DIN CEN/TS 17198:2018-10

Predictive emission monitoring systems (PEMS) are used for continuous monitoring of emissions at stationary sources as an alternative and/or backup for automated measuring systems (AMS). PEMS define the relationship between a number of characteristic process parameters of an emission source and the corresponding emission concentration. If the characteristic process parameters are continuously monitored it is possible with a PEMS to continuously determine the emission concentration of the emission source. PEMS are deduced using process data and, where required, emission data from the emission source. Using these data the PEMS is modelled. PEMS contains an emissions model and a sensor validation system for quality assurance of incoming process data. PEMS are plant-specific emission monitoring systems and vary as regards to methodology and design: - relational models (emission concentration as a function of one or more process parameters): theoretical or empirical relations that are fitted to a plant-specific emission data set; - nonlinear statistical models, e. g. neural network models, or other multiple regression techniques. This Technical Specification gives requirements for the certification of PEMS software and for the performance and quality assurance for a PEMS to prove suitability for its measuring task and to ensure that the PEMS continues to perform within the specified performance during operation of the PEMS.

13.040.40