Best way to calibrate pressure sensor for domed-end horizontal tanks

I just realized that the three new horizontal tanks tanks we're installing this weekend will need a CPT and some in order to allow the provided 4-20mA pressure transmitters to produce an accurate volume reading due to the domed heads. I'm guessing the heads are torospherical, but I can't really tell from the drawing and the tanks aren't on site yet. It's imperative I have the calculations accurate by Sunday afternoon. The tanks arrive on site Friday morning! To make things worse, I'm on vacation and have no access to RSLogix to pre-program anything.

I'm using an SLC 5/05, gathering data from an analog input module in an Automation Direct remote I/O over Devicenet.

First, I'll calibrate the senders against a pressure head using a standpipe, then I'll SCP to arrive at a depth in m (D)

Here's the formula I'm going to use to calculate the volume in the heads, in cubic meters:

V = 0.1618*((3*Tank Diameter*(D^2)^2)-(2*(D^2))

To convert this to CPT, this is what I figure:

CPT N7:49 ( 0.1618 * ( ( 3 * N7:48 ) * ( ( N7:47 ** 2 ) ** 2 ) ) ) - ( 2 * ( N7:48 ** 2 ) )

N7:48 = tank diameter
N7:47 = depth of liquid in the tank
N7:49 = volume of liquid in the heads

Would anyone mind trying this on a 5/05 for me and letting me know if it works?

Thanks!

I am too tired to go over your math, but here is a thread on the subject, and I think there may be another one somewhere that details how to do the domed ends.
http://www.plctalk.net/qanda/showthread.php?t=49969&highlight=horizontal+tanks&page=2

You might also search MrPLC. I am pretty sure the dome end math was discussed one of these two places.

double-check my entries - my eyes are bleary at this time of night ...

that NEGATIVE volume doesn't look too good ...

this is with a 5/04 processor - should be the same as a 5/05 ... if you think it will make a difference I can try one of those when I get to the lab tomorrow ...

tip: give us some reasonable sample values to work with ... I just plugged in 2 and 1 for examples ...

.

Thank for the replies.

I mixed up tank diameter and depth of liquid.

Uh, yeah, now that I am bright-eyed and bushy-tailed, your formula doesn't look anything like pi*r^2h+ends...

EDIT: After some coffee, I realize that the formula for volume of a cylinder may be of little or no use in this case. I have only seen a couple of situations where volume was needed with an odd shaped tank, and both times, look-up tables were used.

This formula just calculated volume of the ends. Not sure how it works, but I used in in a previous project. I'm guessing I dug it up from somewhere.

I think what you want is the area of the filled portion of hte circle times the length to get volume. (forget the ends for now).

So your level transmitter gives you the cutoff point of the chord for which you must account for.

gotta run but here's my 1st gut goole result:
http://www.regentsprep.org/regents/math/geometry/GP14/CircleSectors.htm

AND, if I recall, one thread somewhere used the conical formula for the ends for simplicity, which will get you very close to the math you need for the domes.

I would set it up programmatically for a lookup table. Undoubtedly, you may benefit from being able to:
a) prove the accuracy
b) adjust for alterations, obstacles, irregularities

rguimond said:

....It's imperative I have the calculations accurate by Sunday afternoon....

Click to expand...

You stand a good chance of getting the calculations accurate with the help of the formulae given, but that doesn't mean you will get an accurate result of the volume in the tank....

Rarely is a manufactured tank precise to the design drawings and dimensions. Rarely are they perfectly round, and many tank ends have 2, sometimes 3 different radii. You will also need to account for any "inclusions" in the tank if there are any, and the volume in any open pipework connected to the tank (e.g. to the closest closed valve.

The sure-fire way to get this system calibrated is to progressively fill the tank through a known-to-be-good flow-meter (positive displacement), and log the pressure readings and volumes. This is exactly how brewery vessels in the UK used to be calibrated, when an accurate volume was essential for duty purposes. The height of the liquid was taken with a dip pole from a reference point on the top of the vessel, not an analog signal. Customs and Excise used the dips and dip-tables to calculate volume, and hence the duty payable.

Nowadays, most people opt for weighing the tank on load-cells, and calculating the volume from the density of the contents.

Take a look at this calculator....

http://vps.arachnoid.com/TankCalc/

...and it includes a great discussion of the maths involved.

daba said:

Take a look at this calculator....

http://vps.arachnoid.com/TankCalc/

...and it includes a great discussion of the maths involved.

Click to expand...

That's a golden nugget...outstanding link.

still best way is a table and correct calibration.