Risks to Power Systems Infrastructure

Risks to Power Systems Infrastructure

We spoke with Professor Janusz Bialek, a Professor of Power and Energy Systems at Newcastle University, about his work on power systems infrastructure, including his work on blackouts and renewable energy transitions. Professor Bialek’s work on blackouts received funding from the EPSRC and fell under the global uncertainties umbrella because it was identified as having the potential to contribute to our understanding of threats to infrastructure.

How did you come to work in this research area? 

I was born and educated in Poland, arriving in the UK in 1989. Since then, I have worked at Durham University, Edinburgh University, and now at Newcastle University, where I work in electrical engineering, studying power systems. 

While I have an engineering background, my research is very interdisciplinary – the idea of ‘power systems’ encompasses the whole chain of supply starting from generations, through transmission and distribution, down to the homes and factories of individual customers. To understand those customers, you need the social sciences, which means that a lot of my work relies on collaborations with the social scientists who try to better understand what customers do, and what they want. I also work closely with mathematicians, because at the end of the day power networks are graphs, and graph theory can be used to solve a lot of the problems faced in my field.  

What sort of problems are you grappling with? 

The main challenge now in power systems is decarbonization – the replacement of fossil fuels by renewables can bring a lot of challenges. The obvious problem is that sometimes the wind blows, and sometimes it does not. So, making sure we have enough reserves is important. However, while the problem seems straightforward, there are several related technical challenges. Traditional fossil fuel plants are driven by synchronous generators, big machines with lots of rotating parts which provide stability within the system. There is a lot of inertia in the system, and if there is any disturbance, you can run over it using inertia, which is good for systems stability. When you replace synchronous generators and fossil fuel plants with renewable energy sources, you do not get that inertia – while wind plants are also rotating, they do not provide inertia as they are connected to the rest of the system by means of power electronics – and solar panels have no moving parts, etcetera. That creates technical challenges for how we operate power systems, and my work is exploring how we can overcome those challenges as our power grids take those next revolutionary steps towards the most profound power systems transformation since the end of the 19th century.  

What about your work on blackouts? 

The problem we are here to talk about today is a traditional power system problem, not related to renewables, called cascading outages. Cascading line tripping can lead to blackouts, such as the famous Northeast blackout in 2003 which saw the East Coast of the United States and parts of Canada experience a prolonged outage, with 10 million people in Ontario and 45 million people in the United States affected. There was also an outage in 2003 in Italy which saw more than 50 million people lose power and many other. And the reason for those failures is that transmissions networks are interconnected. When one transmission line is disconnected, the power is redistributed automatically and flows through other transmission lines… and if those lines were on the border of how much they can carry because of thermal limits, they may get overloaded, which means that they could then trip too. This series of tripping, if not contained in time, is what causes widespread blackouts.  

In the United Kingdom, we have never had a wide area blackout because the transmission network is well connected, however in other countries, this is a common problem – with tens of millions of people losing power, or hundreds of millions in the case of India in 2012.  

About a decade ago, I worked on an ESPRC funded project which tried to find new ways of containing these blackouts. We developed this idea of creating an ‘island’, where to avoid the domino effect once a blackout starts to spread, we would try to contain it by creating an island which would be disconnected from the rest of the power system. This way, we might lose supply in that one island, but the rest of the system would stay healthy. There are challenges surrounding how to select those islands, and I worked with mathematicians from Southampton University who suggested that spectral clustering could allow us to identify clusters of islands within a transmission network – densely connected areas which are weakly connected to other parts, and which could be easily delinked.  

We had good results from our project, and a company which is now part of GE, and the National Grid were advising us. However, no country has been willing to implement our system because system operators are afraid that the medicine may be worse than the disease. Our approach involves basically amputating part of the grid, which would be a big shock to the system from a stability point of view. Our approach was seen to be too brutal of an intervention. So, I have subsequently worked with collaborators from Caltech on how to generate the same effect of containing the system disturbance without creating islands. We have come up with a new method called tree partitioning, in which the islands are no longer separated but form clusters connected by tie-lines to form a cluster-level tree. Tree partitioning can be proved to contain cascading failures within a cluster, so that the rest of the system will stay unaffected, but the best way to implement this in practice is still an area of ongoing research.  


Are there things policymakers should be thinking about to minimize the risks to infrastructure as we transition to renewables, and to minimize the risk of blackouts?  

Preventing blackouts is largely a technical problem, and the best thing to do is to build new transmission lines. Practically however, that presents political problems which are difficult to resolve in this country. When there is more power needed to be transferred from Scotland to the South, there is no way people would allow new transmission lines to be built, so we end up with bootstrap subsea connection instead, which are a much more expensive way of transferring power.

Meanwhile with respect to renewable energy transitions and challenges to our grid, there are a few problems. The first is the well understood problem of ‘what happens on days when there is no wind?’. The answer is simply to invest in diverse forms of renewable energy – such as nuclear power, to build connections to other countries so that power can be imported with needed or build energy storage. The second, less well understood, problem however is related to synchronous generators being replaced by renewable sources. There are new technical problems here which we will need to better understand if we are to ensure the stability of our power systems going forward. However, there are few people working on these problems at present, and more research is needed. My colleagues and I have formed a Global Power System Transformation Consortium, which brought together leading System Operators from GB, Ireland, California, Texas, Denmark, Australia, and leading research institutes around the world, however, we all realise that we need more funding for fundamental research now, while we still have time to keep the system together. The pace of change is fantastic, especially with the increased focus on renewable energy in light of the ongoing gas shock, and technical problems specific to renewable energy powered grids are coming. So, the important policy issue is to ensure that the support is there to better understand the problem, and to fund the basic research which will help us prepare for the future. That goes beyond what the existing sources of funding have thus far been able to provide – we need a collective effort, and we need to get better at getting that message through to the non-technical folks such as politicians and civil servants. There is a difficult problem ahead, and it can be dealt with, but it requires focus and a substantial research effort now.


Photo by Matthew Henry on Unsplash