Insulation Coordination

Contamination of Outdoor Insulation: Coordination & Altitude

Part Two of three. Turning pollution severity into a real insulator length: the approximate equivalence between ESDD, salinity and layer conductivity, the deterministic CFO-per-metre and specific-creepage routes, how contamination can raise the required station BIL, the probabilistic method, altitude correction of creepage distance, and how contamination reduces lightning-impulse, switching-impulse and temporary-overvoltage withstand — with the IEC 815 creepage guidance.

Reading time ≈ 40 min · Part Two of Three

Section 1

Relating the Three Contamination Test Methods

Part One built the mechanism of contamination flashover and the three IEC 507 test methods. This second part turns pollution severity into a practical insulator length — through the equivalence between methods, the deterministic and probabilistic design routes, station BIL selection, altitude correction, and the effect of contamination on impulse and temporary-overvoltage withstand.

Continuing from Part One

Part One explained the physical mechanism of contamination flashover and introduced the main quantities used to describe pollution severity — SDD, ESDD, layer conductivity and salt-fog salinity. This second part uses those quantities for insulation coordination: how to select the insulator length, the creepage distance, the number of units, the station insulation class, and the correction factors.

Each test method describes severity with a different parameter: the salt fog method uses salinity in kg/m³, solid-layer procedure B uses SDD or ESDD in mg/cm², and solid-layer procedure A uses layer conductivity in µS. In theory they should relate to one another, because all three describe the same physical question — how conductive the wet polluted surface becomes. In practice the equivalence is approximate, because each method reproduces a different contamination and wetting process.

What Part Two covers

the approximate equivalence between ESDD, salinity and layer conductivity;
the deterministic design — CFO/m and specific-creepage routes;
how contamination can raise the required station BIL;
the probabilistic method and why deterministic design is conservative;
altitude correction of the required creepage distance;
how contamination reduces LI, SI and TOV withstand, and the IEC creepage guidance.

Section 2

Approximate Equivalence Between the Severity Measures

Different standards and laboratory methods express contamination severity using different quantities. The equivalence relationship is useful because it lets test data from one method be compared with another — but it should be treated as an approximate engineering correlation, not an exact physical conversion. Based on CIGRE reference values, the approximate correlation is:

\[ 1\ \text{mg/cm}^2\ ESDD \;\;\equiv\;\; 140\ \text{kg/m}^3\ \text{salinity} \;\;\equiv\;\; 100\ \mu\text{S}\ \text{layer conductivity} \]

Table 1 — Which severity quantity belongs to which test method.

Method	Severity Quantity	Unit	Typical Representation
Solid-layer Procedure B	SDD / ESDD	mg/cm²	Deposited salt on the insulator surface
Solid-layer Procedure A	Layer conductivity	µS	Conductivity of the wetted pollution layer
Salt-fog method	Salt salinity	kg/m³	Salt concentration in the applied fog

Equivalently, \(\text{salinity} \approx 140 \times ESDD\) and \(\text{layer conductivity} \approx 100 \times ESDD\) (ESDD in mg/cm²). For example, with \(ESDD = 0.05\ \text{mg/cm}^2\):

\[ \text{salinity} = 140 \times 0.05 = 7\ \text{kg/m}^3 \qquad\qquad \text{layer conductivity} = 100 \times 0.05 = 5\ \mu\text{S} \]

So \(0.05\ \text{mg/cm}^2\) ESDD corresponds to about \(7\ \text{kg/m}^3\) salinity and \(5\ \mu\text{S}\) layer conductivity — useful when comparing data from different methods.

Section 3

A Warning About the Equivalence

The conversion is not a perfect physical equivalence, because the methods do not reproduce the same process. In the salt fog method, contaminant and wetting arrive together (representative of sea pollution). In solid-layer procedure A, the insulator is contaminated, dried, wetted to maximum conductance, then energised. In solid-layer procedure B, it is contaminated, dried, energised, then wetted — the closest to service, where an energised line or station insulator becomes wet after pollution has accumulated.

Use with care

The same numerical severity may not produce exactly the same flashover behaviour in all three methods. Treat the equivalence as a comparison aid, not an exact conversion.

Section 4

Purpose of Contamination Insulation Coordination

The goal is to select the insulator type, string length, number of units, required creepage distance, the BIL class for station insulation, and any countermeasures — so the insulation withstands the maximum expected contamination stress in service. Two design methods exist, deterministic and probabilistic; in practice the deterministic method dominates because contamination behaviour is so uncertain.

