Designing & building a solar hot water tank

[Daox]

Since doing the concrete work, I had to redesign the tank to fit the new area. It is now based off a 4'x6' sheet of plywood base. The new tank will hold around 500 gallons. Due to the shorter length it will loose a bit less heat, but not a ton. The main benefit is that I'll be able to insulate the tank to around R40-50. The combination of the two should bring the heat loss down from around 4k btu/day down to 2.4k btu/day. The new shorter design also reduces the maximum stress on the plywood down to a peak of 866 PSI which is about 1/5th the maximum rating of plywood (4500 psi). This should account for any imperfections in the wood, aging of the wood, any effects of thermal cycling, etc which might weaken the tank over time.

[AC_Hacker]

Hey Daox,

I just came across this solar thermal storage tank kit. It has a fabric outer part and a water-proof fabric liner part...

The Softank Kit - American Solartechnics, LLC.

They sell you the liner parts all assembled, you build the tank...

What do you think?

-AC

[Daox]

Not a bad idea. I like it and putting it up would probably be quite easy. Being cylindrical heat loss is also minimized which is also good. However, I would need 2.5 of them to get the capacity I desire and I'd rather not try to deal with two tanks.

[AC_Hacker]

Originally Posted by Daox

...I would need 2.5 of them to get the capacity I desire and I'd rather not try to deal with two tanks.

Did you check out the link?

They do have larger capacity solutions...

-AC

[Mobile Master Tech]

Those softanks look great-I looked into them a bit. Like Daox said, you need a lot of tanks (or a lot of $ for a big one). If you have more than one tank, you can't take advantage of the simplicity of stratification. I doubt the advantages of multiple staged tanks outweigh the benefits of a single simple stratified tank-I suppose it would depend on the situation.

The most efficient storage shape, considering volume inside divided by surface area, is the sphere. The next best is the cylinder, like the softanks. The closest easy to fabricate shape is the cube. Heat loss is a function of surface area, temperature differential, and the heat conductivity of the whole package. The heat loss that really matters, though, is how much you lose vs. how much you can store, assuming you have a heat source large enough to bother with storing anyway. A 4x4x4' tank holds 478 gal and has 96 square feet of surface area, for 0.2sf/gal. An 8x8x8' tank holds 3,830 gal and has 384sf of surface area, for 0.1sf/gal. A 16x16x16' tank (sorry, I couldn't convince Charmaine to let me build it!!!) holds 30,640 gallons and has 1536sf of surface area, for 0.05sf/gal-1/4 of the loss area per volume! So if the purpose is to store heat when you can get it to use when you can't, and any heat you lose through the tank does you some good anyway in 6-9 of the 12 months (meaning the tank is within the building envelope), it makes sense to build as large a tank as is practical so your loss area compared to storage capability is as favorable as possible. Multiple 60-120 gal round tanks? Don't even get me started!!

[Mobile Master Tech]

Since it is tough to install anything that won't fit through a doorway, most tanks have been small, collapsible like the softanks, or built of dimensional lumber. Plywood being sold in 4x8' sheets plus the limitations of sheet lining material and concerns about bulging due to water pressure has meant that most homebuilt tanks have been approximately a 4 foot cube or a 4x8' tank 4' high, which is not enough to acheive effective stratification.

Although I personally will build a tank with 7x7' outer dimensions due to space limitations and to keep as close to a cube as possible, this plan will work for up to 87.5x87.5" square by 69" high water dimensions, assuming you don't want to get too far away from a cube shape, use 8' long dimensional lumber for the support perimeter, have 3" headroom from the top of the water to the top of the tank, and have it fit and be maintainable in a room or basement with an 8' ceiling. This will hold 2295 gal and have 0.12sf/gal surface area.

To build the base, use a 2x4 laid flat to join two 4x8' sheets of 3/4" plywood to form an 8x8' square. Scale down all dimensions to make a smaller footprint tank. From the underside, screw down (with lots of weather resistant screws) 4 2x4's in a butt joint pattern to make a band perimeter all the way around the outside edge. The joining 2x4 will be on the bottom. Use construction adhesive anywhere things join together. Place this on top of 2 sheets of ISO (polyisocyanurate foamboard insulation) in your choice of thickness (I recommend 3"), with a cutout for the 2x4 joiner on the bottom made with a router or whatever means you see fit. Even though 2300 gallons of water and the tank will weigh nearly 20,000 lbs (wow!), you have 9,216 square inches to spread this over, for a little over 2psi. Most polyiso is rated for 16-25 psi at 10% compression. You probably would worry more about ground settling due to the weight.

Start assembling on or very close to the tank's final position because it will be hard to move when done. Screw 4 4x8 plywood sheets horizontally to the inside of the base band and to each other. Now you have 8x8 by 4' high tank walls. Screw 3 more bands of 2x4's on edge horizontally around the sides with the first 2 8" on center then the other 16" on center (you could keep doing 8" on center if you doubt the extra supports described later will do the job, but there is no need for a solid stack of 2x4's as seen on one build at builditsolar.com), but this time overlap the end of one 2x4 over the other so the 2x4's are staggered and leave the ends loose. Now screw a band of 4x4s with 2 butt joints so they are at the same perimeter height around the top edge of your plywood box so that half is screwed to your existing box and the other half forms a support for 2'x8' horizontal plywoood pieces making the sides 6' high. Put those in, screw them to the 4x4's and continue to the top with more horizontal bands at no more than 16" on center. The top band of 2x4's needs to be butt jointed flush with the top of the plywood sides so the top can be screwed to it all the way around. Make the top the same way as the bottom, except frame in a removable hatch with flat 2x4's in one corner for a manhole and with the joining 2x4 on top instead of on bottom. Dont cut the corner out of the plywood-move away from the edges a few inches. The top and bottom need to be intact at the edges and supported from one side to the other at the manhole opening so they can be strong in tension to take the strain of the supports that prevent bulging.

