Technical and Economic Approaches to Design Net-Zero Energy Factories: A Case Study of a German Carpentry Factory

Abstract

As many German SMEs approach the end of their photovoltaic (PV) feed-in tariff period, the challenge of maintaining economic viability for these installations intensifies. This study addresses the integration of intermittent renewable energy sources (iRES) into production processes by proposing a method to identify and exploit industrial flexibility. A detailed case study of a German carpentry factory designed as a Net-Zero Energy Factory (NZEF) illustrates the approach, combining energy monitoring with blockchain technology to enhance transparency and traceability. Flexibility is exploited through a three-layer control system involving passive operator guidance, battery storage, and electric vehicle charging. The installation of a 40 kWh battery raises self-consumption from 50 to 70%, saving approximately EUR 4270 annually. However, this alone does not offset the investment. Blockchain-based transparency adds economic value by enabling premium pricing, potentially increasing revenue by up to 10%. This dual benefit supports the financial case for NZEFs. The framework is replicable and particularly relevant for low-automation industries, offering small and medium enterprises (SMEs) a viable pathway to decarbonization. The results align with the European Clean Industrial Deal, demonstrating how digitalization and industrial flexibility can reinforce competitiveness, sustainability, and digital trust in Europe’s transition to a resilient, low-carbon economy.

1. Introduction

In the wake of the European Green Deal’s ambitious target for European climate neutrality by 2050 [1], numerous European nations have hastened their initiatives to usher in the era of net-zero carbon power grids. One particularly effective strategy has involved providing economic incentives to encourage the generation of electric power through intermittent renewable energy sources (iRES), such as wind and solar, which are poised to be the primary contributors to electricity generation. This incentive is typically assured for approximately two decades, after which iRES operators may face significantly diminished compensation or none at all. However, owing to the inherently unpredictable generation patterns of iRES, system operators (SOs) at both transmission and distribution levels will necessitate substantial zero-emission flexible firm capacity for both short-term (minute to minute or hourly) and long-term (weekly and monthly) needs [2,3]. This calls for significant reinforcement of the energy grids (electric [4], thermal [5], and gas [6]).

Industrial systems can play a significant role in supplying a part of the flexibility needed [7,8,9]. To date, most of the research has focused on technical and economic analyses on applying demand response (DR) solutions in highly automated industrial manufacturing systems [10,11,12,13,14,15,16,17]. However, the majority of European businesses are represented by small-enterprises, accounting for over 90% of all European enterprises [18]. These enterprises predominantly operate using job shop manufacturing processes, which are characterized by small-scale, customized production with limited automation. As a result, small-enterprises typically exhibit lower levels of automation compared to medium and large enterprises. There is still a gap in understanding the potential benefits of extending flexibility programs to smaller, less automated enterprises.

In Germany, in the past few decades, a very high iRES capacity has been installed on roofs of the buildings of small and medium-sized enterprises (SMEs). It has been estimated that, by 2030, about 10.6 GW of the photovoltaic (PV) installed capacity on the roofs of SMEs’ buildings will reach the culmination of their incentive period (see Figure 1). Considering that German SMEs paid an average electricity price of 342.4 EUR/MWh in the years 2020–2023 [19], the prospect of operating industrial sites as net-zero energy factories (NZEFs) becomes an appealing avenue for SME operators.

Figure 1. Photovoltaic capacity incentive expiration installed on German small and medium-sized enterprise buildings based on the Bundesnetzagentur Market Master Data Register (MaStr) [20].

Figure 1. Photovoltaic capacity incentive expiration installed on German small and medium-sized enterprise buildings based on the Bundesnetzagentur Market Master Data Register (MaStr) [20].

Nonetheless, seamless integration into industrial processes using iRES may mandate substantial investments, particularly if relying solely on energy storage systems, which remain relatively high-priced [21,22]. Hence, exploring alternative solutions becomes imperative, enabling SME operators to adeptly plan and operate industrial sites as NZEFs. Digitalization of the industrial processes for implementing this is mandatory [23]. Digital technologies and digital-based solutions, such as the internet of things (IoT), wireless sensor networks, artificial intelligence, machine learning techniques, federated learning, blockchain, and the Hyperledger framework, can contribute to the operation of industrial systems and processes more actively, allowing one to exploit flexibility in integrating the volatility of iRES generation into the facility system, thereby contributing to the development of new sustainable business models [24,25]. These digital technologies allow one to perform highly accurate predictions of energy consumption and generation to optimize the resources (humans, energy, material, and machines) in industrial processes and optimally integrate the power generated by renewable energy sources. In addition, interest in net-zero strategies has recently expanded beyond technical and environmental aspects to include the impact on consumer perceptions.

The originality and novelty of this paper lie in its holistic approach to addressing the often overlooked challenge faced by many SMEs with rooftop PV systems nearing the end of their feed-in tariff incentive period. This work uniquely addresses the integration of blockchain solutions to enable transparency in low-automation industrial processes, with a focus on SMEs, going beyond the scope of most current NZEF studies. Most existing studies focus on large, highly automated industrial systems or offer generic solutions for renewable energy integration. By contrast, this paper fills an important research gap by providing a targeted, practical framework specifically designed for small enterprises engaged in discrete manufacturing processes, such as carpentry, which typically lack high levels of automation and standardized flexibility solutions. Unlike previous research, which mainly emphasizes technical optimization or energy storage systems, this study combines process flexibility with blockchain-based solutions to enhance the transparency and traceability of energy sustainability. This dual approach offers a clear economic rationale by not only increasing energy self-consumption but also creating added business value through transparent and verifiable sustainability practices. This makes it possible for SMEs to differentiate their products in the market and justify the investment required to operate as NZEFs. To the best of the author’s knowledge, this is one of the first studies to explore the integration of blockchain as a tool to compensate for the costs of designing industrial flexibility, particularly in low-automation sectors, such as carpentry. The paper presents a detailed case study and cost–benefit analysis that demonstrate how this approach can be practically applied. It offers replicable strategies for integrating renewable energy and digital solutions into various industrial sectors, providing practical guidance for SMEs aiming to remain competitive while contributing to decarbonization efforts.

