OUTLOOK FOR THE COAL VALUE CHAIN: SCENARIOS TO 2040 ... · TECHNICAL REPORT. TECHNICAL REPORT: SACRM SCENARIOS TO 2040 | i DISCLAIMER The statements and views of the South African

!"#$%

&'()*&+

*"&,

("&-.&/

JULY 2013

OUTLOOK FOR THE COAL VALUE CHAIN:

SCENARIOS TO 2040

TECHNICAL REPORT

TECHNICAL REPORT: SACRM SCENARIOS TO 2040 | i

DISCLAIMER

The statements and views of the South African Coal Roadmap are a consensus view of the participants in the

development of the roadmap and do not necessarily represent the views of the participating members in their

individual capacity. An extensive as reasonably possible range of information was used in compiling the roadmap;

all judgments and views expressed in the roadmap are based upon the information available at the time and

remain subject to further review. The South African Coal Roadmap does not guarantee the correctness, reliability

or completeness of any information, judgments or views included in the roadmap. All forecasts made in this

document have been referenced where possible and the use and interpretation of these forecasts and any

information, judgments or views contained in the roadmap is entirely the risk of the user. The participants in the

compiling of this roadmap will not accept any liability whatsoever in respect of any information contained in the

roadmap or any statements, judgments or views expressed as part of the South African Coal Roadmap.

The analysis underpinning the Coal Roadmap contained in this report was prepared by The Green House

(www.tgh.co.za) for the South African Coal Roadmap Steering Committee. The Roadmap and accompanying

scenarios and technical reports are based on information, views and data provided to The Green House,

supplemented by information obtained from the open literature. The views expressed in this document thus do not

reflect those of The Green House.

Cover photos Copyright Sasol and Exxaro, 2013.

TECHNICAL REPORT: SACRM SCENARIOS TO 2040 | ii

TABLE OF CONTENTS

DISCLAIMER ............................................................................................................................................................. I

1 INTRODUCTION ................................................................................................................................................ 1

2 DESCRIPTION OF THE SCENARIOS .............................................................................................................. 1

2.1 Coal use in South Africa .............................................................................................................................. 3

2.1.1 Electricity generation ............................................................................................................................. 3

2.1.2 Coal-to-liquids ..................................................................................................................................... 16

2.1.3 Other uses of coal ............................................................................................................................... 16

2.2 Carbon capture and storage (CCS) ........................................................................................................... 18

2.3 Coal supply ................................................................................................................................................ 23

2.4 Coal exports ............................................................................................................................................... 25

3 IMPLICATIONS OF THE SCENARIOS ........................................................................................................... 28

3.1 Resources and reserves ............................................................................................................................ 28

3.2 Implications of the electricity generation build plans .................................................................................. 31

3.3 Economic implications of the scenarios ..................................................................................................... 32

3.3.1 Electricity generation infrastructure investment and electricity generation cost .................................. 32

3.3.2 Coal price and revenue from coal sales .............................................................................................. 42

3.3.3 Global competitiveness ....................................................................................................................... 45

3.3.4 The cost of climate adaptation ............................................................................................................ 46

3.4 Energy Security ......................................................................................................................................... 46

3.4.1 Reliance on local resources versus energy imports ............................................................................ 47

3.4.2 Technology considerations ................................................................................................................. 47

3.5 Employment and other socio-economic considerations ............................................................................ 48

3.5.1 Mining .................................................................................................................................................. 48

3.5.2 Electricity generation ........................................................................................................................... 48

3.5.3 Coal-to-Liquids .................................................................................................................................... 52

3.5.4 Richards Bay Coal Terminal ............................................................................................................... 52

3.5.5 Transnet .............................................................................................................................................. 52

3.5.6 Results and analysis: Employment under the SACRM scenarios ....................................................... 52

3.5.7 Other socio-economic considerations ................................................................................................. 57

3.6 Water demand ........................................................................................................................................... 57

3.6.1 Mining and beneficiation ..................................................................................................................... 57

3.6.2 Electricity generation ........................................................................................................................... 58

3.6.3 Coal-to-liquids ..................................................................................................................................... 59

TECHNICAL REPORT: SACRM SCENARIOS TO 2040 | iii

3.6.4 Carbon Capture and Storage .............................................................................................................. 59

3.6.5 Communities ....................................................................................................................................... 60

3.6.6 Results and analysis: Water demand under the SACRM scenarios ................................................... 60

3.7 Infrastructure .............................................................................................................................................. 67

3.7.1 Transport infrastructure ....................................................................................................................... 67

3.7.2 Water supply infrastructure and catchment management ................................................................... 70

3.8 Greenhouse gas (GHG) emissions ............................................................................................................ 72

3.8.1 Assumptions relating to GHG emissions ............................................................................................. 72

3.8.2 Results and analysis: Greenhouse Gas Emissions ............................................................................ 75

3.9 Environmental implications ........................................................................................................................ 77

3.9.1 Water provision, land and biodiversity ................................................................................................ 77

3.9.2 Solid waste generation ........................................................................................................................ 80

3.9.3 Non-GHG emissions ........................................................................................................................... 86

4 SENSITIVITY ANALYSES ............................................................................................................................... 89

4.1 Replacing the first nuclear plant with gas CCGT ....................................................................................... 90

4.1.1 Electricity supply infrastructure investment and electricity generation cost ........................................ 91

4.1.2 Greenhouse gas emissions and emissions intensity .......................................................................... 91

4.2 Replacing the first nuclear plant with a coal-fired power station ................................................................ 91



4.2.3 Impact on exports ................................................................................................................................ 92

4.3 Reduced electricity demand post 2030 ...................................................................................................... 92



4.4 Diversion of coal from Eskom to exports ................................................................................................... 94

4.4.1 Coal from the Waterberg to supply Central Basin power stations ....................................................... 94

4.4.2 Transport infrastructure requirements ................................................................................................. 95

APPENDIX A: DETAILS OF COAL-FIRED POWER STATION DECOMMISSIONING 2010 – 2040 ................... 97

APPENDIX B: DETAILS OF ELECTRICITY GENERATION BUILD PLAN 2010 – 2040 ................................... 103

TECHNICAL REPORT: SACRM SCENARIOS TO 2040 | iv

LIST OF TABLES

TABLE 1: COAL-FIRED POWER STATIONS NET MAXIMUM CAPACITY (MW), NOMINAL CAPACITY (MW)

AND AUXILIARY POWER REQUIREMENTS IN 2010 (%) .............................................................................. 4

TABLE 2: COAL-FIRED POWER STATIONS ASSUMED CV REQUIREMENTS .................................................. 5

TABLE 3: NET MAXIMUM CAPACITY (NON-COAL GENERATION TECHNOLOGIES) INSTALLED AT THE

BEGINNING OF 2010 (ESKOM AND NON-ESKOM GENERATION) .............................................................. 6

TABLE 4: CAPACITY FACTORS OF NON-COAL GENERATION TECHNOLOGIES ........................................... 6

TABLE 5: ASSUMED RETURN TO SERVICE SCHEDULE FOR GROOTVLEI AND KOMATI POWER

STATIONS ......................................................................................................................................................... 7

TABLE 6: ASSUMED DECOMMISSIONING SCHEDULE IN MW NET GENERATION CAPACITY TO 2030 ....... 7

TABLE 7: USE OF COAL IN OTHER APPLICATIONS IN 2010 ........................................................................... 16

TABLE 8: GROWTH PROJECTIONS FOR FERROALLOY AND IRON AND STEEL SECTORS ....................... 17

TABLE 9: CRITERIA FOR SUITABILITY OF CCS RETROFIT TO ESKOM’S EXISTING FLEET ....................... 18

TABLE 10: PHASE IN OF CCS UNDER LAGS BEHIND ...................................................................................... 19

TABLE 11: PHASE IN OF CCS UNDER LOW CARBON WORLD ....................................................................... 20

TABLE 12: POTENTIAL CARBON STORAGE SITES IN SOUTH AFRICA ......................................................... 20

TABLE 13: MODEL INPUTS FOR CCS USING MEA SOLVENT CAPTURING 90% OF CO2 EMISSIONS ........ 21

TABLE 14: 2010 CAPITAL AND O&M COSTS ..................................................................................................... 33

TABLE 15: FUEL COSTS IN 2010 RANDS ........................................................................................................... 35

TABLE 16: COSTS OF BUILD PLANS (R BILLION) ............................................................................................ 40

TABLE 17: MINE ESTABLISHMENT COSTS (R/TONNE CAPACITY ROM) ....................................................... 42

TABLE 18: COST OF PRODUCTION EXCLUDING TRANSPORT AND PORTS (R/TONNE) ............................ 42

TABLE 19: COST OF TRANSPORT OF EXPORT PRODUCT FROM MINE TO PORT (R/TONNE) ................... 42

TABLE 20: EXPORT PRICES OF DIFFERENT GRADES OF COAL ................................................................... 44

TABLE 21: FULL-TIME EMPLOYEE YEARS (FTE) FOR CONSTRUCTION OF KUSILE ................................... 48

TABLE 22: EMPLOYMENT INTENSITIES FOR MANUFACTURING AND CONSTRUCTION OF RENEWABLES

......................................................................................................................................................................... 49

TABLE 23: EMPLOYMENT INTENSITIES FOR O&M OF RENEWABLES .......................................................... 50

TABLE 24: WATER PURCHASED FOR MINING AND BENEFICIATION (Ml/MT ROM) ..................................... 58

TABLE 25: WATER DEMAND IN ELECTRICITY GENERATION ......................................................................... 58

TECHNICAL REPORT: SACRM SCENARIOS TO 2040 | v

TABLE 26: CO2 EMISSIONS FACTORS IN ELECTRICITY GENERATION ......................................................... 74

TABLE 27: WASTE GENERATION FACTORS IN ELECTRICITY GENERATION .............................................. 81

TABLE 28: NON-GHG EMISSION FACTORS APPLIED IN ELECTRICITY GENERATION ................................ 86

TABLE 29: MORE OF THE SAME COAL-FIRED POWER STATION RTS AND “LATE” DECOMMISSIONING

2010 - 2040 ...................................................................................................................................................... 97

TABLE 30: LAGS BEHIND COAL-FIRED POWER STATION RTS AND “MID” DECOMMISSIONING 2010 -

2040 ................................................................................................................................................................. 98

TABLE 31: AT THE FOREFRONT COAL-FIRED POWER STATION RTS AND “MID” DECOMMISSIONING

2010 - 2040 ...................................................................................................................................................... 99

TABLE 32: LOW CARBON WORLD COAL-FIRED POWER STATION RTS AND “EARLY”

DECOMMISSIONING 2010 - 2040 ................................................................................................................ 101

TABLE 33: MORE OF THE SAME ELECTRICITY GENERATION DECOMMISSIONING AND BUILD PLAN

(ALIGNS WITH IRP2010 BASE CASE TO 2030) ......................................................................................... 103

TABLE 34: LAGS BEHIND ELECTRICITY GENERATION DECOMMISSIONING AND BUILD PLAN (ALIGNS

WITH IRP2010 BASE CASE TO 2030) ......................................................................................................... 104

TABLE 35: AT THE FOREFRONT ELECTRICITY GENERATION DECOMMISSIONING AND BUILD PLAN

(ALIGNS WITH IRP2010 POLICY ADJUSTED TO 2030) ............................................................................ 106

TABLE 36: LOW CARBON WORLD ELECTRICITY GENERATION DECOMMISSIONING AND BUILD PLAN

(ALIGNS WITH IRP2010 EMISSIONS 3 TO 2030) ....................................................................................... 107

TECHNICAL REPORT: SACRM SCENARIOS TO 2040 | vi

LIST OF FIGURES

FIGURE 1: THE FOUR SOUTH AFRICAN COAL ROADMAP SCENARIOS ......................................................... 2

FIGURE 2: MAIN DETERMINANTS OF THE SCENARIOS .................................................................................... 2

FIGURE 3: ASSUMED PEAK DEMAND (MW) AND ANNUAL ELECTRICTY FORECAST (GWH SO) TO 2030

FROM IRP AND EXTRAPOLATED TO 2040 ................................................................................................... 9

FIGURE 4: CONTRIBUTION OF DSM ASSUMED IN IRP SHOWING NO ADDITIONAL DSM FROM 2017 TO

2030 AND EXTRAPOLATED TO 2040 ........................................................................................................... 10

FIGURE 5: ELECTRICITY GENERATION BUILD PLAN (LAGS BEHIND) .......................................................... 13

FIGURE 6: ELECTRICITY GENERATION BUILD PLAN (LOW CARBON WORLD) ........................................... 13

FIGURE 7: ELECTRICITY GENERATION BUILD PLAN (MORE OF THE SAME) .............................................. 13

FIGURE 8: ELECTRICITY GENERATION BUILD PLAN (AT THE FOREFRONT) .............................................. 13

FIGURE 9: COAL DEMAND FOR ELECTRICITY GENERATION (LAGS BEHIND) ............................................ 15

FIGURE 10: COAL DEMAND FOR ELECTRICITY GENERATION (LOW CARBON WORLD) ........................... 15

FIGURE 11: COAL DEMAND FOR ELECTRICITY GENERATION (MORE OF THE SAME) .............................. 15

FIGURE 12: COAL DEMAND FOR ELECTRICITY GENERATION (AT THE FOREFRONT) .............................. 15

FIGURE 13: PROJECTED LOCAL DEMAND FOR METALLURGICAL COAL ................................................... 18

FIGURE 14: UTILITY COAL SUPPLY FROM EXISTING MINES AND PROJECTS IN CENTRAL BASIN ......... 25

FIGURE 15: UTILITY COAL SUPPLY FROM EXISTING MINES AND PROJECTS IN WATERBERG ............... 25

FIGURE 16: SOUTH AFRICAN COAL EXPORTS (FIVE YEAR ROLLING AVERAGE) ...................................... 27

FIGURE 17: EXPORTS FROM CENTRAL BASIN ................................................................................................ 28

FIGURE 18: EXPORTS FROM THE WATERBERG .............................................................................................. 28

FIGURE 19: DECLINE IN HIGH GRADE (> 24 MJ/KG) ROM RESOURCES IN WITBANK, HIGHVELD AND

ERMELO COALFIELDS .................................................................................................................................. 29

FIGURE 20: DECLINE IN MEDIUM GRADE (22 - 24 MJ/KG) ROM RESOURCES IN WITBANK, HIGHVELD

AND ERMELO COALFIELDS ......................................................................................................................... 30

FIGURE 21: DECLINE IN LOW GRADE (< 22 MJ/KG) ROM RESOURCES IN WITBANK, HIGHVELD AND

ERMELO COALFIELDS .................................................................................................................................. 30

FIGURE 22: DECLINE IN RUN-OF-MINE COAL RESOURCES IN CENTRAL BASIN ........................................ 31

FIGURE 23: DECLINE IN RUN-OF-MINE COAL RESOURCES IN THE WATERBERG ..................................... 31

FIGURE 24: IMPACT OF TECHNOLOGY LEARNING ON OVERNIGHT CAPITAL COSTS OF RENEWABLES

AND NUCLEAR (R/kW) .................................................................................................................................. 36

TECHNICAL REPORT: SACRM SCENARIOS TO 2040 | vii

FIGURE 25: ANNUAL INVESTMENT IN ELECTRICITY GENERATION CAPACITY (LAGS BEHIND) .............. 38

FIGURE 26: ANNUAL INVESTMENT IN ELECTRICITY GENERATION CAPACITY (LOW CARBON WORLD) 38

FIGURE 27: ANNUAL INVESTMENT IN ELECTRICITY GENERATION CAPACITY (MORE OF THE SAME) ... 38

FIGURE 28: ANNUAL INVESTMENT IN ELECTRICITY GENERATION CAPACITY (AT THE FOREFRONT) ... 38

FIGURE 29: INDICATIVE ELECTRICITY GENERATION COST (LAGS BEHIND) .............................................. 39

FIGURE 30: INDICATIVE ELECTRICITY GENERATION COST (LOW CARBON WORLD) ............................... 39

FIGURE 31: INDICATIVE ELECTRICITY GENERATION COST (MORE OF THE SAME) ................................... 39

FIGURE 32: INDICATIVE ELECTRICITY GENERATION COST (AT THE FOREFRONT) ................................... 39

FIGURE 33: FIVE YEAR ROLLING AVERAGE PRICE OF COAL SOLD TO ESKOM ........................................ 43

FIGURE 34: PRICE TRAJECTORIES FOR 27 MJ/KG EXPORT PRODUCT ....................................................... 44

FIGURE 35: LOCAL SALES REVENUE (ESKOM COAL ONLY) ......................................................................... 45

FIGURE 36: EXPORT SALES REVENUE (METALLURGICAL AND THERMAL COAL) .................................... 45

FIGURE 37: EMPLOYMENT IN MINING IN CENTRAL BASIN ............................................................................ 53

FIGURE 38: EMPLOYMENT IN MINING IN THE WATERBERG .......................................................................... 53

FIGURE 39: CONSTRUCTION JOBS FOR POWER STATIONS AND CTL ........................................................ 54

FIGURE 40: EMPLOYMENT ASSOCIATED WITH OPERATION OF POWER STATIONS AND CTL ................ 54

FIGURE 41: OPERATION PHASE EMPLOYMENT UNDER LOW CARBON WORLD IN POWER STATIONS

AND CTL ......................................................................................................................................................... 55

FIGURE 42: TOTAL EMPLOYMENT (LAGS BEHIND) ......................................................................................... 56

FIGURE 43: TOTAL EMPLOYMENT (LOW CARBON WORLD) .......................................................................... 56

FIGURE 44: TOTAL EMPLOYMENT (MORE OF THE SAME) ............................................................................. 56

FIGURE 45: TOTAL EMPLOYMENT (AT THE FOREFRONT) ............................................................................. 56

FIGURE 46: WATER DEMAND FOR MINING IN THE CENTRAL BASIN ............................................................ 61

FIGURE 47: WATER DEMAND FOR MINING IN THE WATERBERG ................................................................. 61

FIGURE 48: NATIONAL WATER DEMAND PER SCENARIO FOR POWER STATIONS ................................... 62

FIGURE 49: WATER INTENSITY OF ELECTRICITY GENERATION ................................................................... 62

FIGURE 50: WATER DEMAND PER COALFIELD FOR ELECTRICITY GENERATION (LAGS BEHIND) ......... 64

FIGURE 51: WATER DEMAND PER COALFIELD FOR ELECTRICITY GENERATION (LOW CARBON

WORLD) .......................................................................................................................................................... 64

FIGURE 52: WATER DEMAND PER COALFIELD FOR ELECTRICITY GENERATION (MORE OF THE SAME)

......................................................................................................................................................................... 64

TECHNICAL REPORT: SACRM SCENARIOS TO 2040 | viii

FIGURE 53: WATER DEMAND PER COALFIELD FOR ELECTRICITY GENERATION (AT THE FOREFRONT)

......................................................................................................................................................................... 64

FIGURE 54: WATER DEMAND IN CTL ................................................................................................................. 65

FIGURE 55: WATER DEMAND IN CENTRAL BASIN (LAGS BEHIND) .............................................................. 66

FIGURE 56: WATER DEMAND IN CENTRAL BASIN (LOW CARBON WORLD) ............................................... 66

FIGURE 57: WATER DEMAND IN CENTRAL BASIN (MORE OF THE SAME) ................................................... 66

FIGURE 58: WATER DEMAND IN CENTRAL BASIN (AT THE FOREFRONT) ................................................... 66

FIGURE 59: WATER DEMAND IN THE WATERBERG (LAGS BEHIND) ............................................................ 67

FIGURE 60: WATER DEMAND IN THE WATERBERG (LOW CARBON WORLD) ............................................. 67

FIGURE 61: WATER DEMAND IN THE WATERBERG (MORE OF THE SAME) ................................................ 67

FIGURE 62: WATER DEMAND IN THE WATERBERG (AT THE FOREFRONT) ................................................ 67

FIGURE 63: PLANNED RBCT PORT AND RAIL LINE EXPANSION WITH TOTAL COAL EXPORTS (5 YEAR

ROLLING AVERAGE) ..................................................................................................................................... 68

FIGURE 64: EXPORTS FROM WATERBERG - LAGS BEHIND (5 YEAR ROLLING AVERAGE) ...................... 69

FIGURE 65: EXPORTS FROM THE WATERBERG – LOW CARBON WORLD (5 YEAR ROLLING AVERAGE)

......................................................................................................................................................................... 69

FIGURE 66: EXPORTS FROM THE WATERBERG – MORE OF THE SAME (5 YEAR ROLLING AVERAGE) . 70

FIGURE 67: EXPORTS FROM THE WATERBERG – AT THE FOREFRONT (5 YEAR ROLLING AVERAGE) . 70

FIGURE 66: TOTAL ROM COAL MINED IN THE CENTRAL BASIN ................................................................... 75

FIGURE 67: TOTAL ROM COAL MINED IN THE WATERBERG ......................................................................... 75

FIGURE 68: CO2 EMISSIONS FROM ELECTRICITY GENERATION .................................................................. 76

FIGURE 69: CO2 EMISSIONS INTENSITY FROM ELECTRICITY GENERATION ............................................... 77

FIGURE 70: CUMULATIVE DISCARD GENERATION IN THE CENTRAL BASIN (DISCARD PRODUCED

LESS DISCARD BURNED IN FBC) ................................................................................................................ 83

FIGURE 71: CUMULATIVE DISCARD GENERATION IN THE WATERBERG .................................................... 83

FIGURE 72: CUMULATIVE ASH GENERATION .................................................................................................. 85

FIGURE 73: CUMULATIVE FGD WASTE GENERATION .................................................................................... 85

FIGURE 74: CUMULATIVE HIGH LEVEL NUCLEAR WASTE GENERATION .................................................... 85

FIGURE 75: CUMULATIVE LOW/INTERMEDIATE LEVEL WASTE GENERATION ........................................... 85

FIGURE 76: TOTAL SO2 EMISSIONS FROM POWER GENERATION ................................................................ 88

FIGURE 77: TOTAL NOX EMISSIONS FROM POWER GENERATION ............................................................... 88

TECHNICAL REPORT: SACRM SCENARIOS TO 2040 | ix

FIGURE 78: TOTAL PARTICULATES FROM POWER GENERATION ............................................................... 88

FIGURE 79: NON GHG POWER STATION EMISSIONS IN CENTRAL BASIN UNDER MORE OF THE SAME 89

FIGURE 80: NON GHG POWER STATION EMISSIONS IN WATERBERG UNDER MORE OF THE SAME ..... 89

FIGURE 81: CHANGE IN GENERATION INFRASTRUCTURE INVESTMENT BY REPLACING ONE NUCLEAR

STATION WITH GAS ...................................................................................................................................... 91

FIGURE 82: CHANGE IN GENERATION COST BY REPLACING ONE NUCLEAR STATION WITH GAS ........ 91

FIGURE 83: CHANGE IN GENERATION INFRASTRUCTURE INVESTMENT BY REPLACING ONE NUCLEAR

STATION WITH COAL .................................................................................................................................... 92

FIGURE 84: CHANGE IN GENERATION COST BY REPLACING ONE NUCLEAR STATION WITH COAL ..... 92

FIGURE 85 ELECTRICITY DEMAND IN AT THE FOREFRONT AND ADJUSTED TO INVESTIGATE

SENSITIVITY TO LOWER DEMAND POST 2030 .......................................................................................... 93

FIGURE 86: INVESTMENT IN GENERATION INFRASTRUCTURE UNDER THE IRP 2010 AND A REDUCED

DEMAND SCENARIO ..................................................................................................................................... 93

FIGURE 87: ELECTRICITY GENERATION COST UNDER THE IRP 2010 AND A REDUCED DEMAND

SCENARIO ...................................................................................................................................................... 93

FIGURE 88: NEW BUILD REQUIRED POST 2030 FOR AT THE FOREFRONT AND WITH LOWER

ELECTRICITY DEMAND POST 2030 ............................................................................................................. 94

FIGURE 89: EXPORTS AND CENTRAL BASIN SUPPLY FROM THE WATERBERG - LAGS BEHIND (5 YEAR

ROLLING AVERAGE) ..................................................................................................................................... 95

FIGURE 90: EXPORTS AND CENTRAL BASIN SUPPLY FROM THE WATERBERG – LOW CARBON WORLD

(5 YEAR ROLLING AVERAGE) ...................................................................................................................... 95

FIGURE 91: EXPORTS AND CENTRAL BASIN SUPPLY FROM THE WATERBERG – MORE OF THE SAME


FIGURE 92: EXPORTS AND CENTRAL BASIN SUPPLY FROM THE WATERBERG – AT THE FOREFRONT


TECHNICAL REPORT: SACRM SCENARIOS TO 2040 | x

NOMENCLATURE

bbl Barrel

BECSA BHP Billiton Energy Coal South Africa

CCGT Combined Cycle Gas Turbine

CCS Carbon Capture and Storage

CO2e Carbon dioxide equivalent

CSP Concentrated Solar Power

CTL Coal-to-Liquid

CV Calorific Value

DMR Department of Mineral Resources

DSM Demand Side Management

EIA Environmental Impact Assessment

EPRI Electric Power Research Institute

ETP Energy Technology Perspectives

FBC Fluidised Bed Combustion

FGD Flue-gas desulphurisation

FOB Free-on-board

FTE Full-time employee

GDP Gross Domestic Profit

GHG Greenhouse Gas

Gt Gigatonne

GWh Gigawatt Hour

IEA International Energy Agency

IGCC Integrated Gasification Combined Cycle

IRP Integrated Resource Plan

IRR Internal Rate of Return

kt kilotonne

kWh Kilowatt hour

l/kWh Litres per kilowatt hour

l/kWhSO Litres per kilowatt hour sent out

LHV Lower Heating Value

LTMS Long-Term Mitigation Scenarios

MIT Massachusetts Institute of Technology

MJ/kg Megajoule per kilogram

MJ/MWh Megajoule per Megawatt hour

Mm3

Million cubic meters

Mt Megatonne

Mtpa Megatonne per annum

MW Megawatt

MWhSO Megawatt hour sent out

NERSA National Energy Regulator of South Africa

O&M Operations and maintenance

OCGT Open Cycle Gas Turbine

PF Pulverised Fuel

PWR Pressurised Water Reactor (nuclear)

TECHNICAL REPORT: SACRM SCENARIOS TO 2040 | xi

RBCT Richards Bay Coal Terminal

RTS Return-to-Service

SACRM South African Coal Road Map

SC Supercritical coal

SNAPP Sustainable National Accessible Power Planning

TFR Transnet Freight Rail

UCG-CCGT Underground Coal Gasification – Combined-Cycle Gas Turbine

USC Ultra-Supercritical Coal

TECHNICAL REPORT: SACRM SCENARIOS TO 2040 | 1

1 INTRODUCTION

The South African Coal Roadmap aims to support coal industry, policymakers and other stakeholders in

navigating an uncertain future in which there are multiple objectives to be met and trade-offs to be made, in order

for the country as a whole to flourish. A scenario led approach has been used to develop the Roadmap – by

identifying four different futures and exploring the implications of following each of these futures, an

understanding can be gained of the contribution of the coal value chain to the South Africa economy and the

well-being of its people and natural environment under different scenarios. The technical analysis and

development of the scenarios and Roadmap was conducted by The Green House (www.tgh.co.za) with input

from a group of Experts, on behalf of the South African Coal Roadmap Steering Committee.

An in-depth quantitative and qualitative analysis of the implications of each of the scenarios thus underpins the

Roadmap development. This technical report, presents the details of the analysis, with a separate document

providing a high-level summary of the implications of the scenarios. This report is divided into two parts: the first

describes the scenarios that were identified and modelled, and the second describes the implications of following

a particular scenario as measured against a set of indicators chosen to capture the possible economic, social

and environmental differences between the scenarios. For each component of the analysis, details are provided

of the approach and assumptions used for the modelling, and results are presented and discussed for each of

the indicators.

2 DESCRIPTION OF THE SCENARIOS

The SACRM scenarios were established by considering a range of local and global drivers including the global

economy and global development needs, the South African economy, the global climate change response, South

Africa’s climate mitigation response, global coal markets, balancing exports and local demand in South Africa,

evolution of local infrastructure and the evolution of technologies including carbon capture and storage. Together,

these often inter-dependent drivers will shape the future and the evolution of the coal value chain in South Africa.

Of these, two primary drivers were considered in framing the scenarios, namely the global and local response to

climate change. These drivers were used to define a set of four distinct scenarios with very different implications

for the development of the coal value chain over the period from 2010 to 2040. Figure 1 shows the framing

drivers, while Figure 2 shows the resulting scenarios: More of the Same, Lags Behind, At the Forefront and

Low Carbon World.


FIGURE 1: THE FOUR SOUTH AFRICAN COAL ROADMAP SCENARIOS

MA

IN D

ET

ER

MIN

AN

TS

LAGS BEHIND

The world decarbonises, but coal remains

a significant energy source in South Africa

and other developing countries. Coal-

based power generation still dominates

local electricity supply, but with a move

towards clean coal technologies, such as

ultra-supercritical power stations, carbon

capture and storage and underground coal

gasification as they become available.

A new coal-to-liquids plant is built in 2027

to meet local liquid fuels demand.

LOW CARBON WORLD

The world decarbonises and moves

towards use of nuclear and renewables for

electricity supply. Funding is available for

South Africa to follow suit, and no new coal-

fired power stations are built beyond

Medupi and Kusile.

Carbon capture and storage is pursued and

no more coal-to-liquids plants are built in

South Africa.

MORE OF THE SAME

Coal use continues globally and locally.

Coal-based power generation using

existing supercritical technologies

dominates the electricity mix, and the life of

existing power stations is extended.

Two new coal-to-liquids plants are built

between 2027 and 2040 to meet local liquid

fuels demand.

AT THE FOREFRONT

Coal use continues globally, but South

Africa aims to diversify its energy mix to

include renewables and more nuclear

generation. New coal-fired power plants

after Medupi and Kusile use ultra-

supercritical technology, with smaller power

stations (including FBC stations) also being

built.

No new coal-to-liquids plants are built.

FIGURE 2: MAIN DETERMINANTS OF THE SCENARIOS

LOW CARBON

WORLD

AT THE FORE-

FRONT

MORE OF THE

SAME

LAGS BEHIND

High global

response to

climate change

Low local

response to

climate change

Low global

response to

climate change

High local

response to

climate change


The implications of these scenarios for local coal demand, coal supply and exports of coal are now described.

2.1 Coal use in South Africa

The scenario models are driven primarily from a local electricity demand perspective. In other words, coal is

mined to ensure local electricity security, but is supplied at a cost to the electricity generator that ensures

adequate return on investment in mine projects and new mines. At the same time, various industry sources were

used to identify coal projects that are likely to come on line between now and 2040. These projects are assumed

to provide the thermal coal required for electricity generation, the thermal coal for non-Eskom domestic demand,

the metallurgical coal for domestic demand, and coal for exports (both thermal and metallurgical).

2.1.1 Electricity generation

The electricity generation mix is at the heart of the scenarios, given the strong reliance on coal for generation in

South Africa currently and the impact on coal use of any increase in electricity from nuclear and renewables in

the grid. The generation mix determines not only coal demand, but, also has implications for energy security,

electricity costs, requirements for infrastructure investment, water needs and pollution impacts, transmissions

grid requirements, land requirements, employment, and greenhouse gas and other emissions to air.

The basis for the electricity build plans under the four SACRM scenarios are a selection of the build plans

developed during the 2010 Integrated Resource Plan (IRP2010) process, which have been extended to 2040.

The published IRP scenarios give MW installed capacity per year by technology type up to 2030.

The SACRM scenarios were matched to what were considered to be the most likely IRP2010 build plans to 2030

to evolve under the different futures as follows:

• More of the Same and Lags Behind: IRP2010 Base Case scenario, under which new power stations

are mostly coal-fired;

• At the Forefront: IRP2010 Policy Adjusted scenario, which provides a diversified mix of new power

stations including coal, renewables and nuclear;

• Low Carbon World: IRP2010 Emissions 3 scenario, under which no new coal-fired power stations are

built after Medupi and Kusile.

While the IRP scenarios were followed closely to 2030, some minor adjustments were made to account for data

becoming available since publication of the IRP2010 draft reports. These updates pertain specifically to Eskom

generation capacity and the realised Return-to-Service (RTS) schedule. This had the effect of reducing the

electricity generation (GWh sent out) in 2010 and 2011 as predicted by the IRP2010, to rather match that

reported in the Eskom Divisional reports for these years. The assumptions made to extend the IRP2010

scenarios to define the build plan for the period 2031-2040 are described in detail below.

