(MINLP) for synthesis and optimal operation of cogeneration seawater desalination plants

This paper presents a rigorous mixed-integer nonlinear programming (MINLP) model for optimal synthesis and design of dual-purpose seawater desalination plants. The proposed superstructure considers more alternative configurations than the model recently proposed by Mussati et al. [1] and all process equipment is modelled in a rigorous way. The MINLP model introduces binary variables in order to select equipment for the cogeneration plant. The detailed model for the MSF desalter developed by Mussati et al. [2] was considered. The MSF mathematical model involves the real-physical constraints for the evaporation process. Nonlinear equations are used to model all plant equipment rigorously in terms of chemico-physical properties (enthalpies, entropies, vapor pressure) and design equations (efficiencies, NEA, BPE, heat transfer coefficients, momentum balances, among others). The proposed model is not only useful for synthesis, but also for analyzing different design alternatives. The model has been implemented in a general algebraic modelling system [4]. Several study cases were successfully solved by applying the MINLP model. A case study is presented in order to illustrate the model’s capabilities.

A MINLP model for optimal synthesis and design of a dual-purpose desalination plant system has been developed by modifying the previous model developed by Mussati et al. [1]. A large number of different possible configurations were included in the superstructure for the MINLP model (additional burners, air pre-heater, stream splitters, among others). Also, a more detailed description of the different equipment (gas turbine, high-pressure and back-pressure steam turbines, deareator) was introduced. The mathematical model for an optimal MSF desalter design developed by Mussati et al. [2] has been introduced in the rigorous model. The resulting MINLP model is not only useful for synthesis, but also for analyzing different configurations. Different examples have been successfully solved (not presented in this paper). The proposed MINLP model is characterized by its robustness and flexibility. The same qualitative results presented in Mussati et al. [1] have been obtained in this paper. However, in all cases, more detailed designs for the power generation cycle and desalter were achieved. The preference of one scheme over another would depend mainly on many factors such as the required power to water ratio, cost of fuel energy charged to the desalting process, electricity sales, capital costs, and local requirements. The costs have a strong influence in selecting the structure. It is not possible to establish a generally valid economical viability of configurations. A very detailed analysis must be done for each situation. For example, back-pressure or/and extraction/ condensation steam turbines coupled to a MSF desalter are preferred for low power/water ratios whereas combined cycles (gas turbine + back- pressure steam turbine and/or gas turbine+backpressure steam+low-pressure steam turbines) coupled to a MSF desalter are preferred for high ratio values.