Effects of total suspended solids, particle size, and effluent temperature on the kinetics of peracetic acid decomposition in municipal wastewater Open Access

PAA

Peracetic acid

TSS

Total suspended solids

PE

Primary effluent

SE

Secondary effluent

CT

Concentration of PAA × contact time (mg.min/L)

COD

Chemical oxygen demand

MLSS

Mixed liquor suspended solids

C_PAA

Residual PAA concentration (mg/L)

C_PAAo

Initial PAA spike concentration (mg/L)

D

PAA demand (mg/L)

t

Contact time (min)

k_d1

First order decay rate constant (min⁻¹)

k_d2

Second order decay rate constant (L/mg.min)

ORP

Oxidation-reduction potential

Peracetic acid (PAA) is an effective disinfectant that has been reportedly used for wastewater treatment as a substitute for chlorine-based compounds (Koivunen & Heinonen-Tanski 2005; Zanetti et al. 2007; De Luca et al. 2008; Beber de Souza et al. 2015). PAA is a strong oxidizing agent with biocidal properties (Luukkonen et al. 2014). The main advantage of PAA disinfection over chlorination is the minimal associated toxic by-products (Kitis 2004; McFadden et al. 2017). The disinfection mechanism of PAA is due to its oxidative nature, causing organism cell membrane damage and thereby inactivation of biological species. Understanding the decomposition behaviour of PAA in municipal wastewater is critical for identifying optimal operating conditions and PAA dose delivery requirements.

In water, PAA decomposes to form acetic acid and oxygen according to the following reaction pathways (Yuan et al. 1997).

It has been shown that wastewater quality parameters such as temperature, pH, organic content, chemical oxygen demand (COD), biochemical oxygen demand, and total suspended solids (TSS) impact the consumption of PAA in water and wastewater (Baldry et al. 1995; Stampi et al. 2002; Wagner et al. 2002; Dell'Erba et al. 2004; Rossi et al. 2007; Kunigk et al. 2012; Pedersen et al. 2013; Kunigk et al. 2014; Luukkonen et al. 2015; Henao et al. 2018a, 2018b).

Henao et al. (2018a) reported that the effect of organics such as proteins on the PAA demand and decay was pronounced within the first five minutes of contact time. In another study, the same authors reported that the PAA oxidative demand was a function of TSS and the decay rate correlated strongly with both the suspended and soluble matter (Henao et al. 2018b). Furthermore, given that 30–85% of suspended organic matter in wastewater has been associated with particles between 0.1 and 100 μm (Levine et al. 1985), it is expected that particle size may play a role in the PAA decomposition kinetics. This effect, however, has not been captured in previous literature. Falsanisi et al. (2008), for instance, reported that TSS particles larger than 120 μm and within the 10–120 μm range had the same PAA consumption (i.e. reduction in PAA concentration after 40 min).

Temperature is also a contributing factor to PAA decomposition. Baldry et al. (1995) and Stampi et al. (2001) reported that the disinfection activity of PAA in secondary effluent was improved in warmer temperatures. Kunigk et al. (2012) reported a positive correlation between increased temperature and PAA decomposition rate in synthetic samples containing organic matter. In two other studies, the same authors reported first order reaction kinetics for the decomposition of PAA in aqueous solutions containing organic compounds, within a temperature range of 25–45 °C and a pH range of 3–7 (Kunigk et al. 2001; Kunigk et al. 2014). However, the effect of temperature on the decomposition kinetics of PAA in wastewater has not been systematically studied.

In this paper, the kinetics of PAA decomposition in the presence of narrowly fractionated primary and secondary effluent suspended solids is investigated. To our knowledge, this is the first systematic study to compare the PAA decomposition kinetics of isolated effluent suspended solids size fractions. In addition, the impact of wastewater temperature on the kinetics of PAA decomposition and its potential implications on the required PAA dose to meet effluent quality targets in primary and secondary effluents are examined.

Primary effluent (PE), secondary effluent (SE), and mixed-liquor suspended solids (MLSS) samples were collected from a large municipal wastewater treatment plant in Toronto, Ontario, Canada, between December 2015 and February 2018. All samples were collected via the grab sampling method and tested within 48 h of collection and were refrigerated in between experiments.

