Fire detection in industrial peanut and cottonseed warehouses: determining pre-combustion and combustion gas species and measuring gas flow dynamics in scale model warehouses

Authors: McIntyre, Joseph; Holt, Gregory; Farley-Talamantes, Erik; Pelletier, Mathew; BUTTS, CHRISTOPHER - Retired ARS Employee; Lamb, Marshall; Todd Jr, Kenneth; Cook, Hunter

Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 5/11/2026
Publication Date: N/A
Citation: N/A

Interpretive Summary: Abstract Only

Technical Abstract: Introduction Almost annually, a catastrophic fire occurs at an industrial peanut or cottonseed warehouse where conventional smoke detectors perform poorly due to the unique and harsh conditions inside the warehouses. Peanuts and cottonseed are oily fuel kernels wrapped in cardboard or cotton lint. One such peanut warehouse fire consumed 23 Gg (25,000 US tons) of peanuts. Tens of millions of dollars of losses occur in a single warehouse fire making insurance prohibitively expensive or unavailable. Warehouse fires often begin as smoldering combustion within the mass of stored material. The smoldering fire can be ignited by a variety of sources; biological activity (spontaneous combustion), mechanical system failures, or hot work. Peanut and cottonseed warehouses combine the problems of smoldering in porous fuel beds and industrial warehouse fire detection. The unburned porous fuel filters smoke and insulates the heat of smoldering [1]. Smoldering combustion in porous fuel beds can give little hint of its presence until close to the surface [1]. The low heat rate of smoldering produces a weakly buoyant smoke plume [1]. Sensors that detect smoke particulates and combustion gases require the concentration of those particles to exceed the detection threshold for the sensors. The active ventilation of industrial storage facilities requires careful placement of smoke/gas detector systems [2]. The insulation of subsurface smoldering interferes with the performance of heat and radiation detectors in addition to the typical problems with warehouse internal structures blocking detection [2]. Methodology Peanut warehouse detection system development began with an airborne combustion emission assay of farmers’ stock peanut materials conducted at the Missoula Fire Sciences Laboratory Large Scale Combustion Chamber [3]. The peanut materials were placed in expanded steel grating burn baskets forming a cube with 30.5 cm long sides. A brass cylinder containing an electronically controlled cartridge heater placed in the center of the volume. The peanut materials were heated at varying rates with pauses at selected cylinder surface temperatures as gas sensor data was collected until flaming ignition occurred. Instruments sampling flow from the combustion chamber chimney measured formaldehyde (H2CO), formic acid (CH2O2), carbon monoxide (CO), carbon dioxide (CO2), sulfur dioxide (SO2), hydrogen cyanide (HCN), total reduced sulfur (TRS), and aerosol particles of 2.5 microns or less (PM2.5). The heated peanuts produced significant smoke while heating to flaming ignition. H2CO, CH2O2, and HCN emissions produce a distinct gas signal that farmers’ stock peanut biomaterials are being heated near to or have begun smoldering combustion. H2CO, CH2O2, and HCN are not produced during normal operation of commercial peanut storage facilities and therefore may make good candidates for trigger chemicals for peanut warehouse fire detection systems. To determine the flow dynamics of pre-combustion and combustion gases a scale model traditional peanut warehouse, 4.2x2.45x2.4m(14x8x8ft), and a scale model flat-storage warehouse, 4.2x3.7x1m (14x12x3.4ft), at the ARS National Peanut Research Laboratory were instrumented to measure carbon dioxide concentration, humidity, and temperature at several locations within the stored peanuts [3]. The carbon dioxide sensors measure over a concentration range of 0 to 10k ppm carbon dioxide with an accuracy of ±40 ppm. Carbon dioxide was then injected at a controlled rate of 100 mL/min through aeration stones to simulate a buried pre-combustion carbon dioxide producing region within the stored peanuts. The carbon dioxide concentration was measured with the scale warehouse ventilation off and at airflow rate of 0.57 m3/min (20 CFM). Carbon dioxide concentrations were monitored for two days with ventilation off