# Garn hydronic design

Posted By Antman, Dec 31, 2015 at 7:18 AM

1. #51

### Antman Member 2. ```NULL ```

Sep 27, 2015
54
0
Loc:
Memphis, TN
Latest iteration for clarification on a way to put both houses in a series with 3/4" ID pex. I cannot visualize it but I am keeping my mind open to what Leon and Dan can share.

File size:
117.7 KB
Views:
127
2. #52

### leon Minister of Fire 2. ```NULL ```

Feb 3, 2013
502
20
Loc:
Southern Finger Lakes Region of New York

Hello and good morning antman, maple1, Still to early and waiting on plumber

Your current illustration is perfect antman. That helps me, because a circulator is just like a trash water pump, run, run, run, push water, push water, push water, push water to a discharge point at the end of a hose.

A circulator with a check valve may be all that is necessary.

They use them in three story building for hot water heat and cooling towers with no issues
in my case I am circulating through 250 feet of base board in a poorly insulated house and I do not have a check valve in either of my circulator's.

Find a local mechanical engineer. The national mechanical engineering society can help you find someone-can remember the groups name right now but a nearby university with an engineering curriculum can direct you to one I am sure

I have to go.

3. #53

### maple1 Minister of Fire 2. ```NULL ```

Sep 15, 2011
8,408
1,578
Loc:
Nova Scotia
I'm quite sure I wouldn't do series, and I would definitely not use 3/4" pipe underground.

For one thing - in the last diagram, the flow to house 2 is limited to what is flowing through the loads in house 1.

Although, I don't think we've seen the exact horizontal relationship between all three points of interest - barn, and two houses. One diagram showed the barn being between the two houses, houses being 230' & 300' in opposite directions (post 48 after looking back). Whereas the one in post 51 notes a 900' loop & 560' loop. I assume there is a triangle of sorts connecting the 3? Depending on that geometry, primary/secondary may even be a consideration.

Stop hovering to collapse... Click to collapse... Hover to expand... Click to expand...
4. #54

### Karl_northwind Feeling the Heat 2. ```NULL ```

Feb 13, 2012
496
85
Loc:
Central Wi.
skimmed the thread, didn't read it all. are the 2 houses in line, IE would you later be running house 2 off the same lines that go to house 1? your heat source is bought, so unless you want to add a second wood boiler, we have to work with what we have.
figuring the 150kbtu/h you calculated, 600' of 1.5" pex round trip at 15 GPM will require about 27' head loss (and 14 gpm=23'). a taco 2400-20 will do that, although I'd consider upping that one size to a 2400-45 if it's going to feed both houses.

I just did my first set of foamed in place lines, (copper) and the price was really reasonable compared to logstor or anything else. the 1.5" uponor with EP fittings would be fine. 2" might be even better, at a couple bucks a foot higher price. the foam ran about \$5 per foot. plus \$8? per foot for 1.5" pex (\$4x2) for \$13 a foot you have durn good insulation, durn good flow.

keep the circulator near the boiler to remove issues with the NPSH on that circulator, and that will help too.
the 2000 is a little small for that load, but most of the year it'll handle all the load, and part of the year it'll handle most of the load.
karl

also, zoning with small circulators in the basement to deliver to the house air coils will help them self purge should a little air get in them.

Stop hovering to collapse... Click to collapse... Hover to expand... Click to expand...
Antman likes this.
5. #55

### leon Minister of Fire 2. ```NULL ```

Feb 3, 2013
502
20
Loc:
Southern Finger Lakes Region of New York

Hello and good morning antman, maple1, Still to early and waiting on plumber

Your current illustration is perfect antman. That helps me, because a circulator is just like a trash water pump, run, run, run, push water, push water, push water, push water to a discharge point at the end of a hose.

A circulator with a check valve may be all that is necessary.

They use them in three story building for hot water heat and cooling towers with no issues
in my case I am circulating through 250 feet of base board in a poorly insulated house and I do not have a check valve in either of my circulator's.

Find a local mechanical engineer. The national mechanical engineering society can help you find someone-can remember the groups name right now but a nearby university with an engineering curriculum can direct you to one I am sure

I have to go.

