setting up for the 2020s - arep · 1 setting up for the 2020s addressing south africa [s...
Post on 10-Oct-2020
3 Views
Preview:
TRANSCRIPT
1
Setting up for the 2020sAddressing South Africa’s electricity crisis and getting ready for the next decade
CSIR Energy CentrePretoria. January 2020v1.0
DR JARRAD WRIGHT JOANNE CALITZ
2
45Mitigation options/solutions5
39What if?4
30Current plans (IRP 2019)3
18The burning platform2
2Executive Summary1
Agenda
3
What is being addressed in this contribution?
1 As announced by the President of South Africa (March 2019); 2 Crises are (a) sources of uncertainty, disruption, and change, (b) harmful/threatening for organisations and
stakeholders, (c) behavioural phenomena (socially constructed by actors involved) (d) parts of larger processes (rather than discrete events); 2 As an individual household, as a
business, as a regulator, as a policymaker/decision-maker; 3 It is not intended to address other topics in the electricity industry in South Africa e.g. financial sustainability & operational
efficiency of institutions, political economy of stakeholders interacting on the basis of relevant interests.
Sources: Bundy
Focus is to analyse, interpret and present an evidence-base that can assist in supporting custodians & stakeholders to ensure short-term system adequacy3
Is South Africa in an electricity crisis1,2?
How long will loadshedding last?
What options are available to solve theelectricity crisis3?
What is happening now and what willhappen in the next 3-5 years?
Assessing whether this is an electricity crisis (andthe extent thereof)
Present two scenarios to understand theexpected intensity & duration of loadshedding
Proposing feasible options/solutions to alleviatethe crisis and testing their impact
What is CSIR addressing?
4
Why is CSIR making this contribution?
1 Public domain information; 2 As with previous analyses in this domain e.g. Draft IRP 2016, Draft IRP 2018, IRP 2019 (amongst others); 3 As announced by the President of South Africa (March 2019); 4 DMRE RMPPPP RfI is in the public domain (closing date 31 January 2020) with RfP to follow (publication for all custodians/stakeholders will inform responses, reduce risk and allow for more informed procurement)
• CSIR is an independent scientific research institution with no vested interests in outcomes
Independence
• CSIR has (over the years) developed & collated necessary data, analytics and modelling frameworks1
• CSIR publish all analysis in this domain and intend to continue (public interest)2
• Providing an evidence-base to inform and aid planning for all custodians/stakeholders
Transparency
• CSIR is highly proficient/competent in this domain with well respected capabilities• CSIR is well positioned to undertake this suite of analysis as a publicly trusted independent institution
Experience/expertise
• South Africa is in an electricity crisis3
• CSIR is already engaging with key custodians/stakeholders & intends to formally engage further to assistin driving expedited responses/actions by various custodians/stakeholders4
Expedience
5
The burning platform & need for an urgent response requires a widely understood evidence-base for all custodians/stakeholders to take action
1 Even if Medupi, Kusile, REIPPPP capacity under construction comes online as expected; EAF – Energy Availability Factor
South Africa had the worst year of loadshedding on record in 2019 (1352 GWh, 530 hours) with up toStage 6 load shedding being implemented having significant impacts on the economy (≈R 60-120 bln)
Loadshedding is expected to continue for 2-3 years depending on key decisions/actions
An urgent response is necessary to ensure short-term adequacy and set South Africa on a path towardslong-term adequacy in the 2020s
Systemic changes in Eskom fleet performance (EAF) expectation and demand forecast requires anupdated understanding of capacity & energy gap relative to IRP 2019 assumptions1 (as will be shown)
6
Notes: Load shedding assumed to have taken place for the full hours in which it was implemented. Practically, load shedding (and the Stage) may occassionally change/ end during a
particular hour; Total GWh calculated assuming Stage 1 = 1 000 MW, Stage 2 = 2 000 MW, Stage 3 = 3 000 MW, Stage 4 = 4 000 MW, Stage 5 = 5 000 MW, Stage 6 = 6 000 MW;
Cost to the economy of load shedding is estimated using COUE (cost of unserved energy) = 87.50 R/kWh
Sources: Eskom Twitter account; Eskom se Push (mobile app); Nersa; CSIR analysis
400
0
200
600
1 600
800
1 000
1 200
1 400
Load shed [GWh]
2007 2008
568
45
123
2009 2010 2011 20202012 2013 2014
874
406
2015 2016
203
2017
80
1 325
130
62
618
43
30
476
93
20192018
192
1 352
143
17126
176
Unknown Stage 4Stage 5Stage 6 Stage 3 Stage 2 Stage 1
Year
2007
2008
….
2014
2015
2016
2017
2018
2019
2020 (YTD)
Duration
of outages
(hours)
-
-
.…
121
852
-
-
127
530
80
Energy
shed
(GWh)
176
476
.…
203
1325
-
-
192
1352
143
2019 was the most intensive year of loadshedding to date in South Africa with Stage 6 being implemented in December 2019
Est. econ.
Impact
(ZAR-bln)
8-15
21-42
.…
9-18
58-116
-
-
8-17
59-118
6-12
7
System inadequacy has been highlighted recently by DMRE and Eskom – but no specific solutions/interventions provided (yet)
Integrated Resource Plan 2019 (IRP 2019)
• Released: October 2019
• Highlighted short-term supply gap between 2019-2022
• Primarily due to lead-time of first new-build capacity in2023
Medium-Term System Adequacy Outlook 2019 (MTSAO 2019)
• Released: November 2019
• Also highlighted lack of system adequacy if Eskom fleetEAF is below 72% for time horizon 2019-2024
• No new investments made in time-horizon
NOTE: Different assumptions made in a few dimensions for IRP 2019 and MTSAO 2019 (demand forecast, EAF, supply) but same conclusions reached.
Sources: DMRE; Eskom
8
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030
70
94
0
68
62
64
86
80
82
72
90
98
78
100
88
66
84
74
96
76
92
67.0%
83.1
EAF[%]
73.6
80.0
75.5
Historical fleet EAF decline seems irreversable, IRP 2018 EAF has not materialised... risk of IRP 2019 or MTSAO 2019 EAF (High) not materialising?
Actual
MTSAO 2019 (MES 1, High)
Low (IRP 2018)
High (IRP 2018)
Moderate (IRP 2018)
IRP 2019
MTSAO 2019 (MES 1, Low)
MTSAO 2019 (MES 2, High)
Planned
EAF – Energy Availability Factor; MTSAO – Medium-term System Adequacy Outlook
NOTE: 2019 EAF actual is YTD
Sources: Draft IRP 2018; IRP2019; Eskom; CSIR Energy Centre analysis
9
Not explored
Worst scenario
IRP 2019
(DMRE)
Updated
(CSIR)
Not explored
Best scenario
What if the demand forecast is lower than in IRP 2019 (very likely) and what if the Eskom fleet EAF is lower than in IRP 2019 (also very likely)?
HighEAF [%]
Sources: CSIR Energy Centre analysis
Demand forecast [TWh]
Low
Low
High
10
Would an updated EAF expectation and demand forecast make any difference to capacity and energy shortages?
