By George Konstantopoulos and Pablo Rodolfo Baldivieso-Monasterios, University of Sheffield
Local energy systems can be characterised as ‘smart’ based on the intelligent and autonomous decisions that take place within the system. Given that the active components of smart local energy systems (SLES) mostly consists of Distributed Energy Resources (DERs), such as renewables, storage and active loads, the intelligence of the system mainly lies in the way we ‘control’ and ‘operate’ these DERs.
Consistent communication between all DERs is essential to meet the objective of a SLES to optimise performance and efficiency. A central monitoring and control unit collects data such as energy generation and consumption, temperature and battery State-of-Charge (SoC), and then executes an optimisation algorithm that describes the desired performance of the SLES. This can be used to dictate the action for each DER component e.g. how much power to inject/draw from the network at each set time interval – these time intervals can be very small, i.e. few seconds. However, this approach is costly, potentially vulnerable to cyber-attacks and requires reconfiguration when additional energy units are added to the SLES.
EnergyREV researchers have therefore responded to the following question:
Can we accomplish the same level of intelligence in a SLES without requiring a centralised control unit and using minimum communication links between DERs?
The answer lies with the concept of ‘distributed control’, which is based on local computations taking place at individual DER units. This information is only available locally and only through neighbouring units, removing the need of central monitoring and control operator. Among the several ‘distributed control’ methods available, the most common ones are based on collaboration between neighbouring DERs to achieve the same goal such as with power sharing, electric grid voltage/frequency regulation, or to solve a desired optimisation problem in a distributed manner.
‘Distributed optimal control’ is preferred when flexibility is required to modify the objective of an SLES, for example when:
- online adjustments are required based on electricity tariffs change, user/operator preferences;
- constraints such as power flow constraints or generator/consumption limits are necessary.
EnergyREV researchers have successfully implemented the ‘distributed optimal control’ concept in practice at the ADEPT SLES demonstrator in South Wales (http://www.infiniterenewables.com/adept-microgrid/).
Originally the control system of the ADEPT was centralised - information from all DER units was collected at a central hub unit, where the control algorithm for the optimal power management of the micro-grid was executed. Real-time data of power injected or drawn by the DER units was transferred to the central controller, together with additional information on the status of the units, such as the SoC of the battery and the price of electricity. Based on all this data, the controller executed an optimisation algorithm to identify the desired set-point for the power that had to be drawn by all the controllable DERs to achieve the desired objective. In this case the objective was to minimise the price of energy consumed by the local load, minimise the power drawn by the main grid, and maximise the lifetime of the battery unit.
The ‘distributed optimal control’ approach developed and tested in this case study is based on the data typical available in other systems such as instantaneous power values, battery SoC. By only using data from neighbouring DER units within the system and with additional capability of handling constraints in the power generation, storage and demand, the system becomes more efficient and resilient. For example, where there is a loss of communication between two individual units, the remaining DERs will continue to operate effectively. This distributed optimal control also enables additional energy resources to be integrated without having to reconfigure the entire system.
With the ADEPT case study, EnergyREV researchers have shown how the ‘smartness’ of a SLES can be integrated with minimum communication links between DERs, enhancing the reliability and resilience of a real SLES, thus paving the way towards implementing distributed control and coordination of actions for energy resources available in the PFER demonstrators and other local energy systems.