Pod Insights a16z Podcast Fraying Wires: The Decentralization of the Electric Grid About this episode The electric grid needs an update.
In this explainer, a16z Partner Ryan McEntush discusses the escalating complexity of the grid, unveils its vulnerabilities, and traces the evolutionary path that has led us to this point. From the surging demands of AI to outdated infrastructure, we delve into the potential roles of cutting-edge technologies such as solar batteries, natural gas, and nuclear power in shaping the grid's future.
Transcript Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.
0:00 Ryan McEntush? We can't afford for the grid to go down at all. One of the key risks for the grid over the next decade is actually policy. If that equilibrium is possible with the free market is, I think, a question on the minds of many people. We have an increasing outages, a lot of this weather-related. The price of electricity is now a lot more volatile. How decentralized can they be? Is it going to be 100%? Is it going to be 50? Is it going to completely fail? The new grid, the decentralized grid, is going to look very different than the old grid, and in many ways, it's going to be better, cheaper, more resilient. That's the future we should be aiming for.
0:37 Steph Smith? I bet you've heard that our grid needs an update. But I equally bet that a majority of people, myself included, don't have the faintest idea of how the grid actually works. We just assume that when we're cold and turn on the heat, our furnace kicks in and we get a bill at the end of the month. But as technologies like AI heat up, so too does their demand for power. This trend happens to coincide with several others, including the rise of intermittent renewable sources, the increasing bloat in the approval process for new grid infrastructure, supply chain delays exceeding 12 months, and more. All of these things make the fact that our grid relying on outdated technology, in some
1:19 Steph Smith? cases from over a century ago, something of a national emergency. And that is why we brought in A16Z American Dynamism Partner, Ryan McEntush, who recently wrote two deep dives on the grid and nuclear power to break down how all this actually works. So in today's explainer, we'll dive into the grid's increasing complexity, what's broken and how we got here, how technologies like solar, batteries, natural gas, and nuclear might play a role in this future, and why solving these problems is essential in allowing our economy and countries' technological dominance to thrive.
1:58 Steph Smith? As a reminder, the content here is for informational purposes only, should not be taken as legal, business, tax, or investment advice, or be used to evaluate any investment or security, and is not directed at any investors or potential investors in any A16Z fund. Please note that A16Z and its affiliates may also maintain investments in the companies discussed in this podcast. For more details, including a link to our investments, please see a16z.com/disclosures.
2:28 Steph Smith? All right, let's kick things off. Maybe an obvious question for some listeners, but why does our grid even need this refresh? And why do some people believe this might actually be one of the most important issues for our country to address?
2:43 Ryan McEntush? I think at the high level, the grid supports electricity transfer or industrial capacity in our country. There's this assumption that when you turn on the light switch, the lights are just going to turn on. But behind that, there's a lot of industrial processes, both in the power generation and moving that power around. And for the last couple of decades, we really haven't seen a growth in energy demand. So the grid has not really changed that much. But increasingly, over the last 20 years, and especially with the increase in data center AI for the energy demand on that side, we're seeing the grid change a lot. Increasingly, people are seeing the effects of that. We have an increasing outages, a lot of this weather related. The price of electricity is now a lot more volatile. And I think people are becoming more aware
3:26 Ryan McEntush? of not only the climate related effects of energy, but how it affects their daily lives.
3:31 Steph Smith? Yeah, absolutely. And you mentioned some of these new technologies. We're talking electric vehicles. We're talking about AI and data centers. But what is that gap really? If we look ahead, are we super far behind the energy requirement that we'll need? Or are we actually pretty close?
3:47 Ryan McEntush? All projections seem to indicate a huge gap. A lot of the current world is powered by things like fossil fuels, like your car, you fill up with gasoline. That's likely going to increasingly shift as more people adopt electric vehicles. A lot of the things in our home, you look at like thermostats, heating with some of these heat pumps. You look at like new induction stoves. Every single device now has a battery connected to the internet. So we were seeing huge demand there. And then also on the industrial side, we're seeing a lot of processes that were previously using fossil fuels as inputs, now increasingly using electricity. Another good example is like data centers, as you noted. These regions where a lot of these data centers are located are expecting huge growth in energy demand. To the point now that some large hyperscalers
4:32 Ryan McEntush? are looking at things like nuclear to support and make sure that the data centers stay online in periods of lower reliability or price volatility. This is a huge concern a lot of people are paying attention to.
4:42 Steph Smith? Absolutely. And it dawned on me that many people, including myself, don't actually have a clue how the grid works. I mean, to your point earlier, we just kind of assume when we flick the light switch on, that the lights turn on and that we get this electricity, not just sometimes, but all the time. And so let's just quickly discuss how the grid actually works. What is it composed of? And how do each of those components contribute to electricity actually being generated, but then also transmitted and distributed to all of our homes?
