Pump Curve/motor sizing hot water heating

elfie

here is a pump curve diagram (interested in better understanding it)



this diagram relates to a pump on a hot water heating system that has a 15 hp motor powering this pump - has the motor been properly sized for system (i.e. maybe a 10 hp motor would work or install a VSD)



reducing motor rpm's (and power) during periods where demand is lower makes sense



sizing a pump/motor for a hot water system seems challenging







 

Zman

Part of the puzzle

The pump curve shows how the different impeller, motor and output combinations will perform across a range of conditions.

The part you don't have is the system curve. Every system has a certain amount of resistance.All systems are different. This resistance will increase as the flow rate increases. This curve will start at the bottom left corner of the chart and curve to the upper right. The point that these 2 curve intersect is the gpm the pump will move.

The third question is what are the design perimeters. How many gpm are required to satisfy the emitters in your system.

To confuse things farther, these numbers will change as the load changes. Zone valves open and close and change the load constantly.

Taco's Flow pro University has some tutorials that may help describe this better.

Carl

elfie

motor size issue

hi, thanks for TACO pump tutorial help link (system uses a taco pump)



according to the pump curve, the 15 hp motor would seem to relate to a system that has a flow rate of around 400-500 gpm (dashed line for 15bhp on right) - a 5 inch pipe can handle 251-450 gpm.



on the system I am looking at, there is a 4 inch pipe extending from the pump.  a 4 inch pipe can handle 121-250 gpm.  so to mean it suggests that a 7.5 hp motor may be a better size;  the gpm flow for a 7.5 hp motor per this diagram has a mid point range of 250 gpm.



one thing needed is the 'head in feet' calc (although the existing 4 inch pipe would seem to limit gpm water flow to 250 gpm).



it would be nice to swap in a new motor to try it out but the mounting stuff beneath the motor needs to be changed for this smaller motor (annoying).



does this make sense

thanks

Gordy

Shooting from the hip

Reread Carl's post. Or another way to think this through is if you were designing the present system. First you would do a heat loss calc. Then decide what emitter you will use then you need to know what flow rate those emitters need to deliver the btus required. Then you size the pipe accordingly. Now select the right size boiler, and the circulator to deliver the btus to the emitters.





Just looking at what size pipe is hooked into the circ now only gives a little clue. Like Carl said you really need to figure the system curve ( head ) to size the circ right. Otherwise you may 1/2 the hp of the motor, and end up with no heat calls.



Could be a mountain of reasons why there is a 15 hp motor on there to begin with.



Gordy

Jean-David Beyer

Could be a mountain of reasons why there is a 15 hp motor on there to begin with.

Right, and it would take only one reason.



I had a two-stage oil pump to pump from my in-ground 1000 gallon oil tank up to my boiler at ground level. It was a two-pipe oil system (supply and return). The reason for a two-stage pump?



The old one-stage pump worked perfectly fine, but with old age, it would leak back to the tank due to internal wear. So the boiler would not start well because of all the air that got in there.



The techie replaced the pump, but all he had on the truck was a two-stage one, so that is what he put in.

Zman

You are likely correct

That does seem like massive circulator for a 4" pipe.

It will also take a significant amount of research to be sure you have resized it correctly. Circulators of that type are quite expensive, you need to be right.

Another approach I like is to replace a large central circulator and zone valves with several smaller ECM circs. It is not possible on all systems but is a great way to reduce circulator energy costs.

Sizing (or oversizing) a circulator for the worst design day then throttling it down for the average day is like hopping in your car, pushing the gas to the floor, then applying the brakes when you go to fast. All the while keeping the gas floored.

It wastes energy and wears out you piping and components.

Carl

elfie

pic of motor and pump

here is a pic of the motor and pump (a backup motor is in the background)



it doesnt seem terribly expensive to change to a lower hp motor (as long as rpm specs for new motor are same as current motor)



the connecting piping to the pump seems to be smaller than the 4 inch pipe above (not sure why)



the current motor is a baldor motor

SWEI

Motor may not be original

Taco paints theirs green, so this one has probably been replaced.  Who knows what was originally there - I'd not be surprised to find a replacement was upsized by one of the all-to-prevalent graduates of the "bigger is better" school of heating.



As you've been advised above, you really do need to assess the entire system before you start changing parts.  Start with a room-by-room heat loss, then a system curve.  Assuming the motor is a 3-phase, I would hook a VFD to it and log RPM versus differential pressure with all valves open and then at a couple of other points to really understand the hydraulics.  Then you need to look at occupant comfort -- are there hot or cold areas in the building?  Factor those in and you know what you want to have when you're done.  A modern smart circ (Taco Viridian, Grundfos Magna, Wilo Stratos, etc.) would allow the pump to run a delta-P algorithm and could reduce pumping costs by 75% or more.



You've posted quite a bit here over the past few months.  Looking at the size and complexity of your system I believe you should hire either a really good contractor or a consulting engineer before you make any changes to your system.

paul_79

pump sizing

yes sizing pumps for systems is a process, you first have to know how many btu's you are going to deliver with the water or water/glycol mix. then you have to set your  temp difference .  td's  that should give you your gpm then the piping and resistance will give you  the head  your pump can pump against at the correct gpm's  then you go to the piping schedules and  get the correct size pipe to handle the gpm's at the correct feet per second  ( to avoid noise, erosion ect.)   the correct friction loss rate should be between 1-4   then you get to the pump chart and find the correct gpm plus head and see where they intersect with the pump curves. on the chart you show it has different pump motors  available if you noticed also the numbers showing  10.40" down to 8.05" these are the impeller sizes so you find your gpm and head and where they intersect will be the correct pump motor and impeller size.  i would start with the idea that the engineer went through all of the steps and picked the correct pump / motor / pipe combination for your system. here are some pictures of the forty or so taco pumps i have to deal with.







the first photo is of a chilled water loop at 465 gpm at 65 tdh  the motor is a 15 hp.



the second pump is a heating  loop at 66gpm  at 64 tdh  and the motor is a 5 hp

elfie

pressure differential

thanks for feedback about VFD and pressure differential;  i like the vfd approach which will probably be alot easier than doing a full scale contractor review (the facility is rather complex - 2 bldgs sharing a heating system)



would be interesting to know the GPM rate at certain areas of facility.  would knowing the pressure at a locations at far end of heating loop (i.e. on a main ) be helpful in determining the gpm flow rate?  it probably would make sense to install a pressure guage in this far end location



i am doing a 'rough head loss calc' (i have all the blueprints) - a complex area

bob_46

Pump

I would try this approach first. One ,make SURE you have the correct curves for the make and model pump. Two, verify diameter of the impeller. To do this you will have to have gages on the supply and return of the pump. It is best to do this with one gage and some piping and valves. With the pump running close the valve on the discharge and read the differential across the pump in psi and multiply by 2.31. This is the maximum head the pump can produce, look at the curves and determine diameter. Three, open the valve and with ALL the valves in the system in the position requiring max gpm, read the differential across the pump in psi. Multiply by 2.31 this is the head loss for this system at these conditions. Four, go to the curve for the impeller that you proved and

find the point on the curve found in step three. This is where the pump is operating . Now look at the HP requirement .

bob_46

Example

Using the curves you posted. With the pump dead headed as in step two you read a differential on the gages of 37.6 psi times 2.31 you get about 87 feet of head which tells you you have a 9" impeller. Let's say in step three you get 32.5 psi differential times 2.31

equals 75 feet of head. Read to the right on the 9" curve to 75' of head and 315 gpm and a 10 HP motor.