Bisphenol A in surface waters in Germany: Part I. Reassessment of sources and emissions pathways for FlowEQ modeling

Bisphenol A (BPA) enters the environment through various industrial and consumer‐related pathways. Industrial sources include BPA manufacturing and secondary industrial uses such as the manufacturing of polymers and other substances based on or containing BPA. However, secondary sources and emissions to the environment, such as those related to the consumer use of articles containing BPA, may be more important than industrial emissions. Although readily biodegradable, BPA is widely distributed in various environmental compartments and living organisms. It is still not well understood which specific sources and pathways are responsible for releasing BPA into the environment. Therefore, we developed FlowEQ, a coupled flow network and fugacity‐based fate and transport model for the assessment of BPA in surface water. The work is divided into two parts. In Part I, inputs needed to support the modeling and model validation were collected. Bisphenol A was measured at 23 wastewater treatment plants (WWTPs) and 21 landfills in Germany. In addition, the BPA content of 132 consumer articles from 27 article classes was analyzed. Bisphenol A concentrations in WWTPs ranged from 0.33 to 910 µg L−1 in influents and from less than 0.01 to 0.65 µg L−1 in effluents, resulting in removal efficiencies of 13%–100%. Average BPA concentrations in landfill leachate ranged from less than 0.01 to approximately 1400 µg L−1. Bisphenol A concentrations measured in consumer articles varied significantly by type, ranging from less than 0.5 µg kg−1 in printing inks up to 1 691 700 µg kg−1 in articles made from recycled polyvinyl chloride (PVC). These concentrations were combined with information on use, leaching, and contact with water to develop estimates of loadings. Together with the results of the FlowEQ modeling presented in Part II, this assessment improves our understanding of the sources and emission pathways of BPA in surface water. The model considers various sources of BPA and can estimate future surface water concentrations of BPA based on changes in use. Integr Environ Assess Manag 2024;20:211–225. © 2023 The Authors. Integrated Environmental Assessment and Management published by Wiley Periodicals LLC on behalf of Society of Environmental Toxicology & Chemistry (SETAC).


INTRODUCTION
Substances occurring in consumer products, such as bisphenol A (BPA), can enter the environment during manufacturing processes or through the use and disposal of products.They can be released via various pathways such as manufacturing outfalls, domestic wastewater, and others.Depending on their biodegradability potential and physicochemical properties, they are either degraded or transformed in aquatic environments or will partition in environmental compartments.Many of these substances are infrequently characterized in aquatic environments due to inadequate or expensive analytical methods or because they are not included in long-term monitoring programs.Furthermore, even a robust monitoring dataset provides only information on the presence of a substance but not on the origin and its contribution to the total concentration of the substance in the environment.
Bisphenol A is an industrial chemical used for the production of numerous products and consumer articles.Depending on the application, it can be used to produce polymers (e.g., polycarbonate [PC], epoxy resins [EpRs]), it can act as an additive to control chain length, facilitate compounding, or it can be used as an antioxidant (polyvinyl chloride [PVC], rubber) or color developer (thermal paper).Despite the demonstrated ready biodegradability (Kleĉka et al., 2001;Ying & Kookana, 2003), BPA is found (albeit mostly at low levels) in various environmental compartments and organisms, for example, in surface water and sediments as well as in plants and animals (Staples et al., 2018).The specific sources and pathways responsible for BPA concentrations in the environment are not well understood.
For decades BPA has been under regulatory and scientific scrutiny because of its high production volumes, ubiquity in the environment, and indications of weak estrogenic activity.Recently, the focus of regulation in Europe has been on environmental concerns.As a consequence of a Registration, Evaluation, Authorisation, and Restriction of Chemicals (REACH) substance evaluation conducted between 2012 and 2016, the German authorities have been preparing a restriction proposal under REACH with the aim of lowering environmental releases of BPA.The proposal was submitted to the European Chemicals Agency (ECHA) in October 2022.Another restriction under REACH, which has been in force since December 2020, limits BPA concentrations in thermal paper to 0.02% (w/w).Although originally aimed at protecting workers, it is expected to also affect environmental releases of BPA (ECHA, 2022).
Fate and transport models are increasingly used to simulate the complex behavior and occurrence of chemical substances in the environment.Multimedia fate and transport models, such as the FlowEQ described by Bock et al. (2023), provide a mechanism by which known chemical sources can be used to predict the concentrations of these chemicals in the environment.The results of these models can be verified by comparison with measured concentrations.This comparison can help determine if the important sources and sinks are identified based on the ability of the model to accurately predict surface water concentrations.If these predictions are sufficiently accurate, the model can be used to estimate concentrations in environmental areas that lack reliable measurement data or predict how changes in manufacturing, use, and disposal may affect environmental concentrations.FlowEQ is a hybrid fate and transport model that combines the flow network of the MoRE/MONERIS model (UBA, 2010) with a Level III fugacitybased model.It is composed of 46 units of 1500-55 000 km² corresponding to coordination areas from the MoRE/ MONERIS.The model considers emissions (loading to soil, water, and air), advection (transport of surface water in a flow network), degradation, and multimedia exchange (see Figure 1).The model domain includes German watersheds and the upstream drainage areas (Bock et al., 2023).
An earlier version of this model was previously used for the assessment of BPA in the environment (Ramboll Environ/ BiPRO, 2015).In these initial studies, best estimate emissions revealed total loadings of approximately 9 t a −1 into surface waters of the model domain.These loadings were superseded by an updated assessment as part of this work, although these results informed the relative importance of different sources.The model results indicated that emissions from municipal wastewater dominate in virtually all geographical areas compared with the contribution from industry and manufacturing processes.However, specific sources could not be identified using this iteration of the model.A follow-up study specifically addressed the contribution of PC and EpR-based articles considering updated market segment data and BPA release rates for interior and exterior use scenarios (Ramboll, 2017).FlowEQ results based on experimental migration data suggested that PC and EpR-based consumer articles contribute to approximately 1% of total BPA loadings to surface water during their use and after disposal (Ramboll, 2017).However, most (99%) of the annual BPA emissions and the subsequent concentrations in surface water could not be attributed to specific sources.Primary and secondary environmental inputs of BPA in consumer articles must be further elucidated to refine the FlowEQ model in a way that allows more useful predictions concerning the importance of specific sources, particularly during use and after disposal and discharges via municipal wastewater, or as input from landfill leachates.
The current article and Bock et al. (2023) are intended to address the uncertainties identified above and to determine the sources of BPA in surface waters more precisely.This article serves to provide the data needed for more realistic and precise modeling in Bock et al. (2023).Due to the lack of sufficient monitoring of BPA in wastewater and landfill leachate, BPA measurements at 23 wastewater treatment plants (WWTPs; influent/effluent, sewage sludge) and 21 landfills (leachate) in Germany were performed.In addition, 132 consumer articles covering 27 article classes were analyzed for their BPA content to provide a more robust basis for estimating BPA content and releases from consumer articles.By addressing these data gaps, it became possible to create a more robust data basis for the FlowEQ model in Bock et al. (2023).

