Xiaoshe Zhai

Determination of opening velocities for vacuum circuit breakers at transmission voltage

The objective of this paper is to propose a quantitative method to determine opening velocities for VCBs at transmission voltage. A couple of 2/3 coil-type axial magnetic field (AMF) contacts, which are used in a 126 kV vacuum interrupter, are arranged in a demountable vacuum chamber. The movable contact is opening with three different kinds of opening travel-curves. Each curve is adjusted by two opening velocities, v1 and v2, respectively. The v1 is an average opening velocity within 1/3 full contact gap, which is with 1.8 m/s, 2.4 m/s and 2.7 m/s, respectively. The v2 is an average opening velocity within 2/3 full contact gap, which is with 2.4 m/s, 2.7 m/s, and 3.0 m/s, respectively. The test current ranges from 4 kA to 40 kA, stepped by 4 kA. When a VCB is used to interrupt a short-circuit current, the preferred opening velocity is determined by reducing the arcing interval tI for the intense arc mode and increasing the arcing interval tII until the formation of high-current anode mode (HAM). The preferred v1 is higher than a threshold value v1_th, at which there is a peak critical contact gap dI_peak, for the evolving from intense arc mode into diffuse arc mode. The preferred value v2 is lower than another threshold value v2_th, at which there also exists a peak critical contact gap dII_peak, for the evolving from diffuse arc mode into the footpoint mode and anode spot mode.

Opening velocity characteristics for vacuum interrupters with cup-type axial magnetic field contacts

The objective of this paper is to experimentally determine an opening velocity for vacuum interrupters, with cup-type axial magnetic field (AMF) contacts. The opening velocity characteristics of the VCB includes an initial opening velocity v₁ and an average opening velocity v₂. The initial opening velocity v₁ is defined as an average velocity in 1/2 full contact gap of 20 mm. The average opening v₂ is defined as an average velocity of full contact gap. Experiment result shows that a higher v₁ always decreases the duration for the intense arc mode. However, a higher v₂ tends to result in an earlier formation of anode spot mode. The preferred initial opening velocity v₁ should higher than a threshold velocity v_{1_th}, at which the peak critical contact gap d_{I_th} is achieved, for the intense arc mode evolving into diffuse arc mode. The preferred average opening velocity v₂ should be at a threshold velocity v_{2_th}, at which the maximum critical contact gap d_{II_th} is achieved, for the diffuse arc mode evolving into the large-gap active-anode mode. The lower of v₂ below the v_{2_th} the better, as the large-gap active-anode mode would be avoided.

Back-to-back capacitor switching performance of vacuum interrupters after high frequency sub-microsecond voltage impulse conditioning

The objective of this paper is to investigate back-to-back capacitor switching restrike characteristics of vacuum interrupters (VIs) after conditioning by using a series of high-frequency sub-microsecond voltage impulses. 7.2 kV VIs are conditioned by several hundreds of batches of sub-microsecond high voltage impulses (1000Hz, 0.1s). The peak value of each impulse reaches 100kV. If a breakdown occurs under a voltage impulse, a high frequency conditioning current of several kA (peak value) would flow through the VI. Then the back-to-back capacitor switching tests are carried out at the inrush currents with peak values of 2 kA and 5 kA, respectively. The peak value of the recovery voltage reaches about 15 kV. The experimental results show that the restrike probability can be reduced significantly by using this high frequency voltage impulses conditioning method. By using VIs which are not conditioned as a benchmark, the restrike probability of the conditioned VIs is reduced from 17.9% to 9.6% at the inrush current of 2 kA. With a higher inrush current of 5 kA, the restrike probability reaches higher to 30.4% with the no conditioned VIs. However, the restrike probability reduces significantly to 7.9% after the conditioning. For VIs which are conditioned, the contact surfaces are fully covered by the molten metal layer resulting from the conditioning current. The contact welding areas caused by the inrush current are much broader and the depths are much shallower than that of VIs in benchmark.

Drawn vacuum arc behaviors of 1/2 coil-type axial magnetic field contact

The objective of this paper is to investigate the transition of drawn vacuum arc anode discharge modes of the 1/2 coil-type axial magnetic field contacts. The diameter of the contact used in the experiment include 60 mm and 63 mm. The opening velocity is 1.5 m/s from the contact separation to the 75% of full contact stroke of 15 mm. A dynamic solid angle Ω_I as well as Ω_II is used to evaluate the transition of anode discharge mode. Either the Ω_I or the Ω_I is defined as a ratio of contact diameter on contact separation, at which the transition from intense arc mode to diffuse arc mode or from the diffuse arc mode to the anode spot mode occurs, respectively. Test results show that the transition from intense arc mode into diffuse arc mode is dominated by the Ω_I, rather than the axial magnetic flux density B_axial. However, the transition from diffuse arc mode into anode spot mode is dominated by the B_axial, rather than the Ω_II. The arcing interval t_I, from the initiation of vacuum arc to the evolving of the intense arc mode into diffuse arc mode, increases with the decreasing of Ω_I, and with the increasing of B_axial. The arcing interval t_II, from the initiation of vacuum arc to the formation of the anode spot, increases with the decreasing of Ω_II, and with the increasing of B_axial.