Existing coal-fired power station capacity in 2010

The following parameters relating to existing Eskom coal-fired power stations are used in the model: Net

maximum capacity (MW), nominal capacity (MW) and auxiliary power requirements (%). The values applied are

shown in Table 1.


TABLE 1: COAL-FIRED POWER STATIONS NET MAXIMUM CAPACITY (MW), NOMINAL CAPACITY (MW)

AND AUXILIARY POWER REQUIREMENTS IN 2010 (%)

Power station Net maximum capacity (MW)

Nominal capacity (MW)

Auxiliary power requirement (%)

Arnot 2,232 2,352 5.1%

Camden 1,450 1,530 5.2%

Duvha 3,450 3,600 4.2%

Grootvlei 380 (1,090) 400 (1,150) 5.2%

Hendrina 1,865 1,965 5.1%

Kendal 3,840 4,116 6.7%

Komati 170 (878) 182 (940) 6.6%

Kriel 2,850 3,000 5.0%

Lethabo 3,558 3,708 4.0%

Majuba 3,843 4,110 6.5%

Matimba 3,690 3,990 7.5%

Matla 3,450 3,600 4.2%

Tutuka 3,510 3,654 3.9%

Medupi 0 (4,332) 0 (4,618) 6.2%1

Kusile 0 (4,338) 0 (4,680) 7.3%

Non-Eskom 435 459 5.2%

Sasol 1 (Sasolburg) 130 137 5.2%

Sasol 2&3 (Secunda) 520 549 5.2%

Notes:

1. For those power stations under refurbishment or construction in 2010, the design capacity is given in brackets. 2. The generation of electricity by Sasol for its own facilities in included in total electricity demand in South Africa in the IRP2010. To retain comparability

with the IRP2010 it is thus included in Table 1, but in the scenario models power generation for CTL is considered separately as Sasol’s generation does not strictly contribute to electricity available for use in South Africa (see the section on CTL for further details on modelling of CTL).

Source: Net maximum capacity and nominal capacity are taken in most cases from the Eskom Divisional Report for the year ended 31 March 2012, with auxiliary power calculated from the difference between the two. For Komati, nominal capacity was taken from Eskom Divisional Report for the year ended

31 March 2012, and includes 3 x 300 MW units in reserve storage. Auxiliary power for this station was adjusted to get the same net max capacity as in IRP Table 12 (2010 Rev 2). Medupi and Kusile net maximum capacity is taken from Integrated Resource Plan (2010 Rev 2), with auxiliary power requirements taken from the IRP technical report (EPRI 2010)2 and nominal capacities calculated from these. Net maximum capacity for non-Eskom power stations and Sasol power stations taken from the IRP2010 Rev 2 Draft report with auxiliary power assumed to be the same as an older Eskom power station.

All power stations are assumed to be operating at their net maximum capacity at the beginning of 2010, with the

exception of the RTS stations Grootvlei and Komati, which have an assumed installed net maximum capacity of

380 and 170 MW respectively at the start of 2010. Medupi and Kusile were both not online in 2010 and therefore

have an installed net maximum capacity of 0 MW. Total coal-fired net maximum capacity installed at the

beginning of 2010 is therefore calculated to be 35,373 MW.

Power station efficiencies and capacity factors for individual Eskom power stations were set in such a way so as

to achieve the average Eskom reported generation load factor value of 65% and coal consumption in 20113 of

125 Mt. Non-Eskom and Sasol power stations were assigned an efficiency and capacity factor similar to an older

Eskom power station. Capacity factors for new coal-fired power stations (both FBC and PF) were assumed to be

85%, with underground coal gasification also assigned a capacity factor of 85%.

1 The auxiliary power of Medupi increases to 7.3% when FGD is added in 2021. 2 EPRI (2010) Power Generation Technology Data for Integrated Resource Plan of South Africa. Palo Alto, CA. 3 Based on figures published in the 2012 annual report for the year ending 31st March 2012


Data constraints were such that power station coal demand was split according to whether the power station

required a high, medium, low or very low calorific value (CV), as shown in Table 2, where high CV = 22-24 MJ/kg,

medium = 20-22 MJ/kg, low =18-20 MJ/kg, and very low = 16-18 MJ/kg (on an air-dried basis). It is noted that

other quality parameters such as ash, sulphur, volatiles and abrasiveness are as important as CV, which together

make up the coal specification for the power station. However, given data constraints and the complex nature of

coal, it was not possible to include these characteristics in these relatively high-level scenario models of the coal

value chain. It is further noted that the CV ranges used in the scenario models are broad, and some power

stations are designed to operate optimally for a CV that might fall at the top of the range, and that there would be

efficiency penalties if a coal with a CV at the bottom end of the range is burned. Similarly, a power station’s

optimal CV may fall at the bottom of the range indicated below, but could burn coal in the lower CV band with

some efficiency losses. These effects are not captured in the models.

TABLE 2: COAL-FIRED POWER STATIONS ASSUMED CV REQUIREMENTS

Power station CV Range

Arnot 22 – 24 MJ/kg

Camden

Tutuka

Non-Eskom

Kriel 20 – 22 MJ/kg

Duvha

Grootvlei

Hendrina

Komati

Majuba

Matla

Kendal 18 – 20 MJ/kg

Matimba

Medupi

Kusile

Sasol 1 (Sasolburg)

Sasol 2&3 (Secunda)

Lethabo 16 – 18 MJ/kg

Source: Eskom personal communication

Existing non-coal generation capacity in 2010

Table 3 shows the net maximum capacity of non-coal generation technologies assumed installed at the

beginning of 2010.


TABLE 3: NET MAXIMUM CAPACITY (NON-COAL GENERATION TECHNOLOGIES) INSTALLED AT THE

BEGINNING OF 2010 (ESKOM AND NON-ESKOM GENERATION)

Technology Net maximum capacity assumed installed at beginning of 2010

(MW)

OCGT 2,490

CCGT 0

Co-generation 407

Nuclear 1,800

Wind 0

CSP 0

Solar PV 0

Import hydro 1,500

Landfill, small hydro 600

Pumped storage 1,580

Coal imports 0

TOTAL 8,377

Source: IRP2010 Rev 2 Draft Report

The capacity factors of these technologies are assumed as shown in Table 4, with the same capacity factors

assumed to also apply to new build, unless specified otherwise.

TABLE 4: CAPACITY FACTORS OF NON-COAL GENERATION TECHNOLOGIES

Technology Capacity Factor (%)

Source/Comment

OCGT 10% pg. 3-7. EPRI (2010)4

CCGT 50% (peak)

90% (baseload)

Peak from pg. 3-7.

Baseload as supplied by Expert Group.

Co-generation 85% Assumed to be IGCC. Typical capacity factor of IGCC, Table 1-7 pg. xv. EPRI (2010).

Nuclear 84% (existing)

92% (new)

Existing nuclear on basis of Koeberg current performance.

New nuclear from IRP2010 Rev 2 Draft report.

Wind 21% Based on BP Statistical Review 2011. Revised down on advice of Expert Group from the 29% used in IRP2010 Rev 2 Draft report (pg. 38).

CSP 44% Assumed 9 hours of storage; pg. 38 IRP2010 Rev 2 Draft Report.

Solar PV 17% Revised down on advice of expert group from 26.8% in IRP2010 Rev 2 Draft report (pg. 38).

Import hydro 100% Imported power is not modelled in relation to a particular power station.

Landfill, small hydro

36% (existing) 46% (new)

The existing figure is for small hydro only, as calculated from Eskom Divisional Report

for year ended 31 March 2012. For a mix of new landfill and small hydro, the weighted average of landfill gas capacity factor from IRP2010 Rev 2 Draft report and existing

small hydro (calculated from Eskom Divisional Report 2012) on a 25:100 ratio was used.

Pumped storage 24% (existing) 20% (new)

Existing based on understanding of capacity factors of existing plant.

New from pg. 3-7. EPRI (2010)

Coal imports 100% Imported power is not modelled in relation to a particular power station.

4 EPRI (2010) Power Generation Technology Data for Integrated Resource Plan of South Africa. Palo Alto, CA.


Return to Service 2010 - 2013

The difference between assumed total net maximum capacity installed in 2010 (43,750 MW) in the model which

was based on an analysis of reported values as described above, and that of the IRP2010 (43,897 MW)

amounted to 147 MW, with the IRP installed net maximum capacity being the greater of the two. The difference is

as a result of discrepancies between those values for Arnot, Camden, Grootvlei, Hendrina and Komati applied in

the IRP2010 and given in the Eskom Divisional Report for the year ended 31 March 2012.

For the RTS programme, it was decided that it was more appropriate in the scenario modelling to use capacities

as reported in the Eskom Divisional reports rather than the values in the IRP2010, as the Eskom values reflect

the actual progress in the RTS programme between 2010 and 2012. The RTS schedules were thus adjusted

from the IRP2010 schedule as shown in Table 5.

TABLE 5: ASSUMED RETURN TO SERVICE SCHEDULE FOR GROOTVLEI AND KOMATI POWER

STATIONS

Year Grootvlei (MW) Komati (MW)

2010 380

2011 190 114

2012 140 296

2013 298

Source: Eskom Integrated Reports 2012, 2011, 2010, 2009 and own calculations

Decommissioning of power stations 2010 - 2030

The IRP2010 shows a total decommissioning of 10,902 MW between 2015 and 2030 across all IRP scenarios.

This was all assumed to be the decommissioning of coal-fired power stations. A decommissioning schedule

provided by Eskom was used to indicate the order of decommissioning. From this the schedule shown in Table 6

was developed to align with the IRP assumption for total amount of capacity decommissioned, with the order of

decommissioning suggested to be as follows: Non-Eskom, Camden, Komati, Grootvlei, Matla and Duvha. Note

that decommissioning is not disaggregated explicitly into individual power stations in the IRP2010.

TABLE 6: ASSUMED DECOMMISSIONING SCHEDULE IN MW NET GENERATION CAPACITY TO 2030

Year Power Station Total

2015 Non-Eskom -180

2016 Non-Eskom -90

2017

2018

2019

2020

2021 Non-Eskom -75

2022 Camden, Komati, Non-Eskom -1,960 (1450, 420, 90)

2023 Grootvlei, Komati, Matla -2,448 (1090, 458, 900)

2024 Matla -1,030

2025 Matla -1,520


Year Power Station Total

2026

2027

2028 Duvha -2,322

2029 Duvha -1,128

2030

TOTAL 10,753

The difference between total MW decommissioned here (10,753 MW) and under the IRP2010 scenarios to 2030

(10,902 MW) amounts to 149 MW which is approximately the same as the difference found between installed

capacity5 as predicted by the IRP2010 and that reported by Eskom (with the latter used in the scenario

modelling). Thus, the assumption that all decommissioning in the IRP2010 is coal coming off line seems to be

substantiated.

Decommissioning of power stations 2030 - 2040

Extending the life of existing power stations is an important variable to consider in the modelling as it not only

affects coal demand but also the emissions profile, with newer coal-fired power stations being more efficient

(using less coal) and having a lower emissions intensity, as well as the investment required to replace existing

power stations that are closed earlier. The decommissioning schedule was not altered prior to 2030 so as to

retain the build plan of the IRP2010 to 2030, other than through the explicit assigning of the stations to be

decommissioned, based on the information obtained about Eskom’s decommissioning schedule. However, the

schedule was altered post 2030 to explore the impact of this variable on the model outputs as follows:

• More of the Same: Late decommissioning of coal-fired power stations after 2030. In this

scenario neither coal usage nor emissions are constrained and thus it makes sense to keep

power stations on line for as long as feasible before investing in new infrastructure.

• Lags Behind: Mid-range decommissioning of coal-fired power stations after 2030. In this

scenario, while coal use is still dominant there is some external pressure to reduce emissions

and so existing coal-fired power stations are replaced with more efficient coal-fired power

stations earlier than in More of the Same.

• At the Forefront: Mid-range decommissioning of coal-fired power stations after 2030. This is in

line with this scenario’s fairly ambitious low carbon build plan.

• Low Carbon World: Early decommissioning of coal-fired power stations. In this scenario,

reduction of GHG emissions is one of the primary drivers of the build plan and so coal-fired

power stations are decommissioned early.

Mid-range decommissioning aligns with Eskom’s current decommissioning plan with power stations having a life

of 50 years, with the exception of Arnot, Kriel and Hendrina power stations having a 60-year lifespan. For early

decommissioning, the power stations were assumed to be decommissioned 4 years earlier and similarly, for late

decommissioning the power stations were assumed to be decommissioned 4 years later. Post 2030 power

stations were decommissioned over a number of years, with a unit coming off line per year. This is considered

5 The 2 MW discrepancy can be traced to the IRP where reported non-Eskom capacity is rounded down from 3,262 MW to 3,260 MW between the draft

IRP 2010 and the final IRP 2010, with this work using the disaggregated non-Eskom installed capacity as reported in the Draft IRP.


more realistic than the decommissioning schedule assumed under the IRP where whole power stations were

assumed to be decommissioned in a single year or over two years.

Power stations partly or fully decommissioned in the 2030 – 2040 period include (in order of decommissioning):

Hendrina, Arnot, Tutuka, Lethabo, Kriel, Matimba and Kendal. Therefore, power stations still operational in 2040

are Majuba, Medupi and Kusile.

Disaggregated decommissioning schedules can be found in Appendix A.

Assumptions underpinning demand projections for electricity between 2030 and 2040

New power station build plans between 2030 and 2040 were based on the installed capacity that is required to

meet projected increases in demand, as well as that required to make up for power stations which have been

decommissioned.

The annual electricity sent out forecast between 2010 and 2030 from the IRP, which is the same across all the

IRP scenarios, was extrapolated linearly to 2040 (see Figure 3). In the IRP, GDP grows almost linearly at 4.5%,

while energy intensity of the economy (the amount of electricity consumed for every one rand of GDP in real

terms) declines linearly. Together the GDP growth and the reduced energy intensity give rise to a net linear

increase in electricity demand, although at a rate that is lower than the GDP growth. In the IRP, all DSM

opportunities are assumed to have been exploited by 2017, and no further opportunities are identified thereafter

(Figure 4).

FIGURE 3: ASSUMED PEAK DEMAND (MW) AND ANNUAL ELECTRICTY FORECAST (GWH SO) TO 2030

FROM IRP AND EXTRAPOLATED TO 2040

10,000

20,000

30,000

40,000

50,000

60,000

70,000

80,000

90,000

200,000

250,000

300,000

350,000

400,000

450,000

500,000

550,000

600,000

2010 2015 2020 2025 2030 2035 2040

Pe

ak D

em

an

d [

MW

]

Fo

recast

Dem

an

d [

GW

hS

O]

Forecast Demand Peak Demand


FIGURE 4: CONTRIBUTION OF DSM ASSUMED IN IRP SHOWING NO ADDITIONAL DSM FROM 2017 TO

2030 AND EXTRAPOLATED TO 2040

Projecting electricity demand so far into the future using linear GDP growth coupled with a linear reduction of

energy intensity of the economy (as has been done for the IRP and is replicated here) is a significant

simplification, and projecting electricity demand growth so far into the future carries a significant level of

uncertainty. This approach does not take into account slowing of GDP growth in response to global economic

trends (as has been seen recently) or as the country becomes more developed. It also does not take into account

reductions in electricity intensity due to step changes in technology, or potential restructuring of the economy

which could come about if the country moves from a resource-led economy towards a more service-based

economy as a result, for example, of resource, climate change or energy cost drivers. Any of these factors could

lead to a slowing in electricity demand growth and hence a requirement for less rapid building of new power

stations, and a consequent slowing of the growth in electricity prices (assuming recovery of cost for new build is

reflected in the electricity price). The use of the same demand across scenarios (as was done in the IRP and is

replicated here) also ignores the fact that other sectors of the economy will likely track the decisions made in the

electricity supply sector – for example, a shift away from coal-fired power could be coupled with a reduced

electricity intensity of the economy.

Technology selection after 2030

The assumptions that determine the generation mix post 2030 to meet the growth in demand and that which

results from decommissioning old power stations are as follows:

• In More of the Same, new supercritical pulverised fuel (PF) coal-fired power stations, new

fluidised bed combustion (FBC) plants with sorbent injection and underground coal gasification

combined cycle gas turbines (UCG-CCGT) are built. It is assumed that 10% of net increased

demand is met with UCG-CCGT and the build plan has been extended to 2040 so that the ratio

of installed PF to installed FBC remains the same as in the 2030 build. FBC units are assumed

to be smaller than PF units (2 x 250 MW units for power stations of 500 MW net max). As noted

earlier in the discussion on the decommissioning schedules, the life of existing power stations in

2030 is extended 4 years beyond the default assumption of 50 or 60 years. Furthermore, open

cycle gas turbines (OCGT) are built such that the proportion of OCGT capacity in the mix is kept

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

2010 2015 2020 2025 2030 2035 2040

MW

Demand Side Management


constant and 3,557MW (net max) of new combined cycle gas turbine (CCGT) capacity is phased

in over four years from 2031.

• In Lags Behind, new ultra-supercritical PF coal-fired power stations, FBC and UCG-CCGT are

built. It is assumed that 10% of net increased demand is met with UCG-CCGT and the build plan

has been extended to 2040 so that the ratio of installed PF to installed FBC remains the same as

in the 2030 build. FBC units are assumed to make up smaller power stations (2 x 250 MW net

max units for power stations of 500 MW net max). As noted earlier in the discussion on the

decommissioning schedules, a mid decommissioning date is selected for coal-fired power

stations operational in 2030. Furthermore, open cycle gas turbines (OCGT) are built such that

the proportion of OCGT capacity in the mix is kept constant and 3,557MW (net max) of new

CCGT capacity is phased in over four years from 2031.

• In At the Forefront, the contribution to the generation mix by OCGT, co-generation, wind, solar

photovoltaic (PV), landfill gas, small scale hydro-electrical generation, pumped storage (for peak

power), imported hydropower and imported coal-based power to meet increased demand and

that required by decommissioning of old coal-fired power stations is kept constant between 2030

and 2040, on a per MWh basis. In other words, if wind power provided x% of the annual

electricity sent out in 2030, it still provides x% of the annual electricity sent out in 2040. Installed

capacity for each technology in each year is then determined by the electricity sent out and the

capacity factors associated with the technology. The limits to imported coal-based power

(Botswana: 1,200 MW, Mozambique: 1,000 MW) and imported hydropower (Mozambique: 2,135

MW, Zambia: 1,230 MW) as suggested in the IRP2010 Rev 2 draft report are taken into account

in the build plan. The remaining demand is met by installing 3,557MW (net max) of new CCGT

capacity phased in over four years from 2031, and a combination of nuclear, smaller FBC power

stations (500 MW net max in 250 MW units) with sorbent injection and large ultra-supercritical PF

stations (4,500 MW net max in 750 MW units). The split between nuclear, FBC and PF is done

such that the ratio of installed PF to installed FBC remains the same as in the 2030 build. As

noted earlier in the discussion on the decommissioning schedules, a mid decommissioning date

is selected for operational coal-fired power stations in 2030.

• In Low Carbon World, no new coal-fired power stations are built after 2030 and no imported

coal-based power utilised. New renewable, co-generation, OCGT, CCGT, pumped storage and

nuclear capacity is built and imported hydropower utilised. Under this scenario, as coal-fired

power stations are decommissioned, new nuclear power stations are built to keep the overall

contribution of baseload power (i.e. nuclear and coal) to MWh sent out constant. The remaining

capacity is made up by installing 6,402 MW (net max) of new CCGT capacity, phased in over

four years from 2031, building new OCGT, co-generation, pumped storage and renewables

capacity, as well as utilising imported hydropower so as to keep electricity generation by these

individual technologies in proportion to their relative MWh contribution in 2030. As noted in the

discussion on the decommissioning schedules, an early decommissioning date is selected for

operational coal-fired power stations in 2030.

It is assumed in all scenarios that all new power stations are built with flue gas desulphurisation (FGD), in

addition to FGD being retrofitted to Medupi in 2021. Kusile is also assumed to be built with FGD.

It is recognised that these four scenarios do not account for the potential for shale gas from the Karoo to play a

role in electricity generation in South Africa. This, and the more extensive use of imported gas (from fields such

as those in Mozambique) could have a substantial impact on the build plans under the various scenarios.


Electricity generation build plans to 2040

The resulting build plans to 2040 under the different scenarios are shown in Figure 5 to Figure 8. Details of

capacities installed in each year are shown in Table 33 to Table 36 in Appendix B to this document.


FIGURE 5: ELECTRICITY GENERATION BUILD PLAN (LAGS BEHIND) FIGURE 6: ELECTRICITY GENERATION BUILD PLAN (LOW CARBON WORLD)

FIGURE 7: ELECTRICITY GENERATION BUILD PLAN (MORE OF THE SAME) FIGURE 8: ELECTRICITY GENERATION BUILD PLAN (AT THE FOREFRONT)

0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

2010 2015 2020 2025 2030 2035 2040

Net

Max C

ap

acit

y [

MW

]

Coal (ex imports) Nuclear Renewables (incl imports) Gas Other

0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

2010 2015 2020 2025 2030 2035 2040

Net

Max C

ap

acit

y [

MW

]


0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

2010 2015 2020 2025 2030 2035 2040

Net

Max C

ap

acit

y [

MW

]

Coal (excl. imports) Nuclear Renewables (incl. imports) Gas Other

0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

2010 2015 2020 2025 2030 2035 2040 N

et

Max C

ap

acit

y [

MW

]



Coal demand for electricity generation

National demand for coal for electricity generation according to the four CV bands considered in the model is

shown in Figure 9 to Figure 12. The rapid growth in demand in More of the Same and Lags Behind in the 18 to

20 MJ/kg category is for new power stations, with all new-build power stations (including Kusile and Medupi)

assumed to burn coal in this range. Of this low quality band of coal, by 2040 approximately 80% of the demand is

for power stations in the Waterberg, 3-5% is for UCG in the Free State and the remainder is for the Central Basin

power stations Kusile and Kendal (with Kendal already partially or fully decommissioned in some scenarios). At

the Forefront shows moderate growth in demand for coal in the 18 to 20 MJ/kg CV band as fewer, more

efficient, power stations are opened in this scenario, while Low Carbon World remains constant to 2028,

increases after that year as Medupi and Kusile are retrofitted with CCS (causing a decrease in efficiency, see

Section 2.2 for the discussion on CCS retrofitting), and then declines towards the end of the period as Kendal

and Matimba power stations are decommissioned.

The other bands of coal used in electricity generation are utilised in the Central Basin. Under all scenarios, there

is an on-going demand for over 50 Mtpa of the 20 to 22 MJ/kg coal band to around 2020 (depending on the

scenario), with the demand for this band of coal starting to drop off after this year as older power stations are

decommissioned – earlier under Low Carbon World and At the Forefront than More of the Same and Lags

Behind. Ensuring security of this grade of coal is critical to meeting on-going electricity demand up to 2040 in all

scenarios, although the volumes required are fairly low by 2040 (around 10 Mtpa in all scenarios other than More

of the Same where it is 19 Mtpa).

Of particular concern is the supply of the 22 to 24 MJ/kg coal band given the decline in high quality resources in

the Central Basin (see Section 3.1) and the competition for this grade of coal with exports and other domestic

users (see Section 2.3). Coal in the 22 – 24 MJ/kg CV band is required up to 2038 in all scenarios, other than

More of the Same, where it is required beyond 2040. Ensuring security of this grade of utility coal is critical to

meeting on-going electricity demand up to 2040, without having to prematurely retire coal assets and build

additional capacity to make up the shortfall.

The small rise in the under 18 MJ/kg coal seen in More of the Same, Lags Behind and At the Forefront seen

from around 2023 onwards is required by the small fluidised bed combustion plants brought on line after this

year. The decline towards the end of the period in some scenarios is caused by the decommissioning of Lethabo

power station.


FIGURE 9: COAL DEMAND FOR ELECTRICITY GENERATION (LAGS BEHIND) FIGURE 10: COAL DEMAND FOR ELECTRICITY GENERATION (LOW CARBON

WORLD)

FIGURE 11: COAL DEMAND FOR ELECTRICITY GENERATION (MORE OF THE SAME)

FIGURE 12: COAL DEMAND FOR ELECTRICITY GENERATION (AT THE FOREFRONT)

0

20

40

60

80

100

120

140

160

180

200

2010 2015 2020 2025 2030 2035 2040

Co

al d

em

an

d [

Mtp

a]

Below 18 MJ/kg 18 - 20 MJ/kg 20 - 22 MJ/kg 22 - 24 MJ/kg

0

20

40

60

80

100

120

140

160

180

200

2010 2015 2020 2025 2030 2035 2040

Co

al d

em

an

d [

Mtp

a]


0

20

40

60

80

100

120

140

160

180

200

2010 2015 2020 2025 2030 2035 2040

Co

al d

em

an

d [

Mtp

a]


0

20

40

60

80

100

120

140

160

180

200

2010 2015 2020 2025 2030 2035 2040 C

oal d

em

an

d [

Mtp

a]



2.1.2 Coal-to-liquids

Sasol is assumed to continue to operate coal-to-liquids operations in Secunda across all scenarios, under similar

levels of coal consumption, although it is recognised that there may be an expansion of production with gas

rather than coal as feedstock (this possibility is not assessed in the scenario models). Coal-fired electricity

demand for utilities at the Sasol facilities for the 2010-2030 period was taken from the IRP (IRP2010 Rev 2). For

the 2030-2040 extension it was assumed that Sasol facilities would continue to operate as before, although it

may be that existing power stations could be replaced with gas-fired electricity generation facilities in the future

(emissions from gas usage, both GTL and for utilities are not included in the scenario models).

The assumptions regarding the opening up of further CTL plants are as follows:

• More of the Same: With no constraints on coal use under this scenario it is assumed that a new

CTL plant with a capacity of 80,000 bbl per day starts to be brought on line from 2027, with a

further plant of the same capacity brought on line from 2037.

• Lags Behind: With continued domestic coal use, but a gradual global shift to a low carbon

future, it is assumed that one new CTL plant with a capacity of 80,000 bbl per day starts to be

brought on line from 2027.

• At the Forefront: With a national ambition for a diversified and low carbon energy mix, no new

CTL plants are built under this scenario.

• Low Carbon World: No new CTL plants are built in this scenario.

New CTL facilities are brought on line over a period of 4 years, with 25% of capacity coming on line in the first

year, and a further 25% at a time in the second, third and fourth years. All future plants are assumed to be

located in the Waterberg. The costs of constructing a new CTL plant are not considered in the study.

Sasol Secunda consumed 39 Mt of coal in 2010 for a 160,000 bbl per day refinery, including both process coal,

which is converted into product, and utility coal (used for steam and electricity generation)6. Coal consumption in

an 80,000 bbl per day refinery is assumed to be half of that used at Secunda.

2.1.3 Other uses of coal

Although the majority of coal use in South Africa is in electricity generation and synfuels production, coal is also

used domestically in other thermal and metallurgical applications. The Department of Mineral Resources (DMR)

collates information on use of coal in such applications, with 2010 consumption shown in Table 7.

TABLE 7: USE OF COAL IN OTHER APPLICATIONS IN 2010

Type Mtpa

Bituminous steam coal 19.8

Bituminous coking coal (metallurgical) 2.4

Anthracite 1.0

Source: Department of Mineral Resources (2012), Personal Communication

Demand for thermal coal has remained relatively flat over the past number of years, and expert opinion is that

6 Sasol 2010 Analyst Book


this will continue to be the case moving into the future. This assumption is used for all four scenarios, with a

demand of 19 Mtpa of coal from the Central Basin from 2010 through to 2040 assumed in the SACRM scenario

models. The demand is assumed to be for coal with a CV greater than 22 MJ/kg.

Growth in metallurgical coal consumption is assumed to differ per scenario under the SACRM. Iron and steel and

ferroalloys are the primary consumers of metallurgical coal in South Africa. Various growth projections for these

two commodities to 2030 were explored in a study conducted for the National Planning Commission and are

used to guide growth in metallurgical coal demand for the SACRM scenarios. These growth projections are

shown in Table 8. To map the growth projections in these two commodities onto the metallurgical coal demand

data shown in Table 7, it was assumed that a nominal 85% of the metallurgical coal is used in iron and steel

production and the remainder in ferroalloys.

TABLE 8: GROWTH PROJECTIONS FOR FERROALLOY AND IRON AND STEEL SECTORS

Commodity Baseline projection Low output projection High output projection

Ferroalloys Ferrochrome and ferromanganese

production grow at 2% per annum

throughout the analysis period to

2030.

Ferrochrome and

ferromanganese production

decline at 2% per annum

throughout the analysis period

to 2030.

Ferrochrome and ferromanganese

production grows by 4% per

annum to 2020 and then at 6%

between 2021 and 2030.

Iron and

steel

Blast Furnace/Basic Oxygen Furnace

(BF/BOF) production of iron and steel

will remain unchanged.

BF/BOF production shrinks by

23%, spread linearly over the

analysis period to 2030.

BF/BOF output as per the baseline

projection.

Source: Cohen, B., Lewis, Y. and K. Mason-Jones (2012), Projections of greenhouse gas emissions trajectories from the South African mining and minerals processing sectors to 2030. Report prepared for the National Planning Commission.

Mapping the above growth projections onto the SACRM scenarios is done as follows:

• More of the Same and Lags Behind: High output projection is followed, as South Africa continues to

pursue an energy- and coal-intensive trajectory;

• At the Forefront: Baseline projection, as South Africa’s economy is diversified following diversification of the

electricity mix;

• Low Carbon World: Low output projection, as South Africa moves away from a coal- and energy-intensive

economy.

For the extension of these growth projections to the period between 2031 and 2040 in the SACRM study, no

further growth or decline is assumed between 2031 and 2040 for the Baseline and Low Output projection, while

under the High Output projection, growth in ferrochrome and ferromanganese production is assumed slow to 3%

per annum between 2031 and 2040, while iron and steel production remains unchanged.


FIGURE 13: PROJECTED LOCAL DEMAND FOR METALLURGICAL COAL

2.2 Carbon capture and storage (CCS)

Coal is an inherently greenhouse gas intensive resource, so the coal value chain is highly exposed to activities to

reduce greenhouse gas emissions both within South Africa and globally. There is a strong signal from

government that it will introduce a carbon tax by 2015, and introduce caps on emissions to various sectors

through the passing of the National Climate Change Response White Paper. Carbon Capture and Storage (CCS)

– as an enabling technology to allow the continued use of coal in an increasingly carbon-constrained world – is

therefore of high interest to the coal industry.

The process streams from coal-to-liquids operations show the most promise for CCS due to their volume, high

concentration of CO2 and ease of capture. For coal-fired power stations, CCS could have application on new

power stations. In terms of potential for retrofitting to Eskom’s existing fleet, guidance was obtained from a report

produced by the International Energy Agency in 20127, which was based on the conclusions from four other

studies. A summary of this analysis and comments to contextualise this for Eskom’s existing power station fleet is

presented in Table 9.

TABLE 9: CRITERIA FOR SUITABILITY OF CCS RETROFIT TO ESKOM’S EXISTING FLEET

Parameter Criterion Comment

Size of unit Minimum of 100 to 300 MW, depending on the study

All Eskom’s units are above this threshold

Efficiency High efficiency (not sub-critical)

Minimum LHV efficiency 29% to 35% depending on the study

Only Medupi and Kusile will be supercritical, with all power

stations before this sub-critical and thus below this threshold.

Age Not for older plants – at least less than 35 years old

Majuba, Medupi and Kusile power stations will be less than 35 years old in 2030 (the assumed commercialisation date)

Distance to storage Less than 40 km All Eskom’s units are at least 250 km away from potential storage sites.

7 Finkenrath, M., Smith, J. and Dennis Volk, D. (2012). CCS retrofit: analysis of the globally installed coal-fired power plant fleet, International Energy Agency.

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

2010 2015 2020 2025 2030 2035 2040

Lo

cal m

eta

llu

rgic

al co

al d

em

an

d

pro

jecti

on

s [

Mt]

More of the same and lags behind At the forefront Low carbon world


Although there are no absolute criteria for suitability of retrofit, retrofitting plants that do not satisfy the minimum

requirements generally will not be economically justifiable. Within this study, therefore, it is assumed that since

Eskom’s existing plants do not meet the majority of the criteria listed in the Table above, none of the existing

power stations would be retrofitted with CCS. However, where substantial funding would be available for

greenhouse gas mitigation, CCS could be retrofitted to Medupi and Kusile in 2030 (currently under construction

without CCS).

