small hydrocarbon flow calibration facilities at nmij
TRANSCRIPT
NATIONAL METROLOGY INSTITUTE OF JAPAN
Small hydrocarbon flow calibration facilities
at NMIJ
Yoshiya Terao
APMP/TCFF Workshop
NATIONAL METROLOGY INSTITUTE OF JAPAN
Original presentations
at 9th International Symposium on Fluid Flow Measurement (ISFFM2015),
April 2015
“Expansion of Calibration Flow Range of Small Liquid
Hydrocarbon Flow Facility at NMIJ”
by Kar-Hooi Cheong, Ryouji Doihara Takashi Shimada and
Yoshiya TERAO
“Development of prototype low-flow rate test rig applicable for
highly volatile liquids” by Ryouji Doihara, Takashi Shimada, Kar-Hooi Cheong and Yoshiya Terao
NATIONAL METROLOGY INSTITUTE OF JAPAN
1.E-06
1.E-05
1.E-04
1.E-03
1.E-02
1.E-01
1.E+00
1.E+01
1.E+02
1.E+03
1.E+04
0.1 1 10 100 Viscosity (mPa·s)
Flo
wra
te (
m3/h
)
Fuel mixer
Fuel efficiency
measurement
Hydraulic machinery
LP
G
Gaso
lin
e
Kero
sen
e
Lig
ht
oil
Heavy
oil
Chemical plant
Tanker
Petroleum terminal
Petroleum refinery
Primary standards for hydrocarbon at NMIJ
Primary
standard for
hydrocarbon
Small
Hydrocarbon
Flow Facility
100 L/h
0.02 L/h
1 L/h
NATIONAL METROLOGY INSTITUTE OF JAPAN
37.0 36.4
33.0 33.0 32.6 30.8 30.4 30.2 30.0 30.0
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
Car models
JC0
8m
od
e f
ue
l eff
icie
ncy
(km
/L)
Top 10 car models with best fuel efficiency in Japan
hybrid vehicle
mini / light vehicle
Source: http://e-nenpi.com
Background and Motivation
NATIONAL METROLOGY INSTITUTE OF JAPAN
Main targets Automotive industry
Engine test bench
Vaporizer for
manufacturing equipment
Fuel flowmeter
Liquid mass flowmeter
Semiconductor industry
NATIONAL METROLOGY INSTITUTE OF JAPAN
Measurement conditions:
• Real gasoline (high volatility), Light oil,
• 200 L/h – 0.02 L/h,
• Wide flow range (over 1000:1),
• Wide temperature range,
• (High pressure over 200MPa).
• Water, Alcohols, Solvents, Toxic liquids,
• 10 L/h – 0.001 L/h,
• Wide temperature range (10 – 85 ℃),
Liquid mass flowmeter
Automotive industry
Semiconductor industry
Fuel flowmeter
NATIONAL METROLOGY INSTITUTE OF JAPAN
Measurement principles Gravimetric calibration:
• Advantage to accuracy and to simple traceability
• Disadvantage to evaporation error at the low flow rate
range with volatile liquid
• Water is mostly used as test liquid.
Volumetric calibration:
• SVPs have been adopted in many calibration facilities for
oil (including real gasoline), because it can make a closed
pipe line system.
• Some of them have a linear encoder to output many
pulses.
Syringe pump:
• Syringe pumps are widely used in medical and analysis
fields for ultra-low flow rate.
• Fluctuation has been observed as a result of the stepper
motor driving mechanism.
