5.6 Prototypical Implementation
151
In summary, the application of the optimization methodology on the fictional
use case can be described as successful. The use of different simulation techniques was proven to be possible. The material flow of the production line was built
following a process-oriented DES logic, while the internal logic of the machines
was created using ABS and then combined with the energy simulation in SD.
Thus, the dynamic depiction of the energy consumption behavior became possible, being able to get an exact consumption value for any point in time. The
reference scenario reflects the current behavior in production, which is basically
to make no efforts to bring machines into a less energy-intensive state during
production interruption periods. By introducing the idle and standby optimization
parameters, production employees receive precise information as of which interruption length a change of the machine states makes sense and saves energy. The
determination of exact starting times of machines that are used later in the production process opens up the possibility of making production more energy efficient.
In addition, the offset parameter gives an exact recommendation for action defining which offset for an energy-intensive machine is required in order to prevent
energy consumption peaks.
The use of the optimization methodology has shown that significant savings in
total energy consumption have been achieved compared to the reference scenario.
For the consumption peaks, a slight improvement could be achieved by using the
offset parameter. It is believed that the peak optimization provides better results
when a larger area of production is considered in the simulation model. The optimization methodology was also tested for the use with mean values instead of
exact energy load profiles. While the optimization of the total energy consumption has delivered results in similar high quality as for the exact load profiles, the
peak optimization was not possible under these circumstances. In addition to the
settings and setups tested in this section, the fictional case model was also tested with varied delivery schedules, machine settings, and production parameters
in further simulation experiments. The results have been evaluated by comparing
the reference and the optimization models as discussed before. Since the results
were similar to the extensively described setup, further explanations are omitted
here. The next section summarizes gained findings to consider peculiarities for
the practical application of the methodology.
151
In summary, the application of the optimization methodology on the fictional
use case can be described as successful. The use of different simulation techniques was proven to be possible. The material flow of the production line was built
following a process-oriented DES logic, while the internal logic of the machines
was created using ABS and then combined with the energy simulation in SD.
Thus, the dynamic depiction of the energy consumption behavior became possible, being able to get an exact consumption value for any point in time. The
reference scenario reflects the current behavior in production, which is basically
to make no efforts to bring machines into a less energy-intensive state during
production interruption periods. By introducing the idle and standby optimization
parameters, production employees receive precise information as of which interruption length a change of the machine states makes sense and saves energy. The
determination of exact starting times of machines that are used later in the production process opens up the possibility of making production more energy efficient.
In addition, the offset parameter gives an exact recommendation for action defining which offset for an energy-intensive machine is required in order to prevent
energy consumption peaks.
The use of the optimization methodology has shown that significant savings in
total energy consumption have been achieved compared to the reference scenario.
For the consumption peaks, a slight improvement could be achieved by using the
offset parameter. It is believed that the peak optimization provides better results
when a larger area of production is considered in the simulation model. The optimization methodology was also tested for the use with mean values instead of
exact energy load profiles. While the optimization of the total energy consumption has delivered results in similar high quality as for the exact load profiles, the
peak optimization was not possible under these circumstances. In addition to the
settings and setups tested in this section, the fictional case model was also tested with varied delivery schedules, machine settings, and production parameters
in further simulation experiments. The results have been evaluated by comparing
the reference and the optimization models as discussed before. Since the results
were similar to the extensively described setup, further explanations are omitted
here. The next section summarizes gained findings to consider peculiarities for
the practical application of the methodology.