- Email: [email protected]
Improved wastewater disinfection by ultrasonic pre-treatment
Ultrasonics Sonochemistry 11 (2004) 333–336 www.elsevier.com/locate/ultsonch
Improved wastewater disinfection by ultrasonic pre-treatment q Torben Blume *, Uwe Neis Technical University Hamburg––Harburg, 21073 Hamburg, Germany Accepted 9 June 2003 Available online 12 August 2003
Abstract The objective of our work is to evaluate the scientific and economic potential of US application as a pre-treatment step in combination with UV to optimise the disinfection process of wastewaters. Ultrasound application of 20 s at low density of 30 W/l changed the particle size distribution (PSD) of the samples, the mean particle diameter dropped from 70 to 11 lm. Generally it is assumed that bioparticles bigger than 50 lm are difficult to disinfect by UV. We observed that the relevant particle size range >50 lm in samples taken from the primary clarifier was reduced by at least three-quarters by low ultrasound doses. As expected, these changes in PSD notably effect the disinfection efficiency of UV. Whereas UV treatment of secondary clarifierÕs effluents alone led to a reduction of fecal coliforms by 2.5 log units, pre-treatment by sonication (only 5 s at densities of 50 and 310 W/l) clearly enhanced the disinfection efficiency: reductions of CFU (colony forming unit) concentration now ranged between 3.3 and 3.7 log units. We noticed an influence of the bacteriaÕs morphology on the disinfection efficiency of the combined process (US plus UV). Gram-positive streptococci seem less vulnerable to ultrasound exposure than thinner-walled gram-negative bacteria like the entire group of coliforms. The application of an ultrasound step might be also useful in terms of cost-effectiveness. In our lab-scale tests 30 s of UV treatment alone were required to reduce the number of fecal coliforms by 3.7 log units. When applied in combination, 5 s of ultrasonic followed by only 5 s of UV irradiation had the same result and energy consumption was only 43%. Ó 2003 Elsevier B.V. All rights reserved.
1. Introduction At present, it is not prescribed by regulations that European sewage treatment plantsÕ (STP) effluents have to meet microbiological criteria. Nevertheless, it is common sense that these effluents have to be disinfected when discharged into sensitive areas. Therefore, in many European countries microbiological guidelines contained in the EU bathing water directive [1] are being adopted and applied to STP effluents [2]. As counts of indicator organisms (such as fecal coliforms) are usually not reduced to tolerable levels in a conventional treatment process, an additional subsequent disinfection step is unavoidable. Types of disinfection techniques are q This paper was originally presented at Applications of Power Ultrasound in Physical and Chemical Processing (Usound3), Paris, December 2001. * Corresponding author. Tel.: +49-40-42878-2416; fax: +49-4042878-2684. E-mail address: [email protected] (T. Blume).
1350-4177/$ - see front matter Ó 2003 Elsevier B.V. All rights reserved. doi:10.1016/S1350-4177(03)00156-1
various: they include physical, chemical and natural/ biological methods. Several studies have shown that the efficiency of disinfection methods is highly dependant on the sampleÕs concentration of suspended solids (SS) [3,4], due to the fact that SS can act as protection to bacteria and viruses [5]. For example, UV applicationsÕ efficiency is limited for samples with high concentrations of suspended matter (Fig. 1) [4]. Recent studies [6] have shown that large particles (bigger than 50 lm in diameter) are hard to penetrate so that the required UV demand is raised drastically. Therefore, it is common practice to install sand filters (e.g. rapid sand filters) to reduce particulate matter prior to the UV step. Rapid sand filters are expensive in construction and maintenance. They are well known in potable water production, but when it comes to wastewater treatment there are many drawbacks (e.g. clogging, algae growth, backwashing). Another attempt to bring down the size of agglomerates of particles is the application of ultrasound (US). High power ultrasound, operated at low frequencies,
334
T. Blume, U. Neis / Ultrasonics Sonochemistry 11 (2004) 333–336
Fig. 2. Flow scheme of the experimental set-up.
Fig. 1. Effects of particles on UV disinfection.
is an effective means for disintegration of bacterial cells: first, at low ultrasound doses bacteria flocs can be deagglomerated by mechanical shear stress. When the US dose is increased, ultrasound cavitation can destroy cell walls. This effect is lethal to the microorganisms. We observed in other tests that long sonication at maximum density (400 W/l) of up to 60 min leads to significant reductions of fecal coliform counts (3 log units). This is also in accordance with the findings of Hua and Thomson [9]. For economical aspects this method can not be considered an alternative to conventional disinfection: energy input is almost factor 500 higher than for an UV treatment with the same reduction rates. The aim of our experiments presented here is to show that ultrasound provides an appropriate means to change the physical composition of the sample so that ‘‘big particles’’ are transformed into smaller ones. Therefore, we tested a combination of ultrasonic pretreatment and UV disinfection to find out whether changes in particle size distribution of wastewater samples facilitate a subsequent UV disinfection.
2. Equipment and analytical procedures The experimental set-up is depicted schematically in Fig. 2. Ten litres of treated municipal wastewater are stored in a glass bottle and mixed constantly by a magnetic stirrer. A peristaltic pump is used to convey the medium through the system: first it passes the ultrasound apparatusÕ flow chamber, then it enters the ultraviolet device to be exposed to UV irradiation. Sampling outlets at various stages facilitate taking of samples after each individual step of treatment.
Fig. 3. Horn sonotrode in a continuous flow chamber.
The ultrasound device (Fig. 3) used was a ‘‘Branson Sonifier W-450’’, a horn sonotrode equipped with a horn tip of 1.3 cm in diameter which is operated at 20 kHz. Electrical power in the range of 41–154 W was applied, and calorimetric measurements exposed that intensities ranged from 1.7 to 60.8 W/cm2 , and densities from 10 to 400 W/l, respectively. The low-pressure mercury arc lamp (manufacturer: ‘‘Pureflow Ultraviolet Inc.’’, nominal length: 20 cm, diameter: 1.3 cm) is enclosed in a tubular floating chamber (volume: 300 ml). A surrounding thin layer of quartz glass shields the lamp from the sample that flows parallel to the orientation of the lamp. Its energy consumption is 14 W of which 3 W are emitted at 254 nm (37 lW/cm2 @ 1 m), the relevant wavelength for bacteria inactivation.
T. Blume, U. Neis / Ultrasonics Sonochemistry 11 (2004) 333–336
Particle size analysis was conducted with a Hiac Royco, Model 8000A (equipped with a sampler, model 3000A and a HRLD-150 sensor). In this automatic particle counter a laser diode functions as the illumination source and a photo diode serves as the detector. Particle counts and size distributions are calculated and displayed automatically. The Spread Plate Technique (for high concentrations of microorganisms) and the Membrane Filtration Technique (for low levels of detectable microorganisms) have been applied according to the ‘‘Standard Methods for the Examination of Water and Wastewater’’ [7]. For the enumeration of total germs, total coliforms, Escherichia coli, fecal coliforms and fecal streptococci, specific types of solid agar have been chosen. Results are presented as CFU per 100 ml.
3. Results and discussion Fig. 4 demonstrates the capability of ultrasound application to eliminate the fraction of big particles: samples (STPs primary effluent) were treated for 20 s at various ultrasound densities. Initially, 63% of the solids in the wastewater sample were characterised by a size bigger than 50 lm. After a sonication of 20 s at 30 W/l, this fraction just accounts for 5% of the total counts. Low ultrasound energy (30 W/l) is already sufficient to provoke a clear change in particle composition. Further increased ultrasonic doses have only a marginal effect. Although the concentration of suspended matter is not reduced by sonication, the particle size distribution is comparable with those of filtered effluents: Herwig et al. [8] reported that sand filters indeed do eliminate preferentially the large particle fraction bigger than 50 lm almost completely from secondary effluents. Fig. 4 shows that ultrasonic treatment of wastewater had a comparable effect: almost no particles >50 lm are left after sonication. The effect of a short sonication on UV efficiency is depicted in Fig. 5. Secondary effluents have been irra-
Fig. 4. Effect of sonication on particle size distribution.
