In collaboration with:
Project Overview
Cellular networks require electrical power to operate, and increasingly, electrical network operators are using cellular networks to communicate. Electrical network operators historically have deployed independent, resilient communication networks to manage their operations, spending on average £58.5m per year on operational telecom networks[1].
The majority of these costs are used to communicate with high voltage (HV) substations, which provide electrical power to a large number of customers. Real-time control and monitoring of these assets is required, and accordingly, these systems require extremely reliable and dual communication links so that in the event of any communication issue, the electrical networks can continue to operate. On the other hand, low voltage (LV) systems traditionally lacked communication network connectivity. However, recently, there has been a dual push to increase network visibility and reduce network outages on the LV voltage network. Accordingly, recent innovations such as Automated Power Restoration Systems (APRS)[2] have been designed and implemented to manage electrical faults and automatically re-route electrical power to restore as many customers as possible. Moreover, as additional low-carbon technologies and distributed energy resources come online connected to LV network, it is also becoming increasingly beneficial to have resilient network connectivity on the LV network.
The Challenge
The challenge this presents is the increased scale of substations on the LV network compared to the HV network. Across the UK, there are approximately 15,000 HV primary/grid sites and an estimated 550,000 LV secondary sites. The higher the percentage of these secondary substations provided with communication, the more fine-grained reporting and control would be possible. However, under existing cost-benefit analysis models, it has been determined that for most of these sites, it would not be economically viable to provide dedicated connectivity as surrounding sites may be able to provide ‘good enough’ visibility and control considering the cost. However, this still results in an estimated 200,000 sites across the UK requiring connectivity. Accordingly, this still represents significant potential operational costs for electrical networks and, ultimately, the bill payer.
During the initial phase of CommsConnect, possible lower-cost methods of providing resilient connectivity to enable this large number of devices to more cost-effectively communicate were explored. Moreover, if the overall cost to provide connectivity to each site could be reduced, then the cost-benefit calculation changes and additional sites and corresponding control and visibility may become economically viable.
One possible method of this is employing existing public mobile networks as these networks feature wide-scale geographic coverage predicted to achieve 95% coverage by 2025[3]. Re-using this existing infrastructure could present a cost-effective solution. The UK Home Office has elected to use such an approach to replace their legacy TETRA-based system, Airwave. This system was used by the emergency services such as Police, Fire and Ambulance to communicate. The challenge the system presented was that the technology was becoming obsolescent, and maintaining this was becoming prohibitively expensive at an estimated cost of £450m per year[4]. As a result, the decision was made to shift to using a prioritised service on the public mobile network operator EE called the Emergency Service Network (ESN)[5]. Once operational, this nationwide network is estimated to save £250m per year.
However, one of the challenges of employing third-party networks for electrical network communication is ensuring this network has sufficient resilience for the safe and efficient operation of electrical networks. This is particularly important as these communication networks are reliant on electrical power to operate. Therefore, when the electrical power goes off, the communication may go down as well, depending on the levels of resilience installed, and this could create a circular interdependency.
Storm Outages
Storm Arwen in 2021 demonstrated this when there were extensive power cuts, affecting approximately 40,000 customers across the UK who were off supply for more than three days[6]. The majority of these faults were caused by strong winds or trees and branches falling onto power lines. These power losses caused public mobile communication networks to lose power and fail, with thousands of mobile cell sites disrupted by power outages, with 130 of these remaining off for 72 hours[7]. This restricted communication, making internal coordination non-standard, and coordination with external organisations, for example, a tree surgeon, more challenging, increasing overall restoration time.
Storm outages, while infrequent, are increasing in the UK due to climate change; meanwhile, the United States of America can provide an established baseline of storms. One such storm which hit New York City was Hurricane Sandy in 2012. This is significant because it represents a dense urban area with various overlapping cell sites and buried electrical infrastructure. During the storm, an estimated 25% of all public wireless communication was lost across an area of 10 states[8]. During the damage assessment between the 3rd and 5th of November, wireless coverage in the western half of Rockaway Peninsula was effectively non-existent. The storm also resulted in certain base stations being without power for 360-400 hours.
