GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Extra-Large Water Heater Expansion Tank

gary__b | Posted in Mechanicals on

Hello and thank you for your time.

I recently purchased a house in which I’m preparing to do some upgrades, including replacing the water heaters. Right now there are 2 small electric tank heaters (in parallel) that are original to the house–late 70s. I plan to install one HPWH to replace them.

In the crawlspace of the house there is a very large water tank upstream of the water heaters. My guess is about 60-80 gallons, situated horizontally on concrete pedestals. The main cold water line branches to this tank, a line exits the tank and feeds only the cold inlets on both WHs. No ordinary expansion tanks on the WHs.

The only explanation I can think of is an expansion tank, but why so large? Has anyone ever encountered something similar? The reason I ask is because I’d like to remove it if unnecessary (add a regular expansion tank with the WH replacement). The house water is spring fed and gravity pressurized to about 80 psi. No backflow valves or anything like that.

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. Expert Member
    Akos | | #1

    The one thing I can think of is if the supply line is small than the large expansion tank was there as a buffer to maintain pressure.

  2. Expert Member
    BILL WICHERS | | #2

    I agree with Akos, this sounds like a pressure tank as would normally be used on a well. These tanks are sized to limit pump cycling. In your case, I'm guessing you're fed by a spring that is quite a ways uphill from you, and then that water source is piped down to your house and that's what you mean by "gravity pressurized". In such a case, I suspect the water volume that is supplied is rather low. To make up for the very low volume of water supplied by the remote spring, you use a large pressure tank to make up the volume so that you have enough water when you need it. The pressure tank acts like a battery, constly recharging at a low rate but able to provide bursts of power (water, in this case) when needed.

    The only time I need to use expansion tanks in the tens of gallons size range is on large glycol systems of thousands of gallons (8" pipes, things like that). I can't see how you'd ever need such a large expansion tank on a residential system, so your tank must be there for a different purpose other than the expansion of heated water.


    1. gary__b | | #3

      Thanks Bill and Akos. Pressure tank crossed my mind but the thing that confounds me is why it'd only be on the hot side (outlet only goes to WHs). I suppose it'd necessarily pressurize the cold too, but it'd have to backflow to do it. That being said, the hot pressure is clearly higher than the cold at the tap.

      Yes Bill, your inferences are correct. A tank about 200 ft above the house collects spring water, which is then piped downhill to 3 houses. The pipe to the house is 1.5" PVC, and between the tank and all the piping there is significant capacity, so I wouldn't think volume would be challenged. But perhaps at one point it was. The incoming line doesn't seem as old as the house and the tank.

      Perhaps I'll try a temporary bypass around the tank to see what it happens. Reason for wanting to take it out is due to age (integrity, build-up concerns) and to make a crawlspace encapsulation easier.

  3. Expert Member
    BILL WICHERS | | #4

    Unless you have check valves in the system, the tank will act to help with water delivery for everything. The tank just "sees" pressure differentials, and flows from high pressure to low pressure. The tank has no concept of DIRECTION of water flow in terms of which fitting water enters or exits.

    You might try looking for valves before removing the tank. See if you can isolate the tank from part of your system, then see what the difference is on the system with the tank isolated. If you have a lot of pressure drop, you'll know what the tank is there for. If the pipe coming to the house from that distant tank is really long, even a 1.5" line will have a lot of pressure drop when water is flowing. It's similar to what happens when you try to run a power tool at the end of a very long extension cord.


    1. gary__b | | #10

      Thanks Bill. I understand all that (I dabble as a mechanical engineer). I guess my point is it's simply a strange place in the system to put it. The obvious choice would be upstream of the whole system, not on the hot branch. I can't show by picture, but trust me that the piping to achieve the placement is a bit convoluted.

      But I will take a look for a valve next time I'm down there. Hoping I can test the effect on the cold output by shutting down the tank feed. Thanks for the idea.

  4. Expert Member
    NICK KEENAN | | #5

    Does the tank have any writing on it that might give a clue to what it is?

    1. gary__b | | #9

      The only label I spotted simply says "Sears Water Systems." No specs or any indication of what it's for.

  5. jonny_h | | #6

    With the description of the cold water entering at one point and water to the water heaters leaving from another point, I wonder if it's a sort of "tempering" tank -- allowing the water to come to room temperature to save on energy required by the water heaters. Is the spring extra chilly? Just a random thought, don't know if anyone actually does this though...

    1. Expert Member
      NICK KEENAN | | #7

      Good answer! This is something I've long fantasized about. Basically it's a poor-man's HPWH. The heat of the house is used to pre-heat water and reduce the load on the water heater. In the summer it contributes to cooling and possibly dehumidification. By allowing the cold water to warm you also reduce issues with pipes sweating.

      But I figure there's got to be a downside I'm not seeing.

    2. Expert Member
      BILL WICHERS | | #8

      That’s an interesting thought. I would think a linear run of copper pipe doubled back a few times as a exchanger would be simpler, but there is no reason a tank couldn’t do the job too. I know the water from my well is COLD so I can certainly understanding the thinking here!


    3. gary__b | | #11

      This is a very interesting thought Jonny! The spring runs at 45F. It's an unvented crawl (cold), so it probably doesn't due to much to temper in the winter, but it certainly could other times of year.

      If my pressure test without the tank doesn't result in much change, this seems like the only other possibility.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |