Safex Newsletter No.84 July 2026


It is remarkable that we are already approaching the midpoint of 2026. This has been a particularly busy and productive period for Safex International, highlighted by our Congress in Lisbon. We once again extend our sincere thanks to our hosts, Orica, for their support and hospitality. In particular, we acknowledge Amanda Santos from the Lisbon office for her dedication, attention to detail, and significant contribution to making the Congress such a success.

The Congress brought together 184 participants representing 26 companies from across the globe. Delegates returned with a shared message: the event provided valuable learning and reinforced awareness of the safety challenges and objectives facing our industry worldwide. These insights will be shared with our membership as part of our continued commitment to achieving zero harm.

I extend my sincere gratitude to all attendees for their participation, engagement, and support. This was one of the most successful Congresses I have had the privilege of attending.

This Newsletter opens with a message from our Chairman, John Rathbun, who reflects on the Congress and the highly successful CEO meeting.

Andy Begg has brought together contributions from the Expert Panel and individual Associates, resulting in the following insightful articles on incidents and safety management:

Also included are Vignettes of Lisbon and Surroundings, photographed and compiled by Noel Hsu, which provide a visual reflection of the Congress venue and its surroundings.

I trust that you will find this Newsletter both interesting and informative. As this is my final edition as Secretary General of this outstanding organisation, I would like to express my sincere appreciation to every contributor who has supported the Newsletter over the past 12 years.

Piet Halliday , Secretary General

Incident and accident recollection over 30 years experience in the explosives production

Maurice Bourgeois
Expert Panel – Safex International

Much before my time I read an accident report involving 5 casualties that concluded that the operator used a beryllium copper flat screw driver as a chisel to widen a crack in comp B filled munition for a repair.

  • Lesson learned: although beryllium copper screw drivers are non-sparking, their tips are pointed and represents a small area concentrating a very high pressure at impact which replicate the lab impact test in the workshop; non sparking tools give a false impression of safety with explosives but their improper use can be fatal like in this case.

An accident happened during the dismantling of a melter used to empty rejected shells that was washed down many times; comp B ended up penetrating bolt threads possibly because of explosive contaminated water ingress; the mechanic during the unbolting operation put his hand above the wrench to prevent it from slipping; an ignition occurred firing the bolt through his hand, hitting the concrete ceiling knocking off a significant concrete chip; had the mechanic’s head been in the bolt trajectory, there would have been most certainly a casualty. We had been told that cold comp B was hardly sensitive to friction; we weren’t aware of a nut-bolt test indicating through a probit analysis that ignition with comp B was very likely.

  • Lesson learned: Consider a possibly explosive-contaminated bolt as a loaded gun; hence don’t have any part of your body in line with the bolt axis nor with any bystander
  • Use penetrating oil before unbolting possible contaminated bolt threads
  • All bolts susceptible to explosive contamination especially equipment that have to be washed down frequently should be sealed with silicon caulking to prevent explosives entering the threads; maybe not very popular with mechanics but it’s for their safety

Primer cap rejects were put in cotton bags submerged in water for desensitizing during transport and handling in the burning ground. Burning-ground personnel spread the caps in a screen covered pan in a burning silo and submerged them in diesel fuel for destruction; when the accident occurred, the worker used as usual a shovel to pick the shrapnel when a detonation occurred under his shovel because of unburned primer caps. He got shrapnel in his face and suffered hearing loss.

  • The desensitizing liquid hindered combustion; diesel floats on water making it difficult to build heat around the caps; the solution: diesel oil replaced water to assure full destruction of the caps
  • Operators in the cap production and small cal. assembly workshops complained about diesel smell; diesel was replaced with less aromatic fuel (diesel engine filter oil less flammable than diesel but still combustible)

Caping machines with protective enclosures had been operating for ages; one day 2000 caps detonated projecting one of the protective-panels against the wall; fortunately, there was no-one in front of the panel.

