object detection in computer vision

Industry Served

Power Manufacturing

object detection in computer vision

Objectives

Streamlining Power Plant Operations for Efficient Energy Costs.

object detection in computer vision

Solution

Ripik Vision

What Calls for a Change

This corporation, which is well-known in the mining and metals sector of India, has a long history of strategic resource management and operational efficiency. It has constantly shown that it is committed to environmental responsibility by emphasizing sustainable methods.

The facility manager was looking to optimize energy consumption in their 120MW power plant operations. They faced challenges with manual processes in the Coal Handling section, leading to inconsistent size and moisture levels. This resulted in unburnt carbon issues and increased heat losses. A real-time tracking solution was needed to provide prompt team alerts for quick remedial measures, enabling efficient operations of CFBC Boilers.

Problem

The efficient operation of CFBC Boilers hinges on crucial coal properties, encompassing size, grade mix, and moisture levels. Despite their significance, the power plant facility manager often handholds with maintaining consistency in these properties due to manual processes dominating the Coal Handling section of the plant. This manual handling not only results in a higher than promised LOI (Loss on Ignition) but also introduces challenges in detecting issues within the Coal Handling Plant (CHP) screens, like breakages or clogging. These problems go unnoticed until variations manifest in the boiler bed, causing size irregularities. Furthermore, undetected variations in grade mix and moisture can adversely impact the fluidized bed, contributing to heightened heat losses. Effectively addressing these challenges is imperative for optimizing boiler performance and ensuring overall plant efficiency.

img

The Solution

We planted our patented Cognitive and Vision AI based technology – Ripik Vision for 90% + accuracy in burden mix optimization for reducing energy consumption. The technology is equipped to provide real-time predictions on coal size distribution, moisture, and colour mix, offering timely alerts for any significant variations. This involves optimization of burden mix components such as Lump, Chips, FRBL, and Briquette percentages and burden chemistry factors like basicity, MgO-by-Al2O3 ratio, and total Cr, among others. The technology extends its optimization to the mix of metallurgical coke and anthracite coal, resistance set point optimization, and the identification of maximum resistance for different burden mixes, all contributing to enhanced efficiency. With the Resistance Set Point Recommender in place, the suite ensures an overarching strategy for operational excellence and resource optimization.

The Impact

The SaaS deployment has led to notable improvements in production and reductions in specific power and reductant consumption, with an annualized benefit of INR 230 (in lacs) observed over the last 6 months. Automated burden mix preparation optimizes efficiency, resulting in a 1% reduction in SPC and specific reductant consumption, translating to an annualized impact on the power industry. Additionally, our comprehensive approach with the development of a Burden Charging Recommender and Resistance Set Point Recommender, aimed at optimizing power and reductant consumption. The solution’s impact is evident in a 1.1% reduction in unburnt carbon in boilers, attributed to proactive actions by the CHP team based on software alerts, highlighting its effectiveness.

img

“The team has good knowledge in the areas of data science & machine learning, and their problem solving skill set is high”

Senior Vice President

Explore how CFBC boiler optimization can reduce unburnt carbon losses for enhancing operational efficiency.