More recently, on 9th July 2024, Hurricane Beryl hit Houston, Texas, United States of America. 28.7% of 4656 mobile masts in the affected counties area went down as a result of this storm. 1,137 of these went down to the lack of power, and a further 490 sites were operational on battery backup power.
Public Mobile Networks
Public mobile network base stations may feature battery backup in the event of mains power loss to continue operating service. However, this is frequently measured in minutes rather than hours, as sustained power autonomy would be cost-prohibitive to install at every site with estimates it would cost approximately £460m to install 72-hour battery backup, the existing power autonomy recommendation for electrical network operators[9], at 15,000 mobile sites[10]. This represents approximately 19% of the total mobile masts within the UK and focuses on coverage sites. Mobile networks often provide overlapping mobile coverage, particularly in low-band frequencies, to enable network handovers and provide a consistent coverage layer. However, by default, mobile devices will connect to the mobile mast, presenting the strongest signal. The benefit of this approach for conventional mobile network users is that the stronger signal reduces energy consumption and increases throughput. For most users of public mobile networks, that is the preference. However, this would require an alternative approach for utility users, which normally have low data rates but high availability requirements.
Some public mobile networks feature reinforced power autonomy at key sites, and accordingly, knowing which sites and being able to steer utility assets to connect to these sites specifically would present a low-cost way of increasing effective resilience. This could be extended to steer the majority of electrical assets within an area onto a single mast, reducing the number of masts requiring reinforcement.
Project Aims
CommsConnect aims to create a detailed understanding of any such dependency issues and accordingly suitably detect and enable the prevention of creating an interdependency, where electrical network operators require communication networks to operate, yet communication networks electrical networks.
This increased understanding of the power autonomy of the public mobile communication networks would enable electrical network operators to make informed decisions. This would further modify a reactive situation of having to respond to loss of communications as and when it happens to a proactive approach, where estimated times to loss of comms are provided. This includes detailed real-time reporting on which electrical assets will lose connectivity and when power is lost. This increased information allows electrical operators to make informed restoration decisions and potentially run automated power restoration programs, ultimately reducing offline time for all electrical consumers, including cellular networks.
The Approach
CommsConnect took two approaches to provide additional visibility over cellular networks to the electrical network – direct integration and active network monitoring. Both proposals were designed, implemented, and tested within a control environment, enabling greater visibility to the state of the communication network.
The advantage of the direct integration approach was it could provide more detailed information. However, this approach was also more complex and would require significant engagement from cellular network operators, which would be more challenging to implement nationwide.
The advantage of the active network monitoring approach was that it could be achieved entirely independently of partnerships with cellular network operators but would not provide the same level of granular information. Both approaches were able to provide real-time status information about the state of cellular networks to electrical network operators.
Proposed Solution – Direct Integration
To mimic a public mobile network, a private mobile network was constructed, including a smart battery enabled mast and a mobile core network. This design was configured to detect a loss of mains power at the mast and report this outage to the core mobile network. The core network would then collect unique identifiers of all the connected devices on that particular mast and report this electrical outage, the list of devices and expected battery depletion to the utility operator. This architecture is visualised in Figure 1.
The benefit this provides is providing the utility operator with advanced warning of loss of communication, detailing which assets will be affected in advance, and provide a prediction of estimated battery life at the site. This would shift the model from a reactive approach responding to issues as they arrive, to a proactive approach where sites with longer remaining battery life can be deferred in favour of prioritising key sites or more at-risk sites.
This increased level of information about the resilience of specific geographic locations of the public cellular network would also enable utilities to make informed decisions about which communication technology should be employed. For example, an electrical network operator can use fibre, BGAN, VSAT, and public or private cellular networks. Each of these has its own benefits and considerations, but public cellular networks are one of the most cost-effective options to disperse assets, particularly for low data rates applications. This information detailing the connectivity status of each device, the mast providing the connection, and its status is visualised in the example dashboard shown in Figure 2.