The following assumptions are thus made about the adoption of CCS for the different scenarios:

• More of the Same: No commercial CCS capacity is installed for either power stations or CTL

plants. Local CCS effort stops early, probably in 2020, after the test injection project being

planned by the South African Centre for Carbon Capture and Storage. This is due to no

requirement for CCS in a world with limited focus on reducing greenhouse gas emissions.

• Lags Behind: CCS continues to be pursued due to the international focus on low carbon

technologies, with a local trial injection by 2017 and an internationally funded demo in about

2025. Commercial-scale CCS is available from around 2030 and is retrofitted to the Sasol

Secunda process stream, the process stream from the new CTL plant built in 2027, as well as

new large PF coal-fired power stations built from 2034 onwards.

• At the Forefront: Local CCS effort stops early, probably 2020, after the test injection as there

will be no global funding to take the effort beyond this point, and South Africa is unable to

develop CCS without this support. Thus no commercial CCS is installed in South Africa.

• Low Carbon World: CCS is adopted actively both globally and locally. CCS is retrofitted to

Sasol Secunda, as well as to Medupi and Kusile, from 2029 onwards.

Phasing in of CCS

The phase in of CCS with CTL under Lags Behind is shown in Table 10. Furthermore, under this scenario, all

new PF power stations from 2034 onwards are built with CCS installed.

TABLE 10: PHASE IN OF CCS UNDER LAGS BEHIND

Plant Year Mt CO2 captured per year

Sasol Secunda (process stream) 2030 1.0



Sasol Secunda (process stream) 2033 onwards

Full capacity (90% of 24 Mt CO2 per annum process stream)

8

New CTL plant process stream 2033 5.0

New CTL plant process stream 2034 onwards

Full capacity (90% of 12 Mt CO2 per annum process stream)

The phase in of CCS at Sasol Secunda under Low Carbon World is shown in Table 11. In addition, under this

scenario, Medupi and Kusile are retrofitted with CCS, resulting in around an 85% drop in the CO2 emissions from

8 Extent of capture from: DEA (Department of Environmental Affairs) (2008). Long Term Mitigation Scenarios. Available online:

http://www.environment.gov.za/hotissues/2008/ltms/ltms.html, accessed July 2012.


these two power stations. CCS is phased in over three years, starting in 2029. No further coal-fired power

stations are built under this scenario.

TABLE 11: PHASE IN OF CCS UNDER LOW CARBON WORLD

Plant Year Mt CO2 captured per year




Sasol Secunda (process stream) 2033 onwards Full capacity (90% of 24 Mtpa process stream)

Carbon Storage in South Africa

The focus of CCS research in South Africa is primarily on geological storage, with the South African Centre for

CCS leading the exploration of the potential for geological carbon storage in South Africa. The Centre has

produced an Atlas on Geological Storage in which potential injection sites are identified. An estimated storage

capacity of 150 Gt of CO2 in South Africa is identified, with almost all of this capacity being located off-shore in

saline formations associated with oil- and gas-bearing formations, with less than 2% of the estimated capacity

occurring on-shore (Table 12).

TABLE 12: POTENTIAL CARBON STORAGE SITES IN SOUTH AFRICA

Off-shore sites: Capacity [Gt]

On-shore sites: Capacity [Gt]

Outeniqua Basin 48 Zululand Basin 0.46

Orange Basin 56 Algoa Basin 0.4

Durban/Zululand Basin 42 SA coalfields 1.2

Alternatives to geological CCS include chemical reaction of the CO2 to form chemically stable compounds

through a process known as mineral sequestration, and algal sequestration, in which the CO2 is pumped into a

pond and is used to grow algae subsequently recovered for their energy value. At present these are not being

explored extensively in South Africa, although there may be merit in further exploration of their potential as an

alternative.

Assumptions applied in modelling CCS

CCS with coal-fired power stations

CCS incurs a thermal efficiency penalty for power stations, and also increases the auxiliary power requirements.

The loss in thermal efficiency arises because the capture system requires a large amount of heat for amine

solvent regeneration, whilst the increased auxiliary power is required for flue gas pre-treatment, blowers, pumps

and compressors. Together this results in an appreciable drop in net plant efficiency. As such, power plants with

CCS use more coal and water than those without, and bigger power stations are required to achieve the same

electricity outputs.

A review of the available literature data on CCS suggested that it would be useful to have two assumptions sets:

One for the retrofit of Medupi and Kusile, and one for new build. For these, both leading CCS technologies were


reviewed: post–combustion capture technology using amine solvent (MEA) and pre-combustion capture

technology (the so-called oxy-fuel process). The oxy-fuel process has a number of significant advantages,

notably no increase in water use, a slightly lower efficiency penalty and a smaller plant footprint. However, the

MEA process is the more proven and established technology, and most of the cost and performance data

available in the literature is for the MEA process. The assumption here is that the MEA technology is the more

likely in the time-frame of this study and also has the more accurate data set, and thus is the technology used to

model both CCS retrofits and new build in the SACRM scenario models. Assumptions used in the modelling of

CCS are shown in Table 13.

TABLE 13: MODEL INPUTS FOR CCS USING MEA SOLVENT CAPTURING 90% OF CO2 EMISSIONS

Input variable Supercritical Ultra-supercritical

Energy penalty

Retrofit: Percentage point drop in net efficiency 12.5

New Build: Percentage point drop in net efficiency 9.6 9.5

Retrofit: Auxiliary power requirements (% of nominal capacity) 27%

New build: Auxiliary power requirements (% of nominal capacity) 20% 20%

Emissions

Retrofit: CO2 net capture efficiency 85%

New build: CO2 net capture efficiency 86% 87%

Retrofit and new build: % decrease in SO2 emissions 99.9% 99.9%

Retrofit: NOx emissions (kg/MWh net) 2.9

New build: NOx emissions (kg/MWh net) 2.7 2.7

Retrofit: particulate emissions (kg/MWh net) 0.20

New build: particulate emissions (kg/MWh net) 0.18 0.15

Water use

Specific water consumption of dry-cooled plant (l/kWh net) 0.67 0.59

Sources: Shuster, E and Hoffmann, J, (2009) Water Requirements for Fossil-Based Electricity Plants with and without Carbon Capture, National Energy Technology Laboratory, 2009 GWPC Annual Forum, Salt Lake City, UT; MIT (2007) The Future of Coal - Options for a carbon-constrained world;

DOE/NETL (2008) Estimating Freshwater Needs to Meet Future Thermoelectric Generation Requirements; Haibo Zhai, Edward S. Rubin, and Peter L. Versteeg (2011), Water Use at Pulverized Coal Power Plants with Postcombustion Carbon Capture and Storage, Environ. Sci. Technol. 2011, 45, 2479–2485 dx.doi.org/10.1021/es1034443; NETL (2007) Cost and Performance Baseline for Fossil Energy Plants, Vol. 1, DOE/NETL-2007/1281, May 2007. B_PC_051507 (Pulverized Bituminous Coal Plants With and Without Carbon Capture & Sequestration); IEA GHG / Foster Wheeler Italiana (2010) Water

Usage And Loss Analysis Of Bituminous Coal Fired Power Plants With Capture, Report: 2010/ 05, March 2011; Finkenrath, M (2011) Cost and Performance of Carbon Dioxide Capture from Power Generation, Working Paper, International energy agency; NETL (2008) Water Requirements for Existing and Emerging Thermoelectric Plant Technologies, DOE/NETL-402/080108, August 2008 (Revised April 2009); DOE/NETL (2006) Carbon Dioxide Capture from Existing Coal-Fired Power Plant, DOE/NETL-401/110907, revised November 2007.

CCS Retrofit

Retrofitting post-combustion CCS, e.g. MEA, has a significant effect on the boiler and steam cycle, and because

of this unbalancing effect, higher net efficiency penalties are seen in the literature for retrofits than for new build.9

For this reason, pre-combustion CCS or oxy-fuel might be a better option for retrofitting since the steam cycle is

less affected and the major impact is the increased electricity requirement for the auxiliaries (primarily the

ASU10

). However, for the reasons stated above, only the MEA process is modelled in the SACRM.

9 MIT (2007) The Future of Coal - Options for a carbon-constrained world; and DOE/NETL (2006) Carbon Dioxide Capture from Existing Coal-Fired Power Plant, DOE/NETL-401/110907, revised November 2007 10 Air separation unit


Retrofit of CCS is assumed to differ from new build CCS only with respect to the energy penalties associated with

the CCS plant. In all other respects the effect of CCS on power station performance (e.g. on water consumption

and stack emissions) is modelled the same for retrofits and new builds.

CCS New Build

The net thermal efficiency penalty for supercritical plants with post-combustion capture of 90% of their CO2 was

found to vary between 9.2 and 11.9 percentage points (see data source list for Table 13), the OECD average of

9.6 % given in the IEA review11

is used in the scenario models. A similar range is found for ultra-supercritical

plants (9.2 to 11.3 percentage points). In the ultra-supercritical case, the average across all data sources was

identical to the OECD average used in the IEA study, so this value is used in the models (9.5 percentage points).

The majority of the data points were for wet-cooled stations, but the efficiency drop for dry-cooled stations was

found to be only marginally higher in a study by Foster Wheeler using South African climate conditions12

(9.4 for

dry-cooled vs. 9.2 for wet-cooled).

Emissions

In all cases, the MEA plants are estimated to remove 90% of CO2 emissions. However, because of the decrease

in power station efficiency, on a net electricity sent out basis, the decrease in CO2 emissions relative to a plant

without CCS is less than 90%. In other words, 90% of CO2 stack emissions are captured, but because more coal

now being burnt to provide the same power output as a plant without CCS - and thus more CO2 emissions are

now being produced – the net effect is a drop in CO2 emissions of between 84 and 87% (depending on the drop

in net thermal efficiency applied). The CO2 net capture efficiency assumed in the models follows from the net

efficiency penalty and auxiliary power increases assumed for each case. These calculated values are given in

Table 13.

Emissions of SO2 are also substantially decreased (to negligible levels) with the MEA CCS process.

NOx and particulate emissions are assumed to be the same as for plants without CCS (on a MWh generated

basis). However, because of the drop in net efficiency, they increase slightly on a net sent out basis. The energy

penalties given in Table 13 are used to adjust the baseline NOx and particulate emission factors for power

stations with CCS. Improvements in boiler design for future build power stations mean that NOx emissions might

not necessarily increase with addition of CCS (as they currently do slightly in the scenario models).

Particulate emissions increase to a higher percentage than NOx emissions with addition of CCS (increases of

35% and 16%, respectively, for a supercritical plant with CCS based on MEA). This is because unlike NOx

emissions, particulate emissions are linked to coal burn rate in the scenario models. The ash content and

particulate collection efficiency are assumed to be constant, and particulate emissions therefore increase in

proportion to the increase in coal feed (on a per MWh sent out basis). The CCS technology is assumed not to

remove a higher proportion of particulates than the equipment in place without CCS, although this assumption

has not been rigorously checked (only one of the literature studies reviewed included data on particulate

emissions13

).

Cost penalties

11 Finkenrath, M (2011) Cost and Performance of Carbon Dioxide Capture from Power Generation, Working Paper, International energy agency 12 IEA GHG / Foster Wheeler Italiana (2010) Water Usage And Loss Analysis Of Bituminous Coal Fired Power Plants With Capture, Report: 2010/ 05, March 2011 13 Shuster, E and Hoffmann, J, (2009) Water Requirements for Fossil-Based Electricity Plants with and without Carbon Capture, National Energy

Technology Laboratory, 2009 GWPC Annual Forum, Salt Lake City, UT


Cost penalties associated with CCS fitted to new power stations and retrofitted to Medupi and Kusile are

presented in Section 3.3.1, along with implications for the operation of these power stations. Fuel cost penalties

are incorporated into the coal consumption figures required to meet electricity demand. It is recognised that the

O&M costs are likely to be substantially underestimated for South African conditions, given the long piping

distances from power stations and CTL plants to the coast where the majority of storage locations are found.

CCS applied to coal-to-liquids facilities

In a CTL facility greenhouse gas emissions arise from the production of the synfuels product (process

emissions), as well as from burning coal (or gas) for steam and electricity generation (utility emissions). The

process emissions, which represent about 50% of overall emissions from CT, are produced in a concentrated

form and are directly suitable for capture via CCS, whilst the utility emissions are in a relatively dilute form, and

their capture is energy and water intensive. It is thus assumed that, both for retrofit to Secunda, and also in any

new CTL facility, 50% of CO2 emissions are suitable for capture via CCS14

.

Unlike the capture of CO2 emissions from power stations, it is assumed that the high concentration process CO2

stream can be captured with negligible impact on process efficiency and water demand. It is thus assumed that

coal consumption does not increase with addition of CCS to the CTL process. Furthermore, it is assumed that the

relatively slight additional energy requirements of the plant (for compression and pumping of the CO2 stream) can

be met either by utilising waste heat in the process, as suggested by Mantripragada and Rubin15

, or would

increase gas consumed to generate electricity (and is thus not captured in the models where the focus is on coal

consumption). These assumptions only hold for the relatively low CO2 capture rates identified above (50%). To

reduce CO2 emissions further would require capture of the low concentration streams, in which case the same

substantial increases in energy costs and water demand would be evident as for CCS applied to coal-fired power

generation.

Costs for building and operating CTL plants are not included in the economic models. Although costs of CO2

capture for CTL are not as extensive as for power generation, they are still appreciable, with the Long Term

Mitigation Scenarios study (LTMS)16

suggesting the cost of capturing CO2 from CTL is to the order of

R 476/tonne CO2 (in 2007 Rands).

2.3 Coal supply

The scenario models are primarily demand driven, in other words it is assumed that sufficient coal is mined to

meet the local demand for utility coal and other domestic coal, as reported above. The model also looks at

projections of coal supply from existing mines and mine projects, which were collated from annual reports, the

Wood Mackenzie coal supply service data and industry experts. For thermal coal projects, the coal from a

particular project was allocated either to Eskom, other domestic supply or to export. If there was no information

on the project, the coal was allocated according to wherever there was shortfall in demand. A number of projects

were identified as not for Eskom, in which case the coal was designated for export, unless it was needed to meet

other domestic demand. Projects supplying coal for domestic use are assumed to come on line when required by

the demand, i.e. whenever there is a shortfall in that grade of coal. This demand-driven timing was applied also

14 Assumption based on DEA (Department of Environmental Affairs) (2008). Long Term Mitigation Scenarios. Available online: http://www.environment.gov.za/hotissues/2008/ltms/ltms.html, accessed July 2012. 15 Mantripragada and Rubin (2011). Techno-economic evaluation of coal-to-liquids (CTL) plants with carbon capture and sequestration, Energy Policy 16 DEA (Department of Environmental Affairs) (2008). Long Term Mitigation Scenarios. http://www.environment.gov.za/hotissues/2008/ltms/ltms.html,

accessed July 2012.


to dual-producing mines, where the export fraction is assumed dependent on the domestic fraction. The timing of

export-only projects is according to industry expert opinion and the Wood Mackenzie database.

Sufficient projects were identified in the Central Basin coalfields to secure coal supply to the existing power

stations and to other non-Eskom domestic users throughout the 2010 to 2040 period, in all scenarios other than

in More of the Same. However, in all scenarios other than in Low Carbon World, supply of high-grade utility

coal is very constrained from the mid-2020s. Whether in reality there is sufficient supply of this grade of coal in

the Central Basin from the mid 2020s onwards will depend on new mines opening on time, whether projects

opening in the Central Basin are willing to supply Eskom, or whether the coal is exported as low-grade exports

(23.5 MJ/kg). If the latter occurs, in all scenarios, other than Low Carbon World (where early decommissioning

of power stations occurs), alternative sources of this grade of coal will need to be sourced, most likely from the

Waterberg. This is demonstrated as part of the sensitivity analysis presented in Section 4. Furthermore, where

the decommissioning of power stations is delayed and 22-24 MJ/kg coal is required beyond 2040, as in More of

the Same, an alternative source of coal will be required from the mid 2030s.

Most of the growth in coal supply occurs in the Waterberg, where it is assumed all future coal-fired power stations

are located. Each future power station is associated with a new dual-producing mine, that produces a low-grade

utility product, together with a high grade export product. In addition to supplying new power stations in the

Waterberg, as per the previous paragraph it is likely that coal will be required from the Waterberg to supply high-

grade coal to existing Mpumalanga power stations, where delays in projects, or projects switching to export

rather domestic supply will leave a shortfall in supply from the Central Basin. The scenario models show that it

would thus be prudent to plan for such a Waterberg coal supply from the early 2020s in all scenarios other than

in Low Carbon World.

Figure 14 shows utility coal supply from existing mines and projects located in Central Basin coalfields, while

Figure 15 shows utility coal supply from existing mines and projects in the Waterberg coalfields. Coal supply from

mines in the Central Basin is similar under all four scenarios, with some variations due to generating load and the

timing of decommissioning of power stations. Coal supply from the Central Basin increases slightly to 2020, and

then declines to 2040.

Coal supply from the Waterberg differs between scenarios, with More of the Same and Lags Behind requiring

the highest volumes of coal from new mines in this coalfield. No new mines are developed under Low Carbon

World, where coal supply increases slightly with the retrofit of CCS to Medupi in 2030, and then declines towards

2040 as Matimba power station is decommissioned. At the Forefront shows limited growth in the Waterberg,

under which fewer mines are brought on line than in More of the Same and Lags Behind.


FIGURE 14: UTILITY COAL SUPPLY FROM EXISTING MINES AND PROJECTS IN CENTRAL BASIN

FIGURE 15: UTILITY COAL SUPPLY FROM EXISTING MINES AND PROJECTS IN WATERBERG

2.4 Coal exports

Global demand for coal differs between the scenarios as follows:

• More of the Same and At the Forefront: global demand for coal grows, with high demand for

Asia for coal including lower grade products for power stations.

• Lags Behind: Although there is some global lock-in to coal given existing infrastructure with long

service lives, growth for the higher grade export products (27 and 25 MJ/kg) slows and is

replaced by demand for low-grade coal for Asian markets (23.5 MJ/kg).

• Low Carbon World: Global demand for coal will decline as the world moves away from fossil

fuels, potentially retiring existing plants early. However, whilst demand for the higher grades of

export coal declines, demand in low-grade coal for Asian markets (23.5 MJ/kg) is assumed to

remain strong throughout the period.

0

20

40

60

80

100

120

140

160

2010 2015 2020 2025 2030 2035 2040

Mtp

a

More of the Same Lags Behind At the Forefront Low Carbon World

0

20

40

60

80

100

120

140

160

2010 2015 2020 2025 2030 2035 2040

Mtp

a



Whilst acknowledging the likely global changes, it is recognised that South Africa makes up a relatively small

proportion of global trade in coal, and hence under the SACRM it is assumed that a market can be found for all

export coal that is produced. All Waterberg mines are assumed to be dual-product mines along the lines of the

existing Grootegeluk mine (the only Waterberg mine operating in 2010), and are assumed to continue to produce

high-grade thermal coal export and a low-grade utility coal product under all scenarios.

Exports from existing mines and projects were projected from a collation of data from annual reports, the Wood

Mackenzie coal supply service data and expert input. Where specific information could not be found about a

project, especially with regards to potential product grades and markets, these were estimated by assuming a

number of indicative yields and product splits according to the grade of the coal resource (spit into high, medium,

low or very low, see Section 3.1). If the deposit was sufficiently high-grade, a dual-producing mine was selected

as the default option, with the mine configuration selected based on the option giving the lowest cost of the

particular grade of utility coal required while still ensuring mine profitability (i.e. that which provides an IRR of

10%). Thus, the export yield is dependent on the utility demand (grade and volume), unless specific information

is known for the project, in which case this was used instead. In some scenarios, project start dates are pushed

out until the utility coal is required (as happens in At the Forefront), whilst if there is no utility demand, as

happens in Low Carbon World, the mine is assumed to produce for export only. Projects utilising lower-grade

deposits and which were labelled specifically as not for utility supply, were assumed to produce low-grade

exports and were brought on line to keep exports reasonably smooth, i.e. to keep export volumes as constant as

possible so as to keep transport infrastructure utilised. The models assume there are no infrastructure

constraints, and that there is sufficient infrastructure to transport export coal to market, as discussed in Section

3.7.1.

Metallurgical exports are estimated from total metallurgical coal production less the predicted metallurgical

domestic demand (see Section 2.1.3). Metallurgical coal production from existing coal mines and projects was

taken from the Wood Mackenzie coal supply service data.

Total South African coal exports (thermal and metallurgical) under the four different scenarios are shown in

Figure 16. Exports are reported as a five-year rolling average. Under all four scenarios, exports grow to around

90 Mtpa and remain at that level until 2025. Exports under More of the Same and Lags Behind increase further

(to 95 Mtpa) as new power stations are opened in the Waterberg, and then from 2031 start to decline to around

current levels (80 Mtpa) as mines in the Central Basin close. On the other hand, exports from At the Forefront

and Low Carbon World start to decline gradually from 2025 to below current export levels (60 Mtpa), as mines

close in the Central Basin, and are not replaced by as many mines opening in the Waterberg as under the other

scenarios. If exports are looked at on a 20 year rolling average rather than a 5 year rolling average, More of the

Same and Lags Behind maintain exports at around 90 Mtpa. between 2020 to 2040, whilst in At the Forefront

and Low Carbon World exports decline from 85 Mtpa to 75 Mtpa over this time.

The export trends are shown more clearly by looking at Figure 17 and Figure 18, which show the exports from

the Central Basin and Waterberg, respectively.

From Figure 17 it can be seen that exports from Central Basin rise as new dual-producing and export only mines

are brought on line, and then begin to decline as resources are mined out and power stations begin to close. If

looked at on a 20 year rolling average, the Central Basin sustains exports at around 75 Mtpa between 2015 and

2030 in all scenarios. Low Carbon World shows slightly higher exports at the end of the period than the other

scenarios. This is because early decommissioning of power stations under this scenario means that certain

projects supplying utility coal in the other scenarios (either as dual-producing mines or dedicated utility coal


supply) are no longer required under Low Carbon World. However, it is assumed that these projects would still

go ahead but as export only mines instead.

An important assumption affecting the predicted export volumes from the Central Basin is that non-Eskom

domestic demand stays at a relatively high and constant 19 Mtpa. This coal competes directly for low-grade

exports, thus exports could potentially be higher should domestic demand decline (this is not investigated under

any of the scenarios).

Turning to Figure 18, exports from the Waterberg increase from 2023 as the first power station post Medupi is

opened in More of the Same and Lags Behind. Thereafter dual product mines start to be opened and a steady

growth in exports is observed. At the Forefront shows modest growth in export production post 2023 as fewer

new power stations are built under this scenario than under More of the Same and Lags Behind. In Low

Carbon World export production from the Waterberg remains low as no further coal-fired power stations are built

after Medupi, and Matimba power station is decommissioned early towards the end of the period.

A very important assumption for the predicted growth in exports is thus the dependency on exports from the

Waterberg on securing a market for a utility coal stream, i.e. that all future Waterberg mines will be dual-

producing (along the lines of the only currently operating Waterberg mine). However, if exports from South Africa

are to be maintained at current levels (or to grow beyond these) further exploration and development is required

in the Waterberg so as to demonstrate the feasibility of export only mines. Export only mines would be required in

the Waterberg from 2030 in At the Forefront and Low Carbon World to keep infrastructure from becoming

stranded (i.e. to keep exports from declining below current levels). Furthermore, the levels of exports seen under

More of the Same and Lags Behind are dependent on the relatively conservative estimate of export fraction in

the dual-producing mines (only one scenario is applied for all new Waterberg mines: 8% high-grade exports and

30% low-grade utility coal). There is far higher export potential should future exploration of the Waterberg

suggest alternative product splits from its mines, such as the production of a low-grade export stream.

FIGURE 16: SOUTH AFRICAN COAL EXPORTS (FIVE YEAR ROLLING AVERAGE)

0

10

20

30

40

50

60

70

80

90

100

2010 2015 2020 2025 2030 2035 2040

Mtp

a

More of the same Lags behind At the forefront Low carbon world


FIGURE 17: EXPORTS FROM CENTRAL BASIN FIGURE 18: EXPORTS FROM THE WATERBERG

3 IMPLICATIONS OF THE SCENARIOS

3.1 Resources and reserves

Remaining run-of-mine coal resources17

in the Witbank, Highveld and Ermelo coalfields is estimated to be around

12,000 Mt (combined reserves and resources across all three coalfields). Various internal industry reports and

expert opinion was used to categorise these remaining resources into four possible grades:

• High (> 24 MJ/kg), making up 4% of remaining resources

• Medium (22 to 24 MJ/kg), making up 26% of resources

• Low (20 - 22 MJ/kg), making up 40% of resources

• Very low (<20 MJ/kg), making up 30% of resources

ROM coal from the existing mines and projects in these coalfields were assigned to one of the resource grades.

The analysis is very approximate because estimates of resources are very uncertain and can differ substantially

between studies, largely because of what can be considered a minable resource, for example, whether resources

are adjusted for environmental considerations (e.g. wetlands), whether pillars from old underground workings are

included and whether the possibility of reworking fines and discard dumps is included (these factors are excluded

from the resource estimate used in the scenario models).

Decline in resources according to these resource grades up to 2040 is shown in Figure 19, Figure 20 and Figure

21 (with the two bottom grades combined). In all scenarios, the two top grades of resources are close to being

depleted over the period. There are still large volumes of lower grade resources, even though these resources

are more than halved over the time period. Low grade resource estimates are more extensive if the Vereeniging-

Sasolburg coalfield is included in the estimate.

Total ROM resource depletion in Central Basin coalfields is shown in Figure 22. All four scenarios show very

similar profiles as they are fed by the same set of existing mines and projects.

17 Resources here should be interpreted as resources and reserves, i.e. the total coal remaining that can be viably extracted.

0

10

20

30

40

50

60

70

80

90

2010 2015 2020 2025 2030 2035 2040

Mtp

a

More of the same Lags behind

At the forefront Low carbon world

0

10

20

30

40

50

60

70

80

90

2010 2015 2020 2025 2030 2035 2040

Mtp

a




Although rather approximate, this top-down analysis of remaining resources seems to tie in fairly well with the

bottom-up analysis conducted by identifying all potential projects and minable deposits remaining in the Central

Basin. If the bottom-up analysis is comprehensive and all potential projects have managed to be captured, then

assuming that all these projects are brought on line over the period 2010-2040, as is done in all scenarios,

should leave very few (if any) deposits remaining to be exploited. This seems to be the case, with the top-down

analysis showing high and medium quality resources dwindling close to zero by 2040. Furthermore, whilst plenty

of low-grade resources remain in the Central Basin at 2040 (5,000+ Mt), these are not currently earmarked for

development because of quality, environmental and/or cost constraints.

Resources in the Waterberg coalfield are understood to be very extensive, and an indicative value of 45,000 Mt

run-of-mine resources in 2010 is used in the analysis. The total run-of-mine resource depletion in the Waterberg

coalfields is shown in Figure 23. The trend is as expected: for More of the Same and Lags Behind there is a

heavy power station build post 2025, so the resources begin to decline rapidly after that year. At the Forefront

and Low Carbon World are less coal intensive, so the decline in resources is less rapid. The resource base of

the Waterberg coalfield is thought to be so large that even with the considerable coal consumption of the high-

coal scenarios, the remaining resources are still very substantial.

FIGURE 19: DECLINE IN HIGH GRADE (> 24 MJ/KG) ROM RESOURCES IN WITBANK, HIGHVELD AND

ERMELO COALFIELDS

0

100

200

300

400

500

600

2010 2015 2020 2025 2030 2035 2040

Mt



FIGURE 20: DECLINE IN MEDIUM GRADE (22 - 24 MJ/KG) ROM RESOURCES IN WITBANK, HIGHVELD

AND ERMELO COALFIELDS

FIGURE 21: DECLINE IN LOW GRADE (< 22 MJ/KG) ROM RESOURCES IN WITBANK, HIGHVELD AND

ERMELO COALFIELDS

0

500

1,000

1,500

2,000

2,500

3,000

3,500

2010 2015 2020 2025 2030 2035 2040

Mt


0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

2010 2015 2020 2025 2030 2035 2040

Mt



FIGURE 22: DECLINE IN RUN-OF-MINE COAL

RESOURCES IN CENTRAL BASIN

FIGURE 23: DECLINE IN RUN-OF-MINE COAL

RESOURCES IN THE WATERBERG

3.2 Implications of the electricity generation build plans

The total new installed capacity needed in At the Forefront and Low Carbon World (98,706 MW net max and

114,359 MW net max, respectively) is higher than in More of the Same and Lags Behind (63,712 MW net max

and 74,724 MW net max, respectively) to meet the same electricity demand projection. The difference is due to

the lower capacity and availability factors for renewable energy technologies, which are used more extensively in

At the Forefront and Low Carbon World. A further contributing factor to increased new build requirement in

Low Carbon World is the earlier retirement of coal-fired power stations post 2030 than in the other scenarios,

which then need to be replaced during the analysis period. The overall build in More of the Same is higher than

that in Lags Behind. The higher build requirement is due to the lower efficiency of supercritical power stations

built in More of the Same than the ultra-supercritical power stations built in Lags Behind – although this

difference is offset somewhat by the fact that certain of the existing power stations operate to beyond the

analysis period in More of the Same but are decommissioned in Lags Behind, and hence need to be replaced

in Lags Behind.

Nuclear power stations require a lead time of nine to ten years, due primarily to the requirements for careful site

selection, public consultation, sourcing of funding and provisions for safety. Under Low Carbon World, the first

nuclear power station is to come on line in 2022, and At the Forefront in 2023, suggesting that it may already

almost be too late to achieve these build plans. Coal-fired power stations have somewhat shorter lead times, of

about eight years. Failing to begin planning with sufficient lead times, particularly for coal and nuclear power

stations, will result in the technologies not being able to be brought on line in time, and the gap needing to be

made up by alternative technologies which have shorter lead times, including gas and some renewable

technologies, such as wind.

The renewables and nuclear builds under At the Forefront and Low Carbon World are ambitious and would

require investment in local manufacturing capacity and skills development in the very short term. Low Carbon

World in particular assumes a substantial roll out of concentrated solar power (CSP) and wind power from 2017

onwards. Given that CSP technology is still in its infancy in South Africa (with 100 MW demonstration plants

being discussed at the time of writing in 2012), the achievement of this CSP build requires rapid technology

advancement. Failing to achieve this level of advancement will require the gap to be filled by alternative

technologies – once again likely to be gas and wind, which have shorter lead times.

0

5,000

10,000

15,000

2010 2015 2020 2025 2030 2035 2040

Rem

ain

ing

reso

urc

es a

nd

re

serv

es [

Mt]

More of the Same Lags Behind

At the Forefront Low Carbon World

0

10,000

20,000

30,000

40,000

50,000

2010 2015 2020 2025 2030 2035 2040

Rem

ain

ing

reso

urc

es a

nd

re

serv

es [

Mt]




Technological and economic challenges associated with CCS

A number of challenges associated with the roll-out of CCS technology that is installed under Lags Behind and

Low Carbon World need to be considered. Capital and operating costs are high; plant efficiency is decreased

and hence fuel usage increases substantially; water requirements are high and the technology is immature (at

least in terms of integrating capture, transport and storage on full-scale plants). Key issues that need to be

resolved for the successful application of CCS in South Africa include:

• Resolution of regulatory and liability matters associated with carbon capture, piping and storage.

• Accessing the substantial funding required to install CCS technologies, the scale of which is

shown in Section 3.3.1.

• Fully determining the implications of high piping infrastructure and pumping costs due to the

distant location of CCS sites relative to power stations and coal resources.

• Demonstrating the long-term stability of CO2 in the storage sites.

• Weighing up the investment in CCS against the other socio-economic imperatives that the

government is required to deal with.

• Overcoming the need for skills for all aspects of CCS (design, operation, maintenance),

particularly if the technology is pursued extensively globally and skills are in global demand.

• Advancing technologies to reduce water required for cooling of CCS.

