SEARCH

SEARCH BY CITATION

Keywords:

  • Biosolids;
  • PhATE™;
  • Pharmaceutical;
  • Model

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. OVERVIEW OF THE PHATE MODEL
  5. MODEL CORROBORATION
  6. API SIMULATIONS
  7. CONCLUSIONS
  8. Acknowledgements
  9. REFERENCES

This article presents the capability expansion of the PhATE™ (pharmaceutical assessment and transport evaluation) model to predict concentrations of trace organics in sludges and biosolids from municipal wastewater treatment plants (WWTPs). PhATE was originally developed as an empirical model to estimate potential concentrations of active pharmaceutical ingredients (APIs) in US surface and drinking waters that could result from patient use of medicines. However, many compounds, including pharmaceuticals, are not completely transformed in WWTPs and remain in biosolids that may be applied to land as a soil amendment. This practice leads to concerns about potential exposures of people who may come into contact with amended soils and also about potential effects to plants and animals living in or contacting such soils. The model estimates the mass of API in WWTP influent based on the population served, the API per capita use, and the potential loss of the compound associated with human use (e.g., metabolism). The mass of API on the treated biosolids is then estimated based on partitioning to primary and secondary solids, potential loss due to biodegradation in secondary treatment (e.g., activated sludge), and potential loss during sludge treatment (e.g., aerobic digestion, anaerobic digestion, composting). Simulations using 2 surrogate compounds show that predicted environmental concentrations (PECs) generated by PhATE are in very good agreement with measured concentrations, i.e., well within 1 order of magnitude. Model simulations were then carried out for 18 APIs representing a broad range of chemical and use characteristics. These simulations yielded 4 categories of results: 1) PECs are in good agreement with measured data for 9 compounds with high analytical detection frequencies, 2) PECs are greater than measured data for 3 compounds with high analytical detection frequencies, possibly as a result of as yet unidentified depletion mechanisms, 3) PECs are less than analytical reporting limits for 5 compounds with low analytical detection frequencies, and 4) the PEC is greater than the analytical method reporting limit for 1 compound with a low analytical detection frequency, possibly again as a result of insufficient depletion data. Overall, these results demonstrate that PhATE has the potential to be a very useful tool in the evaluation of APIs in biosolids. Possible applications include: prioritizing APIs for assessment even in the absence of analytical methods; evaluating sludge processing scenarios to explore potential mitigation approaches; using in risk assessments; and developing realistic nationwide concentrations, because PECs can be represented as a cumulative probability distribution. Finally, comparison of PECs to measured concentrations can also be used to identify the need for fate studies of compounds of interest in biosolids. Integr Environ Assess Manag 2012; 8: 530–542. © SETAC


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. OVERVIEW OF THE PHATE MODEL
  5. MODEL CORROBORATION
  6. API SIMULATIONS
  7. CONCLUSIONS
  8. Acknowledgements
  9. REFERENCES

Many compounds, including pharmaceuticals, are not completely transformed in wastewater treatment plants (WWTPs) and remain in biosolids that may be applied to land as a soil amendment (Kinney et al. 2006; USEPA 2009; Walters et al. 2010). This practice leads to concerns about potential exposures of people who may come into contact with amended soils and also about potential effects to plants and animals in the terrestrial compartment (Clarke and Smith 2011; Higgins et al. 2010; Norwegian Scientific Committee for Food Safety (VKM) 2009; Smith 2009). In addition, the application of sewage sludge as a soil amendment may contribute to increased loadings of trace organics to groundwater and surface water (Golet et al. 2003).

Prompted by these concerns, the PhATE™ (pharmaceutical assessment and transport evaluation) model was expanded to predict concentrations of active pharmaceutical ingredients (APIs) in the sludges and biosolids generated by WWTPs. The PhATE model is available on request to Pharmaceutical Research and Manufacturers of America (MGarvin@phrma.org). In this article, we use the term “sludge” to refer to untreated solids resulting from wastewater treatment and “biosolids” to refer to treated sludge. The original version of the PhATE model estimates concentrations of APIs in surface waters of the United States that result from patient use (or consumption) of medicines (Anderson et al. 2004). The model has been used in a number of case studies (Cunningham et al. 2009, 2010; Hannah et al. 2009; Robinson et al. 2007; Schwab et al. 2005) supporting its utility for estimating predicted environmental concentrations (PECs) in surface waters. Overall, the model uses a mass balance approach to generate PECs in WWTPs in 12 watersheds selected to be representative of the range of hydrologic regions present in the United States. These watersheds incorporate 2712 stream segments that include 44 398 kilometers of streams and 1302 WWTPs. Of these WWTPs, 888 (68%) have information regarding biosolids processing (aerobic, anaerobic, composting, or no treatment) and were used in the present analysis. PhATE estimates the mass of API that enters each WWTP based on the per capita use of a product and the population served by the treatment plant. It also takes into account any nonreversible metabolism that may occur in patients before discharge. For the purpose of the model evaluation reported here, the total annual masses of triclosan (TCS) and linear alkylbenzene sulfonates (LAS) used in the United States were the same as reported in Anderson et al. (2004). Estimates of the total annual masses of APIs sold in the United States for July 2008 through August 2009 were obtained from IMS Health Inc. and used in the model runs (IMS Health 2009). The mass input values reflect sales of all products containing the API (including combination products) to retail pharmacies, nonfederal hospitals, federal facilities, long-term care facilities, clinics, and health maintenance organizations (HMOs). The annual masses were not adjusted for drug product that was sold but not used. By using the total mass of API sold, the analysis assumes that all of the drug product sold is consumed by patients and is available for excretion (i.e., after accounting for metabolism) to wastewater. This approach takes into account unused medicine disposed of down the drain but not any disposed of through solid waste.

The model uses this information to estimate concentrations in influents, effluents, sludge, and biosolids. Because the watersheds included in PhATE are representative of the range of per capita flow, populations served and treatment types used throughout the United States, the predicted concentrations of APIs in WWTP influent, effluent, and biosolids represent concentrations of APIs that would be expected to be present throughout the country. Although there are other excellent models (e.g., Bock et al. 2010) that deal with predicting concentrations in biosolids in wastewater treatment plants, PhATE attempts to model a much larger spatial area, focusing on distribution of concentrations in biosolids in many treatment plants across the United States.

Two primary objectives of the PhATE biosolids module are to determine the potential mass and concentration of API adsorbed to and retained in the sludge generated by WWTPs, and estimate the PECs of APIs in biosolids leaving WWTPs, including the potential loss of APIs during the handling of sludge and its “conversion” to biosolids.

The analysis presented in this article corroborates the module by comparing PECs generated by PhATE with biosolids' measured environmental concentrations (MECs) reported in the literature for US WWTPs.

