Hi WJ I presume that you've worked your way through the CoSHH requirements for LEV and also the other HSE provided Guidance? Given your comments, it would seem clear that something in the LEV isn't working as required. My first thought is that whilst the LEV may not be showing any inherent problems, it would appear that it is inadequate/inefficient for your production needs - not exactly uncommon! May I suggest that you should be checking air-flows - velocity, air-changes - volume, miminimum capture speed for entrainment of visible and sub-visible particles, and have there been any changes to booths and/or working practices. Frank Hallett

OK WJ - in answer to your question:- You have quoted "suspect burn off fumes from the ovens congregating in pockets dependant on wind direction/ doors open etc" in the original post. Forgive me, but I find this statement to be at odds with your subsequent statement that you've worked through the suggestions offered as wind direction and doors being open should not have any noticeable effects on an appropriately well-designed and installed LEV system. If these things do affect the capture of particles, then by definition, the design and positioning of the LEV &/or structure has to be questioned. Smoke bombs may not provide an accurate picture unless they have been selected to match the size and density of the paint particles to be captured and extracted. Are the 2 colours that you've identified formed of larger or more dense particles; or possibly require more heat or greater duration in the ovens thus altering their characteristics? Given that the problem appears rather intractable, until you've actually fixed the generation of local clouds of paint product particulates [lot's of alliteration there!] outside the LEV, you should provide air-sampling to determine what the quantity, form and type of particulates you're dealing with. If it is genuinely not reasonably practicable to fix this [be very sure about this] then following on from the air-monitoring you will have to decide whether any RPE may be appropriate for this problem. Frank Hallett