My Tank construction and experimental Hex

[sweetheat]

Any hardware or building supply would have form banding. It comes in a long roll with holes punched every inch for nailing to the form boards. It works quite well for major concrete pours. Should work OK for storage tanks. It's not rust proof though. Steel wailers around on 4 sides would also work. Lumber banding would work too, their's a special tool used to tighten them up. Rich, I beleive tarm uses EPDM. I have questions about drain holes in the tank. Do they make fittings designed for EPDM? Does the tank water stay put or will it circulate? Is it just the heat storage medium? And the exchangers do the water circulating? Plumbing term for heat exchanger is hx? correct? How far are you from coastal downeast sparke? I would like to come and look at your set up some day. A day closer to clean heating. sweetheat

 

[jebatty]

Sparke - what is the flow rate gpm between boiler and tank?

 

[sparke]

jbatty, I am not sure on gpm. I tried to address this issue in another thread but got no responses. I am currently using a taco 007. I just purchased a 3 speed Grundfos UP26-99. I think some of my heat zones need a bigger pump as well. I eliminated the pump at the wood boiler all pumping is done from heat zone pumps or tank pump. It is working very well. I would like to know the professional way of sizing pumps but know one wants to part with that info...

Sweetheat, I live 20 minutes north of Augusta. You are welcome to check it out anytime. The water in tank is "dead water". Just a transfer medium.

 

[slowzuki]

Maybe I should start a thread on pump sizing? Oh well in short heres a few tips. You size the pump for the friction loss and how many gpm you need to move around a loop. The gpm is set by the heat loss in the loop. So:

1) Say you have a rad loop designed to give a temp drop of 30 F at 5 gpm to get the right BTU's for the coldest day.
2) Say your loop is an inline rad loop using finned copper, you add up for each size the total lenght of pipe and number of each type of fitting. Convert the fittings to equivalent pipe lengths using tables. Use a friction loss calc or table to solve the friction loss at 5 gpm. Lets assume there is 13 feet of head loss (13 feet of head loss is not 13 feet of pipe, it refers to a 13 ft tall water column which is about 0.44 or something psi per foot).
3) Go to a collection of pump curves and mark the point 5 gpms/13 feet. Any pump curve below that will not move enough water. Any curve above will move enough. If you oversize the pump it costs you extra money to run the pump and it will circulate more water than needed, maybe with velocity noise depending on pipe sizes etc.
4) In a backwards way, you can size pipe based on a pump by reversing the process.
5) It is bad practice to size a pump near its static / shutoff pressure if it has a flat curve out to high flow. The pump will not last as long.

 

[Dave T]

Hi sweetheat I am a commercial roofer and work with EPDM on a regular basis .045 and .060 when we make a penatration through a roof that we want water to flow through we ues a drain hub and run the EPDM right across it,using a mastic caulk with constant,consistant compression the EPDM is clamped between two pieces of metal with the caulk sealing the EPDM to the drain hub.This is on a commercial roof I am not fimilar with lined tanks but would think it has to be very similar.

 

[sparke]

Thanks for the interpretation Slowz. I have a book coming from Grundfos so hopefully it will have the tables in them. I will post a schematic when I have the chance to draw it. I have an interesting set up...

 

[Der Fuirmeister]

Sparke,

It's been awhile since this post paused for summer....how did your tank liner and pex hx work out?

I have a lot of 5/8" pex left over and might consider this over copper.

 

[sparke]

It works great!. My hx can easily sink all the BTU's the boiler puts out. The boiler runs wide open until the tank hits about 170* I have had the tank over 200* a few times (not on purpose). Again the jury is out on how long it will last but so far so good! On a side note - I have had problems with the cover. The combination of odd shape & location make it hard to get a good fit...

 

[Der Fuirmeister]

Thanks. I have lots of 5/8 pex left over from the radiant floor installation. Think I'll try your pex hx.

Have you tried using silicone adhesive to adhere foam backer rod (or similar type) to the EPDM on you tank top?

