Electricity – The urgent challenge of improving resilience
This updated article has been written for the National Preparedness Commission by Robin MacLaren, Chair of the Resilience Working Group at the Institution of Engineering and Technology.
Society depends on electricity and that dependency has increased significantly over the last century to a point where it would be difficult for our civilised society to exist without it. As use has grown and technology developed, society has become more and more dependent on functionality driven by electricity, whether that be the internet, water supply or healthcare.
In recent times, our electricity supplies have been extremely reliable and from early days when power outages were common we have now moved to a point where a customer may typically see an interruption less than once in every two years, and for an average duration of around 30 minutes.
However, electricity production and distribution are complex. While many of us have experienced short interruptions, usually locally, rarely are events widespread, except perhaps in circumstances of severely adverse weather where restoration of supplies to small groups of customers may sometimes extend to more than 24 hours. Given the wider distribution of society, however, other parts of the community continue to function and extreme, widespread events are rare.
One such rare event occurred on 9 August 2019 when the near simultaneous loss of an offshore wind farm in the North Sea and a gas-fired power station in East Anglia resulted in a loss of supply to just over a million customers. A national shutdown was averted by protection systems automatically kicking-in to reduce demand immediately through a predetermined sequence of disconnecting supplies to customers until the power system regained stability. Although the power system was restored within 45 minutes, the event resulted in ongoing and widespread disruption, illustrating the dependence of society on continuous and uninterrupted electricity supplies. In the event, society continued to function reasonably normally as the great majority of customers across the country were unaffected.
However, although still highly improbable, there is a trend of escalating risk in a power system which is transitioning away from traditional, centralised generation and becoming increasingly reliant on new technologies associated with weather-dependent generation e.g. onshore and offshore wind and solar PV. These technologies not only make the power system less resilient to large or coincident multiple losses of infeed (i.e. from generators and/or interconnectors), they would also make it much more complex and slower to restart the system should a nationwide shutdown occur.
Thirty years ago it would have been possible nearly to guarantee that such an event might take, at its extreme, 24-48 hours to recover from with supplies to large parts of the country restored far more quickly. Today, such an event is increasingly more challenging as we move to a lower carbon-electricity industry and continue to close further large, traditional generating stations. This is a particular concern given that this has happened concurrently with a greater dependency on electricity including, for example, 5G, data and smart devices and electric transport. This is further exacerbated by the interdependencies between multiple and ‘independently governed’ systems (e.g. energy, communications, digitalisation, etc.)
This conflation of issues is compounded by geographic changes of generation sources from centrally located, controllable, large generation stations to a combination of remote sources with long transmission paths for electrical energy (e.g. from offshore-wind generation) and widely distributed decentralised sources (e.g. onshore wind and solar PV generation). The latter are generally unable to operate until the grid is re-established i.e. they currently lack the ‘grid-forming’ capability to operate as a power island.
A further issue is that improvements in digital communications and control technologies – in the past, dependent on simple analogue systems – have enabled more efficient operation but with many of these systems dependent on electrical supply. There is a circular dependency here and if this circle is broken it further challenges our ability to ‘restart’ the power system following a shutdown.
In summary, for all the reasons given, our ability to restore supplies following a system shutdown has, therefore, become more complex, more interdependent, and hence more challenging.
Basically, impact and time for recovery are currently bounded by best effort, working with ‘unproven’ plans that are informed by modelling and desktop exercises. No framework is set by society as to what time limits should apply against which the necessary technical framework can be implemented. The management of the costs to the economy of total grid failure risk, or societal impact, are not built into regulatory models so cost pressures make technical solutions difficult to justify from a regulatory perspective. This emphasises efficient, co-ordinated and economic network development.
In the past few years, the Cabinet Office’s National Risk Register has positioned the risk and impact to society of ‘supply failure’ second to a pandemic and has elevated the perceived risk by extending the assumed recovery period from a couple of days up to seven days for full restoration. Some in the industry view this as pessimistic, others optimistic. Whilst, historically, we have moved towards a more reliable and efficient electricity system, the potential risk (i.e. probability and impact) of a national shutdown has increased.
Work has now commenced to address these risks, with an objective of achieving a proposed standard for the electricity system to achieve recovery following a national shutdown to, at worst, 24 hours for 60% of the load in each of a number of zones covering the country, and allowing five days at the maximum to achieve 100%, setting what is a first major nationwide resilience standard for the industry to achieve by 2026.
When considering resilience there is obviously a wide range of definitions for electricity, and risk consequences and costs all drive the degree of resilience we can expect. The solutions may range from technical through to social.
One of the significant complexities we are building into our society is the interdependence of systems e.g. between electricity supply, gas supply, the internet, water supply, communications infrastructure and healthcare, where any failure in any of these systems may have widespread consequences. Most sectors have good plans in place to try to minimise the risks in their own sector but often the interfaces are not co-ordinated with others.
Electricity is a good example: few backup arrangements, whether in communications, water, electricity or healthcare, could ride through a seven-day electricity shutdown. We need to consider all the interdependencies between and across sectors when addressing resilience and the degree of resilience required.
The proposed new standard is a first government statement of the standards that society expects. It reflects the fact that while electricity customers require regulatory protection, many indirect users of electricity, from the water industry, other utilities, through to transport control (e.g. traffic lights) also require continual supplies.
In view of this, the key messages are:
- The increasing risk (probability and impact) of a nationwide power system shutdown needs to be managed within an accountable governance framework that recognises cross-sector dependencies.
- Improving resilience and recovery capability will be technically challenging and may be accompanied by considerable costs, requiring examination of practicability and mechanisms for permitting exceptions. The findings of a current ‘discovery’ project – Distributed ReStart – being expedited by the Electricity System Operator in conjunction with a Distribution Network Company – will be highly informative in this respect.
- There is a degree of urgency to address the challenges as national trends (e.g. Net Zero 2050 emission reduction targets) are increasing the resilience challenges year on year.