6. #56

### Antman Member 2. ```NULL ```

Sep 27, 2015
54
0
Loc:
Memphis, TN
The knockout for underground lines coming into the back of the GB is in right on a straight line between the two points of entry into both homes. From that point to house 1 is 230' and to house 2 is 300'. The longest loop happens to be highest point of the system for house 1. For house 2, it happens to be through any one of the 7 radiant floor loops, which I understand for radiant floor shouldn't be more than 300'. I've attached an illustration.

So, to get out of the ground, up into the attic and over to the highest point in the system for house 1, there is an additional 50'. That means 230' + 50' to the furthest load or 280' is half the loop. To get from there back to the GB requires an additional 280' making the longest loop 560' which is the length I used to estimate head loss for house 1:

560' * 1.5 * 0.04 = 34 feet of head

As for house 2, it takes 300' to get into the house and to the supply manifold, 300' through the longest run of radiant floor and back to the return manifold, then another 300' back to the GB which totals 900' which I used to estimate head loss for house 2:

900' * 1.5 * 0.04 = 54 feet of head

File size:
30 KB
Views:
101
7. #57

### leon Minister of Fire 2. ```NULL ```

Feb 3, 2013
502
20
Loc:
Southern Finger Lakes Region of New York
With 54 feet of head is about three and a half stories with tall ceilings in comparison.

the B+G NRF36 without a check valve is in that range using speed three and
reaching 34-35 feet of head pressure at 37 gallons per minute+-.

Measure three times, take aspirin or Tylenol or Advil, measure again, then sit down and
look at the big picture as to where you can park a circulator.

Then you see where the back pressure of the 2,000 gallons of heated water in the Garn boiler
will only be a help to you as the circulator will be blessed with a "flooded suction" condition
wherein the water is always there no matter the pump speed.

As long as you have "flooded suction" at the circulator inlet your circulator will run cool
and quiet with no air bubbles.
=========================================================================
The question is more of "do I want a one inch feeder pex from the Garn to the 3/4 inlet of the
NRF-36 which would be a possibility as flooded suction is always a circulator's best
friend.
==========================================================================
The same applies to a less costly gear pump for a commercial log splitter. as flooded suction
always, always, always helps in keeping a gear pump wet and removing air bubbles.

Lots of ways to do this "BUT" all it is is a simple water pumping problem with a "dumb" low pressure circulator
operating at 1,750 RPM with restrictions from air handlers and radiant in floor heat.

Once you start pushing even heat through the system using speed three you may be able to slow the pump
down after the air bubbles are out and save on heat losses from the Garn with or without a system bypass loop.

The circulator will not overheat from lack of flow as the suction port is always flooded and cooling the impeller
housing and impeller.

Once the air is gone and there is little to no turbulence except for any micro bubbles at the initial start up that
can be dealt with a drop or two of dawn dish soap in the pump inlet.

"Pumping", "Circulator", "water flowing through "X" sized pipe" problems require a mechanical engineers insight for this
situation as you have a potential racetrack for water flow there to take advantage of.

You have a cooling tower with two chilling loads if you look at as a cooling problem, that is the only difference in my
thought anyway.

8. #58

### maple1 Minister of Fire 2. ```NULL ```

Sep 15, 2011
8,408
1,578
Loc:
Nova Scotia
A pump curve chart for a B&G NRF36 shows that at 34 feet of head, the pump will flow exactly ZERO gpm using speed three. If you could knock it down to 30 feet, you might get 5 gpm. You need ZERO head to get 37 gpm.

There may be some useful info in your posts, but there is also some misleading & confusing stuff.

Stop hovering to collapse... Click to collapse... Hover to expand... Click to expand...
Karl_northwind likes this.
9. #59

### leon Minister of Fire 2. ```NULL ```

Feb 3, 2013
502
20
Loc:
Southern Finger Lakes Region of New York
I am sorry if I have confused the issue.

The pump information sheet I have came with the NRF 25 pump I bought listing the four pump models and the specs for them

On the cover sheet/first page they list the operational limits for these pumps as 150PSI with a maximum operating temperature
of 225 degrees. The electrical ratings are for 115 volts and 230 volts

The left side of the graph Lists (kpa) (M) then (FT).

