Steam Drum Inspection: Best NDE Techniques for Cracks

By Johnson on July 28, 2026

steam-drum-inspection-nde-techniques-crack-detection

A steam drum is the thickest-walled pressure vessel on your boiler island and often the component that receives the least systematic NDE attention during routine outages. The assumption that thick walls provide wide safety margins breaks down the moment a 12mm ligament crack is found propagating through a tube hole field that has not been examined with magnetic particle testing in three consecutive inspection cycles. Steam drum cracks do not grow at a constant rate once they initiate in a high-stress concentration zone. They accelerate under cyclic loading from every startup and shutdown, and they can reach critical length between two scheduled inspections if the wrong NDE technique was selected or the wrong locations were examined. Book a demo to see how iFactory structures steam drum inspection programs around actual defect history rather than calendar intervals.

What You Cannot See Inside a Steam Drum Will Eventually Shut You Down

Steam drums operate for decades under cyclic thermal and mechanical stress. The cracks that force unplanned outages almost always start small, in locations that only the right NDE technique can reach.

Four Defect Zones Where Steam Drum Cracks Actually Form

Steam drum cracking is not random. It concentrates at specific geometries where stress, temperature differential, and fabrication history create conditions for crack initiation and propagation.

Zone A

Tube Hole Ligament Field

The web of material between adjacent tube holes carries the highest localized stress in the drum. Thermal cycling during startups causes differential expansion between the tube stubs and the drum shell, loading each ligament in tension-compression reversal. Cracks initiate at the tube hole surface and propagate radially through the ligament toward adjacent holes. This is the single most common location for steam drum cracking across all boiler types and pressure ranges.

Primary NDE: MT, PAUT
Zone B

Nozzle Weld Junctions

Every feedwater, chemical injection, steam outlet, and instrument nozzle creates a weld junction where the nozzle forging meets the drum shell. These joints are subject to thermal shock from temperature differences between incoming feedwater and drum saturation temperature, as well as external loading from piping thermal expansion. Cracking typically occurs in the weld heat-affected zone on the drum shell side, often at the toe of the weld where stress concentration is highest.

Primary NDE: MT, TOFD, PAUT
Zone C

Longitudinal Seam Welds

Older drums manufactured from rolled plate use longitudinal seam welds to close the cylindrical shell. These welds were often made using submerged arc or electroslag processes that can leave lack-of-fusion indications, slag inclusions, or hydrogen-induced cracking in the weld centerline or fusion boundaries. While these are fabrication defects rather than service-induced cracks, they grow under cyclic loading and must be evaluated for propagation during every inspection cycle.

Primary NDE: TOFD, PAUT, UT
Zone D

Internal Surface and Attachments

The internal drum surface is exposed to saturated steam and water chemistry that can cause corrosion pitting, under-deposit corrosion, and fatigue cracking at internal attachment welds such as steam separators, feedwater distribution pipes, and chemical injection quills. These internal surfaces are only accessible during extended outages when manway covers are removed, making visual and magnetic particle examination timing critical to the outage schedule.

Primary NDE: VT, MT

NDE Technique Capability Map: Which Method Detects What

No single NDE technique can detect every defect type in a steam drum. The capability map below shows the relative detection effectiveness of the four primary techniques across the five defect categories most relevant to drum integrity assessment.


Magnetic Particle (MT)

Conventional UT

Phased Array (PAUT)

TOFD
Surface Crack Detection

95%

30%

45%

20%
Subsurface Flaw Detection

35%

75%

90%

85%
Weld Examination

15%

80%

92%

90%
Through-Wall Sizing

5%

65%

80%

95%
Corrosion Mapping

80%

55%

65%

25%

Ligament Cracking: The Defect That Defines Drum Life

Ligament cracking accounts for more steam drum replacements than all other defect types combined. Understanding how it initiates, how it propagates, and how to detect it early is the single most impactful knowledge a boiler inspection team can apply.

68%

of all recorded steam drum crack indications are found in tube hole ligament fields

2–8mm

typical crack depth at first detection when MT is applied on a routine inspection interval

3x

faster crack growth rate in drums experiencing more than 100 cold starts per year versus baseload units

12mm

average critical crack length in a standard ligament that triggers mandatory repair per ASME criteria


Initiation

Micro-cracking begins at the tube hole ID surface where stress concentration from the hole geometry and thermal loading from tube expansion create the highest local stress intensity factor in the ligament.


Early Propagation

Cracks grow radially outward through the ligament thickness at a rate driven by the cyclic stress range. At this stage, cracks are typically 1-4mm deep and only detectable by magnetic particle testing on the prepared tube hole surface.


