GARN piping design

[Jim K in PA]

OK – now that I know my GARN is actually going to be in my possession before the snow flies, I am ready to get down to the task of finalizing the design of the piping layout. Below are a description and a diagram prepared with the Taco HVAC software. Heat loss calcs were done with Slantfin Heat Loss Explorer.

Heat source – GARN WHS-2000
Max input - 425k Btu/hr.
Water storage - 1,825 gallons

Max. heat load(s):
1. House* – 70k Btu/hr. (2700 sq. ft.)
2. Greenhouse** - 30k Btu/hr. (~200 sq. ft.)
3. Shop/Garage*** - 35k Btu/hr. (~1700 sq. ft.)

* - The house currently has a 140k Btu/hr input oil fired hydronic furnace with 4-zones of HWBB throughout the house. Most of the rooms are plumbed with at least 50% more radiator than necessary, and none with less than 20% more than necessary (based on room by room heat loss calcs). DHW is currently via a coil in the furnace. The existing furnace is an EFM unit c.1995 in excellent condition. For this heating season, the furnace will be depressurized and run in series with the GARN. Next year I will have a propane backup unit installed near the GARN, and the oil unit will be taken off-line permanently. At that point an indirect water heater will be added.

** - Greenhouse is not yet built – and will not be on-line for at least 2 years. Type of heating method not yet determined – most likely with water:air HX. The Greenhouse will be attached to the Garage/Shop, and located less than 40’ from the GARN.

*** - Shop/Garage will not be on-line this year – most likely will be for next heating season. Garage (1st floor) will use a pair of water:air HXs, and upstairs (wood shop) will utilize radiant floor heating.

The GARN unit will be located in an enclosed shed behind and attached to the existing Shop/Garage (see pics HERE). Piping will be located within the lower garage area. The house is located approximately 60 feet from the Garage. The buried portion of the s/r lines to the house will be that distance, with the balance of the run located within the Garage.

System design – I intend to build a P/S system. I will break up the design thoughts/questions that way.

Primary loop:

Based on a total calculated max heat loss of 135k Btu/hr., I calculated I need a primary circulator of approximately 14 GPM (not sure why TACO software specs it at 5 GPM- I am probably doing something wrong with the software). I would like to use a variable or at least multi-speed pump that would step up flow rate to match the demand. I will research the Taco site for such an animal, but if someone has a suggestion on a model to use, please do share. The primary circulator will be on the supply branch before the mixing valve. There will be full port ball valves on the circuit to isolate the GARN, the pump and the backup heat source HX (future-propane fired) and the future isolated passive solar input (not included on this sketch). Primary loop piping will be rigid material, either copper or steel (“black iron”). Appropriate drain valves will be included.

The GARN comes with a 2” fitting on the supply side and a 1.5” fitting on the return side. However, do I really need to run the full 2” diameter to the pump?

Should I step the piping down after each set of closely spaced Tees?
Will a mixing valve be necessary on the primary loop? (I included one in the diagram). If so, what size/type?


Secondary loop #1 - House:

I will feed the house demand with the first loop. I am running oxy-barrier PEX, 1.25” diameter. S/R will be via a pair of 1.25” closely spaced Tees. Total pipe length for this loop is ~225 feet. Full port ball valve will be used to isolate the loop. Drain/bleed valves will be included. A 7-8 GPM secondary pump will be installed. I would like to use a variable speed pump with outside air temp feedback (setback?) to ramp the pump rate up/down according to demand. Sound reasonable? If I use the variable rate pump, do I need a mixing valve?

Secondary loop #2 – Greenhouse:

Next demand will be the greenhouse, due to the higher water temps needed (although total calculated max load is slightly less than garage/shop). .75” oxy-PEX will be used for that run, along with a ~3-4 GPM variable speed pump wired like the house circuit. The circulator will be actuated by a T-stat controlled zone valve.

Secondary loop #3 – Garage/Shop:

Last set of Tees will feed a manifold (.75”) for the Garage/shop loops. The garage zone will run direct from the supply to the HXs, but the shop RF circuit will utilize a mixing valve to ensure proper water temps to the floor. This loop will also utilize a variable speed ~3-4 GPM pump with feedback. The two zones will each have a T-stat controlled zone valve, actuating the secondary pump for this loop.

OK – I am sorry that this has gotten real long. I appreciate the input from anyone/everyone willing to read this far. Even if you call me a luddite and a lunatic.

Drawing below.

 

[Jim K in PA]

Did I not include enough info? Too much? Forget to brush my keyboard?

 

[Nofossil]

I can't quite read the GPM numbers on the Garn pump. Looks like an effective design, but I'm pretty sure that the Garn circ wants to be at least as large as the biggest load circ.

 

[Sawyer]

Two questions Jim.
Is it necessary to have the primary flow equal to the sum of all the secondary flows?
It seems the head on the primary loop is quite high, I was of the understanding that the primary lcirculator is sized for the flow and head of the primary loop, not the sum of the secondaries. Is this not true?

I am asking to learn more before my Garn arrives.