335
Fig. 5. Effect of ultrasonic pre-treatment density (sonication time: 5 s) on reduction of fecal coliforms and on number of particle counts >50 lm.
diated by UV light for 5 s so that the concentration of fecal coliforms (FC) was reduced by 2.5 log units. Prior sonication for 5 s resulted in improved UV disinfection efficiency: at a low sonication level, disinfection rate could be improved by 0.8 log units and for the higher ultrasound level efficiency could be enhanced by 1.2 orders of magnitude (compared with the not pre-treated sample). These results coincides with the data we obtained from particle size analysis: dependant on the ultrasound dose applied, particles bigger than 50 lm could be reduced by 25% and 60%, respectively (volume related) by ultrasonic pre-treatment. We did not only focus on fecal coliforms, but also analysed the effect on the less vulnerable group of streptococci, a group of microorganisms that is more resistant to environmental stress than commonly monitored coliforms [10]. Fig. 6 depicts disinfection efficiencies for a secondary effluent (TSS: 5.2 mg/l, mean diameter: 68 lm) that was irradiated with sole UV light and with a combination of US and UV. Thirty seconds of UV irradiation were needed to bring down the number of E. coli to less than 1000 and fecal streptococci
Fig. 6. Reduction of E. coli and fecal streptococci by ultrasonic pretreatment for 10 s at different densities followed by UV irradiation for 30 s.
336
T. Blume, U. Neis / Ultrasonics Sonochemistry 11 (2004) 333–336
ml, an exclusive UV irradiation of as much as 30 s is not capable of achieving this goal and is even more ‘‘costly’’ as the consumed energy of 1500 W s/l is about 66.7% higher.
References
Fig. 7. Improved UV disinfection (density: 50 W/l) by ultrasonic pretreatment (sonication time: 5 s, density: 80 W/l) to meet the EU bathing water guide value.
to less than 100 CFU/100 ml, respectively. Ultrasonic pre-treatment for only 10 s at densities of 170 W/l brought down mean particle size to 35 lm, with an increased US density of 310 W/l the samplesÕ mean diameter could even be lowered to 20 lm. For both microbiological parameters observed, the ultrasonic pre-treatment has a clear beneficial effect; disinfection efficiency is by more than one order of magnitude higher. As expected, the thicker-walled streptococcus species seems to be less vulnerable than E. coli. In order to demonstrate that the combination of a short ultrasonic and a subsequent ultraviolet treatment is even cost-efficient in our lab-scale experiments, Fig. 7 demonstrates that this combination is very meaningful although the specific energy consumption of ultrasound (80 W/l) is higher than the UV lampÕs (50 W/l). Whereas 5 s of ultrasonic pre-treatment and 10 s of UV disinfection consume 900 W s/l to reduce fecal coliforms to a level beneath the critical concentration of 100 FC/100
[1] EEC, Council directive concerning the quality of bathing water (76/160/EEC), Brussels, 1976. [2] T. Nelle, Abwasserdesinfektion mit UV-Licht und Ozon unter Einbeziehung der vorgeschalteten Reinigungsverfahren, Inaugural-Dissertation an der Naturwissenschaftlichen Fakult€at der Universit€at Witten/Herdecke, 1994. [3] N. Narkis, R. Armon, R. Offer, F. Orshansky, E. Friedland, Effect of suspended solids on wastewater disinfection efficiency by chlorine dioxide, Wat. Res. 29 (1) (1995) 227–236. [4] J.L. Darby, K.E. Snider, G. Tchobanoglous, Ultraviolet disinfection for wastewater reclamation and reuse subjects to restrictive standards, Wat. Environ. Res. 65 (2) (1993) 169–180. [5] M.W. LeChevallier, C.D. Cawthon, R.G. Lee, Factors promoting survival of bacteria in chlorinated water supplies, Appl. Environ. Microbiol. 54 (1988) 649–654. [6] G. Sakamoto, C. Zimmer, UV disinfection for wastewater reclamation: the effect of particle size and suspended solids, in: 1997 PNPCA Annual Conference, 26–29 October 1997, Seattle, Washington. [7] American Public Health Association, American Water Works Association, Water Environment Federation, Standard Methods for the Examination of Water and Wastewater, 19th ed., Washington, 1995. [8] V. Herwig, A. Tiehm, U. Neis, Particle size analysis for improved deep bed filtration in waste water treatment, in: World Filtration Congress 8, Brighton, UK, 3–7 April 2000, Proceedings Volume 2, 2000, pp. 827–830. [9] I. Hua, J.E. Thomson, Inactivation of Escherichia coli by sonication at discrete ultrasonic frequencies, Wat. Res. 34 (15) (2000) 3888–3893. [10] M.J. Figueras, F. Polo, I. Inza, J. Guarro, Past, present and future perspectives of the EU bathing water directive, Mar. Pollut. Bull. 34 (3) (1997) 148–156.
