Catalyst with Shayle Kann - What the grid can learn from the internet

Episode Date: March 11, 2022

For nearly two decades, the terms "smart grid" and "grid edge" have been used to define the digital layer of the electricity system that can help integrate more rooftop solar panels, EVs, smart meters, and home batteries to avoid outages and save customers money. But even with massive increases in computing power, utilities are still lagging in technology to communicate with DERs (also known as grid-edge assets) and the computing power to crunch all that data. In this episode, what can the grid learn from the internet? Guest host Lara Pierpoints talks to a person deep in both worlds. Astrid Atkinson is a former senior Google engineer who specialized in distributed networks. She’s now founder and CEO at the grid software company Camus. Lara and Astrid examine where the grid still needs a digital upgrade. They also discuss concerns about giving utilities access to the technology to communicate with DERs and control over consumers’ devices. Plus, Astrid and Lara also cover FERC Order 2222, the incentives that allow DERs to play in electricity markets, and the under-appreciated role of electricity co-ops in testing out new grid-edge technologies.  Catalyst is supported by Antenna Group. For 25 years, Antenna has partnered with leading clean-economy innovators to build their brands and accelerate business growth. If you’re a startup, investor, enterprise or innovation ecosystem that’s creating positive change, Antenna is ready to power your impact. Visit antenna​group​.com to learn more. Catalyst is supported by Nextracker. Nextracker’s technology platform has delivered more than 50 gigawatts of zero-emission solar power plants across the globe. Nextracker is developing a data-driven framework to become the most sustainable solar tracker company in the world — with a focus on a truly transparent supply chain. Visit nex​track​er​.com/​s​u​s​t​a​i​n​a​b​ility to learn more.

from the studios of PostScript Media and Canary Media. I'm Lara Pierpoint, and this is Catalyst. It's never going to make sense to contact a human, really, on a very large scale and tell them, like, the price of electricity today is 13.5 cents a kilowatt hour. Yesterday, it was 13.25 cents a kilowatt hour. So, you know, you should really go ahead and, like, adjust all of your devices to take advantage of that. Like, that is absolutely a job for computers. Solar panels, EVs, smart water heaters, and batteries galore. The number of distributed energy resources, or DERs and energy walk-speak, is multiplying exponentially, and each one is generating loads of data.

In fact, the grid is looking more and more like a totally different distributed system, the Internet. So what can the grid, the most complex piece of machinery on the planet, learn from the Internet, the most complex computer system on the planet? When utilities need flexible capacity they can count on, they turn to Energy Hub. Energy Hub works with more than 170 utilities, coordinating over 2.5 million devices to manage 3.4 gigawatts of flexibility built for the moments when utilities can't afford uncertainty. Energy Hub builds and operates virtual power plants that utilities actually stake their grid planning on, coordinating EVs, batteries, thermostats, and more through a single platform built for utility scale. predictive, verifiable, and designed to perform when it counts.

Learn more at Energy Hub.com. Trillions of dollars are flowing into clean and critical infrastructure, but those investments aren't driven by technology alone. They're shaped by markets, by policy, by capital, and by the institutions that connect them. I'm Alfred Johnson, CEO of Crux, and host of a brand new podcast, Critical Capital. Each episode, I talk with people deploying capital,

shaping policy and building the clean economy. Tune in as we unpack how progress is actually made. Listen to Critical Capital on Spotify, Apple, or wherever you get your podcasts. I'm Laura Pierpoint, filling in for Shale Khan while he's on family leave. Congrats to Shale. I'm the director of climate at Actuate. It's a nonprofit focused on systems innovation to scale greenhouse gas emissions reductions. For decades, the grid has involved large centralized electric power generators

and wire systems that transported electricity one way to customers at the other end. But low-cost distributed energy resources like solar panels, electric vehicles, and opportunities to smartly connect with devices that use energy are changing the game. 20 years ago, California had tens of thousands of grid edge assets. Now, it has millions of them. This includes 1.2 million solar panels providing an estimated 15% of California's electricity and a million plug-in vehicles. By 2030, the number of solar panels could nearly double, and the number of EVs on the road could increase five times to about $5 million.

Additional resource growth will come from grid-edge batteries and connected electrical appliances. Now, not every area in the U.S. is going to see growth like that, but demand for grid-edge resources like Ford's Electric F-150 and Ream's smart water heaters are booming in states not named California. Enough communities are seeing an expansion of these resources that we have a major trend. So what does it mean that the edge of the grid, the consumer side of the meter, is starting to feed electricity back onto the wires? First off, there are some legitimate concerns. The delicate balance of supply and demand gets trickier with grid edge resources. And so far, a lot of these edge resources have been added without giving full visibility to grid operators or utilities. That's fine if there are just a few solar panels out there, but it becomes a major challenge when those resources start to create instabilities that are tricky for grid operators to manage. But it's also true that these developments present huge opportunities.

Resources on the grid edge can be clean, they can be deployed in ways that help decrease system-wide costs, and during a grid failure, they can continue to supply customers with locally generated electricity. If we can get the right kinds of data shared across the right entities, allowing for visibility, control, and the setting of proper incentives, we can do a lot of cool things on the grid. We can better use existing infrastructure, we can shape electricity demand so that that it matches renewable energy generation, and we can make communities resilient to blackout conditions.

But this is a gigantic if. Silicon Valley is great at crunching huge amounts of data to make good decisions. The electric industry is great at delivering power. There are very, very few people out there who are truly deep in both worlds. I chatted with one of them today. Astrid Atkinson, founder and CEO of Camus Energy, knows as much about data and communications infrastructure as she does about electricity markets. I talked to Astrid about what it will take to have a smart, well-orchestrated grid that includes a high level of distributed resources.

Take a listen. Hey, Astrid, how are you doing? I'm awesome. It's really good to see you. It's really good to see you, too. Great to be having this conversation with you. So I want to launch in. We're going to talk about the grid edge, but first I want to tell you a story that I think will highly amuse you.

