Outdoor Insulation Contamination (Part One)

Section 1

When Contamination Governs the Insulation Design

Outdoor insulation is normally sized to withstand lightning impulses, switching impulses, power-frequency voltage and temporary overvoltages. But in polluted, wet environments a fifth stress — contamination — can become the criterion that actually decides the insulator length.

In many insulation-coordination studies, lightning and switching surges receive most attention. For outdoor insulation, however, contamination can become the dominant design criterion: the required insulation length may be governed not by impulse withstand but by the ability to hold normal system voltage under polluted, wet conditions.

This matters for overhead-line insulator strings, post insulators, bus supports, transformer and circuit-breaker bushings, disconnectors and outdoor station equipment. Where pollution is severe, the creepage length needed for contamination can exceed the length needed for lightning or switching performance — so the final design is dictated by contamination.

This first part builds the physical foundation: how contamination flashover starts, how pollution severity is measured, and how ceramic-insulator strength is represented through CFO and creepage equations. Part Two then converts these concepts into insulation-coordination calculations — deterministic design, altitude correction and impulse effects. Part Three continues with composite insulators, RTV coatings, ageing, icing and bird-related flashovers.

What this page teaches

why contamination is a separate insulation-design criterion;
the surface flashover mechanism — wetting, leakage current, dry bands and arcing;
how site severity is measured (SDD, ESDD, layer conductivity, salinity);
the laboratory test methods — salt fog and solid-layer procedures A and B;
CFO and specific-creepage equations for ceramic cap-and-pin insulators;
how to turn severity into a required creepage distance and insulator length.

Section 2

Contamination, Pollution, Creepage and Leakage Distance

Two pairs of words are used interchangeably throughout this topic. Contamination and pollution mean the same thing; creepage distance and leakage distance are synonyms.

Contamination / pollution

The deposition of conductive or semi-conductive material on the surface of outdoor insulation — sea salt, road salt, industrial or cement dust, fly ash, limestone dust, chemical deposits, or gases that become conductive when moist.

Creepage / leakage distance

The distance measured along the surface of the insulator between two conductive parts — not the straight-line air distance. It is the path the leakage current follows over the contaminated surface, and it is one of the most important parameters for contamination performance.

Creepage is not arcing distance

Creepage distance is the path along the insulator surface. It is different from arcing distance, which is the shortest air path between the energised and earthed fittings. Contamination performance is mainly governed by the surface path, because the leakage current flows along the polluted, wet surface — contamination is a surface phenomenon, not simply an air-clearance problem.

Section 3

Why Contamination Can Dominate the Design

Lightning and switching surges are short-duration events. Contamination stress is different: contamination flashover occurs under normal or temporary power-frequency voltage when the surface becomes polluted and wet. Its performance is therefore tied to the continuous operating voltage, the surface condition, weather, moisture, pollution accumulation, creepage distance, and the insulator material and shape.

Key idea

In contaminated areas, sufficient BIL or switching-impulse withstand is not enough. The insulator must also have enough creepage distance and a suitable surface design to hold the operating voltage under polluted, wet conditions.

Section 4

A Deterministic Design Philosophy

A probabilistic method is theoretically possible, as for switching-surge coordination. But contamination behaviour depends on many uncertain variables — pollutant type and amount, wetting and drying pattern, wind direction, rain washing, insulator geometry, ageing and maintenance. Because of this, a conservative deterministic rule is normally applied:

\[ \text{Minimum Strength} \;\ge\; \text{Maximum Stress} \]

The selected insulation should have enough withstand strength for the maximum expected contamination stress. The approach is conservative but practical.

Section 5

Ceramic and Non-Ceramic Outdoor Insulation

Ceramic insulation

Porcelain and glass insulators have been used for many decades, so the practical knowledge and test experience for ceramic insulation is extensive. This page focuses on ceramic insulation.

Non-ceramic insulation

Polymeric or composite insulators — typically silicone rubber — are widely used for lines, substations and bushings. Their contamination performance can be very good, mainly because of hydrophobicity (the surface resists forming a continuous water film). Their long-term ageing and material degradation must, however, be considered.