In addition to the as-received wastewater samples, synthetic samples with a wide range of TSS and samples containing a narrow particle size distribution were prepared according to the following procedures.

To examine the effect of TSS concentration on PAA demand and decay, synthetic wastewater samples with varying TSS were prepared. For synthetic PE samples, 500 mL of the as-received PE was centrifuged at 3,000 rpm for 15 min. The centrifuged samples were then decanted, and the precipitated suspended solids were re-suspended in the supernatant at varying known amounts. For synthetic SE samples, first the TSS of the as-received SE sample as well as MLSS sample collected on the same day were measured using vacuum filtration following APHA standard method 2540D. Then, pre-determined amounts of the MLSS were added to a fixed total volume of the filtered SE to prepare synthetic samples with varying TSS levels. The TSS of these samples was then measured and recorded.

To study the effect of TSS particle size, sieving was used to prepare synthetic samples with isolated particle size fractions according to the procedure described elsewhere (Yuan & Farnood 2010; Azimi et al. 2012; Tan et al. 2017). In brief, for synthetic PE samples, 4 L of as-received PE was sieved successively through standard sieving trays (USA Standard Testing Sieve) with openings of 90, 75, 45, and 20 μm. Collected particles were gently washed with distilled water for 10 min to ensure that the obtained sample contained only particles within the desired size fraction. Each size fraction was then re-suspended in 250 mL of filtered PE. Filtered PE samples were obtained by passing the as-received PE through 1.2 μm Whatman^® glass fiber filters (WHA1822047). For synthetic SE samples, 500 mL of MLSS was sieved through the abovementioned sieving trays, washed thoroughly with distilled water, and the obtained size fractions were re-suspended in 250 mL of filtered SE. TSS of the samples was measured and adjusted to 20–24 mg/L in PE samples, and 1–2 mg/L in SE samples prior to testing.

The TSS of each sample was measured using vacuum filtration and 1.2 μm Whatman glass fiber filters following APHA standard method 2540D. COD was measured using a Hach COD reactor digestion kit (Method 8000). The pH and oxidation-reduction potential (ORP) measurements were obtained using a REED Instruments SD-230 pH/ORP datalogger (Newmarket, ON, Canada), and sample conductivity was measured using an OMEGA Engineering CDH221 meter (Laval, Quebec, Canada). The particle size distribution was measured using a Beckman Coulter MultiSizer 3 particle size analyzer equipped with a 280 μm aperture tube with dynamic range of 5.6–168 μm. NaCl solution with a concentration of 10.9 g/L was used as the electrolyte solution and the particle concentration was adjusted in the range of 5–10%.

The PAA stock solution of 22% wt (Peragreen 22WW) was obtained from EnviroTech PeraGreen Solutions (Modesto, CA. USA). For the PAA decomposition tests, a 250 mL aliquot of each prepared sample was spiked with a dosage of 5–8 mg/L of PAA for a total duration of 40 min at room temperature (20–25 °C). The residual PAA concentration was measured over time using the diethyl-p-phenylene diamine colorimetric method (APHA Standard Method 4500-Cl.G-2000) with a CHEMetrics^® PAA test kit (Rice et al. 2017). Control experiments were also conducted using Milli-Q water (18.2 MΩ.cm resistivity at 25 °C) to compare the decomposition behaviour of PAA in organic-free water. For the effect of temperature on PAA decomposition, PE and SE samples were kept in fixed-temperature water baths during the experiments to maintain a temperature of 10, 20 or 30 °C.

In this manuscript, all reported demand values are the numeric difference between the PAA spike concentration and the PAA concentration after 30 s, and all reported decay values were derived using data collected between 30 s and 40 min.

CT is a common measure for the effective disinfectant dosage. As described earlier, there are a number of reaction pathways that facilitate the gradual decay of available PAA (i.e spontaneous decomposition, hydrolysis, catalytic decomposition). Therefore, using the experimental demand and decay rate constant values to calculate CT will help with determining the delivered concentration of PAA available for disinfection.

