Methods of Maximizing Power System Voltage Stability Margin using Active Distribution Networks

Abstract

The proliferation of distributed energy resources and renewable energy sources that are connected to the grid through power converters, is changing significantly the way that distribution networks operate and consequently also transmission systems. Generation is gradually moving towards the lower voltages, and as a result, ancillary services provided traditionally by thermal power plants, are also vanishing from the transmission system. Despite the challenges, opportunities also appear in terms of service provision from the distribution grids to support the transmission system, thus turning previously passive distribution networks to Active Distribution Networks. Thus, Active Distribution Networks are expected to support the transmission system in dealing with problems such as frequency stability and voltage stability by providing support in both normal and emergency operating conditions. This thesis examines the support that Active Distribution Networks can offer to the transmission system in terms of Voltage Stability and how can this be achieved most efficiently (optimally). Current practices for voltage stability support, usually from big generating units and/or system connected Wind Farms, include maximum reactive power support, but in some cases this may prove inefficient or even worsen the situation if not coordinated properly, especially when reactive power is coming from the distribution grid where load is also connected. For that reason, it is important to build frameworks for support in case of voltage stability problems, so that the support from the distribution grid is efficient and does not violate operational limitations of the distribution network.

In the current thesis, the voltage stability problem is expressed as an optimization problem, or voltage security margin maximization problem, taking into consideration also possible contingencies in the transmission system. In this way, it can be ensured that support will opt to maximize the stability margins and move the transmission as far away as possible from imminent voltage instabilities, taking also into account possible contingencies, such as the loss of transmission lines or generating units. Following that, different approaches or frameworks are proposed for support from the distribution network, assuming some information exchange between the transmission system operator and the ADN operator. This is very critical in the approaches, as usually, limited communication is established between transmission and distribution in real power systems. In the recent years, a lot of effort is being made between Transmission System Operators and Distribution System Operators to improve communication and data exchange in all time frames, including real time operation. First, the effect of system connected RES, usually including Wind Farms or big Photovoltaics, on the maximization of the voltage stability margin is examined. In that direction, requirements already imposed by the European legislation, for fast current injections by renewable energy resources are implemented to deal with imminent voltage instabilities. Good parameterization is deemed critical in order to not have adverse results and limited stability margins. In the direction of support from the distribution side, a corrective (distributed) and a preventive (centralized) optimization framework are created. Both frameworks exploit the ability of the active distribution networks to change the active and reactive power flows at the point of common coupling with the transmission grid, thus supporting the transmission system when facing a voltage stability problem and maximize the voltage stability margins. In order to ensure that the control taken in the distribution grid is optimal for the support of the transmission system, sensitivities are used in the decentralized approach. Sensitivities drive the control taken in the active distribution network into the right direction by linearly associating the changes of both active and reactive power flows at the point of common coupling with the voltage stability margin of the transmission system. In the centralized optimization framework, the flexibility of the ADN is exploited. More specifically, active distribution networks estimate a region in the active and reactive power plane at the point of common coupling and send it to the transmission system operator that takes it into consideration when computing the stability margins of the power system. In that way, the flexibility region that is calculated by the active distribution networks, can be used in the form of additional constraints, in the voltage stability margin problem maximization directly. The calculation of the flexibility region of an active distribution network is a separate challenge and various methods are examined and compared. An optimization based approach is proved to be the most efficient especially when dealing with complex feeders containing multiple decision variables. The optimization based approaches, first solved with non-linear programming solvers, prove to have some limitations. For that reason, further optimization methods are explored, such as the convex relaxation second order cone programming approach is examined in order to accelerate the calculation process. Both frameworks and the flexibility region estimation are implemented in test systems and results show significant support, in terms of increasing the voltage security margins when a critical contingency occurs. Moreover, results are also evaluated using time-domain simulation using the quasi steady state approach and implementing the results that are extracted from the optimization frameworks. In all cases, the proposed frameworks are able to support the transmission system in terms of voltage stability and even tackle imminent voltage collapses. Flexibility region estimation, can in more complex feeder configurations prove to be time consuming. For that reason, an alternative fast method is implemented for the computation of the flexibility region, using a relaxation of the optimization problem. This proves to improve simulation time and still be fairly accurate. Last but not least, an incident in the Hellenic Interconnected System is analysed, using the optimization formulation for the voltage security margins. The implementation on a real system and real incident makes the findings of the thesis and the formulation implemented more concrete. The thesis not only proves that active distribution networks can significantly support the transmission system in cases of voltage stability issues, but also proposes alternate support frameworks to facilitate and optimize the procedure both preventively and correctively.