5:13 Ryan McEntush? I think the most important idea is that the electric grid is not some sort of monolithic entity that is the same across the United States. It's actually made up of three major interconnections. The grid is made up with AC or alternating current, but between these interconnections, there are a different phase, which means that you can't directly connect these interconnections. And they're split up basically as east, Texas, and west. And a lot of that was just for historic reasons, also, as I said, technical reasons with the phase, but also regulatory reasons. Texas, for many reasons, wants to have an independent grid, whereas the west and the east coast are completely fine having these interstate networks. And then beneath these interconnections are things called RTOs and ISOs. The difference is really mainly that RTOs oversee multiple states.
5:58 Ryan McEntush? So these are regional transmission entities, whereas ISOs are really just focused on a state level, which allows that more direct regulatory oversight and funding.
6:08 Steph Smith? Quick note, ISO stands for Independent System Operators, while RTOs are regional transmission organizations. To put it simply, these entities help ensure that the grid is reliable. Practically, that means balancing the grid, enabling efficient markets, coordinating infrastructure, and more.
6:27 Ryan McEntush? A good example is California. California's CalISO is just specifically looking at California, whereas some other RTOs are looking at the entire northwest region. These balancing authorities, as they're called, essentially operate these wholesale markets. They also operate a lot of the feasibility studies and planning and oversight of the entities involved in those local grids.
6:47 Steph Smith? So when we think about it, there's the interconnections, there's different entities that actually generate the power as well. And then there's market and there's regulation. Maybe you could speak to some of those other buckets and how they fit into the broader mix.
6:59 Ryan McEntush? Yeah, so at a high level, there's two types of entities. I'm going to really simplify this, but there's deregulated markets and there's regulated markets. So in regulated utilities, essentially it's top-down controlled by the government. Tennessee Valley Authority is a really famous example. There's a lot of these sort of regulated entities in the Southern United States, like Southern Power Company, or like a lot of these nuclear facilities were built, and a lot of this in the Western United States, spare California. In this case, the power generations, like the actual power plant, the transmission and distribution lines, and actually interfacing with the end ratepayer customer, all of that is controlled by the government with guaranteed profit margins and things like that. The other type are deregulated markets, which started really becoming a thing over
7:43 Ryan McEntush? the last like 20 or so years, largely pushed forward originally by Enron and pushing these policies of these free markets where energy was going to be traded in these wholesale markets. In these areas, the dominant markets would be California, Texas, the East Coast. There's power generators, people who own the plants, and then oftentimes different companies will own the power lines and be in charge of maintenance and operating the congestion and making sure that the power moves and is delivered. And then there's separate companies that actually interface with the customers. These interfacing, providing like a retail experience to the end ratepayer. So oftentimes, these are completely separate, different companies. Sometimes they're the same, sometimes these look like a little bit of a hybrid structure.
8:25 Ryan McEntush? It all gets very confusing, but it's part of the complexity of the grid that there's very, very different structures depending on where you are.
8:31 Steph Smith? Yeah, it certainly sounds complex, and I think there's probably a bunch of misconceptions. I mean, even just the idea that the grid is a monolith, it's clearly not true. Does anything else jump out there in terms of this idea of public misconception around the grid and how it works and ultimately who runs it, who's getting paid, and how people ultimately get their electricity?
8:52 Ryan McEntush? Yeah, I think there's like a technical idea. People think that you plug something into the grid and it's magic. But electricity is not like a commodity. It's certainly not magic. Electricity is generated basically at the moment it's being used. When people say the grid needs to remain balanced, it's measured in terms of frequency. In the United States, it's 60 hertz. In Europe, it's like 50, I believe, and I think in Japan, it's 50 as well. But basically, you need to make sure that the demand for electricity, basically people sucking electricity out of the grid and then other power generators pushing electricity in is always the same to maintain that perfect hertz level. And if that ever falters, when people say like the grid is unbalanced, that means it shifts away from that point.
9:35 Ryan McEntush? In which case, a lot of the electrical equipment hooked up to the grid that is calibrated to a specific frequency is damaged. And if things happen too quick, you have to shed load, meaning you have to pull things off. And if you can't do that, then you have blackouts where essentially equipment's fried and you have to like completely replace it, which could take weeks or months. It can be very dangerous very quickly as the winter storms in Texas over the last couple of years and the East Coast made really very clear. And then also, the other big point is about this reliability. It's very hard to predict the demand for electricity and the supply of electricity that's going to come in the far future. Most of the time, it's really just looking at weather and increasingly so as solar and wind become larger portions of our grid.
10:19 Ryan McEntush? You can assume the sun is going to shine every day, but not always. Like sometimes it can be cloudy and it can decrease the percent. People often say like the wind is always blowing somewhere, but that only matters if you have the transmission lines to move wind from one place to another. So like a nightmare scenario can be you set up a grid with a large portion of wind and it's winter, so everyone has their heaters on. But the wind, for some reason, is not blowing right as like a cold spell hits. So that's when you have this issue of huge spike in demand. But not enough supply for it. And so there's a bunch of scenario studies that these grid operators will do, trying to figure out like, can we ever guarantee the wind will be blowing in a certain time? And it's really, really hard to do that.
