Towards Leveraging Artificial Intelligence for Sustainable Cement Manufacturing: A Systematic Review of AI Applications in Electrical Energy Consumption Optimization

Abstract

Cement manufacturing is known for its significant energy consumption and environmental footprint. As the world strives for sustainability, optimizing electrical energy consumption (EEC) in cement manufacturing is essential for reducing operational costs and minimizing the industry’s environmental impact. This systematic review aims to synthesize and analyze existing scholarly works and industry reports on methods and approaches for EEC optimization in cement production. It examines papers published between 1993 and 2023 in academic databases, scholarly journals, and industry publications to identify open questions and areas where future research may be needed. While challenges remain, continued research and innovation are key to further advancements in energy efficiency in cement production. With the advent of Industry 4.0 digitalization and advancements in data analytics and industrial Internet of Things (IIoT), artificial intelligence (AI) can be leveraged to optimize EEC. This study is a review of the applications of artificial intelligence to EEC optimization in industries that have heavy demand for electric power to highlight the value of directing research to its applications in cement manufacturing. The study posits that with digitalization, applying artificial intelligence to extract operational insights from the data collected from embedded sensors and meters at the plant presents the most cost-effective, high-return, and low-risk opportunity to optimize EEC in cement manufacturing.

1. Introduction

The cement manufacturing industry is vital for global infrastructure development and is at the nexus of environmental sustainability and energy efficiency. The sector, recognized for its significant energy consumption, is therefore a focal point for research and innovation in optimizing energy usage [1]. The main stages involved in cement production are extraction and preparation of raw materials, production of clinker, and cement grinding. In the first stage of the process, raw material extraction and preparation starts with the grounding of limestone (CaCO₃), together with other minor components, and heating it to 900 °C in a kiln with the help of cyclones. The mixture enters into a rotary kiln where it goes through a series of reactions at temperatures of up to 1500 °C, which produces a calcium silicate chemical compound called cement clinker. Finally, it is cooled, grounded, and mixed with gypsum to make cement. Process modifications and energy efficiency improvements offer mitigation pathways to reduce CO₂ emissions caused by process-related and fuel-related emissions. The process-related and fuel-related emissions account for around 40% of the direct emissions experienced during cement manufacture.

Significant research efforts have been directed at carbon-reduction process improvements in cement production, including exploring alternative sources for the energy used in cement production and finding ways to reduce CO₂ emissions due to the calcination process that occurs during clinker formation [2,3,4,5,6,7,8]. However, going by the literature search, the research on energy use in cement has been mostly focused on thermal energy, even though EEC is also significant and leaves a carbon footprint if the direct source of the electrical energy is from fossil fuel power plants or natural gas power plants. A typical modern cement plant is estimated to consume up to 110 kWh of electricity for grinding a ton of cement [9]. About 95 million metric tons of cement was produced in the United States in 2022, using up more than 8.9 billion kWh of electricity with resultant emissions of an estimated 4.47 million tons of CO₂ by the power plants which typically use fossil fuel or natural gas to generate the electric power used for manufacturing [10,11]. This fact highlights the potential value of optimizing EEC in cement manufacturing and its impact on carbon footprints.

Despite its obvious purpose, research on the implementation of electrical energy-saving measures at the plant level has been surprisingly slow. Hence, this systematic literature review aims to articulate the following:

• The research that has been performed and the methods that have been investigated in the optimization of EEC in cement manufacturing.
• The contributions of previous investigations to EEC optimization and the gaps that need to be addressed.
• The application of AI and other methods and approaches applied to EEC optimization in other power-consumption-heavy industries.
• The potential areas of future research in EEC optimization in cement manufacturing, and the inherent opportunities to leverage AI applications.

At the nexus of sustainability and industrial development, cement manufacturing presents a unique challenge in curbing energy consumption without compromising the quality and quantity of production. This study contextualizes the significance of optimizing EEC, addressing the industry’s demand and its implications for environmental sustainability. The rationale behind this study stems from the urgency to reduce the environmental impact of cement manufacturing while ensuring its operational efficiency [5,12]. The electric power the cement plants consume is significant, and the generation of the electricity consumed contributes to the carbon footprint due to the indirect impact of the power plant which uses either fossil fuel or natural gas to generate electricity. There are research efforts on alternative ways to generate electricity; however, there does not seem to be as many research efforts focused on optimizing electricity consumption in cement manufacturing. Analyzing and synthesizing the existing body of knowledge is pivotal to identifying gaps, novel advancements, and future directions for optimizing EEC in this vital industry. This study aims to critically examine and consolidate insights from a wide array of scholarly articles, industry reports, and research studies.

Discovering a dearth of research on the optimization of EEC in cement manufacturing, this literature review explores approaches adopted in other industries with high power demands. It further explores the relatively untapped potential for leveraging artificial intelligence to optimize EEC in cement manufacturing. Thus, it contributes to the evolving discourse on sustainable practices within cement manufacturing and endeavors to identify opportunities for innovation, foster informed decision-making, and propel the cement manufacturing industry toward a more energy-efficient and environmentally conscious future.

2. Methods

The systematic review process described in [13] was adopted for this study. The first stage of the process includes identifying a need for a review and developing a review protocol. In the second stage, we select the papers, assess their quality, and extract and synthesize data and relevant information. The final stage involves analyzing and interpreting the content, reporting the findings, and making recommendations for further research. The article selection process was streamlined into four steps: keyword selection, searching relevant databases, criteria-based shortlisting of articles, and selection of articles for in-depth review.

