Abstract: Employee #1 and a coworker were assigned to dump slag pots. Some of the pots contained loose fly ash and bricks, and some contained flash smelter slag that had been slow cooled. The slow cool process requires that pots sit in the open yard for five days after they have been poured. They are not supposed to be dumped before that, and are to be manually checked for heat before dumping. Employee #1 was assigned to operate a trackhoe that had been equipped with a rock pick, while the coworker was assigned to drive the slag pot hauler. The cooled slag in the pots would sometimes stick inside the pot, and the hauler would turn the pot upside down. If the operator could not get out he cooled slag in the form of a solid "button," the trackhoe was used to tap the pot to loosen the button and get it to fall out. The coworker checked the first two pots by hand, found them to be cool, and dumped them without incident. He then checked the next two pots by hand, putting the slag hauler between them and climbing out of the cab. He felt the top rim of the two pots and did not sense any heat, so he picked up one of the pots to take it to the dumping area. The dumping area is normally wet from the water spray used to cool blister copper slag pots, and from water that accumulates during storms, snow melt, etc. When this pot was turned upside down, the crust that had formed on the surface broke, allowing molten slag to spill out. The slag fell into a puddle of water, which exploded in steam. Molten slag was blown over 20 ft into the air, and horizontally toward the trackhoe, which was in position to tap the pot if necessary. The molten slag blew into the cab of the trackhoe, burning Employee #1. It caused combustible material in the cab, such as the seat cushions, to ignite; the tires on the slag pot hauler also caught on fire. Employee #1 sustained second- and third-degree burns over 50 percent of his body. Employee #2, who was nearby, came to help and saw that a smaller backhoe had started to burn. He used a fire extinguisher to put out the flames and then moved the vehicle from the area. Employee #2 was treated and released for minor smoke inhalation and a related illness. The coworker was able to escape without injury. He had not checked the log book at the operator's station approximately 100 yd from the scene, to verify the date and time the pot had been brought to the area to cool. Subsequent investigation determined that the pot was placed at the site at 10:38 p.m. on December 28, 1998, meaning it had cooled for approximately 33 hours before the accident. The procedure used by the coworker to hand check the pot for heat before dumping was not adequate to detect the molten material inside.