Occurrence of Giardia cysts and Cryptosporidium oocysts in raw and treated water from two water treatment plants in Selangor, Malaysia

The occurrence of Giardia cysts and Cryptosporidium oocysts was determined both in the raw water and the treated waters from two water treatment plants in Selangor, Malaysia between July 1994 and January 1995. At each treatment plant, raw and treated water samples were collected fortnightly on 10 separate occasions. Physical parameters, including dissolved oxygen, pH, temperature, conductivity, turbidity and free and total chlorine were measured, and samples were also taken for faecal coliform analyses. The isolation and enumeration of cysts and oocysts was according to UK standard methods. Whilst Giardia cysts were detected in 90% of the raw water samples (range 0–60 cysts per litre) no Cryptosporidium oocysts were detected. Higher concentrations of cysts were detected in the raw water of the first treatment plant. All raw water samples were faecally contaminated, as faecal coliform counts ranged from 4.6 × 10⁴ to 1.3 × 10⁵ colony forming units per 100 ml. No correlation between the concentration of Giardia cysts and the concentration of faecal coliforms could be demonstrated statistically, nor was there any correlation between the concentration of Giardia cysts and the physical parameters tested for any of the raw waters. No cysts, oocysts or faecal coliforms were detected in samples of treated water.     

Monitoring of Cryptosporidium and Giardia river contamination in Paris area

This study evaluates the protozoan contamination of river waters, which are used for drinking water in Paris and its surrounding area (about 615,000 m(3) per day in total, including 300,000 m(3) for Paris area). Twenty litre samples of Seine and Marne Rivers were collected over 30 months and analyzed for Cryptosporidium oocysts and Giardia cysts detection according to standard national or international methods. Cryptosporidium oocysts and Giardia cysts were found, respectively, in 45.7% and 93.8% of a total of 162 river samples, with occasional high concentration peaks. A significant seasonal pattern was observed, with positive samples for Cryptosporidium more frequent in autumn than spring, summer and winter, and positive samples for Giardia less frequent in summer. Counts of enterococci and rainfalls were significantly associated with Giardia concentration but not Cryptosporidium. Other faecal bacteria were not correlated with monitored protozoan. Marne seems to contribute mainly to the parasitic contamination observed in Seine. Based on seasonal pattern and rainfall correlation, we hypothesize that the origin of contamination is agricultural practices and possible dysfunction of sewage treatment plants during periods of heavy rainfalls. High concentrations of protozoa found at the entry of drinking water plants justify the use of efficient water treatment methods. Treatment performances must be regularly monitored to ensure efficient disinfection according to the French regulations.     

Presence of Cryptosporidium spp. and Giardia duodenalis through drinking water

To evaluate the presence of Cryptosporidium spp. and Giardia duodenalis in the influent and final effluent of sixteen drinking water treatment plants located in a hydrographic basin in Galicia (NW Spain) - in which the principal river is recognised as a Site of Community Importance (SCI) - estimate the efficiency of treatment plants in removing these protozoans and determine the species and genotypes of the parasites by means of a molecular assay. All plant samples of influent and final effluent (50-100 l) were examined in the spring, summer, autumn and winter of 2007. A total of 128 samples were analysed by method 1623, developed by US Environmental Protection Agency for isolation and detection of both parasites. To identify the genotypes present the following genes were amplified and sequenced: 18S SSU rRNA (Cryptosporidium spp.) and b-giardina (G. duodenalis). The mean concentrations of parasites in the influent were 0.0-10.5 Cryptosporidium spp. oocysts per litre and 1.0-12.8 of G. duodenalis cysts per litre. In the final treated effluent, the mean concentration of parasites ranged from 0.0-3.0 oocysts per litre and 0.5-4.0 cysts per litre. The distribution of results by season revealed that in all plants, the highest numbers of (oo)cysts were recorded in spring and summer. Cryptosporidium parvum, C. andersoni, C. hominis and assemblages A-I, A-II, E of G. duodenalis were detected. Cryptosporidium spp. and G. duodenalis were consistently found at high concentrations in drinking water destined for human and animal consumption in the hydrographic basin under study, in Galicia (NW Spain). It is important that drinking water treatment authorities rethink the relevance of contamination levels of both parasites in drinking water and develop adequate countermeasures.     

