(341m) Metaheuristic Optimization for the Structural and Operating Conditions of the Methanol Process

This paper presents a systematic approach for the synthesis, selection, design, and optimization of methanol plants. A metaheuristic optimization approach is proposed to reconcile the economic and environmental objectives of the process while incorporating simulation tools for the reliable modeling of the generated alternatives. The economic objective function consists of maximizing the net profit. The environmental objective function is aimed at minimizing the total annual CO₂ emissions. To perform the optimization of this problem, process data were obtained from the chemical process simulation software ASPEN HYSYS^®. A client-server interface based on Component Object Module technology using Excel-Visual Basic for Applications scripts was developed to call the ASPEN HYSYS^® simulator repetitively for various sets of input variables.

The discovery of substantial reserves of shale gas in recent years has spurred technological innovations and development to monetize the gas into value-added products.^1,2 Methanol is an intermediate product that can be used to produce a vast number of chemicals including formaldehyde, acetic acid and dimethyl ether.³ Shale/natural gas can be converted to methanol via catalytic reforming where syngas with certain hydrogen to carbon oxides ratio is produced.⁴ There are different conventional reforming pathways including steam reforming (SMR), partial oxidation (POx), and auto-thermal reforming (ATR).

The methodological strategy that is followed in this work consists in the construction of a process flow diagram for the simulation process; this way, the operating conditions or design specifications can be manipulated in order to fulfill objective functions. It is necessary to construct the process flow diagram for each configuration in the commercial process simulator

The multi-objective optimization problem states an important degree of difficulty so that a suitable optimization strategy must be used. In this work, it has been used an improved multi-objective differential evolution (I-MODE⁵) algorithm. The values for the parameters associated to the used I-MODE algorithm are the following: population size (PS): 10 individuals, maximum number of generations (MNG): 100, taboo list size (TLS): 5 individuals, taboo radius (TR): 0.01, crossover fraction (CF): 0.5, mutation fraction (F): 0.5.

In general, each reformer can be modified in the reactant ratio (steam-to-methane) and operating conditions (temperature or pressure). Five decision variables were selected in each of the nine analyzed cases (1a, 1b, 2a, 2b, 3, 4, 5, 6 and 7).

The implementation of the global optimization approach involves a hybrid platform, which links ASPEN HYSYS^TM and Microsoft (MS) Excel^TM through a Component Object Module (COM) Technology. During the optimization process, the decision vector of design variables is sent from MS Excel^TM to ASPEN HYSYS^TM, in this process simulator rigorous calculations for the data that identify a design of the process are obtained (e.g., temperature and pressure in the reactors) via resolution of the mass and energy balances in each unit and accounting for the thermodynamic and design equations. These data are returned from ASPEN HYSYS^TM to MS Excel^TM for calculating both objective functions, the values obtained for the objective functions are evaluated and new vectors of design variables are generated according to the stochastic procedure of the used method.

The variables that are considered as well as the search intervals are decisive in the obtained results. In this work, it was decided to select variables whose manipulation definitely has a considerable impact on the performance of the objective functions. Likewise, it was observed that specifying a very large number of MNG as well as of i, generated inflexible solutions that ended with stopping the algorithm. The parameters of the use of the evolutionary algorithm proved to be efficient in solving this particular case. The simultaneous analysis of the possible configurations previously raised in the process flow diagrams and simulated in the software, allows not only to find the best operating conditions of each one, but also to choose among all of them which are the ones that have the best performance, as reported in this work.

REFERENCES

[1] Al-Douri, A.; Sengupta, D.; El-Halwagi, M. M. Shale gas monetization - a review of downstream processing to chemicals and fuels. J. Nat. Gas Sci. Eng. 2017, 45, 436-455. DOI: 10.1016/j.jngse.2017.05.016

[2] Zhang, C.; M. M. El-Halwagi. Estimate the capital cost of shale-gas monetization projects. Chem. Eng. Prog. 2017, 113(12), 28-32. Available at: www.aiche.org/resources/publications/cep/2017/december/estimate-capital-...

[3] Riaz, A.; Zahedi, G.; Klemeš. J.J. A review of cleaner production methods for the manufacture of methanol. J. Clean. Prod. 2013, 57:19-37. DOI: 10.1016/j.jclepro.2013.06.017

[4] Martinez-Gomez, J.; Nápoles-Rivera, F.; Ponce-Ortega, J.M.; El-Halwagi, M.M. Optimization of the production of syngas from shale gas with economic and safety considerations. Appl. Therm. Eng. 2017, 5:110. 678-685. DOI: 10.1016/j.applthermaleng.2016.08.201

[5] Sharma, S., Rangaiah, G. P. An improved multi-objective differential evolution with a termination criterion for optimizing chemical processes. Comput. Chem. Eng. 2013, 56, 155-173. DOI: 10.1016/j.compchemeng.2013.05.004