Acceleration

As crack depth increases, the stress intensity factor at the crack tip rises non-linearly. Growth rate accelerates even if operating conditions remain unchanged. PAUT from the OD surface can now detect and size the crack through-wall.


Critical Length

The crack approaches the critical length where remaining ligament strength can no longer sustain the design pressure with the required safety margin. At this point, the component must be repaired or replaced before the next operating cycle.

See Your Drum Inspection Data Structured for Action

iFactory ingests NDE findings from every outage, correlates them to operating history, and surfaces which zones are trending toward critical defect thresholds before the next inspection window closes.

Technique Selection Guide by Defect Type and Location

Selecting the wrong NDE technique for a given defect type and location is the most common reason cracks go undetected between inspections. This matrix maps the right technique to the right job based on what the inspection needs to accomplish.

Defect Type
Location
Primary Technique
Supplementary
Key Advantage
Surface ligament cracks
Tube hole ID surface
Wet MT
PAUT from OD
Highest sensitivity to surface-breaking cracks under 1mm depth
Subsurface ligament cracks
Ligament mid-wall
PAUT
Conventional UT
Imaging capability shows crack orientation and through-wall extent
Nozzle weld toe cracks
Weld HAZ, OD surface
Wet MT
TOFD
Direct surface access with no couplant limitations on curved geometry
Nozzle weld embedded flaws
Weld volume
TOFD
PAUT
Accurate through-wall height sizing without amplitude dependence
Seam weld lack of fusion
Weld centerline
TOFD
PAUT
Detects planar defects independent of defect orientation relative to beam
ID corrosion pitting
Internal drum surface
VT + MT
UT thickness
Direct visual assessment with MT confirmation of pit-associated cracking
General wall thinning
Full drum circumference
UT thickness scan
Profile mapping
Systematic grid measurements build a thickness map for remaining life calculation

The Steam Drum Inspection Sequence, Step by Step

A thorough steam drum inspection follows a specific sequence that maximizes defect detection while respecting the limited time available during an outage. Skipping any step reduces the confidence level of the entire assessment.

Step 01

Internal Visual Examination

Manways are opened and the internal surface is cleaned to bare metal. A systematic visual walk-through documents corrosion pitting, deposits, attachment weld condition, and any visible cracking on the ID surface. This step establishes the baseline condition and identifies where to focus magnetic particle testing.

Step 02

Magnetic Particle Testing of Tube Holes

Every accessible tube hole in the ligament field is prepared and examined with wet fluorescent magnetic particle testing. Indications are measured for length and recorded by location. This is the most time-intensive step and the one most likely to be shortened under schedule pressure, which is precisely when ligament cracks go undetected.

Step 03

Magnetic Particle Testing of Nozzle Welds

All nozzle-to-shell welds on both the ID and OD surfaces are examined with MT. The nozzle weld toe and heat-affected zone on the shell side receive particular attention because this is where thermal shock cracking initiates. Findings are correlated with any previous inspection records at the same locations.

Step 04

Ultrasonic Examination of Welds

Longitudinal seam welds and nozzle attachment welds are examined from the OD surface using TOFD and PAUT. These techniques detect and size subsurface flaws that MT cannot reach, including lack of fusion, slag inclusions, and embedded cracking that may have propagated from fabrication defects during service.

Step 05

Ultrasonic Thickness Mapping

A systematic grid of thickness measurements covers the full drum shell, with denser spacing in known corrosion zones. Results are compared against previous thickness maps to calculate the corrosion rate and project minimum thickness at the next inspection interval for remaining life assessment.

Step 06

Findings Correlation and Reporting

All NDE findings are compiled, correlated with operating history and previous inspection data, and evaluated against applicable acceptance criteria such as ASME Section VIII or API 579. Defects are classified by severity, and recommendations are issued for repair, continued service with monitoring, or replacement at the next outage.

Defect Severity Classification and Required Response

Not every indication found during a steam drum inspection requires the same level of response. This classification system provides a consistent framework for triaging findings and allocating repair resources during a time-constrained outage.

Level 1

Low Severity: Monitor

Isolated MT indications under 2mm with no evidence of propagation since the previous inspection. No embedded flaws detected by UT. Corrosion rate within design allowance. Action: Document and re-examine at the next scheduled inspection interval. No repair required.

Level 2

Moderate Severity: Engineering Evaluation

MT indications between 2mm and 5mm, or confirmed growth since the last inspection. Small embedded UT indications below acceptance criteria but requiring fracture mechanics evaluation to confirm remaining life. Action: Perform fitness-for-service assessment per API 579. Shorten next inspection interval.

