The aim
of
the research project PowerACE is to develop a realistic
agent-based simulation model of the German electricity system. In this
model, the short-term perspective of daily electricity trading is
combined with long-term investment decisions for power plants. A
special focus is placed on the impact of emissions trading and the
increased use of renewable energy sources on markets and power
generation structures.
By applying agent-based simulations we make
use of
a new method for market simulation which allows for realistically
modelling different aspects that characterise today's electricity
sector:
- Strategic behaviour of market players
- Learning in daily repeated auctions on
the basis of trading success
- The interplay between several
interrelated
markets, e.g. the day-ahead market, markets for reserve power,
certificate markets etc.
The methodology of agent-based simulation is
based
on concepts from computer science and economics in equal measure.
Agents can be equipped with learning algorithms that allow them to
optimise their strategies on the basis of the experience gained from
trading in subsequent rounds. This approach constitutes a promising way
of analysing complex dynamic phenomena with decentralised problem
solving processes. In comparison with centralised optimisation models
agent-based simulations may lead to more realistic results, because
actor specific behaviour and dynamic adaptation of strategies by the
actors can be modelled explicitly.
Systematic simulation runs are supposed to
contribute to a better understanding of the interplay between different
actors and different markets in electricity and emission certificate
trading. The insights gained from the simulation results may be used
for deriving recommendations for a good design of markets and
the regulatory framework in the electricity sector.
|
