Check out the episode:
You can find the shownotes through this link.
Are you interested in how to create an energy system in the 21st century?
Our debate today works with the article titled EnergyNet explained: Internetification of energy distribution from 2025, by Jonas Birgersson, Dr. Marc A. Weiss, Dr. Jimmy Chen, Dr. Daniel Kammen, Dr. Tomas Kåberger, Dr. Franklin Carrero-Martínez, Dr. Joakim Wernberg, Dr. Michael Menser, and Dr. Newsha K. Ajami.
This is a great preparation to our next interview with Jonas Brigersson in episode 480 talking about distributed energy systems and the EnergyNet itself.
Since we are investigating the future of cities, I thought it would be interesting to see how we can implement the thinking of internet to energy systems. This article introduces a groundbreaking blueprint for decentralising electricity distribution.
[intro music]
Welcome to today’s What is The Future For Cities podcast and its Research episode; my name is Fanni, and today we will introduce a research by summarising it. The episode really is just a short summary of the original investigation, and, in case it is interesting enough, I would encourage everyone to check out the whole documentation on the WTF4Cities website where you can find the shownotes and more information on supporting the podcast. This conversation was produced and generated with Gemini LM as two hosts dissecting the whole research.
[music]
Speaker 1: So imagine driving down a highway at, like, sixty miles an hour. Suddenly, your brake pedal just drops its connection to the wheels for half a second.
Speaker 2: Because the network is buffering, right?
Speaker 1: Exactly. Because it’s buffering. Physics doesn’t wait for a reboot. You crash, and for over a century, that unforgiving physical reality is how we’ve had to treat our electrical grid. It is the ultimate impatient physical force. Today, we’re looking at a fascinating paper called Energy Net Explained: Internetification of Energy Distribution.
Speaker 2: Which is a pretty wild concept, honestly.
Speaker 1: It proposes this radical translation of digital networking logic to physical electrical infrastructure. The core thesis is that we should just abandon our massive synchronized real-time power grid and replace it with a software-defined packet-switched architecture, and I’ll be arguing that this actually works. Energy Net uses local buffering through decentralized batteries and this software-controlled physical firewall called galvanic separation. By isolating faults locally, it makes the grid, frankly, infinitely more resilient than the brittle infrastructure we rely on today.
Speaker 2: Well, I’m taking the other side because treating electrical currents like delayable data packets dangerously underestimates the strict real-time thermodynamics of electricity. I’ll be arguing that this architecture risks severe localized failures because physical storage simply can’t always match digital routing commands the way these authors just sort of assume it will.
Speaker 1: So let’s start with what we’re actually replacing here, the traditional alternating current grid, or what the paper calls POGS, the plain old grid system.
Speaker 2: Great acronym, by the way.
Speaker 1: And POGS is a marvel, sure, but it was built for a one-way street. Centralized power plants pushing electricity down radial feeders to passive consumers. Today, generation is totally decentralized. We’ve got rooftop solar, EVs, home batteries. They’re all pushing power backward into a grid that was just never designed for dynamic bidirectional flows.
Speaker 2: No, I don’t disagree that POGS is straining. It absolutely is under the weight of the energy transition. But the internet analogy Energy Net uses, it’s intellectually seductive but physically problematic.
Speaker 1: How so?
Speaker 2: Because we have to separate digital metaphors from thermodynamic realities. Look at the internet. When a data packet hits a congested router, it sits in silicon memory. If the buffer is full, the packet is dropped. Poof. No harm done. The system requests a retransmission.
Speaker 1: Sure.
Speaker 2: But electricity isn’t data. When you drop a, quote, unquote, “packet of unrouted electricity,” it doesn’t quietly disappear. It generates rapid heat. It trips physical breakers. Or you get an immediate blackout. You just don’t have the luxury of dropping physical power.
Speaker 1: But that is exactly why Energy Net shifts us away from that strict real-time synchronous balancing. It moves to a near real-time digital coordination model. It uses local storage, like all those EVs and home batteries, as a data cache. It buffers the imbalances locally inside an ELAN, an energy local area network. Think of it like moving from a rigid, highly pressurized pipe system to a network of local reservoirs. In the rigid pipe, a sudden pressure spike bursts the main line, taking everyone down. In Energy Net, the local reservoirs absorb the shock and hold the energy until the wider system can handle it.
