the ones that will proceed once REMT is in place. But it is to say that REMT can proceed with some confidence that there are real renewable energy project transactions that will be taken up by the private sector in response to the market transformation activities planned for REMT.

5

The sole exception to this is solar water heating (SWH), where the scale of each individual project (in commercial as well as residential applications) is MWh rather GWh. Therefore in this sector, while one can show the results of some representative projects, the contribution to the various targets is based on a topdown, rather than bottom-up approach. 5

DRAFT

2. METHODOLOGY

14. This study follows an approach similar to that of other background studies conducted for World Bank renewable energy studies, notably that for China and Croatia.6 The basic idea is to build up a renewable energy supply curve based on a detailed economic and financial analysis of a sample of actual (or reasonably probable) projects (as opposed to more general qualitative assessments of the size of particular renewable energy resources, and estimated average costs). This approach is particularly appropriate to a sector where the bulk of the projects are expected to be undertaken in the private sector, structured either as non-recourse financed IPPs or under balance sheet financing by large corporations (e.g. as in pulp and paper). 15. Where economic analysis shows a project to be economically attractive – i.e. based on economic costs, shadow priced as appropriate, and benefits assessed on the basis of avoided generation costs priced at economic LRMC – but where financial returns are insufficient for implementation by the private sector, there is a rationale for assistance by public sector development banks tasked with assisting priority sectors, by various forms of private-public partnerships, or even by Government subsidy. Nevertheless, the main question of achieving renewable energy goals remains that of financial feasibility.

Financial analysis 16. For purposes of this study a standard financial model to calculate financial internal rate of return (FIRR) has been developed.7 The financial analysis is conducted from the perspective of the implementing entity, and the calculated FIRR is therefore a function of the capital structure assumed.8 As noted in Box 2.1, the FIRR will be somewhat lower 6

China Renewable Energy Scale-up Program (CRESP): Economic Analysis, Volume I: World Bank, 2003; Frontier Economics, Cost-Benefit Analysis for Renewable Energy in Croatia, Report to the World Bank and the Government of Croatia, May 2003. 7

Obviously the financial model used in this study – based on annual cashflows -- is not as detailed as would typically be used by an IPP for a major project finance application – which would be quarterly during the construction period, and quarterly or semi-annually during operation (depending on the requirements of the lenders). Nor has this study attempted to capture the finer points of non-recourse financing (such as debt service and maintenance reserve escrows, lending facilities for working capital, etc). However, the model is adequate as a first step in assessing the financial feasibility of proposed renewable energy projects from the perspective of the private sector. 8

This is in contrast to the standard World Bank presentation of a financial analysis, which differs from economic analysis only in the adjustment for taxes, duties and transfer payments, and which implies a weighted average financial cost of capital exactly equal to the assumed discount rate (opportunity cost of capital). This definition of financial rate of return is of academic interest only (and really relevant only to 6

DRAFT

2. METHODOLOGY

than the “equity return” or “investment yield” as the latter terms are commonly understood in South Africa. Box 2.1: FIRR v. return on equity Consider the hypothetical project whose cost and benefit streams are shown below. This project has a capital cost of 100, and net annual returns of 20, so the “investment yield” (“equity return”) would be 20%. But the IRR computes to 16.6%, or 3.5% less.

Cost

-1

0

-50

-50

-50

-50

benefit net benefits IRR equity return

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

16.6% 20%

If the construction period were only one instead of two years, the IRR is 18.4%, 1.6% less than the simple “investment yield” of 20%.

17. Capital costs have been estimated by the sector technical experts at 2004 prices, including physical contingencies, but excluding price contingencies. Capital costs are escalated by the applicable inflation rate to the year of assumed construction outlay. The model assumes loans at variable rather than fixed rates.9 The current rate of VAT is 14%. Interest rate and inflation assumptions 18. Currently the interest rate for debt with a relatively low risk could be viewed as the predominant prime overdraft rate of clearing banks, which currently stands at 11.5%. Depending on the definition of inflation (overall consumer price index (CPI) or the CPIX overall price index, excluding interest rates on mortgage bonds), a 4.5% inflation rate for 2004 may be used. Therefore this study infers a current real interest rate of 7%. 19. Given present policies of the Government it may be assumed that the 4.5% inflation rate will persist for some time. Currently the inflation rate of the industrialized countries is in the region of 2.5%, but this can be expected to increase to between 3 and 4% by 2008. 20. The currently very large inflow of short-term foreign capital into South Africa is an indication that current real interest rates of 7% are too high, and this can be expected to will gradually decline over the next few years, stabilizing at 5.5% by 2007. Consequently, the nominal (prime) interest rate will fall from the present 11.5% to 10%. 21. The question is the risk that will be assigned by lending institutions to renewable energy projects. Large corporations in the pulp and paper and sugar industries will doubtless finance projects in their sectors based on their balance sheets, and therefore specific project risk premia (on lending rates) would not apply (though the corporations themselves will doubtless expect substantial rates of equity return on what for them are small projects outside their core businesses).

traditional public sector projects): but in the case of renewable energy projects that are to be undertaken mainly by the private sector, or by local authorities under increasing pressure to evaluate project investment decisions along commercial lines, explicit assumptions about capital structure must be made for the FIRR to be useful guide to actual investment decisions. FIRR is therefore calculated with respect to equity capital. 9

The bulk of DBSA infrastructure lending is now at floating rates, based on JIBOR 3 or 6-month rates. 7

DRAFT

2. METHODOLOGY

22. In the case of IPPs, the situation is different, since projects will doubtless be financed on a non-recourse basis. Moreover, there are significant uncertainties associated with LFG projects or wind farms.10 On the other hand, there are strong indications that the Development Bank of Southern Africa (DBSA) and Industrial Development Corporation (IDC) will provide a significant portion of debt finance, and they may be prepared to absorb some of the credit risk. Consequently this study does not attempt to assess the impact of technology or project specific risks on the interest rate, and uses 10% as the lending rate in all cases (except where expressly noted otherwise). The macroeconomic assumptions used in the models follow as shown in Table 2.1 Table 2.1: Macroeconomic assumptions Real rate of interest, SA Inflation rate Prime rate Escalator [2004=1] Real rate of interest Inflation rate Prime rate escalator[2004=1] exchange rate

[NPV] 2004 [ ] 7.0% [ ] 4.5% [ ] 11.5% [ ] 1 [ ] 1.00% [ ] 2.00% [ ] 3.0% [ ] 1 6.9 [ZAR/$U

[-1] [0] [1] [2] [3] [4] [5] [6] [7] 2005 2006 2007 2008 2009 2010 2011 2012 2013 6.5% 6.0% 5.5% 5.5% 5.5% 5.5% 5.5% 5.5% 5.5% 4.5% 4.5% 4.5% 4.5% 4.5% 4.5% 4.5% 4.5% 4.5% 11.0% 10.5% 10.0% 10.0% 10.0% 10.0% 10.0% 10.0% 10.0% 1.045 1.092 1.141 1.193 1.246 1.302 1.361 1.422 1.486 1.25% 1.50% 1.75% 2.00% 2.00% 2.00% 2.00% 2.00% 2.00% 2.50% 3.00% 3.50% 4.00% 4.00% 4.00% 4.00% 4.00% 4.00% 3.8% 4.5% 5.3% 6.0% 6.0% 6.0% 6.0% 6.0% 6.0% 1.03 1.06 1.09 1.14 1.18 1.23 1.28 1.33 1.38 7.0 7.3 7.6 8.0 8.4 8.9 9.5 10.1 10.9

Investment returns 23. The willingness of the private sector to take up renewable energy projects is a function of the achievable equity returns. Since carbon financing is designed to make possible transactions that would not otherwise be taken up, an assumption must be made in the financial model as to what constitutes the minimally acceptable equity return for a feasible transaction (from an equity standpoint). There is much anecdotal evidence about the equity returns desired by IPPs, but little ex-post evidence about actually achieved returns, particularly for renewable energy projects. 24. To take a view on what value is appropriate is therefore not straight forward, particularly since yields are dependant upon risk perception of the equity investors, which vary from sector to sector and project to project. Unfortunately, access to the financial models prepared for IPP projects currently under development in South Africa has not been possible.11 However, because the development banks take equity positions in priority infrastructure development projects, the following four sources were consulted for guidance: • • • •

Historic yields of companies listed on the JSE securities exchange A major commercial bank Industrial Development Corporation (IDC), and Development Bank of South Africa (DBSA)

JSE Securities exchange 25. The Bureau of Financial Analysis (BFA), calculates financial ratios for companies listed on the JSE securities exchange. Table 2.2 shows average financial returns on

10

These are discussed further in Section 4.

11

For example, the detailed financial model prepared for the Darling Wind Farm is confidential, so no view can be offered here about the equity return expectations of the investors in this project. 8

DRAFT

2. METHODOLOGY

equity (RoE) as reported for 1994 – 2003 for selected sectors, as well as for the years 2002 and 2003. Table 2.2: Financial return on equity for selected sectors Average RoE%

1

2002 IARoE

2

2003

RoE% IARoE% RoE% IARoE%

Coal

13.43

10.24

22.10

14.60

22.10

14.60

Mining Finance

7.81

5.52

-1.23

-2.01

0.02

-1.04

Chemicals

13.87

8.82

20.96

13.58

23.52

15.70

Forestry and paper

9.76

3.86

12.61

2.22

12.61

2.22

Diversified industrials

11.78

7.66

12.71

7.57

12.80

7.66

Average

11.33

7.22

13.43

7.19

14.21

7.83

IARoE=Inflation adjusted RoE Source: BFA

26. The average nominal RoE for the 1994/03 period is approximately 9.43 % and the average inflation adjusted RoE is 5.93% for this same time period. The coal sector could be viewed as the closest investment sector in relation to RE because coal is the main energy source used in generating electricity in SA. Returns in this sector, at 13.43% and 10.24%, are higher than the average. 27. These figures are substantially lower than reported expectations of IPPs that are likely to finance projects on a non-recourse basis, rather than large industrial firms (such as are likely to undertake projects in pulp and paper, or sugar), where balance sheet financing would be most likely. Views of a major commercial bank 28. Investment opportunities – i.e. where banks take equity positions -- are judged purely by the risk attached to the investment. The yield on government bonds, where very little risk is attached, forms the lower bound of their expectations. Currently yields on government bonds, 10 years and over, have a nominal yield of 9.27. This value, adjusted for an expected 4½ % inflation rate, results in a yield of approximately 4.8 %. The attitude of the Bank is that a further 6 to 8 percentage points could be added to this rate where high risk is attached to a specific project. The inflation adjusted yield expected by banks could therefore be as high as 13 % in real terms. Industrial Development Corporation(IDC) 29. The general policy of the IDC regarding its equity participation in projects is that the debt equity ratio should be approximately 50:50. The expected RoE should be in the region of 12 – 18%. Again, the higher the risk, the closer the yield will be to the upper bound of 18%. Depending on the profile of the other equity holders, the IDC will vary its RoE requirement, e.g. in a recent joint venture project with an international development agency that has a strong social development goal, the RoE was lowered to about 6 % in real terms. Development Bank of South Africa (DBSA) 30. The DBSA is mandated to invest in the energy field generally, but also specifically in RE projects. They are already participating in pilot projects (SWH, Darling Wind Farm, LFG). Furthermore, there is already agreement with the DME and 9

DRAFT

2. METHODOLOGY

the Department of Finance that a substantial portion of investment in RE will be channelled through the DBSA. 31. The DBSA, which is very flexible in terms of financing projects, may provide equity as well as debt finance, and may even provide a grant component for capacity building and project preparation for priority sectors. Interest rates would be structured to meet the needs of the project, and may range from 5% to the current prime rate of 11.5%. The DBSA appears prepared to channel a considerable amount of its own funds into RE projects, taking equity positions as well as loan funding. Assumptions for equity returns Table 2.3 presents an example of investment in a RE project by the various parties described in this section. Equity involvement reflects the relative proportion of the equity that will be provided by the participants. In this example, the DBSA contributes the largest portion of equity, followed by the IPP, the IDC and the commercial bank. The expected yields are based on the discussion presented above and represent an average requirement for each equity participant. The weighted average RoE for this mix of participation is 10.9%. Table 2.3: Weighted average RoE

IPP Commercial Bank IDC DBSA (RSA Government) Weighted Average RoE

Equity Participation,%

Expected Real Yield ( Inflation Adjusted) %

30% 15% 20% 35%

18% 13% 9% 5% 10.9%

32. For purposes of this study it is suggested that a RoE of 11% in a real terms, and a nominal yield of 16.5%, based on a 4.5% long-term inflation rate, should be used as a default value. 12 This is much higher than the 7.33% investment yield that is currently reported by companies in the sectors of interest to this study listed on the Johannesburg Stock Exchange. The time profile of the FIRR 33. Another important (but often neglected) point is that the FIRR, as typically calculated as a single figure, indicates the return for the project if it operates to its completion (assumed in this report as 20 years, though in the case of hydro projects, assumed lifetimes may well be longer). Indeed, an understanding of the time profile of FIRR explains why some industries appear to impose very high FIRR requirements. 34. Consider, for example, the FIRR profile for the Ngodwana pulp and paper mill waste to energy project (discussed in more detail in Section 4), shown in Figure 2.1. The 25% target postulated may appear high, but even the actually achieved 23.4% is attained only in the 20th year. If the project terminates after 10 years, the achieved return is only 17%. Therefore, depending on the degree of risk associated not just (in this case), with the sustainability of the waste stream (perhaps this may be sold at some higher value as a 12 The 16.5% used here is slightly higher than the 15% proposed by a 2001 DME study (Capacity Building in Energy Efficiency and Renewable Energy, 2001) .

