DUNAMET D36

Report
Project DUNAMET
D36
Tomáš Valenta
Czech Metrology Institute
no.
Project DUNAMET
no. D36
• the inter-comparison among
Czech Republic (CMI)
Austria (BEV)
Hungary (OMH)
and Slovak Republic (SMU)
• took place from May 2004 to August 2004
• the inter-comparison of low pressure test facilities for
gas meters in the range from 100 m3/h to
1000
m3/h
Project DUNAMET
no. D36
The axial turbine gas meter G650 was used for
the comparison.
Manufacturer:
PREMAGAS
(ELSTER –AMCO
Group)
Type: DAE
Size: G650
DN150
Project DUNAMET
no. D36
Calibration procedure:
The turbine gas meter G650 was tested in each
laboratory by air in horizontal position.
It was necessary to use the upstream straightening pipe
that was long at least 5x DN and the downstream
straightening pipe that was long at least 3x DN.
The reference pressure from the turbine gas meter was
measured from the output “pr”.
Project DUNAMET
no. D36
The reference temperature from the turbine gas meter
was measured in the distance (23)x DN downstream
of the turbine gas meter.
The HF pulse emitter A1S (the pulse number 4579,88
pulses/m3) was used for the tests.
The calibration had to be performed in the laboratory
where the temperature was from 19,5°C to 23,5°C. No
oil lubrication was used.
Project DUNAMET
no. D36
The turbine gas meter was tested in 8 flow rates:
1000 m3/h, 800 m3/h, 650 m3/h, 500 m3/h, 400 m3/h, 300 m3/h,
200 m3/h, 100 m3/h.
Error of the meter was determined. The Error is value which
shows the relationship in percentage terms of the difference
between the volume indicated by the meter and the volume
which has actually flowed through the meter, to the later value.
Vi  Vc
E
.100
Vc
where
(%)
E is the error of the meter
Vi is the indicated volume by the meter (m3)
Vc is the real volume which has actually flowed through
the meter (m3)
Project DUNAMET
no. D36
Determination of the reference value
To determine the reference value tree different approaches were used.
Mean: Mean is the arithmetic average of a set of numerical data.
Median: Median is the middle value of an ordered set of numerical data.
Weighted mean error (WME):
If in a series of observations a statistical weight (1/u2) is assigned to each
value, a weighted mean y can be calculated by the formula:
x1 u 2 x1   ....  x N u 2 x N 
y
1 u 2 x1   ....  1 u 2 x N 
where
xi are values of errors of meter in different laboratories in one flow rate
u (xi) are uncertainties U(k=2) of errors of meter in different laboratories in
one flow rate
Project DUNAMET
no. D36
Results:
Flow rate
in the meter
Mean
Median
Weighted mean
error (WME)
m³/h
%
%
%
1000
-0,17
-0,14
-0,18
800
-0,03
0,01
-0,03
650
0,01
0,05
0,02
500
0,02
0,03
0,03
400
-0,03
-0,04
-0,03
300
-0,03
-0,05
-0,03
200
0,15
0,12
0,14
100
0,93
0,91
0,93
Project DUNAMET
no. D36
Austria (BEV)
DUNAMET No. D36, turbine gas meter G650
1,20
1,00
BEV
0,80
mean
error (%)
0,60
WME
0,40
median
0,20
0,00
-0,20 0
100
200
300
400
500
600
-0,40
-0,60
-0,80
-1,00
flow rate (m 3/h)
700
800
900
1000
1100
Project DUNAMET
no. D36
Hungary (OMH)
DUNAMET No. D36, turbine gas meter G650
1,20
1,00
OMH
0,80
mean
0,60
WME
error (%)
0,40
median
0,20
0,00
-0,20
0
100
200
300
400
500
600
-0,40
-0,60
-0,80
-1,00
flow rate (m 3/h)
700
800
900
1000
1100
Project DUNAMET
no. D36
Slovak Republic (SMU)
DUNAMET No. D36, turbine gas meter G650
1,20
1,00
SMU
0,80
mean
0,60
WME
error (%)
0,40
median
0,20
0,00
-0,20
0
100
200
300
400
500
600
-0,40
-0,60
-0,80
-1,00
3
flow rate (m /h)
700
800
900
1000
1100
Project DUNAMET
no. D36
Czech Republic (CMI)
DUNAMET No. D36, turbine gas meter G650
1,20
1,00
CMI
0,80
mean
error (%)
0,60
WME
0,40
median
0,20
0,00
-0,20
0
100
200
300
400
500
600
-0,40
-0,60
-0,80
-1,00
flow rate (m 3/h)
700
800
900
1000
1100
Project DUNAMET
no. D36
Conclusion:
All participants were successful. All determined
errors of meter with their uncertainties U(k=2)
in every flow rate cover the reference values
Mean, Median and Weighted Mean Error.
New EUROMET project
THIS PROJECT CAN BE REPEATED AMONG
EUROMET COUNTRIES IN 2006.
Note: I would like to coordinate quick and simple intercomparison. That is why the supposed time to make
calibration and to deliver the meter to next laboratory
will be only 1 week.
The form „PROPOSED EUROMET PROJECT“
will be sent to everybody by e-mail.
Please don´t applause
because Mr. Lojek is
sleeping.
Thank you for attention.

similar documents