A ball valve suddenly broke inside a pharmaceutical plant. The synthesis of a chemical product in the developmental phase required sequencing 2 reactions inside a multipurpose reactor, namely: borohydride reduction, and transformation of the alcohol obtained into acetamide through acetonitrile action in the presence of 70% sulphuric acid. The reactor was being fed via a measurement gauge, and the flow rate was set by the gauge bottom valve (DN 40 ball valve, with a PTFE epoxy body and a ceramic ball). Upon completion of the 2nd reaction, since the acetamide transformation rate was inadequate, the technician decided to add more acetonitrile and sulphuric acid. After pouring a litre, he manipulated the valve in order to reduce what he considered to be an excessive flow rate. An ‘explosion’ ensued in the valve body: concentrated acid spewed onto the technician’s face and right hand, and a piece of the valve was ejected 4 m destroying an ADF fluorescent tube. The technician, who was wearing protective glasses, was washed at the accident scene and taken to hospital. Released to return home that evening, he resumed work after a leave of about 10 days. The initial accident analysis suggested that the borohydride solution, which had been loaded first, filled the dead space (2 x 3.1 cm³) between the ball and the valve body. The addition of sulphuric acid onto the acetonitrile heated the valve channel enough for the concentrated acid to be thrust into this dead space, with the transition perhaps being facilitated by the presence of a decompression hole. A spontaneous reaction, capable of reaching 125 bar, was then triggered causing the valve to break (30 bar). The plant operator noted the efficiency of both the safety glasses worn and first responders’ efforts. The operator also planned on: completing standard loading procedures by allocating separate circuits for supplying incompatible products; strengthening the ‘unlisted situation’ procedure; and having the valve examined by an expert.
