Flow through existing Pump 1

[onecolumbyte]

I have been tasked with upgrading an existing process heating loop to a higher temperature. The system currently uses a glycol/water mix (30% by weight) to heat a small (1 lb capacity) mix bowl. The customer wants to upgrade the process to achieve 200 deg bowl temps. I have proposed using an inline circulation heater to heat the incoming fluid from 180 to customer reqs of 225 deg (fluid temp). To size the heater I must know the gpm through the system. So, How do I determine the flow through an existing system?
My thought: find the system head and use the pump flow curve to find the flow. This is closed loop circulation so the head is only the piping losses. right?
Is there another way to measure head in an existing system besides measuring pipe and counting fittings to find equiv. pipe length?

Thanks in advance.

 

[smckennz]

Hi onecolumbyte
the best way is to use accurate pressure gauges to measure the differential pressure across the pump. If the rise is low , you may need a differential pressure gauge or a manometer.
If you are using an in-line heater, pay special attention to heat flux density or you may experience localised boiling.
The normal way of estimating heater size is by estimating (bowl) heat drain and adding a bit for losses and uncertainty.

Cheers

Steve

 

[25362]

One could estimate the water-glycol flow rate by making -if possible- an enthalpy balance around the mix bowl. Ethylene glycol is toxic, why not use just properly treated water ?

 

[ccfowler]

onecolumbyte,

It should not be necessary to use glycol for such temperatures. It should only be necessary to operate the circulating water system at a slightly elevated pressure.

The use of glycol increases the temperature requirement for the circulating fluid since it produces less favorable coduction/convection heat transfer performance at the liquid-solid boundary when compared to just water. (Usually the degradation is in the 10% to 15% range depending on concentration and flow velocities.)

I would not put very much effort into trying to precisely determine the flow rate. I would determine reasonable estimates for all parameters, install a heater with sufficient excess capacity to cover all likely situations including heat losses to surroundings, and install a suitable temperature controller for the heater to assure stable heat transfer to the mixing bowl.

 

[onecolumbyte]

re: why glycol? This is an existing system that also feeds other test rooms so I'm going to use what already exists. As to why they chose glycol to begin with ...?

smckennz: My cameron hydrauics book says to calc head I must also know the velocity head. Do you think this is term can be ingnored for a rough calculation? I will go get a pressure reading this afternoon.

fowler: I agree completely. My problem is what is a reasonable estimate for the flow? I know the pump will operate at the intersection of the system curve and the published flwo curve. But how do I establish a reasonable guess at these?

 

[abeltio]

there are excellent on-line ultrasonic flowmeters. these are attached to the pipe from outside and allow flow measurement without disrupting service.

there are companies that give the service (i.e. rent the equipment) when there is no need for permanent monitoring.


saludos.
a.

 

[scalleke]

I ask myself the question why you are asking yourself questions about the flow rate.

As Ccfowler mentioned the heat loss that you have is the main factor here that will determine the heater sizing.

There are possibly other ways to determine the heat loss of the system than just using the flow rate.

True ultrasonic flow meters exist but I feel they are too inaccurate.

What about using the pump to heat the liquid?

We have an oil flow loop and we can circulate oil at a couple of GPM and heat the oil up from ambient to over 70°C by just throttling the discharge of the pump and running at low efficiency.

What about using the pump info, the E-motor data and the E-power data to determine the flow rate?

Best regards.

Scalleke

 

[onecolumbyte]

Scalleke: I am interested in your suggestions. Another engineer in this department also suggested measuring the power consumption of the motor as a way to estimate flow. Along the lines of if we know the votlage and amperage we should be able to get the horespower ... but it hasn't been fully explained to me. Do you have a more detailed suggestion or a source suggestion?

Thanks for the help.

 

[25362]

I found two articles on the subject or similar.
These are Plant Notebooks published by the ChE magazine in its issues of:

Jan 19, 1987: Check pump performance from motor data, by V. Ganapathy

and of

July 1989: A quick estimate for centrifugal-pump efficiency by Alejandro Anaya Durand.

Good luck.

 

[DLANDISSR]

onecolumbyte:
Get a pump curve
get a test pressure gage and measure the pressure at the suction and at the discharge of the pump''
Measure the voltage and amps at the motor.

Look at the pump curve with this data and it should give you a good estimation of the flow. This should be sufficient to use in selecting a heater.

The piping should be well insulated.

Once the flow is determined you can measure the temperature across the mix bowl. With flow, fluid data and temperature differential you can determine the heat used at the mixing bowl.

Ratio this heat with the present temperature to the new higher temperature.

Select a heater with an output of aabout 1.1 to 1.25 time this heat requirement.

 

[onecolumbyte]

Thanks for the continued input.
Based on the differential pressure on the pump, the flow curve, and the knowledge of system pipe size as 1/2" and 3/8" I've come up with an estimate of the max flow rate as 5 GPM. This estimate is based mostly on the fact that a higher flow rate would require very high velocities especially in the mix bowl area where the pipe is 3/8". Amen to the "reasonable estimate" method. Based on the heater catalog's formula

Q=(F*Cp*dT*S.F.) * 1KW / 3412 Btu/hr

F=flowrate(lbs/hr)
Cp=specific heat (.93 Btu/lb/F)
dT=change in Temp. (170 to 225 F)
S.F.= Safety Factor of 20%

I come up with 52Kw heater.
I'm sure this heater will do the job, it sure is a big one. If anyone out there thinks this unreasonable, opinions are appreciated.

 

[25362]

It's a bit on the high side considering the heat up in your first post was from 180F to 225F. Even w/o this correction -and including the 20% SF- it seems by about 15% too large. I may be wrong, would you re-check and confirm ?