plan.be

The challenge of running 100% renewable energy scenarios

Methodological issues

2nd Ghent Summer SchoolAugust 28, 2014

D. Devogelaer, FPB

plan.be

100% what?

• Mission from 4 energy ministers in 2011:

• 1 federal, 3 regional ministers of Energy

• Concerns on climate, economy and SoS• Time frame, target and consortium fixed:

2050 - 100% - FPB (fed), VITO (Fl), ICEDD (Wl)

manoeuver within that framework

plan.be

“You say you want a [renewable] revolution…

We'd all love to see the plan” The Beatles, White Album,

1968

Relatively easy to calculate number of windmills, solar panels, ... required, less straightforward to have them all available when needed

(V)LT model Scenario analysis

How can we answer a question like this?

Which model?

Which scenarios

?

Which output?

Which input?

plan.be

The model

plan.be

« En économie, un modèle est une représentation simplifiée de la réalité économique ou d'une partie de celle-ci. Comme dans les autres disciplines scientifiques, les modèles économiques utilisent le formalisme mathématique qui permet de représenter le modèle sous forme d'équations. Outre ces équations, les modèles empiriques sont constitués d’une banque de données propre et généralement d’un jeu de paramètres. « 

Source: FPB.

• a description of a system using mathematical concepts and language

• used not only in natural sciences (e.g. physics) and engineering disciplines (e.g. computer science, artificial intelligence), but also in social sciences (e.g. economics)

• may help to explain a system and to study the effects of different components, but also to make projections about future behaviour

• basically, a set of interrelated equations that can assign results either to a variable or to a dimension value

Model: What(’s in a name)?

plan.be

• What is your research question? • Which models are apt to answer the question?

100% study: Model TIMES

1. National energy system2. VITO partner in project3. Quick feedback loops4. Experience with time horizon 2050

How do you define the choice of the model?

plan.be

• Partial equilibrium (energy) model• Bottom-up optimisation model of the national

energy system • Detailed representation of energy-material flows

and technologies (broad sense) • Various alternative technological choices • Up to 2050• Driving factor: fulfilment of energy service

demand (≠ energy demand)

Basic principles of TIMES model

plan.be

Challenge: Dealing with variable renewable sources daily and seasonal fluctuations

Click icon to add chart

Source: Elia.

plan.be

1. Cope with uncertainty of power supply

• Extending the temporal resolution to 78 periods in one year = 26 periods of two weeks x 3 periods a day

• Reserve capacity requirement (sum of nominal power of biomass plants, geothermal and storage facilities)

• Constraint to assure that BE can be self sustained for 14 consecutive days without counting on wind and solar

2. Day-night and seasonal electricity storage options

3. Alternative solutions to increase system’s flexibility• Overproduction - grid disconnection curtailment• Endogenous steel production timing (not ‘just in time’)

4. Endogenous transmission and distribution network

Major model improvements for dealing with variable character of renewable energy

plan.be

The input

plan.be

How do you go about? • Literature review: problem of finding exactly what you

need

• Geographical scope

• Time frame

• “Old” data

• Coherence

• …

• Stakeholders: want to have their say, can deliver valuable input, but

• Often not familiar with model specificities

• Often not aware of time lags of model

• One model cannot solve all

• Expert judgment: more difficult for references/objectivity

Define assumptions

plan.be

You will get attacked on your assumptions… always!

• Because it is no exact science e.g. oil price projections

• Because by the time your study gets published, assumptions can be out-dated -> the curse of the modeller

• By pressure groups, lobbyists, but also peers

Solution: Perform sensitivity analyses to test robustness of results

Often the definition of assumptions is an exercise on its own and can give rise to new studies!

Define assumptions (II)

plan.be

Background information

Surface: 30.000 km2

Population: 11 million (330p/km2)GDP: 350 billion €Final energy consumption: 1800 PJ Per capita final energy consumption:

150% EU27 average or 75% USHydro: limited to 120 MW Domestic fossil energy supply = 0

plan.be

• Belgian GDP: increases at an AAGR of 1.8% in 2010-2050 • Fuel prices: Energy roadmap 2050, CPI, crude oil to some 127

$’08/bbl in 2050

• Carbon price: Energy roadmap 2050, CPI, 15 €/tCO2 in 2020, 51 €/tCO2 in 2050

• CCS technologies: not allowed• Coal: no investments in new coal fired PP’s• Nuclear: current legislation on the phasing out of nuclear PP’s• Electricity imports: limited to 5.8 TWh (average Belgian net

imports 2003-2010)

