Abstract: On June 10, 1998, Employee #1 and a coworker were trying to release clogged battery material from the ram cavity on the #3 reverb furnace. Ram cavity jams occurred occasionally and required an employee to put his upper body into the cavity and physically chip away at the clog. The clog consisted of old battery sections that were normally used to charge up the furnace, which the heat at charge-up had melted into one mass. According to operators of the equipment, the normal cycle time for the ram was between 60 and 90 seconds, but less than 30 seconds for a complete cycle when in the MANUAL mode. The area of the ram was extremely noisy, and it was not possible to hear the ram in motion. At the time of the accident, Employee #1 and a coworker had been working on clearing the ram cavity for two hours. They were later assisted by two supervisors. At the bottom of the cavity was a hinged door, which had to be open when employees were in the cavity. Employee #1 and the coworker were alternating chipping out the clogged battery material, with the ram fully retracted behind them. They used the ram and wooden railroad ties to push the material through the cavity after they had chipped away at it for a period of time, but the ram caused the wood to break against the clog. They tried unsuccessfully using a chain fall to pull the clog out. They also tried five or six times to push the material through with the ram alone, but without success. Employee #1 then went back into the cavity to chip some more, while the coworker remained at the controls on the other side of the machine and out of the line of sight. Employee #1 had just gotten out of the cavity and the ram was cycled once before he re-entered to resume chipping. The coworker then received a call on his radio for assistance nearby, and he left the control panel. Approximately one minute later, the ram apparently cycled on its own, decapitating Employee #1. One of the supervisors called to the coworker to shut off the ram and he rushed to do so, but even with repeated attempts the ram did not stop until it was fully extended. The ram operated at a pressure of approximately 1 kpsi. To operate the ram manually, both switches had to be in 'manual' mode. In normal operation, both switches would be on 'auto'. During the cleanout process ongoing at the time of the accident, the switches would typically be in the 'off' position.