Sensor for Dowtherm G and or Q

Should be a supplier here listed....

http://msdssearch.dow.com/PublishedLiteratureDOWCOM/dh_0036/0901b80380036229.pdf?filepath=heattrans/pdfs/noreg/176-01335.pdf&fromPage=GetDoc

Thanks for the link.

It sounds like you are a specialty OEM.

Or, an instrumentation supplier.

If you are an OEM, enlist the help of your suppliers, or potential suppliers for this instrument system.

If you are supplying the instrument, you are on your own, I think.

As an OEM, you have to insure the equipment you supply will do the intended job, regardless of cost.

I dealt with this in the circuit board industry. Cheap is not always good.

You don't say where you are or what the industry is.

What does Dow have to say?

...or turbidity.

The presence of the one fluid in the other represents a leak, a loss of containment. So wouldn't it be easier, from a loss-prevention perspective, to prevent the two fluids coming into contact in the first place?

Gentlemen,

Thank you for the input. I am working on the design of a tempering loop in a chemical process plant.

The client company had a heat exchanger tube rupture earlier this year and as a result their cooling water was contaminated. Due to the system design nothing was released to the environment, however the clean up was costly.

That being said we are designing a secondary cooling loop to further isolate any leakage that my occur. They want a sensor on each heat exchanger return to warn them of any future leaks so that any leakage can be immediately contained.

Presently they are using Dowtherm G. We are presently investigating a Rosemount Clarity II turbidity sensor for suitability. This sensor is much less costly than what we might normally use in more worrisome applications.

They will back up the sensor for a certain time frame with daily laboratory testing.

My original question was posted to see if anybody had any better ideas.

The idea of die addition my be a real possibility as a further safety precaution as the end customer is very conservative in these situations.

Thank you for your help.

You could also look into double-wall heat exchangers.

Did you check the compatibility of the tube material with the wetting fluids to see if there is a better choice? Did you check the maintenance regime on the exchanger to make sure that everything is being done to prevent the leak? Did you check that the water has been properly treated and maintained, and will remain so? I mean, after all, your instrument proposal only proves the door is shut after the proverbial horse has bolted.

Very good points. The heat exchanger was replaced with a new exchanger last weekend, one designed to be better for the intended application.

They are also very concerned about water treatment due to a problem in one of their other plants where microbiologically induced corrosion of a 316 stainless steel heat exchanger was found during a plate inspection several years ago.

They do better job with maintenance of their equipment than most manufacturers and the unit that failed had been maintained regularly and properly.

I'm not smart enough to know the constituent ingredients of each fluid, but if that is known, there should be a sensor to detect the presence of one in the other fluids.

Can you tell us what the make-up of each fluid is?

Conductivity, color, pH, you get the idea.

Or look at the various liquid sensors out there and what each can sense and go from there.

As I said I did PWB process equipment. We just cooled with water.

Have you talked to Rosemount about other options?

Turbidity can be tricky, especially if the densities of the carrier and the intruder fluids are close.

You may find your solution using a fluorometer, where you might be able to send samples to a manufacturer for proving on their equipment before purchase. Been there, done that (but for a different application). Google 'fluorometer'......

What you need is a sensor that is very sensitive to fouling, however the water itself may exhibit a fouling potential (if conductivity was being utilized to compare clean vs. contaminated (water). I suppose you could make a pressure test loop (small high pressure pump, small orifice/nozzle, digital pressure guage, as a viscosity test: Log the results for "clean" water - if that baseline is flat, you may have something.

Add some Dowtherm to the test loop water: If the viscosity signal becomes extremely erratic, then droplets of the fluid are entering the nozzle, and causing back-pressure to fluctuate wildly. This might be a really great diagnostic. I don't know, since I have not tried that, but I suspect the viscosities of the two liquids are quite different.

I assume the pressure of the Dowtherm side of the heat exchanger is higher than that of the cooling water loop. Can you play with the pressure so that it is higher than that of the Dowtherm so that water will leak into the Dowtherm if there is a failure rather than the other way around? Of course, IF this is a more desirable mode of failure than what you have just described.

Can you consider coating the heat exchanger tubes with a nonconductive layer so that when a tube corrodes through it also destroys the nonconductive layer and therefore allows an electrical signal to pass between the two thermal fluids? This kind of solution may require a larger heat exchanger because of the reduction in exchanger efficiency but the electrical leak sensors should be very inexpensive.

It strikes me that if a failure of the heat exchangers is "costly", then a proven solution of $60K might be worth looking at. I guess the company beancounters are ruling here; what is considered costly? $6K, 60K, $600K? What did the recent failure cost?

Interesting problem! Please keep us posted on what you end up doing with it!

Jon.