The study is structured as follows: Section 2 deals with listing the aims of the work. Section 3 provides a literature review on Net-Zero Energy Factories considering the social, ecological and economic advantages, on state of the art to exploit flexibility within industrial systems and on identifying the research gaps. Section 4 deals with the developed methods to identify, quantify and exploit flexibility within industrial systems, and with the testing the methodology in the design of a German carpentry factory as an NZEF. The results of this testing, including the evaluation of energy sustainability and the cost-benefit analysis of implementing flexibility and sustainability tracking in the considered carpentry factory, as well as the benefits of designing blockchain-based solutions to increase energy sustainability, are depicted in Section 5. The study ends with the conclusion in Section 6.

2. Aim of the Work

The purpose of this study is to develop and validate a replicable methodology for designing a NZEF in the context of a SME. The proposed approach addresses the integration of variable renewable energy sources into industrial processes, aiming to maximize on-site renewable energy self-consumption while ensuring economic viability.

Specifically, this study aims to carry out the following:

• Identify, quantify, and exploiting process flexibility from metered load profiles.
• Evaluate technical integration options, combining process flexibility with energy storage systems and electric vehicle (EV) charging infrastructure.
• Implement blockchain-based monitoring to enhance transparency, traceability, and verification of sustainability indicators.
• Provide a replicable benchmark for other SMEs with similar operational characteristics, ensuring both environmental and economic benefits.

3. Literature Review

3.1. Net-Zero Energy Factories and Their Social, Ecological and Economic Advantages

Various definitions exist for net-zero energy systems (NZESs) in the literature. Typically, NZESs refer to residential, commercial, or industrial systems that can locally generate as much energy as they consume for electric, thermal, and transportation loads over an annual period [26,27]. This definition generally relies on an annual energy balance analysis rather than the integration of energy into the system.

In this study, an NZES is defined differently. NZESs are systems capable of locally producing electric energy from iRES, fully integrated into the systems’ loads, thereby avoiding the need to feed into the external electricity grid. Designing systems as NZES offers several advantages. From a broader perspective, such systems indirectly support grid operators (distribution and transmission SOs) in their decarbonization efforts. This design results in a lower intermittent power that requires balancing, subsequently, reducing balancing service costs. Additionally, all consumers benefit, as the costs of grid balancing are generally distributed among their electricity bills. Grid balancing costs, for instance, reached EUR 4.2 billion in Germany in 2022, causing an increase in CO₂ emissions by 4.9 million t [28]. Most of these costs are borne by small residential consumers, who are charged about three times more than industrial consumers (see Figure 2). It is evident that systems capable of locally producing electricity through iRES and locally consuming it, avoiding feeding the surplus to the external grid, contribute to decreasing the amount of iRES that SOs need to balance. This positively affects the grid charges paid by residential consumers, consequently increasing the social acceptance of the decarbonization process [29,30].

Figure 2. Grid charges distributed among the electricity consumers in Germany, classified according to different yearly electricity consumption. Adaption from [30] (2022).

Figure 2. Grid charges distributed among the electricity consumers in Germany, classified according to different yearly electricity consumption. Adaption from [30] (2022).

Designing industrial systems as NZEFs aligns with all three pillars of sustainability, as formulated by Elkington in [31] (see Figure 3). Power generation by iRES positively impacts environmental conservation by reducing anthropogenic emissions. The full integration of the power generated by iRES into the industrial facility enhances the social acceptance of the decarbonization process. Furthermore, the self-consumption of all the power generated positively influences the economic growth of the facility by stabilizing costs related to the energy resource.

Figure 3. Triple bottom line of sustainable development. Based on [31].

Figure 3. Triple bottom line of sustainable development. Based on [31].

Studies, such as the one conducted in Korea [32], have demonstrated that adopting net-zero initiatives can positively influence consumer trust and product evaluations through mechanisms of reciprocity and transparency. This dynamic is particularly relevant when considering the importance of transparency in industrial operations, such as those within NZEFs. Additionally, according to [33,34], an increasing number of major investors and consumers prioritize the sustainability of industrial facilities and the goods the produce as decision criteria for their investments. In this context, the incorporation of blockchain technology can play a pivotal role in making the sustainability of industrial processes more transparent [35]. Blockchain’s decentralized and immutable ledger ensures that data related to energy consumption, carbon emission, and resource utilization is securely recorded and easily accessible. Blockchain can create a trustworthy and auditable system for tracking the sustainability metrics of industrial processes [36]. The enhanced transparency provided by blockchain technology in tracking sustainability metrics can also unlock economic value for industrial facility operators [37] willing to design their facilities as NZEFs. The ability to demonstrate a commitment to sustainable and transparent operations becomes a market advantage, potentially attracting partnerships, clients, and customers who prioritize environmentally responsible businesses, thereby, contributing to economic growth for facility operators.

3.2. Demand-Side Flexibility in Industrial Systems

The integration of iRES into manufacturing systems requires one to identify and quantify the existing hidden flexibility and to optimally exploit it. In many cases, additional flexibility needs to be designed. Flexibility options, such as storing electricity in energy storage systems [38] or converting it into other energy forms, such as heat [39,40], compressed air [41,42,43], hydrogen, [44,45] or synthetic fuels [46], can provide a high degree of flexibility. However, these options are both economically and energetically expensive. Indeed, all energy conversion processes from electricity to other forms (i.e., heat, compressed air, hydrogen, synthetic fuels) involve energy losses ranging from 20 to 60% [47]. Therefore, it is advisable in the design of NZES to identify, quantify, and exploit the inherent flexibility of the system itself. In cases where the identified and quantified flexibility potential is insufficient to fully integrate the power generated by iRES, other options, such as energy storage systems and/or energy conversion systems through Power-to-X technologies [48,49], should be considered. Industrial systems offer the highest potential for flexibility exploitation among residential, commercial, and industrial systems [50,51]. Small and medium-sized enterprises make up a significant portion of the national economy in Germany, accounting for 99.6% of all businesses [52]. The flexibility that SMEs can exploit depends on their industrial processes. Using the product variety and production volume as primary criteria, the flexibility of the industrial processes can be qualitatively assessed. In general, greater product variety increases schedulable slack, while lower production volume reduces takt constraints. Processes that are customer-oriented and involve job shops or butchering tend to have a wide range of final products but lower production volume, making them more flexible. These processes have the highest potential for flexibility. On the other hand, repetitive and continuous processes, characterized by a high production volume and a smaller variety of products, have lower flexibility potential, as illustrated in Figure 4. In comparison with German heavy industries, which could exploit flexibility of up to hundreds of MW of power for up to 4–6 h [53], the flexibility potential that could be exploited by the industrial processes of a single SME is much smaller. It could be quantified as about 30% of its maximum load (generally from a few hundred kW to a few MW) for up to 15 min [54].