Contamination flashover depends on the actual pollution level, wind direction, wetting rate, fog duration, rain washing, surface ageing, non-uniform pollution, insulator shape, dry-band development and material hydrophobicity. Because of these uncertainties the design follows the conservative rule:

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

The insulation is chosen so the lowest expected withstand strength still exceeds the highest expected service stress.

Section 5

The Maximum Stress — Line-to-Ground Voltage

For contamination, the maximum stress is the maximum continuous line-to-ground operating voltage, because one insulator string is connected between the phase conductor and earth or tower:

\[ V_{LG,\max} = \frac{V_{LL,\max}}{\sqrt{3}} \]

\(V_{LG,\max}\): maximum rms line-to-ground voltage
\(V_{LL,\max}\): maximum rms line-to-line system voltage

The actual voltage across the string is the line-to-ground voltage — which is why the voltage basis (line-to-line vs line-to-ground) must always be tracked carefully.

Section 6

The Minimum Strength — A Withstand from CFO

Minimum strength may be defined from contamination CFO/m, a withstand voltage, a withstand specific creepage, test data or service experience. A common conservative approach starts from CFO and subtracts three standard deviations (with \(\sigma \approx 10\%\) of CFO):

\[ V_3 = CFO - 3\sigma,\quad \sigma = 0.1\,CFO \;\;\Rightarrow\;\; V_3 = 0.7\,CFO \]

\(V_3\): low-probability contamination withstand voltage
\(\sigma\): standard deviation of flashover voltage (≈ 10% of CFO)

0.7 is a statistical assumption

The value \(V_3 = 0.7\,CFO\) is obtained only when the standard deviation is assumed to be 10% of CFO and the withstand is taken as CFO minus three standard deviations. Treat it as a statistical design assumption, not a fixed property of all insulators — the same 0.7 factor reappears in the worked example below.

Section 7

Two Deterministic Design Routes

The deterministic method is normally used because it is simple and conservative. The maximum expected contamination level is selected first; then the required insulation is calculated either from the contamination CFO per metre or from the withstand specific creepage distance. Both routes should give similar results when they are based on the same test data.

The two routes, step by step

Route 1 — CFO per metre: CFO/m → withstand voltage per metre (\(V_3\)) → required string length → number of insulators.
Route 2 — specific creepage: required specific creepage \(L_s\) → total creepage distance → number of insulators → string length.

Section 8

Route 1 — CFO per Metre

Select the maximum contamination level, find the CFO/m, take the withstand \(V_3 = 0.7\,CFO\), then size the string and count units:

\[ L_{\text{string}} = \frac{V_{LG,\max}}{V_3} \qquad\qquad N = \frac{L_{\text{string}}}{s} \]

\(N\): number of insulator units
\(L_{\text{string}}\): required connected string length
\(s\): insulator spacing per unit (0.146 m for a standard cap-and-pin unit)

Section 9

Route 2 — Specific Creepage Distance

Select the maximum contamination level, find the required specific creepage \(L_s\) (mm/kV), then compute the total creepage and count units:

\[ L_{\text{creep,total}} = L_s\,V_{LG,\max} \qquad\qquad N = \frac{L_{\text{creep,total}}}{L_{\text{creep,unit}}} \]

\(L_{\text{creep,total}}\): total required creepage distance
\(L_s\): withstand specific creepage distance (mm/kV)
\(L_{\text{creep,unit}}\): creepage per unit (305 mm for the standard unit)

The physical string length is then \(L_{\text{string}} = N\,s\), with \(s = 0.146\ \text{m}\) for the standard unit.

Section 10

Worked Example — 230 kV V-String, Both Routes

Use the line-to-ground voltage

For one insulator string connected between phase and tower/earthed structure, the applied voltage is the maximum line-to-ground voltage, not line-to-line. So \(V_{LG,\max} = V_{LL,\max}/\sqrt{3}\) must be used — unless the standard value being used has already been expressed on a line-to-line basis.

