Abstract—The most important process of the water treatment plant process is coagulation, which uses alum and poly aluminum chloride (PACL). Therefore, determining the dosage of alum and PACL is the most important factor to be prescribed. This research applies an artificial neural network (ANN), which uses the Levenberg–Marquardt algorithm to create a mathematical model (Soft Jar Test) for chemical dose prediction, as used for coagulation, such as alum and PACL, with input data consisting of turbidity, pH, alkalinity, conductivity, and, oxygen consumption (OC) of the Bangkhen Water Treatment Plant (BKWTP), under the authority of the Metropolitan Waterworks Authority of Thailand. The data were collected from 1 January 2019 to 31 December 2019 in order to cover the changing seasons of Thailand. The input data of ANN are divided into three groups: training set, test set, and validation set. The coefficient of determination and the mean absolute errors of the alum model are 0.73, 3.18 and the PACL model are 0.59, 3.21, respectively. Keywords—Soft jar test, jar test, water treatment plant process, artificial neural network. I. INTRODUCTION ATER supply is fed to Bangkok by three water treatment plants (WTP), namely BKWTP, Mahasawat Water Treatment Plant (MSWTP), and Samsen Water Treatment Plant (SSWTP). All three use raw water from the Chao Phraya River. In this research, BKWTP is set as a case study. With the largest production of 4,400,000 cubic meters per day, there are 18 clarifiers equipped in two production lines with different chemicals (i.e., Alum and PACL) [1]. The giant BKWTP is a huge challenge in terms of operational cost-effectiveness. Optimal chemical dosages are required for optimum cost- effectiveness. As a guideline for optimal chemical dosages, a traditional jar test has been used for a number of years, although there are a lot of disadvantages. Apparently, the behavior of being offline (labor process) among equipped online sensors (e.g., pH meter, flow meter) causes suboptimal operation since the operator cannot receive information on time, resulting in the so-called bottle-neck problem. In order to alleviate the offline problem, the virtual version of the Jar Test is proposed with the help of ANN modeling and the so-called Soft Jar Test (SJT). The primary goals of this research mainly are: (i) to set up ANN models (SJT) for chemical dosage prediction (i.e., alum and PACL based on Jar Test results; and (ii) to evaluate the performance and limitations of JST. II. WATER TREATMENT PROCESS: BKWTP CASE STUDY The BKWTP’s water treatment process is shown in Fig. 1. As shown in the figure, BKWTP receives raw water from Chao Phraya River at the Samlae pumping station and is conveyed through the 18-kilometer-long canal. The water is pumped into the filtration plant through a filter by rough and fine screens and then chlorine is added to kill germs and algae and to adjust the pH; this process is called the retreatment and pH adjustment process. After that, the water flows into the clarification process, Alum and PACL are added for the coagulation process to destabilize colloid and generate small floc. To increase floc concentration, a coagulation aid is included to form a large floc and it is able to precipitate (sedimentation) into the bottom of the clarifier. The water coming out of the clarifier tank is controlled for turbidity at no more than four Nephelometric Turbidity Units (NTU). The filtration process consists of two layers: coal and sand. The turbidity of the filtered water at this point is not more than 1 NTU. Finally, chlorine is added again to meet sanitation hygiene standards before pumping water to the public. A. The Conventional Jar Test Currently, at the BKWTP, the optimum dosage of alum and PACL is obtained by performing a traditional laboratory jar test. Fig. 2 illustrates the jar test equipment used in the method that has been in place since 1979. The objective of this procedure is to simulate three key processes: (1) coagulation; (2) flocculation; and (3) sedimentation. All these processes are physically demonstrated in one-liter containers with a varied set of agitation [2]. Optimal chemical dosage can be obtained by manually changing chemical doses and considering residual turbidity. In other words, it can be accounted as trial-error by an expert. In BKWTP, Jar testing is performed twice a day at approximately 8:00 a.m. and 4:00 p.m. Ninlawat Phuangchoke is a graduate student, Asst Prof. Dr. Waraporn Viyanon is a faculty member in the department of Computer Science, Faculty of Science, and Dr. Setta Sasananan is a faculty member in the department of Civil and Environmental Engineering, Faculty of Engineering, at Srinakharinwirot University in Bangkok, Thailand (e-mail: ninlawat.phu@g.swu.ac.th, waraporn@g.swu.ac.th, setta@g.swu.ac.th). Ninlawat Phuangchoke, Waraporn Viyanon, Setta Sasananan An Approach for Coagulant Dosage Optimization Using Soft Jar Test: A Case Study of Bangkhen Water Treatment Plant W World Academy of Science, Engineering and Technology International Journal of Environmental and Ecological Engineering Vol:15, No:8, 2021 206 International Scholarly and Scientific Research & Innovation 15(8) 2021 ISNI:0000000091950263 Open Science Index, Environmental and Ecological Engineering Vol:15, No:8, 2021 publications.waset.org/10012165/pdf