SIM-98-010

BACKGROUND
Prior to the onset of boiler life extension or unit betterment programs in the early 1980s, superheater and reheater sections were often totally replaced when tube failures began to impact unit availability. While this was effective in preventing additional tube failures, it was akin to throwing out the baby with the bath water, as many tubes with decades of future reliable service were replaced. In order to provide utilities with some planning tools, tube samples were removed and analyzed in the laboratory to determine the extent of damage and degradation. This was state-of-the-art practice at the time for determining remaining life of SH and RH tubing, but often the boiler was back on-line when results were received. At a cost of several thousands of dollars per sample, the number of samples was usually restricted to less than a few percent of the total number of tubes and did not necessarily give a good projection of future behavior.

In the mid-1980s, both Babcock & Wilcox and Dayton Power & Light developed reliable ultrasonic techniques for measuring both the wall thickness and the internal oxide scale thickness. DP&L personnel were testing RH tubes for coal ash corrosion and found that the thicknesses determined by UT were greater than those measured in lab samples. The difference was that an internal oxide scale (built up over years of operation) added to the total wall thickness. Using somewhat higher frequency transducers, DP&L was able to separate the wall and oxide thicknesses.

B&W developed the technique using a much higher frequency transducer, which permitted more precise oxide thickness measurements. They first introduced a testing service in the mid-to-late 1980s, which proved to be of great benefit, as thousands of feet of relatively undamaged tubing were left in service. As hundreds of readings could be obtained in a day, this permitted the development of SH/RH maps showing relative damage versus pendant and tube number. This permitted maintenance personnel to plan cutting and welding of heavily damaged tubes, and minimize the number of good tubes that had to be cut to get to the bad ones.

Using the oxide thickness and total operating time, average tube temperature can be estimated, and using current wall reading, tube wastage can be estimated. In 1984, Dave French and John Alice of GPU Services developed a computer program to estimate the total life of a tube based on these measure-ments. In 1988, Structural Integrity, (SI), received permission from GPU Services to reprogram their Fortran computer code for estimating tube life into a LOTUS 1-2-3 spreadsheet, TUBEPRO. EPRI also develop-ed a code, TUBELIFE, which produced essentially the same results as both the GPU and SI codes. This code is now available to all utilities as part of SI's SH/RH tube life evaluation course. A sample output of SI's code (now in Excel) is shown in Figure 1.

SI APPROACH
In 1989, SI approached the Empire State Electric Energy Research Corp. (ESEERCO) with a project that would combine our unique technology for acquiring digital UT signals and our tube life prediction technology into a single piece of software - LIFECODE. Digital UT has many advantages over conventional analog UT. Foremost, the digital signal is stored to disk and can be read later, as opposed to analog signals that must be analyzed while on screen. The operator can analyze the signals stored on disk automatically, or individually. This system provides the most rapid signal acquisition and analysis available for tube life determination.

SI then combined our focused UT technology with the LIFECODE system. The focused transducer, housed in a custom Plexiglas shoe (Figure 2), provides a much stronger reflection of the sound waves from the metal/oxide and air/oxide interfaces, permitting more reliable tube life prediction. Signals are read directly by the computer for remaining tube life prediction, or the operator can manually override the computer. A typical LIFECODE output screen is shown on Figure 3. Many utilities have purchased LIFECODE or have used SI inspection services.

Figure 2. Custom Focused UT Transducer

Figure 3. Typical Output Screen for LIFECODE, Showing the UT Signal, Various Interfaces and Remaining Life Calculation Results

More recently, SI/ESEERCO have added options to the LIFECODE software, including probabilistic tube life prediction and the capability of projecting the effects of chemical cleaning. These new tools provide the user with the potential for making economic, risk-based forecasts of the impact of various tube maintenance strategies.

LIFECODE also contains a customized Excel spreadsheet for viewing the final results as a map of the boiler tubes (Figure 4). This type of information is very valuable to maintenance personnel for planning tube replacement.

Similarly, SI developed a computer code, DMW-LIFE, for ESEERCO to predict damage accumulation in dissimilar metal welds. At the time, radiography and surface resistance testing were the only means available, other than sample removal, for estimating the physical damage in a DMW. In 1997, SI developed a focused UT technique for finding internal damage in DMWs. Used in conjunction with a positioning scanner, a two-dimensional image of the damage can be projected on a computer screen and saved for analysis. To date, SI has had considerable success with the technique and are continuing to improve its accuracy. Figure 5 shows a scan of a DMW with OD, ID and mid-wall defects and Figure 6 shows the metallurgical results from the same tube.

Figure 5. UT Scan of a DMW Containing OD (Upper Left), Mid-Wall and ID (Lower Right) Damage

Figure 6. Metallographic Cross Section Through a Heavily Damaged DMW

The UT scan technique can find mid-wall damage, undetectable by electrical resistance techniques. It is also much more cost effective, accurate and rapid than radiography, where access to the boiler may be restricted during testing.

The damaged welds in Figure 6 were taken from tubes that were cut out of the boiler after UT inspection results showed considerable damage. As can be seen, all tubes exhibit severe OD damage (oxide notching), but varying degrees of mid-wall and ID damage. This is to be expected, as the OD experiences the highest bending stresses during start-ups and shutdowns, and the fatigue portion of damage in DMWs can be very significant.

CONCLUSIONS
Cost of a single tube failure is between $50,000 to $500,000, depending on unit size and dispatch position. SI's tube life prediction services and products save considerable tube replacement dollars and still permit tube replacement prior to failure. A typical inspection of SH and RH tubing and DMWs is just a fraction of the cost of an average tube failure.

SI also supplies equipment to utilities to perform their own testing, if desired. A two-day training course, which covers design, manufacture, life estimation, and risk evaluation of SH/RH tubes is also available (see adjacent typical course outline).

High-Temperature Tube Life Prediction SIM-98-010

High-Temperature Tube Life Prediction
SIM-98-010