Speaker 2: Yeah, I mean, the reservoir analogy sounds lovely for water, but let’s talk about the physical limits of chemical batteries versus digital memory. Digital buffers in a telecom router are virtually limitless relative to the size of a data packet, and they operate instantaneously.
Speaker 1: Right.
Speaker 2: Chemical batteries are heavily constrained. What happens when an ELAN, say, a local neighbourhood microgrid, experiences a massive unexpected load, and let’s say the local batteries happen to be fully charged from a sunny day?
Speaker 1: The software would negotiate-
Speaker 2: But wait, where does that physical heat actually go if the software says, “Don’t send it to the main grid”? The excess energy cannot simply wait in the wire like a queued email.
Speaker 1: I mean, no, it doesn’t wait in the wire.
Speaker 2: Exactly. The thermodynamic reality demands an immediate physical outlet. In the legacy grid, that surge is absorbed by the sheer size of the interconnected continent. It dissipates across thousands of miles of infrastructure. If you decouple the system and the local battery is full, you are trapped. You have nowhere to dump the heat. You melt wires or trigger a localized blackout to save the hardware.
Speaker 1: Okay, hear you. But that vulnerability is precisely why the paper introduces galvanic separation. It’s not just a software firewall. Let me explain how this actually works. In digital networking, a gateway uses network address translation to protect a local network. Energy Net does this for power, but physically. Galvanic separation literally breaks the physical wire.
Speaker 2: The two pools analogy.
Speaker 1: Exactly. It’s like two separate pools of water. Instead of a pipe connecting them where a massive wave could flood from one to the other, you use a water wheel. The energy transfers across, but the pools never actually touch. Power only crosses to the macrogrid when explicitly digitally negotiated via the energy protocol. A surge in one pool cannot physically cross over to destroy the other.
Speaker 2: The water wheel is a brilliant way to protect the macrogrid. Sure, the wide area network, the EWAN, but what does that mean for the neighbourhood?
Speaker 1: It means the neighbourhood doesn’t take down the whole city.
Speaker 2: Yeah. But in POGS, the grid is a massive continent-wide shock absorber. Heavy spinning turbines at centralized plants provide immense physical inertia. A tree falls, causes a short, those massive turbines literally muscle through the frequency drop to keep the system stable. By galvanically separating the local microgrid, you completely sever it from that continental shock absorber.
Speaker 1: You’re localizing the problem, which is a good thing
Speaker 2: You are localizing it, yes, but you’re forcing the neighbourhood to handle its own volatility entirely alone. If that software negotiation fails or if local storage is inadequate for even a millisecond, the localized system is trapped with its own surge. It shuts down. You’ve protected the wider grid, but you’ve made the local neighbourhood incredibly fragile to its own internal physics.
Speaker 1: I see the risk, but I’d argue this is where EnergyNet proves operationally superior even during a failure. Let’s talk about what a failure actually looks like on the traditional grid. It’s binary. You have one hundred percent power or you experience a zero percent blackout.
Speaker 2: Generally, yes. When the macro grid fails, your house goes dark. There’s no middle ground. EnergyNet fundamentally changes this through the Energy Router Operating System, EROS.
Speaker 1: Which, again, is just software trying to manage physics.
Speaker 2: It is, but look at how it manages it. EROS controls an internal direct current backplane. Think of it like a highly intelligent internal power strip inside the router. All local devices, solar, batteries, they pool their direct current there before it’s routed out to the house. During a crisis, EROS doesn’t just fail. It communicates via the energy protocol to reallocate limited power dynamically.
Speaker 1: Okay, but in a real crisis, it can tell the heavy-drawn HVAC system to shut down while keeping medical devices or emergency lighting online. It prioritizes some power over a total blackout.
Speaker 2: Prioritizing critical loads is a great concept in theory, but it replaces physical certainty with software-defined rules. The traditional grid relies on hardwired synchronous coupling. The physics guarantee the connection. EnergyNet relies on the energy protocol, the EP. In an extreme thermal crisis, the environment is chaotic. Voltage fluctuates wildly. If the EP server logic misfires or local sensors lag by even a few milliseconds, how does the system know what to prioritize?
Speaker 1: Because the hardware is built with massive redundancy. We aren’t talking about a single cheap computer chip. The routers use dual redundant energy supervisors in active-passive failover modes.
Speaker 2: Meaning?
