9.2 HTHA Inspection, Detection Methods & Mitigation

The Inspection Challenge: Physical Stages of HTHA Degradation

1. Why Conventional NDE Fails Early HTHA Detection

Detecting High-Temperature Hydrogen Attack represents one of the most formidable non-destructive examination (NDE) challenges in petroleum refining asset integrity. In early-stage HTHA (Stage 1 sub-micron cavitation and Stage 2 micro-fissuring), equipment experiences zero wall thinning, no macroscopic dimensional swelling, and no surface-breaking defects.

                         HTHA DAMAGE EVOLUTION VS NDE DETECTABILITY

  Stage 1: Incubation / Sub-Micron Bubbles
  [ •  •  •  •  •  • ]  Sub-micron cavities (<0.5 µm) at grain boundaries
                        Conventional NDE: 100% Blind (Zero wall loss, zero RT contrast)
                        Advanced NDE: AUBT backscatter, high-frequency spectral attenuation
                                │
                                ▼
  Stage 2: Micro-fissuring
  [ ──•──•───•──•─── ]  Fissure links (10 to 100 µm) along grain boundaries
                        Conventional NDE: Blind (No thickness change, no backwall drop)
                        Advanced NDE: Velocity Ratio (VL/VS drop), TOFD tip diffraction, PAUT-TFM
                                │
                                ▼
  Stage 3: Macro-cracking & Rupture
  [ ════════════════ ]  Continuous planar intergranular cracks (>1 mm to through-wall)
                        Conventional NDE: Shear wave UT & WFMT (ID) can detect macro-flaws
                        Advanced NDE: TOFD, PAUT-FMC/TFM provide high-precision sizing

---

Advanced NDE Methodologies Described in API RP 941

To overcome the limitations of conventional testing, API RP 941 describes an advanced suite of complementary ultrasonic and metallurgical techniques specifically calibrated for detecting and characterizing HTHA.

1. Advanced Ultrasonic Backscatter Technique (AUBT)

• Operating Principle: When high-frequency ultrasonic waves encounter acoustic discontinuities that are smaller than the sound beam wavelength ($\lambda$), the sound scatters in all directions according to Rayleigh and Mie scattering principles. In undamaged steel, the acoustic wave passes through grain boundaries with minimal scattering. In steel containing methane cavities, the microscopic cavities act as individual point reflectors, scattering a fraction of the high-frequency acoustic energy directly backwards toward the transducer.
• Transducer Setup: Employs high-frequency, highly focused longitudinal or shear wave transducers (typically 5 MHz to 10 MHz or higher) with specialized focal depths set to interrogate the near-weld heat-affected zone.
• Signal Signature: The instrumentation records an amplitude-versus-depth profile. Undamaged steel produces a clean, flat baseline between the initial pulse and the backwall reflection. HTHA-damaged steel produces a distinctive cluster of high-frequency backscatter signals within the material volume.

2. Velocity Ratio (VR) Testing

• Operating Principle: Sound velocity in a crystalline solid depends directly on its elastic modulus and density. In undamaged carbon and low-alloy steel, the acoustic velocity for longitudinal waves ($V_L$) is approximately 5,900 m/s (0.232 in./µs), while the velocity for shear (transverse) waves ($V_S$) is approximately 3,240 m/s (0.128 in./µs). The ratio of longitudinal to shear wave velocity is constant across unattacked ferritic steels: $\text{Normal Velocity Ratio: } \frac{V_L}{V_S} \approx 1.80 \text{ to } 1.82$
• HTHA Velocity Shift: As methane micro-cavities form along grain boundaries, they degrade the material's bulk compressive elastic modulus far more aggressively than its shear modulus. Consequently, longitudinal waves—which propagate via compression and rarefaction—are slowed significantly, while shear waves experience a much smaller velocity decrease.
• Diagnostic Threshold: In HTHA-damaged steel, the velocity ratio shifts downward measurably: $\text{Damaged Velocity Ratio: } \frac{V_L}{V_S} < 1.78 \quad (\text{severe attack: } 1.72 \text{ to } 1.75)$
• Practical Execution: Requires precise simultaneous measurement of longitudinal and shear wave transit times through the identical sound path, or dual-element pitch-catch probes calibrated on baseline reference blocks.