Section 6

The Basic Flashover Mechanism — Two Events

Event 1 — pollutant deposition

A sufficient amount of contaminant must be deposited, and it must contain soluble ionic material (such as salt) so it can become conductive when wet. Dry contamination alone is usually not dangerous — a dry polluted layer has high resistance and may sit on the insulator without causing flashover.

Event 2 — wetting without washing

The polluted surface must then become wet from fog, mist, light rain, drizzle or dew. The wetting must moisten the surface without washing the pollution away.

Why light wetting is the dangerous case

Light rain, fog, mist or dew can be more dangerous than heavy rain, because they wet the pollution layer without removing it. Heavy rain may wash the surface and reduce contamination severity, while light wetting creates a conductive film over the contaminant that is still present — which is exactly the condition that starts the flashover process.

Section 7

Conductive Film and Leakage-Current Heating

When moisture mixes with the deposited contaminant, a conducting film forms on the surface and leakage current flows along the creepage path from the energised end to the grounded end. The insulator now behaves like a high-resistance surface conductor, and the leakage current produces heat:

\[ P = I^2 R \]

\(P\): power dissipated as heat in the wet layer
\(I\): surface leakage current
\(R\): resistance of the wet contaminated layer

The heating is not uniform, because the current density varies along the surface — and that non-uniformity is what triggers the next stage.

Section 8

Dry-Band Formation

Some regions have a smaller diameter or a narrower surface path — the neck of a post insulator, the rib area under a line insulator, narrow shed regions. There the leakage-current density is higher, so the heating is greater. The local heating dries the pollution layer and creates a dry band: a narrow region where the surface layer has dried and now has much higher resistance than the surrounding wet surface.

Section 9

Voltage Across a Dry Band

Once a dry band forms, the surface current path is interrupted. The wet regions stay conductive, but the dry band becomes highly resistive, so a large part of the line-to-ground voltage appears across it:

\[ V_{\text{dry band}} \approx V_{\text{applied}} \times \frac{R_{\text{dry band}}}{R_{\text{total}}} \]

\(V_{\text{dry band}}\): voltage appearing across the dry band
\(V_{\text{applied}}\): applied line-to-ground voltage
\(R_{\text{dry band}}\): resistance of the dry band
\(R_{\text{total}}\): total surface resistance

Because \(R_{\text{dry band}} \gg R_{\text{wet layer}}\), a large voltage can appear across a small dry region — concentrating the stress exactly where the surface is most resistive.

Section 10

Dry-Band Arcing and Complete Flashover

If the stress across the dry band is high enough, the band flashes over and a small arc forms. The arc can grow: as it extends it heats and dries adjacent wet regions, lengthening the arcing path, and the process can become self-sustaining. Eventually several dry-band arcs extend and join; when they bridge the full insulation distance, complete flashover occurs.

The full flashover chain, in one place

dry pollution deposits on the surface;
light wetting (fog, mist, dew, drizzle) turns the pollution into a conductive film;
leakage current flows and heats the layer;
local heating dries narrow regions and forms dry bands;
most of the voltage appears across the dry bands;
dry-band arcs ignite;
the arcs extend along the wet surface;
complete flashover occurs when an arc bridges the insulation.

Contamination flashover is therefore not a puncture failure — it is a surface process driven by leakage current and arcing.

Section 11

Direct Methods of Selecting Insulation

Test stations along the right-of-way

Different insulator types and string lengths are installed along the proposed route and exposed to the actual local environment; their performance is observed over time. The advantage is realism — it naturally includes local pollution, weather, wetting patterns, wind direction, natural rain washing and seasonal effects. The disadvantage is time: a long period is needed to collect enough flashover data, so it is useful but not always practical for project timescales.

Using existing line performance

If existing lines in the same area perform acceptably, a similar design may be chosen for the new line. For a different voltage level the string length may be scaled approximately with voltage — assuming contamination performance is roughly proportional to string length, which appears justified up to at least 345 kV transmission. The limitation: if the reference line is very conservatively insulated with an extremely low flashover rate, it is hard to know how much margin actually exists.

Section 12

Sea, Industrial and Mixed Contamination

Contamination is broadly classified as sea or industrial — though in practice it is often mixed. In some surveys the most frequent type was mixed contamination containing both industrial and sea contaminants, so real-site pollution is often not a pure laboratory pollutant.

Table 1 — Sea versus industrial contamination.

Aspect	Sea Contamination	Industrial Contamination
Source	Wind-borne salt from the sea	Cement / chemical / coal plants, fly ash, limestone dust, road pollution
Deposition & wetting	Often arrive together (salt spray, sea mist)	Accumulates dry; dangerous when later wetted by fog, mist or dew
Build-up	Can wet and deposit simultaneously	Slow accumulation over time
Representative test	Salt fog method	Solid-layer methods (A and B)

Section 13

Distance from the Pollution Source

Severity usually reduces with distance from the source. For sea salt, contamination decreases rapidly beyond about 50 km from the sea; for industrial plants, a noticeable decrease may begin beyond about 500 m from the plant.

These are general guidance, not universal rules. The actual severity depends on wind direction, terrain, rainfall, source strength, humidity, local obstacles and line orientation.

Section 14

Salt Deposit Density (SDD) and Equivalent SDD (ESDD)

Salt Deposit Density (SDD) is the amount of salt contamination per unit insulator surface area, normally in mg/cm²:

\[ SDD = \frac{\text{mass of salt}}{\text{insulator surface area}} \qquad \left[\,\text{mg/cm}^2\,\right] \]

Higher SDD means more soluble conductive material and more severe contamination. Real pollution, however, is rarely pure NaCl. Equivalent Salt Deposit Density (ESDD) standardises this: it is the amount of NaCl that would produce the same conductivity as the actual contaminant after complete dilution. ESDD converts different soluble pollutants to an equivalent NaCl amount, so different contamination types can be compared on a common basis.

Why ESDD is needed

Real pollution is not always pure NaCl — different salts and soluble pollutants produce different conductivities. ESDD converts them into an equivalent NaCl amount, so different pollution types can be compared on a common basis. That is why both SDD (the actual salt present) and ESDD (the NaCl-equivalent) are used.

Section 15

Layer Conductivity and Salt-Fog Salinity

The wet layer’s conductivity (in microsiemens, µS) indicates how easily current flows through it — higher conductivity means more leakage current and higher flashover risk. It relates approximately to ESDD by:

\[ \text{Layer conductivity}\;(\mu\text{S}) \approx 100 \times ESDD \qquad \left[ESDD \text{ in mg/cm}^2\right] \]

For example, \(ESDD = 0.05\ \text{mg/cm}^2\) gives a layer conductivity of about \(100 \times 0.05 = 5\ \mu\text{S}\). For sea contamination, severity is often expressed as salt-fog salinity (salt per volume of water, in kg/m³):

\[ \text{Salt-fog salinity}\;(\text{kg/m}^3) \approx 140 \times ESDD \qquad \left[ESDD \text{ in mg/cm}^2\right] \]

So \(ESDD = 0.05\ \text{mg/cm}^2\) corresponds to a salinity of about \(140 \times 0.05 = 7\ \text{kg/m}^3\). These relations let industrial (ESDD) and sea (salinity) severities be compared. The four severity measures, and which test method each belongs to, are summarised below:

Table 2 — The four contamination-severity measures.

Quantity	Unit	Mainly Used for	Meaning
SDD	mg/cm²	Solid-layer testing	Actual salt deposited on the insulator surface
ESDD	mg/cm²	Industrial / mixed pollution	Equivalent NaCl deposit giving the same conductivity
Layer conductivity	µS	Solid-layer procedure A	Conductivity of the wetted pollution layer
Salt-fog salinity	kg/m³	Salt-fog method	Salt concentration in the applied fog

Section 16

Site Severity Classes

Sites are categorised by ESDD. Different standards set the boundaries slightly differently, but the engineering idea is the same: higher ESDD means more severe pollution and therefore a longer creepage distance.

Table 3 — Approximate site-severity classes by ESDD.

Severity Class	ESDD (mg/cm²)
Very light	0.015 – 0.03
Light	0.03 – 0.06
Moderate	0.06 – 0.12
Heavy	0.12 – 0.24
Very heavy	0.24 – 0.48
Exceptional	> 0.48

Section 17

Why Contaminated-Insulation Strength Is Hard to Define

The strength of contaminated insulation depends on many variables: the type of line or station insulator, the type and amount of contaminant, the inert binder material, the insulator configuration and string length, the type of wetting, the droplet size, the relative contamination on top and bottom surfaces, and the test method. This is why contamination design is less exact than lightning-impulse coordination — the flashover process is strongly dependent on surface condition and environment.

Section 18

The Role of Configuration and Shape

Configuration

The same insulator unit performs differently as a vertical I-string, V-string, horizontal string, post insulator or bushing. A V-string can outperform a vertical I-string under some test conditions — through better natural washing, lower accumulation, a different wetting pattern and improved drainage. Contamination performance is therefore not only a function of creepage distance; the physical arrangement matters too.

Shape

Insulator shape affects wetting, drying, the leakage path, dry-band formation, arc movement, rain washing and pollution accumulation. High-leakage or fog-type profiles may perform better in pollution — but a longer creepage path helps only if the profile lets that creepage be used effectively. Deep sheds, ribs and under-ribs increase creepage but can trap contamination if poorly designed.

Section 19

What a Good Test Method Must Achieve

An ideal contamination test should be representative (reproduce service conditions as closely as possible) and reproducible (give consistent results for the same insulator under the same conditions). Representativeness is never fully achieved because real outdoor conditions are highly variable — but reproducibility must be achieved, because it is essential for standardisation.

The standard discusses three methods from IEC 507: the salt fog method, the solid-layer method procedure A, and the solid-layer method procedure B — each representing a different contamination condition. The sections that follow describe each in turn; the table below compares them on the same points so they can be scanned side by side.

Table 4 — The three IEC 507 methods compared on the same points.

Point	Salt Fog	Solid-Layer A	Solid-Layer B
What it represents	Sea pollution (salt and wetting arrive together)	Industrial; energising a wet contaminated line (“cold switch-on”)	Normal service: an energised line that later becomes wet
Insulator preparation	Clean; no pre-layer	Pre-contaminated, then dried	Pre-contaminated (NaCl + binder), then dried
How wetting is applied	Salt fog of set salinity	Steam fog after drying	Fog / wetting after energising
When voltage is applied	With the fog, simultaneously	At maximum layer conductance	Before wetting, then kept constant
Result obtained	Withstand salinity	Withstand layer conductivity	Contamination CFO (up-and-down) or withstand SDD
Main limitation	Less like dry industrial build-up	Less like the usual energise-then-wet service condition	Regarded as the best representative of the three

Section 20

Salt Fog Method

The salt fog method was developed to represent sea contamination, where salt and moisture arrive together. Voltage and salt fog of a known salinity (kg of NaCl per m³ of water) are applied simultaneously. The procedure: apply test voltage, expose the insulator to the salt fog, continue for about one hour, and record a withstand if no flashover occurs. The salinity for which the insulator withstands three or four tests is the specified withstand salinity.

Originally the goal was to find the salinity an insulator could withstand at a given system voltage; in many modern applications the test instead determines the withstand voltage at a specified salinity. The method is most relevant when contamination and wetting occur together (coastal / marine) and is less representative of industrial pollution that accumulates dry and only becomes dangerous when wetted later.

Section 21

Solid-Layer Method — Procedure A

Procedure A represents industrial pollution. A contamination layer is applied, dried, then wetted, and the voltage is applied when the layer conductance reaches its maximum — sometimes interpreted as energising a contaminated wet line (a “cold switch-on”). The typical steps: apply a uniform layer; dry to ambient; wet with steam fog; wait for maximum layer conductance; apply the test voltage; continue the fog; record withstand or flashover. The voltage is applied for 15 minutes or until flashover.

A typical IEC 507 contamination mixture is 100 g Kieselguhr + 10 g silicon dioxide + 1000 g water; alternatively 40 g kaolin per 1000 g water (rarely used). The layer is intended to create a reproducible polluted surface.

Section 22

Layer Conductivity and the Form Factor

In procedure A the degree of pollution is defined by layer conductivity, obtained by correcting the measured conductance for the insulator geometry:

\[ K_l = G\,K_f \]

\(K_l\): layer conductivity (geometry-independent)
\(G\): measured layer conductance
\(K_f\): form factor of the insulator

The measured conductance of a polluted insulator depends not only on the pollution layer but also on the shape of the insulator. A narrow region gives less surface width for the leakage current, while a wider region gives more surface width — so a large-diameter insulator and a small one carry different conductance under the same pollution. The form factor corrects the measured conductance so the result represents the pollution layer rather than the particular insulator geometry. It is defined as:

\[ K_f = \int_0^{L_f} \frac{dL}{\pi\,d(L)} \]

\(K_f\): form factor
\(L_f\): total leakage / creepage distance
\(L\): position along the leakage distance
\(d(L)\): local diameter at position \(L\)
\(\pi\,d(L)\): local circumference (surface width for current)

The term \(1/\pi d(L)\) is the inverse of the local surface width, so narrow regions raise \(K_f\). The physical basis: for a layer of resistivity \(\rho\) and thickness \(\delta\), a small surface section has resistance \(dR = \rho\,dL /(\delta\,\pi d(L))\), and integrating gives:

\[ R = \frac{\rho}{\delta}\int_0^{L_f}\frac{dL}{\pi\,d(L)} = \frac{\rho}{\delta}\,K_f \qquad\Rightarrow\qquad G = \frac{1}{R} \]

Because the measured conductance depends on both the layer property and the form factor, the layer conductivity is recovered as \(K_l = G\,K_f\). The layer conductivity for which three of four tests are withstood is the specified withstand layer conductivity.

Section 23

Solid-Layer Method — Procedure B

Procedure B (developed by IEEE) better represents the slow build-up of contamination on an energised line. The insulator is contaminated (by spraying or dipping with an NaCl-and-binder mixture), dried, cooled to ambient, then energised at constant voltage while fog or wetting is applied. This is closer to service: an energised line accumulates pollution, it later becomes wet, leakage current flows, and dry bands and arcs may form. The standard regards procedure B as the best-representative of the three. The binder may be kaolin, Tonoko or another clay; IEC 507 specifies 40 g binder per 1000 g water.

The CFO (up-and-down) method

Procedure B can determine the contamination CFO (Critical Flashover Voltage) by the up-and-down method: apply voltage at a chosen level; if flashover occurs reduce the next level, if it withstands raise it; after several tests estimate the 50% flashover voltage. This CFO method is regarded as superior to the three-out-of-four withstand method.

Withstand salt deposit density

Procedure B can also determine the specific withstand salt deposit density (mg/cm²) using the three-out-of-four withstand method — the insulator must withstand three of four tests at the selected contamination level.

Section 24

The Standard Cap-and-Pin Insulator

First, check the voltage basis

Before using any creepage or per-metre value: some requirements are given in mm/kV line-to-line and others in mm/kV line-to-ground. For one phase insulator string the physical voltage across the string is \(V_{LG} = V_{LL}/\sqrt{3}\). Mixing the two bases can create a large design error — on this page, the values are line-to-ground.

The reference unit is approximately 146 × 254 mm (146 mm spacing, 254 mm diameter) with a creepage distance of 305 mm per insulator. Its contamination performance for solid-layer procedure B is expressed as a CFO per metre of insulator length (kV rms line-to-ground per metre). If the relation is linear, the total CFO scales with string length:

\[ CFO_{\text{total}} = \left(\frac{CFO}{m}\right) \times (\text{string length in m}) \]

This linearity is what lets the required insulator length be estimated from a per-metre CFO — the basis of the design equations that follow.

Section 25

CFO-per-Metre Equations for I- and V-Strings

Which CFO is this?

Here CFO means the contamination flashover voltage under the relevant power-frequency test condition, expressed as kV rms line-to-ground per metre of connected insulator length. It is not the lightning-impulse CFO used in lightning insulation coordination — the same abbreviation is used for both, so the basis must always be checked.

For standard vertical I-strings (with contamination level \(C\) in mg/cm²):

\[ CFO\,(\text{kV/m}) = 72.3 + \frac{1.64}{C} \quad (0.02 < C < 0.04) \]

\[ CFO\,(\text{kV/m}) = 64.4 + \frac{1.96}{C} \quad (C > 0.04) \]

For V-strings:

\[ CFO\,(\text{kV/m}) = 106 + \frac{1.22}{C} \quad (0.02 < C < 0.04) \]

\[ CFO\,(\text{kV/m}) = 87.6 + \frac{1.96}{C} \quad (C > 0.04) \]

As contamination increases, CFO per metre falls. The V-string values are higher than the I-string values — V-strings perform better because they are more easily cleaned by fog or rain and accumulate less contamination, with better drainage and natural washing. This does not make V-strings always preferable: mechanical design, tower geometry, conductor movement, clearances and cost must also be considered.

Section 26

From CFO to a Withstand Voltage

The standard deviation of contamination flashover is approximately 10% of the CFO. Taking a three-standard-deviation margin gives the withstand voltage:

\[ V_i = CFO - 3\sigma,\quad \sigma = 0.1\,CFO \;\;\Rightarrow\;\; V_i = CFO - 0.3\,CFO = 0.70\,CFO \]

\(V_i\): contamination withstand voltage
\(\sigma\): standard deviation of the flashover voltage (≈ 10% of CFO)

So \(V_i \approx 0.70\,CFO\), corresponding to a low flashover probability of about 0.135% — equivalent to a three-standard-deviation margin.

The 0.70 factor is an assumption, not a constant

0.70 is not a universal physical constant. It follows from the assumed statistical spread: if the standard deviation is 10% of CFO, then \(CFO - 3(0.1\,CFO) = 0.70\,CFO\). With a different standard deviation or margin, the factor changes — so do not apply 0.70 blindly to all insulators and all tests.

Section 27

Worked Example — A 230 kV System

Take a maximum system voltage \(V_{LL,\max}=242\ \text{kV}\), a vertical string, and \(C = 0.05\ \text{mg/cm}^2\). The line-to-ground voltage is:

\[ V_{LG} = \frac{242}{\sqrt{3}} = 139.7\ \text{kV} \]

Using the vertical-string equation for \(C>0.04\):

\[ CFO\,(\text{kV/m}) = 64.4 + \frac{1.96}{0.05} = 64.4 + 39.2 = 103.6\ \text{kV/m} \]

For a six-insulator string of 146 mm spacing, length \(= 6 \times 0.146 = 0.876\ \text{m}\):

\[ CFO_{\text{total}} = 103.6 \times 0.876 = 90.8\ \text{kV} \;\;\Rightarrow\;\; V_i = 0.70 \times 90.8 = 63.6\ \text{kV} \]

Why contamination drives the length

The withstand \(V_i = 63.6\ \text{kV}\) is well below the operating \(V_{LG} = 139.7\ \text{kV}\) — so six insulators are not sufficient for this pollution. Contamination can require a far longer string than lightning or switching impulse withstand would.

Section 28

Withstand Specific Creepage Distance

Specific creepage distance is the creepage per unit voltage, in mm/kV — best interpreted here as mm/kV rms line-to-ground, because the voltage across a phase insulator is the line-to-ground voltage:

\[ L_s = \frac{\text{creepage distance}}{\text{voltage}} \qquad \left[\text{mm/kV (line-to-ground)}\right] \]

For solid-layer procedure B, the required specific creepage rises with severity as a power law of SDD:

\[ L_s = A\,(SDD)^{b} \]

\(L_s\): required withstand specific creepage distance (mm/kV)
\(SDD\): salt deposit density (mg/cm²)
\(A,\ b\): constants depending on the data source and insulator configuration

Table 5 — Example constants for cap-and-pin insulators (procedure B).

