Heat Gain from Pump Motors 1

[KLH]

The Carrier Load Estimating Manual states that the some of the energy input to a pump motor is dissipated as heat in the motor frame. The rest of the power input is dissipated by the driven machine, the pump.

Carrier goes on to say that this heat gain to the fluid does not show up as a temperature rise because as the pressure reduces around the piping loop, the fluid expands. The fluid expansion is a cooling process which EXACTLY offsets the heat generated by friction.

Does this mean that I do not need to add any pump motor heat load to a chiller load?

Can anyone shed some light on this. Thanks.

 

[electricpete]

I don't believe it. Why would expansion of water take up energy.

The only way expansion factors into heat balance that I can see is expansion accompanied by phase change in a refrigerant system. I think you are talking about the chill water though... no phase change there. Heat in must equal heat out.

 

[KLH]

Good analogy.

Now that I think about it, there is no phase change at a refrigeration system's expansion valve, and that is where the temperature drops. The expansion of the refrigerant absorbs heat. HMMM.

 

[TD2K]

I think the way I would look at this is:

There is heat generated off the motor itself because of its inefficiency. Motors will range, depending on the type and design (high or normal efficiency) from 95%, maybe a bit lower, up to 98% or so. So, if you have a 150 Hp motor, you could be dissipating as much as 5 Hp or so as raw heat off the motor.

Now, a centrifugal pump takes Hp and converts it into pressure head AND temperature rise of the fluid because centrifugal are not 100% efficient (PD pumps tends to run quite a big higher efficiency).

Assuming a centrifugal pump with an efficiency of 75%, 25% of the inputted energy therefore goes into the liquid as heat. However, for most applications this is results in a minor temperature rise across the pump unless you have high head units (this has been discussed on this site several times). But, doesn't this heat ultimately have to handled by the HVAC system which Carrier would be focusing on?

Now, what about the energy imparted to the liquid as pressure? As the liquid travels around the system, this energy is 'consumed' by friction losses. Since energy has to be conserved, would this not show up as heat being the lowest form? Obviously, if the liquid is going through a machine that uses the pressure to do work, that work has to be subtracted from this but for a chilled water system, wouldn't all the pump's inputted energy ultimately show up as heat or am I missed something?

 

[electricpete]

good point TD2K to separate it into a fluid power component and a pump loss component. I was only talking about pump loss.

KLH - I don't think that expansion of a gas in a thermal expansion valve can compare to anything that goes on in a pumped liquid system which remains liquid.

My belief
- pump losses goes immediately into heat. It may be carried to other parts of the building by the fluid system but it is heat.
- fluid power is converted into heat by friction with exception noted by td2k (hydraulic machine like turbine) which produces mechanical work.

Let's say you have a pump pumping continuously in a closed loop against friction, in steady state with constant fluid velocity. You are putting energy into the system. Where does that energy go? To heat, where else?

 

[d23]

KLH:

For what it’s worth TD2K made a very good explanation of pump efficiency or what you will be considering as the pump inefficiency. The heat added to the fluid is as follows:

Fahrenheit Rise
=((BHP-WHP)*42.41)/(lbs/min*Specific Heat)

Where:
BHP = Break HP
WHP = Water (fluid) HP
42.41 = Conversion factor of HP to Btu/min
Lbs/min = GPM*8.33*Ave Specific Gravity

Centigrade Rise:

((BkW-WkW)*14.34)/ (Q in l/m*Specific Heat)

Where:
BkW = Break kW from pump curve
WkW = Water kW = (Meters * Liters/Min)/6000
14.34 = Conversion from kW to kilocalories

Somewhere in your system you will need to account for this heat to be dissipated.

Good Luck

 

[ChasBean1]

KLH - good insight by all above.

BUT (a word that generally discredits any nice things you said previously), you're sizing the system based on an anticipated design condition, e.g., 90°F DB 72°F WB outdoors. That in itself can be the source of significant error as summer heat varies. Account for a 20% system diversity to reduce the chiller size and not waste energy. The error from estimating heat loads based on tables can be very high. Heat addition from pumps is usually negligible for unit sizing, unless you are a manufacturer that needs to give detailed information. The sizing is more along the lines of, "I want to land the Apollo in the Caribbean," whereas your concern is more along the lines of "I want to land the Apollo on Loquilla beach at 6 PM and have a banana daquari waiting." You're talking trees where this is really a forest item... (Hope I didn't just udder nonsense!) -CB

 

[quark]

Concept of no phase change after expansion valve is exactly not right. Some portion of the refrigerant does flash after expansion valve(otherwise your expansion valve may not be working properly) and Electricpete is right.

It is a general practice to check rise in chilled water temperature across pump.If it is not significant you can leave it.For most of the systems I didn't see morethan 0.5⁰C rise from pumping station to the end point.

100 marks to (out of 100 )TD2K and CB.

Temperature drop due to liquid (high boiling point and non compressible) expansion seems to me as a gimmick to prove technical superiority.

 

[KLH]

Thanks all.

I'm still intrigued by Carrier's statements. I think I'll give them a call.

 

[electricpete]

KLH - That sounds like a good idea.

If you don't have luck with them, I would recommend to post the exact words that you have read and the context. Maybe there is something lost in your translation that would become apparent during that process.

 

[25362]

Designers of chillers provided with rotating mixers usually add the whole (mixer) motor loads to the process heat input that the chiller has to remove, including heat losses (albeit having a good thermal insulation) even though these are considered small in comparison with the process load, probably as a safe reserve.

 

[ClydeMule]

Though I wouldn't want to discredit the "expansion heat los theory" immediatley (haven't done the math), I have to say that it seems like a stretch.

In an HVAC system, the ratio of refrigeration capacity to pump HP is really high. Therefore the heat gain to the water from the pump head an inefficieny is almost unmeasureable across the pump. While it is very small at one given instant, it will heat up the water over time. However, because of the dynamics of a building chilled water system, you would hardly ever see it.

In small process chillers, you have to pay attenion to this sort of thing. The pump horsepowers tend to be higher. Also there are more high head centrifugals and PD in this chilled water business. Therefore the Refrigeration to Pump HP is much smaller. The heat load from a pump can acutally put the too much of a load on the chiller and cause it to not keep up in some conditions.

I can't remember the right terms from my thermo class, but I think that it has to do with the idea that pressure drop is and irreversible process, whereas the energy input to moving the water is reversible. This is why PD pumps put more heat into the water even though they are way more efficient.

This is why when sizing hydraulic (fluid power) system they have to be careful about get rid of the heat because most of the time the pump is byassing the fluid and just generating heat. THe same thing happens to machine shops where they do precision work. The high pressure coolant pumps (usually multistage centrifugals) add a lot of heat to the coolant. As it increases over time it can be hard to keep tolerance on close tolerance work like grinding or when working with aluminum

We actually ran a test in our plant with a small 20 gal tank and a 1 HP stainless steel centrifugal. Insualted the tank, and it heated up to over 120 F in an hour or so. I can't rememebr the actual numbers. But the heat balacne almost worked out perfectly when comparing heat rise to BHP and WHP inefficencies.

Someone help me out on the irreversible work thing.


Adios.

Clyde