English Abstract
Abstract :
This research aims to better understand the optimization parameters that
substantially impact the ethylene and propylene output of non-integrated and integrated
plants. To achieve this goal, we propose three single-objective optimization models and
three trade-off scenarios for the intermediate and final materials produced from the oil
refinery, ethylene production plant, and integrated plant and then evaluate the
performance of each plant in terms of profitability, production, income, inventory costs,
and unit operating costs. Profit maximization of plant products is the chief goal of the
single-objective optimization model for each plant. We rerun the single-objective
model multi-periodically and use an artificial neural network (ANN) framework to
predict the future price of crude oil and its derivatives for 2021, 2030, 2040, and 2050.
Three trade-off scenarios are proposed for the multi-objective model. They include the
multi-objective optimization scenario for ethylene-prioritized production, the multi-
objective optimization scenario for propylene-prioritized production, and the single-
objective optimization scenario for the production of ethylene and propylene. All the
models and scenarios proposed in this thesis are formulated as mixed-integer nonlinear
programming (MINLP) problems. After developing and examining all the proposed
models, our findings show a substantial improvement in the intermediate and final
products of the ethylene production plant and the oil refinery after the integration. For
the single-objective model, net profit of the integrated plant increases from 4% ($4,
384, 500) in 2021 to 29% ($32, 330, 000) and 38% ($41, 120, 000) in 2030 and 2050,
respectively. The multi-objective optimization model of an integrated plant is also
found to be superior to the single-objective optimization model of an integrated plant
in terms of overall performance. The net profit per day generated from the single-
objective model, multiple-objective plant (ethylene production prioritized), and multi-
objective plant (propylene production prioritized) is $799,647.6, $821,152.6, and
$8401, 12.2, respectively. The performance comparison of the two multi-objective trade
scenarios reveals that the multi-objective model for propene-prioritized optimization
produces more ethylene and propylene than the multi-objective model for ethylene-
prioritized optimization. Overall, this research results reveal the potential of an
integrated plant based on a multi-optimization model in helping the world boost
petroleum materials' production and profitability.