Figure 1 presents typical demand and decay curves for PAA decomposition for as-received PE and SE samples at room temperature (20–25 °C). For PE samples, PAA decay was found to follow a second order rate law with respect to PAA concentration (R² = 0.98), while both first order and second order kinetics properly described the PAA consumption behaviour in SE samples with R² values of 0.95 and 0.96 respectively. In this manuscript, first order kinetics has been selected for the kinetic modelling of SE.

This result suggests that spontaneous decomposition (Equation (1)) in the presence of organics was likely the dominant mechanism for PAA decay in PE, while hydrolysis (Equation (2)) and/or catalytic decomposition (Equation (3)) – potentially due to residual metal ions from coagulant use in the secondary treatment process – were more important in the case of SE samples. Compared to wastewater samples, the decomposition of PAA in Milli-Q water began with a relatively small demand of 0.28 mg/L, followed by a slow first order decay profile (k_d1 = 0.0004 min⁻¹) over the course of 40 min.

PAA demand and decay rate constant as well as calculated CT for as-received SE and PE samples collected in this study are summarized in Tables 1 and 2, respectively. CT values were calculated using Equations (6) and (7) at a constant initial PAA concentration of 5 mg/L. For comparison, literature data for various types of effluents as well as tap water are also included in Table 1.

From the results, the average PAA demand for the as-received SE samples in the present work was 0.61 mg/L which was about 40% of that of the as-received PE samples (1.56 mg/L). This was likely due to the higher concentration of organic matter in the PE compared to the SE samples (see Table 3). In addition, the PAA demand for PE samples exhibited more variability (coefficient of variation = 75%) than that of SE samples (coefficient of variation = 30%), likely due to the greater daily variations in primary effluent quality.

PE samples exhibited a higher PAA decay rate than SE samples, once again likely due to their higher levels of organics. However, regression analysis showed that the PAA decay rate constants for the as-received PE and SE samples were strongly correlated with the PAA demand with a Pearson correlation of 0.69; i.e. about 47% of variations in k_d could be explained by variations in D. This finding suggests that while PAA demand and decay represented different mechanisms for PAA consumption, they were partly affected by similar water quality parameters.

Comparing the pairwise coefficients of correlation in Table 4 shows that 13–16% of variations in PAA demand and 18–26% of variations in decay rate constant for PE and SE samples could be explained by TSS.

In order to better understand the role of TSS in PAA decomposition kinetics, synthetic PE and SE samples with varying TSS levels were prepared, according to the procedure described earlier. Figure 2 shows that PAA demand on average increased by about 0.042 and 0.034 mg/L in PE and SE samples, respectively, for every 10 mg/L increase in TSS.

PAA decay rate constants for these samples also exhibited a strong dependency on TSS (Figure 3). The PAA decay rate constant in synthetic SE and PE samples increased on average by 0.0014 L/mg.min and 0.00039 min⁻¹, respectively, for every 10 mg/L increase in TSS. This result is consistent with the findings of Henao et al. (2018a) who reported that PAA decay rate increased by almost five times as TSS increased from 40 to 160 mg/L.

The PAA demand in filtered SE (∼0.32 mg/L) was 14.3% higher than the demand observed in Milli-Q water (0.28 mg/L). This increase may be attributed to the presence of dissolved organics in the secondary effluent compared to organic-free Milli-Q water. Another potential explanation is the existence of metal ions from coagulants introduced during the secondary treatment process, enhancing the catalytic decomposition of PAA upon contact with the SE sample.

It is important to note that the dependency of PAA demand and decay rate constant on TSS was effluent-specific. PAA demand and decay rate constants for filtered effluents plotted in Figures 2 and 3 accounted for 57–64% and 38–85% of those of as-received PE and SE samples (see Tables 1 and 2). This finding suggests that in addition to TSS, dissolved and colloidal substances play an important role in PAA decomposition kinetics.

In order to examine the overall effect of TSS on PAA dosing requirements, CT values for the synthetic samples at various TSS levels were calculated for an initial PAA concentration of 5 mg/L at 10, 30 and 60 min contact times using Equations (6) and (7). Figure 4 illustrates that for short contact times (i.e. 10 min), TSS had little effect on the calculated CT values. However, at high contact times (i.e. 60 min), the calculated CT values reduced by about 30% and 25% as the TSS level increased from 0 to 130 mg/L and from 0 to 60 mg/L for synthetic PE and SE samples, respectively. Note that TSS for municipal secondary effluents is typically between 1 and 10 mg/L, in which case the effect of TSS variations on CT can be estimated to be less than 8%. These results suggest that in secondary effluents, typical fluctuations in the TSS alone would have very little effect on PAA performance.

Figure 5 illustrates the size distribution of various PE and SE particle fractions prepared according to the procedure described earlier. Note that the particle size analyser used in this study produces measurements based on the solid volume of particles and does not account for particle porosity, which in turn results in a peak shift to the lower end of the size distribution (Yuan & Farnood 2010).

PAA decomposition of synthetic PE and SE samples with narrow particle size fractions showed that PAA demand and decay rate constant decreased as particle size increased (Table 5). However, this decrease was more prominent for PAA demand. These findings are reasonable since, at the same TSS, smaller particles offer a greater total surface area for reaction, hence increasing the decomposition rate of PAA. Our result, however, contradicts an earlier report by Falsanisi et al. (2008) where PAA consumption of a wastewater sample was not affected by filtration through 120 and 10 μm membranes.

Based on the kinetic parameters provided in Table 5, CT values for synthetic PE and SE samples with narrow size fractions were calculated using Equations (6) and (7). As seen in Figure 6, the calculated CT values were lower for the finer particle size fraction (20–45 μm) and increased with the particle size. Therefore, at the same TSS level, effluents containing smaller particle size fractions will have lower CT values, once again likely due to the higher specific surface area of smaller particles. Similar results were found for the SE samples (see Figure S1, Supplementary Data).

Results showed that raising temperature from 10 to 30 °C had a larger effect on the PAA decomposition kinetics of as-received PE compared to SE (Figure S3, Supplementary Data). As seen in Table 6, this is mainly due to the greater increase in the PAA demand of PE compared to that of SE (0.48 mg/L vs. 0.04 mg/L).

A comparison of activation energy and rate constant of PAA decay from the present work with those reported by Kunigk et al. is provided in Table 7. Kunigk and co-workers observed a lower activation energy when the water samples contained high organic content. The largest activation energy values reported by Kunigk et al. were associated with tap water and aqueous synthetic solutions, which is indicative of the potential effect of organic content on the decomposition of PAA in water.

Impact of temperature on CT values in PE samples is illustrated in Figure 7. In general, increasing effluent temperature resulted in a decrease in CT values. A similar trend was observed for SE samples (Figure S2, Supplementary Data); however, temperature had a greater effect on CT values in PE samples compared to SE. These findings illustrate the need to regulate the dosage of PAA according to the effluent temperature.

Results from this study show that PAA demand and decay rate constant in primary and secondary effluents depend on the TSS level as well as TSS particle size distribution. Examining the overall effect of TSS revealed that smaller particles appear to have a higher PAA demand and decay rate constant, likely due to their higher specific surface area that increases the rate of reaction with PAA. In other words, at a constant TSS, effluents with smaller particle size are expected to require a higher PAA demand. Therefore, better characterization of the source wastewater should be an important consideration in determining the optimal dosing requirement and equipment sizing for wastewater treatment using PAA.

Effluent temperature is another important consideration for PAA dosing. Given that temperature of primary and secondary effluent streams in municipal treatment facilities may vary from 5 to 30 °C (depending on the geographical location and extent of seasonal fluctuations), variability in CT due to changes in the effluent temperature could be significant. For example, in the case of the secondary effluent sample used in this study, increasing effluent temperature from 5 to 30 °C will result in an estimated 17% decrease in CT (initial PAA concentration: 5 mg/L, contact time: 60 min). The above findings suggest that the target PAA dosages should be adjusted by a safety factor to compensate for temperature variations in order to attain optimal dosing parameters and effective equipment sizing.

There are no conflicts to declare.

Financial support from Trojan Technologies is gratefully acknowledged. Authors would like to also thank Dr Yaldah Azimi for her assistance with fractionation experiments.

The Supplementary Material for this paper is available online at https://dx.doi.org/10.2166/wst.2020.047.