Least-cost options for integrating intermittent …...2016/01/22 · Least-cost options for...

Copernicus Institute of Sustainable Development


Least-cost options for integrating intermittent renewables

Anne Sjoerd Brouwer 22-01-2016


Introduction

• Stricter climate policies increasingly likely

• Power sector in a state of change More intermittent renewables: wind and solar PV

Large CO2 emission reductions: only a role for low-carbon power plants

• Effect of intermittent renewables?

• Role of emerging technologies? Carbon Capture and Storage (CCS)

Demand response

Increased interconnection capacity

Electricity storage


Structure

• 1. Background on intermittent RES (iRES)

• 2. Methods

• 3. Results

• 4. Conclusions


Background on intermittent RES

• Intermittent RES important component of low-carbon power systems

• Specific properties1. Intermittent

2. Partly unpredictable

3. Location specific

• Affect the whole power system Operation

Economics

0

10

20

30

40

50

60

70

80

90

100

0

50

100

150

200

250

300

350

400

450

500

ma 17 jun di 18 jun wo 19 jun do 20 jun vr 21 jun za 22 jun zo 23 jun

Ele

ctri

city

Pri

ce (

€/M

Wh

)

Ge

ne

rati

on

(G

W)

Generation during week with minimum residual demand (6.4 GW)

Nuclear Geothermal Wind Offshore Wind OnshoreHydro Solar PV NGCC-CCS NGCCBiothermal Pumped Hydro Gas Turbine Demand ResponseElectricity Price (LWA)


Methods

• Some technologies match better with intermittent renewables

• Goal of our research: identify which technologies may be part of a power system with:

High reliability (LOLP <0.1 day/year)

Low emissions (>96% reduction CO2 emissions)

Low costs

• Power system with high shares of RES and iRES

• Consider a future power system…


Methods

• … in Europe in the year 2050

• Six-node power system simulated with PLEXOS

• Three scenarios evaluated

• Quantify system costs

FR

GE

IT

IB

SC

BR

Scandinavia (SC)

Norway, Sweden, Denmark

Germany & Benelux (GE)

Germany, The Netherlands, Belgium, Luxembourg

Italy & Alpine States (IT)

Italy, Austria, Switzerland

Iberian Peninsula (IB)

Spain, Portugal

France (FR)

France

British Isles (BR)

United Kingdom, Ireland

RES 40% 60% 80%

iRES 22% 41% 59%


1. Define non-fossilcapacity per scenario

2. Define capacities of complementary options

3. Optimize fossil capacityper scenario [PLEXOS]

4. Run hourlysimulations [PLEXOS]

30%20%

10%

6%

9%13%

9%9%

9%

2%2%

2%

6%10%

14%

11% 20% 28%6%

10%14%30%

20%10%

0%

20%

40%

60%

80%

100%

40% RES 60% RES 80% RES

Shar

e o

f an

nu

al p

ow

er g

ener

atio

n (

%)

Generation capacity per scenario

(Fossil capacity)

Solar PV

Wind onshore

Wind offshore

Geothermal

Hydro

Biomass

Nuclear

Study methods

Power generation per scenario


Results – power generation

• Two aspects can reduce total system costs

Least-cost power generation

Efficient system operation

• First: what is the cheapest way to generate power in low-carbon power systems?



• Cheapest technology to generate low-carbon power?

0

10

20

30

40

50

60

70

80

90

100

Elec

tric

ity

sto

rage

ch

argi

ng

cost

/

Elec

tric

ity

pri

ce (

€/M

Wh

)

Capacity factor (%)

Generation option with the lowest LCOE - 70 €/tCO2

NGCC

GT

NGCC-CCS

PC-CCS

DR-Shift

DR-Shed

PHS

CAES

NGCC-CCS

DR

-Sh

ed

NGCC

DR

-Sh

ift

Co

al-C

CS

0

10

20

30

40

50

60

70

80

90

100

Elec

tric

ity

sto

rage

ch

argi

ng

cost

/

Elec

tric

ity

pri

ce (

€/M

Wh

)

Capacity factor (%)


NGCC

GT

NGCC-CCS

PC-CCS

DR-Shift

DR-Shed

PHS

CAES

NGCC-CCS2. GT

1. D

R -

She

d

NGCC

1. D

R -

Shif

t

Co

al-C

CS

At low capacity factors, DR is cheapest, butits potential is limited. GT is the next cheapest