11:03 Ryan McEntush? So that's why it's so important that it's not going to be all nuclear, all solar, all wind, or all natural gas, because there are situations where pipes freeze. And so it becomes very, very complex. And it's not as simple as just connect as much solar as possible. Reliability is a huge issue. In a modern country with our modern luxuries, we can't afford for the grid to go down at all.
11:23 Steph Smith? That's such a good point, because it really does feel pretty much binary, right? It's not like we want uptime of 95%. It's like we want uptime of 99.9999% or 100%, right? Just basically, people, we cannot afford for the grid to go down. And so it does feel like people automatically assume like, let's add more supply. And there are many reasons, it seems like, why that is not such a simple solution. We'll get into the energy mix in a second. But maybe you could just speak to this broad dynamic that you mentioned in your article. That is, that the price to generate electricity has come down over the years with new technology. Yet, the price to actually distribute it or deliver it has gone up. Sounds kind of counterintuitive.
12:07 Steph Smith? Where are things really breaking down there?
12:09 Ryan McEntush? So I think the most important point to understand here is that when you pay your power bill, for me, PG&E in Northern California, you're paying a combination of prices. You're paying, on one hand, the price of the energy itself, which in deregulated markets is known as like the wholesale price. So this is the price of like the power, out the door cost that the power plant is selling the power to these other entities in these power markets. That's one portion of it. And then there's also the cost of the infrastructure and operational costs of these utilities like PG&E, who own the lines and might interface with the customers to deliver that power to you. And so like if you look at your electricity bill, typically they'll separate out these fees. And you can see like many times, and increasingly so in California, the cost to deliver is like over 50% of the bill.
12:55 Ryan McEntush? It's becoming increasingly high. And the big reason for that is that the infrastructure to build new power lines, to build new power plants is incredibly, incredibly expensive. The interconnection queue is something that people might have heard about before. But basically there's over like two, I want to say it's like two terawatts in the queue right now, which means that there's like two terawatts of power generating capacity, a nameplate at least, like what it could max out at, waiting to be connected to the grid.
13:21 Steph Smith? Could you just super quickly ground us on how that compares to how much electricity is on the grid?
13:26 Ryan McEntush? I believe our grid today is about 1.2 terawatts. So this is almost effectively double what our current national generating capacity is. So what this shows to me is that it appears like there's a lot of demand for electricity. Now, when you unpack what's going on, there's some interesting, I would say, disincentives of how this queue works, where because of what I mentioned earlier, like the frequency and grid balancing, there's a lot of feasibility studies that go into can I put a gigawatt of power generating capacity here? And then how does that affect the ability of this localized region as well as the broader grid? And these things take a lot of time, especially when people are sort of, I want to say, spamming projects where there,
14:09 Ryan McEntush? you know, there's a company who puts in like 10 proposals of different power plants, waiting to see which ones are going to have the cheapest interconnection fee. And then they'll pull back all the ones that are too expensive and just end up trying to build the one that's the cheapest. But the problem with that is that all these feasibility studies have to run with the idea that all these plants are legitimate. And so this has created this massive queue of maybe, I think it's like 10 to 20% of projects are actually built.
14:35 Steph Smith? That's wild.
14:36 Ryan McEntush? It's crazy. And I think there's a lot of really interesting new policy coming from the federal level about how to manage these queues. Texas famously has a way that they handle it. They call it, I believe, connect and manage. If anyone has been paying attention to electric grid recently, they'll know Texas has been able to connect a lot of assets. But instead of running these feasibility studies, they'll basically levy that risk of connection to the developer and say, you can connect your solar or wind anywhere, but if it at all risks our grid, we're going to cut you off.
15:02 Steph Smith? Interesting.
15:03 Ryan McEntush? Whereas in California or other places, they're like, we want to make sure it's always going to be delivering power and it's not going to destabilize our connection. So there's a lot of ways to handle this, but that makes any new construction very, very difficult. But the success of Texas and how much wind and solar and batteries they've been able to connect and why a lot of companies are flooding there is because of this policy that makes it very, very easy to connect new assets.
15:26 Steph Smith? If you didn't catch that, basically what Ryan is saying is that ERCOT, the Electric Reliability Council of Texas, is flipping the incentive. Instead of the onus being on them to prove that a project can work, it's on the developer to make it a reality. And if they don't, well, there are consequences. But there are also other approaches being explored in tandem.
15:46 Ryan McEntush? There are also interesting ways that FERC, which is a national body overseeing this, trying to bundle up different projects, providing fees. So instead of like being able to sort of just YOLO projects in, you are now sort of taking on some sort of financial risk by putting in fake projects.
16:03 Steph Smith? Right.
16:04 Ryan McEntush? So trying to disincentivize doing that. But even if you can get these projects approved, things like transformers are crazy lead times, like over a year or more, and not to mention that it's increased in price dramatically. If you're trying to build out like a new housing project or something like that and you have to build like the substation, securing these things is also a huge problem, which are just additional things to worry about. Both cybersecurity and physical security, I think of the last couple of years, people have been shooting substations and it takes down like a local grid. There's very little you can do there besides just paying somebody to just be on site and securing that. And then perhaps lastly, like just the existing power lines.
16:46 Ryan McEntush? I believe the local PG and E fires that people might have seen Northern California. I believe Camp Fire was one of them. But some of these lines were built in the early 1900s. I was up hiking at Muir Woods the other week. And if you're up at the top of the hill, you can see these transmission lines that are going through. And there's trucks that will go and manually check a lot of them. If one of these lines falls over dry grass, like it's likely going to start a fire. And that's what we continue to see, especially if they're that old. When something like that happens, they don't know where the line actually fell. They know like what type of power line it fell, but not specifically the spot. So it actually often takes trucks driving through to see where these damages occur. So there are a lot of interesting like potential upgrades that need to be done.
17:29 Ryan McEntush? And all of this just ends up being pushed onto the consumer. The company has to like handle all these feasibility studies. It has to build upgrades on the lines, prevent fires. A lot of this stuff ends up going back to the right pair and leading to the increased delivery costs.
17:46 Steph Smith? That's so interesting. I did not realize they can't actually tell where things are breaking down at the local level. So I've heard a few things. You mentioned that the actual queue is increasing in terms of adding more power. I'm also hearing things maybe around supply chain, security. Is there anything else you'd mention in terms of just the overall set of challenges that exists as we're trying to improve this grid?
18:10 Ryan McEntush? I think this rise of intermittent power introduces a lot of complexity. And not just intermittent power, but where the power is generated. I think one of the big theses we have is that there's going to be decentralization of the grid. Now, it's not like we're going to start taking offline all of our large power plants. But as the demand for energy is increasing, it's going to become easier for people to generate power on site. Solar is going to be the big reason for that, but also battery storage. And as I mentioned earlier, people are going to have electric vehicles that can store energy and connect to their house. Your stove is going to have a battery likely. Your heating might actually have some energy storage capabilities as well.
18:49 Ryan McEntush? That power is going to be flowing in many different directions, which is very different from a world where power is generated at a big plant and just going in one direction. So people need to rethink how grid infrastructure works and functions. It is not as simple as just pressing the button and the power going backwards.
19:05 Steph Smith? Totally.
19:05 Ryan McEntush? So that's something that the grid operators are thinking about as well.
19:08 Steph Smith? Yeah, that sounds really important. And maybe at the crux of us needing to distribute more of the grid is this idea that a lot of our new sources, as you're saying, are more of that intermittent type. So we're talking a lot of renewables. So maybe we can just start there and address the question, why are renewables alone not enough, especially as we increase the capacity of solar and wind? You spoke a little bit to this, but maybe you could just break that down explicitly.
19:35 Ryan McEntush? I want to be clear, renewables are fantastic, and they are cheaper, even without subsidies. LCOE is often a metric thrown around, which is the levelized cost of energy. Solar and wind are really, really good. The problem is that you cannot guarantee that they're going to be available when you need it. But it's often the case that, for example, if you look at the demand profile for energy in any sort of modernized economy, it's typically in the early mornings, everyone's waking up and the energy use is high. Then there's a trough in the middle of the day, which is right when it's sunny, when the solar panels are generating the most electricity. And then there's a huge spike in the evening when everyone comes home, turns on their TV, starts cooking dinner. And there's a drop as well later into the middle of the night. So the problem with this, obviously,
20:17 Ryan McEntush? and this is a phenomenon known as also as the duck curve, if people are familiar with that, but basically, the middle of the day, when energy is essentially free, when solar, the sun is blaring, and energy is really, really cheap, but the demand is at its lowest point. So you need to find a way to capture that energy in the middle of the day and have it available in the evenings, when the sun is not shining, or when it's dusk. And then similarly, it's something like wind, you cannot guarantee that wind is going to be blowing. Generally, you can look at trends, have this idea of when's going to be blowing, and generally at night, it's windier. But this is really, really hard for operators to rely on. So this is why energy storage is so critical. You're doing
20:58 Ryan McEntush? intermittent sources, you need some way to sort of level it out. So operators know that we're going to have enough energy when we need it. And this is coming from a world where we used to have natural gas plants that could chug along 24/7, say, but nuclear, which is, these are things that are our baseload. Geothermal is another one. Previously, in our country, almost all of our energy storage, I think it's like 90% plus today, is through hydro, through dams. So when there's a shortfall at all, we would open the floodgates a little bit on the dams, and the water would trickle through and be able to make up smaller changes in energy fluctuation, or major ones. And so if we're going to increase the amounts of solar And wind on the grid, we also need proportionally increase the amount of
21:41 Ryan McEntush? storage. And this is going to be very significant challenge.
