Sewage sludge processing systems in Scotland

3 Treatment impacts on hazards of interest

3.1 Overview

Modelling has identified eleven hazards of interest in biosolids:

1. Atenolol

2. Benzothiazole

3. Cyclomethicone 5

4. Cyclomethicone 6

5. Triclocarban

6. Nonylphenol

7. NP2EO

8. PBDE99

9. PBDE209

10. PCB118

11. PCB138

As discussed in Section 1.2.4, several techniques are applied to the treatment of sewage sludges in Scotland, principally:

1. Drying and pelletising / granulation

2. Thermal Hydrolysis and Anaerobic Digestion

3. Liming

4. Anaerobic Digestion

When applied adequately, the first three of these techniques are capable of producing biosolids that meet the pathogen reduction requirements of ‘Enhanced Treated’ biosolids (as defined by the Safe Sludge Matrix), whilst anaerobic digestion is capable of producing ‘Conventionally Treated’ biosolids. However, neither the Safe Sludge Matrix nor regulatory controls apply to organic compound contaminants, such as those listed above. In this section, we report the findings of a Rapid Evidence Assessment (REA) that examined whether the common sludge processing techniques are known to impact on these contaminants. Any reductive impact could reduce exposures below those modelled in Section 3 of the Human Health Risk Assessment Report, although further research would be necessary to understand whether any such reductions were significant – or sufficient to deem these hazards of no further interest.

3.1.1 Atenolol

This compound is partially broken down by biological activity during anaerobic digestion. No evidence could be identified to determine its fate during drying, thermal hydrolysis or liming.

Margot et al. (2015) collated data from multiple sources to illustrate the impact of various wastewater treatment techniques on hazards of interest. This review considered partitioning of hazards between aqueous and sludge phases, as well as genuine attenuation. Atenolol was considered to be broken down by biological activity (rather than sorption to the sludge or solid phase), although this activity was only sufficient to reducing incoming concentrations by 41%. D’Alessio et al. (2015) and Ke et al. (2014) cite higher removal efficiencies – as shown in Table 3-1.

Table 3-1 Removal efficiency of atenolol during conventional activated sludge treatment

Compound	Removal efficiency (%)	Reference
Atenolol	63	D'Alessio et al., 2015
	61	Ke et al., 2014
	41	Margot et al, 2015

Malmborg & Magnér (2015) examined the fate of a number of chemical compounds (both spiked and unspiked) in sewage sludges, through a variety of sludge treatments. These compounds included atenolol, while sludge treatments included thermal hydrolysis and anaerobic digestion. The former treatment had no significant impact on atenolol concentrations, whilst they were significantly reduced in the latter. This suggests that a combination of initial (aerobic) wastewater treatment, followed by anaerobic digestion of sludges would maximise opportunities for removal of atenolol. Further studies would be required to determine the presence and fate during anaerobic digestion of atenolol in Scottish sewage sludges.

3.1.2 Triclocarban

This compound is partially broken down by thermal hydrolysis. Anaerobic digestion has been found to have little impact, while no evidence is available to determine impact during drying. Liming appears to have little or no impact.

Overall removal of triclocarban (TCC) from waste water can be greater than 95% (Gardner et al., 2013), although this removal relies largely on partitioning to biosolids. In a synthesis of multiple sources, Margot (2015) suggests 10% biodegradation during wastewater treatment, with 90% sorption to sludges. Ogunyoku & Young, 2014 indicate that TCC is partly degraded by biological action under aerobic conditions, whole Blair et al. (2015) state a removal efficiency of 11% during conventional activated sludge treatment.

Armstrong et al. (2017) examined the influence of sludge treatment via THP-AD (thermal hydrolysis with anaerobic digestion) on concentrations of triclosan (TCS) and triclocarban (TCC), together with several of their transformation products. The study focussed on sludge treatment centre in the USA that had recently changed its sludge treatment from liming to THP-AD. They found that levels of TCC significantly decreased during thermal hydrolysis. Average concentrations of TCC prior to THP ranged from 6816 to 7368 ngg^-1 dry weight (dw) while concentrations measured after THP treatment ranged from 67.5 to 89.9 ngg^-1 dw. The degradation pathways were not determined. In contrast, concentrations of TCS, methyl triclosan, and 2,4-dichlorophenol increased during anaerobic digestion (Figure 3-1).

