Reducing re-calibrations
is justified by a density analysis
on LNG for Micro Motion
Coriolis meters
by
Aart Pruysen
Approval Director Europe
Emerson in the LNG market chain
Results of master metering at Skangas –
Risavika Havn Stavanger – Norway
over one year
Result of common cause analysis
GIIGNL acceptance of density monitoring
to allow Coriolis flow meters
LNG supply chains supported by Emerson
LNG supply chains supported by Emerson;
many include custody transfer applications
Dispensers
Tank truck
loading/unloading
Application:
LNG bunkering at Risavika Havn, Stavanger, Norway
Application: LNG bunkering of vessels
Vacuo insulated pipes
Coriolis meters also
vacuo insulated
Subsequent verification by master metering
Minimum costs and maximum availability
A 4” Coriolis master meter (CMF400) to verify per year each
of the two installed 4” stream meters (CMF400) ;
Master meter calibrated at NIST on LN2 ; to be re-calibrated on water per 3 years ;
Master meter was manufactured one year later than the two stream meters and comes
from another “heat slot” ;
Only stream 1 operational in first year (stream 2 and master meter stand-by)
2 stream
meters
Master
meter
Proving results after one year ; excellent results
Initial verification April 2015:
Flowrate:
150 t/h
70 t/h
Stream 1 : + 0.016 %
+ 0.084
Stream 2: - 0.016 %
+ 0.043
Subsequent verification April 2016:
Flowrate:
150 t/h (only one flowrate)
Stream 1 : - 0.018%
Stream 2 : - 0.025 %
API repeatability < 0.020%; type A (2k) < 0.005%
Zero stability spec per meter: 0.027% at 150 t/h
0.057% at 70 t/h
Stability over first year at 150 t/h
Stream 1 : - 0.034 %
Stream 2 : - 0.009 %
Not affected by common causes??
In-depth analysis for common causes in mass
Mass and density are both affected by stiffness
determine the “reference”
density and deviation in density to verify common causes
“Reference” density with an uncertainty of 2 kg/m3 (GIIGNL, report 2015)
- temperature and pressure measurements near the meter
- composition from Gas Chromatograph (GC)
- REFPROP/GERG equations
In-depth analysis for stability in mass (background)
Mass flow equation simplified (slope and intercept):
Qm = MF . FCF. (∆Tmeas - ∆Tstoredzero ) . TFmass
Temperature Factor mass Tfmass
-
Stability determined by change in the “slope” part:
- Stiffness at 0°C is part of FCF
- Stiffness changes with temperature
In-depth analysis for stability in density (background)
Simplified density equation:
ρfluid = C1 * TFdensity * TP 2 − C 2
Where:
ρfluid
= measured density
C1
C2
TFdensity
Tp
= sensor constant
= sensor constant
= Temperature factor density
= measured tube period
Stability determined by change in first term
- Stiffness at 0°C is part of C1
- Stiffness changes with temperature
C2 is sensor constant (mass tubes/volume of tubes) and
not likely to change with LNG
In-depth analysis for common causes in mass
Based on “reference” density and Coriolis density, the Young’s modulus
can be calculated from Coriolis meter.
Compare with published NIST data (1980)
Young’s modulus difference between Coriolis density and NIST
Results do not show stability changes for all three meters;
no common causes.
Difference between stream and master coming from uncertainty density
In-depth analysis for stability in mass
Equations to calculate the Young’s modulus difference is not easy for user
easier to perform via density monitoring.
Density outputs virtually adjusted by Meterfactor density as it was April 2015:
Uncertainty sampling
determines stability
All three meters do not shift more than claimed uncertainty of reference
density ; all three meters stable and no common cause
Stream 1 (most used meter) is equally stable as the other two.
What do these values mean in terms of stability for mass?
In-depth analysis for stability in mass
Sensitivity of C1*TFdensity*Tp2 :
ρ fluid = C1 * TFdensity * TP 2 − C 2
For CMF400:
C2= 2250 kg/m3 C1*TFdensity*Tp2 : = 2700 kg/m3 with LNG
high sensitivity:
1 kg/m3 density deviation change corresponds to 0.037 % in stiffness change
Uncertainty in reference density per sample: 2 kg/m3
Estimated uncertainty in difference of two samples also: 2 kg/m3
Density reading has the potential to determine the stability of the slope part of
the mass equation within 0.08%
Zero verification still needed
In-depth analysis for stability in mass
Relation mass stability vs density stability
(not for extreme faults, such as gasket in one of tubes):
∆m% ≤
∆Density
*100%
C2 + ρliquid
∆m% ≤
− 190
*100% = − 5.85%
3250
See mass and density results and their relation for a faulty meter on water:
In-depth analysis for stability in mass
Stability density:
Stability mass proving vs density monitoring:
- no shift detected based on density monitoring;
- mass proving and density monitoring within uncertainty limits,
coming from zero offset limit and uncertainty reference density
In-depth analysis for stability in mass
Conclusion:
A
Monitoring shift of density deviation over time gives “slope part of
stability” in mass
B
Subsequent density monitoring gives actual meter factor density in
relation to initial meter factor density and easy for end user
C
Density and also volume can be adjusted for better performance
D
Monitoring the zero offset gives the “zero part of stability” in mass
Proposal set-up for density monitoring (if justifiable);
BOG (vapor return line)
BOG direct mass measurement
Flow computer
DCS /
host
Bill of Lading Printer
Gas chromatograph
Temp
X-er
Press
X-er
Flow
X-er
Operator
Interface
Transfer Point
LNG sample probe
LNG liquid direct mass measurement
Proving connection
For optimum performance:
- Temperature correction of Coriolis meter for mass and density
based on external temperature measurement
- Pressure correction of Coriolis for mass and density
Stability in mass from density stability
In addition,
Report of GIIGNL, the International Group of Liquefied Natural Gas Importers
(version 2015):
GIIGNL recommends to use flow meters as soon as field proving is solved
density monitoring is a tool to alert the user to initiate mass proving, only
when needed
GIIGNL should accept this density monitoring to allow Coriolis meters
Consider current situation, based on tank gauging:
Uncertainty volume questionable while:
* tanks were calibrated once many years ago
* automatic level gauging hard to verify in situ
Density proving with Coriolis meter is of higher verification level than
current tank gauging
Stability in mass from density stability
In addition,
Report of GIIGNL, the International Group of Liquefied Natural Gas Importers
(version 2015):
- only tank gauging considered as the quantity measurement
- uncertainty in energy between 0.50 and 0.74% (k=2), coming from:
* uncertainty in volume between 0.40 and 0.54%
* uncertainty in density 0.46%
Acc GIIGNL:
Uncertainty in mass: 0.61% - 0.71%
* uncertainty in caloric value 0.08%
Opportunity for improvement
Coriolis meters give better uncertainty;
Time to change to Coriolis meters
Take away
- Introduce density monitoring with Coriolis meters to alert proving when a
certain density shift occurs for big meters (justifiable)
- Implement meter factor density/meter factor volume
- Zero verification needed
- Coriolis meters gives better uncertainty than tank gauging (0.61-0.71%)
- GIIGNL should accept density monitoring method
- Already known:
Traceability to mass is highly needed for the LNG industry
Emerson chosen as supplier for 6 master meters
Thank You
Let’s connect
Giuseppe Bernardelli
LNG business development manager
[email protected]
+39 348 0195 604
Aart Pruysen
Approval Director Europe
[email protected]
+31 653 294 288