The events of 9 August 2019 are indelibly etched on the memories of network engineers and business leaders. On that summer’s day large parts of England and Wales were left without electricity following a major power cut that had a serious impact on rail and road services. Passengers were shut out of some of the country’s busiest train stations during the Friday evening rush hour, while hundreds of thousands of homes were left without electricity.
The investigation that followed found that the combined loss of two large generators after a lightning strike, as well as the smaller loss of generation at a local level, together triggered the subsequent disconnection, loss of power and disruption to more than one million consumers.
A few years earlier on 5 December 2015, Storm Desmond caused unprecedented flooding in north Lancashire and Cumbria. In Lancaster, the main electricity substation was flooded, cutting electricity supply to 61,000 properties. The loss of power quickly affected other vital services. The mobile phone network, the internet, television, DAB radio and so on, were all knocked out.
The energy system in the UK boasts a 99.9 percent resilience and we’ve never experienced a national blackout. But the absence of a plan to deal with low probability high impact events have untold consequences. So though highly unlikely we still need a back-up plan to restart the grid as the energy system undergoes seismic change transitioning to net zero.
For the past three years National Grid ESO, in partnership with Scottish Power Energy Networks and TNEI, has supported an unprecedented innovation project known as Distributed ReStart to explore the potential for black start services to be provided by distributed renewables such as solar, wind and hydro and other low carbon energy generators.
As this project draws to a close, Utility Week and National Grid ESO brought together a range of energy system stakeholders to discuss their hopes and expectations for the future of black start services in a net zero world and provide some feedback on the Distributed ReStart programme.
The far-ranging and insightful discussion also touched on a range of issues: how well are we equipped to restore power quickly as distributed energy becomes more pervasive and weather becomes more turbulent? Where are the weak links? And are we learning the lessons from events in the past?
Here are some of the key themes.
Peter Chandler, power system manager and lead for NIC Project Distributed Restart at National Grid ESO, began with a project recap.
Instead of the traditional ”top down” approach to black start, which creates islands of thermally generated power, Chandler said the initiative explored whether distributed energy resources (DER) alone could do the same job in a “bottom up” way to generate clusters of clean power on local networks. This would scale up to the high-voltage transmission lines, gradually restoring grid power.
The mainly Ofgem-funded project, with partners SP Energy Networks and TNEI, began in 2019 setting out the scope. In 2020 it explored detailed design options and in 2021/22 moves into live trials to demonstrate whether the designs work or not.
Not only was the project a world first, if successful it would help in the drive to deliver net zero, open up the market to distributed energy services, and accelerate regional restoration timescales, said Chandler.
The programme has been split across four workstreams:
- • Power engineering and trials;
- • Organisational, systems and telecommunications;
- • Procurement and compliance;
- • Knowledge and dissemination.
The aims of the project were also aligned to a new resilience standard being issued by BEIS (the Department for Business, Energy and Industrial Strategy) for the restoration of energy services.
Chandler said that the project’s aim also dovetailed with enhanced resilience requirements from BEIS being set out in a new Electricity System Restoration Standard. The standard is intended to reduce restoration time across Great Britain and ensure a consistent approach across all regions and is expected to be operational by 2026.
It will require the national electricity transmission system to have sufficient capability and arrangements in place to restore 100 per cent of the country’s electricity demand within five days. BEIS stipulates that it should be implemented regionally, with an interim target of 60 per cent of regional demand to be restored within 24 hours.
Participants in the virtual discussion were highly supportive of the project, pointing out the need to be prepared for the worst. “You can’t pretend it wouldn’t happen, and if it were to, we’re not readily supported by neighbours, compared to a country like Italy. Interconnectors wouldn’t work at that level. All the UK’s coal-fired power stations will have closed by 2024, yet by then our electricity dependence will have increased by the uptake of EVs [electric vehicles]. So the motivation for the project is brilliant,” commented one.
“If we do get movement in the jet stream it is feasible that there is a period of calm conditions across the UK and this will create a lack of energy feeding into the system,” said another, pointing to the recent events that have seen wind turbines becalmed and gas prices rocket as a potential taster for emergent problems.
“So, there should be a lot of learning from this project, which may be applicable in other circumstances, not necessarily just a total or large area blackout.”
DER and dispatching power
While there was agreement around the screen for the development of a distributed restart solution, there were concerns around technical capabilities and communication protocols. Some participants questioned whether communication protocols could be put in place to guarantee rapid dispatch given distributed assets are not generally geared up for this purpose and since some use communications infrastructure operated from outside the UK.
When using DER it makes more operational sense for smaller networks to operate as close as possible to where the demand is. Thus enabling power islands with dedicated communications arrangements becomes important, explained a guest. Reducing risk requires better local intelligence on the devices, including an ability to understand the potential state of the network and how they must respond.
Chandler said that the ESO control room had 24/7 communication with transmission-connected generation assets. “But it is still an issue for distributed restorations so we’re looking at ways we can mitigate that.”
“For anchor generation sites, where the generation will provide a voltage reference for other resources to latch on to, we’re looking at those anchor generator sites being manned to 24/7.
“For other resources, the ones that would latch on to our anchor to help expand the island, we’re looking at those to provide a data resilient communication channel to enable that sort of linkage.
“We’re hoping to codify a lot of these requirements through modifications to various codes, like the distribution code,” he explained.
As well as communications, there were a number of engineering challenges to be overcome, including persistent protection schemes and the need to provide a source of local earthing, which require network infrastructure reinforcements to make a power island approach to system restoration work.
Learning from previous mistakes
An issue of frustration for some of those taking part was that previous outages had flagged up that the codes in place are not always adhered to when crisis hits.
“Looking back at the August 9th event we found a lot of generation on the distribution network and the transmission network did not operate as it should have done,” said one.
“Where we do have generation connected to the distribution network forming part of a recovery system, my plea is that it meets the compliance requirements of the standards as they are at the time of connection.
“What we don’t want is to have disrupted generation sources feeding back up to the grid, or arranged as island communities to support key customers, only to find that they don’t actually work as they’re supposed to.”
Putting a value on restoration
Funding for high impact, low probability risk remains a thorny issue and there was debate as to whether Ofgem’s evaluation methodology properly considered today’s interdependency of our infrastructure systems.
“Can you really get there by doing a cost-benefit analysis or have you got to just have a precautionary principle that says this would be so unthinkable that we have to deal with it and make sure we have a way of dealing with it?” was the question posed by one expert.
It was felt that the methodology and figures used no longer reflected how the impact of lost load varied between different customers or how the value of lost load changed over time. “If you’re off for 30 minutes then it may not matter but if you’re off for two days it’s maybe more than 48 times as bad because of everything else that will be affected. If Ofgem wanted to use a cost-benefit analysis approach then they have to have the right inputs to feed into it,” our participant concluded.