When I got to Exelon in 2016, I found out that the utility industry was newly starting to use the word customer. And I looked at someone when they said that to me and said, what on earth are you talking about that you're newly using the word customer? And they said, well, historically, in the utility industry, we just talk about meters. You know, for them, basically, everything that happens at the end of the line is about the meter. So you have this many meters in this territory. You have this many meters acting in this manner. So literally, up until maybe five or six years ago, there wasn't a recognition of the human at the other end of the line here when it came to electricity. This was sort of all about a system that was centralized, that was pushing electricity. And so it was sort of all about a system that was

pushing electricity down to a meter, not recognizing there was a person with agency at the other end. Now, of course, we had customer billing and a lot of customer-facing programs that really were meant to help and support customers. This isn't to say that utilities ignored the fact that there were customers. There were a lot of things that they did that were really important around supporting customers and still are. But in the engineering context, the customer wasn't so much recognized until relatively recently. But that's obviously changing. Now people have agency, and we've got distributed energy resources at the edge of the grid. So let's plunge in on what those are.

How do you characterize what's going on at the grid edge and what distributed energy resources really are? Yeah, absolutely. I mean, I think that there's a pretty widely held view of the potential of a future grid that has a really significant role for customers and not just as meters, but as active participants who might be part of a two-way system.

But then there's also a lot of questions about whether or not the grid that we have or the management that we have for the grid is capable of providing that. I think that's really what this conversation is about. It's how do we get from the physical infrastructure that we have today, which many will note is one of the most complicated machines ever built by humans to that dynamic two-way grid in a pretty short time frame and in a way that takes advantage of existing resources to really enable a much more. rapid pace of change. Right. And so when we're talking about what those resources are, I think, you know, the knee-jerk reaction, we all think about solar panels, right? This is the classic example where now you're generating, you know, electricity really at that grid edge and pushing it back onto the grid. But there's so much more than solar. What else do you classify in that bucket of DERs? Yeah, I mean, there's always the sort of standard list of four or five types. So there's

solar, there's battery, whether that's utility or behind the meter. There's EVs. There's various classes of smart customer devices such as thermostats or hot water heaters. But there's also a pretty large category of ones that people don't think about whenever they think about DERs. And this is all the stuff that's in the kind of commercial and industrial type space, things like large-scale refrigeration, things like large-scale fleet EV charging. Heat loads for buildings are often controllable. There's a lot of opportunity for electrification of very large loads. in things like some communities heat their streets, and they typically do that with natural gas today.

Many of these resources participate in existing demand response programs, some of which are very old. Some of these programs have been around for 30 years or more, and it's the kind where they actually call up a program participant on the phone to say, could you turn down your air conditioning next Tuesday for a peak event? But these are also part of the landscape of what we could think of more broadly as controllable customer resources, and their DERs as well.

Right. So, and really cool that, as you say, we've got a lot of this stuff that's already online. So let's talk about why that is. I think, you know, it's clear, obviously, the cost reductions have played a big role.

So solar cells and solar panels are much cheaper than they were, you know, say 30 years ago, which is really enabled, I think, a lot of the expansion of solar. But why else are people putting these kinds of distributed resources on the grid?

Why are we connecting loads? You know, why are we adding these kinds of resources in this particular manner? What are some of the benefits? of this distributed energy resource, you know, environment. Well, I think part of the benefit is definitely economic. A lot of the adoption of rooftop solar has certainly been driven by people looking to improve their

electricity rates or to make a major investment in solar panels at one time and then not have to pay utility bills on an ongoing basis for years after that. Definitely one really big benefit and a big driver of adoption. we're definitely also seeing a really significant acceleration of resilience investments. And so this is really true in any place that started to see a lot of grid disruptions. So in California, in my neighborhood where I live in the mountains, there's a massive waiting list for battery installations because folks in California have seen a lot of scheduled and natural disaster-fueled power outages. and it's not uncommon now, especially in more semi-rural places,

for people to see planned outages of days or weeks in the shoulder season. And in other places like Texas, there's definitely been some major outages in the last few years, which have driven really significant impacts on the people who live in the area. And any time you've been sitting there with the power up for more than a day or so, watching all the stuff in your fridge go bad and potentially, you know, being concerned about freezing to death. I think it really drives people to invest in resources in a way that's not strictly economically driven. At a certain point, you're just sort of looking

at it and you're like, what do I have to pay to make this go away? Everybody really likes having electricity. That seems to be a very important future of human existence. It matters a lot. And you really get to experience a sense of a very personal relationship with nature and civilization when you're sitting there on the third day of a major power outage. Right, which in some ways is lovely, but also when your food is spoiling in your fridge, it feels a little bit less fun. It's definitely changing for people as well. You know, a lot of us are used to storm-driven outages, which are typically only a couple hours. The power company is working really hard to get the lights back on, and you know that, you know, there are folks out there climbing trees and climbing poles in the dark

to help you out. It's actually a little different when you're sitting there on, you know, the third day of a planned outage, and you know that it's actually just going to stay out. until the wind stops blowing and fire conditions are safer. And there's really not a hope that the power is coming back before then. Yeah, that's really tough. Well, and then maybe one other to add to the list of benefits is, of course, decarbonization. It's fundamentally a lot of the solar that we're putting on there. And a lot of folks that are responding to some economic signals,

that's what some of the signals are meant to do, right, is drive some of this decarbonization. Yeah, and so there's a lot of investments that consumers have made and people all around the country, all around the world, in things like programmable thermostats and things like programmable hot water heater in EVs and beneficial electrification. And a lot of the motivation for that, of course, is participating in decarbonization. But there's also an expectation around these devices that either for the person who owns them

or for the company that deploys them, that there may be ways to find an economic benefit out of that as well. And I actually think that's something that's really been a struggle for a DERs. Many people have nest thermostats. Very few participate in programs where they're actually. actually getting paid to do anything on the grid. And so figuring out ways to bridge that gap from kind of the intention of being able to be grid supporting with all of these devices to ways to make that really happen in terms of the

incentives that are out there, the ability to signal, the ability to actually integrate them into the operational and economic landscape of the grid, that's a really big gap today. Totally. So, yeah, we'll definitely get into a lot of the incentives and how those relate to the decisions that people are making. But now that we've outlined some of the things that are beneficial about distributed energy resources. There are also some potential challenges, and we've seen some examples where as you get to higher penetrations of these kinds of resources, you come up against some troubles, and Hawaii presents a pretty interesting example not long ago where this happened.

