The Plan

After Peter Argus' transformation team has several discussions with Ron Erickson, they make a list of the steps they will take to analyze and reduce the refinery's energy demand, or consumption:

  • 1. Real-time energy monitoring. The team will monitor energy usage for one month during seasonable weather and during typical production rates. Once refinery energy usage is baselined through EIDI inline calculations and historical analysis, EIDI run-time analytics will monitor and check that consumption does not deviate from or exceed historical norms for a unit, process, product, and so on. If it does, it will generate real-time alerts via the EIDI to forewarn operations personnel that the process is using excess energy, and personnel should take corrective actions.
  • 2. Energy event management. Detect and analyze process changes that cause consumption to exceed forecasts. This is done via online forensic analysis and by extracting process data for input into advanced analytics, such as Microsoft's Power BI.
  • 3. Peak demand management. Minimize usage in peak demand periods, to avoid triggering higher rates and potentially harsh penalty fees. This is done by classifying the real-time data into event frames, calculating projected usage, and using EIDI real-time notifications to generate warning alerts well before the peak thresholds are reached.
  • 4. Idle state management. Reduce energy usage to absolute mini- mums when running assets in an idle state.
  • 5. Demand/response management. If feasible, offer excess energy capacity back to the power grid when requested in exchange for revenue or incentives.


Integrated Proclndustries refinery diagram with asset data model for continuous improvement.

In addition, management asked Peter to work with refinery teams to implement a plan to better ensure energy resiliency during storms or lightning strikes. Ron reviews their approach and further recommends that they capture energy data (electricity, fuel, and steam flows) for each unit, classifying the information in order to subdivide consumption by unit for subsequent analysis. He also wants to capture unit conditions and losses when the plant is not on target. Ron proposes integrating production information, energy usage, emissions data, and process availability in order to monitor the cost of electricity under specific conditions.

A detailed process flow diagram with the key energy centers of the industrial complex assists in identifying the energy areas to consider. Using a process block diagram adds the necessary context to the process, asset, and energy data.

Figure 7.2 shows a schematic of the South Texas refinery block flow diagram to gather all information from all process areas. All relevant inputs that can be measured should be included for treatment by the data infrastructure system.

< Prev   CONTENTS   Source   Next >