DNVGL.com

Data analytics in the electricity sector

SHARE:
PRINT:
Podcast data analytics
Welcome to the latest DNV GL Talks Energy podcast series. This series focuses on digitalization and the impact of technology and data. Each week notable industry thought leaders join us to discuss AI, data analytics and all the hot topics impacting the strategies of global businesses, communities and nations as we transition towards a more efficient energy system.

Data Analytics in the Electricity Sector 

With new and innovative technologies developing at record pace, DNV GL’s Michael Wilkinson, Global Segment Leader, Energy Digitalization, gives us some advice on how to see through the digitalization hype and better understand the solutions that can help to transform the electricity sector. In this digitally acute episode, Michael gives us an overview of the technologies that are driving the digitalization revolution and introduces DNV GL’s latest whitepaper - Data Analytics in the Electricity Sector, which examines the current role of data analytics, and its likely influence on the energy systems of the future. He gives examples of how digitalization aids the forecasting of renewable energy flows, and so enables operators to better respond to demand; the impact on energy trading and markets; and how technologies such as blockchain are transforming local energy communities. Finally, Michael touches on the regulatory challenges arising from the introduction of these new technologies.

For more information you can download the Data Analytics in the Electricity Sector whitepaper here.

iTunesiTunes

SoundCloudSoundCloud

RSSSubscribe

Read the transcription here

NARRATOR Welcome to the DNV GL Talks Energy podcast series. Electrification, rise of renewables and new technologies supported by more data and IT systems are transforming the power system. Join us each week as we discuss these changes with guests from around the industry.  

MATHIAS STECK    Welcome to a new series of DNV GL Talks Energy, series five of this podcast. And we start with a guest from DNV GL. So, I have here with me Michael Wilkinson who has a new role as Global Segment Leader, Energy Digitalization. Good morning, Michael. 

MICHAEL WILKINSON    Good morning, Mathias, thank you very much for inviting me. 

MATHIAS STECK    Michael, we have a really interesting topic to talk about. We want to talk about data analytics in the electricity sector. Before we start it would be great if you could talk a bit about yourself as a person and your career history within DNV GL before. 

MICHAEL WILKINSON    Yes, absolutely. So, my background is from renewables, and recently stepping into this role of now taking a wider remit across the energy value chain. So, really my job is to drive digitalization in the energy industry and to help customers right across the energy value chain to respond to some of the rapid changes and disruption resulting from new digital technologies. And really I see for DNV GL the opportunity there is about combining digital technology with our domain experience. And I think that’s the real sort of key aspect of it. 

MATHIAS STECK    Right. Actually interestingly, myself, I started more or less in the renewable industry as well. But what I found was, what was very interesting starting as a very, kind of niche business, maybe the tree huggers which nobody really believed in, this is now something which is driving innovation. So, utilities have to wake up to that theme. So, what would you see there? How does the technology help us also to advance on the digitalization theme? 

MICHAEL WILKINSON    Yes, I think that’s absolutely right. For me, I think in the energy industry, there’s three main trends that people talk about. There’s decarbonization. So, this proliferation of renewables. Decentralization, which is also about renewables being embedded more in the grid and about people buying and selling energy at the distribution level. And then digitalization as well. And by digitalization, it’s a bit of a sort of buzz word in many ways, or in there’s a lot of hype around it. But for me I think the best definition I’ve heard recently, someone described digitalization as the sets of business opportunities which are created by computers, large and small. The connectivity between them through the power of the internet. Software. So, analytics and artificial intelligence. But then also sensors and all the data that’s generated from them. So, within the renewables space that’s in many ways sort of born digital. You know, these wind turbines and solar PV panels and so on have always been generating data. You know, the very first wind turbine started with SCADA systems generating lots and lots of data. And about 20 years ago, all this data was, you know, was too much for people. They couldn’t handle it. But now with this progression of digitalization the opportunities to actually gather data connected together and then do something with it is increasingly becoming realized. I mean, it is real. And then the power, electricity, sort of distribution transmission system has seen this all happening within renewables, and they’re now looking at this and saying, well actually, we wouldn’t mind a bit of that. Actually, we need some of that new technology and we need some of that digital technology to come and help us to transform our own part of the electricity sector.

MATHIAS STECK    So, when we talk about data analytics in the electricity sector there’s a lot of vision. There’s a lot of fear as well. So, years back we had the utilities talking about the death spiral of utilities. We talked about rebalancing of the different actors in the grid, so people who did renewables had a lower incentive, people had to integrate it, had a lot of problems. At least that was the perception. The things you just talked about, maybe we can use this podcast to establish or break that down, how do we make these things happen? 