Demand Response has limited potential

Storage can be attractive at low charging costs

0

10

20

30

40

50

60

70

80

90

100

Elec

tric

ity

sto

rage

ch

argi

ng

cost

/

Elec

tric

ity

pri

ce (

€/M

Wh

)

Capacity factor (%)


NGCC

GT

NGCC-CCS

PC-CCS

DR-Shift

DR-Shed

PHS

CAES

Compressed Air Energy Storage

NGCC-CCS2. GT

1. D

R -

She

d Pumped Hydro Storage

NGCC

1. D

R -

Shif

t

Co

al-C

CS


0

10

20

30

40

50

60

70

80

90

100

Elec

tric

ity

sto

rage

ch

argi

ng

cost

/

Elec

tric

ity

pri

ce (

€/M

Wh

)

Capacity factor (%)

Generation option with the lowest LCOE - 70 €/tCO2 NGCC

GT

NGCC-CCS

PC-CCS

DR-Shift

DR-Shed

PHS

CAES

Electricityprice

Compressed Air Energy Storage

NGCC-CCS2. GT

1. D

R -

She

d

NGCC

The dotted line shows the typical charging cost (p10 of 60% RES electricity price)

1. D

R -

Shif

t

Co

al-C

CS


Not realizable for electricity storage due to charging

constraints

Electricity storage charging costs are too high


0

200

400

600

800

1000

1200

1400

ExistingCapacity(2012)

40% RES(22% iRES)

60% RES(41% iRES)

80% RES(59% iRES)

Tota

l In

stal

led

Cap

acit

y (G

W)

OilCoalDemand ResponseGas TurbineNGCCNGCC+CCSSolar PVWind OnshoreWind OffshorePumped HydroGeothermalHydroBiothermalNuclear

Installed capacity GenerationInstalled capacity Generation

0

500

1000

1500

2000

2500

3000

3500

17% RES(0% iRES)

40% RES(22% iRES)

60% RES(41% iRES)

80% RES(59% iRES)

Po

we

r ge

ne

rati

on

(TW

h/y

r)

0

500

1000

1500

2000

2500

3000

3500

17% RES(0% iRES)

40% RES(22% iRES)

60% RES(41% iRES)

80% RES(59% iRES)

Po

we

r ge

ne

rati

on

(TW

h/y

r)

• Fossil capacity optimization: only natural gas capacity

Combined cycle with CCS

Gas turbines



• NGCC-CCS generates power during the night in the summer

0

10

20

30

40

50

60

70

80

90

100

0

50

100

150

200

250

300

350

400

450

500

ma 17 jun di 18 jun wo 19 jun do 20 jun vr 21 jun za 22 jun zo 23 jun

Ele

ctri

city

Pri

ce (

€/M

Wh

)

Ge

ne

rati

on

(G

W)

Generation during week with minimum residual demand (6.4 GW)




• NGCC-CCS baseload generation during winter time

• Gas turbines supply peak demand

0

10

20

30

40

50

60

70

80

90

100

0

50

100

150

200

250

300

350

400

450

500

Mon 14 Dec Tue 15 Dec Wed 16 Dec Thu 17 Dec Fri 18 Dec Sat 19 Dec Sun 20 Dec

Ele

ctri

city

Pri

ce (

€/M

Wh

)

Ge

ne

rati

on

(G

W)

Generation during week with maximum residual demand (388 GW)




0

10

20

30

40

50

60

70

80

90

100

0

50

100

150

200

250

300

40% RES(22% iRES)

60% RES(41% iRES)

80% RES(59% iRES)

Ele

ctri

city

co

st o

r p

rice

(€

/MW

h)

An

nu

al s

yste

m c

ost

s (

bn

€/y

r)

Total system costs per scenario

Operational Costs (bn€/yr)

Fuel Costs (bn€/yr)

FO&M Costs (bn€/yr)

Invest DR (bn€/yr)

Invest NGCC/GT (bn€/yr)

Invest Solar PV (bn€/yr)

Invest Wind (bn€/yr)

Invest NGCC-CCS (bn€/yr)

Fixed scenario costs (bn€/yr)

Avg. electricity price (€/MWh)

Avg. generation cost (€/MWh)

• Effect on costs: intermittent renewables increase total system costs

12% higher capital costs

18% reduction in electricity price



• CO2 emission reduction target is met in all scenarios

Specific emissions of <13g/kWh correspond to a >96% reduction in CO2 emissions

0

5

10

15

20

25

30

0

50

100

150

200

250

300

40% RES 60% RES 80% RES

Spec

ific

CO

2 e

mis

sio

ns

(g/k

Wh

)

CO

₂ em

itte

d /

sto

red

(M

tCO

₂)

CO2 emissions and storage per scenario

CO₂ storedGas TurbineNGCCNGCC-CCSSpecific emissions



Results – system operation

• Which options can decrease system costs by improving “system efficiency”?