So tell us a bit about what that looks like when you start reaching some of those kind of points where things get a little more challenging. Yeah, I mean, I think Hawaii is a really good example of what the role of distributed resources can be, both in terms of benefiting customers and benefiting the grid, but also causing a lot of pressure within the grid environment. Hawaii is a place that has a lot of sun. It also has very expensive energy prices. And so a lot of people in Hawaii have put on rooftop solar, whether that's because they want to really help with decarbonization or just because they really want to save money on their electricity bill. Both of those things are definitely drivers,

but it's led to a really high adoption for rooftop solar in Hawaii. And that's caused problems on the grid as the local utilities start to really struggle with managing grid conditions in the presence of so much rooftop solar. One of the things that we've seen go into place in places like Hawaii and Australia where this is also an issue is that you'll start to see limitations on how much solar can go in, whether that's limiting new installs or limiting the size of the install. you start to see a really significant pushback from the utility on the rate at which people can put these resources into place. That's obviously not ideal because it's limiting the economic benefit that people can get. And it's also really limiting our ability to use renewable energy. Now, ideally, people would put in the largest solar array that they can, and we would get energy

from all of those arrays. But in order for that to be possible, you need a lot of other things to be true. Like, you need to be able to manage the grid impact. You need to be able to potentially put that energy somewhere. somewhere. And so that's really caused a lot of friction between utilities and customers in places where this is the case. And it really does have some operational impacts on the grid that need to be managed. So yeah, this is one of those situations where, you know, at small penetrations of these distributed energy resources, it doesn't so much matter that we don't have a lot of visibility

into what they're doing. You know, the grid is kind of overbuilt enough that it can manage the input from these kinds of things. But that starts to not be true once you get to really high levels of penetration. But what you're pointing out is also true is that it doesn't have to be that way, that there actually are ways that we can really leverage data communications infrastructure, controls, optimizations, and ways that really would enable a lot more DERs to come onto the grid with relatively little investment. So let's talk a bit about that. What did the technologies look like that really kind of enable this sort of communications infrastructure that needs to be overlaid on the grid in order to bring on a lot of distributed energy resources? Well, you know,

One piece of good news is that there have been a lot of investments in instrumentation for the grid and additional data sources in the last probably 10 to 20 years that make this a lot easier. So most to many utilities now have smart meters that are deployed in the field that provide some amount of instrumentation about what's happening at individual customer allocations. And most have instrumentation at the grid level through their SCADA systems that tell them about what's happening. at the substation and the circuit level. So some of that data is already being provided today, but very few utilities have great data on what's happening with the solar generation within their territory. For example, most utilities don't directly meter that.

So they can't really tell exactly what's happening with rooftop solar. They also typically won't have very good access to information about the activity of any DERs. There's not a great way to get that into utility. as it stands today either, because you'd have to connect directly to the devices or have some other source of direct information about what's happening on the customer side. And most utilities don't have that. And most don't have data or software systems that can handle it either.

And that's a really big gap in terms of how we're managing the grid today. Yeah, I mean, that makes a lot of sense. This is sort of the picture I paint for people a lot of the time is to say Silicon Valley, super good at managing data, you know, finding ways to really leverage things like machine learning. Whole industries are changing. Not so knowledgeable necessarily across the board on the energy industry where, you know, there's a lot of resistance to change for some really good reasons, because, of course, you know, this is an industry that's trying to keep the lights on. But then within the energy industry, there's not a whole lot of resources or knowledge on how to really harness and

manage big data, particularly for effective decision-making and kind of infrastructure investments. But so what's going on? I mean, how are things really? really changing, say a little bit more about what it means to actually manage huge amounts of data, because I think, you know, as we've both said now, this is really, this is not the forte of utilities, and they say that it's really hard to manage a whole lot of things, but give a sense for what we actually can do with respect to data management and how that could ultimately affect our ability to do interesting things with distributed energy resources. Yeah, I mean, the state of the art in managing data and managing real-time visibility into big systems has really dramatically

changed in the last probably 15 to 20 years. And most utilities are still working with software and data management systems that are kind of old school. They're still deploying software that goes on a really big server in the server room that's sort of on site with the utility. And that was state of the art in the 90s and the early 2000s. But something that happened in the kind of early 2000s was a really significant transition away from that older model of computing that, you know, put all the data into one big database towards a model that was able to take advantage of lots of computers. And that's really, at the heart of it, that's what cloud computing is. It's really the ability to use a very, very large number of computers to tackle hard problems.

And so that's, you know, that's what my technical background is in. It's how do you use a really large number of computers effectively to manage, very, very large amounts of data and do so in a way that's reliable and trustworthy. And just give a sense of what that transition looks like. Most on-prem software solutions

will typically use sort of two, three, maybe five, or even 10 servers. Whereas the kinds of modern computing solutions are managing really large amounts of data from places like Google or Facebook or Uber or whoever will typically

be using tens to hundreds of thousands to millions of machines. If you ever sort of wondered, like, how is it that Google can give you an answer about anything on the Internet so quickly? The answer is millions of computers. Like, it's actually, like, at its core, that's really the technology leap that made that kind of global-scale computing possible. You know, this comes up when we talk to utilities because a lot of folks at utilities

are really struggling with dealing with scale just from the, systems that they already have in place. And that is really hard with those kinds of on-prem technologies and sort of old school databases. You know, folks at utilities will sort of, or vendor companies will kind of cite metrics around data volume.