Wastewater treatment plants
Representative WWTPs for this measurement campaign were selected to reflect the variety of plants in Germany (i.e., connected industries, different sizes, treatment types, uses).In total, 23 German WWTPs participated in the sampling campaign and provided information on technical properties.These plants either treat the wastewater of only municipal or both municipal and industrial sources.The participating WWTPs run similar processes comprising, at a minimum, both mechanical and biological treatments.The average total volume flow of treated wastewater ranged from 3 to 432 000 m 3 day −1 .Resulting sludge volumes in dry weight ranged from 2 to 87 600 t a −1 .The sludge of the WWTPs investigated is used in agriculture and landscaping or is incinerated.The detailed characteristics of the participating WWTPs are shown in Supporting Information: Table SI-1.Bisphenol A samples were collected in 2018.After an EU-wide restriction of BPA in thermal paper as of January 2020, it is prohibited to place thermal paper that contains ≥0.02% (w/w) BPA on the market in the EU.To investigate the ramifications of these restrictions, measurements at selected WWTPs connected to the paper processing industry and at reference WWTPs (with comparable sizes and treatment) were also conducted two and four years after the first measurements (2020 and 2022).

Landfills
Twenty-one German landfills from North Rhine-Westphalia, Bavaria, and Baden-Württemberg agreed to participate in the study and were sampled in 2018, 2020, and 2022.In most cases, active (hereinafter referred to as cells) and inactive areas existed on-site, with operating years beginning as early as the 1950s.The total waste volume disposed ranged up to 9 Mio m 3 per site or part of the site.Types of waste included municipal solid waste (MSW) and waste from landfill Classes I and II (I = low-pollutant and largely mineralized waste, with a low organic content; II = larger pollutant load, which also has a greater biological content than those in landfill Class I).The leachate volumes ranged from 400 to 60 000 m 3 a −1 .Leachates were discharged either directly or indirectly (via WWTPs) to the surface water after potential pretreatment, if available, at the landfill.Figure 2 provides an overview of the sampling locations and timing (e.g., before or after treatment) where the leachate samples were taken.
Measured leachate samples were grouped into five different categories before analysis.Category 1 included measurements from leachate that is treated preliminarily onsite followed by subsequent external WWTP treatment (offsite).This category was further subcategorized for samples taken either before (1a) or after (1b) on-site pretreatment.Category 2 measures leachate samples that are not treated on-site but are treated by external WWTPs (off-site) before discharge to surface waters.As only on-site sampling was possible, the effluents of respective off-site WWTPs could not be measured.Leachate from Categories 3 and 4 is discharged directly to surface water but Category 3 undergoes on-site pretreatment before discharge, and sampling takes place after this process; Category 4 remains untreated.No information was provided on the treatment of leachate from Category 5.Because one landfill can have several areas with different treatment processes, more than one category can occur at one landfill.In cases where one landfill had several areas of the same category (e.g., Landfill 3 had three areas from Category 1a), mean values for the landfills were calculated.The characteristics and measured BPA concentrations from leachate of the participating landfills are shown in Supporting Information: Table SI-2.