10

DRAFT

2. METHODOLOGY

raw material to another industry), but with other business risks (the plant may close, to be replaced by a completely new plant elsewhere, or operations shifted offshore), 25% FIRR for a 20 year project translates to 15% for a 10 year project, which will appear to many private-sector decision-makers as a more reasonable planning criterion. Figure 2.1: FIRR profile for the Ngodwana Pulp and Paper mill waste recovery project 0.3

ERR target=25%

0.25 ERR to project completion:=

23.4%

FIRR (to equity)

0.2

0.15

0.1

0.05

0 [-1]

[1]

[3]

[5]

[7]

[9]

[11]

[13]

[15]

[17]

[19]

Economic Analysis 35. A standard economic analysis is included for each sample project: NPV and economic rate of return (ERR) are calculated for flows converted to constant 2004 prices. Benefits (of avoided energy at the applicable substation) are calculated at the LRMC (see Section 3). ERR is calculated both with and without environmental benefits. 36. In the economic analysis, avoided greenhouse gas emissions are valued at the reported PCF transaction price for the Durban Landfill gas project of $3.75/tonne CO2 – which is used as the proxy for global willingness to pay for South Africa’s carbon reductions, and which may be taken as the relevant economic benefit.13 The quantity of GHGs avoided is taken at 0.89 Kg CO2/kWh, based on the PCF Durban LFG Emission Reduction Study.14 The discount rate 37. There is no official rate used by the Government of South Africa (GoSA) for public sector investments. A 2002 study on benefit cost analysis for the Water Research Commission, which may be viewed as the closest thing to published official guidance, recommended a rate of 8%:15 this value was used in the original Darling Wind Farm

13

Note that this is different to the GEF methodology (as enshrined in the World Bank OP 10.04) – under which the cost of avoided carbon is based on the undiscounted sum of GHG emissions over the lifetime of the project.

14

PCF, Durban South Africa, Landfill Gas to Electricity, Emission Reduction Study, Revised Draft July 2003.

15

Conningarth Economists, A Manual for Cost Benefit Analysis in South Africa with Special Reference to Water Resource Development, Report to the Water Research Commission, Report TT 177/02, October 2002. 11

DRAFT

2. METHODOLOGY

economic analysis.16 However, the NER has used 10% in its recent electricity sector studies,17 and for this reason, this study adopts a 10% rate. Table 2.4 shows discount rates used in other recent World Bank renewable energy project appraisals: 10% is seen to be a typical value. Table 2.4: Discount rates as used in World Bank renewable energy projects Country Rate Renewable energy technologies evaluated Philippines India China

15% 12% 12%

Vietnam Sri Lanka Cape Verde Croatia

10% 10% 10% 8%

solar homes (PV) solar homes(PV), small hydro Small hydro, wind, bagasse, landfill gas (CRESP); Wind, solar homes (REDP) Village (micro) hydro Small hydro, wind, village (micro) hydro, solar homes Wind Biomass (combined heat and power), wind, small hydro

38. The choice of discount rate, obviously, has no bearing on the financial viability from the perspective of an IPP. But the choice of discount rate does have a significant effect on the economically optimal quantity of renewable energy, which should be the criterion for Government (for example in setting the magnitude of any renewable energy target). Economic theory holds that the optimal quantity of RE is represented by the point at which the supply curve intersects with the LRMC: thus, in Figure 2.2, the renewable energy supply curve S10 (at the 10% discount rate) intersects the LRMC10 at the optimal quantity Q*10. Figure 2.2: The optimal quantity of renewable energy

P

S10

S8

LRMC10

Q*10

Q8

16

Conningarth Economicsts and DBSA, The Darling National Wind Farm Demonstration Project, Report to DME, 22 August 2002 17

National Electricity Regulator, (with ESKOM, and Energy Research Institute, University of Capetown), Integrated National Resource Plan 2003/2004, Reference Case, Pretoria, February 2004. 12

DRAFT

2. METHODOLOGY

39. A lower discount rate will shift downward the economic supply curve, shifting the point of intersection with the LRMC to the right (i.e. implying that a larger quantity of renewable energy is optimal). Thus, in the example of Figure 2.2, the quantity for a discount rate of 8% shifts to Q8. This is not, however, the optimal quantity at the 8% discount rate, because if that same lower discount rate is also used to determine the system LRMC, then this will also decline, say to LRMC8. Therefore, as shown in Figure 2.3, the optimal quantity Q8* is somewhat less than Q8. Note that since capital is a smaller proportion of fossil-based generation, the relative decline of the LRMC will be smaller than the downward shift in the more capital intensive renewable energy supply curve, so the optimal quantity is still larger than the original value. Figure 2.3: The optimal quantity of renewable energy at 8% discount rate.

P

S10

S8

LRMC10 LRMC8

Q*10

Q*8

Q8

40. This is not merely of academic interest, but has significant practical consequences, particularly for wind power. As noted below in Section 4, the use of an 8% discount rate increases the quantity of wind power that is economically optimal.

Sources of concessionary finance. The Central Energy Fund 41. The Central Energy Fund (CEF) is involved in the search for appropriate energy solutions to meet the future energy needs of South Africa, SADC and the sub-Saharan African region. These energy solutions include oil, gas, electrical power, solar energy, low-smoke fuels, biomass, wind and renewable energy sources. CEF also manages the operation and development of the oil and gas assets and operations of the South African government. 42. The Energy Development Corporation (EDC), a division of the CEF, is part of a group owned by Government and controlled by the Minister of Minerals and Energy. It is one of the Government's key institutions for transforming the favourable policy framework on renewable energy into reality. Being close to policy makers, EDC is able to lobby the relevant government departments and institutions for support when necessary. At the same time it operates as a fully commercial entity. 13

DRAFT

2. METHODOLOGY

43. EDC is committed to the objectives of the Government's Black Economic Empowerment strategy. The intention is to ensure that EDC encourages procurement from black or black empowering enterprises and facilitates black ownership. 44. EDC is primarily an equity investor with a long-term commercial perspective, and seeks partnerships with a variety of donor agencies, including Danida (Danish International Development Assistance,) Norad (Norwegian .Agency for Development Co-operation), GTZ (German Technical Assistance), UNDP (United Nations Development Programme) and GEF (Global Environment Facility). EDC is building up a portfolio of investments and projects across the spectrum of energy and low smoke fuels. These include • • • • •



Bethlehen 3.9 MW small hydro project (one of the projects evaluated in Section 4). EDC is one of the anchor equity investors. LFG projects: EDC is teaming up with metropolitan municipalities in pursuit of viable and sustainable investment opportunities in LFG extraction projects. Solar energy: EDC is currently implementing two projects aimed at addressing the barriers that prevent the uptake and full commercialisation of selected solar technologies. Solar Water Heating Project: a joint project with UNDP, focussed on standards for SWH to increase consumer confidence. Solar Cooker Project: Supported by EDC, GTZ and UNDP, the project aims to commercialise the use of solar and other renewable energy cooking technologies. Having demonstrated the potential viability of developing a dedicated industry by creating a 'business case', the project is targeted on activities that reduce time, costs and risks and improve the economies of scale of businesses that wish to participate in the renewable and alternative household energy industry. Wind energy: The South African Government has requested EDC to become a cosponsor of the 5.2MW Darling National Demonstration Wind Farm Project.

Development Bank of Southern Afrcia 45. The Development Bank of Southern Africa (DBSA) is southern Africa's premier infrastructure development finance institution. The DBSA envisions an empowered and integrated southern African region free of poverty, inequity and dependency. Towards this end, the DBSA seeks to catalyze socio-economic development and economic integration in southern Africa, and to become a strategic development partner to the wider African region south of the Sahara. 46. Established in 1983 by the government of the Republic of South Africa, the DBSA is one of five existing development finance institutions in South Africa and has a mandate to accelerate sustainable socio-economic development in the region by funding physical, social and economic infrastructure. In doing so, the DBSA endorses and promotes human resource development and institutional capacity- building. 47. The DBSA finances and sponsors programmes and projects formulated to address the social, economic and environmental needs of the people of southern Africa in improving their quality of life. The Bank adheres to the principles of sustainable development. A recent transformation at the Bank saw the institution moving away from being solely focused on development finance, and becoming a key national development institution having a threefold role as financier, advisor and partner. The Industrial Development Corporation 48. The Industrial Development Corporation of South Africa Limited (IDC) is a selffinancing, state-owned national development finance institution that provides finance to entrepreneurs engaged in competitive industries. The vision of the IDC is to be the 14

DRAFT

2. METHODOLOGY

primary source of commercially sustainable industrial development and innovation to the benefit of South Africa and the rest of the African continent. 49. The IDC’s primary objectives are to contribute to the generation of balanced, sustainable economic growth in Africa and to the economic empowerment of the South African population, thereby promoting the economic prosperity of all citizens. The IDC achieves this by promoting entrepreneurship through the building of competitive industries and enterprises based on sound business principles. IDC’s core strategies include: • • • • •

Providing risk capital to the widest range of industrial projects Identifying and supporting opportunities not yet addressed by the market Maintaining its financial independence Building upon and investing in human capital in ways that systematically and increasingly reflect the diversity of South African society, and Establishing local and global involvement and partnerships in projects that are rooted in or benefit South Africa and the rest of Africa

15

DRAFT

3. POOL PRICES AND LRMC

Institutional arrangements The present setup 50. The general institutional setup is shown in Figure 3.1. ESKOM is one of the largest utilities in the world, with some 39,000 MW of generating capacity, with municipals and private companies adding another 1,800 MW and 1,400 MW, respectively. Total generation in [ ] was [ ] GWh, 93% of which was coal –fired. Figure 3.1: Institutional setup

51. The National Electricity Regulator (NER) licenses all plants with annual generation greater than 5GWh. Potential IPPs (conventional or renewable) are free to enter into PPAs with buyers: indeed, one of the criteria set by the regulator for issuing a generation 16

DRAFT

3. LRMC and POOL PRICES

licence is that the buyer is clearly identified and a PPA is in place. If wheeling over the ESKOM grid is required, no charge is levied (on grounds that customers have already paid for 100% of the transmission grid as part of the NER-approved retail rates). Table 3.1 shows the existing renewable energy plants licensed by NER and the PPA arrangements that are in place. Table 3.1: Renewable energy plants licensed by NER

Friedenheim

IPP

Licensed capacity 3

Maximum Power Produced, MW 2

Energy sent out, GWh 14434

Private consumption 1041

Load Factor,% 73.7

Lydenburg

Municipal

2

2

6000

0

34.2

Ceres

Municipal

1

1

413

0

8.6

Piet Retief

Municipal

1

1

2900

0

33.1

First Falls

ESKOM

6

6

36792

0

70

Second Falls

ESKOM

11

11

67452

0

70

Ncora

ESKOM

2

2

12264

0

70

Gariep

ESKOM

Vanderkloof

ESKOM

TH Amatikulu

Sugar Industry

12

10

43775

43775

51

TH Darnall

Sugar Industry

13

7

27388

27388

44.7

TH Felixton

Sugar Industry

32

22

79935

79935

41.5

TH Maidstone*

Sugar Industry

29

20

79582

44917

45.4

Transvaal Suiker* Sugar Industry

20

-

-

-

-

Source: DME, Bulk Renewable Energy Independent Power Producers in South Africa, January 2001.

Box 3.1: The Friedenheim small hydro IPP The 3MW Friedenheim hydro project can be viewed as the only existing IPP in South Africa. It is privately owned, sells the bulk of the generated electricity through a Power Purchase Agreement (PPA) and is a profitable operation. The plant is operated as a commercially profitable and sustainable business venture. Its successful track record has gone a long way towards dispelling the myths common in the electricity supply industry that it is not possible to compete with Eskom, especially not with small renewable energy producers selling energy wholesale. The project is situated next to the town of Nelspruit in the Mpumalanga province. It is owned by the members of the Friedenheim Irrigation Board and operated by an engineering firm. The plant provides power for water pumping to the Friedenheim Irrigation Board (about 1 GWh), but 93% of the power generated (14.4 GWh) is sold to the Nelspruit local authority through a PPA that sets the tariff at 12% below the price at which Nelspruit buys power from Eskom (its bulk electricity provider). The water supply to the turbines is by means of two penstocks feeding from an irrigation channel. The outlet of the turbines is into the Crocodile river. This set-up is suitable for constant and base load power supply as there is no capacity to store water for peak power generation. After some initial difficulties with the water inlet system and the dedicated power line connecting the plant to the Nelspruit distribution system, the plant has proved to require very low levels of maintenance. Maintenance and operation costs are in the order of R200,000/annum, which is mainly routine maintenance and labour. The reported load factor is 74%. Source: DME, Bulk Renewable Energy Independent Power Producers in South Africa, January 2001.

Restructuring options 52. For some years now, following the 1998 Energy Policy White Paper, the Government of South Africa has been considering significant change in the setup of the 17

DRAFT

3. LRMC and POOL PRICES

electricity supply industry. However, it was stated in the White Paper that “any market restructuring is likely to be delayed for a number of years while the distribution sector is restructured and the bulk of the electrification programme is undertaken.” Thus the pace of reform has been slow, and since the publication of the White Paper there has been considerable debate and analysis on whether and/or how to restructure (or unbundle) Eskom, how to introduce competition into the generation sector, and encourage greater private sector participation in the power sector in general. Figure 3.2: The structure under wholesale competition

source: DME (2002).

53. Most observers envisage some form of wholesale competition structure emerging at some point over the next decade, but with considerable uncertainty about timing (Figure 3.2). This represents a significant uncertainty for renewable energy projects since the structure will be a principal determinant of pool prices. For renewable energy projects, cashflows in the first 5-10 years will determine financial viability, and therefore projects built in the REMT time frame will face pool prices and LRMC conditions as they follow from the existing structure. The higher pool prices that will follow from wholesale competition (and the need for additional capacity, as explained below) can be expected only beyond 2010, so these changes will likely benefit only the second phase of REMT.

