There is a challenge for many refinery and chemical companies to improve the analysis process and raise performance, which can be achieved through real-time visualised data. Nick Milner reports
Making sense of vast volumes of production data is the essence of asset optimisation. If you cannot see what you do not know, then you will not be able to make the right decisions to optimise production and be competitive. The inside story of operational intelligence lies in rich thin visualisation and analytics capabilities that improve production execution, enabling process manufacturers to quickly identify and resolve operational issues.
The challenge for many refineries and chemicals companies in today's market is to improve the analysis process and incorporate product characteristics and other non-time series data along with the usual sensor and meter data being collected by the process historian. It is those data elements that provide additional context for an improved understanding of conditions that limit production operations.
Cutting-edge manufacturing execution systems (MES) integrate the operation, using real-time business performance management to optimally plan, execute, monitor and respond to change immediately across all time horizons. Information must be relevant, timely and collaborative. With discoverable data displayed in a clear, easy-to-understand dashboard, process manufacturers can efficiently resolve operational issues to remain profitable.
Identifying the problem
Production analysis often lacks the ability to visualise data in flexible formats or incorporate operational characteristic data quickly to identify and correct the root causes of operational problems. Analysis of production events, such as individual batches, phases or steps in batch and semi-batch processes or time-delimited events in continuous manufacturing processes (i.e. start-ups/shut-downs, production runs, operating mode switches, the comparison of production shifts), allows manufacturers to analyse multiple responses to operational events and adopt best practices to further optimise operations. To gain a thorough understanding of the asset and its performance requires powerful data visualisation technology, which can put plant information into context, so that operations personnel can apply timely corrective actions to address important operational issues, like bottlenecks, yield efficiency, variability in product quality and overall asset effectiveness.
Visualisation techniques are mainly appropriate for the three types of variability in batch or event processing: average over time, within-batch/event and batch-to-batch/event-to-event. Event data has been difficult and time-consuming to incorporate in data visualisation in the past. Significant time and effort was required to find, extract and include non-time series data like product characteristics previously. This effort has been significantly reduced in modern production visualisation solutions.
Modern manufacturing is like an ecosystem of interconnected software and hardware that helps chemical and petrochemical companies optimise plants and achieve operational excellence. As businesses generate vast amounts of data, efficient decision support solutions are needed to make sense of vital information and ensure operations can adapt quickly to dynamic conditions.
Understanding the data
Automated MES decision support allows manufacturers to make corrective decisions faster to achieve operational efficiencies and deliver greater productivity. Bridging the gap between the technical operation of the plant and the commercial transactions using the latest digital environment (ie, combination of web, mobile, tablet) provides users with vital information at the right time, allowing staff to dynamically keep up-to-date with operational challenges, anytime and anywhere.
aspenONE Process Explorer (a1PE) helps users to understand the story behind the data. This web-based tool converts production and business data into operational intelligence with the ability to access, visualise, analyse and monitor data in a clear, graphical display all in one single platform. The visual graphics help engineering users to understand the patterns in the data and associate them with conditions in the plant. This means that users can quickly identify issues and correct problems in production. The data discovery tools bring the most actionable insights to the surface.
a1PE utilises HTML-5 graphics that contain fast user interfaces and can connect to Aspen InfoPlus.21 and other process data historians on the market. As a web-based product, a1PE does not need to be installed on thousands of individual user desktops across an organisation. The tool is intuitive, easy-to-use and can be customised to allow users to only see what they need. Data discovery is built into a1PE, which addresses the common challenge of trying to quickly find the most relevant production information. Users can search multiple sources of data, identify and analyse the key issues and relate it to the actual physical asset and operations of the plant.
a1PE is able to present several years' worth of data in an instant, along with annotations and other unstructured information such as alarms, events and other items that provide context, enabling the user to swiftly understand what is happening in the plant. Events are flexibly-defined and users can easily plot periods of time where operators have placed comment markers or where alarms exist. The rich discovery tools work seamlessly, whether the process is batch, semi-batch or continuous. Event analysis can overlay multiple events and analytic plots, such as X-Y scatter plots and frequency histograms, can be applied within a batch, batch-to-batch or over time. Many of these capabilities can also be used to evaluate continuous processes (e.g. comparing plant performance across different crude types or operating shifts).
Performance analytics include Overall Equipment Effectiveness (OEE), which allows users to compare the performance of one piece of equipment, or plant area, with another and then take the learnings from the better performing equipment and apply them to underperforming areas to help increase the overall plant performance. Statistical Process Control (SPC) is another analytics capability that is available in the a1PE tool to improve production performance.
These enhanced capabilities improve analysis by capturing context. Users can easily visualise all pertinent information regardless of the type. Instead of simply looking for anomalous trends, contextual data can be viewed alongside process data to show what is happening in production, thereby delivering greater insights into the source of problems.
Context is crucial
Advanced MES software dramatically improves data analytics, based on a dashboard that supports faster and better decisions that will be used to elevate production to the highest standards. With aspenONE Process Explorer, analysis is easier, faster and more effective, allowing operators, process engineers, unit managers and executives to build a complete picture of production performance from continuous to batch. The rich context and analytics tools help drive greater collaboration across the enterprise by enabling benchmarking and sharing of best practices for faster decision-making.
Establishing a centre of operational excellence improves communications and brings operational staff closer together. Those chemical and petrochemical companies that adopt best practices and use the power of data visualisation software tools can better understand the story behind the data and ensure that production outcomes meet with a profitable ending.
Nick Milner, Director, Professional Services, AspenTech.