3.3 Economic implications of the scenarios

Economic implications of the scenarios are considered in terms of:

• Electricity generation infrastructure investment and generation cost

• Revenues associated with local sales and export sales of coal

• Impact on competitiveness

3.3.1 Electricity generation infrastructure investment and electricity generation cost

Electricity build plans were analysed across the different scenarios to ascertain annual investment requirements;

and average electricity generation cost.

The annual investment requirement represents the capital that needs to be raised to fund new build, either from

local or global funders. It is calculated by determining the total investment required for each power station (in

2010 Rands), and spreading the investment over the years assumed to be required for construction of the power

station. The investment spread was obtained from the SNAPP power systems model developed by the Energy

Research Centre at the University of Cape Town18

.

Average electricity generation cost per technology is calculated as the sum of annualised cost of capital (i.e.

spread of capital cost over the life of plant) for each technology, fixed and variable O&M costs, fuel costs as well

as an environmental levy on electricity from non-renewable resources. The costs of carbon transport and storage

are not included in the calculation. The electricity generation cost presented is indicative and is used as a proxy

for electricity price, as a variety of other factors contribute to determining the ultimate price paid for electricity,

18 Energy Research Centre, UCT (2012), Sustainable National Accessible Power Planning. Available online:

http://www.erc.uct.ac.za/Research/Snapp/snapp.htm, accessed August 2012.


such as the costs of transmission and distribution and utility operational costs, which are not modelled in the

study. Imported hydro and imported coal are excluded from the calculation of electricity generation cost as they

are not built in South Africa, and the price paid by South Africa will not necessarily be linked to the cost of

generation.

Model input data to calculate generation infrastructure investment and electricity

generation cost

The data required to calculate the generation infrastructure investment and electricity generation cost includes

capital and O&M costs, the impact of technology learning on capital cost, the phasing in of capital spend in the

years leading up to the commissioning of the power station, fuel costs and the environmental levy.

Capital and O&M costs (in 2010 Rands) were extracted from the promulgated IRP2010, and other literature, as

shown in Table 14.

TABLE 14: 2010 CAPITAL AND O&M COSTS

Technology Overnight

capital cost [R/kW net

max]

Fixed O&M

[R/kW net max/a]

Variable

O&M [R/MWh

SO]

Reference

Existing large coal (subcritical)

N/A 199 8.18 SNAPP model for power generation developed by the Energy Research Centre at UCT

19

Existing small coal (subcritical)

N/A 275 8.18 SNAPP model for power generation developed by the Energy Research Centre at UCT19

Supercritical PF 15,470 348 36.30 IRP2010

Supercritical PF with FGD

17,785 455 44.40 IRP2010

Ultra-supercritical PF

17,931 348 36.30 IEA ETP20

suggests a ratio of capital costs of ultra-

supercritical to supercritical of 2,550: 2,200. This ratio is

applied to the supercritical PF capital costs with no CCS or FGD from IRP2010. For O&M costs, the MIT study on the future of coal

21 cites various studies on

O&M costs of SC and USC. The average values for SC and USC are almost identical, although there is variability between studies. On this basis, the fixed and

variable O&M costs were assumed to be the same for USC and SC.

Ultra-supercritical PF with FGD

20,614 455 44.40 The IEA ETP20

ratio of capex of USC to SC was

assumed to apply in scaling to SC with FGD to USC with FGD. For O&M costs, the MIT study

21 reports on

various studies on O&M costs of SC and USC power

stations. Average values for SC and USC are almost identical, although there is variability between studies. The same was assumed to hold for supercritical with FGD and ultra-supercritical with FGD.

Ultra-supercritical

PF with CCS and FGD

32,983 778 75.92 The IEA ETP 2012 ratio of capex of USC to SC was

assumed to apply in scaling to SC with FGD to USC

with FGD. Hamilton et al22

suggests that a multiplier of 1.6 times be applied to the capex cost of supercritical PF without CCS to account for the cost of CCS

19 Energy Research Centre, UCT (2012), Sustainable National Accessible Power Planning. Available online:

http://www.erc.uct.ac.za/Research/Snapp/snapp.htm, accessed August 2012. 20 International Energy Agency (2012), Energy Technology Perspectives 2012. 21 MIT (2007), The Future of Coal, Massachusetts Institute of Technology, ISBN 978-0-615-14092-6 22 Hamilton, M, Herzog, H, Parsons, J 2009, ‘Cost and U.S. public policy for new coal power plants with carbon capture and sequestration’, Energy

Procedia vol. 1, pp 4487-4494.


Technology Overnight

capital cost

[R/kW net max]

Fixed O&M

[R/kW net max/a]

Variable

O&M

[R/MWh SO]

Reference

equipment. The same factor was assumed for the cost of adding CCS to USC. For O&M, the MIT study

21

suggests that a USC with CCS has, on average, O&M costs of 1.71 times that without CCS. The multiplier of 1.71 was applied to fixed and variable O&M costs of SC with FGD (given that average O&M costs for SC and USC are almost identical as discussed above). Note that the cost given here excludes the costs of

pipelines, transportation and sequestration. However, the impact of the cost of transportation and sequestration on electricity generation price

has been estimated to be R 2.02 per tonne of CO2 captured and is explored as a sensitivity.

Fluidised bed with FGD

16,540 404 99.10 IRP2010

OCGT 3,955 70 0 IRP2010

CCGT 5,780 148 0 IRP2010

Nuclear 26,575 0 95.20 IRP2010

Landfill/small hydro 3,508 99 22.22 Own calculations based on landfill figures in IRP2010. Assumes small hydro comes with negligible costs.

Wind 14,445 266 0 IRP2010

CSP parabolic 50,910 635 0 IRP2010

Solar PV 20,805 208 0 IRP2010

Pumped storage 7,913 123 0 IRP2010

Cogeneration 21,248 436 0.13 NERSA consultation paper on COFIT23

, pg 19. Figures used for coal CHP. Exchange rate = R 8/$.

UCG-CCGT 9,973 148 659 In the absence of local cost data, this data is based on a US study

24. Capital costs are for an oxygen-fired

UCG plant coupled with a combined cycle gas turbine. The fixed O&M costs is for a CCGT plant, the O&M costs associated with the UCG are considered fuel costs as per the cited study.

Retrofitting CCS to Medupi and Kusile

17,920 778 75.92 Cost estimates for retrofitting presented in the literature are highly variable, depending on the power station technology, amount of space available etc. The figure here is from a single 2001 study cited in the MIT study

25. The cost was escalated to 2010 values using

the Chemical Engineering Plant Cost Index (1.4) and converted to Rands using an exchange rate of R 8/$. The cost of running a supercritical PF with FGD and CCS is assumed to be 1.71 times that without CCS, as per the MIT study. Note that the cost here excludes that of pipelines, transportation and sequestration as discussed below.

Retrofitting FGD to Medupi

1,385 As per Supercritical

PF with FGD

As per Supercritical PF with FGD

An Engineering News article in 201026

suggests that it would cost R 6 billion to add FGD to Medupi which has a net maximum capacity of 4,332 MW. This figure is used here.

23 NERSA consultation paper on COFIT, pg 19. Figures used for coal CHP. Exchange rate = R 8/$, available online at http://www.nersa.org.za/Admin/Document/Editor/file/Electricity/Consultation/Documents/NERSA%20Consultation%20Paper%20Cogeneration%20Regulatory%20Rules%20and%20Feed-In-%20Tariff.pdf, accessed November 2012. 24 IU (2011) Viability of Underground Coal Gasification with Carbon Capture and Storage in Indiana. Prepared by the School of Public and Environmental Affairs, Indiana University (Submitted May 4, 2011). 25 MIT (2007), The Future of Coal, Massachusetts Institute of Technology, ISBN 978-0-615-14092-6 26 http://www.engineeringnews.co.za/print-version/medupi-units-to-undergo-air-quality-retrofit-from-2018-onwards-2010-01-22


All operating costs are indicated in the IRP2010 to include the cost of sorbent (in the case of technologies fitted

with FGD), and are assumed to include the cost of water. It is not possible to isolate either the sorbent or water

costs from the IRP study, and hence it cannot be ascertained whether the price trajectories for these inputs are

appropriate. Of the costs listed in the table, the highest uncertainty exists around the costs of CCS, and those of

new ultra-supercritical power stations. All data from international references and have not been adjusted to

account for costs of construction in South Africa, which adds further uncertainty.

Fuel costs were included as follows:

• For coal supplied by existing mines to Eskom power stations, the cost of fuel in 2010 Rands was

based on information presented by Eskom. The price for OCGT, CCGT and nuclear was

extracted from the IRP2010, as shown in Table 15. Cogeneration costs were extracted from

alternative reference sources as shown. No escalation in fuel costs is assumed in the modelling

for calculating generation costs, as per the IRP assumptions.

• For new power stations as well as existing Eskom power stations supplied from new mines, the

cost of coal was determined by calculating the average price at which individual mines would

have to sell coal to Eskom in order to achieve an IRR of 10%, taking into account the cost of

capital to produce the different coal qualities required by Eskom (including mine establishment

costs and costs of single or multiple stage washing plants), production costs, and the revenues

achieved from exporting coal. It is recognised that the debate about an appropriate rate of return

for mining houses supplying coal to Eskom is on going.

TABLE 15: FUEL COSTS IN 2010 RANDS

Technology Fuel cost

Coal for existing power stations supplied by existing mines

R 205.00 per tonne

Discards for fluidised bed combustion Assumed to be zero

OCGT R 2,385 per MWhSO

CCGT R 462.40 per MWhSO

Nuclear R 68.18 per MWhSO27

Cogeneration R 163.4 per MWhSO28

The costs of operating power stations equipped with CCS as indicated in Table 14 exclude capital and operating

costs associated with pipelines, transportation and storage. Various studies internationally have explored these

cost implications. A study by McKinsey29

in 2008 suggested that for early commercial CCS projects, of a total

cost of sequestration of ! 35 – 50 per tonne CO2 sequestered, ! 4 – 6 per tonne is associated with transport

(with the ! 6 per tonne upper estimate referring to offshore transport), and a further ! 4 – 12 per tonne with

storage. Capture thus represents to the order of two-thirds of the costs. These figures account for both capital

and operating costs. Transport distances considered in that study were to the order of 200 – 300 km. In South

27 IRP 2010 suggests a fuel cost of R 6.25/GJ. Assuming a conversion efficiency of 33%, this figure is calculated as 6.25*3600/1000/0.33 28 NERSA consultation paper on COFIT, pg 19. Figures used for coal CHP. Exchange rate = R 8/$. http://www.nersa.org.za/Admin/Document/Editor/file/Electricity/Consultation/Documents/NERSA%20Consultation%20Paper%20Cogeneration%20Regulatory%20Rules%20and%20Feed-In-%20Tariff.pdf 29 McKinsey & Company (2008) Carbon Capture and Storage: Assessing the Economics, available online at http://www.mckinsey.com/clientservice/ccsi/pdf/ccs_assessing_the_economics.pdf, accessed November 2012.


Africa, where storage sites are located offshore and power stations in The Central Basin or the Waterberg, the

longer pumping distances will result in higher transport costs.

A final cost included is that of the environmental levy currently charged on electricity generated from non-

renewable resources. The levy started at 2c per kWh in July 2009. This was escalated to 2.5c per kWh in July

2011 and to 3.5c per kWh in July 2012 (all in nominal terms). For the purposes of modelling, it was assumed that

the levy continues to increase at 1c/kWh/a in nominal terms. The levy can be converted to real terms using a

discount factor of 6%. Note that the impact of the proposed carbon tax on generation cost was not explored as

part of this study.

Although costs of CCS were not included in the results, to provide a reference point it is suggested the upper

ends of the cost estimates could hold in South Africa (! 50 per tonne CO2 sequestered in 2008). Capture costs

are covered by the costs shown in Table 14, and transport and sequestration account for around ! 18 per tonne.

Inflating these figures at 6% p.a. from 2008 to get to 2010 values, and using a Rand/Euro exchange rate of 10 to

convert to Rands, this gives a 2010 cost of around R 202 per tonne CO2 sequestered for transport and

sequestration.

Technology learning

Technology learning will bring down the overnight capital costs of renewable and third generation nuclear

technologies. For the period 2010 to 2030, learning rates presented in the IRP2010 are used. For the period

2030 to 2040, it is assumed that the trend in learning observed between 2025 and 2030 continues. UCG-CCGT

is assumed to follow the IGCC technology learning curve.

FIGURE 24: IMPACT OF TECHNOLOGY LEARNING ON OVERNIGHT CAPITAL COSTS OF RENEWABLES

AND NUCLEAR (R/kW)

Supercritical PF power stations with FGD, FBC power stations and gas technologies are assumed to be mature

and costs do not change over time. The impact of technology learning on the costs of ultra-supercritical PF power

stations was not included, due to uncertainties associated with their costs and a lack of available data to develop

similar trajectories.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

2010 2015 2020 2025 2030 2035 2040

Co

st

rela

tive t

o 2

010

CCS Nuclear Wind CSP Solar PV UCG-CCGT


Results and analysis: electricity generation infrastructure investment cost and

electricity generation cost

Annual investment costs are shown in Figure 25 to Figure 28, while the indicative electricity generation cost is

shown in Figure 29 to Figure 32. The drop in costs at the end of the time period (from about 2035 onwards) on

Figure 25 to Figure 28, indicated by the shaded area, is attributed to the fact that the models only run to 2040,

and hence no upfront capital for plant built from 2041 onwards is accounted for. Hence this drop is a modelling

anomaly, and should not be interpreted literally. Similarly, generation costs do not take into account recovery of

capital for new build post 2040.

In interpreting generation costs, it is important to recognise that generation cost is not the same as

electricity price. Generation cost includes an allowance for capital cost of new build only (and excludes

depreciation of existing capital which is consistent across scenarios), O&M, fuel costs and an

environmental levy on electricity generated from fossil fuel. It thus excludes transmission and

distribution, CCS transport and sequestration and other costs typically taken into account in determining

electricity price.


FIGURE 25: ANNUAL INVESTMENT IN ELECTRICITY GENERATION CAPACITY (LAGS BEHIND)

FIGURE 26: ANNUAL INVESTMENT IN ELECTRICITY GENERATION CAPACITY (LOW CARBON WORLD)

FIGURE 27: ANNUAL INVESTMENT IN ELECTRICITY GENERATION CAPACITY (MORE OF THE SAME)

FIGURE 28: ANNUAL INVESTMENT IN ELECTRICITY GENERATION CAPACITY (AT THE FOREFRONT)

0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

2010 2015 2020 2025 2030 2035 2040

Investm

en

t [R

millio

n]


0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

2010 2015 2020 2025 2030 2035 2040

Investm

en

t [R

millio

n]


0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

2010 2015 2020 2025 2030 2035 2040

Inv

estm

en

t [R

millio

n]


0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

2010 2015 2020 2025 2030 2035 2040

Inv

estm

en

t [R

millio

n]



FIGURE 29: INDICATIVE ELECTRICITY GENERATION COST (LAGS BEHIND) FIGURE 30: INDICATIVE ELECTRICITY GENERATION COST (LOW CARBON WORLD)

FIGURE 31: INDICATIVE ELECTRICITY GENERATION COST (MORE OF THE SAME)

FIGURE 32: INDICATIVE ELECTRICITY GENERATION COST (AT THE FOREFRONT)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

2010 2015 2020 2025 2030 2035 2040

To

tal g

en

era

tio

n c

ost

[R/k

Wh

sen

t o

ut]

Coal Nuclear Renewables Gas Other

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

2010 2015 2020 2025 2030 2035 2040

To

tal g

en

era

tio

n c

ost

[R/k

Wh

sen

t o

ut]


0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

2010 2015 2020 2025 2030 2035 2040

To

tal g

en

era

tio

n c

ost

[R/k

Wh

sen

t o

ut]


0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

2010 2015 2020 2025 2030 2035 2040

To

tal g

en

era

tio

n c

ost

[R/k

Wh

sen

t o

ut]



The total capital investment, expressed as 2010 Rands, over various periods is shown in Table 16.

TABLE 16: COSTS OF BUILD PLANS (R BILLION)

2010-2015 2016-2020 2021-2030 2031-2040* Total

More of the Same

140 120 400 270 930

Lags Behind 140 120 480 500 1,240

At the Forefront 220 210 580 580 1,590

Low Carbon World

260 540 650 610 2,060

* Note that the investment during this time period excludes capital cost of infrastructure build required post 2040

More of the Same offers the lowest requirement for capital investment over the entire period, for the continued

building of supercritical power stations. The substantial capital cost implications of fitting CCS to new power

stations built from 2034 onwards in Lags Behind is highlighted by the figures in Table 16. This results in the

capital requirement for the 2031-2040 period being the highest of all four time periods for this scenario, and

accounts for the substantially higher costs of Lags Behind relative to More of the Same in the 2031-2040

period.

The total upfront capital expenditure associated with pursuing an electricity build plan that relies primarily on

renewables and nuclear under Low Carbon World is substantial. Notable here are the substantial investment

costs required for solar CSP that is brought on line in 2017 to 2021, causing the sharp peak observed in Figure

26, and the high capital costs associated with nuclear. It is highlighted that by 2031, when more solar CSP

begins to be brought online, the technology learning as shown in Figure 24 brings down the overnight costs of

this technology by almost 30% between 2017 and 2040. What this suggests is that using an alternative

technology to CSP (such as CCGT if the gas is available or wind) between 2017 and 2021 would bring down the

overall cost associated with Low Carbon World substantially.

On the other hand, a number of considerations need to be taken into account when interpreting this data, which

could contribute to the capital investment costs for At the Forefront and Low Carbon World as shown

representing an underestimate of what will likely be seen in practice. Firstly, the cost of new transmission

infrastructure is excluded from the analysis. Solar technologies will likely be located in the Northern Cape where

the solar resource is highest, but where there is limited existing grid infrastructure. Similarly, nuclear power

stations will likely be at the coast, so as to allow the use of seawater for cooling, thus requiring grid infrastructure

to transmit power to the inland centres of economic activity. A second factor, in Low Carbon World is that a high

renewables contribution to the grid requires additional infrastructure to manage grid stability. Thirdly, more recent

evidence suggests that the capital investment numbers would be different to those used in IRP2010, with the

costs of some renewables being lower and the cost of building nuclear being substantially higher – with the

impact of the higher cost of nuclear likely pushing the overall investment cost up. Finally, the costs of nuclear

waste management, liability, insurance and plant decommissioning are excluded – these will be substantial.

Although these will only be incurred post 2040, provision needs to be in for these costs in advance.

Moving on to total generation cost, which is different to electricity price in that it includes only the cost of local

generation of electricity and excludes the costs of imports, transmission and distribution, as well as the costs of

CO2 pumping and sequestration; More of the Same demonstrates the lowest overall generation costs starting

from about 21 cents per kWh in 2010 and rising to about 55 cents per kWh in 2040. The increased cost of

generation is attributed primarily to the cost of new generation infrastructure. The impact of cleaner coal


technologies can be seen by comparing More of the Same and Lags Behind. These two scenarios are similar

in terms of generation cost to just after 2030. The additional costs associated with ultra-supercritical PF

technologies is off set to some measure by the higher efficiency of ultra-supercritical as compared to supercritical

technologies used in More of the Same. Post 2034, however, when all new power stations are fitted with CCS

under Lags Behind, the cost of electricity generation increases significantly, ending about 13 cents higher by

2040 under Lags Behind as compared to More of the Same. The cost of the diversified mix built under At the

Forefront is lower than the coal-intensive build under Lags Behind, ending at 66 cents per kWh in 2040. Finally,

the electricity generation cost in 2040 is 14% higher under Low Carbon World than At the Forefront, due to the

cost of renewables and nuclear capacity that is built. However the generation figures for At the Forefront and

Low Carbon World need to once again be considered under estimates for the reasons given in the previous

paragraph.

Access to global funding

The four scenarios differ in terms of both the level of funding required for building of new infrastructure, and the

likely availability of funding from global sources:

• More of the Same requires the lowest level of funding on an annual basis. Given that coal

continues to be pursued as a primary energy source globally, there will be limited pressure on

international funders to move away from funding coal-fired power stations. As such, access to

funding is not expected to be a significant issue. The reduction in the level of sulphur dioxide

emissions from coal-fired power stations using a sorbent could be seen to be a prerequisite for

access to funding as has been observed already with loans to South Africa; the installation of

FGD for new PF and direct sorbent injection for FBC has been accounted for in the costing of the

build plans to account for this requirement.

• Funding requirements for Lags Behind are of a similar magnitude to those for More of the

Same, with the difference being due to installation of ultra-supercritical PF in the former and

supercritical in the latter. This is the case until the introduction of Carbon Capture and Storage

onto power stations built from 2034 onwards, at which point the cost of new power generation

increases significantly. As the remainder of the world has moved away from coal as a primary

energy source while South Africa continues to use coal, access to funding is likely to be a

significant challenge. The introduction of CCS would, however, need to be achieved through

access to international funding sources for greenhouse gas mitigation.

• Under At the Forefront, South Africa will be seeking funding for diversification of electricity

generation infrastructure, while the rest of the world shows limited activity in this regard. Access

to funding for nuclear and renewables could thus be challenging.

• Under Low Carbon World, the world moves away from coal-fired power towards nuclear and

renewables, and strong support is offered globally for this transition. The significant funding

required for CSP in the 2015-2020 period mentioned previously may, however, represent a

challenge if developed countries do not yet have mechanisms in place to provide this support by

then. It is recognised that the investment requirement for this transition is nonetheless very high,

and there could be competition for financial support from other developing countries also

undergoing the transition.


3.3.2 Coal price and revenue from coal sales

Revenue from local and export sales were used as two proxy indicators of contribution of the coal value chain to

the economy. These indicators were calculated by multiplying tonnages of coal by associated selling price (in

2010 Rands).

Local coal prices

As stated previously, coal sold to Eskom by mines that are already in existence is assigned an average price of

R 205.00 per ton in 2010, based on information provided by Eskom. For new mines, the price of coal was

calculated as being that which gave a return of investment of 10%, taking into account the cost of capital to

produce the different coal qualities required by Eskom (including mine establishment costs and costs of single or

multiple stage washing plants), production costs, and the revenues achieved from exporting coal.

Assumptions made to calculate coal prices are shown in Table 17 to Table 19, with the average values for the

Central Basin used in the models for those mines. Sales revenues for local metallurgical coal and non-Eskom

domestic thermal coal have not been included in the models.

TABLE 17: MINE ESTABLISHMENT COSTS (R/TONNE CAPACITY ROM)

Waterberg Central Basin

Surface Surface Underground Average

Mine 235 1,200 1,000 1,100

Single stage wash plant N/A 160

Multi-stage wash plant 230 200

Source: SACRM Expert Group (2013).

TABLE 18: COST OF PRODUCTION EXCLUDING TRANSPORT AND PORTS (R/TONNE)

Waterberg Central Basin

Production cost 56 225

Source: SACRM Expert Group (2013).

TABLE 19: COST OF TRANSPORT OF EXPORT PRODUCT FROM MINE TO PORT (R/TONNE)

Destination Transport cost

Waterberg to RBCT 258, rising to 308 in 2015

to account for cost of building new rail line from

the Waterberg

Mpumalanga to RBCT 126

Waterberg to Central Basin/Vereeniging

132

Port costs 15

Source: Average costs calculated from data presented for individual mines by Woodmac, and McGeorge, N. (2013), Personal Communication.

The five year rolling average price of coal sold to Eskom calculated using this approach is shown in Figure 33. As

identified above, coal supplied by existing contracts remains at R 205 per tonne over the analysis period, and

hence the steady growth in price is attributed to coal supplied through new mines supplying new power stations,


as well as existing power stations in excess of existing contracts. It is clear that the prices of coal sold to Eskom

will grow steadily under all four scenarios. The higher price of coal in Low Carbon World as compared to the

other scenarios is due to the fact that the under the model assumptions Low Carbon World requires

substantially less coal for power generation than the other scenarios, and a very much smaller proportion of this

is supplied from the Waterberg. Waterberg coal is substantially cheaper than Central Basin coal, bringing down

the average under the other three scenarios.

FIGURE 33: FIVE YEAR ROLLING AVERAGE PRICE OF COAL SOLD TO ESKOM

Coal export prices

Trajectories for export selling prices of a 27 MJ/kg export product were set as follows:

• More of the Same and At the Forefront are those in a world in which demand for export coal

remains high and hence coal prices continue to grow. Here export prices are assumed to follow

the trajectory proposed in the IEA “current policies” scenario to 2020, the Woodmac FOB Atlantic

price projection to 2030, and are then assumed to remain flat thereafter in the absence of any

projections.

• Lags Behind follows the coal prices as projected under the IEA Energy Technology

Perspectives’ New Policies scenario, representing a world which makes some progress in

moving away from coal, and takes account of global policy commitments already made in the

form of national pledges to reduce greenhouse gas emissions under the Copenhagen Accord.

The IEA projections are to 2035, prices are assumed to remain flat thereafter in the absence of

further projections.

• Low Carbon World follows the IEA 450 Scenario, which assumes that the world achieves

extensive policy implementation to achieve the upper estimate of greenhouse gas reductions,

with the aim of limiting atmospheric concentrations of CO2e to 450 ppm and global temperature

rise to 2°C. The IEA projections are to 2035, prices are assumed to remain flat thereafter in the

absence of further projections.

The resulting price trajectories are shown in Figure 35.

200

220

240

260

280

300

320

340

360

2010 2015 2020 2025 2030 2035 2040

Co

al p

rice [

R/t

]



FIGURE 34: PRICE TRAJECTORIES FOR 27 MJ/KG EXPORT PRODUCT

The prices of other thermal export products were calculated by calculating the ratio of current export prices to the

27 MJ/kg product as shown in Table 19. These ratios were assumed to remain fixed throughout the analysis

period. Metallurgical coal was priced at US$209 per tonne in 2010 terms30

.

TABLE 20: EXPORT PRICES OF DIFFERENT GRADES OF COAL

Product 2010 price

[US$/tonne]

Ratio of current price to that of 27 MJ/kg product

Low CV exports (23.5 MJ/kg) 64 0.71

Medium CV exports (25 MJ/kg) 75 0.83

High CV exports (27 MJ/kg) 90 1.00

Source: McGeorge, N. (2012), Personal Communication.

Results and analysis: local and export sales revenues

The prices discussed above were used in the calculation of electricity generation cost, as presented in Section

3.3.1. Export sales revenue and local sales revenue under the four scenarios are shown in Figure 35 and Figure

36. Under all of the scenarios, the contribution of thermal and metallurgical coal exports to bringing in foreign

revenue grows until 2030, bringing in almost R 120 billion per annum by 2030 in More of the Same and

R 80 billion in Low Carbon World (in 2010 Rands). Thereafter, however, the export revenue declines under all

of the scenarios. These results do, however, need to be understood in the context of the modelling assumptions

made in predicting future export volumes (see section 2.4), and particularly the assumption that no export only

mines are opened in the Waterberg. Particularly under At the Forefront and Low Carbon World, this stresses

the need for increased exploration of the Waterberg to determine the feasibility of export only mines in that

coalfield should South Africa wish to continue to ensure that coal contributes to earning foreign revenue. The

scenario models only implement one Waterberg mine configuration – that of a small stream of high-grade exports

and a large stream of low-grade utility coal. Export revenues could potentially be very much higher in More of

30 McGeorge, N. (2012), Personal Communication.

0

20

40

60

80

100

120

140

160

180

2010 2015 2020 2025 2030 2035 2040

$/t

More of the Same/Lags Behind At the Forefront Low Carbon World


the Same and Lags Behind should future development of the Waterberg move towards higher volumes of

lower-grade exports and/or export only mines.

FIGURE 35: LOCAL SALES REVENUE (ESKOM COAL ONLY)

FIGURE 36: EXPORT SALES REVENUE (METALLURGICAL AND THERMAL COAL)

3.3.3 Global competitiveness

The implications of following the different scenarios for South Africa’s global competitiveness can be qualitatively

described as follows:

• Under More of the Same, South Africa remains competitive globally. There is no pressure to

reduce dependence on coal, and markets and prices for coal exports remain strong. Electricity

generation cost remains relatively low, and hence electricity price will remain low, enabling

manufacturing to be competitive.

0

10

20

30

40

50

60

70

2010 2015 2020 2025 2030 2035 2040

Lo

cal sale

s r

even

ue [

R b

illio

n]


0

20

40

60

80

100

120

140

2010 2015 2020 2025 2030 2035 2040

Exp

ort

reven

ue [

R b

illio

n]



• Under Lags Behind, South Africa’s export markets could be penalised due to the continued coal

intensity of the economy, given the global drive away from coal towards lower carbon energy

sources. This penalty could be in the form of border tax adjustments on export products.

Electricity generation cost also remains low initially, but grows as CCS is introduced. South

Africa’s manufacturing sector consequently becomes increasingly non-competitive in high export

and energy intensive sectors.

• Under At the Forefront, South Africa does not gain any global competitiveness benefits from

diversification of its supply, except in the case of trade with a select few countries that continue

to pursue a low carbon trajectory. Electricity generation cost is ultimately comparable to Lags

Behind.

• Under Low Carbon World, South Africa’s efforts to reduce greenhouse gas emissions ensures a

place in the global export markets, as the world decarbonises. Electricity generation costs are

high, potentially resulting in uncompetitive energy intensive industries.

3.3.4 The cost of climate adaptation

It is widely accepted that the world is going to experience the impacts of climate change due to increased levels

of greenhouse gas emissions in the atmosphere from human activities. Such impacts as those associated with

extreme weather events (floods, droughts and heat waves), as well as changes in the long-term average climate

are projected to occur. Impacts include those on water resources, agriculture (and hence food security), forestry,

human health etc.

There are substantial costs associated with adapting to climate change impacts. Examples include costs of new

infrastructure, changing agricultural practices, protecting biodiversity, protecting coastlines and improving

resilience of rural and urban communities. Addressing food security issues, particularly in the poorer, least

developed areas, represents a substantial challenge for the country, and may require high levels of government

support.

The extent of climate change, the consequent impacts and the associated costs, including those for the coal

value chain, depend largely on global efforts at mitigation of greenhouse gas emissions, and less so on the

greenhouse gas emissions trajectory followed by South Africa. This is due to South Africa being a relatively small

emitter as compared to the world’s major emitters. On this basis, impacts and the consequent adaptation costs

are expected to be lower in Low Carbon World and Lags Behind, where global action is taken on climate

mitigation, than they are in More of the Same and At the Forefront, where greenhouse gas emissions continue

unabated. The timing of action is critical to determining the scale of impacts and adaptation costs: early action

may imply a greater upfront cost, but will reduce adaptation requirements down the line. Delayed action will in

turn result in a need for greater investment in adaptation and in dealing with the costs of impacts.

It is recognised that estimating the actual costs of impacts and adaptation requirements is challenging, with

requirements and costs being highly localised.

3.4 Energy Security

Energy security refers to the ability to maintain uninterrupted availability of a country’s main energy sources at an

affordable price. Two key factors are suggested to be important in this regard, being reliant on local resources as

opposed to energy imports, and technology related considerations.


3.4.1 Reliance on local resources versus energy imports

Electricity-related imports include imported gas, hydropower and coal-fired power (from Botswana and

Mozambique), while liquid fuels-related imports include crude oil and gas for GTL. More of the Same and Lags

Behind offer the greatest security for electricity provision in this regard, as the overwhelming majority of

electricity is supplied by coal from local resources. A small amount of imported gas, hydropower and coal-fired

power are found in the mix, but this is suggested to be insufficient to impact on domestic energy security. At the

Forefront performs similarly to More of the Same and Lags Behind, although there is an additional 1,150 MW

of imported hydro and 300 MW of imported coal-fired power in At the Forefront. Low Carbon World considers

a somewhat higher reliance on gas (6,000 MW versus 3,000 MW in the other scenarios). This gas may be

imported, and so may offer less security. In general, however, the difference between the scenarios in terms of

gas and electricity imports is suggested not to be high enough to be of considerable concern to the country.

Having said this, Low Carbon World and to a lesser degree At the Forefront build a number of new nuclear

power stations. At present, although uranium is mined locally, processed nuclear fuel is imported from overseas,

which is less desirable from an energy security point of view. Increased local fuel processing would help to

overcome this concern.

In terms of liquid fuels supply, building of two new CTL plants under More of the Same will help to reduce

reliance on foreign crude oil, global demand for which is likely to grow dramatically given the continued global

reliance on fossil fuels under this scenario. Lags Behind will also provide a high level of liquid fuels security.