LRMC estimates 54. The current excess capacity in the ESKOM system derives from the overly large investments in capacity in the late 1970s and early 1980s, which resulted in some 4,000MW of coal-fired capacity being mothballed. Thus the demand growth over the next 6 or 7 years can be met by bringing these mothballed plants back into service, which means that the LRMC will climb only slowly from its currently low value (in effect, at the present time with excess capacity, the LRMC equals the short run marginal cost SRMC, of about R0.04/kWh. 55. NER has estimated the system LRMC in its recent 2003/2004 integrated resource plan (NER, 2004). The capacity expansion plan is shown in Figure 3.3: no new baseload 18

DRAFT

3. LRMC and POOL PRICES

capacity is envisaged until 2012, when the first greenfield fluidised bed combustion (FBC) project is commissioned. In the years 2005-2012 the three moth-balled plants are returned to service, and new open cycle gas turbines are commissioned starting in 2008. Figure 3.3: Capacity expansion plan

Source: NER (2004)

56. Based on the modelling results, (in turn based on the expansion plan shown in Figiure 3.3), NER has estimated the LRMC as depicted in Figure 3.4 – shown is the NER’s “Reference Plan” which is used in this REMT study. NER also developed some alternative plans with somewhat earlier timing (about 1-2 years sooner in the period 2008-2013) in achieving the long-term equilibrium level of R0.25/kWh, but the Reference Plan was judged to be appropriate for REMT. The LRMC slowly climbs as the mothballed plants are brought into service, and by 2009 the LRMC climbs rapidly. As noted in Section 5, this increase in LRMC (and pool price) during the years of REMT-II (expected to be 2009-2012) is of some significance to the likely renewable energy technology mix.18 The figure also shows the Pool prices, discussed in the next section.19

18

FIRR is particularly sensitive to the magnitude of revenues in the early years of a project. Higher Pool prices just a few years earlier into the operation of a project therefore has a disproportionate effect: a project that is not viable in 2008 (during REMT-I) may become so just a few years later, say in 2011, during REMTII. 19

The values of LRMC are presented below in Table 3.2. 19

DRAFT

3. LRMC and POOL PRICES

Figure 3.4: LRMC and Pool prices (at constant 2004 prices) 0.3 LRMC 0.25

R/kWh

0.2

POOL

0.15

0.1

0.05

0 2004

2006

2008

2010

2012

2014

2016

2018

2020

2022

Source: LRMC: NER; pool price, ECON

Pool price scenarios 57. At present, the “Pool” is entirely within ESKOM. Prices paid by the pool for ESKOM generation are regulated by the National Electricity Regulator (NER), and are in effect historical cost accounting prices (based on actual costs plus an allowable equity return). Pool prices are key to this study, for they determine in large part the financial price at which renewable energy projects can sell their energy to local authorities and distributors. 58. A detailed study of pool prices was undertaken by the Norwegian company ECON.20 Two basic scenarios were developed: A. Regulatory Monopoly, with prices essentially cost-plus largely determined by regulatory policy and investment levels; and B. Wholesale competition, where prices are set by the market balance of supply and demand. Both scenarios were modelled making use of financial and market models developed by ECON. 59. It should be stressed that this study examined only wholesale generation prices, and excluded consideration of transmission and distribution costs. While transmission costs are fairly small, distribution charges constitute a large portion of the overall price buildup to end-users, and will vary considerably for different customer categories. Prices in Scenario A: Regulated monopoly 60. Prices in this Scenario will be regulated by the NER on the basis of costs incurred in the industry.20 The key determinants of costs will be the level of investment as well as regulatory policy. The study analysed two different regulatory approaches: •



Historical cost regulation: Asset values based on the historical costs, adjusted for accumulated depreciation. This determines depreciation costs as well as the capital base used for return calculations. A nominal rate of return is applied. Current cost regulation: Asset values are based on their replacement cost, adjusted for accumulated depreciation. This change will affect depreciation as well as the capital base

20 Electricity Price Scenarios For South Africa. A Study Commissioned by the Department of Minerals and Energy, South Africa. Econ Centre for Economic Analysis. 2001

20

DRAFT

3. LRMC and POOL PRICES

for return calculations (although taxation remains the same). A real rate of return is applied.

61. Eskom currently prepares both historical and current cost accounts. Given that Eskom’s rate of return on current cost accounts is low, the study presumed that historical cost accounts are used for pricing. Cost models were based on historical costs that closely match the inflation-adjusted average energy price in the proposed Wholesale Electricity Pricing system (WEPS). The scenario assumes that Eskom gradually increases its rate of return from current levels to a more commercial level, i.e. an increase in the nominal rate of return on capital from 10% to 15% over five years. The real rate of return (used in current-cost regulation) also increases by 1% per annum. 62. The results of the price modelling to 2025 are presented in Figure 3.5. Historical cost accounting results in gradually increasing prices as new capacity is added to the system. It was assumed that, if investment levels stay at a steady state, the price would tend to converge to the cost of new capacity (and higher if over-investment occurs): the decline in prices in the last few years is simply an end-effect of the model.21 Figure 3.5: Price results under Scenario A: Regulated Monopoly Wholesale price (2001 c/kWh)

20

Current cost regulated prices Historical cost regulated prices

15

10

Regulated prices are likely to continue to increase slowly if investm ent after 2025 is taken into consideration

5

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

-

Source: ECON, op.cit.

63. Prices based on current cost accounts show greater stability. Prices start higher (a consequence of the asset revaluation), and increase at a slower rate. Regulation based on current cost accounts will typically provide more stable prices than under historical cost accounts, and the results illustrate this feature. Prices in Scenario B: Wholesale competition 64. This scenario assumes that prices in a competitive market will depend on the balance of supply and demand in the market. It further assumed that the same underlying pressure on demand as in the regulated scenario, with the exception that explicit consideration of demand-side management was excluded. Instead, consideration was taken of demand response to price (through the use of price elasticities), which incorporates demand-side responses.

21 This is corrected in the financial model by assuming that once the maximum pool price of 18.75c/kWh is reached (in 2016, see Table 3.2), it will remain at this level to the end of the planning horizon.

21

DRAFT

3. LRMC and POOL PRICES

65. It was assumed that investments would be triggered solely by price developments. Investment and prices are thus inter-related – investment occurs to keep prices at around the level of new entrants. As demand grows and more peaking plant is required, prices will rise to the level required by peaking plant. Similarly, as demand grows further, prices will rise to the level required by base-load plant. Figure 3.6 shows the projected prices, and compared to the historical cost-accounting trajectory of Scenario A. 66. During the period 2001 – 2006, it was expected that there would be surplus capacity in the market, and as a result, prices would fall to as little as 4c/kWh. These would increase only gradually to a level of around 10 c/kWh, as mothballed plants come into service. Between 2009 and 2015 prices remain fairly stable at this level, as demand growth adjusts to the new price level, and as small amounts of peaking capacity ate added. Only after 2015, as new baseload capacity is required, do prices climb again to reach 15 c/kWh. Figure 3.6: Price results under Scenario B: Wholesale Competition 25

Wholesale price (2001 c/kWh)

C urrent cost regulated prices H istorical cost regulated prices

20

C om petitive m arket prices 15

10

5

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

-

Source: ECON, op,cit.,

67. Based on recent NER decisions that barely account for inflation, the study has taken the view that historical cost regulated prices will prevail for the foreseeable future, and that a sudden jump to a current cost basis is very unlikely. On the other hand, competitive market prices of scenario B will unlikely occur in the near future, because, whatever the Government’s long-term commitment to structural reform, the associated institutional changes will inevitably take considerable time. 68. Therefore, for modelling purposes, the ECON historical cost accounting pool prices are used, and converted to constant 2004 prices. Subsequent assumptions about inflation are endogenous to the model (see Table 2.1). Table 3.2 shows the assumed pool prices at 2004 prices. To this is added a margin for transmission and distribution that is determined as the difference between the present price at which the local authority or local distributor buys from ESKOM (typically between R0.13 and 0.16/kWh), and the generation pool price. This latter margin is location specific, and calculated for each renewable energy project as applicable (and escalated separately for inflation without any further assumptions about changes in the real value of the transmission margin).

22

DRAFT

3. LRMC and POOL PRICES

Table 3.2: Assumed pool and LRMC prices (at 2004 prices) Pool price, LRMC Rcents/kWh Rcents/kWh 2004

9.44

4.55

2005

9.56

4.58

2006

9.74

4.9

2007

10.34

5.66

2008

10.94

7.48

2009

11.54

10.46

2010

12.62

14.53

2011

13.7

17.85

2012

14.06

20.65

2013

15.03

22.77

2014

16.01

23.79

2015

18.03

24.4

2016

18.75

24.7

2017

18.51

25.1

2018

18.39

25.3

2019

18.51

25.4

2020

17.79

25.5

23

DRAFT

4. REPRESENTATIVE PROJECTS AND SUPPLY CURVES

Small Hydro Background 69. The hydro baseline study22 shows substantial hydro power potential in South Africa, as shown in Table 4.1. Table 4.1: Installed hydropower capacity and development potential. Water Management Area Limpopo Luvuvhu/Letaba Crocodile West & Marico Olifants Inkomati Usutu to Mhlatuze Thukela Upper Vaal Middle Vaal Lower Vaal Mvoti to Umzinmkulu Mzimvubu to Keiskama Upper Orange Lower Orange Fish to Tsitsikamma Gouritz Olifants/Doring Breede Berg Total

Installed Capacity (Mw) Pumped storage

Large> 10MW

Firm development potential (Mw) Small< 10 MW

Large> 10 MW

Small< 10 MW

0,6 3,2 3,0 12,4 1,0 0,01 0,3

1 000

53 600

0,1 14,1 0,6

10,5 21 3 700

1 250

5,0 12,7

18,1 10,1

120 7,7 1,0 0,8 0,01

400 180 1 580

653

33,9

1,7 5 091

69,0

Pumped storage 1 000 2 000 1 000 2 000 1 000 7 000

Source: Baseline Study Hydropower in South Africa (2002). Note: The values do not include imported hydropower.

22 B. Barta and D. Stephenson, Baseline Study – Hydropower in South Africa. Water Systems Research Group, 2002.

24

DRAFT

4. PROJECTS

70. Clearly the bulk of the hydro development potential is in large projects and pumped storage. However, pumped storage and large hydro schemes are not of interest to REMT -- the former because (in South Africa) they are net emitters of GHGs, the latter because they do not generally qualify for carbon finance (the largest hydro project financed by PCF to date is 78MW (see Box 4.1). Although the usual definition of small hydro is 10MW,23 hydro generation that can be achieved at existing dams or water transfer schemes (i.e. that do not require construction of new impoundments, and where the environmental costs of large impoundments are sunk) should be included in the definition of qualifying renewable energy under the various targets, including REMT.24 Box 4.1: PCF Deals for small hydro projects The largest small hydro project to obtain PCF funding to date is a 78MW RoR project in Colombia: most are below 45MW. PCF small hydro contracts Location

Description

Chile: Chacabuquito Small Hydro

26 MW run-of-river hydro to replace coal or gas 6.69 in the grid 43 MW peaking run-of-river hydroelectric plant 7.50 in the west coast of Guatemala to displace thermal power plants 39MW run-of-the-river hydroelectric project in 6.30 the lower part of the General River sub-basin in the Chirripó Atlantic basin 78 MW run-of-the-river scheme to displace 7.85 thermal power generation 20 MW run-of-river hydro on the Sabanilla river in Zamora Chinchipe Province 3.02 11.8 MW run-of-river hydro in the province of 1.92 Cotopaxi 18 MW run-of-river hydro on the Toachi river in 2.90 the province of Cotopaxi 64.7 MW capacity run-of-river scheme to 8.25 displace thermal plants in several cities in Peru

Guatemala: El Canada Small Hydro

Costa Rica -Rio General

Colombia - Rio Amoya ROR hydro Ecuador: Sabanilla Small Hydro Ecuador: Pilalo Small Hydro Ecuador: Sigchos Small Hydro Peru: SIIF Andina S.A. Hydro

million US$

Source: PCF

71. The DME study classified hydro schemes by generic type, as shown in Table 4.2. Note that here, “large” means projects in the 10-50MW size class.

23 In China, the world’s leader in small hydro with some 23,000MW currently in place, the criterion is 50MW. All hydro, of whatever capacity, appears to be included in the overall EU targets for renewable energy. 24

Indeed, installed capacity is an exceptionally poor criterion for what constitutes “small” hydro as opposed “large” hydro, and has been adopted purely for administrative convenience, rather than on any rational basis that reflects the actual environmental impact of different types of project. A much better criterion would be by the amount of storage that is required: qualifying for REMT should be any hydro project that does not require the construction of a new impoundment that stores more than some given proportion of annual average flow. Clearly, a 100 MW project at the foot of an existing irrigation dam is likely to have a significantly smaller environmental impact than a new 50 MW run-of-river project that may require a new impoundment. 25

DRAFT

4. PROJECTS

Table 4.2: DME Study classification of hydro projects Size classification

Category Name

Description

1-10MW (small)

Refurbishment

1-10MW (small)

Inter-Basin Transfer

1-10MW (small)

Run of River – High Head

1-10MW (small) 1-10MW (small)

Run of River – Low Head Unconventional

>10MW to 50MW (large)

Refurbishment

>10MW to 50MW (large)

Inter-Basin Transfer

>10MW to 50MW (large)

Run of River – Low Head

>10MW to 50MW (large)

Diversion Fed

>10MW to 50MW (large)

Storage projects

Refurbishment of existing small hydropower infrastructure Adding small hydropower equipment to existing interbasin infrastructure Creating a high head by diverting water flow via a canal or pipeline. Power output fluctuates with the flow of the river Creating a low head with a weir or barrage structure. Power output fluctuates with the flow of the river Small hydropower installations in irrigation canals, reverse pumping in urban water supply and underground pumped storage Refurbishment and upgrading of existing, idle large hydropower infrastructure Incorporating large hydropower equipment installations into future water transfer/inter-basin transfer schemes Creating a low head with a dam impounded storage system. Power output can (but does not necessarily) fluctuate with the flow of the river Creating a high or low head scheme with a long canal/tunnel/pipeline diversion from a river tributary to a main course river Assumption in this category is that powerhouses would be installed at existing multi-purpose projects (and therefore involves minimal civil works.

