Comparative meta-analysis of organic contaminants in sewage sludge from the United States and China

Abstract

Characterizing the occurrence of organic contaminants (OCs) of environmental health concern in municipal sewage sludges is essential for safe handling and disposal of these abundant materials. This meta-analysis aimed to (i) summarize the extent of studies performed on the chemical composition of sewage sludges from China and the U.S., the world’s two largest chemical producers, (ii) identify chemical groups of priority concern, (iii) quantitatively compare chemical abundance in sludge between nations, (iv) determine longitudinal contaminant accumulation trends in sludge, and (v) identify data gaps with regard to OC concentrations in sludge. A literature search was conducted on concentrations of OCs in U.S. sludges produced during treatment of domestic and industrial wastewater and compared statistically to contaminant levels in Chinese sludge abstracted from a recently established database. Longitudinal trends of OC occurrence were interpreted in the context of national chemical production, usage statistics, and regulations. A total of 105 studies on OCs in U.S. sewage sludge were found, while a total of 159 had been found in China. Among 1,175 OCs monitored for, 23% of all analytes had been monitored in both countries (n = 269), 41% (n = 480) in China only, and the remaining 36% (n = 426) in the U.S. only. On average, concentrations of OCs were 4.0 times higher in U.S. than in Chinese sewage sludge, with the highest detection being observed for alkylphenol ethoxylates. Data from a new binational database on toxic OCs in sewage sludges suggest and reiterates the need for additional chemical monitoring in both countries, risk assessments for emerging OCs contained in sludges destined for application on land, and stronger enforcement of sludge disposal restrictions in China, where as much as 40% of sludge is currently being dumped improperly.

Graphical abstract

1. Introduction

Sewage sludge is a solid byproduct of the wastewater treatment process, and generally acts as a sink for organic contaminants (OCs) (Clarke and Smith 2011; McClellan and Halden 2010; Patureau et al. 2021; Venkatesan and Halden 2014). Disposal methods of sludge produced in municipal and industrial wastewater treatment plants (WWTPs) vary significantly depending on the quality of sludge. In the U.S., sludge is disposed of through landfilling, incineration, or as soil amendment (fertilizer), thereby potentially reintroducing these OCs back into the environment. While difficult to estimate, the most recent estimates for wet annual sewage sludge production in the United States and China are 13.84 MT/year and 35.4 MT/year, respectively (Lu et al. 2019; Seiple et al. 2017). Currently, about 47% of U.S. sewage sludges are land applied, and the remaining 53% disposed of either through incineration (15%), or surface disposal (6%) or other methods (32%) (Environmental Protection Agency 2018a). While prior to 2010, about 3% of Chinese sludge was legally applied to land, almost none was incinerated, and over 80% was improperly dumped, China has made significant changes in recent years, with recent data showing about 13% of sewage sludge being used for building materials, 22% being incinerated, 27% landfilled, and just under 40% either land applied or improperly dumped, as it is difficult to distinguish between the two. (Lu et al. 2019; MOHURD 2018; Yang et al. 2015; Zhang et al. 2016). While the EPA has performed two distinct human health and environmental risk assessments on the disposal of toxic metals and a few classes of organic contaminants in sewage sludge (Environmental Protection Agency 1995, 2002a), there remain 352 pollutants that the EPA has been unable to perform risk assessment for, 61 of which are listed as a hazard or a priority on at least one of the Resource Conservation and Recovery Act hazardous waste listings, the EPA priority pollutant list, or the National Institute for Occupational Safety and Health’s list of hazardous drugs (Environmental Protection Agency 2018a). Due to the number of pollutants unevaluated through risk assessment in the U.S. and the continued improper dumping of sewage sludge in China, the monitoring of toxic chemicals in sludge is paramount for environmental practices in both countries.

While many organic compounds degrade easily and have minimal harmful effects on the environment, other more persistent organics have the potential to accumulate in environmental and biological matrices and can eventually cause harm to humans, wildlife, and the environment. In multiple instances, recalcitrant OCs have been found to be uptaken from land-applied sewage sludge into plant roots (Engwall and Hjelm 2000; Wu et al. 2010; Wu et al. 2012; Wu et al. 2015) and wildlife (Gaylor et al. 2013; Rivier et al. 2019) due to the land application of sewage sludge. In an effort to characterize the presence of OCs in sewage sludge, hundreds of studies have been published worldwide. The U.S. EPA has conducted four major national sewage sludge surveys for organic contaminants in 1982 (Environmental Protection Agency 1982), 1987-1988 (Environmental Protection Agency 1996), 2001 (Environmental Protection Agency 2002b), and 2006-2007 (Environmental Protection Agency 2009b). In the 1987-1988 and 2001 studies, the data was used in part for risk assessments to guide potential sludge regulation. The 1987 EPA sludge study examined 411 analytes, but later noted that most of the chemicals identified did not have sufficient toxicity data to conduct human health and environmental risk assessments (Environmental Protection Agency 2018b). Of the 50 analytes selected for risk assessment, nine metals were ultimately determined to have concentrations and toxicity levels that merited regulation in final treated sewage sludge. In 2001, the EPA conducted it’s third nationwide sewage sludge study, examining 12 polychlorinated biphenyl (PCB) congeners and 17 dioxins and furans in 201 samples from 171 WWTPs, and after performing toxicity studies and risk assessment, the EPA decided that the data did not support a need to regulate the disposal of either PCBs or dioxins in land-applied sewage sludge (Environmental Protection Agency 2003b).

