Circulation Pump Too Small?
We recently converted to a NTI Combi Boiler from a cast iron boiler upon purchasing this house built in 1925. Two story home with approx 2500 sq ft and 14 hot water radiators. The two pipe piping runs in a u shape with the boiler in the center of the U and going outward and is one zone for the entire house with the thermostat in the center hall first floor. The longest pipe run is about 75' across the house and up to the second floor, the second longest run is about 50'. There is one Taco07 pump on the system (supply line after the expansion tank). Radiators at the ends of the loop are luke warm while rads closer are overly hot. Installer has bled all rads and tried valve adjustments but no significant change is happening. I believe it is a circulation problem and that the pump is undersized and only getting the flow to go to the 4 or 5 rads closest to the boiler. I have to run the thermostat at 74+ to get any noticeable heat to rooms at the end of the pipe run.
Comments
-
take a picture of the boiler install please, and what model pump was removed from the old system?
1 -
You might need to severely restrict the flow to the radiators closest to the thermostat
Miss Hall's School service mechanic, greenhouse manager, teacher, dog walker and designated driver
2 -
What size is the pipe from the boiler to the system?
What is the model number of your NTI boiler?
Is this the pipe connections on the bottom of your boiler Fig. 10-2 (c)?
Edward Young Retired
After you make that expensive repair and you still have the same problem, What will you check next?
0 -
0
-
We live in a 100-year-old building with original cast iron radiators and a converted gravity hot water system. We have 11 radiators on 2 floors, 2400 sq ft, heated by a cast iron boiler. There's one Taco 007 circulator that does fine. Because the pipes are so big, there is very little flow restriction, and the 007 was pumping 19 GPM before I installed a controller to turn the flow rate down.
So it is possible to run a significant number of cast iron radiators off a single Taco 007, but only if you have big pipes and a big boiler. If you do have the original big pipes, I'm guessing the problem is the flow restriction through your new smaller boiler.
0 -
I think the issue is the lack of flow control at the radiators which can be addressed by TRV at the radators.
I have a similarly configured heating system on 1928 home. The main supply / return lines are 2” with radiator pipe parallel series. The first floor radiators are serviced with 1” pipe, second level is 3/4” pipe.
The boiler is a low head FireTube with a single alpha 2 15 series circulator.Works perfect running at low speed.
If the NTI has high head, you will need to pipe primary secondary and will need pump on both.1 -
We have similar pipe sizes, 2-1/2" near the boiler, then stepping down to 2" and 1-1/2", then to 1" risers for the first floor and 3/4" for the second floor.
But we don't need any TRV's, etc. The system is perfectly balanced. Not everyone is as fortunate, especially if their original old piping has been modified.
2 -
Agree with @PC7060 mod con boiler are restrictive to flow by design. If the Taco 07 is the only circulator in the entire system that could be your issue.
Not familiar with NTI and weather that boiler has its own internal circulator or requires primary secondary piping but that could be your fix if it does not have an internal circulator and requires primary secondary piping it has to be installed in accordance with the MFG instructions to work.
What is the model of your boiler NTI makes many.
4 -
Sounds like a balancing issue more than anything, but we'd need to see the piping to determine this for sure.
0 -
you can direct pipe some of the firetube models, everybody has them now. The manual may show you the available pumping capacity, beyond the boiler
Hydraulic separation is always an option, buy you get some temperature blending as flows vary
NTI, FTN model piping exampleBob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream0 -
Thank you all for the responses. I'm adding photos of the boiler and piping near it.
Installer is coming back tomorrow to look into it again and NTI suggested changing the modulation control - possible to constant flow instead of ramping up and down.
Turning some radiators closer to to boiler to 1/2 to 3/4 closed hasn't made any noticeable changes.
0 -
NTI Boiler is Model TRX199C
0 -
Have you checked the BoilerMag to see if it's sludged up? I know you said this is a recent install, and if the BoilerMag is partially blocked with all the crud that built up in your old radiators and pipes, that could reduce your circulation.
https://www.boilermag.com/na/another-boilermag-saves-the-day/
0 -
It's been checked and cleaned for fall maintenance and on two service calls since.