Source: DME

72. Based on this classification, the DME study results are as indicated in Table 4.3 (sorted by increasing financial cost). Only large refurbishment projects come in at less than typical current purchase rates (at the 11kV level) of R0.12-0.14 /kWh, and in the overall ranking only this type of project falls into the REMT Phase I 1000 GWh target. Table 4.3: DME study results Category

Large: Refurbishment Large: Inter-Basin Transfer Small: Unconventional Large: RoR - LH Large: Diversion Small: Inter-Basin Transfer Small: RoR - HH Large: Storage Regular Small: RoR - LH Small: Refurbishment TOTAL

Annual Output GWh 273 526 205 820 6964 95 77 158 108 19

Financial Target Cost R/kWh 0.11 Within the first 1000 GWh 0.3 Within the 10000 GWh target 0.34 0.34 0.43 0.47 0.48 0.51 0.56 0.58

9244

73. The problem in the DME small hydro potential estimates is that they are based on assumed averages of what are in fact quite broad ranges. For example, load factors in 26

DRAFT

4. PROJECTS

any given category vary by 50%, which results in generation costs that vary by a factor of two. Therefore, in this study a sample of seven small representative hydro projects that provide a more precise picture of what is financially and economically feasible has been selected. The type of project most likely to fall under the REMT 1,000 GWh and 4,000 GWh targets are: • • •

rehabilitation of existing small hydro projects hydro generation at water transfer schemes new hydro stations at existing dams that do not require major civil works

74. Indeed, the first bona fide IPP with an NER generation licence is the 2.5MW Friedenheim small hydro scheme (See Box 3.1),25 and there are many similar types of small hydro projects that would be suitable for the first 4,000 GWh of renewables. Moreover, many of these projects will contribute significantly to the GoSA’s commitment to rural poverty alleviation, either as stand-alone small hydro projects, or as part of a “mini-hybrid” off-grid system that includes other RE technologies such as solar and wind26. As such, it is recommended that this potential be evaluated during subsequent phase of the REMT project. Small hydro rehabilitation 75. The Kougha (formerly Paul Sauer) dam hydroelectric plant was last operated in 1982, but still has in place three 1,750 kW turbines which need refurbishment, and generators which must be completely replaced. The PPA would be with the Gamtoos Irrigation Board, and would replace electricity purchased from ESKOM at R 0.16/kWh. A 3km 11kV transmission line would need to be built to the nearest existing substation. The estimated baseline FIRR is 36.2% (Table 4.4),27 and is clearly viable without any carbon finance.

25

As noted in Table 3.1, there are also three municipally owned hydro schemes, the largest of which is the 2MW Lydenburg project. 26

See, e.g., the notes on Sustainable Villages Africa in Appendix I.

27

Assuming 8 year finance at 12%; 33% marginal income tax rate; all of the equipment and construction costs may be assumed sourced locally. 27

DRAFT

4. PROJECTS

Table 4.4: Financials for the Kougha dam small hydro rehabilitation project [NPV] capital cost capacity capital cost

442 5.25 2319 16.0

disbursement profile total debt:ZAR 0.70 debt:FOREX 0.00 equity 0.30 GEF Grant/Capital subsidy 0.00 auxilary consumption 0.0% transmission loss 1.5% renewable generation 59.1% supplemental generation 0.0% total sendout 59.1% revenues average tariff PPA sales [Table 17] green premium/kWh subsidy CER sales [power][Table 18] CER sales [methane][Table 18] total revenue levelised revenue costs supplemental fuel cost O&M costs[as%capital] 5.0% O&Mcosts 0.435 debt service [Table 16] Equity 4.8 total costs total financial flows, before tax income tax financial flows, after tax FIRR nominal

[$/kW] [MW] [1000$] [Rmill.] [ ] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.]

[GWh] [GWh] [GWh] [R/kWh] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [R/kWh] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [ ]

[-1] [0] [1] [2] [3] [4] [5] [6] [7] 2005 2006 2007 2008 2009 2010 2011 2012 2013

16.0

14.6 9.4 0.0 4.0 0.0

3.34 0.2 3.3 2.2 0.0 1.0 0.0

17.3 0.8 14.0 9.0 0.0 3.8 0.0

191.2 0.0 191.2 62.8 0.0 0.0 0.0 62.8 0.33

27.2 0.0 27.2

0.0

27.2 0.0 27.2

27.2 0.0 27.2

27.2 0.0 27.2

27.2 0.0 27.2

-1.0 -1.0 -1.0

-0.2 -3.8 -4.1 -4.1

-1.0

-4.1

0.00 -0.91 -0.50 -2.5

0.00 -0.95 -0.52 -2.4

0.00 -1.00 -0.54 -2.2

0.00 -1.04 -0.57 -2.1

0.00 -1.09 -0.59 -2.0

0.00 -1.14 -0.62 -1.8

0.00 -1.19 -0.65 -1.7

-3.9 1.1 -0.3 0.8

-3.9 1.4 -0.4 1.0

-3.8 1.9 -0.5 1.4

-3.7 2.5 -0.7 1.8

-3.6 3.1 -0.9 2.2

-3.6 3.8 -1.1 2.7

-3.5 4.7 -1.4 3.3

76. This rehabilitation project has high economic returns as well, as shown in Table 4.5. Before environmental benefits, the ERR is 19.7%; this rises to 23.9% when avoided carbon is valued at $3.75/ton CO2.28

28

High economic and financial returns for small hydro rehabilitation projects are consistent with the experience in other countries. In China, small hydro rehab was found to be one of the most attractive renewable energy options. Motivated largely by the desire to stimulate rural development in remote areas far from grid access, large numbers of small hydro projects were built in the 1960s and 1970s, and which can now be brought to modern standards at relatively low cost. 28

DRAFT

27.2 0.0 27.2

0.184 0.194 0.211 0.228 0.246 0.273 0.301 5.0 5.3 5.7 6.2 6.7 7.4 8.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5.0 5.3 5.7 6.2 6.7 7.4 8.2

-8.8 -9.5 -4.0 -27.1 35.7 -11.5 24.3 36.2%

27.2 0.0 27.2

4. PROJECTS

Table 4.5: Economic returns, Kougha dam small hydro rehabilitation project. [NPV] 2004

[-1] [0] [1] [2] [3] [4] [5] [6] [7] 2005 2006 2007 2008 2009 2010 2011 2012 2013

at current prices Financial capital cost -17.4 -3.3 [Rmill.] 2.4 0.5 taxes&duties [Rmill.] Capital cost -14.9 -2.9 [Rmill.] -12.2 O&M cost [Rmill.] total costs -27.2 -2.9 [Rmill.] at constant 2004 prices -21.6 -2.8 [Rmill.] benefits LRMC transmission 0.05 [R/kWh] generation [R/kWh] total [R/kWh] energy displaced(at substation) [GWh] 36.9 benefits@LRMC [Rmill.] net economic flows 15.3 -2.8 [Rmill.] ERR 19.7% [ ] GWh 191.24 [GWh] levelised cost 0.14 [R/kWh] Environmental benefits 0.00 local [Rmill.] carbon power generation 0.89 [KgCO2/kWh] avoided emissions 170 [1000tons global WTP 3.75 [$/ton CO2] 8.73 0.0 [Rmill.] methane avoided emissions [as CO2 eq] [1000tonsCO2] at global WTP [Rmill.] net economic flows with env.benefits [Rmill.] 23.99 -2.8 ERR 23.9% [ ]

-17.3 2.4 -14.9 -14.9 -13.6

-1.3 -1.3 -1.1

-1.3 -1.3 -1.1

-1.4 -1.4 -1.1

-1.4 -1.4 -1.1

-1.5 -1.5 -1.1

-1.6 -1.6 -1.1

-1.7 -1.7 -1.1

-13.6

0.05 0.05 0.10 27.2 2.6 1.5

0.05 0.05 0.10 27.2 2.7 1.6

0.05 0.06 0.11 27.2 2.9 1.8

0.05 0.07 0.12 27.2 3.4 2.3

0.05 0.10 0.15 27.2 4.2 3.1

0.05 0.15 0.20 27.2 5.3 4.2

0.05 0.18 0.23 27.2 6.2 5.1

27.2

27.2

27.2

27.2

27.2

27.2

27.2

24

24

24

24

24

24

24

0.0

0.7

0.7

0.8

0.8

0.9

0.9

1.0

-13.6

0 0 2.2

0 0 2.3

0 0 2.5

0 0 3.1

0 0 4.0

0 0 5.1

0 0 6.1

Hydro projects at water transfer schemes 77. Two potential small hydro projects at existing water transfer schemes are evaluated. The schemes have low capital costs because all of the water supply and most of the required civil works are already in place: • Orange Fish Tunnel-Teebus Outlet: this is an existing water transfer tunnel of some 75km between Oviston intake in the Garib dam impoundment and the outlet at Teebus, with an historical average annual flow of 17 m3/s. Three turbines of 3MW each could be situated at the outlet to exploit the 65 metre head. The baseline FIRR is 37.4% (nominal). • Fish Sundays chute water transfer scheme: An existing chute of some 55m fall is situated between the diversion canal of Fish Sundays and the Little Fish river. The historical recorded flows (since 1978) are recorded at 7.2 m3/s. A 2.7MW generator can supply local farmers to replace power purchased by the Cookhouse Local Authority from ESKOM. The baseline FIRR is 18%.

29

DRAFT

4. PROJECTS

Table 4.6: Financials, Orange-Fish Tunnel project [NPV] capital cost capacity capital cost

570 9 5130 35.4

disbursement profile total debt:ZAR 0.70 debt:FOREX 0.00 equity 0.30 GEF Grant/Capital subsidy 0.00 auxilary consumption 0.1% transmission loss 8.0% renewable generation 73.5% supplemental generation 0.0% total sendout 73.5% revenues average tariff PPA sales [Table 17] green premium/kWh subsidy CER sales [power][Table 18] CER sales [methane][Table 18] total revenue levelised revenue costs supplemental fuel cost O&M costs[as%capital] 0.0% O&Mcosts 1.24 debt service [Table 16] Equity 10.6 total costs total financial flows, before tax income tax financial flows, after tax FIRR nominal

[$/kW] [MW] [1000$] [Rmill.] [ ] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.]

35.4

32.8 21.5 0.0 9.2 0.0

[GWh] [GWh] [GWh]

407.8 0.0 407.8

[R/kWh] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [R/kWh]

134.0 0.0 0.0 0.0 134.0 0.33

[Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [ ]

[-1] [0] [1] [2] [3] [4] [5] [6] [7] 2005 2006 2007 2008 2009 2010 2011 2012 2013

18.5 0.5 18.5 12.4 0.0 5.3 0.0

37.8 0.5 19.3 12.4 0.0 5.3 0.0 58.0 0.0 58.0

0.0

58.0 0.0 58.0

58.0 0.0 58.0

58.0 0.0 58.0

58.0 0.0 58.0

58.0 0.0 58.0

0.184 0.194 0.211 0.228 0.246 0.273 0.301 10.7 11.3 12.2 13.2 14.3 15.8 17.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 10.7 11.3 12.2 13.2 14.3 15.8 17.5

0.0 -21.6 -9.2 -44.4 89.5 -28.9 60.7 37.4%

58.0 0.0 58.0

-5.3 -5.3 -5.3

-1.3 -5.3 -6.6 -6.6

-5.3

-6.6

0.00 0.00 -1.42 -5.6

0.00 0.00 -1.48 -5.3

0.00 0.00 -1.55 -5.0

0.00 0.00 -1.61 -4.6

0.00 0.00 -1.69 -4.3

0.00 0.00 -1.76 -4.0

0.00 0.00 -1.84 -3.7

-7.0 3.7 -1.0 2.7

-6.7 4.5 -1.3 3.3

-6.5 5.7 -1.6 4.1

-6.3 6.9 -2.0 5.0

-6.0 8.3 -2.4 5.9

-5.8 10.0 -2.9 7.1

-5.6 11.9 -3.5 8.4

New hydro schemes 78. The study has also assessed several new hydropower schemes that avoid major new civil works: • Bethlehem Hydro: Two sites (1.8 and 2.2MW) will be operated as a single project by Bethlemhem Hydro (Pty) Ltd to supply the Dihlabeng Local Municipality in the Free State Province. Water transfer flows from the Lesotho Highlands Water Projects (LHWP) will be used. An NER licence application and water issue permit from DWAF are pending, but the EIA is complete. • Orange River Onseepkans weir and canal: An existing weir and neglected canal at Onseepkans on the Lower Orange River are suitable for a 10MW hydro plant to supply local fruit and vegetable farmers on both river banks (RSA and Namibia). The baseline FIRR is 45.8%. There are 12 other potential sites along the Lower Orange River of a similar nature (this scheme falls under the DME category of small low-head hydro). • Pongolapoort: A 2.7MW project is under investigation at the existing Pongolapoort (Jozinii) Dam administered by DWAF for irrigation (northern Kwazuluu/Natal province). Constant water releases of 5m3/s are available for generation. The FIRR is 21%. • Krokodilpoort: A 3.4MW project at a site on the Crocodile river in Mpumalanga Province is under investigation. An average 12m3/s flow will be conveyed by a 3.5 km pipeline to the drop of 53m to the new powerhouse (this falls into the DME category of small high head). This is the only project among 30

DRAFT

4. PROJECTS

the ones examined here that does not appear to be financially viable (given the information available). 79. The financial returns for the Bethlehem scheme are shown in Table 4.7. FIRR is 14.1%.

