XX IMEKO World Congress

Metrology for Green Growth

September 914, 2012, Busan, Republic of Korea

UNCERTAINTY EVALUATION FOR THE COMPOSITE ERROR
OF ENERGY METER AND INSTRUMENT TRANSFORMER

H. Kato, H. Imai

Japan Electric Meters Inspection Corporation,
4-15-7 Minato-ku, Tokyo, Japan,
E-mail : [email protected]

Abstract: Japan Electric Meters Inspection Corporation
(JEMIC) is a Designated Institute of power and energy
standards in Japan. All the energy meters for electricity
trading in Japan are traceable to the National Standard in
JEMIC. In industrial world, power is supplied by various
voltage and current. However, the energy standard has been
calibrated at 110 V and 100 V, for 5 A. Therefore, it is often
used in combine with the instrument transformer. This paper
describes the effect of composite error of instrument
transformers to uncertainty evaluation of energy meters.
Keywords: AC Power, AC Energy, Energy Meter,
Instrument Transformer
1. INTRODUCTION
Instrument transformer errors are evaluated by two
components (Ratio error and Phase displacement). In
addition, an energy meter is evaluated by instrumental error.
In this paper, the composite error of the instrument
transformer (instrument voltage transformer: VT, current
transformer: CT) is calculated from JIS C 1736-1(2009)
Annex-A, in combination with the instrumental error of
energy meter. Therefore, we examined the amount of
contribution of the uncertainty and error of instrument
transformer for energy meters.
2. CALIBRATION METHOD
The calibration of energy meter is carried out by using
the following equipment.
1) Standard energy meter
2) Power Source
3) Frequency counter
4) Voltage transformer
(This equipment is used only when the voltage is different
from the standard energy meter.)
5) Current transformer
(This equipment is used only when the current is different
from the standard energy meter.)

Figure 1 is a calibration circuit and Figure 2 is a view of
calibration system. The power is supplied on standard and
device under test through an instrument transformer, then
the output pulse can be measured for calibration using the
frequency counter.


Figure 1 The block diagram of calibration circuit


Figure 2 The view of calibration system

3. ABOUT ERROR CALCULATION EXPRESSION
(FUNCTION MODEL)
Described following is an example of error calculation
formula of a single-phase two-wire.


1) Energy meter error calculation formula:
( )
tr s
ws
f s
ws
f s
wx
f x
tr s s f x x
k
p p
k
p p
k
p p
p p p f


+ +
+
+

+
=
=
100
, , , ,
(1)
where,

x
: Energy meter error of device under test (%)
P
x
: Average of the output pulse of device under test
k
wx
: Instrument constant of device under test (pulse/kWs)
P
s
: Average of the output pulse of standard
k
ws
: Instrument constant of standard (pulse/kWs)
p
f
: Correction value of frequency counter

s
: Energy meter error of standard (%)

tr
: Composite error of instrument transformer (%)

2) Composite error calculation formula:
( )
( )

tan 0291 . 0
100 1 cos
100
1
100
1
cos
1
v c c v
c v
c v
tr
+ +

(
(

+
|
|

\
|
+
|
|

\
|
+ =
(2)
where,

tr
: Composite error of instrument transformer (%)

v
: Ratio error of VT (%)

c
: Ratio error of CT (%)

v
: Phase displacement of VT (min.)

c
: Phase displacement of CT (min.)
: Load power factor angle (deg.)
4. SOURCES OF UNCERTAINTY
The sources of uncertainty for the calibration of energy
meter with instrument transformer are described as follows:
1) Uncertainty contributes to the mean value of the output
pulse count of device under test.
A. Repeated measurement : u(p
x1
)
B. Resolution of digital display : u(p
x2
)
C. Difference of input voltage : u(p
x3
)
2) Uncertainty contributes to the correction value of the
frequency counter.
A. Uncertainty of calibration using an upper standard : u(p
f1
)
B. Long-term stability : u(p
f2
)
3) Uncertainty contributes to the mean value of the output
pulse count of standard.
A. Repeated measurement : u(p
s1
)
B. Resolution of digital display : u(p
s2
)
4) Uncertainty contributes to the Energy meter error of standard.
A. Uncertainty of calibration using standard : u(
s1
)
B. Stability : u(
s2
)
C. Linearity : u(
s3
)
D. Temperature dependence : u(
s4
)
E. Due to correlation : u(
s5
)
(Except for single-phase two-wire)
5) Uncertainty contributes to the composite error of
instrument transformer
A. Uncertainty of resultant error : u(
tr1
)
B. Long-term stability (VT) : u(
tr2
)
C. Long-term stability (CT) : u(
tr3
)
5. CALCULATION OF COMBINED STANDARD
UNCERTAINTY
Combined variance [ u
c
2
(
x
)] of function model [
x
]is
estimated by the following equation.
( ) ( ) ( )
( ) ( ) ( )
tr tr s s s ps
f pf x px x
u c u c p u c
p u c p u c u