And this one's really instructive to me. So if you think about how much data do you collect from a million meters every day? If you collect data every 15 minutes from each of those meters, that's $96 million. data points in a day. That sounds like a lot. But for a system like monitoring systems at Google to keep track of Google systems that conduct searches and manage YouTube traffic and provide

maps and all these things, if you look at search alone, publicly available stats on the number of searches per day are $5.6 billion. If you're touching 100 computers for every search, which is probably a low estimate by quite a bit. And you're keeping 10 data points for every interaction that that search has with a computer. That's something like 5.6 trillion data points per day. And again, that's a really, really low-end estimate. And so the size and complexity and sophistication

of the data handling systems that you use for understanding what's happening in a system like that and making sure that it's working the way it's supposed to are just quite different than the predecessors which are still kind of state of the art in the utility space. And that's a really big change. That's super cool. Yeah, I mean, I think that's really instructive to really illustrate that big data management

is certainly not the barrier here, right? This is actually providing a really incredible opportunity to do some very different things with our grid. But before we get into kind of what are some of the things that are going to support and kind of motivate this change versus some of the things that could hold it back, let's just talk high level. Where is this going? You know, especially kind of globally.

Do you think that we're going to get to a point where we've got like 50% or maybe even in some regions of the world higher, you know, like 80% of the electricity is distributed energy resources that are generating it? Or do you think it's likely to be some kind of smaller fraction, you know, that will stay somewhere close to 15% of electricity generated, for example, in California's grid? Where on the spectrum do you think this is going to go in terms of the expansion of distributed energy resources? I mean, I really think the answer is kind of all at the above. there's a lot of work that needs to be done by large-scale and utility-scale generation resources in the energy transition. And I've seen estimates on the share of energy generation from those large-scale resources varying from sort of 60 to 80%. One of the customers that we work with that's a local utility in Colorado, they're looking to get to fully decarbonized electricity supplies by, 2030 and the sort of back of the envelope calculation that they've given me as part of their

planning process is that they expect bulk resources to account for actually about 80% of that transition. But it still leaves 20%. Like if we really want to decarbonize, we got to we've got to get that last 20%. And we also really need more flexibility in the local use of energy and on the customer side then is currently available in order even to get to the 80%. And so I, I think, you know, if you think about just pure percentages, my guess is that probably 20 or 30 percent is probably about right for generation from local resources. But when you think about the role of moving demand around in time and making sure that you can really take advantage of renewable energy at the time and place where it's being generated

and then translate that into true decarbonization across the whole system, the role of flexibility is going to be huge. Virtual power plants are becoming a reliable way for utilities to manage. capacity. But enrolling devices is just the start. What really matters is confidence, knowing those resources will perform when dispatched and being able to prove it from the control room to the living room. Energy Hub's platform handles the full picture, from near real-time forecasting, locational dispatch, and the kind of rigorous verification that holds up when regulators, grid operators, or leadership ask, did it deliver? Easy enrollment creates momentum, proven performance

builds trust. That's why more than 170 utilities rely on Energy Hub to manage over 2.5 million devices delivering 3.4 gigawatts of flexible capacity. See what that looks like at energy hub.com. We're living through a profound economic shift, and energy sits at the center of all of it. Trillions of dollars are flowing into power plants, transmission lines, battery factories, data centers, but the future of energy isn't shaped by technology alone. It's shaped by markets, by policy, by capital, and by the institutions that connect them. I'm Alfred Johnson, CEO of Crux, the capital platform for the clean economy.

Join me for my brand new show, Critical Capital, as I talk with people deploying capital, shaping policy and building projects. Together, we unpack how risk is priced, how incentives are structured, and how progress is actually made. Listen to Critical Capital on Spotify, Apple, or wherever you get your podcasts. So give us some examples of how things are going. So we talked a little bit about California. We talked about one of the challenges that folks met in Hawaii. What about in some other places that you guys have some familiarity, some of the local co-ops within the U.S. or Australia? Yeah, absolutely. We work primarily with co-ops, and co-ops are fascinating. Cooperative utilities are much more

common than people maybe imagine within the U.S. there's about 3,000 utilities overall. And of those, about 150 are the investor-owned utilities that most people are familiar with, like PG&E or ComEd or Con Ed or whoever. The remaining 2,850 or so are primarily publicly owned utilities, and these are either municipal utilities that are owned by a city or a township or their co-ops, which are technically member-owned non-profits. Many rural communities in America are served by cooperative utilities, and most of them were created actually during the New Deal era as part of a widespread government-led program to bring the benefits of electrification to all of the rural communities around the country and make sure that they had equivalent access to the economic

opportunity that electrification provides as folks in cities. So there's actually this kind of really rich history and equity of access to energy that the co-ups are a part of. of. And that's one of the things that actually really love about them just in terms of their role within their communities. You know, you mentioned working at Exelon where they kind of introduced the idea of talking about customers. Co-ups actually hate talking about customers. They firmly insist on only talking about their members. Oh, love it. Because they are member-owned entities. And to them, that relationship of the people they serve being their members is really important. And so there's a few things that this drives in terms of their, their

incentive structure and how they think about their role in the community. One is that they really are very concerned about making sure that their members' interests are represented. And so if their members are very interested in decarbonization or renewable energy, you'll typically see the co-op reflect that in terms of investment in large-scale solar, in large-scale wind, in transitioning their energy supply, putting in programs for eBs and batteries, these kinds of things. And because co-ops are relatively small organizations that can happen pretty quickly. So in some cases, you know, we see co-ops investing in really dramatic goals, like the one that I mentioned, Holy Cross Energy in Colorado, that wants to get to 100 percent decarbonized electricity by 2030. It's really only six

utilities in the country that have set goals that are that short term for full-scale decarbonization. Another example from some folks that we work with in New Mexico, Kit Carson Electric Cooperative that serves the area around Tauas in New Mexico, set a goal to get to 100% of their daytime energy usage generated from local solar by 2022. And they are making that goal. They're going to be there within a couple months probably. And so, you know, these are goals that they've been able to set that are very aggressive and been able to execute on quite quickly. Whereas if you look at the goals that have been sort of set for bigger utilities around the country, it's much more often like fully decarbonized by 2050. which is effectively after a couple of generations of retirement. And it's typically not really linked to that kind of very short-term investment in local resources or in customer-sided resources.