0 -
I don't know what the previous pump was but I think it was a similar size and the heat was balanced. It could be the modulation feature on this boiler and ramping up and down before the full loop is satisfied. Where is your thermostat located? Mine is first floor center hall which catches wind when doors open and other micro changes. May move the thermostat to a room with less fluctuations and get a modulation sensor thermostat from NTI.
0 -
My thermostat is in the dining room, not near the front entry. However, when I first moved here the boiler was short cycling, and it took me a long time to figure out the several factors involved. The sneakiest one was that the thermostat was mounted in front of a pipe chase hidden inside the wall…the same chase containing the radiator riser pipes to the second floor. So when the heat came on, the wall behind the thermostat would heat up and cause the thermostat to end the heat call prematurely.
Once I found that, I moved the thermostat to an adjacent wall and re-routed the wires up from the basement. I also found a cold updraft in the wall from the basement, and had to foam the wire hole to stop the draft.
The other thing I highly recommend is the ecobee smart thermostat. I got two last January, and they have been instrumental in troubleshooting our heating system. You can see when and for how long the thermostat is calling for heat, download the data, etc. And when you make changes to your system, the ecobee data will tell you what the effect was on your fuel consumption, etc.
2 -
EDIT AT TOP:
Your boiler has an internal pump and your boiler is connected to the system with closely spaced tees. Your system pump does not need to deal with the boiler's restriction. The internal pump handles that. System pump is labeled CH for Central Heat as opposed to DHW and Boiler pumps.
Here is your problem. Just using a Rule of Thumb calculation I find that your pump is not powerful enough to move all the heat from your boiler to the entire system.
The way I like to explain this is: Measure the longest run (if the pump can do the longest run it is powerful enough for all the smaller runs. in your original statement above you said:
"The longest pipe run is about 75' across the house and up to the second floor"
To me, that means that that there is about 75 feet of strait pipe to get from the boiler to the farthest radiator. So it is probably going to be another 75 feet back to the boiler from the farthest radiator. That is 150 feet of strait pipe. This rule of thumb states that you should add 50% more piping for the equivalent length of all the fittings between the boiler and the radiator. So the math for that is 75 + 75 = 150. add 50 % of 150 to that number and you get 150 x 1.5 = 225 total equivalent feet. Then you multiply 225 by .04 to get the pump head. That would be 9 feet of pump head.
Now if we look at the performance curve of the Taco 007 you can clearly see that there is almost no gallon per minute (GPM) flow rate at 9 ft head. As you move down the curve to a 10 GPM That is equal to about 100,000 BTU of heat @ a 20°∆T
Your boiler has a NET BTU capacity of 160,000. That would require 16 GPM @ 20°∆T
If you look at the Taco 007 performance chart the total pump heat at 16 GPM is about one foot of pump head.
So you have a pump that can do 9 FT heat and 16 GPM. but not at the same time… WRONG PUMP. to paraphrase Chief Brody. “You’re gonna need a bigger PUMP”
You may need to use a Taco 0012 pump to get what you need. I believe it also comes in a ECM model for more $$$$. A 0010 might be enough but to be sure get a 0012. It all depends on how many BTUh your building requires. 160,000 BTUh = 16 GPM. 140,000 BTUh = 14 GPM
If you look at the 16 GPM @ 9ft head it is clearly above the 007 curve (#5) but is is clearly under the 0012 curve. (#10 in yellow highlight)
Edward Young Retired
After you make that expensive repair and you still have the same problem, What will you check next?
0 -
Do you usually add 50% to the length of pipe one way or 50% to the total run up and back??
0 -
@EdTheHeaterMan what is your opinion on the near boiler piping? For a 160,000 BTU boiler wouldn't larger pipe be more appropriate to move more heat? That appears to be 1" copper, wouldn't 1-1/4" be better?
1 -
So you have a pump that can do 9 FT heat and 16 GPM. but not at the same time… WRONG PUMP. to paraphrase Chief Brody. “
You’re gonna need a bigger PUMP”If the OP can live with 12 GPM, the headloss on 200 feet of 1.25" steel pipe would drop to 5.7 feet. This is well within the capability of the 007. The 20 DT is not mandatory and the need for 160K is really uncertain (and NOT LIKELY).
