Gas Boiler Distribution Pipe Sizing

Reply

  #1  
Old 12-12-12, 07:29 PM
D
Member
Thread Starter
Join Date: Dec 2012
Location: United States
Posts: 9
Received 0 Votes on 0 Posts
Gas Boiler Distribution Pipe Sizing

Weil McClain Gas Boiler CGM 140,000 btu
Existing system has 2 zones each with circulators

I have temporarily added a radiant zone which is added on the end of the 1[SUP]st[/SUP] floor loop (as an initial - worked out nicely as far as transferring warm thru the floor, it’s a staple up app application).

I want to make the radiant lop a separate loop w/ circ pump.

My Question related to the primary piping out of the boiler.
Existing condition: is 1-1/4” immmediatly reduced down to 1” black iron pipe to the air scoop. 1” inch copper to the 1st branch (which is ¾” – baseboards), then ¾” copper to the second branch (which is also ¾” – baseboards).

The radiant loop is 5/8” pex. Right now its only 1 loop. But I plan to add a manifold to feed 2 – 5/8” loops.

I wondering if the initial piping out of the boiler should be upgraded to 1-1/4 pipe to feed 1st and 2[SUP]nd[/SUP] branches, then use 1” to feed radiant (which will ultimately feed 2 -5/8 pex loops – maybe even more)

I realize this thinking is more along the lines of how a domestic water system would be piped which is supplied under a constant pressure.

I understand that pipe size in a closed loop hydronic heating system is sized based on principals of thermodynamics and heat transfer.

The heat loss analysis for the house = 80,000 btus (at 20 deg drop, flow rate =8-9 gpm
1[SUP]st[/SUP] Floor (which includes all radiant loops) = 50,000 btus (at 20 drop, flow rate = 5-6 gpm)

If the pipes are based on flow velocities (2 to 4 ft per second)
Main pipe supply from boiler is be 1” for 9 gpm
1[SUP]st[/SUP] floor supply is ¾” for 6 gpm
2nd floor supply is 3/4" for 5 gpm
According to this the existing system (1" out of boiler) is OKAY to support everything.

The issue that I'm concerned about is that: the main pipe from the boiler is sized based on heat transfer, the branch loop pipe sizes are sized based on heat transfer.

It seems to me that their needs to be a double check (purely based on flow capacity) that the main boiler piping (from where the individual branches return and join together at the boiler to the point where the individual branches separate to supply each of the respective branches) can provide the required flow.

If the:
1st floor baseboard loop is pumping at 6 gpm.
1st floor radiant loop is pumping at 3 gpm and
the 2nd floor baseboard is pumping at 5 gpm

The main distribution piping needs to be able to handle 14 gpm

Is the assumption, that the individual circulator pumps will combine forces to increase the velocity in the main distribution pipe to provide the required flow? (closed loop system)

At 14 gpm the flow in 1" main pipe out of the boiler is 5.7 ft/sec.
Even though a 1" pipe can transmitt the required 80.000 btu's for the whole house, should the 1" pipe be upgraded to 1-1/4" keep the velocity in the main pipe less than 4 ft/sec??

Or is there a different design methodology for the main distribution piping???

Thanks for any assistance

daveb781
 
Sponsored Links
  #2  
Old 12-12-12, 08:20 PM
T
Member
Join Date: Jul 2009
Location: Toronto
Posts: 1,174
Received 0 Votes on 0 Posts
If it is easy to upgrade, I would do it.

I don't think you will get much erosion in the pipe at that velocity nor noise.
Any noise you might get should be contained in the mech room
 
  #3  
Old 12-12-12, 08:28 PM
NJT's Avatar
NJT
NJT is offline
Member
Join Date: Oct 2006
Location: USA
Posts: 23,539
Received 3 Votes on 2 Posts
Hi Dave, I'm getting ready to cut out for the evening but will respond a bit more tomorrow...

I wanted to get some clarification on something though...

Your radiant loop; do I understand correctly that you are feeding that radiant with the same temperature water (up to as high as 180) as you are feeding the baseboards?

That's no good. You should never run more than 120 (and even that is 'pushing it' in some cases) water to a radiant loop. You need a mixing valve to temper the water down to an appropriate temperature for the radiant.