Figure 1: Direct integration network architecture
Figure 2: Direct integration dashboard detailing real-time status information about cellular masts
Proposed Solution – Active Network Monitoring
One of the challenges of the direct integration approach is that it would require significant involvement from third-party cellular operators. To address this, the second proposal for providing utilities information about cellular networks was the independent active network monitoring approach. This method was able to gather information completely independently, using existing utility assets, thereby reducing the reliance on third-party operators and ensuring the reliability of the data.
Similarly to the direct integration proposal, the utility requirements highlighted what utility devices were connected to which mast, providing an understanding of the real-time status that would most benefit utility operators. Accordingly, a modified 4G router was turned into a monitoring device to be able to report unique cell identifiers, received signal strength, network operator code, and frequency of surrounding cellular cells. Using this information, it was possible to monitor the status of the cellular network. All the detected information is sent to a centralised server for reporting.
To demonstrate how this would work in practice, a user-friendly website was created. A number of secondary electrical substations, denoted by the light blue dots, were mapped. A subset of these substations would feature a cellular router/sensor device, and they would detect nearby cellular masts as denoted by the red/orange/green circles. The different colours denoted how recently one of the cellular sensors recorded that particular mast. The mast will appear green if detected within the last 15 seconds, orange if detected between 15 and 60 seconds ago, and red, indicating there may be an issue with that mast due to it not being detected.
This website also enabled users to hover over each substation and see which cellular mast the router is currently connected to. Similarly, hovering over the cellular masts will display all the cellular routers that could connect to that specific mast. This would enable electrical network operators to monitor which of the public masts can provide the most connectivity to their substations and focus on the resilience of those masts.
To demonstrate the functionality, the private mobile mast was rapidly de-energised, simulating a grid power outage. The monitoring device then locally recorded this outage in a database and, through an out-of-band communication link, reported this outage in real-time. Once power was restored to the cellular mast, the device was able to detect it again and report this to the dashboard. This demonstrated that the cellular router modified to become a sensing device would be able to detect the loss of a cellular cell.
These experiments of both the direct integration and active network monitoring approaches would provide electrical network operators with increased information about the status of the cellular network, increasing the viability of using low-cost public cellular networks as an alternative communication mechanism.
Socialisation
Several events were organised to demonstrate these systems and engage with the industry. These included Distribution Network Operators (DNOs), Transmission Operators (TOs), Gas Distribution Networks (GDNs), water companies, regulators, government departments and standards development organisations.
UK Power Networks Innovation Day: Net Zero Networks
UK Power Networks held an event at IET in London on 21 February 2024 for all their SIF round 2 innovation projects to disseminate information about their progress and encourage collaboration. At this event, CommsConnect delivered a presentation, initially outlining increasing reliance on mobile communication technology. This then developed into explaining the two approaches CommsConnect has to increasing information exchange about the status of the mobile network. The event also featured a small number of demonstration stands. CommsConnect took this opportunity to give participants a tabletop demo of the active network management system. This demonstration effectively engaged attendees with the project, and the event offered opportunities for further detailed engagement with organisations who attended.
CommsConnect PNDC Workshop
A workshop was organised and arranged at PNDC to present the project to a wider community of people working in communications for power networks. The workshop was designed around several distinct parts. The first was a more general introduction to the problem and the development of wireless communications and their potential. Secondly, there was a more detailed description and demonstration of the project, including a demonstration of Active Network Monitoring and a tour and demonstration of the mobile network functionality available through the Direct Integration approach. Thirdly, there was a panel discussion session, which was organised as a large round table, to encourage all participants to share their views on what they thought about the project and its potential.
The event was organised as a hybrid event, with participants able to join online and participate in the demonstrations and tours. A PNDC staff member was tasked with following the tours with a camera to stream to virtual participants.
Figure 3: CommsConnect workshop at PNDC
Engineering and Physical Sciences Research Council (EPSRC) Risk and Resilience Day
The CommsConnect project team submitted a project abstract in December 2023 for inclusion as part of the oral presentations at the Risk and Resilience Day event organised by EPSRC Supergen Energy Networks Hub.
