Redistribution Effects of Energy and Climate Policy
Lion Hirth, Falko Ueckerdt
International Energy Workshop
University of Cape Town
2012-06-19
Goal: explain and quantify the redistribution flows
induced by climate and energy policy
RES support
Producers
CO2 pricing
Government
Producers
Government
Consumers
Consumers
• Redistribution: changes in economic surpluses of 3 sectors.
• Three sectors: Producers (existing generators), Consumers, Government
• Two policies:
1.
Renewable support
2. CO2 pricing
• The same methodological framework is applied in two models
1.
2.
Analytical model: understand chain of causality, derive qualitative findings
Numerical model (North-Western Europe): quantify, assess ambiguous results
Conclusions
RES support
CO2 pricing
Government
Producers
Producers
Government
Consumers
Consumers
• Redistribution large relative to welfare effects
• RES support:
– electricity price decreases  producers lose, consumers win
– RES support  State pays
• CO2 pricing:
– electricity price increases  effect on producers depends on technology mix, consumers lose
– auction / tax revenues  government net income increases
• Opposite flows  policy mix allows CO2 mitigation without changing profits
Conclusions
RES support
+
CO2 pricing
Producers
Government
Consumers
• Redistribution large relative to welfare effects
• RES support:
– electricity price decreases  producers lose, consumers win
– RES support  State pays
• CO2 pricing:
– electricity price increases  effect on producers depends on technology mix, consumers lose
– auction / tax revenues  government net income increases
• Opposite flows  policy mix allows CO2 mitigation without changing profits
Connecting two branches of the literature
• Merit-order literature
- How much does subsidized wind generation reduce the electricity price?
- Do consumers gain, even if they pay the subsidy?
- Sensfuss (2007), Sensfuß et al. (2008), Sáenz de Miera et al. (2008), Munksgaard & Morthorst (2008),
MacCormack et al. (2010), Rathmann (2007), O’Mahoney & Denny (2011), Gil et al. (2012)
• CO2 pricing literature
- How do producer profits change when carbon trading is introduced (depending
on different allocation rules for emissions allowances)?
- To what extend can CO2 costs be passed through to consumers?
- Martinez & Neuhoff (2005), Chen et al. (2008), Burtraw et al. (2002), Bode (2006), Sijm et al. (2006)
5
Long-term equilibrium (LTE) without policies
• capital stock endogenous (“green field approach”)
• scarcity prices
 long-term profits zero (free market entry, perfect competition)
methodological framewok
investments sunk
Short-term equilibrium (STE) prior to policy
• capital stock is given (investment is possible)
• no scarcity prices
 short-term profits positive (used to pay back capital cost)
policy introduced
New STE with RES support
 short-term profits changed
CO2
pricing
Effect of a policy (def.):
compare profits
before and after
policy is introduced
Both
policies
capital stock
endogenous
Difference
between policies (def.):
compare profits between
…
…
New LTE with RES support
two new STEs
• LTE changed
 zero LT profits
This framework is applied in an analytical and a numerical model
6
Analytical model
Long-term screening curves
C
•
Two generation technologies: coal and gas
•
Methodology
1.
2.
3.
