Yanche's Toolbox, Piping & Circulator Selection Method
 Yanche
 Member
 Posts: 3030
 Joined: Fri. Dec. 23, 2005 12:45 pm
 Stoker Coal Boiler: Alternate Heating Systems S130
 Coal Size/Type: Anthracite Pea
 Location: Sykesville, Maryland
I’ve had several requests via private emails in the past several months for recommendations on pipe sizing and/or circulator selection. The request is often difficult to answer because not enough detail design information is given to provide an engineered answer. In this “Toolbox Note” I’ll describe my design procedure.
In a closed loop hydronic system the circulator pump only needs to overcome the hydraulic friction loss created by the piping and plumbing components. The design approach is to calculate this loss and then match a circulator pump that will overcome this loss and flow the fluid at a rate needed to meet the BTU load.
Fluids have been conveyed in pipes for many, many centuries, everything from water to poop. Analysis and design engineering techniques have improved considerably from the dawn of modern civilization. Hydronic heating piping is a very special case of full fluid filled pipe with smooth inside walls. This simplifies the design considerably. While many will here will consider my design procedures complex, in the engineering world it’s quite simple.
Common hydronic piping is copper, PEX and steel. Copper has the smoothest interior followed by PEX and steel. You likely haven’t thought about it but the rougher the inside wall of the pipe the more resistance to flow it has. As you will see later the effect is significant. Schedule 40/80 steel pipe has the most resistance because of the way it is made. It’s rolled from flat stock and then welded lengthwise. The seam makes an interior bump, something not present in copper or PEX.
The design procedures I’ll show are promoted by John Siegenthaler, P.E., in his books and various hydronic heating magazine articles. The formula below calculates the piping head loss in feet. It’s a specialized relationship derived from the widely used DarcyWeisbach equation. It only applies to hydronic heating type applications.
Even with this simplified formula there is some complexity. The fluid properties factor varies with fluid temperature. To simplify the equations’ use I have plotted the fluid properties factor(a) for water as a function of water temperature. Hotter water flows easier than cold water, so a conservative design procedure would be to use your design boiler return water temperature. Fluids other than pure water, i.e. some antifreeze mix would require you to calculate a new graph.
The pipe size coefficient is listed in the chart below. Just select the type and size pipe you are using. Pipe fittings and valves have a resistance to flow. For analysis purposes these fittings have resistances expressed in equivalent lengths of pipe. For example the resistance of a 3/4" copper 90 deg elbow is equivalent to 2’ of pipe.
Let’s do an example for a boiler located outside a heated home in a shed or garage. Assume the equivalent total piping (supply + return + fitting equivalent) is 160’ and we want to use 1” PEXALPEX. A 140 deg return water temperature gives a fluid properties factor of 0.0475. Plugging in the numbers gives:
This formula gives us the flow characteristics of our piping. The piping loss (feet of head) will be greater as we attempt to get more flow (f = GPM) through the pipe. The challenge is to find a circulator that will pump our water at the flow rate required to deliver the needed BTU.
A pump curve is provided by the pump manufacturer. It’s a plot of pump head vs. pump flow. The circulator will pump to over come the head resistance it sees. Said another way the head loss will equal the pumping head provided by the circulator. This is known as the intersection of the “pump curve” and the “system curve”. This intersection determines the fluid flow rate in the pipe.
Let’s take an example. The Taco 007 is a very common hydronic system circulator. It’s “pump curve” is given by the equation:
To find the flow rate f you set the two equations equal to each other:
and solve for the flow rate f. This can be solved by many methods, trial and error, pocket calculator, graphically, etc. A graphical solution is shown below. The curve intersection is approximately 9.6 GPM.
Is 9.6 GPM sufficient to flow the BTUs you need? Let’s find out. An approximation to the rate of heat transfer using water as the fluid is give by:
For a 20 degree temperature drop design our example heat flow would be:
Q = 500*9.6*20
Q = 96,000 BTU/hr.
