Extremely corrosive level measurement

[patrussell]

I have been asked to provide continuous level measurement on a tank of molten salt that is kept between 500F and 1000F.

Right now we are using level switches to provide point control but this is a little too course for some of our product and I need something better.

Anyone know of a product that can handle this environment?

Pat Russell

 

[ScottyUK]

How far away from it can you get? If you can get far enough away for the ambient temperature to be more reasonable, use a non-contact measurement such as microwave or radar. I'm not sure how laser systems would perform with a surface radiating large levels of IR radiation, but that might be an option also if you can keep the optics clean.



----------------------------------

Sometimes I only open my mouth to swap feet...

 

[controlnovice]

I think Magnetek has a high temperature rated guided wave radar. I don't know how high though..... Check them out.

______________________________________________________________________________
This is normally the space where people post something insightful.

 

[danw2]

Presumably the ambient is not excessive, it's the radiant or oonvective heat that's of concern for non-contact instruuments.

So the concern is what temeperatures the sensing elements 'see'. The processing electronics will escape most of the heat, being behind the sensor.

Normal ultrasonics can take up to about 200°F, high temp up to about 300°F. Radar horns can take even higher temps.

Radar is sensitive to the dielectric constant of the material, where it works great with high dielectrics (good conductors) and performance falls off for materials that are very good insulators. Even though salts make great conductive solution whne dissolved in water, is molten salt conductive in the absence of water? I honestly don't know. A vendor will take a shot and see.

The salt pots I've seen in heat treat shops always have slaggy stuff floating on the top. If yours does too, then the slag will definitely bounce a radar signal.

The opening you have to shoot at is important because ultrasonic and radar beams expand considerably, forming 'cones' as the signal progresses.

One advantage of laser is it's very narrow beam. I've never used laser so I, like ScottyUK, have no idea what laser is susceptible to. Call a vendor and find out.

Dan

 

[Ashereng]

Hi Pat,

I'm a bit more old school than my counterparts here, and like to go simple. I use very little non-contact microwave, have not used laser (as of yet), definitely not nuclear, and usually try to go cheap (my clients like it that way).

When you said that the switches are a little too coarse, what accuracy do you need, and what denomination do you need?

Usually, the problems with level measurement is temperature, corrosion, a second phase, or uneven surface. In your case, you already have a level switch that works (takes care of heat and corrosion), you didn't mention uneven surface or second phase. So, with that in mind, here are some food for thought:

1) More level switches. The problem often is you don't have any more "taps/nozzles" or you can't drill any more holes. So, maybe not so good.

2) Some point measurement switches have 2 or 3 points in the same unit. Magnetrol's point level switches, the Echotel line, has a dual point level switch (you get two levels in the same unit). You can use these in place of your single point units. The downside to this is that these are not side entry mounts, so you need holes on top. So, maybe something else.

3) A bridle with float and magnetic pickup. What material are you using now for your level switches? Use the same material for the bridle and float, and then use a magnetic pickup to send the 4-20 mA signal to your control system. I am assuming that molten salt has a fairly high density, so floats should work.


You didn't say where you are, but that's okay. I have recently dealt with Magnetrol on some water processing applications, and they have a wealth of knowledge when dealing with severe service. Might be worth a call to their rep/engineers.

Good luck.


"Do not worry about your problems with mathematics, I assure you mine are far greater."
Albert Einstein
Have you read FAQ731-376 to make the best use of Eng-Tips Forums?

 

[itsmoked]

I like Ashereng's point. A float with a long vertical rod would allow many different cheap sensors to note the top-of-rod position.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[nbucska]

I wonder it it wouldn't be simpler and cheaper to weight
the whole tank with a strain gage?

Plesae read FAQ240-1032
My WEB: <

http://geocities.com/nbucska/>

 

[Ashereng]

nbucska,

I did that once - put an entire tank on load cells, custody transfer. I believe that the operations guys still curse me. I wouldn't do it again.

Way too much capital cost. And a lot more maintenance than one would initially think.

"Do not worry about your problems with mathematics, I assure you mine are far greater."
Albert Einstein
Have you read FAQ731-376 to make the best use of Eng-Tips Forums?

 

[HEC]

I like the simplicity of Ashreng's suggestion, simple and cheap. Team that with a LVDT (Linear Voltage Distance Transducer) and a continuous level measurement will be given.
Also need to reinforce the problems associated with the weighing method. There are so many issues with this that it is not worth the bother. I also used this method for a batch tank, pharmaceutical application, and did not want any projections or seals within the tank. The calibration hysteresis for the weighing system was horrendous. Another issue witht his specific application would be, as it is a molten salt bath I presume items are placed into it for heat treatment, in which case the apparent density would alter and provide false level readings.

Mark Hutton

 

[unclesyd]

We use nuclear level controls on 18 jacketed vessels operating at 600F on a continuous basis. The source is mounded off center to miss 16"-20" dia internal shaft. The detector is mounted on top about 6" above the insulation and is air cooled. We have had very few problems with these instruments over 30 years.

http://www.ohmartvega.com/

 

[unclesyd]

Just remembered what was used prior to the use of nuclear level controls on the above mentioned equipment.
They used a dip tube arrangement using N2 heated to the polymer operating temperature(600F). The only problem was this arrangement had to cleaned about once each week. The nuclear device was a much needed improvement.

Still remembering.
A dip tube arrangement was also use on some molten salt reactors operating around 850F. This one got me as the reactors were upflow but they wanted an alarm if for some reason the salt got closer than a certain distance from the
top tube sheet. This system was used until a mechanic recommended that the overflow line to the day tank be
moved to the desired level instead of next to the tubesheet. Though never confirmed I think this was also to be able to change the salt level based on gas input to reactor. Reactors were designed to have a pretty good turndown ratio but this feature was never used as the rates are always wide open and how to get more. I don't know what type of compensation was used for changing physical characteristics of the salt.
I just called to check on the dip tube status on the reactors. It is still on the control panel display but there is no controlling function or alarm associated with it. Someone is too lazy to cleanup the display.

 

[richs]

Hiya-

Someone on another thread mentioned weighing the tank?

A little thought might make the mechanical/thermal isolation a pretty doable thing. Heat shields and levers and load cells, oh my!

Cheers,

Rich S.

 

[richs]

Opps. My bad. Please disregard earlier post. Don't know
why I missed earlier answers to weigh tank.

Sorry.

Cheers,

Rich S.

 

[glowing]

We use Endress Hauser IR for a similar aps...been in service for years...trouble free

 

[glowing]

oh yeah...our total install including transmitter, probe, welding the flange, running 110V, installing trasmitter, programming, running tec cable and laour was about 2600 -2700 bucks...so it is a cheap option that remains fairly maintenence friendly / free

,,V,