Enhancing EBITDA per tonne has been a constant pursuit by cement manufacturers owing to rising input costs. The current quarter performance is better than the last, but the performance is yet to reach desired levels. Companies in the cement sector have been exploring ways to bring in innovative and sustainable ways to minimise the RM inflation impact and protect their EBITDA performance. Many leaders in the sector have turned to Digital methodologies and accelerators to bring in precise improvements in a scalable and sustainable way.
There are many digital enablers such as Artificial intelligence (AI), Machine learning, Video Analytics and recently Gen AI is also a great support. But the strategy to get maximum returns from Digital technologies is to identify those business-critical areas, choose the right digital accelerator and apply them to get the maximum efficiencies.
Cement manufacturing value stream starts from Mining where the limestone is extracted, passing through the crushers, Raw mill, Kiln, where the conversion happens to a clinker, and then to grinding & packing towards dispatch. Asset reliability, Energy consumption and RM consumption control become critical parameters in Cement manufacturing and Digital accelerators are applied in these areas to optimise operations.
This article talks about 10 different areas with its accelerators to bring in sustainable benefits.
Digital Accelerator 1: Intelligent mine mix optimizer powered by AI
Currently in many cement plants, mine planning is manual and is done through offline coordination between mining team, geologists and QA team. There is no realtime analysis to take informed decisions which leads to variation in ore quality which in turn affects clinker quality. Further this also increases rehandling of ore to balance quality and unplanned movements of HEMM. An intelligent mine mix optimiser powered by AI can ingest different data points such as geological data, mine sampling analysis data, past extraction data etc to arrive and guide the optimal mine mix realtime. The AI can provide inputs to the dumpers and loaders on field based on realtime outputs. This mine mix optimiser can help in reducing variability to a significant level.
Digital Accelerator 2: Digital Twin of Raw mill and Cement mill for energy optimisation
Vertical roller mills are one of the core energy consuming equipment in a cement plant. Currently energy consumption is managed by adjusting 15 to 20 parameters manually every shift leading to significant variation in power consumed. Realtime optimisation on 15 to 20 parameters may not be manually possible and is prone to errors. Further no detailed analysis is done correlating output quality with energy consumption. Digital twins are a powerful option to simulate conditions to arrive at the best possible energy management strategy along with increased grinding efficiency.
Digital Accelerator 3: AI powered fuel optimizer for kiln
Energy consumption in kilns forms a significant part of the overall energy costs incurred in cement manufacturing. It is determined based on different parameters including raw meal consistency, nozzle feed rate, burning zone temperature, kiln motor current, ID fan flow rate, etc. Based on experience maturity spectrum, we have seen huge variability in kcal/kg across different plants. A fuel optimiser for kiln powered by AI, learns based on past data and suggests right setting parameters that gets fed into DCS in a cement plant, delivering the best kcal/kg based on nature of input parameters.
Digital Accelerator 4: Machine Learning based Predictive Asset Management
Reliability is a key factor in cement plants that ensures uninterrupted output. Stacker, Gear boxes, ID Fans are some of failure prone equipments affecting Kiln run days. We have seen in some of the plants that the reliability hovers around 68-75% because of these issues affecting performance. A machine learning based Predictive Asset Management program helps in anticipating failures in advance and align preventive maintenance practices accordingly. Data is collected from respective machine using sensors to predict burn wear, shaft misalignments, coupling failures etc. This approach helps in enhancing reliability higher than 90%.