Energy Storage for Control of Distributed Photovoltaic Power Systems
MetadataVis full innførsel
- Institutt for elkraftteknikk 
The residential energy storage market is growing at a rapid pace in many countries. Increasing the consumption of locally produced energy and the use of energy storage as a backup source have been the main reasons for residential customers to be encouraged to install battery-based energy storage systems. The storage capacity connected to the grid increases with the accumulated use of home storage systems. Increased storage capacity has several benefits for the grid operator. A grid with significant storage allows for increased capacity of power generated from renewable sources. That is due to the fact that storage contributes to maintain stable operation of the grid against variations of power generation from these intermittent sources. Storage systems connected to the distribution grid can help the Distribution System Operators (DSOs) to manage power flow and to alleviate power quality issues. However, these potential benefits can only be realized with properly designed control strategies. Therefore, the objective of this thesis is to suggest, develop and verify effective control strategies for customer-owned Battery Energy Storage (BES) systems, which are co-located with photovoltaic (PV) systems. In this work, these systems are assumed to be connected to low voltage grids with high level of PV penetration, to reflect future trends and developments that are already taking place. Distribution grids are usually unbalanced to a certain extent. Therefore, the proposed control strategies are designed for three-phase unbalanced grids while the compatibility is ensured for balanced grids as well. In the proposed methodologies, the primary control objective is to fulfil the requirements of the BES owners. The BES systems should however also support the local grid to alleviate over-voltage problems, which is a common issue experienced by distribution grids with high PV penetration levels. But the use of customer-owned BES units for such purposes should not adversely affect the BES owners’ local objectives. In this thesis, two methods to control the operation of customer-owned BES systems in a grid-supportive manner are proposed, which do not adversely affect local objectives. The first method is a local control strategy, in which the control set points are determined by a local controller based on locally extracted information. A receding horizon control approach is adopted to determine the BES set points. In this approach, an optimization problem is solved repeatedly over a moving time horizon to determine the set points. The objective function of the optimization problem can be locally decided according to the requirement of the BES owner. The Home Energy Management System (HEMS) implements the local objectives. Details of HEMS systems are provided in the thesis and parts of their operation are experimentally tested and verified. The critical period is when the PV production is potentially highest in the middle of the day. The high PV production results in excessive reverse power flow in the grid, which can cause over-voltage problems. Constraining the charging operation of the BES systems to this critical period is a requirement set by the DSO to lower the reverse power flow. The optimization problem is formulated such that the BES systems maintain charging operation over the entire critical period without reaching the maximum state of charge until the critical period is over. Consequently, the reverse power flow in this period will be reduced and the over-voltage issues will be alleviated. The second method proposed to control the BES set points, is based on a distributed control approach. In this approach, the local controllers of the BES systems determine the set points in the same way as in the first method, whereas a central controller may modify these set points if required. The set points computed from the local optimization process are modified if they fail to maintain the voltages within the statutory limits. These modifications are done during real-time operation. The central controller coordinates the charging set points among the three phases. In this work, we assume that the BES units are connected to the grid using three-phase converters and the per-phase set points of the converters are independently controlled. The active power capabilities of the BES systems can be maximally utilized for voltage support through coordination of the charging operation among the three-phases in an unbalanced grid. This will reduce the unnecessary use of reactive power for voltage support and enhance the performance of the overall grid section. Both control strategies mentioned above are accompanied by reactive power control as well. Reactive power support from the converters is utilized only when the cumulative storage capacity available in the grid is not sufficient to mitigate the over-voltage problem completely. For the local control strategy, the proposed reactive power support solution is based on a gradual increment of reactive power support from PV inverters. The PV inverters respond to a broadcast signal received from a monitoring device located at a strategic node in the network. For the over-voltage case, the PV inverters decrease their power factors gradually until the problem is solved. For the distributed control strategy, the central controller determines the optimal power factor set points of the converters that can maintain the voltage profile of the grid within acceptable limits. Simulation studies were carried out to evaluate the performance of the proposed control strategies. The two control strategies were implemented in an IEEE low voltage test feeder and the voltage profiles of two critical nodes were observed. Both control strategies managed to improve the voltage profile of the network effectively. The distributed control strategy was able to maintain the voltage profile within acceptable limits with reduced reactive power support when compared to the local control strategy. A study was also conducted to evaluate the effect of the DSO’s involvement in the control of customer-owned BES systems, on the local objectives of the customers. A rule-based control strategy was used as the base scenario for this comparison. In this rule-based control strategy, the rules were carefully defined to maximize the economic benefits for the customers. From this study, it was discovered that the proposed control strategies do not adversely affect the local objectives. The difference between the cost savings, the self-consumption rates, and the self-sufficiency rates calculated over the period of a month were negligible with or without the involvement of the DSO. Therefore, the local objectives are not adversely affected by the proposed grid-supportive operation of the customer-owned BES systems. The main contributions of this work have been the two suggested control strategies, the accompanying reactive power support solutions, and their validation for three-phase unbalanced grids. In addition, it was demonstrated that customer-owned BES systems can be used by DSOs to alleviate over-voltage problems in distribution grids without compromising the local objectives of the owners. Grid codes and market models can therefore be designed with this in mind.