Maintenance teams access this data via intuitive dash- boards that visualise vibration trends and highlight anom- alies. This setup allows engineers to detect early signs of mechanical issues, such as bearing wear, misalignment, or shaft deflection. By leveraging the capabilities of the vibration sensor and the surrounding digital infrastructure, the system significantly improves reaction times, reduces manual workload, and supports data-driven maintenance decisions. Moreover, the wireless nature of the vibration sensor sim- plifies installation and scalability, making it ideal for retro - fitting existing assets without requiring extensive cabling or infrastructure changes. Its battery-powered design and robust enclosure ensure long-term operation in demanding environments, while its integration into refinery-level mon - itoring systems enables seamless data flow from field to control layers. As seen in the following discussion, this project demon- strates how modern wireless sensing technologies can be effectively applied to critical rotating equipment, paving the way for broader adoption of predictive maintenance frame- works across industrial facilities. By leveraging the capabilities of the vibration sensor and the surrounding digital infrastructure, the system significantly improves reaction times, reduces manual workload, and supports data- driven maintenance decisions System architecture and data flow System architecture is built upon a robust, multi-layered framework designed to ensure secure, scalable, and effi - cient data flow from field-level sensors to enterprise-level monitoring and alerting platforms. This architecture not only supports real-time condition monitoring but also ena- bles intelligent decision-making through automated alert- ing mechanisms and historical data analysis. At the heart of the system are wireless vibration sensors, which are strategically mounted on the tank mixer motor housings. These sensors continuously measure three-axis vibration velocity (in mm/s) and temperature, capturing the dynamic mechanical behaviour of the equipment. The data transmission is handled via the WirelessHART protocol, a secure and interference-resistant communication standard tailored for industrial environments. This protocol ensures that data packets are reliably delivered even in the presence of electromagnetic noise, metallic obstructions, and com- plex refinery layouts. Once collected, the sensor data is transmitted to a WirelessHART gateway, which serves as the first aggrega - tion point. Here, each sensor’s output is mapped to a unique tag name, allowing for precise identification and traceabil - ity throughout the system. The gateway then forwards the
data using the OPC protocol to an industrial OPC server located in a Level 3.5 DMZ. This DMZ server acts as a secure intermediary between the operational technology (OT) and information technology (IT) layers, enforcing strict access controls and ensuring that only validated data proceeds further into the system. A firewall is positioned between Levels 3.5 and 4, allowing only unidirectional traffic. Within the DMZ OPC server, incoming data undergoes source tag validation, where only data marked with a ‘good’ status is approved for transmission. This step is crucial for maintaining data integrity and preventing false alarms or corrupted readings from influencing maintenance deci - sions. Once validated, the data is sent to the PHD, which serves as the central repository for long-term storage, trend analysis, and real-time monitoring. In the PHD, destination tags are created and linked to their corresponding source tags through a process known as tag routing. This enables real-time data mirroring into the distributed control system (DCS), allowing operators to observe equipment conditions directly from their con - trol interfaces. This mirroring is achieved without direct communication between the field sensors and the DCS, preserving network segmentation and minimising cyberse- curity risks. A key innovation in this architecture is the automated email alerting mechanism, which is tightly integrated into the PHD platform. The system continuously evaluates the vibration velocity values (mm/s) against three critical- ity thresholds that have been meticulously defined by the maintenance team based on field experience and equip - ment specifications: • Normal operating range: This range indicates that the equipment is functioning within acceptable vibration lim - its. No action is required, and the system continues passive monitoring. • Warning range: When vibration values enter this range, it suggests potential early-stage mechanical issues. The system automatically triggers an email alert advising main- tenance personnel to perform a field inspection. This level acts as a proactive checkpoint to catch anomalies before they escalate. • Critical alarm range: This is the highest severity level, indicating that vibration levels have reached a point asso- ciated with imminent failure or safety risk. Upon detection, the system immediately sends out high-priority email alerts to designated personnel, prompting urgent intervention such as shutdowns or emergency services. Transforming raw vibration data into actionable insights These bespoke thresholds are not arbitrary; they are the result of collaborative input from the maintenance team, who analysed historical vibration data and correlated it with actual failure modes observed in the field. By embedding this domain expertise into alerting logic, the system ensures that notifications are both meaningful and actionable. The email alerts themselves are structured to include detailed information, such as the affected equipment tag, timestamp, measured vibration value, and the severity level.
64
PTQ Q1 2026
www.digitalrefining.com
Powered by FlippingBook