OVERVIEW OF THE PHATE MODEL

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. OVERVIEW OF THE PHATE MODEL
  5. MODEL CORROBORATION
  6. API SIMULATIONS
  7. CONCLUSIONS
  8. Acknowledgements
  9. REFERENCES

Model context

Figure 1 shows a schematic of the methodology used by PhATE for estimating PECs. Previous versions of the model focused on the aqueous environment, providing PECs for surface and drinking waters. The new version of PhATE (v 4.0) incorporates an additional module to estimate PECs in sludge and biosolids. API loss can occur during sludge treatment processes such as digestion or composting. The loss of API during sludge treatment may be estimated in PhATE as a fixed loss fraction or as a first-order loss term. Sludge treatment processes will also reduce the concentration of dry solids in the sewage sludge. As a result, those APIs that are not efficiently removed by the treatment process and whose relative mass reduction is less than the relative mass reduction of dry solids can have higher concentrations in biosolids (µg/kg dry solids) than in sludge. It should be noted that dilution of the sludges or biosolids with other components to improve porosity has not been taken into account. Such processing would be expected to lower concentrations, from both straight mass dilution and improved aerobic degradation.

thumbnail image

Figure 1. PhATE schematic describes the partitioning and degradation of API within the WWTP.

Download figure to PowerPoint

Biosolids from a WWTP can be disposed of in several ways in compliance with US Environmental Protection Agency (USEPA) regulations (40 CFR Part 503, 2003) and Standards for the Use or Disposal of Sewage Sludge (USEPA 2003). These standards consist of general requirements, pollutant limits, management practices, and operational standards for the final use or disposal of sewage sludge generated during the treatment of domestic sewage.

The Part 503 Rule describes several beneficial uses of WWTP sewage sludge including application to land as a fertilizer (e.g., agriculture, silviculture for forest management, gardens, parks, and golf courses), or as a soil amendment for reclamation areas (e.g., strip mines). API loss can occur during these different practices, but will vary by API and will depend on the API concentration, the land-use type, the frequency of application, and the number of years of application. PhATE generates PECs in amended soils for the land application types described above; however, this article only discusses the use of PhATE to estimate PECs in WWTP wastewater, effluent, sludges, and biosolids.

Model input

Many model inputs required to estimate PECs for biosolids are the same as those described in Anderson et al. (2004) including: human use (kg/year), human loss (nonreversible metabolism), and potential loss of the compound from the aqueous phase during primary and secondary wastewater treatment. Note that many APIs are metabolized by conjugation with glucuronic acid or sulfate to form highly water-soluble compounds that are more easily excreted than the parent. These conjugates are susceptible to hydrolysis by WWTP microorganisms, regenerating the parent compound (Belfroid et al. 1999; Desbrow et al. 1998), and must be considered as the nonconjugated parent to establish a mass balance. The following subsections summarize model inputs required in PhATE to estimate PECs for biosolids.

Fraction sorbed on primary and combined primary–secondary sludge

For the PhATE biosolids module, key input parameters are the fractions of the compound sorbed on primary and combined primary–secondary sludge. If these parameters are not available from laboratory or field data, they may be estimated from the sludge–water partition coefficients (Kp) and the first-order degradation rate constant (k, per hour). The rate constant, k, is a composite rate that is calculated as the sum of all relevant loss rate constants that might be associated with aqueous phase loss mechanisms (e.g., biodegradation, hydrolysis, evaporation). SimpleTreat 3.1 (Struijs 1996; Struijs et al. 1991) is a useful model to predict these PhATE input parameters for neutral compounds. Once the Kp and k values have been determined, by measurement or estimation, the values can be entered into SimpleTreat 3.1 to calculate the fraction sorbed on primary and secondary sludge.

SimpleTreat 3.1 uses empirical equations to estimate the Kp based on the compound's octanol–water partition coefficient (KOW). For ionic compounds, more typical of pharmaceuticals, the SimpleTreat equations do not always adequately estimate the Kp values, especially for compounds with multiple functional groups, such as zwitterions. An alternative estimation of Kp, at least for secondary sludges, may be made using the equations discussed in Cunningham (2008). This Kp estimation method is included as a link in the PhATE model, as is a link to SimpleTreat 3.1.

SimpleTreat 3.1 allows the k value to be entered in 1 of 3 ways: estimated from Organisation for Economic Co-operation and Development/European Union (OECD/EU) standardized biodegradability tests; estimated from degradation results from an activated sludge batch test according to first order kinetics with respect to the slurry phase, implying biodegradation in both aqueous and sorbed phases; and estimated from biodegradation results according to Monod kinetics (Metcalf and Eddy 1979) As noted above, where other first-order degradation mechanisms (e.g., hydrolysis) are known, those can be incorporated into a composite k value.

It should be noted that SimpleTreat 3.1, although allowing degradation to take place in the activated sludge process (i.e., secondary treatment), does not take degradation into account for the material sorbed to sludge during primary treatment. Given that most compounds that sorb to solids to any extent will be predominantly removed during primary treatment, this could lead to an overestimation of the concentration on the resulting sludge or biosolids.

Biosolids depletion mechanisms

In calculating concentrations of trace organics in the biosolids, PhATE takes into account the removal of any material that is sorbed to sludge during treatment of the sludge to create biosolids. The model provides for 4 sludge treatment types including: no treatment, aerobic digestion, anaerobic digestion, and composting. Each WWTP is assigned 1 of these 4 sludge treatment types based on information reported in the 2008 Clean Water Needs Survey (USEPA 2008) or from alternate sources (e.g., telephone contact with the WWTP). Of the 888 WWTPs included in the biosolids module, 446 use anaerobic digestion, 307 use aerobic digestion, 21 use composting, and 114 do not treat the sludge.

Degradation during treatment of sludge can be entered into the model as no loss, fixed loss fractions, or as first-order rate constants (1/d). In the analysis reported here, experimental data were used when available from the literature.

Predicted environmental concentrations

Aqueous phase of sludge and biosolids

Another important aspect of sludge and biosolids is the amount of water in the solids matrix. When considering the actual sorption behavior of a compound, calculations are carried out using the sludge–water partitioning coefficient, referred to as Kp as discussed above. However, this results in a concentration on just the solid phase of the biosolids themselves, and neglects the water component. For compounds with an appreciable Kp, on the order of 100 or greater, the contribution of the aqueous component was assumed to be negligible. However, for compounds that are not highly sorbed (Kp < ≈100), the mass of a compound in the aqueous phase of sludge and biosolids can be considerable, particularly if the compound is used in high volume and is relatively highly water soluble (e.g., acetaminophen). PECs generated by PhATE include the total mass present in both solid and aqueous phases and, thus, represent the total amount of the API that could be introduced into the environment through land application. Note that analytical results of biosolids samples should be interpreted with care. If samples were prepared by techniques such as lyophilization followed by extraction, or by extracting the entire sample with solvent before analysis, the compounds present in both the sorbed and dissolved materials will be measured. However, if samples were physically dewatered before extraction (e.g., by centrifugation), some of the compounds in the aqueous phase may be lost. Because applied biosolids include both the solid and aqueous phases, analytical methods should account for both phases.

Concentration in aqueous streams

PhATE calculates the API mass loading entering a WWTP using the average per capita usage of the API, loss of API resulting from human use (e.g., nonreversible metabolism), and the population served by the WWTP. The average per capita usage is calculated using Equation 1 and the mass loading of API that enters a WWTP is calculated using Equation 2. The WWTP influent concentration is calculated from the influent load using Equation 3. All equations are provided in Table 1.

