Jackrabbit Journal Rotating Header Image

Introducing our Solar Hot Water System

Our new facility houses a 150-gallon vat pasteurizer with which we have begun to make cheese.  Fresh cheeses must be pasteurized, which means raising the temperarture of the milk to 145 degrees and holding it there for 30 minutes. Heating 150 gallons of milk by 110 degrees requires a lot of energy.

In order to cut our heating costs, as well as our environmental impact, we installed a solar hot water system.  The system incorporated used solar panels that someone gave us for free.  Instead of spending thousands on a commercial heat exchanger and storage tank, we elected to build our own using the principles (if not exactly the plans) we found on BuildItSolar.com.

We constructed an EPDM-lined  plywood tank on a 4 x 8 foot plywood base.  We used 60 feet of 3/4″ copper pipe as a heat exchanger to preheat the water we use for washing.  Then we built a system to recirculate the solar-heated water through our vat pasteurizer, with two tankless hot water heaters as a booster.

Because our goal is to heat the tank to 170 degrees for pasteurizing milk– hotter than a typical solar water heating system– we couldn’t use pex for any line under pressure.  We used pex for the plumbing to the collectors and back.  But for the heat exchanger, which carries up to 40 psi of tap water, we chose type M copper pipe.  And for the pasteurization system, we used 3/4″ copper pipe insulated with polyethylene pipe insulation, and covered with 2″ PVC pipe.  The PVC pipe isn’t necessary for structural purposes, but in a food processing area there can be no exposed insulation.

By the way, we used “shark bite” type fittings inside the PVC pipe.  They’re not cheap, but they were the only way we could find to completely encase the hot water plumbing.

The vat heating system is not entirely closed– overflow comes from the vat during operations.  So we also installed a float valve to keep the thermal storage tank full, and a drain to keep it from overflowing.

The result: a system that will heat and store over 300 gallons of water.  It heats our washing water, and our vat.  And it cost less than $2,700.

But we had a challenge: how to drain back the system so it wouldn’t freeze on cold winter nights.  The system is pretty simple: a pump, which is situated outside the tank but below the water line, pumps water up through the roof to the collectors.  Gravity brings the water back down.  But when the pump stopped, there was no vent to allowthe water to drain back.

The obvious solution was to install a vacuum break at the collector, which we did.  That let air into the tubing to allow the system to drain back.  But siphoning action drained it back all the way into the tank, leaving the pump dry.  In order to get the water flowing again, we had to re-prime the pump.  Every time.  That’s not very automated.

If we put in a check valve, the water wouldn’t drain back at all.  I spent hours bugging the folks at my local plumbing supply store trying to figure out how to let the water drain back so far and no farther.

It took weeks of experimentation, but we finally found a solution: two check valves.  Actually we chose swing valves because they were cheaper and gravity worked in our favor.  The first I installed at the bottom on the intake line, preventing water from draining back that way.  Without a second valve, water wouldn’t have drained back at all.

The second, I installed on a tee on the intake side of the pump, but above the water level– and I installed it “backwards.”  It prevents water from being sucked in that leg of the tee by the pump, and when the pump stops, the pressure of gravity on the water allows draining back through this valve– but only down to the level of the check valve.  When water reaches that level, gravity closes the valve and flow stops.  The pump stays wet and primed for the next use.

Here you can see the pump sitting alongside the tank.  At the top left is the drainback swing valve, with a length of 1/2″ pex that returns water to the tank.  The valve you see in the middle is a standard hose bibb, which allows priming the pump the first time it is used (or any time air has gotten into the pipe).  The valve on the right is another hose bibb on the outlet side of the pump.  We use that one to drain the tank when needed.

The system looks more complex than it is.  Here, a series of ball valves allows the vat operator to choose hot water from the solar system to heat the milk, or cold water from a storage tank to cool the milk– and return the water to the location from which it came.

It has taken some doing, but we’re now using the system to make cheese.  In cold weather, we use the tankless hot water heaters to boost the temperature.  But in summer, we expect to have plenty of hot water.


  1. Sean says:

    Hi there,

    Thanks a million for posting details and photos about your drainback solar hotwater system!

    I’m curious about check value #2. You say that gravity opens the swing value when the pump is off. Has this worked in practice? Wouldn’t the pressure created by the volume of water in the piping leading to the collector keep this value closed when the pump is off? In addition, have you had any air in the system?

    Thanks again for posting details about your system.


  2. DJ says:

    Sean, the valve is installed in the other direction, so when the pump is on the valve is closed from the suction of the pump (and gravity). When the pump shuts off, the pressure of the water swings the valve open. When the water drops to the level of the valve, pressure equalizes and gravity swings the valve closed again. This works perfectly, though as I mention in the post, it was a long experimental process to figure out the proper configuration.

    The swing valves have greatly reduced problems with air in the intake side of the system. I did install a hose bib at the uppermost point of the intake line, just upstream of check valve #2. That way, if I do get air in the line, which seems to happen when the system has been shut off for a week or more, I can easily refill it with water by attaching a hose.

  3. Sean says:

    Hi again,

    Thank you so much for explaining how it works. I plan to design the swing values into my system. It’s a nice low-tech (and low cost) solution!

    Regina, SK, Canada

  4. Tim Connolly says:

    I have some questions>
    Are all your pipes sloped 1/4” /ft, back to the tank.?
    Is your return pipe,once inside the tank, above the water line, with a “Tee”?
    This could be your problem #2>…….”But when the pump stopped, there was no vent to allow the water to drain back”…….
    *****The first#1 problem: Your Taco pump is mounted wrong:(A) It should have the Body/motor on a horizonal plane.(See taco installation info.) (B) Its needs to be 3′ or more, below the water level in the tank.
    It needs whats called “Minimum positive suction head”. Taco has a Minimum spec on this.
    Do that and you can get rid of the check valves, and the vacume breaker.
    If you find this important, and of value, send me some Cheeze Please!!

    What size are your feed and return pipes, for the solar loop.?

Leave a Reply