Tuesday, April 28, 2015

Kitchen hot water

A while ago, I wrote about my reasons for wanting a dedicated hot water system for the kitchen. Today, the system is finally operational. It's a big day!

I decided I wanted to do all the plumbing myself, which has made things take a bit longer (I have no previous experience brazing or bending copper pipe), but it has gone reasonably smoothly (albeit slowly), and in the end I'm reasonably happy with the standard of work. I left the most visible parts until last, when I'd had most practice and they look reasonably professional.

Here is a picture of the little hot water system on the roof (with the unfinished space heater next to it). It holds 30L of water, and has 8 x 50cm evacuated tubes. It is a low pressure system, so I can't just connect it to mains pressure with a tap in the kitchen (as is the norm). This system cost me $300 on ebay. It came with the aluminium racking immediately underneath it, but all the rest of the aluminium on the roof I installed myself.

Because the unit can't run at mains pressure, I needed a system whereby the tap to it controls its supply (before that supply reaches the tank). Then the tank is just sitting there at atmospheric pressure. To do this, I used an electric solenoid valve, which cost about $20 on ebay, with a 12 V transformer that I installed in the roof. We had an existing powerpoint in the roof that is controlled by a switch in the kitchen. That switch now energises the solonoid valve, switching it from normally-closed to open and allowing more water to flow into the tank of the hot water system.
Because the tank is always basically full, this causes it to overflow. The overflowing water exits the tank via the (hard-to-see) pipe at the top-left in the first image, and that pipe leads straight down to a (tapless) faucet in the kitchen below. Here is the last bit of pipework that directs that overflow through the kitchen window and out the spout in the kitchen.

How to use it

In the kitchen, flick the power switch, and wait. That switch activates the solonoid valve, opening it. Water enters the tank and causes it to overflow, and come out of the kitchen faucet (separate to the existing tapware). A bit before we have enough water, flick the switch back off -- the rest of the water will drain from the pipe into the sink before it stops.


Yes, this is quite an unusual system.

The main quirk is that, because the "overflow" pipe is basically empty of water most of the time, there is quite a latency between flicking the switch and getting water out of the tap. There is also quite a latency between un-flicking the switch and the water stopping. There's more testing to go here, but I think it's about 20 seconds.

Also, because the faucet is open to the tank, it is possible that the tank will boil in summer and lead to dripping in the kitchen. This remains to be seen.

Reflection on goals

The goal was to have "instant solar hot water" available at the flick of a switch. I now think this was a bit naive, but I still think it will be handy for two main reasons:
  1. Because it is a shorter run of hot water (it is basically hot immediately with no wastage), it decreases the load on the main solar HWS during winter. 
  2.  For the same reason, less water is wasted in summer (when we really need it!) -- we've been catching that water in jugs and drinking it later, but it is a bit of a pain, and still results in some wastage (from the main HWS, there is just over 2 L wastage to get hot water in the kitchen)
If I was doing this again, I'd consider installing an additional full-size solar HWS down that end of the house and just run the kitchen from that. One day, I may do that (depending how we go for hot water in winter).


  1. Hi Angus. Blogger is hungry it just ate my comment.
    Nice work and very good value too at only $300 for the evacuated tubes and header tank. Wow, that is good value. How much heat is it producing at this time of year? The solar hot water panels here don't produce much heat at all over winter. Mind you it has been very cloudy. How are you going for cloud? Hey what sort of flow rates do you get from that system?
    PS: A bit of lagging over the copper pipe may help improve the winter effeciency a little bit (not too much though)

  2. Hi Chris,
    Thanks, yeah I'm pretty happy with it. Having said that, if I did it again, I'd probably go the whole hog and just put a second 240 L solar HWS there (which is what I have at the other end of the house) and know that I always have plenty of hot water. It's was a bit of pain though making sure the roof was strong enough to bear the weight, and I also wanted to increase the tilt of the other system which made it harder -- by comparison 30 L is just so easy! ;-)

    It seems to be performing pretty well. We had about 1 hour of clear sunny sky yesterday, and the water came out at about 40 degrees (guesstimate only). On the day I installed it, the water was too hot to touch, and it looks sunny again today. We're yet to boost our 240L system this year, but then we don't use a lot of hot water. We'll see how we go over winter... note that the little system lacks a booster.

    The pressure is high. There's about 5 - 6 m of head pressure and no taps in between (there are two sharp right-angles in the pipe, and another 3 gentle curves). Really, I'd be happier if the water came out more slowly! A consideration, though, is that the tank itself can only tolerate 5 psi, so I really want to avoid a situation where the overflow can't keep up with the injection of water and I blow up the tank -- for that reason I won't be occluding the pipe at all.