That's number 1...

I highly doubt you are getting that kind of flow rate in the loops.

In general, a 3/4" pipe loop should not run more than 4 GPM, or you will end up with 'velocity noise' in the system (should be silent), and piping will begin to erode from the inside out and pinhole leaks will begin to appear over time (years usually)

So I think you are over estimating the flow in the loops...

More tomorrow.
 
  #4  
Old 12-13-12, 04:19 AM
T
Member
Join Date: Jul 2009
Location: Toronto
Posts: 1,174
Received 0 Votes on 0 Posts
Troop,

He is running staple up. He can run a high water temp thru that. But 180 is rather high, it would respond fast though :-)
 
  #5  
Old 12-13-12, 05:55 AM
NJT's Avatar
NJT
NJT is offline
Member
Join Date: Oct 2006
Location: USA
Posts: 23,539
Received 3 Votes on 2 Posts
TO, ? , I dunno... pushing to the MAX I think maybe 140... but that's still too much IMHO.

Let's check with Dave though that he described his system fully...

Dave, please define what you mean by 'staple up' ... no plates, right, just tubing on the bottom of the floor?

TO, you aren't confusing with the suspended tubing in the joist cavity, are you?
 
  #6  
Old 12-13-12, 07:25 PM
D
Member
Thread Starter
Join Date: Dec 2012
Location: United States
Posts: 9
Received 0 Votes on 0 Posts
Gas Boiler Pipe Sizing

Currently, the 5/8” pex radiant loop is held to the underside of the 1[SUP]st[/SUP] floor kitchen with ¾”pipe straps. I cut tightly fitting 2” rigid insulation between the floor joist and then covered the rigid insulation with R-19. No heat transfer plates were used. The joists are 24” o.c. so I was able to place 3 runs of pex between each sat of joists. The 1[SUP]st[/SUP] floor consists of tile over 2 courses of 5/8” plywood.

The radiant loop is piggybacked onto the end of the 1st floor loop. Based on a few temperature gages that I installed through the system, highest temp in the pex is 160. I realize this is pushing it a bit. But I wasn’t sure how well the staple up would transmit throught the floor – so this started as a bit of an experiment. The kitchen has no baseboards or other heat emitters – so the purpose of this experiment is to “help” take the chill off kitchen. With no other emitters in the kitch, the floor is pretty chilly. The kitchen is surrounded by rooms with baseboard.

I realize I’m trying something a little unconventional – but my goal is “take the curse of the cool kitchen floor without a huge investment.

So the experiment began – right or wrong here was the logic –

Throw the radiant loop on the end of the1st floor loop - this will put slightly lower water temp (after the 1[SUP]st[/SUP] flr baseboards scrub some heat out) thru the pex. The TACO 007 really can’t handle the extra headloss in the pex. So for $80, I spliced in a Grunfos 15-58 3 speed pump in line to push the rest of the way thru the pex (I know –2 pumps in series – a bit unconventional). See how it works, then possibly go to next step. Its actually works very nicely – floors are pretty evenly warmed. My wife even says the kitchen is significantly warmer. Granted the radiant cycles on& off with the 1[SUP]st[/SUP] floor thermostat (I know - not the preferred radiant setup).

Bottom line – for $150 in pex and $100 pump – I’ve noticed a dramtic improvement of the comfort in the kitchen. Additional running on the boiler is slightly more, but I really didn’t want to start the project by running the boiler more often (including during night time hours) to heat water to 180, just so I could cool it down with a mixing valve.

I realize I might wind up with the mixing valve setup with the pex on a separate loop – but again, I wasn’t so sure how well the thing was going to work in the first place.