3. Spectral Analysis and Frequency Attenuation

• Operating Principle: Rayleigh scattering cross-section is proportional to the fourth power of frequency ($\text{Scattering} \propto f^4$). Therefore, higher ultrasonic frequencies attenuate far more rapidly than lower frequencies when traversing micro-fissured material.
• Diagnostic Implementation: A broadband acoustic pulse containing frequency components from 2 MHz to 15 MHz is transmitted through the component. By performing a Fast Fourier Transform (FFT) on the received waveform, inspectors analyze the frequency spectrum. HTHA-damaged material filters out the upper frequencies, shifting the center frequency downward and producing a distinctive high-frequency spectral dropout.

4. Time-of-Flight Diffraction (TOFD)

• Operating Principle: TOFD utilizes a pair of angled longitudinal wave probes placed on opposite sides of a weld seam in pitch-catch mode. Rather than relying on specular reflection from flaw faces, TOFD detects the faint diffracted waves emitted from the upper and lower tips of discontinuities.
• Application to HTHA: While TOFD cannot resolve isolated sub-micron methane bubbles (Stage 1), it is exceptionally powerful for detecting and through-wall sizing of Stage 2 micro-fissure clusters and Stage 3 macroscopic intergranular cracks. TOFD provides continuous B-scan cross-sectional imaging of the weld heat-affected zone, enabling inspectors to measure crack depth, vertical extent, and remaining ligament with an accuracy of $\pm 0.5\text{ to }1.0\text{ mm}$.

5. Phased Array Ultrasonic Testing (PAUT) with FMC / TFM

• Full Matrix Capture (FMC): The instrument sequentially fires each individual element in a multi-element phased array probe (e.g., 32 or 64 elements) and records the raw, unsummed time-domain A-scan received by every element across the entire array, capturing the complete acoustic dataset.
• Total Focusing Method (TFM): Advanced post-processing algorithms divide the inspection volume into a high-density grid of pixels. For each pixel, the algorithm calculates the exact round-trip transit time from every transmitting element to the pixel and back to every receiving element, summing the amplitudes to synthesize a perfectly focused sound beam at every point in the grid.
• HTHA Capabilities: FMC/TFM delivers spatial resolution superior to conventional phased array, enabling inspectors to resolve fine intergranular crack clusters in the weld HAZ, map crack branching, and distinguish true HTHA micro-fissures from fabrication flaws (such as lack of side-wall fusion, porosity, or slag inclusions).

6. In-Situ Metallographic Replication (Surface Replica)

• Procedure: Field polishing of the external (or accessible internal) steel surface through progressive grinding stages down to 1-micron diamond paste, followed by chemical etching with 2% to 5% nital. A thin cellulose acetate film softened with acetone is pressed onto the etched surface to preserve the microscopic topography.
• Laboratory Examination: The peeled replica is analyzed under an optical metallurgical microscope or Scanning Electron Microscope (SEM) at 100x to 1,000x+ magnification.
• Diagnostic Value: In-situ replication clearly reveals surface decarburization (pure ferrite grains devoid of pearlite) and grain boundary micro-fissuring.
• Critical Limitation: Replication evaluates only the outermost metallurgical surface. It cannot detect subsurface, mid-wall, or inner-diameter HTHA. A clean surface replica provides zero assurance that subsurface weld HAZ micro-fissuring is not actively propagating.

Comparative Evaluation of HTHA Inspection Techniques

Because no single NDE method provides 100% probability of detection across all stages of HTHA, API RP 941 emphasizes a multi-technique approach. Combining high-frequency backscatter with velocity ratio measurements and diffraction-based imaging minimizes false negatives.

