Washing machines, software and the solar sponge
Free midday solar power renders decades of energy-efficiency engineering obsolete, making cheap, dumb appliances the most economically rational choice.
By Hugh Askell
Sparked by Australian energy retailers must offer three hours of free daytime electricity · discussion
If you look at the back of your refrigerator or washing machine, you will see a sticker boasting about its energy efficiency. For thirty years, we have optimised the entire global appliance industry for this exact metric, training consumers to pay a high premium upfront to save a few pennies a day on their utility bill. But look at a recent mandate in Australia for free daytime electricity, or the accompanying Hacker News debate, and you realise we might now be measuring the completely wrong variable.
We have arrived at the undeniable moment when distributed solar reaches the top of its S-curve. In doing so, it breaks the fundamental physics of the grid by removing the cost constraint entirely at lunchtime.
Since the EPA introduced Energy Star in 1992, the foundational assumption has been that the total volume of energy consumed matters most. We have spent three decades engineering for scarcity—layering complex variable-speed compressors, intricate firmware, and delicate circuitry into domestic boxes merely to shave tiny percentages off an annual kilowatt-hour draw. Silicon Valley, naturally, tends to look at the current energy transition and assume it is a software problem. The default tech-industry view expects that we will balance the chaotic new renewables grid with elegant AI, smart algorithms, and connected homes dynamically trading electrons on a micro-market.
But physics moves faster.
The Australian Energy Market Operator recently reported that wholesale spot prices fell below zero during 19% of dispatch intervals. When power is so wildly abundant that generators are financially penalised for pumping it into the wires, micro-optimising a compressor motor becomes an exercise in missing the point. The relevant framework shifts from a test of total-use to a test of time-of-use.
When a foundational physical constraint vanishes, it creates unexpected ripples in entirely adjacent, profoundly boring industries. Think of how the rise of autonomous cars might wipe out the impulse-tobacco industry because nobody needs to visit a gas station, or how 24/7 robotic trucking alters the real estate geography of rural motels. Here, the casualty is the economic logic of premium consumer hardware.
Imagine a simple two-axis chart comparing the combined capital and operating cost of two domestic machines over a ten-year lifespan. On the top left side, you have a $4,000 hyper-efficient, internet-connected heat pump running overnight on standard grid tariffs. On the bottom right, you have an $800 dumb, incredibly inefficient resistive water heater turned on mechanically at 1 PM. Operating as a domestic 'solar sponge', this cheap, terrible device happily absorbs 15kWh of thermal energy using free midday power.
The brute-force inefficiency of the resistive heater actually makes it structurally superior. It acts as a massive thermal battery, rendering the micro-optimisation of the heat pump economically worthless. You do not need a machine-learning algorithm to parse dynamic pricing if the energy is reliably free for four hours a day; you just need a really big, stupid tank of water connected to a basic mechanical timer.
This breaks a core hardware assumption: companies are trying to sell a complex software solution to a problem that sheer physical oversupply has already solved. Big capital is already reacting to this structural shift. Australia has rapidly emerged as the third-largest grid-scale battery market in the world, underscoring that the new physics of energy are entirely about time, not volume. The 30-year economic architecture is moving from rationing a trickle of expensive power to gorging on a brief daily flood, and the industrial supply chain for domestic hardware will have to reorganize itself entirely around that flood.
So, we have a new physics of the grid where the dumbest possible appliance is the most economically rational choice, provided you simply turn it on at lunchtime. Will that logic scale globally as the rest of the world hits their own solar S-curves, or will we simply fill the sponge and require a completely new layer of hardware? The structural economics have clearly inverted, but mapping out exactly what sits in our basements a decade from now remains entirely unpredictable.