← Open Circuit![The great divergence: Can the grid catch up to transportation innovation? [partner content] — Open Circuit cover](/_next/image?url=https%3A%2F%2Fmegaphone.imgix.net%2Fpodcasts%2F7ec68a0c-e24b-11ef-b277-7b3d3c3c73b9%2Fimage%2F046bd19a73ac412f06e9fd707b709b72.png%3Fixlib%3Drails-4.3.1%26max-w%3D3000%26max-h%3D3000%26fit%3Dcrop%26auto%3Dformat%2Ccompress&w=1920&q=75)
Open Circuit
The great divergence: Can the grid catch up to transportation innovation? [partner content]
00:27:50
Kraken’s Devrim Celal on how utilities can unlock the same innovations that shaped transportation systems.
Loading summary
Transcript36 lines
- [00:03]
Narrator
This is partner content from Latitude Studios. Devram Jalal is Chief Flexibility and Marketing Officer at Kraken, the only proven AI powered operating system for utilities. His mission revolves around one central theme.
- [00:18]
Devram Jalal
At Kraken, we're in the business of using technology to accelerate the energy transition. My role is to ensure that that the system we built gives utilities and consumers the maximum possible output from existing infrastructure and devices that they own.
- [00:36]
Narrator
That means making the most of infrastructure such as poles and wires, as well as consumer technologies including EVs, solar panels and batteries. But it also means thinking about the intellectual framework for the energy transition. And recently he's been learning lessons from the past.
- [00:51]
Devram Jalal
I'm trying to learn how things developed in other industries and to see how we can apply those to electricity and transportation was an interesting one.
- [01:04]
Narrator
One question that Devram is exploring why have transportation and electricity taken such radically different innovation paths? The answer begins in the mid to late 1800s. That's when in 1882, Thomas Edison opened the first commercial power plants in London and New York, right around the same time that the first major train stations and commercial automobiles were being built.
- [01:23]
Devram Jalal
When I take a step back and look at a little bit holistically, structurally they look very similar in those days.
- [01:29]
Narrator
But both were structured the same way topographically. A series of central hubs with spidery networks coming off them, transmission and distribution networks for energy and main and local train lines and roads for transportation.
- [01:41]
Devram Jalal
But then I started looking at how things change, and that's where you start seeing some major differences between the two.
- [01:47]
Narrator
Over the course of the 20th century, even before the rise of the Internet, transportation became increasingly decentralized and intelligent, with a lot more actionable data and dynamic pricing.
- [01:58]
Devram Jalal
You got telemetry through camera systems to say there's a congestion here and panels to display back to drivers, saying to your next exit, you will have to wait x amount of time or you will encounter congestion. At certain point. It allowed it to give preferential routes for people that had more passengers in their cars.
- [02:19]
Narrator
As the world became more digitally connected, new services were built to manage congestion and the driver experience.
- [02:25]
Devram Jalal
You had cameras on main arteries coming into London that would read the number plate and say, you've just entered the congestion area. I've read your number plate, I know where you live and I'll send you a bill to your home or to your mobile phone. This all happened in transportation because of the need to get better utilization over an existing infrastructure.
- [02:47]
Narrator
Over time, the electric grid got much more connected as well, but it mostly served the dispatchability of Large centralized power plants, not the consumer experience on the distribution system. That divergence only increased during the modern Internet era.
- [03:00]
Devram Jalal
And Google Maps came. And one of the interesting things Google Maps is it's user generated content. So user generated content allowed us a tool to tell me in the morning, should I take the Metro or the underground, should I walk or should I take a little E scooter to get to work? Then we went even further. Uber came and completely transformed the taxi domain in multiple geographies despite different regulations, each one of those and brought up the utilization of the passenger car from about 5% to 60, 70 or even higher percentage. Now we're moving transportation next level with autonomous vehicles. When I look at electricity, we've just left the Victorian era and we've got a long way to go.
- [03:53]
Narrator
This week, are we finally entering a golden age of innovation for the grid? Kraken's Devram Jalal talks with Stephen Lacy about how distributed tech, consumer engagement and integrated software are creating a new flexibility paradigm in electricity. Can utilities capture the same magic that has transformed transportation?