This is likely too small a pump. Let’s try a Taco 0012
The equation for a Taco 0012 is:
Plotting the Taco 0012 on the system piping graph gives:
Reading the curve intersection shows the flow rate as approximately 11 GPM. Again for a 20 degree temperature drop design the heat flow would be:
Q = 500*11*20
Q = 110,000 BTU/hr.
Still not enough flow. Try another pump.
The equation for a Taco 0013 is:
Plotting the Taco 0013 on the system piping graph gives:
Reading the curve intersection shows the flow rate as approximately 17 GPM. Again for a 20 degree temperature drop design the heat flow would be:
Q = 500*17*20
Q = 170,000 BTU/hr.
Yes, it’s a winner. It can meet the heat flow requirement for a 130,000 BTU/hr. boiler.
There are other solutions. Increase the pipe size or change the pipe type. Copper pipe has less resistance, so with a given circulator copper of a similar size will flow more BTUs than PEXALPEX. Schedule 40/80 steel pipe will flow more BTUs than PEXALPEX but not as much as copper. As you have perhaps noted pipe coefficients are not listed for schedule 40/80 steel pipe. That’s because a different analysis technique is used. One that’s more complicated and difficult to calculate. What I do is use the copper based procedure to select a pipe size and if I want to pipe it in schedule 40/80 use the next larger pipe size.
To complete the example the student should repeat the design procedure using 1” or 11/4” copper pipe. It will show you can use a smaller pump and still get adequate heat flow. Long term, this would be the more economical solution. You are trading up front costs (more expensive pipe) for lower operating cost (less electric consumption). As you can see there can be an endless combination of choices that will work.
As the complexity of your piping system increases so does the analysis complexity. Parallel pipes split the flow and require more complex analysis. My intent here is to give you some simple analysis tools to select you piping choices wisely. Some straight forward analysis can save you money. In my own case when I first installed my boiler, I simply piped it in the same size as the tappings in my boiler. A much smaller pipe size would have done equally well. The larger pipe size was a needless expense. More design techniques are available in magazine articles authored by John Seigenthaler. Do a web search on his name to find them.
In a closed loop hydronic system the circulator pump only needs to overcome the hydraulic friction loss created by the piping and plumbing components. The design approach is to calculate this loss and then match a circulator pump that will overcome this loss and flow the fluid at a rate needed to meet the BTU load.
Fluids have been conveyed in pipes for many, many centuries, everything from water to poop. Analysis and design engineering techniques have improved considerably from the dawn of modern civilization. Hydronic heating piping is a very special case of full fluid filled pipe with smooth inside walls. This simplifies the design considerably. While many will here will consider my design procedures complex, in the engineering world it’s quite simple.
Common hydronic piping is copper, PEX and steel. Copper has the smoothest interior followed by PEX and steel. You likely haven’t thought about it but the rougher the inside wall of the pipe the more resistance to flow it has. As you will see later the effect is significant. Schedule 40/80 steel pipe has the most resistance because of the way it is made. It’s rolled from flat stock and then welded lengthwise. The seam makes an interior bump, something not present in copper or PEX.
The design procedures I’ll show are promoted by John Siegenthaler, P.E., in his books and various hydronic heating magazine articles. The formula below calculates the piping head loss in feet. It’s a specialized relationship derived from the widely used DarcyWeisbach equation. It only applies to hydronic heating type applications.
Even with this simplified formula there is some complexity. The fluid properties factor varies with fluid temperature. To simplify the equations’ use I have plotted the fluid properties factor(a) for water as a function of water temperature. Hotter water flows easier than cold water, so a conservative design procedure would be to use your design boiler return water temperature. Fluids other than pure water, i.e. some antifreeze mix would require you to calculate a new graph.
The pipe size coefficient is listed in the chart below. Just select the type and size pipe you are using. Pipe fittings and valves have a resistance to flow. For analysis purposes these fittings have resistances expressed in equivalent lengths of pipe. For example the resistance of a 3/4" copper 90 deg elbow is equivalent to 2’ of pipe.