NOTE: “Updated” scenario is a test scenario developed by CSIR; Demand forecast is based on Eskom MTSAO demand forecast (until 2024) and IRP 2019 growth rates thereafter;
Updated EAF based on MTSAO MES 1 (Low); EAF – Energy Availability Factor
Sources: IRP 2019; MTSAO; CSIR
Demand forecastEAF
2000 2005 2010 2015 2020 2025 2030
80
60
85
0
65
70
90
95
75
100
64.7
Energy Availability Factor (EAF)[%]
67.0
75.5
2000 2005 2010 2015 2020 2025 2030
150
200
50
250
300
350
0
400
100
246
267
306
Electrical energy demand [TWh]
298
284
Updated
IRP 2019
Actual
IRP 2019
Updated
Actual
11
4
1
0
3
2
5
6
7
8
9
102
02
1
[GW]
20
18
20
19
20
20
20
22
20
23
20
24
20
25
Shortage from IRP 2019 indicating a dominant short-term capacity gap & small energy gap until planned new-build capacity comes online
* Estimated Eskom Demand Response (DR) capability (mostly industrial & energy limited); NOTES: Energy & capacity shortage is demand that cannot be served due to a lack of
capacity (including OCGTs, pumped storage & Eskom DR); Outcomes shown are from deterministic simulations - thus indicative; 99th percentile of capacity & energy shortage is
reported; All IRP 2019 capacity is assumed to come online as planned (Step 3 is always considered implemented); Cost of load shedding is estimated using COUE (cost of unserved
energy) = 87.50 R/kWh; Sources: CSIR Energy Centre analysis
Capacity (shortage) Energy (shortage)
2 000
0
500
1 000
1 500
2 500
3 000
3 500
4 000
4 500
5 000
20
25
20
20
20
19
[GWh/yr]
20
18
20
21
20
22
20
23
20
24
Supply gap (from IRP 2019)
Supply gap (from IRP 2019)
IRP 2019 EAF & IRP 2019 demand forecast(EAF recovery from ≈67% in 2019 to 75.5% by 2024) (Demand forecast immediately growing to 284 TWh by 2025)
Stage 1
Stage 6
Stage 5
Stage 4
Stage 3
Stage 2
Eskom DR*
Loadshedding (actual)
2019: 1352 GWh
(R60-120bln cost to the economy)
NOTE: Understatement of energy shortages
in simulation (conservative estimates)
Stage 7
Stage 8
12
500
3 000
0
1 000
1 500
2 000
2 500
3 500
4 000
5 000
4 500
20
21
20
24
20
23
[GWh/yr]
20
18
20
25
20
19
20
20
20
22
Updated EAF & Updated demand forecast(EAF from ≈67% in 2019 to ≈64% by 2024) (Demand forecast initially flat & growth to 267 TWh by 2025)
6
0
1
2
3
4
5
7
8
9
10
20
23
20
19
20
21
20
18
[GW]
20
20
20
22
20
24
20
25
Updated EAF & demand forecast indicates further shortage relative to IRP 2019 requiring capacity and significantly more energy
Capacity (shortage) Energy (shortage)
Supply gap (from IRP 2019)
Supply gap (from IRP 2019)
* Estimated Eskom Demand Response (DR) capability (mostly industrial & energy limited); NOTES: Energy & capacity shortage is demand that cannot be served due to a lack of
capacity (including OCGTs, pumped storage & Eskom DR); Outcomes shown are from deterministic simulations - thus indicative; 99th percentile of capacity & energy shortage is
reported; All IRP 2019 capacity is assumed to come online as planned (Step 3 is always considered implemented); Cost of load shedding is estimated using COUE (cost of unserved
energy) = 87.50 R/kWh; Sources: CSIR Energy Centre analysis
Loadshedding (actual)
2019: 1352 GWh
(R60-120bln cost to the economy)
NOTE: Understatement of energy shortages
in simulation (conservative estimates)
Stage 1
Stage 6
Stage 5
Stage 4
Stage 3
Stage 2
Eskom DR*
Stage 7
Stage 8
13
Criteria to decide on options to meet the short-term gap should be informed and applied consistently to ensure reasonable cost & timeous delivery
Criteria for choice of short-term options available can ensure a portfolio of options:
Can be supply-side, demand-side and/or storage
Can be delivered in 1-2 years
Will not require extensive procurement process (lead-time)
Can meet capacity (MW) and/or energy needs (MWh)
Can be contracted for 1-3 years (or more if aligned with long-term energy mix)
Ease of implementation (does not require extensive regulatory reform/change)
Aligned with long-term energy mix pathways (technology choices)
Does not require extensive network expansion or augmentation for interconnection
1
2
3
4
5
6
7
8
Sources: CSIR Energy Centre analysis
14
01/2020 01/202401/2021 01/2022 01/2023 01/2025
Critical decisions/actions needed now along with accelerated processes to ensure timeous implementation allowing RSA to ramp into 2020s successfully
RfP (response)PBs (COD)
IPPPP1 (constr.)
RfP (release)
SSEG2 & DG regulations3
REIPPPP PPA neg.REIPPPP PPA (operations)
Comms/incentives (SSEG)
RfI (response)
PBs (FC)
Min. Determinations
IPPPP1 (release)
PBs
IPPPP1 (response)
IPPPP1 (PBs)
IPPPP1 (constr.)
PBs (constr.)
IPPPP1(FC)
PBs (operations)
IPPPP1 (operations)
Cu
sto
me
r re
spo
nse
at
sca
leSt
ep 1
DM
RE
RM
PP
PP
Step
2
Imp
lem
ent
IRP
20
19
4
Step
3
Capacity1
Decision/action
Stakeholders/custodians
Capacity1
Decision/action
Stakeholder/custodian
Capacity1
Decision/action
Stakeholder/custodian
NOTES: Timelines are estimated and in no way prescriptive; PBs – Preferred bidders, PPA – Power Purchase Agreement; RfI – Request for Information; RfP – Request for Proposal; FC – Financial
Close, COD – Commercial Operations Date; 1 Total additional installed capacity; 2 Requires adjusted SSEG regulations (proposed lifting licensing requirement for SSEG & only requiring registration
with NERSA - from 1 MW to 10 MW (or more)); 2 SSEG (res.) does not require regulatory changes (just communications rollout but could be further incentivised by Eskom/municipalities); 3 Will require
Ministerial Determination - generators expected >10 MW (technologies aligned with IRP 2019); 4 Unlikely to get capacity online before 2023-2024 (risk of misalignment with IRP 2019)
1.2 GW (supply/storage) 2.6 GW (supply/storage) 4.9 GW (supply/storage)
1.3-2.8 GW (supply) 0.1-1.9 GW (supply) 0.1-1.6 GW (supply) 0.1-1.6 GW (supply) 1.2-3.9 GW (supply)
1.6 GW (wind), 1.0 GW (PV), 0.5 GW (storage)
3.2 GW (wind), 2.0 GW (PV), 0.5 GW (storage),
0.5 GW (DG/EG), 0.75 GW (coal)
4.8 GW (wind), 2.0 GW (PV), 0.5 GW (storage),
1.0 GW (DG/EG), 0.75 GW (coal)
6.4 GW (wind), 3.0 GW (PV), 0.5 GW (storage),
1.5 GW (DG/EG), 0.75 GW (coal)
Custodians: IPPO/DBSA, DMRE, NERSA, Eskom (DPE), NT Stakeholders: Customers (businesses, households), municipalities, IPPs
4.9 GW (supply/storage) 4.9 GW (supply/storage) 4.9 GW (supply/storage)
Custodians: IPPO/DBSA, DMRE, NERSA, Eskom (DPE), NT Stakeholders: Customers (businesses, households), municipalities, IPPs, financial institutions
Custodians: IPPO/DBSA, DMRE, NERSA, Eskom (DPE), NT Stakeholders: Customers (businesses, households), municipalities, IPPs
This timeline is ambitous (but achievable) - means mostcapacity will only come online after it is needed (in 2021) –accelerated processes are necessary
15
Step 1 is the only immediate feasible response as Step 2 & 3 are expected from 2022/2023 only (best case) but still critical to ensure system adequacyKey recommendations
SSEG – Small-Scale Embedded Generation; DG – Distributed Generation; EG – Embedded Generation; RMPPPP – Risk Mitigation Power Purchase Procurement Programme
