Pollution flashovers of insulator strings have historically been one of the main causes of line outages under normal operating conditions. Indeed, years ago this was a relatively common occurrence across many countries suffering from heavy industrial pollution, for example much of Eastern Europe. The situation began to improve over the past two decades as new ways were developed to successfully combat air pollution. As a result, the main pollution sources impacting overhead lines these days have become mostly natural – either of the marine or desert type. While industrial pollution remains a problem in some developing countries, it typically impacts only those sections of lines closest to heavy industry.
Fortunately, there is already a vast body of international experience on polluted insulators. Therefore, it is fair to say that the problem of pollution and its impact on line insulators is well understood as are the measures by which it can be combated.
Bird-induced line outages have also become more and more an issue in recent years. However, as is the case of pollution, there is now quite a body of knowledge on how these can best be minimized.
By contrast, today there seem a high proportion of outages that cannot easily be attributed to either pollution or birds – or indeed any specific cause. One can describe the same typical scenario for these outages, including affecting clean or only lightly polluted insulators and occurring most often in the early morning, in the presence of dew. Similarly, these outages tend to be single phase rather than three phase with no evidence of pollution during visual inspection after successful re-closing.
This INMR article from 2011, contributed by Igor Gutman in Sweden, Evgenij Solomonik of NIIPT in Russia and Wallace Vosloo of Eskom in South Africa, looked to share experience on this category of outage commonly referred to as due to ‘unknown causes’. By all accounts their percentage is quite high, as for example in Russia where they make up some 30% of all line outages at 220 kV and 500 kV.
Resolving such unexplained outages is increasingly important as power companies the world over try to establish norms on acceptable levels of outages at each voltage class. To accomplish this, however, they need to better understand what is really be happening to insulators on their lines.
Table 1 left summarizes the unusually high level of outages due to flashover that cannot be directly attributed to either pollution or birds. Depending on country and voltage class, these often represent the single major cause of service interruptions.
Renewed interest in this area began in Sweden with the unexpected flashover of a porcelain support insulator on the test transformer at STRI’s high voltage hall in Ludvika. The affected insulator’s specific creepage distance was 8 mm/kV while relative humidity (RH) at the time was about 80%.
This same phenomenon was later successfully simulated during a laboratory test, where high humidity was created in the vicinity of the insulator’s ‘foot’ (i.e. the support insulator) using a continuous spray of water but with no direct spraying to its surface. As such, the presence of water on the insulator could not be visually observed and wetting would rather take place only at the microscopic level. Under these test conditions, flashover occurred at 400 kV, which correlated well with this previous incident.
To gain more insight into the problem, the insulator’s E-field profile was measured using special probe at both 100 and 200 kV (see Figure 1).
It is obvious that at 80% RH the electrical field stress is concentrated near the top of the support insulator. Higher temperature was also observed here, indicating that moisture was evaporating due to the presence of small leakage currents. Later, a leader-type discharge was initiated, effectively short-circuiting this section of the insulator and resulting in the 400 kV flashover.
A similar outcome, however, could not be simulated on strings of glass cap & pin insulators and a possible explanation is their different E-field orientation, i.e. not along the insulator (see Figure 2).
With these considerations in mind, STRI began to collect service data on several 110 kV to 130 kV overhead lines in Sweden and Finland covering several years. Below are some of the main findings that seem to always apply to outages attributed to ’unknown’ causes:
• These usually take place in areas of very light contamination, i.e. ESDD/NSDD of only 0.01/0.06 mg/cm2
• Outages tend to coincide with an average RH of 91%
• The affected glass cap & pin insulators all have an insulation level of 11-12 mm/kV.
Fingrid, the transmission network operator in Finland, conducted their own investigation into these outages and concluded that light pollution was the best explanation. This is because data from pollution monitoring confirmed that such ’unknown’ faults generally took place when leakage currents were highest – which also correlated with periods of high humidity. Most of the outages, in fact, occurred during the early morning hours, while some of the flashed over insulators were located close to highways.
Indeed, pollution tests of insulator strings in the laboratory revealed that, for ESDD levels comparable to those seen in service, flashovers were recorded at only 87 kV (for 110 kV strings). This meant that the safety factor between the test voltage and the maximum operating voltage was only 1.2 and suggested that Finngrid should avoid relying on only six standard profile insulators for its 110 kV lines.