Hazardous location environments carry a high risk of explosion due to the presence of flammable and explosive substances which can range from combustible gas to fine airborne, particulate matter. We are well aware of these risks, as our bread and butter involves the design and production of a broad range of LED luminaires that operate in this unforgiving environment. Our inspiration for innovation sometimes comes out of the most unexpected places. A corporate environment that supports curiosity and fun allows interesting questions to emerge. One such case that captivated us was thinking about the optical properties of ordinary bottled water. We started with the simple question: “Can ordinary bottled water pose a threat in a hazardous location?” The thought occurred to us that bottled water can act as a lens to focus sunlight onto a combustible material located at the focal length. Our curiosity led us to considerations that have important but relatively little known ramifications for LED luminaires operating in hazardous locations.
So just how dangerous is bottled water?
A quick search on Youtube on this topic returns videos such as this that demonstrate how bottled water can be used as a lens to ignite paper. A recent news story from Boise, Idaho reported how a car seat caught fire because a bottle of water left on it, a result later replicated by the local fire department. So it is indeed possible to ignite combustible material with bottled water. The question is, could this pose a threat in a hazardous location environment?
The technical term relevant to this investigation is optical ignition, a pretty self-explanatory term. It is defined as the energy possessed by a stream of photons sufficient to ignite a combustible material and is measured in units of mW/mm2.
A lot of early work in the field of optical ignition was motivated by the use of fibre optics for data communications and process control in hazardous locations. Chemical, petrochemical, pharmaceutical and polymer manufacturing plants use Fourier Transform Infrared (FT-IR) and Four Transform Near Infrared (FT-NIR) to measure polymers to control its polymerization process online. Online analyzers used fiber optics to place probes directly into the process pipe. Fiber optics are used extensively in hazardous process areas due to their immunity to electrical noise. It was assumed that because they were optical and not electrical, they bypassed electrical arcing plaguing electrical equipment. The cables were often run in process areas and through underground vaults with other electrical cables. ISA optical radiation committee SP12.21 chairman Tom Dubaniewicz constructed a lab experiment in which he caused films of coal dust and methane gas to combust when ignited from the optical radiation coming off the end of a broken fibre optic cable, proving the threat is real. Engineers began to realize that if optical fiber broke in an underground vault that was filled with methane gas, a sufficiently high optical radiation radiating off the end of the fiber optic cable could cause ignition. Polishing the fibre while the fiber optic laser was on could be another potential source of ignition. Recognizing the safety risk brought up by these experiments, and the legacy rollout of fiber optic in process areas, from 1995, the CENELEC and IEC proposed safe optical radiation limits of 35 mW and 5 mW/mm2. The EN50303: 2000. Group I, Category M1 standard proposes less stringent values of 120mW and 20mW/mm2. EN 60079-28: 2007 was adopted in Europe. In 2006 IEC released a standard covering optical ignition hazards, since updated in 2015, IEC 60079-28, Explosive atmospheres – Part 28: Protection of equipment and transmission systems using optical radiation. For the purpose of this article, we use the IEC 2015 standard, which adheres to the more conservative 5 mW/mm2threshold.
Back to the bottled water. Bright sunlight has a photon flux of 1 to 2 mW/mm2at the surface of the earth. This alone is insufficient to ignite materials. However, a water bottle that acts as a lens can potentially magnify the sunlight by a factor that scales it close to or beyond the 5 mW/mm2threshold. The figure below is a ray optics simulation using the open source tool found at https://ricktu288.github.io/ray-optics/ with the assumption of beam source for sunlight, a circular glass lens with refractive index of 1.33 (for water). From visual inspection alone, the magnification appears to fall somewhere between 2 and 3. It is therefore conceivable that it could exceed the minimum threshold.