How to measure and display liquid in a vessel?

[avz]

Hi
I hope that you will be able to provide me some advise. I have a vessel that contains about 14 cubic meters of liquid. (canola oil, in this case) and I would like to add a device that will be able to measure and display the amount of the liquid in the vessel. (ultrasonic sensor?). the problem is getting worse, since the upper part of the vessel is cylindrical and the lower part is conical - so that the amount in the vessel is not linear.
waiting for your ideas.
Thanx.

 

[rumpfy]

Dont really know what the application is and hence the problem, but a 'mass' measurement using load cells would be my first suggestion. Problem is that the mass measurement is affected by the density of the material, so this approach might not be suitable.
The mass of 14 m*3 of cooking oil could be around 12 tonne so the solution is not mickey mouse. This is a seriously complex manufacturing problem for which proper engineering advice is called for. Its unlikely your normal fish and chip shop has such a vessel.
hope this helps.

 

[ChrisP58]

Can you weigh the vessel?

If you use some sort of level sensor you will have to come up with some sort of non-linear conversion table.

 

[dr pepper]

Allthough oil is not as volatile as petrol, for safety whatever you make ought to be designed so as not to create a spark for obvious reasons.

How about something like a vehicle fuel guage with a float, you can get around the non linearity by marking on a panel meter the no. of litres in the tank by pouring in set amounts.

 

[cowboybob]

Presumably the vessel can be emptied by a simple bottom drain. If access to the drain is possible, how about adapting a resistive type pressure sensor/readout (perhaps an automotive engine oil type) or, simpler yet, a low pressure mechanical gauge.

You'd have to calibrate the readout to the quantity of oil in the vessel, of course. This might be done with an empty tank that you then fill in known increments.

 

[ericgibbs]

Hi
I hope that you will be able to provide me some advise. I have a vessel that contains about 14 cubic meters of liquid. (canola oil, in this case) and I would like to add a device that will be able to measure and display the amount of the liquid in the vessel. (ultrasonic sensor?). the problem is getting worse, since the upper part of the vessel is cylindrical and the lower part is conical - so that the amount in the vessel is not linear.
waiting for your ideas.
Thanx.

hi avz,

I guess as its a food product oil, the vessel is made from food quality stainless steel, so cutting into is not an option.??

Your ultrasonic sensor would work OK, looking down from the top inner face of the vessel and reflecting from the oil surface.

You will have to choose a suitable ultra sound sensor frequency that will give the resolution you require.

The change from cylindrical to conical volume would not a problem for a micro based system.
The MCU maths could be applied to calculate the 'amount' in the vessel as it could be programmed to change over its calculations at the boundary between cylindrical and conical , as it will be programmed to 'know' the depth at which this occurs.

E.

 

[nsaspook]

You might want to look at bubble tube liquid level system to measure the oil. I've used them mainly with thick or corrosive liquids at work.

Bubble and other level technologies:
https://blog.lesman.com/2012/01/04/bubbler-systems-for-measuring-liquid-level/
https://www.lesman.com/unleashd/catalog/level/level_siemens_air-bubbler-system.html

 

[kinarfi]

A load cell can be added to each of the leg the container and then calibrated to indicate gallon, pounds, even ounces. depends on your budget. Properly mounted and set up, load cells are very accurate.
Can you post a photo of your tank?

 

[avz]

Hi Guys
Thanks for your response. I have to "digest" all the info provided. at a first glance, the bubble method is not applicable here because adding air to canola oil shortens its life. in fact, there will be a "nitrogen blanket" that will isolate the oil surface from the air in the vessel.
Load cells - its quite a problem now, since the vessels are already positioned. so what seems to be most appealing are the pressure method - if a suitable sensitive enough sensor is available or the ultrasonic one.
Please keep on providing ideas.
p.s. - as per kinarfi's request, I'll post pics.

 

[misterT]

Ultrasound through "nitrogen blanket".. hmm. I have no idea if it is a problem or not, but it does effect the speed of sound. If the thickness of the "nitrogen blanket" changes then you would need to re-calibrate the system. Well.. maybe the effect is very small.. How accurately you need to measure the liquid level?

