Calculating Heat Load for Cooling Tower...

[cfordyce]

I am replacing a Cooling Tower. I would like to calculate our current cooling load (heat load). This is the info that I have...

Temp in is 85DegF
Temp out is 74DegF
Flow is 2800GPM
...through a 10 inch pipe ( I don't think U need this info though.

I would like to know the BTUs?
or whould it have to be BTUs per minute???

I know 1 BTU is the amount of heat to raise the temp of 1 pound of water by 1 degF.

How is this calculated?

 

[TD2K]

2800 gpm is about 1.4 e6 lb/hr of water.

You have the heat capacity and the temperature change of the water. This is a simplified number as a portion of the incoming water evaporates to provide the cooling but it will give you a close number for your cooling load.

About 15.4 MMBTU/hr is my rough estimate.

 

[krb]

Check out ctdoc.com for more info.

 

[TBP]

One USGPM of water is very nearly equal to 500 lbs/hr. (1 gal/min X 8.33 lb/gal X 60 min/hr = 499.8 lbs per hour.) The delta-T is 11*F, and the flow is 2,800 GPM.

2,800 X 500 X 11 = 15,400,000 BTU/hr, which is the same as TD2K's number.

 

[stressriser]

I have 15,393,840 BTU per hour.

heh heh, couldn't resist, sorry.

 

[TD2K]

I did say it was an approximate answer stressriser ;-)

 

[quark]

Cfordyce!

Use the formula Q = mCpdT. (excuse me all. I know you already did it )

But Iam getting 15,474,921.97 Btu/Hr. This may be because I converted the values first to SI units and then reconverted back. (if so what will be the error?)

(2800gpm/60sec)*(3.785lit/gal)*4.2kJ/Kg. degC*(85-74)*5/9Deg.C = 4533.58 kW*(56.89btu/min/kW)*60 = the above figure.

The above figure is equal to 1289.57 TR (1TR = 12000BTU/HR)

This tower is a pretty big one. What is the application?

Regards,

 

[cfordyce]

Thanks for the calculations. Some of my design data may have changed. I have been looking at the pump curves and now realize that one of the pumps is operating way out of it's design conditions. It is operating way up on the pump curve and is therefore not moving as much water as first thought.

Tell me if I am reading this thing right...

I measure the actual pump head (psi x2.31 = ft head).
now I go to the pump curve and move right across the graph until I hit the appropriate curve.
then I go down to the bottom of the graph from that point to see the flow rate at that head pressure! Right?

Now is the pump curve still accurate when reading way over to the right of the graph?

maybe I can email this curve to someone so you could confirm this?

 

[stressriser]

I think it is beneficial to lay the system curve over the pump curve. This way you identify the extreme operating points, which may be off to the right.

 

[KenRad]

cfordyce,

First, you say "until I hit the appropriate curve." Unless this is a variable speed pump, there is only one curve that applies to your pump. If the info you have shows multiple curves, you are probably looking at different curves for different impeller diameters. Look at your pump nameplate to get the impeller diameter for your pump, and this will lead you to the right curve.

Also, you want to measure the head developed by the pump, so to be accurate, you need to take a pressure reading at the pump suction as well as the pump discharge. The head difference between the two will be the head developed by the pump.

And since you are an electrical type, you can verify your operating point by taking motor amperage, calculating BHP, and seeing if it matches the BHP curve at your calculated operating point.

---KenRad

 

[TD2K]

Cfordyce, you can email me the pump curve if you want a double check on your approach (testdog2000@yahoo.com). As KenRad points out, you need to measure the differential pressure across the pump to come up with its head.

The pump curve should be accurate for the range shown by the vendor. If you have to start extrapolating the curve past those end of curve points (BTDT), you can have problems. On one pump I remember, I was out more than 20% on flow from the curve versus what I got with a portable ultrasonic meter (trying to remember the exact numbers, might have been higher). When I talked to the rotating equipment engineer, he wasn't surprised at the difference given where the pump was operating (and he wasn't happy with its operating point).

 

[stressriser]

I still feel it necessary to consult the system curve. You can not pick out a pump without knowing the operating point. Since the operating point defines the system head and system flow rate, can we be sure the pump is even sized properly?

Or maybe I am not thinking straight. Just my 2 cents.

 

[vanstoja]

The manufacturers pump curve is, strictly speaking, only assuredly good for the manufacturer's test loop which presumably met some pump code like Hydraulics Institute or ASME Power Test Code. If his inlet piping is complex (eg. with several pipe bends) he's allowed to put flow straighteners in the inlet piping to assure good flow inlet conditions to meet the "guarantee point". If you've got inlet piping exactly like his test loop you may not have to run a baseline performance test in your loop to determine your facility's pump head-flow, power, current, etc. curves. If you're playing for high stakes, don't bet capriciously that your pump facility will be hydrodynamically identical to the manufacturer's test loop. RETEST THE PUMP IN YOUR LOOP! If you haven't done this then you may be miles off using the manufacturer's curves. vanstoja