Let’s do an example for a boiler located outside a heated home in a shed or garage. Assume the equivalent total piping (supply + return + fitting equivalent) is 160’ and we want to use 1” PEXALPEX. A 140 deg return water temperature gives a fluid properties factor of 0.0475. Plugging in the numbers gives:
This formula gives us the flow characteristics of our piping. The piping loss (feet of head) will be greater as we attempt to get more flow (f = GPM) through the pipe. The challenge is to find a circulator that will pump our water at the flow rate required to deliver the needed BTU.
A pump curve is provided by the pump manufacturer. It’s a plot of pump head vs. pump flow. The circulator will pump to over come the head resistance it sees. Said another way the head loss will equal the pumping head provided by the circulator. This is known as the intersection of the “pump curve” and the “system curve”. This intersection determines the fluid flow rate in the pipe.
Let’s take an example. The Taco 007 is a very common hydronic system circulator. It’s “pump curve” is given by the equation:
To find the flow rate f you set the two equations equal to each other:
and solve for the flow rate f. This can be solved by many methods, trial and error, pocket calculator, graphically, etc. A graphical solution is shown below. The curve intersection is approximately 9.6 GPM.
Is 9.6 GPM sufficient to flow the BTUs you need? Let’s find out. An approximation to the rate of heat transfer using water as the fluid is give by:
For a 20 degree temperature drop design our example heat flow would be:
Q = 500*9.6*20
Q = 96,000 BTU/hr.
This is likely too small a pump. Let’s try a Taco 0012
The equation for a Taco 0012 is:
Plotting the Taco 0012 on the system piping graph gives:
Reading the curve intersection shows the flow rate as approximately 11 GPM. Again for a 20 degree temperature drop design the heat flow would be:
Q = 500*11*20
Q = 110,000 BTU/hr.
Still not enough flow. Try another pump.
The equation for a Taco 0013 is:
Plotting the Taco 0013 on the system piping graph gives:
Reading the curve intersection shows the flow rate as approximately 17 GPM. Again for a 20 degree temperature drop design the heat flow would be:
Q = 500*17*20
Q = 170,000 BTU/hr.
Yes, it’s a winner. It can meet the heat flow requirement for a 130,000 BTU/hr. boiler.
There are other solutions. Increase the pipe size or change the pipe type. Copper pipe has less resistance, so with a given circulator copper of a similar size will flow more BTUs than PEXALPEX. Schedule 40/80 steel pipe will flow more BTUs than PEXALPEX but not as much as copper. As you have perhaps noted pipe coefficients are not listed for schedule 40/80 steel pipe. That’s because a different analysis technique is used. One that’s more complicated and difficult to calculate. What I do is use the copper based procedure to select a pipe size and if I want to pipe it in schedule 40/80 use the next larger pipe size.
To complete the example the student should repeat the design procedure using 1” or 11/4” copper pipe. It will show you can use a smaller pump and still get adequate heat flow. Long term, this would be the more economical solution. You are trading up front costs (more expensive pipe) for lower operating cost (less electric consumption). As you can see there can be an endless combination of choices that will work.
As the complexity of your piping system increases so does the analysis complexity. Parallel pipes split the flow and require more complex analysis. My intent here is to give you some simple analysis tools to select you piping choices wisely. Some straight forward analysis can save you money. In my own case when I first installed my boiler, I simply piped it in the same size as the tappings in my boiler. A much smaller pipe size would have done equally well. The larger pipe size was a needless expense. More design techniques are available in magazine articles authored by John Seigenthaler. Do a web search on his name to find them.
 009to090
 Member
 Posts: 5101
 Joined: Fri. Jan. 30, 2009 10:02 am
 Stoker Coal Boiler: EFM 520 HighBoy
 Hot Air Coal Stoker Stove: DVC500 x 2
 Coal Size/Type: Anthracite Rice
 Location: Warrenton, NC
Has anyone ever used PVC? Just curious....Yanche wrote: Common hydronic piping is copper, PEX and steel.
 europachris
 Member
 Posts: 993
 Joined: Sat. Dec. 09, 2006 5:54 pm
 Stove/Furnace Make: EFM 350/Iron Fireman
 Stove/Furnace Model: Custom bituminous burner
 Location: N. Central Illinois
Nice job, Yanche!
 steamup
 Member
 Posts: 1209
 Joined: Fri. Oct. 03, 2008 12:13 pm
 Stoker Coal Boiler: AxemanAnderson AA130, Keystoker K6
 Hand Fed Coal Boiler: HS Tarm 502 Wood/Coal/Oil
 Coal Size/Type: pea, buck, rice
 Location: Napoli, NY
PVC pipe has a maximum working temp of about 140 deg. F. It starts to sag pretty good at 120 deg. f. You need hangers every 3' to support it. No good for heating sytems but ok for most water source heat pump applications. CVPC on the other hand can go to about 180 deg. f. I would not use it on any boiler applications but can be used for domestic hot water. CPVC is commonly used in pool heater piping.DVC500 at last wrote:Has anyone ever used PVC? Just curious....Yanche wrote: Common hydronic piping is copper, PEX and steel.
 Yanche
 Member
 Posts: 3030
 Joined: Fri. Dec. 23, 2005 12:45 pm
 Stoker Coal Boiler: Alternate Heating Systems S130
 Coal Size/Type: Anthracite Pea
 Location: Sykesville, Maryland
As has been pointed out PVC is not generally suitable for heating applications. Some have used it with success embedded in concrete for radiant heat. It can work because the circulated water temperature is low. It does have the oxygen migration problem.DVC500 at last wrote:Has anyone ever used PVC? Just curious....Yanche wrote: Common hydronic piping is copper, PEX and steel.
As far as using my outlined design procedure with PVC tubing for some cold water pumping application it will work. The difficulty will be getting the pipe coefficients. I would suggest measuring the ID of your PVC tubing and using the closest size PEX or PEXALPEX tubing. Realize I'm talking actual measurements, not nominal or trade tubing sizes. Look up the PEX or PEXALPEX inside diameters from a pipe manufacturer's data sheet and compare to your PVC tubing measurements. You will need to use the appropriate lower water temperature. Also, if it's an open loop cold water pumping design, remember to add the static head to the pump requirements.
 SMITTY
 Member
 Posts: 11912
 Joined: Sun. Dec. 11, 2005 12:43 pm
 Stoker Coal Boiler: Patriot Coal  (custom built by Jim Dorsey, Taunton MA  RIP 4/18/13)
 Hand Fed Coal Stove: Harman Mark III (SOLD!)
 Coal Size/Type: Rice / Blaschak anthracite
 Other Heating: Oil fired Burnham boiler
 Location: WestCentral Mass
Great stuff here! Maybe Richard could "sticky" this one so it doesn't disappear in the pile of threads.
 Yanche
 Member
 Posts: 3030
 Joined: Fri. Dec. 23, 2005 12:45 pm
 Stoker Coal Boiler: Alternate Heating Systems S130
 Coal Size/Type: Anthracite Pea
 Location: Sykesville, Maryland
So far all I’ve described is a design procedure for sizing pipe and selecting the circulator. Obviously the piping has fittings, elbows, tees, valves, etc. These components present resistance to flow in the same way a pipe presents resistance to flow. To quantify the amount of resistance the flow resistance is tabulated as the equivalent length of straight pipe. The equivalent length of a fitting is usually a measured pressure loss value express as the equivalent length of a straight pipe. Said another way a fluid is pumped through the fitting and the inlet and outlet fluid pressures measured. The equivalent length is that length of pipe that would give the same pressure loss. Therefore, when designing a seriespiping layout you total up all the pipe lengths and to it add an equivalent length of all the fittings. This length (L) is what you use in the piping head loss equation described in my first post. A table of equivalent lengths is shown below. Units are in feet. It is for copper tubing fittings. For fittings with threaded ends double the listed value. The data in the table is from Seigenthaler’s published work.
