Journal of Environmental Management 278 (2021) 111497 Available online 29 October 2020 0301-4797/© 2020 Elsevier Ltd. All rights reserved. Research article Response surface methodology and artificial neural network modelling for the performance evaluation of pilot-scale hybrid nanofiltration (NF) & reverse osmosis (RO) membrane system for the treatment of brackish ground water Alka Srivastava a , Aghilesh K a , Akhil Nair b , Shobha Ram b , Smriti Agarwal c , Jahangeer Ali d , Rajneesh Singh e , Manoj Chandra Garg a, * a Amity Institute of Environmental Sciences, Amity University Uttar Pradesh, Sector 125, Noida, Uttar Pradesh, 201313, India b School of Engineering, Gautam Buddha University, Greater Noida, 201308, India c Department of Electronics and Communication Engineering, Motilal Nehru National Institute of Technology Allahabad, Prayagraj, Uttar Pradesh, 211004, India d Biological System Engineering, University of Nebraska, Lincoln, United States e Nebraska Water Center, University of Nebraska, Lincoln, United States A R T I C L E INFO Keywords: Brackish groundwater Nanofiltration Response surface methodology Reverse osmosis Artificial neural network ABSTRACT Artificial neural network (ANN) and response surface methodology (RSM) were employed to develop models for process optimisation of pilot scale nanofiltration (NF) and reverse osmosis (RO) membrane filtration system for the treatment of brackish groundwater. The process variables for this study were feed concentration, tempera- ture, pH and pressure. The performance of NF/RO was assessed in terms of permeate flux, water recovery, salt rejection and specific energy consumption, which were considered as responses. The experimental design was employed to develop both RSM and ANN models. RSM model was validated for the whole range of experimental levels, while the ANN model was considered for the whole range of experimental design. RSM and ANN models were statistically analysed using analysis of variance (ANOVA). Further, the models were graphically compared for its predictive capacity. Numerical optimisation of NF and RO pilot plant to determine the optimum conditions were verified. Finally, using the optimum conditions, three hybrid configurations of NF and RO were studied to determine the best mode for the treatment of brackish groundwater. It was found that parallel NF-RO had a recovery of 57.18% and rejection of 44.89%, for RO-concentrate-NF (RO-C-NF) recovery was 49.55% and rejection of 38.64% & for NF-concentrate-RO (NF-C-RO), the recovery of 39.53% and rejection of 49.66% was obtained. Results obtained also suggested that the mode of configurations and the feed concentration affect the performance of the hybrid system. 1. Introduction Water, despite being one of the most necessary natural resource for the survival, is continuously being highly exploited and contaminated by anthropogenic activities. Groundwater, as one of the major water resources, fulfils 40% of the nation’s water demand (Arfanuzzaman and Atiq Rahman, 2017; UNESCO, 2019; Wakode et al., 2018). In fact, at some places, the amount of water contamination is at such a high level that the use of it has been restricted for daily amenities. With the in- crease in water demand due to the overburdened population, the existing freshwater resources are overwhelmed. Water pollution is generally seen with respect to the surface waters, but with the increase in pollution and anthropogenic activities, groundwater is also no less affected. Salinity, nitrate and nitrite are the most significant pollutants, presents in groundwater, drawing serious attention from environmen- talists. Presence of salinity, along with nitrate and nitrates, can render the waters unfit for consumption. Nitrite causes serious health impacts like blue baby syndrome (methemoglobinemia) in humans (Guo et al., 2018). The reduction of nitrates and nitrites from groundwater has been achieved by many treatment alternatives, including application of * Corresponding author. E-mail addresses: alka.srivastava@s.amity.edu (A. Srivastava), kaghilesh@amity.edu (A. K), akhilgbu@gmail.com (A. Nair), Shobharam@gbu.ac.in (S. Ram), smritiagarwal@mnnit.ac.in (S. Agarwal), Jahangeer@unl.edu (J. Ali), rsingh15@unl.edu (R. Singh), mcgarg@amity.edu (M.C. Garg). Contents lists available at ScienceDirect Journal of Environmental Management journal homepage: http://www.elsevier.com/locate/jenvman https://doi.org/10.1016/j.jenvman.2020.111497 Received 10 July 2020; Received in revised form 20 September 2020; Accepted 7 October 2020