Table 1. Equations
  1. CF1 = 1 000 000 mg/kg; CF2= 365 day/y; CF3 = 1000 µg/mg; Cin = total WWTP influent concentration (mg/L);equation image = aqueous phase effluent concentration (mg/L); Cs = concentration of API sorbed to sludge (µg/kg); Cs(a) = concentration of API in the aqueous phase of sludge, normalized to the concentration of dry solids (µg/kg); Cs(t) = total concentration of API, including API sorbed to the sludge and API in the aqueous phase normalized to the concentration of dry solids (µg/kg); D = hydraulic retention time during sludge treatment (days); f = reduction in solid mass during sludge treatment (0.38 for treated sludge or 0.0 for untreated sludge); ks = first order loss rate constant (1/d); La = fixed loss of API from the aqueous phase during wastewater treatment (fraction) due to degradation and sorption; Lm = human loss of API due to nonreversible metabolism (fraction); Ls = fixed loss of API from solids during sludge treatment (fraction); Min = mass loading into a WWTP (mg/day); PUS = US population; Pwtp = population served by the WWTP; Qin = influent flow rate (L/day); qs = quantity of dry sludge solids produced per volume of wastewater treated (kg/L), 0.000135 for primary and 0.000220 for secondary WWTPs (Tchobanolous and Burton 1991); S = API removed via sorption to sludge (fraction); Sales = annual US sales of the API (kg/y); TSS = total suspended solids in sludge (0.037 kg/L for untreated; 0.06 kg/L for aerobically and anaerobically digested; 0.4 kg/L for composted sludge [Tchobanolous and Burton 1991]); U = average per capita usage in the United States (mg/capita/y).

  • equation image(1)
  • equation image(2)
  • equation image(3)
  • equation image(4)
  • equation image(5)
  • equation image(6)
  • equation image(7)
  • equation image(8)
  • equation image(9)

PhATE assumes there are no regional differences in per capita use across the United States and that all API excreted from human use is treated in a WWTP. The assumption of uniform use was examined in Anderson et al. (2004) and found to be realistic.

Losses that may occur within the WWTP are specific to both the compound and the treatment type (Anderson et al. 2004). PhATE provides for 7 wastewater treatment types: no discharge, raw discharge (i.e., no treatment), primary, advanced primary, secondary, advanced secondary I, and advanced secondary II (USEPA 2008). Each of the 1302 WWTPs in PhATE is assigned 1 of these wastewater treatment types based on information reported in the 2008 Clean Water Needs Survey (USEPA 2008) or from alternate sources, e.g., telephone contact with the WWTP. As noted above, information on sludge treatment could only be obtained for 888 of the 1302 WWTPs in the PhATE model. Therefore, only results from these WWTPs are included in this analysis.

PhATE considers potential WWTP losses from the aqueous phase via degradation (e.g., biodegradation, hydrolysis, etc.) and/or sorption to sludge during wastewater treatment. The mass loading from a WWTP to a river segment is calculated using Equation 4. If the API loss in sludge treatment is estimated with a fixed fractional loss, the concentration of API in the aqueous phase of the sludge is calculated using Equation 5 by applying the sludge treatment removal efficiency and normalizing by the concentration of dry solids in the sludge. If the API loss in sludge treatment is estimated with a first order loss rate, Equation 6 is used to calculate the API concentration in the aqueous phase of the sludge.

Concentration in sludge and biosolids

The model allows the user to specify unique loss terms (i.e., fixed fraction or first order loss rate) for aerobic digestion, anaerobic digestion, and composting sludge treatment types. No loss is assumed for those WWTPs categorized as having no sludge treatment. If the API loss in sludge treatment is estimated with a fixed fractional loss, the concentration of API in sludge is calculated using Equation 7. If the API loss in sludge treatment is estimated with a first order loss rate, the concentration of API in sludge is calculated using Equation 8. The total API concentration in sludge, including sorbed API and API in the aqueous phase, is calculated using Equation 9.

MODEL CORROBORATION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. OVERVIEW OF THE PHATE MODEL
  5. MODEL CORROBORATION
  6. API SIMULATIONS
  7. CONCLUSIONS
  8. Acknowledgements
  9. REFERENCES

For this article, the term corroboration is used to discuss evaluation of the model in terms of independent experimentally determined data. It recognizes that a model can never be shown to be absolutely free of flaws (validated) or true (verified), but that confidence in the ability of the model to make useful predictions can be based on a variety of comparisons of model predictions with available data under a range of conditions (Oreskes et al. 1994). We corroborated PhATE by comparing PECs generated by the model to MECs reported by other researchers. Although literature data from WWTPs in other countries are available, only data from US WWTPs were used in the comparisons because PhATE is a model based specifically on the United States. Given that PhATE was developed as nationwide screening level model, we consider agreement of PECs and MECs within an order of magnitude as demonstrating successful corroboration.

Use of surrogate corroboration compounds

TCS and LAS were used as corroboration compounds because a considerable amount of measured physical property data and measured mass balance data from WWTPs in the United States, including sludge treatment processes, were available in the literature and from regulatory sources. Although not APIs, some of their chemical and/or physical properties are similar to those of APIs. Additionally, like APIs, their primary entry point to the environment is via WWTPs after human use, specifically “down the drain.” These similarities make quantitative comparisons of PECs and MECs valuable for corroboration of the PhATE biosolids module.

Methodology for comparing model results with field data

PhATE model inputs for TCS and LAS are summarized in Table 2. Total mass and concentration of a compound were estimated for solids in each WWTP and combined across WWTPs by sludge treatment type (i.e., untreated sludge, sludge digested aerobically, sludge digested anaerobically, and sludge treated by composting). Each set of PECs was analyzed separately. In addition, a composite data set, consisting of the sludge or biosolids PECs from each WWTP (after no treatment or final treatment, as applicable) was analyzed. This was to allow more appropriate comparison with some of the experimental data sets.

Table 2. Model input parameters for corroboration compounds
ParameterTCSLAS
  • LAS = linear alkylbenzene sulfonates; NA = not applicable; TCS = triclosan.

  • a

    Data taken from Anderson et al. (2004).

  • b

    Anderson et al. (2004) reported 314,000,000. However, Feijtel et al. (1995) reported 40%–60% loss in the sewer before the WWTP. A correction of 0.5 was used to calculate input volume.

  • c

    Data taken from Federle et al. (2002).

  • d

    Data taken from HERA (2009), which provided composite removal data for EU exposure calculations after the TGD and EUSES.

  • e

    Calculated from results provided in McAvoy et al. (2002).

  • f

    Calculated from McAvoy et al. (1993).

  • g

    Data taken from McAvoy et al. (2002).

  • h

    Data taken from Mogensen et al. (2003).