Although only one CTL plant is built here, global demand for crude oil will reduce as the remainder of the world

decarbonises. The lowest liquid fuel security is under At the Forefront under which no new CTL plants are built

locally, and global demand for fossil fuels continues to grow, given the lack of focus on reducing fossil fuel

consumption. Finally, Low Carbon World sees a shift away from fossil fuel dependence globally and locally.

Here there will be a focus on alternatives in mobility to liquid fuels, which in turn is dependent on technology

development.

It is noted that Project Mthombo, the proposed crude oil refinery at Coega, could result in new CTL plants not

being built under More of the Same and Lags Behind. Although local refining capacity will then be available,

energy security will be impacted by global crude oil availability.

3.4.2 Technology considerations

A strong reliance on foreign companies to supply and service power generation plant could become a concern for

energy security moving into the future. The extensive roll out of nuclear power stations under Low Carbon

World, and to a somewhat lesser degree At the Forefront, is of particular importance to consider here. South

Africa has historically been dependant on foreign suppliers to provide support for Koeberg. Unless a significant

training programme for local nuclear skills development is launched, the reliance on foreign service providers will

continue to grow. If the uptake of nuclear power continues to grow in other countries, particularly under Low

Carbon World, these skills will become increasingly scarce. A similar situation of foreign dependence on support

for renewables could arise under these two scenarios, unless there is a substantial development of a local

renewables industry in South Africa.


3.5 Employment and other socio-economic considerations

Employment data is expressed as intensities, or jobs per tonnes coal, per MWh, or per MW capacity installed etc.

Employment is then calculated by multiplying activity data by the employment intensities.

3.5.1 Mining

An average employment intensity was calculated from data obtained from the Chamber of Mines for the industry

for 2010. This data suggests 73,817 employees are employed for 254,529 kt saleable production, giving an

average intensity of 0.29 employees/kt saleable production31

. For comparison, BHP Billiton Energy Coal South

Africa (BECSA) has approximately 6,000 full time equivalents for 34 Mt of sales i.e. 0.17 employees/kt sold, while

Anglo American has 15,560 employees for 58.5 Mt sold, or 0.27 employees/kt saleable production.

The following limitations relating to mining employment data used in the models are identified:

• The data does not distinguish between employment on opencast and underground mines;

• No data on employment intensity of beneficiation was found. As such, this is assumed to be included in

the employment intensity of coal mining;

• Data excludes employment in mine construction phases as no data was found;

• Data is assumed to include contractors;

• No data on indirect employment could be found; and

• The models assume that employment intensity does not change over time. The impact of increased

mechanisation could reduce employment intensity over time, and hence employment in mining could be

over estimated.


Employment associated with construction and operation of electricity generation plants is presented below. As for

mining, indirect employment was not considered for electricity generation.

Coal-fired power stations

Power station construction

The Eskom Factor Project32

reports employment over a period of eight years for the construction of Kusile power

station as shown in Table 21.

TABLE 21: FULL-TIME EMPLOYEE YEARS (FTE) FOR CONSTRUCTION OF KUSILE

Year FTE

2011 6,400

2012 8,600

2013 10,000

2014 11,000

2015 9,100

31 Department of Mineral Resources (2011) Facts and Figures 2011. Available online: www.bullion.org.za/, accessed August 2012. 32 Eskom (2011) Eskom Factor Report 2011. Available online: www.eskomfactor.co.za, accessed August 2012.


2016 5,200

2017 4,200

2018 1,900

TOTAL 56,400

The nominal installed capacity of Kusile is 4,800 MW, suggesting total construction jobs of 11.75 FTE-years/MW

nominal installed capacity for new PF coal-fired power stations. This ratio is used for all new large PF coal-fired

power stations. Half this employment intensity (5.9 FTE-years/MW nominal installed capacity) is assumed for

small FBC plants (500 MW net max) in the absence of better information, given the shorter construction times.

Coal-fired power station operation

No consistent data sets could be found to provide an indication of employment for operation and maintenance of

the coal-fired power station fleet. In personal communication, Eskom has suggested that the average

employment is about 900 people per power station (including both permanent employees and contractors), which

is the same as the average calculated from data presented on employment at individual power stations on the

Eskom website33

. The figures for the individual power stations are thus used for the existing power stations. For

new power stations, Eskom has suggested that employment would be lower, at 800 people per power station.

This is the employment used for the operation of new PF coal-fired power stations.

No figures were found for employment at FBC power stations. FBC installations in the model are approximately

12% of the installed capacity of a PF power capacity. It was assumed that employment at FBC power stations is

10% of that at a PF station, or 88 employees.

Renewable electricity generation technologies

Employment intensities for construction and manufacturing of wind, CSP and solar PV are shown in Table 22.

TABLE 22: EMPLOYMENT INTENSITIES FOR MANUFACTURING AND CONSTRUCTION OF

RENEWABLES

Technology FTE-years/MW installed capacity

Wind Energy

Employment intensity for construction 1.5

Employment intensity for manufacturing 4.5

Concentrated Solar Power (CSP)



Solar PV



33 Eskom, Power station data. Available online: http://www.eskom.co.za/c/12/power-stations/, accessed August 2012. Note that no information is given as

to the year for which this data applies.


Source: Maia et al (2011) and others34

It is recognised that manufacturing of renewables could very likely be conducted overseas. Under More of the

Same and Lags Behind, which focus on coal-fired power, it is assumed that renewables build uses imported

technologies, and hence no manufacturing jobs are created in South Africa. Construction jobs are assumed to be

local. For At the Forefront and Low Carbon World, In line with the assumptions used in developing the IRP35

,

30% of the jobs created in wind manufacturing are assumed to be local, 50% of those in solar thermal and 30%

in solar photovoltaic. Construction jobs are assumed to be local.

Employment intensities for operation and maintenance of renewables are provided in the same report from which

manufacturing and construction employment intensities are taken (Maia et al.) and reproduced in Table 23.

TABLE 23: EMPLOYMENT INTENSITIES FOR O&M OF RENEWABLES

Technology Employees/MW installed capacity

Wind 0.5

CSP 0.54

Solar PV 0.7

Nuclear

Employment data for the construction phase of nuclear power stations is based on an estimate made by Arcus

Gibb in the Environmental Impact Assessment (EIA) for “Eskom Nuclear 1”, the proposed first nuclear power

plant to be built 36

. This study suggests that 8,737 people will be employed every year for the 9-year period

required to build a 4,000 MW power station. The employment intensity is thus assumed to be 8,737*9/4,000 =

19.65 FTE-years/MW. However, in line with the EPRI data used in the IRP, only 40% of the labour is assumed to

be local labour hence a value of 7.86 FTE-years/MW nominal installed capacity is assumed 37

.

In the operational phase, Koeberg currently employs approximately 1,200 employees, for a power station with a

nominal installed capacity of 1,910 MW, suggesting 0.62 employees per MW installed capacity38

. The Arcus Gibb

EIA study suggests 1,385 permanent employees for a 4,000 MW power station, or an employment intensity of

0.35 employees per MW installed capacity. The latter figure is assumed in the models, given firstly that nuclear

power stations may be larger than Koeberg, and secondly that more recent designs may be more automated and

less labour intensive.

34 Construction and manufacturing employment for wind and solar PV, manufacturing employment for CSP and operational phase jobs, were taken from Maia, J., Giordano, T., Kelder, N., Bardien, G., Bodibe, M., Du Plooy, P., Jafta, X., Jarvis, D., Kruger-Cloete, E., Kuhn, G., Lepelle, R., Makaulule. L., Mosoma, K., Neoh, S., Netshitomboni, N., Ngozo, T. and Swanepoel, J. (2011) Green jobs: an estimate of the direct employment potential of a greening

South African economy. Industrial Development Corporation, Development Bank of Southern Africa, Trade and Industrial Policy Strategies. The figures for solar CSP in this report appeared, however, implausable at 21.6 jobs per MW capacity. As such the average figure from a US study was used, which reported 0.85 to 4.65 peak construction jobs per MW. A value of 2.75 jobs per MW at the peak was thus used. The IRP assumes CSP is built over 4 years, with 10% of capital expenditure occurring in year 1, 25% in year 2, 45% in year 3 and 20% in year 4. These same ratios were used to scale jobs, i.e.

10/45*2.75 = 0.6 jobs per MW in year 1, 25/45*2.75 = 1.53 obs/MW in year 2 and 20/45*2.75 = 1.22 jobs/MW in year 4, giving a total of 6.1 jobs/MW (http://webservices.itcs.umich.edu/drupal/recd/?q=node/64). 35 EPRI (2010) Power Generation Technology Data for Integrated Resource Plan of South Africa. Palo Alto, CA. 36 Arcus GIBB, Environmental Impact Assessment for the Proposed Nuclear Power Station (‘Nuclear 1’) and Associated Infrastructure: Social Impact

Assessment. Available online : http://projects.gibb.co.za/portals/3/projects/201104%20N1%20DEIR/27.%20APP%20E2%20to%20E30%20Specialist%20Reports/Rev%20DEIR%20APP%20E18%20Social%20Impact%20Assessment.pdf, March 2011, accessed August 2012 37 EPRI (2010) Power Generation Technology Data for Integrated Resource Plan of South Africa. Palo Alto, CA 38 Eskom (2012) Koeberg Power Station. Available online: http://www.eskom.co.za/c/74/koeberg-nuclear-power-station/, accessed August 2012


OCGT, CCGT and UCG-CCGT

The Draft EIA for the Mossel Bay OCGT (Gourikwa) power station suggests the creation of 358 employment

opportunities during the construction of the OCGT power plant and transmission substation. Although not clear

from the report, this is assumed to be in FTE-years. The EIA was for a 150 MW plant, suggesting 358/150 =

2.39 FTE-years/MW during construction.

For CCGT, construction will likely take place over 3 years (see discussion below on phase-in of capital).

Employment in construction is assumed to be scaled by multiplying the OCGT construction figure by 3/2, giving a

total of 537 FTE-years for a similar size plant. This provides an employment intensity of 537/150 = 3.58 FTE-

years/MW nominal capacity installed.

Operational phase employment at OCGT power stations is small, with the same EIA for a 150 MW plant

suggesting 20 jobs being created during the operational phase, or an employment intensity of 0.133 jobs/MW

installed capacity. The same figure is assumed for OCGT plants.

No data was found for employment associated with construction of UCG-CCGT power stations. In the absence of

further information, it was assumed that employment to construct a UCG plant was 50% higher than for CCGT –

in other words 3.58*1.5 = 5.37 FTE-years/MW.

A study conducted by Indiana State University39

reports a wide range of figures for employment during the

operational phase of UCG-CCGT. Based on the figures presented in that report, a nominal 100 employees for a

250MW UCG power station, or an employment intensity of 0.4 employees/MW installed capacity is assumed in

the models.

Local hydro, pumped storage and cogeneration

Small hydropower build and cogeneration form a very small component of the long-term electricity build plan and

are not particularly employment intensive, and hence employment associated with these categories is excluded

from the models. In the absence of any available information, employment at pumped storage facilities is also

excluded. This is expected to be small.

Imported electricity from coal and hydro

No employment was allocated to import coal and hydro as construction and operation occurs outside of the

borders of South Africa and hence has no employment implications for South Africa.

Distribution of jobs over the power station construction period

It was assumed that the employment profile for construction of power stations and renewables follows the

phasing of investment capital, as per the IRP2010. However, where power stations were built over shorter

periods of time, the distribution of jobs was manually adjusted to take this into account.

39 IU (2011) Viability of Underground Coal Gasification with Carbon Capture and Storage in Indiana. Prepared by the School of Public and Environmental

Affairs, Indiana University (Submitted May 4, 2011)


3.5.3 Coal-to-Liquids

In terms of CTL construction, a 2008 Engineering News article40

suggests that 37,000 jobs could be created over

five years for the construction of a 80,000 bbl/day CTL plant. In the absence of any further data, it is assumed

that this figure represents FTE-years of employment, and that these are spread over the five years as follows:

Year (where year 0 is when the plant begins to produce output)

-2 -1 0 1 2

Percentage of jobs (FTE-years) 10% 28% 33% 18% 11%

Employment at Sasol Secunda for 2010 was 5,362 employees for a 160,000 bbl/day CTL refinery41

. A new

80,000 bbl/day CTL refinery was assumed to employ half of this number.

3.5.4 Richards Bay Coal Terminal

Employment at RBCT is included in the model. In 2012 RBCT employed approximately 500 people42

, and it is

assumed that this figure will not change with an increase in coal handled from current levels to the maximum

capacity of RBCT. It may be possible that there will be a small amount of growth if coal volumes increase.

However, given this contributor to employment as compared to those of other elements of the value chain is

small, this assumption is not expected to have any effect on the results.

3.5.5 Transnet

Transnet Freight Rail has approximately 25,000 employees43

. It is impossible to allocate a proportion of these

employees to coal transport specifically, based on information contained in the public domain. For the purposes

of the model, therefore, an assumption was made that employment is proportional to the tonnes of each

commodity transported by TFR. In 2009-2010 TFR transported a total of 178.6 Mt of freight. Coal exports were

61.8 Mt and coal transported for Eskom was 30.5 Mt in 2010, suggesting 92.3 Mt or 51.7% was made up by

coal44

. Employment associated with coal transport is thus assumed to be 12,925 in 2010. It is not clear how this

will increase with increased volumes of coal being transported, so for modelling purposes it was assumed that for

every 10% increase in coal transported to export markets, there will be a 5% increase in employment. No

allowance is made for jobs during construction or upgrading of rail lines or RBCT.

3.5.6 Results and analysis: Employment under the SACRM scenarios

Employment in mining

Employment in mine construction was not included due to the lack of data available in the public domain, and the

wide variability across different types of mines and mine locations.

Employment in coal mine operation in the Central Basin and Waterberg coalfields is shown in Figure 37 and

Figure 38. Under all scenarios, mining reaches a peak in the Central Basin in 2020 when Kusile is fully brought

40 Kolver, L. (2008) Sasol making progress on the globalisation of its coal-to-fuels technology. Available online: http://www.engineeringnews.co.za/article/sasol-making-progress-on-the-globalisation-of-its-coaltofuels-technology-2008-03-28, accessed August 2012 41 Sasol (2011) Sasol Integrated Report 2011. Available online: http://www.sasol.com/sasol_internet/frontend/navigation.jsp?navid=21100001&rootid=3; 18 June 2012. Pg 83, accessed August 2012 42 RBCT (Richards Bay Coal Terminal) (2012) Economic overview. Available online: http://www.rbct.co.za/jit_default_1108.Economic_overview.html, accessed 27 June 2012 43 Transnet website (2012). Available online: http://www.spoornet.co.za/, accessed August 2012 44 Transnet Annual Report 2009-2010, Eskom Annual Report 2010.


on line, and then begins to decline as power stations are decommissioned. From the 2020 peak to 2040,

between 31,000 and 35,000 jobs are lost in this coalfield, depending on the scenario. In the Waterberg, however,

employment creation depends much more strongly on the scenario followed: under More of the Same and Lags

Behind there is on-going substantial employment creation in that coalfield, reaching about 20,000 employees by

the mid-2020’s, and over 50,000 employees by 2040. Job creation in the Waterberg under At the Forefront is

more moderate, growing from 10,000 in 2025 to 21,000 by 2040. Low Carbon World is the only scenario with

limited growth in mining employment in the Waterberg, and as with At the Forefront, nation-wide there is an

overall decline in coal mining employment under these scenarios.

FIGURE 37: EMPLOYMENT IN MINING IN

CENTRAL BASIN

FIGURE 38: EMPLOYMENT IN MINING IN THE

WATERBERG

Employment in construction of power stations and CTL plants

Figure 39 shows the total number of construction jobs created for power stations and CTL plants for the different

scenarios. This figure demonstrates little distinction between the scenarios, especially taking uncertainties in

employment intensities into account. The key outlier here is the jobs associated with the intensive build of solar

CSP and wind under Low Carbon World between 2014 and 2021. Supplying this level of skilled employees

within the relatively short term in South Africa could well present a substantial challenge.

A further concern here is availability of skilled personal for building of nuclear power stations under At the

Forefront and Low Carbon World. An estimated 96,100 FTE-years of employment is created between 2010

and 2030 under At the Forefront and 122,000 FTE-years of employment under Low Carbon World in nuclear

alone. There is reportedly a global shortage of skilled people to build nuclear power stations, so sourcing skills

could be challenging.

0

20,000

40,000

60,000

80,000

2010 2020 2030 2040

Em

plo

yees


0

20,000

40,000

60,000

2010 2020 2030 2040

Em

plo

yees



FIGURE 39: CONSTRUCTION JOBS FOR POWER STATIONS AND CTL

Employment in operation of power stations and CTL plants

Employment in operation of power stations and CTL plants shows a different picture under the four scenarios to

that of construction employment. Figure 40 shows that Low Carbon World has the highest number of permanent

employees, with 57% higher employment than Lags Behind in 2040. The majority of the job creation in Low

Carbon World is in the renewables sector, as shown in Figure 41. Employment associated with CTL in More of

the Same is about 34% of the combined employment of power stations and CTL plants by 2040, in Lags Behind

it is 27%, in At the Forefront it is 13% and in Low Carbon World it is 11%. Separating out power station

employment only thus suggests an even greater differential between the scenarios.

FIGURE 40: EMPLOYMENT ASSOCIATED WITH OPERATION OF POWER STATIONS AND CTL

0

10,000

20,000

30,000

40,000

50,000

60,000

70,000

2010 2015 2020 2025 2030 2035 2040

Nu

mb

er

of

FT

Es


0

10,000

20,000

30,000

40,000

50,000

2010 2015 2020 2025 2030 2035 2040

Op

era

tio

nal p

hase e

mp

loyees



FIGURE 41: OPERATION PHASE EMPLOYMENT UNDER LOW CARBON WORLD IN POWER STATIONS

AND CTL

Summary: total operational phase employment

Total employment for the operational phase of mining, power stations and CTL is shown in Figure 42 to Figure

45. These figures show the overwhelming dominance of mining jobs in determining the overall employment

profile for the coal value chain, and that around 35,000 more jobs are created by 2040 in the coal dominant Lags

Behind, compared to Low Carbon World. Once again, employment figures exclude indirect employment.

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

45,000

2010 2015 2020 2025 2030 2035 2040

Op

era

tio

nal p

hase e

mp

loyees



FIGURE 42: TOTAL EMPLOYMENT (LAGS BEHIND) FIGURE 43: TOTAL EMPLOYMENT (LOW CARBON WORLD)

FIGURE 44: TOTAL EMPLOYMENT (MORE OF THE SAME) FIGURE 45: TOTAL EMPLOYMENT (AT THE FOREFRONT)

0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

2010 2015 2020 2025 2030 2035 2040

Nu

mb

er

of

em

plo

yees

Mining Electricity generation CTL Transnet

0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

2010 2015 2020 2025 2030 2035 2040

Nu

mb

er

of

em

plo

yees


0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

2010 2015 2020 2025 2030 2035 2040

Nu

mb

er

of

em

plo

yees


0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

2010 2015 2020 2025 2030 2035 2040 N

um

ber

of

em

plo

yees



3.5.7 Other socio-economic considerations

Employment and contribution to export and local sales revenue (as discussed in Section 3.3.2) represent a proxy

for the contribution of the coal value chain to South Africa’s economy and society. There are, however, other

considerations that need to be taken into account when assessing the overall impacts of the value chain. Notable

here is indirect employment – apart from the direct employees quantified above, the coal value chain requires a

vast array of material and service inputs which would no longer be required in the case where coal production

reduces (At the Forefront and Low Carbon World). At the same time new opportunities open up through

development of a renewables and nuclear sector under these scenarios. Quantification of the relative job losses

and gains is not undertaken here.

At the same time, it needs to be recognised that socio-economic impacts are localised. More of the Same and

Lags Behind, which will involve development of mines, power stations and CTL plants in the Waterberg, will

contribute to economic growth in these areas, through the increase in the sizes of communities in these areas.

Effective planning is required to ensure that suitable facilities are established in these areas to support growing

communities.

Provision also needs to be made for the post-2030 period in the Central Basin, when many mines are closed and

power stations are decommissioned, potentially resulting in significant unemployment in this region. Strategic

planning is required to ensure that interventions such as relocation of employees to the Waterberg (in More of

the Same and Lags Behind) and reskilling of employees contribute to minimising this impact.

Significantly, the economic costs of mining and coal-fired power generation on the environment and society have

not explicitly been quantified in this study.

3.6 Water demand

As with the other impacts considered, water demand is expressed on an intensity basis (e.g. kl or m3 per activity),

and overall water consumption is then calculated by multiplying the intensity with the activity such as tonnes coal

produced or MWhSO.

3.6.1 Mining and beneficiation

Estimates used in the models for calculation of water demand in mining are shown in Table 24. The literature that

was reviewed for this study (data from individual mines and a 2001 survey of consumption in the mining industry)

showed a very wide range of water usages for different mines, with no clear trends or differences between

opencast and underground mining. Much of the variability arises from whether or not the mine has excess water,

e.g. through seepage and rainwater collection. To remove this variability to some degree, the values in Table 24

represent purchased water only (river and municipal water sources). Note that the wider impact of coal mines on

water catchment is discussed qualitatively in Section 3.9.1.


TABLE 24: WATER PURCHASED FOR MINING AND BENEFICIATION (Ml/MT ROM)

Mine type Value applied

Central Basin underground (mine only) 25

Central Basin opencast (mine only) 50

Central Basin combined or unknown 40

Central Basin wash plant only 85

Waterberg mine and wash plant 65

Source: Pulles and Notten45


Table 25 presents the water intensity factors and assumptions used in the models to calculate water demand in

electricity generation. Actual data for Eskom power stations was used where available. For future build and

renewables, water intensity factors were taken from the supporting technical documentation for the IRP46

. The

water factors used in Table 25 were also checked and found to be in line with those in the Harvard Kennedy

School review47

. The increased water requirements as a result of inclusion of CCS are discussed in Section

3.6.4.

TABLE 25: WATER DEMAND IN ELECTRICITY GENERATION

Application Value applied Units Comments Reference

Fossil fuels

PF wet cooled power stations: existing

1.70 – 2.01 l/kWhSO Values for existing power stations

chosen so as to get Eskom’s overall water intensity figure quoted in the 2012 Divisional Report (1.34

l/kWhSO). Grootvlei is the one station that falls outside of this range with a demand of 1.38 l/kWhSO.

Eskom

PF dry cooled power stations: existing

0.12 l/kWhSO Eskom

New coal: supercritical PF, dry cooled with FGD

0.36 l/kWhSO Value for Medupi and Kusile, as well as future supercritical build

Eskom plans

(personal communication)

Above value comprised of:

Dry cooled supercritical PF 0.11 l/kWhSO

FGD 0.25 l/kWhSO Average figure, actual water use for

FGD will depend on performance and technology at each station.

New coal: ultra-supercritical PF, dry cooled with FGD

0.31 l/kWhSO Water usage as for supercritical PF

with FGD, but assumes 5% increase in thermal efficiency.

Retrofit supercritical PF, dry cooled with FGD and CCS

0.67 l/kWhSO Calculated taking into account cooling

load of CCS (assuming wet cooled) and net efficiency drop of plant. This

figure applies to Medupi and Kusile when retrofitted with FGD and CCS.

See Section 3.6.4

New coal: ultra-supercritical PF, dry cooled with FGD and CCS

0.59 l/kWhSO Calculated taking into account cooling

load of CCS (assuming wet cooled) and net efficiency drop of plant.

See Section 3.6.4

45 Range of values from: Pulles, W., Boer, R.H. and Nel, S. (2001) A Generic Water Balance for the South African Coal Mining Industry. WRC Report No 801/1/0145 and Notten, PJ. (2001) Life Cycle Inventory Uncertainty In Resource-Based Industries - A Focus On Coal-Based Power Generation, PhD

Thesis, University of Cape Town 46 EPRI (2010) Power Generation Technology Data for Integrated Resource Plan of South Africa, Palo Alto, CA 47 Water consumption of energy uses, resource extraction, processing and conversion, from Energy Technology Innovation Policy Research Group, Harvard Kennedy School, 2010



Fluidised bed combustion

(FBC): dry cooled with sorbent injection

0.11 l/kWhSO Assumed to have same water

consumption as dry cooled supercritical PF station

Eskom

Nuclear at coast 0.055 l/kWhSO Consumption only, excludes sea water which is returned

Eskom (personal communication)

UGC-CCGT 0.26 l/kWhSO Figure is for IGCC. The water

consumption thus excludes water for UGC component, which is highly variable depending on the resource,

i.e. can either be a net producer of water (because of need to maintain hydraulic gradient) or a net consumer of water.

EPRI (2010)

OCGT 0.020 l/kWhSO Air cooled EPRI (2010)

CCGT 0.013 l/kWhSO Air cooled EPRI (2010)

Renewables

CSP parabolic trough, 9 hours storage (dry cooled)

0.3 l/kWhSO Air cooled condensers, primary use of water is for mirror washing

EPRI (2010)

Solar PV 0.024 l/kWhSO Water for washing PV panels EPRI (2010)

3.6.3 Coal-to-liquids

Secunda used to the order of 88.4 Mm3 of water in 2010 and this level of consumption is assumed to continue

throughout the analysis period48

.

Water supply is needed for new coal-to-liquids plants in More of the Same and Lags Behind. Demand is

estimated at 8 – 10 m3/tonne for an 80,000 bbl/day plant, lower than Secunda due to dry cooling and extensive

effluent re-use and recycling49

. The total annual water consumption of a new 80,000 bbl/day plant is thus

estimated at 37 Mm3/a. The water demand figure for new CTL plant (approximately 9.6 m

3/ton) is assumed to

include that required for CCS (see next section).

3.6.4 Carbon Capture and Storage

It is clear in the literature that post-combustion CCS results in an increase in water use for electricity generation.

This is primarily due to the large cooling water demand of the amine process, with a smaller amount of water

required in the scrubber. There is also an increase in the water required in the FGD process. The majority of the

literature is for wet-cooled stations, and shows a large range in increases with post-combustion CCS (from 83%

to 290%). However, it is reasonable to assume that in South Africa the additional cooling water demand of the

amine process would be dry-cooled, rather than incur the substantial water increases cited in the literature for

wet-cooled plants. The water requirements of the post-combustion CCS plant are therefore estimated using the

following information:

• Approximate doubling in cooling duty50

.

• Approximately 45% increase in FGD water use (net sent out basis)15,51

48 Sasol (2011) Sasol Integrated Report 2011. Available online: http://www.sasol.com/sasol_internet/frontend/navigation.jsp?navid=21100001&rootid=3; 18

June 2012. Pg 83, accessed August 2012 49 Meyer, A. (Sasol) (2012) Personal communication. 50 NETL (2008) Water Requirements for Existing and Emerging Thermoelectric Plant Technologies, DOE/NETL-402/080108, August 2008 (Revised April 2009) 51 HZhai, H., Rubin, E.S. and Versteeg, P.L. (2011) Water Use at Pulverized Coal Power Plants with Postcombustion Carbon Capture and Storage, Environmental Science and Technology, 45, pp. 2479–2485


• Approximately 90 litres/MWh required in amine scrubber 16

The values in Table 25 are derived from the above three assumptions and the value provided by Eskom for a

dry-cooled supercritical plant (0.36 l/kWh net sent out, comprised of 0.11 l/kWh net sent out general plant water

consumption and 0.25 l/kWh net sent out FGD water consumption). These values are assumed to be 20% and

11% lower, respectively, for an ultra-supercritical plant.

No additional efficiency penalty is applied for the use of dry-cooling rather than wet-cooling, i.e. the net thermal

efficiency drops given in Table 13 are straight from the wet-cooled literature and are likely therefore to be an

underestimate.

3.6.5 Communities

Provision needs to be made for water usage in communities surrounding new power stations and mines,

especially for the development in the Waterberg. The work done around Phase 2 of the Mokolo and Crocodile

Water Augmentation Project explores the need for new water supply for new mines and power stations to the

Waterberg. This work suggests a requirement for municipal water associated with a coal-fired power station

(4,368 MW) and associated mine of 1.5 Mm3/a. This demand has been included in the models. No allowance has

been made for increased municipal demand in the Central Basin as it is suggested this will not change

significantly.

3.6.6 Results and analysis: Water demand under the SACRM scenarios

Water demand in mining

Water demand for mining in the Central Basin (Figure 46) peaks between 2015 and 2020 in More of the Same,

Lags Behind and At the Forefront, and in 2015 in Low Carbon World at about 12 Mm3/a and then decreases

to between 15% and 30% of 2010 levels by 2040, depending on the scenario. In the Central Basin, the scenarios

follow very similar trajectories, since water demand tracks existing mining projects which are common to all

scenarios. More of the Same ends slightly higher than the other scenarios. Water stress is already experienced

in most of the water catchments supplying the Central Basin (see Section 3.9.1). To supply the increased water

demand up to 2020, therefore, solutions will have to be found, which include the costly, but technologically

proven, desalination of contaminated mine water and much higher reuse of effluents (including sewage). Some

relief will be obtained with older power stations being decommissioned, although this occurs after the 2020 water

demand peak has been reached. Furthermore, where water from the Upper Vaal is used to supplement water

requirements in Limpopo for development of the Waterberg, the water freed up by decommissioning power

stations and mines will be used in the Waterberg and will do little to relieve the water stress in the Central Basin.

Mining water demand in the Waterberg (Figure 47) is very different for the four scenarios. More of the Same and

Lags Behind show the considerable water required in the mines supplying the new Waterberg power stations (to

the order of 25 Mm3 by 2040). Low Carbon World shows essentially constant water demand as the mine

supplying the existing Waterberg power stations first ramps up to full coal supply for Medupi power station, but

then ramps down again towards the end of the period when Matimba power station is closed. At the Forefront

shows moderate growth in water demand particularly after 2030.


FIGURE 46: WATER DEMAND FOR MINING IN THE

CENTRAL BASIN

FIGURE 47: WATER DEMAND FOR MINING IN THE

WATERBERG

Under More of the Same and Lags Behind, water demand for mining in the Waterberg by 2040 is substantially

higher than that currently used in the Central Basin. This water is needed for the extensive washing required of

the Waterberg coals, and thus advances in mining technologies, especially dry beneficiation techniques, will be

required to alleviate the water demand.

Water demand in electricity generation

Figure 48 shows total national water demand for electricity generation, while Figure 49 shows water intensity of

electricity supply. More of the Same is more water intensive than Lags Behind due to supercritical power

stations being built in the former as opposed to ultra-supercritical in the latter. Both however show a very

significant decline in water intensity over the period as wet-cooled power stations are replaced by more efficient

dry-cooled power stations. At the Forefront shows an on-going decline in water demand as coal-fired power

stations are brought off line and are replaced by renewables and nuclear which are less water intensive. Of the

four scenarios, Low Carbon World is the least water intensive, from about 2018 onwards, again due to early

decommissioning of coal-fired power replaced with lower water intensive nuclear and renewables, although water

demand and intensity in Low Carbon World exceeds that of At the Forefront for a short period after 2029,

which is attributed to the water penalty from the retrofit of Medupi and Kusile with CCS.

The retrofit of Medupi with FGD in 2021 under all of the scenarios also causes a noticeable spike in total water

consumption in all scenarios.

0

2

4

6

8

10

12

14

2010 2020 2030 2040

Wate

r d

em

an

d [

Mm

3/a

]



0

5

10

15

20

25

30

2010 2020 2030 2040

Wate

r d

em

an

d [

Mm

3/a

]




FIGURE 48: NATIONAL WATER DEMAND PER SCENARIO FOR POWER STATIONS

FIGURE 49: WATER INTENSITY OF ELECTRICITY GENERATION

Water demand by coalfield

Given the importance of water demand at a regional level rather than a national level, water demand is also

considered disaggregated by coalfield in Figure 50 to Figure 53. Notable here is the following:

• The dramatic drop-off in demand for water in the Central Basin as coalfields are

decommissioned, which occurs earlier and to a greater degree in At the Forefront and Low

Carbon World than it does in More of the Same and Lags Behind as coal-fired power stations

are decommissioned earlier under the former two scenarios than the latter two.