China and the United States are the leading consumers of chemicals and together account for 50 percent of worldwide chemical sales (The European Chemical Industry Council 2014), and therefore are likely to have higher concentrations of OCs in sludge when compared to other countries. Understanding the concentrations of chemicals in sludge from these two countries can also give insight into global trends of OCs in sludge. Past reviews of OCs in sewage sludge have not had access to data from Chinese sewage sludge studies, as over one-third of those studies were published in Mandarin Chinese and were therefore not available to the international research community. In 2016, a comprehensive review of OCs found in Chinese sewage sludge was published (Meng et al. 2016). In this review the sludge data originally published in Mandarin Chinese was compiled and translated into English, allowing the international community access to the data. We have utilized this database by juxtaposing it with a U.S. OC sludge database that we created for this review, thereby providing a tool by which researchers can compare the OC burden in sewage sludge from the world’s two largest chemical consumers.

2. Methods

2.1. Identification

We expanded an existing single-country municipal sewage sludge database and performed a comparison to similar data from a previous study done in China (Meng et al. 2016). To expand on this database, using a modified version of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, we used three different approaches to find studies that fit within our criteria (Figure S1). In June of 2019, we performed searches of Web of Science, Google Scholar, and Arizona State University’s OneSearch database using the format (“sewage sludge” OR “biosolids”) AND (“pollutant 1” OR “pollutant 2” OR “pollutant 3” …), in which the pollutants used were each of the 749 compounds found in the Chinese OC database. Secondly, additional records identified from an existing database on sludge-related articles not captured by the search were included in the preliminary identification. Any study was excluded if it did not include at least one statistical data point for an organic contaminant concentration in U.S. municipal sewage sludge. No limitations on date of publication were utilized. Duplicates between any of the databases searched and the internal sludge article database were removed. As the third approach for finding articles, each of the remaining journal articles were forward- and back-citation checked for additional articles on OCs in U.S. sewage sludge.

We are aware of the potential for a lack of rigorous search criteria and methodology to lead to publication bias, wherein the selected journal articles are not representative of the whole field of literature (Sutton 2005). This could be due to the authors choice of databases to search, due to the choice of search terms, or due to the methods for filtering out unrelated publications. While it is certainly not possible for a body of literature to be fully indexed within one search, we have taken certain precautions, including combing through each initial search result by hand to determine eligibility for inclusion. Confirmation bias may also play an important role (Dickersin 2005), although this is potentially mitigated because only the raw chemical data from sewage sludges were used for this publication. Additionally, no statistical tests performed by the studies were included in the final statistical analysis from this study.

From each of these publications, we recorded statistical, temporal, and spatial data as available. If the sampling time was not reported, a sample collection date equal to the publication date minus three years was applied, as three years was the average time between sample collection and publication for the publications that included this information. For OCs with non-detects (ND), we substituted a concentration of zero for use in mean and median equations. Typically, a more sophisticated method of interpreting NDs, typically by dividing the method detection limit (MDL) by 2 or the square root of 2, is employed. However, as many of the older studies did not include details on the MDL, we decided not to include it in the meta-analysis in instances where provided. The sample locations were recorded at the highest spatial resolution available. All concentrations were converted to units of micrograms per kilogram of dry weight for uniformity, however, the 1982 first EPA nationwide sludge study only provided wet weight concentrations. To address this, a water content of 90% was assumed to convert to dry weight. Detection frequencies that were missing from the publication were labelled NA. The median was used as a substitute for the mean in a few instances where the mean was not provided. These chemicals were grouped into thirty-two primary groups of chemicals, and the remaining chemicals were placed into an “Other” category. The geometric mean of the U.S. to China average concentration for each OC was used to calculate an overall ratio between the two countries. An estimate of total global publications related to OCs in sewage sludge and biosolids was performed by a year-by-year search of Google Scholar for “organic AND (biosolids OR ‘sewage sludge’)” Information on U.S. and Chinese sludge regulations was retrieved from the EPA website and from a review paper on Chinese sludge regulations (Yang et al. 2015).

2.2. Statistical Analysis

Welch’s t-test was performed in Microsoft Excel 2016 to compare the means between U.S. and Chinese sewage sludge OC concentrations for each contaminant, which served as the summary statistic for this analysis. This test is a specific application of the Student’s t-test that is more robust when the sample groups have potentially unequal variances and unequal sample sizes. While significant values of α = 0.05, 0.01, and 0.001 were used to evaluate the difference in mean concentrations of priority pollutants between the U.S. and China, the raw p-values were also provided directly in Figure 4.

Figure 4.

Comparison of concentrations of major contaminants in sludges from the U.S. (blue) and China (gray), (x, y) indicates a total of x studies that contain data for a total of y samples for that chemical. Standard error bars shown in black.