Take \(V_{LL,\max} = 242\ \text{kV}\), a standard 146 × 254 mm V-string, and \(C = 0.10\ \text{mg/cm}^2\). First the line-to-ground voltage:

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

Route 1 — CFO/m. The V-string equation gives \(CFO = 107.2\ \text{kV/m}\), so:

\[ V_3 = 0.7 \times 107.2 = 75.0\ \text{kV/m} \;\;\Rightarrow\;\; L_{\text{string}} = \frac{139.7}{75.0} = 1.86\ \text{m} \;\;\Rightarrow\;\; N = \frac{1.86}{0.146} = 12.8 \to 13 \]

Route 2 — specific creepage. The required \(L_s = 28.0\ \text{mm/kV}\), so:

\[ L_{\text{creep,total}} = 139.7 \times 28.0 = 3912\ \text{mm} \;\;\Rightarrow\;\; N = \frac{3912}{305} = 12.8 \to 13 \]

The two routes agree

Both give \(N = 13\) insulators (string length \(13 \times 0.146 = 1.90\ \text{m}\)). The important point is not only the answer of 13 units — it is that two independent routes give the same result when the underlying test data are consistent, which gives confidence in the selected string length. The specific-creepage route is often more convenient, because many standards give pollution requirements directly in mm/kV.

Section 11

Station Insulation and the BIL It Forces

For post insulators and bushings the process is slightly different: the required creepage may force a physically longer insulator, and a longer insulator usually carries a higher impulse withstand rating. So contamination can indirectly raise the required BIL class of station insulation — even when lightning or switching impulse alone would not require it.

Why does a power-frequency problem change the BIL?

Contamination does not directly increase the lightning-impulse stress. But it may require a longer post insulator to provide enough creepage distance, and standard post insulators with longer creepage usually correspond to higher BIL classes. So contamination indirectly forces the selection of a higher-BIL unit — the BIL rises as a side effect of needing more creepage, not because the impulse stress changed.

Section 12

Worked Example — A 230 kV Post Insulator

Take the same system (\(V_{LG,\max} = 139.7\ \text{kV}\), \(C = 0.10\ \text{mg/cm}^2\)) and an average post-insulator diameter \(D_A = 200\ \text{mm}\). The station-insulator equation gives a required total creepage of about \(L_{\text{creep,total}} = 5.3\ \text{m}\). Checking standard post-insulator classes, either a 1425 kV BIL Class I or a 1050 kV BIL Class II unit can satisfy that creepage. For a 230 kV system the usual BIL would be 900 or 1050 kV, so the practical selection is 1050 kV BIL, Class II.

Key lesson

The BIL may be selected not because impulse stress requires it, but because contamination requires a longer creepage. A designer might think 900 kV BIL is enough for 230 kV on impulse grounds — yet pollution forces the longer 1050 kV unit. Contamination can govern the physical insulation selection.

Section 13

The Probabilistic Method — Concept

The probabilistic method asks a different question from the deterministic method. Instead of assuming maximum contamination and minimum strength occur together, it estimates how often a severe contamination condition and a weak insulation condition actually coincide. This can give a much lower expected flashover rate — but it requires reliable statistical data.

Formally, it mirrors switching-surge coordination: it combines a stress distribution (how wetted contamination severity varies over time) with a strength distribution (the withstand capability as a function of contamination level). The flashover probability is obtained by convolving the two:

\[ P_{\text{flashover}} = \int P_{\text{flashover}\,|\,C}\;\cdot\;f_{\text{contamination}}(C)\;dC \]

\(P_{\text{flashover}}\): total probability of flashover
\(P_{\text{flashover}\,|\,C}\): conditional flashover probability at contamination level \(C\)
\(f_{\text{contamination}}(C)\): probability density of the wetted contamination level
\(C\): contamination severity (e.g. mg/cm²)

Section 14

The Stress Distribution

The stress distribution describes how the wetted contamination level varies through the year, often approximated as Gaussian. If the maximum level \(0.10\ \text{mg/cm}^2\) sits at \(\mu_c + 3\sigma_c\), with the standard deviation as large as about 150% of the mean, the spread shows how highly variable contamination severity is.

Section 15

The Strength Distribution and Reformulating the CFO

Test data usually give flashover probability versus voltage for several contamination levels, but the probabilistic method needs flashover probability versus contamination level at the operating voltage. Reading a vertical line at the operating voltage gives that. The strength characteristic can also be obtained by rearranging a CFO equation — for the V-string form:

\[ V = 87.6 + \frac{1.96}{C} \;\;\Rightarrow\;\; C = \frac{1.96}{V - 87.6} \]

This gives the contamination level corresponding to a given voltage. At the operating voltage \(V = 139.7\ \text{kV}\) (a 0.50 flashover probability) the level is \(C = 0.038\ \text{mg/cm}^2\); for a lower probability of 0.023 the voltage is taken as \(139.7 \times 0.9\), giving \(C = 0.0514\ \text{mg/cm}^2\). Repeating builds the full probability-versus-contamination relationship.

Section 16

Approximating the Strength Distribution

The full strength characteristic cannot easily be captured by one simple continuous distribution. But in the low-probability region — especially below \(P \approx 0.20\) — it can be approximated by a cumulative Gaussian with a standard deviation around 18% to 22% of the mean. That approximation is what makes the probabilistic calculation tractable.

Section 17

Contamination Flashover Rate — Example

Take a minimum strength of \(0.10\ \text{mg/cm}^2\) located at \(\mu_s - 3\sigma_s\) with \(\sigma_s/\mu_s = 0.18\), giving a mean strength \(\mu_s = 0.2174\ \text{mg/cm}^2\). After accounting for the number of towers and insulators, the adjusted mean becomes \(\mu_{sn} = 0.1065\ \text{mg/cm}^2\), and the contamination flashover rate is:

\[ CFOR = 1 - F\!\left[\frac{0.1065 - 0.01818}{0.02727}\right] = 0.000600 \]

Why the deterministic method is conservative

That is about 6 flashovers per 10,000 years. The maximum-contamination and minimum-strength conditions are each rare, so the probability of both occurring together is extremely small. This is why probabilistic methods can sometimes justify shorter strings.

The very low calculated rate does not mean the deterministic method is wrong. It means the deterministic method intentionally combines severe assumptions to give a conservative design — which is exactly what is wanted where contamination data are uncertain or where service reliability is critical. That is why deterministic design remains in wide use.

Section 18

Why Different Methods Give Different Numbers

Different sources give significantly different insulator counts, because they use different test methods, assumptions, insulator types, severity definitions, safety margins and experience-based corrections. For a 550 kV system under light contamination, the number of standard vertical-string units can range from about 27 to 37 (or 27–31 if one outlier source is discounted); under medium contamination, 35 to 49. These are large differences.

Table 2 — IEEE service-experience minimum unit counts (low contamination).

System / Configuration	Units
345 kV, I-string (minimum)	15
345 kV (more standard value)	18
500 kV, V-string (minimum)	22
765 kV, V-string (minimum)	30

These service values were consistent with IEEE-based calculations at low contamination. The practical conclusion: utility service experience is essential — test data should guide the final number of insulators, not dictate it.

Section 19

Effect of Altitude on Strength

Air density falls with altitude, and lower air density reduces the flashover strength of external insulation — including contamination performance. An insulator adequate at sea level may have a lower withstand at high altitude. The CFO is corrected by:

\[ CFO_A = CFO\,\delta^{m} \]

\(CFO_A\): CFO at altitude
\(CFO\): CFO at sea level
\(\delta\): relative air-density factor (\(\delta < 1\) at altitude)
\(m\): exponent: 0.5 for standard insulators, 0.8 for fog-type

Since \(\delta < 1\), \(\delta^m < 1\), so \(CFO_A < CFO\): the withstand is lower at altitude, and the required creepage must increase to compensate. The exponent \(m\) represents how sensitive the particular insulation arrangement is to the air-density reduction — different insulator types have different values (0.5 for standard, 0.8 for fog-type are typical), so the correct value should be taken from the applicable standard or test guidance rather than assumed universal.

Section 20

Altitude Correction of Specific Creepage

The required specific creepage is scaled up by the same factor:

\[ L_{s,A} = \frac{L_s}{\delta^{m}} \qquad (\delta^m < 1 \Rightarrow L_{s,A} > L_s) \]

\(L_{s,A}\): required specific creepage at altitude
\(L_s\): sea-level specific creepage

Example at \(A = 2000\ \text{m}\), where \(\delta = 0.8174\), starting from \(L_s = 20\ \text{mm/kV}\):

\[ \text{standard: } \delta^{0.5} = 0.904,\;\; L_{s,A} = \frac{20}{0.904} = 22.1\ \text{mm/kV} \quad(+10.5\%) \]

\[ \text{fog-type: } \delta^{0.8} = 0.851,\;\; L_{s,A} = \frac{20}{0.851} = 23.5\ \text{mm/kV} \quad(+17.5\%) \]

The increase is moderate but should not be ignored for high-altitude substations and lines.