The adaptability of load consumption within industrial systems offers various avenues for optimization. Demand-side management programs, exemplified in Figure 5, present options such as valley filling, load shifting, strategic load growth, conservation, peak clipping, and flexible load shaping [8,55]. Typical technologies involved in demand-side management programs are energy storage (electric, thermal and mechanical), advanced building envelopes, electric heat pumps, air compressors, electric vehicles, controllable machines, energy management systems, and material storages [56]. All these technologies allow one to align load consumption with iRES generation. The integration of demand-side management technologies within NZEFs has significant implications for operational and economic performance. Through innovative demand-side strategies and the adoption of flexible energy solutions, NZEFs are equipped to optimize resource utilization while maintaining production efficiency. These advancements position industrial facilities at the forefront of energy innovation, creating a strong foundation for sustainable economic growth and competitiveness in the evolving energy market.

3.3. Research Gaps

Despite significant advances, three main gaps remain in the literature:

• Limited focus on SMEs: Most NZEF studies address large, automated facilities, leaving a lack of practical methodologies or small-scale, semi-automated production systems.
• Integration of flexibility and transparency: Few works combine process flexibility measures with blockchain-based traceability in a unified framework.
• Economic replicability: Existing studies often overlook the cost–benefit trade-offs for SMSe, particularly in the context of post feed-in tariff scenarios where revenue streams from iRES generation are reduced or are zero.

This study addresses these gaps by proposing and testing a replicable methodology for NZEF design tailored to SMEs, combining technical flexibility measures with blockchain-enabled transparency, and assessing the economic feasibility in a real-word case.

4. Methods and Materials

4.1. Methodology for Identifying, Quantifying, and Exploiting Flexibility in Industrial Processes

The flexibility potential within the NZEF is defined by the industrial process’s ability to adapt its energy demand in response to energy generation by iRES. The flexibility can be exploited by varying the demanded energy (compared to a prescheduled value) and by anticipating or postponing the prescheduled energy demand. From this perspective, the flexibility can be expressed as a flexibility of power (φ_P) and as temporal flexibility (φ_T). Let T = {t_i, t_j,…t_f} represent the timeframe of electricity generation, divided into equal intervals tj (each lasting 15 min, for example). The flexibility of power (φ_P) throughout the entire time frame T can be mathematically expressed as in Equation (1). It is constrained by Equation (2), meaning that the total energy consumed with a flexible load does not change:

$${\phi }_P=\Delta \mathrm{P}=P_s-P_f$$

(1)

where ΔP represents the change in demand during the time interval t_j. It is calculated as the difference between the scheduled power (P_s) and the newly flexible consumed power (P_f).

$${\int }_{t_i}^{t_f}{P}_s\left(t\right)dt={\int }_{t_i}^{t_f}{P}_f\left(t\right)dt$$

(2)

Graphically, the flexibility of power is represented in Figure 6. Within the considered time interval, the power demand profile changes without affecting the total energy consumed.

The temporal flexibility (φ_T) describes the system’s ability to adjust its energy usage to occur sooner (advance) or later (defer) compared to a set consumption schedule. Although the total energy remains constant, the flexible power consumption profile can differ significantly from the initially scheduled profile (see Figure 7). To clearly define temporal flexibility, the scheduled energy and flexible energy must remain equal (see Equation (3)).

$$E_{scheduled}={\int }_{t_i}^{t_f}{P}_s\left(t\right)dt={\int }_{t_i\_ant}^{t_{f\_ant}}{P}_{f\_ant}\left(t\right)dt={\int }_{t_{i\_post}}^{t_{f\_post}}{P}_{f\_post}\left(t\right)dt$$

(3)

The temporal flexibility (φ_T) can then be expressed as in Equation (4):

$${\phi }_T={\displaystyle \frac{{\int }_{t_i\_ant}^{t_{f\_ant}}{\left(t_{i\_ant}-t\right)·P}_{f\_ant}\left(t\right)dt}{E_{scheduled}}}+{\displaystyle \frac{{\int }_{t_i\_post}^{t_{f\_post}}{\left(t-t_{i\_post}\right)·P}_{f\_post}\left(t\right)dt}{E_{scheduled}}}$$

(4)

where

• $P_{f_{ant}\left(t\right)}$ is the flexible power consumption profile anticipated to an earlier timeframe {t_{i_ant}, t_{f_ant}} relative to the originally scheduled interval;
• $P_{f_{post}\left(t\right)}$ is the flexible power consumption profile postponed to a later timeframe {t_{i_post}, t_{f_post}} relative to the originally scheduled interval.

The temporal flexibility (${\phi }_T$) quantifies the time shift in energy consumption, distinguishing between anticipated (negative shift) and postponed (positive shift) scenarios. Figure 7 shows the graphical representation of the temporal flexibility.

Figure 8 illustrates the various values related to the exploitation of flexibility. The flexibility that industrial processes can practically exploit is generally significantly less than the theoretical flexibility. Theoretical flexibility is the upper bound obtained from the power and temporal metrics (φ_P, φ_T) in Equations (1)–(4). This discrepancy is influenced by the nominal load of each individual machine in the industrial process and the nature of the industrial process (see Figure 4). The technical flexibility filters this potential by applying machine and process level limits (e.g., rated power, idle, constraints, warm-up/setup times). Practical flexibility further narrows what can be executed on site once facility realities are considered (e.g., operator availability, shift calendars, material buffer, production sequencing). Economic flexibility pertains to the ability to generate positive cash flows, such as minimizing energy costs, maximizing self-consumption, battery degradation. Viable flexibility reflects the cost–benefit analysis for exploiting flexibility, incorporating a multi-criteria approach that considers factors such as the risk policy, overall energy efficiency of the industrial facility, and associated costs.

Figure 8. Different values of flexibility. Based on [57].

Figure 8. Different values of flexibility. Based on [57].