And so, looking at the kind of progress that you can make with these community-based organizations, the kind of progress that they are making, I think there's a really big opportunity there to try out a lot of the work involved in decarbonizing the grid, both in terms of transitioning energy supply, but also in terms of transitioning into new technologies, trying those in the field and demonstrating that they can be really effective and that they can actually operate the grid in this kind of new way. Yeah, no, these are really great points. And I think, like, you're absolutely right,

that there's a really interesting trade-off, which was that, you know, at Exelon, there was a big commitment to sustainability and obviously to serving customers in some really important ways, but also, you know, a lot of resources that we could put behind this with a really big company. So that's one thing that you are seeing is that a lot of the bigger utilities are actually able to kind of make some big changes

and have innovation teams that can push on some of these sorts of things. But that tradeoff is then, you know, Exxon was a 34,000 person company. So there's a lot more that you need to do to kind of get over the finish line with some of these kinds of changes than potentially some of these co-ups that can be more experimental. So that's a really great point. Yeah, I mean, that's definitely the case. I mean, I think it's always easier to move faster. smaller organization than a big one. But you touched earlier on the fact that utilities tend to be

conservative about technology adoption and change because they have some good reasons for that, right? Most utilities are not necessarily rapid adopters of new technology. And the reason for that is primarily that they need to keep the lights on. You know, rapid technology innovation is awesome, but it's not necessarily associated with having really reliable and trustworthy quality of service. And so, you know, bigger utilities especially are responsible for, you know, energy supply to a lot of customers. And they need to do that really reliably. It's a critical part of everybody's lives. And so, you know, they tend to be, you know, both a little slower moving because they're large, but also because they're making sure that the changes that they make actually

work the way that they intend and are safe. Right. No, that's right. And this is really a great segue into the conversation about, you know, how we talked about some of the really big benefits of having lots of distributed energy resources and how that can be helpful to customers, but there are a lot of big barriers to really enabling that future. And so this is certainly one of them, you know, that you've got utilities that, again, for some really good reasons, are trying to keep the lights on and therefore not necessarily moving at light speed when it comes to making massive changes to grid architecture. But let's talk a bit more about some of the things that are, you know, some challenges

that need to get overcome if we're really going to realize this well-orchestrated future with lots of distributed energy resources. And you mentioned earlier in the conversation that incentives need to actually support this kind of transition. So say a bit more about that. What are some of the incentives that you're seeing kind of across the country or maybe even in some other parts of the globe that you think are really helpful in pushing towards this kind of future in a productive way? What are some of the things that we really need to work on with respect to setting up the right kinds of market and incentive structures to make this happen well? Yeah, I mean, you mentioned one of the biggest ones earlier, which is net metering. Net metering is an incentive structure that basically pays people money in order for them to install local solar on their rooftops and generate energy. So that's a really well-understood incentive structure, but it doesn't necessarily extend to any other kinds of distributed energy resources that people install today.

And there's not really very good standard mechanisms for doing that. So let's say that you install a smart thermostat today. It might be with the intention of helping to decarbonize or providing grid support with it. But in order for that to be actually happening in practice, the utility needs a program in place that can use those thermostats in a way that actually does support decarbonization or does provide some grid benefit. There's a few different ways that they can do that. But there's not really any standard ways that that happens locally today. one thing that is starting to happen very broadly is that more and more utilities have programs that they've set up that customers can register into that do various classes of demand response so that the utility can kind of call on customers to turn down their use of energy during times when energy supply may be restricted or there are other problems. And as I mentioned earlier, like this is a kind of technology that's actually very old.

Most utilities have had some form of demand response program in place for 10, 20, 30 years. But typically, up until now, they were really only participated in by very, very large customers because the mechanism actually was calling people on the phone to tell them to make a change. And still is sometimes today, right? It sure is. And so, you know, wise-scale computing technology provides the opportunity to potentially do this at scale, but it requires actually being able to use the software technology that makes this possible. So, you know, some of those incentives have been around for a long time.

But if you think about all those resources that are coming onto the grid today, whether it's people putting in batteries in California to stop the power from going out or people investing in smart thermostats or whatever, there's kind of two ways that a utility might be able to use them or that we collectively may be able to use them to drive decarbonization. One is to directly reach out and control those reasons. The other is to provide some sort of signal that that's something that we might want people to do. Now, you know, most people are a little bit reluctant to have the utility reach out and directly

control assets that they own, especially if those assets have another job. Like, if your battery's job is to make sure that the power doesn't go out at your house, you may or may not want to sign up to a program where the utility just does stuff to it sometimes. You know, maybe actuates it for grid support purposes, like, how do you know if you're still going to have battery charge available when you need it? And why would you hand that over also to an asset that you own, unless they pay you for it? And once you open up the idea of paying people for it, that's a much different conversation, right? Can you touch my battery? Well, I don't know. Like, why would I let you do that? Will I let you pay me to use my battery when I'm not using it?

Well, definitely, that's something that makes it easier to afford a battery. And so I think, like when we look at how you get customer resources more generally to do the things that you want them to do, paying people for it is a very time-honored tradition that is proved effective in many, many formats over and over again. But then the question is how? Like, whose job is it to do that? Who pays for it? Is that a market thing? Is that a utility thing? And how do you make sure that those financial incentives are aligned with the behavior that you want to get? Right. And that's all very tricky, right? Because also, you know, certainly I think you're right, The point is to pay people for it, but I remember we had the statistic we talk about, which is that people typically spend less than three minutes per year thinking about their

electricity bill. So it's not a lot of headspace that you have to kind of work with a customer and assess their preferences and invite them to make choices about their energy usage. So it's kind of, you know, you have to do all those things you said and also make it unbelievably easy and almost invisible for the majority of people who don't really want to think about electricity, right? Yeah, and that's why this is a job for computers. Like, it's never going to make sense to. contact a human really on a very large scale and tell them, like, the price of electricity