MICHAEL WILKINSON    Yes, exactly. So, you’re right, and there’s a lot of hype around digitalization. But it’s not an end in itself. Digitalization is not an end in itself. It’s a means to an end. And it’s helping us to achieve our ambitions as an industry. So, we’ve written this white paper recently called data analytics in the electricity sector, and there we talk about the applications of data analytics, and make it a bit more real and say... So, I think there’s three main categories that you can group the application of data analytics, and it’s forecasting and operations, and then energy trading and markets. And all of those different elements of the energy value chain have got, or, different processes within the electricity sector have got really nice examples of applications where digitalization is helping this transformation. 

MATHIAS STECK    Yes, that would maybe be interesting if you just, you know, stick to those. So, what about forecasting? How is digitalization, connectivity, how is it helping?

MICHAEL WILKINSON    Yes, so, with forecasting, really it’s all about the ability to predict power flows minutes ahead. We talked about renewables before, and renewables of course are intermittent. They generate power when there is natural resource available to generate that power. One of the really interesting things about wind, for example, and solar, but wind... They are intermittent, wind turbines are intermittent in what they generate. But also you can forecast that really, really accurately. So, you can forecast the electricity that’s going to be produced a day ahead or, an hour ahead even, really accurately. So, that’s where data analytics really comes into its own. And the ability to take data from wind turbines, wind farms, feed that into your analytics models in real time and use that to refine your power forecast for your renewable generation. So, that’s one example where forecasting is really helping to integrate renewables into the power generation sector. 

MATHIAS STECK    I was wondering, when we talk about this forecasting, we can do that on the generation side. But should it not also help us to adopt the loads accordingly? I think that’s not happening at the moment. At the moment we do conventional generation to ride through the bumps, but in future could we use that technology to kind of balance the grid?

MICHAEL WILKINSON    Exactly. And I think that’s where there’s huge scope for forecasting. To be able to understand, so, first of all to measure the actual power at many more different levels within the grid, right down to the distribution level. But then also use that and feed it into your model, into your digital twin of your power grid, and use that to properly understand what’s happening right now, but also what’s going to happen in the future. There’s a nice example from, IBM Watson, applied their IoT platform onto Fingrid, I think it was. And there they fitted a load of IoT devices along the electricity lines and into the distribution grid. And then used that to gather the data, feed it into their models, and give real time understanding of what was happening, but also combine that with forecasting to understand what is going to happen in the next hour ahead, the next 30 minutes ahead, the next five minutes ahead even. 

MATHIAS STECK    So, as you mentioned that this paper you have just published is also looking at operations, and a little bit, I think you talked about operations just now. But how can utilities grid operators utilize new technology, or what are your covering in this operations chapter?

MICHAEL WILKINSON    Yes, so with operations I think there are some really nice examples of how data is being used by grid operators, for example, to detect fraud using smart meters. You know, they can now gather much more data in real time, and then predict or understand what should be used versus what is being used, and use that to identify overuse and fraud, for example. There’s a nice example from using data within the renewable generation world as well. So, asset optimization. We’ve got this really nice tool that we’ve been developing recently called WindGEMINI, which is a digital twin of an operating wind turbine. And we’ve been working in renewables for many years and we’ve got a lot of domain experience there. But the really interesting part of that has been to take it from an offline, do an analysis, and then send a report, and then three months later you might implement some changes; to take it from that into: let’s take real data from wind turbines in real time, feed it into your model, take your load calculations, data of those load calculations in real time, and then use that to give your wind farm owner / operator, but then also maybe your network operator as well, your grid operator, real-time information about what’s happening and what is going to happen on that wind farm as well. So, there I think there’s huge applications for analytics to help improve the efficiency of our renewable generation. 

MATHIAS STECK    Yes, so you mentioned WindGEMINI digital wind turbine, digital wind farm, digital twin. I always wonder, up to now we have, well, we design wind turbines towards a certain wind class and a certain standard, of course. And then every turbine in a farm around the world, if it’s the same time, the type has kind of a certain controller making sure that the wind turbine’s operating the way it was certified. With digitalization, with this digital twin technology, could we be more smarter and have individual control functions for each individual turbine basically to optimize for generation over lifetime, or for the lifetime itself, or whatever it is?