We selected search terms from the scoping study and held discussions within the review team to conduct a comprehensive and unbiased search. The keywords and term adopted was “electric energy consumption optimization in cement manufacturing” to identify contributions related to electric power optimization, specifically in cement manufacturing. The keywords selected stem from the main concept of the area of study and are intended to aid the identification and location of the relevant literature and resources from various databases and search engines.

All contribution types, including articles, conference papers, and literature reviews published between 1993 and 2023, were considered, and relevant ones were included in the study. The search was conducted across the Google Scholar, Scopus, IEEE EXplore, and PhilPapers databases. These databases contain scholarly resources relevant to the search topic.

The most relevant 120 from a sample of peer-reviewed publications matched on Google Scholar (each with at least 100 citations) were assessed. Most of the matches came from a match on power optimization; however, most were not cement manufacturing applications. Other relevant articles were selected from the Scopus, IEEE EXplore, and PhilPapers databases. The literature survey indicates that studies on power consumption optimization in cement manufacturing are limited in number and scope, and that this paper can contribute to a better understanding of strategies for optimizing EEC in cement manufacturing. The selection criteria are summarized in

Table 1. Literature review selection criteria.

Items	Criteria for Selection
Keywords/Term	“Electric energy consumption optimization in cement manufacturing”
Databases	Google Scholar, Scopus, IEEE Explore, and PhilPapers
Source	Article title; Abstract; Keywords
Time frame	1993–2023
Document typology	Article; Conference review; Literature Review; Conference paper

below.

Research articles were identified and selected based on their focus and relevance to power optimization in cement manufacturing. Then, we performed a backward–forward search by reviewing the references of the articles in our initial selection and identifying relevant articles that cited the ones in our initial sample [14]. After screening titles and abstracts, full-text articles were reviewed for eligibility, and 25 papers were finalized for detailed analysis. These papers’ full texts were evaluated in detail to identify the main themes and perspectives. Figure 1 highlights the steps adopted for selecting the literature reviewed.

In comparison to applications in cement manufacturing found in the literature, we reviewed the applications of AI in EEC optimization in manufacturing industries that have a heavy demand for electric power, including metals and mining, chemicals and petrochemicals, electronics and semiconductors, food and beverage processing, pulp and paper, automotive and transportation equipment, and textiles and apparel.

The quality assessments adopted include exploring and comparing different sources of knowledge for data quality, reviewing theoretical adequacy, and confirming that the claims made are generalizable logically and theoretically from the data [13]. Our analysis followed the descriptive design highlighted in the structuring content analysis approach of [15], which analyzes the texts to find the power optimization categories and register the occurrence of the categories. For instance, the adoption of power-efficient equipment occurred frequently in the literature, suggesting significant cumulative power-saving effects from implementing multiple strategies.

We organized the discussion on power-efficient equipment around processes in the different heavy power consumption subsystems of cement manufacturing. These include general electrical power management, gas handling and pneumatic systems, motors and transmissions, and comminution and separation. Other major categories in the literature include automation and process control along with adopting Industry 4.0 concepts. These categories formed the key themes around which we frame our findings in the following section.

3. Results

The literature review findings will be presented and discussed in the following sub-paragraphs, identifying opportunities for EEC in electric power management, gas handling and pneumatic systems, motors and transmissions, and comminution and separation. First, we provide a descriptive analysis of the relevant literature. Subsequently, we describe energy-saving opportunities from the cumulative effect of adopting power-efficient equipment. Then, we highlight the opportunities within other categories, such as automation and process control, digitalization, and adoption of Industry 4.0 concepts. We also examine how other industries that have heavy demands for electrical energy implement EEC optimization by leveraging AI. The potential role of advanced analytics and AI in achieving electric power optimization in cement manufacturing specifically is further discussed, focusing on its potential to be leveraged for energy data analysis, predictive modeling, process optimization, and integration with energy management systems.

3.1. Descriptive Results

The research trend (Figure 2) indicates recent interest in electrical energy optimization in cement manufacturing.

There is a recent increase in awareness of the importance of energy conservation and environmental protection. With EEC accounting for more than 60% of the total consumption and the higher cost of electricity linearly increasing the cost of production [5], energy conservation in cement manufacturing is key to the economical and sustainable production of cement. Optimizing electrical energy consumption in the production process can improve energy efficiency, reduce operational costs, and mitigate environmental impacts. It is noteworthy that all nine of the shortlisted publications that focused on digitalization and the application of Industry 4.0 concepts were all published between 2018 and 2023, and involved the use of simulation, modeling, or analytics.

Table 2. Categories of electric-energy-saving opportunities in the literature.

Electricity Saving Category	Source Types	Number of Papers	Literature References
Energy-efficient equipment and technologies	Conference proceedings, Journals	6	[16,17,18,19,20,21]
Pyro-processing, grinding, and milling opportunities	Journals	2	[22,23]
Raw material, recycling, and the circular economy	Journals	2	[24,25]
Waste heat recovery and power generation	Journals	3	[26,27,28]
Automation and process control	Journals	5	[20,21,25,29,30]
Digitalization (Industry 4.0)	Conference proceedings, Journals	8	[30,31,32,33,34,35,36,37,38]

groups the literature [16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38] into the different electrical energy-saving categories elaborated on in the papers. Many of the papers covered multiple categories, compared potential energy-saving opportunities, and aligned with the concept of accumulating power savings by adopting multiple power-saving methods.

3.2. Energy-Saving Opportunities

The preparation process of raw materials in cement manufacturing includes quarrying, crushing, grinding, blending limestone with other materials, calcination in kilns, and clinker cooling. These, along with the final cement grinding, are all electrical-energy-intensive processes. Figure 3 is a schematic representation of the cement manufacturing process, showing the electrical-energy-intensive processes and emissions. The following subsections describe some of the energy-saving opportunities applicable to the cement production process.

3.2.1. Energy-Efficient Equipment and Technologies

Advocacy for adopting energy-efficient equipment was found to be common in the literature within the period in consideration. As cement manufacturing equipment technology evolved, the advances were applied to curtail electricity costs by reducing specific power usage [16,17,18,19]. Before the first period (1993–1999), there were improvements in fuel energy requirements, and the adoption of alternative fuels increased, while specific electrical energy consumption increased by as much as 40%. Many cement plants in Europe and North America at the time had higher electrical bills than fuel bills. The high bills drove innovations that created power reduction opportunities within electrical power management, gas handling and pneumatic systems, motors and transmissions, and comminution and separation [16].

With modern electrical power management technology, cement plants are digitalized, and computerized monitoring of motors and starters is implemented so that all the data collected can be analyzed to develop control strategies. These data-driven power management strategies include production planning to shift loads and maximize on-peak and off-peak demand power rates, preventing unnecessary equipment idling, and studying the causes of power variation.

Worrell et al., in their research [17], examined strategies for decreasing energy consumption and carbon dioxide emissions in US cement production by gathering data on the costs, energy conservation, and reductions in carbon dioxide emissions associated with various technologies and approaches. These methods were classified into two groups: existing measures widely adopted in cement plants globally and innovative measures that were either minimally implemented or on the brink of commercial viability. Using an energy conservation supply curve, the authors assessed the cost-effectiveness of the potential for improving energy efficiency. Energy conservation supply curves compare investments in energy conservation against investments in energy supply in order to take the least-cost approach to meeting energy needs. According to the findings, switching products to blended cement resulted in total fuel savings of 1.41 GJ/t of blended cement but an estimated increase in electricity consumption of 17 kWh/t from grinding the blending materials. Estimated electricity savings from energy-efficient technologies for cement making in the US in 1994 ranged from 0.03 GJ/t from using high-efficiency roller mills to 0.07 GJ/t from adopting heat recovery for power generation. The International Energy Agency reported that, from 2010 to 2020, the thermal energy intensity of clinker decreased by 0.2% annually [39]. However, it has since plateaued at around 3.6 GJ/t. This decline brought about a rise in the sector’s electricity intensity, estimated at 100 kWh/t cement since 2022. This indicates that some of the advancements in thermal energy efficiency in cement manufacturing resulted in increased electrical energy consumption within the sector.

Common to many of the papers reviewed is an advocacy for the adoption of methods that have the potential to save energy. Al-Mansour et al. [18] submitted that prompt activation of savings opportunities yields economic advantages and reduces both direct and indirect greenhouse gas emissions. The analysis of Slovenia from 1997 to 2020 illustrated that promptly implementing energy efficiency measures in the industrial sector is economically justified. Using analytical methods with a ‘least cost’ economics analysis approach, the authors opined that enhancements to internal industrial conversion systems, such as the co-generation of electricity and heat, contribute significantly to overall efficiency improvements. The ‘least cost’ method identifies the scope for cost-effective energy savings as a function of energy cost and investment criteria.

The report in [19] is a guide for energy and plant managers that covers more than forty energy-efficient technologies implemented in cement plants worldwide. It assesses the technologies’ effectiveness in terms of energy and carbon dioxide savings, investment costs, and operation and maintenance expenses. It recommends a corporate-wide energy management program that advocates for changes in staff behavior and attitude, energy efficiency training programs, and a formal management structure for managing energy.

Energy-efficient technologies are often combined with process controls to reduce energy consumption and carbon emissions [20,21]. Madlool et al. [20] surveyed energy efficiency measures for raw materials preparation, clinker production, products and feedstock changes, and finish grinding.

Table 3. Aggregation of estimated electrical energy savings and emission reductions reported by Madlool et al. [20]. The item with * is the only item in the table measured in GJ/t. All others are measured in kW h/t.

Energy–Efficiency Measures	Electrical Energy Savings (kW h/t, *GJ/t)	CO2 Emission Reduction (kg/t)
Efficient transport systems for raw materials preparation—pneumatic systems, mechanical conveyors, and pipe conveyors	2–3.4	0.41–3.22
Raw meal homogenizing systems	1–4.3	0.26–2.73
Raw meal process control for vertical mills	1.02–1.7 *	0.16–1.45
Use of roller mills	6–11.9	1.24–10.45
High-efficiency classifiers/separators	3.18–6.3	0.51–5.23
Energy management and process control systems for clinker making in kilns	2.35–5	2.48–16.61
Adjustable speed drive for kiln fan for clinker making in all kilns	4.95–6.1	1.4–6.27
Low-temperature heat recovery for power generation for clinker making in rotary kilns	20–35	4.6–31.66
High-temperature heat recovery for power generation for clinker making in rotary kilns	17.84–22	3.68–9.25
Low-pressure drop cyclones for suspension preheaters for clinker making in rotary kilns	0.66–4.4	0.16–2.67
Efficient kiln drives for clinker making in rotary kilns	0.45–3.9	0.16–0.9
Process control and management in grinding mills for finish grinding	1–4.2	0.9–4.11
Vertical roller mill for finish grinding	10–25.93	8.82–26.66
High-pressure (hydraulic) roller press for finish grinding	8–28	1.28–25.09
High-efficiency classifiers for finish grinding	1.9–7	0.4–2.07
Improved grinding media (for ball mills)	1.8–6.1	0.29–6.27
High-efficiency motors and drives	0–25	0–47
Adjustable or variable speed drives	0.08–9.15	1–9.41
Changing product and feedstock—blended cements		0.3–213.54
Use of waste-derived fuels		12–76.31
Changing product and feedstock—Limestone Portland cement	2.8–3.3	8.4–29.86

summarizes the estimated electrical energy savings and emission reductions reported in the paper from energy efficiency measures at each stage of the cement manufacturing process. The study indicated that as of 2013, the largest recorded amount of electrical energy savings from the adoption of energy efficiency methods was 35 kW h/t.

One factor that should be addressed when implementing efficiency improvements in the manufacturing process is the economic viability of the proposed improvements. In a study that presented energy-saving measures with varying degrees of complexity and investment requirements, cost savings achieved surpassed the investment cost of implementing actions by USD 0.9 million [21]. Implementing energy-efficient technologies led to a 10.45% reduction in electrical energy consumption between 2008 and 2010. Notably, the improvements in process control techniques introduced during the same period resulted in substantial energy savings, with a 22.6% improvement in specific electrical energy consumption. The authors employed the cumulative sum of differences technique to evaluate the effectiveness of the implemented actions. Therefore, with this know-how, there is an argument to be made that the cement manufacturing industry will gain much in cost savings and reduce the indirect impact of CO₂ by looking at innovative ways to optimize electrical usage at the plant level.

The articles reviewed indicate that the adoption of energy-efficient equipment and technologies in cement manufacturing has the potential to reduce energy consumption, electricity costs, and carbon dioxide emissions. In addition, it is important to ascertain the economic viability of efficiency improvements to ensure the manufacturer’s motivation to implement.

3.2.2. Pyro-Processing, Grinding, and Milling Opportunities

Quarry crushing is the start of the preparation of the raw material needed for the manufacturing of cement. The crushing process which also consists of conveying blasted limestone material consumes electricity because of motor drives. The process accounts for a small percentage relative to the other processes. Raw material milling, which follows the quarry crushing to reduce the material to fines for cement manufacturing, also consumes electricity because of the raw mill motor drive. However, compared to the other processes, it is a small percentage. Pyro-processing is a term used to describe the high-temperature thermal treatments that occur during cement manufacturing. The processes involve using intense heat to transform grounded raw materials into clinker, the primary component of cement. Considerable electrical energy is required to power the motors and gears of the kiln drive system, the fans and blowers of the preheater and pre-calciner systems, the clinker cooler system, and the numerous auxiliary equipment such as air compressors, pumps, lighting, and control systems. In a study, raw material processing, clinker burning, and clinker grinding accounted for 28%, 25%, and 32% of electrical energy consumption, respectively [21]. Demand-side energy management measures can provide immediate and short-term improvements in energy efficiency. Examples of these measures include leveraging free and low-cost options such as motor, compressed air, and process heater optimization software tools, energy management training programs, and energy assessments and audits.

After the clinker has been cooled, it is stored in clinker domes or silos, in storage bins, or in heaps outside. It is conveyed to storage, and to the finish mill for further grinding, using belt conveyors, deep bucket conveyors, or bucket elevators, which can also dump it directly into silos or bins for shipment. In order to produce powdery cement, the nodules are ground to a fine powdery consistency. Approximately 3–5 percent gypsum is added to regulate the setting properties of the cement. These powders, the clinker and gypsum, are ground in ball mills or a combination of ball mills and roller presses, in roller mills, or roller presses. Vertical roller mills are technically feasible, but they have yet to be widely accepted in the US. Coarse material generated by all these systems is screened and returned to the mill for further grinding to ensure a uniform surface area of the final product. The electrical power required to grind the raw material depends on the required surface area of the final product, the additives used, the hardness of the materials used (principally the limestone, clinker, and pozzolana extenders), and the desired fineness of the cement [19].

In a study of a methodology to shift electricity consumption to economical off-peak periods, various physical components such as the raw mill, the kiln, crushers, cement mill, and auxiliary equipment were integrated into a simulation model to accurately predict their influence on production and cost [22]. Utlu et al. [23] examined pyro-processing energy-saving opportunities in a Turkish cement plant using actual operational data. They proposed a tool for the analysis of energy and exergy utilization, the development of energy policies, and the provision of energy conservation measures. However, delivering energy-efficient technologies to consumers effectively is a challenge.

The pyro-processing, grinding, and milling processes involved in cement manufacturing are energy-intensive due to the significant electrical energy consumed by equipment such as motors, fans, mills, and conveyors. Some of the potential energy management measures recommended in articles include the use of optimization software, energy audits, shifting electricity consumption to off-peak periods, and the analysis of energy and exergy utilization data for the development of conservation measures.

