Short Notes On PEC
Short Notes On PEC
PEC scope limited to ferromagnetic materials. E.g. - carbon steel and low alloyed steel.
Ferromagnetic material is material with magnetic permeability of µr»1
PEC method based on Lenz’s law of electromagnetic induction
Lenz law- an electrically conductive object when subjected to change in magnetic field will produce
eddy current that oppose change in magnetic field.
‘Lift-off’- Distance between probe and steel surface.
After primary field has been switched off, eddy current decay and at same time diffuse into the steel.
Change in sec. field inside receiver coil induces voltage over its terminal, this voltage as function of
time is called ‘PEC signal’.
Point between straight section and curved section in PEC signal is called ‘transition point’. Occurs
when PEC signal reaches back wall of steel.
‘Amplitude’(A)- signal strength at the beginning of PEC signal
‘Slope’ (n) - rate of decay at the straight section. Slope parameter n≈1.5; n>1.5
‘Transition Time (τ) – onset of fast decay.
PEC signal is displayed on log-log graph.
𝑊𝑇 = √𝜏 ÷ 𝜎𝜇, 𝜎- electrical conductivity, 𝜇- permeability of steel is electromagnetic properties of
steel.
The velocity of diffusion is higher for a thin wall than for a thick wall.
If the wall thickness at reference spot is unknown the readings can be presented as percentage relative
to reference measurement and are ‘relative wall thickness reading.’
If the wall thickness at reference spot is known then it can be expressed in mm (inches) and wall
readings are ‘absolute.’
Electromagnet properties may vary across test specimen if one part of test sample has received heat
treatment different than other (e.g. furnace tube).
Variation of electromagnetic properties limits accuracy of PEC, to typically ±10% of wall thickness.
PEC does not show internal or external wall loss, it is only sensitive only to wall thickness.
Time scale is expressed in milliseconds (ms).
Limiter line method- for the determination of transition time 𝜏 ref
Starting point at normalisation time is 30% below the reference signal, with amplitude of 0.7V ref.
Normalisation means multiplying the PEC signal with a factor.
The strength of the primary field is largest when the PEC sensor is closest to the surface.
Lower lift-off- higher amplitude and vice versa.
Real PEC signals are lift-off dependence.
Degree of lift-off dependence varies with wall thickness and geometry.
Typical dependence varies from 0.1% decrease to 0.25% increase in PEC reading per mm lift-off.
PEC readings vary strongly with distance.
Insulation covers
Material between lift-off can affect PEC signals if they are electrically conductive or magnetic.
Aluminium covers
Electrical conductivity of aluminium 5 times higher than steel. Therefore strong contribution of
aluminium covers.
Aluminium covers results in early tail in time in the PEC signal.
Strength of early tail depends on the thickness of insulation cover.
If less than 0.2mm it will hardly influence
If 1mm thick then will influence signal.
The effect is stronger where there is an insulation overlap.
Steel reinforcement bars
Inside concrete, have same effect as aluminium covers.
Strength of early tail depends on the density of the reinforcement mesh.
For both aluminium covers and steel bars, the normalisation time should be chosen after the early tail.
For Al covers, the norm time should be chosen when effect of Al has become weaker and signal is
straight again. It is after 6-8ms.
Prolonged and strong tail- for thick Al sheet over lap and dense reinforcement.
S.S covers
Electrical conductivity of Stainless steel 5 times less than steel. Are weakly magnetic.
No noticeable effect on PEC signal.
Galvanised insulation sheeting
Made of carbon steel.
Magnetic and good electrical conductor.
Distorts PEC signal
Chicken wire mesh
No influence to signal if made of S.S.
No influence if made of carbon steel but if it is properly fixed that it does not move.
Influence signal if magnetic and free to move i.e. loose.
Corrosion product
Ferrous oxide (Fe2O3) - highly magnetic but do not conduct electric current.
Magnetised by primary field but no eddy currents are generated.
Instead field decays very slowly
Corrosion product results tail very late in time.
Fe3+ (Fe2O3) are magnetic, Fe2+ (FeS) are non magnetic.
Corrosion product thinner than 5mm can be ignored.
Stand off – distance between bottom of PEC probe and corrosion product.
Used for suppressing the influence of corrosion product by placing plastic shims in between.
Contribution of corrosion product strongly reduced by higher stand offs (at least 20mm).
