| With the growing environmental concern in China, It is inevitable to develop clean and renewable energy resources nowadays. As a promising source of clean fossil fuel, liquid natural gas has been widely used because of its clean burning characteristics. Natural gas liquefaction plant is the core part of production of liquefied natural gas. However, previous studies show that some problems of natural gas liquefaction plant exist such as object independency, the process can't get the preferred operating parameters quickly, and also can't assess the dynamic response of the liquefaction process after disturbance. To address these problems, this paper works on natural gas purification process and two propane precooled mixed refrigerant cycle natural gas liquefaction processes. The key elements are listed as follows:(1) In order to avoid severe corrosion of vessel or freezing in cryogenic exchangers, feed gas from the well heads containing various contaminants and acid gases cannot be sent to the liquefaction unit directly. This process must pass through purification units to remove the undesirable components such as CO2, H2S and H2O. In this article, we couple DEA and TEG methods to remove those undesirable components and obtain natural gas purification process in order to reach the purification target. We firstly generate a process superstructure, then develop a simulation-assisted mixed linear programming (MLP) to optimize the capital cost and utility cost generated by the interconnection streams. Subsequently, we use pinch point technology to perform heat exchange network (HEN) design and analysis to accomplish the maximum energy saving target. On the other hand, considering the multi-objective demand of the process, AHP-Fuzzy Comprehensive Evaluation is utilized to comprehensively analyze the purity levels, energy efficiency, the height of packed tower and other assessment indicators in this process, then optimal process parameter is selected.(2) The purified natural gas is cooled to liquid natural gas (LNG) through liquefied process. In this paper, we propose a novel doubled propane precooled mixed refrigerant cycle for base-scale natural gas plant considering the plant scale and energy consumption of different natural gas liquefaction processes. The proposed process is then simulated by Aspen HYSYS. Next, the energy consumption of compressor as an objective function is optimized with these key parameters including the composition of mixed refrigerant, the inter-temperature between multi-heat exchangers, the low pressure of mixed refrigerant and the high pressure of mixed refrigerant. Genetic algorithm (GA) in Matlab is selected as the optimization method for the process. By using HYSYS server, Matlab can interface with Aspen HYSYS by using HYSYS server. Furthermore, the effect of operating variables on energy consumption and minimum temperature difference approach of LNG exchangers are analyzed. A better direction of this process is finally proposed combined with Thermodynamic Analysis.(3) The liquefaction unit is isolated as an independent research object by controlling multiple variations in the compressor and propane precooled units. Besides, the operating variables and controlled variables under different operating conditions form the multi-input and multi-output process systems correspondingly. Based on the steady optimized result and dynamic performance evaluation, the most favorable control structure is obtained by calculating condition number for steady Gain Array and Relative Gain Array. Finally, we use the process simulator Aspen HYSYS for a rigorous dynamics simulation of the C3MR process. Better flexibility of the optimized control structure under various operating conditions is shown through dynamic response evaluation of the optimal control structure after disturbance in raw natural gas... |
PDF Full Text Request