21:45 Steph Smith? Yeah, I mean, to your point, it sounds like we need some balancing mechanism. We'll talk about natural gas and nuclear in a second, but let's speak directly to batteries. Something that I found really fascinating from your article is the point you just made about hydro. When I think about storage and many people think about storage, they automatically think of batteries. And only 2% of our US storage in 2022 actually came from lithium ion batteries, which was surprising to me. And 94% from pumped hydro, like talk about misconceptions from earlier. I did not know that in any capacity. And so can you just speak to where we are today? And the storage demand, whether that comes from battery or otherwise?
22:26 Ryan McEntush? Sure.
22:26 Steph Smith? How do we get to where we need to be? Is it just a matter of creating a bunch more batteries or hydro capacity? Is this all really feasible to do in the next few years?
22:36 Ryan McEntush? Yeah, so it's difficult to make projections here because a lot of this comes down to the cost of batteries.
22:41 Steph Smith? Yeah.
22:42 Ryan McEntush? If batteries are $10, $30 a kilowatt hour, then a lot of this stuff becomes meaningless. Obviously, there's critical mineral concerns about where we access this stuff. This newer generation of batteries that are less energy dense, like sodium ion becomes less of a concern. Regardless, that's going to be the key thing to pay attention to. What I will say is in the near term, batteries have already dominated this very, very short timeframe markets. So I mentioned these wholesale markets earlier, but they're sort of structured in different time durations. So depending on how quick you need to supply energy. So I mentioned sort of the fact that you need to maintain a certain frequency. There's a market that's called broadly ancillary services, which is really just focusing on power quality.
23:23 Ryan McEntush? So if it moves from 60 hertz a little bit, the grid will sort of seek more energy or less energy to balance it out. Batteries already dominate that because they're very, very good at moving very, very fast. But it's a very, very small amount of energy. Increasingly, as batteries enter the grid, which as you've noted from the chart, it's not that significant penetration yet. They will start to be able to cycle over a daily timeframe. So as I mentioned earlier, taking energy from the middle of the day and allowing its use in the evenings when people actually need it.
23:55 Steph Smith? If you didn't catch that, batteries currently only cycle from minutes to hours. But the hope is that their storage duration will become hours to days, just like hydro, which composes the lion's share of US current storage capacity.
24:09 Ryan McEntush? Now, the problem with batteries is they can't solve these longer duration spells. So I mentioned this case where if your grid has a huge wind component, it might be not windy for weeks. Unlikely, but it's possible. Batteries are not very good at storing energy for long periods of time. When I say long, days or weeks. You'll need alternative methods of storing that energy. There are a lot of things proposed, a lot of companies trying novel solutions, like synthetic natural gas or hydrogen. There's gravity batteries, there's sort of iron-air batteries, a bunch of different ways to do this. But these are essentially trying to do almost weeks to months to seasonal storage. And that's also going to be a component as well that I'm excited to see continue to build out.
24:53 Ryan McEntush? Batteries in the energy sense, but not the same batteries you'd see in cars.
24:57 Steph Smith? And then let's jump to natural gas and nuclear, which are two other, you could say, balancing mechanisms that have been used in different capacities in different places. They have their pros and cons. At least with nuclear has a level of public discourse around the use of it. And so, yeah, just tell me a little bit more about where natural gas and nuclear currently play a role, and then also the potential for them to play a different role as we look ahead.
25:22 Ryan McEntush? So natural gas is dominant on our grid. The United States is a major natural gas producer. Fracking was huge for this country in terms of energy independence. Today, there are two different types of natural gas plants. There's some that are running 24/7. There's some plants called peaker plants that essentially, if there's a shortfall or if there's no other sort of generating capacity, peaker plants were bid to fulfill that gap. And it's more expensive. Let's say you're building a natural gas plant. You can amortize it 24/7 and it's going to be cheaper to run that power plant as opposed to some of these peaker plants might run at a 1% capacity factor. It was effectively 1% of the time that they could be running. And those are going to be way more expensive to generate that power,
26:04 Ryan McEntush? but it'd probably be cheaper than using solar and storing it in batteries today. Now, natural gas is also good because you don't really rely on the weather, so you can always turn it on. And you can typically ramp it up pretty quickly. There's a lot of benefits there as well. Also, I would note that the relationship between natural gas and nuclear is a historic one. Natural gas was, I don't want to say the reason that nuclear found a lot of competition, but from a price point, as we looked at nuclear in the late 20th century, was the same time that natural gas became very big.
26:35 Steph Smith? Interesting.