• The rise in water demand in the Waterberg as new power stations are built in that coalfield under

More of the Same and Lags Behind, with a significant increase in Lags Behind post 2034 as

new power stations coming on line are fitted with CCS. Given low water availability in that

0

50,000

100,000

150,000

200,000

250,000

300,000

350,000

2010 2015 2020 2025 2030 2035 2040

To

tal w

ate

r d

em

an

d f

or

po

wer

sta

tio

ns

[Mm

3/a

]


0

0.2

0.4

0.6

0.8

1

1.2

1.4

2010 2015 2020 2025 2030 2035 2040

Wate

r in

ten

sit

y o

f p

ow

er

gen

era

tio

n [

Ml/

GW

hS

O]



coalfield, sufficient infrastructure will need to be built to ensure this water can be supplied –

assuming it is available from elsewhere in the country.

• A relatively small but notable increase in water demand in the Waterberg when FGD is fitted to

Medupi in 2021 under all scenarios.

• Some growth in demand for water in other parts of South Africa to supply nuclear and renewable

technologies (particularly in At the Forefront and Low Carbon World). Solar power plants are

likely to be located in the Northern Cape where the solar resource is highest, so even although

the water volumes required are low relative to the demand for coal-fired power stations, this

demand could be important in the local context.


FIGURE 50: WATER DEMAND PER COALFIELD FOR ELECTRICITY GENERATION (LAGS BEHIND)

FIGURE 51: WATER DEMAND PER COALFIELD FOR ELECTRICITY GENERATION (LOW CARBON WORLD)

FIGURE 52: WATER DEMAND PER COALFIELD FOR ELECTRICITY GENERATION (MORE OF THE SAME)

FIGURE 53: WATER DEMAND PER COALFIELD FOR ELECTRICITY GENERATION (AT THE FOREFRONT)

0

50

100

150

200

250

300

350

2010 2015 2020 2025 2030 2035 2040

To

tal w

ate

r d

em

an

d f

or

po

wer

sta

tio

ns [

Mm

3/a

]

Waterberg coalfield Mpumalanga coalfields Other coalfields Other parts of SA

0

50

100

150

200

250

300

350

2010 2015 2020 2025 2030 2035 2040

To

tal w

ate

r d

em

an

d f

or

po

wer

sta

tio

ns [

Mm

3/a

]


0

50

100

150

200

250

300

350

2010 2015 2020 2025 2030 2035 2040

To

tal w

ate

r d

em

an

d f

or

po

wer

sta

tio

ns [

Mm

3/a

]


0

50

100

150

200

250

300

350

2010 2015 2020 2025 2030 2035 2040 To

tal w

ate

r d

em

an

d f

or

po

wer

sta

tio

ns [

Mm

3/a

] Waterberg coalfield Mpumalanga coalfields Other coalfields Other parts of SA


Water demand in CTL

Water demand for CTL is shown in Figure 54. Water demand remains constant at Secunda (Mpumalanga), whilst

the new CTL plants cause step increases in water demand in the Waterberg as they are brought on line (one

80,000 bbl/day plant in Lags Behind and two 80,000 bbl/day plants in More of the Same).

FIGURE 54: WATER DEMAND IN CTL

Total water demand

Total water demand for the Central Basin is shown in Figure 55 to Figure 58 and for the Waterberg is shown in

Figure 59 to Figure 62 for the different scenarios.

The water demand profiles in the Central Basin are similar for all scenarios, although the drop in demand occurs

earlier in At the Forefront and Low Carbon World than in the other two scenarios:

• Water demand for CTL stays constant at 88 Ml per annum (around a fifth of total water demand in the

Central Basin in 2010), increasing to 43% of total water demand in 2040, because of declining water

demand for electricity generation as coal-fired power stations are decommissioned.

• Water for mining contributes around 3% to the total water demand in 2010, and peaks in 2019, after

which it declines to around 22% of 2010 levels.

• Water demand for the coal-fired power stations dominates water demand in the Central Basin, but

declines steadily over the period as power stations are decommissioned, the rate of decline depending

on the particular scenario.

• Water demand is highest in More of the Same at the end of the period, due to later decommissioning of

power stations under this scenario. Similar, the demand in Low Carbon World is lowest by the end of

the analysis period due to more power stations having been taken off line.

Water demand in the Waterberg, however, is very different for the different scenarios:

• More of the Same and Lags Behind – water demand for electricity generation, mining and CTL ramp

up to give a total demand comparable to that required in the Central Basin in 2040, and just over half of

what is required in the Central Basin in 2010.

0

20

40

60

80

100

2010 2015 2020 2025 2030 2035 2040

Wate

r d

em

an

d [

Mm

3 p

er

an

nu

m]

Waterberg - More of the Same

Waterberg - Lags Behind

Waterberg - At the Forefront/Low Carbon World

Central Basin - All Scenarios


• At the Forefront – relatively low water demand is seen in the Waterberg, with no CTL demand. The

retrofit of FGD to Medupi causes a slight increase in demand but levels are very low compared to the

other scenarios.

• Low Carbon World – lowest water demand of all the scenarios.

When viewed together, the following are evident:

• The decline in water demand in the Central Basin is taken up by new water demand in Waterberg under

More of the Same and Low Carbon World, which is significant if transfers from the Upper Vaal are

contemplated for supplying the Waterberg.

• CCS increases water demand very considerably when applied to power stations, and from a water

demand view is far better suited to CTL. Nonetheless, electricity generation in Lags Behind still has

lower water intensity than More of the Same due to its more efficient ultra-supercritical power stations.

• Further solutions may well have to be found for supplying the Waterberg at the levels required in More

of the Same and Low Carbon World, including desalination, dry-cooling CCS, dry coal beneficiation

techniques etc.

FIGURE 55: WATER DEMAND IN CENTRAL BASIN

(LAGS BEHIND)


(LOW CARBON WORLD)


(MORE OF THE SAME)


(AT THE FOREFRONT)

0

50

100

150

200

250

300

350

400

450

2010 2015 2020 2025 2030 2035 2040

Wate

r d

em

an

d [

Mm

3/a

]

CTL Mining Power stations

0

50

100

150

200

250

300

350

400

450

2010 2015 2020 2025 2030 2035 2040

Wate

r d

em

an

d [

Mm

3/a


0

50

100

150

200

250

300

350

400

450

2010 2015 2020 2025 2030 2035 2040

Wate

r d

em

an

d [

Mm

3/a

]


0

50

100

150

200

250

300

350

400

450

2010 2015 2020 2025 2030 2035 2040

Wate

r d

em

an

d [

Mm

3/a



FIGURE 59: WATER DEMAND IN THE

WATERBERG (LAGS BEHIND)


WATERBERG (LOW CARBON WORLD)


WATERBERG (MORE OF THE SAME)


WATERBERG (AT THE FOREFRONT)

3.7 Infrastructure

3.7.1 Transport infrastructure

The development of appropriate transport infrastructure is at the centre of the scenarios – both to transport

export coal to the markets, and to move coal around South Africa for electricity supply and other applications.

With respect to transport infrastructure, consideration is given here to:

Port capacity and rail capacity on the rail line from the Central Basin to RBCT; and

Capacity on the Waterberg line to transport coal for use in Mpumalanga power stations, and for

exports via RBCT.

0

50

100

150

200

250

2010 2015 2020 2025 2030 2035 2040

Wate

r d

em

an

d [

Mm

3/a

]

CTL Mining Power stations Municipal

0

50

100

150

200

250

2010 2015 2020 2025 2030 2035 2040

Wate

r d

em

an

d [

Mm

3/a

]


0

50

100

150

200

250

2010 2015 2020 2025 2030 2035 2040

Wate

r d

em

an

d [

Mm

3/a

]


0

50

100

150

200

250

2010 2015 2020 2025 2030 2035 2040

Wate

r d

em

an

d [

Mm

3/a

]



Port capacity and rail capacity on the RBCT line

South Africa’s total exports were shown previously in Section 2.4. The plot shown in Figure 16 is repeated in

Figure 63, with the addition of lines indicating a possible evolution of the port capacity at RBCT and capacity on

the Central Basin to RBCT rail line. The latter two lines were developed based on the Transnet 2013 National

Infrastructure Plan, as well as personal communication with individuals at Transnet. The Transnet National

Infrastructure Plan suggests the TFR coal line from the Central Basin to RBCT will be expanded to 81 Mtpa by

2019, but could be further expanded to as high as 97.5 Mt per annum, post 2019. An expansion of RBCT from a

capacity of 91 Mtpa to 97.5 Mtpa to match that expansion would then also be required.

FIGURE 63: PLANNED RBCT PORT AND RAIL LINE EXPANSION WITH TOTAL COAL EXPORTS (5 YEAR

ROLLING AVERAGE)

The following important observations are highlighted from this plot:

Given that not all exports are transported through RBCT (with small amounts exported through Matola,

the Richards Bay dry bulk terminal and Durban), there is likely to be sufficient port capacity in South

Africa for exporting coal, even without the expansion of RBCT or Matola.

The challenge to South Africa in achieving the exports predicted under the four scenarios relates to

the on going limitation in provision of rail infrastructure to transport coal from the Central Basin to

RBCT, rather than port capacity. The results of the analysis suggest that even with Transnet’s planned

upgrade to this line, exports will still be constrained by rail infrastructure limitations at least till the early

2020’s. This matter needs to be urgently addressed.

There is, however, potential for stranded port and rail infrastructure post about 2032, unless existing

capacity is used for export of other bulk commodities or the potential of the Waterberg as a viable

source of exports is realised.

Within this context, it is critical that the justification for expansion of the port facility at Matola be fully

explored, unless this facility is to transport coal from other countries such as Botswana and

Mozambique.

It is reiterated that the coal export projections under the different scenarios are based on various assumptions

about allocation of coal to Eskom in the Central Basin, that future mines open on time and at the capacities

0

20

40

60

80

100

120

2010 2015 2020 2025 2030 2035 2040

To

tal exp

ort

s f

rom

So

uth

Afr

ica [

Mtp

a]



RBCT Capacity Capacity on RBCT rail line


captured in the models, that domestic non-Eskom demand remains strong and that no export only mines can

operate profitably in the Waterberg. If any of these assumptions do not hold, the above picture will change. In

particular, if the country wishes to continue to grow exports, the feasibility of establishing export only mines in the

Waterberg needs to be explored.

Coal transport on the Waterberg to Central Basin line

Coal potentially needs to be transported from the Waterberg to Central Basin for supply to Eskom power stations,

and for export coal to be transported onwards to RBCT.

Under the assumptions made in the SACRM scenarios, from a purely resource-supply point of view, coal is only

required to be transported from the Waterberg to the Central Basin power stations under More of the Same,

from after 2035, where 4 Mtpa will be required in 2035, rising to about 10 Mtpa in 2040. Under the other

scenarios, supply from the Central Basin is sufficient to sustain the existing power stations over their lifetimes.

However, supply options are very constrained from 2025, and managing to meet the demand of the Central

Basin power stations will require that all of the Central Basin mines are opened on time, to the capacities

specified in the SACRM study, and to the markets allocated (i.e. to supply Eskom and not to supply other

domestic users or exports). Already it appears that some of these projects are not on track to being operational

by the time coal is required by Eskom. The impact of just one mine diverting its coal from Eskom supply to

exports is explored in the sensitivity analysis in Section 4. Thus from a risk and cost point of view, provision

should be made for transport of coal from the Waterberg to Central Basin from the early 2020s in all scenarios

other than Low Carbon World.

Exports from the Waterberg under the four scenarios are shown in Figure 63 to Figure 66. Also shown in these

plots is the expansion on the Waterberg to Central Basin line, as proposed in the National Infrastructure Plan.

The additional 5-10 Mtpa coal required to be transported from the Waterberg to the Central Basin by 2025 in all

scenarios other than Low Carbon World should be kept in mind when interpreting these figures.

FIGURE 64: EXPORTS FROM WATERBERG - LAGS

BEHIND (5 YEAR ROLLING AVERAGE)

FIGURE 65: EXPORTS FROM THE WATERBERG –

LOW CARBON WORLD (5 YEAR

ROLLING AVERAGE)

0

5

10

15

20

25

30

35

40

2010 2015 2020 2025 2030 2035 2040

Mtp

a

Exports

Capacity on Waterberg to Central Basin rail line

0

5

10

15

20

25

30

35

40

2010 2015 2020 2025 2030 2035 2040

Mtp

a

Exports




MORE OF THE SAME (5 YEAR ROLLING

AVERAGE)


AT THE FOREFRONT (5 YEAR ROLLING

AVERAGE)

The need for new rail infrastructure from the Waterberg to support access to export markets thus differs between

the scenarios. Under More of the Same and Lags Behind, as new power stations are opened in the Waterberg,

the volume of exports that is produced in multiproduct mines will rise, with the new capacity being planned by

Transnet starting to be required by 2017. The currently planned capacity will be exceeded by the late 2020s, after

which new capacity is required to support exports. If, this infrastructure is also required to transport coal to

Central Basin power stations as seems likely, the capacity on this rail line will be exceed by 2025, and additional

upgrades on the RBCT line will be required earlier.

Under At the Forefront and Low Carbon World, export levels from the Waterberg are significantly lower,

reaching a peak of just over 10 Mt per annum under At the Forefront and then declining. As such, new

infrastructure built from the Waterberg to the Central Basin coalfields would need to be carefully considered, in

order to avoid it becoming a stranded asset. Under At the Forefront, the currently planned upgrade would be

required to transport coal from the Waterberg to the Central Basin power stations, as seems likely, from the

2020s, although the planned capacity of the line could be reduced. Under Low Carbon World the upgrade is not

required. However, under these two scenarios the development of export only mines in the Waterberg are

essential to maintaining South Africa’s coal export revenues and to avoiding port capacity becoming stranded.

Should export only mines become economically feasible in the Waterberg, then the upgrade on the Waterberg

rail line will be required to get this coal to market. Thus even in Low Carbon World there is a likelihood of the

upgrade still being required, but only from the 2030s.

In summary, what is clear from this analysis is that overall transport requirements differ substantially between

scenarios, and as the sensitivity analysis shows are highly assumption dependent. Early strategic planning is

thus required to ensure that there is not overinvestment in transport infrastructure. At the same time,

underinvestment should be avoided to avoid a situation which capacity requirements are quickly exceeded.

3.7.2 Water supply infrastructure and catchment management

All scenarios show comparable water demand in the Central Basin (see Figure 55 to Figure 58). Sufficient water

infrastructure is assumed to be in place to meet these demands, which increase to 2020, and decline thereafter.

0

5

10

15

20

25

30

35

40

2010 2015 2020 2025 2030 2035 2040

Mtp

a

Exports Capacity on Waterberg to Central Basin rail line

0

5

10

15

20

25

30

35

40

2010 2015 2020 2025 2030 2035 2040

Mtp

a

Exports



Having said this, it is recognised that the water catchments of the Central Basin are already under considerable

stress, as witnessed by the fact that water qualities in this area are too poor for power station use, and Kusile

power station will thus need to be supplied by transfers from the Upper Vaal52

.

Water demand in the Waterberg is considerable in More of the Same and Lags Behind (at about 200 Mm3/a by

2040), and less, although still appreciable, in At the Forefront and Low Carbon World (60 and 17 Mm3/a,

respectively by 2040). Expert opinion is that water will be able to be provided to the Waterberg, but that the issue

is rather one of infrastructure cost, and who will pay for the pipeline53

.

In planning for water supply to different areas of the country, consideration needs to be given to where the water

is to be sourced from. The intention, should the required pipelines be put into place, is to transfer water from the

Upper Vaal. In implementing water transfer projects, water resource planning needs to be accounted for. This

process attempts to balance the availability of water with the needs of the economy, whilst ensuring sufficient

remaining water flow for reasonably functional ecosystems. In South Africa this has been implemented through

the declaration of 19 Water Management Areas (WMAs), which generally correspond to catchment areas or sub-

sections of river basins, each of which need to maintain a water balance that allows for an acceptable ecological

reserve54

.

The Upper Vaal WMA is one of the most important from a coal industry perspective, providing water to the

Central Basin coalfields and power stations and Sasol facilities, and represents a potential source of additional

water for the Waterberg area. After accounting for the ecological reserve, as of 2000 the Upper Vaal was only

marginally within the desired range of positive water balance, with 19 Mm3/a still available after withdrawals of

1,045 Mm3/a. This situation is only made possible by artificial water transfers of 1,311 Mm

3/a from other

catchment areas, an amount which exceeds the local water yield in the Upper Vaal WMA55

. For comparison,

water requirements for development of the Waterberg coalfields as contemplated in More of the Same will

involve a growth in annual water demand of up to 130 Mm3/a by 2040, although annual water demand in the

Central Basin would decline by around 176 Mm3/a by this time from 2010 levels.

Lesotho and the Thukela WMA in KwaZulu-Natal are the main sources for interbasin transfers supplying water to

the Upper Vaal56

. In 2000 the Thukela WMA was already extracting water above the ecologically desirable level

in all four of its sub-catchment areas, in large part due to the artificial transfer of water to other areas57

. The

Lesotho Highlands Water Project, on the other hand, draws from the upper reaches of the Orange River

catchment58

where water is relatively abundant. However, any additional withdrawals from the Orange / Vaal river

system, including from Lesotho, would reduce the river flow through the extremely water-scarce Lower Orange

WMA in the Northern Cape, impacting on a water resource that is important to the economy of that region as well

as the ecologically important Orange River mouth59

.

Solar energy generation will require water resources in the area of Upington, which falls within the Lower Orange

WMA. This WMA was fully exploited in 200057

and additional withdrawals in this area would therefore also be

expected to reduce flows in the Orange River.

52 Yolandi Groenewald (2012) Coal’s Hidden Water Cost to South Africa, Greenpeace Africa 53 SACRM Expert Group water discussion, June 2012 54 DWAF (2004) National Water Resource Strategy. Department of Water Affairs and Forestry 55 DWAF (2004) National Water Resource Strategy. Department of Water Affairs and Forestry 56 DWAF (2004) National Water Resource Strategy. Department of Water Affairs and Forestry 57 DWAF (2004) National Water Resource Strategy. Department of Water Affairs and Forestry 58 Middleton, B.J. and Bailey, A.K. (2005) TT 382/08 Water Resources of South Africa (WR2005): Book of Maps. Water Research Commission 59 DWAF (undated) Environmental Issues. Available at: http://www.dwaf.gov.za/orange/default.aspx, accessed August 2012


Pipeline costs

The planning for Phase 2 of Mokolo and Crocodile Water Augmentation Project, being that which will deliver the

interbasin transfers described above to the Waterberg, provides an indication of the costs of building water

pipelines. This work suggests that the cost of a 42 Mm3/a pipeline would be to the order of R 11.42 billion, while

that of a 90 Mm3/a pipeline would be R 13.94 billion (values in real 2012 Rands).

3.8 Greenhouse gas (GHG) emissions

Greenhouse gas emissions from the use of coal in South Africa at present make up approximately 60% of a total

of 510 – 520 Mt per annum as assessed in 200960

. The great majority (229 Mt per annum) is from coal-fired

power generation, followed by about 47 Mt per annum from CTL. The direct emissions from coal mining itself are

much lower, estimated at 4.2 Mt per annum61

. With possible future constraints on the amount of GHG that can be

emitted, the estimation of future GHG emissions from possible coal futures is important.

3.8.1 Assumptions relating to GHG emissions

In the model of the coal value chain developed for the SACRM, of interest is both the total greenhouse gas

(GHG) emissions arising from the value chain as well as the GHG intensity of electricity supply (expressed in

kg CO2e/MWh SO). Quantitative estimates of GHG emissions are limited in this study to CO2 emissions (and do

not include other greenhouse gases such as methane). Emissions of CO2 are calculated by multiplying carbon

dioxide intensity factors by the relevant activity data in the models.

Mining and beneficiation

No GHG emissions are considered from the mining stage in the models as these are small relative to the

remainder of the value chain, with direct emissions from coal mining estimated to account for less than 2% of the

total GHG emissions of the coal value chain. Having said this, methane emissions account for the majority of the

GHG emissions of an underground mine (around 50 – 60%), with the remainder being made up of GHG

emissions associated with electricity used in mining (indirect GHG emissions), and those from burning liquid

fuels. On opencast mines, however, methane emissions account for only a small portion of the total GHG

emissions (less than 2%), with emissions associated with electricity provision or liquid fuel use accounting for the

majority of GHG emissions (depending on the mining method employed).

The other GHG emissions associated with mining are those from spontaneous combustion (commonly referred to

as “sponcom” in the industry). Where sponcom is occurring, this typically dominates the GHG emissions from the

mine. However, sponcom is very difficult to quantify reliably, with estimates based on expensive thermal imaging

techniques or on quantifying the reserves lost. In South Africa, the most severe sponcom problems are

experienced in surface mines in areas previously mined by bord and pillar methods62

. Thus a number of strip

mines in the Witbank coalfield have considerable GHG emissions from sponcom. This is a particular problem for

large operations that rely on blasting multiple pillars simultaneously to achieve high outputs. Small-scale

operations of previously mined areas can avoid the problem by excavating pillars one at a time. The other control

60 Jongi Witi, DEA (2012) Personal communication 61 Resarch undertaken for Coaltech Research Association. J Beukes (2012) Personal communication 62 Coaltech (2011) Prevention and Control of Spontaneous Combustion: Best Practice Guidelines for Surface Coal Mines in South Africa. Available online:

www.coaltech.co.za, accessed August 2012


mechanism is to put sufficient buffers (i.e. un-mined areas) in place, although this represents a trade-off with loss

of reserves. Many strip mines also experience severe spontaneous combustion in the spoil heaps.

Having said this, the problems with spontaneous combustion of surface dumps are reported to have largely been

brought under control and can be prevented through good management. Fires in underground mines are also

now relatively rare.

Electricity generation

Table 26 presents the carbon dioxide intensity factors and assumptions used to estimate GHG emissions from

electricity generation. Emission factors for the existing Eskom power stations for 2010 were used so as to ensure

alignment with Eskom’s reported annual CO2 emissions. For future build, CO2 emission factors were taken from

the supporting technical document for the IRP63

.

Where CCS is employed (either as a retrofit to Medupi and Kusile or with new coal build), 90% of the CO2

emissions are assumed to be captured. However, as indicated earlier, since CCS substantially affects the net

plant efficiency, the CO2 emission intensities for PF stations with CCS are around 86% less than those of the

corresponding power stations without CCS (see Table 13). The emission factors for PF stations with CCS were

based on those of the corresponding plant without CCS, but adjusted for the increase in auxiliary power required

to run the CCS process, and also for the increase in coal burn to produce the same electrical output due to the

drop in thermal efficiency.

63 EPRI (2010) Power Generation Technology Data for Integrated Resource Plan of South Africa


TABLE 26: CO2 EMISSIONS FACTORS IN ELECTRICITY GENERATION


Fossil fuels

Existing PF power stations 1.07 – 1.13 kg/kWhSO Values for existing power stations

chosen so as to get Eskom’s overall CO2 figure quoted in the 2012 Divisional Report (231.9 Mt)

Eskom

New coal: supercritical PF, dry cooled

0.924 kg/kWhSO Value for Medupi until FGD is retrofitted

EPRI (2010)


0.936 kg/kWhSO Value for Kusile, Medupi once FGD is

retrofitted, as well as future supercritical build with FGD


0.822 kg/kWhSO CO2 intensity as for supercritical PF,

but adjusted for decrease in coal burnt as a result of 5.3% increase in thermal efficiency between the two.

Based on EPRI (2010)

New coal: supercritical PF, dry cooled with FGD and CCS

0.142 kg/kWhSO Value for Medupi and Kusile under

Low Carbon World scenario where

CCS is retrofitted from 2029. Calculated taking into account drop in net plant efficiency and 90% CO2 removal.


New coal: ultra-supercritical PF, dry cooled with FGD and CCS

0.106 kg/kWhSO Calculated taking into account drop in

net plant efficiency and 90% CO2 removal.




0.997 kg/kWhSO EPRI (2010)

UCG-CCGT 0.857 kg/kWhSO Excludes UCG component EPRI (2010)

OCGT 0.622 kg/kWhSO EPRI (2010)

CCGT 0.376 kg/kWhSO EPRI (2010)

Renewables

Landfill gas and small scale hydro

0.161 kg/kWhSO

Coal-to-liquids

Sasol Secunda reported emissions of 47.2 Mt of CO2 for the production of 7.38 Mt of synfuels in 201064

. This

translates to an emission factor of 6.4 tonnes CO2/tonne synfuels. The same emission factor is assumed for a

new CTL plant, which results in around 23.6 Mt CO2 per annum for an 80,000 bbl/day plant.

In a CTL facility GHG emissions arise from the production of the synfuels product (process emissions), as well as

from burning coal (or gas) for steam and electricity generation (utility emissions). The process emissions are

produced in a concentrated form and are directly suitable for capture via CCS, whilst the utility emissions are in a

relatively dilute form, and their capture is energy and water intensive. It is thus assumed that, both for retrofit to

Secunda, and also in a new CTL facility, only 50% of CO2 emissions are suitable for capture via CCS65

. For the

purposes of modelling, it is thus assumed that the process emissions, which are those more readily captured,

account for 50% of the total CO2 emissions from CTL facilities and 50% of the total CO2 emissions are power

generation emissions which are not captured. The CO2 emission factor for CTL plants with CCS applied in the

64 Sasol (2011) Integrated Report. Available online http://www.sasol.com, accessed August 2012. 65 Assumption based on DEA (Department of Environmental Affairs) (2008) Long Term Mitigation Scenarios.

http://www.environment.gov.za/hotissues/2008/ltms/ltms.html, accessed July 2012.


models is thus half that of a plant without CCS (i.e. 3.2 tonnes CO2/tonne synfuels). Retrofitting CCS to Sasol

Secunda which produces 160,000 bbl/day would ultimately result in 26 Mt CO2 emitted per year (23.6 Mt CO2

from power generation and 2.4 Mt CO2 process emissions, as CCS only captures 90% of the process emissions,

as stated previously). For a future 80,000 bbl/day plant with CCS the annual CO2 emissions with CSS are

ultimately 13 Mt CO2 per year (11.8 Mt CO2 from power generation and 1.2 Mt CO2 process emissions).

3.8.2 Results and analysis: Greenhouse Gas Emissions

Mining

GHG emissions from mining are only inferred from the available quantitative data because the mining operations

were not modelled at a sufficiently detailed level to allow a meaningful quantification of emissions.

Figure 68 and Figure 69 show the total extent of mining in the two major coalfields across the four scenarios.

Mining in the Central Basin declines in the period to 2040 in all scenarios, after a peak between 2015 and 2020 in

More of the Same, Lags Behind and At the Forefront, and a peak in 2015 in Low Carbon World. With this

relatively high level of mining in all scenarios, GHG emissions from spontaneous combustion in the Central Basin

surface mines are anticipated to remain a problem, and there might even be potential for the number of

incidences to increase towards the end of the period when the Central Basin coal reserves have been

significantly drawn down and there is temptation to reduce the buffers and mine out the pillars of previous

underground operations.

Mining in the Waterberg ramps up very substantially over the period to 2040 in More of the Same and Lags

Behind. Waterberg mines are more likely to be surface operations, with the nature of the resource such that

underground mines are likely to be difficult and expensive. Coal-bed methane emissions are thus anticipated to

be low. Furthermore, the Waterberg surface mines do not have the potential for spontaneous combustion that the

Central Basin mines have (where this is primarily caused by intersecting old underground workings). However,

experience at the only operating Waterberg mine has shown Waterberg mines could be prone to spontaneous

combustion in the spoil heaps. Furthermore the low yields in the Waterberg (only around 50%) lead to high

discard volumes and extensive waste dumps that could also pose a spontaneous combustion risk.

FIGURE 68: TOTAL ROM COAL MINED IN THE

CENTRAL BASIN

FIGURE 69: TOTAL ROM COAL MINED IN THE

WATERBERG

0

50

100

150

200

250

300

350

400

2010 2015 2020 2025 2030 2035 2040

Mtp

a



0

100

200

300

400

500

600

2010 2015 2020 2025 2030 2035 2040

Mtp

a





Annual greenhouse gas emissions from power generation are shown in Figure 70, while the GHG intensity of

electricity supply is shown in Figure 71. The scenarios in which coal-fired power generation dominates (More of

the Same and Lags Behind) are, as expected, more CO2 intensive than At the Forefront and Low Carbon

World – with the former diversifying electricity supply and the latter building no further coal-fired power stations.

The impact of installing CCS on power stations built from 2034 under Lags Behind and retrofitting Medupi and

Kusile in Low Carbon World in helping to reduce emissions can clearly be seen.

South Africa’s National Climate Change Response White Paper presents a stated intention to reduce national

greenhouse gas emissions by 2050. The White Paper sets out a so-called benchmark National GHG Emissions

Trajectory Range, under which emissions peak in the period 2020 to 2025 in a range with a lower limit of

398 Mt CO2e and upper limits of 583 Mt CO2e and 614 Mt CO2e for 2020 and 2025 respectively. Emissions then

plateau for up to ten years after the peak within the range with a lower limit of 398 Mt CO2e and upper limit of

614 Mt CO2e. From 2036 onwards, emissions decline in absolute terms to a range with lower limit of

212 Mt CO2e and upper limit of 428 Mt CO2e by 2050. The extent to which the given levels of mitigation will be

achieved depends on support from developed countries.

There is no resolution as to the allocation of emissions to the electricity generation sector under this trajectory,

although at present electricity emissions are approximately half of total emissions. If the current contribution of

electricity generation to emissions is to remain constant, More of the Same and Lags Behind will contribute to

the country exceeding the upper limits of the electricity generation component of the trajectory after about 2025.

At the Forefront will fall within the trajectory to 2035, although will begin to exceed emissions during the decline

period post 2035. Only Low Carbon World affords the possibility of remaining within the proposed emissions

trajectory.

FIGURE 70: CO2 EMISSIONS FROM ELECTRICITY GENERATION

0

50

100

150

200

250

300

350

400

450

500

2010 2015 2020 2025 2030 2035 2040

CO

2 e

mis

sio

ns [

Mtp

a]



FIGURE 71: CO2 EMISSIONS INTENSITY FROM ELECTRICITY GENERATION

Greenhouse gas emissions from CTL

CO2 emissions from Sasol Secunda remain constant over the period to 2040 at 47.2 Mt CO2 per annum in More

of the Same and At the Forefront. The two additional CTL plants built in More of the Same add an additional

23.6 Mt CO2 per annum each, resulting in total GHG emissions from CTL of 94.4 Mt CO2 per annum by 2040.

In Lags Behind, from 2030, Secunda is fitted with CCS, so total GHG emissions from Secunda drop from

47.2 Mt CO2 to 26 Mt CO2 per annum by 2033. However, the single new CTL plant built in 2027 introduces

23.6 Mt CO2 by 2030, leading to a peak in CO2 emissions of 69.8 Mt CO2 in 2030. As this CTL plant is retrofitted

with CCS in 2033, the emissions from this CTL plant reduces to 13 Mt CO2, leading to a drop in total emissions

from CTL to 38.9 Mt CO2 from 2034 to 2040.

In Low Carbon World, no additional CTL plants are built, whilst the CCS retrofit to Secunda decreases

emissions from CTL to 26 Mt CO2 per annum by 2033.

3.9 Environmental implications

The environmental implications of the four scenarios are considered in terms of water supply (as well as land and

biodiversity) considerations, solid waste generation and other emissions.

3.9.1 Water provision, land and biodiversity

Water supply, land and biodiversity are discussed together here because they are interconnected, with impacts

on one inevitably leading to impacts on the others. Water demand is discussed in Section 3.6.6. The discussion

here focuses on the ability of the land to provide the resources needed to meet the demand. The complexity of

these topics, with a large number of factors playing a role, means that they are dealt with qualitatively, and the

broad implications of the scenarios discussed.

0

200

400

600

800

1000

1200

2010 2015 2020 2025 2030 2035 2040

CO

2 in

ten

sit

y o

f ele

ctr

icit

y s

up

ply

[kg

/M

Wh

SO

]



Water provision, land and biodiversity in mining

The impact of mining on water catchments is well documented66

, with the Olifants Catchment clearly showing the

historic impacts of mining, with a large proportion of the degradation attributable to coal mines. Coal mines have

the potential to affect water catchments through both reducing the volume of water available, as well as

degrading the quality of the water. The former is partly due to the mine’s own water requirements, primarily for

dust suppression and in coal washing (requiring about 150 litres/tonne coal washed), but more significantly due

to the amount of water extracted by the mine when aquifers are intersected (e.g. seepage into underground

mines and opencast pits), and water that needs to be contained once it has come into contact with coal discard

dumps, spoil heaps and stockpiles. Thus coal mines frequently have excess poor quality water that needs to be

disposed of in some way, although whether mines are a net consumer or generator of water depends on the

particular geological setting of the mine.