Articles
The term "article" as it is used in this work follows the definition of the REACH legislation.It is defined as "an object which during production is given a special shape, surface, or design which determines its function to a greater degree than its chemical composition" (European Parliament and Council of the European Union, 2006).The aim of the campaign was to identify articles that could contribute significantly to BPA environmental emissions.It should be emphasized that only the content of BPA in the articles was measured, and no leaching studies were done.The articles were also assessed based on market analysis, contact with water, material, and expected BPA content.Environmental releases of BPA had previously been estimated in the EU-Risk Assessment report on BPA (European Commission, 2010a), which served as one important information source to select relevant articles for this measurement campaign.Therein, it has been demonstrated that the use of BPA in thermal paper (via paper recycling) as well as the use of BPA in various stages of PVC manufacturing and use were previously important uses with large environmental emissions of BPA.Environmental releases from articles made of or containing PC or EpR have already been assessed in previous studies (Ramboll, 2015(Ramboll, , 2017)).It was concluded that both are minor contributors to the estimated loading of BPA to surface water (ca.1% of the total).Details regarding the original methodology, refinements, and updated results are provided in the Supporting Information: Attachment 1 and Table  No new measurements of articles from PC and EpR were performed in this measurement campaign.

Sampling and analytics
Wastewater treatment plant influent and effluent grab samples were collected every 3 h over a 9-h period and later combined into composites.Bisphenol A was identified in the pooled composites within 48 h after sampling (DIN 38407-27: 2012-10).The limit of quantitation (LoQ) for BPA in the aqueous phase was 0.01 μg L −1 .For each sample, an analytical determination was run in duplicates.Sludge samples were collected after the last processing step after dewatering, and BPA was identified within 48 h of sampling (DIN ISO 14154, 2005-12/ISO 14154:2005).The analysis of landfill leachate was conducted in the same way as for the WWTP samples.However, these were separated first between filterable matter and the aqueous fraction according to the German Technical standard DIN 38409-2:1987-03.
For the analysis of all articles, methanol was added to the crushed sample, and extraction was conducted at room temperature, applying ultrasonic treatment in an ultrasonic bath.Subsequently, samples were centrifuged, and an aliquot of the supernatant was analyzed by LC-MS/MS.Quantification was done using an isotopically labeled standard.For each sample, sample preparation and analytical determination was run in duplicates, and mean values for replicates were calculated.The LoQ of the applied method was 0.5 μg kg −1 .Additional details are provided in the Supporting Information under Sampling and analytics.

Calculations
Wastewater treatment plant BPA removal efficiencies were calculated by comparing BPA influent and effluent concentrations: FIGURE 2 Overview of the landfill sampling scheme and derived classification.Arising leachate in landfills or parts of landfills can be pretreated, discharged directly into surface water, or treated externally by wastewater treatment plants (WWTPs) or others.Based on the fate of the leachate, measured bisphenol A (BPA) values were assigned to five categories.Category 1 was further subcategorized to reflect the amount of BPA before (1a) and after (1b) on-site pretreatment The term "loadings" refers to the amount of BPA that enters (influent) or leaves (effluent) the WWTP per year or that is deposited in the sewage sludge.However, only the loadings of the effluent come into contact with surface water and the part of the sewage sludge that is applied to soil.Annual BPA loadings arising in each WWTP were calculated on the basis of BPA concentrations and annual average volume flow: or sewage sludge volume: In the context of landfills, the term "loadings" refers to the amount of BPA that is deposited in the landfill leachate in one year.As described above, one landfill can have several areas with different types of waste and treatment.The operators indicated the volume flow rate per landfill area.Annual BPA loadings were calculated on a basis of BPA concentrations in the leachate sampled in the area and the annual average leachate volume of the area as described above.
Bisphenol A loadings to surface water and wastewater from the consumer use of articles were calculated using concentrations in articles, annual article use, fraction of BPA leached to water, and the population of Germany.The approach to each article class is presented in Supporting Information: Additional details for the derivations of loadings to surface water and wastewater from consumer use and Supporting Information: Tables SI-5-SI-10.

