Pattern Recognition in Industry

In essence a petroleum refinery or chemical plant may be thought of as a series of tubes through which feedstock is converted to more valuable products through chemical reactions and by adding and removing heat and different material components at various points along the tubes. Hydrocarbons constitute the bulk of the material flowing through these tubes. As the flowing material is heated, coke is formed and, along with other foulants, is deposited inside the tubes. This phenomenon is prevalent in the tubes of fired heaters and heat exchangers. Fired heater tubes are most vulnerable to the formation and deposition of coke as certain sections of the tubes experience very high heat flux and consequently high temperatures. As the coke layer builds up with time, the heat transfer resistance of the composite tube wall and coke layer increases, thereby requiring increasing skin, or tube metal, temperature (TMT) in order to maintain the amount of heat transferred to the flowing fluid. Ultimately, the tube metal temperature reaches the pipe's metallurgical limit and the unit has to be shut down and decoked. Unplanned unit shutdown can be prevented by predicting coke or foulant deposition rates and consequent TMT increase. This enables the plant operator to develop coordinated operating strategies and make operating decisions for maximizing asset utilization.
This case study deals with a fired pipestill heater that experienced high local TMT increases in one of its tubes. This resulted in shorter run lengths, as the heater tubes needed frequent decoking in...