The University of Newcastle hosted the 2024 iteration of this event. CommsConnect’s application was successful, and the project was presented on as part of the Industrial Innovation session. There were high levels of interest in the project from those in attendance, and an engaging questions and answers session was held following the presentation. Keynote speakers at the event included Professor of Complex Systems Liz Varga from University College London and Chief Engineer (Interim) at Ofgem Martin Queen. Poster sessions and oral presentations were conducted throughout the day by various academic institutions and industry parties, all of which focused on the topic of risk and resilience within the electricity network space.
Figure 7: CommsConnect Presentation at Risk and Resilience Day
Conclusion
The CommsConnect project aimed to provide electrical network operators with more information about the state of cellular networks. This was achieved through first understanding which information would be most beneficial to electrical network operators. This was defined as information about which specific cellular mast utility devices connect to and information about those masts, including their location, unique identifier, and available power autonomy.
The CommsConnect project then proposed two methods of sharing this information, the first through direct integration and the second through active network monitoring. Both solutions were implemented and tested in a controlled environment. This demonstrated that both could provide the required information.
Direct Integration provided the most detailed and immediately actionable information, as it can provide the exact remaining battery backup and an estimated time to full depletion. In contrast, active network monitoring would have to infer each site’s supplied power autonomy based on historical and future performance. This means that uncertainty regarding power autonomy will remain until an electrical outage occurs, during which time the system could be naturally profiled.
The benefit is that implementing a wide-scale active network monitoring could be undertaken independently, whereas direct integration would require significant involvement and assistance from third-party cellular network operators. The project aims to continue developing this information exchange process through a staged approach, first developing a data-sharing agreement with relevant third parties.
Testimonials
Nokia was very pleased to be able to work with the University of Strathclyde, PNDC, and UK Power Networks to enable the successful delivery of the Alpha phase of the CommsConnect project. Nokia’s AVA platform was central to this success, leveraging AVA’s flexible APIs in order to make visible, understand and process key battery metrics from the mobile radio sites. Nokia sees this as a light, but powerful example of the benefits of integrating AI and Analytics capabilities into a power utility’s operational toolkit to enable them to build further resilience in the power network.
The CommsConnect project demonstrated PNDC’s productive participation and valuable engagement in successfully delivering a solution that provides distribution network operators with critical insights into mobile network status through both direct integration and active network monitoring, showcasing the importance of such a solution and its successful integration for enhancing the stability and resilience of the grid. Working with UK Power Networks, SSEN Transmission, and Nokia in such a challenging and important area to ensure resilience was a valuable and collaborative experience for all project parties.
Reference List
1 https://www.ofgem.gov.uk/decision/riio-ed2-final-determinations
2 http://cired.net/publications/cired2017/pdfs/CIRED2017_0871_final.pdf
3 https://commonslibrary.parliament.uk/research-briefings/sn07069/#:~:text=According%20to%20Ofcom’s%202023%20Connected,7%25%20had%20no%204G%20coverage
4 https://questions-statements.parliament.uk/written-questions/detail/2024-01-10/9008
5 https://committees.parliament.uk/publications/40869/documents/199069/default/#:~:text=In%202015%2C%20the%20Department%20awarded,has%20but%20which%20commercial%20mobile
6 https://www.ofgem.gov.uk/sites/default/files/2022-06/Final%20report%20on%20the%20review%20into%20the%20networks%27%20response%20to%20Storm%20Arwen.pdf
7 https://www.ofcom.org.uk/siteassets/resources/documents/research-and-data/multi-sector/infrastructure-research/connected-nations-2022/connected-nations-uk-report.pdf?v=321647
8 https://www.ofgem.gov.uk/decision/riio-ed2-final-determinations
9 https://www.spenergynetworks.co.uk/userfiles/file/App8_2012_ENA_EngRecommendG91BlackStartRes.pdf
10 https://www.jrc.co.uk/Plugin/Publications/assets/pdf/ICT-Economic-rationale-for-enabling-Smart.pdf
Get in touch
For further information on this case study and to discuss collaborative opportunities, please get in touch.
Established in 2013, PNDC is one of the University of Strathclyde’s industry-facing innovation centres and focuses on accelerating the development and deployment of novel energy and transport technologies through multiple collaboration models and open access facility provision.
Find out more about PNDC partnership opportunities and follow us on LinkedIn.