Harnessing data to maximise refinery efficiency and reduce downtime

Unplanned downtime and operational inefficiencies significantly impact refinery profitability. Refinery managers must continually balance the need to maximise uptime with the requirement to carry out essential maintenance.

In practice, this is a complex task. Too often, unplanned downtime occurs because there are limited economical options to address root causes before they escalate. The result is reduced profitability and higher CO₂ emissions, as inefficiencies in the refining process go unresolved.

Identifying and implementing effective solutions to reduce inefficiency and unnecessary downtime remains a priority for the oil and gas sector.

Unplanned refinery shutdowns are one of the most visible causes of operational inefficiency. While short shutdowns are sometimes unavoidable, there is evidence that outages are lasting longer than necessary, even with today’s available technology.

These events are rarely planned for in budgets and often indicate a serious underlying issue. Between January and September 2023, there were 124 unplanned maintenance shutdowns at refineries in the United States, a 53% increase compared with the same period in 2022.

A growing challenge is the extension of planned maintenance shutdowns. In many cases, additional issues are only identified once a refinery is offline. This can lead to delays while replacement components are sourced, delivered, and installed, extending downtime beyond original plans.

Even large-scale planned maintenance events can be affected. Recent years have seen prolonged shutdowns at some of the world’s largest refining facilities, including activity across multiple units within a major refining complex in India.

The focus on uptime alone can create additional risks. During the pandemic period, concerns were raised that deferring planned maintenance to maintain output contributed to a rise in unplanned downtime.

The key challenge is not simply maximising uptime, but achieving it while maintaining effective and timely maintenance practices.

A central limitation is visibility. In many cases, refinery managers and engineers lack real-time insight into the condition of critical assets. Improving monitoring capabilities is therefore essential to reducing both unplanned and extended shutdowns.

Proactive monitoring of critical components is an important step towards reducing inefficiencies. In practice, however, this is challenging due to the scale and complexity of refinery operations.

A typical refinery can contain hundreds of thousands of individual components. Manual inspection programmes often rely on contractors working continuously throughout the year to assess tens of thousands of assets.

Some components are also difficult or unsafe to access, meaning they may be excluded from routine inspection processes.

Even when comprehensive, manual inspections only provide a point-in-time view. A component may fail shortly after inspection, with the issue remaining undetected until the next scheduled visit, often up to 12 months later.

Real-time asset monitoring provides continuous visibility into equipment performance, enabling issues to be identified as they occur. This approach supports predictive maintenance strategies that reduce both downtime and inefficiency.

Real-time monitoring typically uses wireless sensors to track asset condition. Earlier generations of these systems often relied on battery-powered devices using wireless protocols, which could reduce battery life to under 3 years. This can increase operating costs and limit feasibility in hard-to-access locations where battery replacement is difficult.

The latest generation of systems uses Low Power Wide Area Network (LPWAN) technology, including LoRaWAN (Long Range Wide Area Network). This can extend battery life to up to 14 years, reducing maintenance burden and improving long-term viability.

The value of real-time monitoring increases when combined with appropriate bypass and redundancy strategies.

These approaches allow components to be replaced or isolated without waiting for planned shutdowns or risking unplanned outages. Effective implementation depends on understanding the criticality of assets and applying suitable mitigation measures where required.

Steam traps illustrate how component-level failures can significantly affect refinery performance. A typical refinery may contain more than 10,000 steam traps, with annual failure rates ranging from 10% to 25%.

Failure commonly leads to steam leakage, reducing efficiency and increasing CO₂ emissions. In some cases, failure can be catastrophic, creating serious safety risks, including exposure to high-temperature steam.

Despite their importance, steam traps are often only inspected annually. In some cases, especially where access is difficult, they may not be monitored at all.

The scale of the issue is significant. In one refinery example, annual steam loss equated to approximately 6,000 trees' worth of carbon offset. While an individual steam trap may cost around $380, a leaking unit can result in losses of around $2,300.

Blockages also pose safety hazards, as water can travel through pipework at high speed, creating dangerous conditions for personnel.

Monitoring steam traps in real time provides a clearer view of overall boiler performance. Without accurate data, refinery operators may believe boiler efficiency is around 82%, when in reality it may be closer to 70% due to undetected steam losses.

A 12-percentage point reduction in efficiency can represent substantial operational and financial impact.

Real-time data from steam trap monitoring enables operators to understand the relationship between asset condition and boiler performance, supporting more informed maintenance planning.

IMI’s NEON Sonic steam trap monitoring solution provides real-time alerts when a steam trap fails, delivering immediate visibility of associated inefficiencies and financial impact.

The system uses LoRaWAN (Long Range Wide Area Network) technology for long-range, low-power connectivity, enabling straightforward deployment and reduced operating costs.

Unlike many competing solutions, NEON Sonic uses edge diagnostics. Steam trap condition is assessed through intelligent anomaly detection at the sensor level, removing the need for additional software to interpret results. This simplifies integration with existing plant systems, including Distributed Control Systems (DCS) and Supervisory Control and Data Acquisition (SCADA) platforms.

IMI also provides an end-to-end solution through SolidRed cloud-based dashboards, enabling maintenance teams to identify issues and initiate corrective action quickly.

Real-time asset monitoring provides a practical route to improving refinery efficiency while maintaining essential maintenance standards. By increasing visibility of asset performance, operators can reduce unplanned shutdowns, minimise extended outages, and improve overall operational reliability.

The latest generation of self-powered monitoring technologies represents an important step forward in supporting more resilient and efficient refinery operations.

Further Reading

Related IMI Content

IMI TWTG NEON Sonic for Steam Trap Monitoring (Ex)

IMI TWTG NEON Sonic Steam Trap Monitoring (Non-Ex)

SolidRed Asset Monitoring Software

A version of this article was originally published on Engineerlive.com in 2024.