Multisensor data fusion for safety instrumented systems as optimization of online-based safety models for the protection of chemical reactors
SmartHIP (Smart Technology for High Integrity Overpressure Protection) is a CSE research project for the investigation of protection of chemical reactors with mere safety instrumented systems (SIS) by online-based safety models. In comparison to mechanical safety devices like safety valves or bursting discs, no emission will occur (zero emission) and downstream plants are no longer required (i.e. flares or cyclones). Thereby, the acquisition and operating cost can be reduced extremely.
Current online-based safety models were studied and it was shown, that these models can represent badly the trajectories of the chemical process, due to their simplifications and required data of the chemical components. Therefore, new models, which are based on the measurement of speed of sound or density, were developed, which were presented at the 4. CSE-Sicherheitstage on Wangerooge in April 2018.
To improve the accuracy and the availability of the online-based safety models, a combination of these different models is strived. Therefore, a huge amount of produced data must be interpreted, meaningful conclusions must be drawn from the multitude of observations. For this purpose, multisensor data fusion should be used, whose basic concept was presented 2017 at the Collaborative European Research Conference (CERC) in Karlsruhe and at the European Conference on Process Safety and Big Data (CCPS) in Frankfurt.
Currently, the project SmartHIP is following up with transferring multisensor data fusion to safety technology and online model-based reactor safety system. Various aspects, such as the implement ability in an SSPS, the traceability of the algorithms used must be taken into account as a condition for the certification as safety technology.