ǂ Unable to do statistical comparison due to China having only one sample available

3. Results

3.1. Temporal Distribution of Studies

The preliminary literature search yielded over 1,500 sludge-related journal publications for OCs. After filtering using the prior stated methodology, a total of 105 peer-reviewed publications remained (Figure S1), spanning over 40 years of research and including four national EPA sewage sludge studies (Figure 1A). In contrast, existing international sludge review papers have included no more than 20 USA sludge studies each (Clarke and Smith 2011; Harrison et al. 2006). The earliest available U.S sludge publication on organic contaminants was a 1976 study that analyzed municipal sewage sludges for dieldrin and total PCBs (Furr et al. 1976). The U.S. publication rate for sludge studies has tapered off in recent years, being surpassed by China’s publication in the mid-2000s. Between 2011-2015, China maintained a publication rate of 23 studies per annum, almost four times greater than the U.S. annual publication rate (6) for sludge studies on OCs over the same time period. These publications that have data on OC concentrations in sludge were preceded by a general publication interest in OCs in sewage sludge, which increased from 1994 to 2001, and has remained relatively steady since then.

Figure 1.

A) Temporal distribution for sewage sludge studies for organic contaminants, and international trendline for biosolids related studies.

B) Timeline for U.S. and Chinese sewage sludge regulations.

NEBRA North East Biosolids and Residuals Association.

*Found searching Google Scholar for “organic AND (biosolids OR ‘sewage sludge’)”

**No information provided for Chinese studies. U.S. studies through July 2019

***Chinese regulations sourced from Yang et al, 2015

3.2. Spatial Distribution of Sample Locations

Of the 105 studies found with data from the U.S., about one third (n = 37, 35%) were either national studies or did not provide sample locations. The remaining 64 studies were focused on specific geographic areas, with a significant portion from EPA Region 5: the Mid-Atlantic (n = 18, 17%), some from EPA Region 3: the Midwest (n = 12, 11%), smaller portions for both EPA Region 2 and EPA Region 9: the Southwest (n = 11, 11%; n = 11, 11%, respectively), and the few remaining studies from each of the other EPA regions. The national studies provide some sample coverage of the less-reached EPA regions, although no studies with samples from Alaska, Hawaii, Puerto Rico, or the U.S. Virgin Islands have been reported. Spatial distribution of the Chinese sewage sludge studies have already been reported in a previous study (Meng et al. 2016).

3.3. Sewage Sludge Regulation

The only federal regulations in the U.S. that affect sewage sludge disposal at the federal level are the 1972 Clean Water Act, the 1988 Ocean Dumping ban, and most recently the 1993 40 Code of Federal Regulations Part 503 rule that set the framework for sewage sludge regulations (Environmental Protection Agency 1988, 1994; United States 1972). To inform these as well as potential future regulation, the EPA conducted four national sewage sludge studies (1982, 1987, 2001, 2006/2007), but only the 1987 study’s data was used in creating the 1993 regulations (Figure 1B). The 2001 study concluded in a decision not to institute limits for PCBs in applied sewage sludge, and the 2006/07 study has had no formal risk assessment performed. In contrast, China has had 36 separate regulations regarding sewage sludge (Yang et al. 2015), but has not had a government-run national survey to inform regulation.

3.4. Comparison of Contaminants Studied in US and Chinese Sludge

In U.S. sludge, 310 (43%) of OCs analyzed were non-halogenated (Figure 2A). This is comparable to the Chinese study, in which 335 (45%) of the analytes tested for in sludge were non-halogenated. Almost half of the U.S. chemicals were chlorinated (n = 315, 43%), of which more than half (n = 158) were PCB congeners tested for in the 2001 EPA sludge study (Environmental Protection Agency 2002a). In contrast, a smaller percentage of chemicals were fluorinated (n = 38, 5%) or brominated (n = 67, 9%) when compared to the Chinese study (n = 83, 11%; n = 103, 14%, respectively). This is partially due to the U.S. lack of data for hydroxyl- and methoxylated-polybrominated diphenyl ethers (OH-PBDEs and MeO-PBDEs, respectively), and due to having comparatively less data available for perfluoralkyl and polyfluoroalkyl substances (U.S.: n = 21; China: n = 37). In addition, a total of eight chemicals (citalopram, fipronil, naled, leptophos, and four of their metabolites), contain more than one kind of halogen and have been detected in U.S. sludge.

Figure 2.