Source / Config	SDD range (mg/cm²)	\(A\)	\(b\)
IEEE, I-strings	0.02 – 0.1	86.5	0.374
IEEE, I-strings	0.1 – 0.3	51.4	0.158
CIGRE, I-strings	0.02 – 0.4	66	0.223

Different sources give different curves — reflecting uncertainty and differences in test method. The other two methods use the same power-law form: for procedure A, \(L_s = A\,(K_l)^b\) with layer conductivity \(K_l\) in µS (e.g. \(A=14.2\)); for the salt fog method, \(L_s = A\,(S)^b\) with salinity \(S\) in kg/m³. In every case, higher severity demands more creepage.

Section 29

Station Insulation and the Diameter Effect

Contamination affects not only line insulators but station insulation — post insulators, bus supports, transformer and breaker bushings, disconnector supports and instrument-transformer bushings. In substations, contamination may dictate the insulation length, sometimes forcing a higher BIL class simply because a longer insulator is needed for sufficient creepage.

For post insulators and bushings, the average diameter \(D_A\) matters: smaller average diameters give higher insulation strength, because leakage-current distribution, drying and arc development depend on surface geometry. The specific creepage again follows \(L_s = A\,(SDD)^b\), with \(A\) and \(b\) depending on the average diameter.

Table 6 — IEC 815 creepage increase for large diameters.

Average Diameter	Creepage Increase
300 – 500 mm	≈ 10%
> 500 mm	≈ 20%

Large-diameter insulators may have poorer contamination performance for the same nominal creepage, so the creepage is increased to compensate.

Section 30

Further Important Results

Linearity with string length

Many equations assume contamination withstand grows linearly with string length (\(CFO_{\text{total}} \propto\) length). This is usually practical, but for very long strings — high-voltage or UHV — some tests show possible saturation, where extra length gives less benefit than expected. As a practical correction, calculated lengths may be increased by 2–6% for 765/800 kV lines and by about 10% for UHV.

Creepage is important but not sufficient alone

Two insulators with the same creepage can perform differently because of shed profile, diameter, under-rib design, material, surface hydrophobicity, washing behaviour and dry-band behaviour. Specifying creepage alone may not fully define performance.

Contamination uniformity

Natural contamination is usually non-uniform — rain washes upper surfaces more effectively, so the lower side is often more severe. Laboratory tests often coat uniformly, which can be conservative (more severe than natural), giving a lower withstand specific creepage.

Natural versus laboratory tests, and binder effects

Natural contamination tests often show strength equal to or greater than fog-chamber tests, so laboratory tests can be conservative — yet they remain necessary because they are controlled and reproducible. Finally, the binder amount matters: relative to a reference, less binder may raise the CFO and more binder may lower it, so the exact test mixture must be standardised carefully.

Section 31

Engineering Interpretation

Contamination flashover is not mainly an impulse phenomenon — it is a surface leakage-current and dry-band arcing process requiring pollution, moisture, leakage current, dry-band formation and arc extension. Contamination design is therefore tied to surface performance, not only air clearance.

Contamination may control the design when the site is near the sea or industrial plants, where road salt is present, where pollution accumulates, where fog or mist is common, where rainfall is insufficient to wash the insulators, where equipment is outdoors, and where the system voltage is high enough to sustain leakage-current arcs. In such cases the required creepage may exceed the value set by lightning or switching impulse.

Section 32

Design Countermeasures for Severe Pollution

Where contamination is severe, the options include: increasing string length; using high-leakage-distance or fog-type insulators; using non-ceramic (silicone rubber) insulators; applying RTV silicone coatings; using resistance-glazed insulators; periodic washing; applying silicone grease; selecting better shed profiles; and improving station layout to reduce pollution accumulation. The choice depends on site severity, voltage level, maintenance philosophy, cost and the required reliability.

Section 33

A Simple Design Logic

Contamination design steps

identify the contamination severity;
express it via ESDD, SDD, layer conductivity or salinity;
select the applicable test method or standard curve;
determine the required specific creepage \(L_s\);
compute the total required creepage and choose insulator type and number;
check mechanical, impulse, switching and power-frequency requirements;
apply corrections for diameter, length, material and configuration, and consider maintenance and ageing.

