Effect of Cracking Furnace Parameters through Transfer Line Exchanger Outlet Temperature

摘要

In the cracking furnace operation, many process variables involve in the unit, each of which has a particular characteristic and impact to the process. These variables are usually set based on past experience, which lead to inefficient operations such as having a short operating cycle and increasing the probability of severity in the process. Furthermore, frequent process shutdown for decoking furnace with short cycle lengths leads to high operating cost and coil deterioration. In this study, stepwise regression models are developed to determine the relationship between ten input process variables of the cracking furnace and the transfer line exchanger outlet temperature as the output so that key process parameters could be controlled to maintain the desired operating conditions and to predict the outlet temperature three and seven days in advance so that the maintenance schedule for decoking could be planned. By using stepwise regression, the model of process relationship reveals that only eight from ten process variables and their 2^nd order interactions best explains the output variable, with the R-squared adjusted of 93.63%. For the 3-day and 7-day ahead prediction models, fewer process variables and their interactions are used to explain the output variable, with MAPE of 4.65% and 1.83% respectively.