The process of identifying and quantifying flexibility potential within industrial processes lacks standardization and is influenced by several factors. Diverse models and tools have been developed in recent years to represent industrial processes, encompassing energy and material flows, human resources, and external factors, such as ambient temperature, electricity price signals, or weather conditions. Models such as agent and multi-agent [58,59], algebraic logical [60], regression [61,62], state-space [63,64], data mining methods [65,66], and graphical models [67,68] are employed for manufacturing system analysis and performance indicators.

This study introduces a methodology based on measuring the active power consumption of manufacturing machines, adapting the methodology proposed in [69,70,71]. The approach rests on the premise that hidden flexibility potential can be identified by continually measuring consumption profiles over an extended period (e.g., different months) and observing nonidentical load patterns. The application of this methodology to discrete manufacturing processes, typical of carpentry processes, is investigated in this study. It involves eight steps:

I.

Describe the manufacturing process;

II.

Measure the electricity consumption of all devices and machines in the factory;

III.

Analyze the metered data qualitatively;

IV.

Classify electric loads into controllable (P_con) and noncontrollable (P_ncon);

V.

Evaluate the difference between P_con and P_ncon for each time interval (t_j);

VI.

Sort positive values calculated in step V in decreasing order and negative values in increasing order;

VII.

Evaluate the flexibility duration curve; and

VIII.

Perform statistical analyses.

The value calculated at step V corresponds to the potential to exploit the flexibility of power (see Equation (1)). If this value is positive, it indicates that the load can be reduced. Conversely, a negative difference suggests that the load can be increased. This type of flexibility also involves either postponing or anticipating processes that need to be rescheduled to align with the time when consumption matches the power generated by iRES. This methodology enables the scheduling of manufacturing processes to enhance the matching effect between the power demanded and the power generated by iRES, as outlined in Equation (4). Figure 9 summarizes the flowchart of the proposed methodology.

The matching effect within the NZEF concept should never exceed 1, and is calculated only when iRES plants generate power (P_g(t) > 0) and the factory is in operation (P_d(t) > 0).

$${ME}_{NZEF}(P)={\displaystyle \frac{{\int }_{t_i}^{t_f}{P}_g\left(t\right)dt}{{\int }_{t_i}^{t_f}{P}_d\left(t\right)dt}}, { P}_d\left(t\right)\cup { P}_g\left(t\right)>0$$

(5)

The quantification of flexibility is visually explained through the representation of flexibility duration curves, which combine the flexibility of power and the temporal flexibility that a machine or a group of machines can exploit (see Figures 13 and 14 in Section 5.1).

4.2. Material for Designing a Sustainable Net-Zero Energy Factory: A Case Study of a German Carpentry Factory

A specific case study, namely, a carpentry factory located in the city of Magdeburg (Germany) has been analyzed. A PV plant capable of generating up to 125 kW of electric power has been installed on the roof of the carpentry factory. By designing the carpentry factory as an NZEF, the exploitation of the flexibility needed to fully integrate the power generated by iRES is fundamental. However, the economic and social sustainability that the industrial facility has to possess must also be considered. Generally, NZEFs comprise four essential layers which should be designed or upgraded (see Figure 10):

• Infrastructures: Encompassing energetic and industrial aspects, such as energy generation, storage (for energy and materials), energy conversion systems (for example, power to heat using industrial heat pumps), load (electric, compressed air, heat at different temperatures), and digital infrastructure.
• Sensorics: Involving devices to measure energy and material flows within the facility, as well as parameters such as wind speed, solar irradiation, ambient temperature, room temperature, and the rotating speed of machines.
• Data Platform: A centralized platform for processing and analyzing data collected by sensorics to enable various services. It will support decision-making processes for flexibilizing the industrial processes.
• Services: Offered in two primary categories—energetic services, optimizing the integration of power generated by iRES into industrial processes, and digital services, such as utilizing blockchain for tracking the energetic footprint of industrial processes [72].

A holistic approach should be considered in the design phase of NZEFs. The design of the carpentry factory as an NZEF involved, firstly, the installation of sensorics. They were installed to measure the electric power consumption of all equipment as well as the power generated by the PV plant. The metering of the data forms the basis for estimating the theoretical flexibility that industrial processes can exploit. The network of sensors provides real-time monitoring and detailed measurements of energy flows within the carpentry factory. Smart meters were installed on all machines to collect data at 1 s intervals. These sensors are complemented by WAGO smart meters, a class of industrial-grade digital energy meters produced by WAGO GmbH designed for high-precision monitoring of electrical parameters in commercial and industrial systems, which monitor the total electrical load, and a Solar-Log system, which records the PV plant’s power generation every 30 s. All devices communicate via the Modbus protocol, and publish to the central database via message queuing telemetry transport protocol/hypertext transfer protocol (secure) (MQTT)/HTTP(s). This infrastructure integrates real-time data acquisition with active and passive control strategies to ensure the optimal use of renewable energy sources. The system plays a crucial role in collecting data, forecasting energy generation and consumption, and making informed decisions to align production processes with renewable energy availability. The monitoring and control system is structured into two main layers: monitoring and active and passive control (see Figure 11):

• Monitoring layer: This layer is responsible for continuously collecting and analyzing data from key energy-related components of the factory, including the PV system, load portfolio (sawing, drilling, milling, sanding), battery storage, and electric vehicle charging stations. Smart meters are used to monitor the real-time energy consumption, while the Solar-Log system records the PV generation. Data from all devices are transmitted through various communication protocols—HTTP(S), MQTT, file transfer protocol secure (FTP(S)), and Modbus transmission control protocol/user datagram protocol (TCP/UDP)—and stored in a centralized database. The data are analyzed to evaluate historical trends, forecast future energy generation and consumption (both electric and thermal), and support decision-making processes.
• Active and passive control layer: This layer uses the data collected by the monitoring system to implement control strategies that optimize energy consumption. Passive control provides machine operators with real-time visual feedback (e.g., traffic light signals) to guide process adjustments during periods of high solar power generation. If the suggested adjustments are not made, active control intervenes to manage battery storage and electric vehicle charging, ensuring maximum integration of the renewable energy into the production process.

The combination of real-time monitoring and automated control enables the factory to dynamically adapt its energy demand, maximize self-consumption of renewable energy, and reduce operational costs, ultimately increasing the economic and environmental sustainability of the facility.

