Energetic material deposits on scrap metal at military training and testing installations range from trace surface residues to bulk explosive formulations, including partially detonated or complete unexploded ordnance (UXO) items. The current practice of visually screening large amounts of scrap materials is inefficient and has misidentified mass detonating quantities of explosives, causing significant equipment losses and injuries to personnel during scrap recycling operations. The concept of an automated screening process using chemical sensing technology has the potential to provide an efficient low-cost method to discriminate scrap metal containing mass detonating quantities of energetic materials before entering a treatment process and to verify decontamination of surface residues after treatment.

The objective of this project was to determine if a chemical sensor could identify mass detonating quantities of explosive formulations and UXO in scrap materials.

Technical Approach

Environmental restoration technology has been demonstrated that can reduce the volume of contaminated soil by sorting on a conveyor system linked to a sensor array that directs clean soil to one pile and contaminated soil to another pile. The sensor requirements for applying this concept to range scrap include high sensitivity and selectivity for the chemicals found in the energetic materials and fast response time to allow real-time sorting. Mass detonating quantities of energetic materials are of principal concern for safety during treatment of scrap metal. A minimum critical diameter or thickness is necessary for detonation to propagate through the energetic material.


A novel sensor engineered by Nomadics Inc. from a sensor developed in the Defense Advanced Research Projects Agency (DARPA) Dog's Nose program was evaluated by measuring and comparing the absolute sensitivity of the sensor in the laboratory to vapors derived from water solutions. The vapor-sensing threshold for the chemical sensor was determined to be 10 to 20 parts per trillion (ppt) for TNT and 150 to 200 ppt for DNT. Bin tests were conducted by loading scrap metal and various sized pieces of energetic materials into bins for a sufficient time to establish a steady-state vapor concentration. Headspace vapors then were sampled and analyzed with the sensor followed by active vapor sampling for comparison with the sensor value. In addition, non-mass detonating quantities of explosive residues found on artillery/mortar packaging were placed into bins with scrap metal. The magnitude of the sensor response was compared with values obtained from bin tests with mass detonating quantities of energetic materials. Field tests with the sensor demonstrated the proof-of-concept that energetic material residues can be identified with vapor sensing in enclosed scrap bins. Items such as low-order detonation debris, demolition block granules, and unused 81-mm mortars were detected quickly and with minimum effort.


This project established that mass detonating quantities of energetic materials can be identified by vapor signature in static bin headspace with the Nomadics sensor. Conceptual designs for field-screening scrap metal for energetic material residues include handheld vapor sensing systems, batch scrap sensing systems, continuous conveyor sensing systems, and a hot gas decontamination verification system. (Project Completed – 2002)