Volume, kg/y600 000a157 000 000b
Fraction removed via primary sorption0.365cNA
Fraction removed via secondary sorption0.021cNA
Fraction removed via biodegradation during activated sludge process0.534c0.79d
Total fraction removed from aqueous phase0.92c0.99d
Total fraction removed via sludge0.386c0.2d
Fraction removed during aerobic digestion0.9153e0.98f
Fraction removed during anaerobic digestion0g0h

Comparison of measured versus modeled concentrations for corroboration compounds

For each of the predicted and measured data sets, the 5th percentile, the 25th percentile or first quartile (q1), the 50th percentile or median, the 75th percentile or third quartile (q3), and the 95th percentile were derived. An exception to this was for LAS experimental data for which these values were calculated using the reported means and standard deviations, assuming a normal distribution. For the purpose of comparing PECs with MECs, the median or 50th percentile values were used, except for LAS, where the mean value was used as an estimate of the median. Under the assumption of a normal distribution, the mean and the median values would be the same.

Comparison of the median MECs with median PECs for TCS shows excellent agreement (within a factor of 2) (Figure 2). The box plot labeled “EPA Final Sludge” summarizes data from all of the samples analyzed in the USEPA Targeted National Sewage Sludge Survey (USEPA 2009). They are considered the “final” sludge product from the WWTP samples, and represent a variety of sludge treatments. The box plots labeled “Literature” represent statistics from a compilation of data reported by several studies (Cha and Cupples 2009; Lozano et al. 2010; McAvoy et al. 2002; Xia et al. 2010).

thumbnail image

Figure 2. Comparison of PhATE output and measured concentration data for TCS.

Download figure to PowerPoint

LAS was not part of the USEPA survey, so only literature data were available (McAvoy et al. 1993). These were reported as percentiles by assuming a normal distribution and using the reported means and standard deviations from samples from anaerobic and aerobic digestion processes (Figure 3). The results again show excellent agreement (within a factor of 4) between the median values predicted by PhATE and those reported in the literature (Figure 3).

thumbnail image

Figure 3. Comparison of PhATE output and measured concentration data for LAS.

Download figure to PowerPoint

API SIMULATIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. OVERVIEW OF THE PHATE MODEL
  5. MODEL CORROBORATION
  6. API SIMULATIONS
  7. CONCLUSIONS
  8. Acknowledgements
  9. REFERENCES

Selection of APIs

The USEPA Survey, a comprehensive survey where USEPA reported 84 measured concentrations of APIs in sewage sludge at 74 publicly owned treatment works in the United States (USEPA 2009), was used as a source of MECs for comparison with PECs generated by PhATE. From the data set of compounds studied by USEPA, 18 APIs were selected for comparison with model PECs (Table 3). These APIs were selected to represent a broad range of conditions, including: compounds detected at high concentrations and compounds detected at low concentrations, compounds that were frequently detected and some that were not detected, and compounds with comprehensive fate data and some with limited fate data.

Table 3. US sales and metabolism loss for APIs
CompoundUS salesMetabolism loss
(kg/y)Study reference(%)Study reference
  1. APIs = active pharmaceutical ingredients.

17α-Ethinyl estradiol78IMS Health 200950.0Hannah et al. 2009
Acetaminophen11 199 552IMS Health 200910.0PDR 2004
Albuterol4041IMS Health 200972.0Cunningham et al. 2009
Azithromycin82 989IMS Health 200944.0

Page et al. 2001

; Pfizer 2009

Carbamazepine127 790IMS Health 200983.4Cunningham et al. 2010
Cimetidine36 546IMS Health 200952.0PDR 2004
Ciprofloxacin221 566IMS Health 200915.0PDR 2004
Digoxin97IMS Health 200916.0PDR 2004
Diltiazem172 036IMS Health 200996.0PDR 2004
Erythromycin31 555IMS Health 200975.0

Burrows 1980

; Griffith and Black 1970

Fluoxetine25 711IMS Health 200990.0Anderson et al. 2004
Gemfibrozil327 636IMS Health 200924.0Anderson et al. 2004
Ibuprofen2 738 386IMS Health 200985.0PDR 2004
Metformin3 857 747IMS Health 20090.0Anderson et al. 2004
Norethindrone706IMS Health 200995.0Anderson et al. 2002
Ranitidine231 812IMS Health 20096.0Anderson et al. 2004
Tetracycline49 693IMS Health 200940.0

Barr et al. 1972

; Meyer et al. 1974

; Jaffe et al. 1973

; Tuomisto and Mannisto 1973

Trimethoprim76 437IMS Health 200920.0PDR 2004

Model input parameters for APIs

Model input parameters for the APIs are presented in Tables 3 and 4. The annual use for each API reflects US sales of all products containing the API (including combination products) to retail pharmacies, nonfederal hospitals, federal facilities, long-term care facilities, clinics, etc. (Table 3). The sales volumes also include estimates of API mass from the sale of over-the-counter drug products. The analysis assumes all of the API sold is consumed by patients and is available for excretion after accounting for nonreversible metabolism. Human metabolism data reported in Table 3 assume that conjugated metabolites are deconjugated in the environment and converted to the active parent.

Table 4. PhATE input data for APIs
CompoundKp 1° sludgeKp 2° sludgeRemoval from aqueous phase via sorption (%)Removal from aqueous phase via sorption plus other mechanisms (%)Removal from solid phase via treatment (%)
1° only1°/2°1° only1°/2°Aerobic digestionCompostAnaerobic digestion
  • APIs = active pharmaceutical ingredients; NA = data not available (value of zero was used as a conservative assumption [i.e., no loss]); 1° = primary treatment; 2° = secondary treatment; 3° = tertiary treatment.

  • a

    Data taken from Anderson et al. (2004).

  • b

    Data taken from Radjenovic et al. (2009).

  • c

    Calculated. Radjenovic et al. (2009) reported higher value, but could be due to high biodegradation.

  • d

    Data taken from Cunningham et al. (2009).

  • e

    Based on Log Dow = 1.68 (Cunningham et al. 2010) and equation log Kp (neutral) = 0.7125 * log Dow + 0.9835 (V.L. Cunningham, Sustainability Sciences LLC, Hatboro, PA).

  • f

    Based on Dow = 23.1 (Cunningham 2008).

  • g

    Based on Dow = 61.9 (Cunningham 2008).

  • h

    Based on log Dow = −1.09 (Cunningham 2008).

  • i

    Calculated with SimpleTreat V.3.1.

  • j

    Calculated with SimpleTreat V.3.1, biodegradation rate = 3.0/h (Joss et al. 2006).

  • k

    Primary removal is mean of 2 studies (Gobel et al. 2007; Yasojima et al. 2006), assume all sorption occurs in primary treatment; secondary removal from Yasojima et al. (2006).

  • l

    Calculated with SimpleTreat V.3.1, biodegradation rate = 0.01/h (Cunningham et al. 2009).

  • m

    Primary removal is mean of 2 studies (Lindberg et al. 2006; Golet et al. 2003), assume all sorption occurs in primary treatment; secondary removal is median of 7 studies (Terzic et al. 2008).

  • n

    Calculated with SimpleTreat V.3.1, biodegradation rate = 0.0035/h (Cunningham et al. 2009).

  • o

    Calculated with SimpleTreat V.3.1, hydrolysis half-life = 1 day at pH 7.3 (M. Buzby, Merck & Company, Whitehouse Station, NJ).

