(48c) How Do You Choose One Among 280 Million Scenarios?

We have 6 cracker in Brazil and one of them was designed considering the price of the crude oil less than U$10 a barrel and because the low utility integration with the downstream a large number of our compressor had been designed to use a condensing steam turbine.

Since the first start up until nowadays we did four revamps that increase our capacity from 180 to 700 kta of ethylene and include one ROG unit that operate in parallel with the naphtha cracker. Also many improvements were done to increase its energy efficiency but most of these then were related with process and few modifications were done to adjust the steam balance or to reduce the amount of steam that was used by condensing steam turbines.

As this specific unit is the oldest one and with the lower energy efficiency, we decide to start a new feasibility study to know what must be done and how much money we will save.

During the operational and technical feasibility we defined fifteen equipments that could be optimized, but even so, each person that had the opportunity to make any comment show us a different scenario to be checked. A combinatory analysis using only preselect options give us almost 2,8x108 scenarios. Due this huge number of scenarios, we developed a computer program using visual basic that allowed test these scenarios and define the most efficient one.

This paper will show how we did this work and how useful it was.