AGA3 flow calculations


Thread Starter


AGA -3 based natural gas flow measurement at custody transfer point.

Situation: For flow measurement of natural gas the following transmitters are used.
1) Gauge pressure transmitter
2) Diffrential pressure transmitter
3) Temperature transmitter
4) Solatron make Specific gravity transmitter.

A flow computer computes the standard volumetric flow after taking the above mentioned inputs. This computer has an entry of Local baromatric pressure which is a keypad entry.

My question: Since this barometric pressure is used to calculate the absolute pressure from the transmitted gauge pressure, as and when the local barometric pressure varies the absolute pressure calculations become wrong. This gives rise to inconsitencies (according to me, am I right in this matter?) in the flow calculations. Is there a solution for this situation.


Johan Bengtsson

The error is probably not very big, but if you feel it is an issue it can be corrected.

You need an absolute pressure transmitter, either you measure the absolute pressure directly in the pipe, or you use one to measure the absolute air pressure and replace the keypad entry with that one. It would work either way, in the first case you will of course have to remove the formula for calculating the absolute pressure since you
measure it instead.

/Johan Bengtsson

Do you need education in the area of automation?
P&L, Innovation in training
Box 252, S-281 23 Hässleholm SWEDEN
Tel: +46 451 74 44 00, Fax: +46 451 898 33
E-mail: [email protected]
I don't know about the specific computer you are using, but it may allow actual measured barametric pressure. A pressure transmitter measuring absolute pressure (psia or inHg absolute) could be used to provide input.

For most applications the absolute pressure does not change enough to make a difference. But I remember back in my power plant performance testing days when we had on-line barametric pressure measurement, and for high accuracy measurements we used a temperature compensated mercury column barameter to get the data.

You might try measuring, with a barometer, the range of pressure normally encountered, and then entering each end of that range into the keypad to see how much difference it makes, and then determine if it matters. Very likely a single entry and an occassional change during times of extreme weather will be sufficient.

The error is quite significant as a change in .003 kg/cm2 pressure results in a change of around 3000 scmh.


Automation Linse

Actually, "standard formulas" like AGA-3 exist to side-step such questions. You need to review the standard and FOLLOW it exactly. They define what to measure and how to calculate the "legal value". If barometric pressure is important, they will mention it (remember, you are NOT the first person to measure absolute pressure!) There is no wiggle room for you to second-quess the AGA standard.

Standards like AGA exist to solve the classic problem of people and watches - "Someone with 1 watch knows what time it is - someone with 2 watches is never really quite sure".

You are doing "custody transfer" - which means 2 businesses are exchanging gas and cash. They will have 2 meters and the only way they can obtain roughly the same value on both meters is if both sides follow the AGA standard to the letter. Likely the AGA-3 values are never 100.000000% correct at all temp/pres/density combinations, but they are an agreed upon convention and both side will have built this into their pricing structure.

Best regards
- LynnL,