ResearchTSSG News

RESERVE – Renewables in a Stable Electric Grid

By 24th July 2020 No Comments

In September 2016, as part of a consortium of partners from across Europe led by Ericsson, the TSSG attended the kick off meeting for the RE-SERVE project in Aachen, Germany.  Fast forward to summer of 2020 and the technology developed by TSSG within this project has been listed on the European Commission’s Innovation Radar under the topic ‘Smart & Sustainable Society’ – a significant achievement for the TSSG developers.

The aim of the RE-SERVE project was to find ways to stabilise energy systems in the face of a shift away from fossil fuel based synchronous power generation towards less predictable but more renewable energy sources like Solar and Wind. From a power systems perspective Frequency and Voltage Control are the most researched and best implemented methods to provide that stability. However a large part of the RE-SERVE project was centred on ICT and how low latency requirements could be overcome with the use of 5G and also how interoperability issues could be overcome with the use of defining appropriate architectures to further refine frequency and voltage control in a 100% renewable environment.  For all the technological aspects of the RE-SERVE project it was seen that these aspects would not be capable of being realised without effective regulatory investigations and changes to ensure that these techniques would have the desired impact.

Primarily, the role of the TSSG in the project was based on five main activities:

  1. working closely with ESB Networks, the owners of the trial sites in Ireland, enabling the deployment of the voltage control algorithms in both laboratory and trial site settings;
  2. the gathering of the ICT requirements for all aspects of the project to ensure that the techniques could be implemented;
  3. the development of a set the appropriate communications mechanisms so that the relevant data could be retrieved from the trials and the monitoring of the devices could be achieved
  4. publishing a paper titled Deriving policies from connection codes to ensure ongoing voltage stability, which was centred on monitoring the impacts that a voltage control algorithm has on the power system from a regulatory perspective using policies extracted from regulatory documents.
  5. and also the definition of 3 main architectures that enabled the implementation of the techniques overcome a significant interoperability hurdle.

The four aspects of this role had the overarching goal of effectively implementing a voltage stability technique called Active Voltage Management (AVM), which was developed by University College Dublin, on a range of trial sites across Ireland and to achieve this several investigations were required in terms of the data transfer protocol required, the capability of the inverters at the trial site, the level of access available at the trial sites and also the network capabilities at the trial site.  The outcome of these investigations highlighted  already known interoperability issues that are common in the deployment of such techniques in grid systems. To help overcome these issues three common architectures were developed with the Volt Var Curve (VVC), the output of the AVM algorithm being generated and executed at different points in the system and with each architecture suited to scenarios with varied levels of access, connectivity levels and technological capabilities. The three architectures identified were Centralised, Decentralised and Hybrid-Edge and the following presents these architectures and a description of the scenario that they aim to address.

Centralised

In modern grid systems the presence of Aggregators and their control of the RES devices via bespoke Distributed Energy Resource Management Systems (DERMSs) is becoming more common. This is acting as a type of firewall in terms of control of the device from a centrally controlled grid system that looks to ensure the stability of the overall network from a voltage perspective by deploying algorithms like AVM across the grid. To overcome this, cooperation from both the aggregator and the grid operator is required and it is necessary that both systems communicate to ensure that the AVM gets deployed and voltage stability is maintained. In this case, more often than not in modern systems, this communication will be cloud-to-cloud and a centralised architecture for the AVM execution would be most applicable. This architecture would afford the aggregator with a layer of data agnostics with only the values applicable to the AVM execution being sent to the system operator for the generation of the reactive power set-point. Given that very low latency is not a high priority requirement for AVM and that the packet size and message frequency is low, core to the ICT deployment is the security of the communication and privacy with how the data is treated. Security and privacy are key and the responsibility for this is on both parties. The video below demonstrates the implementation of the AVM technique in a centralised setting where the required measurements are sent to the cloud so a calculation is made and the result sent down to the device as a setpoint.

Figure 1. Centralised ICT Architecture 

Decentralised

With the Distributed Energy Resource deployment becoming common and with the potential of some of these units being deployed to geographical regions with limited network connectivity comes the need for the control mechanisms, like AVM, to be deployed in a way that is not dependant on having a full-time dedicated communications link. This would mean that any deployment, as far as possible, should be stand alone and the most applicable architecture for this would be the decentralised architecture.  This method of deployment would rely heavily on the hardware and software capabilities of the inverter for AVM deployment but from an ICT perspective it would only require communications to send the VVC to the inverter and if the VVC were changed to update it on the inverter.  Inverter capability in this case is a key factor with the level of intelligence provided by the inverters ranging from being able to receive and execute the algorithm on the device to just being capable of receiving the setpoint as a command on the device via a supplementary device like a RaspberryPi.   Latency, frequency and message size are not an issue in the deployment of this architecture as all the data needed is either on or beside the inverter. Protection and device access either in a physical or virtual way must be ensured to maintain the integrity of this deployment but from the perspective of network security and privacy the scope for attack is very narrow due to minimal external data transfer and communications.  The video below demonstrates this approach and details how the VVC is sent to the device as a configuration payload and the full execution of the algorithm occurs either on or very near the inverter.

Figure 2 Decentralised ICT Architecture

Hybrid-Edge

In urban areas the scope to deploy DER is increasing with battery storage and rooftop solar PV becoming more common and easier to integrate into the power grid. This scenario is centred around providing voltage stability to an urban area where there is a high saturation of DER. The issue to be addressed with this scenario would be that the DERs connected to the grid may be owned by different entities and may have different software and hardware capabilities but given that their proximity may be close their combination may impact negatively on stability, thus viewing, analysing and controlling these as one entity may be required. To achieve this using a centralised architecture would require a great degree of complexity and would involve a large volume of network traffic potentially travelling over large distances. While a decentralised approach would be an option it would require frequent updating of the VVC at each DER due to the saturation of instable RES sources causing a real time lack of system awareness thus this approach would not provide an adequate level of dynamism to cater for the complexity of the system. Therefore, the most suitable solution would be the Hybrid-Edge Computing architecture. It would add a layer of control between the central controller and the DER to hand off the system awareness at a granular level to software hosted at communications base stations. This secondary layer would handle the control of DERs in its immediate geographical region whether this control be the direct propagation of the VVC to the inverter or the receipt of readings from the RES device and the execution of the AVM algorithm at the base station with the reactive power setpoint being returned. This method of deployment of the AVM is heavily reliant on robust and available communications on the base station to ensure that voltage stability can be maintained across the cluster in unison and security at this point is also key due the impact of a malicious attack at a secondary control would be service effecting.  The video below demonstrates the how the Hybrid-Edge architecture distributes the execution of the VVC in a distributed type setting.  Note here the implementation of both the decentralised and centralised execution of the VVC at the base station.

Figure 3 Hybrid Edge ICT Architecture

Impact

While the three architectures presented above and their implementation are set in the energy domain and specifically in the management of voltage stability, given the real-time aspects shown in the decentralised and hybrid-edge architectures, these architectures could be transferrable to other applications in the energy sector like autonomics, self-healing, energy trading and device control and indeed could be used in other domains like Agriculture, Industry 4.0 and Health.

Indeed this approach could be taken in any application where the proliferation of IoT devices and the data they generate may need to be processed, analysed or controlled in a low cost and real-time way with the distributed computing power of the edge applications taking the place of Big Data solutions that are expensive to implement.

Central to this approach is, for the most part, the need to have a lightweight and flexible messaging system to securely transfer the data in as close to real-time as possible between the applications both at the edge and in the cloud. In the RE-SERVE project, MQTT was used to enable these communications as it both has a broker element that can be secured and also a client that can be deployed to devices with varying levels of computing resources.  In investigating MQTT as a solution, at the time, it was noticed that there wasn’t an implementation of the MQTT broker readily available that could be configured at the time of deployment with varying levels of security.  To facilitate the different security requirements across the trial sites we extended the implementation of MQTT available at the time by developing a Docker based deployment that was parameterised to enable authentication and encryption configurations to be deployed within the container.  In keeping with the open source philosophy held within the organisation the extended implementation was openly published and is freely available on the DockerMQTT project on the TSSG GitHub page.

This novel concept of extending the MQTT capabilities by developing a secure docker based deployment awarded the TSSG researchers a spot on the European Commission Innovation Radar; an initiative to identify high potential innovations and innovators in EU-funded research and innovation framework programmes. The Innovation Radar is divided into four market maturity categories – Tech Ready, Business Ready, Exploring and Market Ready – this innovation developed by TSSG is listed in the latter as it addresses the needs of existing markets. As the RE-SERVE project investigates Renewables in a Stable Electric Grid, the ‘Docker environment for the Eclipse Mosquitto MQQT  broker’ technology is listed under the Smart & Sustainable Society category. The TSSG and Waterford Institute of Technology have been involved in a number of European projects which have innovation technology solutions also listed on the Innovation Radar. Click here for further details.

Not only has the RE-SERVE team welcomed the success on the Innovation Radar, the consortium have welcomed a wide range of accomplishments since its completion in March 2020. The project passed the final review with very positive feedback and impressive results in terms of the papers published   , research carried out   and impact achieved with reviewers from the European Commission.  A measure of this success is seen in the decision to publish an article about the RE-SERVE project in the ‘Results in Brief’ section of the European Commission’s CORDIS website. A further measure of this success is the funding awarded to a new project called EdgeFLEX , which kicked off on April 1st 2020. It will take some of the outputs  from the RE-SERVE project and look to apply the research and applications developed in the context of Virtual Power Plants in an effort to further advance the digitisation of the energy sector and aid the transition to a less carbon reliant energy sector.