More efficient use of power plants

Less curtailment


225

230

235

240

245

250

255

260

265

Current(37 GW)

Min(86 GW)

Low(123 GW)

Med*(189 GW)

High(257 GW)

Max(349 GW)

Tota

l Sys

tem

Co

sts

(€b

n/y

)

Interconnection Capacity

Effect of CAES and Interconnection Capacity on Total System Costs

95 GW CAES

0 GW CAES

95 GW CAES

0 GW CAES

95 GW CAES

0 GW CAES

80%RES

60%RES

40%RES


Interconnects reduce costs up to a ‘sweet spot’

Storage is expensive

225

230

235

240

245

250

255

260

265

Current(37 GW)

Min(86 GW)

Low(123 GW)

Med*(189 GW)

High(257 GW)

Max(349 GW)

Tota

l Sys

tem

Co

sts

(€b

n/y

)

Interconnection Capacity

Effect of CAES and Interconnection Capacity on Total System Costs

95 GW CAES

47 GW CAES

24 GW CAES

5 GW CAES

0 GW CAES

60% RES, 95 GW CAES

60% RES, 47 GW CAES

60% RES, 24 GW CAES

60% RES, 5 GW CAES

60% RES, 0 GW CAES*

40% RES, 95 GW CAES

40% RES, 47 GW CAES

40% RES, 24 GW CAES

40% RES, 5 GW CAES

40% RES, 0 GW CAES*

95 GW CAES

0 GW CAES

95 GW CAES

0 GW CAES

95 GW CAES

0 GW CAES



• Demand response also reduces costs by 2-3%

Effect stronger at high RES penetration

-

50

100

150

200

250

300N

o D

R

23

GW

DR

*47

GW

DR

95

GW

DR

No

DR

23

GW

DR

*47

GW

DR

95

GW

DR

No

DR

23

GW

DR

*47

GW

DR

95

GW

DR

40% RES 60% RES 80% RES

Tota

l Sys

tem

Co

st (

€B

n/y

)

Effect of DR on total system costsGeneration Costs Fixed Costs


Conclusions

• iRES affect the operation of power systems, but their impacts are manageable.

Larger reserves, effect on capacity factors of other generators

• Low-carbon power systems can be realized with various generation portfolios

iRES increase total system costs

Natural-gas fired generation important in all scenarios

DR & interconnections least-cost options

Electricity storage too expensive


• Future power systems will become increasingly complex

New technologies

More decentralized generation

Cross-sectoral integration

More uncertainty for investors

• Necessary investments in power plants will not be made with the current energy-only market design

Business cases are unsound.

Generation adequacy may become a key issues of future power systems

The bigger picture


Thank you for your attention

Any questions?

Anne Sjoerd [email protected]

Will Zappa Machteld van den Broek

[email protected] [email protected]

Anne Sjoerd Brouwer, Machteld van den Broek, William Zappa, Wim C. Turkenburg, André Faaij. Least-cost options for integrating intermittent renewables in low-carbon power systems. Applied Energy 161, pp 48-74.(2016)

mailto:[email protected]




6. Supplementary slides


Overview of method

Input data Method Results

1. Low-carbon scenarios from previous studies

2. Projected load patterns

4. Projected power plant flexibility parameters

6. Projected power plant techno economic data

5. Projected balancing reserve patterns

3. Projected iRES production patterns

3. Run PLEXOS LT plan to

optimize fossil generation capacity

The full generation mix is

now defined

8. Projected demand response potential and costs

9. Projected techno-economic specifications of electricity storage

7. Projected interconnection capacity

2. Define capacities of complementary

technologies (excluding fossil)

1. Define plausible non-fossil

generation scenarios

4 Run PLEXOS MT and ST

schedules to optimize hourly

unit commitment and economic

dispatch model for western Europe

6. iRES integration costs: - Balancing costs - Profile costs

3. Costs and CO2 emissions per scenario

2. Overview of full generation mixes

7. Profitability of generators

8. Sensitivity analyses

4. Effect of demand response

5. Effect of electricity storage and inter-connection capacity

1. Comparison of complementary technologies


Key input parameters

€/GJ Coal Natural gas

Uranium Biomass

Fuel price 1.7 6.5 1 7.2

€/tCO2 Transport and storage

CO2 costs 13.5

TWh/yr Britain Scandi-navia

France Ger-many+

Iberian pen.