Bisphenol A concentrations in WWTPs
Bisphenol A concentrations were measured in WWTP influents, effluents, and sewage sludge from March to May 2018.Based on the measurements, the BPA removal efficiency of these WWTPs was calculated.Figure 3 gives a summary of the measured BPA concentrations.Bisphenol A concentrations in the influent varied from 0.33 to 910 μg L −1 with a mean of 41 μg L −1 .The measured effluent concentrations ranged from below the LoQ to 0.65 μg L −1 (mean 0.16 μg L −1 ).Bisphenol A concentrations in sewage sludge were from 0.068 to 200 mg kg −1 dry matter with a mean of 9.2 mg kg −1 dry matter.Table 1 shows mean influent and effluent BPA concentrations and corresponding removal efficiencies for the aqueous phase.
The influent values of WWTP Nos. 4 and 8 were elevated compared with the other measured influents (Figure 3; Table 1).Wastewater treatment plant No. 8 is fed by discharges from several paper processing plants including two large paper recyclers, manufacturers of hygiene and baking paper, and manufacturers of thermal paper (together >60% of the wastewater).Similarly, more than 50% of the FIGURE 3 Mean bisphenol A (BPA) concentrations in wastewater treatment plant (WWTP) influent, effluent, and sewage sludge.Influent and effluent are presented as μg L −1 and sewage sludge as mg kg −1 dry matter wastewater at WWTP No. 4 originates in a paper processing plant.The paper processing industry may have discharged larger BPA loadings resulting from the manufacture of thermal paper containing BPA as an additive in its free, unbound form.Additionally, recycling of wastepaper comprising BPA-based thermal paper may also contribute BPA.Finally, handling of recycled paper that contains unintentional BPA might also lead to BPA loadings into paper processing industry wastewaters.Most WWTPs exhibited removal efficiencies greater than 90% (Table 1), which is consistent with literature values (Clara et al., 2005;Fürhacker et al., 2000;Jonkers et al., 2009;Musolff et al., 2010;Tran et al., 2015).In cases where significantly lower removal rates were found (WWTP Nos.5: 65%, 9: 13%, 10: 63%, 11: 57%), operators were contacted to investigate possible explanations.The operator of WWTP No. 5 reported three possible reasons for low BPA removal efficiencies: low water temperature, use of aluminum salts to counteract contamination with a detrimental bacterium (Microthrix parvicella), and the installation of a new equipment in the aeration tank.All of these factors could have led to decreased microbial activity and limited mineralization of BPA.The operators of WWTP Nos. 9, 10, and 11 could not explain the observed low removal efficiencies.The operator of WWTP No. 9 reported greater removal efficiencies for BPA (85% and 97%) during an earlier internal measurement campaign in 2015.A potential reason for the observed extraordinarily low removal rates could be that influent and effluent samples taken represented different water masses.As described above, influent and effluent were sampled at the same time but the residence time of wastewater in these plants varies from hours to days.Investigations revealed heavy rainfall in the days prior to sampling in the area upstream of WWTP No. 9 that could have affected calculated removal efficiencies.Rapid changes in water loadings can result in a decoupling of the water masses, meaning that the effluent water sample could have had a different mix of wastewater and stormwater than the influent sample.A dilution of influent without an equal dilution of effluent would lead to inaccurate estimates of removal efficiencies.Additionally, heavy rain could overwhelm WWTP capacity, leading to bypass events and therefore lower removal efficiencies.It is likely that the heavy rainfall influenced the calculated removal efficiency, although it is not possible to determine if differential dilution effects or a bypass event, or both, were the reason for the values observed.The mean removal efficiency of all WWTPs was 86%.After exclusion of the anomalous results for No. 9, mean removal efficiency was calculated at 89%.For further analysis, the data were divided based on the dominant influent source of the WWTP.As stated above, the two WWTPs, Nos. 4 and 8, which had substantially higher concentrations of BPA in influents than other facilities, could be explained by their influents being dominated by paper-industry wastewater (>50% and >60%, respectively).The removal efficiency of these two paperindustry dominated WWTPs was 99%.We hypothesize that the greater removal efficiency in paper-industry dominated WWTPs is the result of acclimation of the microbial community to BPA as an energy and carbon source.
Bisphenol A concentrations in German WWTPs are not routinely measured.However, data are available for 28 WWTPs in Saxony where effluents were sampled and analyzed from 2001 to 2007 (Engelmann, 2008).Although the maximum value found in the 2008 measurement campaign was higher (5.3 μg L −1 ), the highest 90-percentile of 1.2 μg L −1 in the same measurement campaign suggests that the overall BPA concentrations may be comparable to the results of the present study.However, Engelmann calculated WWTP effluent values without distinguishing between types of WWTPs or wastewater.In another study, effluents of nine municipal WWTPs in Hesse were sampled and analyzed for BPA in 1999 and 2000 (Leisewitz, 2003).Although BPA was not detectable in four WWTP samples (<0.05 μg L −1 ), in the remaining five WWTP effluent samples, the reported BPA concentrations were between 0.051 and 0.37 μg L −1 and thus, had a range similar to that found in the present study.In a study conducted in 2010, BPA concentrations in the influent of WWTP were 0.03-21 μg L −1 and, in the effluent, up to 0.9 μg L −1 (Musolff et al., 2010).In the present study, BPA loadings for each WWTP were calculated based on the measured BPA concentrations and volume flow.Bisphenol A loadings for the influent were between 0.00036 and 18 000 kg a −1 (kg per annum), for the effluent between 0.00013 and 24 kg a −1 , and for sewage sludge between 0.00016 and 1700 kg a −1 .
For the plants that provide sewage sludge for use on agricultural fields or in compost plants, the highest BPA loading was 1.4 kg a −1 .

