Nitrite Quantification by Second Derivative Chemometric Models Mitigates Natural Organic Matter Interferences under Chloraminated Drinking Water Distribution System Conditions

Abstract

Nitrite (NO2−) production in chloraminated drinking water distribution systems (CDWDSs) is among the first bulk water indicators of a nitrification event and is typically quantified using ion chromatography (IC) or colorimetric techniques. NO2− can also be quantified using chemometric models (CMs) formulated using molar absorptivity (Ɛ) and/or ultraviolet absorbance (UVA) spectra, but concerns exist regarding their accuracy and generalizability because of varying source water natural organic matter (NOM), monochloramine (NH2Cl), bromide (Br−), and other species in CDWDSs. We demonstrate that the impact of NOM was mitigated in the second derivative molar absorptivity (Ɛ″) and UVA spectra (UVA″) between 200–300 nm and developed a generalizable CM for NO2− quantification. The Ɛ″+UVA″ CM was calibrated with daily NO2− measurements by IC from five biofilm annular reactor (BAR) tests with feedwater from Fayetteville, Arkansas, USA (FAY1, n = 275) and validated with eight BAR tests (n = 376) with another Fayetteville water (FAY2) and two waters from Dallas, Texas, USA (DAL1 and DAL2). The Ɛ″+UVA″ CM used Ɛ″ for NO2−, nitrate (NO3−), Br−, and NH2Cl at wavelengths of 213-, 225-, 229- and 253 nm, had an adjusted R2 of 0.992 for FAY1 and 0.987 for the other waters, and had a method detection limit (MDL) of 0.050 mg·L−1-N. NO2− challenge samples with three reconstituted NOM types and Br− indicated the Ɛ″+UVA″ CM was generalizable at NOM concentrations like those in the BAR tests (≤ 2.5 mg·L−1-C). The Ɛ″+UVA″ CM accurately simulated NO2− in field tests from two CDWDSs undergoing nitrification, including one with NOM at 3.5 mg·L−1-C, illustrating a practical application of the CM for identifying biological ammonia oxidation.