The

Table 4.7: Financials for the Bethlehem hydro project [NPV] capital cost capacity capital cost

1504 4 6014 41.5

disbursement profile total debt:ZAR 0.70 debt:FOREX 0.00 equity 0.30 GEF Grant/Capital subsidy 0.00 auxilary consumption 0.0% transmission loss 6.0% renewable generation 75.1% supplemental generation 0.0% total sendout 75.1% revenues average tariff PPA sales [Table 17] green premium/kWh subsidy CER sales [power][Table 18] CER sales [methane][Table 18] total revenue levelised revenue costs supplemental fuel cost O&M costs[as%capital] 1.4% O&Mcosts 0.372 debt service [Table 16] Equity 12.5 total costs total financial flows, before tax income tax financial flows, after tax FIRR nominal

[$/kW] [MW] [1000$] [Rmill.] [ ] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.]

[GWh] [GWh] [GWh] [R/kWh] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [R/kWh] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [ ]

[-1] [0] [1] [2] [3] [4] [5] [6] [7] 2005 2006 2007 2008 2009 2010 2011 2012 2013

41.5

38.1 24.8 0.0 10.6 0.0

14.7 0.3 14.7 9.9 0.0 4.2 0.0

44.7 0.7 29.9 19.2 0.0 8.2 0.0

185.2 0.0 185.2 60.8 0.0 0.0 0.0 60.8 0.33

26.3 0.0 26.3

0.0

26.3 0.0 26.3

26.3 0.0 26.3

26.3 0.0 26.3

26.3 0.0 26.3

26.3 0.0 26.3

0.184 0.194 0.211 0.228 0.246 0.273 0.301 4.8 5.1 5.5 6.0 6.5 7.2 7.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 4.8 5.1 5.5 6.0 6.5 7.2 7.9

-6.2 -24.9 -10.6 -45.8 15.1 -6.9 8.1 14.1%

26.3 0.0 26.3

-4.2 -4.2 -4.2

-1.0 -8.2 -9.3 -9.3

-4.2

-9.3

0.00 -0.64 -0.42 -5.3

0.00 -0.67 -0.44 -5.1

0.00 -0.70 -0.46 -4.8

0.00 -0.73 -0.48 -4.6

0.00 -0.76 -0.51 -4.4

0.00 -0.80 -0.53 -4.1

0.00 -0.83 -0.55 -3.9

-6.4 -1.5 0.0 -1.5

-6.2 -1.1 0.0 -1.1

-6.0 -0.5 0.0 -0.5

-5.8 0.2 0.0 0.2

-5.6 0.9 0.0 0.9

-5.4 1.7 0.0 1.7

-5.3 2.7 0.0 2.7

80. The Bethlehem project typifies the kind of small hydro project that would benefit from carbon finance. Figure 4.1 shows a sensitivity analysis of FIRR as a function of carbon finance available and the capital cost. At 3.75$/ton the FIRR improves to 17.1%, with total carbon finance requirement (in NPV terms) of R8.4 million.

31

DRAFT

4. PROJECTS

Figure 4.1: Sensitivity analysis: FIRR v. value of CER: Bethlehem Hydro 0.3

1353$/kW [-10% ]

0.25

FIRR

capCost:=1504$/kW

0.2

1729$/kW [-15% ]

hurdle rate:=15%

0.15

0.1

source:

0

1

2

3

4

5

6

7

8

9

10

11

12

carbon price,$/tonCO 2eq.

The small hydro supply curve 81. Table 4.8 summarises the results for FIRR and ERR for small hydro projects. Only Krokodilpoort seems unlikely, but this mainly a function of the low tariff with which it must compete: the present tariff of R0.124/kWh is the lowest of any of the small hydro projects. Table 4.8: Summary of small hydro projects GWh MW FIRR ERR, /year with environ ment 22.8%

cumul GWH

OrangeRiver OnSeepkans

57

10.0

45.8%

57

OrangeFishTunnel

58

9.0

37.4%

21.2%

115

KoughaDam

27

5.3

36.2%

23.9%

142

Poongalapoort

21

2.7

21.0%

13.4%

163

FishSundays

21

2.7

18.0%

11.8%

185

Bethlehem hydro

26

4.0

14.1%

9.5%

211

Krokodilpoort

22

4.0

3.4%

5.5%

233

82. Figure 4.2 shows the cumulative GWh available at a given FIRR. About 210 GWh appears likely (from the seven projects studied), of which only 26 GWh (Bethlehem) may need require additional carbon finance.

32

DRAFT

4. PROJECTS

Figure 4.2: Cumulative GWh v. FIRR 0.5

0.4

FIRR

0.3

0.2 FEASIBLE WITHOUT CARBON FINANCE (FIRR>15.0%) 0.1 UNLIKELY (FIRR8.5 8.0-8.5 7.5-8.0 7.0-7.5 6.5-7.0 6.0-6.5 5.5-6.0

0.37 0.35 0.31 0.27 0.24 0.2 0.17

63 78 167 5,109 24,841 31,139 2,705

0.38 0.40 0.45 0.51 0.58 0.70 0.82

total

financial

economic

No Yes, @9,056 GWh No

64,102

Source: DME, op.cit.

46

DRAFT

4. PROJECTS

Baseline economic analysis 119. The economics of wind farms are quite straightforward. Since operating costs are quite small, there are really only two critical variables: capital costs, and average annual load factors. 120. Capital costs for wind power are largely a function of the global market for wind turbines (which has seen dramatic decreases over the last decade, with turbine costs declining from 1,300$/kW to 800$/kW as typical machine sizes have increased from 225kV in 1990 to the present 2MW). Table 4.20 shows various cost estimates for wind power in South Africa. Based on these (and other international estimates).the baseline capital cost for wind farms as may be built during the REMT Phase I period (2005-2008) can be assumed at $1,050/kW. Load factors for each site class are taken from Table 4.17. Table 4.20: Estimates of wind-power costs in South Africa 2000 [Danish Study] Project size

MW

Investment in windmill. Other investment Total

ZAR/kW ZAR/kW ZAR/kW $US/kW

Lifetime Load factor Fixed capital Fixed capital cost – other investment O&M Total (economic) cost

23.4% Rcents/kWh Rcents/kWh Rcents/kWh Rmillion/year Rcents/kWh

(2020) [Danish Study]

2003 [NER]

Darling 20

5,628 1,860 7,688 1085 20 years 23.4% 22.0 7.3

4,680 1,638 6,318 916 20 years 28.5% 15.0 5.3

8.2

4.1

7,724 1,119

2.8 37.5

4x1300kW

1227 25years 28%

1.4

24.4

Sources: (1) Tyge Kjaer et al., Options for Green Electricity in South Africa, Roskilde University, 2002. (2) NER, Integrated Resource Plan, 2004, Appendix 3. (3) Conningarth Economicsts and DBSA, The Darling National Wind Farm Demonstration Project, Report to DME, 22 August 2002. Capital cost is estimated at R63million, converted here at R10/US$ (for 2001)

121. As shown in Table 4.21, for the reference capital cost of $1,050/kW and 37% load factor, the FIRR is 8.4%. The economic analysis shows an ERR of only 5%, increasing to 6.4% when GHG benefits are costed at $3.75/ton. To bring the FIRR to 12% requires carbon finance equivalent to $23.5/ton CO2, which is six times higher than for landfill projects (as discussed below).

47

DRAFT

4. PROJECTS

Table 4.21: Financials, class 1 wind site, current capital costs [NPV] capital cost capacity capital cost

1050 20 21000 144.9

disbursement profile total debt:ZAR 0.70 debt:FOREX 0.00 equity 0.30 GEF Grant/Capital subsidy 0.00 auxilary consumption 0.0% transmission loss 1.0% renewable generation 36.6% supplemental generation 0.0% total sendout 36.6% revenues average tariff PPA sales [Table 17] green premium/kWh subsidy CER sales [power][Table 18] CER sales [methane][Table 18] total revenue levelised revenue costs supplemental fuel cost O&M costs[as%capital] 2.0% O&Mcosts 0 debt service [Table 16] Equity 43.5 total costs total financial flows, before tax income tax financial flows, after tax FIRR nominal

[$/kW] [MW] [1000$] [Rmill.] [ ] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.]

144.9

[GWh] [GWh] [GWh]

451.5 0.0 451.5

[R/kWh] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [R/kWh]

154.3 0.0 0.0 0.0 154.3 0.34

[Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [ ]

130.8 83.8 0.0 35.9 0.0

[-1] [0] [1] [2] [3] [4] [5] [6] [7] 2005 2006 2007 2008 2009 2010 2011 2012 2013

0 158 0.0 1.0 0.0 158.2 0.0 101.4 0.0 0.0 0.0 43.5 0.0 0.0 64.2 0.0 64.2

0.0

64.2 0.0 64.2

64.2 0.0 64.2

64.2 0.0 64.2

64.2 0.0 64.2

64.2 0.0 64.2

0.193 0.209 0.226 0.250 0.276 0.293 0.321 12.4 13.4 14.5 16.0 17.7 18.8 20.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 12.4 13.4 14.5 16.0 17.7 18.8 20.6

-31.9 -83.8 -35.9 -151.6 2.7 -11.6 -8.9 8.4%

64.2 0.0 64.2

0.0 0.0 0.0

0.0 -43.5 -43.5 -43.5

0.0

-43.5

0.00 -3.31 0.00 -16.9

0.00 -3.46 0.00 -16.2

0.00 -3.61 0.00 -15.6

0.00 -3.77 0.00 -14.9

0.00 -3.94 0.00 -14.2

0.00 -4.12 0.00 -13.5

0.00 -4.31 0.00 -12.8

-20.2 -7.8 0.0 -7.8

-19.7 -6.3 0.0 -6.3

-19.2 -4.7 0.0 -4.7

-18.7 -2.6 0.0 -2.6

-18.1 -0.5 0.0 -0.5

-17.6 1.2 0.0 1.2

-17.2 3.4 0.0 3.4

122. Indeed, the corresponding supply curves are stark: as shown in Figure 4.6, the big tranches of wind capacity, which occur in wind speed classes 5 and 6, have FIRR of minus 10% to minus 20%. Figure 4.6: GWh v. FIRR 0.2 FEASIBLE WITHOUT CARBON FINANCE (FIRR>15%)

0.15

FIRR

0.1 UNLIKELY (FIRR 7.50%)

0.05

0

-0.05

-0.1 0

20

40 Thousands

60

80

D elivered energy,G W h/year

48

DRAFT

4. PROJECTS

123. The conventional economic supply curve is no more encouraging, even at significantly lower capital costs. Figure 4.7 shows the ERR for the baseline capital cost of $1,050/kW as a function of load factor, together with the corresponding curves for $850/kW and $650/kW. Figure 4.7: ERR as a function of capital cost and load factor 0.2

$650/kW 0.15 $850/kW 0.1

]

$1050/kW class1

[

class2 class3 0.05

class4 class5 class6

0

-0.05 0.15

class7

0.2

0.25

0.3

0.35

0.4

load factor

124. At present capital costs, even Class 1 sites are not economic (i.e. ERR is less than the opportunity cost of capital taken here at 10%). By 2012, prices should decline to around $850/kW, at which point class 1 and class 2 sites are economic. At the very low cost of $650/kW (which would imply turbine prices of $450/kW, given site costs of around $200/kW), class 3 sites are also economic.38 125. Figure 4.8 shows the same sensitivity analysis but for FIRR. If the minimum rate is set at 15% (nominal), then again even at $850/kW, wind is not financially feasible (without carbon finance).

38

It has been suggested that such low capital costs might be achieved by use of rehabilitated turbines (with a limited performance warranty and manufacturer rehabilitation) now available as European wind site operators upgrade to larger and more efficient machines. This analysis shows that if indeed such turbines can be sourced at 450$/kW, then sites in class 1 and 2 would become both economic and financially feasible. The actual availability of such equipment to South Africa (and the willingness of IPPs to use second hand equipment) should be studied further. In any event, according to the DME study, the GWh available in site Classes 1 & 2 is only 140 GWh (Table 4.17), so the ability to make a large contribution to the REMT targets would appear limited. 49

DRAFT

4. PROJECTS

Figure 4.8: FIRR as a function of capital cost and load factor 0.3

$650/kW 0.2

]

$850/kW 0.1

[

$1050/kW class1

class2 class3 class4 class5

0 class6 class7 -0.1 0.15

0.2

0.25

0.3

0.35

0.4

load factor

126. This can be analysed in a different way, by asking what must be the cost of a CER to achieve 10% ERR. The results are shown in Figure 4.9, where the cost of avoided carbon is also compared to the actual PCF carbon price of 3.75$/ton CO2. At 1,050$, even class 1 sites require more than 3.75$; and at 650$/kW, sites up to class 4 would be economic with carbon finance. Phrased differently, say in the case of $850/kW, from South Africa’s perspective only a class 1 site (i.e. with a load factor of 37%) is economically attractive with carbon finance. Figure 4.9: Cost of avoided carbon 40

avoided cost of carbon, $tonCO2Eq.

class7 30 class6 20

class5 class4 class3

10

$1050/kW class1 $850/kW

0

-10 0.15

$650/kW

0.2

0.25

0.3

0.35

0.4

load factor

Capacity penalties 127. Unlike the other renewable energy options considered in this study, wind is entirely non-dispatchable, and therefore gives rise to capacity penalties to the buyer. This

50

DRAFT

4. PROJECTS

issue is controversial,39 and there is a growing literature on the subject.40 The general rule of thumb that the capacity credit of non-dispatchable renewables may be approximated by the ratio of capacity factors has recently been tested in a detailed study of the North China and Zhejiang grids of China, where detailed WASP model runs showed levels of capacity displacement that were roughly in this proportion (Box 4.4).41 Box 4.4: Capacity penalties for wind energy in China The figure below shows an example from China, in which the impact of some 2,600MW of wind power (located mainly in Inner Mongolia) was forced into the North China Grid (which serves a series of major metropolitan areas including Beijing). Using a capacity expansion optimisation model (WASP), it was found that this 2600 MW of wind capacity displaces only 1132MW of coal and oil-fired CCCT capacity, for an overall capacity credit of around 43%. A similar study was conducted for a small hydro scenario in Zhejiang province, where small hydro was found to have a slightly higher capacity credit of 46%. 3

2632

change in capacity, 1000 MW

2

1

0

-256 -1

-876

-2

coal

CCCT oil

wind hydro

pumped storage

Source: China Renewable Energy Scale-up Program (CRESP): Economic Analysis, Volume I: World Bank, 2003.