Immediate focus on customer response at scale(self-supply) in all customer segments via enablingregulations (easy to implement)Driven by SSEG (residential), EG (commercial/agricultural), EG/DG(industrial/mining), municipalities & storage, REIPPPP ‘power-up’
Accelerate DMRE RMPPPP process to addressremaining capacity & energy gap and ensure capacitycan come online timeouslyAn accelerated process necessary due to immediate shortages &should be based on estimated required capacity (complementingStep 1)
Immediate focus on Ministerial Determinations forall technologies in IRP 2019 followed byprocurement process to ensure timeousimplementationDue to procurement processes & technology specific lead-times thisaction/decision is required now
NOTE: Even with short-term interventions, if EAF does not recover to IRP 2019 levels, shortage is expected in 2020-2021 depending on capacity thatcan feasibly come online, structural load shedding may still need to be considered for 2-3 years (shortages exaggerated in Updated scenario)
• Immediate reduced load shedding as capacity cancome online in 2020 already
• Further assistance from 2021 onwards as morecapacity comes online
• Capacity online from 2022 only (mid-2021 withaccelerated DMRE RMPPPP process)
• Further reduction in load shedding once capacitycomes online
• First capacity online from 2023 only (best case)but required in 2022 (as per IRP 2019)
• Adequate power system into mid-2020s asexisting capacity is decommissioned if IRP 2019planned new-build capacity comes online
DECISION/ACTION IMPACT
Cu
sto
me
r re
spo
nse
at
sca
leSt
ep 1
DM
RE
RM
PP
PP
Step
2
Imp
lem
ent
IRP
20
19
4
Step
3
16
This is a crisis - get capacity under construction online, recover Eskom plant performance AND implement all steps with urgencyKey recommendations
Ensure capacity under construction is delivered as planned (Medupi/Kusile, REIPPPP)
Recover Eskom fleet EAF to realistic levels whilst ensuring value for money relative to alternatives
Implement/enable 3 steps concurrently and with urgency:
• Step 1: Intentionally drive a customer response at scale (with enabling regulations) driven by SSEG(residential), EG (commercial/agricultural), EG/DG (industrial/mining) & storage1
• Step 2: Address remaining capacity/energy gap via an accelerated DMRE RMPPPP process to ensurecapacity is online when required
• Step 3: Continued implementation of IRP 2019 as an immediate focus to ensure sufficient lead-timefor procurement processes and technology specific construction lead-times
1 As shown, these are the only options/solutions that would assist in mitigating load shedding in the next 2-3 years (other options/solutions would take further time to implement).
SSEG – Small-Scale Embedded Generation; DG – Distributed Generation; EG – Embedded Generation; RMPPPP – Risk Mitigation Power Purchase Procurement Programme;
NOTE: For further details – please refer to the remainder of this presentation; SSEG could include a range of technologies but would be dominated by solar PV.
17
Urgency required on actions/decisions now to ensure capacity comes online timeouslyKey recommendations
In order to enable and implement recommended steps, first immediate decisions/actions are needed:
• DMRE/Nersa to update regulations to enable streamlined & expedited self-supply options1
• DMRE/Nersa to publish Ministerial Determinations and/or update EG/DG regulations toenable DMRE RMPPPP additional capacity
• DMRE/Nersa to publish Ministerial Determinations aligned with IRP 2019
• IPPO engage & implement feasible existing REIPPPP PPA negotiations (REIPPPP ‘power-up’)
• Various stakeholders to drive intentional communications program and/or incentives forSSEG deployment (res./com./agri. focus)2
• IPPO to run bid windows for new-build procurement (technologies aligned with IRP 2019)and undertake annually going forward
IPPO – Independent Power Producers Office; DMRE – Department of Mineral Resources and Energy; res. – residential; com. – commercial.1 Adjusted/updated SSEG regulations (proposed lifting licensing requirement for SSEG & only requiring registration with NERSA - from 1 MW to 10 MW (or more)); 2 SSEG (res./com.)
does not require regulatory changes (just communications rollout but could be further incentivised by Eskom/municipalities with appropriate tariffs).
Q1-2020
18
45Mitigation options/solutions5
39What if?4
30Current plans (IRP 2019)3
18The burning platform2
2Executive Summary1
Agenda
19
Notes: Load shedding assumed to have taken place for the full hours in which it was implemented. Practically, load shedding (and the Stage) may occassionally change/ end during a
particular hour; Total GWh calculated assuming Stage 1 = 1 000 MW, Stage 2 = 2 000 MW, Stage 3 = 3 000 MW, Stage 4 = 4 000 MW, Stage 5 = 5 000 MW, Stage 6 = 6 000 MW;
Cost to the economy of load shedding is estimated using COUE (cost of unserved energy) = 87.50 R/kWh
Sources: Eskom Twitter account; Eskom se Push (mobile app); Nersa; CSIR analysis
400
0
200
600
1 600
800
1 000
1 200
1 400
Load shed [GWh]
2007 2008
568
45
123
2009 2010 2011 20202012 2013 2014
874
406
2015 2016
203
2017
80
1 325
130
62
618
43
30
476
93
20192018
192
1 352
143
17126
176
Unknown Stage 4Stage 5Stage 6 Stage 3 Stage 2 Stage 1
Year
2007
2008
….
2014
2015
2016
2017
2018
2019
2020 (YTD)
Duration
of outages
(hours)
-
-
.…
121
852
-
-
127
530
80
Energy
shed
(GWh)
176
476
.…
203
1325
-
-
192
1352
143
2019 has been the most intensive year of loadshedding to date with Stage 6 being implemented in December 2019
Est. econ.
Impact
(ZAR-bln)
8-15
21-42
.…
9-18
58-116
-
-
8-17
59-118
6-12
20
Hourly distribution of actual load shedding January to June 2019
January February AprilMarch May June
Day 1
Day 31
0h00 24h00 0h00 24h00 0h00 24h00 0h00 24h00 0h00 24h00 0h00 24h00
No load shedding
Notes: Load shedding assumed to have taken place for the full hours in which it was implemented. Practically, load shedding (and the Stage) may occassionally change/ end during a
particular hour; Total GWh calculated assuming Stage 1 = 1 000 MW, Stage 2 = 2 000 MW, Stage 3 = 3 000 MW, Stage 4 = 4 000 MW, Stage 5 = 5 000 MW, Stage 6 = 6 000 MW;
Cost to the economy of load shedding is estimated using COUE (cost of unserved energy) = 87.50 R/kWh
Sources: Eskom Twitter account; Eskom se Push (mobile app); Nersa; CSIR analysis
21
Hourly distribution of actual load shedding July to December 2019
July August OctoberSeptember November December
Day 1
Day 31
0h00 24h00 0h00 24h00 0h00 24h00 0h00 24h00 0h00 24h00 0h00 24h00
No load shedding
Notes: Load shedding assumed to have taken place for the full hours in which it was implemented. Practically, load shedding (and the Stage) may occassionally change/ end during a
particular hour; Total GWh calculated assuming Stage 1 = 1 000 MW, Stage 2 = 2 000 MW, Stage 3 = 3 000 MW, Stage 4 = 4 000 MW, Stage 5 = 5 000 MW, Stage 6 = 6 000 MW;
Cost to the economy of load shedding is estimated using COUE (cost of unserved energy) = 87.50 R/kWh
Sources: Eskom Twitter account; Eskom se Push (mobile app); Nersa; CSIR analysis
22
0
5
20
15
10
25
30
35
40
GW
5
3
0
1
4
2
6
GW
