ADVANCED BOILER TUBE LEAK DETECTION

LeakTrac^TM is a collaborative effort between Automation Applications Inc, LLC and Hercules Pulp & Paper Division (formerly BetzDearborn Inc). This state-of-the-art system was designed with the combined expertise of Hercules, Automation Applications Inc, LLC and Union Camp Incorporated. The LeakTrac system combines all of the most successful industry methods of detecting boiler tube leaks and integrates them into one extremely accurate system.

Of the numerous leak detection schemes applied in the past, three types are currently being used today:

• Acoustic
• Conductivity
• Mass Balance

Each approach has strengths and weaknesses.

The acoustic detection method uses sensors and software to detect tube leaks much the same way as operations personnel. While this can be a reliable method, the software must learn which sounds are considered leaks and which are considered normal operating sounds. Until the software is properly trained, this method can be susceptible to false tube leak indications and alarms.

The conductivity detection method is based on accurate measurement of known concentrations of substances that do not escape when water is turned into steam. When a change in concentration is recognized, this can be attributed to a boiler tube leak. This method is capable of identifying very minute leaks, but usually requires twelve to twenty-four hours for detection.

A water mass balance has most often been used to detect tube leaks, but in the past it has been nearly impossible to find small leaks. Using mass balances is difficult because of the accuracy of the flow measurement devices and the considerable amount of process noise which is normal. With this difficulty, the method has either produced a significant number of false alarms or the alarm limits are relaxed to the point of being able to detect only large leaks.

With the use of moving averages, rule based logic, and statistical methods however, the mass balance method does not require ultra-accurate flow measurements. All that is required is repeatability of the flow measurements, which is normally obtainable. With this concept, small changes in the boiler's mass balance can be separated from process noise within a reasonable amount of time, and the probability of a tube leak is easily detected.

AAI and Hercules have extensively developed techniques for obtaining a meaningful value for changes in boiler mass balance averages and have also incorporated patented statistical algorithms which reliably calculate these deviations. These methods are packaged in the AAI/Hercules Boiler Leak Detection System. The LeakTrac System is a software based solution which runs in a computing module of the boiler Distributed Control System (DCS) or PC environment. LeakTrac utilizes instrumentation which is typical of a Boiler Process as inputs to the software programs. Operator interface to LeakTrac is provided through graphic displays at the DCS control console.

The LeakTrac Detection System has been proven to effectively identify tube leaks in short periods of time. To some extent, system tuning will determine how quickly an event is brought to the attention of process operators and at what level a leak will be alarmed. An example of system performance is an 8-10 gpm leak being identified in twenty minutes.

An advantage of the mass balance technique for leak detection is that the system can be tested through simulation of a leak by opening a blow down valve. For more information please contact Automation Applications Inc. or your Hercules Pulp and Paper Account Representative.