Abstract: Paper pulp mill chemical recovery boiler steam piping is monitored for leaks by an acoustic leak detection system responsive to atmospherically carried sound transmissions. Energy level amplitudes of respective analog electrical signals generated sequence multiplicity of microphones are converted in a rapid time sequence to a first electric pulse signal sequence representative of corresponding digital values. Such digital values are temporarily stored in a data memory bank to be subsequently processed by a Fast-Fourier Transform spectral analysis into a second electrical pulse signal sequence representative of corresponding digital values in the amplitude vs. frequency domain. Such amplitude vs. frequency signal values respective to each microphone are further refined for comparison to historical threshold data for a probability determination of a leak status.

Abstract: Paper pulp mill chemical recovery boiler steam piping is monitored for leaks by an acoustic leak detection system responsive to atmospherically carried sound transmissions. Energy level amplitudes of respective analog electrical signals generated by a multiplicity of microphones are digitized in a rapid time sequence. Such digital values are temporarily stored in a data memory bank to be subsequently processed by a Fast-Fourier Transform analysis into amplitude vs. frequency domain data. Such amplitude vs. frequency data respective to each microphone is further refined for comparison to historical threshold data for a probability determination of a leak status.