- [04:15]
Stephen Lacy
So what do you see happening right now along the electric grid that leads you to believe that we are going to catch up with the acceleration that we saw in transportation?
- [04:27]
Devram Jalal
In some of the work that we've been doing, we're seeing some incredible results that consumers are willing to participate. So we've got two dimensions here that we need to change. One is market structures. So we need to work closely with regulators and we've been doing so to open up the possibility to engage consumers with the energy system so they would allow us to decide when is the best time to charge their cars or how to use their batteries at home to maximize the benefit to the grid. I'll give you three use cases where why I'm feeling quite excited about this. And first one of those was a group, Octopus Energy Group. In this case, we wanted to prove that NIMBYism was something that we can actually help people get over. And to prove that, we actually went out and bought two wind turbines in rural areas that already existed. And we said to people in the vicinity of those wind turbines, these are your fans. If you subscribe to them and they're rotating, you'll get a discount on your electricity. If they're rotating really fast, you'll get a bigger discount on your electricity. They were massively oversubscribed immediately. So we opened up the proposition to the whole of the island and now we've repeated it in Germany and other places and said to people, would you like to have a wind turbine in your backyard if you were to drive a tangible benefit from it. And the answer was overwhelmingly positive. In the UK we had over 35,000 applications in a very short amount of time. The other signal to that is they actually changed their behavior significantly and they engaged to do that. We started publishing 48 hour wind forecasts to these people so that they can schedule their lives around when was their electricity going to be cheaper. Those apps publishing had over 80% constant engagement rate and significant shift in their loads, which sells testimony consumers are willing to participate. We just need the market structure to incentivize them properly. The second one was something called saving sessions. And again we work closely with our regulator in the uk, used to be National Grid, now is the national energy system operator. And we engaged them and said look, with the energy crisis as a result of Ukraine, we're going to have an expensive winter. And during that winter if we rely on gas and coal especially, prices will go up, so affordability will go down. Instead of calling on these power plants to operate, why don't you come to us and we'll ask our consumers to use less electricity, reducing demand per generation that produced over a million and a half customers. Almost 40% of those who are eligible to actually participate in programs that were based on a signal to them they had saying hey, tomorrow at 4 o' clock we're getting into a congested period. Normally you would have used I'm making numbers up here, but 2 kilowatt hours of electricity during that period, if you reduce that, the money that would have gone to a coal fired power plant will come to you. And we had million and a half customers engaging and earning about 15 million pounds. So or let's say 22.5 million dollars. The money per consumer wasn't that big. It was $15 to $20 per consumer on average. But the willingness to participate, to have an impact, we were effectively shifting load from what would have been the most carbon intensive periods to, to others purely by customers saying yes, I will help.
- [08:05]
Stephen Lacy
Those are such great examples and I think more evidence that behavioral demand response really works, that you can get a lot of capacity out of consumer participation. So what's next? What comes after we've proven all of that?
- [08:18]
Devram Jalal
So with behavioral response, you ask consumers to do something and we've seen they're actually doing it. But even better is if they don't have to do anything, and in this case what we've been doing for a little bit over three years, actually coming up to four years now, since January 2022 is saying to consumers, you may have an electric vehicle or a battery or heat pump. Here's a tariff, a product for you to sign up to with using a mobile app. Then register your device and I'll explain the electric vehicle example in this case. So register your electric vehicle and or your charger with us. Give us control to decide on your behalf when to charge your car. Give us permission, but tell us what would you like us to do for you? It's a strange question for utility to often ask for the customer and in this case the consumer's response to us is I need you to charge my car at least up to 80% or to 80% by 7am tomorrow morning. Now those two metrics are totally up to the consumer to decide what it should be. What we have just done is what we call unlock inherent flexibility. Because a utility who has access to that kind of mechanism has just unlocked a 9 to 14 hour period. The time when the consumer gets home till when they need their car ready by to do typically two to three hours of charging two to three times a week. The utility can then take that information and say look, you've given me your metrics based on this I'm going to find the cheapest periods of electricity which also happen to be when renewables are in abundance. I'm going to look at the state of the network and choose the time when it's not congested so we're not having to upgrade it all the time, which is expensive, but give the consumer what they want. That tariff went live January 2022 with 811 Tesla drivers on it. They were the Tesla drivers that came from the alpha and beta phases without any marketing. Today the tariff has exceeded 300,000 customers on it. The net promoter score is 20% higher than the next best in class. And next best in class isn't a bad thing. It's another really popular tariff. The churn rate on the tariff OR program is 0.5%. That includes people who leave the country or sell their electric vehicle. And that to me is that's absolutely nirvana. It's what I call the self automated, self healing system. That once you unlock the flexibility, the trader or the system operator, depending on if we're working on a vertically integrated region or a competitive one, could set up an initial schedule. But things might change. I may suddenly have a wire trip somewhere and I have a congestion or I won't be able to charge those cars during that period or electricity prices might spike. When I thought I could charge cheaply, it Gives me the ability to, in real time, respond and push the demand on the grid to different times and different locations. And to me, that's the future of where electricity should be heading.