26:36 Ryan McEntush? Natural gas quickly became very, very cheap, and nuclear became very expensive. So, natural gas became a very big piece of our grid and will likely continue to be a major piece of our grid as long as our domestic production continues and there's no sort of radical change in energy source. Now, if we can make nuclear fission incredibly cheap, we can have nuclear fusion become incredibly cheap, then a lot of this need for dispatchable or ready-to-go energy sources will change. But until then, natural gas will probably remain a significant piece of our grid because it is reliable.
27:08 Steph Smith? We've been working on nuclear for decades, right? And it would be surprising, I know there's a long chartered history of what's happened in those decades, but it does seem like even just purely from the technological perspective, we have new reactor designs coming online or at least being developed, and maybe in a new renaissance of public discourse around it.
27:27 Ryan McEntush? Absolutely. I think the main thing to understand for nuclear, particularly the historic nuclear, which are these gigawatt scale power plants, the newest of which the designs, the AP1000, this is Vogel, this is Southern Power Company, that is a regulated utility that saw a lot of government support to build this. In fact, the local companies, a lot of times when people are building nuclear power plants, the utilities will go out of business, the government needs to step in to help them. It's been a really, really difficult journey for utilities to buy large nuclear power plants. So I think that's the primary reason you haven't seen it done. And those who have tried, like Vogel, you see these horrible cost increases and a lot of problems. And I think I go into that in detail in my essay of what those causes are.
28:12 Ryan McEntush? It is sort of rational, like why would a utility risk bankruptcy? Or even more so, like, I don't know when the power plant's going to be up. Is it going to be 10 years, it's going to be 20 years? You don't really know. And so it's really hard to make that $10 billion bet. So that's why it's very, very hard for especially deradified utilities. It really takes a government, which is why... Nationally, Japan and France, and even like I said, our regulatory ties in the United States, a lot of our nuclear power plants are government-supported. It's very, very hard to take on that risk as a private entity. Now, the future of nuclear, I think what we're seeing is the rise of SMRs in smaller reactors. I think that the reliability of power, I think, is going to become increasingly relevant for a lot of people. In the news, you've seen Microsoft, just some of the other hyperscalers, talk about,
28:58 Ryan McEntush? we want to do nuclear-powered data centers, we want to have access to clean energy that is available 24/7. Now, you could achieve this with a large solar farm and batteries. You could do it with geothermal, which some of these hyperscalers are also looking at. But nuclear is one of the most compelling. If you're one of these data centers, you want to ensure reliability. You want something that you plug it in, it's going to last 15, 20 years. Nuclear sort of uniquely provides that. So in terms of this decentralization trend, solar and batteries are going to be a piece of it, residentially, and perhaps like these local communities. It's likely that these off-the-grid or this focus on reliability over sort of cost, nuclear is going to be a very compelling option.
29:39 Ryan McEntush? Now, all that's to say that we solve some of the issues on the regulatory side and some of these new designs get approved and built.
29:46 Steph Smith? Since we're on the topic of policy, does anything jump out at you in terms of where policy can help kind of usher this new grid forward or help solve some of the gaps that we just talked about in terms of like the interconnection queue or the different energy mix that we're discussing?
30:03 Ryan McEntush? FERC is the federal body that oversees a lot of the transmission and grid infrastructure. The stuff I've read, at least, seems very positive. People are aware of a lot of these issues. And there's a separate group called the NERC, which is a semi-government-associated, but they run these risk reports. And they have highlighted that one of the key risks for the grid over the next decade is actually policy. So a lot of people will say things like, we need to go completely renewable. We need to be completely solar and wind expeditiously. The problem with that is, as I said, like the reliability concerns, it's going to create a lot of, lot of issues. But I think when we think about what are our climate goals, like these are human goals, like we want to live in a better world. That better world near term is not a world with increased blackouts.
30:46 Ryan McEntush? There are a lot of countries that were once had great grids. South Africa, great example. They have nuclear power, but much of the day in South Africa, the grid is actually completely down.
30:56 Steph Smith? Really?
30:56 Ryan McEntush? So dysfunction from a policy perspective, there are worlds where it can revert.
31:02 Steph Smith? Right.
31:02 Ryan McEntush? And the reliability that a country once had disappears. And I don't think it'll get that bad in the United States, but there are a lot of interesting warning signs and people in utilities and people who have been in the power industry for a long time are beginning to speak up. The last thing I'll mention is the IRA. So the Inflation Reduction Act put a lot of policies forth for energy. So a lot of companies, for example, batteries, manufacturing credits, various incentives to both bring onshore a lot of this production, also make available a lot of funding for grid enhancement, both locally and sort of regionally. So there's a lot of funding available for this, which is very exciting. And so a lot of the things that need to be done, it seems like are beginning to happen. So I'm optimistic.
31:46 Steph Smith? Me too. And maybe one other area of optimism can come from this idea of a smarter grid. I mean, something you spoke to earlier is just this idea that transmission lines are a century old or they're not updated to the point that we are at today in 2024, where we are in terms of other technologies. And so what does that even mean? When people talk about a smart grid?
