You are here:
Multivariate and Molar Absorptivity-Based Chemometric Models for Quantifying Nitrite and Nitrate in Chloramine Systems
Citation:
Do, T., J. Fairey, AND D. Wahman. Multivariate and Molar Absorptivity-Based Chemometric Models for Quantifying Nitrite and Nitrate in Chloramine Systems. 2022 Water Quality & Technology Conference, Cincinnati, OH, November 14 - 17, 2022.
Impact/Purpose:
The environmental or health problem addressed by the study: Measurement of nitrite and nitrate in drinking water A general description of the work and results: Experiments have been conducted to evaluate a new methodology to measure nitrate and nitrate in drinking water The long term importance or significance of the findings: Provides online monitoring for nitrite an nitrate in drinking water Who would be interested in or could apply the results (e.g. program or regional partners, general public, local communities): Researchers and drinking water utilities trying to prevent nitrification in drinking water systems.
Description:
Temporal nitrite (NO2−) and nitrate (NO3−) profiles can be used to assess biological oxidation of free ammonia and NO2−, respectively, which define a nitrification event in chloraminated drinking water distribution systems (CDWDSs). Several NO2− and NO3− chemometric models (CMs) have been formulated using (1) multivariate regression of ultraviolet absorbance spectra between 190–400 nm (UVA190–400) and (2) simple linear regression of the second derivative in wavelength between 215–226 nm (UVA″215–226). UVA-based CMs do not require reagents and can be collected in the field by grab sampling or continuously as part of sensor systems. Concerns with the generalizability of multivariate CMs exist due to differences in source water natural organic matter (NOM) and inorganics such as bromide which may manifest in many significant independent variables (SIVs). UVA″-based CMs are insensitive to differences in NOM, but their accuracy for tracking NO2− and NO3− under conditions typical in CDWDSs has not been demonstrated. In this presentation, we will introduce a novel CM based on the second derivative of molar absorptivity (¿″) which was circa zero for three drinking water NOM types, in contrast to the ¿″ spectra for NO2− and NO3− that have maxima’s of about 12- and 31 L•mol−1•cm−1•nm−2 at 232- and 226 nm, respectively. Thirteen biofilm annular reactor (BAR) tests were completed using four feedwaters with varying bromide (0.08–0.25 mg¿L−1-N), two from Fayetteville AR (FAY1 and FAY2) and two from Dallas TX (DAL1 and DAL2). Monochloramine and free ammonia were manipulated to stimulate cycles of nitrification onset and arrest. Daily measurements of NO2− and NO3− by UV spectroscopy and ion chromatography were used to develop multivariate UVA″-based CMs and four-point spectral deconvolution was used to develop an ¿″-based CM for NO2−, NO3−, bromide, and monochloramine. Each CM was calibrated using NO2− from FAY1 (n = 275) and validated against the other three feedwaters (n = 376). The multivariate UVA″-based CMs were strong, with root-mean square errors (RMSEs) of 0.029- and 0.051 mg¿L−1-N for NO2− (SIVs = 8) and NO3− (SIVs = 5), respectively. The ¿″-based CM was similarly strong with RMSEs of 0.042- and 0.046 mg¿L−1-N for NO2− and NO3−, respectively. The findings indicated that the ¿″-based CM could be used to accurately track NO2− and NO3− under conditions relevant to CDWDSs and are broadly generalizable given the second derivative is insensitive to NOM.