Page 317 - Decision Making Applications in Modern Power Systems
P. 317
278 Decision Making Applications in Modern Power Systems
accepted approach or a unified basis to quantify the trade-off between a
secure and an economic performance. As a consequence, decision-making in
the presence of uncertainty has often resorted to ad hoc or rule-based meth-
odologies, leading either to a design that is conservative in terms of cost for
the desired security level or, if uncertainty is ignored in the design phase, to
a solution where security might be at stake. Performance with ignored uncer-
tainty might be acceptable if the level of uncertainty of the system is rela-
tively low. But with the increase in renewable generation due to
environmental concerns and the pursuit of sustainable energy sources system,
uncertainty has generally increased. The main challenges with renewable
energy sources (RESs), such as wind and photovoltaic power generation,
are that they are nondispatchable, fluctuating, and uncertain. As a result,
employing suboptimal measures to account for the stochastic nature of
RES may result in an undesirable economic effect. On the other hand a
deterministic design where uncertainties are ignored will lead to an
unacceptable reliability level.
It is, therefore, necessary to develop a mechanism for optimal decision-
making in the presence of uncertainty that takes into account the multiobjec-
tive nature of the problem, that is, the trade-off between security and
economic operation. The increased share of RES results in an increased
amount of required reserves, which may have an opposite effect both from
an environmental and economic point of view. The latter raises the need of
cheaper and environmental friendlier reserves providers. Demand-side
resources have already been used to provide certain control services, but the
full exploitation of their controllability has recently become an emerging
research topic.
Demand response and storage resources could be utilized to offer ancil-
lary services including reserve provision. Promising technologies, such as
electric vehicles and thermostatically controlled loads, could contribute with
a large amount of reserve capacity and hence allow for the integration of
high shares of RERs.
However, these technologies include uncertainty, mainly introduced due
to human behavior and weather conditions, rendering their successful exploi-
tation challenging. Taking the uncertainty into account in the decision-
making mechanism introduces additional operational costs compared with a
deterministic solution. To alleviate this the controllability of certain network
components other than the loads could also be exploited. These components
could be utilized for preventive and corrective control actions. Some exam-
ples of controllable components are flexible alternative current transmission
system (FACTS) devices, HVDC lines and transformers. These components
do not provide reserve capacity, but their set point can be modulated in a
post disturbance situation, thus leading to lower operating costs.
This dissertation deals with the problem of developing a unified stochas-
tic framework for optimal decision-making, taking the uncertainty due to