Comparison of A) halogenation of organic contaminants found in U.S. and Chinese sewage sludge. F_x-R : fluorinated compounds. Cl_x-R: chlorinated compounds. Br_x-R: brominated compounds. B) number of chemicals found in U.S. and Chinese sewage sludge

A total of 730 organic chemicals belonging to 29 classes in 105 U.S. sludge studies have been found in the literature (Figure 2B). The most studied classes of chemicals were triclosan, triclocarban, and their metabolites (n = 34, 32% of studies), polybrominated diphenyl ethers (n = 20, 19%), pharmaceuticals (n = 16, 15%), alkylphenol ethoxylates (n = 11, 10%), antibiotics (n = 10, 9%), synthetic musks (n = 10, 9%), steroids and hormones (n = 9, 9%), polychlorinated biphenyls (n = 8, 8%), and perfluoralkyl and polyfluoroalkyl substances (n = 9, 9%). Due to many studies including sludge concentration data for more than one chemical class, the sum of n is greater than N = 105. However, of the 730 chemicals detected in U.S. sludge, only 206 (30%) of these chemicals have been detected both in the U.S. and in China. Chemical groups for which sewage sludge data exists in China but not in the U.S. include the following: OH-PBDEs, MeO-PBDEs, n-heterocyclic carbenes, and quaternary ammonium compounds (QACs), although it is important to note that MeO-PBDEs, while analyzed, have not been detected in Chinese sewage sludge to date. Chemicals and chemical groups for which no data from the past five years exists for U.S. sewage sludges include the following: dechlorance plus, hexabromocyclododecane, most polybrominated diphenyl ethers (PBDE), phenol compounds, steroids and hormones, synthetic musks and fragrances, PCBs, phthalic acid esters (PAEs), volatile aromatic hydrocarbons (VAHs), aromatic amines (AAs), polycyclic aromatic hydrocarbons (PAHs), and dioxins and furans.

3.5. Concentration Levels

Five chemicals, octamethylcyclotetrasiloxane (D4), decamethylcyclopentasiloxane (D5), and dodecamethylcyclohexasiloxane (D6), ketoprofen, and PCB-11, were each found to have a mean concentration more than two orders of magnitude higher in Chinese municipal sewage sludge than in the United States. D4, D5, and D6, (U.S.: 6.0, 0.8, 9.4; China: 1189, 1876, 1678 μg/kg, respectively) are cyclic siloxanes, and are widely used in biomedical and cosmetic applications. Ketoprofen, an analgesic and antipyretic, is the only pharmaceutical of the 18 found in both countries that had a difference in mean sludge concentration of more than two orders of magnitude (U.S.: 8.6 μg/kg; China: 1578 μg/kg, respectively). PCB-11 was the only one out of 44 shared PCB congeners that varied by more than three orders of magnitude (U.S.: 11.4 μg/kg; China: 4478.4 μg/kg), having concentrations much higher in Chinese sewage sludge.

Several chemicals were found to have significantly higher concentrations in U.S. sludge than in China. One of the main components of OctaBDE, BDE-196, had mean concentrations of 132.7 μg/kg in the U.S. and 0.8 μg/kg in China. One novel brominated flame retardant, 1,2-Bis(2,4,6-tribromophenoxy)ethane (BTBPE), had an average concentration of 1762 μg/kg in the U.S. and 0.95 μg/kg in China. Norgestrel, a steroid used in female contraceptives, had average concentrations in the U.S. of 344.5 μg/kg, while concentrations were significantly lower in China (0.091 μg/kg). Chlordane, an insecticide, had an average concentration of 244.5 μg/kg in the U.S. and 1.08 μg/kg in China. Lastly, two musks, galaxolide (HHCB) and tonalide (AHTN), had concentrations over three orders of magnitude greater in the U.S. than in China (U.S.: 25,300 μg/kg and 20,333 μg/kg; China: 27 μg/kg and 6.2 μg/kg, respectively). On average, OC concentrations in U.S. sludge were 4.0 times greater than in Chinese sludge. The resulting power regression equation of the log-log data for Figure 3 was y = 0.9054x^0.6772 (x = U.S. concentration, y = Chinese concentration, R = 0.60). The exponent on x is less than 1, which indicates that on average, OC sludge concentrations are higher in the U.S. than in China (Figure 3).

Figure 3.

Log-scale plot of the average concentrations for 206 OCs found in U.S. and Chinese sewage sludge

3.6. Comparison of Total Concentration Across Chemical Classes

The chemicals with the greatest mean concentrations in U.S. sewage sludge were alkylphenol ethoxylates (APEOs), phthalic acid esters (PAEs), and polycyclic aromatic hydrocarbons (PAHs), while dioxins and furans were the lowest in concentration. Out of the 34 representative chemicals (Figure 4), the United States had higher mean concentrations than China for 26 of these chemicals and was more than an order of magnitude greater than China in 8 of these. Of these 34, the mean concentrations between U.S. and China were significantly different with p < 0.05 for seventeen chemicals, p < 0.01 for fourteen chemicals, and p < 0.001 for ten chemicals. The representative chemicals that had the greatest difference between U.S. and China were testosterone, BDE-47, and PFDA (U.S.: 131, 610, 26.1; China: 2.3, 13.6, 1.0 μg/kg, respectively).

