Decision agility


While unplanned downtime will always impact productivity and profitability in an oil refinery or petrochemical plant, the effects go beyond the financial. Safety is critical, but so is environmental efficiency. A single, unplanned shutdown can release a years’ worth of emissions into the atmosphere – not to mention the losses in profitability, reduced productivity, higher maintenance costs and increased waste.

Gas flaring, or the combustion of excess product, is one such consequence of an unplanned shutdown. Excessive flaring is a visual sign that something is wrong, meaning an increased risk to safety.

According to the World Bank’s Global Gas Flaring Reduction (GGFR) programme, each year, 145 billion cubic metres (BCM) of gas is released into the atmosphere from gas flaring. That’s equivalent to 270 million tonnes of CO2 emissions per year.

These figures paint a grim picture, but with the power of machine learning and predictive analytics, companies can begin to reduce unplanned upsets and capture all the associated benefits: minimising dangerous conditions, reducing emissions and maximising uptime.

So, what if it was possible to know which equipment would fail, and when – so repairs could be performed as part of a managed shutdown?

Today’s asset performance management technology can deliver advanced warning of failures through a combination of predictive and prescriptive analytics, enabled by integrated software that incorporates artificial intelligence (AI) and machine learning – providing a detailed view of all equipment, systems, facilities and networks, thereby enabling a capability we call ‘decision agility’.

With time to simulate event outcomes and plan holistically around predicted downtime, plant personnel can see exactly how each decision impacts planning, scheduling, inventory and sales.

When the outcome is known in advance, operators and engineers can collaborate on a plan to make the safest, most profitable decisions across all regions, scaling the technology to better understand the co-dependencies of separate facilities and maximising return on investment. So, when there is an issue in one location, the software can show how it will affect the pipeline coming in, the ships going out and whether the facility is at risk of defaulting on any contracts.

As such, digital transformation is knocking down the data silos and delivering the tools necessary to make sense of the data available at the enterprise scale.

Operational insights
High performance computing, artificial intelligence and advanced analytics can now generate deeper insights from operating data, with leading-edge simulation programmes using these insights to enable operators to quantify the value or cost of any renovation or improvement project, maintenance change, operations improvement or supply chain constraint. Utilising statistical sampling techniques, it’s possible to predict a system’s future performance, analysing equipment behaviour for a ‘time to failure’ estimate.

As a result, plant personnel can be alerted to impending failures and understand the potential wider impacts to discover exactly what’s costing money, negatively impacting performance or affecting environmental efficiency.

At a refinery, for example, the software might indicate the impending failure of a fluid catalytic cracker or part of a cooling tower, which would cause wider business disruption. But with advance notice and time to plan, scheduling models can determine the best time to take that equipment offline, and even insert additional maintenance activities to maximise the planned downtime – or load the information into a longer-term planning model that can account for sales impacts and operations planning, maintaining customer commitments and minimising emissions.

Through those two models, personnel are making informed decisions across a multi-network supply chain – and the greater the window of predictability, the more powerful the business options.

The right technologies for the right results
The right technologies also help maintain safe operations and meet environmental goals.

Unplanned downtime and transient conditions lead to flaring, which releases gases into the atmosphere – and this is where predictive analytics can have a major impact by alerting to pending problems. And with planning model integration, this technology can provide recommendations on the most eco-conscious actions to take.

Increased uptime for optimised operations
Beyond sustainability, companies also stand to gain from increased uptime. Optimised maintenance processes reduce unplanned shutdowns and help companies realise incredible payback on their investment in predictive analytics.

For example, one refinery suffering from repeated hydrogen compressor failures was able to reduce shut-down time by eight days due to a 35-day time-to-failure prediction. In addition, the cost for planned maintenance was less than 30 per cent of the cost of emergency repairs.

One source estimates that equipment failures cost oil and gas companies an average of $42 million a year. The US Department of Energy reported 1700 shutdowns at refineries between 2006 and 2017; 46 per cent were due to mechanical breakdown.

Just eliminating some productivity-reducing events can add millions of dollars to the bottom line. And when companies are able to quantify exactly how much any particular event affects revenue, they know where to target their technology strategy for maximum impact.

Leading from the front
As companies face growing pressures from shareholders, regulators and consumers alike, the need for agility is greater than ever but so too is the need for sustainability. By reducing risk and uncertainty through the implementation of advanced technology solutions, companies can put themselves in the best position to win in the marketplace of tomorrow.

John Hague is Executive Vice President, Operations at Aspen Technology (AspenTech), a leading software supplier for optimising asset performance. Its products thrive in complex, industrial environments where it is critical to optimise the asset design, operation and maintenance lifecycle. The company’s purpose-built software platform automates knowledge work and builds sustainable competitive advantage by delivering high returns over the entire asset lifecycle. As a result, companies in capital-intensive industries can maximise uptime and push the limits of performance, running their assets safer, greener, longer, faster.
For further information please visit: