(66a) Optimization Approach for Production Planning in Olefins and Aromatic Plant, Cracking Vs. Reforming

In the cyclical petrochemical industry, there is a more competitive and globalized market than ever. Feedstock can come from any place in the world and products are also available in any part of the world. Associated with Globalization, geopolitical tensions and speculation make the oil market extremely volatile, directly influencing the petrochemical market.

On the eve of another ?low cycle? in the petrochemical industry, with large capacities and high technology projects, the result is an offer surplus and a reduction in the profitability of the companies.

Competitiveness is a key success factor for the companies that are preparing themselves for this challenge. Among other actions, an efficient choice of feedstock, the best use of its assets and the ?exploration? of the most profitable markets are the best answers to boost results.

The volatility of the prices, sometimes benefiting the olefins, sometimes the aromatics products and the complexity and diversity of Braskem's processes which include Reforming and Cracking units, strengthen the necessity of searching the best point between feedstock, market and operations.

To solve this highly complex problem, the acquisition of an advanced optimization application gives companies a competitive advantage. A year and a half ago Braskem, the largest petrochemical company in Latin America, acquired Optience SCMart Suite as the optimization application platform to help their planning and scheduling process.

The main model, developed in SCMart, is a nonlinear multiperiod monthly planning model which includes all plant constraints and uses a nonlinear correlation yield model based on SPYRO simulator for the cracking furnaces. It also includes a nonlinear correlation to estimate the reformer yields as well as a complex model for a naphtha fractionator. The planning model is supported by a scheduling model to achieve optimal results at the daily level. The system also uses a feedstock database library that tracks the available feedstock qualities and properties.

In this work, we discuss the planning module that is used to support the technical and economical decisions at Braskem. The nonlinear planning model consists on 5,000 variables per period including flows, inventories, properties and other process variables. It covers two ethylene plants, one reformer plant, plus other C4s, C5s and aromatics plants. The main products are ethylene, propylene, butadiene, benzene, para-xylene and other aromatics. These decision process variables that are optimized include COTs (Coil Outlet Temperatures), reforming and cut temperatures, possible co-cracks, furnace and reformer utilization, ethane recycling flows and loads to other downstream units.

We present different case studies and scenarios which have been running during these last months that demonstrate how the tool has been helping us in the decisions as the amount and quality of the Naphtha to be bought and processed in each unit. It has also helped us to identify the ideal markets and in the exploration of trade offs and ideal yields, as well as the optimal use of the assets, through concepts of shadow price and profit variability.