-
•
“classical approach” to investment planning:
screening curve
load duration curve (LDC)
price duration curve (PDC)
assumptions: inelastic demand, no
externalities, perfect competition, perfect
foresight, no intertemporal constraints, no
trade, no storage, energy-only markets
Long-term equilibrium (derived in the paper)
-
market equilibrium is cost-minimum
long-term profits of all technologies are zero
scarcity prices assure that there is no “missing
money”
total costs
(€/MW-year)
Gas
Coal
T1
q
T
(hours per year)
(Residual) load duration curve
Load
(MW)
p
Price duration curve
(€/MWh)
ps
Δ
T1
T
(hours per year)
7
The short-term equilibrium
Long-term screening curves
Short-term screening curves
C
total costs
(€/MW-year)
• investments are sunk
 no capital cost
Gas
Coal
Coal
• capacity is constrained
T1
T
T1
T
(hours per year)
• no scarcity prices
q
(Residual) load duration curve
Load
(MW)
T
p
Price duration curve
(€/MWh)
• Results
ps
Δ
- base-load technology
makes ST profits
T1
T
T1
T
(hours per year)
8
C Short-term screening curves
(€/MW-year)
CO2 pricing
Coal
T1
T
q
(MW)
RES support
T
p
(€/MWh)
T1
T
9
Without support
Wind Support
With wind support
C Short-term screening curves
(€/MW-year)
•
•
C Short-term screening curves
(€/MW-year)
changes the LDC to
RLDC
Gas
Coal
strictly reduces
producer rents
T1
Coal
T
T1
q
q
(MW)
(MW)
T
RES support
T
p
T
p
(€/MWh)
(€/MWh)
T1
T
T2wind T1
T
10
CO2 pricing
C Short-term screening curves
(€/MW-year)
•
•
Is more complex and
shown in paper
Effect on producers
depends on
technology and CO2
price
CO2 pricing
Coal
T1
T
q
(MW)
T
p
(€/MWh)
T1
T
11
Numerical model
Model & scenario setup
• why numerical modelling?
- quantitative estimates for North-Western
Europe (orders of magnitude)
- ten technologies (wind, solar, eight
dispatchable, pump hydro)
- interconnectors, storage, CHP, ancillary services
• Same framework applied
1. long-term equilibrium
2. short-term equilibrium
3. policy shocks
• integrated dispatch and investment
- hourly time steps for a full year
- existing plant stack, storage and
interconnectors
- endogenous (dis-)investments in generation,
storage and interconnectors via annualized
investment costs
• 1M equations, 4M non-zeros, solving time ½ h
How big is the redistribution effect of wind support?
PANEL 1: REDISTRIBUTION (€/MWH) WHEN INCREASING THE
WIND SHARE FROM ZERO TO 30 %.
70% of
Nuclear, 60%
of coal, 50% of
gas profits are
taken away
Conv Producers
Nuclear Rents
Coal Rents
Gas Rents
- 13
-9
-1
Producer Surplus
- 22
Consumer Surplus
Electricity market
Heat market
AS market
Interconnectors
+ 28
-2
-0
-0
Cons Surplus
+ 25
Effect on Government Budget
CO2
/
Wind
- 18
Gov’t Budget
- 18
Welfare
Consumers
Producers
Government
+ 25
- 22
- 18
Externalities
ignored
Welfare
- 15
Redistribution
effect is large
Consumers gain even if
they pay for subsidies
13
Technology
dependence
Wind support
PANEL 1: REDISTRIBUTION (€/MWH) WHEN INCREASING THE
WIND SHARE FROM ZERO TO 30 %.
Conv Producers
Nuclear Rents
Coal Rents
Gas Rents
- 13
-9
-1
Producer Surplus
- 22
Consumer Surplus
Electricity market
Heat market
AS market
Interconnectors
+ 28
-2
-0
-0
Cons Surplus
+ 25
Effect on Government Budget
CO2
/
Wind
- 18
Gov’t Budget
Welfare
Consumers
Producers
Government
Welfare
- 18
+ 25
- 22
- 18
- 15
CO2 pricing
PANEL 2: REDISTRIBUTION (€/MWH) WHEN INCREASING
THE CO2 PRICE FROM ZERO TO 100 €/T
Conv Producers
Nuclear Rents
Government
+ 21
CO2
+ 20
Coal Rents
- 10
Wind
/
Gas Rents
+0
Prod Surplus
+ 12
Gov’t Budget
+ 20
Consumer Surplus
Electricity market
- 43
Welfare
Consumers
- 49
Heat market
-6
Producers
+ 12
AS market
-0
Government
+ 20
Interconnectors
-0
Cons Surplus
- 49
Welfare
- 17
Existing generators’
ST profits increase
Wind support and CO2 pricing induce opposite redistribution flows
Questions?
Comments?
Ideas?
[email protected]
How big is the redistribution effect of CO2 pricing?