32:10 Ryan McEntush? Yeah.
32:11 Steph Smith? What would that really look like?
32:13 Ryan McEntush? Yeah, so I think at a high level, a lot of this big power plant to customer, you plug something in the wall, it's there. All that is very dumb, for lack of a better word. There's no signals coming back. They'll monitor the frequency, they'll monitor some of the local sort of substations to make sure that everything's okay, but, but they don't really know very much. Similarly, like I described, looking at energy demand, like trying to say in a week, how much energy will be demanded of a certain region? A lot of that is just guessing the weather because heating right now is like the biggest component and increasingly on the supply side as well, like will there be sun or will there be wind?
32:48 Steph Smith? Mm hmm.
32:48 Ryan McEntush? So that is growing in complexity. The Spark grid is really just introducing a lot more data and a lot more connectivity across the different nodes on the grid. So it's not just using AI to supply and demand. It's also like making sure everything is producing data, there's telemetry, you're able to connect to your EV, the smart thermostat, your whatever it may be, your stove in your house. And that sends signals back to the grid, back to other entities that help not only figure out what is happening live, but also what will continue to happen. What are the behaviors of individuals? How can we better model and predict future use? How can we then know what to build? So these are all very, very interesting things that are changing. And this happens both behind the meter.
33:33 Ryan McEntush? So like I said, all this like home energy storage type solutions. Also ahead of the meter, like how do we, you know, enhance these grid enhancing technologies? Dynamic line rating, like making sure infrastructure is more set up, better transformers, all that. And then obviously there's a huge software component. Forecasting, making sure that these feasibility studies are done a lot more efficiently. It's a lot of things to think about, and increasingly so, the complexity is growing. And frankly, I think we need our best builders to go and tackle these problems today.
34:03 Steph Smith? So it's clear that there is a lot we can do in this new age of technology with new software. But there's also this idea of decentralization. And I feel like I've heard of all of these different kind of early experiments where we're thinking about the like virtual power plant that has a bunch of Teslas on the road and they can actually send electricity back to the grid. Is that what people mean when they talk about decentralization? Or is there some larger trend here when we're talking about the full grid? What's your take on this future of the potential to decentralize the grid versus the very centralized grid today?
34:35 Ryan McEntush? So this is, I think, the philosophical question that a lot of people working in the energy space and the grid in particular are debating, which is the centrally planned, top-down electric grid that like France or Japan or a lot of other countries utilize that that might enable more nuclear power because the government steps in as willing to take on that risk. Versus one that's deregulated, that is like a free market approach, which seems to incentivize a lot more of the lowest marginal cost generation of electricity because you're constantly competing over generation, which might incentivize solar and wind, for example. I think what's interesting and why I personally think, and one of our theses is decentralization, because a lot of the advancements are coming at the edge. It's leaning into this decentralization trend.
35:19 Ryan McEntush? So you mentioned the virtual power plants. Obviously, we've talked about these home energy storage, a lot of the interesting software, trying to solve a lot of this growing complexity. But a lot of that is innovation that can happen in spite of whatever the broader grid does. The energy demand is increasing, so someone needs to meet that demand. Solar, batteries, SMR, nuclear reactors, these are things that are inherently decentralized. They don't necessarily take advantage of the scale of large thermal power plants, which have increased economies of scale as they get bigger. They're able to be closer onsite to where the power is needed. And by doing so, you also avoid having to build these incredibly expensive grid infrastructure.
36:00 Ryan McEntush? No longer do you need transmission lines crisscrossing the states or going into your woods. You can generate it in your local community or next to a sort of like a factory area where many people are joining together to pay for an SMR or maybe even data centers or something like that. And so as this demand increases, I think we're going to see increased amount of energy being built at this edge. Now, what I'm not saying is that people will completely defect. I think there are arguments to be made that perhaps certain people, particularly wealthy people, can perhaps go complete defection, have their own solar and energy storage systems. I'm not saying that, some people are. But I think what's going to happen is the amount of power that you're going to be buying from the central sort of utility
36:42 Ryan McEntush? versus generating on your own will be increasingly shared. And I think there's a lot of opportunities for entrepreneurs to help facilitate that. At the end of the day, I think that's going to be creating a more decentralized resilience grid and hopefully a cleaner and cheaper one as well.
36:56 Steph Smith? Yeah, so I think you're right. There are a bunch of trends that are kind of inching us towards this more decentralized grid. But I think it's worth just Reflecting on the fact that the grid historically has been pretty centralized, how have we benefited from that more centralized grid to date? And do you really see that changing, that kind of benefit from having the economies of scale?