Improved wastewater disinfection by ultrasonic pre-treatment q Torben Blume *, Uwe Neis Technical University Hamburg––Harburg, 21073 Hamburg, Germany Accepted 9 June 2003 Available online 12 August 2003
Abstract The objective of our work is to evaluate the scientific and economic potential of US application as a pre-treatment step in combination with UV to optimise the disinfection process of wastewaters. Ultrasound application of 20 s at low density of 30 W/l changed the particle size distribution (PSD) of the samples, the mean particle diameter dropped from 70 to 11 lm. Generally it is assumed that bioparticles bigger than 50 lm are difficult to disinfect by UV. We observed that the relevant particle size range >50 lm in samples taken from the primary clarifier was reduced by at least three-quarters by low ultrasound doses. As expected, these changes in PSD notably effect the disinfection efficiency of UV. Whereas UV treatment of secondary clarifierÕs effluents alone led to a reduction of fecal coliforms by 2.5 log units, pre-treatment by sonication (only 5 s at densities of 50 and 310 W/l) clearly enhanced the disinfection efficiency: reductions of CFU (colony forming unit) concentration now ranged between 3.3 and 3.7 log units. We noticed an influence of the bacteriaÕs morphology on the disinfection efficiency of the combined process (US plus UV). Gram-positive streptococci seem less vulnerable to ultrasound exposure than thinner-walled gram-negative bacteria like the entire group of coliforms. The application of an ultrasound step might be also useful in terms of cost-effectiveness. In our lab-scale tests 30 s of UV treatment alone were required to reduce the number of fecal coliforms by 3.7 log units. When applied in combination, 5 s of ultrasonic followed by only 5 s of UV irradiation had the same result and energy consumption was only 43%. Ó 2003 Elsevier B.V. All rights reserved.
1. Introduction At present, it is not prescribed by regulations that European sewage treatment plantsÕ (STP) effluents have to meet microbiological criteria. Nevertheless, it is common sense that these effluents have to be disinfected when discharged into sensitive areas. Therefore, in many European countries microbiological guidelines contained in the EU bathing water directive [1] are being adopted and applied to STP effluents [2]. As counts of indicator organisms (such as fecal coliforms) are usually not reduced to tolerable levels in a conventional treatment process, an additional subsequent disinfection step is unavoidable. Types of disinfection techniques are q This paper was originally presented at Applications of Power Ultrasound in Physical and Chemical Processing (Usound3), Paris, December 2001. * Corresponding author. Tel.: +49-40-42878-2416; fax: +49-4042878-2684. E-mail address: [email protected] (T. Blume).
1350-4177/$ - see front matter Ó 2003 Elsevier B.V. All rights reserved. doi:10.1016/S1350-4177(03)00156-1
various: they include physical, chemical and natural/ biological methods. Several studies have shown that the efficiency of disinfection methods is highly dependant on the sampleÕs concentration of suspended solids (SS) [3,4], due to the fact that SS can act as protection to bacteria and viruses [5]. For example, UV applicationsÕ efficiency is limited for samples with high concentrations of suspended matter (Fig. 1) [4]. Recent studies [6] have shown that large particles (bigger than 50 lm in diameter) are hard to penetrate so that the required UV demand is raised drastically. Therefore, it is common practice to install sand filters (e.g. rapid sand filters) to reduce particulate matter prior to the UV step. Rapid sand filters are expensive in construction and maintenance. They are well known in potable water production, but when it comes to wastewater treatment there are many drawbacks (e.g. clogging, algae growth, backwashing). Another attempt to bring down the size of agglomerates of particles is the application of ultrasound (US). High power ultrasound, operated at low frequencies,
334
T. Blume, U. Neis / Ultrasonics Sonochemistry 11 (2004) 333–336
Fig. 2. Flow scheme of the experimental set-up.