Overall, concentrations of TCC in THP-AD biosolids (Class A) was significantly lower than that of limed biosolids (Class B) (P < 0.01). The authors noted that, during the THP-AD start-up period, TCC concentrations in the resulting biosolids continuously decreased until the process had stabilized and been fully commissioned (February 2015), after which concentrations remained relatively steady with a range of 102-294 ngg^-1 dw (Figure 3-2).

Heidler et al. (2006) undertook a sampling campaign at a ‘typical’ U.S. wastewater treatment plant sized to treat 680 million litres of wastewater a day. A mixture of primary and secondary sludges were subjected to anaerobic digestion at between 35 and 37°C for an average period of 19 days. Samples were collected from influent, final effluent (before and after polishing in a sand filter) and de-watered digested sludge (biosolids). Overall, the authors concluded that anaerobic digestion for 19 days did not promote TCC transformation, resulting in its accumulation in biosolids to concentrations of 51 ± 15 mg/kg dry weight. 76 ± 30% of the compound entering the plant underwent no net transformation and instead partitioned into and accumulated in the final biosolids.

Ogunyoku and Young (2014) examined TCC concentrations in sludges and (resulting) biosolids following a range of treatments that included sludge liming – but were unable to reach any clear conclusions over the impact of this practice on the compound of interest.

3.1.3 Benzothiazole

Published literature suggests that partitioning of this compound to sludges is not an important component of its removal from wastewater, with significant proportions instead removed by biological action during activated sludge treatment. No evidence was identified for the fate of this compound in sewage sludges subjected to the different treatments of interest.

While benzothiazoles (BTHs) are used in a variety of drugs, it is unlikely that these are the primary source in wastewater. BTHs and their derivatives are high-production-volume industrial chemicals – their main use being the vulcanisation of rubber – although they are also used as corrosion inhibitors in antifreeze and cooling liquids, in wood preservation and in industrial processes (Margot, 2015). Given the possible sources, it is thought that BTHs in wastewater mainly originate from urban runoff (tyre abrasion on roads) and household sources (e.g. washing clothes). In addition to BTH, the main BTH derivatives that have been measured in wastewater and sludge are benzothiazole-2-sulfonic acid (BTSA), 2-OHBTH, 2-methylthiobenzothiazole (2-MeBTH) and 2-amino-benzothiazole (2-ABTH).

Removal rates (concentration in influent compared with concentration in effluent) for BTHs have been reported as 50-90% for BTH, 40-66% for 2-OHBTH and 50-80% for 2-MeSBTH (Asimakopoulos et al., 2013; Stasinakis et al., 2013; Karthikraj & Kannan, 2017).

Asimakopoulos et al. (2013) reported that removal from wastewater due to sorption onto biosolids and accumulation in sludge was insignificant for BTH, 2-OHBTH, 2-MeBTH and 2-ABTH. Mass balance analysis suggested that a major portion of BTHs was instead lost or bio-transformed during activated sludge treatment processes.

The predicted boiling point of benzothiazole is 227.0±9.0°C (www.chemspider.com), which could render it susceptible to loss during sludge drying (Section 1.2.2), although no evidence was found to support this suggestion. Further studies would be required to determine the presence and fate during drying, liming, thermal hydrolysis or anaerobic digestion of benzothiazole in Scottish sewage sludges.

3.1.4 PBDE 99 & 209

These compounds are partially broken down by anaerobic digestion, with impacts on BDE 209 greater than for BDE 99. No evidence is available to determine impact during thermal hydrolysis or liming. Thermal treatment has been demonstrated to have no impact on BDE 209.

