I’ve been a fan of Low-Tech Magazine’s solar-powered website since its inception. Low-Tech Magazine takes a unique approach to scaling and sustainability: it argues for downsizing, downtime, and lifestyle changes. Often, this involves combining modern advancements with past technology.
I’m quite fond of this style of problem-solving because it seems like an application of the suckless philosophy to meatspace to areas besides software.
Yeah, but that’s in Barcelona. Try that in Stockholm or Montreal and you don’t have enough solar radiation. Try that in Lagos and you may not have a good enough connection to serve pages reliably. It only works in a rather narrow range of places that have lots of sunlight and good internet connectivity at the same time.
Don’t get me wrong, I do think it’s cool and I admire their efforts.
To keep the website online during four days of low or no power production, we would need a 440 watt-hour lead-acid battery – the size of a car battery. We include this configuration to represent the conventional approach to off-grid solar power.
It takes a 50W Solar panel 6h45m to fully charge a 440Wh battery. I think even during cloudy winters a server admin will be able to collect enough energy to keep their site going for a few days. Also, given the “lifestyle” changes suggested by the website with regular downtime etc. I don’t see why a person wouldn’t be able to sustain a website in Stockholm or Montreal using solar power.
It’s not about clouds. Compare these charts for Stockholm and Barcelona. Barcelona gets at worst nine hours of sunshine, while Stockholm gets six. That’s not all though.
At high latitudes, the Sun doesn’t rise high over the horizon in winter. Even in an open field your panel will not get much direct sunlight.
Clouds make the situation even worse of course.
As someone who is rather sick of the cloudy and dark winters where I live… solar power is still feasible, and even usually economical here. Clouds and shorter days are variables in the “how much capacity do I need” question, which get offset by larger batteries and panels in not-necessarily-simple but also not terribly complicated ways. There’s also the possibility of having panel pointing hardware, which seems annoyingly uncommon for reasons I don’t fully understand. You never “don’t have enough sunlight”, it’s “enough solar capacity to keep going is not worth it”.
Naturally, at the extremes “enough solar capacity” gets pretty tricky. I don’t think anyone is going to want to buy enough batteries to run a solar-only webserver with 99% uptime in Tromsø any time soon. Stockholm is easy in comparison.
Edit: actually, I want to make a calculator to figure some of this stuff out now. Thanks!
It’s a really neat idea, and they’ve put a lot of effort into it, and I’ve come back to that site from time to time and it seems to work well. But I have to call them out:
318 kb website, of which 25 kb is content and style, 25-50 kb is useful figures (which may well be smaller if they were SVG’s and/or not dithered), and the rest of which is magazine-header fluff images.
If I have to think about whether a website is bloated, I just try to imagine what it would be like to download over the 33.6k dialup I had as a child.
While I agree with you that the site is somewhat bloated, I don’t see how that invalidates their point. At some point the savings of minifying the input data compared to the serving of that data is going to end up costing more power than it’s worth. Converting 50kb of images to SVG would help, but if the input files are already PNGs or JPGs you might spend more power doing the conversion than the server uses to serve the files for a year. Also, with the footprint of this server, does it matter that they want to have a fancy header image? It loaded quick enough for me…
You’re right, they are trying to reduce the size of the images but also processing power. See here what they did.
Yes, they’re not doing a perfect job there and they have a lot of feedback on the topic already. Still, there’s a website running of of a toaster. I find that really cool.
This project is wonderful and very well documented. Right down to their study on the effect of different size pairings between solar panel and battery.
Wonder how feasible augmenting the power with wind would be - might help uptime over here in the US PNW.
I’ve been a fan of Low-Tech Magazine’s solar-powered website since its inception. Low-Tech Magazine takes a unique approach to scaling and sustainability: it argues for downsizing, downtime, and lifestyle changes. Often, this involves combining modern advancements with past technology.
I’m quite fond of this style of problem-solving because it seems like an application of the suckless philosophy to
meatspaceto areas besides software.Yeah, but that’s in Barcelona. Try that in Stockholm or Montreal and you don’t have enough solar radiation. Try that in Lagos and you may not have a good enough connection to serve pages reliably. It only works in a rather narrow range of places that have lots of sunlight and good internet connectivity at the same time.
Don’t get me wrong, I do think it’s cool and I admire their efforts.
It takes a 50W Solar panel 6h45m to fully charge a 440Wh battery. I think even during cloudy winters a server admin will be able to collect enough energy to keep their site going for a few days. Also, given the “lifestyle” changes suggested by the website with regular downtime etc. I don’t see why a person wouldn’t be able to sustain a website in Stockholm or Montreal using solar power.
It’s not about clouds. Compare these charts for Stockholm and Barcelona. Barcelona gets at worst nine hours of sunshine, while Stockholm gets six. That’s not all though. At high latitudes, the Sun doesn’t rise high over the horizon in winter. Even in an open field your panel will not get much direct sunlight. Clouds make the situation even worse of course.
As someone who is rather sick of the cloudy and dark winters where I live… solar power is still feasible, and even usually economical here. Clouds and shorter days are variables in the “how much capacity do I need” question, which get offset by larger batteries and panels in not-necessarily-simple but also not terribly complicated ways. There’s also the possibility of having panel pointing hardware, which seems annoyingly uncommon for reasons I don’t fully understand. You never “don’t have enough sunlight”, it’s “enough solar capacity to keep going is not worth it”.
Naturally, at the extremes “enough solar capacity” gets pretty tricky. I don’t think anyone is going to want to buy enough batteries to run a solar-only webserver with 99% uptime in Tromsø any time soon. Stockholm is easy in comparison.
Edit: actually, I want to make a calculator to figure some of this stuff out now. Thanks!
It’s a really neat idea, and they’ve put a lot of effort into it, and I’ve come back to that site from time to time and it seems to work well. But I have to call them out:
If I have to think about whether a website is bloated, I just try to imagine what it would be like to download over the 33.6k dialup I had as a child.
While I agree with you that the site is somewhat bloated, I don’t see how that invalidates their point. At some point the savings of minifying the input data compared to the serving of that data is going to end up costing more power than it’s worth. Converting 50kb of images to SVG would help, but if the input files are already PNGs or JPGs you might spend more power doing the conversion than the server uses to serve the files for a year. Also, with the footprint of this server, does it matter that they want to have a fancy header image? It loaded quick enough for me…
You’re right, they are trying to reduce the size of the images but also processing power. See here what they did.
Yes, they’re not doing a perfect job there and they have a lot of feedback on the topic already. Still, there’s a website running of of a toaster. I find that really cool.
This project is wonderful and very well documented. Right down to their study on the effect of different size pairings between solar panel and battery.
Wonder how feasible augmenting the power with wind would be - might help uptime over here in the US PNW.
Low Tech Magazine has sent mixed messages about small wind turbines: in the past, they said “small windmills are a swindle”: https://www.lowtechmagazine.com/2009/04/small-windmills-test-results.html
However, recently, they suggested that building small wind turbines out of wood could negate many of their sustainability flaws: https://solar.lowtechmagazine.com/2019/06/small-wooden-wind-turbines.html
I suspect the answer depends on: