Column: Tap-changer know-how

Column: Tap-changer know-how

Vol. 5 Issue 1

Tap-Changer DGA – Uncovering an Enigma

 

Part 3: Interpretation

Abstract

Gas generation in OLTCs is influenced by operational and design parameters. There are multiple gas generating components present depending on the specific OLTC model or class resulting in a superimposed gas pattern thus making its interpretation for the purpose of fault detection more difficult. In many cases, suspicious behaviour can be revealed, and in some. the normal bandwidth of observed gas pattern can fully mask a thermal or electrical fault. The DGA results can be misleading or fail to identify an actual fault, so they should be confirmed by additional testing and never be used as a single criterion for shutdown or repair decisions.

 

Keywords: on load tap changers (OLTC), dissolved gas analysis (DGA), gas pattern, fault detection

Part 2 of this article [19] showed different pathways how to evaluate the DGA patterns observed in OLTCs. The numerical approach is based on a statistical evaluation of as much as possible DGA data of a selection (class) of OLTCs, while the phenomenological approach starts by identifying the gas generating components inside an OLTC and assigns typical gassing patterns to each component. The ppm values for each gas generating component vary with the actual operating conditions. Depending on the specific OLTC model or class, there may be more than one gas generating component present, so the respective gas patterns superimpose and give a resulting mix gas pattern which may be clearly interpretable – or not. But before we discuss these mix gas patterns, we should first identify the parameters which influence the gas generation and assign possible tap-changer faults to fault gas patterns.

  • Parameters influencing gas generation

The factors which determine the amount of gases generated by the respective component have been listed in Fig. 11 of Part 2 of this article [19]. A few more words are advisable.

OLTC type

The main difference in gassing appears between non-vacuum and vacuum type OLTC models. While non-vacuum type models can generate more than 100,000 ppm of combustible gases, vacuum type models only generate very low amounts of gases, usually in the same range as in the transformer. See also of Part 1 of this series, section 2.2, for more details [9]. Also the influence of the oil volume is very evident: gases generated by the same energy amount cause only the half ppm amount if the oil volume is doubled.

Switching principle

Several different switching mechanisms are available to perform the load switching operation. In case of arc-switching models for example, the switching contacts can break the current by one to four switching arcs connected in series, and two or four transition contacts can be used to smooth the load switching operation. Vacuum type models show an impressive variety of possible switching principles, using one to four vacuum interrupters per phase. Depending on that, auxiliary and commutation contacts may be necessary to allow a reliable load switching operation without interruption of the load current. For details see [20]. The sparking intensity on the commutation contacts is mainly determined by the internal impedances of the tap-changer.

Design of transition impedance

In combination with the step voltage, the ohmic value of the transition resistors or the impedance of the transition reactance determines the circulating current, which has to be broken by the transition contacts. The transition impedance values are usually set in a way that the transition contacts see a similar load like the main switching contacts. The cross section of the transition resistors is optimized to the individual application so that surface temperatures <300 °C are achieved for normal operation. Temperatures beyond 450 °C are only exceeded in very severe overload cases which are rare.