---

Engineering Prevention and Mitigation Strategies

1. Materials Selection & Upgrades

The primary defense against HTHA is selecting an alloy with sufficient thermodynamic resistance to withstand the maximum operating temperature and hydrogen partial pressure across the unit's lifecycle.

• Upgrading from Carbon Steel: In operating circuits where temperature and $p_{H_2}$ approach or exceed the non-PWHT or PWHT carbon steel curves, equipment should be upgraded to 1.25Cr-0.5Mo (ASTM A387 Gr 11) or 2.25Cr-1Mo (ASTM A387 Gr 22).
• Vanadium-Modified Low-Alloy Steels: For modern heavy-wall hydroprocessing reactors operating above 750 °F (399 °C) and 1,500 to 2,500 psia $p_{H_2}$, the industry standard is 2.25Cr-1Mo-0.25V (ASME SA-336 F22V / SA-542 Type D). The addition of approximately 0.25 wt% vanadium precipitates fine, coherent vanadium carbides ($V_4C_3$) throughout the grains. These carbides are exceptionally stable, tying up carbon and suppressing methane synthesis, while also providing superior resistance to high-temperature hydrogen embrittlement and temper embrittlement.
• Austenitic Stainless Steel Cladding / Weld Overlay: Heavy-wall hydroprocessing reactors typically specify 2.25Cr-1Mo or 2.25Cr-1Mo-0.25V base metal with an internal roll-bonded cladding or weld overlay of stabilized austenitic stainless steel (Type 347 or Type 321), typically 3/16 inch (4.8 mm) thick. Austenitic stainless steel has a face-centered cubic (FCC) crystal structure, in which hydrogen diffusivity is orders of magnitude lower than in ferritic steels ($D_{\gamma} \ll D_{\alpha}$). The stainless cladding acts as a hydrogen diffusion barrier, reducing the hydrogen concentration reaching the Cr-Mo base metal interface. Follow RP 941 guidance before taking any credit for cladding in an HTHA assessment.

2. Post-Weld Heat Treatment (PWHT)

PWHT is a key mitigation for carbon steel and low alloy steels in high-temperature hydrogen service (RP 941 has a separate, lower curve for carbon steel welded without PWHT):

• Stress Relief: PWHT relaxes peak residual tensile stresses resulting from weld shrinkage, dropping stresses from yield magnitude (>50 to 70 ksi) down to benign levels (<15 to 20 ksi). Because tensile stress supplies the mechanical driving force for cavity dilation, stress reduction significantly extends the incubation time.
• Carbide Stabilization: In Cr-Mo steels, PWHT at proper temperatures (1150 °F to 1350 °F / 621 °C to 732 °C depending on alloy grade) tempers the hard, brittle martensitic/bainitic weld microstructure, causing chromium and molybdenum to precipitate into stable complex alloy carbides ($M_{23}C_6$, $M_7C_3$).

3. Integrity Operating Windows (IOWs) per API RP 584

Refiners must establish and rigorously enforce Integrity Operating Windows (IOWs) for all equipment in hydrogen service:

• Critical Upper Temperature Limit: Temperature must be continuously tracked via skin thermocouples and process instrumentation. Process excursions—such as loss of cold recycle gas quench in hydrocracker beds, fired heater overfiring, or bypass of feed-effluent heat exchangers—can push equipment into the HTHA danger zone.
• Cumulative Exposure Tracking: Because HTHA incubation is cumulative, temporary operating excursions above the Nelson curve consume incubation life irreversibly. An excursion of 50 °F (28 °C) for a few weeks can initiate sub-micron cavitation that continues to propagate once normal operating temperatures are restored.
• Monitoring Hydrogen Partial Pressure: Refiners must monitor recycle gas hydrogen purity. Changes in feed sulfur/nitrogen levels or reformer operation that alter hydrogen purity impact $p_{H_2}$ proportionally.