NATIONAL METROLOGY INSTITUTE OF JAPAN
Cheong Kar-Hooi’s
gravimetric system
NATIONAL METROLOGY INSTITUTE OF JAPAN
Calibration Facility
Weighing section
(liquid collection) Test section
(DUT mounting) Flow generation section
(tank, pump)
NATIONAL METROLOGY INSTITUTE OF JAPAN
Calibration Facility Weighing section
(liquid collection)
Test section
(DUT mounting) Flow generation section
(tank, pump)
100g weighing system
(liquid collection for
0.02~1 L/h)
2kg weighing system
(liquid collection for 1~100 L/h)
Test meterWeighing
vessel
Weighing
scale, ms
Bypass to
storage tankMeasurement point of temperature, Ttm
and pressure, ptm
Pulse counter,
Number of pulses, Ip
Pulse signal
DUTTest meterWeighing
vessel
Weighing
scale, ms
Bypass to
storage tankMeasurement point of temperature, Ttm
and pressure, ptm
Pulse counter,
Number of pulses, Ip
Pulse signal
DUT
Calibration method ・ Gravimetric (static weighing)
・ Standing-start-and-finish
thermostatic chamber
(for better stability of liquid
temperature controllable
at any value between
15℃~35℃)
NATIONAL METROLOGY INSTITUTE OF JAPAN
Uncertainty Budget
Uncertainty sources Light oil Kerosene
Pulse count 8.2×10-5 8.2×10-5
Estimation of liquid density 2.6×10-4 3.0×10-4
Measurement of liquid mass 1.8×10-4 1.8×10-4
Dead volume effect 1.2×10-4 1.3×10-4
Relative combined standard uncertainty 3.5×10-4 3.8×10-4
Uncertainty budget for 0.02 L/h ~ 1 L/h using 100 g weighing system
Uncertainty sources Light oil Kerosene
Pulse count 8.2×10-5 8.2×10-5
Estimation of liquid density 2.6×10-4 3.0×10-4
Measurement of liquid mass 3.3×10-5 3.3×10-5
Dead volume effect 6.1×10-6 6.4×10-6
Relative combined standard uncertainty 2.8×10-4 3.2×10-4
Uncertainty budget for 1 L/h ~ 100 L/h using 2 kg weighing system
NATIONAL METROLOGY INSTITUTE OF JAPAN
Uncertainty Budget (Kerosene)
2 kg weighing system
(1 L/h~100 L/h)
100 g weighing system
(0.02 L/h~1 L/h)
pulse count 0.008%
dead volume effect 0.001%
measurement of liquid mass
0.003%
estimation of liquid density
0.030%
pulse count 0.008%
dead volume effect 0.013%
measurement of liquid mass
0.018%
estimation of liquid density
0.030%
NATIONAL METROLOGY INSTITUTE OF JAPAN
pulse count 0.008%
dead volume effect 0.013%
measurement of liquid mass
0.018%
estimation of liquid density
0.030%
Uncertainty budget for 0.02 L/h ~ 1 L/h using
100 g weighing system
Dead volume effect
The value is evaluated in relative to
the smallest amount of liquid
collection which is 10 g.
10 g of liquid collection is performed
at 0.02 L/h and it takes about 40
minutes for one collection.
Some measures to cut down the dead
volume effect:
Dead volume is made as small as
possible.
Temperature variation of liquid in the
dead volume is controlled and
stabilized (±0.025oC in one hour
duration).
NATIONAL METROLOGY INSTITUTE OF JAPAN
pulse count 0.008%
dead volume effect 0.013%
measurement of liquid mass
0.018%
estimation of liquid density
0.030%
Contributing factors to uncertainty of
liquid mass measurement
correction factor of weighing scale
0.003% reading of weighing scale 0.001%
buoyancy effect correction 0.001%
liquid evaporation
0.015%
leakage possibility
0.009%
NATIONAL METROLOGY INSTITUTE OF JAPAN
Conditions for liquid collection
Liquid temperature:15℃~35℃.
Smallest liquid collection: 10 g (about 40 minutes at 0.02 L/h )
Method of evaluation
An evaluation method thought to give the closest value to the actual
condition was performed.
Results
At 35℃, the evaporation rate is 0.0015 g per hour for kerosene and
0.0007 g per hour for light oil.
In relative to 10 g of liquid collection, this produces 1.5×10-4 (0.015
%) in terms of relative uncertainty.
