Making measurements in closed tanks ?

[ArFa]

Hi all,
Can you give me some ideas about how to measure fuel level and oil level in a closed tank ?
It means we can't make any hole, so I don't think that ultrawave sensors will work.

 

[milind.a.kulkarni]

One simple way if possible..... make the weight of the tank for different fuel level and do the correlation ( Here my assumption is tank is small enough such at it can be put on weighting scale) .... if that is not possible mostly I think you need to go for non contact approach but that too require some kind modification to the tank as you said those are not doable.....so need some more input.....

Good Luck

 

[ArFa]

One simple way if possible..... make the weight of the tank for different fuel level and do the correlation ( Here my assumption is tank is small enough such at it can be put on weighting scale) .... if that is not possible mostly I think you need to go for non contact approach but that too require some kind modification to the tank as you said those are not doable.....so need some more input.....

Good Luck

Thanks a lot for your ideas.
Do you mean that weight measuring will work?
I had same idea but I'm still projecting the project and I need some ideas to start with.
Thank you

 

[milind.a.kulkarni]

This is not very accurate way of measurement of level but it works for some resolution and measuring range-

refer -

**broken link removed**
**broken link removed**

Good Luck

 

[ArFa]

This is not very accurate way of measurement of level but it works for some resolution and measuring range-

refer -

**broken link removed**
**broken link removed**

Good Luck

Thank you very much for your help.
Can you suggest any sensor or any device for this method:
Thump the water tank and measure the frequency at which it rings. The higher
the water level, the lower the pitch.

 

[kripacharya]

The thump-and-measure sounds like a fun project. And it will probably work too, once you can calibrate it & filter out the noise components. Oh yes, it'll require a bit of signature analysis as well. But give it a try, it sounds like a fun approach.

Alternately you can try a capacitive method. You attach plates outside your container, and measure the capacitance value. This of course depends on what material the container is made of. Metal container means no-can-do. Others materials - possible.

So a little more details about the container, where its used, whats it made of, can you take it elsewhere and put it on a scale, and so on. Would be useful.

 

[PA4TIM]

Sometimes TDR is used to measure the level on a tank. HP has a appnote or benchbrief about this. (can not find it so quick but it is somewhere from the 60's. They even measure how much fluid is floating on an other fluid. But you must be able to make a coaxial pipe in the tank. This has a certain impedance and the fluid in the pipe will work as a dielectric and change that impedance. The reflection then will change. The time between pulse and reflection then can be measured and you can calculate the level from that.
https://www.hpmemory.org/an/pdf/an_62.pdf about the basics

 

[kripacharya]

Sometimes TDR is used to measure the level on a tank. HP has a appnote or benchbrief about this. (can not find it so quick but it is somewhere from the 60's. They even measure how much fluid is floating on an other fluid. But you must be able to make a coaxial pipe in the tank. This has a certain impedance and the fluid in the pipe will work as a dielectric and change that impedance. The reflection then will change. The time between pulse and reflection then can be measured and you can calculate the level from that.
https://www.hpmemory.org/an/pdf/an_62.pdf about the basics

The OP had a requirement to not make any hole or intrusion into the tank. Doesn't your method require such a 'hole' ?

 

[PA4TIM]

I do not know. There is not much info on the tank. But if it has oil or fuel it could be they design the tanks. On the otherhand it is possible he talks about the petrol and oiltanks of a twostroke moped, or the tanks of a mammuttanker (ship) or underground tanks of a petrol station.

A friend of mine desings tanks for a living and I told him about the TDR and that gave him ideas and they now implement it in some tanks.

If the tank has a fill cap (many tanks have) you can enter the "sensor" through there or make an adapion to the fillercap with a TDR line or capacitive sensor, or a floater or whatever sensor sensor mounted to it. All depents on what kinda tank this is. I was just thinking outside the box ;-)

 

[kripacharya]

...I was just thinking outside the box ;-)

good one.. LoL

 

[Venkadesh_M]

And also you can fit a capacitive transducer on the cap itself, try searching liquid level measurement using capacitive transducer, that can be only used for non conductive liquids..

