The Electricity Network's Dilemma
How distributed generation and storage may change the economics of central electricity networks and create difficult feedback loops.
Model & Argument · Initial edition, November 2026
Electricity networks are expensive shared infrastructures. Their costs do not fall in direct proportion to the amount of electricity flowing through them. If households generate and store more power themselves, the economic relationship between the network and its customers changes.
This creates a possible network dilemma.
The feedback mechanism
Imagine a network recovering substantial fixed costs from charges linked to consumption. As household solar and batteries reduce purchases, remaining costs must still be met. Higher tariffs or fixed charges may then make local alternatives more attractive, further reducing network usage.
In extreme cases, this can become a self-reinforcing exit problem: those who can afford to disconnect do so, leaving a smaller and potentially less wealthy customer base carrying shared costs.
It is tempting to call this a death spiral, but that phrase can conceal important complexities. Network charging can be redesigned; distributed resources can provide network services; some industrial loads may grow; and most households still benefit greatly from grid reliability.
What actually matters
The key variable is not just the cost of panels or batteries. It is the total practical cost of independence, including seasonal reliability, backup, maintenance and changing tariffs.
If independence becomes cheaper and good enough for significant classes of users, then traditional pricing and investment assumptions may fail. That could produce political conflict over access, fairness and stranded assets.
The network's continued usefulness is not the same as its ability to charge customers under the old model.
Why it matters more broadly
The same pattern may appear in other incumbent infrastructures: systems with high fixed costs whose remaining users face rising charges as capable alternatives spread. See When Systems Unravel.
This is an empirical hypothesis. Research is required into tariff design, storage performance, network costs, real disconnection behaviour and the economics of maintaining universal service.