- [11:31]
Stephen Lacy
Yeah. And so now you're not just talking about positive grid outcomes, but a broader positive business outcomes as well. And do the utilities value those business outcomes? Are you seeing them understand? Certainly we've seen this language change from thinking about ratepayers as electricity off takers and more as actual consumers with needs and wants and preferences. Is that cultural shift really translating inside the business of utilities that you're working with?
- [12:02]
Devram Jalal
Yes, it is. And what we're seeing is a number of different drivers. There are some utilities with a burning platform who need to do something. And this is an incredible tool to have to start solving your generation issues, your network issues. The second class is they want to be leaders, to forge ahead, to innovate and use that as, especially in competitive markets, as their competitive advantage to win with consumers and regulators who are basically starting to see this as a viable, if not better, tool to manage the grids with.
- [12:38]
Stephen Lacy
This brings us to another question about customer control as well. I think there's a debate within the industry about how much customers should give up control or whether they should be able to override the signal. Where do you imagine the customer in this transaction, in this relationship? How much control should they have? Are they willing to give up that control, or do they need some ability to adjust based on their needs?
- [13:03]
Devram Jalal
The customer should always have control.
- [13:06]
Stephen Lacy
And what does that look like exactly?
- [13:09]
Devram Jalal
We should give the customer choice and say things like, if you defer when you use electricity, you will get a benefit out of it if you allow the utility to decide when to charge your car. And that was the part that I missed in my earlier example. You. You'll get a benefit in this case. In the case that I was describing, the utility actually gives its customers a 60 to 70% discount on the cost of charging the car over their regular rate. So their regular rate could say it's 30 cents per kilowatt hour. Their EV charging in that case would be 7 and a half or 5 cents per kilowatt hour. So we need to define clear benefits to the consumer and then let the consumer decide if they're attractive enough and if they would participate. On top of that, we should definitely give the consumer to override any control that they've given to the utility without penalty. So the consumer could have said, charge my car by 7am but something changes in their life. They need to go somewhere Urgently and they need the car charge quicker. They should always have a button that overrides and says charged now, no penalty except you will get charged at your regular tariff as opposed to the discount that you would have had otherwise.
- [14:25]
Stephen Lacy
One thing that I've also seen evolve is that utilities are really taking seriously the capacity that these programs can provide. Just five years ago, if you went to the operations team, they would not take it seriously in the way they are now. So how are you actually seeing them evolve in their thinking about what this can deliver, the grid as a meaningful resource?
