4 ms·
Less efficient but more reliable, like software. The future of the grid should be many small generators and energy storage installations spread throughout the c
by ok_dad 4y ago
Less efficient but more reliable, like software. The future of the grid should be many small generators and energy storage installations spread throughout the communities they service, not many large farms with requirements for huge transmission infrastructure. Spreading generation and storage will achieve more reliability than efficiency, sure, but when you're generating off solar or wind, who cares if it's efficient?
- luckylion 4y ago> when you're generating off solar or wind, who cares if it's efficient Whoever pays for the electricity, most likely. If you don't connect neighborhoods and each block or even house is an off-grid island, your reliability will suffer because you'll need large margins everywhere. If you do connect them, you have that "transmission infrastructure". Rooftop is great because it gives you a feeling of security and being self-sustaining, but if you want impact, it's probably much more efficient to have technicians install panels en masse at large installations than do one-off jobs on all those custom homes.
- jacquesm 4y agoThis is true, but with electric cars becoming more and more widespread local generation is having a measurable impact on daytime generation. It also offloads the grid because every KWh generated locally and consumed relatively locally doesn't need long haul transmission, as opposed to 'large installations'.
- ok_dad 4y agoYou can reduce transmission infra considerably, probably not eliminate it though. I am not advocating for islanding, I am advocating for something like the internet's smartly distributed network for the power grid. I do work in the industry, so while I'm not completely knowledgeable I do have knowledge that makes me think this is the way for the future. You're not going to eliminate all centralized generation, of course, especially for solar! However, smaller installations of solar, wind, batteries, and etc. are useful for reducing the transmission grid requirements and increasing reliability and distribution capabilities. If you do end up with transmission lines disconnected, it's a good idea to have some capability to island and generate some portion of the power locally. I say "who cares" about efficiency, because the efficiency gains on the technologies involved with distributed grids and the cost reductions over time will make it "not matter much" because it will be relative pennies compared to today's infrastructure. It will be worth it to overbuild the distributed infrastructure in order to achieve better reliability. Energy is a necessity today, and will be a literal life-saver when climate change causes massive shifts in the climate where people currently live and won't want to abandon. We have to make it more reliable than it is with today's centralized system. tl;dr: I am advocating for more distributed hardware + software that will make the grid more reliable in the face of disconnections and transmission infra failure, not distributing generation completely to the end-user thus requiring them to generate their own power, only a portion of it.
- jacquesm 4y agoThat makes very good sense, and I would be very happy to participate in such a scheme if it were on offer.
- hot_gril 4y agoThe efficiency matters cause of the cost and waste of panels and maintenance. I don't know exactly how expensive transmission is, but there's probably a reason wind/solar power plants are so big and centralized rather than being smaller and spread out closer to homes, which must still be far more efficient than homeowners doing the job. About reliability, the transmission network still needs to handle peak load during a cloudy week. At least A/C usage would be lower then. Speaking of software, crowd computing via widely distributed CPU-heavy jobs to home PCs is fading away. It's just so much more efficient to do that stuff in a datacenter, especially if there are specialized accelerators involved.
- ok_dad 4y agoOn the last note, most companies in the building management software industry these days run a computer at the edge so that if there is no connectivity they can still process the control signals on-site. It was common a few years ago for us to be doing it in the cloud, but today it's starting to cost a lot more due to the massive numbers of buildings installing BMSes, and therefore these same companies are shifting to "edge computing" for as much of the site control as possible. Only analytics that cut across customers or aggregations and data storage are done in the cloud, and even then we're all trying to push as much as we can to the edge. My company is specifically rebuilding our infra to push the control loop and everything else we can to the site edge computer because doing it in the cloud is not efficient for our cost structure, and we have a $1000 industrial PC sitting there already, so why not use it? As far as distribution goes, the goal is to not need long-distance transmission infra in the future, via a mix of several types of renewables, energy storage, and even adding some emergency carbon generation in the mix if required. Today we need the transmission infra, but the goal is to reduce that by an order of magnitude which will reduce costs and maintenance by at least the same amount.