New orders for Powitec

New orders for Powitec
Published: 03 August 2007

New and important orders for Powitec, producer of NMPC optimization systems for the cement & lime industry, power plants and waste to energy plants. NMPC stands for Non-linear Model Predictive Control and is used to optimize production processes with neural nets (artificial intelligence).


Holcim Apasco Mexico decided to buy 4 PiT Indicator Systems and to install 2 PiT Predictor. The PiT Indicator is an intelligent thermography system. With digital image processing, patented “electronic dust filter” and dynamic RGB-Thermography it shows a much more sharp and detailed picture than the usual video or conventional pyrometric camera. Characteristic and dynamic behavior of the flame is visualized by Polylines, delivering precise information about flame, flame shape, clinker bed and coating. 


The PiT Predictor is a soft sensor, using a self-learning adaptive software, based on artificial neural nets, for online prediction of essential process parameters like free lime content. Holcim uses this optimization system, consisting of optical sensor and software based on highly sophisticated neural net technology, to analyze and control the clinker production process.  


At Grasim’s Gujarat Cement Works, India shortly a PiT Navigator will be commissioned. The PiT Navigator consists of PiT Indicator, Soft Sensor free lime and Closed-Loop process-control. The PiT Navigator uses self-learning adaptive neural nets with NMPC (Non-linear Model Predictive Control) to optimize calciner, kiln, burner and clinker cooler.


PiT Navigator allows an active process-control and fast adaptation e. g. on changing fuel characteristics. Grasim expects the following advantages:  Increased availability and production capacity up to 3%.


Constant and increased clinker quality – elimination of off-spec clinker Long-term reduction of maintenance and repairs costs (refractory material)