Choice- Lochinvar WBN051 or Cadet CDN040?

Eastman

what formula?

?



Holy smokes, didn't realize that about the vv circs!

ced48

100' Run x 1.5 x.04 = 6' Head

This formula is for 3/4" pipe-

Eastman

That's at roughly 6gpm

That's at about 6 gallons per minute.  Head loss is an exponential function of fluid flow.  It is not the hydraulic resistance.



www.taco-hvac.com/uploads/FileLibrary/SelectingCirculators.pdf



Head loss = k * c * L * (f^1.75)



k and c are constants that depend on the diameter of the pipe and type of fluid.  L is the length of pipe.  f is the the flow rate and as you can see it varies exponentially with gpm.  The product (k * c * L) is what is typically considered to be the hydraulic resistance and assumed to be constant.  The head loss would vary --for example if you installed a variable speed pump, it would go up and down in relation to changes in fluid flow.  You are in the design phase here:  pick a target flow rate, estimate the head loss at the target flow rate, and then go and try to find a pump that can meet or exceed your requirements.  Look over the Taco tech pdf above to get an over view of this process.



The rule of thumb you are using gives one an idea of the head loss in X feet of 3/4 at the maximum recommended velocity of 4 feet per second for copper.  (Copper suffers from a phenomenon called erosion-corrosion at high fluid velocities.)  That's about 6 gallons per minute.  What is the head loss estimate at your target flow?





Edit:  Caught a typo ced!



Head loss = k * c * L * (f^1.75)    that's f to the 1.75 power

ced48

Okay Eastman, I Think I Tried it Your Way?

I used this formula I got from Taco, and you, HL = k x c x L x (f1.75) and I come up with about .5 feet of head loss at the target flow, 1.5 GPM per circuit, 3 GPM, total. Sound right? I will probably design at 2 GPM per circuit, 4 GPM, total. When I add in things like a Flow-chek valve, I will have to adjust. So having said all of that, how can I possibly need such a tiny circulator, still must be doing something wrong!

Eastman

Sounds about right, but...

did you see my typo?  It's f to the 1.75 power.  (f^1.75)  Not f times 1.75. 



Here's an example head loss calc for 100 feet of 3/4 @ 1.5gpm  --I rounded off some of the digits.



H loss = k x c x L x (f^1.75) = 0.003 x 1 x 100 x (1.5^1.75) = 0.3 x 2 = 0.6 feet of head loss



Now, without knowing all the elbows and tees etc used, most people would use 150 feet of equivalent straight copper as the starting point.  That works out to a head loss of 0.9 feet.



The total head loss is the sum of this value with the head loss through the boiler and supply and return lines at 3gpm.  (And whatever else you add to the system.)  The pump performance would have to exceed total head loss at 3gpm.



But honestly, without knowing the particulars of your home, I would have to advise you to design with a smaller delta t across the baseboard.  15k btu of baseboard in series can heat unevenly if it spans multiple rooms, particularly if there's a big window or some other major heat loss in the room at the tale end of loop.  With a 10 degree delta your target flow would be 3 gallons in each loop; 6 though the boiler.  3 gpm in the baseboard loop is around 2 to 3 feet, 6gpm though the firetube boiler is something under 1 foot.  So you're looking at a total of 4 feet or less at 6gpm through the pump.  If you look at the bumblebee's fixed power (fixed speed) pumping curves, you should see that it happens do 6gpm @ 3.75 feet on a speed setting of 2.  And it will only use 22 watts.  So we should expect a flow of around 6 gpm if you used this pump at speed 2.  If our estimates are off a bit, you could set it to something else to achieve more or less flow, reduce system delta, or save a few watts.  In summary, most systems are way over pumped and large delta t's may be uncomfortable in your home.

ced48

For Those Following This Insanity-

we have forgotten that the common piping is 1", not 3/4". To review, the two circuits are piped 3/4", but their common supply- return is 1". So if the flow was 3GPM in the 1", flow in the 3/4" would be 80 percent , or 2.4 GPM.

Eastman

Double check

I understood the situation like this:  You have a 1 inch supply from the boiler.  This tees off to 3/4 on one side of the house and another to the other side.  If there is a 3gpm supply in the 1 inch, logic dictates that the sum of the flows in the 3/4 must also add to 3gpm.  If the flow is split evenly, that would be 1.5 + 1.5 = 3.

ced48

I Think You are Right

in calculating required flow for btu's, you would divide the total by two, in a system with 2 equal circuits. But I think the physical flow would be calculated differently. If we were dividing a 1" pipe in half, the flows would be half, but if we split the 1" pipe into two, 3/4" pipes, i think the flow rate would drop by only a small percentage.

Eastman

It sounds like...

It sounds like you are trying to condense the problem down to a single head loss calculation.  Something like: What is my HL at F gpm for X feet of 3/4 copper equivalent length.  And then proceeded to try to come up with something for F and X so the Head loss could be computed.



Looking at the Taco pdf, I can see why someone would think that.  The short discussion on parallel circuits really requires some more context.