Did you ever find the submersible temp sensors? If not I might be able to locate a source.

 

[jebatty]

I am currently using a taco 007. I just purchased a 3 speed Grundfos UP26-99.

Something that works also has to mean that the physics are pretty good too. I did a line loss calc on 200' - 1/2", and pump head at 2 gpm is about 27'-28'; at 1.5 gpm pump head is about 16'-17'; at 1.1 gpm pump head is a little over 9'. Head at over 2 gpm quickly gets extremely high. Mutlitply the gpm x 4 for the four coils. At 9' head a 007 will move about 5 gpm, so it appears that you are moving about 4.5 gpm. At a delta-T of 20, your gross btuh is about 45000 with the 007.

Based on the above, for its pump curve my estimate is that a UP26-99 (single speed) likely will move about 2 gpm or a little more, x 4 = 8 gpm total = 80,000 btuh at delta-T = 20.

Your btuh will drop off as tank temp closes to within 20 and less of your boiler input temp. If boiler input is 190 and tank is 180, at 100% heat transfer your delta-T is only 10, and maximum btuh is 22500 at 4.5 gpm or 40000 btuh at 8 gpm.

Are you still using the 007? Do these calcs seem to fit what you are experiencing?

 

[sparke]

I am just using the one probe in the pictures. It has a 4' lead on it which gets me down to the lower 1/3 of the tank. I charge that part of the tank to 180*. I have measured the difference before. It is about 10* hotter at the top. So I charge the tank 170/180/190. This is a subjective guess based on readings at different lvls in the tank. As far as the cover goes - I have it fixed for the time being. I used Hardy board(concrete board used on floors and shower spaces for tile work). I then sealed the cracks with high temp silicone. This cover is inside the tank held up by supports. Over that - across the top is bubble wrap. Since the cover is in the tank the bubble wrap has a 2" air void. 2" polyiso. and r -24" fiberglass insulation over the top. Working well for a month now.

 

[sparke]

jebatty, It is funny you mention the pumps. I just posted yesterday in a different thread about us hearth.com folks always seeking better efficiency. The pumps are my next area of concentration.

I have more concerns about my heat loop pumps then anything else at the moment. Can you dumb it down a bit for me? I believe the way to size a pump properly is to figure out run length, size pipe, friction in fittings, etc... Is there an easier way to figure head loss? If you have temp gauges on supply and return and know the ambient temp of rooms being heated... I assume you can figure out whether or not your pumps are adequate? I have a difference of 10-20* from return and supply. depending on the room temp. Any quick rule of thumb? Maybe we should start another thread about pump sizing. I know there are other threads kicking around but since we are on the subject...

 

[jebatty]

Any quick rule of thumb? Maybe we should start another thread about pump sizing. I know there are other threads kicking around but since we are on the subject...

There are some rules of thumb which have appeared in various threads. For heat transfer, maximum gpm: 1/2" - 1.5 gpm, 3/4" - 4 gpm, 1" - 8 gpm, 1-1/4" - 14-17 gpm, 1-1/2" - 22-25 gpm, 2" - 45-50 gpm. For btuh calculation: gpm x delta-T x 500. For line head, total pipe run length (round trip + equivalent for fittings, etc.) and then use the calculator at Pipe Friction Loss. For calc of additional pipe length to add for fittings: see attachment. Once you've calculated pump head at target gpm, then look at the pump curves from Taco, Grundfos, B&G;etc., and find a pump that provides the target gpm at the calculated pump head, and you want your target gpm/pump head to fall about in the middle of the curve.

If your pump head is too high, you need to go to larger pipe size or shorten your run. As a practical matter, pump head 5'-30' is about the operable range, I think, for most home applications, and the lower the pump head the more flexibility you have for system design.

Hope this helps. If I made an error or someone can do this more simply, add what you can.

 

[sparke]

That is a big help. Thank You very much !

 

[Duetech]

Yes that is a lot of help thanks jebatty.