Trends in BPA concentrations in WWTPs
For two WWTPs connected to the paper processing industry (WWTP Nos. 4 and 8) and two typical WWTPs (Nos.9 and 10, serving as a reference in terms of size and treatment but without connection to the paper processing industry) sampling and analysis were repeated in 2020 and 2022.These measurements were intended to investigate the effects of the REACH restriction on the use of BPA thermal paper put in place in January 2020.Table 2 shows the BPA concentrations measured in influent, effluent, and sewage sludge samples.We observed a strong reduction in BPA influent concentrations in WWTPs connected to the paper processing industry.We assume this reduction is a direct consequence of the restriction of BPA in thermal paper.This restriction specifies that no new thermal paper containing BPA will enter the wastepaper stream, which is expected to lead directly to reductions in BPA in influents of WWTPs connected to the paper recycling industry.However, BPA is still present in recycled paper and in thermal paper already in use and will continue to be found in wastepaper and recycled paper streams for some time.Hence, a slow decay of BPA in recycled paper is expected, but it is not fully understood how long it will take for BPA to be completely absent from the paper recycling stream.Nevertheless, the observed reduction in BPA influent concentrations in reference WWTPs could be explained by such a delayed indirect effect.A comparison of the measured values between 2018 and 2022 clearly indicates a reduction in all measured values, which supports both assumptions.

Bisphenol A in landfill leachates
As described in the Materials and Methods section and in Table 3, measured BPA concentrations in landfill leachate were categorized according to the implementation of an onsite pretreatment step, subcategorized by the sampling spot (before or after the on-site pretreatment) and fate of the leachate (WWTP, direct discharge or unknown).The highest concentrations were detected in Category 1a: sampled before on-site pretreatment and WWTP.The concentrations ranged from 62 to 11 000 μg L −1 (mean 1900 μg L −1 ).As this leachate undergoes an on-site pretreatment, it does not discharge directly to surface water.After pretreatment, it is transported to a WWTP for further treatment.The measured BPA concentrations directly after the initial on-site pretreatment step (Category 1b) decreased by three to four orders of magnitude and ranged from 0.026 to 2.1 μg L −1 (mean 0.76 μg L −1 ).In Category 2 landfill leachates, where no on-site pretreatment is present and the leachate is directly transported to a WWTP, the mean BPA concentration was 15 μg L −1 and ranged from 0.072 to 70 μg L −1 .Only one measurement was available for Category 3 landfill leachates (on-site pretreatment, direct discharge).The measured BPA concentration was 0.010 μg L −1 (after on-site pretreatment).Leachate of landfills that do not perform an on-site pretreatment and discharge directly to surface water (Category 4) revealed BPA concentrations between 0.033 and 4.3 μg L −1 (mean 1.5 μg L −1 ).In Category 5, for which the fate of the leachate is unknown, BPA concentrations varied between 78 and 1300 μg L −1 (mean 690 μg L −1 ).
For active landfills, the year of landfill opening was investigated (Figure 4).In recently opened landfills (later than 2005), BPA was detected in low concentrations near the LoQ.In recent years, only landfills of Classes I and II (Deponieklasse) were opened in Germany because the disposal of MSW has been legally prohibited since 2005.Only inert waste is disposed of in Class I and II landfills.Articles made of ER, PC, PVC, and thermal paper as well as wastepaper, including recycled paper, are not inert waste.It is therefore reasonable to assume that this legal change for landfills will result in lower BPA concentrations in leachate in Based on the mean BPA concentrations calculated per category and the volume flow rate, loadings per category were estimated.The calculated loadings were 0.017 kg a −1 (Category 1b), 0.38 kg a −1 (Category 2), 0.00012 kg a −1 (Category 3), 0.0077 kg a −1 (Category 4), and 6.3 kg a −1 (Category 5).These values form the basis for calculating loadings per landfill used as input to the FlowEQ model, described in Bock et al. (2023).The calculated loadings for Class 1a (49 kg a −1 ) do not enter the surface water or WWTP directly, because they undergo an on-site pretreatment first.
The loadings of Category 1b were based on measurements after the pretreatment and were therefore considered for further calculations.There is no central register for all landfills in Germany, only some federal states (e.g., North Rhine-Westphalia) can provide reliable data.Additionally, there are numerous "wild" landfills.It was only in 1972 that the Law on the Disposal of Waste came into force in Germany, prohibiting wild and/or unregulated deposition of waste.Prior to 1972, waste of any kind was allowed to be disposed of.Without knowing the number and treatment methods of landfills in Germany, it is not possible to make a reliable extrapolation of landfill loadings.As discussed above, new landfill parts (e.g., Nos.6 and 3) that opened after 2005 belong to Category 4. The leachate is not treated but the loadings are comparably low.Older landfills of other categories (e.g., Category 1b, 2, or 5) are expected to be less significant in the future, although old deposits will continue to affect surface waters.Because of the new regulations, it is expected that loadings from older landfills will also decrease slowly.