Thanks for your inputs
 
  #7  
Old 12-14-12, 03:06 AM
rbeck's Avatar
Member
Join Date: Jul 2007
Posts: 2,482
Received 5 Votes on 4 Posts
Just a few things I see might be a problem long term.
Are the pipe fasteners metal? If so, in a few years how many leaks will you have as pex really expands and contracts a lot.
Secondly, did you leave a two inch airspace between the floor and the insulation?
Thirdly, you probably did not need the second pump. A Taco 007 probably would have handled the flow OK. What is the total run in feet of pipe do you have on the first floor from the boiler around the loop and back to the boiler? Is the radiant tubing on a manifold?
What size boiler? The maximum flow through a boiler is a 20f delta-T. So a 100,000 btu boiler can move 10 gpm. Try to run faster and you will pick up less temp and a better chance of condensing if the boiler is cast iron or steel.
 
  #8  
Old 12-14-12, 04:58 PM
D
Member
Thread Starter
Join Date: Dec 2012
Location: United States
Posts: 9
Received 0 Votes on 0 Posts
Pipe Sizes

  1. Used plastic pipe clamps (like used for running domestic water)
  2. Left 1-1/2 air space between rigid insulation and underside of floor (used spacer blocks to keep uniform space), When I said tightly fitting – I meant “between the joists”
  3. As far as the TACO 007 - I calced though at 5-6 gpm thru 200 (equivalent feet - with fittings, check valves etc) I get 7.4 to 10.4 ft of headloss, which is right on target with the 007, but when I piggybacked 250 ft og 5/6" pex I get an another 14 ft of headloss - that's why I added the Grunfos 15-55.

    Now I realize if I put the PEX on a separate loop, the 007 would be fine - but I I figured in either senario (piggy back or separate loop) I needed a another pump

Right now with PEX piggybacked – no radiant manifold. But the radiant seems to work so well, my next move is to add another loop in the entry foyer – so as a result- I would be adding a 2 zone manifold. That is the maximum I plan to add

Quick note on flow rate - as in earlier post heatloss for 1st floor is 50,000-60000 btu's - based on 20 deg delta- that's how I wound up at 5-6 gpm.

Whole house is 80000 to 90000 btus


I really appreciate all the feedback ....

But I'm back to my earlier question about the whether to upgrade the pipe size (currently 1") from the boiler to where the branches separate.

An earlier post mentioned – if its’ easy to access do it. Which I kinda agree with – for price of 5-6 ft of 1-1/4 black iron pipe, an upgraded air scoop (1” to 1-1/4”) and a few copper fittings – why not, its not a big expense. To be honest – my biggest fear was unscrewing the 1-1/4” to 1” reducer at the boiler – its looks in good shape, but is 15 years old. Not looking to turn my radiant experiment into a boiler replacement, when I crack something trying to take apart fittings.

By the way Boiler is WM CGM series 140000 btu (115000 DOE, call it 98000 btu with 15% off for operating in a unheated basement).

I'm curious about getting a practical view on sizing the pipes
Pipes are sized only off heat transfer principals btu’s = 500 * Flow Rate * Delta Temp (for a heating fluid=water) and preferred flow velocities. The reality is that baseboards generally get ¾” pipe. Pumps pump water along the respective pump curve which is a function of headloss. So what finally happens inside the boiler/heating system may not necessarily match up with the mathematical equations.

So that’s why my gut was questioning if a 1-inch main pipe could adequately supply 2-¾” branches plus what will eventually be 2-5/8 pex runs (manifold feeding 1 at 250 ft, plus the newer second run will probably be 150 ft).

I suppose I could just splice on ¾” copper to the existing 1-inch feed. Run ¾” to the new radiant manifold to supply 2-5/8” pex. From what I’ve heard as feedback the only drawback of an undersized supply line is excess velocity which could lead to pipe erosion.

Sorry to be long winded, but as an practical engineer, I don’t like to blindly follow the equations- it can get you in just as much trouble sometimes, as when a tradesman designs by rule of thumb. Don’t get me started on the limitations of heatloss analysis in a residential house. The secret is that fine line between analysis and what works in practice
 
  #9  
Old 12-14-12, 05:36 PM
NJT's Avatar
NJT
NJT is offline
Member
Join Date: Oct 2006
Location: USA
Posts: 23,539
Received 3 Votes on 2 Posts
Pumps pump water along the respective pump curve which is a function of headloss. So what finally happens inside the boiler/heating system may not necessarily match up with the mathematical equations.
True...

And I'm sure that you are aware that the 'operating point' of any pump will be at the intersection of the pump curve AND the 'system curve', yes?