The same table in a pdf file format
Let’s do an example. Suppose we have a primarysecondary piping system and we want to know the flow rate around the primary loop. See the diagram for the piping. The intent of this primarysecondary design is for two or more boilers to provide the boiler water to loads located in two locations. This is my situation where both my oil boiler and coal stoker boiler are located in my garage/workshop. The secondary loads are in my workshop and via underground piping, my home. Because a primary loop is generally short it’s resistance to flow is slight and if the loop is close to the boiler it is effectively an extension of the boiler.
Assume 11/4” copper pipe.
For a 2 x 6 ft loop.
Total pipe length = 2x2 + 2x6 = 14 ft.
Equivalent length of tee (straight through run) = 6x(0.45) = 2.7 ft.
Equivalent length of 90 degree elbows = 6x2.5 = 15 ft.
Assume air scope equivalent length is twice it’s length = 2 ft.
Total = 14 + 2.7 + 15 + 2 = 33.7 feet.
This is the L value used in the pipe loss equation.
So the values for the Piping Loss equation are:
fluid properties factor (140 deg water) = 0.0475
pipe size coefficient (11/4 Copper) = 0.0068082
pipe length = 33.7 feet
Putting it all together gives a pipe loss equation:
Plotting this on a Taco 007 pump curve give the following graph:
Again the intersection of the pump and pipe curves gives the flow rate. In this example reading from the graph gives 21 GMP. This is the flow rate around the primary loop piping. This is a substantial water flow. There will be no problems delivering boiler water to the two load connection closely spaced tees.
The same table in a pdf file format
Let’s do an example. Suppose we have a primarysecondary piping system and we want to know the flow rate around the primary loop. See the diagram for the piping. The intent of this primarysecondary design is for two or more boilers to provide the boiler water to loads located in two locations. This is my situation where both my oil boiler and coal stoker boiler are located in my garage/workshop. The secondary loads are in my workshop and via underground piping, my home. Because a primary loop is generally short it’s resistance to flow is slight and if the loop is close to the boiler it is effectively an extension of the boiler.
Assume 11/4” copper pipe.
For a 2 x 6 ft loop.
Total pipe length = 2x2 + 2x6 = 14 ft.
Equivalent length of tee (straight through run) = 6x(0.45) = 2.7 ft.
Equivalent length of 90 degree elbows = 6x2.5 = 15 ft.
Assume air scope equivalent length is twice it’s length = 2 ft.
Total = 14 + 2.7 + 15 + 2 = 33.7 feet.
This is the L value used in the pipe loss equation.
So the values for the Piping Loss equation are:
fluid properties factor (140 deg water) = 0.0475
pipe size coefficient (11/4 Copper) = 0.0068082
pipe length = 33.7 feet
Putting it all together gives a pipe loss equation:
Plotting this on a Taco 007 pump curve give the following graph:
Again the intersection of the pump and pipe curves gives the flow rate. In this example reading from the graph gives 21 GMP. This is the flow rate around the primary loop piping. This is a substantial water flow. There will be no problems delivering boiler water to the two load connection closely spaced tees.