Insulation components ensures this stand offs.
Such case observed when inspecting annular ring of tank floor from outside.
Corrosion product on far side of the steel has no effect on the signal
PEC signal are sensitive to corrosion thickness layer above 5mm, ensure a standoff of at least 20mm or
if the insulation thickness exceeds 40mm.
ADC sample time at least 600ms if corrosion product more than 5mm and lift-off less than 40mm.
Clipping or Saturation
ADC has max input amplitude.
ADC output is max for a signal with max signal input amplitude.
ADC output will not increase if signal input increased above the max output. It will remain max
This is referred as clipping.
PEC signal cannot be used before clipping time.
Gain of amplified signal inside probe adjusts such clipping occurs around 1.5ms
High frequency noise
Always superimposed on PEC signal.
This noise comes from amplifiers, environment and steel itself.
High frequency noise can be reduced by HF filters.
50/60 pick up
PEC probes picks up mains-related signals.
This pick comes from electromagnetic fields from power cables, electric motors of pumps or other
electrical equipment.
Frequency 50Hz in Europe and 60Hz in USA.
Pick up is often accompanied in harmonics of 100, 150, 200Hz etc, for 50Hz and 120Hz, 180Hz,etc for
60Hz
Filter used in software that uses frequency analysis (Fourier filter) to remove 50/60 and higher
harmonics.
Fourier filter suppress mains related noise provided it is symmetrical.
Probe movement
Movement causes erratic signal that is superimposed on the PEC signal.
PEC signal verified for signal distortion if PEC measurement differ by more than ±10% from the
average over the other PEC readings.
Criteria for an acceptable PEC signal
A straight and a curved section after smoothing and filtering.
Amplitude of distortion is less than the signal amplitude at intersection point.
Amplitude of PEC signal at norm point is within a factor 4 of the amplitude of that of the reference
signal.
Repeatability of signal should be preferably within 1%.
Footprint
Twice of the insulation thickness and steel thickness.
PEC readings are averages over footprint.
Time of onset of fast decay, i.e. the transition time, is measure of thickness of test specimen.
PEC wall thickness readings slightly affected by lift-off. Requires lift-off correction.
PEC footprint size increases with lift-off.
The electromagnetic properties also depend on steel grade and can vary from one batch of steel to next.
PEC measurement also depends on temperature of steel, because electromagnetic properties σ & µ also
depends on temperature.
Wall thickness reading decrease by 0.1% per ˚C. This can be ignored for wall thickness mapping but
need to be corrected for corrosion mapping.
Deviant readings at insulation overlaps, separately calibrate insulation overlaps with normalisation late
in time.
.
Corrosion product
Corrosion i.e. iron oxide, poor electrical conductor.
Does not conduct eddy current.
Contributes to signal if between probe and steel, due to magnetic permeability of rust.
Effect is a ‘tail’ on the PEC signal.
Apparent transition point increased due to it.
This results in increase in PEC wall thickness reading.
Magnitude of increase depends on:
a) Thickness of corrosion product.
b) Distance of probe to corrosion product.
Sampling Time- PEC signal sampled during a pre set time
Sampling frequency- ADC records PEC signal at a pre set frequency, typically 20Hz, i.e. one ADC
sample every 20µs(micro second)
Measurement Repeats- referred to as current pulses. Each repeat results in one PEC signal, which can
be added to produce one PEC signal, which is sent to computer.
Cycle Delay- time between end of one cycle and beginning of next cycle.
PEC signal from trailing edge (when the current is off) is recorded. Signal from trailing
discarded as its not stable though
INSTRUMENT PROBES
Four standard PEC probes:
1. CP (contact probe)
2. AP-50 (probe with coils without ferrite, ‘Air probe’)
3. FP-50 (probe with ferrite inside the coils)
4. TB (for tank bottom inspection)
Substantial piece of steel, such as support or nozzle, affects PEC readings if PEC probe is closer than
twice lift-off to the inspection plate.
Notes:
1. The criteria for an acceptable PEC signal are:
a) A straight and a curved section of the PEC signal can be observed after filtering and
Smoothing;
b) No signal clipping after 3ms;
c) The amplitude of any distortions of the PEC signal is less than the signal amplitude at the
Intersection point;
d) The amplitude of the PEC signal at the normalization point is within a factor 4 of the amplitude
Of that of the reference signal, this is automatically verified by the PEC software.