How satellite monitoring is shedding new light on refinery operations

By Berend Heringa, Partner at McKinsey

Advances in satellite monitoring provide insight into operational interruptions and maintenance activities at major refineries to give traders a competitive edge.

In the increasingly complex and interconnected oil and gas value chain, any disruption anywhere in the world can have far-reaching implications.

Unplanned or planned shutdowns, driven by unanticipated incidents or simply maintenance, can yield unforeseen opportunity for some or negative consequences for others. Given the impact of such events on commodity flows and energy markets, knowledge is power.

For oil traders and others working in the trading arms of oil companies, hedge funds, etc., having up-to-date information on production and supply—meaning as real-time as possible—can mean the difference between making profitable decisions and loss-making ones. For refinery owners, timely and accurate data is crucial to achieve the best margins available by using knowledge of how the market may be impacted by any competitor actions. For NOCs, it is simply about energy security. Yet, the oil and gas sector has long struggled with patchy, time-lagged and sometimes incorrect information pieced together from different and insular sources, some more reliable than others.

That’s where satellite monitoring comes in. Thanks to advances in digitisation, automation and advanced analytics, coupled with exponential growth in the number of commercial satellites in Earth orbit, high-resolution optical images from space can be procured, processed and mined for insights in close to real time. As an indicator of refinery activity, ground-level temperatures can be measured via infra-red imaging. Similarly, the concentration of atmospheric pollutants can be assessed to gauge SOx and NOx emissions near refineries. The possibilities are many, especially when combined with other data sources. Oil tanker movements can be monitored using automatic identification system (AIS) vessel tracking. Oil volumes can be estimated from the rise and fall of floating roofs on storage tanks.

By analysing and synthesising activity indicators like these through a combination of algorithms, machine learning and expert market knowledge, it is possible to predict—sometimes as much as six weeks in advance—the likely timing and extent of a shutdown and assess its potential impact on crude capacity and commodity flows. This capability is already uncovering new information, as the monitoring of recent shutdowns at two major refineries attests.

Satellite monitoring at play

At ExxonMobil’s Singapore refinery, satellite monitoring confirmed that a pre-announced shutdown was under way. Signs of activity were first observed at the Jurong Island complex in mid-February 2019. By early March, it was clear that multiple units—crude trains, hydrocracker, aromatics, lube oils—were affected. While the market had been informed about the planned shutdown, the monitoring provided valuable intelligence on its timing, scope, and capacity outage.

At Shell’s Singapore refinery, by contrast, satellite monitoring revealed that an unannounced shutdown had taken place. Observation on 13 March revealed that materials and cranes were accumulating next to the crude units and that the chimney of the furnaces connected to the crude trains had stopped smoking. By early April, images showed that the clearance of the materials and cranes had begun, suggesting that the refinery would soon be back online.

Insight into operational interruptions and maintenance activities at major refineries gives traders a competitive edge, allowing them to modify their trading strategy and make better and faster decisions based on objective data.