Keywords

1 Introduction

The manufacturing industry is responsible for one third of global CO2 emissions [1]. To contain and minimize the negative consequences of human-caused environmental impacts, companies and legislators are attempting to counteract the harmful trend that is already emerging [2]. Due to political regulations e.g., the German Climate Protection Plan legislations such as the Climate Protection Act, CO2 pricing, or the EU taxonomy are increasing pressure on companies to measure and report their environmental impact. Various guidelines provided by different rating and reporting organizations are available to direct companies in such measurements. However, these existing guidelines show shortcomings [3,4,5,6,7]. The core problem is the low granularity of measurement dictated by these guidelines. Often only the overall environmental impact of the company is required to be measured, rather than allocation to single product units or machines. Standards such as ISO 14001 or the CDP framework tend to address external stakeholders rather than to support internal identification, analysis and optimization of operational environmental impacts [3]. Especially not at the production machine level. The result is that energy consumption and environmental impact of the entire factory is known, while which production step or machine is responsible and where the impact can be reduced remains unknown.

The goal of the research presented in this paper is to develop a guideline for manufacturers to determine the data needed to identify improvement potentials, while assessing the ecological gate-to-gate impact of their production. The intended use is for a quick initial assessment of which data should be measured with which granularity for the main processes in the facility. With this knowledge a more informed decision can be made where to invest resources to collect data.

2 Brief Overview of Common SRTs and Weaknesses

To measure and report their environmental impacts, companies rely on various reporting and disclosure frameworks and standards, which are collectively known as corporate sustainability reporting tools (SRTs). A direct comparison of the most common SRTs shows that they focus mainly on information at the company level. Only ISO 14001 addresses the process and product levels. Here however, the entire product life cycle is considered and not just gate-to-gate production. The comparison of common SRTs is summarized in Table 1, below.

Table 1. High-level qualitative comparison of the three most common SRTs.
Full size table

The shortcomings of the existing SRTs have already been shown. The existing efforts to improve STRs can be summarized into three groups. Firstly, efforts looking for an approach to make SRTs more accessible and easier to apply [3, 8, 9]. Secondly those aiming to use more data and a higher level of granularity to uncover their own optimization potential [1, 10]. Thirdly, those developing concepts to reduce uncertainties and assumptions in life cycle assessments (LCAs), to make more robust statements about environmental impacts [11,12,13]. The authors found no approaches including all three of these areas, which is a gap to guideline aims to fill.

3 Development of Guideline

3.1 Guideline Structure

Based on the shortcomings of existing SRTs and the resulting challenges, the following four requirements (R1-R4) are defined for the guideline: R1: Indicators must be measurable, R2: System boundaries to gate-to-gate production, R3: Identify optimization potential & allow prioritization, R4: Easy to understand and apply.

Firstly, the required data that the guideline recommends to be collected is derived based on the main ways production systems cause ecological impact [14, 15]. As a result, the following data types were selected: energy demand (electricity, heat), water demand, material use, process time (machine times in the four statuses off, start-up, idle, operation), and scrap amount.

Secondly, a structure for the guideline is developed. Long-term, the guideline should be developed for all manufacturing processes, but for the initial concept version the authors have focused on discrete production processes. A classification according to DIN 8580 is adopted, which is a widespread standard that provides detailed main groups, groups and subgroups coving all discrete processes. To manage the scope, not all processes from the standard are included in the guideline; a selection is made based on which processes a) are relevant for the later application example and b) are frequently used in the context of discrete production. The following subgroups are included in the guideline: 2.1.1 Rolling, 2.1.4 Indentation, 2.2.2 Deep drawing, 3.1.1 Shear cutting, 3.1.2 Knife cutting, 3.6.4 Solvent cleaning, 5.4.2 Electrostatic coating. See DIN 8580 for descriptions [16].

Thirdly, to define levels of data granularity and to evaluate where information about the production steps in a manufacturing company is generated RAMI 4.0 is chosen. RAMI 4.0 is a popular conceptual architectural model created to provide companies with a framework for approach and deploying Industry 4.0 initiatives [17]. Specifically, the hierarchy dimension of RAMI 4.0 is adopted into the guideline, as shown in Fig. 1. RAMI 4.0, rather than the classical automation pyramid from DIN EN 62264 [18] was chosen because RAMI 4.0 was conceived specifically with manufacturing systems of the future in mind; with increasing connectivity across value, fewer companies will have a rigid hierarchy according to DIN EN 62264 or other standards, but rather take on a more flexible network structure [19].

Fig. 1.
figure 1

Definition of guideline data granularity levels, based on RAMI 4.0 hierarchy levels [17].

Full size image

3.2 Determining Data Granularity Levels

Based on the guideline requirements, the required data, the seven selected manufacturing processes and the data granularity levels, an evaluation matrix is developed, which dictates a granularity level for each manufacturing process. The matrix consists of a list of process characteristics, their qualitative relative values, an evaluation of each value and the resulting recommended granularity level. Evaluating all characteristics together results in an overall granularity recommended for the process.