George

 

[Jim K in PA]

Nofossil - it is small; it says 4.8GPM. I agree, it makes sense that it should be larger.

Nofossil & Sawyer - my concern is that if ALL the districts and all zones call for heat simultaneously, I need to be able to feed the output that the three secondary pumps are going to draw down from the primary loop. If I do not, then I can imagine that the three secondary pumps combined draw down will pull flow from primary loop at a rate greater than the primary pump can support. Flow will be reduced, and perhaps cavitation will result?

However, thinking this through out loud, if you size the primary pump at least as large as the largest secondary draw, you will have 100% of the primary flow entering and leaving the largest circuit, and as long as that return flow exceeds the COMBINED flows of the rest of the circuit, then all should work, yes?

For my secondaries, the largest is 7.5GPM. The combined flow of the other two secondary loops is 6.7GPM. Therfore, if I install an 8GPM primary pump, I should have more than enough flow to satisfy all loops simultaneously. Am I missing something?

I am sure I sound like an idiot. My background is in mechanical engineering, but most of what I do is civil related.

Thanks for the input.

 

[Nofossil]

My background is in mechanical engineering, but most of what I do is civil related.

Basic thermodynamics apply, civil or uncivil.

When calculating the primary loop head loss, the secondary loops do not enter into the equation - only the straight-throughrestriction (very low) of the tees.

If primary loop flow is just a bit larger than the biggest secondary flow, then the return temperature will be lower and there will be a bigger drop between the first load and the last when more than one load is active. You show a 60 degree drop around the loop. That's large - most systems aim for less than half that. Th implication is that your last load would be seeing a water temp of 160 or less coming in.

The Garn is only 135,000 BTU/hr?

 

[Jim K in PA]

Basic thermodynamics apply, civil or uncivil.

Oh how true!

When calculating the primary loop head loss, the secondary loops do not enter into the equation - only the straight-throughrestriction (very low) of the tees.

Understood. I have read both of Dan H's excellent books on the subject. The head loss calcs shown in the drawing are not mine - they are inserted automatically by the Taco software, and I do not know how it comes up with any of the numbers. I am in the middle of doing some calcs right now, and I have "Primary-Secondary Pumping Made Easy" on my desk as I type this.

If primary loop flow is just a bit larger than the biggest secondary flow, then the return temperature will be lower and there will be a bigger drop between the first load and the last when more than one load is active. You show a 60 degree drop around the loop. That's large - most systems aim for less than half that. Th implication is that your last load would be seeing a water temp of 160 or less coming in.

Again, that is what the Taco software calculated. I am trying to hammer out the calcs myself right now. I thought that was too much of a drop as well.

The Garn is only 135,000 BTU/hr?

No - up in my monologue above I listed the GARN's max input of 435k Btu/hr. The 135k figure is inserted by the Taco software based (apparently) on the sum of the loads placed upon the furnace. Clearly this software is not designed to do what I am asking it, but I was just trying to use it as a tool to build a graphic example of what the design should look like.

 

[Willman]

Nice planning for your loads. Forward thinking and due diligence never hurt anyone willing to do them. My question would be are you concerned with boiler being over sized to begin with ? I know from reading posts here that one should
try to match boiler size with load requirements to prevent idling. Would you just use less wood per load to regulate ?
A suggestion on your future greenhouse. I have been looking into that for myself. I have been told and saw a GH that uses radiant tubing in the root zone. Easy install on foam board and cover with soil. Plants like warm feet too ;-)
Will

 

[Jim K in PA]

Nice planning for your loads. Forward thinking and due diligence never hurt anyone willing to do them. My question would be are you concerned with boiler being over sized to begin with ? I know from reading posts here that one should
try to match boiler size with load requirements to prevent idling. Would you just use less wood per load to regulate ?
A suggestion on your future greenhouse. I have been looking into that for myself. I have been told and saw a GH that uses radiant tubing in the root zone. Easy install on foam board and cover with soil. Plants like warm feet too ;-)
Will

Hi Will,

Thanks for the comments and kind words. I am learning at an exponential rate. Later this AM I will post a revised sketch with updated figures for flow and temps. I spent a couple of hours yesterday doing heat balance calculations and I think I am 99% there on determining my flow rates and total delta T.

As to the oversize question, overall yes, the 2000 gallon unit is much bigger than I would need for just the house at this point. However, GARNs do not idle per se. Fewer burns will likely be the case for just the house load, and I may have shorter burn times as well. I have A LOT to learn on that process, which cannot begin until I have it and start using it.

For the greenhouse, I am not planning on having any direct rooted plants in the floor of the house, just benches and hanging stock. I am considering radiant combined with a water:air HX. I plan on a lean-to style house about 12'x16', long side against the garage. I should have room for three narrow rows of benches with two ailes, which will have concrete pathways. I would put the radiant tubing in the pathways. For fall and spring, the radiant slab should provide most of what the house will need, while in winter the blower will kick in to keep the house above 55 deg F at night. I have had greenhouses before, with far less sophisticated heating systems. This one will be a treat.

 

[Willman]

Jim, Keep us posted on your progress. The Garn seems to be quite the unit.

Will