Speaker 1: Meaning one supervisor actively runs the logic, and the second silently shadows it, ready to take over in milliseconds if the primary fails. It’s designed precisely to prevent those single points of failure.
Speaker 2: I just don’t buy that active-passive failover is enough during a violent electrical event. You’re trading a known physical limitation for an unpredictable software limitation. When voltage goes haywire, software crashes. DEP server logic could fail entirely, leaving the local network just as dark as the legacy grid, but with way more complex proprietary failure points for a technician to diagnose.
Speaker 1: But we manage software limitations all the time in telecom. The paper outlines the energy network management system, which provides real-time diagnostics on temperature and voltage across the fleet. If one physical port in a rack overheats, the backplane automatically routes energy through an alternative port. This distributed redundancy is exactly how the internet achieved its resilience Look at the Baltic Sea cable cuts in 2024. The internet simply routed around the physical damage with zero significant packet loss. The legacy power grid cannot do that. Rigid architecture is why targeted disruptions cause catastrophic cascading national blackouts. Energy net’s modularity isolates damage and maintains function.
Speaker 2: But the Baltic Sea example perfectly illustrates my point about data versus thermodynamics. Data routed around those severed cables seamlessly because data is massless. A router in journey simply told a packet to take a slightly longer path. It added maybe 20 milliseconds of latency. Nobody even noticed. But if a major transmission line is severed in an electrical grid, the megawatt hours of physical energy flowing through it cannot simply be asked to take a detour through a smaller local microgrid without melting the infrastructure.
Speaker 1: Which is why it doesn’t route it through a smaller wire. It buffers it locally.
Speaker 2: But physical copper limits are absolute. The energy router’s internal DC back plane is phenomenal engineering, but it has strict thermal limits. Software redundancy is not physical capacity. If the remaining physical pathways don’t have the heavy copper capacity to carry the load, all the intelligent routing in the world will not save you. You cannot software update your way out of a melting wire.
Speaker 1: I completely acknowledge that physical limits of copper and thermal constraints exist, obviously. But that is exactly why Energy Net focuses on balancing supply and demand at the most local level possible. Look at the real-world examples they bring up, like the Tammarinden neighbourhood in Sweden. By keeping the generation physically close to the storage and consumption, you dramatically reduce the need to push those massive megawatt hours across long distances in the first place. The protocol actively avoids the scenario where huge amounts of power need to be suddenly rerouted across the country.
Speaker 2: And when the sun is shining and local batteries are perfectly cycled, the system is beautifully elegant. My concern remains focused on the edge cases, the severe weather events, a sudden unseasonal cold snap where everyone’s heat pumps kick on simultaneously. The local batteries drain in an hour, and that galvanic separation prevents the massive inertia of the larger grid from riding to the rescue.
Speaker 1: But in that exact cold snap scenario under POGS, a single failing substation takes down the entire city Energy net isolates the failure. A neighbourhood might have to aggressively shed non-critical loads, sure, shutting down EV chargers to keep the heat on, but that neighbourhood survives the night. Under the old system, the whole city freezes. We are trading a brittle, centralized architecture that fails catastrophically for a flexible one that degrades gracefully.
Speaker 2: Or, as I would argue, we are stripping away the massive physical safety net of the synchronous grid and relying on the hope that small-scale local storage will always perfectly match digital routing commands.
Speaker 1: Even if we disagree on the specific implementation risks, it seems we both recognize the underlying reality. The legacy one-way alternating current grid is reaching its absolute limit. It was a miracle of the twentieth century, but it simply cannot scale to manage millions of bidirectional electric vehicles and decentralized solar arrays. A structural shift is inevitable.
Speaker 2: Absolutely. The era of passive consumption is over. The traditional grid is straining under the weight of technologies it was never designed to support. The transition to active decentralized generation demands a new paradigm, without question.
Speaker 1: This is why deeply examining the tension between digital network theories and physical thermodynamic realities is so essential. As we attempt to modernize our most critical infrastructure, understanding both the immense promise of software-defined routing and the unforgiving nature of physical electricity ensures safer, more robust innovations. We’re gonna leave it there for today. We encourage you to read the material and draw your own conclusions because ultimately, whether we’re waiting for a digital file to transfer or trusting our car’s brakes at sixty miles an hour, the architecture underneath matters.
[music]
What is the future for cities podcast?
Episode and transcript generated with Descript assistance (affiliate link).
Leave a Reply