*3.2. Simulation*

A dynamic annual simulation in Matlab/Simulink [57] is used to determine the effects of the volatile steam demand profile on energy demand and costs. The modelling is based on the energy balance of the system, whereby for the technologies the part load efficiency in each operating state and the thermal inertia must be determined. The technical characteristics of the models are described in Section 2. The Runge-Kutta method with a variable step size is used to solve the energy balance. In the energy model, the main utility and the backup boiler are connected in series, each with an internal storage component. The stored energy is calculated from the masses of water and steam and the respective enthalpies at the given pressure. To determine energy demand, costs and *CO*2,*e*, the operating point of the utility is determined at the current output. On the basis of the operating point, the efficiency is determined, which leads to the energy demand for the thermal output. The absolute costs and *CO*2,*e*-emissions are calculated using the current values of the energy carrier demand and

the related costs and *CO*2,*e*-emissions. Energy losses and restrictions in the start up time or part-load performance of the different utility systems are considered in the modelling.

Due to the strongly fluctuating demand profile of the considered application, the steam storage size is optimised. The accumulator is dimensioned based on the average maximum overload and the maximum overload. These measures aim at offsetting all surplus and overload intervals by analysing all recorded intervals for one year. For all concepts including a steam accumulator, it is charged if there is a surplus of produced steam and the pressure in the storage is below 13 bar. It is discharged, if the produced steam of steam supply utility is less than the demand and the pressure in the storage is higher than 9 bar. In case the total produced steam by both one steam utility and the accumulator is less than the demand, the backup boiler supports.

For the electricity provided by the FC or MGT, both a cost-oriented and an emission-oriented variant are investigated. For this purpose, the costs and emissions of the generated electricity are allocated via the Finish Method. These values are compared to the current market values. Whether the electrical energy generated is used for own consumption or sold and fed into the grid depends on the aforementioned trade-off between minimising emissions or costs. The corresponding process type of electricity generation results in accounting effects for the recorded *CO*2,*e*-emissions as well as economic effects for sale and purchase of electricity.
