Question

Summarize the various conditions and minimum clearances required for the accurate measurement of peak voltage using sphere gaps.

Answer 1

R B D S A

S R A
The breakdown voltages of uniform and slightly non-uniform electric fields, as
e.g., those between sphere electrodes in atmospheric air, show high stability and low dispersion. Schumann (1923) proposed an empirical criterion to estimate the criticalfield strength at which self-sustaining electron avalanches are ignited. modified, this criterion can also be used to calculate the breakdown voltage Vb of uniform fields versus the gap spacing S. For a uniform electric field in air at standard conditions the breakdown voltage can be approximated by the empirical equation
Vb=kV ¼ 24:4 S þ S
13:1 cm 0:5 " #: ð2:33Þ
This equation is applicable for sphere gaps if the spacing is less than one-third
of the sphere diameter .
Based on such experimental and theoretical results, sphere-to-sphere gaps are
used for peak voltage measurement since the early decades of the twentieth century and led to the first standard of HV testing, the present IEC 60052:2002.
Meanwhile it is fully understood that this applicability is based on the so-called
streamer breakdown mechanism, e.g., Meek (1940), Pedersen (1967), and break- down voltage-gap distance characteristics of sphere gaps can also be calculated with sufficient accuracy (Petcharales 1986).
For a long time measuring sphere gaps with gap diameters up to 3 m formed
the impression of HV laboratories. But the voltage measurement by sphere gaps is connected with the breakdown of the test voltage therefore their application is not. simple Furthermore they need a lot of clearances (see below), well maintained clean surfaces of the spheres and atmospheric corrections for measurement according to the standard.
Today they are not used for daily HV measurement and do not play the same important role in HV laboratories as in the past. Their main application is for performance checks of AMSs or linearity checks
For acceptance tests on HV apparatus the inspector may require a check of the applied AMS by a sphere gap to show that it is not manipulated. For these applications mobile measuring gaps with sphere diameters D B 50 cm are sufficient.
The IEC Standard on voltage measurement by means of sphere gaps has been
the oldest IEC standard related to HV testing. Its latest edition IEC 60052
Ed.3:2002 describes the measurement of AC, DC, LI and SI test voltage with
horizontal and vertical sphere-to-sphere gaps with sphere diameters D = (2 …200 cm) and one of the spheres earthed. The spacing S for voltage measurement is required S B 0.5 D, for rough estimations it can be extended up to S = 0.75 D. The surfaces shall be smooth with maximum roughness below 10 lm and free of irregularities in the region of the sparking point. The curvature has to be as uniform as possible, characterized by the difference of the diameter of no morethan 2 %. Minor damages on that part of the hemispherical surface, which is not involved in the breakdown process, do not deteriorate the performance of the measuring gap. To avoid erosion of the surface of the sphere after AC and DC, breakdowns pre-resistors may be applied of 0.1–1 M9X.
Surrounding objects may influence the results of sphere gap measurements.
Consequently the dimensions and clearances for standard air gaps are prescribed in IEC 60052 and shown in Figs. 2.19 and 2.20. The required range of the height A above ground depends on the sphere diameter, and is for small spheres A = (7 … 9) D and for large spheres A = (3 … 4)- D. The clearance to earthed external structures depends on the gap distance S, and shall be between B = 14 S for small and B = 6 S for large spheres.
The dispersion of the breakdown voltage of a measuring gap depends strongly
from the availability of a free starting electron, especially for gaps with
D B 12.5 cm and/or measurement of peak voltages Up B 50 kV. Starting elec-
trons can be generated by photo ionization (Kuffel 1959; Kachler 1975). The
necessary high energy radiation may come from the far ultra-violet (UVC) content of nearby corona discharges at AC voltage, or from the breakdown spark of the openswitching gaps of the used impulse generator, or a special mercury-vapour UVC lamp with a quartz tube.
Note In the past, even a radioactive source inside the measuring sphere has been applied.
For safety reasons this is forbidden now.
Table 2.7 gives the relationship of the measured breakdown voltage Ub
depending on the distance S between electrodes for some selected sphere diameters D B 1 m which are mainly used for the mentioned checks, for other sphere diameters see IEC 60052:2002. A voltage measurement with a sphere gap means to establish a relation between an instrument at the power supply input of the HVG (e.g., a primary voltage measurement at the input of a test transformer) and the known breakdown voltage of the standard measuring gap in the HV circuit depending on its gap distance D. This is similar to the calibration by comparison.
For AC voltage measurement a progressive stress test delivers 10
successive breakdown voltage readings by the instrument. Their mean value and the relative standard deviation are determined. The voltage shall be raised sufficiently slowly to allow accurate readings. The mean value characterizes the breakdown voltage according to the gap parameters (D, S).
When the standard deviation is B1 %, one can assume that the measuring gap was correctly maintained and the relative expanded uncertainty of measurement is B3%.
Note With n = 10 measurements and a standard deviation of 1 % one gets a standard uncertainty of u = 0.32 % (Eq. 2.14). This means there are about 1.2 % for the othercontributions to the standard uncertainty when the expanded uncertainty (k = 2) shall be
B3 % (Eq. 2.29).
For LI/SI voltage measurement, the pre-selected breakdown voltages (D, S in
Table 2.7) are compared e.g., with charging voltage of the impulse voltage gen-
erator. The 50 % breakdown voltages U50 are determined in a multi-level test of
m = 5 voltage levels with n = 10 impulse voltages each (see Sect. 2.4), and the
corresponding reading is taken as the pre-selected reading. When the evaluated standard deviation is within 1 % for LI and 1.5 % for SI voltages it is assumed that the measuring gap works correctly.
For DC voltage measurement, sphere gaps are not recommended because
external influences as dust or small fibres are charged in a DC field and cause a
high dispersion. Therefore, a rod–rod measuring gap shall be applied if the
humidity is not higher than 13 g/m3 (Feser and Hughes 1988; IEC 60052:2002).
The rod electrodes of steel or brass should have a square cross section of
10–25 mm for each side and sharp edges. When the gap distance S is between 25 and 250 cm the breakdown is caused by the development of a streamer discharge of a required average voltage gradient e = 5.34 kV/cm. Then the breakdown voltage can be calculated by
Vb=kV ¼ 2 þ 5:34 -
S=cm: ð2:34Þ
The length of the rods in a vertical arrangement shall be 200 cm, in a horizontal
gap 100 cm. The rod–rod arrangement should be free of PD at the connection of the rods to the HV lead, respectively to earth. This is realized by toroid electrodes for field control. For a horizontal gap the height above ground should be C400 cm.
The test procedure is as that for AC voltages described above.