4. Discussion

4.1. Binational Differences in Publication Rate

A few possible reasons could explain the difference in total number of sludge studies between U.S. (n = 105) and China (n = 159). First, the lack of a Chinese equivalent to the EPA-sponsored national surveys could have driven more studies to be published at the university level to understand OCs in Chinese sludge. Additionally, the trends in government funding for environmental protection are different. Chinese funding for environmental protection projects has increased steadily over the past two decades, increasing from less than 15 billion USD per year prior to 2001, to over 150 billion USD in 2013 (Meng et al. 2016). In contrast, U.S. funding for the EPA has stagnated at about seven to nine billion USD per year for the past 15 years, and is projected to drop to below six billion USD per year for the following five years (White House Office of Management and Budget 2017). A combination of these can partially explain why China has published over 50% more studies on sewage sludge than the U.S has published. Similarly, differences in government structure result in a difference between the number of federal sludge-related regulations in the U.S. (n = 3) and in China (n = 36). In total, 4 laws, 10 national standards, 15 ministry standards, and 7 technical regulations and guidelines for sludge management have been created by five administrative agents that oversee sludge disposal in China (Yang et al. 2015). This high number is due in part to China having five separate sewage sludge administrative agents, while in the U.S. sludge disposal is regulated primarily by the EPA.

4.2. Binational Differences in Analytes Detected

Halogenation of organic chemicals is a relevant sorting mechanism due to halogenated chemicals generally being more toxic and recalcitrant than their non-halogenated counterparts (Ghosal et al. 1985). The U.S. and China have detected approximately the same percentage of analytes when organized by halogenation. In both countries, about 40% of the chemicals with sludge data are non-organohalogens and belong to a variety of the chemical classes used in this study. China has studied about twice as many fluorinated OCs (U.S.: n = 38, 5%; China: n = 83, 11%), most of which are accounted for in a greater number of antibiotics and per- and polyfluorinated chemicals (PFCs) studied (U.S. n = 41, 21; China: n = 86, 59 respectively). China has also studied 38 more brominated compounds, the majority of which are QACs, OH-PBDEs, and MeO-PBDEs (China: n = 27, 9, and 9 respectively). Of these, MeO-PBDEs have been tested for but not detected in Chinese sludge, but OH-PBDE sludge concentrations have been found within the same order of magnitude as their parent PBDE congeners. QACs are used as disinfectants, and have been detected in Chinese sludge at concentrations between 38,000 to 154,000 μg/kg (total QACs) (Meng et al. 2016). Even though China has not undertaken a national sludge survey before, China has studied a greater number of OCs in sludge, likely due to China’s greater raw number of studies. Just under half (n = 206) of the 461 U.S. chemicals that have not been tested in China belong to just a few classes: pharmaceuticals, steroids and hormones, organochloride pesticides, and PCBs. In contrast, China’s unique chemicals are more spread evenly across the chemical classes, even including six classes of chemicals, synthetic phenolic antioxidants (SPAs), short-chain chlorinated paraffins (SCCPs), QACs, n-heterocyclic carbenes, napthenic acids, and organometals, none of which have been tested in U.S. sludge before.

The lack of data on these classes can be partially explained through the EPA national sewage sludge studies, the two most recent of which targeted specific classes of chemicals considered to be higher risk than the above listed classes. In 2001, the EPA national sludge survey targeted dioxin and dioxin-like compounds, and in 2006/07 focused on antibiotics, steroids and hormones, and flame retardants, (Environmental Protection Agency 2002b, 2009b). Thus, while the sample sizes for these studies are significantly higher than other studies in either the U.S. or in China, the breadth of chemical classes analyzed in the U.S. is narrower than in China. This highlights an important difference between the U.S. and Chinese sludge data that is available—China has a greater number of studies (U.S.: n = 105, China: n = 159) and has tested more chemical classes than the U.S. has (U.S.: n = 29, China: n = 35)., but the U.S. has the benefit of having four national campaigns that offer strong geographic sample distribution and sample sizes ranging from n = 50 to 174, while the largest Chinese study had n = 60 samples (Ruan et al. 2012).

4.3. Priority Chemical Classes for the United States

As sewage sludge continues to be applied on land in the United States, it is important to understand as completely as possible the constituents that make up the sludge before application. The United States specifically would benefit from studies characterizing these six chemical classes, SPAs, SCCPs, QACs, n-heterocyclic carbenes, napthenic acids, and organometals, for which the U.S. currently has no data on their presence in sewage sludge. SPAs, and specifically 2- and 3-tert-butyl-4-hydroxyanisole, tert-butylhydroquinone, and 3,5-di-tert-butyl-4-hydroxytoluene, are used extensively in several industries, but primarily the food industry (Perrin and Meyer 2002). SPAs have been demonstrated to be toxic to some animal tissues (Horvathova et al. 1999; Yu et al. 2000), may have additive carcinogenic effects (Hirose et al. 1998), and several of them are listed in the EPA’s high-production volume (HPV) database (Liu and Mabury 2018). Organometals would likely be detected in U.S. sewage sludge, as the 2015 U.S. total consumption of organometallics was more than China’s total consumption, and the U.S. leads the world particularly in organoaluminums, which are mostly used as catalysts (IHS Markit 2016). While the toxicity of most organometals is not well characterized, several organometals have been shown to be potentially genotoxic, carcinogenic, and neurotoxic (Dopp et al. 2004). SCCPs are used as lubricants and coolants in metal, and have been demonstrated to have chronic toxic effects in both humans and wildlife, and particularly aquatic organisms (Ali and Legler 2010). As of 2007, the total mass of SCCPs and medium-chain chlorinated paraffins produced in the U.S. was estimated to be 100 million pounds, of which about one-third was SCCPs (Environmental Protection Agency 2009a). Until these are characterized, little risk assessment can be done to understand the potential risk they may pose to humans and the environment through the application of sludge onto agricultural land.

