← Back to Programme
Conditioning 1 Room 0D oral

Real Time Monitoring of Water-Glycol Hydraulic Systems: Early Failure Detection Using Optical Particle Counter and Data Analytics Platform in the Caster Main Hydraulic System

Date Wednesday, 03 June 2026
Time 10:30 – 10:50
Topic maintenance
Authors
Rafael Diaz Presenter
Rafael Antonio Diaz Giron
Israel Sanchez
Marco Suárez (Presenting Author)
Affiliations
RMSH
Unión y fraternidad No.8, Int. 3. Col. La amistad
Tula de Allende, Mexico
Abstract

Water-glycol hydraulic systems are widely used in the steel industry due to their fire-resistant properties, yet they present unique challenges in fluid degradation, component wear, and contamination control. In continuous casting operations—such as the Caster Main Hydraulic System—the reliability of critical components depends on continuous and accurate monitoring of both fluid condition and system performance.

This study presents the implementation of a comprehensive digitalization and real-time monitoring strategy combining the optical wear-particle sensor Atten2 OilWear S120 LCD with a custom Python-based data analytics platform. The system integrates multiple data streams, including:

  • Pump discharge pressure and pressure stability at each hydraulic circuit
  • Temperature of the water-glycol fluid
  • Hydraulic reservoir level
  • Pump casing-drain flow as an indicator of internal wear
  • Differential pressure across return and pressure-line filters to identify saturation events
  • Real-time ISO 4406 particle codes and wear-particle trends from the S120 sensor

Using this integrated monitoring platform, an abnormal and progressive increase in 4, 6, and 14-micron particle counts was detected, correlating with changes in casing-drain flow and filter ΔP. The Atten2 S120 identified a distinctive rise in wear-particle generation, allowing early detection of a developing internal component failure before it manifested through pressure, temperature, or operational symptoms.

The study demonstrates how combining optical wear monitoring with a unified, Python-driven analytics environment enables faster diagnostics, earlier anomaly detection, and data-driven decision-making. This approach reduced unplanned downtime, increased system reliability, and provided a scalable model for digital transformation in hydraulic systems using water-glycol in steel plants.