today is 13.5 cents a kilowatt hour. Yesterday, it was 13.25 cents a kilowatt hour. So, you know, you should really go ahead and, like, adjust all of your devices to take advantage of that. Like, that's, that is absolutely a job for computers. You know, the other thing is that it's not necessarily just about incentive consumers to make those changes. Because if it was, that would actually be quite difficult. Like, you'd need systems in everybody's houses. everybody would need to sign up for everything. But every one of those device manufacturers or device vendors or aggregators is interested in finding ways to make some money from the resource

or to find ways to provide extra economic benefits from the resources that they deploy. They all have computer systems that talk to all of those devices all the time. They all know exactly what every thermostat and hot water heater and EV charger and battery is doing every moment of every day. They all have an interest in making more. more money or allowing their customers to make more money off those resources. And so, you know, when we think about, like, how do you drive this change in a very large scale way, actually aligning with the interests of the people who either install or manage those devices and helping them to

accomplish their goals around helping people make more money or making more money themselves is the best way that I can think of to really make sure that all those resources are doing what you want. Right. Of course, at this point, though, now we're stacking up a lot of different things that really kind of need to come together to make this work, right? It's kind of you need the right collections of incentives. So certainly things that are actually going to get people to change their behavior. So sure, net metering, but you also want to make sure that any potential costs are distributed equitably across the system. So there's a whole package of things that you want to do. You want to make sure that you're including all the right stakeholders, which in this case is

including things like device manufacturers. And then to complicate all of that, not only do you need to get this package of incentives right, but ideally you need to do it in a way that across the country or even globally, there's at least a little bit of similarity. among how this stuff actually gets done and how it plays out. And I think maybe some folks may not realize that that's really tricky because there are a lot of different jurisdictions. They overlap. There are a lot of different ways in which regulation and policy ultimately affects how these things happen. So can you say a bit about that? How do you handle this environment, you know, within your company, within this space? How do you deal with the fact that really there are just so many different players when it comes to the regulatory landscape and things look really different in the middle of the country than they do say in California? I think this is a place where there's a lot of leverage in a small number of entities, if it's used correctly. So within the U.S., there are seven independent system operators or regional transmission operators. And they set a lot of the rules around how energy services can be rewarded within their territory. So that's like kind of one leverage point that can be common across many states and can provide common opportunities for many jurisdictions. Another is the role of federal. regulation, there's only so much that FERC can do, but there are some places where that can be really meaningful. So with recent regulatory changes at the federal level within the U.S., such as FERC ruling 2222, which mandates that local resources should be allowed to participate

into wholesale markets in aggregate, that's starting to put some structure in place that's nationwide that doesn't necessarily tell any utility how that's to happen or any aggregate or how that can happen. But it definitely says that it should. And so that's starting to really drive collective conversations between independent system operators and distribution utilities and aggregators and customers and other interested parties. And I think there's potential for driving collective change with that sort of very broad strokes. federal leverage that can make the problem a little bit simpler. No, I think that makes a lot of sense.

And I know that, you know, Shale loves to talk about how this is really a wonky show. And I think, you know, when it comes to electricity markets, it doesn't get a whole lot wonkier than FERC orders. So I'm glad that you brought this up. And just to really underscore that, I mean, it's a really amazing thing that they've done to basically say, you know, hey, we've got this electricity system that's built on these, you know, big centralized bulk resources. And the markets that are set up to govern those, you actually now have to let

aggregations of distributed energy resources play in those markets. That's a pretty big deal to have that be a statement that's made really nationwide. Yeah, and it requires some changes to happen in a really broad way within the industry as a whole. So that then requires the independent system operators to sit down with their stakeholders and figure out a mechanism for that. It actually mandates that those aggregations are allowed to cross multiple connection points between the transmission system and local grids. And so that actually puts quite a bit of pressure onto distribution utilities to figure out mechanisms that would allow them or enable them to be able to support that kind of interaction, which really today is not something that they would

comfortably agree to or really have systems in place to deal with. But it's something that we need if we're really going to transform the system that we have into the system that we want, where everybody can participate, and we really can leverage all these additional technologies. And so I think it's just a really helpful additional leverage point within the system as a whole

that starts to drive a lot more conversations downstream. That makes a lot of sense. Okay. Well, so now we've talked about a couple of things that are sort of pushing us in the right direction and some things, some hurdles that we're going to need to overcome. Let's talk a sec about the data piece of this

because, you know, we've mentioned a couple of times that big data is obviously going to have to underpin these transitions. And it certainly is true that managing huge amounts of it is not necessarily the biggest tech challenge that we're going to encounter here. But let's talk about getting the data. That's a bit of a trickier piece of this, right? Because you need to actually have the right kinds of data coming in. So maybe, you know, to start, how much data do we really think we need, right? Because you can imagine trying to connect every device under the sun all the way down to the coffee pot in my house. But you may not need to do that. there may be cases where you can use modeling, where you have some way of kind of estimating

or assessing what's happening on the grid. Where do you think we need to be in terms of data collection points? How many of them do we need? Where does the data need to come from in order to make this really work effectively? I mean, you know, I come from a big data and a big systems background. So, you know, in general, the more you can get, the better. But it's also true that the more you collect, the more challenges it poses about how to use it effectively and what to use it for. it's normal and state-of-the-art for any very large-scale distributed system like Google systems or Facebooks or Ubers or SpaceX or any company that deals with highly instrumented systems that have a lot of points you might want to know something about to collect all of the data from all of those points and keep it forever. Any worthwhile large-scale distributed system has real-time monitoring infrastructure that collects all that data in real-time, typically to the same. sub-second granularity level, and then makes that available for both management of the system and

for human operators to look at and reason about. Now, our transmission systems today are pretty heavily instrumented and actually have a really pretty good data collection and pretty precise real-time management. That's what our independent system operators do, and any utility that's managing transmission infrastructure typically has a good real-time view of the state of their system. Now, that's less the case on the distribution side. Until recently, it was pretty normal for distribution utilities to not have any data at all about what was happening at an individual customer location, except for the billing information that they might pick up monthly about how much energy, a particular meter used over the last month. That's why they thought about their