In other words, you can't really get a completely accurate picture of system flow by doing a rough calc of the headloss of a circuit and then saying that it falls at this point or that point on the curve.

You've got TWO moving 'targets' to reconcile.

I wonder have you seen Siggy's article about the above subject? If not, I'll see if I can find it again...

Found it, this should give yer slide rule something to chew on!

Determining Flow Rates in Parallel Piping Systems Constructed of Smooth Tubing - Archives - PMEngineer

I built a spreadsheet with these formulas in it that I would share if I could find it. I suspect that it got lost in a recent HD replacement, but I'll look and share if I can find it.

Dave, have you actually measured the delta T and know it to be 20°?
 
  #10  
Old 12-14-12, 05:43 PM
NJT's Avatar
NJT
NJT is offline
Member
Join Date: Oct 2006
Location: USA
Posts: 23,539
Received 3 Votes on 2 Posts
Pump flow rate can be empirically measured by means of a pressure gauge on both suction and discharge as close to pump as physically possible.
 
  #11  
Old 12-14-12, 06:37 PM
D
Member
Thread Starter
Join Date: Dec 2012
Location: United States
Posts: 9
Received 0 Votes on 0 Posts
right now .. with 250 ft pex piggybacked I'm measuring 25 deg delta +/-.

Now keep in mind this will all change if this radiant loop gets reconfigured to independent loop - I'll also be running hotter water thru the pex ... I know the next comment ... I'm 1 step away from the mixing valve , which I'm fine with.

In the end, I'm just trying to get to the most economical way of powering the radiant loops - preferrably without sending gas consumption thru the roof by running the boiler constantly to power the radiant.

Once I decide on the how to finish building the basics - I think it will come down to how to tweak the controls - maybe moving the thermostat might help achieve that sweet spot. If I put the thermostat in the kitchen - I think i'll be overheating the rest of the 1st floor.
 
  #12  
Old 12-14-12, 06:45 PM
NJT's Avatar
NJT
NJT is offline
Member
Join Date: Oct 2006
Location: USA
Posts: 23,539
Received 3 Votes on 2 Posts
measuring 25 deg delta +/-.
Using what tools? IR ? clamp on thermometers?

IR devices can be notoriously inaccurate in their measurements, depending on the emissivity of the object being measured, but even so, if you measure the same material on both ends, the delta should be somewhat accurate.

If you are using IR and you have a case on your boiler where you can meausre a rusty old cast fitting and a copper pipe right next to it, check out the difference between the two materials.

Infrared Thermometer

Emissitivity of Common Materials

Being a enjinear and all, I think you know this, but just in case not...

Oh... don't try to measure the temperature of the sun!
 
  #13  
Old 12-14-12, 06:55 PM
D
Member
Thread Starter
Join Date: Dec 2012
Location: United States
Posts: 9
Received 0 Votes on 0 Posts
Wow ... read that article you suggested ... it hits that nail on the head. I'm trying to analyze what happens in that short section of "common piping" - the analysis gets hairy - so that's why - I'm doing what I've done - when the engineer realizes he is trying to analyze something so complicated, with so many variables - you go ask the guys that install the stuff every day (and more importantly, see the problems that arise from the wrong assumptions).
 
  #14  
Old 12-14-12, 07:13 PM
D
Member
Thread Starter
Join Date: Dec 2012
Location: United States
Posts: 9
Received 0 Votes on 0 Posts
IR .. really? not a chance

spliced in temp gages on tee fittings
 
  #15  
Old 12-14-12, 07:40 PM
NJT's Avatar
NJT
NJT is offline
Member
Join Date: Oct 2006
Location: USA
Posts: 23,539
Received 3 Votes on 2 Posts
IR .. really? not a chance
Ha! I knew that already!

Glad you liked that article... most of the guys that do it every day know this stuff only intuitively though. Them ciphers in that article are cornfuzin! Thanks to calculators and computers though, well within reach of most, if they try. "...spiral out, keep going..."
 
Reply
Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Thread Tools
Search this Thread
 
Ask a Question
Question Title:
Description:
Your question will be posted in: