Reactive power control with the distributed energy resources

Future distribution networks must operate according to the ADN operational targets, even independently as a microgrid. One essential goal is to provide local and system-wide AS by DERs. A potential local AS is the reactive power control by the power electronics connected DER. This section presents a case study of the DER units offering local AS for the Sundom Smart Grid. The Sundom Smart Grid is a local smart grid pilot in Finland, Vaasa, located in a suburban/rural area, offering a novel research platform to develop ADN solutions.

Figure 20 outlines the Sundom smart grid. The primary substation connects the 110 kV and the 21 kV grids, and the secondary substations connect 0.4 kV LV distribution networks to the power system. A 3.6 MW WT is connected to the MV bus with its own short feeder, and a 33.6 kW PV unit is located in the LV grid.

Around 2500 metering points are consisting of residential and small commercial electricity users. The peak power was about 8 MW in 2018 and is increasing as housing in the area grows. The DG units, electric vehicles (EVs) and BESS, are assumed to increase at the customer premises. Customers use the electric heating systems, which can be direct, partially storing, storing, or heat pumps. Also, other energy sources are used for heating.

Figure 20. Outline of the Sundom Smart Grid.

Publication III presents the different requirements for reactive power flow between the distribution and transmission grids considered for the Sundom Smart Grid. The “future reactive power window” for the Sundom Smart Grid is based on the requirements for reactive power management for the transmission grid-connected distribution systems (EU, 2016), conditions set by Finnish TSO, Fingrid (Fingrid, 2017), and non-detection-zone requirements for microgrids (Laaksonen

& Hovila, 2017, 2016; Uebermasser et al., 2017). TSO’s reactive power window specifies the volume of reactive power delivered or received through the distribution/transmission connection points without penalties or separate compensation.

The future reactive power window for the Sundom Smart Grid was utilised for a reactive power control scheme formulation. The TSO’s requirements and the requirements for reliable islanding detection for the microgrids were considered.

The future reactive power window for Sundom Smart Grid (Publication III, Figure 3) is updated in Figure 21 because of the revision of Fingrid’s requirements (Fingrid, 2021) which relaxed requirements when feeding in active power. Also, estimating the peak power (or energy in one year) increase (forecast 2030) in Sundom is considered. In the active power consumption situation (Poutput), the reactive power output limit QD and input limit QD1 are applied. In the active power production situation (Pinput), the reactive power output limit QG and input limit QG1 are applied. The QDmin is 2 MVAr in a transmission line connection and 4 MVAr in a substation connection. The QDmax is 50 MVAr.

Figure 21. Future reactive power window for the Sundom Smart Grid.

The reactive power control algorithm for the WT’s full-scale converter was developed to limit the reactive power flow at the POI, according to the yellow section in Figure 21. The algorithm used the year data generated based on one-month measurement data from the Sundom Smart Grid. The results show (Publication III, Figures 4 and 6) that a coordinated reactive power management scheme across the different voltage levels by utilising control of DG could benefit the voltage support AS and a microgrid reliable islanding.

Further research interests arose based on the results. More reliable results would be obtained through implementing one-year measurement data from both consumption and WT power generation. The cabling degree increases in the Sundom Smart Grid. Therefore, the adequacy of the WT converter’s reactive power capacity should be verified. The number of PV units increases, so the PV inverters could be utilised in the reactive power management accordingly.

Publication IV (Prospects and Costs for Reactive Power Control in Sundom Smart Grid) presents feasibility case studies for reactive power management in the developing Sundom Smart Grid by utilising MV and LV network connected DER units, developed control algorithm (Publication III), and the nine-month one-hour average measurement data. The main challenge in reactive power flow management is low consumption time since the cables generate reactive power when they are lightly loaded. The simulations showed that coordinated reactive power management across the voltage levels utilising various DERs (PV and WT) is technically and economically significant for developing future distribution networks. The case studies demonstrated that the effect of the LV grid-connected PV units with 4000 m² of panels (or 40 units) and inverter power factor control (cosϕ = 0,8ind), the PV inverters consumed reactive power was up to 250 kVAr, which is minimal considering the reactive power flow at the transmission and distribution networks POI. The panel area should increase notably to affect the reactive power flow at POI via the LV level PV units.

Further development of the reactive power flow management scheme is necessary.

A coordinated operation with voltage control devices is needed. Therefore, they are included in the total management scheme. A possibility of active power control affecting reactive power control could be established using an energy storage system, through DR, or by limiting the generation from the WT and PV units.

Modelling the LV distribution networks, energy storage systems, and DR actions and developing the active and reactive power control algorithms are essential for voltage control AS. The developed simulation model could be utilised in future studies of the Sundom Smart grid. More accurate results could be obtained by implementing one-year measured data. Further, modelling a 21 kV-side reactive

power window is essential for reactive power control studies of the microgrids, in which more frequent measured data can be implemented. Adapting the offline models for the control algorithm and the power system for the real-time simulation platform builds a frame for HIL testing of the developed reactive power controller.

Publication III and IV provide the means for further studies of the voltage control AS provided by the Sundom Smart Grid or utility grid-connected microgrids.

Publication IV has been cited in Hafezi & Laaksonen (2019), which presents a soft open point (SOP), the power electronic devices installed in the normally open points of the distribution networks, for the voltage control ASs. SOPs can provide active power flow control, reactive power compensation and voltage regulation in the network normal operating conditions, and fast fault isolation and supply restoration in the disturbances.

The management of the reactive power flow at POI of distribution and transmission networks has been topical. Retorta et al. (2020) presents a local market and a mechanism for managing reactive power flow according to the TSO’s reactive power profile release, to which market agents (managing the DERs) can send bids. Stanković et al. (2021) presents a reactive power flexibility measure through flexibility coefficients based on the DSO capability and TSO desirability surfaces.

5 REALISING CONCEPTS FOR SMART GRIDS

This section focuses on the fourth research question: What kind of development and testing platforms are necessary for developing control functions for the microgrids or the active distribution networks?

New solutions for realising ADNs and ADNM scheme are being developed universally. An essential driver for the implementation of the microgrids is the more efficient DERs integration involving several stakeholders and actors. Hence, the control and protection of microgrids require novel technical solutions and applications. Standardisation is critical in developing multi-vendor ADN and microgrid systems where interoperability is mandatory, providing compliance of different and various vendors’ solutions.

This section presents the development procedure of a reactive power control ADNM scheme from a control algorithm to a hardware controller for developing control solutions and related testing methods. First, however, the relationships among standardisation and product development and the simulation and testing methods are presented. Conclusively, this section discusses the significance of testing and piloting in living laboratories.