Chief Brody is in error.
0 -
If you must move 16 GPM through 1" you're looking at a velocity of 6 FPS which is really ridiculous. Now you certainly need a bigger pump.
If you stay to 12 GPM, the loss in 10' of 1" is less than 1 foot.
0 -
One of DanH's earlier books was actually written for B&G when he was a Manuf. Rep. This is it:
https://s3.amazonaws.com/s3.supplyhouse.com/product_files/108119-Reference%20Guide.pdf He does not get credit in the book because B&G paid him to write it. He didn't write it on his own.
When I taught classes for EH-CC.org I used it as my text book for Hydronics and modeled my class around the seminars I attended when Dan was still teaching. I took every class That Dan taught when in the Philadelphia area. That said, I have read this book at least 60 times over the last 30 years. It is my go to source for the basics. piping and pumping questions like this one.
Yes The total run. You take the total amount of straight pipe and multiply it by 1.5 to get the total equivalent length. It might be a little overkill, but not so much that you oversize the pump. I have had years of success doing it this way.
As far as the near boiler piping, I’m not sure that the total load is 160,000 BTUh All I know is that the boiler has that capacity. If the home only needs 140,000 BTUh or less then 1-¼” copper is fine @ 20°∆T. You won’t find out until you get close to design temperature if the total 160,000 is needed. This particular "not enough heat" problem is happening now. I don’t think that we are experiencing design temperatures right now. At least @Notanusername has not indicated that yet. So the problem is more likely the pump being too small. The pipe size may be ok for now.
And replacing the pump is easier that re-plumbing all the near boiler piping to 1-1/2" copper
Edward Young Retired
After you make that expensive repair and you still have the same problem, What will you check next?
0 -
The load on the radiation dictates as @hot_rod always correctly points out.
I almost never used the 50% add on. It is conservative and is a good rule of thumb. I usually count or estimate the number of fittings if they are not visible. Doesn't have to be perfect.
0 -
Nice piping work. A little insulation wouldn't hurt.
8.33 lbs./gal. x 60 min./hr. x 20°ΔT = 10,000 BTU's/hour
Two btu per sq ft for degree difference for a slab0 -
And replacing the pump is easier that re-plumbing all the near boiler piping to 1-1/2" copper
Definitely.
I'd still like to see the DT and fully understand what the boiler is doing during this observation. If the DT is 35 and the boiler is modulating down because the flow rate will not allow it to deliver 160K, you have your confirmation.
In the case of CI radiation, more is always better when you have a cold start. If you have 160K available, you might as well use it.
0 -
I just used @DanHolohan's text from the book I referenced.
It is easy to read and I have referred to it many times…
Some of you may have already read it, but for those of you that have never clicked on the link to that book, these are the paragraphs I am referencing.
Figuring Pump Head
So now that you’ve figured out your longest equivalent run, it’s time for another “Rule of Thumb”. Up to this point we’ve sized everything based on “The Law of theMaximum.” We want to get the maximum, trouble-free performance from the piping and components we select. That makes sense, doesn’t it? Let’s get the job done correctly, and at the same time, economically. That’s just plain good business. So when we consider Flow, we size the pipes for the most water that will flow through them without giving us velocity noise. It makes no sense to oversize pipe; that only adds unnecessary expense to the job. And it makes no sense to undersize pipe either because then we’ll get noise. So, based on the maximum flow that will flow quietly through our piping, we came up with this next “Rule of Thumb.”
Pump Head “Rule of Thumb”
If you’re sizing the pipe for the maximum flow rate that will fit through the pipe without making velocity noise, the required pump head will be 4 feet for every 100 feet of total equivalent length. So:
- Measure the Longest Run in Feet
- Add 50% To This
- Multiply That By .04, and
- That’s the Pump Head!
