Brine and Cooling Water Mixup

took the words right out of my mouth. Interlocking the supply and return valves is the way to go. Mechanical interlock is somewhat more fool-proof, but air operated or PLC controls might do the same thing.

Good not on cross contamination. Can the vessel be drained between utility switch-overs?

I agree with bigg. Valving and controls are easy enough to implement.

I'd add a purge or diverter valve and return the brine or water to their respective systems, or to drain.

1.) As others have mentioned by having them controlled with a PLC.

2.) Or if manual (budget) by having a flow plate.

But you will still have a problem with operator era.

Or

3.) Train your personnel and hold them accountable. But to be accountable, they have to be competent.

Dear members,

If you could notice there is ball valve in the picture at the bottom, operator before changing to any utility opens this valve and applies air pressure through one of the inlet valves and flushes the reactor jacket.

We are more concerned about contamination or mixup during utility is in operation. Secondly we have some system which have pneumatic valves running on PLC, but here too the we are facing this problem but with less intensity. PLC operation doesnt negate the possibility of mixup if the valve is slightly passing or if the valve isnt closed 100%. If if there is a slightest passing, at about 5-6 Bar pressure the loss is huge over a period of time.

We have trained the operators on using the system, but the pointer stops at what if the valves are passing or not completely closed? I guess some thing that alerts operator on mixup would be ideal.

Have to agree here. Look at your maintenance schedule to prevent leakage from the valves. Maintain them before they leak. And never forget there is a live to everything. A valve that is in service so many years or hours will even with good maintenance fail at some point. So a replacement plan for equipment has to be implemented.

Bingo,

i would also add a leak detect, if a seal is going bad to avoid contamination.

Many process reactor jackets have an air inlet, a drain outlet and an air vent. Correct sequencing of valves can be carried out using any programmable device according to a pre-agreed protocol that has passed through a HazOp Study. Responsibility for correct operation then lies with the Utilities Designer, the Controls Engineer and the Maintenance Manager, which is where it should be.

No-one comes to work to be chided. It just doesn't happen in a well-run company.

I wonder how are others addressing this problem, after all they must also be facing this kind of problem. What about any ideas on detecting the mixup any known techniques in place?

You've seen several good plans for preventing the mixup.
Since "S**t Happens" you are wise to seek also to detect a mixup.

Specific gravity measurements have not served you well, so it's worth a look at a conductivity sensor, for example:

http://www05.abb.com/global/scot/scot203.nsf/veritydisplay/4e2c7fbcb4be8725c1257958004f8da2/$file/DS_TB404-EN_C.pdf

Such a sensor will betray the presence of brine where it's not expected.

Mark Bingham
Relativity PL

Have you looked into a conductivity sensor for the fluid in the jacket? This might reveal (1) whatever the active fluid is in the jacket, and (2) if there is a trend taking place during one of the services. You could intentionally add something that is conductive (not necessarily corrosive) to the different services to distinguish them, the other is to use colorants in the different services and monitor them with a flow through spectrophotometer.

Is passing one of the fluids through a coil wrapped around the equipment ( between the equipment and outer shell) such as it will be completely isolated from the other fluid feasible? This will spare you from draining at least the brine mixture.

I am making many assumptions here. you are trying to conserve energy and not the OH solution. You are like many batch reactors you want to heat to temperature quickly and control at temperature. Cooling costs more than heating and refrigeration is limited more than heating. you neglected to show a recirculating loop on the diagram.

If i am right:

I suggest you watch the return temperature (not the outlet temp, but they may be the same). You should find that when it switches from heating to cooling' there is a lot of cold brine that goes into the jacket before cold brine comes out of the jacket. Switch the return flow on dropping temperature. There will be a mass imbalance because now the brine system will be droopping while hot liquid is returning to the hot system. No it will not always balance out on the backside because heating will use different volumes and return on different temperature conditions. Some of the optomization will require you to understand how the brine storage system works. If there is a warm brine return tank and a cold brine storage tank that overflows back to the warm tank is typical.

Not enough information to be more help.

Greetings Srao,

i agree with the other posters as to the mechanism and their application.

Still it seems you are trying to make the system "fool proof".

If that is your goal then your only option after implementation of the above posted control systems is Not To hire Fools!

I can't imagine not automating the piping and valving to solve the bulk of your issues, but if you still want to determine which fluid is in the pipe, 2 methods come to mind.

Refractive index is used to determine concentrations in aqueous solutions. A quick test by a prominent vendor like K-patents could determine how much the index changes and whether the magnitude of difference is sufficient to trigger an output.

The Siemens clamp-on ultrasonic flowmeter is used to differentiate grades of diesel, kerosene or gasoline, because the speed of sound (sonic velocity) is unique to each grade. In fact the flowmeter is sometimes used not as a flowmeter but to calculate API gravity of the fluid in the pipe.

A test would determine whether the sonic velocity for you fluids is different enough [over operating temperature range(s)] to be useable as a fluid identifier.