Section 21

Effect of Contamination on LI, SI and TOV

Contamination mainly affects power-frequency and long-duration withstand, but it can also reduce lightning-impulse (LI), switching-impulse (SI) and temporary-overvoltage (TOV) strength — most severely when the surface is polluted, leakage current is flowing, and dry bands (or dry-band arcs) have already formed. Because contamination flashover is a time-dependent surface process, the longer the voltage lasts, the more it is affected:

\[ \text{TOV reduction} \;>\; \text{SI reduction} \;>\; \text{LI reduction} \]

A TOV lasts long enough for leakage current, heating, dry-band formation and arc development; an SI is shorter, so less develops; an LI is much shorter, so the effect is usually least.

Section 22

Worst-Case Strength Reductions

Under the worst condition, where dry bands have already formed, the reductions can be severe:

Table 3 — Approximate worst-case strength reduction (dry-band stage).

Stress	Strength Reduction
Lightning impulse (LI)	≈ 20% to 30%
Switching impulse (SI)	≈ 30% to 60%
Temporary overvoltage (TOV)	Approaches the AC contamination strength

So TOV performance under contamination may fall to the normal AC contamination flashover strength.

Section 23

Why These Reductions Are Not Applied Blindly

These factors are important but should not be applied mechanically. Lightning impulse: a stroke often occurs with rain, which washes the surface and reduces severity, and the chance that the impulse coincides exactly with the dry-band stage is low — so a severe LI reduction is often too conservative. Switching impulse: surges may follow reclosing after a lightning flashover, but a contaminated insulator presents a shunt resistance that can reduce the switching overvoltage, especially under heavy pollution — so the real risk may be lower than expected.

Notably, there has been no reported flashover caused by lightning or switching surges on contaminated insulators. This does not mean contamination has no effect on LI/SI strength — it means applying severe reduction factors is questionable. The engineer must weigh the probability of transient overvoltage, of wet contamination, and of dry-band formation, plus the surge magnitude, rain washing, the shunting effect of contamination resistance, and service experience.

Practical position

Take TOV seriously (its duration allows surface flashover to develop); assess SI against probability and system conditions; treat LI reduction as a worst-case laboratory result whose service coincidence is unlikely. Contamination design is normally governed by AC / power-frequency withstand rather than LI/SI reduction.

Section 24

IEC 815 Suggested Specific Creepage Distances

IEC 815 gives suggested withstand specific creepage distances by pollution level (here in mm/kV rms line-to-ground):

Table 4 — IEC 815 specific creepage distance by pollution level.

Pollution Level	\(L_s\) (mm/kV line-to-ground)
Light	27.7
Medium	34.6
Heavy	43.3
Very heavy	53.7

In exceptional pollution even 53.7 mm/kV may not be adequate, and higher creepage, washing, greasing, coatings or alternative insulator technologies may be required.

Section 25

Typical Environments for Each Pollution Level

Table 5 — Typical environments and corresponding specific creepage.

Level	Typical Environment	\(L_s\) (mm/kV)
Light	No industry; low-density heated housing; agricultural or mountainous; frequent wind/rain; ≥ 10–20 km from the sea, no direct sea wind	27.7
Medium	Non-heavily-polluting industry; average heated housing; or dense housing/industry but frequent wind/rain; exposed to sea wind but not very close to coast	34.6
Heavy	High industrial density; suburbs of large cities with polluting heating plants; close to the sea with relatively strong sea winds	43.3
Very heavy	Conductive dust or thick industrial deposits; very close to coast with sea spray; deserts with sand, salt, strong winds and regular condensation	53.7

In very lightly polluted areas a lower value — around 20.8 mm/kV — may be justified by service experience.

Section 26

A Step-by-Step Selection Workflow

Deterministic selection workflow

System voltage — find \(V_{LL,\max}\), then \(V_{LG,\max} = V_{LL,\max}/\sqrt{3}\).
Severity — classify the site (light…exceptional) or quantify via ESDD, SDD, layer conductivity or salinity.
Method — choose the CFO/m or specific-creepage route.
Required insulation — \(L_{\text{string}} = V_{LG,\max}/V_3\) or \(L_{\text{creep,total}} = L_s V_{LG,\max}\).
Number of units — \(N = L_{\text{creep,total}}/L_{\text{creep,unit}}\) or \(L_{\text{string}}/s\).
Round up — e.g. \(N = 12.8 \to 13\).
Corrections — altitude, large diameter, profile, long strings, pollution type, service experience.
Check the rest — LI, SI, TOV withstand, mechanical loading, clearances, maintenance access.