3.2.3. Raw Material, Recycling, and the Circular Economy

More recent research efforts indicate that value can be derived from adopting recycling and the circular economy concept. The findings in [24] provide an estimate that recycled cement production using cement paste from concrete waste consumes 30% less energy than clinker production. The electrical energy consumption for an industrial setup is estimated by extrapolating laboratory results through simulations and analogies. The energy requirement is approximately one-third lower than conventional methods, as the concrete waste is already pre-crushed, thus reducing the need for extensive crushing to obtain fine aggregates. Additionally, the weaker nature of concrete waste compared to natural rock further contributes to the lower energy demand. However, compared to clinker, using concrete waste as a substitute has limitations that necessitate further investigation to comprehend the differences fully. While concrete waste exhibits lower reactivity and potentially lower strength characteristics, it has demonstrated acceptable results when used to replace up to 40% of Portland cement. Ambient air temperature and moisture content of the raw material also affect the performance of the raw mill and, in effect, the energy consumption. First and second law efficiencies of raw mill increase as ambient temperature increases and the moisture content of the raw materials decreases [25].

In summary, research suggests that cement production from recycled concrete waste has the potential to reduce energy consumption when compared to conventional clinker production; however, its possible limitations in strength and reactivity are worthy of further investigation.

3.2.4. Waste Heat Recovery and Electric Power Generation

Researchers have explored using waste heat recovery to generate power for cement manufacturing, as the process is energy-intensive and generates significant amounts of waste heat [26,27,28]. Mirhosseini et al. [28] confirmed the feasibility and economic viability of generating electricity using thermoelectric generators (TEG) at the cement rotary kiln. The researchers proposed installing a metallic frame to act as an absorber around the cement kiln, maintaining a specific gap from the kiln surface. This configuration serves two purposes: firstly, it facilitates heat recovery by thermoelectric generators, and secondly, it prevents the kiln from experiencing excessive temperature rises and additional weight. The study presented an optimal TEG system design that maximizes power generation by dividing the arc-shaped absorber into ten distinct sections. For each section, the study calculated the optimal design parameters for the TEG and corresponding heat sink, ensuring efficient heat transfer and electricity production.

Sanaye et al. [26] highlighted procedures for the modeling and optimum design of waste heat recovery and power generation systems. The modeling covered energy, exergy, economic, and environmental aspects. They reported on two parallel lines of cement production that were new and specific to the cement plant studied and leveraged a genetic algorithm for multi-objective optimization. The study’s Organic Rankine cycle results showed that using water as a working fluid provided a 9.14 MW power output with an exergy destruction ratio of 47.9%. In contrast, using toluene as a working fluid provided a 6.56 MW power output with an exergy destruction ratio of 53.5%.

Sani et al. implemented waste heat recovery by installing an air-quenching chamber at the clinker cooler’s output and incorporating suspension preheater boilers in the preheating stage [27]. A genetic algorithm was employed to determine the optimal energy and exergy performance evaluation parameters, considering three different working fluids: water, refrigerant R123, and refrigerant R245fa. By optimizing the use of refrigerant R123, the power generation capacity increased significantly from 5 MW to 9 MW, whereas utilizing water as the working fluid only resulted in a modest increase from 4.8 MW to 5 MW. Furthermore, the analysis revealed that refrigerant R123 exhibited a 4.1% decrease in total exergy loss, indicating its suitability as an efficient working fluid for cement plants’ waste heat recovery cycle.

In summary, the feasibility and optimal design of waste recovery systems for generating electricity in cement manufacturing have been investigated by researchers. The approaches they have proposed include using TEG around the rotary kiln, incorporating an air-quenching chamber at the clinker cooler’s output, and suspension preheater boilers in the pre-heating stage.

3.2.5. Automation and Process Control

Automation and advanced process control systems are critical for optimizing electrical energy usage in cement manufacturing. The cement production process is highly energy-intensive, with massive equipment consuming vast amounts of electricity. About two-thirds of the total electrical energy is used to reduce the particle size of the raw materials and clinker. Practical methods to optimize the efficiency of the cement ball mill (CBM) were explored in a study of its energy and exergy in a new-generation cement plant [29]. The electrical energy consumption of the CBM unit was initially specified as 37.9 kWh/t. The study investigated the impact of factors such as ball charge pattern, cement fineness, and the addition of limestone and pozzolan on the performance of the CBM unit and the quality of the produced cement. By modifying the ball charge pattern, the first and second law efficiencies of the CBM increased to 81.8% and 20.6%, respectively, while the electrical energy consumption of the CBM unit decreased to 36.5 kWh/t. Furthermore, the results demonstrated that as the cement fineness decreased from 3250 cm²/g to 2820 cm²/g, the cement production rate increased from 185 t/h to 224 t/h, and the electrical energy consumption decreased from 41.1 kWh/t to 33.1 kWh/t. However, reducing the cement fineness affected the cement compressive strength (at 3, 7, and 28 days), which decreased while the cement setting time (initial and final) increased.

Gangwar et al. [30] developed a simulation optimization method for scheduling energy-intensive industries under the uncertainty of the spot electricity market. This method, which can be adopted by the cement manufacturing industry for optimizing electricity costs, uses a Monte Carlo scenario generator to generate future electricity prices and calculates the potential outputs for each of the plausible forecasted electricity price scenarios independently. Overall, the model identified the most profitable ways to operate an air separation plant in Spain under the uncertainty of electricity prices in the future.

