Horizontal Tank Level Measurement & Logging

[Roy Matson]

Radar would be more accurate than ultrasonic because it's not effected by the speed of sound in different vapour however a magneto-restrictive is even more so. If they really want to know how much they are using "Measure It" A bent tube coreolis meter, No straight pipe requirement and not effected by the tank profile.

Roy

 

[TConnolly]

The formula for the volume of fluid in a horizontal and level cylindrical tank with flat ends

V = l * (pi*r^2/2 - r^2*arcsin(1-h/r) - (r-h)*sqrt(h(2r-h)))

where V is volume,
l is the interior horizontal length of the tank cylinder
r is the interior tank radius,
h is the height of fluid in the tank as measured from the bottom of the horizontal cylindrical tank interior and must be less than 2*r.

(If the tank isn't level then all bets are off)

Once you have the formula any method that accurately gives you the height of the fluid will suffice, including a yardstick.

Since right now the customer is already measuring the fluid and entering it in a log manually and it seems that the issue is with using the look up table then the cheapest and simplest solution is a spreadsheet with the above formula built in. Customer enters the fluid height, and it spits out the result. The spreadsheet becomes the usage log. You could use a formula in excel using excels built in functions, or you could also program a custom function in Excel VBA, eg tankvolume(l,r,h) and that way deliver software and a billable service (or I'll do it for you for a small fee ).

If you want to automate the measuring then there are already some good suggestions. Magnetostrictive sensors will probably give you the best resolution for clean inert fluids. If the tank is actually elliptical or not level then load cells are an excellent way to measure the tank contents.

You will have the least amount of resolution in your measurement when the tank is half full, or in your tank, at 32 inches. At that level you need a resolution of .025 inches to resolve 1 gallon. .025 inches is easy for a magnetostrictive sensor. Its also easy for a 100" WC pressure sensor mounted at the bottom of the tank. IFM efector makes a submersible one you can insert from the top of the the tank and just let it lay on the bottom of the tank. Even a 12 bit A/D will give you the needed resolution with one of those, although I would recommend 14 bit or better. If the fluid has a specific gravity different from water its a simple multiplication to scale a 100" WC sensor to compensate.

 

[daba]

Once you have the formula any method that accurately gives you the height of the fluid will suffice, including a yardstick.

That comment, and all the replies so far, have made the assumption that the tank is perfect itself.

In my experience the only way to be certain of the volume of liquid in a tank is to fill-guage it, logging the transducer (whatever that may be) readings at known fill volumes, to build up a "dip-table". From the dip-table, any look-up table can be derived.

Here in the UK, Customs and Excise used to insist that brewery and distillery vessels were calibrated this way, rather than using tank design dimensions.

 

[shooter]

with 100 points the accuracy is better as 0.2 % so it is not the calculation but the sensor is never this accurate.
when he does it now with a stick then i would say use a float with a weight over a wheel and have the rotations counted.
The echo things have problems with the walls and if tank is almost empty they will not be accurate enough.
Every gasstation has something to measure the tankvolumes in the ground.
Some electronic height some with the old stick.
Pressure is also fine, however due to the density and temp it will not be accurate enough for you.

btw, if you give me the tanktable of the 500 points in excel i will give you a 100 point tanktable and a formula to calculate.
Probably the formula is faster and a lot more accurate.

If he really wants exact then why not use a small day tank
or a watermeter or a flow sensor.

I know a company that can measure microns on a waterlevel (DB pruftechnik)

Our local waterboard uses the echo for the big ones, and use a wheel with float.

 

[daba]

To add to my previous post, Customs & Excise used to use a measuring stick to dip excise vessels, then use the dip-table for that vessel (one was produced for each and every vessel they needed to dip). Since they were interested only in the volume in the tank, no compensation for density was required, the dip-measurement produced volume directly.

They would not have accepted any other method, because it would always depend on factors outside of their control (resolution, calibration accuracy, correct installation, drift, temperature compensation etc. etc.).

These dip-tables had thousands of points for each vessel, i would say a volume at each mm in dip-height. They were produced (and at great cost) by filling each tank through an excise registered flowmeter, and dipping at various intervals. I do not think they did the "thousands" of points mentioned above, intermediate values would have been calculated by interpolation.

The dip method measures the liquid height from a guage point on the vessel top door, and because the tables were produced by filling the vessel with known volumes, the method is entirely suitable for vessels with odd shapes, external pipework and/or heat exchangers, "inclusions" such as heating coils, CIP sprayballs etc.

Therefore any method which can accurately measure the liquid height will result in an accurate volume measurement, providing care is taken to produce an accurate dip-table.