Has anyone figured out how to reduce the energy demand for refining aluminum or manufacturing ammonium nitrate? Efficiency gains in things like lighting, heating, and ground transport are nice but those only go so far. Heavy industry will always require enormous energy inputs.
Smelting aluminum requires between 12,500 kWh and 15,000 kWh of electricity per ton of yield.
My lights use 8.5 Wh apiece.
The notion that consumer savings will make a dent in our problem has a few problems with decimal places. This is, by-and-large, a capital-driven problem. It's probably true that a reduction of individual consumption of stuff would be more beneficial than our energy use, too...but then that creates a new problem for that capital, too. So not only is avoiding the collective-action problem probably easier, but it's more effective, too.
Something I don't see acknowledged in discussions like these, is that industrial energy usage isn't something that is distributed equally, and never will be, and never should be, because some industries use more or less, and people and geographic areas specialize.
If you are going to run an aluminum smelter, you need a lot of power. An area where they don't have those, doesn't need anywhere near the amount of power per capita. But somebody has to produce aluminum if society is going to have it.
Reducing energy consumption: Cutting production because it's no longer economically viable with the increased energy prices, and maybe to a much smaller extent switching to more efficient methods
If you're only looking at one energy source: Switching to another. For example, gas was scarce and expensive while oil was less affected, so some factories switched from gas to oil (some systems can be switched relatively easily).
Well, shutting down the heavy industry certainly is a way to reduce energy consumption. And I guess this is even a preferred way for local politicians, even though this industry is surely running elsewhere afterwards...
Al processing already consumes vast amounts of energy, and the plants are already being used for intraday load balancing. There was a story recently about an incident in Germany where due to weather and strange market mechanisms, the grid was in danger or collapsing. The operator’s third line of defense (after gas peakers and hydroelectric storage, IIRC), was to remotely turn off the Al plant 600 km away, something they can contractually (and technically) do up to four times per day and six times per week for an hour each.
That plant uses 1 % of Germany’s electricity. It’s not the only such plant, and Hermany generally frowns on frivolous uses of Aluminum: cans are far less popular, for example, and will earn you dirty looks.
The Audis may make up for it I part, but in the US it should be possible to shift some non-trivial fraction of the electricity consumption by, say, doubling processing capabilities and running them seasonally. Al isn’t too expensive, and the plant is a small part of the costs anyway, with energy making up the bulk.
Of course there are countless opportunities to optimize consumption patterns once consuming devices all have data connections: heat and cooling, washing machines, dryers, dishwashers, roombas, notebooks... all these devices could shift their consumption on the hours-to-one-day scale and easily cushion even large variability in production. Once electric cars are ubiquitous, they‘ll make up the bulk of consumer electricity usage and they could even feed power back into the grid.
The only real problem with Al is, obviously, that we keep misreading it as AI. The confusion will drive some lesser Al insane, and either that or the singularity will cause a bloodbath, henceforth known as Al Gore.
Industrial consumption dwarfs the rest of us. We can fool around with 'energy theatre' turning off lights and walking to the market. But unless we can do 'big industry' more efficiently then it's all for show.
And then the things you are trying to get energy out of need to be used frequently enough to recoup the cost and energy demands of refining and building a more efficient device.
Yes, its astounding how much energy is used by industry. I was involved in a sad industrial energy debacle, at a job long ago.
We were making electric arc furnace controllers. An electric arc furnace is commonly used to recycle metal. Think of large pencil leads (size of your thigh) stuck through an insulated lid over a large vat of old bicycles, water softeners, bedframes etc - recycled metal. By creating an arc between the three phases of the electrodes and the pile of metal and controlling the current and length of the arc, considerable energy can be transferred to the metal as heat. Pretty efficient.
Out beta test site complained that they'd be running a shift and notice the arc was misbehaving every morning - failing to moderate the arc length properly. Resulting in most of the energy going into the electrodes and eroding them away. What was going on?
By having an engineer sit there on an overnight shift, we discovered that the desktop computer running the algorithm was going into 'power save' mode, stopping the app from responding to inputs and leaving the furnace electrodes stalled.
We estimated that this 30MW furnace wasted as much energy over the week, as 'power save' mode on PCs saved in all of America for a year.
Think about it. It would have been better if Microsoft had never invented Power Save mode. More electricity would have been saved that year.
That's how much industrial power matters, and how little your app matters.
reply