Eddy current testing (ECT) offers a non-destructive method to measure remaining PFA wall thickness on installed heaters without removing them from the tank. The technique works because the metal core (Incoloy, titanium, or stainless steel) is conductive, while the PFA sheath is non-conductive. An eddy current probe placed on the PFA surface induces currents in the metal core; the phase lag of the return signal is proportional to the distance between the probe and the metal – i.e., the PFA wall thickness. Calibration with known thickness standards achieves accuracy of ±0.05–0.10 mm for PFA thicknesses of 0.5–3.0 mm. ECT detects thinning from erosion, localized wear, or chemical attack, allowing predictive replacement before perforation. The method is rapid (seconds per point) and can be performed by trained technicians without tank draining (probe inserted through a sight port or via a wet-contact technique).
Principle of Operation
An eddy current probe generates an alternating magnetic field (typical frequency 100 kHz–2 MHz). This field induces eddy currents in the metal core. The eddy currents create a secondary magnetic field that opposes the primary field. The probe measures the impedance change. The phase angle difference between the transmitted and received signal is a function of the lift-off distance – the gap between the probe and the metal. For non-magnetic metals (Incoloy, titanium, stainless steel), the lift-off effect is linear with distance for gaps of 0–3 mm. By calibrating with PFA shims of known thickness, the measured phase angle converts directly to PFA thickness.
The method works through wet PFA (the liquid does not interfere because it is non-conductive and non-magnetic). The probe can be applied to a wet heater surface in service, though best accuracy is achieved with a clean, dry surface.
Calibration Procedure
Obtain a calibration block: a piece of the same metal core (Incoloy, etc.) with a flat surface. Alternatively, use the heater's cold end (where thickness is known from manufacturing).
Place PFA shims of known thickness (0.5 mm, 1.0 mm, 1.5 mm, 2.0 mm, 2.5 mm) between the probe and the metal block.
Record phase angle for each thickness. Plot phase vs. thickness (should be linear).
Verify linearity R² > 0.99. For titanium cores, the linear range is 0.5–2.5 mm. For Incoloy, 0.5–3.0 mm.
Create lookup table or linear equation: Thickness = m × (phase) + b.
Measurement Procedure
| Step | Action |
|---|---|
| 1 | Clean the PFA surface at the measurement point (remove scale, biofilm) – a soft wipe is sufficient. |
| 2 | Apply a thin layer of water or gel couplant (to ensure acoustic contact – ECT does not require couplant, but consistent contact improves repeatability). |
| 3 | Place probe perpendicular to heater surface. For curved heaters (typical 25 mm OD), use a contoured probe tip or take multiple measurements around circumference and average. |
| 4 | Record phase angle (3–5 readings). |
| 5 | Convert to thickness using calibration curve. |
| 6 | Mark measurement locations with permanent marker to track changes over time. |
Accuracy and Limitations
| Parameter | Performance | Notes |
|---|---|---|
| Thickness range | 0.3–3.0 mm | Outside this range, non-linear |
| Accuracy | ±0.05–0.10 mm | Calibration-dependent |
| Repeatability | ±0.03 mm | On same spot |
| Minimum spot size | 5 mm diameter | Smaller spots have edge effects |
| Metal core effect | Calibration required for each alloy | Incoloy, Ti, SS have different conductivity |
| Temperature effect | Drift >60°C | Perform at room temperature (shut down and cool) |
| Curvature effect | Error up to 0.1 mm for 25 mm OD | Use curved probe or average 3 positions |
Application Examples
| Service Environment | Expected Thinning Rate | Inspection Frequency | Action when thickness < |
|---|---|---|---|
| Clean water | <0.05 mm/year | Annually | 1.0 mm (replace) |
| Abrasive slurry (silica, 2 m/s) | 0.2–0.5 mm/year | Quarterly | 1.2 mm (plan replacement) |
| Acidic chemical (30% HCl, 80°C) | 0.05–0.15 mm/year (permeation not erosion) | Semi-annually | 1.0 mm (monitor closely) |
| High-temperature (>150°C) | Creep thinning | Semi-annually | 1.5 mm (risk of rupture) |
| Ultrasonic cleaning tank | 0.1–0.3 mm/year erosion | Quarterly | 1.0 mm |
Field Example
A mining slurry loop used PFA heaters in silica sand service (2.5 m/s, 30% solids). Previously, they replaced heaters every 12 months preventively, often finding some heaters still thick (1.8 mm) and others nearly perforated (0.4 mm). They implemented ECT thickness measurement every 3 months. After 9 months, one heater measured 0.7 mm (from 2.0 mm new) while others were 1.4–1.6 mm. The thin heater was replaced. The others continued to 18 months. The plant saved $5,000/year in unnecessary replacements and avoided a catastrophic failure from the thin heater.
Equipment Recommendation
Portable eddy current thickness gauges suitable for PFA-on-metal include:
Olympus (now Evident) 45MG with lift-off calibration
Fischer FMP20 with probe ECP10
Elcometer N800 (configured for non-conductive coating on metal)
Cost: $3,000–6,000 for a basic gauge with probe. Training: 1–2 days. The investment pays for itself after 1–2 years in reduced unnecessary heater replacements and avoided failures.
Conclusion: ECT Provides Accurate, Non-Destructive Thickness Measurement
Eddy current testing allows non-destructive measurement of remaining PFA wall thickness on installed heaters with ±0.05–0.10 mm accuracy. The technique works through water and scale, requires no tank draining, and takes seconds per point. Calibrate with known thickness standards, correct for curvature, and track measurements quarterly or annually. Use ECT to replace heaters based on actual remaining wall thickness, not on calendar time. This extends heater life where possible and prevents unexpected failures where erosion is severe. For abrasive, corrosive, or high-temperature service, ECT is the most cost-effective predictive maintenance tool. Measure thickness, predict replacement, save money. The probe knows how much PFA is left. Let it tell you when to replace. Don't guess. Measure.