Commercial penta-BDE and commercial octa-BDE were added to Annex A of the Stockholm Convention in May 2009 , while commercial deca-BDE was added in May 2017. A recent review of PBDE concentrations noted a downward trend in sludge concentrations which is likely to be a reflection of the on-going efforts to phase-out PBDEs (Kim et al., 2017).

PBDEs enter wastewater treatment plants from contaminated indoor dust, leachate from landfilled PBDE-containing products, and discharge from industrial sites processing PBDE-containing material. Mass balance calculations suggest that 96% of PBDEs are sorbed to sewage sludge (North, 2004).

Shin et al. (2010) investigated the fate of common PBDE congeners (BDE 47, 99, 100, 138, 153, 154, 183 and 209) in sewage sludges under anaerobic conditions (at 37°C). In batch cultures over a period of 238 days, the concentrations of BDE 47, 99, 100 and 209 decreased significantly (by 22–40% from their initial concentration), while concentrations of the other congeners (BDE 138, 153, 154 and 183) remained stable. However, in a parallel study conducted in a pilot-scale anaerobic digester, loss of all eight congeners was observed. The pilot digesters each had a working volume of 535 litres, and their contents were subjected to continuous mixing. The impacts for two different hydraulic retention times were explored: 15 and 30 days (referred to as PSR1 and PSR2, respectively). Data from PSR2 were not presented by the authors, since the digesters did not reach a steady operating state during the experiments. Data from PSR1 are presented in Figure 3-3.

These data show that concentrations of highly-brominated congeners (such as BDE 138, 183, and 209) were significantly lower in the digested sludge when compared with the feed, with reductions ranging from 30.7 to 64.4% by mass. Reductions in the less brominated congeners (BDS 47, 99 and 100) were rather less, at 21.4 and 24.0%.

Mailler et al. (2014) examined the fate of a range of micropollutants at three different wastewater treatment works in Paris, all with different treatment processes. These included thermal drying (at 260°C) and anaerobic digestion. Thermally-dried samples were collected on six different occasions over a period of three months, whilst digested samples were collected on each day over a period of three days. There is a large spread in the resulting data – ascribed to low concentrations and analytical variance, but overall the authors concluded that thermal drying had little to no impact on PBDE 209, and variable impact on nonylphenol and related compounds (Figure 3-4) (see Section 3.1.7 for further discussion of this impact). BDE 209 was significantly removed (>50%) by anaerobic digestion (Figure 3-5).

Further studies would be required to determine the presence and fate during drying, liming, thermal hydrolysis or anaerobic digestion of BDE 99 and 209 in Scottish sewage sludges.

3.1.5 PCB 118 & 138

These compounds are partially degraded during anaerobic digestion. No evidence was found that considered their fate during drying, thermal hydrolysis or liming.

Patureau & Trably (2006) examined anaerobic (and aerobic) biodegradation of six PCBs in continuous stirred anaerobic digesters fed with naturally contaminated sewage sludge. PCB removals were about 20% irrespective of PCB molecular weight or degree of chlorination (Figure 3-6). This corresponds well with later data from Siebielska & Sidełko (2015).

Siebielska & Sidełko (2015) examined changes in PCB concentrations in a mixture of sewage sludge and the organic fraction of municipal waste during composting and during anaerobic digestion at laboratory scale. They concluded that anaerobic digestion was much more effective than composting at removing PCBs from a mixture of sewage sludge and the organic fraction of municipal waste (Table 3-2). Anaerobic digestion was undertaken in 50litre reactors at 39°C over 21 days. Composting was undertaken in 60litre reactors over cycles of 182 days. Concentrations of the more highly-chlorinated PCBs were less impacted by composting than the less chlorinated PCBs. The chlorination level of the PCB congener did not affect the decrease observed during anaerobic digestion.

Table 3-2 Decreases in PCB concentrations during composting and AD. From Siebielska & Sidełko (2015).

