I have a project I'd like to attempt, and I wanted to see if I'm thinking about things right.
Background: It's taking a long time (60 seconds+) for some of my faucets to produce hot water. After some trial and error and thinking about how stuff works, I'm 99% sure the problem is the size/length of the hot water line combined with the flow rate of the faucets. New faucets seem to limit the flow rate to like 1 gallon/minute (gpm) based on my empirical tests. The hot water line is approx 30 feet of 1", followed by 1/2" to the faucets (one run is 10' and another is 40'). By my math, it's the 1" pipe that is killing me, as I think it takes ~30 seconds to flush alone. After doing some research, the standard solutions seems to be either adding an instant hot water heater at the faucet, or running a recirculation line from the faucet back to the hot water tank. I don't like either. In the former case, I don't have the room for even the smallest instant hot unit. In the latter, it doesn't seem very green (plus it will be difficult to do).
My "arduino powered" solution: I want to put a "smart solenoid value" at each faucet in a "T", one end of the T going to the faucet and the other connected to the drain directly. The drain connection is the part that opens/closes of course, and it opens based on this general piece of logic:
((flow is on) && (temperature < hot))
Yes, this "wastes water", but I'm already dumping the water into the drain myself waiting for it to get hot. But my manual ability to flush the line using the faucet happens much much much more slowly (*)
Make sense? Any recommendations? I see my project being rather similar to: http://labs.teague.com/?p=722
Though in terms of actual supplies, I was thinking of things more like:
(*) I did a test, and even with the 1/2" line I can get just under 6 gpm through the pipe itself, which is a lot faster than 1 (6x to be exact).
I don't see how a recirculation pump is less environmentally friendly than dumping water down the drain.
Put insulation on your hot water pipes. It makes a big difference.
Recirc pumps are typically low power. They can be on a timer. There can be a trigger, so when you want hot water you push a button and it will start running. Of course it still need a minute or so, but push the button when you go in and when you're ready to wash your hands it will be there.
I don't see anything technically wrong with your proposal. Just make sure there are no leaks, the valve is normally closed and there is kill switch.
You have not said how you get the hot water. That is critical. If you have a tank of hot water that is already hot then how fast it gets to you is down to flowrate. If you have a boiler that heats the water when there is a demand then that is a completely different problem, and in fact a slow flowrate (at least initially) may be your friend.
Imagine you have a boiler full of cold water. You open your faucet as far as it will go wanting hot water. The boiler switches on and tries to heat the water to the desired temperature but there is a huge flow of cold water coming into it and it struggles to raise the temperature.
What you really need is a valve near the boiler that only opens fully when the water is warm. So: you turn on your faucet, boiler heats water (only a little flows out), water get warm, valve opens hot water flows down pipe, boiler manages to keep inflow upto temperature. A radiator valve on a vehicle gives the general idea.
To do it manually turn the faucet on slightly so that the flow is enough to start the boiler, wait until you get hot water, then turn up the flow. If you time this I am sure you will find you get hot water faster but wasting less water and power.
Thanks for the feedback! There were a few points brought up I thought I'd address.
Recirc pump vs. "drain dump": I probably shouldn't have mentioned the loaded "green" word. My overall thought is that right now I turn on the hot water and stand there for 60 seconds waiting for it to get hot. Based on the arduino system & high level logic I described I'd wait a lot less. The amount of water down the drain remains the same (as the valve would shut once the line is hot). As for solving the problem with timers/buttons, I personally feel each works against the normal way of using a faucet (e.g. turning it on). Timers only work at certain times of the day. And a new button my the sink seems confusing to me (and will annoy my wife), and my guests will have a poor hot water experience not knowing what it does (and/or I'll have to post instructions). I'm not saying I don't appreciate the tips! I just still like my solution better
Style of hot water tank: I do in fact have a tankless heater. And they seem to get a lot of blame for having slow hot water response. But I think that's a (partial) red herring. Based on a ton of reading, and trial and error tests, I'm pretty sure the tankless unit adds a fixed overhead of about 10 seconds of cold->hot. I agree the "vary the flow" rate trick would work to produce hotter water faster by giving the water more time in the heating coils (*) while the coils are still getting hot. But, that will only cut down on the 10 seconds it takes the coils to get hot, and I still have that 50 seconds of pipe clearing.
(*) I haven't taken the tankless apart, yet, but I assume it uses something like a coil of copper tube that heats the water as it flows through.
Don't get me wrong - I completely understand why you want to do your own (I was thinking about building one myself as well..). Off the shelf means most likely expensive, and obviously not as fun, just wanted to point out they exist.
They recirculate back into the water heater and usually have a sensor so that it shuts off once hot water reaches the farthest point you need it at. No real reason you couldn't time how long it takes and assume that it's always completely cold to skip the temp sensor though.
What I don't understand is why they don't run really small diameter, extra-hot water lines, and a tempering valve at each location to mix it with cold at the far end of the system - that way you get hot water almost instantly because of the size of the line, and since you're storing and transporting a smaller volume, you can insulate it much better for the same price. Smaller tank = less surface area, same amount of insulation to offset the higher temperature differential. Only problem is as the temperature gets lower, the tempering valve would make you consume hot water at a faster rate in an exponential curve, and the temp would just completely drop-off after a certain point. Also not sure about the cost of the tempering valves. Just an idea..
