Inspection planning and execution framework covering effectiveness grading, NDE method selection, CML strategy, thickness trending, and remaining life projection.
API 581 defines five levels of inspection effectiveness that quantify how well an inspection reduces uncertainty about the equipment's condition. Level A (Highly Effective): Inspection techniques that provide near-certain detection of the damage mechanism in question, with accurate sizing and characterization. For thinning, this means automated UT scanning with full coverage. For cracking, this means TOFD or PAUT with full volumetric coverage. A-level inspections provide the greatest reduction in the damage factor, potentially reducing PoF by up to 90%. Level B (Usually Effective): Inspection techniques that detect most damage with reasonable accuracy. For thinning, this means manual UT at a high density of TMLs. For cracking, this means WFMPI or angle-beam UT at key locations. B-level inspections typically reduce the damage factor by 60-80%. Level C (Fairly Effective): Inspection techniques that detect significant damage but may miss early-stage or localized damage. Spot UT at widely-spaced TMLs for thinning, or visual inspection supplemented with limited UT for cracking. Damage factor reduction of 30-50%. Level D (Poorly Effective): Inspection techniques that provide minimal confidence in the equipment's condition. Visual inspection alone for internal corrosion, or spot UT at a very limited number of locations. Damage factor reduction of 10-20%. Level E (Ineffective): No inspection, or inspection methods that are not applicable to the damage mechanism in question. The damage factor continues to accumulate without reduction. Reliatic assigns effectiveness levels based on the NDE method, coverage area, and the damage mechanism being targeted. The same inspection method can achieve different effectiveness levels depending on the specific damage mechanism.
Selecting the right Non-Destructive Examination (NDE) method is critical to achieving the target inspection effectiveness. Reliatic guides method selection based on the active damage mechanism. For general and localized thinning: Ultrasonic thickness measurement (UT) is the primary method. Automated UT scanning (AUT) with corrosion mapping achieves the highest effectiveness for both general and localized thinning. Profile radiography is effective for piping where UT access is limited. Long-range guided wave UT (LRGUT) can screen long pipe runs for areas of significant thinning. For external corrosion under insulation (CUI): Pulsed eddy current (PEC) can screen through insulation without removal. Neutron backscatter detects trapped moisture under insulation. Thermography identifies wet insulation areas. Targeted insulation removal with visual and UT examination provides definitive assessment. For cracking: Phased array UT (PAUT) and time-of-flight diffraction (TOFD) provide the best volumetric detection and sizing accuracy for subsurface cracks. Magnetic particle inspection (MPI) detects surface-breaking cracks in ferromagnetic materials. Dye penetrant inspection (DPI) detects surface-breaking cracks in non-ferromagnetic materials. Acoustic emission (AE) can monitor for active crack growth during operation. For high-temperature damage: In-situ metallography (replication) reveals microstructural changes from creep, graphitization, and embrittlement. Hardness testing detects temper embrittlement and decarburization. Advanced UT techniques (backscatter, velocity ratio) detect HTHA damage. The platform's inspection work scope generator automatically recommends NDE methods based on the credible damage mechanisms and the target effectiveness level.
Condition Monitoring Locations (CMLs), also called Thickness Monitoring Locations (TMLs), are the permanent measurement points where thickness readings are taken over the life of the equipment. Strategic placement of CMLs is essential for effective thickness monitoring. Reliatic supports CML strategy through several capabilities. Placement Guidance: The platform recommends CML locations based on damage mechanism susceptibility, flow patterns, and equipment geometry. For example, on a pipe bend in erosive service, CMLs should be placed on the extrados (outer radius) where erosion is most severe. On a vessel in a corrosive service, CMLs should cover the liquid-wetted zone, the vapor space, the liquid-vapor interface, and areas near nozzles and attachments. Numbering and Identification: Each CML receives a unique identifier (typically formatted as equipment tag + sequential number) and is associated with specific coordinates on an isometric drawing or equipment sketch. This ensures that successive inspectors measure at exactly the same location, producing meaningful trend data. Grouping and Circuits: For piping systems, CMLs are organized into corrosion circuits — groups of piping components that share the same process conditions and are expected to experience similar corrosion rates. The worst CML in a circuit governs the remaining life calculation for the entire circuit. Reliatic automatically identifies the governing CML and highlights it in reports. Coverage Assessment: The platform evaluates whether the current CML layout provides adequate coverage for the identified damage mechanisms. If a new damage mechanism is identified (e.g., through a MOC review or an incident), the platform can recommend additional CML locations to address the gap.
Thickness trending is the process of analyzing successive thickness measurements to determine the rate of material loss and project the remaining useful life of the equipment. Reliatic implements two trending methods. Linear Regression (Least Squares): A straight line is fitted to all available thickness data points using least squares regression. The slope of the line represents the average corrosion rate, and the x-intercept (where thickness equals the minimum required thickness) represents the projected retirement date. This method works best when 5 or more data points are available and the corrosion rate is relatively steady. Minimum-of-Data: When fewer data points are available, or when the corrosion rate is accelerating, the trending uses the minimum measured thickness and the most recent corrosion rate to project remaining life. This method is more conservative and is the default when fewer than 5 data points exist. Remaining life is calculated as: Remaining Life = (t_measured - t_min) / CR, where t_measured is the most recent thickness measurement, t_min is the minimum allowable thickness per the applicable design code (ASME, EN, etc.), and CR is the corrosion rate. Reliatic displays remaining life on the asset detail page and uses it to generate alerts when remaining life falls below configurable thresholds (typically 2 years for a warning and 6 months for a critical alert).
Reliatic calculates two corrosion rate metrics from thickness data. Short-Term Corrosion Rate: Calculated from the two most recent thickness measurements. Short-term rate = (t_previous - t_current) / (time between measurements). This rate captures recent changes in corrosion behavior and is sensitive to process upsets, chemical treatment changes, or new damage mechanisms. It is useful for detecting trend changes but can be noisy if measurement uncertainty is significant relative to the thickness change. Long-Term Corrosion Rate: Calculated from the original (as-built or furnished) thickness and the most recent measurement. Long-term rate = (t_original - t_current) / (total time in service). This rate smooths out short-term variations and provides a stable baseline for remaining life calculations. It is less responsive to recent changes but more robust against measurement scatter. Reliatic uses the higher of the two rates for remaining life calculations, providing a conservative projection. When the short-term rate significantly exceeds the long-term rate, the platform flags this as a potential corrosion rate acceleration and generates an alert for engineering review. Conversely, when the short-term rate is significantly lower (which may indicate measurement error or genuine improvement from a process change), the platform notes the discrepancy but does not automatically reduce the design corrosion rate without engineering approval. Confidence intervals are computed for both rates, accounting for measurement uncertainty (typically +/- 0.1 mm for manual UT). Wide confidence intervals — common when few data points exist or when the corrosion rate is close to the measurement accuracy — are flagged as a data quality concern that may warrant increased inspection frequency.