Assumptions

plan.be

Assumptions on renewable costs

plan.be

Assumptions on electricity storage costs

plan.be

The scenarios

plan.be

Definition of scenarios REF Fossil Benchmark scenario

plan.be

The output

plan.be

Results for REF

plan.be

While presenting the results to the stakeholder committee, they noticed that costs for REF were relatively low -> in the search for an answer, it came about that investments in coal fired power plants were still allowed

• Relatively low prices in 2050 due to lack of oil indexation

• No GHG target so not penalised

Changed that, so no new coal investments in power generation

Analysis of results

plan.be

Results for REN scenarios

plan.be

• Model has tendency to postpone investments: bulk of investments during the final decade

• Makes economical sense, but from a societal point of view non-sense

• Imposition of targets:35% of primary energy in 2030, 65% in 2040, 100% in 2050

Analysis of results

plan.be

Energy mix: Primary energy, 2050

Source: TIMES.

REF DEM GRID BIO PV WIND

0

200

400

600

800

1000

1200

1400

1600

1800

Solar Bioenergy (domestic and import)Total ambient heat (air + ground) HydroWind onshore Wind offshoreWind offshore, non-Belgian territory Electricity - import

PJ

plan.be

Energy mix: Final energy, 2050

Click icon to add chart

REF DEM GRID BIO PV WIND

0

200

400

600

800

1000

1200

1400

1600

Electricity Biomass Hydrogen Heat (air & ground & direct) Coal Gas Oil

PJ

Source: TIMES.

plan.be

REF DEM GRID BIO PV WIND0

50

100

150

200

250

300

350

Biomass (incl. CHP)GeothermalHydroPVWind offshoreWind onshoreGasWind offshore - non Belgian territoryImported electricity (other)Of which excess electricity

TW

h

REF DEM GRID BIO PV WIND0

50

100

150

200

250

Biomass (incl. CHP)GeothermalHydroPVWind offshoreWind onshoreGasWind offshore - non Belgian territory

GW

e

Source: TIMES.

Energy mix: Power generation, 2050

plan.be

Energy mix: Power generation capacities (MW), 2020-2050

DEM GRID PVBIO WIND

plan.be

Storage capacities: Electricity (GWh), 2020-2050

DEM GRID PVBIO WIND

plan.be

Energy mix: Energy flows in PV scenario (PJ), 2050

Source: http://www.emis.vito.be/artikel/naar-100-hernieuwbare-energie-belgi%C3%AB-tegen-2050-video

plan.be

Costs: Energy system costs (M€2005), 2050

Source: TIMES.

REF DEM GRID BIO PV WIND

0

20000

40000

60000

80000

100000

120000

Variable costsInvestment and fixed costsDemand reductions

REF DEM GRID BIO PV WIND

-20000

-10000

0

10000

20000

30000

40000

50000

Variable costsInvestment and fixed costsDemand reductions

plan.be

Costs: Additional cost wrt REF (% of GDP), 2050

DEM GRID BIO PV WIND

0.0%

0.5%

1.0%

1.5%

2.0%

2.5%

3.0%

3.5%

4.0%

4.5%

5.0%

Additional total cost Additional energy system cost

Source: TIMES.

plan.be

Costs: Additional cost incl. avoided GHG damage cost (M€2005), 2050

• Total annual add. cost wrt REF, when (global) benefit of avoided GHG in 2050 is included

• Lord Stern@Davos: ‘I got it wrong on climate change – it's far, far worse’ The Observer, Jan 26, 2013 DEM GRID BIO PV WIND

-15000

-10000

-5000

0

5000

10000

15000

20000

Low case CO2 damage (130 €/ton)

High case CO2 damage (300 €/ton)

No longer a cost...

... but a benefit

plan.be

Costs: Additional investments wrt REF (M€2005)

DEM GRID BIO PV WIND

-50000

0

50000

100000

150000

200000

250000

300000

350000

400000

450000

Agriculture CHP Commercial Conversion

Electricity Industry Other sectors Residential

Transport

DEM GRID BIO PV WIND

-2000

3000

8000

13000

18000

23000

28000

Agriculture CHP Commercial Conversion

Electricity Industry Other sectors Residential

Transport

Cumulative for 2013-2050

Investments in 2050

plan.be

Costs: Cumulative additional investment expenditures in the electricity sector (M€2005)

DEM GRID BIO PV WIND

0

50000

100000

150000

200000

250000

300000

Conventional Geothermal Grid expansion Others Solar Storage Wind

Source: TIMES.