PANEL 2: REDISTRIBUTION (€/MWH) WHEN INCREASING
THE CO2 PRICE FROM ZERO TO 100 €/T
Technology
dependence
Existing
generators’ ST
profits increase
Conv Producers
Nuclear Rents
Government
+ 21
CO2
+ 20
Coal Rents
- 10
Wind
/
Gas Rents
+0
Prod Surplus
+ 12
Gov’t Budget
+ 20
Consumer Surplus
Electricity market
- 43
Welfare
Consumers
- 49
Heat market
-6
Producers
+ 12
AS market
-0
Government
+ 20
Interconnectors
-0
Cons Surplus
- 49
Welfare
- 17
Wind support and CO2 pricing induce opposite redistribution flows
16
CO2 pricing: short-term screening curves pivot
(a) Rents are generated by coal power
plants when gas power plants are pricesetting.
(b) The difference of variable costs
decreases, thus the coal rents decrease.
The dispatch remains unchanged.
(c) No rents occur because variable costs of
coal and gas power plants are equal. (65€/t
CO2)
(d) Now the dispatch changes: Gas power
plants now have least variable costs and
cover base load. Coal power plants only
cover the remaining base, mid and peak
load. Gas power plants generate rents when
coal power plants are price-setting.
(e) The screening curve of coal touches the
screening curve of new gas power plants.
The rents of gas power plants reach a
maximum. (80€/t CO2)
(f) Now, new investments in gas power
plants lead to decommissioning of existing
coal capacity. Old gas power plants are the
only plants that generate rents. These rents
remain at their maximum value.
C
C
total costs
(€/kW - year)
Gas
Coal
T1
(a)
Gas
Coal
T
C
T1
(b)
C
total costs
(€/kW - year)
New Gas
Coal
Gas
Gas
Coal
T1
(c)
C
T
T2CO2 T
T
(d)
Coal
C
New Gas
Coal
Gas
total costs
(€/kW - year)
New Gas
Gas
T2CO2 T
(e)
Assuming variable costs of 25 €/MWhth (gas) and 12 €/MWhth (coal), efficiencies of 48% (gas) and 39% (coal), carbon
intensities of 0,24 t/MWhth (gas) and 0,32 t/MWhth (coal) and investment costs of 100€/kWa (gas).
T2CO2 T
(f)
The effect of CO2 pricing
No CO2-Pricing
CO2-Pricing
• With high CO2 price:
Short-term screening curves
C Short-term screening curves
Shift of rents only
Coal
C
(€/MW-year)
New Gas
(€/MW-year)
depends on the initial
Gas
long-term capacity mix
𝑔𝑎𝑠
𝑅2
𝑔𝑎𝑠
− 𝑅1𝑐𝑜𝑎𝑙 = 𝐼𝑔𝑎𝑠 𝑞1
− 𝑞1𝑐𝑜𝑎𝑙
General results
• Total producer profits
depend on long-term
capacities and CO2
price
– Large redistribution
within producers
depending on
technologies
– More low-carbon
technology  total
producer rents tend to
increase
Coal
T1
T
T1 T2CO2 T
q
q
(MW)
𝑐𝑜𝑎𝑙
𝑞2
(MW)
𝑔𝑎𝑠
𝑞1
𝑐𝑜𝑎𝑙
New gas
replaced
𝑞1
𝑔𝑎𝑠
𝑞2
T
p
T
(€/MWh)
• Consumers pay
• State benefits
North-Western Europe?
 Numerical model
𝐶𝑂2
𝑐 𝑐𝑜𝑎𝑙
𝑐 𝑔𝑎𝑠
𝐶𝑂2
𝑐 𝑔𝑎𝑠
𝑐 𝑐𝑜𝑎𝑙
T1
T
T1 T2CO2
T
Policy Mix: redistribution can be minimized
Expenditure of the electricity industry
Redistribution effects of policy changes
Rents of conventional generators (with numbers)
Generation Costs w/o CO2
CO2 payments
Rents needed to recover investment costs
24
60
82
19
40
25
16
55
20
27
0
0
0 €/t
17 €/t
33 €/t
66 €/t
100 €/t
CO2
0%
5%
10%
20%
30%
wind
25
34
0
0
-25
-34
-50
-69
Consumer Producer RentState Revenue Economic
Rent
Welfare
€ bn p.a.
