Level measurement problem 1

[gilden]

We have a level measurement problem that seems trivial but until now no valuable solution has been found. We need to measure the level in a small glass reactor (jacketed) of 20 L (diam. 20 cm, height 70cm). The reactor has filterplate at the bottom. Solvent (DMF) is pumped in the reactor at the top by means of a distribution ring with tiny holes, all directed towards the reactorwall. At the same time liquid is drained from the bottom by means of a 2nd positive displacement pump. The level is to be kept constant by controlling the speed of the pumps by means of a level probe. The medium to measure is a polystyrene swollen in DMF. Differential pressure measurement does not work (the depression created by the pump disturbs the measurement). Capacitance measurement does not work since the medium is not conductive enough. Radar and guided radar level measurement have the disadvantage that dead zones are present at the top (and the bottom), which is inconvenient in such a small reactor, and that the liquid droplets passing the radarwaves might also disturb the measurement. Does somebody know another technique that could work in this application? I am personally thinking about a float level sensor, but I do not know whether such small floats exist and I am not sure whether the float itself will not get stuck by the swollen polystyrene beads (getting in between the float and the guide tube. Does somebody have experience with such systems? Please keep in mind the dimensions of the reactor.

 

[hacksaw]

floats can be made to work, ditto optical or accoustic methods.

the cheapest, however, is to control it manually.

 

[jay165]

Slap on a bridle on that puppy and use either a K-Tek or a Fisher level transmittter:

http://www.ktekcorp.com/

http://www.emersonprocess.com/fisher/

 

[aspearin1]

Sight glass? Is this simply a measurement problem, or are you expecting to tie it to automated controls?

ChemE, M.E. EIT
"The only constant in life is change." -Bruce Lee

 

[khan101]

How about multiranger's ultrasonic or may be a level via weight measurement if the reactor is not very big.

 

[controlnovice]

For these type of conditions, I'd recommend weigh cells for measuring the level. It's non intrusive. But, can be expensive for an existing vessel.

 

[jsummerfield]

Although the ultasonic sensors may work, at 200 mm x 700 mm I kinda like the load cell approach. My sailboat bilge has a level switch but industrial float level sensors do not seem suitable for such a small reactor application.

John

 

[gilden]

To Aspearine1: yes, the level measurement is to be tied to automated controls, so a sight glass is no solution.

 

[moltenmetal]

Anything you can see, you can get an optical device to detect. Bet you can use a series of optical level switches on an external level glass to give you what you want here. Only one problem- it's no good if the level tube is plugged with swollen beads...

A weighscale will do the job too. For such a small unit, don't use loadcells per se- just put the whole unit on a platform weighscale of the correct range and precision. And make sure all connections to the vessel are made sufficiently flexible to avoid affecting the weighscale's readings. If there's an agitator, don't attempt to mount the agitator separately to a frame, as this will definitely affect weight readings. Weigh the agitator, motor, gearbox, stand and all. Subtract all this invariate dead load from the reading.

 

[btrueblood]

Umm. Won't the weight of a given level of liquid change depending on how much the beads swell?

In other words, I'm not sure that keeping the mass/weight of the container constant will keep a volume level constant for this case. I think a float-type level sensor would do better if the absolute volume/level must be kept constant.

Ben T

 

[m777182]

I would take an optical level control system from my Cole Parmer catalog; it works on the basis of back up reflectance of an IR source that is incorporated into the probe.Small and cheap.
m777182

 

[steveen]

gilden,

I assume the speed of the pd pump is adjusted to control level and unit is under vacuum.

MOltenmetal and m777182 are on the right path

Two options for furthur research.
1. I recall a switch that was based on light transmission. If the fluid is clear, more light will pass a liquid filled portion of the vessel than the vapor space. Should be able to set up a simple control system with two switches. Think fiber optics to get under the insulation. A good glassblower should be able to hang a couple of fibers on the vessel -- if one is still alive.... Hopefully the vessel/fiber interface will not distort the signal quality. This is a suggestion for furthur investigation only. I have never seen this attempted. Suggest several trials on inexpensive similar glassware.
2. Also, I vaguely remember the transmittion rate through a fiber decreases greatly if the fiber is wetted. Prehaps you can place a couple of loops at different heights within the vessel. You may have to try different light frequencies, IR or UV.

Good luck