The lower portion of the graph lists GPM from 0-60 GPM
then the metric conversion scale is 0 to 13 (cubic meters/hr)
then below that it lists 0 to 3 (l/s) liters per second

For the NRF 36
speed three peaks at 33 FT+- and on the lower graph the flow peaks near 38 G.P.M. +-
speed two peaks at just under 30 FT and on the lower graph the flow peaks just at 32 G.P.M.+-
speed one peaks at a hair over 25 FT and on the lower graph flow peaks just over 25 G.P.M.+-

10. #60

### maple1 Minister of Fire 2. ```NULL ```

Sep 15, 2011
8,408
1,578
Loc:
Nova Scotia
Right - so looking at the curve itself, and seeing when the vertical or left side of the graph is at 33ft, if you try to follow that 33ft level over to the right you are immediately at the pump curve line, so then if go vertically down from there to the bottom scale you're at zero on the horizontal flow numbers. (From memory, give or take - don't have it in front of me right now). Meaning at 33ft of head, the pump doesn't pump anything. Or, another example from memory, if you start at the 30ft head on the vertical axis on the left side, and go horizontal to the right until you hit the curve, then go vertically down to hit the horizontal flow axis, you come to around 5 gpm. Meaning at 30 ft of head, it flows 5 gpm.

Just trying to alleviate some possible confusion.

And I am, as always, also open to the same alleviation if I get mistaken about something - par for the forum & what makes it a good resource.

Stop hovering to collapse... Click to collapse... Hover to expand... Click to expand...
Karl_northwind likes this.
11. #61

### jebatty Minister of Fire 2. ```NULL ```

Jan 1, 2008
5,594
830
Loc:
Northern MN
A design such as yours, Antman, I find to be very challenging. I haven't added much input because I'm not sure I have the expertise to lead you to a satisfactory result. From my experience I do know that undersized pipe and long distances are major fundamental contributors to disappointment in or failure of installations.

Stop hovering to collapse... Click to collapse... Hover to expand... Click to expand...
12. #62

### Antman Member 2. ```NULL ```

Sep 27, 2015
54
0
Loc:
Memphis, TN
Yes, am putting a pressurized design together using 1-1/2" logstor for the primary loop as the head pressure appears to be half that of the 1-1/4". The Taco 2400 circs should be able to handle this high head situation presented by the long underground run. The head in the secondary loops should be able to be handled using circulators for those zones.

13. #63

### maple1 Minister of Fire 2. ```NULL ```

Sep 15, 2011
8,408
1,578
Loc:
Nova Scotia
I was thinking this as well - oddly enough, when laying in bed this morning while only half awake. Funny how & when the mind works.

There are quite a few challenges here. Two long underground loops. An unpressurized boiler. Several WAHXs in the furthest reaches of the system, which typically require the hottest supply water. More heat exchangers at the boiler to go from unpressurized to pressurized, and every time you go through a HX you usually lose some off your top end temps. A boiler design that lends itself very well to low temp emitters but is I think somewhat limited when it needs to supply consistent high temp water (from reading), as well as maybe also not doing the stratification game as well as others, as well as maybe not as fast to supply its hottest water in the earlier stages of burn.

I think you would be on the right track with the last post. Keep the two main undergound runs as short as possible with fat pipe. Just get them in the house to a manifold as soon as possible (manifolds just inside the house). Then have more pumps inside, one for each load. Then on call for heat, you would have the main undergound pump doing its thing while also getting assist from the load pump(s). If you can find pumps for each that sound like they might be bigger than needed, but are multi-speed, you could have some tunability. Big pumps are expensive though - both to buy, and to keep running (power consumption). That's where the investment in bigger underground will pay back. I was thinking primary/secondary setup might work - primary underground loops to get to the houses, then secondary flows through close Ts for the loads inside. But the secondary close T flows would not provide any assist for the main loops, whereas a manifold setup should. And with 3 load HXs , each successive one would see decreasing supply temps if a close T setup were used.

Has Garn been spoken with re. the issue of using HXs to go from unpressurized to pressurized? And the elevations involved? I would seek their input before plunging fully in. Just wondering about the returns of doing that vs. the bit of loss of top end temps. That's one aspect of the design there hasn't been a whole lot of input on. Stepping back a bit, I am quite sure that there are OWB owners that have somewhat the same situation - an open boiler with a long undergound run. I am sure this could be done, just not sure about the best way to do it - very big pumps are also common in some OWB setups.