Collection 100 g discharge Measurement of evaporation
for 60 minutes
Evaluation of liquid evaporation
NATIONAL METROLOGY INSTITUTE OF JAPAN
Flow range, Q / Weighing system Volumetric flowrate Mass flowrate
1 L/h Q 100 L/h (2 kg weighing system) 0.032 %
(0.064 %)
0.010 %
(0.020 %)
0.02 L/h Q < 1 L/h (100 g weighing system) 0.039 %
(0.078 %)
0.025 %
(0.050 %)
Calibration and Measurement Capability
Ultimate uncertainty of calibration
(Calibration and measurement capability in italic, coverage factor: k=2)
Note:
(ultimate standard uncertainty)2 = (uncertainty due to facility)2 + (uncertainty due to flowmeter)2
Uncertainties due to flowmeter are as follows:
Response of flowmeter to standing-start and finish
Repeatability
NATIONAL METROLOGY INSTITUTE OF JAPAN
Doihara’s weighing and syringe
system
NATIONAL METROLOGY INSTITUTE OF JAPAN
Schematic diagram of the prototype test rig
NATIONAL METROLOGY INSTITUTE OF JAPAN
Schematic diagram of the prototype test rig
NATIONAL METROLOGY INSTITUTE OF JAPAN
Calibration procedure
• The syringe pump is calibrated using
the gravimetric weighing tank.
Calibration
NATIONAL METROLOGY INSTITUTE OF JAPAN
Calibration procedure • Then a DUT flowmeter is calibrated
using the syringe pump.
• At a DUT calibration stage, the flow rig is able to make a closed pipe line system.
Calibration
NATIONAL METROLOGY INSTITUTE OF JAPAN
Standing method and flying method
• The syringe pump is calibrated with a
standing start and stop method.
• What the weighing tank needs to
calibrate is only the syringe pump
(over 10 000 pulses).
• Characteristic of the syringe pump is
essentially linear to flow rate.
• Thus the standing method is not
significant problem in this syringe pump
calibration.
Standing method
Calibration
NATIONAL METROLOGY INSTITUTE OF JAPAN
• DUT flowmeter is calibrated at a constant flow rate with a flying start and finish method.
• If the number of pulses of the flowmeter would be small, the pulse counting error can be reduced using a double chronometry pulse interpolation technique in the flying method.
Flying method
Calibration
NATIONAL METROLOGY INSTITUTE OF JAPAN
Calibration results of the syringe
-0.01
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.01 0.10 1.00 10.00 100.00Flow rate [ L/h ]
Light oil: Jan. 2015
SYSY
SY
PK
V [Pulse/mL]
• They all agree within ±0.02%.
• Influence of the standing method was enough small.
Rela
tive d
evia
tion f
rom
nom
inal valu
e [
% ]
±0.02%.
NATIONAL METROLOGY INSTITUTE OF JAPAN
Evaporation effect and correction
Condition Tank lid with
small holes
Evaporation
trap pool
Evaporation rate
[ mg/h ]
Static A ✓ ✓ 0.6
Static B ✓ 3
Static C ✓ 4
Static D 6
• The weighing tank system has some evaporation controls to reduce evaporation. – The weighing tank was surrounded by
small liquid pools.
– The weighing tank has a lid with small holes to insert tubes.
• The changing of weight from an initial weight was measured in several conditions for one hour (Static).
NATIONAL METROLOGY INSTITUTE OF JAPAN
• The syringe pulse factor was increased at low flow rate.
• Results with evaporation correction (Static A) were similar.
-0.01
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.01 0.10 1.00 10.00 100.00
Flow rate [ L/h ]
Without evaporation correction
With evaporation correction (Static A)
Rela
tive d
evia
tion f
rom
nom
inal valu
e [
% ]
NATIONAL METROLOGY INSTITUTE OF JAPAN
Evaporation effect and correction
Condition Tank lid with
small holes
Evaporation
trap pool
Evaporation rate
[ mg/h ]
Static A ✓ ✓ 0.6
Static B ✓ 3
Static C ✓ 4
Static D 6
Actual E ✓ ✓ 2 – 4
• The atmosphere inside the wind shield was ventilated when the working liquid in the weighing tank was discharged.
