A leading Central and Eastern European integrated refinery and petrochemicals operator sought a non-capital solution to boost margins and minimize product quality giveaway. During a 28-week engagement, Trindent implemented 12 targeted process enhancements that reduced gasoline quality giveaway by 38%, unlocking $9,200,000 in annual margin improvement.
How we Made it Happen™
Planning and Optimization
- Installed a process to adjust targets, and planning buffers using site-specific blend control variability and root cause analysis (RCA), improving client from a 4th quartile performer to a 2nd quartile performer.
- Established a structured LP weekly process for gasoline blending performance, enabling effective cross-functional follow-up and alignment between LP outputs and scheduling outcomes, which led to the identification of additional optimization opportunities >$500,000.
- Installed a volatility forecasting tool for downstream base terminals, and established a structured planning process to drive continuous improvement in phase-down planning while reducing planning buffer.
- Installed grade and seasonal specific octane and RVP blend laws, improving property prediction accuracy and decreasing recipe variance by 20%.
Execution Excellence
- Implemented a restoration plan for the Blending Optimization System, re-establishing system capabilities, rebuilding team expertise, and standardizing blend execution processes.
- Developed technical guidelines for the blending optimization system, standardizing operator responses and documentation procedures for operational alarms.
- Established a weekly cross-functional Blend Review Meeting with a root cause analysis framework to comprehensively identify and historize causes of quality giveaway and develop a proactive approach to resolve underlying issues.
- Developed giveaway training modules, and Peer-2-Peer observation programs across functional groups to enhance knowledge sharing, and strengthen accountability and ownership.
- Installed statistical quality control for online RVP and NIR analyzers to continuously monitor prediction variance and manage bias.
Measurement Accuracy
- Implemented a process leveraging rundown GC analysis of light blending components to develop a quality estimation tool, reducing blending recipe variability by 5–10%.
- Installed an NIR Model Validation Tool, defining model prediction behavior and establishing a baseline for continuous performance improvement.
- Established a weekly cross-functional NIR team meeting to review online NIR performance, plan model updates, and identify improvement actions, resulting in a >50% increase in prediction accuracy.
- Developed NIR model update guideline to strengthen internal expertise and enable continuous in-house model prediction improvement.
"Your team worked collaboratively and seamlessly integrated with our team as technical experts to implement industry best practices for quality giveaway reduction […]. Working with Trindent has been a truly collaborative experience, fostering cross-functional cooperation for us to implement and sustain meaningful change […]. All financial metrics are trending positively, and we are confident that the annualized savings from the project will continue to surpass the original targets […]”