Severely constrained system from 4 December 2019, compounded by further breakdowns up to 9 December 2019 resulting in Stage 6 loadsheddingActual hourly production from all power supply sources in RSA for December 2019
NOTES: Pumping load excluded; UCLF – Unplanned Capacity Loss Factor; OCGTs – Open-Cycle Gas TurbinesSources: Eskom; CSIR Energy Centre analysis
Day
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Coal
Solar PV
Load Shedding
Wind
Hydro + PS
CSP
Imports, Other
Diesel
Nuclear
Stage 6
Stage 5
Stage 4
Stage 3
Stage 2
Stage 1
High UCLF (conveyer supply, mine flooding and unitoutages) combined with depletion of emergency reserves(diesel OCGTs and pumped storage)
23
Other(Gx)
Other(Gx)
Munics.(Dx2)
South African supply-demand balance is maintained by the System Operator (Eskom) with some supply/demand responsive customers
Eskom(Tx2)
Eskom(Dx2)
Eskom1
(Gx)Imports ExportsIPPs1
(Gx)
EG = Embedded Generation; Gx = Generation; Tx = Transmission; Dx = Distribution1 Power generated less power station load; Minus pumping load (Eskom owned pumped storage); 2 Transmission/distribution networks incur losses before delivery to customers
Largecustomers
(Com/Ind)
EG(Gx)
Customers(Dom/Com/Ind)
EG(Gx)
Customers(Dom/Com/Ind)
1 1Production1
2 Imports
3 Exports
32
1 1
1 1
24
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030
66
74
100
78
84
86
80
72
90
88
82
76
98
0
94
70
96
68
92
EAF (annual)[%]
67.0%
71.5%
Historical fleet EAF decline since 2001 with apparent recovery in 2016-2017 but trend continued thereafter and is at ≈67% for 2019
Actual
Planned
EAF – Energy Availability Factor
NOTE: 2019 EAF actual is YTD
Sources: IRP2019; Eskom; CSIR Energy Centre analysis
25
70
1-Jan-16 1-Jan-17
75
0
1-Jan-18
90
1-Jan-19 1-Jan-20
60
95
80
65
1-Jan-21
85
100
EAF (weekly)[%]
61.4%
Recent Eskom weekly fleet EAF has been declining with unfortunate consequences of a highly constrained power system
NOTES: EAF - Energy Availability FactorSources: Eskom; CSIR Energy Centre analysis
2016: 76.4% 2017: 78.7% 2018: 71.9% 2019: 67.0%
26
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030
72
70
66
82
80
84
86
88
76
64
90
74
78
94
62
0
96
98
100
68
92
73.6
EAF (annual)[%]
67.0%
80.0
83.1
75.5
Historical fleet EAF decline seems irreversable... IRP 2018 EAF has also not materialised, risk of IRP 2019 or MTSAO 2019 EAF (High) not materialising?
IRP 2019
High (IRP 2018)
Low (IRP 2018)
Moderate (IRP 2018)
Actual
MTSAO 2019 (MES 1, Low)
MTSAO 2019 (MES 1, High)
MTSAO 2019 (MES 2, High)
Planned
EAF – Energy Availability Factor; MTSAO – Medium-term System Adequacy Outlook
NOTE: 2019 EAF actual is YTD
Sources: Draft IRP 2018; IRP2019; Eskom; CSIR Energy Centre analysis
27
The supply-demand balance must be maintained for every hour of every day utilising available supply/demand options (dispatchable & self-dispatched)
60
10
0
20
30
40
50
[GW]
20
18
53.6
Other Storage
Biomass/-gas
DG
Solar PV
CSP
Wind
Hydro
NuclearPS
Peaking
Gas
Nuclear (new)
Coal
Supply(Installed capacity)
Demand (typical week)
0
50
10
40
20
30
60
[GW]
Peak demand (35 GW)
Min. demand (20 GW)
Mo
nd
ay
Tuesd
ay
Wed
nesd
ay
Thu
rsday
Friday
Saturd
ay
Sun
day
Summer
Winter
28
Example – Lower EAF (lower coal fleet availability) would mean a more constrained power system and high-risk of loadshedding
20
0
10
30
40
50
60
[GW]
20
18
37.1
CSP
Wind
Solar PV
Other Storage
DG
Biomass/-gas
Hydro
PS
Peaking
Gas
Nuclear (new)
Nuclear
Coal
Supply(Available capacity – high outages)
Demand (typical week)
Mo
nd
ay
Tuesd
ay
Wed
nesd
ay
Thu
rsday
Friday
Saturd
ay
Sun
day
EAF
40
0
60
10
20
30
50
Min. demand (20 GW)
[GW]
Peak demand (35 GW)
Summer
Winter
29
Example – Higher EAF (higher coal fleet availability) would mean a less constrained power system and low-risk of loadshedding
0
10
20
30
40
50
60
20
18
[GW]
43.6
CSPOther Storage
Biomass/-gas
Solar PV
DG Hydro
Wind
PS
Peaking
Gas
Nuclear (new)
Nuclear
Coal
Supply(Available capacity – low outages)
Demand (typical week)
Mo
nd
ay
Tuesd
ay
Wed
nesd
ay
Thu
rsday
Friday
Saturd
ay
Sun
day
EAF
60
10
30
0
40
20
50
[GW]
Peak demand (35 GW)
Min. demand (20 GW)
Summer
Winter
30
45Mitigation options/solutions5
39What if?4
30Current plans (IRP 2019)3
18The burning platform2
2Executive Summary1
Agenda
31
System inadequacy has been highlighted recently by DMRE and Eskom – but no specific solutions/interventions provided
Integrated Resource Plan 2019 (IRP 2019)
• Released: October 2019
• Highlighted short-term supply gap between 2019-2022
• Primarily due to lead-time of first new-build capacity in2023
Medium-Term System Adequacy Outlook 2019 (MTSAO 2019)
• Released: November 2019
• Also highlighted lack of system adequacy if Eskom fleetEAF is below 72% for time horizon 2019-2024
• No new investments made in time-horizon
NOTE: Different assumptions made in a few dimensions for IRP 2019 and MTSAO 2019 (demand forecast, EAF, supply) but same conclusions reached.
Sources: DMRE; Eskom
32
CSIR simulated South African power system model benchmarked against actuals shows good alignment (2018 as a reference - annual)
200
0
60
160
180
20
80
220
40
140
240
100
260
280
120
6(2.3%)
3(1.4%)6(2.7%)
2018 (Simulated)
11(4.6%)
1(0.4%)
2(0.8%)
6(2.4%)
10(4.1%)
11(4.8%)
3(1.4%)
1(0.4%)
6(2.7%)2(0.8%)
11(4.4%)
Annual electricity production [TWh]
2018 (Actual)
199(82.5%)
241 241
203(84.3%)
Imports, Other
Wind
Hydro + PS
Solar PV
CSP
Diesel
Coal
Nuclear
Supply Sources
PS – Pumped Storage; HVDC – High Voltage Direct Current;
NOTES: Includes generation for pumping load; Simulation in PLEXOS: production cost model with hourly temporal resolution and public input data; “Imports, Other” includes mostly
imported hydro generation (via HVDC imports from Mozambique);
Sources: Eskom; CSIR Energy Centre analysis
33
2
0
10
14
4
8
6
12
16
18
20
JulJan May Jun
Monthly electricity production from Eskom coal in TWh
OctFeb Mar Apr Aug Sep Nov Dec
2018 (Actual) 2018 (Simulated)
CSIR simulated South African power system model benchmarked against actuals shows good alignment – coal focus (2018 as a reference - monthly)
NOTES: Simulation in PLEXOS: production cost model with hourly temporal resolution and public input data;
Sources: Eskom; CSIR Energy Centre analysis
0
60
40
80
120
100
140
160
180
200
20
220
Annual electricity production in TWh
2018 (Actual)
2018 (Simulated)
203 199 -4(-2.1%)
34
350
0
50
400
300
100
200
450
250
150
20
26
20
22
20
20
202.8169.1
20
30
29.653.0
Electricity production [TWh/yr]
3.36.514.5
3.41.9
13.4
8.2
20
18
14.92.9
20
24
13.40.92.2
20
28
0.5
11.8
20
32
20
34
20
36
20
38
20
40
20
42
20
44
20
46
20
48
20
50
248.0
311.3
IRP 2019 has been gazetted indicating an immediately inclining demand forecast, decommissioning coal fleet but with improving EAFInstalled capacity and electricity supplied from 2018 to 2050
1 Projection based on optimisation of 2030-2050 energy mix utilising input assumptions from DMRE IRP 2019 (not unconstrained least-cost);
DSR – Demand Side Response; DG = Distributed Generation; VRE – variable renewable energy;
NOTE: Energy share is a best estimate based on available data.
