Skip to main content

Making waves with MPC

Published by , Editorial Assistant
World Cement,

Rockwell Automation explains how the optimisation of manufacturing processes can maximise plant yield, improve quality and enhance sustainability.

Energy-intensive industries are feeling the pressure from all sides. Compounding much stricter international sustainability legislation and targets, recent sharp rises in energy prices are pushing companies to examine their products and processes more closely to ensure that they are using energy in the most efficient way possible. Simple processes may require basic analysis and fine-tuning. Still, plants that produce multiple products and rely on a wide range of high-tolerance parameters are far more challenging to tune and will often see significant and immediate benefits from digital solutions.

One technology that positively affects cement manufacturers’ sustainability is model predictive control (MPC). This technology optimises the current performance, including future prediction periods. This is achieved by optimising a finite horizon but only implementing the recent result. MPC anticipates future events and can take control actions to maintain long-term stability. MPC is a closed-loop technology that receives current information about the process. It uses its dynamic model to predict the process response to a sequence of future optimised moves over a specified horizon.

The MPC controller then configures the sequence of moves for each manipulated variable to predict and control the effect on the controlled variables to simultaneously meet and satisfy multiple targets and objectives. This technology is also known as multivariable model predictive control, given its ability to handle several manipulated variables and control several goals simultaneously.

The controller compensates for differences in the predicted and actual responses by tracking the process response to the first moves and comparing it to the prediction trajectory. This compensation enables the controller to function robustly even when the model is imperfect or unmeasured disturbances impact the process, such as varying quality of coal.

Cement manufacturer improves energy and environmental efficiency

Aalborg Portland was founded in 1889 and is now the world’s largest producer and exporter of white cement, producing more than 1.5 million tpy. Aalborg Portland sells and distributes cement worldwide with production units in Denmark, China, Egypt, Malaysia, and the United States.

Intending to continue its pursuit of environmental and energy improvements at the facility, Aalborg Portland focuses on minimising energy used in production and reducing environmentally harmful emissions such as CO2, SO2, and NOx. This directive has resulted in investments by Aalborg Portland into projects to address energy savings and emissions reduction.

Aalborg selected the Rockwell Automation FactoryTalk® Analytics Pavilion8® Model Predictive Control (MPC) solution to reduce a rotary kiln’s specific energy consumption and environmental impact.

Aalborg also desired to leverage process experience to maximise production, improve quality consistency, and improve plant stability and operation.

The Pavilion8 application uses MPC and soft sensor technology to reduce overall variability and operate at peak process performance. The MPC technology assesses current and predicted operational data, compares the data to desired results, and then computes and updates supervisory online setpoint targets.

Under the scope of the MPC project, the rotary kiln system consists of a two-string, two-stage preheater semidry pre-calciner kiln with a tertiary air duct and kiln bypass. Kiln slurry is a pre-blended mixture of chalk, sand, and water. Around 83% of kiln slurry and several other raw materials are fed directly to the kiln. Petcoke is indirectly fired on the main burner and the calciner. The calciner can also burn solid alternative fuels (RDF) on both strings and dried sewage sludge on string A. The main burner can also burn solid alternative fuel, meat, bone meal, and liquid alternative fuel, such as glycerin. Tyre chips are fed to the kiln inlet, providing additional heat for the calciners and the kiln.

The pre-calciner and main burner oil burners are only used for startup situations. O2, CO, NOx, and SO2 are measured at the kiln inlet, and O2 and CO are also measured at the exit of each preheater string. NH3, CO, NOx, and SO2 are measured at the stack, and kiln amps are also measured. Product-free lime is measured by lab samples every two hours.

The primary project objectives for the MPC on the cement kiln were to minimise specific fuel consumption for required calcination and clinker-free lime levels, minimise electrical energy consumption from ID fans, and excess O2 in the system while helping prevent high CO excursions. The aim was also to maximise alternative fuel usage in the pre-calciner to the allowed constraints set and reduce calciner temperature variability and bed depth fluctuation in the grate cooler.

The MPC controller was designed to cover the calciner, rotary kiln, and cooler controllers to meet these objectives. When dealing with large MPC applications such as cement kilns, splitting the solution into several subsystems is typical. This strategy allows more operational flexibility and robustness while generating benefits where only a part of the unit must be excluded.

The project was commissioned after the completion of the controller audit period. This audit demonstrated substantial improvement in energy recovery from the kiln cooler, specific energy consumption in the calciner, improved emission control and optimal chemical usage, and ultimately, a more stable process and high operator acceptance. A summary of Aalborg Portland’s evaluation report shows 12.3 GWh/y of energy savings. This increased stability has led to the opportunity to increase production. Still, it has also dramatically eased operation through all typical changes caused by feed chemistry variations, coal quality variations, and operational changes.

The project took approximately nine months to execute, resulting in sustained benefits to production, stability, and quality. Aalborg Portland has followed up its successful implementation of model predictive control on its Semidry MPC kiln with a new MPC project on two finish mill circuits due to the success of this project.

Enjoyed what you've read so far? Read the full article and the rest of the February Issue of World Cement by registering today for free!

Read the article online at:

You might also like


Embed article link: (copy the HTML code below):