Cryptosporidium and Giardia detection in water bodies of Galicia, Spain

The objective of this study was to determine the mean concentration (per litre) of Cryptosporidium oocysts and Giardia cysts in recreational river areas (n = 28), drinking water treatments plants (DWTPs; n = 52) and wastewater treatment plants (WWTPs; n = 50) in Galicia (NW Spain). Water samples from rivers and from the influent (50–100 l) and the treated effluent (100 l) of the water plants were filtered using Filta-Max filters (IDEXX Laboratories, Inc., Westbrook, ME, USA). A total of 232 samples were processed and the (oo)cysts were concentrated, clarified by IMS and then detected by IFAT. The viability was determined by applying fluorogenic vital dye (PI).In the recreational areas, infective forms of Cryptosporidium and Giardia were detected in 16 (57.1%; 1–60 oocysts per litre) and 17 (60.7%; 1–160 cysts per litre) samples, respectively. In the water flowing into the water treatment plants, oocysts were detected in 21 DWTPs (40.4%; 1–13 oocysts per litre) and cysts were observed in 22 DWTPs (42.3%; 1–7 cysts per litre). In the effluents from the treatment plants, Cryptosporidium oocysts and Giardia cysts were identified in 17 DWTPs (32.7%; 1–4 oocysts per litre) and in 19 DWTPs (36.5%; 1–5 cysts per litre), respectively. The highest concentrations of (oo)cysts were found in the WWTPs; specifically, oocysts were detected in 29 (58.0%; 1–80 oocysts per litre) and cysts in 49 (98.0%; 2–14.400 cysts per litre) WWTP effluents. Cryptosporidium and Giardia were detected in 32 (64.0%; 1–120 oocysts per litre) and 48 (96.0%; 2–6.000 cysts per litre) WWTP effluents, respectively. The percentage viability of the (oo)cysts ranged between 90.0% and 95.0%. In all samples analysed. Moreover, it was found that the effluents from coastal WWTPs were discharged directly into the sea, while inland WWTPs were discharged directly into rivers. The concentrations of both enteropathogens detected in effluents from WWTPs therefore represent a significant risk to human and animal health.These results demonstrate the wide distribution of Cryptosporidium and Giardia in the environment, the ineffectiveness of treatments in DWTPs and WWTPs in reducing/inactivating both protozoa and the need to monitor the presence, viability and infectivity of Cryptosporidium and Giardia in water bodies. In conclusion, the findings suggest the need for better monitoring of water quality and identification of sources of contamination.     

Cryptosporidium spp. and Giardia duodenalis in two areas of Galicia (NW Spain)

The aim of the present study was to investigate the environmental dispersal of Cryptosporidium spp. and Giardia duodenalis in two distinct areas (coastal and inland) in Galicia (NW Spain). Faecal samples were collected from healthy asymptomatic domestic (cows and sheep) and wild animals (deer and wild boars) in the selected areas. In each of the selected areas, samples of untreated water (influent) and of treated water (final effluent) were collected from each of the 12 drinking water treatments plants (DWTPs) and 12 wastewater treatment plants (WTPs) under study. Analysis of a single sample from each of the 635 (coastal) and 851 (inland) domestic and wild animals selected at random revealed that the prevalences of cryptosporidiosis and giardiosis in coastal area were 9.2% and 15.9% respectively, and in inland area, 13.7% and 26.7% respectively. In the coastal area, Cryptosporidium spp. oocysts were detected in influent and effluent samples from 2/12 (16.6%) DWTPs and 8/12 (66.6%) WTPs, while G. duodenalis cysts were detected in influent and effluent samples from 3/12 (25.0%) DWTPs and 12/12 (100%) WTPs. The concentrations were notably higher in WTPs; the mean parasite concentrations in the final treated effluent were 10 oocysts per litre and 137.8 cysts per litre for Cryptosporidium and Giardia, respectively. The mean concentration of G. duodenalis cysts per litre was significantly higher (P < 0.05) than the mean concentration of Cryptosporidium spp. oocysts per litre in both the influent and the effluent samples from all the treatment plants. In the coastal area, C. parvum, C. hominis and G. duodenalis assemblages A (I and II) and E were most repeatedly detected. In the inland area, C. parvum, C. andersoni and G. duodenalis assemblages A (I and II), B and E were most frequently identified.     