Sources: IRP 2019. CSIR Energy Centre analysis
Installed capacity Energy mix
40
0
20
200
140
60
180
120
80
160
100
6.2
0.6
2.1
20
20
3.4
1.5
5.1
1.9
2.1
Total installed capacity (net) [GW]
20
32
4.68.3
17.8
38.0
20
18
30.9
20
30
20
34
20
36
20
38
20
40
20
42
20
44
20
46
20
24
0.3
1.0
20
26
0.512.0
20
28
6.3
7.6
0.7
0.4
32.0
20
48
7.6
1.9
14.4
1.5
6.5
19.2
20
22
8.5
20
50
53.6
82.2
108.8
Other Storage
Biomass/-gas
DG Hydro+PSCSP
WindSolar PV Peaking
Gas
Nuclear (new)
Nuclear
Coal (New)
CoalFirst new-builds:
Wind (2022) 1.6 GWPV (2022) 1.0 GWStorage (2022) 0.5 GWCoal (2023) 0.75 GW Gas (2024) 1.0 GW
Gazetted IRP 2019 Projection1Gazetted IRP 2019Projection1
35
50
300
0
400
450
100
350
150
200
250 6.514.5
3.4
3.3
20
28
20
20
202.8
20
18
20
22
169.1
20
24
20
26
13.4
53.0
0.5
8.21.9
14.9
11.8
2.9
2.2
29.60.9
Electricity production [TWh/yr]
13.4
20
30
20
32
20
34
20
36
20
38
20
40
20
42
20
44
20
46
20
48
20
50
248.0
311.3
0
140
80
40
20
160
60
100
120
180
200
38.030.9
3.41.9
0.7
5.1
8.3
20
48
20
22
20
50
0.5
Total installed capacity (net) [GW]
4.6
1.52.1
6.2
2.1
0.617.8
1.9
20
26
0.4
7.6
20
30
20
32
20
34
20
36
20
40
20
38
20
42
20
44
20
46
20
28
1.0
20
24
53.6
12.06.3
1.5
32.0
20
18
7.6
19.2
0.3
14.4
20
20
6.5
82.2
8.5
108.8
IRP 2019 has also indicated an expected gap between 2018-2022 following which expected new-build options can come online (lead time dependant)Installed capacity and electricity supplied from 2018 to 2050
1 Projection based on optimisation of 2030-2050 energy mix utilising input assumptions from DMRE IRP 2019 (not unconstrained least-cost);
DSR – Demand Side Response; DG = Distributed Generation; VRE – variable renewable energy;
NOTE: Energy share is a best estimate based on available data.
Sources: IRP 2019. CSIR Energy Centre analysis
Installed capacity Energy mix
Biomass/-gas
Other Storage CoalCSPDG
Solar PV Wind
Hydro+PS
Peaking
Gas
Nuclear (new) Coal (New)
NuclearFirst new-builds:
Wind (2022) 1.6 GWPV (2022) 1.0 GWStorage (2022) 0.5 GWCoal (2023) 0.75 GW Gas (2024) 1.0 GW
Gazetted IRP 2019 Projection1 Projection1
Supply gap
Supply gap
Gazetted IRP 2019
36
4
1
0
3
2
5
6
7
8
9
102
02
1
[GW]
20
18
20
19
20
20
20
22
20
23
20
24
20
25
Shortage from IRP 2019 indicating a dominant short-term capacity gap & small energy gap until planned new-build capacity comes online
* Estimated Eskom Demand Response (DR) capability (mostly industrial & energy limited); NOTES: Energy & capacity shortage is demand that cannot be served due to a lack of
capacity (including OCGTs, pumped storage & Eskom DR); Outcomes shown are from deterministic simulations - thus indicative; 99th percentile of capacity & energy shortage is
reported; All IRP 2019 capacity is assumed to come online as planned (Step 3 is always considered implemented); Cost of load shedding is estimated using COUE (cost of unserved
energy) = 87.50 R/kWh; Sources: CSIR Energy Centre analysis
Capacity (shortage) Energy (shortage)
2 000
0
500
1 000
1 500
2 500
3 000
3 500
4 000
4 500
5 000
20
25
20
20
20
19
[GWh/yr]
20
18
20
21
20
22
20
23
20
24
Supply gap (from IRP 2019)
Supply gap (from IRP 2019)
IRP 2019 EAF & IRP 2019 demand forecast(EAF recovery from ≈67% in 2019 to 75.5% by 2024) (Demand forecast immediately growing to 284 TWh by 2025)
Stage 1
Stage 6
Stage 5
Stage 4
Stage 3
Stage 2
Eskom DR*
Loadshedding (actual)
2019: 1352 GWh
(R60-120bln cost to the economy)
NOTE: Understatement of energy shortages
in simulation (conservative estimates)
Stage 7
Stage 8
37
IRP 2019 daily shortage profile shows capacity need in the morning/evening peak combined with daytime energy needs
0
3 000
6 000
9 000
12 000
15 000
18 000
21 000
24 000
27 000
30 000
33 000
36 000
0
1 000
2 000
3 000
4 000
5 000
6 000
7 000
8 000
9 000
10 000
11 000
12 000
13 000
14 000
15 000
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Hour of day
Hourly shortage profile[MW]
Typical Demand profile[MW]
Summer weekend (RHS)
2019
2018
2020 2024
2022
2021
2023
2025
Winter weekday (RHS)
Winter weekend (RHS)
Summer weekday (RHS)
NOTE: Energy & capacity shortage is demand that cannot be served due to a lack of capacity (including OCGTs, pumped storage & Eskom DR); Outcomes shown are from
deterministic simulations - thus indicative; 99th percentile of capacity & energy shortage is reported.
Sources: CSIR Energy Centre analysis
38
0
40
10
30
20
50
60
[GW]
Week 1 Week 2
Representative 2 weeks - constrained/unconstrained showing capacity needs along with daytime energy needs during some weekdaysSimulated hourly generation stack of the total power supply in RSA for 2 weeks in April 2021
Eskom fleet EAF (week) = 64% Eskom fleet EAF (week) = 79%
System Demand
Available Capacity
Unserved Energy
DG
GasStorage
Solar PV
CSP
Wind
Biomass/gas
PS
Hydro
Diesel
Coal
Nuclear
40
0
10
20
30
50
60
20
21
Installed Capacity [GW]
57.0
Energy shortage (daytime)
Capacity shortage (morning/evening)
Pumped storage (pumping mode)
NOTE: Representative week(s) shown but entire time horizon is simulated at hourly resolution in production cost model applying unit commitment and economic dispatch principles;
Sources: CSIR Energy Centre analysis
39
45Mitigation options/solutions5
39What if?4
30Current plans (IRP 2019)3
18The burning platform2
2Executive Summary1
Agenda
40
Not explored
Worst scenario
IRP 2019
(DMRE)
Updated
(CSIR)
Not explored
Best scenario
What if the demand forecast is lower than in IRP 2019 (very likely) and what if the Eskom fleet EAF is lower than in IRP 2019 (also very likely)?
HighEAF [%]
Sources: CSIR Energy Centre analysis
Demand forecast [TWh]
Low
Low
High
41
Would an updated EAF expectation and demand forecast make any difference to capacity and energy shortages?
