Calcined Petroleum Coke Quality Optimization
Challenges
- The delayed coking unit converts residual oils into lighter products and green coke; the calcining unit then calcines that coke into calcined petroleum coke.
- Calcined coke must meet vibrated bulk density and other specifications; off-spec production leads to heavy penalties, rejected shipments, and direct loss of saleable product.
- The feed became lighter, requiring slurry addition and tighter blend and severity control to preserve coke quality — and because coking output drives calcining quality, the decisions have to be optimized across both units at once.
Approach
A coupled digital twin links the two units: models predict the coking unit's product yield split and green coke properties from feed and blend properties and operating setpoints, and further models predict calcined coke qualities from those green coke parameters and the calcining setpoints.
A joint optimization engine then searches the controllables of both units at the same time to maximize the economic objective while holding quality constraints and operational feasibility bounds.
Key features
- DCU simulation — yield distribution and green coke parameters predicted from feed and setpoints
- CCU simulation — calcined coke qualities such as VBD and real density predicted
- Coupled optimization across both units simultaneously
- Quality constraints and feasibility bounds enforced inside the optimization
- Auxiliary-product margin improved alongside coke quality
Impact
3%
Value increase on DCU products
up to 95%
Reduction of off-spec product days
Decisions optimized across the full DCU → CCU chain
As stated in the company deck