Real Time Operational Optimization of a Seawater Desalination System Based on Rolling Prediction of Hourly Freshwater Demand

Abstract

Because the optimal operation trajectory obtained under a fixed freshwater demand and known operational conditions could not achieve cost savings when these parameters change uncertainly over time, in this paper, a real time operational optimization method based on rolling prediction of hourly freshwater demand is proposed. First, via analysis of the historical data of hourly water consumption in a seawater desalination system, a new method for predicting the daily water demand in the next 24 h is proposed, and the predicted trajectory is continuously corrected and updated with present freshwater consumption. Then, with the well-established seawater desalination system model combined with the dynamics of a storage tank, a real time optimal operational problem with the strategy for its solution is given to minimize the daily operational cost. The optimization problem with differential and algebraic equations is discretized into a nonlinear programming problem by finite element collocation, and then a rolling optimization solution strategy based on simulation is used. Finally, a case study is used to verify the proposed method. The results show that the proposed optimal operation method can achieve significant cost savings and can also overcome the water level violation problem under a fixed freshwater demand or via the conventional method.

Simultaneous Optimization of Size and Operation for Seawater Reverse Osmosis Network with Permeate Split and Interstage Permeate Split Designs

“…33 A total of 24 subintervals with three collocation points for 1 day are addressed, which is suitable for the dynamic process of the water tank. 33,34,37 The optimal flow structure and control trajectories in 0−24 h can be obtained by solving the formulated MINLP problem.…”

“…Because the transient time of the spiral wound element is normally in the range of few seconds to several minutes, the manipulated variables could not be changed too frequently . A total of 24 subintervals with three collocation points for 1 day are addressed, which is suitable for the dynamic process of the water tank. ,, The optimal flow structure and control trajectories in 0–24 h can be obtained by solving the formulated MINLP problem.…”

“…Uncertainties exist during the operating of a seawater RO plant. Many uncertainty situations, such as the fouling factor, energy, and financing costs, as well as the fluctuation of product water requirement during operation, are considered in the optimization model of the RO system. ,, A two-stage stochastic approach with recourse is utilized for the optimization of size and operation of the RO network under uncertainties. The two-stage stochastic model can be expressed as follows where The first two equations and the last two equations are the first-stage problem and the second-stage problem, respectively.…”

“…For the research studies provided by refs − , significant cost savings for the RO system could be achieved using the time-of-use power price and buffering effect of the water tank, and the operating conditions can also be adjusted according to the variations of the freshwater demand and seawater temperature. Zhang et al proposed an actual time operation optimization model. The uncertainty of large fluctuation of product water requirement with time was considered; thus, significant cost saving was achieved.…”

“…More cost and energy reductions are expected. Some of the operation optimization research studies predefined the size of the system (the number of pressure vessels and the number of membrane modules inside the pressure vessel) and the volume of the water tank, which might not be a better solution. − , The size of the RO system might influence both the water cost and energy consumption . On another hand, the water tank in the RO system should give enough buffer capacity.…”