- [14:50]
Devram Jalal
I think that's really encouraging signs of that. Like California and other other states in the US taking this really seriously. Look, the way I talk about this to regulators and power traders is when you do consumer flexibility, and I'm focusing on the consumer side, but similarly applies to commercial industrial as well. But consumer flexibility, your biggest friend is statistics. Because when you have hundreds of thousands of people with cars or batteries, statistically you can predict very accurately their behavior. And today for the traders and also the system operator of the 300,000 plus devices, we run a forecast and we say at noon, typically each day for 11pm to 11pm so that 24 hour window, when will people get home? What will be the state of charge of their cars? What will they ask us to do for them and by when? And we get that accurate over 90% of the time, which is I would say probably more reliable than a large power plant. And also when a power plant goes down, you get a significant point of failure. So that's the first element. It's highly reliable once you get your statistics right. The second thing is it's the cheapest form for balancing electricity because consumer devices weren't purchased to provide a particular service because they're not looking for an economic return on it. Because if I build a battery, I got a formula that says I got to get X return on my investment. Consumer devices have a primary purpose of meeting what the consumer needs them to do. If we can combine what the grid's needs are and the consumer needs together, we've just unlocked an incredibly valuable, cheap, accessible form of flexibility that helps us balance the grid.
- [16:44]
Stephen Lacy
So it's very clear that a lot of utilities are taking distributed resources, customer empowerment, demand response, grid flexibility much more seriously. But many of them run into another challenge which is that they have a lot of different software systems running along their operations. And this is really problematic for utilities that are trying to make fast changes. Talk about this actual software integration challenge and how it can hinder utility innovation.
- [17:16]
Devram Jalal
The hodgepodge of technology solutions at utilities is not new. And that's been one of the problems that the utility sector is facing because it slows the move and innovation. I'll give you a favorite quote, recent quote from one of the utility executives that we're working with. He said quite openly in an open room, my IT system prevent me from dreaming. Every time I have a new idea, it's a million bucks. It's a number of months, and by time it's realized, it's too late. Market has moved on. That's a major challenge. And that's one of the challenges that at Kraken we help solve. When customers implement a Kraken solution, we typically retire over a dozen systems and converge everything in a single one, which gives our client, the utility, a single view of all their data, from the consumer data to their metering data to the asset data. That allows you to simplify and augment your customer experience. We develop a new operating model which we call the universal agent, where there are no longer departments. A traditional utility would have a billing issue resolution, debt collection department after department. With a Kraken enable model, there's a single department and every call is meant to be resolved at the first call by the same person without the need to transfer or send the back office to come back to the person. That creates incredible efficiencies because you're simplifying things. Creates incredible customer satisfaction because problems get solved. You have informed, enabled agents responding to your queries. But the one thing that always doesn't get elevated, it creates incredible employee satisfaction because employees who have good tools to solve problems to make customers happy tend to be happy themselves and gives them development opportunities. The customer relationship, customer management part of the platform is going in and replacing a number of legacy systems. It's existing solutions. When we look at optimization, we're actually going into typically green fields because utilities are just starting to flex consumer assets. But in the US as some regions have been quite advanced. When we go into a utility, we realize that they have multiple demand response programs, 3, 5, sometimes higher. And when you have all those programs without a central brain to help orchestrate them, you could have scenarios where different providers are actually actioning things that are perfectly in line with their service requirements, but they may have contradictory actions to what other service providers are doing. So it's a world where everyone is doing what they're meant to do, but not achieving the overall result for the utility because they may have contradictory actions. And what I mean by that is the utility may be trying to reduce load at a substation. But someone is charging a card because the energy signal says so, but somebody else is discharging at the same a battery at the same time. One is plus, the other one is minus. It cancels it out. So where we see our role developing as Kraken is becoming the asset registry for all the service providers for the utility. Because utility has a single view of what's happening in their area. The ability to send dispatch signals to each provider to say I need you to do this now or I need you to do this during this day. It could be a 24 hour signal or a shorter term depending on the environment, so that all the providers are harmonized, delivering the ultimate result for the utility. It gives them the ability to measure and verify that things have actually transpired. And the most important, because all the device data is coming to the utility, it creates that rich data source for building grid aware optimization capability.
- [21:17]
Stephen Lacy
This also brings us to another challenge, which is the implementation of artificial intelligence. And in order to successfully implement machine learning and AI, you need a unified view of data. And so with these disparate systems, they're not able to access the data in these silos in order to create an AI that actually functions well. So I guess the question is, how are you actually using machine learning or AI in the platform? And, and how does it give you the opportunity to layer on additional AI capabilities inside a utility?