  • p

    Calculated with SimpleTreat V.3.1, biodegradation rate = 1.0/h (Joss et al. 2006).

  • q

    Calculated with SimpleTreat V.3.1, biodegradation rate = 0.2/h (Trautwein and Kummerer 2009).

  • r

    Primary removal is from Gulkowska et al. (2008), assume all sorption occurs in primary treatment; secondary removal is median from 3 studies (Gulkowska et al. 2008).

  • s

    Primary removal assumed to be via sorption only; secondary removal from Caldwell et al. (2010).

  • t

    Primary removal assumed to be via sorption only, secondary removal is median from 3 studies (Radjenovic et al. 2009) and tertiary removal is median of 5 studies (Jacobsen et al. 2004).

  • u

    No literature data available, assumed total removal equal to removal via sorption.

  • v

    Primary removal assumed to be via sorption only and secondary removal from Cunningham et al. (2010).

  • w

    Primary removal assumed to be via sorption only and secondary removal from Choi et al. (2008).

  • x

    Literature studies indicate minimal removal (Gulkowska et al. 2008), assume removal is from sorption only.

  • y

    Primary removal assumed to be via sorption only and secondary removal from Vasskog et al. (2008).

  • z

    Primary removal assumed to be via sorption only and secondary removal is median from 7 studies (Stumpf et al. 1999).

  • aa

    Primary removal assumed to be via sorption only, secondary removal is median from 20 studies (Paxéus 2004), and tertiary removal is median of 18 studies (Joss et al. 2005).

  • bb

    Primary removal assumed to be via sorption only and secondary removal is from Scheurer et al. (2009).

  • cc

    Primary removal assumed to be via sorption only and secondary removal is mean from Terzic et al. (2008) and Castiglioni et al. (2006).

  • dd

    Primary removal assumed to be via sorption only and secondary removal is median from 11 studies (Choi et al. 2008).

  • ee

    Removal efficiency in aerobic digestion and composting estimated from nonsorption removal in secondary treatment.

  • ff

    Estimated from data provided by Radjenovic et al. (2009).

17α-ethinyl estradiol5250a5250a46.862.9i46.884.0s2121eeNA
Acetaminophen6.7b21c0.20.2j0.298.4t9898ee80ff
Albuterol4.0d4.0d0.10.2i0.10.2uNANANA
Azithromycin  29.029.0k29.049.0k2020ee80ff
Carbamazepine151e151e4.25.6i4.29.0v33ee20ff
Cimetidine501d501d12.315.6l12.335.0w1919eeNA
Ciprofloxacin  47.047.0m47.077.0m3030eeNA
Digoxin60d60d1.82.3n1.82.3nNANANA
Diltiazem1042f1042f21.327.1o21.339.3o1212eeNA
Erythromycin309b74b8.18.8i8.18.8xNANA50ff
Fluoxetine1884g1884g30.640.6i30.655.0y1414ee6ff
Gemfibrozil23b19b0.70.9i0.767.0z6666eeNA
Ibuprofen9.5b0.0b0.30.3p0.687.0aa8787ee50ff
Metformin31d31d0.91.0q0.989.3bb8888eeNA
Norethindrone1320a1320a24.832.9i24.832.9iNANANA
Ranitidine35.1h35.1h1.01.4i1.030.0cc2929eeNA
Tetracycline  8.08.0r8.072.0r6464eeNA
Trimethoprim75.9d75.9d2.22.9i2.221.0dd1818ee80ff

Sorption of API to WWTP sludge was estimated with SimpleTreat 3.1 using partition coefficients (Kp) and degradation rate constants (i.e., biodegradation or hydrolysis) reported in the literature (Table 4). Removal efficiencies in primary, secondary, and tertiary treatment were based on WWTP data reported in the literature when available (Table 4). If measured WWTP data were not available, then removal efficiencies were estimated with SimpleTreat 3.1 (Table 4). Limited data were available to estimate the removal of API during sludge treatment. Removal of APIs via anaerobic digestion was estimated from Radjenovic et al. (2009). No measured data on removal efficiencies for APIs via aerobic digestion or composting were available; therefore, these removal efficiencies were estimated by using the same removal efficiencies reported for aerobic activated sludge systems (Table 4).

Comparison of measured versus modeled concentrations of APIs

The 18 APIs were divided into 2 categories based on the detection rate reported in the USEPA Survey: 1) APIs with a relatively high frequency of detection (i.e., >25%), and 2) APIs with a relatively low frequency of detection (i.e., <25%). The first category contains APIs with a substantial number of detects (24–83 of 84 samples), allowing for a meaningful comparison of PECs to MECs. The low frequency of detection category contains very few detects (0–6 of 84), limiting comparison of PECs to MECs.

For 9 of 12 APIs with a high frequency of detection (i.e., azithromycin, carbamazepine, cimetidine, ciprofloxacin, erythromycin, fluoxetine, gemfibrozil, ibuprofen, and tetracycline), the ratio of the median PEC to the median MEC ranged from 0.29 to 2.5 indicating very good agreement between the PECs from PhATE and MECs (Table 5 and Figure 4). The median PEC was greater than the median MEC by more than 1 order of magnitude for the remaining 3 APIs (diltiazem, ranitidine, and trimethoprim). The ratio of the median PEC to the median MEC ranged from 11 for diltiazem to 22 for ranitidine (Table 5). PEC–MEC ratios of >10 may indicate the existence of depletion mechanisms for these APIs that are not considered by this analysis. For example, information on the removal of diltiazem or ranitidine in anaerobic sludge treatment processes was not available so a removal efficiency of zero was assumed for the model analysis for these 2 APIs (Table 4).

Table 5. Comparison of PECs with MECs for APIs with a high frequency of detection
CompoundMDLTotal samples (n)Detects (n)Frequency of detection (%)Median MEC (µg/kg)Median PEC (µg/kg)Ratio PEC–MEC
  1. APIs = active pharmaceutical ingredients; MECs = measured environmental concentrations; PEC = predicted environmental concentrations; MDL = method detection limit (µg/kg).

PECs in agreement with MECs
 Azithromycin108480952414581.9
 Carbamazepine1084809538.995.62.5
 Cimetidine48474881722211.3
 Ciprofloxacin358383100588066901.1
 Erythromycin284779219.742.62.2
 Fluoxetine1084799415683.50.53
 Gemfibrozil108476901021431.4
 Ibuprofen10084546414341.60.29
 Tetracycline408481965832100.36
PECs are greater than MECs
 Diltiazem284698213.915511
 Ranitidine484425012.026122
 Trimethoprim1084242911.016715
thumbnail image

Figure 4. APIs in agreement with PhATE PECs.

Download figure to PowerPoint

For the 6 APIs with a low frequency of detection, PECs are considered consistent with MECs if the median PEC is less than or equal to the method reporting limit (MRL) for the API. This is the case for 5 of these 6 APIs (17α-ethinyl estradiol, acetaminophen, albuterol, digoxin, and norethindrone, Table 6). For these 5 APIs, PhATE is capable of providing estimates of environmental concentrations even though available analytical methods are not sufficiently sensitive to detect these APIs in biosolids (i.e., actual environmental concentrations are below analytical reporting limits). For metformin, the median PEC of 2415 µg/kg is greater than the median reporting limit of 542 µg/kg. Thus, the PhATE model would suggest that metformin should be detected at higher frequencies than reported in the USEPA Survey or that an unknown depletion mechanism may exist. The earlier corroboration of the surface water portion of the PhATE model also reported PECs greater than MECs for metformin, providing further evidence of the potential existence of an unknown depletion mechanism.

