Process control in water desalination industry: an overview

Process control is an essential part of the desalination industry that requires for operation at the optimum operating conditions an increase in the lifetime of the plant and reduction of the unit product cost. A review is presented for the commonly used and newly developed control and instrumentation in MSF and RO plants. Process control may be as simple as an on/off valve that is triggered upon offset of the system measured parameters from the desired set-point. More classical and commonly used controllers have combined proportional, integral, and derivative (PID) systems. Also, proportional/integral (PI) controls are used in industry. Controls selection aims at fast response, high stability, and minimum disturbance to the system. Less common controllers include fuzzy logic-based systems. Early testing of such systems shows the need for mathematical analysis of various control loops within the plant, development of control rules, and development and testing on industrial scale. Supervisory systems such as model predictive control are also considered to obtain an integrated control systems of the whole plant. It should be stressed that although the desalination plants are highly complex, accurate and detailed mathematical models for steady state and process of various desalination processes are found in the literature. Such models are necessary for studying plant performance, various control strategies, and forms an essential part for any serious analysis and development of novel and new control systems.

A review is presented for the control loops and instrumentation used in the two major desalination industries, MSF and RO. The following conclusions are made in light of the above review: Current practice adopts the classical control system, which includes PID and PI controllers. The major part of the control systems is based on single input/single output configuration. Instrumentation, indicators, and alarms form an integral part of the control system. The corrosive and fouling nature of the seawater, brine, and distillate necessitate protection of the sensor part of the instrument and proper location far from highly turbulent or dead zones. Alarms are found as limited for the most sensitive parameters, which may include the TBT or brine level in the last stage. Fuzzy logic and supervisory systems are in an early stage of development for the MSF system. However, actual realization of such control systems still requires exhaustive mathematical analysis and more involved field-testing. Accurate and detailed steady-state and dynamic models for both processes form an essential part for development of an efficient and optimum control system. The literature includes various types of models, varying from the simple closed-form formulas to the more detailed numerical models. In summary, the desalination industry is highly conservative because of the strategic nature of the product. Therefore, adoption of novel and advanced control systems will require intimate cooperation between research establishments and the desalination industry. This is necessary to perform the necessary research, development, and field-testing.