Decision-Making through Plant Data Hierarchy
Peter knew from his discussions with the advising companies that it was critical to begin EIDI deployment by mapping sensor data sent from the refinery's plant control systems (distributed control systems [DCSs], programmable logic controllers [PLCs], SCADA, and IoT devices) to a physical or process entity that everyone understood. The EIDI database needed to be represented as a hierarchical data model that represented the refinery, its units, and its equipment. There are three advantages to doing it this way:
- 1. Users know what a data stream represents, for example, crude unit feed-water temperature or pump number 2 intake pressure, as opposed to some instrument nomenclature that are only known by a few people.
- 2. Humans and software systems are able to consume information represented as an asset, comprised of numerous data points, versus having to search for the individual data points that comprise the asset.
- 3. These assets are defined in the EIDI through configurable reusable asset templates. This means that for the many heat exchangers, pumps, or distillation column tray temperatures, one single template per asset type is defined and easily repopulated. This eliminates the need to configure each asset anew and reduces deployment time. These templates consist of:
- • Representing the data points that comprise a particular asset;
- • Defining an inline calculation to provide a derived value for that asset (e.g., degrees Celsius to degrees Fahrenheit, or heat exchanger fouling value);
- • Triggering a real-time alert if a specific metric or value falls outside of acceptable operating ranges; and
- • Including other non-time-series-related data, such as the last time the asset was serviced or the maximum allowable rate at which the asset can be run.
Figure 3.4 displays the components of an asset data template.
Asset framework object: template components.
EIDI Deployment and Configuring the EIDI Templates
The team began undertaking EIDI deployment. They successfully achieved the first milestone, which was to ingest and consolidate refinery data from the control systems into the EIDI, creating a duplicate of essential refinery information in a safe analysis environment. This information is now accessible by knowledge workers. Refinery personnel no longer had to search for data on the control system network or try to find it in some hidden silo that took valuable time to collect. The team had built several master versions of dashboards and displays, with the understanding that these standard displays would enable people to start analyzing data quickly. There were a few old-timers who initially resisted, but soon thereafter, every employee realized the refinery had to change, and most everyone got on board and used the EIDI.