Refinery optimization: a multi-faceted concept
The Oil industry is one of the pioneers in the widespread adoption of optimization techniques. The term “Refinery optimization” has consequently evolved to become a multi-faceted concept.
At an abstract level, refinery optimization is about getting the most from the existing assets; by extension, it can also refer to the most profitable investments in new equipment or new assets. This first distinction gives rise to 2 horizons of application to refinery optimization: long term, structural optimization, and short term operations optimization.
Moreover, operations management itself can be addressed in a variety of ways, depending on whether the approach taken to refinery optimization is bottom-up, or top-down:
- The top-down approach is based on hierarchical optimization: the refinery optimization problem is addressed as a succession of optimization problems of different time magnitudes and granularities. The first problem is a mid-term planning problem and is concerned with the refinery as a whole and a time frame of one to several months. Subsequently, the time frame is subdivided into finer time intervals. At the lowest level, the theoretical goal is to achieve operations scheduling on a day to day basis
- The bottom-up approach is one that starts with local problems and tries to aggregate them into a consistent solution at the refinery level
Our approach to refinery optimization
Moreover, operations management itself can be addressed in a variety of ways, depending on whether the approach taken to refinery optimization is bottom-up, or top-down.
In contrast to traditional bottom-up approaches – which attempt to build an overall solution by aggregating local solutions – we start by achieving a feasible, “good quality” overall solution using our smart simulation engine. Then, guided by the consistent overall solution, we seek to optimize local operational challenges.
Our innovative approach to refinery optimization takes account of 2 crucial facts:
- Day-to-day operations are subject to uncertainty,
- Regardless of the inherent formal difficulty of achieving an overall refinery schedule optimization, an overall optimized schedule is of little use.
This is simply because an overall solution achieved that way is inherently unstable: the slightest change or unforeseen event means that the whole schedule must be rebuilt. And given the huge amount of inertia in such a large system, any sudden change of trajectory consumes an equally huge amount of energy.
Under this pragmatic perspective, refinery optimization appears as the search for a maintainable globally feasible, locally optimized schedule. Optimization is applied to sub-systems where a high economic impact is achievable.
To be realistically achievable, effective and efficient, refinery optimization therefore should be considered as an orchestrated effort that needs to be put in practice under uncertain assumptions, and this approach should therefore provide the means to confidently and rapidly implement day-to-day adjustments to subsystems to achieve locally optimized schedules.
PRINCEPS solutions materialize refinery optimization through a right combination of smart simulation and local optimization modules.