Due to thermal “downshocks” and “upshocks” in the boiler, associated thermal expansions and contractions can lead to axial stresses that cause thermal fatigue cracks on the fireside of water wall tubing. This condition is often exacerbated by corrosion products near the crack tips that accelerate crack growth. Because of these phenomenon, through-wall circumferential cracking has occurred in the furnace wall tubes of both subcritical and supercritical boilers. While this type of cracking can be easily detected by visual inspection, managing the cracking requires the proper tools to accurately monitor the through wall extent and determine growth rates. The best means for monitoring these cracks is with Nondestructive Examination (NDE) methods. As illustrated here, not any NDE method will succeed in overcoming the numerous challenges in sizing this type of cracking in the boiler. We have successfully overcome these challenges by performing a combination of Eddy Current Testing (ECT) and Phased Array Ultrasonic Testing (PAUT) to locate and size the deepest circumferential cracks in boiler tubes. However, even proven techniques must be updated periodically to address new challenges, and, as described here, a recent experience with rifled boiler tubes required a technique modification to reliably size the cracking.
Figure 1. Cracking Severity
As shown in Figure 1, the visual severity of fire side circumferential cracking can vary considerable; from numerous closely spaced cracks numbering 20-30 cracks per inch, to sparse cracking that numbers only 2-3 per inch. In addition to the numerous and closely spaced cracking, additional challenges for NDE include hundreds of tubes with damage and many linear feet of tubing to examine (Figure 2).
As a result, Structural Integrity developed a two-stage approach that has proven effective in identifying and sizing the deepest cracks; ECT to screen for cracks greater than 0.050” deep and PAUT to size those greater than 0.050” cracks. The advantage of this approach is that ECT is fast and requires no couplant so it provides a means to rapidly screen numerous feet of tubing to isolate the most severe (deepest) cracking. This approach typically leaves a much-reduced area to perform detailed sizing using PAUT. Sizing with PAUT is typically performed by bouncing or skipping sound waves off the tube bore as shown in Figure 3. While this works effectively for smooth bore tubing, rifle bore tubing creates another unique challenge, as illustrated in Figure 4. The existence of rifle bore tubing in the boiler at a client site provided the impetus for reevaluating the current PAUT approach to determine the feasibility of sizing cracks in this tubing.
To evaluate the ultrasonic effectiveness for rifled tubing, EDM notches of depths ranging from 0.025″ to 0.150″ were machined into field removed tubes provided by the client. Accurate through wall sizing of cracks is best performed using the crack tip diffraction method, that is, resolving the small amplitude response from the crack tip and measuring from the tip to the component surface (corner trap) to determine the through wall height. This is illustrated in Figure 5. PAUT has many benefits when it comes to characterizing defects in power plant components and one of them that is corrected for distance and depth in the material. When implemented correctly, PAUT can provide the through wall size directly from the 2-dimensional sector scan as depicted in Figure 5 for a 0.100” deep notch in a rifled boiler tube.
Historically, this application was performed using a small footprint probe that did not require contouring the wedge of the probe to couple to the radius of the tube. These probes, which had been successful on smooth bore applications, are easy to manipulate and skew on the tube surface. However, because of their small aperture (total element surface area) they are limited in focusing ability. As previously mentioned, most sizing must be performed on the second leg of sound to ensure the sound beam reaches the crack and crack tip. The focal range of these small aperture probes is shorter than the metal path required for focused sound to reach the cracked Outside Diameter (OD) surface of the tube. To further complicate the examination, the geometry of the Inside Diameter (ID) surface of rifled bore tube caused the sound beams to bounce off the ID bore at irregular angles, as seen in Figure 4.
Therefore, not only were the small aperture probes not able to focus sound at the depths necessary to reach the cracked OD surface, but the rifled ID bore caused this limited sound field to bounce in multiple directions, thereby only allowing small “windows” of unfocused sound to reach the OD crack area.
To improve the PAUT sizing technique, larger aperture probes were evaluated using a combination of software simulation and laboratory testing to prove that the longer focal zone could effectively reach the cracked OD surface after bouncing off the ID surface. This testing confirmed the larger aperture provided a greater amount of sound energy and the geometry of the ID bore surface of a rifled bore tube would have less negative impact on the amount of sound energy reflected to the cracked OD surface.
Laboratory testing confirmed, at best, the smaller aperture probes could only resolve notch tips at the shortest possible metal path (i.e. the lowest angles), but did not have sufficient aperture to detect tip signals at higher metal path distances (i.e., higher angles). While using small aperture probes and low angles worked effectively for smooth bore tubing, the limited “windows” to skip sound in rifled tubing combined with the limited effective angles and short range of the smaller aperture probes resulted in the possibility of missing crack tips and erroneous sizing measurements.
Although the use of a larger aperture probe requires a wedge that must be contoured to fit the curvature of the tube outside diameter, this combination of larger probe and contoured wedge provides range of angles which enhances the ability to size cracks. As an example, on a field removed tube using the larger aperture probe, the tip of a notch imbedded in a group of shallow cracks is readily detectable over an angular range of 35° to over 60° in the sector scan with indications that travel for over 1/2-inch as the probe is indexed axially (Figure 6). For the smaller probes, effective angular range was typically 35° to 45° with potential axial travel less than 1/4-inch and was truncated by the rifling which greatly reduced the chances of locating and resolving the crack tip.
Based on software simulations and laboratory testing on field removed samples with known and unknown defects, it was concluded sizing cracks in rifled tubing is feasible with some adjustments to our existing techniques and approach. While ECT screening is unaffected by the rifled tubing, for PAUT, effective techniques require probes with extended focal ranges and the use of contoured wedges to couple to the tube. Since tube types and sizes can vary considerably from plant to plant, it is a good idea to have representative tubing to use for calibration. When properly applied, NDE can provide a huge benefit for managing cracking in boiler tubing and other components.