Once all 6 sides are assembled tightly, screw the butt joints and lap joints for all the horizontal banding together. Now, use chainlink fencing end posts (the nearly 3" diameter ones that would take a truck to bend) and lag bolt them with modified chainlink clamps or whatever you choose made into a "u-strap" to all horizontal bands. I would suggest 2 posts per side, but one may be enough if you are making a smaller tank. Use 2x4 spacers to support a joining 2x4 perpendicularly across the top of the 2x4 that joined the 2 sheets of plywood forming an "x" or tic-tac-toe grid of supports across the top, depending on whether you have 1 or 2 supports per side. Now the top and bottom are strong diaphragms that will act in tension against any bulging of the sides that are transmitted to them via the chainlink posts.

Follow Sani-tred's instructions for sealing a water tank(they use this on municipal waste water tanks) with their Permaflex/LRB system and don't skimp! Be sure to coat everything inside, don't paint yourself into a corner, ensure you have plenty of ventilation piped in through a duct in the manhole, and that all screws are countersunk enough to not cause a screwpop through the Permaflex. Check for splinters, debris, etc that might compromise the thickness of the Permaflex membrane.

Add a 1" to 1.5" CPVC pickup pipe, depending on your system, that goes within 2" of the bottom, preferably with pipe insulation on it and a reducer on the end that will act like a pickup bell. Any pipe that goes from the top of the tank to the bottom can conduct heat from the top to the bottom and dilute your stratification. Add a 1" to 1.5" return that expands to at least 3" sanitary tees through a mesh screen or somesuch to slow the return flow down. If you have 3 sanitary tees around 1-1.5' apart, the water has 3 choices to enter at its own density. If it is really hot, denser/cooler water lower in the tank will prevent it from flowing down to the lower exits. If it is cooler than the hottest water at the tank, it will continue to sink down to the lower levels. Remember that an open system like this needs a circulator pump with enough head to get the water to the top of the system in case it loses it's prime. If this goes to solar collectors, you need to have a UPS backup for your controller & collector pump so warm water can be moved to the collectors if the power goes out in freezing temps. Alternatively, you could use an antifreeze solution and a separate heat exchanger in the bottom of the tank, but you will lose a lot of your stratification ability. This would be better insurance in extreme cold climates, but not much of an issue in Georgia. Antifreeze increases pumping power due to the viscosity increase and decreases the thermal conductivity, but 20% propylene glycol won't cause too much problems, will provide freeze protection down around 20F, and provide burst damage protection everywhere but where Santa lives.

Add another pickup to get hot water from the top (such as for underground heat injection that I will describe in another post) and a heat exchanger for your working fluid in the upper 1 foot of the tank (DHW in my case for both the house plumbing and radiant flooring), a floating swimming pool thermal blanket on top of the water once full (I will use my reverse osmosis system to fill so there is no significant mineral content) and you are good to go inside. Make these penetrations through the top, not the sides. Insulate the outside as you see fit. I recommend 3" of polyiso cut to fit inside the banding, then another 3" routed to go over the tops of the banding with joints staggered to a different location than the first run and extending to the floor. If the insulation is at least 4" thick, there will be no exposed thermal bridges except the contact points between the horizontal bands and the chainlink posts. If needed, you could build up an insulation "box" around the posts as well. Studies I have seen show there is very little benefit to insulating thicker than 6". The bottom is thinner, but also will be a lower temperature and eventually will warm the ground beneath it so the deltaT will be low.

Thanks to all on ecorenovator, builditsolar and many other places that helped me formulate my ideas. I will detail my own construction once I get started. I've got the Permaflex already!

Craig
The MMT

[Mobile Master Tech]

Daox, are you sure about the 866 psi? A 6ft column of water exerts about 2.5 psi.

Craig

[AC_Hacker]

Originally Posted by Mobile Master Tech

Daox, are you sure about the 866 psi? A 6ft column of water exerts about 2.5 psi.

Craig

Going by THIS formula...

P(static fluid) = ρgh

where...

ρ = fluid density
g = acceleration of gravity
h = depth of fluid

ρ for water = 62.30 lb/foot cubed = 0.0360532407407407 lb/in cubed

g = 32.2 ft/sec squared = 2.683 in/sec squared

h = 6 ft = 72 in

P(6' static water column) = 0.0360532407407407 * 2.683 * 72

P = 6.965 lb/in squared

... or there abouts.

-AC

[Mobile Master Tech]

I think we've got an unit conversion error in that calculation. One foot of gauge water pressure (or head), disregarding absolute pressure since the atmosphere is working on all surfaces, is 0.433psi. 6*0.433=2.598psi. Since you got me thinking, I pulled out my dual pressure differential manometer and stuck it in a bottle of water to be sure-I didn't bother with many decimals but 12" of air in the tube below the surface read 12" of water column and 0.43psi.

[Daox]

Well, my tank is 4 ft tall. So 4/6 * 2.6 = 1.7. That is the pressure I'm using at the bottom of the tank.