110
€/MWh
25
€ bn p.a.
€/MWh
80
Wind penetration from 0 to 30%
CO2 price from 0 to 100 €/t
Both policies simultaneously
This paper brings together two branches of literature
Merit-order literature
CO2 pricing literature
• Decrease of spot market prices due to
• How do producer profits change
renewable electricity generation  savings
(depending on different allocation rules for
for the consumer
emissions allowances)?
• Sensfuss 2007, 2008, de Miera et al. 2008,
• To what extend CO2 costs can be passed
Munksgaard & Morthorst 2008
through to consumers?
• Martinez & Neuhoff 2005, Chen et al. 2008,
Burtraw et al. 2002
Our work adds to the literature in three ways.
• effects of both policies in a consistent framework with the long-term equilibrium as
benchmark
• focus on redistribution effects: evolution of effects at different levels of policy
intervention and comprehensive accounting of all flows
• analytical model to trace causal mechanisms and numerical model for quantifications
CO2 pricing within a nuclear system tends to increase
conventional rents
• xx
No CO2-Pricing
CO2-Pricing
Short-term screening
Short-term screening
C curves
(€/MW-year)
C curves
Coal
New
Gas
(€/MW-year)
Gas
Coal
q
T1 (a)
T
(MW)
𝑔𝑎𝑠
𝑞1
q
CO2 T
(b) T1 T2
(MW)
𝑐𝑜𝑎𝑙
𝑞2
𝑔𝑎𝑠
𝑞1
𝑐𝑜𝑎𝑙
𝑞1
𝑛𝑢𝑐
𝑛𝑢𝑐
𝑞1
𝑞1
(c) T
p
p
(€/MWh)
(d)
T
(€/MWh)
𝐶𝑂2
𝑐 𝑐𝑜𝑎𝑙
𝐶𝑂2
𝑐 𝑔𝑎𝑠
𝑐 𝑔𝑎𝑠
𝑐 𝑐𝑜𝑎𝑙
𝑐 𝑛𝑢𝑐
T1
(e)
T
(f)
T1 T2CO2 T
Long-term equilibrium with nuclear
• Back-up slide
C
Long-term screening curves
total costs
(€/MW-year)
~𝑐 𝑔𝑎𝑠
Gas
Coal
~𝑐 𝑐𝑜𝑎𝑙
𝐼 𝑐𝑜𝑎𝑙
𝐼 𝑔𝑎𝑠
T1
T
(hours per year)
(a)
q
Load
(MW)
𝑔𝑎𝑠
𝑞1
𝑐𝑜𝑎𝑙
𝑞1
𝑛𝑢𝑐𝑙𝑒𝑎𝑟
𝑞1
T
(b)(hours per year)
p
(€/MWh)
ps
𝑐 𝑔𝑎𝑠
𝑐 𝑐𝑜𝑎𝑙
Δ
T1
T
(hours per year)
(c)
Model & scenario setup
•
why numerical modelling?
–
–
–
•
quantitative estimates for North-Western Europe (orders of
magnitude)
ten technologies (wind, solar, eight dispatchable, pump hydro)
interconnectors, storage, CHP, ancillary services
Same framework applied
1. long-term equilibrium
2. short-term equilibrium
3. policy shocks
•
integrated dispatch and investment
–
–
–
•
•
•
hourly time steps for a full year
existing plant stack, storage and interconnectors
endogenous (dis-)investments in generation, storage and
interconnectors via annualized investment costs
stylized electricity market model
– total system costs are minimized with respect to investment and
dispatch decisions under a large set of technical constraints
– no market power, externalities or other market imperfections
 cost minimization is equivalent to profit-maximizing firms
– electricity price is set by variable cost of marginal plant
– no load flow, NTCs between market areas
back-tested and calibrated to market prices
1M equations, 4M non-zeros, solving time ½ h
23/14