MICHAEL WILKINSON    Yes, absolutely. And we’re seeing people actually doing that as well. We’re seeing pilot examples of people individually controlling turbines as part of a whole wind farm control system, which they can then use that to respond directly to requirements from the grid to maybe turn down generation. But to turn down generation when you need to curtail the grid, but then do that in a really smart way so that your turbines... You don’t just turn off one row of turbines, you turn some of them off and you do that in such a way that you actually balance the loads across the wind farm. And therefore your considerations are not just “can I reduce the power output being generated by my wind farm?”, but also, “what’s the impact of the remaining life of these turbines?” as well. And that’s a really neat bit to say, actually there’s more than just the reduce the power, but reduce the power and do it in a smart way so that my wind farm can potentially last longer. So, the money lost in the lost generation can then be recovered by allowing you to operate your wind farm for extra years later on in life.

MATHIAS STECK    Yes, exactly. I want to come back to this pilot, as such, which you mentioned, but before we go there maybe let’s look into the energy trading and markets which you mentioned, as another chapter of the paper. We had in another podcast series the Brooklyn Microgrid as a guest. So, whether you can talk a bit about what technology can help us there. 

MICHAEL WILKINSON    Yes, absolutely. So, I think energy trading in markets for me is one of the really interesting, new ways that digitalization is really coming to bear. So, in various countries we’ve got this roll-out, the smart meter roll-out programme. When those programmes were launched in different countries there was an economic assessment that was done that made an assessment of what is the financial benefit to a particular country of the roll-out of smart meters. You know, this is not a cheap thing. You can’t just go and change every meter in everybody’s house at zero cost. There’s a real, very substantial cost associated with that. So, it was seeing this roll-out of smart meters, and part of the economic assessment really has been around what can you do with the data that’s being generated from those smart meters. So, you mentioned Brooklyn Microgrid, there’s other examples where data from consumers is really being used in a really more active way. So, for example, in Cornwall, in South West UK, there’s an example of a local energy market which is being created and managed by Centrica. And what they’re doing there is that Cornwall is a particular part of the UK where it’s a lot of renewable generation, it’s pretty windy. But also it’s at the end of some long transmission lines. And that makes for a pretty tricky situation where you’ve got a lot of curtailment, constraint, on the grid. So, this local energy market is intending to allow consumers and generators to buy and sell electricity from the transmission lines, but also between themselves as well. And also to buy flexibility within that whole system. So, that’s being enabled by all this data being generated by the generation, but also all this data being generated by smart meters as well. So, connect them together. And then also use technologies like blockchain, for example, to allow that trading of buying and selling to be done in a really secure way that people have got trust in it. 

MATHIAS STECK    Talking about these things, it kind of becomes clear that making this reality, you’ve mentioned pilots before, we have to get from pilot stage to, industrial use and roll-out. But to do this we also have to cut, for example, through industry silos. There is another regulation. [We’re] coming to an end now of this episode already, but maybe you can kind of wrap up how do we move from all these different interesting developments, which work on the small scale, to really make that happen on a global level.

MICHAEL WILKINSON    Exactly. So, someone said recently to me that we should be working in the airline industry, there's so many pilots, pilots everywhere, right? So, the challenge is to take these little test examples and make it real and roll it out. There’s a good example of that from our Smart Cable Guard product, which is being used to predict and detect failures within distribution lines. There we’ve developed that in combination with the university over many years, refined the technology, but now it’s being rolled out and, you know, it’s an industrialized solution which is being rolled out in the thousands of units. And we’re seeing examples of that sort of around other parts of the industry. They’ve found a solution that works and now they’re rolling it out in the wide-scale. You mentioned regulation there as well and I think that’s all this nice talk about these digital technologies, is great and very exciting, but when it comes to actually making it real I think there is a significant challenge around regulation. One of the features of digitalization is that it lowers the barrier to entry. So, we should see a lot of new players, but the regulations are pretty heavyweight and geared towards established players. So, there’s some real challenges there. And for example, back to the UK again, Ofgem, the UK regulator, their stated aim is to protect the interests of consumers. And I think we’re at a critical point now where regulators around the world need to understand, is that protect the interests in the short term to ensure that they’ve got lower energy prices this winter, or is it to ensure that we’re able to take the benefits of digitalization? And to do that we need to make sure that the regulators are working with start-up companies and allowing them to actually implement some of these new technologies in the grid.

MATHIAS STECK    Thank you, Michael, for these very interesting insights around data analytics in the electricity sector. I understand this report is available for download from today onwards. 

MICHAEL WILKINSON    It is, yes. 

MATHIAS STECK    Yes. So, once again, thank you for having the time and talking about these interesting topics. To the listeners, thank you very much for listening in. This was Michael Wilkinson, the Global Segment Leader, Energy Digitalization, on data analytics in the electricity sector. 

NARRATOR    Thank you for listening to this DNV GL Talks Energy podcast. To hear more podcasts in the series, please visit dnvgl.com/talksenergy.