Distribution of high-voltage and high frequency impulse breakdown sparks in conditioned contact surface

The objective of this paper is to experimentally investigate the distribution of high-voltage and high-frequency (HVHF) impulse breakdown sparks in conditioned contact surface. A 10 kV commercial vacuum interrupter (VI) is used, which is machined with a glass envelope and an observing window on the main shield. The distribution of the breakdown position in contact surface follows the Gaussian probability density. The breakdown sparks gradually concentrate in the center of contact surface, with a continue applying of the impulse voltage after a sufficient conditioning process. The frequency of the impulse voltage has slight impact on the distribution of breakdown sparks. However, a higher amplitude of the impulse voltage results in a more concentrative distribution of the breakdown sparks. The larger of the contact gap, the more concentrative of the breakdown sparks in contact surface. The mean value μ of the breakdown position in contact surface shifts to the lower h value of the contact edge gap. The higher of the non-parallelism of the facing contact surface, the more concentrative of the breakdown sparks.

Research on 40.5kV two-break vacuum circuit breaker for switching capacitor banks

Restrike caused by interrupting capacitor banks with vacuum circuit breakers will lead to severe switching overvoltage, which may cause great damage to power system. Reducing the probability of restrike during capacitive current breaking is the ultimate solution to limit overvoltage. In this paper, a new design of vacuum circuit breaker with a very low probability of restrike is proposed based on the analysis of restrike mechanism. A kind of vacuum circuit breaker with two breaks is presented, to carry out the duty of capacitor bank switching, and further meet the requirement of class C2 in IEC terms. A capacitive current switching test circuit is developed, which consisted of a high frequency inrush current circuit and a power frequency circuit, to perform back-to-back capacitor banks switching test. The test results show that two-break vacuum circuit breakers can reduce the probability of restrike significantly; short-circuit making and breaking capacity of the vacuum circuit breakers is validated by synthetic test. The optimum value of the shunt grading capacitor is determined to ensure the stress of each break is approximately equal.

The reignition characteristics of double vacuum interrupters in series after interrupting vacuum arcs

Vacuum circuit breakers are extending their applications from low/medium voltage level to high voltage level due to their good properties. Multi-break vacuum circuit breakers are used in special applications because thry have better breaking capability and withstand higher breakdown voltage compared with single-break vacuum circuit breakers. The objective of this paper is to investigate the reignition characteristics of double VIs in series after interrupting vacuum arc. Two 24kV VIs, which have axial magnetic field contacts and can interrupt a rated short circuit current of 31.5kA, were connected in series. The experiments were carried out in the synthetic circuit providing transient recovery voltage (TRV) with a peak value of 85 kV and a interrupted current of 5 kA. The arcing time was 1ms corresponding to the contact gap 1.2mm. The TRV across VIs were measured by voltage dividers. The results showed that the upper VI was reignited at first due to its higher TRV caused by stray capacitances and then led to the reignition of the lower one. The breaking capability of the double VIs in series was determined by the dielectric recovery of the upper VI and the minimum recovery time was about 2μs.

Transfer Function and Network Synthesis of Electrical Fast Transient/Burst Generator Based on Latent-Roots Method

This paper presents a method to construct a transfer function (TF) of the pulse-forming network by using the latent-roots method for modeling the electrical fast transients/burst (EFT/B) generator. In the TF, two negative real roots characterize the nonoscillation process of the network, while one complex-conjugate pair characterizes the damping-oscillation process. All the four latent roots of the TF are determined according to the open-circuit specification by the manufacturer. The equivalent circuit of the EFT/B generator is synthesized with the Cauer method, and renormalized according to the matching circuit specification. Based on the aforementioned procedures, a parameterized circuit is realized, whose circuit parameters can be adjusted conveniently to meet the specifications identified by different manufacturers. Finally, the modeling method is validated experimentally.

Hybridized 3D-FDTD and circuit simulator for analysis of PCB via's signal integrity

In this paper, the equivalent circuit model containing two pairs of vias is established, and a hybrid method is applied, which combines the full-wave finite difference time domain (FDTD) with an S-parameter based macromodel by using the rational function approximation and circuit simulation. The electromagnetic feature of the model is described by S-parameters, result of the simulation of FDTD. The vector fitting method is adopted to perform the rational function approximation of complex pole pairs. This method is robust and can supply results precise enough for the procedure of macromodeling while the resonances on PCB under very high frequency are being considered. Based on above steps, the SPICE compatible equivalent circuit can be established and be connected with other circuit model in the same pattern. In this way, the analysis of signal integrity for a complicated interconnecting system can be performed more efficiently