Stop hovering to collapse... Click to collapse... Hover to expand... Click to expand...
14. #64

### Karl_northwind Feeling the Heat 2. ```NULL ```

Feb 13, 2012
496
85
Loc:
Central Wi.
If you do the Logstor 1.5" you have a better situation than 1.5" Uponor. 1.5" logstor has an ID of 1.6" ID, better than the number I was using. if you do to and from the GB with 1.5" to house #1, you have a head loss of 11' at 600LF of pipe and 15 GPM.
that's way better.
house 2 at 600 feet would need lower flow rate most of the time, as the floor loops can mix the temp down to 100F probably.
you won't use one pump to pump the entire 900 feet with house 2, you'd use one pump to move the 600' of 1.5" pipe, and do as maple suggested, connect a pair of closely spaced tees with another circulator pushing thru the floor loops when there is a heat call for any particular zone. or you use a flat plate heat exchanger in the building (could be either the GB or house 1) and run the heating loops pressurized. (as a contractor this is what I would do, because if there are air problems in the floor loops, I have to go back and fix it and that costs the price of a heat exchanger the first time I have to go back.)

on a separate topic, while you're waiting for Dan's books (and they're worth the wait, if nothing else, for the pickle story) read Caleffi's Idronics series. google them (they're free PDFs) and read them in order, and you'll have a really decent handle on what is going on in hydronics. the ones that deal with topics you don't particularly care about, just skim and look at the nice pictures.

cheers,
karl

Stop hovering to collapse... Click to collapse... Hover to expand... Click to expand...
15. #65

### Karl_northwind Feeling the Heat 2. ```NULL ```

Feb 13, 2012
496
85
Loc:
Central Wi.
that would put you in the range of a taco 0014, using 174 watts/hr versus 228 watts/hr for a 2400-20: that would save you about 1.5KWH/day, or 15 cents a day, and \$55 per year. for the next 20 or 30 years, that adds up. you might find that you'll save \$ by going with the smaller PEX, and running a high efficiency ECM circulator from wilo or grundfos. you'll have to do the math on that.

Stop hovering to collapse... Click to collapse... Hover to expand... Click to expand...
16. #66

### Antman Member 2. ```NULL ```

Sep 27, 2015
54
0
Loc:
Memphis, TN
I originally had planned to do that until I couldn't find high-flow manifold units with larger than 3/4" outlets. I haven't put that much thought in how to build one yet, because, up until recently, I have been trying to build an open system. It seems that pressurizing may solve several problems for me:

1. oxidation (able to use cast iron pumps)
2. raising water to the highest point of the system where the loads are the greatest
3. less pressure drop across the circulators (less energy required to move water)

Can black pipe tees and nipples be used to build that manifold? I guess I may as well go ahead sweat a copper manifold as I'll already need to be doing that for the WTAHX? Are there any other options for manifolds?

17. #67

### Antman Member 2. ```NULL ```

Sep 27, 2015
54
0
Loc:
Memphis, TN
If my secondary circs can better assist the primary circ, I guess a manifold serves me a lot more than the easier to build close tees. In building the manifold, is there anything in particular that lends itself to assisting the primary circ?

18. #68

### Karl_northwind Feeling the Heat 2. ```NULL ```

Feb 13, 2012
496
85
Loc:
Central Wi.
you can certainly build it out of copper or black iron. you'll find out that black iron fittings are way cheaper than copper in the larger sizes.
if you put a swing check valve between the two secondary tees, you will get a little ghost flow, so use a flow check in the pump, but the secondary pump will assist the primary pump move the fluid along, but ONLY if the secondary flow is higher than the primary loop's flow.

Stop hovering to collapse... Click to collapse... Hover to expand... Click to expand...
19. #69

### Antman Member 2. ```NULL ```

Sep 27, 2015
54
0
Loc:
Memphis, TN
thanks for following my thread, I appreciate your input. I think I'll put the exchangers in the GB so I can use a smaller, less-expensive circ on the open side (the Garn side) which will need to be stainless steel

20. #70

### Antman Member 2. ```NULL ```

Sep 27, 2015
54
0
Loc:
Memphis, TN
when you say smaller PEX, I'm assuming you mean for the secondary? I was planning on using a Taco bumble bee in set point mode with temperature sensor in the supply duct after forced air goes through WTAHX

21. #71

### Antman Member 2. ```NULL ```

Sep 27, 2015
54
0
Loc:
Memphis, TN
So house 1 is getting 1-1/2 logstor but I still need to determine logstor size for house 2. I'll get around to running the headloss after kids go to bed but I'm hoping 1" logstor gets it done

File size:
132.7 KB
Views:
93
22. #72

### Antman Member 2. ```NULL ```

Sep 27, 2015
54
0
Loc:
Memphis, TN
Read Dan's book pumping away and tried to further refine the schematic. Is there any reason why house 1 couldn't be pressurized and house 2 left unpressurized?