Table 6. Comparison of PECs with MECs for APIs with a low frequency of detection
CompoundMDLTotal samples (n)Detects (n)Frequency of detection (%)Median reporting limit (µg/kg)Median PEC (µg/kg)
  1. APIs = active pharmaceutical ingredients; MECs = measured environmental concentrations; PEC = predicted environmental concentrations; MDL = method detection limit (µg/kg).

17α-Ethinyl estradiol21840025.41.95
Acetaminophen4008422397265
Albuterol284115.000.446
Digoxin10084001010.182
Metformin20084675422035
Norethindrone21845622.41.02

CONCLUSIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. OVERVIEW OF THE PHATE MODEL
  5. MODEL CORROBORATION
  6. API SIMULATIONS
  7. CONCLUSIONS
  8. Acknowledgements
  9. REFERENCES

The results of the model corroboration with TCS and LAS, as well as the good agreement of median PECs and MECs for the majority of APIs included in the evaluation, demonstrate the ability of the PhATE model to predict screening level concentrations of personal care products and pharmaceutical compounds in biosolids. This is an essential first step toward assessing potential environmental impact from soils amended with biosolids. Importantly, in those instances where PECs and MECs were not in good agreement (i.e., diltiazem, ranitidine, trimethoprim, metformin), PECs were greater than MECs meaning that if PhATE was used to estimate potential risks associated with concentrations of APIs in biosolids, the estimates of potential risk would be biased high rather than low. In other words, the predicted concentrations are conservative and would tend to over predict rather than under predict potential risks, as is appropriate for a screening level model.

One might ask whether the 18 corroboration APIs are representative of a broad range of compounds or whether these results may be limited because APIs with a fairly narrow range of physical and chemical characteristics were used to corroborate the biosolids module. A data domain analysis shows that the 18 APIs selected from the EPA data set provide very good coverage of the range of possible model inputs used to estimate PECs (Table 7). Detection frequency ranged from 0% to 100%. Human use ranged from 78 to >11 000 000 kg/y, metabolism from 0% to 95%, sorption from 0.1% to 62.9%, and total WWTP removal from 0.2% to 99.7%. As long as the characteristics and inputs for other APIs that may be modeled using PhATE are close to the ranges used in the corroboration data set, relatively high confidence exists that actual concentrations of APIs in biosolids will be similar to or lower than the PECs generated by PhATE.

Table 7. Summary of various characteristics and attributes of APIs
CompoundDetection frequency (%)MDL µg/kgSales kg/yMetabolism loss (%)Kp secondaryRemoval from aqueous phase via sorptionRemoval from aqueous phase via combined mechanismsRemoval from solid phase via solids treatment (%)
PrimaryPrimary and secondaryPrimaryPrimary and secondaryPrimary, secondary, and tertiaryAnaerobic digestion
  1. APIs = active pharmaceutical ingredients; MECs = measured environmental concentrations; MRL = method reporting limit; NA = data not available; PEC = predicted environmental concentrations; MDL = method detection limit (µg/kg).

PEC ∼ MEC
 Ciprofloxacin10035221 56615NA4747477777 
 Tetracycline964049 69340NA8887272 
 Carbamazepine9510127 79083.41514.25.64.29920
 Azithromycin951082 98944NA292929494980
 Fluoxetine941025 71190188430.640.630.655556
 Erythromycin92231 55575748.18.88.18.88.850
 Gemfibrozil9010327 63624190.70.90.76767 
 Cimetidine88436 5465250112.315.612.33535 
 Ibuprofen641002 738 3868500.30.30.68795.550
PEC > MEC
 Diltiazem822172 03696104221.327.121.339.339.3 
 Ranitidine504231 812635.111.413030 
 Trimethoprim291076 4372075.92.22.92.2212180
PEC > MRL
 Metformin71003 857 7470310.910.989.389.3 
PEC < MRL
 Norethindrone62170695132024.832.924.832.932.9 
 Acetaminophen240011 199 55210210.20.210.298.499.780
 Albuterol1240417240.10.20.10.20.2 
 Digoxin01009716601.82.31.82.32.3 
 EE20217850525046.862.946.88484 

Further analysis shows that there was good agreement between PECs and MECs for compounds with high detection frequencies and relatively high annual human use (Table 7). For all compounds with a moderate to high detection frequency (29% to 100%), the PECs and MECs either were in agreement or the PECs were greater than the MECs, which is desirable to conservatively estimate environmental concentrations. For compounds with low annual usage or very high biodegradation removals, the PECs were less than the MRLs. For those compounds where the PECs were greater than the MECs or the MRLs, more occurrence measurements or characterization of possible depletion or removal mechanisms would be expected to result in better agreement. Nonetheless, from a risk prioritization perspective, PhATE can be a valuable tool to generate PECs for APIs without the constraints of analytical method detection limits or low usage levels.

The results also point to the need for fate data in biosolids processing. For compounds that biodegrade aerobically or anaerobically, aerobic, anaerobic and composting biosolids treatments may represent important depletion mechanisms. In addition, data on depletion mechanisms in soils will be needed to run and corroborate the soil portion of the model. Knowledge of these mechanisms will be essential if risk assessments are to use representative concentration estimates. For example, 3 of 4 APIs for which PhATE overpredicted PECs, diltiazem, ranitidine and metformin, had no depletion information for anaerobic digestion (Table 4). The overprediction may be the result of depletion mechanisms that are not characterized using the existing methodology for fate testing.

Finally, evaluations under different processing scenarios are possible, allowing for exploration of potential mitigation effects, such as sludge segregation, improved wastewater and/or sludge treatment, alternative composting methods, etc. Because PhATE incorporates WWTPs with different technologies and servicing different populations, the model can provide a more realistic characterization of the range of concentrations in biosolids, and ultimately in amended soils, rather than presenting only average or worse-case conditions. These concentrations are better suited for assessing environmental exposure for the purpose of conducting human health or environmental risk assessments.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. OVERVIEW OF THE PHATE MODEL
  5. MODEL CORROBORATION
  6. API SIMULATIONS
  7. CONCLUSIONS
  8. Acknowledgements
  9. REFERENCES