 

[alec_t]

Ultrasound through "nitrogen blanket".. hmm. I have no idea if it is a problem or not, but it does effect the speed of sound.

Speeds are 343m/s and 349m/s respectively in air and nitrogen at 20°C. Not a huge difference. Temperature has a significant effect so would have to be compensated if the tank is subject to large temperature changes (e.g. speed in air drops to 331m/s at 0°C).

 

[misterT]

Speeds are 343m/s and 349m/s respectively in air and nitrogen at 20°C. Not a huge difference. Temperature has a significant effect so would have to be compensated if the tank is subject to large temperature changes (e.g. speed in air drops to 331m/s at 0°C).

Nice facts.. thanks. Like I said I was just wondering what the effect would be, but I was too lazy to find out the numbers myself

Now I'm wondering how this nitrogen blanket works.. correct me if I am wrong, but I think nitrogen is lighter than air (which includes plenty of nitrogen).. Nitrogen is lighter than oxygen.. is the tank filled with nitrogen? Otherwise the nitrogen would float to the top.. not at the surface of the liquid.

 

[cowboybob]

The static pressure of water is 0.433 PSI per foot of height (assuming 1 cubic foot of water). Canola oil would be slightly higher.

If your 14 cubic meter tank is 6 ft tall (and full to the brim with oil), irrespective of its volume or shape, the static pressure at its base when full would be 6 X 0.433, or 2.6 PSI.

If the tank were 7 ft tall, the static pressure would be about 3.0 PSI. And so forth.

Any decent low pressure liguid gauge with an appropriate range (in this case, 0-3PSI) would give you a linear estimate of the total volume in the tank, such as:

https://www.coleparmer.com/buy/product/6192-pressure-gauge-1-4-npt-m-process-connection-0-3-psi.html (It's brass components would not, I should think, be of particular concern for the volume of product you indicate.)

The ratio of current pressure to full pressure, multiplied by the total volume of oil, would give you the current volume of oil.

Obviously as the tank neared empty the estimate would become a little harder to deduce.

 

[kinarfi]

Nitrogen could be use with the bubble method, but strain gauges can be glued to the legs of the tank and then calibrated.
https://www.omega.com/subsection/transducer-quality-strain-gages.html

 

[misterT]

Nitrogen could be use with the bubble method, but strain gauges can be glued to the legs of the and then calibrated.
https://www.omega.com/subsection/transducer-quality-strain-gages.html

Strain gauges are painful to get working and calibrated properly if you try to build the system from scratch. Maybe a good idea if you buy complete sensor modules, but then it becomes expensive method just to measure liquid level.

 

[alec_t]

Any chance of fitting the tank with a sight glass tube?
https://encrypted-tbn0.gstatic.com/...HKkljpDSw4Vb7mBZiVgfhV485HhG5zUBR-DhmHWxB8Jkf

 

[misterT]

Any chance of fitting the tank with a sight glass tube?
https://encrypted-tbn0.gstatic.com/...HKkljpDSw4Vb7mBZiVgfhV485HhG5zUBR-DhmHWxB8Jkf

That is a good idea.. easy to replace and fix etc. You can fit all kinds of sensors in that and also see visually the liquid level. If that is something that can be installed.. of course.

 

[avz]

Really thanx for all the input. very helpful. as per the sight glass - its not a good idea in this case because the vessel is placed inside a sort of a shed that is cooled in order to keep the oil in an ambient temp lower than 24deg C, so the access to it is not too comfortable - that's why I want to use a sensor and a display that will be mounted on the outer side of the shed. another con point for this method is the fact that the vessel is almost 6meter (ca. 18ft) high, but thanx anyway.

 

[alec_t]

The displacer method shown in this link (which also rubbishes the sight glass ) looks promising and relatively simple to implement:-
**broken link removed**

 

[kinarfi]

photos, especially of the tank support system, please