4.4. Temporal Trends of Contaminant Concentrations

In some cases, the regulation of OCs can be shown to affect the sludge concentrations over time. The clearest example of this is in total PCBs (Figure 5A). Due to their recalcitrance and toxicity, the EPA banned the production of all PCBs in 1979, and regulated any substance that contains more than 50 ppm of PCBs (Environmental Protection Agency 1979). We demonstrate the effects of this policy in Figure 5A, where a steady decrease can be seen from the first PCB sludge study (Furr et al. 1976), to the most recent PCB data from the third EPA national sludge survey (Environmental Protection Agency 2002b). Another example of production management decreasing sludge concentrations is with PFOS. The United States primary producer of PFOS, 3M, announced a voluntary phase-out of PFOS in their products starting in 2000 due to PFOS’s persistence in human and animal tissues (Environmental Protection Agency 2000). As of 2002, 3M had ceased production of PFOS, which is demonstrated by a decrease in U.S. sludge PFOS concentration over time. However, recent data comes from only one Mid-Atlantic WWTP (Armstrong et al. 2016), and thus is not representative of the country as a whole. More recent studies with samples from more locations are required in order to fully understand the current status of PFOS in sludge. Although PFOS has not been reported to be mass-produced in the U.S. since 2002, it may still be produced in lesser quantities that are not required to be reported, and may be present due to previously made or imported PFOS products (Environmental Protection Agency 2015).

Figure 5.

Temporal trends of certain OC concentrations in U.S. sewage sludge (dashed line is trendline; dots are individual study results). BDE, bromodiphenyl ether; PCB, polychlorinated biphenyl, PFOA, perfluorooctanoic acid; PFOS, perfluorooctanesulfonic acid

However, sometimes regulation has either not made an impact on sludge concentration, or has not had enough time or enough studied samples to see a significant decrease. In 2006, the eight major producers of PFOA in the U.S. signed an agreement to participate in the EPA’s PFOA Stewardship Program, with the goal of reducing PFOA production and emission by 95% by 2010, and end of production and emission by 2015. With the exception of one producer, the participating companies achieved more than 99% reduction in PFOA production and emission by 2014 (Agency for Toxic Substances and Disease Registry 2015). Unfortunately, data for PFOA in sludge is sparse, with only one study being published in the past decade (Figure 5B) (Armstrong et al. 2016). Thus, while data prior to PFOA regulation seem to maintain fairly constant concentrations, the impact of regulation is not clear from the available data. As with PFOS, this could be affected by already existing PFOA products as well as imported PFOA. Similarly, BDE-47 is the primary constituent of pentaBDE, making up between 38 to 42% of the total mass (La Guardia et al. 2006). In 2006, the EPA enacted the Significant New Use Rule, which required any new pentaBDE production pass EPA evaluation prior to manufacture (Environmental Protection Agency 2006). As of 2017, 10 U.S. states have banned or restricted production of pentaBDE (Safer States 2017). Despite these efforts, sludge concentrations of pentaBDE have continued to remain relatively constant since the late 1990s when it was first introduced. This could be due to a combination of its recalcitrance as well as it being formed by the breakdown of higher ordered brominated diphenyl ethers. A previous study has found that through aerobic degradation of PBDE/BDE in sludge, BDE-47 concentrations can increase slightly for the first several months, likely due to debromination of higher-ordered congeners (La Guardia et al. 2006). In one sample, BDE-47 did not significantly decrease over the study period, while other congeners decreased in concentration (Stiborova et al. 2015). The EPA phase-out for decaBDEs was later than pentaBDE and octaBDE, and took place over a 3-year period ending in 2013 (Environmental Protection Agency 2009c). Thus, it may take more time for PBDE congeners to be shown to decrease in sludge. Lastly, triclosan (TCS) and triclocarban (TCC) are antimicrobials that have shown relatively consistent sludge concentrations over the past decade. In 2016, the FDA issued a ban on these as well as other antimicrobials in hand soaps and body washes (Food and Drug Administration 2016). The ban went into effect in September of 2017, and further tracking in sludge would help elucidate how effective regulation is at mitigating the presence of persistence chemicals such as TCS and TCC. Thus, while existing sludge data can provide insight on the efficacy of regulation of older chemicals such as PCBs, future studies involving these chemicals are needed to shed light on the effect of more recent regulation on their presence in sludge.