The authors acknowledge Petr Masopust of AMEC Earth and Environmental and Don Deluca of ARCADIS for their important contributions. The technical assistance of Randy Watts of Merck and Andrew Riddle formerly of Astra-Zeneca are also gratefully acknowledged. The Pharmaceutical Research and Manufacturers of America provided financial support for the continued development of the biosolids module in the PhATE model and for the preparation of this manuscript. Additional support was provided by the Pharmaceutical EHS Sustainability Council.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. OVERVIEW OF THE PHATE MODEL
  5. MODEL CORROBORATION
  6. API SIMULATIONS
  7. CONCLUSIONS
  8. Acknowledgements
  9. REFERENCES
  • Anderson PD, D'Aco VJ, Shanahan P, Chapra SC, Buzby ME, Cunningham VL, DuPlessie BM, Hayes EP, Mastrocco FJ, Parke NJ. 2004. Screening analysis of human pharmaceutical compounds in U.S. surface waters. Environ Sci Technol 38: 838849.
  • Anderson PO, Knoben J, Troutman W. 2002. Handbook of clinical drug data. 10th ed. New York: McGraw-Hill. 1148 p.
  • Barr WH, Gerbracht LM, Letcher K, Plaut M, Strahl N. 1972. Assessment of the biologic availability of tetracycline products in man. Clin Pharmacol Ther 13: 97108.
  • Belfroid AC, Van der Horst A, Vethaal AD, Schafer AJ, Rijs GBJ, Wegener J, Cofino WP. 1999. Analysis and occurrence of estrogenic hormones and their glucuronides in surface water and waste water in the Netherlands. Sci Total Environ 225: 101108.
  • Bock M, Lyndall J, Barber T, Fuchsman P, Perruchon E, Capdeville M. 2010. Probabilistic application of a fugacity model to predict triclosan fate during wastewater treatment. Integr Environ Assess Manag 6: 393404.
  • Burrows GE. 1980. Pharmacotherapeutics of macrolides, lincomycins, and spectinomycin. J Am Vet Med Assoc 176: 10721077.
  • Caldwell DJ, Mastrocco F, Nowak E, Johnston J, Yekel H, Pfeiffer D, Hoyt M, DuPlessie BM, Anderson PD. 2010. An assessment of potential exposure and risk from estrogens in drinking water. Environ Health Perspect 118: 338344.
  • Castiglioni S, Bagnati R, Fanelli R, Pomati F, Calamari D, Zuccato E. 2006. Removal of pharmaceuticals in sewage treatment plants in Italy. Environ Sci Technol 40: 357363.
  • Cha J, Cupples AM. 2009. Detection of the antimicrobials triclocarban and triclosan in agricultural soils following land application of municipal biosolids. Water Res 43: 25222530.
  • Choi K, Kim Y, Park J, Park CK, Kim M, Kim HS, Kim P. 2008. Seasonal variations of several pharmaceutical residues in surface water and sewage treatment plants of Han River, Korea. Sci Total Environ 405: 120128.
  • Clarke BB, Smith SR. 2011. Review of emerging organic contaminants in biosolids and assessment of international research priorities for the agricultural use of biosolids. Environ Int 37: 226247.
  • Cunningham VL. 2008. Environmental exposure modeling: application of PhATE™ and GREAT-ER to human pharmaceuticals in the environment. In: Kuemmerer K, editor. Pharmaceuticals in the environment: sources, fate, effects and risks. Berlin, Heidelberg: Springer-Verlag.
  • Cunningham VL, Binks SP, Olson MJ. 2009. Human health risk assessment from the presence of human pharmaceuticals in the aquatic environment. Regul Toxicol Pharmacol 53: 3945.
  • Cunningham VL, Perino C, D'Aco VJ, Hartmann A, Bechter R. 2010. Human health risk assessment of carbamazepine in surface waters of North America and Europe. Regul Toxicol Pharmacol 56: 343351.
  • Desbrow C, Routledge EJ, Brighty GC, Sumpter JP, Waldock M. 1998. Identification of estrogenic chemicals in STW effluent. 1. Chemical fractionation and in vitro biological screening. Environ Sci Technol 32: 15491558.
  • Federle TW, Kaiser SK, Nuck BA. 2002. Fate and effects of triclosan in activated sludge. Environ Toxicol Chem 21: 13301337.
  • Feijtel TCJ, Matthijs E, Rottiers A, Rijs GBJ, Kiewiet A, de Nijs A. 1995. AIS/CESIO environmental surfactant monitoring programme. Part I: LAS monitoring study in “de Meer” STP and receiving river “Leidsche Rijn.Chemosphere 30: 10531066.
  • Gobel A, McArdell CS, Joss A, Siegrist H, Giger W. 2007. Fate of sulfonamides, macrolides, and trimethoprim in different wastewater treatment technologies. Sci Total Environ 372: 361371.
  • Golet EM, Xifra I, Siegrist H, Alder AC, Giger W. 2003. Environmental exposure assessment of fluoroquinolone antibacterial agents from sewage to soil. Environ Sci Technol 37: 32433249.
  • Griffith RS, Black HR. 1970. Erythromycin. Med Clin North Am 54: 11991215.
  • Gulkowska A, Leung HW, So MK, Taniyasu S, Yamashita N, Yeung LWY, Richardson BJ, Lei AP, Giesy JP, Lam PKS. 2008. Removal of antibiotics from wastewater by sewage treatment facilities in Hong Kong and Shenzhen, China. Water Res 42: 395403.
  • Hannah RE, D'Aco VJ, Anderson PD, Buzby ME, Caldwell DJ, Cunningham VL, Ericson JF, Johnson AC, Parke NJ, Samuelian JH, Sumpter JP. 2009. Exposure assessment of 17α-ethinylestradiol in surface waters of the United States and Europe. Environ Toxicol Chem 28: 27252732.
  • [HERA] Human and Environmental Risk Assessment. 2009. Human and environmental risk assessment on ingredients of European household cleaning products. LAS Linear Alkylbenzene Sulphonate. Version 4. Available from: http://www.heraproject.com/RiskAssessment.cfm?SUBID=48, 2009.
  • Higgins CP, Sharp JO, Sepulvado JG, Littrell B, O'Connor G, Snyder EM, McAvoy DC. 2010. Trace organic chemicals in biosolids-amended soils, state-of-the-science review. WERF Report SRSK5T09. Alexandria, VA: Water Environment Research Foundation.
  • IMS Health. 2009. IMS Midas Quantum July 2008-August 2009. Collegeville, PA: IMS Health.
  • Jacobsen BN, Kjersgaard D, Winther-Nielsen M, Gustavson K. 2004. Combined chemical analyses and biomonitoring at Avedoere wastewater treatment plant in 2002. Water Sci Technol 50: 3743.
  • Jaffe JM, Colaizzi RI, McDonald RH. 1973. Effect of altered urinary pH on tetracycline and doxycycline excretion in humans. J Pharmacokinet Biopharm 1: 267282.
  • Joss A, Keller E, Alder AC, Gobel A, McArdell CS, Ternes TA, Siegrist H. 2005. Removal of pharmaceuticals and fragrances in biological wastewater treatment. Water Res 39: 31393152.
  • Joss A, Zabczynski S, Gobel A, Hoffmann B, Loffler D, McArdell CS, Ternes TA, Thomsen A, Siegrist H. 2006. Biological degradation of pharmaceuticals in municipal wastewater treatment: Proposing a classification scheme. Water Res 40: 16861696.
