An optimization model for refinery production scheduling (designing a software using genetic algorithms & MILP)
Due to increased complexity, quality and environmental requirements in process manufacturing operations, many leading companies have identified Process Systems Engineering (PSE) as a strategic technology. A new method for optimizing process networks is presented in this work. Optimization models deal with planning and scheduling of several subsystems of the petroleum supply chain such as oilfield infrastructure, crude oil supply, refinery operations and product transportation. The focus of the present work is to design general software for modeling and optimization of refinery operations and product blending. This decomposition approach is derived from analysis of the mathematical structure of a general overall plant model, which contains common elements and independent elements. This understanding forms the basis for decomposing the overall plant model into two levels, namely a site level (master model), and a process level (sub models). The master model determines common issues among processes, such as allocation of raw materials and utilities, etc. With these common issues determined, sub models then optimize individual processes.
In this work units assume as boxes and we can use any method for modeling. For modeling of blending operation use MILP model with property index method. for overall optimization we use genetic algorithms and MILP methods together. Therefore we don’t limit in units modeling and we can use every white, black or gray box model.
This work outlines a visually interactive graphical modeling approach for refining process type production systems, with hidden generation of complex optimization models for production planning. The proposed system lets the users build a graphical model of the production system with one-to-one clones of its production units through its interactive visual interface, accepts production-specific data for its components, and finally, internally generates and solves its mathematical programming model without any interaction from the user.
The production planning includes the aggregated production planning deciding where and when production should occur, the shipment planning where customer demand is transformed to schedules for the tankers transporting the products, the scheduling of the processing units, and finally the realization of the plans and the schedules with respect to the utilization of tanks and pipes. In this work we focus on the process scheduling problem, on the question of how to utilize the processing units in an optimal given a specified demand. The scheduling is strongly related to the planning at the other levels, and it affects many types of decisions in the company. The ability to efficiently construct high-quality (low-cost) schedules is therefore crucial for the refinery in order to be competitive. The scheduling problem concerns the question of which mode of operation to use in each processing unit at each point of time, in order to satisfy the demand for a given set of products. A main characteristic of refinery scheduling is that a set of processing units concurrently produce multiple products, and a product obtained as output from one processing unit can be used as input to another processing unit. A mode of operation for a processing unit is specified by the combination of products consumed and produced in the process, and by the yield levels for each of the products.
Processing unit model:
Processing unit is defined as a piece of equipment that is able to physically or chemically modify the material fed into it. According to this definition, processing units are all those that compose the refinery topology. In this types of optimization the modeling of unit seriously dependent on overall optimization algorithm. For example if we use MILP for overall optimization, we should use linear model for this work. The nonlinear modeling method are correctly but difficult to solve. Therefore we select genetic algorithm for overall optimization. Therefore we can use any model for units as white, black and grey box (e.g. ANN).
contact: Mr.mehdi azarmi