Best practice of Level transmitter calibration

C

Thread Starter

CombustionKen

Which is the best practice for calibration of level transmitter of a gas vessel and oil tank. The used type of transmitter are of Rosemount.
 
T
You'll need to be more specific. Is it guided wave radar or a differential pressure transmitter in a pressurized vessel? Are we talking about open vessels. Rosemount makes more than 1 type of level transmitter for a variety of applications & calibration techniques vary significantly depending on the technology employed.
 
C

CombustionKen

It is differential pressure transmitter for oil tank and differential pressure transmitter in a close pressurized vessel. What are the best practice for the calibration for the transmitters?
 
Hi Ken,

If you are going to measure the level of oil in a closed pressurized vessel you need to install a capillary type DPT. If you try and use a normal piped DPT the impulse lines will block so don’t even try it.

Below is a write-up I have done for someone else on how to calibrate a capillary type level DPT. His question was for measuring a hazardous and corrosive liquid in a vessel under constant vacuum, but the procedure on a pressurized vessel is the same.

Previous write-up, ignore what is not applicable to you.

The supplier will ask you the length capillary length needed. The capillary lengths should always be as short as possible to prevent drifting, so look at the installation and decide where you are going to install the new cap DPT and double the distance from the transmitter to the top tap off point on your vessel to get the right capillary lengths. Do the calibration as normal for a capability type DPT and forget about the vacuum inside since it is irrelevant in your calibration.

IN YOUR CASE IGNORE THE PRESSURE IN THE VESSEL AND WORK WITH YOUR TX AS IF THERE IS NO PRESSURE IN THE VESSEL. THE PRESSURE IN THE VESSEL CANCELS OUT SINCE THE SAME PRESSURE IS ON HP AS WELL AS LP LEGS.

After the installation, open both pad cells to atmosphere and have a look at the displayed value. This value is the atmospheric zero value.

When you do the process zero you might find that the displayed zero have shifted slightly from the previously displayed atmospheric zero. It is therefore important to check the process as well as the atmospheric zero especially if you are working on a small span. This small shift will cause big inaccuracies if you do not work from your a process zero. (yes, similar thing to the old pneumatic DPT static alignment problem)

To do the process zero you need to install a piece of tubing temporary between the HP and LP tap off points in order to get a equal pressure on HP and LP side of tx. That is if you cannot drain your vessel and just keep the pressure inside the vessel. We normally install 316SS flushing rings for this purpose before the pad cells.

Use the displayed value as your process zero reference and just add your measured mm, multiply by 1,08(sg for acetic anhydride), to get your L/URV's.

IN YOUR CASE USE THE SG OF THE OIL YOU ARE MEASURING.

For zero (LRV) measure (mm) from the transmitter to the zero position on the vessel and multiply that with the sg of the product plus the process zero value. The same for the span(URV), from the transmitter to the 100% position on the vessel and again multiply that with the sg plus the process zero value. Your answer will then be in mmH2O so modify the units of measure in your DPT to that and put the values in directly as calculated.

If there is a difference between atm zero and process zero always use process zero for the calibration since the DPT will measure the level while under vacuum (PRESSURE)and not at atmospheric pressure.

Something else you might encounter in a application like this is that this process zero could fluctuate slightly due to the variance in vacuum (PRESSURE) and you might find it difficult to read. I had similar experiences before on a FPSO where the whole vessel moves all the time and therefore the process zero keeps on changing all the time. I have compensated for that by standing there and watch the reading fluctuating for about 15 minutes and write done the lowest and highest readings and then use the average in the middle as my final process zero. This is called splitting the error. Your final level indication after your calibration will fluctuate as well, but since you have split the process zero error, the reading will be pretty accurate and you can apply some damping on the tx to stabilize it some more.

You can use this method to minimize normal drift as well. Normal drifting in capillary type DPT's occurs mainly due to long capillaries and temperature changes.It's just one of the things we have to live with but it can be minimized with this error splitting method.
 
Top