You need to do 3 head loss calcs:  1 for one of the baseboard loops, 1 for the common supply and return, and 1 for the boiler.  And then add them all together to get the total head loss.  Now, the boiler manual already has the head loss for the boiler at a variety of flows.  So we don't need to calc that one, just look it up in the table.



The head loss for the common supply and return needs to be calculated with the k value for 1 inch copper and a flow rate of 3 gpm.  ***I forgot to add this step in the last example, I hope I didn't add to the confusion.***  20 feet of 1 inch at 3gpm is about 0.1 feet of head.



The head loss for one of the base board loops needs to be calculated with the k value for 3/4 inch at flow rate of 1/2 of the 1 inch.  (1.5 gallons).



The total head loss is the sum:  H total = H baseboard + H common pipe + H boiler.



Drawing on some of our previous work for 1.5 / 3 gpm target flow:



H total = 0.6 + 0.1 + 0.15 ish (have to look at the boiler table)



for a grand total of under 1 foot.  1.5 feet if you add a 50% smudge factor.  For a 3 / 6 gpm target flow it works out to be about 4 feet of head.



Why is it so low?  Half of it is because you spec'd a firetube hx, the other reason is because you have a very small load that is piped in 3/4 and 1 inch.  You're not dealing with 100's and 100's of feet of small diameter pex tubing.  If you wanted to use a water tube hx, the heads are much higher.  Still doable, but you might be forced to run at relatively high delta Ts that may heat the home unevenly.

Paul48

Just Wondering- Taco?

How can 2 different formulas for estimating head loss (both provided by Taco), wind up with answers that far apart? At FloPro U., they caution that you could over-size by as much as 50%, BUT 500%!!. It would seem...Hydronics Step by Step, becomes Mis-step by Mis-step.

ced48

100' Run x 1.5 x.04 = 6' Head

After trying lots of different formulas, this simple one seems to agree with most of the others, as long as I plugged in all of the right numbers. I am going to stick with it. All my confusion stemmed from not calculating the length of the longest circuit, and basing the loss on it. Instead, I was basing everything on the entire loop, all 2 circuits. This was wrong.

Eastman

One is based on the other

The formula: 



Head loss = equivalent pipe length * 0.04



is assuming a fluid velocity of 3 ft/sec.  3 ft/sec in 3/4 copper corresponds to about 4.8 gpm.



The 0.04 value is 0.04 feet of head per foot.  It represents the head loss @ about 4.8 gpm for one foot of 3/4 copper tubing.



Does this seem reasonable?  Let's do a head loss calculation for 1 foot of 3/4 @ 4.8 gpm.  That would be:



Head loss = k * c * L * (f^1.75)  with k= 0.003, c= 1, L= 1, and f= 4.8

Head loss = 0.003 * (4.8^1.75)

Head loss = 0.046 feet



So with a fluid velocity of 3 ft/sec corresponding to a flow rate of about 4.8 gpm, we do get about 0.04 feet of head per foot.

Pughie1

HDS-2

Have you guys taken a look at Siggy's Hydronic Design Studio-2. It uses all your formula's and will give you an answer in 2 minutes.

                                                                   John Pughe

Paul48

yes

.046 x 150 = 6.9' You are solving for 1 ft

CMadatMe

One Thing Not Taken

Into consideration is a fixed speed pump is going to operate on it's curve no matter how down to the science you get the head loss.

Paul48

If

you've figured correctly, what is the down side to that?

Eastman

rule(s) of thumb

Here is that general rule recalculated for a variety of different fluid velocities for 3/4 copper tubing:



Head loss @ 1.0 ft/sec = equivalent pipe length * 0.0069

Head loss @ 1.5 ft/sec = equivalent pipe length * 0.014

Head loss @ 2.0 ft/sec = equivalent pipe length * 0.023

Head loss @ 2.5 ft/sec = equivalent pipe length * 0.034

Head loss @ 3.0 ft/sec = equivalent pipe length * 0.047

Head loss @ 3.5 ft/sec = equivalent pipe length * 0.062

Head loss @ 4.0 ft/sec = equivalent pipe length * 0.078



So the head loss for 100 feet of 3/4 tubing would range from 0.69 feet @ 1.0 ft/sec through 7.8 feet @ 4.0 ft/sec

Paul48

Pipe Sizing

As long as you're within pipe sizing guidelines for flow, those calcs are not necessary.

CMadatMe

Paul

I was referring to this situation. The basic formula of Run x 1.5 x. 04 gets you the same pump in the end..I'm sure you like me haven't seen too many 003 or 005 Taco circs installed out there. In this case the 003 would be the right pump at twice the cost or more then a 007. This is where a Bumble Bee would fit rather nicely.

Paul48

Agreed

But then the Bumblebee is twice the price. It gives you way more bang for your buck. Which brings us back to P/S piping and the ability to run a 10* DeltaT on the system side(as suggested), while staying with at least a 20 through the HX..