Wastewater 'decant' tank 2

[controlnovice]

We have a wastewater collection tank in a Chemical plant. From here it is pumped to a mixing tank before pumping to an incinerator.

We are trying to determine the interface between the water and organic layer (on the bottom). I was thinking a float that would 'sit' at the interface, but we never know the specific gravity of either the water mixture or the organic layer.

The waste products are:

Methanol 15-25%
Formaldehyde 12-15%
Phenol 5-8%
Glycols 1-2%
Solvents 1-2%
Water: balance

Everything except the Phenol is soluble in water. The Phenol will settle to the bottom of the tank.

Can anyone recommend a method to detect the level? It does not have to be a continuous level reading. If we can insert something on the side of the tank to determine if we are above or below the interface layer, that would be OK. UV, NIR, float, Conductivity? Anything else?

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

 

[bimr]

Investigate capacitive or radar (expensive and accurate) level transmitters.

Aquasant makes a product that may work.

http://www.aquasant-mt.com/Interface_detection.61.0.html?&L=1

http://www.aquasant-mt.com/continuous_level.105.0.html?&L=1

More suppliers are Ohmart, Endress+Hauser, Thermomeasuretech, Ronan.

 

[elguero]

Finding the interface is not all that easy. On paper, many methods look easy; in practice the separation is quite difficult. You probably can’t do it with simple decantation, but since that is most straight forward, you should give it a try.

Here is the major problem with decantation: the interface between phases can take days to develop, then once developed, it is not easy to find. Formation time can be shortened by removing all solids (filtration is the indicated treatment); emulsions can be broken this way.

All detection methods to find the interface are dismally bad. Conductivity looks good on paper, but an electrode can become coated with organic and give false readings. The best method might be to use two batch vessels, side by side, (with a pressure leaf filter, precoated with diatomaceous earth, in the line upstream). Here, one vessel would accumulate new material while the other allows undisturbed separation time. Each vessel should have tap points, to take samples (at least 10, but 20 would be better) along the vertical sides. Then the operator takes several samples where he(she) thinks the interface is, and subjects it to visual and other external tests (conductivity, index of refraction, density, etc.). This way the interface is placed between two tap points. The measuring instruments can be cleaned and recalibrated each time.

The rest is easy.

PS: Methods other than decantation might be more effective. Let me know if these other methods might be of interest.

 

[dirkbouma]

Does the density change in each of the phases signficantly? Are there suspended solids? elguero notes separation issues without mentioning parallel plate clarifiers or centrifuges, both which are used for more rapid approaches to separation if needed. bimr's approach is sensible, but there might be less costly methods:

One thought, would be to try a 'sludge judge'. If you have something suspended in either phase, the difference would show up in transmissivity.

A 'fish-finder' (sonar) might be a relatively cheap approach, if the response off the contact between phases is good.

Let us know what you come up with!

 

[elguero]

Clarifiers are not reliable when three phases are present: two liquid, one solid. Centrifuges might work but are expensive and require a good deal of maintenance -- never the less centrifuge might be worth looking at. Probably a rough estimate should be done (including discounted cash flow for maintenance)of filtration versus centrifuging.

As dirkbouma said: Let us know progress. Good luck.

 

[controlnovice]

Well, we were finally told that there really is no separation in the decant tank happening, unless it's there for a few days. Which it's not.

We've gone to temporarily drumming off the waste from the product we know we have problems with, and will slowly introduce them into the incinerator.

Eventually, we will have separate tanks feeding into the incinerator.

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

 

[stanier]

Would a sight glass work? A glass panel inserted intot he side of the tank?

 

[elguero]

I thought that's what I said in my first communication: "...the interface between phases can take days to develop, then once developed, it is not easy to find. Formation time can be shortened by removing all solids (filtration is the indicated treatment); emulsions can be broken this way."

If you can't wait for a period of say 1-3 days, then consider filtration followed by evaporation. It must be controlled evaporation in which the vapors are totally condensed and captured. Refer to the evaporation chapter in "Unit Operations," by McCabe and Smith. Hire a chemical engineer, then the wheel not not be reinvented.

Evaporation will leave you with a volatile portion consisting of solvents, methanol, phenol, formaldehyde, and water. The not so volatile portion will consist mostly of glycol and water. Then the volatile portion you might be able to sell to a cement company as fuel for their kilns, and the not so portion can be redistilled or perhaps recycled.

 

[controlnovice]

elguero - thanks for the information. Your first response is what led me to ask the chemical engineers to verify if there was even any separation.

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