Damage detection in hazardous waste storage tank bottoms using ultrasonic guided waves

Detecting damage in storage tanks is performed commercially using a variety of techniques. The most commonly used inspection technologies are magnetic flux leakage (MFL), conventional ultrasonic testing (UT), and leak testing. MFL and UT typically involve manual or robotic scanning of a sensor along the metal surfaces to detect cracks or corrosion wall loss. For inspection of the tank bottom, however, the storage tank is commonly emptied to allow interior access for the inspection system. While there are costs associated with emptying a storage tank for inspection that can be justified in some scenarios, there are situations where emptying the tank is impractical. Robotic, submersible systems have been developed for inspecting these tanks, but there are some storage tanks whose contents are so hazardous that even the use of these systems is untenable. Thus, there is a need to develop an inspection strategy that does not require emptying the tank or insertion of the sensor system into the tank. This paper presents a guided wave system for inspecting the bottom of double-shelled storage tanks (DSTs), with the sensor located on the exterior side-wall of the vessel. The sensor used is an electromagnetic acoustic transducer (EMAT) that generates and receives shear-horizontal guided plate waves using magnetostriction principles. The system operates by scanning the sensor around the circumference of the storage tank and sending guided waves into the tank bottom at regular intervals. The data from multiple locations are combined using the synthetic aperture focusing technique (SAFT) to create a color-mapped image of the vessel thickness changes. The target application of the system described is inspection of DSTs located at the Hanford site, which are million-gallon vessels used to store nuclear waste. Other vessels whose exterior walls are accessible would also be candidates for inspection using the described approach. Experimental results are shown from tests on multiple mockups of the DSTs being used to develop the sensor system.