Bisphenol A content in articles
The use of articles containing BPA represents another entry pathway of BPA into the environment.Articles that come into contact with water during use can potentially leach BPA into the environment.The aim of this part of the measurement campaign was to close data gaps, as it was previously unclear what types of articles contribute to BPA emissions.We aimed to understand potentially important article types by measuring their BPA content.In the next step, the measured values should be compared with other factors such as production quantities, uses, and so forth to identify potential important article groups.The exact process of how these data were used is included in the Supporting Information in Part II (Bock et al., 2023).
Preliminary studies (Ramboll, 2015(Ramboll, , 2017) ) concluded that PC uses are a minor contributor to the total environmental loadings of BPA.For the present work, we further revised the estimates of the BPA releases from articles based on estimated total installed surface area accumulated over service life for uses that may release BPA through water contact.For main market segments, downstream industry associations provided respective information.Conservative annual surface-related release rates have been determined for diverse use scenarios reflecting the scenario-specific typical stressors and environmental conditions (e.g., UV radiation, temperature, pH value, etc.) for outdoor and indoor uses (SEA, 2021).Estimated annual loadings from uses of PC in various market segments are 12.2 kg a −1 ) directly to surface water and 11.4 kg a −1 to municipal wastewater treated by WWTPs.Additional details and supporting information are provided in Supporting Information: Table SI-7 and  Attachment 1.As with PC articles, it was concluded that EpR uses contributed a minor fraction of the total environmental loadings of BPA based on market data, annual use, residual BPA content, and leachability.Leaching estimates were provided by Plastics Europe, CEFIC: Epoxy Europe, Ramboll (2021).Leaching was calculated based on surface area rather than mass as was previously done in Ramboll (2017).Plastics Europe, CEFIC: Epoxy Europe, Ramboll (2021) also completed a limited simulation for BPA loss over an article's service life resulting from exposure to physical stressors (e.g., ultraviolet radiation, rain, etc.).Estimated releases account for the fraction of surface area exposed to physical stressors caused by loss of the protective top coating, as unprotected areas account for most of the releases.Estimated annual loadings from uses of EpR in various market segments are 348 kg a −1 directly to surface water and 188 kg a −1 to municipal wastewater treated by WWTPs.Additional details and supporting information are provided in Supporting Information: Table SI-8 and Attachment 2.

Bisphenol A content in tested articles
The BPA content in selected articles was measured in samples obtained from the market.An assessment of the total annual amount of BPA in these article classes has been performed as a basis for release assessments used in the present study.Table 4 lists the article groups and sample numbers of this study, which were identified as potentially containing BPA. Figure 5 and Supporting Information: Table SI-3 provide a summary of BPA contents found in articles investigated in this study.Measured BPA contents in articles tested in this study varied strongly from values below the LoQ to 1 691 700 μg kg −1 .In the following, the measured values of important article categories are discussed.
Bisphenol A was found in all socks tested apart from one (2.5-41μg kg −1 ).No clear trend regarding material composition and BPA content was observed.Colored socks exhibited higher BPA levels than white or black socks.Bisphenol A contents in textiles were investigated in a study from 2017 (Xue et al., 2017).The values measured for socks in the Xue et al. study were much higher (186-13 285 μg kg −1 ) than the values measured in this campaign.However, the overall median BPA value of 10.7 μg kg −1 that Xue et al.
found for all their textile samples (n = 77, not only socks) was similar to the mean value (10.5 μg kg −1 ) measured in the 20 sock samples chosen for this study.
The BPA content in granulates from used tires from four different European producers ranged from 247.7 to 431.7 μg kg −1 (mean 369.7 μg kg −1 ).Literature reports BPA concentrations in tire shred in the range 20-60 μg kg −1 (Håøya, 2002) cited in Aabøe et al. (2004).All four tire granulates analyzed in this study exhibited higher BPA concentrations.
Bisphenol A levels in printing inks varied between values below the limit of quantification and 4 605 000 μg kg −1 .The highest BPA contents of 3 331 000 and 4 605 000 μg kg −1 were determined in two tested thermochromic inks, respectively, which are considered a niche application.Additionally, inks in toners and waste toner containers were tested, revealing BPA concentrations between less than LoQ in the toners and 85.3 μg kg −1 in the containers.By comparison, BPA concentrations in thermal paper were estimated to be 13 920 000 μg kg −1 , based on tonnage of BPA used as a developer divided by tonnage of thermal paper placed on the EU market in 2018 (ECHA, 2019).
Several articles made of PVC were tested.Bisphenol A values found in garden hoses ranged between 24 and 1148 μg kg −1 .Cables made of PVC from different producers were tested.Although three samples contained 10 μg kg −1 BPA or less, the two remaining samples contained 1243 and 6701 μg kg −1 .The values measured are lower than the values found in the literature.Yamamoto and Yasuhara (1999) found BPA values of 71 000 μg kg −1 for insulation of electrical wires.In 2010, the German Environmental Agency reported that BPA contents in cables are in most cases less than 0.1% (1 000 000 μg kg −1 ).The highest BPA content in a cable in the current sampling campaign was 0.00067 w/w% (6701 μg kg −1 ).Another article from the PVC group was shower curtains.Bisphenol A values in shower curtains ranged from 3.2 to 10 914 μg kg −1 .The results indicate that BPA is still intentionally used in production of PVC because the concentration levels found in some samples cannot be explained by impurities.The values found in tarpaulins/ banners ranged from undetectable to 298 μg kg −1 .Products made from (mostly soft) recycled PVC were, besides thermochromic inks, the product group with the highest BPA concentrations found among all products investigated.Bisphenol A concentrations ranged from less than 0.5 μg kg −1 in recycled PVC granules from a hose to 1 691 700 μg kg −1 in a beacon base.Bisphenol A contents of up to 615 800 μg kg −1 (riding area mats), 29 340 μg kg −1 (roofing and flooring mats), and 1 691 700 μg kg −1 (traffic management systems, i.e., beacon bases) were determined.
As representative paper samples, toilet paper, toilet paper tubes, newspaper, and posters were investigated for their BPA content.Two samples (printed and not printed/blank) from three newspapers from the Munich area were analyzed for BPA.Bisphenol A values were higher in the blank samples (13 989 μg kg −1 ) than in the printed samples (12 564 μg kg −1 ).The measured BPA concentrations in toilet paper made of recycled paper ranged from 5035 to 48 409 μg kg −1 , whereas concentrations in toilet paper tubes ranged from 27 583 to 47 561 μg kg −1 .Compared with data reported in the literature, toilet paper concentrations in the present study are on the same order of magnitude.Mean BPA concentrations were reported to be 28 550 μg kg −1 in the EU (Huntsman, 2017), 7300 μg kg −1 in Denmark (Vinggaard et al., 2000), 31 600 (Gehring et al., 2005) 10 200 μg kg −1 in Germany, 34 500 μg kg −1 in Austria, and less than 0.1 μg kg −1 in China (Gehring et al., 2009).In compost bags, two of the three products with unknown recycled paper content did not contain BPA in measurable amounts, whereas the two products made with recycled paper contained BPA at levels of 5643 and 7984 μg kg −1 , respectively.The third product with unknown recycled paper content had the highest BPA concentration (11 473 μg kg −1 ).The results for compost bags provide further evidence of the assumption that BPA can enter paper products via the recycling paper stream.The last paper article type was posters.Only two out of the seven samples (Posters 5 and 6) had measurable BPA levels.This is most likely explained by the fact that posters are usually made from virgin fibers.The highest concentration was measured in a printed sample (Poster 6: 5734 μg kg −1 ).A subsample from the unprinted area of the poster did not contain any measurable BPA indicating that the BPA originated in the ink.

