Advanced Condition Monitoring - Steam Trap Dynamic Analysis
by Raphael Danjoux, PhD.
ITC France
What is a steam trap and how does it work? Why does it fail? All these subjects are thoroughly documented on the internet. Because steam traps affect the performance of whole systems when they fail they require regular inspections. Thermal imaging with thermal imaging cameras is a great way to predict failure and plan ahead to avoid costly setbacks.
How do you use a thermal imager to detect a steam trap fault?
According to A. Bandes and B. Garelick (in Inspect Steam Traps for Efficient System). Faulty or inoperative steam traps can cause losses of hundreds of thousands of dollars. Proper thermal imaging inspection relies on upstream/downstream temperature variations in a trap.
In other words, you need to store and further analyze a sequence to get full data. FLIR’s thermal imagers allow you to do this. For now, most inspections rely on static analysis of a single thermogram. First of all, few cameras are equipped with such a feature (S60, S65 and others in option). Then, although sequences can be displayed in ThermaCAM Reporter 7.0, the software does not allow for time plotting. Reporter 2000 had a trending function that could do it on a series of single images but not on a sequence. It’ss best to use ThermaCAM Researcher!! Here's the basic methodology:
Place an analysis area on the output pipe of the trap.
Choose an analysis function, for instance a rectangle, and place it on the thermogram. We will record the average time vs. temperature data for this area. In the case of the steam trap, the initial signal is rather noisy, so a linear smoothing with a (5X1) operator gives a better result (this can be performed after the data is exported to Excel).
A graph of the smoothed data after exporting to Excel
For the duration of the acquisition the mean value is 120.26°C, the amplitude (peak to peak) however varies a bit. This latter may be due to all external factors including; condensate load, steam pressure variations, or probably the wind in our case. The period, however, looks rather stable: about 4.2 seconds. The fact that the trap is cycling provides good evidence that it is operating correctly.
A trap that is not cycling may have failed open, failed closed (in this case the temperatures may be significantly below saturated steam temperature at the outlet, or may be undersized for the condensate load (in this case the temperature upstream of the trap may have a lower than expected temperature). - Editor
A second sequence was grabbed with 4 steam traps in the field of view. Since there was no tripod used, a slight movement between each picture taken with the thermal imager was noticeable. The complete results, however, are equivalent to those presented before, and only one thermal image shown here.
A gang of steam traps.
Conclusion
Considering the steam pressure, the fluid temperature, the type of equipment (discharge volume & duration), and the production settings, etc, it should be possible to define for each steam trap a standard set of working factors for a reference baseline. Then, monitoring would allow for tracing a trending curve. This would allow for\ condition based maintenance and predictive maintenance.