Reactor Heating and Cooling System Flow Problem

[ChipFuller]

I've install a closed loop heating and cooling loop to a 500 gallon reactor. The system has three 3-way control valves in series with 3 separate heat exchangers. As the PLC calls for different temperatures the flow will either go through the heat exhanger or around it depending on the situation. Under normal conditions the glycol will flow through one heat exchanger and then through the contol valves. Here's my problem the pump that I purchased is for 100 gpm at 100 feet of head. According to the Rosemount mass flowmeter my flowrate is 40 gpm. I used a pressure differential gauge across a B&G circuit setting valve and it says 70 gpm which is acceptable to the end user. Which one do I believe?

The pressure drop across one of the control valves is 6 psig. Samson, the valve manufacturer say the valve has a Cv value of 47 when one of the legs of the 3-way is closed. They said this equates to 100 gpm. How accurate are the valve manufactures in their Cv numbers?

With my valves in series and going through one heat exchanger my pressure drop is (6*3)+11(pressure drop across Heat exchanger)=29. The reactor and the miscellaneous piping is about another 20 psi. For a total pressure drop across the system of 50 psi which is about what I'm getting. Am I missing something? According to the pump curve at 40 gpm and a 50 psi drop, I'm on the flat line of the curve.

I've notice when I close the B&G valve that the Rosemount flowmeter barely changes until the valve is 1/4 shut. Also the user wants turbulent flow through the reactor. I spoke with Tygon the reactor manufacturer. They said to calculate the Reynolds number through the reactor to use inlet piping id which is 1.5". Does this sound right? According to my calculations I be fully turbulent at 40 gpm using 50% Ethylene Glycol. Thanks

 

[erpeterson]

I am unable to answer your questions without more details. However, let me offer the following.

When installing flow measurement devices, make sure you allow several pipe diameters of straight pipe upstream of the instrument. If not, inconsistent and incorrect measurements may result.

Pump curves are generally very good when the physical properties used to develop the curves are correct. You may wish to double check your physical properties.

General equations for Cv are notoriously poor, but within a specific manufacturer, are quite good. Of course, the will also need good physical property data in order to give you the right answer.

Flow measurement is one of the most unsure measurements we make. Calibration with the fluid used is very important. Since you are using 50% ethylene glycol, I suggest trying a bucket test, where you force the flow into a tared bucket or tank for a set period of time then weigh the full tank. Your accurate weight and time measurement will let you correlate th actual flow to what your instrument reported.