2 2 2 2 2 2
2 2 2 2 2
c
+ + +
+ =



Sensitivity coefficient (partial derivatives : c
i
) is as follows.
s wx x
px
p k p
f
c

=
100

s wx f
pf
p k p
f
c

100

2
) ( 100
s ws
f x
s
Whm
p k
p p
p
f
c

+
=

=

1
s
s
=

f
c

1
tr
t
=

f
c
r


Therefore, the combined variance [ u
c
2
(
x
)] will be the
following equation.
( ) ( ) ( )
( )
( ) ( ) ( )
tr s s
s ws
f x
s
s wx
x
s wx
x c
u u p u
p k
p p
p u
p k
p u
p k
u

2 2 2
2
2
2
2
2
2
2
100
100 100
+ +
(
(

+
+

|
|

\
|

+
|
|

\
|

=


In addition, the combined standard uncertainty is composed
of the following equation.
) ( ) ( ) (
) ( 100
) (
100
) (
00 1
) (
tr
2
s
2
s
2
2
2
2
2
2
2

u u p u
p k
p p
p u
p k
p u
p k
u
s ws
s x
f
s wx
x
s wx
x c
+ +
(
(

+
+

|
|

\
|

+
|
|

\
|

=

(3)
The effective degree of freedom [
eff
(
x
)] is determined by
the following equation.
) (
) (
) (
) (
) (
) (
eff
4 4
eff
4 4
4
eff
s
s ps
x
x px
x c
x
p
p u c
p
p u c
u

+
=

(4)
6. DETERMINATION OF COMPOSITE ERROR DUE
TO INSTRUMENT TRANSFORMER
Using the formula for calculating the composite error
function model, we evaluated the uncertainty due to
composite error. In the calculation, we used the expansion
expression.
( )
( )

tan 0291 . 0
, , , ,
v c c v
v c c v tr
f
+ + =
=


Combined variance [ u
c
2
(
tr
)] of function model [
tr
] is
estimated by the following equation.
( ) ( ) ( )
( ) ( ) ( )



2 2 2 2 2 2
2 2 2 2
tr
2
c
u c u c u c
u c u c u
v v c c
c c v v
+ + +
+ =


Sensitivity coefficients (partial derivatives : c
i
) are as follows.
1 =

=
v
v
f
c

1 =

=
c
c
f
c

tan 0291 . 0 =

=
c
c
f
c

tan 0291 . 0
v
v
=

=
f
c

( )

2
cos
0291 . 0
v c
f
c

=

=


Therefore, the combined variance [ u
c
2
(
tr
)] will be
described by the following equation.
( ) ( ) ( ) ( ) ( )
( ) ( ) | |
( )
( )

2
2
2
2 2
2 2 2 2 2
cos
0291 . 0
tan 0291 . 0
tan 0291 . 0
u u
u u u u
v c
v
c c v tr c
(
(

+ +
+ + =

In addition, the combined standard uncertainty [ u
c

(
tr
)] is
composed of the following equation.
( )
( ) ( ) ( ) ( )
( ) ( ) | |
( )
( )

2
2
2
2 2
2 2 2 2
cos
0291 . 0
tan 0291 . 0
tan 0291 . 0
u u
u u u
u
v c
v
c c v
tr c
(
(

+ +
+ +
=
(5)

We examined the influence of uncertainty of composite
error by three types of transformers. It should be noted that
the uncertainty of the following is a coverage factor k = 2.
1) Uncertainty of calibration using a high accuracy transformer .
A. Inductive voltage divider (IVD).
Uncertainty of ratio error : 0.2 (ppm)
Uncertainty of phase displacement : 3.7 (ppm)
B. Current transformer.
Uncertainty of ratio error : 20 (ppm)
Uncertainty of phase displacement : 29 (ppm)
2) Uncertainty of calibration using a general transformer.
A. Instrument voltage transformer.
Uncertainty of ratio error : 50 (ppm)
Uncertainty of phase displacement :0.3 (min.) 87 (ppm)
B. Current transformer.
Uncertainty of ratio error : 60 (ppm)
Uncertainty of phase displacement :0.2 (min.) 58 (ppm)
3) Uncertainty of calibration using a high voltage and high
current transformer.
A. Instrument voltage transformer.
Uncertainty of ratio error : 200 (ppm)
Uncertainty of phase displacement:0.6 (min.) 175 (ppm)
B. Current transformer.
Uncertainty of ratio error : 70 (ppm)
Uncertainty of phase displacement:0.3 (min.) 87 (ppm)

