Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
flow compensation from flow meters
- Thread starter 200272522
- Start date
API publishes a technical paper for correction of meters metering gas. To hope to anwer your question more accurately it is necessary to know what type of meter you are using. A propelle meter is corrected differently from an orifice plate or a rotameter.
Vstd = Vflow * Pflow/Pstd *Zstd/Zflow * Tstd/Tflow Where
V = volume, P= absolute pressure, Z = supercompressibility factor, T = absolute Temperature, std = standard conditions or compensated for flowing conditions, and flow= the flowing conditions.
There are other compensations made to the meter itself that adjust readings, such as reynolds number, expansion of the steel, compensating for average density or density changes, coefficient of discharge adjustments, and on. All of this is covered in API MPMS standards.
V = volume, P= absolute pressure, Z = supercompressibility factor, T = absolute Temperature, std = standard conditions or compensated for flowing conditions, and flow= the flowing conditions.
There are other compensations made to the meter itself that adjust readings, such as reynolds number, expansion of the steel, compensating for average density or density changes, coefficient of discharge adjustments, and on. All of this is covered in API MPMS standards.
Michael2006
Chemical
As Jim says, the correction depends on the operating principle of the gas flow meter.
If the measurement is based on differential pressure (such as with orifice plate, venturi or rotameter type) the correction is approximately:
VTRUE,O = VI,O*SquareRoot(RhoCAL/RhoO)
Where:
VTRUE,O = actual operating gas flow in m3/h ft3/h or whatever
VI,O = flow rate indicated by flow meter in actual m3/h, ft3/h, etc.
RhoCAL = density of the gas at which the flow meter shows the flow correctly
RhoO = density of the gas at the current operating conditions
The gas densities are calculated from:
Rho = MolWt*P*T/z/R/T
With
MolWt = Molecular weight of the gas
P = Absolute pressure
T = Absolute temperature
z= Compressibility factor of gas at T and P
R = Ideal gas constant (If P is in bar abs, T in Kelvin, Rho in kg/m3 then R = 8314.472)
If you just have T and P and the conditions aren't too different from the calibration conditions, assume z doesn't change much and use:
VTRUE,O = VI,O*SquareRoot(PCAL/PO*TO/TCAL)
with the O subscript being for the current operating P and T conditions and CAL the P and T conditions at which the flow meter reads correctly.
There are standards covering some kinds of meters (e.g. ISO 5167 for orifice plates and venturi tubes) and manufacturers, e.g. of rotameters-like devices, may have their own more sophisticated corrections.
If the measurement is based on differential pressure (such as with orifice plate, venturi or rotameter type) the correction is approximately:
VTRUE,O = VI,O*SquareRoot(RhoCAL/RhoO)
Where:
VTRUE,O = actual operating gas flow in m3/h ft3/h or whatever
VI,O = flow rate indicated by flow meter in actual m3/h, ft3/h, etc.
RhoCAL = density of the gas at which the flow meter shows the flow correctly
RhoO = density of the gas at the current operating conditions
The gas densities are calculated from:
Rho = MolWt*P*T/z/R/T
With
MolWt = Molecular weight of the gas
P = Absolute pressure
T = Absolute temperature
z= Compressibility factor of gas at T and P
R = Ideal gas constant (If P is in bar abs, T in Kelvin, Rho in kg/m3 then R = 8314.472)
If you just have T and P and the conditions aren't too different from the calibration conditions, assume z doesn't change much and use:
VTRUE,O = VI,O*SquareRoot(PCAL/PO*TO/TCAL)
with the O subscript being for the current operating P and T conditions and CAL the P and T conditions at which the flow meter reads correctly.
There are standards covering some kinds of meters (e.g. ISO 5167 for orifice plates and venturi tubes) and manufacturers, e.g. of rotameters-like devices, may have their own more sophisticated corrections.
200272522, the answer depends on whether the flow meter reading is express in mass flow, actual volume flow or standard volume flow.
You can easily derive the correct equation from the basic orifice equation. For a gas flow reading in standard ft3/h, the correction is:
Qstd = Qstd,DCS * Sqrt[(Mw.T/P)DCS * (P/Mw.T)actual]
where "actual" refer to current operation and "DCS" is the numbers used for sizing/calibrating the flow meter on the DCS. I assumed Z is constant.
You can easily derive the correct equation from the basic orifice equation. For a gas flow reading in standard ft3/h, the correction is:
Qstd = Qstd,DCS * Sqrt[(Mw.T/P)DCS * (P/Mw.T)actual]
where "actual" refer to current operation and "DCS" is the numbers used for sizing/calibrating the flow meter on the DCS. I assumed Z is constant.
Several people above have talked about a correction factor for orifice meters. There is no such animal. Orifice meters are what they are.