Derivations of article loadings to surface water and wastewater
The measurements provided above are used as the basis for calculating per capita or per facility (paper recycling and landfills) loadings to surface water or wastewater for input to the FlowEQ model described in Bock et al. (2023).Briefly, BPA loadings from the consumer use of articles were calculated using concentrations in articles, annual article use, fraction of BPA leached to water, and the population of Germany.The calculation approach by loadings pathway (e.g., recycled PVC articles, tires, toilet paper, etc.) and calculated loadings are provided in Table 5, and detailed calculations for each are presented in Supporting Information: Additional details for the derivations of loadings to surface water and wastewater from consumer use and Tables SI-5-SI-10.Note that the future recycled PVC value in Table 5 assumes that the BPA content of recycled PVC drops in the future as described in detail in the Supporting Information: Recycled PVC.

DISCUSSION OF POTENTIALLY RELEVANT SOURCES AND PATHWAYS OF BPA IN SURFACE WATERS: BUILDING HYPOTHESES FOR FlowEQ MODELING
In a preliminary assessment by substance flow analysis (Ramboll, 2015), it was concluded that, compared with contributions from industry and manufacturing processes, emissions associated with consumer uses were the largest source of surface water BPA in Germany and associated watershed areas.Because the subsequent findings indicated that PCand EpR-based articles contribute to approximately 1% of total loadings to surface water during their use phase and disposal (European Commission, 2010b;Ramboll, 2017), the main sources remained unclear.The objective of the present pathway and source investigation was to address the data gaps that have been raised during the previous surveys and use the new and refined information as input parameters for a more robust FlowEQ modeling.
Based on the values from the measurement campaigns on selected WWTPs in Germany (including those influenced by discharges from different paper processing plants) and the data from (recycled) paper article measurements, it has been demonstrated that the current paper recycling streams in Germany contain BPA.This is presumably caused by thermal paper, such as labels or cash receipts, in wastepaper, which in Germany is subject to paper recycling.In 2021, 23 123 kt of paper was produced in Germany.A share of ~79% (18 297 kt) can be attributed to recycled material (Verband deutscher Papierfabriken e.V., 2021).Given the high production volume of recycled paper, the release of BPA associated with the use and recycling of paper products may currently also represent an important input pathway for WWTPs and eventually surface waters.However, the restriction on the use of BPA in thermal paper has entered into force in January 2020, and it is expected that this will have a significant effect on the circulating BPA in the paper recycling system and eventually on the future BPA concentration in surface waters.So, in the companion paper (Bock et al., 2023), attention should be paid to different scenarios for estimating surface water concentrations based on the different extent of future contributions from this important pathway.
The current importance of the paper recycling pathway is strikingly indicated by the BPA levels in toilet paper.All tested toilet paper samples based on recycled paper consistently contained high levels of BPA, whereas toilet paper produced from virgin paper did not contain any BPA.Similar findings for BPA levels in recycled toilet paper have been reported by Vinggaard et al. (2000).The release of BPA from used toilet paper into wastewater is not hindered by timedependent leaching, phase transitions or other physicochemical constraints, slow weathering processes-or even by measures to completely suppress even those, such as coatings and other protective mechanisms, which are applied to many PC or EpR polymer materials.
At this stage, it is difficult to predict both the kinetics of the potential reduction in BPA in the paper life cycle resulting from the phasing out of BPA in thermal paper and the associated prospective depletion of BPA levels in recycled paper and finally in surface waters.Wastewater treatment plants can be seen as a link between these two reservoirs, paper recycling streams on one hand and surface water concentrations on the other.Thus, the accompanying long-term monitoring of WWTPs can also be understood as a method to monitor the temporal dynamics of environmental BPA concentrations resulting in temporal trends in the paper streams, as well as using such a link for calibrating stresses to the model.The measurements of the present study may reflect a decline in BPA used in thermal paper because of its current restriction.
