| dc.description.abstract | Denmark aims for a 50% wind integration in 2020. In the current scenario, where
wind represents around 34% of domestic consumption, mothballing of conventional
power plants and heavy dependency in interconnection lines is already the trend.
One of the strategies from the Transmission System Operator is to integrate a big
volume of flexible loads to adapt to wind production and mitigate some of its draw
backs. EVs, even when a significant increase is expected in the following years, will
not represent a volume of consumption that can really impact the load curve by
2020 and this type of response will rely in the short term in other flexible loads like
Heat Pumps.
Due to its configuration and advanced technology, EVs can participate to other
services vital to the correct operation of the Electric Power System as it is provision
of frequency reserves to maintain balance between consumption and generation.
This work presents a solution using the adaptive charging capabilities of an EV
to get the best respond in both the day-ahead market and the regulation market.
The adaptive scheme will achieve: lower price for purchased electricity in the dayahead
market, with higher levels of wind energy penetration, and the possibility to
participate to the frequency regulation market and get revenues. All this features
are gained without affecting the normal operation of the vehicle. Two different
configurations for the battery of the EV are compared in this work: unidirectional
and bidirectional.
A fleet of 400 EVs have been modeled based on statistical survey data for EVs
users driving profiles in weekdays and weekends. This fleet is managed by the figure
of an aggregator who purchases electricity in the day-ahead market and bid on
the frequency regulation market. The reference charging profile is a non-controlled
consumption scheme of plug-and-charge. This reference model is compared first
with a basic adaptive models based on weight coefficients varying according to the
State of Charge of the battery and the level of wind penetration. Later on, the
adaptive model is optimized, following the same indicators, seeking to maximize
wind penetration while bidding to frequency regulation market the most number
of times. The optimization algorithms used are Gradient Search, Genetic and
Differential Evolution. The decision factor for the adaptive charging strategies is the
forecast wind penetration signal with is the coefficient between the level of forecast
wind production and the level of forecast consumption. The idea behind using
this signal is that it will yield typically lower cost of electricity and high net wind
penetration. Allowing high net wind penetration will reduce the presence of energy
from other generation facilities and thus the CO2 content in the battery charge.
Results show that the owner of an EV with bi-directional capabilities and Genetic
Algorithm can reduce the final expenses on the EV by 20% in one year. If GSA is
used instead, 36% more wind energy will be integrated in the vehicle. In addition, because
currently upward regulation is provided by coal and gas fired units, 60% of the
current emissions by providing this service could be cut with a GSA charging scheme. | |