4.5. Binational Comparison

Chinese sludge concentrations were less than U.S. sludge concentrations for 74% (n = 153) of the 206 chemicals which both countries had detected in sewage sludge. The most notable exception to that is Bisphenol A, which had a mean concentration in Chinese sludge of 10,900 μg/kg, and only 830 μg/kg in the United States. These data came from a few studies (U.S.: n = 5; China: n = 10) with samples collected between 2003 to 2015. Looking at other factors, we would expect the opposite mean concentration ratio to be the present. In 2003, the estimated net usage of Bisphenol A in the U.S. was 1.02 million tonnes, while in China it was only 139,000 tonnes (Gao 2003), a ratio opposite to what the sludge concentrations show. The disparity in mean BDE-196 concentration (U.S.: 132.7 μg/kg; China: 0.8 μg/kg) can be largely explained by the fact that China has never produced or used OctaBDE, for which BDE-196 is one of the primary constituents (Chen et al. 2012). Siloxanes were another group of chemicals for which Chinese sludge had greater concentrations than the U.S. As of 2002, the average annual production for cyclic siloxanes, including D4-D6, was almost twice as large in China as in the U.S. (800,000; 470,000 tons, respectively) (Tran et al. 2015; Xu et al. 2015). The sludge shows an exaggerated version of the same relationship – Chinese sludge concentrations were between 2-3 orders of magnitude greater than the U.S. However, the sludge data for the U.S. was limited to one study in south Florida, and was calculated using measured gaseous phase concentrations and air-water and water-sludge partition coefficients (Surita and Tansel 2014). PCB-11, the congener in which China’s concentration is over three orders of magnitude greater than the U.S., is an exception, not a rule for PCBs. In the case of PCB-11, the Chinese concentration is from only one sample near a chemical industry zone, while the U.S. concentration is the average of 94 samples across the U.S. Other PCB congeners tested from the same Chinese sample were similar to U.S. concentrations, and had more samples available for calculations. In addition, recent studies have suggested that PCB-11 can be inadvertently created during the production of yellow pigments (Shang et al. 2014; Vorkamp 2016). Future studies should be concentrated on similar chemicals and should have an increased sample size as well as geographic region covered, before more analysis can be made.

Chlordane, for which the U.S. mean sewage sludge concentration was about 200 times that of China, was banned by the EPA in 1988 (Agency for Toxic Substances and Disease Registry 1994). It makes sense then that U.S. concentrations may be greater than Chinese for chlordane, especially since the U.S. samples tested for chlordane come from when chlordane was still in use. We found that norgestrel, used in female hormonal contraceptives, is over three orders of magnitude greater in concentration in U.S. than Chinese sludge. While data about individual contraceptives is hard to obtain, recent pill contraceptive use in women has been about 10 times greater in the U.S., with 1.2% of married or in-union Chinese women using pill contraception, while 16% of U.S. married or in-union women use pill contraception methods (United Nations 2015).

4.6. Statistically Significant Differences Between U.S. and China

The power regression equation used to show a linear log-log relationship between these two countries demonstrates one of the main findings of our study: that sludge concentrations in the U.S. are about four times higher than in Chinese sludge. There are several compounding factors that can begin to explain this difference. One possible explanation for the variation in the sludge samples between China and the U.S. is that in most of the U.S. studies, the sludge was taken from domestic wastewater. As of 2004, only about 40 million, (14%) of the U.S. population lived in an area with a combined sewer (Environmental Protection Agency 2004), which would incorporate either rainwater runoff or industrial wastewater into the municipal sewer. However, Chinese municipal wastewater is more commonly mixed with industrial wastewater, with industrial contributions making up about 30% of the total contribution (Feng et al. 2015). This may cause the Chinese sludge concentrations to have a higher level of certain anthropogenic organic pollutants than U.S. sludge. Additionally, the U.S. and China have different sludge treatment methods. As of 2004 in the U.S., about 50% of WWTPs anaerobically digested their sludge before disposal (Environmental Protection Agency 2007). Conversely, while in China anaerobic digestion is acknowledged as a preferred sludge treatment method, only a few dozen out of 2600 WWTPs nationwide had implemented the anaerobic digestion process for sludge, and only about 20% of sludge is treated at all (Feng et al. 2015). OCs can be degraded through the anaerobic treatment process, and anaerobic treatment removes 50–75% of the volatile organic carbon which can reduce the total dry solids weight, and therefore artificially increase other OC concentrations (Environmental Protection Agency 1978). Thus, the higher concentrations in U.S. sludge when compared to China can be as a result of higher inflow wastewater concentrations, or due to more advanced sludge treatment techniques which sequester a larger portion of OCs into sludge and out of the treated effluent.

4.7. Limitations

Comparisons between chemicals from the 1982 EPA study were avoided due to the 1982 national study not providing mean concentrations, and concentrations being reported using wet weight instead of dry weight. Only minimum and maximum concentrations are available from that study, and not enough data was provided about the concentration distribution to accurately estimate the mean concentration. Additionally, except for the EPA studies, limited information was available on the sludge digestion method used for the WWTPs from which the samples were collected. Thus, it is not possible with current data available to make comparisons between various sludge digestion methods, or to account for digestion methods in the statistical analyses.

Another limitation arises from the variation in usage and regulation of chemicals inventoried in this binational database. The United States and China do not have the same regulations in place regarding chemical production and usage, and the regulations currently in place have not been in place across the whole study period. For this reason, care should be taken in comparing for example, an older study from China to a more recent one from the United States.