• The actual evaporation rate was measured in the calibration sequence.
• The measured evaporation rate (actual condition E) was between static A and static C.
Since actual evaporation rates were not constant, the
evaporation effect was corrected at each calibration.
NATIONAL METROLOGY INSTITUTE OF JAPAN
• Corrected results (Actual E) showed almost constant value over full
flow rate range.
-0.01
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.01 0.10 1.00 10.00 100.00
Flow rate [ L/h ]
Without evaporation correction
With evaporation correction (Actual E)
Rela
tive d
evia
tion f
rom
nom
inal valu
e [
% ]
NATIONAL METROLOGY INSTITUTE OF JAPAN
5mL
Position dependency of syringe pulse factor The standing stop position was varied from the cylinder end position
to 35 mL, keeping the calibration collection volume of 5 mL.
-0.01
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0 5 10 15 20 25 30 35 40
Standing stop position from cylinder end [ mL ]
5mL
0
Rela
tive d
evia
tion f
rom
nom
inal valu
e [
% ]
NATIONAL METROLOGY INSTITUTE OF JAPAN
10mL
Position dependency of syringe pulse factor
0 5mL
-0.01
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0 5 10 15 20 25 30 35 40
Standing stop position from cylinder end [ mL ]
5mL collection (light oil)
10mL collection (lightl oil)
Volume domain of the cylinder has been fixed.
•Standing stop position: 5m L,
•Collection volume: 10 mL
Rela
tive d
evia
tion f
rom
nom
inal valu
e [
% ]
NATIONAL METROLOGY INSTITUTE OF JAPAN
Calibration results of the syringe
-0.01
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.01 0.10 1.00 10.00 100.00Flow rate [ L/h ]
Light oil: Jan. 2015
Industrial gasoline: Feb. 2015
Industrial gasoline: Mar. 2015
Rela
tive d
evia
tion f
rom
nom
inal valu
e [
% ]
• Difference depending on kinds of liquids has not been observed.
NATIONAL METROLOGY INSTITUTE OF JAPAN
Preliminary uncertainty budget
Uncertainty source Standard uncertainty
Combined
standard
uncertainty
Expanded
uncertainty
(k=2)
Syringe pulse count 0.002%
Mass measurement 0.008%
Density at syringe 0.010%
Change of mass in dead volume 0.007%
Random effect of syringe calibration 0.008%
Calibration for syringe pulse factor 0.017% 0.034%
Linearity and long term reproducibility
of reference syringe pulse factor 0.017%
Density at syringe 0.010%
Change of mass in dead volume 0.015%
Random effect of DUT calibration 0.005%
Calibration for mass flow rate 0.030% 0.060%
Density at DUT 0.016%
Calibration for volume flow rate 0.034% 0.068%
NATIONAL METROLOGY INSTITUTE OF JAPAN
Specification • Flow rate:
60 L/h – 0.02 L/h (1000 mL/min – 0.3 mL/min).
• Working liquid (as the first step):
Light oil (diesel oil),
Industrial gasoline (Flash point ≥ 40 ℃ )
• Calibration principle:
Gravimetric weighing tank,
Syringe pump.
• Temperature and pressure (as the first step):
20 ℃ (Stability ±0.08 ℃)
0 – 0.3 MPa (at test section)
• Preliminary expanded uncertainty (k=2):
0.068 % (for volumetric flow),
0.060 % (for mass flow) .
NATIONAL METROLOGY INSTITUTE OF JAPAN
Summary
Cheong’s gravimetric system
• 0.02 L/h to 100 L/h
• Kerosene and light oil
• Liquid temperature: 15℃ to 35℃
• Open for public calibration service
• Will be peer-reviewed in 2017
Doihara’s gravimetric and syringe system
• 0.02 L/h to 60 L/h
• Applicable to volatile liquids
• Liquid temperature: 20℃
• Flexible structure
• Used for research purpose
NATIONAL METROLOGY INSTITUTE OF JAPAN
Thank you for your attention.