AI for effective control of scale deposition on liquid ring vacuum pumps

In this case study, we present and discuss HABER's approach and role of AI in addressing one of India's leading integrated paper mill’s concerns in controlling scale deposition in the vacuum pump.

Introduction 

In paper manufacturing, water removal is a highly energy-intensive process. Once dewatering through gravity is done, the vacuum created using a liquid ring vacuum pump is used in the next stage. An efficient vacuum process would mean a lower cost of water removal as the cost of removing water is higher in the dryer section.

One of the frequently observed operational issues is the deposition of mineral salts, mostly calcium carbonate, on processing equipment because of the supersaturation of mineral ions. Major contributors for scaling in vacuum pumps are the fluctuations in the process parameters such as pH, temperature, and pressure. These salts have a negative solubility temperature coefficient, which means the solubility reduces with an increase in temperature. Scaling reduces the efficiency of dewatering, in turn, increases the load on the downstream water removal process. Apart from downtime, this also results in higher power consumption and shortens the lifespan of the pump.

The scale deposits can be removed chemically or mechanically. Chemical scale removal is usually a cheaper and easier way, but the effectiveness of the removal depends on the surface-to-volume ratio of the scale. These chemical inhibitors help hinder scale growth. There are fundamental differences between different scales that can lead to more scaling if the wrong choices are made in selection of the chemicals.

Avoiding mineral salts deposition completely is not possible. But the costs and impacts can be minimized by thorough risk analysis and a focused approach on scale prevention.

HABER's Approach 

HABER’s ELIXA® platform provides real-time artificial intelligence and machine learning-based closed-loop control for vacuum pump scale control programs. The process goes as follows:

Real-time data of different process parameters affecting the vacuum pump performance is collected through the IoT sensors, and the big data analysis happens in the ELIXA® cloud. Depending on the process conditions, ELIXA® software predicts the requirement of anti-scalant in the system in real-time. This predicted output is pushed as a set point to the final control element.

AI/ML-based automation eliminates bottlenecks involved in the manual processes such as sampling frequently on daily basis, and hence delay in taking the necessary actions. Also, data points were not sufficient to generate insights or actionable intelligence. Eliminating these problems results in the precise, efficient, and reliable control of scale deposition.

The impact created after the implementation

•  Anti-scalant chemical consumption reduced by 20%, resulting in significant annual savings
•  50% reduction observed in vacuum pump cleaning
• The vacuum required in the suction box(s) is maintained consistently, without many variations for a long duration

In short, ELIXA® helps to achieve consistent better-quality products with a minimum amount of investment and an easy maintenance cycle.