By implementing automated control systems, parameters such as equipment speeds, temperatures, and material flow rates can be dynamically adjusted based on real-time feedback to operate at peak efficiency. This prevents inefficient manual operation that wastes energy. Additionally, machine learning and artificial intelligence capabilities can be integrated to predict ideal operating setpoints and detect anomalies that increase energy drain. Automated monitoring also enables predictive maintenance, reducing unplanned outages and optimizing planned downtime for maintenance, avoiding wasteful energy consumption from faulty equipment. With rising energy costs and pressures to decarbonize, investing in automation and process control yields compounding paybacks through electrical energy savings while enhancing the product quality, throughput, and environmental sustainability of cement production.

3.2.6. Digitalization

Potential benefits of digitalizing cement manufacturing include energy optimization, process automation and control, predictive maintenance, supply chain optimization, remote monitoring and control, product quality control, and data-driven decision-making. While digitalization requires upfront investments in technologies, systems, and skills, the potential benefits from efficiency improvements can result in significant cost savings, improved quality, and enhanced sustainability. Another potential benefit of digitalization is the standardization of data which allows the exchange, comparison, and integration of content between different parties in the industry. These provide valuable opportunities for cement manufacturers to remain competitive and adaptable to changing market demands.

Several research findings confirm the potential value of digitalization, subsequent data collection, and the application of advanced analytics to the cement manufacturing process. In one instance, implementing model predictive control (MPC) technologies drove power consumption reductions ranging from 3% to 8%, simultaneously enabling increased production output and reduced fuel consumption rates [31]. In this study, two cement-related industrial MPC applications were presented—a cement raw-mix blending application and a cement mill grinding MPC application. A linear model predictive control algorithm with soft constraints was applied. By leveraging predictive modeling and optimization algorithms, MPC technologies can optimize various process parameters, improving electrical energy efficiency and leading to a higher throughput, contributing to cost savings and sustainability goals.

Tong et al. [32] introduced the digitization work of a smart cement plant in China, which implemented intelligent factory technology using special integrated robots, a software-as-a-service-based cement process digital applicant, artificial intelligence (XGBoost classifier), and advanced process control. Introducing these technologies gave potential energy savings ranging from 2% to 5% from improved process control. The primary production process in cement manufacturing is continuous, making process operation optimization or real-time optimization a crucial aspect. In their paper, Zhu et al. [33] presented a real-time optimization method based on system identification, which is particularly well-suited for energy and utility systems in cement plants. The proposed approach does not require rigorous mathematical models or specialized performance tests, making it a cost-effective solution. By leveraging system identification techniques, this method can optimize processes without complex modeling or extensive testing, thereby providing an efficient and economical way to enhance energy and utility systems in cement production.

In another study that shows the role that simulations and modeling can play in improving production, simulations and modeling studies were conducted to evaluate modifications to the existing cement grinding circuit flowsheet [34]. The proposed modifications involved redirecting the mill filter stream to the final product silo instead of sending it to the classifier feed [Ibid.]. The simulations indicated that this modification would lead to a 4.45% increase in production rate, resulting in corresponding energy savings of 4.26%. This optimization approach not only improved energy efficiency but also enhanced the quality of the end product. Extensive investigations, laboratory validations, and post-production surveys were undertaken to validate the effectiveness of the modifications and ensure that product quality was not compromised. The results confirmed that the proposed changes did not adversely impact the quality of the cement product while delivering the anticipated benefits of increased production capacity and energy savings. Similarly, in another study, the specific electricity consumption of the raw mill workshop of a cement plant in Morocco was optimized while ensuring mill stability and quality control [35]. A multilinear regression model was used to test the effect of the independent variables (mill throughput, hydraulic pressure, and mill outlet temperature) against the target variables of mill-specific electricity consumption and mill vibration.

Ye et al. [36] focused their study on modeling the primary energy-consuming equipment in cement manufacturing plants by utilizing industrial load characteristics and implementing a Markov process model. Subsequently, they used reinforcement learning algorithms to design demand response scheduling methods tailored for industrial applications. By combining the equipment modeling approach with advanced machine learning techniques, the study demonstrated intelligent energy management strategies that can optimize electrical energy consumption and enable demand response capabilities in energy-intensive cement manufacturing facilities. Artificial neural networks and genetic algorithms have also been applied to reduce the cost of electricity by optimizing different variables in the production process and regulating electricity market costs [38].

The best way for the cement industry to confront the challenges facing it and improve operational excellence is to embrace technological advances and innovations [37,40]. The cement industry’s major challenges include cost reduction, environmental protection, and energy/capital efficiency. Technological advances under the “Industrial Internet of Things” paradigm can be leveraged to address these challenges. These include process automation, engineering software, data collection/analysis platforms, secure data transportation, and storage solutions that provide real-time information for informed decision-making, help improve asset utilization, and optimize relationships with customers and suppliers.