The following characteristics are considered: machine power, process time (off, start-up, idle, run-time, operation), auxiliary materials (e.g., chemicals or lubricants), tool properties (weight and dimensions of the tool), heat (thermal energy demand), water demand and scrap. Machine power describes the maximum power that the machine can deliver. With a high maximum power, the influence on the granularity is generally high. Process time refers to the different electrical energy demand of a machine depending on the time of measurement. Typically, the energy demand increases sharply when a machine is started and then decreases again when it reaches the operating state. However, the energy consumption also varies when the machine is idle, i.e., between two work steps, and when the work is being performed. Thus, the process times have an influence on the actual energy consumption of a machine. Auxiliary materials such as coatings and lubricants are considered. These are typically used to prevent damage to the workpiece during processing such as forming. However, they must be removed again when the product is finished. The more auxiliary materials are used during the production of a product, the more effort (mechanical work, as well as use of water or other solvents) is necessary in the end to separate the auxiliary materials from the workpiece or product. Tool properties refer to the tool to be moved in the machine. This can be, for example, a small sharp saw blade or several punches or bolts moving up and down behind each other to deform a metal sheet. Depending on the tool properties, the energy requirement, and the duration of the work to be performed change. Thus, the impact on granularity is also high. Heat, water consumption and scrap can clearly have a high influence on the required data granularity and vary among production processes. The scope of applicability considers whether a manufacturing process is usually used for different materials or products, or only one or few specific ones. If the scope of applicability is versatile and the influence on other characteristics is large, a higher granularity should be selected.

Following the above approach, each characteristic is given a corresponding granularity level, resulting in a list of various granularity levels. If different granularity levels have been determined, the characteristics tool properties and scope of applicability are given higher weight. The following rules should be observed when selecting the granularity. If the influence on the granularity of all characteristics is the same, then this granularity level is selected as the overall required granularity for the process. If the influence on the granularity of all characteristics is not the same, the granularity level selected for tool properties and scope of applicability determines the granularity, though strong variations tend to indicate that the assessment may have faults. In Table 2. Below, the evaluation matrix for deep drawing is shown as an example.

Table 2. Data granularity evaluation matrix, applied on the deep drawing process.
Full size table

This procedure was carried out for all seven previously selected process types and resulted in a respective granularity level for each process type (Fig. 2).

4 Application Use Case Evaluation

For evaluation purposes, the guideline was applied to the production of a 50 by 20 cm aluminum canister (further details omitted for confidentiality). First, the production steps from the production of the canister were allocated to the DIN 8580 processes in the guideline (Fig. 2). Thus, the required granularity for each production step of the aluminum canister can directly be read from the guideline. To evaluate whether the granularities recommended by the guideline are appropriate, the following question is posed for each production step: “would a higher data granularity level provide additional insights, or would a lower level provide similar insights with less effort?”.

Fig. 2.
figure 2

Application of the guideline to the use case production steps, to determine appropriate data granularity level for each step.

Full size image

For the production steps marking and cutting, the guideline recommends the granularity low. In this context low granularity consists of process step information derived from high-level information from the enterprise resource planning system (ERP), rather than any direct measurements of the process and its machines. Closer evaluation reveals that marking and cutting have only very small contributions to the environmental indicators. Therefore, no significant optimization potentials are to be expected from analyzing the production steps in detail using medium or high granularity data. Thus, low measurement granularity is an appropriate recommendation for these two steps.

For the production steps necking, threading, and washing, the guideline specifies a granularity level of medium. This means that the data should be collected at the production process level, for example using the manufacturing execution system (MES) or supervisory control and data acquisition (SCADA) systems. The medium granularity level for washing, for example, is justified because this is where most of the water needed for the cylinder production is used. A high granularity level is not necessary for washing since almost all the water consumption for the cylinder production occurs at this one step; few additional insights could be gained by measuring water consumption per product being washed, since this can simply be calculated by dividing total water consumption by number of products (of the same type) washed.

For the production steps blanking, deep drawing and painting, the guideline recommends a high granularity. This is confirmed by the high influence of the three production steps on several of the environmental indicators examined. These steps have the highest energy requirements (in order: painting, deep drawing, blanking) of all steps, painting has a long process cycle time, all these steps use many auxiliary materials and heavy tools (blanking, deep drawing) and blanking is responsible for the most scrap. The additional effort for measuring the data on field device or product level is justified by the insights that would otherwise be missed. For example, machine power, as well as scrap, in relation to the sheet thickness for blanking, could only be measured at high granularity, and could reveal insights on whether machine settings are over-dimensioned and thus wasting energy and material.

5 Conclusion

The associated pressure from society and politics has led to companies having to measure, disclose and ultimately improve their environmental impact. Various standards and frameworks already exist for this purpose. However, it has become apparent that these are more suitable for informing external stakeholders than for disclosing optimization potentials within the company's own production. The main problem identified by the authors is a lack of granularity in the SRTs for allocating environmental impacts to production lines, machines or even products. Thus, the authors developed a guideline that specifies a level of granularity for seven manufacturing processes to not only determine the environmental impact of gate-to-gate production, but to also to identify opportunities to reduce this overall impact. The required data were defined in terms of energy, water, and material demand. To provide structure to the many different production techniques, the focus was initially limited to seven processes from the DIN 8580 standard. RAMI 4.0 was then used to define the data granularity levels low, medium, and high. Finally, the required granularity level for each production step was determined by following a structure approach in the form of an evaluation matrix. An evaluation of the guideline on a real-life manufacturing system has shown that appropriate granularity levels are recommended by the guideline, confirming that it has potential to be a useful tool for manufactures wanting to reduce their environmental impact.

Even though the guideline has recommended the right level of granularity in places where the production steps have a high impact, there is still room for improvement. Firstly, the same manufacturing process can vary strongly depending on the application and industry, potentially resulting in different ideal measurement granularities. Thus, even within the detailed catalog of DIN 8580 further differentiation would be valuable. Secondly, only a small selection of certain manufacturing processes has been considered, and further processes need to be added in order for the guideline to be applicable to more manufacturers.

The authors find three main efforts most relevant for future research. Firstly, is to obtain more company data to validate the existing recommendations for the granularity levels for the reasons mentioned above. Secondly, additional production techniques within DIN 8508, and also outside the realm of discrete production, should be investigated and be incorporated using the same approach which this paper has shown to be effective. Thirdly, expanding the guideline to also recommend the required data accuracy, in addition to the granularity, for different processes and situations could be of great value to manufacturers.