- [21:53]
Devram Jalal
So I was at a event in Germany yesterday and we're talking to some incredible new startups. We're talking about deploying AI for utilities. And they were talking mostly from a forecasting perspective, forecasting generation, forecasting demand, forecasting grid congestion. And one of the brilliant mindset, the 80% of what we do is data cleansing. Bringing data together, piping it and doing it over and over again. And then once they do that, they have 20% of their energy goes actually doing the really imaginative stuff. With Kraken, the data is there and the way we've deployed it is in a number of different parts of the platform. So on the customer side we have developed a tool which we call the AI augmented communication capability. The tool, using large language models, looks at a customer's full history with Kraken. And this could be any utility using Kraken. It also looks at anything that may be happening at time, looks at our blogs regulation. At the first step, if a customer query comes in and everything is transcribed, they could have written an email message or it could be a transcribed voice and it summarizes it for the energy specialists to quickly see what has happened and what the customer is really asking for. And at that point it gives the energy specialist the ability to ask AI to respond to that. And it devises a email which then the energy specialist could go and edit. And it learns from the style of the person, so it's writing actually in the tone of voice of that person and when the person is happy, sends the email back. Typically our customers satisfaction rating is 20 to 25% above market. We've taken customers from in the UK, Trustpilot, JD Power is in the US I know, but with Trustpilot we've taken customers from 1.2 rating to 4.4 during the first year of migration. So our customers are usually very happy. AI enabled communications is achieving 20% higher customer satisfaction ratings because simply because it's quicker, it's comprehensive, it's accurate and it's also action driven. It could tell the customer to solve your problem. Here's a link that takes you to the right place to alleviate the issue you're facing. The second part, it obviously improves the energy specialist efficiency. They'll be able to get through a lot more messages using AI. Interestingly, this comes at a time that with devices, consumer owned devices, there are more complex problems to solve for energy specialists. So we're freeing up time from basic tasks so they can focus on more value add tasks to enable utilities to maintain that high level of satisfaction. The second area where we use deeper machine learning and AI in some use cases is predicting behavior. So of the 300,000 so consumers with devices, we have an algorithm that helps us forecast their behavior. It tells us when will they get home, what will be the state of charge of their cars, what will they ask us to do for them and by when. So we're able to use machine learning to forecast those parameters with over 90% accuracy. That gives the system operator an incredible tool because it could say to us, oh, that looks interesting. Here are my problems that I'm trying to solve. It could be a congestion problem in a certain area. It could be balancing energy, maximizing renewable generation. Then we take that constraints they've given to us. We look at the profile of what we can flex and come up with a charging behavior that can meet all the challenges, but also give the users or consumers what they had to ask for in the first place. And that's an algorithm that's been running for three years now and constantly improving its capability.
- [25:58]
Stephen Lacy
So we started this conversation talking about why transportation and electricity systems diverge so much over the course of history. And I wonder now that we have this sophisticated real time data in the transportation system, how close do you think we're getting to that in the electricity system today? Are we finally going to see those two systems convergence?
- [26:19]
Devram Jalal
Well, electric vehicles are actually forcing us to converge them. But yes, I'm very hopeful, especially as we start building capabilities like grid aware optimization by pulling the data from electric vehicles, heat pumps, batteries, meters in real time. By monitoring generation, which the large scale generation is relatively well monitored, the transmission is relatively well monitored. But building that last stage of monitoring in the distribution grid and AI now the sudden prevalence of large language models, making it available to utilities and tech providers like us, will give us the tools to build a system that's at least as good as transportation works today, but hopefully significantly better.
- [27:06]
Stephen Lacy
Devram, this is a fascinating look at history, the future and everything in between. Really appreciate it. Thank you.
- [27:12]
Devram Jalal
No, thank you. It's been fun.
- [27:18]
Narrator
Devram Jalal is the Chief Flexibility and Marketing Officer for Kraken. Kraken removes the outdated, siloed tech that's holding back most utilities. Their unified operating system streamlines and enhances operations, meaning happier teams and happier customers for a fraction of the cost. Join leading utilities across the globe and redefine the sector with Kraken. Click the link in the show notes or go to Kraken Tech to learn more.