Designing an industrial facility as an NZEF can be a solution to accelerate the decarbonization process, increase the social acceptance of the energy transition, and indirectly support SOs. However, it might not always be the best solution for the economic sustainability of the industrial facility. Indeed, the economic benefits related to having more constant energy costs should be compared with the investment needed to design additional flexibility options. The blockchain solution making transparent the sustainability of the manufacturing processes can help the facility operator to create a business model that generates new income, justifying the investment in planning the facility as an NZEF. Blockchain technology is a decentralized, distributed ledger system in which data are stored in sequentially linked blocks secured through cryptographic algorithms. Each block contains a set of time-stamped transactions that, once validated by the network participants (nodes) through a consensus mechanism, become immutable and transparent to all authorized parties. This architecture ensures data integrity, traceability, and tamper resistance without requiring a centralized authority. In the NZEF context, blockchain can be used to record and verify energy generation, consumption, and material flow data, thereby enabling transparent and auditable sustainability tracking. Transparency, traceability, and data security are, therefore, attributes and features on which new business models can also be developed within NZEFs [73,74]. The transparency of sustainable industrial processes builds trust, which is fundamental to the relationship between facility operators and both new and existing consumers. Increasing the transparency of sustainability products manufactured using green energy documented on the blockchain can demand a premium price on the market. Customers are increasingly willing to pay more for products that have verifiable sustainability credentials, which blockchain can provide. A blockchain-based solution has been designed to track the energy sustainability of manufactured items. Figure 12 presents the digital architecture used to assess the energy sustainability in the carpentry factory being studied.

Data is collected to observe both the energy flows—covering both generation and consumption—and the flow of materials. WAGO smart meters record the electricity used by all machines, while Solar-Log systems monitor the output from the solar panels. Communication between the battery system, the electric vehicle charging station, and the passive manufacturing control is managed through the Modbus protocol. Material tracking is facilitated by portable printers and scanners capable of handling barcodes, which are connected to Raspberry Pi units stationed next to each machine. Raspberry Pi is a low-cost, small–sized single-board computer developed by the Raspberry Pi Foundation. It is widely used in industrial and research applications as a compact, energy-efficient platform for data acquisition, device control, and IoT (internet of things) integration. When tracking the material flow, all raw wood blocks receive different barcode numbers before entering the production hall. The barcode numbers and their connection to the raw wood are stored in the database. The first step in the manufacturing process involves cutting a raw piece of wood into several blocks that will subsequently form the furniture parts. Before the raw wood is divided into multiple blocks, the barcode on the raw wood is scanned. After the division, the worker requests a new barcode for each piece and prints it out. Newly generated barcode numbers create a “mother–child” relationship with the raw material’s barcode, and this relationship information, along with the time of the barcode generation, is stored in the database. Each piece of wood is passed on to the next production process. Depending on the furniture piece being produced, each piece of wood goes through various machines for different manufacturing processes. In the subsequent production steps, the barcode on the piece of wood is scanned upon receipt. This scanning process records the time at which a specific piece of wood is received at a particular machine. After completing the processing, the wood is scanned again. If the barcode is lost or destroyed during processing, a new barcode is requested, and the same barcode number as before the processing is used. The scan after processing records the time at which the wood leaves a particular machine. The piece of wood is then passed on to the next processing stage, which must follow the same steps, i.e., pre-machine scan, wood processing, and post-machine scan. Internal protocol addresses for managing the network communication are allocated to the WAGO smart meters, Solar-Log, battery systems, Raspberry Pi, and the main data collector. The data collector broadcasts data to all the connected clients (net-zero check, predictor, and data persister) using the message queuing telemetry transport protocol (MQTT).

Regarding blockchain development, the Hyperledger Fabric is used [75], which is a permissioned blockchain. Only known entities can participate in a permissioned blockchain network. An MQTT hub runs on the central computer for data collection. Upon receiving data, it sends them to all connected clients. WAGO uses the MQTT protocol to transmit the energy consumption of all machines every second. The energy generation data is collected by a program that sends an HTTP request to the Solar-Log every 30 s and analyzes the response. The energy generation data analyzed is transmitted using the MQTT protocol. For the battery, a Python script was developed and executed on the central management system, which controls the charge/discharge rate and monitors the state of charge every second. Each time an employee scans the barcode, the scan time and barcode information are sent from the Raspberry Pi directly to the central computer via a WebSocket connection. Finally, a data collector listens to the energy consumption, generation, and battery status via MQTT and stores the information in a database. The data collector also listens to the scan time and barcode information received via WebSocket and stores them in the database. Finally, when all the wooden parts that make up the furniture piece are collected for assembly, the barcodes of all parts of the furniture piece are scanned, the barcode on each part is removed, and all parts are assembled. After assembly and packaging, the assembler requests a QR code. A QR code with a hyperlink address is attached to the packaging depicting the energetic sustainability of the manufactured processes.

5. Results and Discussions

5.1. Identification and Quantification of the Flexibility of Industrial Processes

This comprehensive monitoring system plays a crucial role in constructing the flexibility duration curves, which quantify the power and duration of flexibility that can be exploited. The data collected help identify periods when production processes can be postponed or adjusted to maximize the self-consumption of renewable energy. This forms the basis for optimizing production schedules and implementing control strategies that align energy demand with renewable energy availability.

The carpentry factory operates as a discrete job shop process since all the furniture manufactured is designed for individual customers (see Figure 4). Such industrial processes are characterized by a low volume of manufactured items but higher flexibility compared to continuous and repetitive processes. Five typical processes are involved in the carpentry factory:

• Sawing;
• Milling;
• Drilling;
• Sanding;
• Banding.

In addition to these, compressed air and dust suction are activated when compressed air is required or one of the sawing, milling, drilling, sanding, and banding machines is used (for suctioning the wood’s dust). Thermal energy for the wood-drying processes is supplied by burning the wood dust. Since this process does not require any electric power, it has not been considered in the analysis of the flexibility potential. The processes for sawing, milling, drilling, and sanding are classified as controllable loads; all the other loads are classified as noncontrollable loads. The classification of controllable loads is based on those processes which are operated by workers. The viable flexibility that those processes can exploit is, therefore, constrained by the flexibility of machine operators (human flexibility). The typical loads of offices are also included among the noncontrollable loads. The total number of workers ranges between 12 and 14. Among them, four workers are not directly involved in the industrial processes and do not operate any kind of machine.

