|dc.description.abstract||The motivation behind this thesis is to utilize complementarity conditions to assess
the applications of energy storage in an energy system with high levels of renewable
energy sources. A model of a perfectly competitive power market has been
developed based on complementarity theory. The model includes firm demand,
system operator, power producers and storage units. In addition, four different
capacity remuneration schemes have been implemented: Energy only, strategic reserves,
capacity market and capacity payment. These models have been subjected
to a scenario with high levels of renewable energy in order to assess the impact of
energy storage in these circumstances.
The methods used in this work are theoretical and highly dependent on the assumptions
and parameters used. The documentation should give a good insight
to complementarity modeling in general as well as describing the entire model all
the way up to the full mixed complementarity problem formulation.
A large part of this work has consisted of improving the existing model from the
project work and develop and integrate the energy storage with the rest of the
The findings show that the system can benefit from energy storage options. In
general, energy storage will give better conditions to base load units such as nuclear
power. The energy storage will smooth over variations and increase the amount
of capacity a base load unit can run for a large portion of the year. Utilizing
Norwegian hydro reservoirs for pumped hydro applications was found to be the
most effective storage unit, increasing the amount of nuclear power by 18.8%.
Pumped hydro was found to shift energy between large periods while the battery
was found to arbitrage prices on a shorter term.
This thesis can give an insight into which barriers need to be overcome in order to
ensure investments in energy storage. For example high fixed costs of the storage
and unfavorable conditions compared to thermal plants in a capacity remuneration
mechanism can be the determining factor resulting in low or no installed storage