Reactive Power From Distributed Generators : Characterisation and Utilisation of the Resource

Both transmission and distribution system operators must increasingly accommodate renewable generators on their respective networks. This presents numerous technical challenges, amongst which is reactive power management. Failure to appropriately manage reactive power can drive up connection costs for new generators, can harm power system voltage security, and can necessitate onerous must-run generation constraints or procurement of costly dedicated reactive power sources. Additionally, transmission and distribution system operators may wish to deploy reactive power in conflicting ways. For instance, the former may stipulate reactive power regimes for distributed generators that minimise connection costs, while the latter may focus on enhancing power system voltage security. Notably, the literature does not offer extensive insight on resolving these objectives, or on the inclusion of distributed reactive power resources into holistic power system operation and planning activities. Three broad research questions emerge: what is the extent of the reactive power resource offered by distributed generators?; how can these resources be included in power system planning?; and how should distributed generators be operated to balance the needs of the distribution and transmission systems? To frame the first problem, a novel recasting of the traditional capability chart is provided. This chart displays the reactive power capability of an entire distribution network, aggregating together the disparate contributions available from the various generators present. This capability chart can be inferred from time series load flow simulations, as the numerous active and reactive power flows calculated at the transmission node imply the underlying relationship. Notably, the range of reactive support available is heavily dependent on the magnitude, and disposition, of active power flows internal to the distribution network. A more rigorous treatment of these effects demands novel optimal power flow techniques. These techniques deploy non-linear programming in an atypical role, functioning as a search technique to find network operating points that are not optimal in the conventional sense of desirability, but rather indicate the extent to which network conditions may align to hinder reactive support provision. Such aggregate capability charts facilitate novel transmission system reactive power planning techniques. These techniques employ a delineation of how much reactive power will be required by a test generator, which may represent a collection of distributed generators, to adequately control the voltage at its connection node. This requirement varies as the generated active power rises, due to displacement of other generators as well as altered network flows. At each transmission system node, the reactive power requirements can be compared to the capability chart which shows the reactive power availability from the various locally available sources, so deficits can easily be identified. Time series validations demonstrate this planning technique to be usefully predicative in anticipating voltage control problems. Finally, it is apposite to examine how the capabilities of distributed generators may best be harnessed. Further enhancements to the optimal power flow tool become essential here. A tailored implementation of a terminal voltage control mode for distribution generators permits the inclusion of generator voltage control settings as free variables within the optimization framework, whose optimal selection presents many exciting potentialities. The optimal power flow framework must be extended to ensure the suitability of the optimized settings across the full gamut of load and generation conditions that may arise. New objective functions are thus made tractable. Notably, it is possible to maximise the voltage responsiveness of a distribution network, such that a decline in transmission system voltage is met by an injection of reactive power, absent of any supervisory control. Under this scheme, the transformer is locked at an optimized static tap setting which exposes the distributed generators to the voltage fluctuations arising at the transmission level, thus invoking a coherent reactive power response.