Detection of enteric viruses, Giardia and Cryptosporidium in two different types of drinking water treatment facilities

In this study, two types of drinking water treatment facilities (two conventional drinking water treatment plants (DWTPs) and two compact units (Cus)) were compared referring to their production capacity. Water samples were collected from three main points: (a) different water treatment steps (b) washings of sand filters and (c) distribution system at different distances from the water treatment plants. Both viruses and protozoa were concentrated from each water sample by adsorption and accumulation on the same nitrocellulose membrane filters (0.45 microm pore size). Enteroviruses were detected by plaque infectivity assay in BGM cells and HAV, HEV and Norovirus were detected by RT-PCR. Giardia and Cryptosporidium were detected by conventional staining methods and PCR. The results revealed that enterovirus load at the intake ranged between 10-15 PFU/L for the two compact units and between 4.5 and 75 PFU/L for the two conventional DWTPs. The virus load in distribution system of the first type DWTPs at 1 km from the plant was the same as that of the intake. Viruses in the other type of treatment plants CUs at 1, 5 and 7 km, were much reduced. Investigation of raw water sediments of the two DWTPs showed enterovirus counts between 12 and 17.5 PFU/L. Virus count was reduced in sand of filters after washing. Giardia cysts were equally detected by microscopy and PCR in only intake samples of EL-Hawamdia CU (33.3%) and Meet Fares DWTP (50%). Cryptosporidium oocysts were equally detected by microscopy and PCR in intake samples of Abo EL-Nomros CU (100%), EL-Hawamdia CU (66.7%) and Fowa DWTP (50%). At Meet Fares DWTP three positive intake samples for Cryptosporidium were detected by PCR, compared with only two positive samples by microscopy. Giardia cysts and Cryptosporidium oocysts were detected in raw water sediment and sand of filters before washing. Only one sample from Meet Fares DWTP sand of filters after washing was positive for both Giardia and Cryptosporidium. It can be concluded that the poor microbial quality of the water may be due to improper operational skills and management of the various water treatment plants (especially at the two high capacity treatment plants).     

Prevalence of Cryptosporidium oocysts and giardia cysts in the drinking water supply in Japan

A one-year monitoring of Cryptosporidium oocysts and Giardia cysts was conducted at a water purification plant. A total of 13 samples of 50 L river source water and 26 samples of 2,000 L-filtered water, treated by coagulation flocculation, sedimentation and rapid filtration, were tested. Prior to conducting a survey of a water purification plant, we developed a method for concentrating Cryptosporidium oocysts from a large volume of raw or filtered water using a hollow fiber ultrafiltration (UF) membrane, and this procedure was adapted to survey a water purification plant. Cryptosporidium oocysts were detected in all of the 13 raw water samples. The geometric mean concentration was 40 oocysts 100 L. Giardia cysts were detected in 12 of 13 raw water samples (92%) and the geometric mean concentration was 17 cysts/100 L. Probability distributions of both Cryptosporidium oocyst and Giardia cyst concentration in raw water were nearly lognormal. In filtered water samples, Cryptosporidium oocysts were detected in 9 of the 26 samples (35%) with the geometric mean concentration of 1.2 oocysts /1,000 L and Giardia cysts in 3 samples (12%) with 0.8 cysts/1,000 L. The estimated log10 removal efficiency of Cryptosporidium oocysts and Giardia cysts by rapid-sand filtration was 2.47 and 2.53, respectively. Empty particles were removed at a higher log10 than intact oocysts and cysts. The efficiency of particle removal in the rapid sand filtration process tends to be reduced under cold-water conditions. Close management is necessary in the winter when the water temperature is low.     