39

For example, in the UK, there are significant differences in the cost of backup: the estimate of a recent study by the Royal Academy of Engineering of 1.6p/kWh is eight times the cost in the UK Government’s white paper of 0.2p/kWh. However, even at the Government’s low figure, this represents a capacity penalty of R0.28/kWh.

40

Early estimates of the capacity credit in a range of US systems has been reviewed by M.Grubb and N.Meyer (Wind Energy: Resources, Systems and Regional Strategies, in T. Johansson et al., eds. Renewable Energy: Sources for Fuels and Electricity, Island Press, Washington DC 1992). The capacity credit was found to generally decrease as the level of wind in the system increases. For example, in the Kansas Gas&Electric system the capacity credit falls from 50% at 5% wind penetration (wind MW as % of system peak) to 30% at 20% penetration. At low penetration levels (5-10%), most estimates of capacity credit are between 20-50%, though one study of the PG&E system suggested a capacity credit of 70%. (Grubb and Meyer, Figure 11) More recent analyses (e.g. M. Milligan and B. Parsons, A Comparison and Case Study of Capacity Credit Algorithms for Intermittent Generators National Renewable Energy Laboratory, NREL/CP-44022591, April 1997 (presented at “Solar 97”, Washington, DC, April 27030, 1997); M. Milligan, Measuring Wind Plant Capacity Value, National Renewable Energy Laboratory, 2003), note the difference between capacity penalties for planning and for operations. The former looks at the problem from the perspective of investment planning: the capacity value of a wind plant must be quantified so that investment of conventional generating capacity can be partially offset by the capacity value of the wind plant. The latter looks at the problem from the perspective of the dispatcher, who must schedule conventional resources to ensure adequate generation to meet the load on an instantaneous basis. 41

China Renewable Energy Scale-up Program (CRESP): Economic Analysis, Volume I: World Bank, 2003. 51

DRAFT

4. PROJECTS

128. For the next decade or so, the ESKOM system has adequate spare capacity, so capacity penalties may not apply. However, over time, capacity penalties would begin to apply as new peaking capacity in the system is required. This will to some extent offset the decline in wind turbine prices that can be expected over the same interval. Risk factors 129. Windpower is subject to two additional risk factors that are unlikely to be a major issue in the other renewable energy forms considered in this report: the uncertainties of wind speed measurement upon which wind farm designs are based, and year-on-year variability. It appears that estimates of load factors at project design are significantly biased – which is a problem not just in developing countries (see Box 4.5), but in European countries as well. For example, Figure 4.10 shows a statistical analysis of 1,080 Danish wind turbines for which there is more than six years of annual production data, which plots the actually achieved average load factor since commissioning against the design load factor predicted at the time of the project design. The actual average annual production is 12% lower than predicted. Figure 4.10: The Danish experience of forecasting load factors for wind projects

production error: annual average actual MWh/design MWh

0.4

0.2

0

-0.2

-0.4

-0.6

-0.8 0

0.1

0.2

0.3

0.4

0.5

design capacity factor

Source: Windstats database42 Note: Production error is defined as the ratio of actual annual load factor divded by that estimated at the time of project design.

130. Moreover, when one looks at the time trend, the picture is even more disturbing: Figure 4.11 shows error as a function of start year – which indicates that the more recent the wind farm, the greater the extent of overestimation.

42

Windstats is well-known for its Magazine Windpower Monthly, and for its database (samples of which may be viewed at www.windstats.com). 52

DRAFT

4. PROJECTS

Box: 4.5: Forecasting wind speeds in Sri Lanka The figure shows the annual average wind speeds reported in the feasibility study for a 3MW pilot project in Southern Sri Lanka, estimated at 5.74m/s. The winning bidder for the turbine contract estimated the long-term average, at the actual site, and at the 46metre hub height suitable for its 600kW machine, at 6m/s. The actual average wind speed over the first four years of production was 5.29m/s, which resulted in an average capacity factor of only 14.2%, as opposed to the 18.7% estimated at time of contract award. Power from this project to date is therefore 35% more expensive than anticipated. 6.5

6.1

annual average windspeed, m/s

6.1 6 5.8 5.7 5.6

5.6

5.7 5.6

5.6

5.7 5.6

Feasibility S tudy (H am bantota) 5.5

5.4 5.3

A ctual,@ site 5

4.9

4.5 1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1999

2000

2001

2002

source: Sri Lanka Energy Services Delivery Project: Economic Analysis, Implementation Completion Report, World Bank, May 2003.

Figure 4.11: Deviation from expected load factor v. start year. 0.4

0.2

production error

0

-0.2

-0.4

-0.6

-0.8 1975

1980

1985

1990

1995

2000

2005

startyear

Source: Windstats database

131. There are three hypotheses that can explain this trend: first, that developers and turbine manufacturers have become more optimistic over time; that machines have

53

DRAFT

4. PROJECTS

become significantly more unreliable;43 or, third (and arguably most likely), that there has been a fundamental change in wind regime over the past decade. Yet if changes of wind regime occur over such relatively short time horizons, then wind power is subject to a significant additional risk factor. 132. In South Africa, it may be noted that the Eskom Klipheuwel demonstration turbines (the only large turbines in South Africa that have been operational for between 6 months and 1 year), have shown an average load factor of 18% against an expected load factor of 22% that was based on a mean annual wind speed observed at Koeberg at a height of 10m of approximately 4.25m/s.44 Conclusions on windpower 133. The avoided costs of carbon in a wind project are so high as to make it extremely unlikely that wind can compete in the competitive market for sale of CERs. The Prototype Carbon Fund has signed no deals for wind projects. Countries that have implemented successful wind power programmes have done so only because of very high “Feedlaw” prices (such as in Germany, where Feedlaw prices are around 10 UScents/kWh (R0.69/kWh), or as a result of a combination of mandated market share (MMS) and production tax credits (as in Texas).45 However, the real reason for the successful Texas program is an exceptionally favourable wind regime, with many sites having average annual wind speeds of over 8m/sec, and annual load factors in the 3540% range. However, in this wind speed category there are only some 140 GWh available in South Africa: and if even at the best such sites, carbon finance at 23$/ton is unlikely, then these will be very expensive to otherwise subsidize. 134. Therefore, wind power is not seen as a component of either the first 1,000 GWh or the first 4,000 GWh under REMT. How much wind would be in the first 10,000 GWh of renewable energy is also uncertain, given alternatives such as solar water heating that could be implemented on a very large scale at much lower economic (and 43

There is little evidence of this being so in the case of Danish wind farms. The Windstats database reveals very few hours lost due to maintenance problems: in 2000, the estimated loss of output attributed to this cause was less than 1%. 44

Smit, 2003.

45 The Texas Renewable Portfolio Standard, introduced in 1999, has been one of the most successful programmes for inducing wind power. Legislation requires 2,000 MW of eligible new renewables to be in place by 2007. Eligible are power production from solar, wind, geothermal, wave, tidal, LFG, and biomassbased waste products, as well as renewable energy sources that do not produce but rather replace electricity, such as solar hot water and geothermal heat pumps. Intermediate targets were set at 400MW by 2003, 850 MW by 2005, and 1400 MW by 2007. In fact, by end 2001, the 2005 requirement was already exceeded, of which 915 MW was installed just in that year alone. All electricity retailers in competitive markets (about 80% of the total load) share the obligation based on their proportionate share of yearly electricity sales; publicly owned utilities must only meet the RPS if they opt into competition. The Texas Public Utilities Commission establishes the RPS rules and enforces compliance, while the Independent System Operator (ERCOT) administers any trading in renewable energy certificates (RECs). The certificates are issued on production, in 1 MWh units. Banking is allowed after the year of issuance, and borrowing up to 5% of the obligation in the first two compliance period is allowed. Penalties for non-compliance are stiff, being the lesser of 5UScents/kWh or 200% of the mean REC trade value for each missing kWh. Most of the renewable energy induced by this system are wind plants in west Texas, where mean wind speeds of 8metres/sec and capacity factors of 40% are common. The sizable purchase obligation makes possible large wind projects that can exploit economies of scale. The key, however, is the Production Tax Credit: with a credit of 1.7 UScents/kWh (R 0.12/kWh), wind projects are able to deliver power to the grid for less than 3 cents/kWh, which is very close to the avoided cost in the grid, and very close to generation from new natural gas projects.

54

DRAFT

4. PROJECTS

financial) costs. However, as discussed in Section 5, it is assumed that the DME assessment of wind energy in sites of class 1, 2 and 3 (totalling 308 GWh) would be part of the 10,000 GWh target.

Pulp and Paper Industry 135. Sappi and Mondi own all of the pulp and paper mills in South Africa. Table 4.22 lists the nine major pulp and paper mills owned by these two organisations. 46 Table 4.22: Major pulp and paper mills in South Africa Mill Name

Pulp Products

Mondi

Richards Bay Piet Retief Felixton Merebank

Sappi

Ngodwana

Hardwood and softwood kraft pulp Hardwood and softwood pulp Unbleached bagasse pulp Thermo-mechanical pulp Groundwood pulp Hardwood and softwood pulp Goundwood pulp Unbleached softwood pulp Hardwood NSSC pulp Bleached bagasse pulp Bleached hardwood pulp Dissolving pulp (hardwood)

Tugela Stanger Enstra Saiccor

2002 output (1000tons /year) 576 60 70 220 66 410 100 230 120 60 90 600

136. Pulp mill waste potentials include partially combusted boiler ash, bark, sludge, sawdust and black liquor. There are two basic processes that the major mills use to convert wood into pulp: a mechanical process and a chemical process. Table 4.23 summarises the pulp yields of these two processes. The mechanical process is applicable only to softwood (i.e. pine) and this process generates about 5% waste at the mill, along with the bark that is stripped of the logs at the mill. Together, these are known as “hog waste.” Table 4.23: Pulping processes Mechanical (Soft Wood only) Pulp Fibre Wood Waste (hog waste) Black Liquor

90% 5% NA

Chemical Soft Wood 40% - 50% NA 50% - 60%

Hard Wood 75% NA 25%

137. The chemical process, which is mostly used for hardwoods, does not generate any hog waste. However, it does produce a fluid, called black liquor, that is formed from the chemical processing. This is burned in a special boiler to recover the chemicals. These boilers raise steam for power generation and production steam. Hardwood (eucalyptus) is stripped of its bark in the forests and generates no wood waste at the mill site.

46

In addition to these major mills, there are at least seventeen finishing mills that use pulp supplied from the major mills to produce a range of paper products such as crepe tissue, linerboard and fine papers. These mills do not have waste products in significant quantities that could be used for electricity generation. 55

DRAFT

4. PROJECTS

138. This study has not attempted to model electricity generation from black liquor as a separate resource. Although it appears that there is considerable energy waste in terms of black liquor, in reality most mills burn the black liquor to recover the chemicals and extract the waste heat. Where this is not being done, the resource is considered (by the mills) to be too expensive to recover the energy or the chemicals. This is a limited (internal to the mills) resource of energy. 139. Some of the above pulp mills already burn part of their waste material (hog waste) to generate steam for the generation of power and for production steam. From the above information, it may be deduced that the only mills that have any hog waste in usable quantities will be: Richards Bay

Hardwood and softwood kraft pulp

Piet Retief Ngodwana

Hardwood and softwood pulp Hardwood and softwood pulp Groundwood pulp

Tugela

Unbleached softwood pulp Hardwood NSSC pulp

140. Mondi and Sappi were asked to indicate the quantity of hog waste available at each of their mills. Both companies were willing to provide the information but it is clear that not all mills keep an accurate record of their hog waste. Mondi advised that they plan to burn all the available hog waste at their Richards Bay mill and are proceeding with a project to increase steam production and electricity generation and to make use of carbon credits. All electricity generated will be used in the plant. 141. This study is based on a detailed analysis of the costs of upgrading Sappi’s Ngodwana and Tugela mills. These are the only mills for which reasonably accurate, reliable input data could be obtained. Sappi Ngodwana 142. 100,000 tons of hog waste are assumed to be available per year (which has an estimated calorific value of 9.24 MJ/kg). The conversion efficiency is assumed at 27%. Given continuous year round operation, the load factor will be high (assumed at 90%), resulting in an exportable capacity of 8.3MW (total MW at the plant increases from 85 to 93 MW). Since the only equipment required is additional waste heat boiler capacity (as the turbo-alternators and ancillary plant exists) the capital cost is a low $733/kW. 143. The estimated financial rate of return (nominal) is estimated at 23.4% (Table 4.24), which makes it a prime candidate for inclusion in the REMT phase I target. 47 There appear to be few risk factors: the main one would be any conversion to hardwood (which produces less waste), but this is considered unlikely over the next decade and could be mitigated by using material from large quantities of waste disposed in landfills.

