the second leg of indoor climate control
Something I've always done is try to keep houses - my house - as cool as possible in summer with just ventilation. Coming from Seattle, nobody used to have AC, and that included us, so if you've got the skills why not keep using them? Saves energy and money. Everybody wins.
There's nothing mysterious about the system. You cool the house down as far as possible overnight with evening ventilation, then close the place up during the day during the heat. Do it right, and you can keep it cooler inside than out. Or, if you have AC, you can run it less - a lot less, if you do things right.
It works better if your house is situated well. And it works even better than that with indoor/outdoor thermometers, so you know when to open and close windows in what parts of the house. And yes, it's different, and different enough to really matter. If it's cooler indoors than out, close the windows. If it's cooler outdoors than in, open the windows.
I've always had a little collection of these indoor/outdoor thermometers, no two alike. So I thought it'd be good to standardise a bit. So I bought a more capable setup, a central unit with eight sensors all displayed on a little panel, with an accessory to let me see it via my phone on the web.
(I made a completely new network for this. Two, if you want to be technical about it. Terrible Slower Things (2.4Ghz), Terrible Faster Things (5.0Ghz), and it's not just separate wifi, it's a truly separate network with different root node to our real LAN, because fuck the Internet of Things. But I digress.)
I've got the sensors set up in pairs, which gives me four indoor/outdoor comparison points, and with all the graphing software, I've learned a few useful things already.
One is that the east side of the building gets even hotter even sooner than I realised. It's really just a matter of scale, not type, but good to know. A second is that the west side is indeed much cooler and stays much cooler thanks to the hill, but I was pretty sure the old sensor placement was exaggerating exactly how much, and I was right. Fortunately, it's not dramatic - just a couple of degrees, no big deal.
But the big surprise is how quickly the north side of the building heats up. I was thinking it was more like the west side than the east, and wow, that was completely wrong. It's not wildly over-temperature like the east, but it's no coolness buffer like the west either.
That boils down to, "I have been completely wrong about keeping that north window open all the time." It needs to be closed, like, in the morning. It's great overnight in that we can keep it fully open without worrying - it's screened and well above the ground, no animal or break-in issues - but it needs to be closed at like 9 or 10am. Otherwise you're pulling in warmer air than is already inside the building.
We're seeing a meaningful difference. With AC off, it's cooler inside than out on three sides of the building literally all day. And the west side, we can continue to use correctly as a coolness source/heat sink. I've been able to keep the house cool this way before, but getting the north side wrong has really been hampering my effectiveness. Now I can just glance at a panel and know for sure what needs to be flipped, and how.
It's funny, when the new kit arrived I was briefly all, "oh, why the hell did I do this, I've already got most of this data (via three separate indoor/outdoor setups) and this really only adds one more and it's a waste of money" but then I set it up anyway and... yeah, that was wrong.
And it's made no-AC temperature management meaningfully more effective. Energy not used is energy not paid for, after all. Between the new HVAC system using less energy and this sensor array's new data, we may do a lot better than I'd even hoped. This hardware actually should pay for itself in a few years!
Here's the sensor array as viewed from the web. The graphing is pretty nice, but I don't think you can use that functionality without an account. Senstors 1-4 are the outdoor set (east, west lower, west upper, north), 5-8 are the matching indoor sensors. If you set your browser's zoom correctly, they pair up nicely in the display.
So. Yeah. That's the other component of the big house cooling project.
The last component is a thermostat-controlled exhaust fan in the server room. Everything else we've done may have made it irrelevant - it's a lot cooler in there this summer than it ever was before, even on crisis temperature days! But I'll probably still get that done too.
Maybe next summer.
Also posted to ソ-ラ-バ-ド-のおん;
comments at Dreamwidth. Please comment there.
There's nothing mysterious about the system. You cool the house down as far as possible overnight with evening ventilation, then close the place up during the day during the heat. Do it right, and you can keep it cooler inside than out. Or, if you have AC, you can run it less - a lot less, if you do things right.
It works better if your house is situated well. And it works even better than that with indoor/outdoor thermometers, so you know when to open and close windows in what parts of the house. And yes, it's different, and different enough to really matter. If it's cooler indoors than out, close the windows. If it's cooler outdoors than in, open the windows.
I've always had a little collection of these indoor/outdoor thermometers, no two alike. So I thought it'd be good to standardise a bit. So I bought a more capable setup, a central unit with eight sensors all displayed on a little panel, with an accessory to let me see it via my phone on the web.
(I made a completely new network for this. Two, if you want to be technical about it. Terrible Slower Things (2.4Ghz), Terrible Faster Things (5.0Ghz), and it's not just separate wifi, it's a truly separate network with different root node to our real LAN, because fuck the Internet of Things. But I digress.)
I've got the sensors set up in pairs, which gives me four indoor/outdoor comparison points, and with all the graphing software, I've learned a few useful things already.
One is that the east side of the building gets even hotter even sooner than I realised. It's really just a matter of scale, not type, but good to know. A second is that the west side is indeed much cooler and stays much cooler thanks to the hill, but I was pretty sure the old sensor placement was exaggerating exactly how much, and I was right. Fortunately, it's not dramatic - just a couple of degrees, no big deal.
But the big surprise is how quickly the north side of the building heats up. I was thinking it was more like the west side than the east, and wow, that was completely wrong. It's not wildly over-temperature like the east, but it's no coolness buffer like the west either.
That boils down to, "I have been completely wrong about keeping that north window open all the time." It needs to be closed, like, in the morning. It's great overnight in that we can keep it fully open without worrying - it's screened and well above the ground, no animal or break-in issues - but it needs to be closed at like 9 or 10am. Otherwise you're pulling in warmer air than is already inside the building.
We're seeing a meaningful difference. With AC off, it's cooler inside than out on three sides of the building literally all day. And the west side, we can continue to use correctly as a coolness source/heat sink. I've been able to keep the house cool this way before, but getting the north side wrong has really been hampering my effectiveness. Now I can just glance at a panel and know for sure what needs to be flipped, and how.
It's funny, when the new kit arrived I was briefly all, "oh, why the hell did I do this, I've already got most of this data (via three separate indoor/outdoor setups) and this really only adds one more and it's a waste of money" but then I set it up anyway and... yeah, that was wrong.
And it's made no-AC temperature management meaningfully more effective. Energy not used is energy not paid for, after all. Between the new HVAC system using less energy and this sensor array's new data, we may do a lot better than I'd even hoped. This hardware actually should pay for itself in a few years!
Here's the sensor array as viewed from the web. The graphing is pretty nice, but I don't think you can use that functionality without an account. Senstors 1-4 are the outdoor set (east, west lower, west upper, north), 5-8 are the matching indoor sensors. If you set your browser's zoom correctly, they pair up nicely in the display.
So. Yeah. That's the other component of the big house cooling project.
The last component is a thermostat-controlled exhaust fan in the server room. Everything else we've done may have made it irrelevant - it's a lot cooler in there this summer than it ever was before, even on crisis temperature days! But I'll probably still get that done too.
Maybe next summer.
Also posted to ソ-ラ-バ-ド-のおん;
comments at Dreamwidth. Please comment there.