After analyzing the data collected, it turns out that the maximum power consumption varies between 80 kW (in winter) and 100 kW (in summer) based on seasonal fluctuations. Figure 13 shows a typical power demand profile of a working week in the winter season. The manufacturing processes of the controllable loads typically take less than 300 s. The processing time affects the energy consumption of the machines. Figure 14 and Figure 15 show two examples of the energy consumption distribution for the banding and saw machines. The sawing process generally takes a few seconds. Therefore, statistically, about 25% of the time that this process is active, it consumes less electricity (about 1.2 kWh). On the other hand, the banding process takes different minutes. When this process is used, it generally demands about 2–3 kWh.

Figure 16 illustrates the variations in power flexibility (${\phi }_P)$ across different hours of a casual day for each season—spring, summer, fall, and winter when the carpentry factory was operating without NZEF solutions. In the case of ${\phi }_P>0$, the demand decreases. In the case of ${\phi }_P<0,$ the demand increases. Negative values during morning in the spring and summer indicate a better alignment between demand and solar generation, highlighting the potential for optimization.

The theoretical flexibility that combines both the flexibility of power and the temporal flexibility is graphically represented by flexibility duration curves. They provide a graphical representation of the power and duration of flexibility that can be exploited in an industrial process. This tool is crucial for identifying how much electrical load can be postponed (see Figure 13) or, in some specific cases, anticipated (see Figure 14) to align the energy demand with the power generated by iRES. Flexibility in the analyzed carpentry factory is mainly offered by the drilling and sanding processes, which are typically executed in the final stages of production. These processes allow significant flexibility because intermediate products can be stored temporarily without affecting the overall production flow. Drilling operations, for instance, can be postponed, aligning them with periods of higher solar generation. Similarly, sanding operations, which follow drilling, offer comparable flexibility. To understand the flexibility duration curves, the density of the curves must be considered. The higher the density in a particular part of the diagram, the higher the probability of exploiting power flexibility (y-axis) by postponing or anticipating the load (x-axis) to another moment. For example, the flexibility duration curves for all controllable loads show that it is possible to postpone up to 30 kW for 3 h (see the red dots in Figure 17) or up to 10 kW for 8 h, ensuring better alignment with renewable energy availability while maintaining production efficiency. Conversely, the flexibility to anticipate processes is more limited. It probably accounts for 0.5–1.0 kW up to 1 h before being scheduled (see red dots in Figure 18). Typically, sawing is the first operation of the day, leaving little room for rescheduling in advance. A potential exception could involve the use of a smart air compressor. In such a scenario, the compressor could start loading the compressed air tank early in the morning, as soon as the PV system begins generating electricity—often as early as 06:00 during the summer months in Germany—well before the production processes (sawing, drilling, milling, and sanding) commence. This strategy would allow the facility to store energy in the form of compressed air and reduce peak electricity demand later in the day.

5.2. Exploitation of Identified and Quantified Flexibility

Since the flexibility of the processes at the carpentry factory analyzed is human-constrained, two other flexibility options have been designed. The first considers a stationary battery with a power capacity of 20 kW and an energy capacity of 20 kWh. The second option considers a charging station for electric vehicles enabling one to charge at the same time up to 22 kW (2 × 11 kW). Flexibility exploitation in carpentry follows a three-layer hierarchy (refer to Figure 19, left). The first layer employs a passive control strategy based on decision tree (DT) algorithms and visual communications. The DT advises machine operators to adjust specific processes to align the power generated by the PV plant with the total load. The advice is communicated to the machine operators visually using a communication system similar to a traffic light (see Figure 19, right). The second layer introduces two Li-ion batteries (20 kW, 20 kWh) which can be activated if the DT-based passive control is not followed by the machine operators. A period of 100 s is given to the machine operator to follow the DT-formulated advice, after that, the stationary battery is activated. A charging station controlling an electric vehicle represents the third layer, employing an on-off strategy to manage power (2 × 11 kW). The electrification of the factory’s logistics expands the available flexibility options. Electric vehicle batteries can act as additional, controllable energy storage units, capable of absorbing excess photovoltaic generation during periods of high solar output and, potentially, supplying power back to the facility or shifting charging to align with renewable availability. This capability supports the integration of volatile renewable energy generation into the factory’s energy balance, thus reinforcing the NZEF objective.

Figure 20 illustrates the identified and designed total theoretical flexibility of each individual option within the carpentry factory. The manufacturing processes, leveraging the passive control developed, can harness up to 40 kW of power for a few minutes, while the batteries can provide up to 20 kW for one hour at their maximum power rate flexibility capacity. At a lower power rate flexibility capacity, the flexibility time increases. The vehicle-to-grid option through the charging station offers a maximum power flexibility of 2 × 11 kW for four hours. In this case, the time flexibility depends on the battery capacity of the unique electric vehicles considered (44 kWh). Figure 21 shows the combined theoretical flexibility (yellow area) to be exploited, considering all the options designed.

5.3. Energy Sustainability

The proposed solutions were tested for manufacturing two different wooden chairs (see Figure 22) in the analyzed carpentry factory. The chairs were produced according to two different operational modes: (i) business-as-usual and (ii) Net-Zero Energy Factory (NZEF). In the business-as-usual mode, the designed flexibility options were not exploited. The integration of the power generated by the PV plant installed on the roof of the carpentry factory building was casual and uncontrolled. In contrast, in the NZEF mode, all the designed flexibility options were managed to fully integrate the power generated by the PV plant.

The energetic sustainability of manufacturing these chairs was recorded in the developed blockchain solution. The results for the business-as-usual scenario and for the NZEF mode are reported in reference [75] and reference [76], respectively. The tests were conducted in summer. In the business-as-usual scenario, 79.68% of the electricity demand was supplied by the on-site PV generation. In the NZEF mode, the chair was manufactured entirely using the power generated by the PV plant.

Figure 22. Chairs manufactured in the business-as-usual mode. References [75,76] show the energetic sustainability in the business-as-usual and in the NZEF modes, respectively.

Figure 22. Chairs manufactured in the business-as-usual mode. References [75,76] show the energetic sustainability in the business-as-usual and in the NZEF modes, respectively.

