In A Flue Gas Desulfurization Unit Reheater, Why Does The Presence Of Fly Ash Particles Preferentially Erode The Titanium Tube's Upstream Side At Bends?
Leave a message
Flue gas desulfurization (FGD) units in coal-fired power plants use reheaters to raise the temperature of cleaned flue gas before it enters the stack. Titanium is the material of choice due to its corrosion resistance in the acidic, chloride-rich environment. However, FGD flue gas contains fly ash particles-angular, hard (mainly silica and alumina), with sizes ranging from 1 to 100 µm. These particles travel with the gas at velocities of 15–25 m/s. When the gas stream encounters a bend in the titanium tube, the heavier particles cannot change direction as rapidly as the gas. They continue along their original trajectory, impacting the tube wall at the outer radius of the bend. This particle impact causes erosion, preferentially removing material from the upstream side of the bend. Over time, wall thinning leads to perforation and leakage of flue gas or condensate.
Mechanism of Fly Ash Erosion at Tube Bends
The erosion mechanism is a combination of impact wear and cutting. Fly ash particles strike the titanium surface at angles typically 20–50 degrees relative to the surface tangent. Titanium, while tough, has relatively low hardness (180–220 HV for Grade 2) compared to silica (approximately 1,200 HV). Each particle impact displaces a small volume of titanium-on the order of 1–10 µm³ for a 20 µm particle at 20 m/s. The erosion rate follows a power law: Erosion ∝ (V)^n × (d)^m × (particle concentration), where V is velocity, d is particle diameter, n is approximately 2.3 for titanium, and m is approximately 2.0. At bends, the effective impact velocity is higher than the bulk gas velocity due to particle acceleration around the curve. The upstream side of the tube receives the highest density of impacts because particles have not yet been deflected by previous collisions.
Quantitative Erosion Rates at Different Bend Locations
Field measurements and laboratory testing (20 m/s gas velocity, 50 mg/m³ fly ash, 20 µm mean particle size) have established the following erosion rates for Grade 2 titanium:
Straight tube section, uniform flow: Erosion rate of 0.05–0.10 mm per year. Material removal is relatively uniform around the circumference. Acceptable for 10–15 year life with 1.5 mm wall thickness.
Bend outer radius, 90° elbow, upstream side: Erosion rate of 0.6–1.2 mm per year. Maximum wall loss occurs at the point where particle trajectory first contacts the tube. Localized thinning can perforate a 1.5 mm wall within 1–2 years.
Bend outer radius, downstream side (after impact): Erosion rate of 0.2–0.4 mm per year. Particles have lost kinetic energy after initial impact. Secondary erosion still significant.
Bend inner radius (intrados): Erosion rate of 0.02–0.05 mm per year. Lower particle impact density. Generally protected by the gas boundary layer.
Multiple bends in series (e.g., serpentine reheater): Erosion rates increase by 20–50% on each subsequent bend because particles become more angular and aggressive after initial impacts.
Erosion Mitigation and Design Selection Guide
The following table provides a decision framework for managing fly ash erosion at titanium tube bends based on particle loading and gas velocity:
| FGD Reheater Condition | Recommended Erosion Protection | Expected Erosion Rate at Bend | Wall Thickness Recommendation |
|---|---|---|---|
| Low ash loading (<20 mg/m³), low velocity (<15 m/s) | Standard Grade 2, no protection | 0.2–0.4 mm/year | 1.5 mm |
| Moderate ash (20–50 mg/m³), velocity 15–20 m/s | Oversized bend radius (≥3D), Grade 7 titanium | 0.3–0.5 mm/year | 2.0 mm |
| High ash (50–100 mg/m³), velocity 20–25 m/s | Tungsten carbide overlay (0.5 mm) on upstream bend surface | <0.05 mm/year (coated) | 1.5 mm with coating |
| Severe ash (>100 mg/m³), abrasive fly ash (high silica) | Replaceable wear plates or sleeves at bends | N/A (wear part replaced) | 1.2 mm tube with 2 mm wear plate |
| Existing FGD with known bend erosion | Install turning vanes upstream of bend | Reduces erosion by 60–80% | N/A |
Engineering Beyond Erosion Resistance
The titanium alloy grade affects erosion resistance. Grade 7 (palladium-stabilized) has similar hardness (180–220 HV) and erosion rate to Grade 2. Grade 12 (molybdenum-nickel) is slightly harder (200–240 HV) and reduces erosion by 15–20%. Surface treatments can improve erosion resistance: nitriding produces a TiN surface layer (500–800 HV) that reduces erosion by 70–80% but adds cost. The tube wall thickness at bends should be increased beyond the straight section requirement. A typical design uses 1.5 mm wall for straight sections and 2.5 mm wall for bend outer radii. Bend geometry is critical: long-radius bends (R = 5D) experience 50% lower erosion than standard-radius bends (R = 1.5D). The gas flow pattern can be modified with turning vanes upstream of bends to redirect particles away from the outer radius.
Making an Informed Specification
When specifying a titanium reheater tube for FGD service with fly ash, identify all bends in the gas path. For each bend, specify an increased wall thickness (2.0–2.5 mm) on the outer radius, either by using a seamless bent tube with thicker wall or by applying a weld overlay of titanium or tungsten carbide on the upstream side. Request computational fluid dynamics (CFD) modeling of particle trajectories through the reheater to identify high-erosion zones. During operation, perform ultrasonic thickness testing annually at bend outer radii. Establish a critical wall thickness threshold (e.g., 0.8 mm remaining) at which the tube should be replaced or repaired. For new installations, specify a minimum bend radius of 3D (three times tube outer diameter). If space constraints require tight bends (R = 1.5D), require tungsten carbide coating or replaceable wear sleeves. By addressing preferential erosion at bends as a distinct design consideration, the engineer extends the service life of titanium FGD reheaters from 1–2 years to 10–15 years in fly-ash-laden flue gas streams.