File size:
147.1 KB
Views:
103
23. #73

### leon Minister of Fire 2. ```NULL ```

Feb 3, 2013
502
20
Loc:
Southern Finger Lakes Region of New York
Good morning antman,

You could do that provided you isolate the flow completely using a shut off valve and
pressure regulator to add water "IF" and only if needed and using the heat exchanger and use
a 30 gallon steel compression tank for more water in the system.
(the only time you add water is if you have a LEAK.

I like the steel compression tanks simply because there is nothing to go wrong with
them nor the worry about the bladder losing pressure through osmosis over time.
A thirty gallon compression tank would be more than adequate for this and makes
the secondary boilers job easier and less taxing on the system

Right now with my system is at 12 PSI using the 15 gallon steel compression tank
and there is no air scoop large piping runs nor crawling around bleeding radiators.

As long as each and every loop has a boiler drain to fill the loop and a second
boiler drain to vent the water your going to be fine.

As I said I would chat with a mechanical engineer pay for an hour of his or her time and
if the same engineer is a "HVAC qualified" all the better as all your wanting to do is pipe water
from two heating loads and back to a common storage tank using a heat exchanger in one loop
and an open system in the other loop "in this scenario.
================================================================================
Just so you know and understand this:

basement/sub basement of the hotels in a common feeder pipe with a return loop to the basement
and the hot water that is not used is returned in the same loop to the basement into a
common hot water storage tank to be reheated.
================================================================================

2. dumb pumps either single speed or three speed are "simpler" and easier to work with.

3. flooded suction for circulator's always, always always lets them run cooler and more efficiently,
but putting the circulator in the return side will create issues with turbulence and air bubbles in a
closed system, but put together properly paying attention to the circulators location in regard to
the point of no pressure change causes less problems as Dan has illustrated in his books.

a. a steel expansion tank above the boiler keeps two thirds of the tanks low pressure water volume
always flooding the boilers steam jacket and pushing water to the circulator.

4. check valves work to do one job and one job only, they keep flow moving in one direction.

5. simple one zone controls are what they are "SIMPLE".

6. steel expansion tanks have no moving parts, no air scoop, no automatic air vents in the top of the air scoop, no Schraeder Air Valves.

7. bladder expansion tanks have a "bladder sheet" that is sandwiched in between the upper and
lower parts of the bladder tank that flexs in relation to pressure and temperature.

Please keep us updated and let us know if you chatted with a Mechanical Engineer and what their thoughts and recommendations are.

I want you to succeed.

24. #74

### heaterman Minister of Fire 2. ```NULL ```

Oct 16, 2007
3,375
629
Loc:
Falmouth, Michigan
I'll just chime in and say that when it comes to pumping power required in a main circulating loop, there is no substitute for large bore pipe and tube.
The "entry fee" is higher initially but the dividend pays back forever.

The true system requirements will depend on the heat emitters more than anything else. (Think large and low temp with corresponding wide T.)
10GPM @ a 20* drop = about 100,000btu
5 GPM @ a 40* drop = the same.
Radiant floors, big radiators, oversized water x air heat exchangers, etc. These all allow less GPM to do the same "work".

Stop hovering to collapse... Click to collapse... Hover to expand... Click to expand...
25. #75

### Antman Member 2. ```NULL ```

Sep 27, 2015
54
0
Loc:
Memphis, TN
I would like to know how to find the best WTAHX for my system (one with a wide delta T) but there doesn't seem to be a lot of options. I was considering Valutech but I was I haven't had a chance to call them to inquire about the design delta T for those units. I am under the impression from other threads that, in general, WTAHX have narrow delta T specs, like 10 or 15. Anyone know where to get 40? Stack 2 on top of each other?