Page 288 - Decision Making Applications in Modern Power Systems

248 Decision Making Applications in Modern Power Systems

adverse effects and lead to an insignificant load margin increase. This
scenario is improved when losses-reduction actions are performed in the
critical regions.
In the literature, there are different approaches for the reactive power
dispatch using particle swarm optimization (PSO). A comprehensive learn-
ing PSO for reactive power dispatch is developed in [3]. A combination of
PSO and a feasible solution search is applied in wind farms by [4],whereas
in [5] the PSO is integrated into multiagent system. The reactive power
compensation of radial distribution systems based multiobjective planning
algorithm using PSO is developed in [6], while in [7] a multiobjective
optimization problem is formulated to relieve the overvoltage caused by
large PV penetration and to minimize total line loss. Refs. [8 10] proposed
the loss reduction by shunt compensation employing different optimization
techniques, the first being solved by PSO and the second by primal-dual
method of interior points. In [11] the PSO was employed in training
artificial neural networks.
The current chapter presents a reactive power dispatch technique based
on the PSO with the objective function of minimizing electrical losses,
associated with a dispatched generator selection performed by means of the
tangent vector sensitivity analysis. In this way the contribution of this study
consists of the union of two methods to solve this problem considering the
intermittency of renewable sources. First, the generators are identified using
the tangent vector and then supplied to the PSO to perform the dispatch in
order to reduce electrical losses. The PSO is employed in this study due to
its applications for reactive power control. Moreover, for this study, voltage
collapse indices, PV and QV curves, are applied facilitating the evaluation of
the operating conditions related to the voltage safety margin, aiding in the
system planning. Using the PV curve, it is possible to quantify how close
the system is from voltage collapse, while the QV curve indicates how far
each bus is from the instability region.
The case studies are performed using the IEEE 118-bus modified system
considering penetration of renewable energy sources (RES), wind and solar,
and a variable demand profile. The analyses are divided into scenarios with
the presence and absence of renewable sources considering different periods
of the day. Furthermore, the influence on the active and reactive load
margins is also analyzed given that the system configurations change
throughout the day. Results are analyzed and discussed seeking to assist the
process of decision-making for voltage stability margin at planning. Further,
a comparison can be made for different system settings through the dispatch
benefit, which represents the system operational gain for additional incre-
ments in the generation dispatch.
The chapter structure is defined as follows: first, the voltage collapse
index determination is presented in Section 10.2. Section 10.3 discusses the
active power losses. Section 10.4 introduces the proposed identification of