Wha's that Smell? analytical Measurements for Odorous Compounds in Pulp and Paper Mill Wastewaters

The National Council for Air and Stream Improvement, Inc. (NCASI) has developed, evaluated, and applied a variety of analytical methods to investigate sources of odor in samples from pulp and paper mill wastewater treatment plants (WWTPs). Accurate, reproducible measurements of sulfide and other reduced sulfur compounds (RSCs) in wastewaters are of considerable importance to the industry. NCASI Method RSC-02.02 was developed for quantification of total (inorganic) sulfide and the organic reduced sulfur compounds (ORSCs) methyl mercaptan (MeSH), dimethyl sulfide (DMS), dimethyl disulfide (DMDS), and dimethyl trisulfide (DMTS). To determine overall odor strength and identify odorants not targeted by RSC-02.02, gas chromatography/olfactory-mass spectrometry (GC/O-MS) was utilized in combination with purge and trap on samples from nine pulp and paper mill WWTPs. In addition, evaluations of a colorimetric methylene blue technique and sulfide ion detector tubes were completed. This paper presents information regarding development, validation, and application of these analytical tools for investigation and reduction of odors at pulp and paper mill WWTPs. These details will also be useful to industries and municipalities that wish to conduct analyses of aqueous wastewaters in support of odor investigation and reduction activities.RSC-02.02 utilizes separate preservations and injections for determination of total sulfide (zinc acetate at pH 10) and ORSCs (ascorbic acid at pH 2.5). All samples are acidified (pH <2.5) prior to direct injection on a GC equipped with a pulsed flame photometric detector (PFPD). During NCASI's study, daily calibration verifications yielded average recoveries of 106% for total sulfide (n=94) and average recoveries ranging from 95 to 102% for ORSCs (n=42). Method blanks were free of the target analytes. Precision and accuracy were assessed using surrogate and matrix spike recovery experiments and replicate analyses. Surrogate recoveries for total sulfide, MeSH, DMS, DMDS, and DMTS in more than 1077 samples ranged from 73 to 131%, with an average recovery of 106%. Matrix spike recoveries averaged 93% (total sulfide), 106% (MeSH), 102% (DMS), 112% (DMDS), and 96% (DMTS). Precision results as reflected by pooled relative percent differences (RPDs) for duplicate analyses ranged from 2.1 to 5.3%. Storage stability studies indicated stability of samples for up to 14 days. Injection pH significantly impacted recovery of total sulfide, with pH 2.5 yielding the highest recovery (96%). Studies to assess matrix and sampling variability yielded average relative standard deviations of 38.9% (total sulfide), 29.8% (MeSH), 20.6% (DMS), 34.2% (DMDS), and 41.0% (DMTS), well above the variability of ∼5% observed for the analytical method.Investigations conducted in conjunction with odor reduction studies at 22 mills yielded a wide range of results for RSC concentrations from similar locations within WWTPs. Median concentrations at primary clarifier outlets were 3.5 mg S/L (total sulfide), 38 μg S/L (MeSH), 66 μg S/L (DMS), 22 μg S/L (DMDS), and <20 μg S/L (DMTS). Median concentrations at the front ends of aerated stabilization basins (ASBs) were 2.9 mg S/L (total sulfide), 60 μg S/L (MeSH), 68 μg S/L (DMS), 68 μg S/L (DMDS), and <20 μg S/L (DMTS). Median concentrations at mid-points of treatment were 0.29 mg S/L (total sulfide) and <20 μg S/L (ORSCs). In final effluents, sample medians were <20 μg S/L for all target analytes.The GC/O-MS method utilizes a combination of purge and trap, GC, olfactory, and MS detection. An odor wheel was created to assist in standardizing odor descriptions for pulp and paper mill samples. The odor wheel was formulated based on descriptions obtained during assessments of standard stocks and pulp and paper mill samples and was utilized by human sniffers throughout the study. Odor intensity was described as threshold (t), weak (w), moderate (m), or strong (s). An odor standard of geosmin was injected into all the samples analyzed in order to provide a reference for odor intensity and consistent descriptors. GC/O-MS analyses were conducted in duplicate and blind using two individual sniffers. The compounds most commonly associated with medium to strong odors during GC/O-MS analyses of samples from nine WWTPs were RSCs, terpenes, ketones, aldehydes, and alcohols.A single laboratory evaluation of a colorimetric total sulfide method based on methylene blue (Hach Method 8131) was conducted on aqueous samples collected from primary clarifiers, ASBs, final effluents, and other mill wastewaters at ten bleached and unbleached kraft pulp and paper mill WWTPs. It indicated that the method was effective in the range of 0.04 to 2 mg S/L. The cost and ease of application make this method a reasonable alternative for measuring total sulfide levels in wastewater samples with concentrations above ∼40 μg S/L. In conjunction with these studies, sulfide ion detector tubes were also examined for applicability in aqueous samples collected at four pulp and paper mills. Tubes evaluated were Sensidyne 200AB (0.5 to 10 ppm) and 200SB (2 to 1000 ppm), and Gastec 211 (1 to 100 ppm) and 211LL (0.5 to 20 ppm). There was good linearity across the working ranges of the tubes, and good precision for the individual tubes was observed for replicates of spiked reagent grade water and in sample matrices. Precision was limited due to the demarcations on the tubes, analysts' interpretations of color development, and the type of tubes employed. Accuracy as indicated by matrix spike experiments was biased high for Gastec 211LL and 211 tubes. Interference due to methyl mercaptan (5 ppm) was not observed in the samples tested. The low-level tubes, Gastec 211LL and Sensidyne 200SB, gave low biased results and some false negative readings in WWTP samples.