Shop/Barn Radiant design

arcticcontrolsgoose

Hello,

Just wanting to confirm I’m on the right track with heating design, and if anyone has input on how I should do the recirc on the primary/ secondary circuits.

Background:
5000 sqft building, total heat loss 90000BTU/H.  Currently 6 zones, the biggest being:

- 5-loop 1120sqft 1/2” pipe at 270-300ft per loop.  Calculated flow per loop: 0.8gpm, velocity 1.5ft/s, ft of head at 17.

- 8-loop 2240sqft 1/2” pipe at 260-280ft per loop.  Calculated flow per loop: 0.6, velocity 1.2, ft of head 16ish. 

The smaller zones that are only 1-3 loops have low flow and higher head given pipe length.  But I’ve been told velocity is more important than head which is why it was sized to 1/2”.

With all that, all circuits combined I get 12gpm and 18-20ft of head.  Can’t really find a pump that sits nicely on the efficiency curve for that. Would be nice to get the flow about double…but that muchu big bucks for a single pump then.   

So then if I go and do individual zone pumps (which I think would be a better design anyways) the low-flow circuits of <1gpm and high head are being difficult to find a good pump for.  Bit stuck as I thought my initial piping design was decent for the areas.  

What would you do in this circumstance?

hot_rod

The head of the highest zone is what you use, is that 17'?

Is this a single temperature system?

Looks like the Grundfos Alpha 26-99 fits nicely. Other brands probably match closely. A delta P circulator will adjust output as zones open and close, you need to cover the design load. 12 gpm?

How are you zoning? Is each manifold zoned with a zone valve, or a few dozen manifold actuators?

Rich_49

Please see page 211 of the attached manual . I think somewhere there must be a mistake . What temp water are you basing this off of , what are the lengths of your S & R piping from mechanical room to manifolds and what size pipe are those runs ?

I come up with 7.88' hd , 1.45 FPS and 4 GPM for you 5 loop manifold and 4.29' hd , 1.1 FPS and 4.8 GPM for your 8 loop . These are using a 140* Temp .

Add the head of your S & R piping from boiler to manifolds to the head of the circuits , only total head of manifolds , highest head circuit and S & R piping get added to figure head . You could probably do this with one circ .

Here are some performance curves from Taco for you . There re also high efficiency ECM circs you can find on their site .

https://www.tacocomfort.com/documents/FileLibrary/curves003-008.pdf

https://www.tacocomfort.com/documents/FileLibrary/curves009-0014.pdf

arcticcontrolsgoose

Appreciate the feedback guys.

I’m basing it off lower temp water.  The main large bays in particular are to 18degC indoor design temp…I don’t have my calcs in front of me but it was something in the low 80’s F for average loop temp.  Then higher for small zones. 

The head calc I used was 3.7ft per 100ft tubing for 1/2” pipe.  So for a 300ft loop, 300*0.037=11.1.  Then multiplied that by 1.5 for estimated total with fittings, etc.  

This was from an HRAI design manual.  Seemed a bit strange, very basic rule of thumb value. Is the uponor table pretty well industry standard, or should I go with whatever specific pipe I have locally?  (Which might be uponor…or not lol).  I planned to calculate actual fitting loss afterwards, just used the 17ft for initial cost estimates of different system elements….

hot_rod

You pressure drop numbers sound a bit high. .8 gpm is the high side of a 300' loop flow rate, .65 gpm/loop is more typical.

I use the RadPad slide rule calculator for quick checks. probably a bit conservative in it's results.

It shows a 300' loop at .8 gpm = 6.8ft head.

Circulators have a lot of adjustment these days, but no sense in buying more pump than you need.

arcticcontrolsgoose

Exactly, want to size appropriately.  

Higher flow rate is probably due to my lower than average design zone temp in this instance? 

Thanks for the sanity check fyi, neat slide calculator. 

hot_rod

arcticcontrolsgoose said:

Exactly, want to size appropriately.  

Higher flow rate is probably due to my lower than average design zone temp in this instance? 

Thanks for the sanity check fyi, neat slide calculator. 

The SWT shouldn't have anything to do with the pressure drop of the circuit. That is all about the resistance of all the tube and components in the loop. The circulator really doesn't know or care about temperature, it just want to move the water along.

When you size a circulator, it doesn't ask you for water temperature. Type of fluid, gpm and head of the circuit.

The surface temperature is what moves the heat energy, a simple rule for radiant floors is about
2btu/ sq foot for every degree difference between ambient room and floor surface.

For example if the floor surface was 82F and the ambient air was 70F. 82-70=12 x 2 btu= 24 btu/ sq ft output.

The 80F you refer to is that floor temperature or supply temperature?
Average loop temperature is not what the floor surface will be, you need to account for the concrete, floor covering, or subfloor and final covering if it is a transfer plate install

arcticcontrolsgoose

Ya, my apologies I was mistaken in my last post.  Looking back my calculated gpm came from this equation {Q = heat loss / 60*Df*Cf*dT}. In my instance I’m using a glycol mix, so not your regular ~500*dT for the denominator.  

I was just trying to grasp the broad strokes of why mine would’ve come out to higher flow design, but I should’ve looked back and realized the discrepancy is mainly the glycol mix.  0.6gpm/loop would be the value for the 8-loop bay on water, which is close to your prediction.  My theory (from memory of all the calcs) was it came to higher flow rate to reduce said heat transfer since calling for lower temps.  But really it’s the dT of supply and return that drives the flow of any given fluid. Somewhat on track, just didn’t have anything to do with the specific average fluid temp of my loop. 

The 80F (which is actually 83 from calcs) was average fluid, with surface being 70.   Concrete floor and 12” spacing. dT = 10. My values for flux were all over the place for the zones.  This bay was 10.5 btu/sqft given 23465 BTUH and 2240 sqft. This area is somewhat well insulated with smaller bays adjoining.  Not much of it exposed to unconditioned air.  Hope the heat loss is right or this is all a wash.  

Been very interesting learning this, appreciation given to well designed and implemented systems…balancing act to the extreme. Anywho, this is a bit side tracked from my circulator question.  You’ve found the mistake in my loop resistance values, which is good.  Needed that.  I like rule of thumbs for sanity checks but also just like to dig to find exacts and why these values are used…and if they are over or under compensating. 

hot_rod

Correct, glycol throws a wrench in the calcs. It is tougher to pump when cold, especially a 50%. The program probably adds the correction factor.
40% should be adequate for most application, it varies from brand to brand, the freeze protection %

70F surface at demand condition? What is the calculated room temperature, 65F? 70-65 X 2=10 btu/sq ft.
70F surface will not transfer much heat to a 70F space.

arcticcontrolsgoose

Haha no doubt re 70vs70.  

Yes, it’s 65F design temp. It’s a maintenance shop on a farm, on propane.  Not needing to go through 1500gal+ of propane each season, so the design is lower in the main areas and the few offices are at usual comfort levels. I’m hoping it doesn’t cause too many issues with the physical running of the plant; ie too low of load…  

Really, anything above 45 is adequate for the need - just enough to keep things from freezing.  Zone 6, 4800 HDD area with -25 as seasonal 99% temp.  I used 65 for design to have some resemblance of a load…”worst case” if you will.   

I’m not using a program, just good old fashioned excel. And a trial of loopcad to get the layout (not on auto cause that seems glitchy).