NOTE: “Updated” scenario is a test scenario developed by CSIR; Demand forecast is based on Eskom MTSAO demand forecast (until 2024) and IRP 2019 growth rates thereafter;
Updated EAF based on MTSAO MES 1 (Low); EAF – Energy Availability Factor
Sources: IRP 2019; MTSAO; CSIR
Demand forecastEAF
2000 2005 2010 2015 2020 2025 2030
80
60
85
0
65
70
90
95
75
100
64.7
Energy Availability Factor (EAF)[%]
67.0
75.5
2000 2005 2010 2015 2020 2025 2030
150
200
50
250
300
350
0
400
100
246
267
306
Electrical energy demand [TWh]
298
284
Updated
IRP 2019
Actual
IRP 2019
Updated
Actual
42
500
3 000
0
1 000
1 500
2 000
2 500
3 500
4 000
5 000
4 500
20
21
20
24
20
23
[GWh/yr]
20
18
20
25
20
19
20
20
20
22
Updated EAF & Updated demand forecast(EAF from ≈67% in 2019 to ≈64% by 2024) (Demand forecast initially flat & growth to 267 TWh by 2025)
6
0
1
2
3
4
5
7
8
9
10
20
23
20
19
20
21
20
18
[GW]
20
20
20
22
20
24
20
25
Updated EAF & demand forecast indicates further shortage relative to IRP 2019 requiring capacity and significantly more energy
Capacity (shortage) Energy (shortage)
Supply gap (from IRP 2019)
Supply gap (from IRP 2019)
* Estimated Eskom Demand Response (DR) capability (mostly industrial & energy limited); NOTES: Energy & capacity shortage is demand that cannot be served due to a lack of
capacity (including OCGTs, pumped storage & Eskom DR); Outcomes shown are from deterministic simulations - thus indicative; 99th percentile of capacity & energy shortage is
reported; All IRP 2019 capacity is assumed to come online as planned (Step 3 is always considered implemented); Cost of load shedding is estimated using COUE (cost of unserved
energy) = 87.50 R/kWh; Sources: CSIR Energy Centre analysis
Loadshedding (actual)
2019: 1352 GWh
(R60-120bln cost to the economy)
NOTE: Understatement of energy shortages
in simulation (conservative estimates)
Stage 1
Stage 6
Stage 5
Stage 4
Stage 3
Stage 2
Eskom DR*
Stage 7
Stage 8
43
Updated scenario shows initial morning/evening capacity need combined with all-day energy, shifting to morning/evening need only in later years
0
3 000
6 000
9 000
12 000
15 000
18 000
21 000
24 000
27 000
30 000
33 000
0
1 000
2 000
3 000
4 000
5 000
6 000
7 000
8 000
9 000
10 000
11 000
12 000
13 000
14 000
15 000
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Hour of day
Hourly shortage profile[MW]
Typical Demand profile[MW]
2018
2019
2022
2021
2020
Winter weekend (RHS)2023
2024
Summer weekend (RHS)2025
Winter weekday (RHS)
Summer weekday (RHS)
NOTE: Energy & capacity shortage is demand that cannot be served due to a lack of capacity (including OCGTs, pumped storage & Eskom DR); Outcomes shown are from
deterministic simulations - thus indicative; 99th percentile of capacity & energy shortage is reported.
Sources: CSIR Energy Centre analysis
44
0
10
20
30
60
50
40
[GW]
Week 1 Week 2
Eskom fleet EAF (week) = 58% Eskom fleet EAF (week) = 63%
System Demand
Available CapacityGas
Unserved Energy Solar PV
Hydro
Storage
Biomass/gas
CSP
DG Wind
PS
Diesel
Coal
Nuclear
0
20
10
40
30
60
50
Installed Capacity [GW]
20
21
57.0
Example – Updated scenario showing significant capacity & energy shortages during representative 2 weeksSimulated hourly generation of the total power supply in RSA for 2 weeks in May 2021
Extensive daily energy shortages
Capacity shortage (morning/evening)Minimal pumped storage
generating/pumping
NOTE: Representative week(s) shown but entire time horizon is simulated at hourly resolution in production cost model applying unit commitment and economic dispatch principles;
Sources: CSIR Energy Centre analysis
45
45Mitigation options/solutions5
39What if?4
30Current plans (IRP 2019)3
18The burning platform2
2Executive Summary1
Agenda
46
Criteria to decide on options to meet the short-term gap should be informed and applied consistently to ensure reasonable cost and timeous delivery
Criteria for choice of short-term options available can ensure a portfolio of options:
Can be supply-side, demand-side and/or storage
Can be delivered in 1-2 years
Will not require extensive procurement process (lead-time)
Can meet capacity (MW) and/or energy needs (MWh)
Can be contracted for 1-3 years (or more if aligned with long-term energy mix)
Ease of implementation (does not require extensive regulatory reform/change)
Aligned with long-term energy mix pathways (technology choices)
Does not require extensive network expansion or augmentation for interconnection
1
2
3
4
5
6
7
8
Sources: CSIR Energy Centre analysis
47
Choice of the suite of solutions/interventions to fill the gap is highly dependant on lead-times and limits resulting options available
Supply-side options
Demand-side options
NOTES: Storage deployment is considered as a demand-side option for simplicity and ease of understanding (is a net user of energy); SSEG could include a range of technologies but
would be dominated by solar PV; Gx – Generation; com. – Commercial; ind. – Industrial; min. – mining; agri. – Agricultural; res. – Residential; DSR – Demand Side Response; 1
Requires adjusted SSEG regulations (proposed lifting licensing requirement for SSEG & only requiring registration with NERSA - from 1 MW to 10 MW (or more)); 2 Does not require
regulatory changes (just communications rollout but could be further incentivised by Eskom/municipalities); 3 Will require Ministerial Determination - generators expected to be >10 MW
(technologies should be aligned with IRP 2019); 4 Will require fast-tracked procurement process; 5 In addition to capacity under construction.
