I obtained a 1st class master’s degree in chemical and nuclear engineering. Since graduation, I have worked as a technical consultant in the UK. During university, I also took non-engineering modules such as finance and subjects in the humanities, which I enjoyed tremendously as well. I also have a passion for languages. I speak four languages fluently: English, Mandarin, Cantonese and Japanese. I currently work part time as an interpreter at the local police station and community centre. I also do voluntary work as a Chinese/English translator in the local church.
Process Control Project – Control of an Industrial Furnace
The aim of the project was to devise a suitable control strategy for an industrial furnace; to maintain a steady outlet temperature for the crude stream Tout (at 560K) and oxygen concentration in the fuel gas Co2 (at 5%) under various conditions. The furnace is modelled in Paragon 5.4, in which all the simulations are done. The degrees of freedom analysis is first done to give insight into the process. Transfer function models relating to changes in Tout and Co2 to changes in flowrates of oil (Foil), gas (Fgas) and air (Fair) as well as flowrate and temperature of the crude supply (Fin and Tin) are then identified using a first order plus time delay (FOPTD) model. Using relative gain analysis (RGA), the best pairing is found to be Fgas controlling Tout and Fair controlling Co2. Keeping the Co2 loop open, the best Pband and integral time were found to be 3.0 and 1.4 respectively for the temperature loop, which exhibits the best compromise among small overshoot, short settling time and small sum of squared error. A similar approach is applied to the Co2 loop, for which 3.7 and 0.4 are adapted as optimum Pband and integral time. To reject disturbances and further reduce loop minimizations, detuning, decoupling and feedforward control schemes are implemented. The best strategy found is to detune Tout loop, decouple Co2 loop and feedforward Fin. Finally, the performance of the devised strategy is tested on set-point and disturbance changes set by supervisors. Results suggest that a second decoupler on Tout loop and a feedforward controller on Tin might be useful.
Please click here to see this writer’s full sample paper