Level 3

High Severity: Plan Repair or Replacement

MT indications exceeding 5mm or approaching critical ligament length. Embedded flaws exceeding code acceptance limits. Confirmed crack growth rate that projects critical length before the next inspection. Action: Schedule repair by weld excavation and repair or drum replacement at the next planned outage.

Level 4

Critical Severity: Immediate Action

Indications at or exceeding critical crack length for the remaining ligament or weld joint. Through-wall cracking detected. Wall thickness at or below minimum required thickness with no remaining corrosion allowance. Action: Do not return to service. Evaluate for emergency repair or immediate drum replacement.

Frequently Asked Questions

Why is magnetic particle testing the primary method for ligament cracking instead of ultrasonic testing?

Magnetic particle testing provides the highest sensitivity to surface-breaking cracks, which is exactly what ligament cracks are when they initiate at the tube hole ID surface. MT can detect cracks as small as 0.5mm in length on a properly prepared surface, whereas conventional UT from the OD surface cannot reliably detect cracks that have not yet propagated a significant distance through the wall thickness. PAUT can detect subsurface ligament cracks from the OD, but it is used as a supplementary technique to size cracks already found by MT, not as the primary screening method. Book a demo to see how iFactory integrates MT and PAUT findings into a single defect map.

What is the difference between TOFD and PAUT for steam drum weld examination?

TOFD uses diffracted signals from the tips of defects to measure through-wall height directly, making it highly accurate for sizing but less effective at detecting surface-breaking defects near the scanning surface. PAUT uses multiple focused ultrasonic beams that can be steered electronically to image the weld volume from multiple angles, providing better detection of surface and near-surface flaws but slightly less precise height sizing than TOFD. For steam drum inspections, the standard practice is to use both techniques together: TOFD for accurate sizing of embedded flaws and PAUT for surface and near-surface coverage that TOFD cannot provide. The combination gives complete volumetric coverage of the weld with the highest confidence level.

How often should a steam drum be inspected for cracking?

Inspection intervals depend on the regulatory code governing the unit, the drum's service history, and findings from previous inspections. Under API 510, pressure vessels typically require internal inspection every 10 years or at the half-life corrosion allowance interval, whichever is shorter. However, for steam drums with known ligament cracking or high cycling duty, most owner-operators reduce this to every 5 years or even every 3 years for the most aggressive cycling profiles. The key principle is that the interval should be shorter than the time it takes for a detectable crack to grow to critical length at the observed growth rate. Talk to a specialist about determining the right interval for your drum.

Can a steam drum with ligament cracks continue in service without repair?

Yes, but only if a formal fitness-for-service evaluation demonstrates that the remaining ligament strength satisfies the applicable code requirements with an adequate safety margin. The evaluation per API 579 Level 3 assessment uses fracture mechanics to calculate the critical crack length for the specific ligament geometry, material toughness, and operating stress, then confirms that the detected crack length is below that critical value with sufficient margin to reach the next inspection. If the evaluation is favorable, the drum can continue in service with a shortened inspection interval and a commitment to re-examine the affected ligaments at the next outage. If the evaluation shows insufficient margin, repair or replacement is mandatory before return to service.

What causes ligament cracks to initiate even in relatively new drums?

Ligament cracking is not exclusively an aging phenomenon. New drums can develop cracks within the first few years of service if the tube hole geometry has sharp corners from drilling rather than reaming, if the tube-to-drum weld creates excessive residual stress in the ligament, or if the unit experiences frequent thermal cycling from two-shifting or load-following operation. Cold starts produce the most severe thermal shock because the temperature differential between the incoming feedwater and the drum shell is at its maximum. Drums in cycling service have been found to initiate ligament cracks in as few as 20,000 operating hours, compared to 100,000 or more hours in baseload units. Book a demo to see how iFactory correlates startup cycles to ligament crack initiation risk for your drum.

The Bottom Line on Steam Drum Inspection

Steam drum cracks do not appear overnight, but they are found overnight because that is when the inspection finally happens. The difference between a planned drum replacement and a forced outage is not better steel or a thicker wall. It is the discipline to apply the right NDE technique to the right location at the right interval and then act on what the data shows before the crack reaches critical length. Every ligament that goes unexamined, every nozzle weld that gets a visual check instead of magnetic particle testing, and every seam weld that is skipped because the outage schedule is tight is a calculated risk that eventually calculates against you.

Stop Guessing About Your Drum Condition

Book a 30-minute scoping call and bring your last drum inspection report. iFactory correlates your NDE findings with operating history and shows exactly which zones are trending toward your defect thresholds.


Share This Story, Choose Your Platform!