Section 27

Summary of Key Equations

Equation Summary

Severity equivalence

\( 1\,\text{mg/cm}^2 \equiv 140\,\text{kg/m}^3 \equiv 100\,\mu\text{S} \)

Line-to-ground voltage

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

Withstand from CFO

\( V_3 = 0.7\,CFO \)

String length (CFO/m)

\( L_{\text{string}} = \dfrac{V_{LG,\max}}{V_3} \)

Units from length

\( N = \dfrac{L_{\text{string}}}{s} \)

Total creepage

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

Units from creepage

\( N = \dfrac{L_{\text{creep,total}}}{L_{\text{creep,unit}}} \)

Altitude correction (CFO)

\( CFO_A = CFO\,\delta^{m} \)

Altitude correction (creepage)

\( L_{s,A} = \dfrac{L_s}{\delta^{m}} \)

Strength reduction ranking

\( \text{TOV} > \text{SI} > \text{LI} \)

Section 28

Final Understanding

Part Two is about converting pollution severity into practical insulation requirements. The methods describe severity differently — ESDD, salinity, layer conductivity — related approximately by the CIGRE equivalence \(1\ \text{mg/cm}^2 \equiv 140\ \text{kg/m}^3 \equiv 100\ \mu\text{S}\), though this is not a perfect conversion because each method represents a different wetting process.

Design is normally deterministic (Minimum Strength ≥ Maximum Stress), with the maximum stress \(V_{LG,\max} = V_{LL,\max}/\sqrt{3}\). The insulation can be sized from CFO/m or from specific creepage — the latter often more practical, \(L_{\text{creep,total}} = L_s V_{LG,\max}\) then \(N = L_{\text{creep,total}}/L_{\text{creep,unit}}\). Altitude lowers strength, so creepage is increased by \(L_{s,A} = L_s/\delta^m\); and contamination can reduce LI, SI and TOV withstand, but most severely only once the surface has reached the dry-band stage (TOV > SI > LI).

Reader should remember

Part Two converts pollution severity into insulation design. The practical route is usually deterministic: select the maximum contamination level, determine the required creepage or CFO per metre, calculate the required number of insulators, then apply correction factors such as altitude. The probabilistic method explains why deterministic design is conservative, but it requires reliable statistical data. Contamination mainly affects AC and TOV performance; LI and SI reductions are only severe when dry bands have already developed.

Part Three continues from this design basis and considers practical improvement methods: silicone-rubber composite insulators, RTV coatings, resistive glaze, ageing tests, and the icing and bird-streamer flashover problems.

This is Part Two of a three-part self-study series on the contamination of outdoor insulation. Part One covers the mechanism, severity assessment and test methods; Part Three covers nonceramic insulators, coatings, icing and birds.

Section 29

Glossary

Table 6 — Key terms used on this page.

Term	Meaning
ESDD	Equivalent Salt Deposit Density (mg/cm²) — NaCl giving the same conductivity as the actual pollutant.
Layer conductivity	Conductivity of the wet pollution layer (µS); \(\approx 100\times ESDD\).
Salt salinity	Salt per volume of water (kg/m³); \(\approx 140\times ESDD\).
Deterministic method	Conservative design: Minimum Strength ≥ Maximum Stress.
\(V_3\)	Low-probability contamination withstand voltage, \(\approx 0.70\,CFO\).
Specific creepage (\(L_s\))	Required creepage per unit voltage (mm/kV line-to-ground).
CFOR	Contamination Flash-Over Rate (flashovers per unit time).
\(\delta\)	Relative air-density factor (< 1 at altitude).
Exponent \(m\)	Altitude exponent: 0.5 (standard insulators), 0.8 (fog-type).
BIL	Basic Lightning Impulse insulation level — can be forced upward by contamination.
LI / SI / TOV	Lightning impulse / switching impulse / temporary overvoltage.

Part Two Reading now

Coordination, Altitude & Impulse Effects

Turning severity into insulator length — the method equivalence, the deterministic CFO/m and specific-creepage routes, station BIL, the probabilistic method, altitude correction, and LI/SI/TOV effects.

Series progress 2 of 3

Part Three

Nonceramic Insulators, Coatings & Icing

Composite insulators and the hydrophobic mechanism, RTV silicone coatings, ageing and strength tests, the six improvement methods, resistive glaze, and the icing and bird-streamer problems.