The other considerable impact coal mines have on water catchment is degradation through acid mine drainage

(AMD), i.e. water runoff with elevated sulphates and heavy metals caused when water comes into contact with

pyrite, present in coal, spoils, discards etc. The Water for Growth and Development Framework identifies AMD

as the most important threat to water quality in South Africa67

. Whilst AMD can be partially controlled through

careful water management on the mines (e.g. catching and containing runoff from stockpiles and dumps), it is

difficult to control in surface coal mines because of the extensive surface disruption and the destruction of

regolith layers that naturally regulate the surface-groundwater interface68

. The eMalahleni water treatment plant

has shown that solutions are possible, and a further five treatment plants are planned for the Central Basin.

Problems to be overcome include getting cooperation between the parties, high treatment costs, and the

problems of brine disposal (currently primarily done in evaporation ponds). The considerable effect AMD has had

on the Olifants catchment is clearly evident in that the water is now too polluted for industrial use, with Eskom

needing to transfer water from other catchments to supply certain of its power stations (most notably the new

Kusile Power station).

Land use is strongly connected to water catchment, and in South Africa only 4% of the land area comprises high

water yield areas. This is particularly notable for the grasslands in which large portions of the Central Basin

coalfields fall, where 23% of the area is high water yielding. Furthermore, the headwaters of some of South

Africa’s most economically important river systems are fed from this area; these rivers subsequently supply the

Midmar, Zaaihoek, Heyshope and the Vaal dams – all of which are vital to the water security for Gauteng and

Durban69

. The competition between land for water security and coal mining is therefore likely to increasingly

become an issue.

Land disruption is frequently used as a proxy for biodiversity impact. The grasslands biome in which the Central

Basin coalfields sit is well recognised for its biodiversity, with large portions of it under threat70

. The Waterberg

falls primarily under sour bushveld habitat, an area of high biodiversity and intact habitats, in recognition of which

it was been declared a UNESCO Biosphere Reserve in 2001. The impact on biodiversity in the Waterberg has

the potential to be as extensive as it has been in the grasslands of the Central Basin coalfields. This is because

of the potentially massive surface disruption of surface mining in the Waterberg, coupled with the very extensive

area required for waste dumps (due to the low yield and attendant high level of discard production) and the

66 WWF (2011) Coal and Water Futures in South Africa. 67 Dept of Water Affairs and Forestry (2009) Water For Growth and Development Framework 68 WWF (2011) Coal and Water Futures in South Africa 69 WWF (2011) Coal and Water Futures in South Africa 70 SANBI website (2012). Available online: http://Sanbi.org, accessed August 2012


consequent land sterilisation this implies. For example, the currently operating Waterberg mine covers an area of

some 2,200 Ha.

Water provision, land and biodiversity in electricity generation

Coal-fired electricity generation does not have the scale of disruption on water catchments as mining does,

although ash dumps and dams can have similar issues to AMD (high salinity seepage). As the power stations will

be in the same area as the mines (Central Basin or the Waterberg), much of the above discussion applies to

power station land use as well.

Land occupation by the power stations is fairly extensive (300 – 600 Ha), with much of that due to the land

requirements of dumps. Where ash can be backfilled into mining pits, this will decrease the overall land

disruption of the power station-mine complex. The addition of flue gas desulphurisation to coal-fired power

stations will increase the land required for solid waste dumps, assuming this will be produced in excess of what

can be sold. In particular, if less reactive limestone is used (which might be the case in South Africa, where FGD

is yet to be demonstrated on the very large scale’s required by Eskom) the solid waste volumes will be more

extensive that that currently predicted in the scenario models. Gypsum dumps have similar problems of high

salinity seepage and possible impact on water quality in the area.

Renewables also have extensive land footprints. A 100 MW solar installation (either parabolic trough or central

receiver) is estimated to occupy a similar land area to a 1000+ MW coal-fired power station (350 Ha). The land

requirements of wind are even more extensive, with a 200 MW installation estimated to occupy 2000 Ha.

However, even though they have high land footprints, the renewables are not anticipated to have as significant

impacts on water catchments as power stations and coal mines do. Furthermore, wind land requirements can

potentially be balanced with other land requirements, for example, cattle grazing. Nonetheless, depending on the

manner of their construction, solar and wind installations could potentially have negative impacts on biodiversity.

But if care is taken during construction, they might potentially have positive biodiversity effects, e.g. if land

beneath wind turbines is kept in a natural state with service corridors rather than converted to grassland.

Nuclear power stations have similar land footprints to coal-fired power stations (520 Ha for a 1,600 MW plant),

although these can be much larger if one includes the land requirements of nuclear waste disposal and the

exclusion zones required around such sites. However, as with solar and wind, they are associated with less

surface disruption and land sterilisation and thus have lower impacts on water and biodiversity than coal-fired

power stations. They may well have positive biodiversity impacts by reserves being created in the exclusion

zones around nuclear power stations and nuclear waste disposal sites (although with the very low probability of

very considerable impacts should an incident occur).

Results and analysis: Water provision, land and biodiversity

Mining

Under all scenarios the water catchments of the Central Basin are anticipated to come under increasing stress.

Water demand peaks in around 2020 and then decreases to about one quarter to one third of current levels by

2040, depending on the scenario. Water stress was already experienced in the Olifants, Inkomati, and Thukela

water management areas in 200071

, whilst the Upper Vaal was operating only marginally within its ecological

71 Van Rooyen, J.A. and Versfeld, D.B. (2009) Strategic planning for water resources in South Africa. A situation analysis 2009. report No. P RSA

000/00/7809, Department of Water and Environment Affairs, Pretoria


reserve (sustained only by inter-basin transfers). The increase in water demand predicted by the scenarios is

probably sufficient to place these catchments further under stress, but especially when coupled with the

anticipated industrial growth in these areas, the potential degradation in supply due to the increased levels of

mining, and the potential impacts of climate change on water availability.

Large-scale development of the Waterberg, as occurs in More of the Same and Lags Behind, is associated with

considerable water demand (mines, power stations, industry (CTL) and communities) and will require

considerable infrastructure (see Section 3.7). To meet demand in this water-scarce area will require substantial

inter-basin transfers, most likely from the Upper Vaal, so the question becomes more one of cost and

infrastructure than water provision in the Waterberg. Nonetheless, considerable water stress will likely be felt in

the area, with dams and construction disrupting river flow, as is already being felt with the construction of Medupi

power station72

.

Land and biodiversity impacts follow the trends of water catchment stress, with extensive mining in More of the

Same and Lags Behind likely to cause significant biodiversity impacts in the Grasslands and Bushveld biomes.

These impacts are softened in At the Forefront because of the decline in coal mining for domestic supply, whilst

Low Carbon World has the lowest biodiversity impacts, and no impact on the Bushveld (the Waterberg is not

developed in this scenario).


The high-coal scenarios (More of the Same and Lags Behind) follow similar trends as discussed for mining,

where the power stations essentially follow the mines in terms of location and impact, although they are attributed

a smaller share of the land and water catchment impact than the mines.

The scenarios with high renewables content (At the Forefront and Low Carbon World) have high land

requirements. Nearly 1,500 km2 is required for the wind power built under At the Forefront, and just under

500 km2 is required for the solar power build (compared to around 40 km

2 required for the new coal build under

More of the Same). Nonetheless, as discussed above, the impacts of this land use are very different, and At the

Forefront and Low Carbon World are anticipated to have lower water provision impacts than More of the

Same and Lags Behind, especially Low Carbon World, where the Waterberg is not developed further.

However, finding suitable sites for the number of new nuclear power station units required under At the

Forefront and Low Carbon World is likely to be problematic. There is significant public resistance to nuclear

power stations being located close to residential areas, as has already been found by Eskom during site

selection studies for a new nuclear station. There is some similar resistance to wind farms due to the visual

impacts.

3.9.2 Solid waste generation

Tonnages of solid wastes produced from mining and electricity generation are calculated in the model, using

yield factors for mining and ash intensity factors for power generation.

Solid wastes from mining and beneficiation

Spoils and discards produced during mining and beneficiation are the major sources of solid wastes on mines.

The former is only a consideration in surface mines, whilst the latter are a by-product of coal washing. Discard is

72 Groenewald, Y. (2012) Coal’s Hidden Water Cost to South Africa, Greenpeace Africa


quantified in the models based on washing yields. Product yields for existing mines and projects were taken from

Wood Mackenzie coal supply service data, whilst for projects that were not well specified, indicative yields were

assumed based on the quality of the coal resource.

Solid waste from electricity generation

Table 27 presents waste intensity factors and assumptions used to estimate total waste tonnages from electricity

generation. Ash factors for the existing Eskom power stations were applied so as to get Eskom’s reported total

annual mass of ash produced. For future build, waste emission factors were taken from the supporting technical

document for the IRP73

. The waste intensity factors for PF stations with CCS are based on those of the

corresponding plant without CCS, but adjusted for the increase in auxiliary power required to run the CCS

process, and also for the increase in coal burn to produce the same electrical output due to the drop in thermal

efficiency.

TABLE 27: WASTE GENERATION FACTORS IN ELECTRICITY GENERATION


Fossil fuels

Existing PF power stations Ash: 140 – 280 kg/MWhSO Values for existing power stations

chosen so as to get Eskom’s overall

ash figure quoted in the 2012 Divisional Report (36.21 Mt)

Eskom


170 kg/MWhSO Value for Medupi until FGD added in 2021

EPRI (2010)


Ash: 172

FGD: 24

kg/MWhSO Value for Kusile and Medupi once FGD is added in 2021, as well as future supercritical build with FGD


Ash: 151

FGD: 21

kg/MWhSO Intensity as for supercritical PF, but adjusted for decrease in coal burnt as

a result of 5.3% increase in thermal efficiency between the two.


New coal: supercritical PF,

dry cooled with FGD and CCS

Ash: 260

FGD: 36

kg/MWhSO Calculated taking into account drop in

net plant efficiency. Applies to Medupi and Kusile after 2029 in Low Carbon World.


New coal: ultra-supercritical

PF, dry cooled with FGD and CCS

Ash: 195

FGD: 27

kg/MWhSO Calculated taking into account drop in

net plant efficiency. Applies to new PF

stations build after 2034 in Lags Behind.




Ash/’FGD”: 239

(includes solids from in-situ

desulphurisation)

kg/MWhSO EPRI (2010)

UCG-CCGT 0 kg/MWhSO Excludes UCG component EPRI (2010)

Renewables

Co-generation 30 kg/MWhSO Municipal solid waste (MSW) or forestry residue

EPRI (2010)



Solid wastes from coal-to-liquids

Sasol Secunda (160,000 bbl/day) reported 1.45 tonnes of ash per tonne synfuels produced74

, giving a total of

10 Mt for 2011. The same waste factor is assumed for a new 80,000 bbl/day plant, which translates to an annual

solid waste production of 5.4 Mt. As discussed above for the GHG intensity, it is assumed that adding CCS to the

plant will not appreciably increase the amount of solid waste generated.

Results and analysis: solid waste generation

Solid wastes from mining

Cumulative discard volumes produced over the time period in the Central Basin are shown in Figure 72. This is

the total discard produced by the mines operating in the Central Basin less the discard burnt in FBC to 2040

(ranging from 216 Mt in More of the Same to 389 Mt in At the Forefront). Solid waste generation from the

Central Basin coalfields is much the same for each scenario. Low Carbon World has the highest cumulative

discards in 2040 (at 1.8 billion tonnes) as it produces more for export towards 2040, so has higher beneficiation

losses than the other scenarios where mines are producing for Eskom at high yields. At the Forefront has the

lowest cumulative discards in 2040 (at 1.5 billion tonnes) as it has the highest installed capacity of FBC power

stations (and so the highest use of discard). Although mining steadily declines after 2020, the fact that the lower

quality resources are increasingly mined from around 2025 means that discard volumes do not decline

proportionally.

In More of the Same and Lags Behind, the decline in mining in the Central Basin is countered by steadily

increasing mining in the Waterberg (see Figure 69).

Figure 73 gives the total mass of discards generated in the Waterberg in the different scenarios. The low yields of

the Waterberg coal resource means that these are very extensive in the scenarios where development of the

Waterberg is extensive (More of the Same and Lags Behind). These scenarios generate just under 3,500 Mt of

discards, which need to be contained in some way. Whilst there will likely be some backfilling of mining pits,

these huge tonnages indicate massive surface disruption, and if not very carefully managed, potentially very

significant environmental impacts (such as emissions of greenhouse and toxic gases through spontaneous

combustion). These tonnages also hint at the need for much larger uptake of fluidised bed combustion (FBC) in

future power stations.

At the Forefront and Low Carbon World have much lower discard production (between 1,000 and 1,500 Mt),

showing the far lower levels of Waterberg development in these scenarios. Low Carbon World sees no future

mine development in the Waterberg beyond that needed to supply Matimba and Medupi power stations.

The Waterberg coal deposits are anticipated to be better suited to surface mining techniques, thus in addition to

the high volume of discard, spoil heaps are potentially also a significant source of mining wastes.

74 Sasol (2011) Sasol Integrated Report 2011. Available online: http://www.sasol.com/sasol_internet/frontend/navigation.jsp?navid=21100001&rootid=3; 18

June 2012. Pg 83, accessed August 2012


FIGURE 72: CUMULATIVE DISCARD GENERATION IN THE CENTRAL BASIN (DISCARD PRODUCED

LESS DISCARD BURNED IN FBC)

FIGURE 73: CUMULATIVE DISCARD GENERATION IN THE WATERBERG

Solid wastes from electricity generation

Solid wastes from electricity generation fall into four categories:

• Ash (fly ash and bottom ash from PF and FBC)

• Gypsum waste from flue gas desulphurisation (FGD)

• High level nuclear waste (spent fuel)

• Low/intermediate level nuclear waste

Plots of solid waste generation under the four scenarios are shown in Figure 74 to Figure 77. Note that these

figures are for cumulative wastes to 2040, and the graphs have different scales on the y-axes. Ash from coal-

fired power stations is produced in the largest quantities, with over 1.6 billion tonnes having been produced

0

500

1,000

1,500

2,000

2010 2015 2020 2025 2030 2035 2040

Dis

card

[M

t]


0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

2010 2015 2020 2025 2030 2035 2040

Dis

card

[M

t]



between 2010 and 2040 in More of the Same and Lags Behind. Ash volumes are considerable as the new-

build power stations are assumed to burn poorer quality coals with high ash contents. FGD waste production

under these scenarios is also high, although this will present less of a management issue if a market for these

products can be found. Markets for fly ash in South Africa have historically been small (around 6%) due to

concerns of heavy metal contamination. The management and containment of solid wastes are thus an important

consideration in More of the Same and Lags Behind, with a consequent increase in land footprint of the power

station (unless backfilling of surface mines occurs), as well as possible high salinity effluent issues (especially

with the wet FGD waste).

Nuclear waste generation under Low Carbon World, as well as At the Forefront, is also of concern not due to

the volumes but rather due to the management requirements for high and medium level nuclear wastes. Secure

waste disposal sites will need to be identified as, at present, all high-level nuclear waste is stored at Koeberg (i.e.

SA does not have a dedicated high-level nuclear waste storage facility).


FIGURE 74: CUMULATIVE ASH GENERATION FIGURE 75: CUMULATIVE FGD WASTE GENERATION

FIGURE 76: CUMULATIVE HIGH LEVEL NUCLEAR WASTE GENERATION FIGURE 77: CUMULATIVE LOW/INTERMEDIATE LEVEL WASTE GENERATION

0

200

400

600

800

1,000

1,200

1,400

1,600

1,800

2010 2015 2020 2025 2030 2035 2040

Cu

mu

lati

ve w

aste

[M

t]


0

10

20

30

40

50

60

70

80

90

2010 2015 2020 2025 2030 2035 2040

Cu

mu

lati

ve w

aste

[M

t]


0

2,000

4,000

6,000

8,000

10,000

12,000

2010 2015 2020 2025 2030 2035 2040

Cu

mu

lati

ve w

aste

[t]


0

5

10

15

20

25

30

35

40

2010 2015 2020 2025 2030 2035 2040

Cu

mu

lati

ve w

aste

[M

l]



3.9.3 Non-GHG emissions

Emissions of SO2, NOx and particulates are also calculated in the model, so as to get a wider feel for the

potential environmental impacts of the coal value chain under the different scenarios. As with GHG emissions,

these are expressed on an intensity basis (e.g. kg/MWh sent out), with the emissions calculated by multiplying

the emission intensity factors by the relevant activity data in the models.

Mining and beneficiation

In terms of non-GHG emissions from mining, dust is a significant potential impact, but is difficult to quantify and

generation is highly site specific. The models do not go to the level of the mining technology employed at the

different mines, which would be needed to be able to estimate the emission profiles of the mines.

As discussed above for GHG emissions, emissions from spontaneous combustion are also not considered in the

models. Apart from CO2, spontaneous combustion is associated with significant aerial pollutants, including CO,

H2S, SO2, NOx and particulates, and where it does occur, is easily the most visible and damaging local

environmental impact.


Table 30 presents the emission intensity factors and assumptions used to quantify the emissions of SO2, NOx

and particulates from electricity generation. Emission factors for the existing Eskom power stations were applied

so as to get Eskom’s reported annual emissions. For future build, emission factors were taken from the

supporting technical document for the IRP75

The emission factors for SO2 and particulate emissions from PF stations with CCS were based on those of the

corresponding plants without CCS, but adjusted for the increase in auxiliary power required to run the CCS

process, and also for the increase in coal burn to produce the same electrical output due to the drop in thermal

efficiency. NOx emissions are primarily a function of boiler conditions. NOx emission factors for PF plants with

CCS are therefore not affected by increase in coal burn, but just corrected for the drop in net plant efficiency.

TABLE 28: NON-GHG EMISSION FACTORS APPLIED IN ELECTRICITY GENERATION

Application SO2 NOx Particulates Units Comments Reference

Existing PF power stations

5.2 – 12 3.2 – 6.7 0.15 – 1.1 kg/MWhSO Values for existing power

stations chosen so as to get the overall figures quoted in Eskom’s 2012 Divisional Report

Eskom (2012)


8.93 2.26 0.12 kg/MWhSO Value for Medupi until FGD added.

EPRI (2010)


0.45 2.3 0.13 kg/MWhSO Value for Kusile, and for Medupi once FGD has

been retrofitted in 2021, as well as future supercritical build with FGD

New coal: ultra-

supercritical PF, dry

0.4 2.29 0.11 kg/MWhSO Emissions intensities as

for supercritical PF, but adjusted for decrease in

Based on

EPRI



Application SO2 NOx Particulates Units Comments Reference

cooled with FGD coal burnt as a result of 5.3 % increase in thermal

efficiency between the two.

(2010)

New coal:

supercritical PF, dry cooled with FGD and CCS

0.014 2.93 0.2 kg/MWhSO Calculated taking into

account drop in net plant efficiency. Applies to Medupi and Kusile in Low Carbon World.

Based on

EPRI (2010)

New coal: ultra-supercritical PF, dry

cooled with FGD and CCS

0.01 2.66 0.15 kg/MWhSO Calculated taking into account drop in net plant

efficiency. Applies to new PF installed after 2034 in Lags Behind.


Fluidised bed

combustion (FBC): dry cooled with sorbent injection

0.19 0.2 0.09 kg/MWhSO EPRI (2010)

UCG-CCGT 0.21 0.01 0 kg/MWhSO Excludes UCG component EPRI (2010)

OCGT 0 0.28 0 kg/MWhSO EPRI (2010)

CCGT 0 0.29 0 kg/MWhSO EPRI (2010)

Cogeneration 0.78 0.61 0.16 kg/MWhSO EPRI (2010)

Landfill gas and small scale hydropower

0 0.12 0.48 kg/MWhSO EPRI (2010)

Coal-to-liquids

Emissions of SO2, NOx and particulates are reported in Sasol’s Sustainable Development Report (2011). The

following emission factors are derived for the CTL process from the reported figures:

• 25 kg SO2 per tonne synfuels

• 20 kg NOx per tonne synfuels

• 1.3 kg particulates per tonne synfuels

The same emission factors are assumed for a new CTL plant, which, for a plant half the size of Secunda, results

in around 94 kt SO2, 75 kt NOx and 4.9 kt particulates emitted per year.

For the reasons discussed above, the addition of CCS to capture the high-concentration stream of CO2 is

assumed not to appreciably increase the emissions of SO2, NOx and particulates per tonne of synfuels.

Results and analysis: other emissions

Emissions from mining

Although not quantitatively addressed, the same trends for the scenarios can be inferred as for the emissions of

GHGs.

Emissions from electricity generation

Figure 78 to Figure 80 give the emissions of SO2, NOx and particulates for the four scenarios.


FIGURE 78: TOTAL SO2 EMISSIONS FROM POWER

GENERATION

FIGURE 79: TOTAL NOX EMISSIONS FROM POWER

GENERATION

FIGURE 80: TOTAL PARTICULATES FROM POWER

GENERATION

Key observations relating to the other emissions from electricity generation are as follows:

• Emissions of SO2 and particulates drop across all scenarios. When viewed on a coalfield basis it

is evident some of the decline is due to the decommissioning of the older Mpumalanga power

stations (see for example Figure 81 for More of the Same). In Low Carbon World the drop-off

occurs earlier with earlier retirement of power stations.

• The drop in SO2 is also a result of installation of flue gas desulphurisation on new-build power

stations (for example, in More of the Same, SO2 emissions in 2040 are about 60% of current

levels). This can also be seen in Figure 82, which shows the SO2 emission profile for the

Waterberg under More of the Same. SO2 emissions rise steadily as new power stations are

opened in this coalfield, and then a drop sharply when FGD is installed at Medupi power station.

• Particulates and NOx on the other hand show a steady rise as new power stations are opened in

the Waterberg coalfield (in comparing emissions between the Central Basin and the Waterberg,

note should be taken of the scales on the vertical axes).

• The fact that particulates decrease marginally between 2010 and 2040 indicates the improved

particulate collection efficiency of the new-build power stations. NOx on the other hand increases

in More of the Same and Lags Behind.

0

500

1,000

1,500

2,000

2,500

2010 2015 2020 2025 2030 2035 2040

SO

2 [kt/

a]



0

200

400

600

800

1,000

1,200

1,400

2010 2015 2020 2025 2030 2035 2040

NO

x [

kt/

a]



0

20

40

60

80

100

2010 2015 2020 2025 2030 2035 2040

Part

icu

late

s [

kt/

a]




• Particulates and NOx both decrease considerably in At the Forefront and Lags Behind,

showing the lower NOx and particulates emission rates of technologies employed in the

diversified electricity build plan.

FIGURE 81: NON GHG POWER STATION EMISSIONS IN CENTRAL BASIN UNDER MORE OF THE SAME

FIGURE 82: NON GHG POWER STATION EMISSIONS IN WATERBERG UNDER MORE OF THE SAME

4 SENSITIVITY ANALYSES

Sensitivity analyses were applied to further interrogate some of the assumptions made within the scenario

modelling. In the sensitivity analyses, consideration was given to changes to the electricity supply infrastructure

build plan, as well as to the impact on local utility coal supply of diverting the coal from one big mine in the

Central Basin from Eskom supply to export markets.

0

10

20

30

40

50

60

70

80

90

100

0

500

1,000

1,500

2,000

2,500

2010 2015 2020 2025 2030 2035 2040

Part

icu

late

s [

kt/

a]

SO

2 a

nd

NO

x [

kt/

a]

SO2 NOx Particulates

0

5

10

15

20

25

30

35

40

45

0

100

200

300

400

500

600

700

2010 2015 2020 2025 2030 2035 2040

Part

icu

late

s [

kt/

a]

SO

2 a

nd

NO

x [

kt/

a]

SO2 NOx Particulates


In terms of the electricity supply infrastructure build plan, the impacts of the following changes to the build plan

under At the Forefront were explored:

• The IRP 2010 policy adjusted scenario has an ambitious nuclear and renewables build

programme. The nuclear build, in particular, is considered by many to already be running behind

the proposed schedule in light of the numerous issues that require resolution, including issues of

costs and financing mechanisms, localisation of manufacture and possible uranium mining and

enrichment, safety and environmental implications and skills availability for construction and

maintenance. The National Planning Commission has suggested that if building the first nuclear

power stations as required by IRP 2010 is unachievable, gas-fired power generation be put

forward as a so-called “plan b”, with potential sources of gas including off-shore natural gas,

shale gas from the Karoo and liquefied natural gas (LNG) imports. The first sensitivity explored in

the SACRM models is thus to change the build plan under At the Forefront to replace the first

nuclear build with gas build in the form of CCGT that delivered the same GWhSO. The

remainder of the build plan remains the same

• Of relevance to the South African Coal Roadmap was a comparison of the above situation (i.e.

gas replacing the first nuclear build) with a new coal-fired power station replacing the first nuclear

build. For the second sensitivity, therefore, the build plan under At the Forefront was adjusted to

replace the first nuclear build with coal-fired power at the scale required to deliver the same

GWhSO. The remainder of the build plan remains the same.

• Demand side management can significantly influence the build plan required and thus the coal

demand. To investigate the effect of reduced demand, in the third sensitivity, annual electricity

demand was reduced by a total of 10% total between 2030 and 2040 against the demand

projections extrapolated from the IRP 2010 demand projections which were used in this study.

The build plan between 2030 and 2040 was adjusted to account for this lower demand.

The impact on the following parameters were explored in the sensitivity analysis, where relevant:

• Investment requirements for new power generation infrastructure;

• Electricity generation cost;

• Coal exports (as these are linked to coal demand for power generation);

• Greenhouse gas emissions; and

• Greenhouse gas emissions intensity of power generation.

In the sensitivity analysis relating to diversion of coal supply to Eskom, the implications of switching a large scale

mine delivering 5 Mtpa of high quality (22-24 MJ/kg) middlings coal to Eskom to a low-quality export mine, which

no longer produces a middlings product was explored. This sensitivity is applied to a project planned to

commence production in the early 2020s, so that the implications for the requirement for coal from the Waterberg

to feed Central Basin power stations can be explored.

4.1 Replacing the first nuclear plant with gas CCGT

In this sensitivity analysis the first 4,800 MW nuclear plant commissioned under the IRP 2010 between 2023 and

2025 is replaced with a 4,908 MW CCGT plant, this being the capacity required to replace the nuclear plant on a

sent out basis given capacity factors applied in the study (92% for nuclear and 90% for gas). Note that impact on

exports is not considered under this scenario, as there is no change to the build plan for coal-fired power stations

and hence exports remain unaffected.


4.1.1 Electricity supply infrastructure investment and electricity generation cost

A comparison between the investment required under At the Forefront, and that under which CCGT replaces

the first nuclear plant, is shown in Figure 83. The replacement of only one nuclear plant saves a total of

R 129.4 billion of the investment required for power station infrastructure (in 2010 Rands) over the 2010-2040

analysis period. Note once again that this excludes any provision for power stations built after 2040, explaining

the drop in investment post about 2035. There is, however, no discernable effect on the cost of generating

electricity despite the lower investment requirement, as shown in Figure 84.

FIGURE 83: CHANGE IN GENERATION

INFRASTRUCTURE INVESTMENT BY

REPLACING ONE NUCLEAR STATION

WITH GAS

FIGURE 84: CHANGE IN GENERATION COST BY


WITH GAS

4.1.2 Greenhouse gas emissions and emissions intensity

The use of gas in place of nuclear for the next base load power station causes an additional 236 Mt of

greenhouse gases to be emitted over the analysis period, due to electricity from gas having higher GHG

emissions than nuclear. The greenhouse gas emissions intensity of electricity supply is predicted to be about 6%

higher by 2040 if the next power station is gas-fired rather than nuclear (as in IRP 2010).

4.2 Replacing the first nuclear plant with a coal-fired power station

In this sensitivity analysis, the first 4,800 MW nuclear plant commissioned under the IRP 2010 between 2023 and

2025 is replaced with a 5,220 MW coal fired plant, this being the capacity required to replace the nuclear plant on

a sent out basis given the capacity factors applied in the study (92% for nuclear and 84.6% for coal PF).


A comparison between the investment required under At the Forefront, and that under which coal replaces the

first nuclear plant, is shown in Figure 85. The replacement of only one nuclear plant saves a total of R 37 billion

of the investment required for power station infrastructure (in 2010 Rands) over the 2010-2040 analysis period.

Note once again that this excludes any provision for power stations built after 2040, explaining the drop in

0

20,000

40,000

60,000

80,000

100,000

120,000

2010 2015 2020 2025 2030 2035 2040

Ele

ctr

icit

y b

uild

pla

n in

vestm

en

t p

er

year

(R M

illio

n)

At the Forefront

At the Forefront CCGT replaces nuclear

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

2010 2015 2020 2025 2030 2035 2040

Ele

ctr

icit

y g

en

era

tio

n c

ost

[R/k

Wh

sen

t o

ut]

At the Forefront



investment post about 2035. There is, however, no discernable effect on the cost of generating electricity despite

the lower investment requirement, as shown in Figure 86.

FIGURE 85: CHANGE IN GENERATION

INFRASTRUCTURE INVESTMENT BY


WITH COAL

FIGURE 86: CHANGE IN GENERATION COST BY


WITH COAL


The use of coal in place of nuclear for the next base load power station gives rise to an increase in emissions of

500 Mt over the analysis period, due to electricity from gas having higher emissions than nuclear. Notably,

building the next power station as coal-fired instead of nuclear results in an additional 27 to 32 Mtpa GHG

emissions on a year-by-year basis. This differential is substantial when considered in the light of South Africa’s

intention for emissions to peak, plateau and then decline. The greenhouse gas emissions intensity of electricity

supply is predicted to be about 12% higher by 2040 if the next power station is coal-fired rather than nuclear (as

in IRP 2010).

4.2.3 Impact on exports

The impact of building an additional coal fired power station on exports is an additional 4.2 Mtpa from 2025,

which is as a result of a new dual-producing coal mine in the Waterberg to supply the new power station. Thus,

exports are predicted to be 8% higher in 2040 if the next power station is coal-fired rather than nuclear (as in IRP

2010), resulting in increased export revenue of some R 5 billion per year.

4.3 Reduced electricity demand post 2030

Under this sensitivity analysis, the electricity demand between 2030 and 2040 was reduced each year, to end up

10% lower in 2040 than was assumed in At the Forefront. The resulting demand profile is shown in Figure 87.

Note that coal exports are not considered under this sensitivity analysis since the impact is negligible (less than

2 Mtpa difference in any one year).

0

20,000

40,000

60,000

80,000

100,000

120,000

2010 2015 2020 2025 2030 2035 2040

Ele

ctr

icit

y b

uild

pla

n in

vestm

en

t p

er

year

(R M

illio

n)

At the Forefront


0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

2010 2015 2020 2025 2030 2035 2040

Ele

ctr

icit

y g

en

era

tio

n c

ost

[R/k

Wh

s

en

t o

ut]

At the Forefront

At the Forefront PF replaces nuclear


FIGURE 87 ELECTRICITY DEMAND IN AT THE FOREFRONT AND ADJUSTED TO INVESTIGATE

SENSITIVITY TO LOWER DEMAND POST 2030


A comparison between the investment required under At the Forefront, and under a scenario in which demand

is reduced, is shown in Figure 88. Note that even though demand only begins to decline post-2030, the impact is

seen from before 2025, due to the long lead times for base load power stations. The reduction in demand of 10%

over a 10 year period saves a total of R 189 billion (or about 12%) of the investment required for power station

infrastructure (in 2010 Rands) over the 2010-2040 analysis period. Note once again that this excludes any

provision for power stations built after 2040, explaining the drop in investment post about 2035. This stresses the

significant importance of electricity demand side management in contributing to reducing the investment required

in new power stations. There is, however, little overall effect on the cost of generating electricity despite the lower

investment requirement, as shown in Figure 89.