The total required creepage follows from the specific creepage and the operating voltage:

\[ L_{\text{total}} = L_s \times V_{LG} \]

\(L_{\text{total}}\): total required creepage distance
\(L_s\): required specific creepage distance (mm/kV)
\(V_{LG}\): maximum system line-to-ground rms voltage

Watch out — voltage basis

Some requirements are stated in mm/kV line-to-line, others in mm/kV line-to-ground. For one phase insulator string the actual voltage is \(V_{LG} = V_{LL}/\sqrt{3}\), so confusing the two bases can cause a large error. For physical understanding of a single string, the line-to-ground basis is the directly relevant one.

Section 34

Summary of Key Equations

Equation Summary

Deterministic rule

\( \text{Min Strength} \ge \text{Max Stress} \)

Line-to-ground voltage

\( V_{LG} = \dfrac{V_{LL}}{\sqrt{3}} \)

Leakage-current heating

\( P = I^2 R \)

Layer conductivity from ESDD

\( \mu\text{S} \approx 100 \times ESDD \)

Salt-fog salinity from ESDD

\( \text{kg/m}^3 \approx 140 \times ESDD \)

Form factor

\( K_f = \displaystyle\int_0^{L_f}\dfrac{dL}{\pi\,d(L)} \)

Layer conductivity

\( K_l = G\,K_f \)

CFO/m, vertical I-string

\( CFO = 64.4 + \dfrac{1.96}{C}\;\;(C>0.04) \)

Withstand from CFO

\( V_i = 0.70\,CFO \)

Specific creepage

\( L_s = A\,(SDD)^{b} \)

Total creepage

\( L_{\text{total}} = L_s\,V_{LG} \)

Section 35

Final Understanding

Contamination performance is a critical part of outdoor insulation design. Unlike lightning or switching impulse design, contamination flashover occurs along the surface under wet, polluted conditions: pollution deposits, moisture wets the layer, leakage current flows, local heating forms dry bands, dry-band arcs ignite and extend, and the arcs finally bridge the insulator.

The most important design parameter is creepage distance — but creepage alone is not enough; shape, material, configuration, diameter, wetting pattern and pollution type all influence performance. Severity is expressed through SDD, ESDD, layer conductivity or salt-fog salinity, and the design is usually deterministic: Minimum Strength ≥ Maximum Stress.

Reader should remember

In contaminated environments, the insulation length required for pollution performance may exceed the length required for lightning or switching impulse withstand. Contamination must therefore be checked as a separate insulation-design criterion.

Part One has established the physical mechanism and the basic quantities used to describe contamination. Part Two puts these quantities to work in design calculations: the equivalence between test methods, deterministic sizing, the probabilistic interpretation, station BIL selection, altitude correction, and the influence of contamination on lightning-impulse, switching-impulse and temporary-overvoltage withstand.

This is Part One of a three-part self-study series on the contamination of outdoor insulation, covering the mechanism, severity assessment, ceramic insulators and test methods. Part Three continues with composite insulators, RTV coatings, ageing, icing and bird-related flashovers.

Section 36

Glossary

Table 7 — Key terms used on this page.

Term	Meaning
Contamination / pollution	Conductive or semi-conductive deposit on an outdoor insulator surface.
Creepage / leakage distance	Distance along the insulator surface between conductive parts (not the air gap).
Dry band	A narrow dried region of high resistance formed by local leakage-current heating.
Dry-band arcing	Arcing across dry bands that can elongate and bridge the insulator.
SDD	Salt Deposit Density — salt mass per surface area (mg/cm²).
ESDD	Equivalent Salt Deposit Density — NaCl giving the same conductivity as the actual pollutant.
Layer conductivity	Conductivity of the wet pollution layer (µS), geometry-corrected via the form factor.
Salt-fog salinity	Salt per volume of water (kg/m³) used to set sea-contamination severity.
Form factor (\(K_f\))	Geometry factor \(\int dL/\pi d(L)\) that removes insulator shape from conductance.
CFO	Critical Flashover Voltage — 50% flashover probability for the contamination condition.
Specific creepage (\(L_s\))	Required creepage per unit voltage (mm/kV), usually line-to-ground here.
Hydrophobicity	Surface property resisting a continuous water film; gives non-ceramic insulators good pollution performance.

Contamination of Outdoor Insulation