Emissions trading and investment decisions in the power sector a case study in Finland

1
Emissions trading and investment decisions in the
power sector - a case study in Finland

• Harri Laurikka, Helsinki University of Technology
• Tiina Koljonen, VTT Processes
• International Energy Workshop 2004, IEA, Paris

2
Introduction

• EU ETS creates a new price risk for power
  capacity investments in Europe
• new plants, retrofits, acquisitions and
  divestments
• Implications have been studied on a European
  scale
• e.g. de Leyva and Lekander 2003, Reinaud 2003
• This paper
• Analysis in a more detailed regional setting
  through a hypothetical case study
• Extension of the standard NPV analysis with two
  real options
• Real option a right, but not the obligation, in
  a capital investment project

3
Investment environment

• Finland is a part of the integrated Nordic
  electricity market and the EU ETS
• consumption in total 383 TWh in 2002
• characterized by a large share of hydropower and
  a high volatility in spot prices
• dominated by large companies
• National Allocation Plan for 2005-2007 ready
• reduction of 3 in emissions compared to
  business-as-usual
• opt-in for boilers less than 20 MW, if connected
  to a network with large boilers
• no guidance for 2008-2012
• No major changes to energy taxation expected
• Wholesale spot market price expected to increase
  4-16 /MWh, when the allowance price level is
  5-20 /tCO2
• Partial equilibrium energy systems model (VTT
  2003, 2004) suggest fuel and technology changes
  from coal/peat in the direction of a lower carbon
  content (e.g. natural gas, biomass, additional
  CHP, wind power).

4
Case study

• a hypothetical 250 MWe condensing power plant
  using either coal or gas
• a simple NPV is extended to a dynamic DCF
  analysis
• Value of the investment opportunity (NPVopp)
  NPV Value of the option to alter operating
  scale Value of the option to wait
• Note NPVopp always gt 0
• Starting point a manager applying a subjective
  discount rate rather than a theoretically
  objective discount rate (i.e. consistent asset
  pricing)

5
Model (1/3)

• (Potential) investment to be made within
  2005-2007
• Two stochastic variables (price of electricity -
  pe, allowance price pCO2) two deterministic
  variables (fuel price - pf, number of free
  allowances - N)
• Stochastic variables follow discrete-time
  continuous-state processes. Time period is 1
  year.
• Form of the annual duration curve for the price
  of electricity is constant (note not the
  absolute values!)
• Price of an emission allowance is constant within
  a year
• There are no switching costs (start-up /
  shut-down)

6
Model (2/3)

• Direct relationship between the price of
  electricity (pe) and the allowance price (pCO2)
• The marginal plant (type) in the power system and
  ? is a function of pe,base

Stochastic baseline price i.e. without ET
Emission factor of the marginal plant
7
Model (3/3)

• Spark spread (S) thus given as
• And the cashflow (CF)

Plant-specific OM costs
Thermal efficiency
CO2 emission factor
Thermal power
Fixed cost
8
Stochastic processes

• The natural logarithms of pe,base and pCO2 are
  modelled as mean-reverting stochastic processes
  (Ornstein-Uhlenbeck)

(pe,base ? pe,base, pCO2 - ? pCO2 )
(pe,base ? pe,base, pCO2 ? pCO2 )
Mean reverting stochastic change 1
Each arrow is a transition with a state-dependent
probability!!
Initial position (pe,base, pCO2)
Mean reverting stochastic change 2
(pe,base - ? pe,base, pCO2 - ? pCO2 )
(pe,base - ? pe,base, pCO2 ? pCO2 )
9
Data

• Scenarios for emissions trading No / Yes
• Scenarios for prices
• Low (average) starting from 7 /tCO2 in 2005 and
  reaching 1 /tCO2 in 2012
• High (average) starting from 7 /tCO2 in 2005
  and reaching 20 /tCO2 in 2012
• Scenarios for allocation
• ET with free of charge allocation forever
• ET with a tighter free of charge allocation only
  until 2012
• Scenarios for volatility of allowance prices 10
  and 40
• Scenarios for correlation of allowance price with
  electricity price 0 and 0.5
• Deterministic scenarios for coal and gas prices

10
Results coal plant

NPVopp zero in all scenarios!!
11
Results gas plant
Positive values for NPVopp
Invest now!
12
Conclusions

• Investment decisions in liberalised electricity
  markets depend in particular on the stochastic
  market prices of electricity, fuel and
  (potential) emission allowances
• A simple NPV approach ignores the value of
  options
• In addition to the expected allowance prices,
  behaviour of the allowance market (e.g.
  volatility, correlations to electricity and fuel
  prices) plays a role in investment decisions
• Uncertainty regarding the number of free
  allowances is critical for a quantitative
  investment appraisal

13
Contact

• Harri Laurikka
• Laboratory of Energy Economics and
• Power Plant Engineering
• Helsinki University of Technology
• Email. harri.laurikka_at_hut.fi
• Tel. [phone]