This chapter discusses insulating materials and dielectrics.

The scope of the glass-insulator manufacturing processes is very much narrower than for porcelain. At present the use of toughened glass is confined to cap-and pin insulators or those types, such as railway pedestals and multiple-cone posts, which can be assembled from disc-like modules. In practice, therefore, the manufacture of toughened-glass insulators is con fined to the following stages: mixing the ingredients; melting the glass; forming and heat-treating the discs; elimination of defective pieces; attachment of metal fittings. It is evident that such simplicity cries out for long runs of standard pieces, and that, when these conditions are fulfilled, cheap and good insulators may be expected.

This chapter discusses housings for polymeric insulators.

An insulator comes to the end of its working life either when it fails mechanically, flashes over at unacceptably high frequency or gives evidence of deterioration to a condition likely to lower its factor of safety in service. All insulators are affected to some extent by impact, cycling both thermal and mechanical, ablation from weathering and electrothermal causes, flexure and torsion, ionic motion, corrosion and cement growth. There are, however, strong differences between ceramic and polymeric insulators, as classes. In general, a ceramic insulator will be vulnerable to impact damage, since its dielectric is a brittle material, and to processes which cause concentrations of tensile or shear stress.

The terms 'contamination' and 'pollution' have special meanings when applied to the condition of insulators. An insulator so heavily polluted by marine deposits that it flashes over immediately on energisation may appear to be perfectly clean, even on close inspection. On the other hand, one which is black with industrial soot, or which has some of its surfaces caked with cement, may have an electrical performance indistinguishable from that of a freshly installed counterpart. The reason for this apparent paradox is that values of surface electrical conductivity which are sufficient to cause flashover are quite trifling in absolute terms. They are readily achieved by the presence of soluble electrolytes, such as common salt or industrial acids, at densities of some 0.1 mg/cm2, provided water is available to dissolve them. They are not readily achieved by layers of carbon particles, which make only intermittent point contacts with each other, or by aggregates of mineral dusts which are free of ionic components (although combinations of such aggregates with soluble salts, giving a 'blotting paper' effect, have caused severe flashover problems in North Africa and in Cornwall, England).

In this chapter, attention is largely concentrated on predictive tests, aimed at determination of orders of merit, in ability to operate under contamination and at those made for purposes of research into insulator behaviour. Natural and artificial tests are covered separately.

Remedies are called for, evidently, when the flashover frequency rises above acceptable levels. What is 'acceptable' depends on the importance of the line or substation and on the required quality of supply, in terms of outage time per year. Standards thus vary widely: for lines, a rate of about one flashover per 150 km per year is general for industrialised countries in Europe, while much higher rates are tolerated elsewhere, e.g. in rural parts of the USA. Flashovers in substations often have serious consequences, and rates lower than one per year per station would normally be called for. The causes of flashover are part systematic and part random. In service, an insulator will carry a resident layer of contamination, accumulated since in stallation or the last cleaning operation, which fluctuates as the resultant of depositing and purging events, but is quasi-stable. The insulator is also challen ged by random occurrences like condensation, frost and onshore gales. These add water, ionisable material or both, which, depending on the design of the insulator, either will or will not carry the surface conductivity into a range where flashover can develop.

Interference with radio and television (RI and TVI) may arise when electrical discharges run on insulators and inject high-frequency currents into associated conductors, which radiate electromagnetic waves. Audible noise (AN) is generated either by electrical discharges or by an entirely different process, resonance in cavities of insulators, excited

The principal shapes of insulator are shown. Vertical-loaded strings of cap-and pin discs are known, in England, as 'suspension sets' and when horizontal loaded as 'tension sets'. For horizontal duty the shape is usually modified towards lesser convolution. 'Pedestal posts' are either assembled from single units of metal and porcelain, bolted together, or in the 'polyped' form. Sub-assemblies using continuous metal parts and embodying several units of metal and porcelain are bolted together. The 'solid-core' post is of continuous porcelain with metal fittings, like the 'longrod' and 'motor' but of much larger diameter. Its appearance is much the same as that of the 'multiple cone' which, as shown, comprises both porcelain and cement.

This appendix contains a selective bibliography of 18 references on the live washing of insulators.