PCB congener	Average decrease in concentration during composting (%)	Average decrease in concentration during AD (%)	F-statistic value
PCB 28	15.7	18.4	5.70
PCB 52	11.8	18.1	9.38
PCB 101	8.4	20.0	7.83
PCB 118	9.6	18.5	8.40
PCB 138	10.5	20.0	7.09
PCB 153	9.0	23.9	31.28
PCB 180	6.9	21.1	14.25

Further studies would be required to determine the presence and fate during drying, liming, thermal hydrolysis or anaerobic digestion of PCB 118 and 138 in Scottish sewage sludges.

3.1.6 Cyclomethicone 5 & 6

Cyclomethicone 5 is readily volatilised during aerobic and anaerobic treatments, while cyclomethicone 6 is thought to remain in sludges. The impacts on these compounds of drying, thermal hydrolysis or liming are unknown.

Cyclomethicone 5 & 6 are volatile silicon compounds (siloxanes), normally referred to as D5 (decamethylcyclopentasiloxane) and D6 (dodecamethylcyclohexasiloxane).

The presence of volatile silicon compounds in biogas is problematic, since the combustion of contaminated biogas can result in crystallisation of silica within engines – causing significant damage and eventual failure. Activated carbon filters are normally used to scrub siloxanes from biogas before combustion. García et al. (2015) reviewed published data in the presence of siloxanes in biogas from anaerobic digestion of sewage sludges and reported that 93.7% of siloxanes in wastewater are D4 or D5. A significant proportion of these compounds was lost through volatilization during aerobic treatment (58.6%), with the remainder retained in sludge – a proportion of which is then volatilised during AD. D6 was stated to remain in the sludge.

This same finding was reported by Appels et al. (2008), who provide an overview of the occurrence and fate of various siloxane compounds during wastewater treatment. They report that the cyclic siloxanes D4 (octamethylcyclotetrasiloxane) and D5 are detected in significant amounts in biogas from anaerobic digestion of sewage sludges, while larger molecules such as D6 do not volatilise during digestion, remaining in the sludge.

Xu et al. (2013) present a contradictory picture, having examined the occurrence and fate of four cyclic and two linear volatile siloxanes in a municipal wastewater treatment plant in Beijing. Through in vitro biodegradation experiments, they concluded that D6 was eliminated mostly by volatilization, while D5 was eliminated by a combination of volatilization and degradation. However, de Arespacochaga et al. (2015) state that biodegradation of siloxanes is thought to play a minor role in their loss during sludge treatment.

With boiling points of 211°C (D5) and 245°C (D6) (de Arespacochaga et al., 2015), it is possible that both compounds would be lost during sludge drying – although no evidence was found to confirm this. Further studies would be required to determine the presence and fate during drying, liming, thermal hydrolysis or anaerobic digestion of cyclomethicone 5 & 6 in Scottish sewage sludges.

3.1.7 Nonylphenol (NP) and NP2EO

Evidence for removal of NP and NP2EO during anaerobic digestion are contradictory, due to potential for transformation of NP2EO to NP. Thermal hydrolysis has no impact on removal of these compounds, although may reduce the potential for biological transformation of NP2EO to NP. No evidence could be found for impacts of liming on these compounds, which may be partly or completely removed by drying.

Published data on the impacts of sludge treatment process on NP are extremely variable. Stasinakis (2012) reviewed publications exploring the fates of emerging contaminants during sludge processing. Data for reductions of NP and related compounds during anaerobic digestion are summarised in Table 3-3.