Effect of pathogen concentrations on removal of Cryptosporidium and Giardia by conventional drinking water treatment

The presence of waterborne enteric pathogens in municipal water supplies contributes risk to public health. To evaluate the removal of these pathogens in drinking water treatment processes, previous researchers have spiked raw waters with up to 10(6) pathogens/L in order to reliably detect the pathogens in treated water. These spike doses are 6-8 orders of magnitude higher than pathogen concentrations routinely observed in practice. In the present study, experiments were conducted with different sampling methods (i.e., grab versus continuous sampling) and initial pathogen concentrations ranging from 10(1) to 10(6) pathogens/L. Results showed that Cryptosporidium oocyst and Giardia cyst removal across conventional treatment were dependent on initial pathogen concentrations, with lower pathogen removals observed when lower initial pathogen spike doses were used. In addition, higher raw water turbidity appeared to result in higher log removal for both Cryptosporidium oocysts and Giardia cysts.     

Contribution of treated wastewater to the contamination of recreational river areas with Cryptosporidium spp. and Giardia duodenalis

Samples of the influent and final effluent from 12 wastewater treatment plants from Galicia (NW, Spain) were analyzed for the presence of Cryptosporidium spp. oocysts and Giardia duodenalis cysts. All of the plants discharge effluent to a hydrographic basin in which there are numerous recreational areas and fluvial beaches. The samples (25-50 liters) were collected in spring, summer, autumn and winter of 2007. A total of 96 samples were analyzed using techniques included in the US Environmental Protection Agency Method 1623. To identify the genotypes present, the following genes were amplified and sequenced: 18S SSU rRNA (Cryptosporidium spp.) and beta-giardina (G. duodenalis). Both parasites were detected in influent and effluent samples from all treatment plants (100%) throughout the year, and G. duodenalis always outnumbered Cryptosporidium spp. The mean concentration of G. duodenalis per liter of influent was significantly higher (P<0.05) than the mean concentration of Cryptosporidium spp. per liter of influent. The mean concentrations of parasites in influent samples ranged from 6 to 350 Cryptosporidium spp. oocysts per liter and from 89 to 8305 G. duodenalis cysts per liter. In final treated effluent, the mean concentration of parasites ranged from 2 to 390 Cryptosporidium spp. oocysts per liter and from 79 to 2469 G. duodenalis cysts per liter. The distribution of results per season revealed that in all plants, the highest number of (oo)cysts were detected in spring and summer. Cryptosporidium parvum, Cryptosporidium andersoni, Cryptosporidium hominis and assemblages A-I, A-II, E of G. duodenalis were detected. The risk of contamination of water courses by Cryptosporidium spp. and G. duodenalis is therefore considerable. It is important that wastewater treatment authorities reconsider the relevance of the levels of contamination by both parasites in wastewater, and develop adequate countermeasures.     

Influence of land use and watershed characteristics on protozoa contamination in a potential drinking water resources reservoir

Relative changes in the microbial quality of Lake Texoma, on the border of Texas and Oklahoma, were investigated by monitoring protozoan pathogens, fecal indicators, and factors influencing the intensity of the microbiological contamination of surface water reservoirs. The watershed serves rural agricultural communities active in cattle ranching, recreation, and is a potential drinking water source. A total of 193 surface water samples were tested over a 27-month period to determine levels of parasite contamination. The overall occurrence of Cryptosporidium oocysts was higher in both frequency and concentration than Giardia cysts. Cryptosporidium oocysts were found in 99% and Giardia cysts in 87% of the samples. Although Cryptosporidium and Giardia occurrence were significantly but not strongly correlated, all other correlation coefficients including turbidity and total dissolved solids were non-significant. Statistically supportable seasonal variations were found suggesting that Cryptosporidium and Giardia were higher in summer and fall than in other seasons of the year. While Cryptosporidium levels were correlated with rainfall, this was not the case with Giardia. The maximum numbers for both protozoan parasites were detected from a site impacted by cattle ranching during calving season. Restriction fragment length polymorphism analysis was used for confirmation of Cryptosporidium in surface waters influenced by agricultural discharges. As we had expected, oocysts were of the bovine type indicating that the Cryptosporidium parvum detected in surface waters perhaps came from cattle living in the watershed.     