Due to a lack of reporting in a significant portion of the sources, the effect of Method Detection Limits (MDLs) was not incorporated into the study analyses. While methods such as the Maximum Likelihood Estimator (MLE) or treating non-detects as the MDL dividied by the square root of 2 are preferable (Newton and Rudel 2007), non-detects had to be treated as zero in order to maintain uniformity across the study, and chemicals that had significant numbers of non-detects were not used for detailed comparisons due to the uncertainty that such a comparison would have. While this is a limitation of the study, the margin of error from using zero as a substitute value for the MDL is relatively minimal when the number of non-detects is less than approximately 25% of the total number of samples (Environmental Protection Agency 2003a).

While a spatial analysis of sewage sludge OC concentrations would be valuable to researchers, it is currently difficult to perform due to a lack of available spatial data for any one OC or group of OCs. The largest contributors to this are 1) the lack of spatial variety within and between studies, 2) the fact that the EPA has kept private the key to match sample ID with the wastewater treatment plant location as a measure to protect the participating treatment plants, and 3) the fact that in some cases, a large number of samples were composited prior to analysis, thus losing the spatial component of the samples collected.

5. Conclusions

This study has combined a database of U.S. OC sludge concentrations with an existing Chinese database to perform a quantitative review of sewage sludge in the two countries that produce the largest amount of chemicals in the world. For the past few years, the United States annual sewage sludge OC publication rate has been about one-third of China’s and concentrations in the U.S. average 4.0 times higher than China, although factors such as combined sewer systems, contribution from industry, and sludge treatment technology complicate the issue. Prominent chemicals present in higher concentrations in U.S. sludge than in Chinese sludge include chlordane, norgestrel, HHCB, and AHTN, while chemicals found in much greater concentration in Chinese sludge than in U.S. sludge include Bisphenol A, ketoprofen, and some siloxanes. In some cases, these differences can be partially explained through production, usage, or other outside data sources, but in other cases, unknown information such as the type of digestion, sampling methods, and detection methods may be more influential. The data presented in this publication are of value to multiple stakeholders. Policymakers can utilize these data to understand the adequacy of current WWTP technologies in eliminating toxic and recalcitrant OCs from the environment. Those working in the field of public health, agriculture, and wastewater treatment may find this information useful in order to assess the potential risk posed to those handling sludge, although this risk assessment remains to be completed. We used sludge concentrations to measure the efficacy of regulation of chemicals such as PCBs and PBDEs in the U.S., and can continue to be used in the future to study the effect of new regulations such as the FDA antimicrobial ban in the U.S. In order to address the aforementioned data gaps, the U.S. should increase their breadth of studied chemicals to encompass OCs such as novel-BFRs, n-heterocyclic carbenes, quaternary ammonium compounds, and chemicals affected by regulation such as triclosan and triclocarban. Conversely, China should focus on increasing the sample sizes of their studies, for example by conducting a national sludge survey. Additionally, this new binational database can be leveraged by future researchers to perform risk assessments on the OCs that have been detected in U.S. or Chinese sludge. As the two largest chemical producers in the world, both the U.S. and China need to continue to ensure that their methods for the disposal of sludge preserve both public health and the environment.

Table 1.

Classes of Organic Contaminants (OCs) Analyzed in U.S. Municipal Sewage Sludge

Compound Class	Number of Studies (N)	Number of Analytes (n)
Antibiotics	10	41
Pharmaceuticals	16	65
Steroids and Hormones	9	35
Benzophenone and Benzotriazole Compounds	2	6
Siloxanes	2	6
Synthetic Musks	10	15
Triclosan, Triclocarban, and Derivatives	33	13
Parabens	2	10
Organochloride Pesticides	6	36
Organophosphate Pesticides	1	45
Polychlorinated Biphenyls	8	158
Polychlorinated Napthalenes	2	1
Alkylphenol Ethoxylates	11	16
Phenol	5	10
Bisphenols	6	18
Polybrominated diphenyl ethers	21	47
Novel brominated flame retardants	4	8
Dechlorance-plus	2	2
Perfluoroalkyl and polyfluoroalkyl substances	8	21
Phthalic acid esters/plasticizers	6	6
Volatile aromatic hydrocarbons	2	23
Aromatic Amines	2	6
Polycyclic Aromatic Hydrocarbons	10	23
Chlorinated Hydrocarbons	2	4
Polychlorinated dibenzo-p-dioxins and - dibenzofurans	6	33
Polybrominated dibenzo-p-dioxins and - dibenzofurans	1	7
Nitrosamines	4	11
Fungicides	6	2
Solvents	2	13
Melamine-based resins	1	4
Other	11	45

Highlights.

• Sewage sludge contains organic contaminants that pose potential threat to biosphere

• Developed U.S. – China database of organic contaminants in sewage sludge

• 1,175 analytes monitored in sewage sludge between the U.S. and China

• Concentrations in China four times lower than in the U.S. for the same analyte

• National surveys and increased monitoring can address data gaps