customers as meters. As those distribution utilities have started to roll out smart metering system, they have more information about what's happening at the customer location, but not always much more. It's certainly helpful to have data points every 15 minutes for a customer, but that's actually pretty close to best case for utilities that have really made a significant investment in data integration and acquisition. They might have 15-minute intervals for an individual customer. It typically lags by several hours. That's the time that it takes for data to get from the meter to the utility, sometimes by a day or more. And so it's really normal for distribution utilities today

to have a view of their grid that's at least two hours old and often more than a day old, at least as regards to what's happening at the individual customer level. Now, they'll pretty much always have data about what's happening right now at the substation level, but that's, you know, a substation might serve, you know, 100 miles of wires. It might serve tens of thousands of customers. And so that gap around what's actually happening at all the points in the grid below the substation, which is, again, nearly all of them, right?

Like all the wires, all the transformers, all the meters, et cetera, they're below the substation. And so traditionally, distribution utilities really haven't had great visibility into what's happening in all of those locations. But as more distributed energy resources come online, that's actually where they get plugged in. They get plugged in in in all the parts of the grid that the utility is effectively blind to. And so, you know, when we got started in this space, we were coming in from a background of having absolutely state-of-the-art, you know, sub-second instrumentation for these massive millions of computers, you know, dozens of data centers, globally distributed systems, and sort of took that for granted. That was something that you just had. It was something that many of my team members built and kind of led the development of. So it was new technology that we've been involved in,

but it was also just table stakes, right? Like if you're going to manage a big system, of course you need to be able to see it. And so when we got started in the grid space, we had initially, when we were kind of evaluating, putting the company together, expected that we might be working in managing all of the distributed resources

or bringing kind of next generation, you know, really smart software systems to optimizing and dispatching all of those distributed resources. But when I was researching the space before starting this company, I talked to a lot of people in utilities and in the energy industry more broadly. And over and over again, what I heard was basically manage it. We can't even see it. Step one.

I know that it's there. What we came to realize was that actually that that set of infrastructure around collecting and making sense of all the data was an important place to start. it was a table stakes kind of capability that many utilities actually don't really have in place and really need. And making sure that you actually can bring kind of state-of-the-art management

of large amounts of data collection, processing, analytics, et cetera, to really find those few sort of high-level, really critical pieces of information about the behavior of the system and where you might have problems. That was something that turned out to be, really important in order to set the kind of technology capabilities in place that would allow

us to do anything more sophisticated around like detailed management of resources within the broader distribution utility environment. Yeah. Okay. So that makes sense. So basically I think the upshot, you know, as you're saying, is that we need more data. We need more information coming in from these endpoints of the system. Is that right? Yeah. But we also need systems that are able to make sense of it. more data alone is not enough. You need to be able to figure out which data is important, and that's actually the hard part. Of the 15,000 or 150,000 transformers that I have on my system right now, which ones are potentially having problems because of the installation of solar panels or an uptick in the adoption of EVs? Well, good question.

Typically, many utilities will have one set of systems that can answer that question, but it's not the same set of systems that can answer that question. but it's not the same set of systems that can tell you about voltage distribution across the system. It's also often not linked to understanding of what's happening at the substation or the circuit level. And when you're looking to change a system

in really significant ways, you really need all that data to be in one place and you need to be able to keep track over time of times when perhaps unexpected interactions are happening or unexpected system impacts are occurring. And that's where being able to to pull all that data together and make sense of it really quickly, matters a lot.

Yeah. I mean, but one piece of that, right, is how you collect it and from whom you collect it and who gets access and all that sort of thing. So let's talk about that for just a second because, like, you know, I think as, for example, folks are connecting a new EV charger, you know, in their homes. Probably there's going to be, you know, some incentives for some connectivity and for some information to potentially be transferred with a utility. In some cases, maybe not.

But how does this play out? I mean, do you see a world where, you know, as people buy major. appliances, there are requirements for connectivity and then requirements for certain kinds of data sharing either with utilities or other kinds of players in the system. Do we need to go in and retrofit a whole bunch of already installed electricity infrastructure in people's homes? Or what exactly is the path going to look like to get to a place where there is some kind of centralized repository of data that can meaningfully support this kind of optimization and decision making? I think there's a couple of layers to the answer to that question. One of them is a concern around

privacy. Do you really want a central repository of data about every activity you conduct in your home? There's already pretty good technology around being able to identify individual houses by looking at their meter data. Providing additional data sources to a central repository about every single activity that you take from charging your car to turning on your TV is not necessarily something that you want to do in a wholesale or careless way. So making sure that there's programs in place that protect privacy and security of that data is really important. You know, one thing that could be really helpful in this space is a better understanding on the utility side and potentially on the regulatory side of which data needs to be shared and what doesn't. For most utilities today, if you were to connect data about your thermostat to data about your meter, that would also be connected to information about your billing history, about your name, your address, your relationship with your data.

your landlord, all of these kinds of things. But you don't really need that, actually, in order to potentially have your thermostat participate in an ISO market. Your utility needs to know about your billing history. Your ISO doesn't necessarily. Your aggregator doesn't need to know about your utility billing history either. If, you know, a smart thermostat vendor is going to be dispatching those resources, they need to know maybe about, you know, the temperature in your house and the set point for the thermostat, but they don't need to know about your financial history. So making sure that we're really careful about what data is shared is something that I actually think many state regulatory commissions are doing a pretty good job of. California has recent

regulations around this that do a lot of work around user data protection. But once that's taken care of, you also still need to be able to get to that data. How do you get to it? Well, like I mentioned earlier, every single device that out there is exporting telemetry to somebody. And that's typically to the manufacturer of the device or potentially an aggregator. So every Tesla battery is telling Tesla what it's up to. Every Tesla charger is telling Tesla what it's up to. Every charge point charger is telling charge point what it's doing. And so the question then is, like, for the data that you need about each of these devices, which is basically, you know, where are they located, what's their

role in the grid, what services can they provide, what impact might they have, you know, way to get to that data and that's where the incentives come in. So it's one of the reasons why I really like market incentives or pricing incentives for integration of these resources into the broader grid because it gives a utility or another entity like a DSO or an ISO, basically a way to say this data is valuable to me. I want you to share it with me. I will provide you a valuable consideration for sharing that data. And you can have that conversation with a smaller number of entities.