上海市浦东自来水厂净水工艺对两虫去除效果的研究

采用免疫磁分离及荧光染色法（EPA1623法）对上海市浦东地区4家自来水厂的原水、出厂水、管网水中贾第鞭毛虫和隐孢子虫的分布状况进行了分析，并考察了水厂净水过程中不同操作单元对两虫的去除效果。研究表明：原水中存在贾第鞭毛虫与隐孢子虫，但其密度分别维持在0～8个／10L和0～6个／10L的低水平，而在出厂水与管网水中均未检出两虫；水厂的混凝、沉淀单元对两虫具有明显的去除效果。     

Detection of Cryptosporidium spp. and Giardia duodenalis in surface water: A health risk for humans and animals

The objective of the present study was to determine the degree of contamination by Cryptosporidium spp. and Giardia duodenalis in a river basin in a livestock farming area in Galicia (NW, Spain). Water samples (50 l) were collected at 22 points in the main basin (including 5 recreational areas), and at the source and mouth of the 3 most important rivers and at the mouth of a smaller, secondary river. Faecal samples were collected from dairy cattle selected at random from 18 herds farmed in the area. A total of 139 neonatal calves, 480 heifers and 697 cows were sampled. The prevalence, intensity of infection and the risk associated with the spread of infection by both enteropathogens were determined. Water and faecal samples were collected in spring, summer, autumn and winter of 2007. The species and genotypes of these parasites present in the water samples were identified. In both water and faecal samples, more parasitic stages were collected in spring and summer than in autumn and winter. In spring, Cryptosporidium spp. oocysts were detected in 33 (9.4%) cows from 13 (72.2%) herds, and G. duodenalis cysts were detected in 56 (16.0%) cows from 15 farms (83.3%); the intensity of infection ranged from 5 to 7895 G. duodenalis cysts per gram of faeces. Infective stages of Cryptosporidium spp. and G. duodenalis were also detected in respectively 26 (89.6%) and 27 (93.1%) water samples, in spring. The mean concentrations of parasites ranged from 2 to 1200 Cryptosporidium spp. oocysts per litre and from 2 to 400 G. duodenalis cysts per litre. Cryptosporidium parvum, C. andersoni, C. hominis and assemblages A-I, A-II, E of G. duodenalis were detected. The presence of both protozoans must be monitored in cattle, in sources of water used for recreational purposes and in artificial waterways used by farmers (water channels, animal drinking water and drainage systems).     

Riverbank filtration for control of microorganisms: results from field monitoring

Microbial monitoring was conducted over a period of more than 1 year at three full-scale riverbank filtration (RBF) facilities, located in the United States along the Ohio, Missouri, and Wabash Rivers. Results of this study demonstrated the potential for RBF to provide substantial reductions in microorganism concentrations relative to the raw water sources. Cryptosporidium and Giardia were detected occasionally in the river waters but never in any of the well waters. Average concentrations and log reductions of Cryptosporidium and Giardia could not be accurately determined due to the low and variable concentrations in the river waters and the lack of detectable concentrations in the well waters. Average concentrations of aerobic and anaerobic spore-forming bacteria, which have both been proposed as potential surrogates for the protozoans, were reduced at the three facilities by 0.8 to > 3.1 logs and 0.4 to > 4.9 logs, respectively. Average concentrations of male-specific and somatic bacteriophage were reduced by > 2.1 logs and 3.2 logs, respectively. Total coliforms were rarely detected in the well waters, with 5.5 and 6.1 log reductions in average concentrations at the two wells at one of the sites relative to the river water. Average turbidity reductions upon RBF at the three sites were between 2.2 and 3.3 logs. Turbidity and microbial concentrations in the river waters generally tracked the river discharge; a similar relationship between the well water concentrations and river discharge was not observed, due to the low, relatively constant well water turbidities and lack of a significant number of detections of microorganisms in the well waters. Further research is needed to better understand the relationships among transport of pathogens (e.g., Cryptosporidium, Giardia, viruses) and potential surrogate parameters (including bacterial spores and bacteriophage) during RBF and the effects of water and sediment characteristics on removal efficiency.     

Occurrence and removal of pharmaceuticals and hormones through drinking water treatment

The occurrence of fifty-five pharmaceuticals, hormones and metabolites in raw waters used for drinking water production and their removal through a drinking water treatment were studied. Thirty-five out of fifty-five drugs were detected in the raw water at the facility intake with concentrations up to 1200 ng/L. The behavior of the compounds was studied at each step: prechlorination, coagulation, sand filtration, ozonation, granular activated carbon filtration and post-chlorination; showing that the complete treatment accounted for the complete removal of all the compounds detected in raw waters except for five of them. Phenytoin, atenolol and hydrochlorothiazide were the three pharmaceuticals most frequently found in finished waters at concentrations about 10 ng/L. Sotalol and carbamazepine epoxide were found in less than a half of the samples at lower concentrations, above 2 ng/L. However despite their persistence, the removals of these five pharmaceuticals were higher than 95%.     

A Practical Assessment of the Use of Particle Counters for Monitoring the Risk of Cryptosporidium and Giardia

The paper describes a project which was carried out to establish whether particle counts can be used to predict the risk of Cryptosporidium and Giardia breakthrough at a large water-treatment works. The project was based at a direct river-abstraction works which treats up to 120 MI/d. On-line particle counters were installed following three of the rapid-gravity filters, and a fourth counter was used to measure particles in spot samples of raw and clarified water.
There was a clear relationship between particle counts and the occasional low-level (oo)cyst (crypto oocysts and giardia cysts) occurrence in the filter samples, and no (oo)cysts were found in the final water. There appear to be complex inter-relationships between particle counts and other parameters such as raw-water conditions, environmental factors and treatment-plant variables.     

Impact of bathers on levels of Cryptosporidium parvum oocysts and Giardia lamblia cysts in recreational beach waters

Recreational beach water samples collected on weekends and weekdays during 11 consecutive summer weeks were tested for potentially viable Cryptosporidium parvum oocysts and Giardia lamblia cysts using the multiplexed fluorescence in situ hybridization (FISH) method. The levels of oocysts and cysts on weekends were significantly higher than on the weekdays (P<0.01). Concentrations of oocysts in weekend samples (n=27) ranged from 2 to 42 oocysts/L (mean: 13.7 oocysts/L), and cyst concentration ranged from 0 to 33 cysts/L (mean: 9.1 cysts/L). For the samples collected on weekdays (n=33), the highest oocyst concentration was 7 oocysts/L (mean: 1.5 oocysts/L), and the highest cyst concentration was 4 cysts/L (mean: 0.6 cysts/L). The values of water turbidity were significantly higher on weekends than on weekdays, and were correlated with the number of bathers and concentration of C. parvum oocysts and G. lamblia cysts (P<0.04). The study demonstrated positive relationships between number of bathers and levels of waterborne C. parvum oocysts and G. lamblia cysts in recreational beach water. It is essential to test recreational waters for Cryptosporidium and Giardia when numbers of bathers are greatest, or limit the number of bathers in a recreational beach area.     

上海市区原水及自来水中两虫分布调查

对上海市区15座水厂的原水和自来水的调查表明,长江原水及黄浦江上游原水中均有贾第鞭毛虫和隐孢子虫存在,以贾第鞭毛虫为主,原水中两虫的总体污染程度较低;在所有自来水样品中均未发现两虫存在;根据调查结果提出了去除两虫的手段和预警监测方法.     

Life cycle cost analysis of a double-effect solar still

In this present paper, the study focuses on the development of a double-effect solar still with evacuated tubes for water desalination considered for small-scale applications at remote locations where only saline water or brackish water is available. In this present paper, the potable water cost per litre has been evaluated. The payback periods for different conditions of the distribution of distilled water, namely at the cost it is produced and at the selling price on market rate, have been evaluated. The cost of water per litre was a minimum of Rs. 0.19 with the average distillate output being about 20?l/day, when the interest rate and the lifetime of the solar still are taken as 4% and 50 years, respectively. The lowest payback time of 45 days was obtained when the selling price of water was Rs. 20?per litre.     

A nationwide survey of NDMA in raw and drinking water in Japan

A nationwide survey of N-nitrosodimethylamine (NDMA) in both raw and finished water samples from drinking water treatment plants (DWTPs) in Japan was conducted. NDMA was analyzed by solid-phase extraction (SPE) followed by ultra performance liquid chromatography (UPLC) coupled with tandem mass spectrometry (MS/MS). NDMA was detected in 15 of 31 raw water samples collected in the summer at concentrations up to 2.6 ng/L, and in 9 of 28 raw water samples collected in winter at concentrations up to 4.3 ng/L. The NDMA concentrations were higher in raw water samples collected from treatment plants with catchment areas that have high population densities. The NDMA concentrations were higher in river water samples collected from the east and west of Japan than in those collected from other areas. NDMA was detected in 10 of 31 finished samples collected in summer at reduced concentrations of up to 2.2 ng/L, while 5 of 28 finished samples collected in winter showed NDMA concentrations up to 10 ng/L. The highest NDMA levels were detected in finished water samples collected from the Yodo River basin DWTP, which uses ozonation. Furthermore, evaluation of the process water produced at six advanced water treatment plants was conducted. Influent from the Yodo River indicated that the NDMA concentration increased during ozonation to as high as 20 ng/L, and then decreased with subsequent biological activated carbon treatment. To our knowledge, this is the first nationwide evaluation of NDMA concentrations in water conducted in Japan to date.     

Determination of caffeine as a tracer of sewage effluent in natural waters by on-line solid-phase extraction and liquid chromatography with diode-array detection

A new liquid chromatographic (LC) method with automated on-line solid phase extraction was developed to determine caffeine at sub-microgram per litre concentrations in waters. The filtered sample was pre-concentrated in a pre-column, which was backwashed with acidic water at pH of 2.70. The concentrated caffeine was separated using a C18 column with a gradient of water-acetonitrile and detected by diode array detection (DAD) at 210 nm. Four different pre-columns: C18, PRP-1, PLRP-s and Env were evaluated for the on-line solid phase extraction of caffeine. The PLRP-s pre-column allowed the enrichment of up to 100 mL of environmental water sample with highest recovery. The procedure was validated by recovery experiments in water spiked at 0.5 1.0 and 4.0 microg/L. Average recoveries were between 92.1 +/- 5.2% and 97.8 +/- 2.6%. Detection limits as low as 0.1 microg/L from 50 ml of sample were achieved. The proposed method has the advantages of higher reliability and sensitivity, simpler sample preparation and shorter analysis time in comparison with off-line solid-phase extraction. The utility of the method was demonstrated at two field sites: Bolivar and Halls Head (Australia). At Bolivar, the treatment process included 6-week lagoon storage which is believed to have attenuated caffeine, and thus limited its use as an environmental tracer of reclaimed water. At the Halls Head site, where the storage period is shorter, caffeine was detected in both the treated sewage effluent and in groundwater near ponds where the reclaimed water is at similar concentrations. These results suggest that the environmental conditions under which caffeine is conservative require better definition.