The impact of the uncoordinated local control of decentralized generation on the reactive power margin

Abstract

The penetration of decentralized generation entails considerable challenges to keep the reactive power flow at the required levels and to ensure the voltage stability in high voltage grid (Lund in IEEE power engineering society general meeting, 2007). The use of the uncoordinated local control in distributed generation level (e.g. in medium voltage level) creates an uncontrolled reactive power flow on the higher-level grid (e.g. high voltage grid) (Ilo et al. in CIRED, 23rd international conference on electricity distribution, 2015; Ilo in Sci. Res. 7:217–232, 2016). Additionally, it systematically modifies the lumped load behavior seen from high voltage grid and thus also its voltage stability phenomenon.

The main focus of this Paper is the impact assessment of the boundaries modelling of the high voltage grid in presence of a large decentralized generation share on the reactive power margin. Traditionally the modelling of the high voltage grid boundaries has been done using lumped loads and the corresponding ZIP model which is not accurate enough for this investigation.

We start with the description of the methodology used for the adequate modelling of the high voltage grid boundaries in case of a large penetration of decentralized generation, which is already retrofitted with local control. Different simulation scenarios, e.g. varying the load situation, varying the amount of conventional power production and distributed generation production have been defined. The simulations are being performed on a test grid, which is based on the European Power Grid configuration (CIGRE task force C6.04, 2014). This test grid consists of a meshed 380 kV and 220 kV transmission grid, a slightly meshed 110 kV and a normally radial operated 20 kV grid. One of the 20 kV sub-systems, which includes decentralized generators, is modelled in detail. The other 20 kV sub-grids are modelled through adequate equivalent lumped loads and injections at the power supply substation busses. The low voltage sub-grids, which also include a high share on distributed generators, are modelled through the adequate equivalent lumped loads and injections at the distribution transformer buses. All decentralized generators connected to it are equipped with $Q(U)$ local control to keep the voltage in the distribution grids within the limits. The modelling of boundaries is extended by the equivalent injection and local control. The static voltage stability is being investigated by creating $V$–$Q$ curves and using the $\mathrm{d}\Delta Q/\mathrm{d}V$ criterion for all selected grid nodes (Clark in Power technologies, 1990). This method provides an indicator of the high voltage grid stability in total as well as the individual reactive power margin for each analyzed grid node. Finally, the node, which is closest to an instable operating point and the corresponding simulation scenario, is being detected.

Appendix

There are Tables 1, 2, 3, 4 and 5.

Table 1. Node data of the HV simulation grid

Table 2. Branch data of the HV simulation grid [11]

Table 3. Node and branch data of the MV grid connected to HV grid node $(0,1,9)$ [11]

Table 4. Node and branch data of the MV grid connected to HV grid node $(1,1,3)$ [11]

Table 5. Node and branch data of the MV grid connected to HV grid node $(3,1,8)$ [11]