Human Reliability Studies

Methodology

Rob King, Human Factors Director Coretex Consulting attended a processing facility to conduct a SCTA for the plant start up process using Systematic Human Error Reduction Predictive Approach (SHERPA) methodology. This included:

  • Walking the plant to identify equipment and design factors associated with relevant steps within the startup process.
  • Working with operators to identify the critical steps.
  • Working through each critical step of the startup process with HSE and Operational personnel to identify:
    • Critical process subtasks;
    • Potential human failure;
    • Failure consequences;
    • Performance Shaping Factors (PSF); and
    • Proposed additional control measures.
  • Capture findings and recommendations within a report for Hexion.

Findings and Recommendations

This section summarises the general findings for SCTA, specific recommendations from the analysis are include for each relevant operational step in the SHERPA analysis in Appendix A.

Procedure Design

There were multiple steps where requirements can be described with greater clarity. For example:

  • Valves position required to be 0% can be clarified as 0% (closed).
  • Arbitrary instructions such as ‘carefully in small stages’ can be described with greater clarity and also note the indication the operator can expect should the rate of change be incorrect.
  • Some steps were deemed redundant and have been removed.
  • Some steps have been reordered to align with work as done.
  • Including personal gas detectors during inspections.

Control HMI Display and Indications

There were opportunities identified to increase situation awareness within the control room by including additional indications of plant conditions. For example:

  • RO tank levels to be displayed in control room.
  • Opportunities for automating processes were identified.
  • Data logs relevant to temperature control being readily available through the HMI.

Plant Engineering Design / Controls

Additional engineering controls and redundancies in plant design were identified. For example:

  • Install sumps with capacity to fill lines to ensure if associated pumps are not started the system will still function as intended.
  • Install additional interlocks and/or control room alerts if certain plant conditions are not met prior to future steps progressing.
  • Additional flame detection systems to support effective start-up.
  • Additional temperature indicators to be installed and improve location for viewing.
  • Alarms to be set ‘skinny’ to provide earlier alert for potential over pressure event.
  • Potential to include continuous monitoring capability in place of testing.

Leave a Reply

Your email address will not be published.