Cumulative for 2013-2050

plan.be

Some things ‘outside’ the model

plan.be

• Part of the assignment was to look at the socio-economical impact of the transformation

• Problem: not within modelling environment• So you start again with

• a literature overview

• looking for an adequate model or instrument

• defining your data

• analysing your results

• …

Some things outside the model

plan.be

• Article of Wei et al. (2010)• Look at ways to adapt to Belgian situation

• Capacity factors

• Lifetime

• Domestic production

• …

• Gather data: define the input that you need and find sourcesSources have to be as coherent as possible with each other and with the previous exercise

• Make spreadsheet model• Run the model

Employment

plan.be

Employment: Some estimations

• The RES trajectories all create more job-years or FTE’s than REF

• REF already integrates a lot of renewables

• PV creates the most FTE’s in any given year

• BIO and DEM are the second highest job generating scenarios

2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030

0

10000

20000

30000

40000

50000

60000

70000

DEM GRID BIO PV WIND

Source: Wei et al. (2010), Federal Planning Bureau (2013).

Annual job-years generated over REF due to the RES trajectories, 2020-2030 Total FTE’s

plan.be

• Going from average to min-max ranges • Going from all jobs to types of jobs• Results in ranges of CIM and O&M and fuel processing

jobs for the years 2020 and 2030

Employment: Some estimations (II)

2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

100000

CIM min O&M min Indirect min2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

100000

CIM max O&M max Indirect max

Source: Federal Planning Bureau (2013).

plan.be

Employment: Some estimations (III)

• National macrosectoral model: HERMES<FPB

• WIND scenario• Results in % wrt

REF

Source: Federal Planning Bureau (2014).

Click icon to add chart

Confidential

plan.be

Other things ‘outside’ the model

plan.be

• 6 critical areas of government action/intervention1. Defining a clear institutional framework

2. Improving energy efficiency

3. Supporting renewable energy production

4. Improving energy infrastructure

5. Supporting research and development

6. Facilitating the electrification of the society• Principles for designing policies Cost effectiveness Fairness Competitiveness

PAMs

plan.be

Technically, a 100% renewable energy system is feasible without having to change the economic paradigm.

However, such a radical society transformation implies that:

o A highly ambitious renewable target goes hand in hand with a trend towards electrification: a doubling/tripling of power production is noted, curtailment is necessary

o Energy imports strongly diminish but remain important: imports tumble from 83% (REF) to [42%-15%] depending on the scenario

o Society shifts from a fuel intensive to a capital intensive society

o It seems cost efficient to maintain overcapacities, both in industry and power generation new paradigm in energy perception

Conclusions (1/2)

plan.be

Conclusions (2/2)

o This comes at a significant cost: in 2050, energy system costs increase by 20% wrt REF, BUT…

o When including disutility costs, the total add. cost is even higher (30%)

o With disutility + GHG damage net positive effect of some scenarios +/- 10 B€/year (highly dependent on GHG damage cost assumptions)

o 300 to 400 billion € of additional investments are needed o Sensitivity to fuel prices and PV costs

o PV costs from 371 – 1000 €05/kWp => variation of 0.5% of GDP2050

o Variant of REF scenario with higher oil prices (250 $08/boe in 2050) additional costs decrease

o Creation of additional employmento 20 000 to 60 000 additional full-time jobs in 2030

o Cost efficiency of adapting to energy flow variability

o Further research is certainly needed

plan.be

• Presentations National International

• TED conference• Summer school (1&2)• Other studies

Central Economic Council: Construction of a model capable of analysing the socio-economical impacts of ‘quelconque’ energy policy

Climact, VITO: Scenarios for a low carbon Belgium by 2050

ICEDD: Le coût d’une transition énergétique postposée en Wallonie

…

• Political First Common Commission on Energy in 2013 Flemish Government’s demand to study 2030

implications

Follow up

plan.be

Thank you!

•Contact:•Danielle Devogelaer, dd@plan.be •Dominique Gusbin, dg@plan.be •Jan Duerinck, jan.duerinck@vito.be •Wouter Nijs, wouter.nijs@vito.be •Yves Marenne, yves.marenne@icedd.be •Marco Orsini, marco.orsini@icedd.be

plan.be

Space requirements: Surface (km2), 2050

Results

REF DEM GRID BIO PV WIND0

5000

10000

15000

20000

25000

30000

35000

40000

45000

Solar Wind offshore (BE) Wind onshore Biomass (domestic and imported)

Belgian land surface Belgian Continental Plate

km

2

plan.be

External fuel bill

Click icon to add chart

2020 2030 2040 2050

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

Bioenergy Electricity from trade Gas Nuclear Oil products Solid fuels

Total energy import costs, Reference scenario (M€2005)

Source: TIMES.

+53%