3.3. Leveraging AI for EEC Optimization in Other Manufacturing Industries

Research efforts have been directed at reducing electrical energy consumption in other manufacturing industries whose operations have a high demand for electrical energy; however, this is lacking in the cement manufacturing industry. There is evidence in the literature that AI can improve energy efficiency by increasing technological efficiency [41,42,43]. For instance, in the metals and mining industry, the furnaces, electrolytic processes, and other energy-intensive operations required for smelting and refining metals such as steel, copper, and aluminum consume considerable amounts of electrical energy. Artificial intelligence solutions were leveraged in an initiative that helped a steel manufacturer drive higher profit margins while meeting quality and quantity demand at reduced energy consumption levels. Pattern recognition and predictive modeling were applied to signals from sensors embedded in the mill to predict demand flow, asset flow, wear rates, and high-expense consumables used in production [44].

The production of various chemicals, plastics, and petrochemicals uses electric motors, pumps, fans and blowers, compressors, and compressed air systems. Motor systems use about 57% of the total electricity used in the chemical industry. Specifically, 26% of the electricity is used by pumps, 23.6% is used in material processing, 27.7% by compressed air systems, 11.9% by fans, and 7.7% by refrigeration systems. Researchers have developed and tested a method that leverages descriptive analytics and neural network modeling to optimize the use of energy resources in chemical production [45]. Similarly, data analytics has been applied to optimize EEC in smart food processing, achieving a 12% reduction in energy consumption and CO2 emissions [46]. Optimization algorithms, mixed data sampling regression, and genetic algorithms have been used to reduce the electrical energy consumed in mechanical pulp production in the papermaking industry [47]. It is essential that the cement industry explores this novel idea to optimize its electrical consumption. No work in the literature was found during this work which shows that AI has been adopted by the cement industry to help optimize electrical consumption as other industries have done.

4. Discussion

Infrastructural changes required at the cement plant to implement energy optimization often require significant capital investments [37]. Replacing aging equipment with more modern energy-efficient equipment and technologies often requires significant financial and human capital. Cement quality requirements are a constraint to swapping out or recycling raw materials. The adoption of automation and process control in cement manufacturing is getting attention. However, the fully autonomous cement plant is still a distant reality [48]. With innovation in sensors, interconnection of machines, automation devices, techniques based on IIoT, and advancement in cloud data storage and computing, large volumes of data on the operations of the cement plant can be collected. Even a staged implementation of digitalization in cement manufacturing can yield immediate, cost-effective results by applying advanced analytics to the data to generate transformative insights.

Table 4. Analytics methods applied to power optimization in cement manufacturing.

Literature and Country	Themes	Analytical Method	Insights
Worrell, Martin, N., and Price, L. (2000) [17]. United States	Energy-efficient technologies	Energy conservation supply curve	Demonstrated that utilizing blended cement is a crucial and economical approach for enhancing energy efficiency and reducing carbon dioxide emissions in the US cement sector
Al-Mansour et al. (2003) [18]. Slovenia	Energy-efficient technologies	Analytical methods with a ‘least cost’ economics analysis approach	The ‘least cost’ method identifies the scope for cost-effective energy savings as a function of fuel prices and investment criteria. Prompt activation of savings opportunities yields economic advantages and reduces direct and indirect greenhouse gas emissions. Enhancements to internal industrial conversion systems, particularly co-generation of electricity and heat, contribute significantly to overall efficiency improvements.
Afkhami et al. (2015) [21]. United States	Process control and energy-efficient technologies	Cumulative sum of differences technique	Process control improvements resulted in 22.6% improvements in electrical energy consumption.
Sanaye et al. (2020) [26]. Iran	Waste heat recovery and power generation	Genetic algorithm for multi-objective optimization	Results for the study’s Rankine (Organic Rankine) cycle showed that using water (toluene) as a working fluid provided a 9.14 (6.56) MW power output with an exergy destruction ratio of 47.9% (53.5%).
Sani et al. (2020) [27]. Iran	Waste heat recovery and power generation	Genetic algorithm to determine optimal parameters for energy and exergy performance evaluation	Optimizing fluid R123 increased power generation from 5 MW to 9 MW, while water only increased power generation from 4.8 MW to 5 MW. R123 fluid also showed a 4.1% decrease in total exergy loss, indicating that it is suitable for the cycle.
Gangwar et al. (2023) [30]. Spain	Automation and process control, digitalization (modeling, industry 4.0, analytics)	Monte Carlo scenario generator (MCSG), ARIMA (Autoregressive integrated moving averages)	Simulation optimization method for scheduling energy-intensive industries under the uncertainty of the spot electricity market
Zhang et al. (2021) [31]. Turkey	Digitalization (control and optimization)	Linear model predictive control (LMPC) algorithm	Demonstrates that model predictive control (MPC) technologies can reduce power consumption by 3–8% while also increasing production and reducing fuel consumption.
Tong et al. (2023) [32]. China	Digitalization, Industry 4.0 concepts	SAAS-based cement process digital applicant, Artificial Intelligence, and Advanced Process Control (APC)	2–5% potential energy savings from improved process control.
Zhu et al. (2022) [33]	Digitalization (modeling, industry 4.0, analytics)	Real-time optimization	A real-time optimization method based on system identification is especially suitable for energy and utility systems.
Altun, O. (2018). [34]	Digitalization (modeling, industry 4.0, analytics)	Simulation and modeling	This optimizes the energy efficiency and the quality of the end product by modeling and modifying the existing flowsheet of the cement grinding circuit.
Belmajdoub, F., and Abderafi, S. (2018) [35]. Morocco	Digitalization (modeling, industry 4.0, analytics)	Multilinear regression model	Optimization of specific electricity consumption
Ye et al. (2023) [36]. China	Digitalization (modeling, industry 4.0, analytics)	Markov process model and reinforcement learning	Design of industrial demand response scheduling methods using a reinforcement learning algorithm

is a synthesis of analytics methods applied in the literature for power optimization in cement manufacturing. The methods involve modeling and/or simulation of the electrical energy-intensive cement manufacturing process and applying a statistical method, algorithm, or optimization method to the model. Researchers have explored model predictive control applications [31], intelligent factory technology applications [32], real-time optimization [33], multilinear regression models [35], and reinforcement learning [36].