Uncertainty of composite error due to instrument
transformer is determined using composite error calculation
formula by the following expanded formula (5). The result
is as follows Table 2.

Table 1 Uncertainty of transformers (ppm)
Classification (1) (2) (3)
IVD or VT

v
0.2 50 200

v
3.7 87 175
CT

c
20 60 70

c
29 58 87

Table 2 Uncertainty based on composite error (ppm)
Composite error
Power factor (1) (2) (3)
PF 1 20.00 78.10 211.90
PF 0.5 Lagging 20.29 78.36 212.40
Note: PF 0.5 Lagging is 60 angle of the load power factor.

From this result, the effect of uncertainty due to the load
power factor angle is changed by the combination of the
transformer, it can be seen that the amount of that affect is
very different.
In this calculation, the number we used were converted to
ppm. Therefore, it can also be seen by the exerted influence
of significant digits.
7. UNCERTAINTY BUDGET
Described in Table 3 and Table 4 are examples of
uncertainty budget. Energy meter setting is single-phase
two-wire, 100 V, 5 A, and combination of the general
transformer.
Note: Usually, when the calibration is made by 100 V and 5
A, an instrument transformer is not used, but the calculation
is done by assuming it.
8. CONCLUSION
The uncertainty due to the composite error of instrument
transformer contributes to the energy meter greatly because
of the uncertainty of itself. It is necessary to be careful that
the difference between ratio errors and phase displacement
lead to change of uncertainty by the change of power factor
angle. In addition, it cannot ignore the impact on the
calculation results of significant digits.

9. REFERENCES
[1] JIS C 1736-1 : 2009 : Instrument transformers for metering
service- Part 1: General measuring instrument
[2] H. Kato, M. Miyakoda : Re-establishment of Power and
Energy Standards Calibration System , JEMIC Technical
Report, Vol.45, No.4 pp.71-77 (2010)
[3] H. Kato, H. Miyamura and T. Ogawa : Re-establishment of
Energy Verification Standards Calibration System , JEMIC
Technical Report, Vol.46, No.3 pp.38-49 (2011)

[4] M. Kogane, N. Yamazaki, S. Kusui : A Very Dynamic
Range Standard Watthour Meter Using a High Speed
Circulating Mark Space Type Multiplier, JEMIC Technical
Report, Vol.15, No.1 pp.7-23 (1980)
[5] A. Hashimoto : Three Phase Standard Watthour Meter
using Modified Precision Self-Calibration Multiplier,
JEMIC Technical Report, Vol.34, No.2 pp.1-9 (1999)
[6] M. Oku, K. Takahashi, R. Yasuda, H. Kato : Calibration
System of Standard Watthour Meters Using VXIbus
Instrument Unit, JEMIC Technical Report, Vol.30, No.3
pp.77-81 (1995)
Table 3 An example of uncertainty budget for PF 1
Single-Phase two-wire (100 V 5 A PF 1)
Sources of uncertainty Note
u (x
i
) u (x
i
)
u (Px )
Uncertainty contributes to the average value
of the output pulse count of device under test
1.04E-02 Hz 0.02 2.08E-04 % 1.06E+01
u (Px
1
) Repeated measurement 1.00E-02 Hz Normal 9 Calibration data
u (Px
2
) Resolution of digital display 2.89E-03 Hz Rectangular 0.01 Hz
u (Px
3
) Difference of input voltage - - Negligible
u ( P
f
) Uncertainty contributes to the correction value of the frequency counter 5.77E-03 Hz 0.02 1.15E-04 %
u ( P
f 1
) Uncertainty of calibration using an upper standard 5.00E-06 Hz Normal Calibration certificate
u ( P
f 2
) Long-term stability 5.77E-03 Hz Rectangular Record data
u (P
s
)
Uncertainty contributes to the average value
of the output pulse count of standard
1.04E-02 Hz -0.02 2.08E-04 % 1.06E+01
u (P
s 1) Repeated measurement 1.00E-02 Hz Normal 9 Calibration data
u (P
s 2) Resolution of digital display 2.89E-03 Hz Rectangular 0.01 Hz
u (
s
) Uncertainty contributes to the Energy meter error of standard 2.77E-03 % 1 2.77E-03 %
u (
s 1) Uncertainty of calibration using standard 2.50E-03 % Normal Calibration certificate
u (
s 2
) Long-term stability 1.15E-03 % Rectangular Record data
u (
s 3
) Linearity 1.15E-04 % Rectangular Experimental data
u (
s 4
) Temperature dependence 2.89E-04 % Rectangular Experimental data
u (
s 5) Due to correlation - Not affected
u (
tr
) Uncertainty contributes to the resultant error of instrument transformer 3.90E-03 % 1 3.90E-03 %
u(
tr1
) Uncertainty of resultant error 3.90E-03 % Normal Calibration certificate
u(
tr2
) Long-term stability (VT) 5.77E-05 % Rectangular Record data
u(
tr3
) Long-term stability (CT) 5.77E-05 % Rectangular Record data
Combined variance u
c
2
(
x
) = Su
i
2
(
x
) = 2.30E-05 %
2
Combined standard uncertainty u
c
(
x
) = 4.79E-03 %
Effective degree of freedom
eff
(
x
) = 1484357.0
Coverage factor k = 2
Expanded uncertainty U = k u
c
(
x
) = 0.0096 %
Calibration uncertainty U