It doesn't have to be with every customer. It can be with the vendors. That doesn't necessarily completely solve the problem because there's a lot of vendors out there and you'd have to have a lot of conversations. But this is where I think folks like the ISOs can play a really important role because they're a large central location

where you can potentially start to manage data exchange between different interests within the grid. If you, as an aggregator, want to have 10,000 thermostat to participate in an ISO market, what does the ISO need to know about that? Well, it's probably just like how much resource can you provide and when. What does the distribution utility need to know about that? Well, it might be which customers are signed up

and might I expect a different usage profile from. It's not everything about you. It's just specific things. And so I think, like, using those leverage points, both for technological integration of data. So I, as a software provider, can go to a charge point or a Tesla or somebody and say, like, hey, what's your API? I can collect that data.

But then also having the business or customer relationships in place that are providing incentives to customers for sharing it in ways that they get rewarded for, I think, can help a lot because that starts to drive incentives collectively in a way that can really drive change. Yeah. No, that makes a lot of sense. I think, you know, Astrid, one of the reasons I love talking to is because, you really think about the full systems view on these things. Because I think that's pretty much in all climate tech conversations. We come to this point, right? Where you're like, okay, we just

need to solve like, you know, five things about this. Data privacy and sharing. We need to get the right market incentives in place. We need to make sure that the business models are all aligned. But what's so cool about you and your work is that you're seeing pass forward through all of these things in ways that like there actually is, you know, something at the end of this rainbow that we actually may see a future where there's, you know, really well-coordinated DERs on the grid. So to that And what's the next step? What's, like, highest on your list for things that need to happen to enable this future? Well, the good news about really dramatic change is that it actually happens all the time.

You know, I'm really fortunate in my career to have participated in one or two of these already. You know, I was working for Google when they first implemented their large-scale cloud computing technology platform and was able to see firsthand, you know, sort of how quickly you can really tackle huge real-world problems, like bringing together all the data or mapping every street in the world or providing shared infrastructure for, you know, sharing how-to videos about how to groom your cat. These are, you know, these are really hard problems and they've really materially transformed how people use everything from like the physical infrastructure of the internet to, you know, devices in their home. these big changes happen all the time. They just haven't happened in the grid yet, and that's what we're here for. And so thinking about, like, how can we apply the patterns that we know work well for driving technology change in other industries

is something that I think can be really helpful in the grid space. And so some places where we've seen that happen, the Internet's obviously a really big one of those, the infrastructure systems that underlie Internet technology, and communications have undergone really rapid change in the last 10 to 20 years. And there are some things we've learned about that, right? Like, we've learned that if you want to effectively drive a large-scale change, getting it into the field with some set of trusted partners and iterating quickly in the wild is very helpful to making that happen quickly and safely.

We know that in some ways driving technology change more quickly. It helps you adapt more effectively when you find problems with it. So, for example, we see this really driving a massive change in the aerospace industry. We went from it taking 10 years to build a rocket and launch it to SpaceX being able to do that once every couple weeks and really saw a very dramatic innovation once they started to move more quickly. So one of the things that I find really encouraging about this is that there's a lot of good data and precedent within many traditional, very safety-oriented industries that you actually can make change quite quickly and safely. And in some ways, those two things are linked. Like the faster you move, the faster

you can find problems, the faster you can really look at behavior of the system in practice and in the wild. And so long as you're doing that in ways that are effectively structured to mitigate damage to humans and physical infrastructure, that's something that can really be accelerated. And so I find that very heartening personally. The other thing that I would say is that we've really seen a lot of that change being driven by thoughtful and successful partnerships between industries. So I don't actually think that you get like really successful large-scale change from just throwing software at things. But I think we do see really successful, large-scale change when you bring together folks that have specific like software expertise with

people who actually understand, like, how drug research works or how the medical system works or how government technology works or how aerospace or rockets work. And making sure that you can put together the right teams of folks who really have deep domain expertise from both sides of the fence has a really solid record of driving large scale positive change successfully. And so, you know, that's more of a kind of squishy answer to your question. really. It's not a specific policy change. It's not a specific technology change. But finding ways to material and the interests of people who bring these points of expertise from different fields, I think makes a big difference quickly. Yeah. No, I think the point you're making is hugely

important. You know, I remember back in the mid-2000s during that clean tech boom that one of the buzzwords everyone kept using is disruption. And everybody was like, we're going to go disrupt the electric utility industry. And well, now we're all pretty clear. that that didn't work. And I think this is exactly the point that there's a real, actually, concrete reason to be hopeful now is that now there are folks like you, folks, you know, within the space who really understand what it looks like to make some of these dramatic transitions, but in ways that, you know, happen within legacy industries, where there's installed infrastructure, where you've got a lot of challenges where, you know, it's literally people's lives that

are on the line if you get it wrong. So it's really good news, and it's really great to hear about folks like you who are working on making these transitions happen. So Astrid, I really just want to thank you. Thanks for joining the podcast and for having this conversation with me today. Yeah, thanks for inviting me. I really enjoy talking about all of these topics, as you well know, and I hope it's been helpful. Astrid Atkinson is the founder and CEO at Camus. Catalyst is normally hosted by Shale Khan. He'll be back after he returns from Family Leave. The show is a co-production of PostScript Media and Canary Media. Find Canary and PostScript on Twitter.

You can find me on LinkedIn. Tag us if you want to provide feedback on this episode or want to suggest future topics. You can find links for this episode's topic and guest in the show notes, or go to canarymedia.com. Our producers are Daniel Waldorf and Stephen Lacey, theme song and mixing by Sean Markwind. I'm Laura Pierpoint, and this is Catalyst.