The science of acoustics touches several areas of our existence. For example, engineers have used acoustic monitoring to identify fatigue crack propagation in aircraft fuselage and bridge structures. Specific to electric power generation, acoustics has earned a role in boiler tube leak detection.

Boiler Tube Leak Detection

Boiler tube failures (BTFs) consistently rank as a leading cause of forced outages among fossil fuel power plants, impacting unit availability by around 3% worldwide. BTF-induced high pressure steam leaks frequently causing collateral damage as broken parts cut through adjacent tubing. Repairs can be extensive and peak-hour downtime can result in high replacement power costs (sometimes exceeding $1 million per day). Traditional online tube leak detection methods (including makeup water use, opacity and temperature monitoring) cannot identify the area of the leak and are usually limited to identifying large leaks that already may be causing damage. Moreover, small undetected leaks are likely to go unrepaired during outages.

Illustration of a 700 MW unit with 46 individual sensor locations. Image source: Mistras GroupIllustration of a 700 MW unit with 46 individual sensor locations. Image source: Mistras GroupAcoustic monitoring has become an essential part of early tube leak detection for utility boilers. Non-invasive sensors that mounted on the boiler wall listen for changes in the normal background noise of the boiler that indicates a steam leak. One such sensor, the MISTRAS | Triple 5 AMS system, provides continuous real-time data that tracks the progression of a tube leak. It helps to indicate a leak sooner than traditional methods and also helps plant operators locate the leak based on sensor location and signal amplitude.

The Science

Tube leaks emit sound waves in the form of white noise consisting of all frequencies. The sound frequencies can be attenuated as they travel through the boiler gasses and steel tubing pathways. Noise also reverberates through the boiler walls, sounding rods and sensors. Sensors absorb these sound waves and send their corresponding signatures to a monitoring system where the raw sound data is processed and stored. This sort of leak detection and warning system is intelligent in the sense that it amplifies both metal-borne and airborne frequencies that depart from normal operational background sounds such as sootblowers, valves and rotating equipment. The detection system continuously measures boiler internal noise and signals alarms when sounds exceed a preset threshold for a predetermined amount of time. Data trends are used for visual comparison where “noise signatures” are analyzed and identified.

Business Justification

Detailed acoustic information and early tube leak detection provides plant operators, maintenance leaders and fleet managers an opportunity to mitigate potential lost revenue from a tube leak event. This enables the following:

  • Improved outage planning efforts through knowledge of leak location and pro-active resource mobilization
  • Informed business decisions to schedule repairs based upon the fleet availability and the energy market
  • Lower power replacement costs
  • Collateral tube damage control
  • Multiple leak detection capability
  • Focused boiler inspections and scheduled repairs during planned outages, preventing future forced outages.

A new application involves the recent promulgation of the U.S. Environmental Protection Agency’s Maximum Achievable Control Technology Standards (MACT) regarding utility boilers. Regulations require hydrochloric acid, sulphur dioxide and mercury to be mitigated to low concentrations. Most existing air pollution control systems may not be sufficient to satisfy the standards. Direct dry injections of activated carbon and/or sodium bicarbonate (Trona) are lower-cost solutions that are being applied to meet these new challenges.

However, plugging of the feed pipes for these processes can directly impact environmental emissions or, in the case of coal pipes, lead to safety and fire concerns. That is because the dry particles strike the inner walls of the transport pipe as they travel inside the pipe. This creates acoustic noise, which can vary with the quantity, speed and weight of the particles.

Acoustic sensor locations illustrated with brief descriptions. Image source: Mistras GroupAcoustic sensor locations illustrated with brief descriptions. Image source: Mistras GroupTo address this, Mistras Group developed an acoustic monitoring system dry particle module to measure acoustic activity generated by the flow of suspended process particles in transport pipes. Acoustic trends are generated along with specific plant that provides input parameters. Trends for each pipe are tuned to a baseline during the transport system’s normal operations. With this information, the system can detect and alarm loss of flow and provide the user a general balance of flow indication between each transport pipe.

Although acoustic monitoring originally was used for power boiler tube leak identification, it is now delivering value elsewhere among power plant equipment. Applications have been successful in detecting steam leaks in other pressurized vessels to include recovery boilers, feedwater heaters and heat recovery steam generators. Moreover, systems have been expanded to include coal flow, valve operation, pulverizer mechanicals and temperature monitoring. Once baseline sound signatures on “healthy” equipment or “normal” processes are gathered, trended and studied, subsequent monitoring can provide value when abnormalities arise. When irregularities occur against preset targets or thresholds, it may be time to investigate.

As the industry recognizes the value of condition-based maintenance, the potential applications of acoustical monitoring continue to expand.