Evaluation of Performance of Existing RO Drinking Water Stations in the North Central Province, Sri Lanka

Abstract

Reverse osmosis (RO) drinking water stations have been introduced to provide safe drinking water for areas with prevailing chronic kidney disease with unknown (CKDu) etiology in the dry zone of Sri Lanka. In this investigation, RO drinking water stations established by community-based organizations (CBO) in the North Central Province (NCP) were examined. Water samples were collected from source, permeate, and concentrate in each station to determine water quality and performance. Furthermore, the operators of the systems were interviewed to evaluate operational and maintenance practices to identify major issues related to the RO systems. Results show that the majority (>93%) of RO systems had higher salt rejection rates (>92%), while water recovery varied from 19.4% to 64%. The removal efficiencies of hardness and alkalinity were averaged at 95.8% and 86.6%, respectively. Most dominant ions such as Ca²⁺, Mg²⁺, K⁺, Na⁺, Ba²⁺, Sr²⁺ Cl^-, F^-, and SO₄^2- showed higher rejections at averaged values of 93.5%, 97.4%, 86.6%, 90.8%, 95.4%, 96.3%, 95.7%, 96.6%, and 99.0%, respectively. Low recovery rates, lower fluoride levels in product water, and membrane fouling were the main challenges. Lack of knowledge and training were the major issues that could shorten the lifespan of RO systems.

Keywords: performance; permeate water recovery; reverse osmosis; salt rejection; selectivity.

Figure 1

Locations of CBO-operated RO stations in North Central Province, Sri Lanka.

Figure 2

Permeate recovery rates of each RO stations in NCP.

Figure 3

Performance of RO stations: (a) the salt rejection variation with the feed water EC; (b) hardness rejection efficiency with the feed water hardness; (c) alkalinity rejection rates with corresponding feed water alkalinity; (d) rejection of individual ions and silicon.

Figure 4

Rejections of monovalent ions with membrane age: (a) K rejection, (b) Na rejection, (c) salt rejection with membrane age, and (d) K rejection with operating pressure.

Figure 5

Average rejection of individual ions by three types of RO membranes.

Figure 6

Variation of permeate recovery (left) and the salt rejection rates (right) with the common RO configurations in NCP. Configurations indicated by (a,b) and (d–f) are illustrated in Figure S2.

Figure 7

(a) Ca/Mg selectivity and permeability; (b) cation selectivity of monovalent/divalent (M/D) RO TFC membranes and (c) M/D ratio with groundwater EC; (d) Ca/Mg ratio with groundwater EC in NCP.

Figure 8

Major cation concentrations in the permeate with their feedwater cation concentrations, (a) variation of Mg²⁺ in the permeate with Mg²⁺ in the feedwater, (b) variation of Ca²⁺ in the permeate with Ca²⁺ in the feedwater, (c) variation of Ba²⁺ in the permeate with Ba²⁺ in the feedwater, (d) variation of Sr²⁺ in the permeate with Sr²⁺ in the feedwater, (e) variation of Na⁺ in the permeate with Na⁺ in the feedwater, and (f) variation of K⁺ in the permeate with K⁺ in the feedwater.

Figure 9

Chloride permeability through RO membranes with the concentration of (a) Na⁺ in the permeate, (b) total monovalent cations in the permeate, and (c) divalent cations in the feedwater.