The IEC TR 60890 provides an empirically based method for calculating the air temperature inside LV switchgear. One might assume that the method could be applied for estimating the temperature rise of the air inside MV equipment too. However, the IEC method assumes the uniform distribution of the power input, which is not normally the case for MV equipment. This paper explores what happens when the IEC TR 60890 requirement of uniform heat input is violated.The presented experiments and simulations show that changing the height of the heat source significantly affects the cooling conditions of the enclosure and therefore the air temperature distribution. The temperature distribution factor should be adjusted to apply the IEC method if the heat source is located in the upper part of the enclosure.

Electrical equipment will experience a rise in temperature during normal operation. During a development process, prototypes and laboratory tests will be required to make sure the temperature rises are within acceptable limits as defined by standards. The aim of having a tool to predict the temperature rise, is to reduce the number of prototypes and test loops needed in the laboratory during a development period. Advanced simulation tools such as CFD can give valuable results, however, they require expertise user and extensive compute and manpower allocation. This paper presents a practical design approach developed for providing a first, quick and rough estimate of the temperature rise of the most critical parts in an air insulated switchgear. The main idea behind the method is to first use the method described in IEC 60890 to estimate the temperature rise of the gas inside the switchgear. Then, simplified heat transfer calculations are used to estimate the over-temperature of critical parts relative to the surrounding gas. The accuracy of the temperature estimates will depend on how well the power input is known, especially the contact resistances. Further, it may be challenging to predict the influence of large metallic construction elements that may function as heat sinks.

MV switchgear experiences a rise in temperature during normal operation due to ohmic losses in conductors and contacts. If the temperature rise is too high, the switching device may be degraded. The focus of this paper is to find a value for the total heat transfer coefficient that may be applied to estimate the temperature of critical parts (open/close contact) of the load break switch in an enclosed MV switchgear, relative to the surrounding air (inside the enclosure) for future design. The values for the total heat transfer coefficient (including all transfer mechanisms) showed a relatively strong dependence on the surface emissivity and the actual design of the switch, but was less dependent on temperature changes within the relevant temperature range. Based on our findings, it is reasonable to assume that the total heat transfer coefficients may be applied in a first approximation of the temperature rise of a load break switch contacts relative to the surrounding air inside an enclosure. Further refinement could be obtained by taking the actual design of the switch into consideration, especially details influencing the emissivity and design elements influencing the heat conduction to adjacent conductor parts.

Arc fault tests of medium voltage switchgear have been performed with reduced volume and arc energy. This has been done in order to investigate if small scale experiments can be used to predict the pressure build-up during full scale arc fault test. Between 40 and 50 % of the arc energy was transferred to the gas in both small scale and full scale tests. The results show that it is reasonable to assume that small scale testing down to 10 % can be used to predict the pressure rise in a full scale test with a single phase arc with about 10 % precision. However, the scaling method of the arc energy seems to be important.

An arc fault inside metal enclosed switchgear will cause the pressure to rise and vaporization of electrode material may contribute to the pressure rise. An experimental study of high current arc erosion on copper electrodes in air has been performed, with an evaluation of fraction lost by gross melting and vaporization. All experiments were performed at NEFI High Voltage Laboratory in Skien, Norway. The measured mass loss from vaporization in our experiments seems to be negligible compared to erosion by gross melting.

No abstract

No abstract

Full scale arc fault experiments are expensive, time consuming, and demand high power laboratory facilities. This paper presents a literature study performed to evaluate the possibilities of performing representative experiments on small scale by reducing the size of the test object. Previous model experiments have shown that the arc energy must be scaled down by the same factor as the volume in order to produce the same pressure build-up during the tests. The arc energy is given by the arc voltage and current, which both are functions of the arc length. In arc fault experiments one usually assumes a linear voltage drop along the length of the arc and experiments have shown that a voltage gradient of 20-30 V/cm can be assumed for arcs in switchgears filled with air. This value will however depend on volume, pressure, and electrode material. Experiments also indicate that there is a limit for short arcs where the linear voltage drop no longer can be assumed. To reveal possible consequences this may have on small scale experiments, a more systematic experimental study is required.