    I've put some lagging right up where the pipe is full of water, but I haven't bothered further down -- I guess I thought that because the pipe is mostly empty of water it doesn't matter so much? I guess it will still lose heat to the air as it flows though... I'll do it ;-)

    How's your PV going? We're just breaking even on average (grid import ~= grid export) though there are days when we draw more. Our panels are only at about 23 degrees, which is too low for good winter production. I'd like to try and increase their tilt without getting a sparky out. That should be fun ;-)

    Cheers, Angus

  3. I forgot to mention, one downside of this tank: because it is so little (30 L) the ratio of its surface area to volume is higher than that for a larger tank. For this reason, it loses heat more quickly. When I said it was 40 C this morning, we'd just had a night down to 9.2 C, and I'm sure it lost some heat. Sometime, I'll try to quantify its heat loss.

    Cheers, Angus

  4. Very interesting and a good explanation with the photos too. Well done.

    At first glance I was agreeing with Chris about the insulation on the down pipes, but then you evacuate that tube entirely don't you, when you get hot water? How cold does it get there?

    I've been looking into building a home for my retirement on some property. One of the things I've been doing as I look at different room layouts is shortening the distance between hot water heater and end use spots, like kitchen, laundry and bath. I think the current apartment I live in, a duplex (two rentals in one building), my hot water heater is something like 18 feet from the bath. All that distance to lose heat, lol. I'm insulating the pipes with foam but still thats cooled water in the pipes that must be forced out before the hot water gets to the shower.

    Much better to start with a main power core room, with heater, cooling and hot water together, then cluster the rooms around it. I'll post some designs sometime over on the Green Wizard forum, if people are interested.

    Again good job.

  5. Hi dtrammel,

    My way of thinking is that there are two problems with a long run of pipe:
    the first is the loss of heat as the water flows through it. This can be reduced by lagging (insulating) the pipe.
    The second is that (normally) when you're finished with hot water you're left with a pipe full of hot water that cools down (is wasted) and must be flushed before you get hot water next time. This wastes both water and heat.
    What I've build here doesn't have the second problem because the pipe empties when no water is flowing in it, but hot water still loses heat in transit so maybe I should insulate it! (a cold winter night just gets to freezing where I live)

    It's all trade-offs. We pump our rainwater at a lower pressure than town water here (about 25 psi) and have 1/2" copper pipe. Narrower gauge pipe is good because there's less water in it. But, then you get less flow off course ;-)

    For a new build, I'd be tempted to put all water-using rooms together or, failing that, do what I've done here and install separate solar hot water systems. Of course, this calculation depends on whether you think you can get by totally on solar or will need to boost. You don't want to be electrically boosting two hot water systems! Also, having two systems increases their surface area and hence their heat loss (versus a single larger hot water system).

    I have been meaning to join up green wizards for about 2 years. I'll send you an email.

    Cheers, Angus

  6. Hi Angus,
    Good stuff. Yeah, the general principle for the angle of solar panels to get the best average between winter (tilted) and summer production (flat) is to match the latitude of your area. I reckon you're about 34'S and I'm about 37'S. Mind you the roof was 34' so the PV and HW panels are at the same angle - a few degrees makes little to no difference. :-)!
    The solar PV was not good a week or two back as there were days on end where the best it would generate was 1kWh / day, so every day the batteries got a little bit lower and reached into the high 60% full. That was about as bad as it gets. They're back up to 90% this morning, but the batteries are chemical reactors are and don't fill up like a water tank! It can take many days to get from 85% full to 100% full regardless of how much energy the PV generates.
    Cheers. Chris

  7. Great job and what a cool idea. You certainly are an inspiration to all the do it yourself-ers out there. From the pictures you make it seem really easy but I can only imagine what it must have been in reality. Hope it works out for you in the future.

    Carmelo @ PRO Hot Water Service

  8. Thanks Carmelo,

    It really was a very do-able project and is still working well a year later. I find that in winter its performance is worse -- this is partly due to the lack-of and weakness-of sunshine, and partly due to the fact that the injected water is so much colder, which rapidly cools the water in the tank. A better setup would be for the solenoid to just drain the tank by gravity pressure (that hot water comes out the tap), and then separately refill it (this would avoid new, cold water chilling the tank). But that's getting a bit complicated...

    But really, the thing I like about this system is that it's user-installable and user-servicable. I think a lot of people underestimate what they can do themselves...

    Cheers, Angus


Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.