Italy+

Load 415 368 602 811 358 525


Input parameters of generation technologies

Generator Investment (TCR, €/kW)

Fixed O&M (€/kW/yr)

Variable O&M (€/MWh)

Efficiency Remarks

Nuclear 4841 103 1 33%

PC-CCS 2847 33 5.6 41% 90% CO2 capture

NGCC-CCS 1349 15 2.1 56% 90% CO2 capture

NGCC 902 11 1.2 63%

Biothermal 1949 37 3 45% 100% biomass fired

Geothermal 2657 44 0

Hydro 3037 52 0

Wind onshore 1402 37 0 CF: 22-26%

Wind offshore 2655 83 0 CF: 40-43%

Solar PV 700 17 0 CF: 11-21%

Gas turbine 438 10 0.8 42%

Pumped hydro 2079 58 0.23 80% 80% round trip η

Demand resp. 3-100 1-11 0 100% 1-2 hrs of “storage”


Demand response potential

0 2000 4000 6000 8000 100001200014000

British Isles

France

Germany & Benelux

Scandinavia

Iberian peninsula

Italy and Alpine states

DR Potential (MW)

Demand Response potential used in this studyShed - Electrolysis

Shed- Chloralkali

Shed - Paper

Shed - Steel

Shed - Cement

Shift - 2h by 1h

Shift - 2h by 2h

Shift - Airconditioning

Shift - Washing machines / dryers

Shift - Fridge and freezer

Shift - Space/water heating


Detailed total system costs

0

10

20

30

40

50

60

70

80

90

100

0

50

100

150

200

250

300

17% RES(0% iRES)

40% RES(22% iRES)

60% RES(41% iRES)

80% RES(59% iRES)

Elec

tric

ity

cost

or

pri

ce (

€/M

Wh

)

An

nu

al s

yste

m c

ost

s (

bn

€/y

r)

Interconnection Cost (bn€/yr)

DR shedding cost (bn€/yr)

CO₂ Costs [tax/storage] (bn€/yr)

Start/stop costs (bn€/yr)

VO&M Costs (bn€/yr)

Fuel Costs (bn€/yr)

FO&M Costs (bn€/yr)

Invest DR (bn€/yr)

Invest NGCC/GT (bn€/yr)

Invest Solar PV (bn€/yr)

Invest Wind (bn€/yr)

Invest NGCC-CCS (bn€/yr)

Invest RES (bn€/yr)

Invest Nuclear (bn€/yr)

Reserve costs (bn€/yr)

Avg. electricity price (€/MWh)

Avg. generation cost (€/MWh)


Capacity factors per month


2. Methods

• Interconnection capacity

Deployment based on previous studies

Costs of 28 k€/MW/yr

• Demand response

Potential based on other studies (11 categories)

Crude costs of 200-5000 €/MWh (shedding) and 2-100€/kw (shifting)

0 2 4 6 8 10 12 14 16


Iberian peninsula

Scandinavia

Germany & Benelux

France

British Isles

DR capacity (GW)

DR capacity per region

Load shedding

Load shifting

0 25 50 75 100 125 150


Iberian peninsula

Scandinavia

Germany & Benelux

France

British Isles

Interconnection capacity (GW)

Interconnection capacity per region

Current (37 GW)Minimum (86 GW)Low (123 GW)Medium (189 GW, default)High (257 GW)Maximum (349 GW)

1. Define non-fossil capacity per scenario


3. Optimize fossil capacity per scenario [PLEXOS]

4. Run hourly simulations [PLEXOS]

0

2,000

4,000

6,000

8,000

10,000

12,000

Pumpedhydro

storage

Compressedair storage

NaS-battery Vanadiumflow battery

Li-ionbattery

Inve

stm

ent

cost

s (€

/kW

)

Electricity storage investment costs

Year 2015

Year 2050

Storagecapacity: 8 hours

-2% -39% -65% -69%-71%


2. Methods

• Two optimizations with PLEXOS

Power system simulation and optimization tool

• Optimization of fossil capacity

IEA projections of specifications

6 generator types

• Hourly simulations

Chronological simulations with 8760 steps

Account for flexibility constraints

Auxiliary reserves also included

Curtailment of RES allowed

1. Define non-fossil capacity per scenario


3. Optimize fossil capacity per scenario [PLEXOS]

4. Run hourly simulations [PLEXOS]


3. Results – power generation

• Total system costs are also increased by the integration costs of intermittent RES:

Profile costs

Balancing costs, grid costs

0

5

10

15

20

25

0% 10% 20% 30% 40% 50% 60%

Pro

file

Co

sts

(€/M

Wh

iRES

)

iRES Penetration (%)

Profile integration costs of intermittent renewables

Residual System Utilization Costs* Curtailment Costs Profile Costs


(3. Results – economics)-3

3 23

-23

-21

-17

36

40

53

27

32

41

32

41

54

19

22

30

50

60

30

16

14

15

15

7

10

1

68

10

04

11

62

44

9

26

4

12

0

60

5

-2

-13

11

62

19

4

63

-200

-150

-100

-50

0

50

100

150

200

40%

60%

80%

40%

60%

80%

40%

60%

80%

40%

60%

80%

40%

60%

80%

40%

60%

80%

40%

60%

80%

40%

60%

80%

40%

60%

80%

40%

60%

80%

40%

60%

80%

40%

60%

80%

40%

60%

80%

Hydro GeothermalWind

OffshoreWind

Onshore Solar PV Nuclear NGCC NGCC+CCS Biothermal Gas TurbinePumped

HydroDemand

ResponseCAES

(4.7GW)

Req

uir

ed C

apac

ity

Pay

men

t (€

/(kW

.y))

40% RES 60% RES 80% RES Capacity Factor

Cap

acit

yFa

cto

r (%

)

0

100

75

50

25


(4. Sensitivities)

• Variations in input parameters are explored, as the input parameters determine the outcome of the study.

• Four tornado graphs depict the effect of variation in input parameters on:

Total system costs

Fuel use

Generation capacities

Electricity generation per generator type


(4. Sensitivities – key findings)

• Results are overall robust. Results are onlyconsiderably affected by:

High gas price (7.8 €/GJ) shift of NGCC-CCS to PC-CCS

Cheaper biomass (5.5 €/GJ) biomass early in merit order

Higher CO2 cap (180 Mt) shift of NGCC-CCS to NGCC

Lower investment costs of iRES lower system costs

• More flexibility in systems resulting frominterconnections, DR or CAES leads to

More base-load generation

Less GT capacity being installed and used

• Investment costs of iRES are a key factor.

Reduction in iRES investment costs has a large impact

Overall iRES cost reduction of >31% required to make 80% RES the cheapest scenario


4. Discussion – scope, assumptions

• This study only considers snapshots of possible future power systems in 2050. No consideration is given to the dynamic transition from the current system

• Starting points of this study are pre-set emission reduction and reliability targets.

• Heating (demand, generation, storage) is not included.

• No transmission constraints are simulated within regions

• Assumed properties of DR capacity

49 GW of DR potential

Shedding costs in this study: 200-5000€/MWh

Shifting investment cost in this study: 2-100 €/kW


4. Discussion - caveats• The scenario approach fixes 50-65% of the costs exogenously leaving

less room for optimization. Thus systems are plausible, not necessarily optimum. This may be reflected in the costs.

• Capacity credits of iRES are fixed. This assumption:

Underestimates the benefits of interconnections (interconnectors can increase the capacity credit by spatial smoothing);

Underestimates the profile costs in the 80% RES scenario compared to the 40% RES scenario (capacity credits decrease with higher iREScapacity, requiring more firm capacity with low capacity factors).

• Significant uncertainties remain about the potential and costs of DR

• The model does not include specialized VOLL-values, or a detailed representation of super-peak generators (e.g. backup generators). These could improve the profitability of power plants by causing price spikes, which lead to big profits in a small amount of time.

Least-cost options for integrating intermittent …...2016/01/22 · Least-cost options for...

Documents

Transcript of Least-cost options for integrating intermittent …...2016/01/22 · Least-cost options for...