FIGURE 88: INVESTMENT IN GENERATION

INFRASTRUCTURE UNDER THE IRP

2010 AND A REDUCED DEMAND

SCENARIO

FIGURE 89: ELECTRICITY GENERATION COST

UNDER THE IRP 2010 AND A REDUCED

DEMAND SCENARIO

0

100,000

200,000

300,000

400,000

500,000

600,000

2010 2015 2020 2025 2030 2035 2040

Ele

ctr

icit

y d

em

an

d G

Wh

SO

lower demand unchanged

-

20,000

40,000

60,000

80,000

100,000

120,000

2010 2015 2020 2025 2030 2035 2040

Ele

ctr

icit

y b

uild

pla

n in

vestm

en

t p

er

year

(R M

illio

n)

At the Forefront Lower Demand

At the Forefront

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

2010 2015 2020 2025 2030 2035 2040

Ele

ctr

icit

y g

en

era

tio

n c

ost

[R/k

Wh

sen

t o

ut]

At the Forefront Lower Demand

At the Forefront


The different between the build plans in At the Forefront and the scenario with lower demand is shown in Figure

90. Notable here is that the reduced demand avoids building of one full PF power station, and about one unit at a

nuclear power station.

FIGURE 90: NEW BUILD REQUIRED POST 2030 FOR AT THE FOREFRONT AND WITH LOWER

ELECTRICITY DEMAND POST 2030


The reduced demand profile gives an emissions savings of 136 Mt (or 2%) over the 10 year period from 2031-

2040. While this is arguably small, given it is over a 10 year period, it is also worth noting that in the year 2040

emissions are 12% or 30 Mt lower in the reduced demand case than in At the Forefront. Continued reduction in

electricity demand will further reduce the emissions. The GHG emissions intensity of electricity supply is also

about 2% lower in the reduced demand case.

4.4 Diversion of coal from Eskom to exports

As stated previously, in this sensitivity analysis the implications of switching a large scale mine delivering 5 Mtpa

of high quality (22-24 MJ/kg) middlings coal to Eskom to a low-quality export mine, which no longer produces a

middlings product was explored. This sensitivity is applied to a project planned to commence production in the

early 2020s, so that the impact on the requirement for coal from the Waterberg to feed Central Basin power

stations, as well as infrastructure needs on the Waterberg line, is explored.

4.4.1 Coal from the Waterberg to supply Central Basin power stations

In Section 2.3, it was identified that under the assumptions used in the models, coal would only be required to be

transported from the Waterberg for Central Basin power stations under More of the Same, starting from 2036.

The loss of Eskom supply in the Central Basin from just one mine, however, has a significant impact on high-

quality utility coal requirement from the Waterberg as follows:

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000 M

W n

ew

bu

ild

po

st

2030

At the Forefront Reduced demand


• More of the Same, Lags Behind and At the Forefront all require coal from the Waterberg for

Central Basin power stations from 2025 onwards, up to a level of just over 4 Mtpa.

• Demand for Waterberg coal in the Central Basin begins to grow again under More of the Same

from 2036, reaching a level of 13.2 Mtpa by 2040.

• The total coal required from the Waterberg for More of the Same under this sensitivity is 100 Mt.

Lags Behind and At the Forefront each require a total of 61.2 Mt.

• Low Carbon World never requires coal from the Waterberg for Central Basin power stations.

Clearly, if planned mining projects either don't commence operation on time as per the schedules used in the

scenarios, or supply at different product yields to those assumed in the models (most notably if exports are

increased at the expense of Eskom supply), further coal will be required from the Waterberg for Central Basin

power stations.

4.4.2 Transport infrastructure requirements

Figure 91 to Figure 94 below are analogous to Figure 64 to Figure 67 from Section 3.7.1, but have been updated

to reflect new export volumes from the Waterberg under the sensitivity, as well as to include coal required to be

delivered to Central Basin power stations. Once again, these plots suggest that Transnet’s currently planned

expansion of the Waterberg to Central Basin line will be sufficient to about the mid-2020s under More of the

Same and Lags Behind, where after a further upgrade will be required. Under At the Forefront, some

expansion of capacity is necessary although not to the full 23 Mtpa level – a total capacity of 15.5 Mtpa is

required. Under Low Carbon World expansion is required to even a lower limit, at 10 Mtpa for a short period in

the early 2030s, after which this need is no longer seen.

It needs to be reiterated that this assumes that all other mines planned for the Central Basin open on time, and at

planned supply volumes to Eskom. If these conditions are not met, additional coal will be required from the

Waterberg for Central Basin power stations, with a concurrent requirement for infrastructure built early and with

higher capacities. Furthermore, as stated previously, if the opening of export-only mines in the Waterberg is

proven to be feasible, additional rail capacity will be required to move this coal to the ports.

FIGURE 91: EXPORTS AND CENTRAL BASIN

SUPPLY FROM THE WATERBERG -

LAGS BEHIND (5 YEAR ROLLING

AVERAGE)


SUPPLY FROM THE WATERBERG –

LOW CARBON WORLD (5 YEAR

ROLLING AVERAGE)

0

10

20

30

40

50

60

2010 2015 2020 2025 2030 2035 2040

Mtp

a

Waterberg to Mpumalanga power stations Exports Capacity on Waterberg to Central Basin rail line

0

10

20

30

40

2010 2015 2020 2025 2030 2035 2040

Mtp

a

Waterberg to Mpumalanga power stations

Exports




SUPPLY FROM THE WATERBERG –

MORE OF THE SAME (5 YEAR ROLLING

AVERAGE)


SUPPLY FROM THE WATERBERG – AT

THE FOREFRONT (5 YEAR ROLLING

AVERAGE)

0

10

20

30

40

50

60

2010 2015 2020 2025 2030 2035 2040

Mtp

a


0

10

20

30

40

50

60

2010 2015 2020 2025 2030 2035 2040

Mtp

a



APPENDIX A: DETAILS OF COAL-FIRED POWER STATION DECOMMISSIONING 2010 – 2040

TABLE 29: MORE OF THE SAME COAL-FIRED POWER STATION RTS AND “LATE” DECOMMISSIONING 2010 - 2040

Arnot Camden Duvha Grootvlei Hendrina Kendal Komati Kriel Lethabo Majuba Matimba Matla Tutuka non-Eskom

2010 0 0 0 380 0 0 0 0 0 0 0 0 0 0

2011 0 0 0 190 0 0 114 0 0 0 0 0 0 0

2012 0 0 0 140 0 0 296 0 0 0 0 0 0 0

2013 0 0 0 0 0 0 298 0 0 0 0 0 0 0

2014 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2015 0 0 0 0 0 0 0 0 0 0 0 0 0 -180

2016 0 0 0 0 0 0 0 0 0 0 0 0 0 -90

2017 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2018 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2019 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2020 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2021 0 0 0 0 0 0 0 0 0 0 0 0 0 -75

2022 0 -1450 0 0 0 0 -420 0 0 0 0 0 0 -90

2023 0 0 0 -1090 0 0 -458 0 0 0 0 -900 0 0

2024 0 0 0 0 0 0 0 0 0 0 0 -1030 0 0

2025 0 0 0 0 0 0 0 0 0 0 0 -1520 0 0

2026 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2027 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2028 0 0 -2322 0 0 0 0 0 0 0 0 0 0 0

2029 0 0 -1128 0 0 0 0 0 0 0 0 0 0 0

2030 0 0 0 0 0 0 0 0 0 0 0 0 0 0


2031 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2032 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2033 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2034 0 0 0 0 -187 0 0 0 0 0 0 0 0 0

2035 -94 0 0 0 -280 0 0 0 0 0 0 0 0 0

2036 -282 0 0 0 -280 0 0 0 0 0 0 0 0 0

2037 -282 0 0 0 -280 0 0 0 0 0 0 0 0 0

2038 -282 0 0 0 -280 0 0 0 0 0 0 0 0 0

2039 -282 0 0 0 -280 0 0 0 -58 0 0 0 -403 0

2040 -282 0 0 0 -280 0 0 -530 -700 0 0 0 -690 0

TABLE 30: LAGS BEHIND COAL-FIRED POWER STATION RTS AND “MID” DECOMMISSIONING 2010 - 2040


2010 0 0 0 380 0 0 0 0 0 0 0 0 0 0

2011 0 0 0 190 0 0 114 0 0 0 0 0 0 0

2012 0 0 0 140 0 0 296 0 0 0 0 0 0 0

2013 0 0 0 0 0 0 298 0 0 0 0 0 0 0

2014 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2015 0 0 0 0 0 0 0 0 0 0 0 0 0 -180

2016 0 0 0 0 0 0 0 0 0 0 0 0 0 -90

2017 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2018 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2019 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2020 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2021 0 0 0 0 0 0 0 0 0 0 0 0 0 -75

2022 0 -1450 0 0 0 0 -420 0 0 0 0 0 0 -90


2023 0 0 0 -1090 0 0 -458 0 0 0 0 -900 0 0

2024 0 0 0 0 0 0 0 0 0 0 0 -1030 0 0

2025 0 0 0 0 0 0 0 0 0 0 0 -1520 0 0

2026 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2027 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2028 0 0 -2322 0 0 0 0 0 0 0 0 0 0 0

2029 0 0 -1128 0 0 0 0 0 0 0 0 0 0 0

2030 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2031 -94 0 0 0 -467 0 0 0 0 0 0 0 0 0

2032 -282 0 0 0 -280 0 0 0 0 0 0 0 0 0

2033 -282 0 0 0 -280 0 0 0 0 0 0 0 0 0

2034 -282 0 0 0 -280 0 0 0 0 0 0 0 0 0

2035 -282 0 0 0 -280 0 0 0 -58 0 0 0 -403 0

2036 -282 0 0 0 -280 0 0 -530 -700 0 0 0 -690 0

2037 -282 0 0 0 0 0 0 -795 -700 0 -79 0 -690 0

2038 -282 0 0 0 0 -183 0 -795 -700 0 -942 0 -690 0

2039 -164 0 0 0 0 -731 0 -729 -700 0 -942 0 -690 0

2040 0 0 0 0 0 -731 0 0 -700 0 -942 0 -345 0

TABLE 31: AT THE FOREFRONT COAL-FIRED POWER STATION RTS AND “MID” DECOMMISSIONING 2010 - 2040


2010 0 0 0 380 0 0 0 0 0 0 0 0 0 0

2011 0 0 0 190 0 0 114 0 0 0 0 0 0 0

2012 0 0 0 140 0 0 296 0 0 0 0 0 0 0

2013 0 0 0 0 0 0 298 0 0 0 0 0 0 0

2014 0 0 0 0 0 0 0 0 0 0 0 0 0 0


2015 0 0 0 0 0 0 0 0 0 0 0 0 0 -180

2016 0 0 0 0 0 0 0 0 0 0 0 0 0 -90

2017 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2018 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2019 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2020 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2021 0 0 0 0 0 0 0 0 0 0 0 0 0 -75

2022 0 -1450 0 0 0 0 -420 0 0 0 0 0 0 -90

2023 0 0 0 -1090 0 0 -458 0 0 0 0 -900 0 0

2024 0 0 0 0 0 0 0 0 0 0 0 -1030 0 0

2025 0 0 0 0 0 0 0 0 0 0 0 -1520 0 0

2026 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2027 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2028 0 0 -2322 0 0 0 0 0 0 0 0 0 0 0

2029 0 0 -1128 0 0 0 0 0 0 0 0 0 0 0

2030 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2031 -94 0 0 0 -467 0 0 0 0 0 0 0 0 0

2032 -282 0 0 0 -280 0 0 0 0 0 0 0 0 0

2033 -282 0 0 0 -280 0 0 0 0 0 0 0 0 0

2034 -282 0 0 0 -280 0 0 0 0 0 0 0 0 0

2035 -282 0 0 0 -280 0 0 0 -58 0 0 0 -403 0

2036 -282 0 0 0 -280 0 0 -530 -700 0 0 0 -690 0

2037 -282 0 0 0 0 0 0 -795 -700 0 -79 0 -690 0

2038 -282 0 0 0 0 -183 0 -795 -700 0 -942 0 -690 0

2039 -164 0 0 0 0 -731 0 -729 -700 0 -942 0 -690 0

2040 0 0 0 0 0 -731 0 0 -700 0 -942 0 -345 0


TABLE 32: LOW CARBON WORLD COAL-FIRED POWER STATION RTS AND “EARLY” DECOMMISSIONING 2010 - 2040


2010 0 0 0 380 0 0 0 0 0 0 0 0 0 0

2011 0 0 0 190 0 0 114 0 0 0 0 0 0 0

2012 0 0 0 140 0 0 296 0 0 0 0 0 0 0

2013 0 0 0 0 0 0 298 0 0 0 0 0 0 722

2014 0 0 0 0 0 0 0 0 0 0 0 0 0 722

2015 0 0 0 0 0 0 0 0 0 0 0 0 0 1444

2016 0 0 0 0 0 0 0 0 0 0 0 0 0 722

2017 0 0 0 0 0 0 0 0 0 0 0 0 0 722

2018 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2019 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2020 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2021 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2022 0 -1450 0 0 0 0 -420 0 0 0 0 0 0 0

2023 0 0 0 -1090 0 0 -458 0 0 0 0 -900 0 0

2024 0 0 0 0 0 0 0 0 0 0 0 -1030 0 0

2025 0 0 0 0 0 0 0 0 0 0 0 -1520 0 0

2026 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2027 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2028 0 0 -2322 0 0 0 0 0 0 0 0 0 0 0

2029 0 0 -1128 0 0 0 0 0 0 0 0 0 0 -286

2030 0 0 0 0 0 0 0 0 0 0 0 0 0 -286

2031 -94 0 0 0 -467 0 0 0 -58 0 0 0 -403 -286


2032 -282 0 0 0 -280 0 0 -530 -700 0 0 0 -690 0

2033 -282 0 0 0 -280 0 0 -795 -700 0 -79 0 -690 0

2034 -282 0 0 0 -280 -183 0 -795 -700 0 -942 0 -690 0

2035 -282 0 0 0 -280 -731 0 -729 -700 0 -942 0 -690 0

2036 -282 0 0 0 -280 -731 0 0 -700 0 -942 0 -345 0

2037 -282 0 0 0 0 -731 0 0 0 0 -785 0 0 0

2038 -282 0 0 0 0 -731 0 0 0 0 0 0 0 0

2039 -164 0 0 0 0 -731 0 0 0 0 0 0 0 0

2040 0 0 0 0 0 0 0 0 0 0 0 0 0 0


APPENDIX B: DETAILS OF ELECTRICITY GENERATION BUILD PLAN 2010 – 2040

TABLE 33: MORE OF THE SAME ELECTRICITY GENERATION DECOMMISSIONING AND BUILD PLAN (ALIGNS WITH IRP2010 BASE CASE TO 2030)

RT

S C

ap

ac

ity

Me

du

pi

Ku

sil

e

De

co

mm

iss

ion

ing

OC

GT

CC

GT

Co

-ge

ne

rati

on

Nu

cle

ar

Win

d

CS

P

So

lar

PV

Imp

ort

Hy

dro

La

nd

fill

, h

yd

ro

Pu

mp

ed

sto

rag

e

Co

al

Imp

ort

s

CC

GT

ba

se

loa

d

Co

al

(PF

)

Co

al

FB

C

Co

al

UC

G

To

tal

ne

w b

uil

d

To

tal

sy

ste

m

ca

pa

cit

y

MW MW MW MW MW MW MW MW MW MW MW MW MW MW MW MW MW MW MW MW MW

Existing 2010 43,750

2010 380 0 0 0 0 0 260 0 0 0 0 0 0 0 0 0 0 0 0 640 44,390

2011 304 0 0 0 0 0 130 0 0 0 0 0 0 0 0 0 0 0 0 434 44,824

2012 436 0 0 0 0 0 0 0 400 0 0 0 100 0 0 0 0 0 0 936 45,760

2013 298 722 0 0 1020 0 0 0 400 0 0 0 25 333 0 0 0 0 0 2,798 48,558

2014 0 722 0 0 0 0 0 0 0 100 0 0 0 999 0 0 0 0 0 1,821 50,379

2015 0 1444 0 -180 0 0 0 0 0 100 0 0 0 0 0 0 0 0 0 1,364 51,743

2016 0 722 0 -90 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 632 52,375

2017 0 722 1446 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2,168 54,543

2018 0 0 723 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 723 55,266

2019 0 0 1446 0 460 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1,906 57,172

2020 0 0 723 0 805 0 0 0 0 0 0 653 0 0 0 0 0 0 0 2,181 59,353

2021 0 0 0 -75 805 474 0 0 0 0 0 1023 0 0 0 0 0 0 0 2,227 61,580

2022 0 0 0 -1960 805 948 0 0 0 0 0 283 0 0 600 0 0 750 0 1,426 63,006

2023 0 0 0 -2448 0 711 0 0 0 0 0 0 0 0 600 0 1500 750 0 1,113 64,119

2024 0 0 0 -1030 0 474 0 0 0 0 0 0 0 0 0 0 1500 250 0 1,194 65,313

2025 0 0 0 -1520 345 0 0 0 0 0 0 0 0 0 0 0 3000 0 0 1,825 67,138

2026 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1500 0 0 1,500 68,638


2027 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1500 0 0 1,500 70,138

2028 0 0 0 -2322 460 237 0 0 0 0 0 0 0 0 0 0 3750 0 0 2,125 72,263

2029 0 0 0 -1128 0 237 0 0 0 0 0 0 0 0 0 0 2250 0 0 1,359 73,622

2030 0 0 0 0 0 237 0 0 0 0 0 0 0 0 0 0 1,500 0 0 1,737 75,359

2031 0 0 0 0 229 0 0 0 0 0 0 0 0 0 0 711 750 250 250 2,191 77,550

2032 0 0 0 0 115 0 0 0 0 0 0 0 0 0 0 711 750 0 0 1,576 79,126

2033 0 0 0 0 229 0 0 0 0 0 0 0 0 0 0 711 750 0 0 1,691 80,816

2034 0 0 0 -186.5 229 0 0 0 0 0 0 0 0 0 0 1,423 0 0 250 1,716 82,532

2035 0 0 0 -373.73 115 0 0 0 0 0 0 0 0 0 0 0 1,500 0 0 1,241 83,773

2036 0 0 0 -561.69 229 0 0 0 0 0 0 0 0 0 0 0 1,500 0 250 1,418 85,191

2037 0 0 0 -561.69 229 0 0 0 0 0 0 0 0 0 0 0 1,500 250 250 1,668 86,858

2038 0 0 0 -561.69 229 0 0 0 0 0 0 0 0 0 0 0 2,250 0 0 1,918 88,776

2039 0 0 0 -1022.8 229 0 0 0 0 0 0 0 0 0 0 0 1,500 0 250 957 89,733

2040 0 0 0 -2482.3 459 0 0 0 0 0 0 0 0 0 0 0 3,000 0 250 1,226 90,959

TABLE 34: LAGS BEHIND ELECTRICITY GENERATION DECOMMISSIONING AND BUILD PLAN (ALIGNS WITH IRP2010 BASE CASE TO 2030)

RT

S C

ap

ac

ity

Me

du

pi

Ku

sil

e

De

co

mm

iss

ion

ing

OC

GT

CC

GT

Co

-ge

ne

rati

on

Nu

cle

ar

Win

d

CS

P

So

lar

PV

Imp

ort

Hy

dro

La

nd

fill

, h

yd

ro

Pu

mp

ed

sto

rag

e

Co

al

Imp

ort

s

CC

GT

ba

se

loa

d

Co

al

(PF

)

Co

al

FB

C

Co

al

UC

G

To

tal

ne

w b

uil

d

To

tal

sy

ste

m

ca

pa

cit

y



2010 380 0 0 0 0 0 260 0 0 0 0 0 0 0 0 0 0 0 0 640 44,390

2011 304 0 0 0 0 0 130 0 0 0 0 0 0 0 0 0 0 0 0 434 44,824

2012 436 0 0 0 0 0 0 0 400 0 0 0 100 0 0 0 0 0 0 936 45,760

2013 298 722 0 0 1020 0 0 0 400 0 0 0 25 333 0 0 0 0 0 2,798 48,558


2014 0 722 0 0 0 0 0 0 0 100 0 0 0 999 0 0 0 0 0 1,821 50,379

2015 0 1444 0 -180 0 0 0 0 0 100 0 0 0 0 0 0 0 0 0 1,364 51,743

2016 0 722 0 -90 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 632 52,375

2017 0 722 1446 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2,168 54,543

2018 0 0 723 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 723 55,266

2019 0 0 1446 0 460 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1,906 57,172

2020 0 0 723 0 805 0 0 0 0 0 0 653 0 0 0 0 0 0 0 2,181 59,353

2021 0 0 0 -75 805 474 0 0 0 0 0 1023 0 0 0 0 0 0 0 2,227 61,580

2022 0 0 0 -1960 805 948 0 0 0 0 0 283 0 0 600 0 0 750 0 1,426 63,006

2023 0 0 0 -2448 0 711 0 0 0 0 0 0 0 0 600 0 1500 750 0 1,113 64,119

2024 0 0 0 -1030 0 474 0 0 0 0 0 0 0 0 0 0 1500 250 0 1,194 65,313

2025 0 0 0 -1520 345 0 0 0 0 0 0 0 0 0 0 0 3000 0 0 1,825 67,138

2026 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1500 0 0 1,500 68,638

2027 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1500 0 0 1,500 70,138

2028 0 0 0 -2322 460 237 0 0 0 0 0 0 0 0 0 0 3750 0 0 2,125 72,263

2029 0 0 0 -1128 0 237 0 0 0 0 0 0 0 0 0 0 2250 0 0 1,359 73,622

2030 0 0 0 0 0 237 0 0 0 0 0 0 0 0 0 0 1500 0 0 1,737 75,359

2031 0 0 0 -560.48 344.1 0 0 0 0 0 0 0 0 0 0 711.3 1500 250 250 2,495 77,854

2032 0 0 0 -561.69 114.7 0 0 0 0 0 0 0 0 0 0 711.3 750 0 0 1,014 78,868

2033 0 0 0 -561.69 229.4 0 0 0 0 0 0 0 0 0 0 711.3 750 0 250 1,379 80,247

2034 0 0 0 -561.69 229.4 0 0 0 0 0 0 0 0 0 0 1422.6 750 0 0 1,840 82,088

2035 0 0 0 -1022.8 344.1 0 0 0 0 0 0 0 0 0 0 0 2250 0 250 1,821 83,909

2036 0 0 0 -2482.3 344.1 0 0 0 0 0 0 0 0 0 0 0 3000 250 250 1,362 85,271

2037 0 0 0 -2546.2 458.8 0 0 0 0 0 0 0 0 0 0 0 2250 0 500 663 85,933

2038 0 0 0 -3592.7 458.8 0 0 0 0 0 0 0 0 0 0 0 3750 250 250 1,116 87,049

2039 0 0 0 -3957.5 573.5 0 0 0 0 0 0 0 0 0 0 0 3750 0 500 866 87,915

2040 0 0 0 -2718.7 458.8 0 0 0 0 0 0 0 0 0 0 0 3000 250 250 1,240 89,155


TABLE 35: AT THE FOREFRONT ELECTRICITY GENERATION DECOMMISSIONING AND BUILD PLAN (ALIGNS WITH IRP2010 POLICY ADJUSTED TO 2030)

RT

S C

ap

ac

ity

Me

du

pi

Ku

sil

e

De

co

mm

iss

ion

ing

OC

GT

CC

GT

Co

-ge

ne

rati

on

Nu

cle

ar

Win

d

CS

P

So

lar

PV

Imp

ort

Hy

dro

La

nd

fill

, h

yd

ro

Pu

mp

ed

sto

rag

e

Co

al

Imp

ort

s

CC

GT

ba

se

loa

d

Co

al

(PF

)

Co

al

FB

C

Co

al

UC

G

To

tal

ne

w b

uil

d

To

tal

sy

ste

m

ca

pa

cit

y



2010 380 0 0 0 0 0 260 0 0 0 0 0 0 0 0 0 0 0 0 640 44,390

2011 304 0 0 0 0 0 130 0 0 0 0 0 0 0 0 0 0 0 0 434 44,824

2012 436 0 0 0 0 0 0 0 400 0 300 0 100 0 0 0 0 0 0 1,236 46,060

2013 298 722 0 0 1,020 0 0 0 400 0 300 0 25 333 0 0 0 0 0 3,098 49,158

2014 0 722 0 0 0 0 0 0 400 100 300 0 0 999 0 0 0 500 0 3,021 52,179

2015 0 1,444 0 -180 0 0 0 0 400 100 300 0 0 0 0 0 0 500 0 2,564 54,743

2016 0 722 0 -90 0 0 0 0 400 100 300 0 0 0 0 0 0 0 0 1,432 56,175

2017 0 722 1,446 0 0 0 0 0 400 100 300 0 0 0 0 0 0 0 0 2,968 59,143

2018 0 0 723 0 0 0 0 0 400 100 300 0 0 0 0 0 0 0 0 1,523 60,666

2019 0 0 1,446 0 0 0 0 0 400 100 300 0 0 0 0 0 0 250 0 2,496 63,162

2020 0 0 723 0 0 237 0 0 400 100 300 0 0 0 0 0 0 250 0 2,010 65,172

2021 0 0 0 -75 0 237 0 0 400 100 300 0 0 0 0 0 0 250 0 1,212 66,384

2022 0 0 0 -1,960 805 237 0 0 400 100 300 1,143 0 0 0 0 0 250 0 1,275 67,659

2023 0 0 0 -2,448 805 0 0 1,600 400 100 300 1,183 0 0 0 0 0 250 0 2,190 69,849

2024 0 0 0 -1,030 0 0 0 1,600 800 100 300 283 0 0 0 0 0 250 0 2,303 72,152

2025 0 0 0 -1,520 805 0 0 1,600 1,600 100 1,000 0 0 0 0 0 0 250 0 3,835 75,987

2026 0 0 0 0 0 0 0 1,600 400 0 500 0 0 0 1,000 0 0 0 0 3,500 79,487

2027 0 0 0 0 0 0 0 0 1,600 0 500 0 0 0 0 0 0 250 0 2,350 81,837

2028 0 0 0 -2,322 690 474 0 1,600 0 0 500 0 0 0 0 0 1,000 0 0 1,942 83,779

2029 0 0 0 -1,128 805 237 0 1,600 0 0 1,000 0 0 0 0 0 250 0 0 2,764 86,543


2030 0 0 0 0 0 948 0 0 0 0 1,000 0 0 0 0 0 1,000 0 0 2,948 89,491

2031 0 0 0 -560 344 0 130 1,600 400 100 280 500 25 375 500 711 750 250 0 5,405 94,896

2032 0 0 0 -562 229 0 0 0 200 0 260 0 25 0 0 711 0 0 0 864 95,760

2033 0 0 0 -562 229 0 0 0 400 100 250 0 25 0 0 711 750 0 0 1,904 97,664

2034 0 0 0 -562 229 0 0 0 200 0 260 0 0 375 0 1,423 750 0 0 2,675 100,339

2035 0 0 0 -1,023 229 0 0 0 400 0 310 0 25 0 0 0 750 250 0 942 101,281

2036 0 0 0 -2,482 459 0 130 1,600 600 100 470 0 50 0 0 0 2,250 0 0 3,176 104,457

2037 0 0 0 -2,546 344 0 0 0 400 100 480 0 25 375 0 0 1,500 250 0 928 105,385

2038 0 0 0 -3,593 574 0 130 1,600 600 100 600 0 50 0 0 0 2,250 250 0 2,561 107,946

2039 0 0 0 -3,958 574 0 0 0 800 100 650 0 50 375 0 0 3,000 250 0 1,841 109,787

2040 0 0 0 -2,719 459 0 0 1,600 600 0 510 0 25 375 0 0 2,250 250 0 3,350 113,137

TABLE 36: LOW CARBON WORLD ELECTRICITY GENERATION DECOMMISSIONING AND BUILD PLAN (ALIGNS WITH IRP2010 EMISSIONS 3 TO 2030)

RT

S C

ap

ac

ity

Me

du

pi

Ku

sil

e

De

co

mm

iss

ion

ing

OC

GT

CC

GT

Co

-ge

ne

rati

on

Nu

cle

ar

Win

d

CS

P

So

lar

PV

Imp

ort

Hy

dro

La

nd

fill

, h

yd

ro

Pu

mp

ed

sto

rag

e

Co

al

Imp

ort

s

CC

GT

ba

se

loa

d

Co

al

(PF

)

Co

al

FB

C

Co

al

UC

G

To

tal

ne

w b

uil

d

To

tal

sy

ste

m

ca

pa

cit

y



2010 380 0 0 0 0 0 260 0 0 0 0 0 0 0 0 0 0 0 0 640 44,390

2011 304 0 0 0 0 0 130 0 0 0 0 0 0 0 0 0 0 0 0 434 44,824

2012 436 0 0 0 0 0 0 0 400 0 0 0 100 0 0 0 0 0 0 936 45,760

2013 298 722 0 0 1,020 0 0 0 400 0 0 0 25 333 0 0 0 0 0 2,798 48,558

2014 0 722 0 0 0 0 0 0 0 100 0 0 0 999 0 0 0 0 0 1,821 50,379

2015 0 1,444 0 -180 0 0 0 0 1,600 100 0 0 0 0 0 0 0 0 0 2,964 53,343

2016 0 722 0 -90 0 0 0 0 1,600 0 0 0 0 0 0 0 0 0 0 2,232 55,575


2017 0 722 1,446 0 0 948 0 0 1,600 1,500 0 0 0 0 0 0 0 0 0 6,216 61,791

2018 0 0 723 0 0 948 0 0 1,600 3,125 0 0 0 0 0 0 0 0 0 6,396 68,187

2019 0 0 1,446 0 805 948 0 0 1,600 3,125 0 0 0 0 0 0 0 0 0 7,924 76,111

2020 0 0 723 0 805 948 0 0 1,600 3,125 0 1,110 0 0 0 0 0 0 0 8,311 84,422

2021 0 0 0 -75 805 474 0 0 1,600 375 0 0 0 0 0 0 0 0 0 3,179 87,601

2022 0 0 0 -1,960 0 0 0 1,600 1,600 0 0 0 0 0 0 0 0 0 0 1,240 88,841

2023 0 0 0 -2,448 0 0 0 1,600 200 0 0 0 0 0 0 0 0 0 0 -648 88,193

2024 0 0 0 -1,030 0 0 0 0 1,600 0 0 0 0 0 0 0 0 0 0 570 88,763

2025 0 0 0 -1,520 0 0 0 1,600 0 0 0 0 0 0 0 0 0 0 0 80 88,843

2026 0 0 0 0 805 0 0 1,600 400 0 0 0 0 0 0 0 0 0 0 2,805 91,648

2027 0 0 0 0 805 0 0 0 1,400 0 0 0 0 0 0 0 0 0 0 2,205 93,853

2028 0 0 0 -2,322 805 0 0 1,600 0 0 0 0 0 0 0 0 0 0 0 83 93,936

2029 0 0 0 -1,128 805 0 0 1,600 400 0 0 0 0 0 0 0 0 0 0 1,677 95,613

2030 0 0 0 0 805 0 0 0 800 0 0 0 0 0 0 0 0 0 0 1,605 97,218

2031 0 0 0 -1,022 574 0 130 1,600 1,200 700 0 500 50 375 0 1,423 0 0 0 5,530 102,748

2032 0 0 0 -2,482 574 0 0 0 1,200 700 0 0 25 375 0 1,423 0 0 0 1,814 104,561

2033 0 0 0 -2,826 574 0 130 1,600 1,200 900 0 500 50 0 0 1,423 0 0 0 3,550 108,111

2034 0 0 0 -3,872 688 0 0 0 1,600 1,000 0 0 50 375 0 2,134 0 0 0 1,975 110,086

2035 0 0 0 -4,355 918 0 130 3,200 1,800 1,100 0 0 50 375 0 0 0 0 0 3,218 113,304

2036 0 0 0 -3,280 574 0 0 3,200 1,400 900 0 500 50 0 0 0 0 0 0 3,343 116,647

2037 0 0 0 -1,798 574 0 130 1,600 1,200 600 0 0 50 375 0 0 0 0 0 2,730 119,377

2038 0 0 0 -1,013 344 0 0 1,600 600 500 0 0 25 0 0 0 0 0 0 2,056 121,433

2039 0 0 0 -896 344 0 0 1,600 800 500 0 500 25 375 0 0 0 0 0 3,248 124,681

2040 0 0 0 0 229 0 0 0 600 300 0 0 0 0 0 0 0 0 0 1,129 125,810