 

[kripacharya]

And also you can fit a capacitive transducer on the cap itself, try searching liquid level measurement using capacitive transducer, that can be only used for non conductive liquids..

and non-conductive tanks as i mentioned in #6

 

[Venkadesh_M]

yeah It is possible with metal tanks also with only one electrode in the cap and another one as tank...

 

[FvM]

The original question is unclear in one point, it doesn't say if the measurement might use existing "holes" on the tank top, as the capacitive sensor according to post #13 does. Obviously every tank has a filling port, but it's not necessarily accessible.

If this isn't an option, ultrasonic measurements through the tank wall can be considered too. There's a number of industrial sensors using this method. Either continuous time-of-flight tank bottom to liquid surface, or multiple level switches on the tank side. The latter can work if the tank construction or the nature of the filler doesn't allow a t.o.f. measurement.

For both designs, a certain knowledge of ultrasonic sensor design e.g. regarding acoustic impedance matching is required. More a field of profound industrial design than for a DIY project. The same is true for other techniques mentioned in this thread.

 

[ard]

I assume your tank is non-metallic. In that case, the capacitive sensor can be formed by sticking 2 aluminium/copper tapes to the the outside of the tank.

 

[ArFa]

I'm sorry for my bad question :S
My issue is to measure fuel in a genset tank. So I assume that this tank is from metal and it's very big size.
Can I use capacitive sensor ?

 

[kripacharya]

I'm sorry for my bad question :S
My issue is to measure fuel in a genset tank. So I assume that this tank is from metal and it's very big size.
Can I use capacitive sensor ?

Riiight.

So answer is No, can't use capacitive.
And coupled with the limitation that you cannot have a 'new' opening through which you can mount/ insert something, leaves almost no choices.
Your 'thumping' option also seems moot, since the genset would drown out whatever you can hear or automatically sense.

A possible candidate is the weight measurement method, which might still work. But you have to decide whether its viable to place a pressure transducer someplace under the tank. It could be near the edge too, as long as the tanks weight actually rests on the transducer, and not some other support.

Also a possible, and which might actually work/ be made to work, it to use ultrasonics externally. Here's a nice link to show how :https://www.olympus-ims.com/en/applications/ultrasonic-measurement-liquids/
If practical, this imho meets ALL of your requirements !!

i suggest you explore the option to make an opening someplace to be able to attach/ insert something at least, if you want good readings.

 

[FvM]

Also a possible, and which might actually work/ be made to work, it to use ultrasonics externally. Here's a nice link to show how :**broken link removed**

The Olympus AN uses a thickness gage instrument for the measurement, which is a nice idea, if you have one, but "a bit" expensive otherwise. As I already mentioned, industrial "through tank wall" level detectors have been made for the application. Unfortunately I don't remember a brand, I have seen it on an industrial control trade show (Interkama) 10 or 15 years ago. So there there's no doubt that the method can be implemented.

 

[kripacharya]

The Olympus AN uses a thickness gage instrument for the measurement, which is a nice idea, if you have one, but "a bit" expensive otherwise. As I already mentioned, industrial "through tank wall" level detectors have been made for the application. Unfortunately I don't remember a brand, I have seen it on an industrial control trade show (Interkama) 10 or 15 years ago. So there there's no doubt that the method can be implemented.

Yes, i saw your earlier post in this thread on the same lines, and it got me thinking. Maybe such a solution can be implemented now with looser specs and hence at a more reasonable cost. That might be sufficient for the op's requirements, and fit all his constraints.

 

[FvM]

You'll find a lot of commercial sensors under the keywords "non-invasive ultrasonic level measurement" respectively "non-invasive ultrasonic level switch"

e.g.: