In the course of the energy transition, energy experts are increasingly discussing infrastructures that transmit electrical energy using direct current. High-voltage direct current transmission is becoming established for long distances and high outputs because it is significantly less lossy. However, the real revolution - direct current grids for low and medium voltages too - is yet to come. Data centers, production lines, cooling systems and LED lights work with DC voltage anyway; eliminating the need to convert between AC and DC voltage would save considerable energy and costs. The same applies to households: Many small appliances actually require DC voltage, which they have to convert with losses using switching power supplies, for example. Electric cars and photovoltaic systems also work with direct voltage. An energy supply with direct current is therefore not a fiction, but could be an imperative of common sense in the coming decade.

Time lapse in a water bath

However, there are still a few technical issues to be resolved before this can happen. One challenge is the development of circuit-breakers for DC voltage, as conventional AC voltage switches are not suitable for DC voltage to the same extent. What has hardly been discussed to date, however, is the question of whether low-voltage AC cables are also suitable for DC voltage. There have been no research results on this - presumably because most experts were of the opinion that this was possible without further ado. A mistake, as long-term tests by a research group in the "Electrical Devices and Systems" department at the Technical University of Ilmenau have now proven for the first time!

The DC voltage was supplied via a feed line to the terminal blocks in which the individual wires were attached on one side.

© TU Ilmenau

Over a period of 2590 hours, individual conductors with different insulation materials were subjected to a 1 kV DC voltage in a water bath at 80 °C in order to understand the effects in fast motion. The test equipment and the cables were provided by Lapp.

The results are striking. Of the 238 test specimens, 44% failed under DC load, while only one specimen failed under AC load. Cables with PVC insulation were particularly affected - all test specimens failed under DC load. The failure rate was just as high with polyolefin insulation. Cables with TPE insulation passed the test very well. One third of the defects were caused by insulation faults. These occur when the insulation resistance falls below a critical value of 10 MΩ. In two thirds of the defects, there was a visible breakdown of the insulation.

The failures occurred much earlier in cables that were subjected to additional mechanical stress due to tight bending radii. Before ageing began, new cables withstood approximately the same voltages, regardless of whether they were DC or AC. Only the ageing process caused the cables with DC stress to fall behind. The wall thickness of the insulation also played no decisive role in the frequency and timing of the failures.

DC aging behavior different

The research results are an important indication that the electrical field caused by the DC voltage has a different effect on the ageing behavior of insulation materials than an AC voltage field. Many experts have not yet commented on this. In a classic AC cable, the electric field strength is highest directly on the surface of the copper conductor and decreases with the radius outwards. The same applies approximately at high temperatures because the dielectric constant is almost independent of the temperature. With DC cables, on the other hand, the electrical conductivity of the insulating material determines the electric field strength distribution. If the conductivity of the insulation increases with increasing temperature, the electric field decreases only slightly with the radius to the outside - at high temperatures, a field inversion can even occur: The electric field strength is higher on the outside of the insulation than on the inside of the conductor.

Only single wires tested

Of the 238 cables tested, 44% failed under DC load. The picture shows a hole due to a breakdown and the stressed surface due to ageing.

© TU Ilmenau

In order to rule out any mutual influence between the individual wires and the sheath during the tests, the laboratory tests were first carried out on individual wires in water. This meant that only their insulation strength was tested. This does not correspond to the conditions of use in practice, as there are usually several cores in a cable that are encased in a sheath. In future, measurements will be carried out on complete cables.

The visual phenomena raise the question of the suitability of the test method used. It is true that the method of accelerating ageing in a water bath complies with the specifications of DIN VDE 0276-605, has proven itself for AC cables and provides good indications of the actual ageing behavior over long periods of years and decades - even for PVC cables. However, it is questionable whether this also applies to operation with DC voltage. Here, water seems to play a different, unforeseen role, in that it obviously accelerates the time lapse considerably.

Further research needed

Further research is needed to come to reliable conclusions here. The working group at TU Ilmenau is therefore planning ageing tests that do not require a water bath, but would then have to take longer.

On the other hand, the group is interested in understanding what happens chemically and physically in the plastic when the aforementioned reversal of the electric field occurs. Possible causes could be the degradation of the polymer, swelling in water, the release of additives or the formation of 'water trees'.

However, until reliable data is available, there is no reason to dispense with cables with PVC insulation in DC applications. However, the prerequisite is that these cables are laid firmly - i.e. without movement - and without mechanical stress, for example due to excessively tight bending radii. In addition, the environment should always be dry.

If these conditions are not met, for example in moving applications in energy chains, users should switch to other insulation materials. TPE has performed excellently in water bath tests.

Author:
Prof. Dr. Frank Berger is head of the "Electrical Devices and Systems" department in the Faculty of Electrical Engineering and Information Technology at Ilmenau University of Technology.