NEPTUNE: Above ground Storage-Tank Inspection Robot System

Abstract

The Neptune system is a mobile robot system used to remotely inspect above-ground storage tanks (ASTs) with petroleum product in place, in order to ascertain the state of corrosion of the floor and side-walls using video and ultrasonics, according to the guidelines laid out by the American Petroleum Institute (API). The robot system allows unmanned entry and sensor data collection in ASTs without the need to empty or clean the tanks nor the required human walk-through inspection which results in a very sparse data set from which the tank's state is statistically extrapolated. The complete system is comprised of (i) a specially designed robot crawler vehicle suitable for classified locations which carries visual and ultrasonic sensors, (ii) a deployment pod atop the tank to lower and retrieve the crawler, (iii) an in-tank acoustic positioning system to chart and control the location of the robot, and (iv) an external remote control console utilizing commercial and custom software for display, planning and control tasks. The crawler, pod and navigation systems are all designed for certification in hazardous environments, classified as Class I, Division 1, Group D by such agencies as the National Electric Code (NEC) and the National Fire Protection Agency (NFPA). The system will be completed by the fall of 1994, demonstrated in a water tank, and then submitted to a certification agency such as Underwriters Laboratories (UL) or Factory Mutual Research Corporation (FM) for testing and certification. Initial tank inspections are anticipated to begin in the summer of 1995. This article describes the design aspects of the system which is currently undergoing testing and evaluation. Emphasis is placed on the overall system configuration and the approach used to obtain certification for classified environments, and other newly developed and engineered mechanical and electrical solutions to meet the stringent environmental requirements of such an inspection task.