Statistical regression and modeliing analysis for reverse osmosis desalination process Sobana Subramani, Rames C. Panda ⁎ Department of Chemical Engineering, CSIR-CLRI, Adyar, Chennai 600 020, India HIGHLIGHTS • Correlations indicate that the permeate characteristics depend on feed flowrate. • The model is also validated using plant data. • The model for permeate flow-rate sup- ports the use as a 2nd degree prediction model. • Modeling 2-input, 3-output will help in developing control strategies. GRAPHICAL ABSTRACT abstract article info Article history: Received 22 May 2014 Received in revised form 23 July 2014 Accepted 26 July 2014 Available online xxxx Keywords: Desalination Reverse osmosis Statistical model Regression Multivariate ANOVA Experiment is conducted, across different units of a typical desalination plant, to build correlations between inputs/ outputs. Steady-state statistical models of reverse osmosis (RO) are developed using stream characteristic data (flow rate, concentration and pH) over a period of time. Data-driven models are useful for deciding real time oper- ational control strategies of desalination plant. The statistical analysis of correlations obtained indicates that perme- ate characteristics depend on feed stream flow rate by a second degree polynomial. Significance of regression was evaluated based on multivariate ANOVA analysis, on visual standardized residuals distribution and their means for confidence levels of 95% and 99%, clearly validating these models. Sensitivity of parameters is found from interaction and co-relation studies. These models will help in safe operation and control of RO-desalination plant. © 2014 Elsevier B.V. All rights reserved. 1. Introduction Desalination is used to separate salts from raw water for use in boiler feed, thermal power generation, electronic industries, chemical indus- tries, textile industries, and leather industry and also for production of portable water. The process is carried out through distillation, multiple effect vapor compression, evaporation, or membrane processes such as electro-dialysis reversal, nano-filtration, and reverse osmosis (RO). Being advantageous over other separation techniques, reverse osmosis is basically a pressure driven process where no energy phase change (or) potentially expensive solvents (or) adsorbents are needed. It is sim- ple to design and operate compared to other traditional separation pro- cesses. Separation of inorganic, organic and microbiological impurities is simultaneously done by RO. A schematic of the process is shown in Fig. 1. There are mainly two inputs, namely, pump pressure (ΔP) and Desalination 351 (2014) 120–127 ⁎ Corresponding author. Fax: +91 44 24911589. E-mail address: panda@clri.res.in (R.C. Panda). http://dx.doi.org/10.1016/j.desal.2014.07.038 0011-9164/© 2014 Elsevier B.V. All rights reserved. Contents lists available at ScienceDirect Desalination journal homepage: www.elsevier.com/locate/desal