5.4. Cost–Benefit Analysis of Implementing Flexibility and Sustainability Tracking in Net-Zero Energy Factories

The transition to NZEFs for facilities with legacy PV installations that no longer benefit from feed-in tariffs (typically available for 20 years in Germany) presents a compelling opportunity for both economic and environmental gains. To assess whether designing an NZEF with the necessary infrastructure for flexibility and sustainability tracking is an attractive investment for facility operators, a detailed cost–benefit analysis was conducted. The values of the business-as-usual scenario were considered as a benchmark. This scenario assumes the facility operates without any flexibility design or blockchain-based digital solutions for tracking energy consumption and sustainability. By metering the energy generation of a 125 kW PV plant installed on the roof of the carpentry factory and metering the electricity consumption of the carpentry factory, in the business-as-usual scenario, 31 MWh of the electricity generated is self-consumed annually, meeting 50% of the factory’s total electricity demand of 62.4 MWh. This corresponds to an average daily consumption of about 250 kWh on working days. Assuming an average electricity price of EUR 342.4/MWh, which reflects the average cost incurred by German SMEs (see [19]), the carpentry factory achieves annual savings of EUR 10,680. Table 1 summarizes the main energetic parameters of the analyzed carpentry factory.

Table 1. Energetic parameters of the analyzed case study.

Installed PV Power (kW)	Yearly Generated Electricity (MWh/Year, After 20 Years) *	Yearly Consumed Electricity (MWh/Year)	Self-Consumed Electricity (MWh/Year)
125	105	62.4	31

* A yearly degradation rate of 0.8% was considered [77].

Table 1. Energetic parameters of the analyzed case study.

Installed PV Power (kW)	Yearly Generated Electricity (MWh/Year, After 20 Years) *	Yearly Consumed Electricity (MWh/Year)	Self-Consumed Electricity (MWh/Year)
125	105	62.4	31

* A yearly degradation rate of 0.8% was considered [77].

The amount of savings is achieved without additional investment and does not account for the electricity generation costs of the PV system, whose investment has been fully amortized within the first 20 years of the plant’s operation. The remaining 74 MWh of electricity produced includes both the excess generated during the working days and all electricity generated on weekends when the factory is not operational. Since the cost–benefit analysis assumes that the facility does not benefit from any feed-in tariffs, this electricity is exported to the grid without remuneration, offering no additional economic benefit.

Additional investment is required to design the carpentry factory as an NZEF.

Table 2. Summary of the investment, installation, and implementation costs to design the carpentry factory as an NZEF.

Component	Specification	Cost (€) for Investment and Installation
Infrastructure for Flexibility
Battery storage systems	Configuration: 20 kW; 20 kWh	15,000
Electric vehicle charging stations	22 kW (2 × 11 kW)	7000
Digital Infrastructure
IoT sensors and energy monitoring	Smart meters and barcode scanners	35,000
Data management platform	Blockchain implementation	20,000
Training Costs
Training personnel	Initial training	2000

summarizes the investment, installation, and implementation costs for the carpentry facility. While the investment and installation costs for the hardware components listed reflect actual expenditures from 2022, the costs of implementing blockchain solutions and operational expenses were estimated considering the man-months required for designing the blockchain implementation at the carpentry factory.

To evaluate the incremental benefit of storage on self-consumption, a single daytime PV peak and limited anticipation of production load are considered. Accordingly, one full charge–discharge cycle per working day (250/year) and a round-trip efficiency of 80% for the battery are used. The battery is charged exclusively from on-site PV surplus (no grid charging and no PV capacity increase).

Under these conditions:

• With a 20 kWh battery (power 20 kW), self-consumption increases by about 4.0 MWh/year, reaching 35 MWh/year (about 56% of the demand).
• With a 40 kWh battery (power 20 kW), self-consumption increases by 8.0 MWh/year, reaching 39.0 MWh/year (about 62% of demand).
• With a 160 kWh battery (power 20 kW), the annual increase in self-consumption is 32.0 MWh (100% of its annual electricity demand). Seasonal/weekend surpluses may still occur; the claim refers to the annual energy balance.

Table 3. Summary of the economic benefits due to the self-consumption.

Benefit Type	Scenario	Value (€)	Additional Savings (€) *
Energy Savings	Business-as-usual (50% self-consumption, 31 MWh)	10,680	0
	20 kWh battery, 56% self-consumption (35.0 MWh)	12,050	1370
	40 kWh battery, 62% self-consumption (39.0 MWh)	13,419	2739
	~160 kWh battery (100% self-consumption, 62.4 MWh)	21,431	10,751

* Monetary values computed at EUR 342.4/MWh.

reports the economic outcomes due to the self-consumption gains to operate the carpentry as an NZEF.

A sensitivity analysis was conducted to evaluate the cost–benefit trade-offs of varying battery storage capacities, considering an investment and installation cost of EUR 750 per kWh of storage capacity and the other costs listed in Table 2. The analysis highlights that increasing self-consumption by adding energy storage capacity is not economically viable under the given assumptions, as the investment costs exceed the financial benefits derived from increased self-consumption. The results indicate that even for a smaller storage capacity of 20 kWh, the associated costs outweigh the benefits of self-consumption (see

Table 4. Cost–benefit analysis considering self-consumption only.

	Digital Infrastructure and Training–Without Energy Storage	Energy Storage Capacity 20 kWh	Energy Storage Capacity 40 kWh	Energy Storage Capacity 160 kWh
Costs	EUR 64,000	15,000	EUR 30,000	EUR 120,000
Total costs	EUR 64,000	79,000	EUR 94,000	EUR 184,000
Total benefits due to the self-consumption	EUR 10,680	12,050	EUR 13,419	EUR 21,431
Sum of costs and benefits (net balance in year 1)	EUR −53,320	−66,950	EUR −80,581	EUR −162,569

).

The analysis reveals a crucial factor that can improve cost–benefit outcomes: the market premium enabled by blockchain-based solutions. These solutions provide transparency regarding the energy sustainability of production, which can significantly enhance product value and revenue. The results suggest that a 10% increase in revenue (market premium EUR 150,000 per year on EUR 1.5 million turnover) can offset costs for battery capacities up to 40 kWh, making the investment economically justifiable. For larger capacities, such 160 kWh, the required market premium to achieve economic feasibility would need to exceed 10%, implying that the added value of transparency would need to be significantly higher to justify the investment. While the direct economic benefits of increased energy storage are limited, the integration of blockchain technology and its potential to drive substantial market premiums emerges as a pivotal factor for achieving cost-effectiveness. Combined with the advantages of process flexibility, which does not require upfront investment, a strategic approach focusing on these elements can significantly improve the overall economic and environmental performance of NZEFs.

