Knowing how to locate and sample boiler tubes that leak or fail, to enable a proper root-cause analysis (RCA), is critical to the long-term reliability of your heat-recovery steam generator (HRSG). Periodic tube sampling also is highly recommended by the industry’s top chemists and metallurgists, as a preventive measure, to identify early on if damage mechanisms are at work on the steam/water side of your units. These experts can help select the sample locations and perform the analyses to assure meaningful results.

Tube sampling was the focus of the first presentation, by Craig Dube, at HRST Inc’s “F-Class HRSG Spotlight Session” at the 24^th annual meeting of the 7F Users Group at the Sheraton Downtown (Denver) in mid-May. He began by answering this question: “Why sample when you find a tube leak and not just weld over the leak?”

The answer: A meaningful RCA is not possible without examining the inside surface of the tube. Some of the first things to determine, Dube continued, are these: Did the failure initiate from inside or outside the tube? Is cracking in evidence? Is pitting involved? If you find cracking, what is the shape of the cracks? Are the cracks in or near a weld? Do the cracks follow the grain boundaries or do they cross them? Regarding the condition of the tube internal surface, is scale present? If bare metal, is it polished?

The speaker said owner/operators will get best results from an investigation that provides both a macro view of the failure location, by a boiler engineer, and a micro view of the damage, by a metallurgist. It’s important, he added, to carefully document the failure (sample) location. Using an economizer panel as an example, Dube asked: Is the location of the distress at a tube-to-header joint? Might thermal shock be involved? Corrosion fatigue? For a multi-pass economizer, is the failure near a splitter plate? Again, might thermal shock be involved? What about buoyancy instability? If the failure is in the middle of a panel, is the affected tube up-flow or down-flow?

For an evaporator panel, is the failure near a feeder, near sidewalls, near coil-to-coil baffles? Is high-velocity wear in evidence? Is heat flux higher than the design value? For superheaters and reheaters, is the failure downstream of an attemperator? If so, is there overspray or leak-by? If the damage is at the lower header and near the ends of the panels, might undrained condensate be involved? Is the damage directly across from header nozzles?

To be sure there’s no misinterpretation of field findings, it might be in your best interest to retrieve a drawing of the panel and show exactly where the damage was found. Also provide design data if available, and any relevant operating information from the plant historian.

Case study. Dube next ran through several tube-leak case histories. Perhaps the case of greatest interest to this F-class audience, and the one best illustrating the need to follow all the evidence to assure a proper conclusion, was one involving a panelized feedwater heater that had experienced many tube failures.

HRST performed the failure analysis for the plant owner, using both macro and micro assessments of the damage. A tube sample and failure sample photos were received by HRST from its customer to initiate work. Plant personnel had kept meticulous records and reported the following stats and observations:

• There were 40 failures in up-flow passes, 117 in down-flow passes.

• Buoyancy instability was considered the most likely cause of the down-flow failures. Note that down-flow tubes absorb more heat and are more susceptible than up-flow tubes to buoyancy instability.

Here’s the mixed-bag of evidence the HRST sleuths had:

• Photos of the sample site revealed internal corrosion with crack-like indications.

• Lab analysis of the sample received showed pitting, but no cracks.

They concluded, based on a macro view of failure locations and micro view from photos and sample lab analysis, the most likely cause of failure is a combination of the following:

• Axial stress caused by flow instability in down-flow tubes.

• Pass-to-pass stress caused by temperature differences between up-flow and down-flow sections.

• Internal corrosion of tubes.

Investigators agreed that (1) not having an accurate sample location could have led to a misleading conclusion, (2) care must be taken to avoid attributing failures to pitting only, and (3) repairs to the tube failure area should involve redesign of economizer up-flow and down-flow characteristics.

Another case study began with the finding of “raccoon tails” in the LP economizer. Non-destructive examination (NDE) of the tube in place revealed crack indications, and a tube sample was removed. Lab analysis identified microscopic intergranular cracks, highly branched within a narrow band, which is indicative of stress-corrosion cracking. With this information, HRST engineers concluded weld repair of the cracks would be unreliable and that an upgraded material should be specified for replacement tubes.

Sampling of high-pressure (HP) evaporator tubes was the next topic covered by Dube. He told attendees these tubes can be damaged beyond repair from waterside-deposits. Under-deposit corrosion and overheating are the primary damage mechanisms of concern, Dube said. He cautioned users responsible for HRSGs operating more than 10 years without having had a representative HP-evap tube sample analyzed by a qualified metallurgist that they were taking an unnecessary risk. Deposits form slowly, he reminded, and you can have no problems for years. But if a water-chemistry upset were to occur you could have big trouble quickly.

The speaker continued, providing a backgrounder on the following forms of under-deposit corrosion:

• Hydrogen embrittlement typically occurs at operating pressures above 1500 psig when the system is subjected to a low-pH excursion. Deposits exacerbate the damage mechanism because hydrogen concentrates under the deposit. This condition can occur in only a few hours.

• When used for pH control, caustic can concentrate on the tube-side of deposits and attack the base material. Thinning of the tube wall is a result.

• If a blended phosphate is part of your plant’s water-chemistry regime, and an upset occurs, attack by hydrochloric or sulfuric acid can occur under a deposit. Local tube-wall thinning is a result.

Dube said HP evaporator tubes in the highest-risk group are in F-class HRSGs where duct burners often are used, there’s a history of flow-accelerated corrosion (FAC) and/or water-quality issues, or a borescope inspection reveals waterside tube-wall deposits of concern. He suggested the following three-step process to assure proper sampling:

1. Perform a heat-flux analysis of the evaporator to help identify areas at highest risk.

2. Borescope the high-risk zones. This is not as easy as it might sound. Before moving forward, it’s important to carefully select entry points to the waterside, specify the hardware and mechanical team required to gain access as well as the borescope best suited for the task, and then finalize work plan and procedures with the borescope technicians assigned to your job.

3. Use borescope results to influence the urgency of sampling and where to cut out the samples. Don’t forget gathering the macro information before sending the samples to the lab for metallurgical analysis.

Two additional points made by the speaker:

• Be sure you have the proper tube material on hand to replace the section removed for the sample.

• In some cases, tubes will fail at locations virtually impossible to reach and tube plugging is the practical solution. Happens. Problem, of course is that there’s no sample to analyze to determine the failure mechanism. If you go this route, highly experienced crafts persons are recommended.