In the example we were looking at before, we came up with a Total Equivalent Length of 300 feet (200 feet of actual run, plus 50%). Our“Rule of Thumb”, tells us that the circulator we select for that job would have a Pump Head of 12 feet.Get it? We’re allowing four feet of Pump Head for every hundred feet of equivalent run. There are three hundred feet of equivalent run, so what we’re doing is multiplying 4 (feet per hundred) x 3 (hundreds) to get12 feet of Pump Head. We saved an arithmetic step by using the .04 factor instead of dividing everything by 100 to get to feet of Pump Head. Nothing to it!
This has worked for me just fine and according to this rule of thumb formula, the 007 is not powerful enough for this system. If it was powerful enough for this system, @DerekLH would not be asking about the cold radiators.
I then ask myself, "After I spend time and money on increasing the pipe size of all that copper near the boiler, and I still have the same problem, What will I check next?" I might take 15 minutes to swap out a pump and just see what happens. Especially after measuring the pump head as I have done for the last 30 years. But I could be wrong. Only @DerekLH can tell by the next repair they choose to make.
Edward Young Retired
After you make that expensive repair and you still have the same problem, What will you check next?
0 -
If
The OP has already got 1-1/4" copper for near boiler piping, and then larger iron pipes for the system. I don't think the pipe size is the problem. and you state that "IF the OP can live with 12 GPM", but they have a 160,000 output boiler, we can't know that they can live with 12 GPM. We also don't know the ∆T of the system. That information is not given either. What we do know is that the radiators are not getting hot enough the farther away from the boiler you go. So that ∆T on those under heating radiators is less than 20° All the heat is going to the closer radiators. I will bet my reputation on the larger pump to solve the problem.
Edward Young Retired
After you make that expensive repair and you still have the same problem, What will you check next?
1 -
@EBEBRATT-Ed said:
I almost never used the 50% add on. It is conservative and is a good rule of thumb. I usually count or estimate the number of fittings if they are not visible. Doesn't have to be perfect.
As one data point on the Taco 007, we have a 100-year-old converted gravity system with all original large-diameter piping (except for the 1-1/4" copper near-boiler piping).
I've estimated our head loss a number of different ways. By Dan's method,
Longest run = 80 feet
Add 50% = 40 feet
Total = 120 feet
Multiply by .04 = 4.8 feet head
From the Taco 007 curve, that gives about 15 gpm.
Based on our known boiler net output and measured delta T, I calculated an actual flow rate of about 19 gpm, which from the Taco 007 curve implies a head loss of about 3 ft.
So Dan's method over-estimates our head loss by about 50%, and if we did the calculation without adding his 50% factor, the result would be quite accurate for our system.
Frank "Steamhead" Wilsey posted the following guide on sizing circulators for old gravity hot water systems about 10 years ago, in which he said to use 3.5 feet of head as the design point for such systems. At first I found it hard to believe that you could just assume 3.5 feet of head without counting up the hundreds of feet of piping, or their diameters, but interestingly, for our converted gravity system, Frank's 3.5 ft head "rule of thumb" turned out to be surprisingly accurate.
Having said that, I realize that in sizing a pump for a system with unknown head loss, it's best to be conservatice. I think Dan said in his book that when a heating contractor called his supply company looking for a replacement pump, Dan would ask the contractor for the dimensions of the house footprint, and then figure the longest possible pipe run in his head, add the 50%, and multiply by .04 to get the head loss, and when he sized a pump that way, without ever having seen the system it was serving, he never had a contractor return a pump.
1
Categories
- All Categories
- 86.5K THE MAIN WALL
- 3.1K A-C, Heat Pumps & Refrigeration
- 53 Biomass
- 423 Carbon Monoxide Awareness
- 94 Chimneys & Flues
- 2K Domestic Hot Water
- 5.5K Gas Heating
- 101 Geothermal
- 156 Indoor-Air Quality
- 3.5K Oil Heating
- 64 Pipe Deterioration
- 927 Plumbing
- 6.1K Radiant Heating
- 383 Solar
- 15.1K Strictly Steam
- 3.3K Thermostats and Controls
- 54 Water Quality
- 41 Industry Classes
- 48 Job Opportunities
- 17 Recall Announcements