Short-term (2020-2022)
Medium-term/ Long-term
(≥2023)
REIPPPP ‘power-up’200-500 MW (2020)
SSEG1 (com./ind./agri.)400-750 MW/yr
IRP 2019 (first new-build):Wind: 1 600 MW (2022)Solar PV: 1 000 MW (2022)Coal: 750 MW (2023)EG/DG: 500 MW (2023)Gas: 1 000 MW (2024)
Structural loadsheddingStage 1: 1 000 MWStage 2: 2 000 MW
IRP 2019 (first new-build):Storage: 513 MW (2022)
SSEG2 (res.)125-200 MW/yr
DG/EG3 (ind./min.)250-750 MW/yr
Standby Gx aggregator200-1 000 MW/yr
Emergency Gx4
500-2 000 MW/yr
IRP 2019 (by 2030)Wind: 14 400 MWSolar PV: 6 000 MWCoal: 1 500 MWEG/DG: 4 000 MWGas: 3 000 MW
Short-term (2020-2022)
Medium-term/ Long-term
(≥2023)
Storage - SSEG2 (res.)30-100 MW/yr / 90-300 MWh
Storage - SSEG1 (com./ind./agri.)100-375 MW/yr / 300-1125 MWh
Storage (munics)30-120 MW/yr / 90-360 MWh
DSR50-300 MW/yr
IRP 2019 (by 2030)Storage: 2 088 MW
Technical potential5
Prescribed in IRP 2019
Technical potential5
Prescribed in IRP 2019
48
Considering the options available – a customer response at scale would form the only options with fastest lead-times (likely mostly self-dispatched)
• Some solutions can be quickly implemented by customers & at scale (short lead-times)
• Immediate requirement is adjusted SSEG regulations
• Adjusted PPAs with already existing REIPPPP generation capacity
• Communications/incentives/financing for SSEG for further deployment (with requisite Eskom/municipal tariffs)
• DG/EG (ind./min.) expected to be >10 MW each, requires licensing & ministerial approval*
• Options include (from 2020 onwards):
o REIPPPP ‘power-up’: 200-500 MW
o SSEG (com./ind./agri.)1: 400-750 MW/yr
o Storage - SSEG (com./ind./agri.)1: 100-375 MW/yr / 300-1125 MWh
o SSEG (res.)2: 125-200 MW/yr
o Storage - SSEG (res.)2: 30-100 MW/yr / 90-300 MWh
o DG/EG (ind./min.)3: 250-750 MW/yr
o Storage (munics)2,3: 30-120 MW / 90-360 MWh
o Structural loadshedding: Stage 1: 1000 MW; Stage 2: 2000 MW; (or more)NOTES: Gx – Generation; com. – Commercial; ind. – Industrial; min. – mining; agri. – Agricultural; res. – Residential; DSR – Demand Side Response; * Technologies should be aligned
with IRP 2019; DSR (EWHs) could also contribute 50-150 MW/yr in the short-term but has not been explicitly considered as yet due to uncertainty; 1 Requires adjusted SSEG
regulations (proposed lifting licensing requirement for SSEG & only requiring registration with NERSA - from 1 MW to 10 MW (or more)); 2 Does not require regulatory changes (just
communications rollout but could be further incentivised by Eskom/municipalities); 3 Will require Ministerial Determination - generators expected to be >10 MW (technologies should be
aligned with IRP 2019)
Step 1 (customer response at scale)
49
Manage short-term needs with the range of options from Step 1 whilst RfI process runs in Step 2 (many dispatchable options likely to be available)
• DMRE Risk Mitigation Power Purchase Procurement Programme (RMPPPP) RfI
• RfI (and resulting RfP) should assist to fill gap following Step 1 (customer response at scale)
• Reality: Only likely to come online from 2021 onwards
• Capacity providers expected to be >10 MW - licensing & ministerial approval (technologies aligned with IRP 2019)
• Can also include some of the options in Step 1
• Options could include (from 2021 onwards):
o Standby Gx aggregator1: 200-1000 MW/yr
o Emergency Gx2,3: 500-2000 MW/yr
• Structural loadshedding: Stage 3: 3000 MW; Stage 4: 4000 MW
NOTES: Gx – Generation; com. – Commercial; ind. – Industrial; min. – mining; agri. – Agricultural; res. – Residential; DSR – Demand Side Response; 1 Requires adjusted SSEG
regulations (proposed lifting licensing requirement for SSEG & only requiring registration with NERSA - from 1 MW to 10 MW (or more)); 2 Will require Ministerial Determination -
generators expected to be >10 MW (technologies should be aligned with IRP 2019 but could include temporary/permanent engines, OCGTs, cogeneration, imports etc.);3 Will require fast-tracked procurement process.
Step 2 (addressing short-term adequacy and setting up for long-term expected energy mix
50
DMRE RfI is needed & important but is only part of the puzzle and a component of the supply/demand options that will be needed
DMRE Request for Information (RfI)
• Request for Risk Mitigation Power Purchase Procurement (Generation)
• Released: 13 December 2019
• Briefing Session: 8 January 2020
• Response: 31 January 2020
• Specification:
o Quantity: 2000 – 3000 MW
o Can include supply-side and demand-side options
o Lead-time: 3-6 months (2000 MW), 6-12 months (3000 MW)
o PPA tenure: 3 years, 5 years, 10 years, 15 year, 20 years
o Type: “Baseload Energy”, “Peaking Energy”, Mid-Merit Energy”
o Regime: “Dispatchable”, “Self-Dispatchable”
• RfP to follow once RfIs have been assessed
51
Step 3 to setup for the 2020s - immediately run procurement rounds aligned with IRP 2019 & consider reductions based on Step 1 & 2 deployments
• Immediate Ministerial Determinations aligned with IRP 2019
• Begin next procurement rounds for new capacity (lead-times necessitate this)
• Reality: Only likely to come online from 2022 onwards (more likely 2023-2024)
• Options as defined by IRP 2019 (should likely be subtracted from capacity already deployed in Step 1 and Step 2):
o Wind: 1 600 MW (2022) Wind: 14 400 MW (2030)
o Solar PV: 1 000 MW (2022) Solar PV: 6 000 MW (2030)
o Coal: 750 MW (2023) Coal: 1 500 MW (2030)
o EG/DG: 500 MW (2023) EG/DG: 4 000 MW (2030)
o Gas: 1 000 MW (2024) Gas: 3 000 MW (2030)
o Storage: 513 MW (2022) Storage: 2 088 MW (2030)
Step 3 (continued implementation of IRP 2019)
52
50
0
10
20
40
30
60
[GW]
Week 1 Week 2
Eskom fleet EAF (week) = 64% Eskom fleet EAF (week) = 79%
System Demand
Available Capacity
Unserved Energy
Hydro
Storage
Biomass/gasSolar PV
DG
CSP
Wind Coal
Nuclear
PS
Diesel
Gas
50
0
10
40
20
30
60
Installed Capacity [GW]
20
21
59.6
IRP 2019 scenario and roll out of step 1, daytime energy shortages & diesel burn substantially reduced, but load shedding still requiredIRP 2019 with customer response at scale options (Step 1) options: Simulated hourly generation for 2 weeks in 2021
Significantly reduced daytime shortages
Increased ability to utilise pumped storageLower peaking capacity
utilisation (reduced diesel burn)
NOTE: Lower end of total estimated capacity in Step 1 is considered in scenarios); Representative week(s) shown but entire time horizon is simulated at hourly resolution in production
cost model applying unit commitment and economic dispatch principles; Sources: CSIR Energy Centre analysis
53
30
20
0
10
40
50
GW
30
0
40
10
20
50
GW
IRP 2019 comparison with customer response at scale options (Step 1) revealing notably less constrained power systemSimulated hourly generation of the total power supply in RSA for 1 week in April 2021
Mon Tue Wed Thu Fri Sat Sun
Unserved Energy
Storage
CSPDG
Solar PV Wind
Biomass/gas
PS
Hydro
Diesel
Gas
Coal
Nuclear
Mon Tue Wed Thu Fri Sat Sun
IRP 2019 IRP 2019 with customer response at scale (Step 1)
System Demand
Available Capacity
Eskom fleet EAF (week) = 64% Eskom fleet EAF (week) = 64%
NOTE: Lower end of total estimated capacity in Step 1 is considered in scenarios); Representative week(s) shown but entire time horizon is simulated at hourly resolution in production
cost model applying unit commitment and economic dispatch principles; Sources: CSIR Energy Centre analysis
Significantly reduced daytime shortages
54
IRP 2019 scenario (where improved EAF occurs) that is supplemented by customer response (Step 1) will then require mostly capacity (less energy)
What is still needed to ensure system adequacy (utilising DMRE RMPPPP RfI/RfP process)? (Step 2)
• Capacity1: 1.3 GW (2021), 0.1-1.2 GW (2022-2025)
• Capacity factor <3%
NOTE:
• This scenario assumes IRP 2019 EAF & IRP 2019 demand forecast:
o EAF recovery from ≈67% in 2019 to 75.5% by 2024
o Demand forecast immediately growing to 284 TWh by 2025
• If lower customer response at scale (Step 1), additional energy will be needed (higher capacity factor)
• Structural loadshedding (Stage 3 & Stage 4) may also be necessary during 2020-2021
1 It is assumed that capacity from Step 2 will only come online in 2021 (owing to procurement process & technology lead times);
Sources: CSIR Energy Centre analysis
55
10
40
0
30
20
50
GW
10
0
40
30
20
50
GW
IRP 2019 comparison with customer response (Step 1) and DMRE RMPPPP (Step 2) revealing capacity needs met (notably less constrained system)Simulated hourly generation of the total power supply in RSA for 1 week in April 2021
Mon Tue Wed Thu Fri Sat Sun
Unserved Energy
Solar PVStorage
DG CSP
NuclearWind
Biomass/gas
PS
Hydro
DMRE
Diesel
Gas
Coal
Mon Tue Wed Thu Fri Sat Sun
IRP 2019 with customer response at scale (Step 1)
IRP 2019 with customer response (Step 1) and DMRE RMPPPP (Step 2)
System Demand
Available Capacity
Eskom fleet EAF (week) = 64% Eskom fleet EAF (week) = 64%
NOTE: Representative week(s) shown but entire time horizon is simulated at hourly resolution in production cost model applying unit commitment and economic dispatch principles;
Sources: CSIR Energy Centre analysis
56
30
0
40
60
20
10
50
[GW]
Week 1 Week 2
Eskom fleet EAF (week) = 58% Eskom fleet EAF (week) = 63%
System Demand
Available Capacity
Unserved Energy
Storage PS
DG
CSP
Solar PV
Wind Hydro
Biomass/gas
Gas
Diesel Nuclear
Coal
0
10
40
20
30
50
60
Installed Capacity [GW]
59.6
20
21
Reduced daytime shortages but still extensive gap
Only evening capacity shortage
NOTE: Lower end of total estimated capacity in Step 1 is considered in scenarios); Representative week(s) shown but entire time horizon is simulated at hourly resolution in production
cost model applying unit commitment and economic dispatch principles; Sources: CSIR Energy Centre analysis
Updated scenario - with updated EAF & demand, with customer response (Step 1) reduces daytime shortages & diesel burn, but load shedding remainsUpdated scenario with customer response at scale options (Step 1): Simulated hourly generation for 2 weeks in 2021
57
30
10
0
40
20
50
GW
0
40
10
30
20
50
GW
Updated scenario shows significantly less constrained system with the roll out of customer response at scale (Step 1) Simulated hourly generation of the total power supply in RSA for 1 week in 2021
Mon Tue Wed Thu Fri Sat Sun
DGUnserved Energy
Solar PVStorage
CSP
Wind
Biomass/gas
PS
Hydro
Diesel
Gas
Coal
Nuclear
Mon Tue Wed Thu Fri Sat Sun
Updated scenario Updated scenario with customer response at scale (Step 1)
System Demand
Available Capacity
Eskom fleet EAF (week) = 58% Eskom fleet EAF (week) = 58%
NOTE: Lower end of total estimated capacity in Step 1 is considered in scenarios); Representative week(s) shown but entire time horizon is simulated at hourly resolution in production
cost model applying unit commitment and economic dispatch principles; Sources: CSIR Energy Centre analysis
Extensive daily energy shortages
Capacity shortage (morning/evening)
58
Updated scenario will require significant additional capacity & energy through to 2025 in addition to customer response at scale (Step 1)
Now… What is still needed to ensure system adequacy (utilising DMRE RMPPPP RfI/RfP process)? (Step 2)
• Capacity1: 2.8 GW (2021), 1.9 GW (2022) and 1.6-3.9 GW (2023-2025)
• Capacity factor 8% (2021), 10% (2022), 1-3% (2023-2025)
NOTE:
• This scenario assumes an Updated EAF & demand forecast:
o EAF from ≈67% in 2019 to ≈65% by 2025
o Demand forecast initially flat & growth to 267 TWh by 2025
• If lower customer response at scale (Step 1), additional energy will be needed (higher capacity factor)
• Structural low-level loadshedding (Stage 3 & Stage 4) may also be necessary during 2020-2021
1 It is assumed that capacity from Step 2 will only come online in 2021 (owing to procurement process & technology lead times);
Sources: CSIR Energy Centre analysis
59
30
20
0
10
40
50
GW
0
10
40
30
20
50
GW
Mon Tue Wed Thu Fri Sat Sun
Storage
Unserved Energy DieselHydroDG CSP
Solar PV Wind
CoalBiomass/gas
PS GasDMRE Nuclear
Mon Tue Wed Thu Fri Sat Sun
Updated scenario with customer response (Step 1) and DMRE RMPPPP (Step 2)
Eskom fleet EAF (week) = 58%
Updated scenario with customer response at scale (Step 1)
Eskom fleet EAF (week) = 58%
System Demand
Available Capacity
DMRE RMPPPP providing peaking/mid-merit capacity
Updated scenario comparison of customer response (Step 1) and additional DMRE RMPPPP (Step 2) revealing notably less constrained power systemSimulated hourly generation of the total power supply in RSA for 1 week in 2021
NOTE: Lower end of total estimated capacity in Step 1 is considered in scenarios); Representative week(s) shown but entire time horizon is simulated at hourly resolution in production
cost model applying unit commitment and economic dispatch principles; Sources: CSIR Energy Centre analysis
Capacity shortage (morning/evening)
60
Thank you
61
References
Loadshedding data
• Eskom twitter account: https://twitter.com/Eskom_SA
• ESP (Twitter app): https://sepush.co.za/
• National Energy Regulator of South Africa (NERSA). (2008). Inquiry Into the National Electricity Supply Shortage and Load Shedding. Retrieved from http://pmg-assets.s3-website-eu-west-1.amazonaws.com/docs/080528nersa.pdf
COUE
• Department of Energy (DoE). (2019). Integrated Resource Plan (IRP 2019). Retrieved from http://www.greengazette.co.za/pages/national-gazette-37230-of-17-january-2014-vol-583_20140117-GGN-37230-003
Eskom fleet Energy Availability Factor (EAF)
• Eskom Annual Reports: http://www.eskom.co.za/OurCompany/Investors/IntegratedReports
• MTSAO: Fabricius, C., Sigwebela, N., Damba, S., Mdhluli, S., & Rambau, P. (2019). MEDIUM-TERM SYSTEM ADEQUACY OUTLOOK. Retrieved from http://www.eskom.co.za/Whatweredoing/SupplyStatus/Documents/MediumTermSystemAdequacyOutlook2019.pdf
• Eskom System Adequacy Reports: http://www.eskom.co.za/Whatweredoing/SupplyStatus
• Department of Energy (DoE). (2019). Integrated Resource Plan (IRP 2019). Retrieved from http://www.greengazette.co.za/pages/national-gazette-37230-of-17-january-2014-vol-583_20140117-GGN-37230-003
• Department of Energy (DoE). (2018). INTEGRATED RESOURCE PLAN 2018. Retrieved from http://www.energy.gov.za/IRP/irp-update-draft-report2018/IRP-Update-2018-Draft-for-Comments.pdf
62
References
Capacity and energy mix to 2030 (IRP 2019)
• Department of Energy (DoE). (2019). Integrated Resource Plan (IRP 2019). Retrieved from http://www.greengazette.co.za/pages/national-gazette-37230-of-17-january-2014-vol-583_20140117-GGN-37230-003
DMRE Risk Mitigation Power Purchase Procurement Programme (RMPPPP) RfI
• Department of Mineral Resources and Energy (DMRE). (2019). REQUEST FOR INFORMATION IN RESPECT OF THE DESIGN OF A RISK MITIGATION POWER PROCUREMENT PROGRAMME. Retrieved from http://www.energy.gov.za/files/docs/MTPPP%20Generation%20-%20DMRE%20Request%20For%20Infromation%20(13%20December%202019).pdf
63
Disclaimer
© CSIR 2020. All rights to the intellectual property and/or contents of this document remain vested in the CSIR.
This document is issued for the sole purpose for which it is supplied. CSIR will take no responsibility or incur any
liability for the use of the information contained herein if used for any other purpose. No part of this publication
may be reproduced, stored in a retrieval system or transmitted, in any form or by means electronic, mechanical,
photocopying, recording or otherwise without the express written permission of the CSIR. It may also not be
resold, hired out or otherwise disposed of by way of trade in any form of binding or cover than that in which it is
published.
top related