  • Kinney CA, Furlong Et, Zaugg SD, Burkhardt MR, Werner SL, Cahill JD, Jorgensen GR. 2006. Survey of organic wastewater contaminants in biosolids destined for land application. Environ Sci Technol 40: 72077215.
  • Lindberg RH, Olofsson U, Rendahl P, Johansson MI, Tysklind M, Andersen BAV. 2006. Behavior of fluoroquinolones and trimethoprim during mechanical, chemical, and active sludge treatment of sewage water and digestion of sludge. Environ Sci Technol 40: 10421048.
  • Lozano N, Rice CP, Ramirez M, Torrents A. 2010. Fate of triclosan in agricultural soils after biosolid applications. Chemosphere 78: 760766.
  • McAvoy DC, Eckhoff WS, Rapaport RA. 1993. Fate of linear alkylbenzene sulfonate in the environment. Environ Toxicol Chem 12: 977987.
  • McAvoy DC, Schatowitz B, Jacob M, Hauk A, Eckhoff WS. 2002. Measurement of triclosan in wastewater systems. Environ Toxicol Chem 21: 13231329.
  • Metcalf and Eddy. 1979. Wastewater engineering: Treatment disposal reuse. New York (NY): McGraw-Hill. 414 p.
  • Meyer MC, Dann RE, Whyatt PL, Slywka GWA. 1974. The bioavailability of sixteen tetracycline products. J Pharmacokinet Pharmacodynam 2: 287297.
  • Mogensen AS, Haagensen F, Ahring BK. 2003. Anaerobic degradation of LAS. Environ Toxicol Chem 22: 706711.
  • Norwegian Scientific Committee for Food Safety (VKM). 2009. Risk Assessment of contaminants in sewage sludge applied on Norwegian Soils. Opinion from the panel on contaminants. Oslo, Norway: VKM. Docnr. 05/511.
  • Oreskes N, Shrader-Frechette K, Belitz K. 1994. Verification, validation and confirmation of numerical models in the earth sciences. Science 263: 641646.
  • Page RL, Ruscin JM, Fish D, Lapointe M. 2001. Possible interaction between intravenous azithromycin and oral cyclosporine. Pharmacotherapy 21: 14361443.
  • Paxéus NA. 2004. Removal of selected non-steroidal anti-inflammatory drugs (NSAIDs), gemfibrozil, carbamazepine, b-blockers, trimethoprim and triclosan in conventional wastewater treatment plants in five EU countries and their discharge to the aquatic environment. Water Sci Technol 50: 253260.
  • Pfizer. 2009. Pfizer Zmax® prescribing information. New York (NY): Pfizer. LAB-0384-3. 0, 2009.
  • [PDR] Physicians' Desk Reference. 2004. Product Information. 58th ed. Montvale (NJ): Thomson PDR.
  • Radjenovic J, Petrovic M, Barcelo D. 2009. Fate and distribution of pharmaceuticals in wastewater and sewage sludge of the conventional activated sludge (CAS) and advanced membrane bioreactor (MBR) treatment. Water Res 43: 831841.
  • Robinson PF, Liu Q-TL, Riddle AM, Murray-Smith R. 2007. Modelling the impact of direct phototransformation on predicted environmental concentrations (PECs) of propranolol hydrochloride in UK and US rivers. Chemosphere 66: 757766.
  • Scheurer M, Sacher F, Brauch H-J. 2009. Occurrence of the antidiabetic drug metformin in sewage and surface waters in German. J Environ Monitoring 11: 16081613.
  • Schwab BW, Hayes EP, Fiori JM, Mastrocco FJ, Roden NM, Cragin D, Meyerhoff RD, D'Aco VJ, Anderson PD. 2005. Human pharmaceuticals in U.S. surface water: A human health risk assessment. Regul Toxicol Pharmacol 42: 296312.
  • Smith SR. 2009. Organic contaminants in sewage sludge (biosolids) and their significance for agricultural recycling. Philos Trans R Soc Ser A Math Phys Eng Sci 367: 40054010.
  • Struijs F. 1996. SimpleTreat 3.0: A model to predict the distribution and elimination of chemicals by sewage treatment plants. National Institute of Public Health and Environmental Protection Report 719101025. Biltoven, The Netherlands: National Institute of Public Health and Environmental Protection.
  • Struijs F, et al. 1991. SimpleTreat: A spreadsheet-based box model to predict the fate of xenobiotics in a municipal waste water treatment plant. National Institute of Public Health and Environmental Protection Report 670208002. Biltoven, The Netherlands: National Institute of Public Health and Environmental Protection.
  • Stumpf M, Ternes T, Wilken R-D, Rodrigues SC, Baumann W. 1999. Polar drug residues in sewage and natural waters in the state of Rio de Janeiro, Brazil. Sci Total Environ 225: 135141.
  • Tchobanolous G, Burton F. 1991. Wastewater engineering treatment, disposal and reuse. 3rd ed. Boston (MA): Metcalf and Eddy.
  • Terzic S, Senta I, Ahel M, Gros M, Petrovic M, Barcelo D, Muller J, Knepper T, Marti I, Ventura F, Jovancic P, Jabucar D. 2008. Occurrence and fate of emerging wastewater contaminants in Western Balkan Region. Sci Total Environ 399: 6677.
  • Trautwein C, Kummerer K. 2009. Incomplete degradation of the type II antidiabetic Metformin, University Medical Center Freiburg. Available from: http://www.epilepsiezentrum.uniklinik-freiburg.de/iuk/live/umweltforschung/veroeffentlichungen/Poster_Metformin_Knappe.pdf, Accessed 3/21/10.
  • Tuomisto J, Mannisto P. 1973. Cross-over study of ten tetracycline preparations. Eur J Clin Pharmacol 6: 6468.
  • [USEPA] US Environmental Protection Agency. 2003. Control of pathogens and vector attraction in sewage sludge (including domestic septage) under 40 CFR Part 503. Cincinnati, OH: USEPA. EPA/625/R-92/013.
  • [USEPA] US Environmental Protection Agency. 2009. Targeted national sewage sludge survey sampling and analysis technical report. Washington, DC: USEPA. EPA-822-R-08-016.
  • [USEPA] US Environmental Protection Agency. 2008. Clean Watersheds Needs Survey 2008 Report to Congress. Washington, DC: USEPA. EPA-832-R-10-002.
  • Vasskog T, Anderssen T, Pedersen-Bjergaard S, Kallenborn R, Jensen E. 2008. Occurrence of selective serotonin reuptake inhibitors in sewage and receiving waters at Spitsbergen and in Norway. J Chromatogr 1185: 194205.
  • Walters E, McClellan K, Halden RU. 2010. Occurrence and loss over 3 years of 72 pharmaceuticals and personal care products from biosolids-soil mixtures in outdoor mesocosms. Water Res 44: 60116020.
  • Xia K, Hundal LS, Kumar K, Armbrust K, Cox AE, Granato TC. 2010. Tricarban, triclosan, polybrominated diphenyl ethers, and 4-nonylphenol in biosolids and in soil receiving 33-year biosolids application. Environ Toxicol Chem 29: 597605.
  • Yasojima M, Nakada N, Komori K, Suzuki Y, Tanaka H. 2006. Occurrence of levofloxacin, clarithromycin and azithromycin in wastewater treatment plant in Japan. Water Sci Technol 53: 227233.