= 0.010 %
Standard
uncertainty
Value of standard
uncertainty c
i
f /x
i
u
i
(
x
)
c
i

u (x
i
)
Degree of
freedom

i
Type of
distribution

Table 4 An example of uncertainty budget for PF 0.5 Lagging
Single-Phase two-wire (100 V 5 A PF 0.5 Lagging)
Sources of uncertainty Note
u (x
i
) u (x
i
)
u (Px )
Uncertainty contributes to the average value
of the output pulse count of device under test
1.04E-02 Hz 0.02 2.08E-04 % 1.06E+01
u (Px
1
) Repeated measurement 1.00E-02 Hz Normal 9 Calibration data
u (Px
2) Resolution of digital display 2.89E-03 Hz Rectangular 0.01 Hz
u (Px
3) Difference of input voltage - - Negligible
u ( P
f
) Uncertainty contributes to the correction value of the frequency counter 3.54E-06 Hz 0.02 7.07E-08 %
u ( P
f 1) Uncertainty of calibration using an upper standard 2.50E-06 Hz Normal Calibration certificate
u ( P
f 2) Long-term stability 2.50E-06 Hz Rectangular Record data
u (P
s
)
Uncertainty contributes to the average value
of the output pulse count of standard
1.04E-02 Hz -0.02 2.08E-04 % 1.06E+01
u (P
s 1
) Repeated measurement 1.00E-02 Hz Normal 9 Calibration data
u (P
s 2
) Resolution of digital display 2.89E-03 Hz Rectangular 0.01 Hz
u (
s
) Uncertainty contributes to the Energy meter error of standard 3.00E-03 % 1 3.00E-03 %
u (
s 1
) Uncertainty of calibration using standard 2.70E-03 % Normal Calibration certificate
u (
s 2) Long-term stability 1.15E-03 % Rectangular Record data
u (
s 3) Linearity 2.31E-04 % Rectangular Experimental data
u (
s 4) Temperature dependence 5.77E-04 % Rectangular Experimental data
u (
s 5) Due to correlation - Not affected
u (
tr
) Uncertainty contributes to the resultant error of instrument transformer 3.90E-03 % 1 3.90E-03 %
u(
tr1
) Uncertainty of resultant error 3.90E-03 % Normal Calibration certificate
u(
tr2
) Long-term stability (VT) 5.77E-05 % Rectangular Record data
u(
tr3
) Long-term stability (CT) 5.77E-05 % Rectangular Record data
Combined variance u
c
2
(
x
) = Su
i
2
(
x
) = 2.43E-05 %
2
Combined standard uncertainty u
c
(
x
) = 4.93E-03 %
Effective degree of freedom
eff
(
x
) = 1659378.7
Coverage factor k = 2
Expanded uncertainty U = k u
c
(
x
) = 0.0099 %
Calibration uncertainty U

= 0.010 %
Standard
uncertainty
Value of standard
uncertainty c
i
f /x
i
u
i
(
x
)
c
i

u (x
i
)
Degree of
freedom

i
Type of
distribution