  • Because of the relative low explosive load, ignition test in the burning ground were conducted to certify the new reinforced protective enclosure with distance of center of explosives to protective panels recorded in case of future modifications like the introduction of vibrating bowls for which new test would have to be performed because the explosive center distance to the protective panels could be shorter.
  • The caping machine table had to be extended to ensure full enclosure making access to caping feed mechanism a little more cumbersome for which the operators complained.
  • A second detonation occurred and this time the operator was sitting in front of where an opening would have been if the machine table had not been extended; it prevented an accident and the operator was very grateful, hence sometimes despite operator’s backlash, it is advisable to insist that their safety is at risk if the modification isn’t made.

A barrel full with primer-contaminated water was stored for an extended period; when the operator opened the cover there was a deflagration but there was no injury; it was concluded that with evaporation, crystals had formed.

  • Disposal of wastes should be done as soon as possible; when stored and sometimes forgotten, like in this case, they can become sensitized or can spill creating dangerous conditions.

While pouring comp B in grenades, it was decided that a quality depth gauge was no longer required because the cavity depth was constant well within the quality criteria range. It remained on the line, hence not being used no-one was looking for it. Comp B risers were collected for remelt and unfortunately the gauge fell off the workbench into the risers with no-one noticing or looking for it. The operator decided contrary to the SOP to put the risers in the melter mixer without putting them through the metal detector wrongly thinking that risers couldn’t possibly be contaminated with metal (Murphy’s law). All of a sudden during the melt, the mixer impeller hit the gauge stuck in the pouring spout; fortunately, the gauge fell hardened steel nose down into the hole while the aluminum handle was sticking up impacting the impeller; the emergency button  or the over-torque sensor stopped the impeller. The aluminum handle was badly damaged and thank God, no ignition.

  • Lessons learned: When a tool is no longer required, remove it from the workshop because you never know where it will end up. It demonstrates the usefulness of shadow boards a visual sign that all tools are stored in a secure area. Strict compliance with SOP’s is an essential part of safety programs
  • The importance of setting mixer over-torque sensor properly so it doesn’t cause continuous stoppages but if ever a foreign body falls in the mixer, it stops the mixer promptly.

While taking the temperature in a small molten comp B holding kettle, the operator dropped a bimetal thermometer in the kettle jamming the mixer impeller. The over-torque sensor stopped the impeller at first contact

  • Lessons learned: The impeller should be stopped before introducing any object in a comp B mixer which may not have been described in the SOP. This raises the question: How granular or detailed should a SOP be. Sometimes simple motions could be critical. You don’t want to clutter the SOP with large amounts of superfluous details submerging the important ones but a thorough risk analysis checking every possible operation can highlight the ones that can be consequential, like WHAT IF the operator drops the thermometer in the mixer while it’s rotating.
  • The importance of well-adjusted over-torque sensors.

There were two remote granulator explosions in a pyro cubicle. The first time, it was concluded that there may have been a foreign body, so tighter screening control was put in place; the second detonation within a month indicated there was another uncontrolled factor; all granulator components were checked for wear. Finally, it was noticed that the granulator was not level which ensued migration of paste in friction areas where it could accumulate, dry and become sensitive to friction

  • The granulator was properly leveled
  • The cleaning frequency was increased from once a week to the end of each shift.

2100 CFM venturi scrubber fan bearing block became lose because of vibration in the shell melt-pour department. The fan was in a rarely visited remote area. The fan wheel impacted severely the steel casing bending one of the 3/8’’ thick blades. The wheel had some comp B deposits due to crystal formation through evaporation. Also, another possibility could be water seal breakage and carry over due to the scrubber drain open to atmosphere. Fortunately, the fan was AMCA type B with an aluminum wheel.

  • The use AMCA type proved to be a savior. Any evacuation ventilation system susceptible to contamination relying on a filtration system should be at least AMCA type B (aluminum wheel-steel casing) and in extreme conditions type A (Aluminum wheel and casing)
  • Because the fan was in a rarely visited remote area, vibration sensors were installed on bearing blocks and motor bearings which interrupted the fan at the detection of any unusual vibration.