Quote from Graeme Weston on 26 January 2026, 1:46 pm

Electricity in New Zealand is not expensive because energy is scarce. It is expensive because the system is organised around an outdated assumption that no longer matches reality. We still design and price electricity as if everyone might need maximum power at the same moment, all the time. That assumption drives oversized poles and wires infrastructure, high fixed charges, and endless justification for network upgrades.

Telecommunications once worked the same way. Before the internet, phone networks were rigid and circuit-switched. A dedicated path had to exist before a call could be made - the single line all the way to London was expensive, charged by the minute. Capacity was reserved whether it was used or not. The entire system was sized for worst-case simultaneity, and consumers paid accordingly. Then packet switching arrived and quietly overturned the logic. Instead of reserving single connected line, information was broken into packets, sent only when space was available, and routed dynamically around congestion. The same copper carried vastly more traffic, costs collapsed, and reliability improved. Nothing changed about the wires. What changed was how they were used.

Electricity has not made that transition, yet. Despite every appliance becoming electronic, despite the rise of solar panels, batteries, and electric vehicles, the electricity system still behaves like a circuit-switched network. Capacity is permanently reserved. Peaks are assumed rather than managed. Intelligence is centralised. Consumers are charged as if the network is always on the brink of overload, even when it is not. This is why fixed daily charges keep rising even as technology gets cheaper, why rooftop solar is treated as a nuisance, and why electric vehicles are framed as a threat instead of an asset.

This is where Jonas Birgersson enters the story. Birgersson was instrumental in proving that the internet worked, not as theory but as communications infrastructure. He has since applied the same thinking to energy, and the result is uncomfortable for incumbents because it does not rely on speculative breakthroughs. It uses off-the-shelf technology already manufactured at scale for electric vehicle charging. It uses DC power electronics, software control, and local buffering to show that electricity, like data, does not need to be delivered at peak power continuously to provide the same service.

The phrase “packets of electricity” tends to provoke reflexive dismissal, so it is worth being precise. No one is boxing electrons. What is being managed is power over time. Households and businesses do not need unlimited instantaneous power; they need energy delivered within tolerable windows. Most loads can wait milliseconds or minutes without anyone noticing. By controlling power in short bursts, smoothing demand, and using batteries and EVs as buffers, peak demand can be collapsed while total energy use remains unchanged. The shower still runs. The meal still cooks. The difference is that the network no longer has to be built for a fantasy moment when everything happens at once.

This only works because modern power electronics operate natively in DC. Solar panels are DC. Batteries are DC. EVs are DC. Even most “AC” appliances convert power to DC internally before doing anything useful. DC allows precise current control, instant limiting, and prioritisation in ways that AC, constrained by waveforms and synchronisation, simply cannot. In this model, AC becomes a compatibility layer for the legacy grid, while DC becomes the control plane where intelligence lives.

The practical embodiment of this idea is the energy router. An energy router does for electricity what a broadband router did for data. It enforces a hard cap at the grid connection, uses local solar, batteries, and EVs to cover short bursts above that cap, and presents the network with a predictable, well-behaved load. From the grid’s perspective, the home or business becomes boring. From the consumer’s perspective, nothing feels constrained, except the bill. Capacity appears to increase without building new poles or wires, because the existing ones are no longer being abused by unmanaged peaks.

This is where the discomfort sets in. Electricity distribution (poles and wires) companies are paid under a regulatory model that rewards capital accumulation. Build more assets, earn a regulated return. Energy internetification does the opposite. It avoids new builds, increases utilisation of existing assets, and exposes the marginal cost of delivery. This is not a technical threat to networks. It is a threat to a revenue model that has grown accustomed to equating reliability with perpetual expansion.

New Zealand is, quietly, in an ideal position to make this shift. Rooftop solar penetration is still low. The network is largely overbuilt relative to average demand. EV adoption is rising. The risks of transition are manageable, and the upside is large. Australia shares the same electrical standards, doubling the effective market for router manufacturers willing to think ahead. What is missing is not technology, but permission: permission to cap demand intelligently, to price delivery based on distance and congestion rather than fixed charges, and to recognise energy routers as infrastructure rather than appliances.

If electricity were invented today, no engineer would design the system we currently have. We already know how to share infrastructure efficiently. The internet taught us. Jonas Birgersson has shown that the same principles apply to energy using technology already on the shelf. The barrier is no longer technical. It is institutional, political, and economic. And every year we delay, consumers continue to pay for a system built for a peak that almost never happens.

References

Internet-inspired power distribution and sharing system starts in Sweden https://www.enlit.world/library/internet-inspired-power-distribution-and-sharing-system-starts-in-sweden

EnergyNet Taskforce https://www.energynettaskforce.org/

How POTS worked

Volts Podcast interview, Jonas Birgersson of EnergyNet explains it here.