A FRIEND of mine passed along an article on Wednesday morning, a report by Politico from about two weeks ago that explained a sweeping proposal by the European Commission (EC) to cut household electricity use at peak hours, “leaving more power available for industry, transport and energy-hungry computers that perform advanced AI functions.” The key to this is a new law to be filed later this year “to accelerate the rollout of AI-powered smart meters, which give consumers ‘greater control over when they use electricity, allowing them to shift consumption to cheaper hours and lower their bills.’”
Cue the eye roll and muttered profanities in three different languages. This is not the week to come at me with this kind of pretentious garbage, especially not before I’ve even turned on the coffee pot.
Europe, along with most of the rest of the developed world, is rapidly trying to electrify everything, and is further challenged by the boom in artificial intelligence and its voracious demand for power, estimated to be about 2.5 percent of Europe’s total generating capacity now. That is expected to more than double over the next four years, according to the EC’s brief, and so the implication is that in order to meet it, some other segment of demand has to cut back, in this case households. But of course, the authorities do not really want to send the message that ordinary citizens will have to accept the degradation of their lifestyles to support the AI industry, so they style it as a cost-saving measure for consumers. The irony that this will be made possible by using “AI-powered” smart meters evidently escapes them.
If you are sitting there wondering, “What in the hell is an ‘AI-powered smart meter’?” congratulations, you are a normal human with a functioning brain. An “AI-powered smart meter” is, according to the definition even I had to look up, is a meter that “uses integrated artificial intelligence (AI) and machine learning algorithms to analyze electricity (or gas, or water) consumption in real time, transforming raw data into actionable insights rather than just passively recording usage.”
So apparently, “turn the light off when you’re not in the room” and “don’t stand there with the refrigerator door hanging open” and other energy-saving bon mots of that nature are no longer good enough, we need “actionable insights.” That, in all likelihood, requires several multiples of the amount of energy we’d otherwise save in order to process the data. We are truly living in history’s stupidest timeline.
The other thing that struck me about the EC proposal is that, on paper at least, it’s a bit of a fallacy. Europe is not actually facing an energy shortage. According to the Energy Information Agency and other sources, generation growth across the EU is expected to be between 30 and 37 percent between now and 2030, whereas demand growth, including AI data centers, is expected to be between 6 and 9 percent during the same time period. Almost all of the new generation capacity is renewable energy (RE), but to compensate for its intermittency, the EU is building battery storage along with it, so that there will still be a reasonable positive margin between supply and demand.
It was easy enough for me to dismiss most of this as typical European goofiness, but there are two factoids in the story worth taking seriously for the sake of the Philippine context. The first is the aforementioned push to build energy storage to complement RE generation sources, and the second is the concept of “time of use.” Both of these are apparently crazy ideas here, although there has been a recent hint that the Department of Energy (DOE) is beginning to catch on to the first one, though in an as-yet completely inadequate way.
Earlier this year, the DOE issued Department Circular (DC) 2026-02-0008 mandating that all large-scale — i.e., over 10 megawatts (MW) capacity — variable RE projects (solar and wind) include at least 20 percent of their capacity in some kind of energy storage system. It is assumed that these would be batteries, but the DC does not limit the requirement to that technology. Further, RE developers are “strongly encouraged” to utilize grid-supporting capabilities in their systems, such as grid-forming inverters.
I have yet to hear an explanation as to why the DOE has not mandated an equivalent capacity in energy storage for RE generation plants. The one suggestion that I have heard, that “we don’t want to discourage investors,” is an answer, but not an explanation, especially in light of domestic RE developers being willing to take the bull by the horns and do it themselves. The best example is the massive Terra Solar project straddling Bulacan and Nueva Ecija, which will have 3,500 MW of solar capacity and 4,500 MW of battery storage when it is completed. On the other end of the scale, the small solar developer that I occasionally advise has been writing battery energy storage systems into all its project proposals, even though none of them are large enough to fall under the DOE requirement.
Second, “time of use” is something that should be implemented here; it is a common rate scheme in other places, and only requires the use of normal “smart meters,” not the “AI-powered” ones and their “actionable insights.” The latest version digital meters Meralco and a few of the other larger distributors and electric cooperatives use may in fact already be capable of time of use recording, but I will have to check on that.
Time of use simply means that an electric consumer is billed different rates depending on what time of the day the electricity is being used; rates are lowest during midday when net demand is lowest (due to the availability of a lot of solar power), and highest during the evening peak of demand. This allows the consumer to moderate usage at expensive times to save costs; for the power system in general, it helps to ease the supply pinch when the evening peak begins, and would reduce the instances of yellow and red power alerts.
ben.kritz@manilatimes.net
Bluesky: @benkritz.bsky.social
Website: www.badmannersgunclub.com