For other meters there is some sort of "K" factor or "meter factor" that is usually provided by the manufacturer to try to convert the measured parameter (e.g. angular velocity of a turbine meter) into a flow rate more accurately. I've often thought of these factors as "multiply by zero and add the right answer", but that is just my prejudice--many meters with meter factors result is very accurate (i.e., low uncertainty, high repeatability) measurement.
The technique and magnitude of any meter factors is very much dependant on both the technology and manufacturer of the device, but there is no meter factor for square-edged orifice measurement.
David
For other meters there is some sort of "K" factor or "meter factor" that is usually provided by the manufacturer to try to convert the measured parameter (e.g. angular velocity of a turbine meter) into a flow rate more accurately. I've often thought of these factors as "multiply by zero and add the right answer", but that is just my prejudice--many meters with meter factors result is very accurate (i.e., low uncertainty, high repeatability) measurement.
The technique and magnitude of any meter factors is very much dependant on both the technology and manufacturer of the device, but there is no meter factor for square-edged orifice measurement.
David
Come on Zdas04, the OP wants compensation factors. I believe we are giving a general idea and where to get the absolute answer.
API does allow a meter factor for an orifice, you can take the meter to a lab (or bring the lab to the meter) and flow calibrate it against a traceable standard. Then you can apply the calibration meter factor in real time along with the other compensation factors.
API does allow a meter factor for an orifice, you can take the meter to a lab (or bring the lab to the meter) and flow calibrate it against a traceable standard. Then you can apply the calibration meter factor in real time along with the other compensation factors.
Where? I've read the standards pretty closely for the last few iterations (hell, I was a consultant to the calibration sub-committee) and I sure can't find any place where you can legally apply any sort of "calibration factor" to a square-edged orifice meter. There used to be (mid-70s) a procedure for "adjusting" the C prime, but that adjustment has gone the way of the dodo.
David
David
For volume meters you also need to be able to determine if you are better to make a PTZ correction or if it is better (more accurate) to use density measurement.
A good source of data is to visit Emerson Mobrey's web site and request a copy of their manual for the gas flow computers. Lots of good stuff there.
JMW
A good source of data is to visit Emerson Mobrey's web site and request a copy of their manual for the gas flow computers. Lots of good stuff there.
JMW
zdas04, we are talking of two different issues.
By compensation I mean:
- The orifice was sized based on the physical properties estimated by the process engineer.
- Based on these properties, the dP cell ranges and meter max flow are determined and put into the DCS.
- Thus, the flow displayed on the DCS (or whatever) is still based on the estimated fluid physical properties.
- If the actual fluid has a different composition and at different temp+pres, then the flow displayed must be "compensated" or "corrected" to yield the real flow.
- It has nothing to do with the C factor. In fact, we typically do not adjust the C factor when we do this "compensation".
By compensation I mean:
- The orifice was sized based on the physical properties estimated by the process engineer.
- Based on these properties, the dP cell ranges and meter max flow are determined and put into the DCS.
- Thus, the flow displayed on the DCS (or whatever) is still based on the estimated fluid physical properties.
- If the actual fluid has a different composition and at different temp+pres, then the flow displayed must be "compensated" or "corrected" to yield the real flow.
- It has nothing to do with the C factor. In fact, we typically do not adjust the C factor when we do this "compensation".
So you are saying that you simply apply the proper gas analysis in the calculation? How is that an adjustment? To tell a flow computer what the maximum flow would be is unusual, but I've seen it. Typically you just tell it the fluid properties, the physical properties of the equipment (material, pipe ID, plate bore OD, etc) and the range of each of the three sensing elements (press, dp, and temp). When it gets a 20 mA signal from dP, it has reached its maximum flow rate.
David
David
zdas04, the only difference is in how the calc is done.
Method 1.
Example: Assume dPcell range = 100 inH2O for 4-20 mA.
dP of 50 inH2O equals 12 mA equals 0.5 signal equals 0.707 after square root. The DCS multiply it with the meter max of say 100 BPSD to give you 70.7 BPSD.
If your composition and conditions are different from that used to calc the 100 BPSD, then you have to adjust the flow.
Method 2.
What you are saying is that you take the 50 inH2O and stick it into the orifice calc and get the right answer directly.
Both methods will give the same answer, it is just that in my limited experience method 1 is implemented in refineries.
Method 1.
Example: Assume dPcell range = 100 inH2O for 4-20 mA.
dP of 50 inH2O equals 12 mA equals 0.5 signal equals 0.707 after square root. The DCS multiply it with the meter max of say 100 BPSD to give you 70.7 BPSD.
If your composition and conditions are different from that used to calc the 100 BPSD, then you have to adjust the flow.
Method 2.
What you are saying is that you take the 50 inH2O and stick it into the orifice calc and get the right answer directly.
Both methods will give the same answer, it is just that in my limited experience method 1 is implemented in refineries.
- Status
Similar threads
- Locked
- Question
