(301m) A Computer Aided Approach to Hazop Based on Functional Models

Ensuring the safety of process operations may be significantly facilitated through the usage of functional models. In this presentation a functional approach to HAZOP studies in which the plant is divided along strictly functional lines starting from the purpose and function of the plant and continuing down to the function of a single node or pipe section is presented. The approach leads to nodes with basic functions such as liquid transport, gas transport, liquid storage, gas-liquid contacting etc. From the functions of the nodes the relevant process parameters and deviation variables for a HAZOP study follow directly. Thereby a foundation for development of a systematic approach to a computer asisted HAZOP study is provided. The knowledge required to perform the pre-meeting task of dividing the plant along functional lines is that of chemical unit operations and transport processes plus familiarity with the plant a hand. Thus the preparatory work may be performed by a chemical engineer with just an introductory course in risk assessment. In any given plant there will be several nodes with the same function, e.g. gas transport. These nodes are analysed together, which removes some of the repetitive aspetcs of a HAZOP meeting. The approach is exemplified by a HAZOP study of an industrial scale heat integrated distillation pilot plant at the Dept. of Chemical Engineering. The purpose of the plant is to separate a binary mixture into pure products using the least amount of energy. This plant is divided into 8 sections, which are further divided into 20 nodes. Typically in this pilot plant a section is divided into 1-5 nodes. Several of the nodes have a similar function, e.g. transport of liquid. A 5 step functional modelling based procedure to HAZOP is introduced, which allows any chemical engineer to contribute meaningfully to a HAZOP study. The approach reduces the work involved in a HAZOP of a plant by dividing the plant along functional lines and analysing nodes with the same function ones. This appears to significantly reduce the manpower effort involved. Furthermore functional models of chemical plants are demonstrated to constitute a useful approach for development of a systematic methodology for computer assisted HAZOP analysis. On a simple model of a part of a pilot plant the computer assisted HAZOP is shown to find the same causes as a traditional HAZOP study. It is furthermore demonstrated, that a computer assisted HAZOP indeed is able to find causes far from the site of the deviation.