47

Sappi (and Mondi) are multinational companies well versed in sophisticated financial hedging, and are probably using a variety of swaps and hedges to lower their domestic lending cost through offshore loans: therefore an interest rate of 6% is used here. 56

DRAFT

4. PROJECTS

Table 4.24: Financials for the Ngodwana Pulp Mill hog waste utilisation project [NPV] capital cost capacity capital cost

733 8.3 6087 42.0

disbursement profile total debt:ZAR 0.50 debt:FOREX 0.00 equity 0.50 GEF Grant/Capital subsidy 0.00 auxilary consumption 0.0% transmission loss 0.0% renewable generation 90.0% supplemental generation 0.0% total sendout 90.0% revenues average tariff PPA sales [Table 17] green premium/kWh subsidy CER sales [power][Table 18] CER sales [methane][Table 18] total revenue levelised revenue costs supplemental fuel cost O&M costs[as%capital] 2.0% O&Mcosts 0.5 debt service [Table 16] Equity 21.0 total costs total financial flows, before tax income tax financial flows, after tax FIRR nominal

[$/kW] [MW] [1000$] [Rmill.] [ ] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.]

42.0

39.5 18.7 0.0 18.7 0.0

[GWh] [GWh] [GWh]

460.4 0.0 460.4

[R/kWh] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [R/kWh]

128.6 0.0 0.0 0.0 128.6 0.28

[Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [ ]

[-1] [0] [1] [2] [3] [4] [5] [6] [7] 2005 2006 2007 2008 2009 2010 2011 2012 2013

35.1 0.8 35.1 16.8 0.0 16.8 0.0

44.3 0.2 9.2 4.2 0.0 4.2 0.0 65.4 0.0 65.4

0.0

65.4 0.0 65.4

65.4 0.0 65.4

65.4 0.0 65.4

65.4 0.0 65.4

65.4 0.0 65.4

0.147 0.161 0.176 0.198 0.221 0.236 0.261 9.6 10.5 11.5 12.9 14.5 15.5 17.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 9.6 10.5 11.5 12.9 14.5 15.5 17.1

-9.2 -18.8 -18.7 -52.3 76.3 -32.0 44.3 23.4%

65.4 0.0 65.4

-16.8 -16.8 -16.8

-1.8 -4.2 -6.0 -6.0

-16.8

-6.0

0.00 -0.96 -0.57 -4.7

0.00 -1.00 -0.60 -4.5

0.00 -1.05 -0.62 -4.2

0.00 -1.09 -0.65 -3.9

0.00 -1.14 -0.68 -3.7

0.00 -1.19 -0.71 -3.4

0.00 -1.25 -0.74 -3.2

-6.3 3.4 -1.2 2.2

-6.1 4.5 -1.6 2.9

-5.9 5.6 -2.0 3.7

-5.7 7.3 -2.5 4.7

-5.5 9.0 -3.1 5.8

-5.3 10.1 -3.5 6.6

-5.1 12.0 -4.2 7.8

Sappi Tugela 144. It is estimated that at least 60,000 tons of hog waste are produced annually at the Tugela mill. In this case, the equipment required is a complete facility including a waste heat, wood chip burning boiler, turbo-alternator and ancillary plant, which results in a much higher capital cost of around $1,650/kW. The resulting FIRR is a low 9.5% (Table 4.25), which is not feasible. To bring the FIRR to 21% would require $21.2/tonCO2 : at the available $3.75/ton the FIRR increases to only 12% (nominal).

57

DRAFT

4. PROJECTS

Table 4.25: Financials for the Tugela pulp mill hog waste utilisation project [NPV] capital cost capacity capital cost

1652 5 8261 57.0

disbursement profile total debt:ZAR 0.50 debt:FOREX 0.00 equity 0.50 GEF Grant/Capital subsidy 0.00 auxilary consumption 0.0% transmission loss 0.0% renewable generation supplemental generation total sendout revenues average tariff PPA sales [Table 17] green premium/kWh subsidy CER sales [power][Table 18] CER sales [methane][Table 18] total revenue levelised revenue costs supplemental fuel cost O&M costs[as%capital] 2.0% O&Mcosts 0.75 debt service [Table 16] Equity 28.5 total costs total financial flows, before tax income tax financial flows, after tax FIRR nominal

[$/kW] [MW] [1000$] [Rmill.] [ ] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.]

[GWh] [GWh] [GWh]

[-1] [0] [1] [2] [3] [4] [5] [6] [7] 2005 2006 2007 2008 2009 2010 2011 2012 2013

57.0

52.4 24.4 0.0 24.4 0.0

20.9 0.4 20.9 10.0 0.0 10.0 0.0

61.3 0.6 40.4 18.5 0.0 18.5 0.0

277.4 0.0 277.4

[R/kWh] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [R/kWh]

78.1 0.0 0.0 0.0 78.1 0.28

[Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [ ]

-12.5 -8.3 -24.4 -24.4 -69.6 8.5 -10.1 -1.6 9.5%

39 0 39

0.0

39 0 39

39 0 39

39 0 39

39 0 39

39 0 39

39 0 39

0.150 0.159 0.173 0.189 0.205 0.230 0.257 5.9 6.3 6.8 7.4 8.1 9.1 10.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5.9 6.3 6.8 7.4 8.1 9.1 10.1

-10.0 -10.0 -10.0

-1.1 -18.5 -19.6 -19.6

-10.0

-19.6

0.0 -1.3 -0.9 -6.4

0.0 -1.4 -0.9 -6.1

0.0 -1.4 -0.9 -5.7

0.0 -1.5 -1.0 -5.3

0.0 -1.6 -1.0 -5.0

0.0 -1.6 -1.1 -4.6

0.0 -1.7 -1.1 -4.3

-8.6 -2.7 0.0 -2.7

-8.3 -2.1 0.0 -2.1

-8.1 -1.2 0.0 -1.2

-7.8 -0.4 0.0 -0.4

-7.6 0.5 0.0 0.5

-7.3 1.8 0.0 1.8

-7.1 3.0 0.0 3.0

Conclusions on pulp and paper 145. Pulp and paper mill waste will be a difficult sector to undertake renewable energy projects. It is clear that only in special circumstances – as exist at the Sappi Ngodwana Mill – will there be attractive projects for inclusion in REMT. Certainly, the Ngodwana project would be a strong candidate for REMT Phase I, to which it would contribute 6.5% of the total 1,000 GWh target. Given the likely evolution of pool prices (even under deregulation), no additional pulp&paper industry projects in Phase II of REMT are envisaged. 146. However, uncertainty in the volume of available waste means that conservative figures have been used for these mills. None of the mills appear to keep accurate records of waste as this is of little significance to them under the present circumstances. A detailed study of the actual waste available and the accumulated waste stored in landfills is a top priority and could result in significantly different figures. 147. Clearly a new technical approach will need to be found to convert pulp&paper industry waste to energy at lower capital costs. One possibility would be recover the energy in landfills. While it is known that the industry has land-filled large quantities of waste in the past, there is little information about quantities or methane recovery, in part because the industry has had little incentive to keep records about what it has traditionally regarded as a pure waste product. It is recommended that an investigation of the energy potentials of pulp and paper mill landfill sites should be undertaken in the detailed resource assessment to be conducted under REMT.

58

DRAFT

4. PROJECTS

148. It is important to note that the waste remaining in the forests amounts to more than three times the volumes at the mills. This waste is difficult to access, costly to transport and expensive to convert to electricity. It can be converted to charcoal but, unlike most countries to the North, there is no market for charcoal in South Africa (except for barbeque purposes). However, the ‘Working for Water’ project has shown that these problems are not insuperable and employment can be generated from these labour intensive activities. The forest owners would also benefit from removal of the waste as unauthorised collection of firewood is a potential fire hazard in the forests. It is recommended that a more detailed resource assessment should be conducted as part of REMT.

Solar water heating Background 149. South Africa experiences some of the highest levels of solar radiation in the World. The average daily solar radiation in South Africa varies between 4.5 and 6.5 kWh/m2 (16 and 23 MJ/m2), compared to about 3.6 kWh/m2 for parts of the United States and about 2.5 kWh/m2 for Europe and the United Kingdom. Figure 4.12 shows the geographical distribution of incoming solar radiation. 48 Figure 4.12: Annual Direct and Diffuse Solar Radiation

48

South African Renewable Energy Resource Database. CSIR. 1999 59

DRAFT

4. PROJECTS

SWH market transformation 150. A scenario for the transformation of the solar water heater market over a 20-year period has been developed, based on a proactive government SWH policy as successfully implemented in Austria (where the climate is colder and sunshine less), Cyprus (where climatic conditions are comparable) and Botswana. 151. As shown in Figure 4.13, the predicted South African population growth tapers off slightly beyond 52M. The inland SWH market reaches saturation at 0,86m² collector area/person, stagnating in 2017. The maximum annual shipment rate reaches a maximum of 6Mm² where it stabilises, supported by the export and local replacement markets. The average life expectancy of SWHs is assumed at 20 years. The annual growth of new SWHs moves out of a linear mode when about 16% of potential market is reached. A rapid growth phase follows until annual addition peaks at 6Mm² before replacement and export markets set in. The total accumulated installed collector area shows the characteristic Gomperz curve. Figure 4.13: SWH market transformation 50,000,000

40,000,000

Total population Total installed collector area m² Annual addition m² Annual addition export m² Annual replacement Total annual output

30,000,000

20,000,000

10,000,000

0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021

Assumptions: • •

• •

South African population projection by van Aard, C.J. UNISA.. Historic SA SWH data from W. Cawood, 2002. SWH Discussion Paper. Durban Projections of total (glazed and unglazed) area of 0,86m²/person achieved in 20 years based on Kalogirou, S.A. 2002. (Parabolic trough collectors for industrial process heat in Cyprus. Energy 27:815). Cyprus has similar elimate conditions to South Africa, but is farther from the equator than Cape Town .The total South African SWH production twenty years ago has been twice the current figure.

Source: D. Holm, Market Penetration of Commercial and Residential SWHs in South Africa, and Energy Output, June 2004.

152. Assuming an average of 5,79kWh/m²/day solar radiation on solar collectors tilted at an average altitude angle of 30°, and with an average annual conversion efficiency of 45%,49 the equivalent electricity yield is 951 kWh/m²/year. When superimposed on the total installed collector area of Figure 4.13, the corresponding scenario for energy output is as shown in Figure 4.14. It appears that the 10,000 GWh/year RE target stated in the White Paper can be reached in less than ten years by 49

W. Cawood, 2002. SWH Discussion Paper. Durban, p. 5. 60

DRAFT

4. PROJECTS

actively promoting the use of residential and commercial Solar Water Heaters in South Africa Figure 4.14: Energy from SWH (in GWh/year) GWh/a & '000m² 50 000

Energy of SWH [GWh/a] 45 000 40 000 35 000 30 000 25 000 20 000 15 000 10 000 5 000

Annual energy output of SWH [GWh/a] Tot installed collector area [ '000m²]

19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 20 17 20 18 20 19 20 20 20 21

0

Contribution to the 1000 GWh target 153. As noted in the introduction, the SWH sector differs from the other technologies examined in this report because the scale of individual transactions is an order of magnitude smaller. The representative projects examined above are financially feasible, but displace only small amounts of electricity: the displacement of any particular project will be measured in MWh, not GWh. On the other hand, the number of potential transactions is very large, as suggested in the macro-market assessment presented in the previous section. Therefore, a top-down approach is needed to establish a lower bound for the contribution of SWH. 154. In 2004 total shipments of collectors for the South African Market were estimated at 35,000m2, of which 70% are commercial. Based on assumption of 951 KWh/year per m2 of collector area, and a market growth rate of 6%, then under such business-as-usual conditions the incremental energy produced from collectors installed in 2005 to 2008 is 154GWh (Table 4.26).

61

DRAFT

4. PROJECTS

Table 4.26: SWH forecast 2004 2004 actual lower bound, business as usual postulated growth rate additions per year additional collector area installed cumulative new collector area efficiency displaced energy REMT postulated growth rate additions per year additional collector area installed cumulative new collector area efficiency displaced energy

[m2/year]

2005

2006

2007

2008

0.060 37100 37100 37100 951 35

0.060 39326 39326 76426 951 73

0.060 41686 41686 118112 951 112

0.060 44187 44187 162298 951 154

0.200 42000 42000 42000 951 40

0.300 54600 54600 96600 951 92

0.400 76440 76440 173040 951 165

0.400 107016 107016 280056 951 266

35000

[ ] [ ] [m2/year] [m2] [kWh/m2/year] [GWh] [ ] [ ] [m2/year] [m2] [kWh/m2/year] [GWh]

35000

35000

155. Under REMT the growth rates may be much greater. At the moment there is substantial excess manufacturing capacity, and it is estimated that production could be tripled without any additional capital expenditure or extra shifts. The sharply higher growth rates in the REMT scenario are therefore not unreasonable, leading to a 2008 GWh equivalent of 266 GWh. Moreover, state-of-the art collectors have efficiencies approaching 55%, as opposed to the 45% assumed here. The study makes the conservative assumption that commercial solar water heating would account for 175 GWh to be included in REMT-I.