Well, I've got two questions, one of which is sort of related to this. I been working on a heating system in a church. The heating system is hot water and at present is using a circulator pump to move the water. It's all a refit from what was originally a single furnace/boiler which at one time was coal fired. Judging from the size of the piping, I'm guessing that the original system was a gravity feed system, but could be wrong. We are having problems in the winter with getting enough heat throughout the building, mostly because the building was built in the early 60's when energy conservation was of no concern, and the many windows throughout are all single pane. So here's the thing. The original output and return lines are either 3" or 4". The output of the pump is 1", and return is the same as the original pipe. The output goes to a reducer (in reverse) to the larger pipe. Even though it is being pumped rather than gravity feed, I'm wondering if that small diameter at the output is affecting anything. There's only about 1"  2" between the pump housing and the reducer. The next question is a bit different. I was trying to clear out the rusted threads on the housing by running the bolts in with "never sieze" and broke one of the bolts off in the hole. I had used a tap on the other holes but could not get the tap into this hole at the bottom because of the slab that the motor sits on. The gasket is a paper one, and I'm wondering if I can get away with just putting it back together with only 5 of the 6 bolts, or whether it's likely to leak or cause further damage. I can get the bolt out, and so forth but that requires removing the pipes at the shut off valves (flange connections) and the bolts here are in bad shape. I'm concerned about the possibility of cracking the pipe or flanges while trying to free up the bolts. BTW, the gaskets between the flanges are to the best of my knowledge rubber. If anything happens to the lower flange of the valve, I'm forced then to drain the entire system as I have no other way to isolate the pump from the system. I guess I should also tell you that the system uses 1 of 2 pumps at a time.
 Yanche
 Member
 Posts: 3030
 Joined: Fri. Dec. 23, 2005 12:45 pm
 Stoker Coal Boiler: Alternate Heating Systems S130
 Coal Size/Type: Anthracite Pea
 Location: Sykesville, Maryland
Assuming the pumps are properly chosen the short length of smaller pipe and reducer coupling will not cause any problem. Think of it this way, if you were designing the system from scratch and the design works with the pump you have chosen and 1" pipe, increasing the piping size only reduces the pipe resistance friction improving the flow.
Have you done a heat loss analysis to determine if your radiation system can supply sufficient heat? With old gravity coal systems you had continuous heat flow, perhaps at a much lower level when it wasn't needed. Now it's likely much lower when not needed and more radiation is needed to bring it back to comfortable temperature quickly. More radiation can come from more radiators or operating what you have at a higher temperature.
I'd put the system back together minus one bolt. If you would feel more comfortable, adapt something like a "C" clamp as a substitute. Obviously you can't use a "C" clamp because the threads don't have fine enough pitch. But you get the idea.
What is the purpose of the two pumps? Different zones? Only one operate at a time? Perhaps a smaller under sized boiler was installed and the compromise made to operate only one zone at a time.
Have you done a heat loss analysis to determine if your radiation system can supply sufficient heat? With old gravity coal systems you had continuous heat flow, perhaps at a much lower level when it wasn't needed. Now it's likely much lower when not needed and more radiation is needed to bring it back to comfortable temperature quickly. More radiation can come from more radiators or operating what you have at a higher temperature.
I'd put the system back together minus one bolt. If you would feel more comfortable, adapt something like a "C" clamp as a substitute. Obviously you can't use a "C" clamp because the threads don't have fine enough pitch. But you get the idea.
What is the purpose of the two pumps? Different zones? Only one operate at a time? Perhaps a smaller under sized boiler was installed and the compromise made to operate only one zone at a time.

 Member
 Posts: 149
 Joined: Fri. Nov. 20, 2009 8:11 pm
 Stove/Furnace Make: Hitzer
 Stove/Furnace Model: Model 82
 Location: South Central, Alaska
I'm math stupid some times.
But I think I have it simpler.
I have an almost 50* drop in return temp. That to me says its moving to slow...
Am I right?
But I think I have it simpler.
I have an almost 50* drop in return temp. That to me says its moving to slow...
Am I right?
 Yanche
 Member
 Posts: 3030
 Joined: Fri. Dec. 23, 2005 12:45 pm
 Stoker Coal Boiler: Alternate Heating Systems S130
 Coal Size/Type: Anthracite Pea
 Location: Sykesville, Maryland
It could be good or bad, just depends on the flow rate. My examples used a 20 degree temperature difference, you have 50. You must know the flow rate and solve this equation: