Environ. Sci. Technol. 2006, 40, 1042-1048
Behavior of Fluoroquinolones and Trimethoprim during Mechanical, Chemical, and Active Sludge Treatment of Sewage Water and Digestion of Sludge R I C H A R D H . L I N D B E R G , * ,† ULRIKA OLOFSSON,† PER RENDAHL,‡ MAGNUS I. JOHANSSON,§ MATS TYSKLIND,† AND BARBRO A. V. ANDERSSON† Environmental Chemistry, Umeå University, SE-901 87 Umeå, Sweden, UMEVA, SE-901 84 Umeå, Sweden, and Astra Zeneca R&D, Mo¨lndal, SE-431 83, Sweden
The behavior and fate of three fluoroquinolones (norfloxacin, ofloxacin, and ciprofloxacin), one sulfonamide (sulfamethoxazole), and trimethoprim were investigated at a sewage treatment plant in Umeå, Sweden, in 2004. This plant uses conventional mechanical, chemical, and activated sludge methods to treat the sewage water and digest the sludge; the dewatered digested sludge is pelleted (dry weight > 90% of total weight). Raw sewage water and particles as well as effluents and sludge from specific treatment areas within the plant were sampled. In addition to quantifying the antibiotics within the plant, we characterized the sample matrixes to facilitate evaluation of the results. Of the five substances examined, only norfloxacin, ciprofloxacin, and trimethoprim were present in concentrations higher than their limits of quantification. Norfloxacin and ciprofloxacin sorbed to sludge in a manner that was independent of changes in pH during sewage treatment, and more than 70% of the total amount of these compounds passing through the plant was ultimately found in the digested sludge. The results suggest that fluoroquinolones undergo thermal degradation during pelleting, but more studies are needed to confirm this. Trimethoprim was found in the final effluent at approximately the same concentration and mass flow as in the raw sewage, and could not be quantified in any solid sample. Predicted environmental concentrations, based on consumption data for Umeå municipality, correlated well with the results obtained, especially when the predicted concentrations were corrected to account for the amount of each active substance excreted in urine. The results obtained were compared to those of previous studies of these three substances’ behavior and fate and were found to be similar, although some of the other plants studied employed the various treatment steps in different orders.
* Corresponding author phone: +46 90 786 6669; fax: +46 90 12 81 33; e-mail:
[email protected]. † Umeå University. ‡ UMEVA. § Astra Zeneca R&D. 1042
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 3, 2006
Introduction Recently, issues related to pharmaceutical residues in the environment have been considered by the Swedish government, and in 2004 the Swedish medical product agency (MPA) concluded that our knowledge of the levels, fate, and effects of pharmaceuticals in the environment is inadequate (1). The MPA also suggested that further research should be conducted to determine environmental levels of pharmaceuticals, to develop models of their environmental transport and fate, and for educational purposes. Because many of the active substances are excreted in patients’ urine and/or feces in significant quantities (2), they may enter sewage treatment plants (STP) via raw sewage water and/or particles. These substances must be degraded during the treatment of the sewage, must be sorbed on sludge, or must undergo some combination of the two if they are to be kept out of recipient waters. If they are sorbed to sludge, the environmental distribution of the substances will be dependent on the ultimate fate of the STP’s solid products (use in fertilizers, disposal in landfills, etc.). Data on accurately measured or predicted levels of antibiotics and their effects on target organisms are required for accurate Environmental Risk Assessments (ERAs). Studies on the identification and quantification of pharmaceuticals in various environmental matrixes can be found in recent literature (3-18). However, there are little data on their fate during sewage treatment and in the environment, or on their effects on ecosystems. Few test systems are suitable for investigating their long-term toxicity, so only acute toxicity was considered when ERAs of 30 pharmaceuticals were conducted by the MPA (1). Ku ¨ mmerer et al. (19) investigated the suitability of a standardized short-term inhibition test for antibiotics, disinfectants, and cytotoxic compounds, concluding that it could not be recommended for ERAs and that every substance or class of substances with the same mode of action should be investigated separately. Antibiotics are pharmaceuticals that are commonly used in human and veterinary medicine to prevent or treat bacterial infections. Their potential as environmental pollutants is due to their potency to develop and/or maintain antibiotic resistant bacteria, and some studies reveal findings that give such indications (20). Investigations about their toxic effects, for example, in aqueous matrixes have also been made (6, 21). Municipal sewage water is usually treated in three steps: a mechanical (screening and removal of sand and fat); a chemical (flocculation of phosphorus with ferrous sulfate or ferrous chloride); and a biological step (degradation of organic material by microorganisms). The order of the two latter treatment steps is reversed in some STPs. Solids (sludge) are removed from the water using clarifiers (in some cases by filtration) and are further processed by digestion (i.e., degradation of organic material) and dewatering. The fate of certain pharmaceuticals in sewage during treatment has been studied to some extent by sampling effluents and (in some cases) particles and sludge from STPs, for example, Golet et al. (11), Carballa et al. (17), Andersen et al. (22), Go¨bel et al. (23, 24), and this study. The extent to which pharmaceuticals are removed from the aqueous phase by degradation and/or sorption can be determined via analysis of the STP’s solid products. Another approach was employed by Ternes et al. (25), who conducted batch experiments by fortification of pharmaceuticals in primary and secondary sludge slurries to be able to predict their sorption capacity in STPs. Holbrook et al. (26) used a full-scale activated sludge system and a 10.1021/es0516211 CCC: $33.50
2006 American Chemical Society Published on Web 12/15/2005
FIGURE 1. Schematic diagram of the treatment processes in the Umeå STP. pilot-scale membrane bioreactor to investigate the sorption behavior of estrogenic substances on colloidal material. Because almost every STP is unique in terms of its catchment area and treatment processes, a careful investigation of the sample matrixes could be useful for the construction of predictive models for use in ERAs. In the course of a screening study in 2002-2003 (18) at the Umeå STP in Sweden, norfloxacin, ofloxacin, and ciprofloxacin (fluoroquinolones, FQs), sulfamethoxazole (sulfonamide), and trimethoprim were quantified in the raw sewage water entering the plant, its final effluent, and the digested dewatered sludge. The results from that study suggested that the fluoroquinolones were sorbed to sludge and that trimethoprim and sulfamethoxazole were typically unaffected by the treatment. In the work presented here, the behavior and fate of these antibiotics were investigated in detail within the STP. Samples were taken from raw sewage water/particles as well as effluents and sludge following each treatment step to determine the concentration levels and mass flows of the antibiotics throughout the process. In addition, calculations of predicted concentrations and mass flows were conducted using consumption data of the antibiotics within Umeå municipality.
Experimental Section Sample Site Description. The sewage treatment plant investigated receives water from Umeå and its surroundings, an area with a total population of approximately 82 000 people. Process schematics and other information on the STP, including sampling locations and detailed retention times (hydraulic and solid), are provided in Figure 1. The total hydraulic and solid retention times are approximately 8 h and 20 days, respectively. In Umeå STP, the incoming raw sewage water consists mainly of domestic water (20% surface water). The catchment area includes a University hospital, which serves approximately 900 000 people. The major local industries are food processing and engineering. The incoming raw sewage is mechanically treated by passing the flow through a screen (3 mm split) followed by removal of sand and fat (together with approximately 3% of the organic matter). During chemical treatment, a flocculating agent is added (FeSO4, 690 kg Fe d-1) under stirring prior to primary clarification. This step reduces the biological oxygen demand (BOD) and phosphorus content of the aqueous phase to 60-
70% of their original values. The primary effluent is then subjected to a biological activated sludge treatment, which reduces its organic content and, after a final secondary clarification step, removes most of the remaining phosphorus. Most of the sludge produced during the secondary clarification is returned to the beginning of the activated sludge treatment to retain and recycle the microorganisms (recycled activated sludge). The age of this sludge, that is, the duration of time it is being recycled, is approximately 3 days. A minor proportion of the sludge from the secondary clarifier is sent to the beginning of the treatment process (as excess sludge) when the activated sludge content exceeds a specific level. Excess sludge is allowed to flow for 30 s approximately four times per hour. Sludge from the primary clarifier (chemical and biological sludge) enters the thickener together with external sludge (approximately 10 000 “person-equivalents”) and is subjected to anaerobic digestion at 38 °C. The digested sludge is stored in a sludge silo prior to the addition of a cationic polyelectrolyte polymer (5 g kg-1 dry weight) and dewatering. If possible, the digested dewatered sludge is pelleted via passing it through a thin layer drier at a constant temperature of 180 °C and then through a moving belt dryer at 105 °C. Sample Collection and Extraction. Sewage water and sludge samples were collected on three consecutive days in November and December 2004 (except for pellet samples, which were collected in September 2004). On each occasion, flow proportional sampling of the sewage following the various treatment steps was performed over 24 h using automatic samplers (approximately 3 L) at the STP. 750 mL of each sample was extracted to quantify its antibiotic content. These samples were immediately filtered to remove particles greater than 0.45 µm in diameter and stored (for 4 days) in the dark at -18 °C prior to extraction/analysis. Solid-phase extraction and liquid chromatography/tandem mass spectrometry were used to extract and determine the antibiotics in the aqueous samples. Details of the method used, such as recoveries and quantification procedure with internal standards, can be found elsewhere (18). Limits of quantification (LOQ) were as follows: norfloxacin, 7 ng L-1; ofloxacin, 6 ng L-1; ciprofloxacin, 6 ng L-1; sulfamethoxazole, 80 ng L-1; and trimethoprim, 8 ng L-1. All sludge samples were grab sampled. Those with a high water content (>90%) were immediately centrifuged at 5000 rpm for 30 min, as was a sample of raw sewage water, to concentrate the particulate matter. Solid particles were extracted from 1 L of each sludge sample and 8 L of the raw sewage water. The decanted sludge and raw sewage particles were air-dried in a fume hood for 2 days. All solid samples were stored (for 5 days) in the dark at -18 °C prior to extraction/analysis. Liquid-solid extraction and liquid chromatography/tandem mass spectrometry were used to extract and determine the concentrations (corrected for recovery yield) of the antibiotics in the solid samples; the methodology involved is described in detail elsewhere (18). LOQ (dry weight) were as follows: norfloxacin, 0.1 mg kg-1; ofloxacin, 0.1 mg kg-1; ciprofloxacin, 0.1 mg kg-1; sulfamethoxazole, 1.1 mg kg-1; and trimethoprim, 0.1 mg kg-1. Potential blank contamination was checked by solid-phase extraction of tap water, and liquid-solid extraction of an empty vial. Characterization of Sewage Water and Sludge. Samples of the different matrixes were collected during the second day of sampling. Selected physical and chemical properties of the sewage water and sludge were determined either at the STP laboratory or at a commercial laboratory (AlControl AB, Umeå, Sweden). The measured properties of the aqueous matrixes were their suspended solid (SS) content, pH, temperature, phosphorus content, chemical oxygen demand (COD7), concentration of ferrous species, and nitrogen levels. VOL. 40, NO. 3, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1043
TABLE 1. Concentrations of the Antibiotics in Sewage Water and Sludge sewage water