37:19 Ryan McEntush? Yeah, so this is really the big question, particularly if you look at it on the international level. China does not have these like free deregulated markets of trying to balance supply and demand with these own firms competing. They all just build hundreds of nuclear power plants, a ton of solar, a ton of high voltage DC lines, like large scale transmission lines crisscrossing the country. This is something they've needed to ensure sort of industrial competitiveness. China previously had a bunch of issues with blackouts in the early 2000s. They've solved a lot of those. They've been able to build up the infrastructure very quickly. That's because the government was able to step in and spend a lot of money to do this. When we look at like France or Japan, countries with a ton of nuclear, as I mentioned previously, that's government.
38:01 Ryan McEntush? Government is the one who's ensuring that those things are funded and supported. I think the interesting question that a lot of people ask, particularly, I would say, people who've been in these sort of energy markets for a long period of time, are how can deregulated, how can the free market incentivize reliability? Right now, if I'm a solar developer, I want to make as much money as I can. And so I'm only going to really care about the profitability of my generation selling it into the wholesale market. I'm not really paid to ensure that the grid as a whole is reliable. So people often use like levelized costs of energy, like what does it cost to generate? But on a system level, like if someone adds solar, someone else has to add batteries or you have to put a peak your plant somewhere and build that to make sure that if something happens to the solar,
38:46 Ryan McEntush? you're still going to have that energy to be able to be delivered and keep the lights on. That's like a system level thing that the grid operators have to think about. And like in ERCOT in Texas, you might see like the price spike to $10,000 a megawatt hour during these storms, which is crazy. They have no price ceiling for this stuff. And that plant may run one day a year, but on that one day, it'll make hundreds of millions of dollars and pay itself off. And so there's like market incentives that will try to balance this out over time. In theory, these price spikes will introduce people to that market and increase competitiveness to drive down those spikes to the point where these margins are healthy, but but nothing crazy. How long that equilibrium takes to reach? And if that equilibrium is possible with the free market is, I think, a question on the minds
39:31 Ryan McEntush? of many people in this world. It's going to be very hard for, as I mentioned earlier, utilities to want to build large scale nuclear power plants. Generators are likely going to look at solar and winds that because they're low marginal cost production and very cheap, it's very easy to throw those on. And as batteries get really, really cheap, throwing those on as well. And maybe that does work out, but it does introduce a lot of complexity. You do have to figure out exactly how much extra wind should I build? How much extra solar should I build that then I have to store to prepare in case the wind goes down? Now, these are a lot of models you can do to try to figure out how that works. Or you could build an extra natural gas plant that was always going to be there or build a nuclear power plant that's always going to be online.
40:13 Ryan McEntush? Like I said, the power generators are not necessarily going to care about the reliability, but the grid does, all the citizens do. So you're seeing in like ERCOT, for example, building this emergency reserve market that sits sort of on top of their free market trading system.
40:28 Steph Smith? Right.
40:29 Ryan McEntush? And so where this ends up, I'm going to be looking very, very closely to ERCOT to see how much intermittent power can they have? How much batteries can they have? How decentralized can they be? Is it going to be 100%? Is it going to be 50? Is it going to completely fail? And we're going to have to figure out how do we build more large scale utility solar or whatever it may be. These are the questions that people are interested in. Personally, and I think as a firm, we're very interested in technology and solving these problems. As I've noted previously, and as historically very obvious, immense problems are solved by technology going after and basically reducing the complexity and making things a lot easier to get done. And so where are we able to build at the edge?
41:11 Ryan McEntush? We're not going to be competing with gigawatt scale power plants, but hundreds of homes with solar on their roof or or thousands of EVs might end up having this gigawatt battery storage that is distributed. And the new grid, the decentralized grid, is going to look very different than the old grid. And in many ways, it's going to be better and cheaper, more resilient. So I think that's the future we should be aiming for. I'm excited for it.
41:34 Steph Smith? Absolutely. And to your point, a lot of people are looking to ERCOT. At least we have a testing ground. We can see the results. And it's not to say that that is the final destination. I'm sure there will be many other experiments on the horizon, but I'm equally excited about how technology can be applied to this space.
41:51 Ryan McEntush? Likewise.
41:54 Steph Smith? All right, that is your call to build. Clearly, the grid is no monolith, and solutions can't stand alone. They need to be thoughtfully interwoven into the matrix of an already complex energy mix, data infrastructure, changing regulation, and more. And if you did enjoy this episode, be sure to check the links in our description to two of Ryan's articles that break down the grid and nuclear energy even more deeply. We'll see you next time.
42:21 Steph Smith? If you liked this episode, if you made it this far, help us grow the show. Share with a friend, or if you're feeling really ambitious, you can leave us a review at ratethispodcast.com/a60z. You know, candidly, producing a podcast can sometimes feel like you're just talking into a void. And so if you did like this episode, if you like any of our episodes, please let us know. We'll see you next time.
Previous Fixing Higher Education & New Startup Opportunities with Marc and Ben Next Text to Video: The Next Leap in AI Generation Page updated October 26, 2026 · transcript and summary are machine-generated