Energy data can be analyzed to uncover opportunities to reduce EEC. Cement plants generate massive amounts of data from various sources, including energy meters, process sensors, and control systems. Advanced analytics techniques can be employed to analyze this data and uncover patterns, trends, and insights related to energy consumption. This analysis can help identify energy hotspots, inefficiencies, and opportunities for optimization. By leveraging historical data and machine learning algorithms, predictive models can be developed to forecast energy demand, production rates, and equipment performance. These models can assist in optimizing energy usage by anticipating fluctuations in demand and adjusting operations accordingly, minimizing energy waste and peak loads. The output of the models can be incorporated into energy management systems that monitor and control energy consumption across the plant.

The use of digital twins and simulation in manufacturing is gaining attention. A high-fidelity digital twin of the cement grinding circuit can be developed using physics-based models and machine learning. Simulations can then be performed to test strategies for energy optimization before implementing them in the actual plant. Digital twins can also be used for virtual commissioning and operator training to improve human–process interaction efficiency.

The applications of AI to process optimization should be explored further for electric-energy-saving opportunities in cement manufacturing. Advanced analytics can be used to optimize various processes within the cement plant, such as raw material grinding, clinker production, and cement grinding. By analyzing data from sensors and meters across the grinding circuit (mill, separators, fans, and conveyors), we can determine the optimal setpoints and operating parameters (mill speed, feed rate, and airflow) that minimize specific energy consumption while maintaining product quality.

The production line of modern cement plants can be partitioned into these four sub-processes: crushing, kiln feed preparation, clinker production, and finish grinding. The cement manufacturing process heavily relies on electricity for various operations, such as crushing and grinding the raw materials, transporting substantial volumes of gases and materials throughout the facility, and grinding the final cement product. The electricity consumption of a grinding workshop is typically modeled as a function of the power demand of the equipment and the throughput of the material in the mill feed [35]. The objective is to optimize specific electricity consumption while keeping the mill operation stable and the product quality and production quantity according to plan. Effectively reducing electricity consumption reduces the carbon footprint due to the indirect impact of the electricity-generating power plants which use either fossil fuel or natural gas. At a high level, the design of the optimization experiment will include the following:

• Identifying and modeling the electrical-energy-intensive components of the production process [49].
• Data measurement and collection of mill-controlled and manipulated parameters such as crushers, conveyors, mill motor and fan power, vibration, bed material height, residue fineness, mill throughput, mill output temperature, mill roller hydraulic pressure, mill fan speed, and separator speed. This will require the installation of sensors and meters where necessary.
• Exploratory data analysis is used to find patterns and uncover insights, helping to determine the linear and non-linear relationships of variables with the objective function and the optimization methods to experiment with.
• Experimental application of optimization algorithms on historical data to determine the values of the input variables that optimize specific electricity consumption under the product quantity, quality, and mill stability constraints.

With the trend of the pricing of the components of IIoT declining over the years, collecting data and leveraging advanced analytics to implement economically viable power optimization solutions in cement plants are more feasible. Technological advancements and economies of scale in production have steadily decreased the costs of various sensors for temperature, pressure, vibration, and other sensing devices. Similarly, wireless connectivity, hardware, and cloud computing costs have also been dropping. Adopting open-source software has helped reduce the cost associated with software development and deployment. All of these suggest that collecting manufacturing process data and applying analytical techniques offer highly cost-effective and low-risk opportunities to optimize EEC and thus improve sustainability in cement manufacturing. Testing this hypothesis with the experimentation of optimization methods on data collected at a cement plant and validating the methods’ economic viability within the plant’s context is the next logical step of this study.

5. Conclusions

This evaluation of EEC in cement manufacturing and the significant carbon footprints associated with generating that electrical energy have underscored the importance of focusing research efforts on optimizing electrical energy usage in the industry. There have been advancements in research to leverage artificial intelligence in other manufacturing industries with high demands for electrical energy, such as the metals and mining, chemicals and petrochemicals, food and beverage processing, and pulp and paper industries. Research in the cement manufacturing industry has been more focused on thermal energy optimization. This highlights the relatively unexplored opportunities to reduce electrical energy consumption and resultant carbon footprints in the cement industry.

The literature reviewed followed electrical energy optimization within the following themes in cement manufacturing: automation and process control, power generation from waste heat recovery systems, energy-efficient equipment, and digitalization with the application of artificial intelligence. Even though many of the papers reviewed did not examine the economic viability of the energy savings methods proposed, it is obvious that these methods will require infrastructure improvements that are capital-intensive and have extended payback periods. By embracing the integration of relatively cost-effective intelligent digital technologies into production processes and decision-making, cement manufacturers can harness the power of data analytics and artificial intelligence to achieve significant electrical energy savings and a resultant reduction in the carbon footprint of the cement manufacturing process.