Table 5. Cost–benefit analysis considering self-consumption and market premium.

	Without Energy Storage	Energy Storage Capacity 20 kWh	Energy Storage Capacity 40 kWh	Energy Storage Capacity 160 kWh
Benefit for market premium 5% in €	EUR 0 *	EUR 75,000	EUR 75,000	EUR 75,000
Benefit for market premium 10% in €	EUR 0 *	EUR 150,000	EUR 150,000	EUR 150,000
Total benefit including self-consumption and market premium 5%	EUR 10,680	EUR 87,050	EUR 88,419	EUR 96,431
Total benefit including self-consumption and market premium 10%	EUR 10,680	EUR 162,050	EUR 163,419	EUR 171,431
Sum of costs and benefits considering market premium 5% (net balance in year 1)	EUR −53,320	EUR +8050	EUR −5581	EUR −87,569
Sum of costs and benefits considering market premium 10% (net balance in year 1)	EUR −53,320	EUR +83,050	EUR +69,419	EUR −12,569

* Market premium is applied only to configurations that include the blockchain layer and NZEF-enabling measures (≥20 kWh storage and process flexibility); the digital-only case does not qualify.

summarizes the cost-benefit analysis when self-consumption and market premium are considered.

5.5. Discussion

The results of the analyzed case study show that meaningful flexibility exists even in human-operated, low-automation job shop processes when it is

• Made visible through fine-grained metering;
• Organized with simple, human-centric scheduling cues;
• Buffered by small, targeted storage and EV charging.

The three-layer control (operator guidance→stationary storage→EV charging) addresses different time scales and uncertainties: operator guidance resolves frequent, short mismatches; storage covers residual intra-day gaps; and EV charging adds opportunistic absorption capacity when vehicles are present. Together, these layers convert intermittent PV into actionable self-consumption without restructuring the production system. In addition, the results also suggest that in designing NZEF a different path must be used: a measurement-first, scheduling-first strategy where storage is sized to complement (not replace) process flexibility, and where digital transparency creates an additional value stream. Rather than pursuing theoretical optimality, the approach privileges operability and low organizational friction—crucial in craft-based settings. The addition of product-level, permissioned-blockchain traceability reframes NZEF from a cost-reduction problem to a value-creation opportunity, making small storage tiers economically viable when paired with traceable sustainability claims.

The results support several practitioner-oriented design rules:

• Meter first, then schedule: sub-meter machines, classify loads, and derive flexibility-duration curves before investing in assets; use simple visual cues to align flexible steps with PV-rich windows.
• Right-size storage: size batteries to daily surplus patterns and operator adherence, not to annual energy balancing; treat EV charging as an opportunistic sink rather than a guaranteed resource.
• Monetize transparency: if sustainability is tracked at the product level (QR-linked proofs), premiums can materially change the business case; governance and data quality are essential for credibility.
• Co-design energy and operations: exploit material buffers and late-stage processes (e.g., finishing) as primary levers; avoid rescheduling that jeopardizes takt or delivery reliability.

6. Conclusions

Designing industrial facilities as net-zero energy factories (NZEFs) supports decarbonization, strengthens public acceptance of the energy transition, and assists system operators. A practical path for low-automation SMEs combines process flexibility, small, targeted storage, and blockchain-based transparency to create market value while keeping costs under control.

The study presents a straightforward method to design NZEFs by identifying and quantifying flexibility. The steps include: describing the manufacturing process, mapping each step on the line, metering electricity use with smart devices, analyzing the data, classifying loads into controllable and non-controllable, comparing scheduled and flexible loads, assessing how much load can be advanced or postponed, and building flexibility-duration curves. Using these steps, the method estimates both theoretical and practical flexibility.

A German carpentry factory serves as an example of a typical low-automation SME. The same approach is replicable in many similar firms. The implementation logic can be identified as follows:

• Identify flexibility in the process;
• Add small storage only where it is effective;
• Use blockchain to make sustainability transparent and to capture a market premium.

The main results of the analyzed case study are summarized as following:

• Process flexibility: The carpentry factory can shift ≈30 kW for ≈3 h or ≈10 kW for ≈8 h (about 30% of peak load). Peak power is 80–100 kW; most operations last for <300 s (about 2 kWh for sawing; 2–3 kWh for banding). This enables better matching with on-site PV without affecting throughput.
• Self-consumption: From a business-as-usual baseline (50%, 31 MWh, 10,680 EUR/year), adding storage raises self-consumption to 56% with 20 kWh (12,050 EUR/year) and to 62% with 40 kWh (13,419 EUR/year). A ~160 kWh battery can reach ~100% (21,431 EUR/year), but requires high investment.
• Economics at current costs: With installed storage at about 750 EUR/kWh, storage alone is not attractive under the adopted assumptions; large batteries (e.g., 160 kWh) do not pay back from self-consumption savings only.
• Transparency and revenue: Tracking product sustainability on-chain can unlock a 5–10% market premium (75,000–150,000 EUR/year on EUR 1.5 M turnover). At 10%, small-to-medium storage (e.g., 20–40 kWh) becomes economically justified, while 160 kWh remains marginal unless premiums are >10% or additional value streams (e.g., demand-charge reduction) are added.

As feed-in tariffs will expire for many SMEs, NZEF adoption is expected to progress naturally: for firms that no longer receive remuneration feeding-in electricity generated by iRES, maximizing on-site self-consumption becomes the winning strategy. This requires creating operational flexibility so that variable renewable generation aligns with production; small, targeted storage can be added only where it closes residual gaps. Integrating on-site renewables directly into the process improves the ecological sustainability of the firm and reduces purchased energy. That improvement also has market value when it is measured and verified: a light digital layer—such as blockchain-based tracking and product QR codes—can document energy provenance and enable price premiums and preferred procurement. In this way, flexibility, modest storage, and transparent reporting provide a continuous, market-driven path for SMEs to scale NZEF practices.