Table 3-3 Removal of NP, NP1EO and NP2EO during mesophilic anaerobic digestion of different sewage sludges

Compound	% removal during AD	Solids Retention Time (days)	Type of sludge	Reference
NP	0	30	Primary	Paterakis et al., 2012
NP	100	30	Mixed primary and activated	Paterakis et al., 2012
NP1EO	3.76	20	Activated	Hernandez-Raquet et al., 2007
NP2EO	2.63	20	Activated	Hernandez-Raquet et al., 2007

Mailler et al. (2014) examined the fate of a range of micropollutants at three different wastewater treatment works in Paris, all with different treatment processes. These included thermal drying (at 260°C) and anaerobic digestion. Thermally-dried samples were collected on six different occasions over a period of three months, whilst digested samples were collected on each day over a period of three days. Overall, the authors concluded that thermal drying had variable impact on nonylphenol and related compounds (Figure 3-4). NP compounds (nonylphenol monoethoxylate (NP1EO) and nonylphenol diethoxylate (NP2EO)) were significantly removed during AD (>50%) while NP was moderately removed (40%) by this process.

This is contrast to McNamara et al. (2012), who compared the impacts of thermal hydrolysis and mesophilic anaerobic digestion on NPEs (a collective term for nonylphenol monoethoxylate (NP1EO), nonylphenol diethoxylate (NP2EO) and nonylphenol (NP)), as compared with conventional MAD and aerobic digestion. Three thermal hydrolysis-mesophilic anaerobic digestion (TH-MAD) reactors were operated:

1. Reactor one (TH150-MAD) received sludge that underwent thermal hydrolysis at 150°C followed by MAD with an SRT of 15 days;

2. Reactor two (TH170-MAD) received sludge that underwent thermal hydrolysis at 170°C followed by MAD with an SRT of 15 days;

3. Reactor three (TH150-MAD20) received sludge that underwent thermal hydrolysis at 150°C followed by MAD with an SRT of 20 days.

A conventional mesophilic anaerobic digester (MAD) was fed untreated sludge and served as a control.

Following THP-AD, an aerobic/anoxic reactor was operated under two different conditions:

1. Aerobic and anoxic phases were alternated every 20 min. The influent and effluent to this reactor are called TH-M-AER 20/20 Inf and TH-M-AER 20/20 Eff, respectively.

2. Aerobic and anoxic phases were alternated every 12 min. The influent and effluent to this reactor are called TH-M-AER 12/12 Inf and TH-M-AER 12/12 Eff, respectively.

In all anaerobic reactors, with one exception (TH150-MAD20), the total masses of NPE in the influent and effluent were within 10% of each other. The authors interpret this as a demonstration that NPEO was generally transformed to NP, and that substantial loss of NP, NP1EO and NP2EO did not occur during AD, irrespective of prior thermal hydrolysis. Data are presented in Overall, thermal hydrolysis did not reduce NPE loadings in the digested sludges, but did influence the potential for transformation from NPEO to NP – the authors speculate that this may be due to thermal hydrolysis rendering NPEO less bioavailable, and therefore less susceptible to biologically-mediated degradation. However, an aerobic step after digestion did reduce NPE loadings in digested sludges – whether they had first been treated by TH or not.

Figure 3-7.

Overall, thermal hydrolysis did not reduce NPE loadings in the digested sludges, but did influence the potential for transformation from NPEO to NP – the authors speculate that this may be due to thermal hydrolysis rendering NPEO less bioavailable, and therefore less susceptible to biologically-mediated degradation. However, an aerobic step after digestion did reduce NPE loadings in digested sludges – whether they had first been treated by TH or not.

Transformations were also explored by Paterakis et al. (2012). Laboratory scale anaerobic digesters (1.5 litre working volume) were operated in duplicate, with an hydraulic retention time (HRT) of 30 days d at 35ºC. Concentrations of NPEs (including NPECs (nonylphenol ethoxycarboxylates)) were measured at the beginning and end of six retention times for different types of sludge (primary and mixed sludge, the latter comprising primary and waste activated sludge at a ratio of 60% (v/v) primary and 40% (v/v) WAS). Overall, greater removal of ΣNPEOs was observed for the mixed sludge >50% in comparison to primary sludge. However, results were inconsistent, as illustrated in Figure 3-8 and Figure 3-9.

Further studies would be required to determine the presence and fate during drying, liming, thermal hydrolysis or anaerobic digestion of NP and NP2EO in Scottish sewage sludges.