Informing decisions in industry: how to evaluate and communicate resource efficiency opportunities

To understand the options available to improve resource efficiency in industrial plants, four key components are required: accessible material and energy flow data; a method to balance and reconcile these; an appropriate... [ view full abstract ]

To understand the options available to improve resource efficiency in industrial plants, four key components are required: accessible material and energy flow data; a method to balance and reconcile these; an appropriate definition of resource efficiency; and, means to visually interpret the results. The raw data collected by control systems has become an attractive source to identify these improvement opportunities. However, the volume and variety of data generated is vast, and continues to grow as the number of metered-facilities and the coverage of existing metering systems expands. For this reason, it is difficult to understand, evaluate and therefore to communicate the relevant information required to inform resource efficiency decisions. In this study we focus on the latter stages: we define resource efficiency and devise effective ways of communicating improvement options.

Data visualisations, in particular Sankey diagrams, have proliferated in industry as a powerful tool to visualise material and energy flow analyses. These diagrams commonly depict static representations of system structures and resource-use scale, and are often used to analyse yearly flows of individual resources at regional or sectoral levels.  However, in view of the time-scales, scope, and type of decisions involved in site-level resource management, the design of Sankey diagrams must be adapted to provide actionable insights.

Following a review of the literature and conversations with industrial practitioners, we identified three main aspects required by plant-level analysis that are currently missing from conventional Sankey diagrams. Firstly, a better understanding of resource use can be gained by conjointly visualising energy and materials with a single metric. Secondly, it is important to be able to show changes in the temporal dimensions: both variation of resource use over time and in time resolution (e.g. aggregating data into months, days or hours). Thirdly, improving the flexibility in the choice of content makes Sankey diagrams more effective at presenting the relevant information required to make specific decisions. For example, being able to select: from multiple metrics, such as CO2 emissions, energy, or exergy; the amount of information, such as visualising specific compounds (e.g. carbon or hydrogen) or; the spatial boundaries. If not designed appropriately, these visualisations can be overwhelming and therefore fail to generate transparency and additional insights on resource efficiency.

From this, we produced Sankey-type diagrams at different temporal scales and scope, using different amounts of data for three case studies: a simulated ammonia production site; control data from a steel-making site; and averaged data from 20 integrated steelworks worldwide. By collecting results from a series of semi-structured interviews with operators, plant managers, and environmental analysts, the discussion will focus on answering three questions: first, which time-scales are appropriate for the different scopes; second, how much information is useful to inform decisions on resource efficiency; third, in what format is this made most accessible to decision-makers?