**5. Discussion**

The discussion of the results is conducted in two areas: the first area was the analysis of the quantitative results obtained from the simulation experiments; and the second area was the evaluation of the proposed method in the context of applications for manufacturing companies and applications for conducting further research of the problem.

In the first area, a simulation model of the problem was developed in FlexSim software. All practically important elements of the problem concerning the managemen<sup>t</sup> of power and electricity consumption in the investigated production process were represented in the model. Research scenarios were prepared, adequate to the needs of solving the problem. Simulation experiments were carried out according to the assumptions. Each experiment provided the expected quantitative results. Analysis of the results showed that the best results in terms of managing power consumption when a constraint occurred were obtained in scenario 3. In this scenario, production was subordinated to the bottleneck of the process. By increasing the throughput of the process and continuity of material flow (continuity of work) at the bottleneck, the lowest electricity consumption in the whole system was obtained (Table 9).

When limiting the available power by 20 kW, a stoppage of the production process was obtained in every scenario studied. All machines were running in the "Idle" state (i.e., waiting for the possibility of carrying out operations). The question to be analyzed was whether it is worthwhile keeping the machines in the "Idle" state. It is also possible to switch machines to the "Power Off" state. Then, when the restriction is removed, all machines must be restarted. The answer to this question depends on whether the electricity consumption in the "Warm Up" state covers the electricity consumption in the "Idle" state. In addition, it has to be assessed whether stopping and starting the machines increases their running costs and makes them wear out faster. In addition to electricity consumption, the need for increased overtime costs for employees must also be considered. These issues should be considered together when deciding how to respond to a reduction in an available capacity for a production process.

In the second area of discussion, it can be concluded that the proposed method was successful. The required data were collected, models of power consumption by the enterprise resources were prepared, experiments were carried out, and practical conclusions were formulated. However, it should be noted that for the analyzed production process, the authors of the paper managed to adequately prepare the research environment.

From a practical point of view, it is the ability to prepare the research environment within the enterprise that determines the ability to manage power and electricity consumption. For the company, this means that good internal preparation is required beforehand. Above all, the machinery and equipment must be metered. For decision-making, the measurements must be available virtually online. For possible long-term analyses, the measurements should be stored in databases, which requires that the company has qualified personnel. In addition, the company must have computer programs for simulations. Such programs require the purchase of licenses and the training of personnel to build models and conduct simulation experiments, which are costly and time-consuming activities. Other elements, indirectly related to the analyzed problem, are the type of manufactured products, the industry in which the company operates, and the technologies used. Not in every case is it possible to take sufficiently flexible actions to be able to react to problems related to the limitation of available power or the high prices of electricity.

On a general level, to mitigate these challenges, manufacturing companies can take the following actions:


Implementation of the above activities requires special care and the assessment of risks associated with their operational execution. In practice, the managemen<sup>t</sup> of power and energy consumption in most manufacturing enterprises, especially those for which energy costs are not significant, requires measures of a very basic nature. Thus far, the perception of energy as a public good has not been conducive to both the building managers' awareness of possible problems and to prepare enterprises in the technical layer. There is also a lack of work in the literature to help identify and better understand the reasons for this situation. The assessment of the convergence of the obtained results with the knowledge in enterprises on power and electricity consumption managemen<sup>t</sup> requires further research.

In the research area, the formulated problem seems to be extremely interesting. The proposed method should be developed quantitatively and qualitatively. Simulation experiments conducted quantitatively require the construction of formal mathematic models of individual production resources. Additionally, optimization models should be formulated, taking into account the various decision-making criteria, not only in the field of power and electricity costs. The investigated problem has a multi-criteria character. Such models and possibly dedicated algorithms should help decision-makers in enterprises to make well-reasoned business decisions.

The advantage of the applied method is that it supports the decision-making process in manufacturing companies. Currently, in many of them, the decision-making process is based on simple quantitative data, intuition, and the experience of production managers as they do not have efficient and effective supporting tools at their disposal.

The method allows one to verify the possibility of carrying out the adopted production schedule in specific conditions of the power possessed in a given production cell, which also enables determining the adverse effects of reducing power consumption in a given time. Simulation modeling enables quick verification of different event scenarios in the scope of the occurring power reduction. It also provides a basis on which to assess the consequences of using different solutions in response to the limitation. Carrying out the simulation takes a short time, and after its execution, a detailed production schedule is obtained. Other data are also available concerning the extension of the order execution time and the amount of machine load. Depending on the model built and the adopted characteristics/parameters of the production process, these data can be varied and dedicated to a specific company. This information is taken into account in the built mathematical model, which shows the relationships between variables. Then, the described relations are translated into a simulation model.

It can be concluded that the main advantage of the proposed solution is the possibility of the experimental verification of different energy consumption managemen<sup>t</sup> strategies. Complex systems do not have such functionality. Simulation modeling software is intuitive and easy to install, does not require large hardware resources, and can be used in both large and small enterprises. Simulation models are scalable and their functions can be defined by users. A certain limitation in their use is their difficult integration with other enterprise systems (no API).

In qualitative terms, it is important to note that the problem under study is related to a much broader research area that concerns the use of modern information technologies including the field of communication IoT systems and data analysis AI algorithms. Manufacturing companies strive to build digital twins of the production systems they own. Regardless of the choice of a given software or tools, the application of the presented method will have a versatile character.

The changes should be considered in the context of the functioning of power markets, which primarily concerns a new definition of products and services in these markets and the necessity of the instantaneous valuation of these products.
