Heat Exchanger With Refrigerant As Cooling - How To Control 3

[vds002]

I was wondering if anyone at this forum has ever came across a heat exchanger which uses refrigerant as a cooling medium.
I had a few question about how to control the cooling.
1. Do I need to measure the refrigerant return temperature and control the refrigerant inlet valve?
2. Do I need to measure the heat exchanger process side outlet temperature and control the refrigerant inlet valve?
3. Do I need to cascade the refrigerant inlet flow with temperature measurement.
4. Do I need to cascade the refrigerant inlet pressure with temperate measurement.

 

[Homayun]

I am not a process control engineer, but I thin option 3 would be the best way to control it.

 

[dcasto]

There is a simple system called a superheater control for refrigerants. A small portion of the refrigerant is placed in a cylinder with a pressure tap. The cylinder is placed in the refrigerant stream leaving the exchanger. What happens is the refrigerant in the cylinder will have the same temperature as the exit stream. The pressure in the cylinder will then be the bubble point pressure. You then measure the pressure on the refrigerant leaving the exchanger. That pressure is compared to the pressure in the cylinder. To assure you have no liquid refrigerant returning to the compressor, the pressure of the refrigerent must be lower. If it is higher, then a valve that allows the rerigerant to enter the exchanger is closed.

I've also set a seperator on the outlet of the exchanger for the refrigerant. I then set a back pressure valve on the scrubber outlet. You fix the pressure in the separator this way. Then you look at the refrigerant PH diagram and you know that it is a saturated gas at X temperature. You measure the temperature in the separator. If the temp in the separator is lower than the dew point then close the valve that allows the refrigerant to enter the exchanger.

 

[CH5OH]

I came across a few systems :
most common is the setup with an expansion valve:
the pressure is controlled with the compressor (on/off or unloading/loading cylinders on multi cylinder compressor.the pressure controls the temperature in the cooler(since there is a vapour/liquid mix the temperature is constant at a given pressure).the expansion valve puts more liquid, when needed.It is controlled by the temperature on the outlet of the cooler.typically it opens when vapour on outlet is super heated by about 4 degC.

another system uses a level control in the cooler.
the cooler is completely flooded with liquid refrigerant and the boil off is collected at the top in a drum.the pressure is kept constant by means of the compressor.from the outlet of the compressor the gas goes to the condensor.
the liquid of the condensor is fed to the control valve .If the level in the cooler drops, the valve opens.
The level control system has the advantage of a better heat transfer in the cooler.but I only came across this setup once (on an inert gas plant)

 

[vds002]

I really appreciate everyone taking time to respond. Sorry for not making the question clear. I am not talking about refrigeration machine (complete system). The plant has a refrigeration system. I am getting the refrigeration from a header (from the refrigeration system) going to shell and tube heat exchanger where it is used to cool process vapors. I have a control valve on the refrigerant inlet line which does the expansion of the refrigerant. What I am trying to figure out is how should i control this valve. Should I use temperature or a temperature-pressure cascade or a temperature-flow cascade.
thanks for you help!!!!!!!!!!!

 

[CH5OH]

well, a refrigeration cycle with a header remains a refrigeration cycle.
header is kept on ambient temperature (high pressure):inlet cooler is liquid, outlet cooler is vapour (super heated)
the outlet pressure of the cooler(suction compressor) defines the boiling temperature. if the required cooling temperature is higher than you place a pressure control valve on the outlet of the cooler, to keep the pressure in the cooler constant (in case you have different cooling temperature requirements throughout the plant and the compressor control is tuned to the lowest temperature requirement)
the amount of liquid evaporated depends on the load of the cooler.therefore the expansion valve is regulated by measuring the super heated temperature on the outlet of the cooler (before the pressure control valve)
so all together PC=pressure controlled valve outlet,TC=temperature controlled valve inlet:
pressure cooler drops>PC closes
temperature outlet cooler drops>TC closes

 

[toabhijeet]

Hi vds002

I guess you are in coneceptual design stage.


Please figure out some critical issues like
a) Allowable DelP for Control Valve
(Concern: JT Effect & Back Pressure constraints)
b) The phase & conditions of refrigerant available
c) What service is expected from the listed HE
d) Any provision is available to measure the T/P of refrigerant outlet.

Thanks,

 

[sshep]

Hello Vds002,

What I envision as typical for what you describe is a variable area type exchanger (refrigerant level varies). The operating pressure on the refrigerant side is determined by the compressor suction (which is controlled by the machine, typically via speed). Liquid refrigerant will pass through the inlet control valve, some will flash cooling refrigerant to bubble point temperature. The remaining refrigerant liquid will vaporize against the process. The refrigerent leaves the exchanger as vapor. The required flow of refrigerent is determined by the process duty, as is the required operating level.

The refrigerant can be in either the shell or tubes. If refrig is in the tubes, a level pot is often attached to the channel. If refrigerent is on the shell, level if needed can be by instrument (or bridle) on the shell. In both cases the refrigerant level varies with process requirement, and the level is self regulating based on the exchanger duty.

The control depends on disturbances. If no disturbances, a process outlet temp to refrigerent control valve can be sufficient (TC->valve). If there are other users on the system a TC->FC->valve (refrigerant flow) is recommended. If the process side has disturbances or variable back pressure on the vapor return, a TC->LC->FC->valve cascade is needed (refrigerent level in the exchanger is measured). If you are in design, I would go for the latter, but surely other users on the system can be a guide to your new installation. In any event, the consequence of control failure can lead to disaster if liquid returns to the compressor- hopefully there is a drum in front of the compressor otherwise there may need to be high level controls to cater to failure of the basic controls.

I think this is the type of system you refer to.

best wishes,
sshep