If the FlowEQ modeling in Bock et al. (2023) of this series confirms that a major input pathway for BPA to surface waters is related to paper use and recycling, the question remains as to what will be the subsequent dominant source and pathway after the thermal paper restriction has fully promulgated through the paper cycle.After concentrations reflecting changes in inputs from the paper recycling pathway are reached, inputs from landfills are expected to become relatively more important, at least temporarily.However, it was indicated by our measurements that BPA levels in leachates from landfills opened more recently were low.It is therefore expected that, due to the current restrictions on landfilling in Germany, there are only marginal new loadings from future landfilled material and a further slow decay of releases from older (legacy) landfills.
Apart from the contribution of end-of-life consumer products deposited in landfills and the direct release of BPA from recycled toilet paper in WWTPs and eventually surface waters, the assumption is confirmed that the input from the use of consumer goods could be considerably high.Although modeling is reserved for Bock et al. (2023), it must be emphasized that, in contrast to PC and EpR (Ramboll, 2017), high levels for BPA have been detected in recycled PVC products (up to 1 691 700 μg kg −1 ; this study).These findings suggest that BPA is also circulating in the recycling streams of PVC, presumably originating, to a significant extent, in the use as an additive in PVC cable sheathing.Macroscopic examination of the articles analyzed in our study revealed remainders of detached cable sheathing.Because most of these recycled PVC articles are manufactured in large volumes and include outdoor use (traffic management systems and/or beacon bases, outdoor flooring, and riding arena mats), which are furthermore not sealed by protective coatings, it is assumed that this article class could represent a significant contribution.
There are hints that point to the importance of a pathway that has received little attention so far: the observations of reduced WWTP purification capacities in the course of heavy rainfall events.The modeling results could provide quantitative predictions (Bock et al., 2023).However, it is beyond the capabilities of FlowEQ modeling to distinguish whether this occurred as a result of dilution effects or bypass events.Further investigations are necessary here.However, it can be assumed that, if confirmed as an important input pathway by modeling, these effects are more likely to increase along predictable higher frequencies and intensities of extreme weather events caused by climate change.Adapting stormwater infrastructure of municipal WWTPs to respond to these events and the impacts of these upgrades is an important area for future research.

CONCLUSION AND OUTLOOK
To the best of our knowledge, this is the most comprehensive study of BPA concentrations in WWTPs, landfill leachates, and consumer articles for Germany.As part of the determination of input parameters for the FlowEQ model in Bock et al. (2023), annual total BPA loadings entering surface waters in the model domain will be calculated based on the results of this study.The FlowEQ model presented in Bock et al. (2023) uses these inputs to attribute a significant part of the annual BPA emissions and the resultant concentrations in surface water to specific sources.This research thus serves as an important tool to model alternative emission scenarios and explore how concentrations may differ as a result of changes in usage patterns and treatment technologies.

LIMITATIONS
The objective of this monitoring campaign was to identify and measure BPA in wastewater, landfill leachate, and various articles.Although the study provides critical insights, certain limitations might be considered when interpreting the results.Data were collected from only 23 wastewater plants and sampled (mostly) once, which may not capture BPA temporal variations, thus affecting the extrapolation accuracy for all of Germany.Moreover, the substantial variation in the amount of treated wastewater, across different plant sizes and processes, introduces further complexity to the interpretation of our results.Similarly, the study considered many articles, but not the full spectrum of potential BPA-containing articles.This could lead to an under-or overestimation of the true BPA levels.Finally, the study measured BPA in the articles, not considering the actual leaching of BPA from these articles.However, our aim was to get an overview of which articles contain large amounts of BPA.In a qualitative analysis of the relevance of article types, we have chosen some types of articles (e.g., recycling paper) for further consideration in the model.Although these limitations must be kept in mind, the study provides an essential foundation for understanding BPA's environmental presence and sets the stage for more comprehensive future research.

FIGURE 4
FIGURE 4 Bisphenol A (BPA) concentrations measured in nontreated landfill leachate plotted against year of opening of cells

TABLE 3
Bisphenol A (BPA) concentrations measured in landfill leachate

TABLE 4
Article groups and number of articles or samples per group analyzed in the measurement campaign FIGURE 5Mean bisphenol A (BPA) concentrations in consumer articles (NP, newspaper printing; PV, protective varnish; PVC, polyvinyl chloride; RPVC, recycled PVC; SO, synthetic-oil-based; TC, thermochromic; UV, ultraviolet; VO, vegetable-oil-based)