FEATURE INTERVIEW

Interview with BIDESH KAR

DIRECTOR OF PRODUCT MANAGEMENT AT SENTIENT ENERGY

AR

Underground makes sense, especially in fire prone areas, because then you don’t risk fire ignition when you run the lines underground.

There's a cost to undergrounding. What is the cost differential between overhead and undergrounding? I used to hear that it was about four times. Is that still the case or not?

BK

Well, there's a difference. You cannot look at the cost of just undergrounding at once. You need to look at the total lifecycle cost. So, Sentient Energy is a partner member at PDi2 (Power Delivery Intelligence Initiative). It's a nonprofit organization. We have done a lot of research and models where utilities can look at it and say, “What's my total cost of ownership of underground?” It's not just the $200,000 or a million dollars per mile of underground conversion, but imagine once you have undergrounded and there is a storm or a tornado, it's not getting those lines down three times in ten years. So, you add that up, you add up the safety aspect of it, then it makes sense in most of the cases. Otherwise, PG&E wouldn't have gone for 10,000 miles of commitment.

AR

Most utilities are state regulated, right? They have to go to the Public Utilities Commission, and a lot of rate-based things are based on capital spending. Is most of the cost of undergrounding covered by capital?

BK

Yes, because you're improving your reliability of service. So as a customer myself for PG&E…well, the area I’m in is already undergrounded, but imagine not losing power anymore when there is a high fire threat and there are critical patients and those who need life support. I think it makes sense ratepayers wouldn't mind that increase over time as long as of course it has to be kind of vetted out with different parties.

AR

So that is a great connection between what you do and the focus on reliability. At the same time, utilities are dealing with the hockey stick of the demand curve which they have not had to deal with in the last 15 years. There's going to be competing priorities for that. How do you address that? I mean, you're going to compete with people saying let's just get as much power to as many people as we possibly can. Forget about underground. How do you answer those people to say, no, we are a solution that is going to be a better solution long term than short-term reactionary solutions.

BK

I think the change in demand profile has immense impact on how you do your asset management strategy. What it means is your typical duck curve doesn't exist anymore, and here is why: we are pushing for more EV. That means that when you come back home from work, at five, six in the evening, you are plugging in the EV, so your load profile just increased dramatically. Let's say you don't plug in at 6:00 p.m. You have it programmed to run late at night. The power consumption of an EV to fully charge is equivalent to a fully running home, which means your load didn't drop after midnight. It actually stayed at the same level, because now, even though the lights are off, your EV is charging, so it's still pulling the load. How do you balance that? The typical grid, the traditional grid, was planned for single directional power flow. If I pick a transformer, a 50 kVA transformer typically would have seven customers. Now all of a sudden, all the seven homes plug in EVs at the same time. That wasn't what it was planned for.

While a transformer can operate at beyond its nameplate rating, you do not want to do it day over day, week after week, year after year. So that's where we come in as an end-to-end solution. We have one of our newest products UM1 – it monitors for fault in our underground system, but also monitors for your loading of the transformer. One of our customers was recently flagged as having the transformer overloaded 150 times over eight months. And the peak loading was 212%. Average was 126%. That's something you don't want. Now with this data they can upgrade it to a 75 kVA transformer before it fails. This is what we call data-driven asset management strategy – don't just wait for the transformer to fail.

AR

So given the fact that you're in product management, you have to be looking ahead at what's coming. What are the issues that the industry is facing that you're solving?

BK

There are a couple of things that come to mind. Number one, utility companies. I feel for them because they are under a resource constraint. They have a ton of projects to work on and they can only focus on so many things. A solution to that is as they are dealing with all these different software platforms and data points from all different kinds of devices, they need to look at more at what cloud capabilities they can leverage, so that the in-house team doesn't need to focus on managing, doing the cookie-cutter task of system management, but they can focus on how to optimize their business processes and have other vendors kind of take care of cloud-based solutions and deal with the data. And the second thing, very linked to that, is an immense amount of data. Just having gigabytes and petabytes of data means nothing unless it's actionable. So, one of the challenges I think needs to be solved over the years is what are the actionable insights. Don't just show me the data. Tell me in a nutshell, what action do I need to take today, three weeks from now, a month out — so that something that's likely to get impacted further along the line I can delay, and then I can prioritize my workforce.

AR

Does that not increase the risk for cybersecurity?

BK

Of course, it does. But we do have the technology to make it cybersecure. You can have them all communicate within the utilities. Private LTE is the utilities network. They can manage it as long as all these devices talk within their network, that's absolutely fine.

AR

EVs and EV charging – most of that is underground. You don't see overhead lines going to you. So, they've already adopted the charging technology, but they haven't adopted a standardized approach to plugins. I mean, I rented an EV, and I could not charge it at the Tesla charging station because my connector didn't work with the Tesla. So how does Sentient work within the industry to do open architecture, open protocols? Everybody has a part of the solution. Where do you come together to create standardization of those kinds of things?

BK

There are areas where things are very well defined. DNP3 protocol is a very common protocol used for communicating across all these devices. That's just at the device level, however, how different systems interact. For example, our analytics platform can ingest data from a completely unknown device. I think that's the kind of standardized interaction or integration that must be in place. We are thinking ahead in terms of building open application programming interfaces, which is called APIs, so that anybody can consume the data. We are thinking along those lines where we publish a message, anybody within the secured network can actually subscribe to it. I think we'll see more of those changes coming in down the road. And to your point about EVs, yes, absolutely. We want it to be like a gas station. You know exactly how it works. So, on the EV side of things, all these manufacturers have to go towards the standard.

AR

Excellent. Bidesh, thank you for joining us. It's been a delight and I wish you good success at Sentient.

BK

Thank you, Alan. Nice speaking to you.