Landfill Gas 156. The 57 landfill sites modelled in the DME study (see Table 4.26, overleaf) were divided into four categories (micro, small, medium and large), based on the estimated methane gas yield of each site. Table 4.27 shows the key statistics of each group. In calculating the kW capacities of the sites modelled, a very conservative approach was adopted. The DME Study noted that it is quite possible that the estimates may be only 50% of the true potential that could be available from these sites. Table 4.27: DME study categories Category 1 Category 2 Category 3 Category 4 micro small medium large KW capacity per SEGP unit Number of SEGP units Total kW capacity Annual load factor Auxiliary consumption Annual production output (GWh)

646 37 23,900 92% 1% 191

2,000 10 20,000 92% 1% 160

3,000 9 27,000 92% 1% 215

4,000 1 4,000 92% 1% 32

Total

57 74,900

598

62

DRAFT

4. PROJECTS

Table 4.28: Landfill sites considered by the DME study province Site Name General Start 2003 gas waste production WC WC WC WC WC WC WC FS FS FS FS FS FS FS GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT GT MP MP MP EC EC EC EC KZ KZ KZ KZ KZ KZ KZ KZ NC

Bellville Park Bellville South Vissershok - Wtech Swartklip waste site Vissershok waste site - CCC Coastal Park Landfill Mossgas regional waste site Bothaville Northern Bloem Southern Bloem Bethlehem Welkom Kroonstad Sasolberg Hatherly Garstkloof Kwaggasrand Onderstepoort Valhalla Ga Rankua Soshanguve Temba Derdepoort Mamelodi Bon Accord Roslynn Waldrift Palm Springs Borrow Pits KwaggaStroom Zuurfontein Luipaardsvlei West Wits Site Holfontein Weltevreden Platkop Simmer & Jack Rooikraal SAPPI Enstra Springs Rietfontein Springs Scaw Metals Robinson Deep Goudkoppies Linbro Park Marie Louise Randburg/ Kya Sands Northern Works/Ennerdale Nuffield TSB Malelane Graspan Middleburg Leuupoort Witbank Koedoeskloof Uit/Dispatch Arlington PE Aloes Second Creek/Reg EL Bisasar Road Pinetown Tongaat/La Mercy Richards Bay Alton Site Alusaf Industrial New England Shongweni Bulbul Drive Kimberley

167700 340000 243200 245000 201000 176000 34060 177762 70200 140000 46800 70400 36000 45000 120444 421080 323856 336396 345192 153816 110400 88356 342540 60000 91366 32000 160000 45000 37000 94000 38000 234000 234000 117260 300000 100000 242000 340000 80000 203056 260000 318000 347000 316000 288000 260000 83000 163000 152000 35000 85000 53313 195000 33341 104000 416000 156000 66000 130000 50000 140000 75000 128030 130000

1985 1970 1974 1976 1976 1985 1990 1990 1990 1990 1990 1990 1990 1990 1998 1980 1965 1998 1979 1995 1995 1995 1998 1980 1980

552 1069 1487 1475 1210 936 156 815 322 642 215 323 165 206 292 1822 1485 815 1489 527 378 303 830

1990 1990 1990 1990 1990 1990 1997 1992 1995 1992 1983 1988 1990 1997 1 1980 1989 1989 1991 1 1988 1 1990 1990 1990 1989 1982 1990 1984 1981 1997 1992 1990 1 1981 1997 1989 1990

734 206 170 431 174 1073 1073 491 1027 418 1338 1672 367 566 1837 1651 1504 1265

KWe 2003 1048.8 2031.1 2825.3 2802.5 2299 1778.4 296.4 1548.5 611.8 1219.8 408.5 613.7 313.5 391.4 554.8 3461.8 2821.5 1548.5 2829.1 1001.3 718.2 575.7 1577 0 0 0 1394.6 391.4 323 818.9 330.6 2038.7 2038.7 932.9 1951.3 794.2 2542.2 3176.8 697.3 1075.4

697 160 390 254 1096 153 564 2372 435 414 596

3490.3 3136.9 2857.6 2403.5 0 775.2 0 1324.3 304 741 482.6 2082.4 290.7 1071.6 4506.8 826.5 786.6 1132.4

798 209 587 596

1516.2 397.1 1115.3 1132.4

408

Boldface indicates site studied in this report

63

DRAFT

4. PROJECTS

Assessed Projects 157. In this study the following projects have been assessed: •

• •

Durban (Ethekwini) Project. The three landfill sites at Bisasar Road, Mariannhill and La Mercy are treated here as a single project. This project is very close to obtaining PCF carbon financing at $3.75/ton CO2, plus an additional 0.20$/ton for social benefits. The project has been extensively documented elsewhere.50 Eersterust (Pretoria).51 Pietermaritzburg: The New England Road Landfill site situated on the flood plain of the uMsundusi River SE of the City of Pietermaritzburg is owned and operated by the Msundusi Municipality. Operations started in the early fifties with poor management and little control. The site was permitted in the early 1980s,and greatly improved control and management introduced. An initial gas extraction scheme was installed of 8 barrier wells and a 300 Nm3/h flare in June 2001. Two additional production wells were also installed. Previously, a 4 well gas pumping trial had been conducted in 1999. The 8 wells yielded some 300 Nm3/h or an average of 38 Nm3/h/well. A total of 20 wells of this capacity are required to supply gas for generation of 1 MW of power. DBSA is currently undertaking a feasibility study on power generation/gas utilisation at this site. This study is being undertaken by a consortium of Palmer Development Group/ Envitech/Rob Short/Lawyers. The report is not yet available. Based on the previous pumping trial and recorded yields from the 2001 scheme up to November 2002 and the 1999 pumping trial, the current gas yield is predicted as 1,100 Nm3/h rising to 2,500 Nm3/h at completion.



Eastern Gauteng1: The EMM Weltevreden Landfill is situated in the Eastern Gauteng area on Main Reef Road Brakpan. Coordinates (WGS 84) are 26o18.51” S and 28o15.88”E. The site accepts a mix of domestic (63%) and industrial waste. Present input rate is around 240,000 tonnes per annum. The site started in 1995, and is projected to close in 2024 with and estimated input of around 6.7 million tonnes at closure. Projected collectable gas yield and maximum power generation based on limited waste data available are 3600 Nm3/h 6MWe. Agaricus Trading Company is currently installing a gas pumping trial at this site. The trial is due to commence pumping in July 2005. The results of this trial are likely to be available by Feb 2004. An initial project design would see 3 x 1MW engines installed in 2005/2006 plus further 3 x 1 MW engines installed in 2007/200/, 2010/2011 and 2016/2017.



Eastern Gauteng 2. The Simmer & Jack Landfill is situated in the Eastern Gauteng area in Germiston (EMM2). Coordinates (WGS 84) are 26o12.26” S and 28o08.41”E. The site accepts a mix of domestic (43%) and industrial waste. Present input rate is around 270,000 tonnes per annum. The site started in 1983 and is projected to close in 2009, with an estimated input of around 6.1 million tonnes at closure. Projected collectable gas yield and maximum power generation based on limited waste data available are 2,100 Nm3/h and 3Mwe, respectively. Agaricus Trading is currently installing a gas pumping trial at this site. The trial is due to commence pumping in July 2004. The results of this trial

50

See e.g. Prototype Carbon Fund, Durban, South Africa Landfill Gas to Electricity, Baseline Study, July 2003; Emission reduction Study, July 2003; Project Design Document, July 2003. 51

ENVIROGAS, Investigation of Eersterust disposal site for landfill gas generation potential and the benefit of gas management, Report to Greater Pretoria Metropolitan Council, 2003. 64

DRAFT

4. PROJECTS





are likley to be available by Feb 2005. An initial project design would see 2 x 1MW engines installed in 2005/2006 plus a further 1 x 1 MW engine installed in 2010/2011. Eastern Gauteng 3: The EMM Rietfontein Springs Landfill is situated in the Eastern Gauteng area off Tonk Meter Road, Springs. Coordinates (WGS 84) are 26o18.02” S and 28o25.59”E . The site accepts a mix of domestic (62%) and industrial waste. Present input rate is around 140 000 tonnes per annum. The site started in 1997, and is projected to close in 2017 with an estimated input of around 3.1 million tonnes at closure. Projected collectable gas yield and maximum power generation, based on its limited waste data available are 2,000 Nm3/h and 3Mwe, respectively. Agaricus Trading Company is currently installing a gas pumping trial at this site. The trial is due to commence pumping in July 2004. The results of this trial are likely to be available by Feb 2004. An initial project design would see 1 x 1MW engines installed in 2005/2006, plus a further 2 x 1 MW engine installed in 2009/2010 and 2017/2018. Eastern Gauteng 4: The EMM Rooikraal Landfill is situated in the Eastern Gauteng area off Barry Marais Road, Brakpan. Coordinates (WGS 84) are 26o18.58” S and 28o15.88”E . The site accepts a mix of domestic (65%) and industrial waste. Present input rate is around 300,000 tonnes per annum. The site started in 1988 and is projected to close in 2026 with an estimated input of around 11.7 million tonnes at closure. Projected collectable gas yield and maximum power generation based on the limited waste data available, are 5,200 Nm3/h and 7Mwe, respectively. Agaricus Trading is currently installing a gas pumping trial at this site. The trial is due to commence pumping in July 2005. The results of this trial are likely to be available by Feb 2005. An initial project design would see 5 x 1MW engines installed in 2005/2006 plus further 2 x 1 MW engine installed in 2008/2009 and 2016/2017.

158. Table 4.29 shows the financials for the Eersterust Landfill (near Pretoria). Given the uncertainties involved, a 15-year life is assumed for all LFG projects.52 The FIRR is 10% nominal, which would not suffice to undertake the project without carbon finance. 159. A significant uncertainty is the evolution of gas flows and power generation: in the above case this study assumes the yield begins to fall in year 6, with an annual rate of decline of 7.5% (and thus electricity generation declines from 6.4GWh in year 1 to 2.9GWh in year 15). It should however be noted that the study has simplified the financial model for LFG projects, and have not incorporated the complicated phasing and well development programmes that would actually apply. For example, CERs from the Bisasar Road well (Durban) are as shown in Figure 4.15.

52 However, note that the Durban contract with the PCF envisages a 21-year contract period. However, the Durban project is at this point extremely well researched.

65

DRAFT

4. PROJECTS

Table 4.29: Financials for the Eersterust LFG site (near Pretoria) [NPV] capital cost capacity capital cost

966 0.9 870 6.0

disbursement profile total debt:ZAR 0.70 debt:FOREX 0.00 equity 0.30 GEF Grant/Capital subsidy 0.00 auxilary consumption 5.0% transmission loss 0.0% renewable generation 80.8% supplemental generation 0.0% total sendout 80.8% revenues average tariff PPA sales [Table 17] green premium/kWh subsidy CER sales [power][Table 18] CER sales [methane][Table 18] total revenue levelised revenue costs supplemental fuel cost O&M costs[as%capital] 0.0% O&Mcosts 0.5 debt service [Table 16] Equity 1.8 total costs total financial flows, before tax income tax financial flows, after tax FIRR nominal

[$/kW] [MW] [1000$] [Rmill.] [ ] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.]

[-1] [0] [1] [2] [3] [4] [5] [6] [7] 2005 2006 2007 2008 2009 2010 2011 2012 2013

6.0

5.4 3.5 0.0 1.5 0.0

[GWh] [GWh] [GWh]

34.2 0.0 34.2

[R/kWh] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [R/kWh]

10.0 0.0 0.0 0.0 10.0 0.29

[Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [Rmill.] [ ]

0 0.0 0.0 0.0 0.0 0.0 0.0

6.55 1.0 6.6 4.2 0.0 1.8 0.0 6.4 0.0 6.4

0.0

6.4 0.0 6.4

6.4 0.0 6.4

6.4 0.0 6.4

5.9 0.0 5.9

5.4 0.0 5.4

0.193 0.209 0.226 0.250 0.276 0.293 0.321 1.2 1.3 1.4 1.6 1.8 1.7 1.7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.2 1.3 1.4 1.6 1.8 1.7 1.7

0.0 -3.5 -1.5 -9.6 0.4 -0.4 0.0 10.0%

6.4 0.0 6.4

0.0 0.0 0.0

0.0 -1.8 -1.8 -1.8

0.0

-1.8

0.00 0.00 -0.57 -0.9

0.00 0.00 -0.60 -0.9

0.00 0.00 -0.62 -0.8

0.00 0.00 -0.65 -0.8

0.00 0.00 -0.68 -0.7

0.00 0.00 -0.71 -0.7

0.00 0.00 -0.74 -0.6

-1.5 -0.3 0.0 -0.3

-1.5 -0.2 0.0 -0.2

-1.5 -0.0 0.0 -0.0

-1.4 0.2 0.0 0.2

-1.4 0.3 0.0 0.3

-1.4 0.3 0.0 0.3

-1.4 0.4 -0.1 0.3

Figure 4.15: CERs at Bisasar Road Landfill, Durban

1000 CERs from electricity production

60

50

40

30

20

10 2004

2006

2008

2010

2012

2014

2016

2018

2020

2022

2024

Source: PCF, Emissions reduction study

160. Figure 4.16 shows the relationship between the rate of decline and ERR and FIRR: the impact on FIRR is seen to be much greater than on ERR.

66

DRAFT

4. PROJECTS

Figure 4.16: Rate of decline of gas production v. FIRR and ERR 0.4

ERR

]

0.3

[

0.2 FIRR(before carbon finance)

0.1

0 0

0.05

0.1

0.15

0.2

annualrate ofdecline afteryear6

LFG supply curve 161. Table 4.30 summarizes the results for the LFG projects, for which the total energy is 242 GWh. Figure 4.25 shows the relationship between cumulative energy and FIRR. Table 4.30: Summary of results for LFG projects GWh MW FIRR ERR ERR, /year before with environment environment

cumul GWH

Pietermaritzburg

27

3.5

0.0%

4.2%

27.7%

27

Eersterust

6

0.9

10.0%

5.0%

30.6%

33

DurbanLFG

62

8.0

11.3%

5.6%

29.4%

95

EEM2

23

3.0

12.2%

6.0%

31.5%

119

EEM3

23

3.0

12.2%

6.0%

31.5%

142

EEM1

46

6.0

14.3%

7.0%

32.3%

188

EEM4

54

7.0

14.9%

7.2%

32.5%

242

67

DRAFT

4. PROJECTS

Figure 4.17: Landfill gas Cumulative GWh v. FIRR 0.2

FEASIBLE WITHOUT CARBON FINANCE (FIRR>15%)

0.15

FIRR

0.1

0.05

0 UNLIKELY (FIRR

Suggest Documents