Fig. 1. Effects of particles on UV disinfection.
is an effective means for disintegration of bacterial cells: first, at low ultrasound doses bacteria flocs can be deagglomerated by mechanical shear stress. When the US dose is increased, ultrasound cavitation can destroy cell walls. This effect is lethal to the microorganisms. We observed in other tests that long sonication at maximum density (400 W/l) of up to 60 min leads to significant reductions of fecal coliform counts (3 log units). This is also in accordance with the findings of Hua and Thomson [9]. For economical aspects this method can not be considered an alternative to conventional disinfection: energy input is almost factor 500 higher than for an UV treatment with the same reduction rates. The aim of our experiments presented here is to show that ultrasound provides an appropriate means to change the physical composition of the sample so that ‘‘big particles’’ are transformed into smaller ones. Therefore, we tested a combination of ultrasonic pretreatment and UV disinfection to find out whether changes in particle size distribution of wastewater samples facilitate a subsequent UV disinfection.
2. Equipment and analytical procedures The experimental set-up is depicted schematically in Fig. 2. Ten litres of treated municipal wastewater are stored in a glass bottle and mixed constantly by a magnetic stirrer. A peristaltic pump is used to convey the medium through the system: first it passes the ultrasound apparatusÕ flow chamber, then it enters the ultraviolet device to be exposed to UV irradiation. Sampling outlets at various stages facilitate taking of samples after each individual step of treatment.
Fig. 3. Horn sonotrode in a continuous flow chamber.
The ultrasound device (Fig. 3) used was a ‘‘Branson Sonifier W-450’’, a horn sonotrode equipped with a horn tip of 1.3 cm in diameter which is operated at 20 kHz. Electrical power in the range of 41–154 W was applied, and calorimetric measurements exposed that intensities ranged from 1.7 to 60.8 W/cm2 , and densities from 10 to 400 W/l, respectively. The low-pressure mercury arc lamp (manufacturer: ‘‘Pureflow Ultraviolet Inc.’’, nominal length: 20 cm, diameter: 1.3 cm) is enclosed in a tubular floating chamber (volume: 300 ml). A surrounding thin layer of quartz glass shields the lamp from the sample that flows parallel to the orientation of the lamp. Its energy consumption is 14 W of which 3 W are emitted at 254 nm (37 lW/cm2 @ 1 m), the relevant wavelength for bacteria inactivation.
T. Blume, U. Neis / Ultrasonics Sonochemistry 11 (2004) 333–336
Particle size analysis was conducted with a Hiac Royco, Model 8000A (equipped with a sampler, model 3000A and a HRLD-150 sensor). In this automatic particle counter a laser diode functions as the illumination source and a photo diode serves as the detector. Particle counts and size distributions are calculated and displayed automatically. The Spread Plate Technique (for high concentrations of microorganisms) and the Membrane Filtration Technique (for low levels of detectable microorganisms) have been applied according to the ‘‘Standard Methods for the Examination of Water and Wastewater’’ [7]. For the enumeration of total germs, total coliforms, Escherichia coli, fecal coliforms and fecal streptococci, specific types of solid agar have been chosen. Results are presented as CFU per 100 ml.
3. Results and discussion Fig. 4 demonstrates the capability of ultrasound application to eliminate the fraction of big particles: samples (STPs primary effluent) were treated for 20 s at various ultrasound densities. Initially, 63% of the solids in the wastewater sample were characterised by a size bigger than 50 lm. After a sonication of 20 s at 30 W/l, this fraction just accounts for 5% of the total counts. Low ultrasound energy (30 W/l) is already sufficient to provoke a clear change in particle composition. Further increased ultrasonic doses have only a marginal effect. Although the concentration of suspended matter is not reduced by sonication, the particle size distribution is comparable with those of filtered effluents: Herwig et al. [8] reported that sand filters indeed do eliminate preferentially the large particle fraction bigger than 50 lm almost completely from secondary effluents. Fig. 4 shows that ultrasonic treatment of wastewater had a comparable effect: almost no particles >50 lm are left after sonication. The effect of a short sonication on UV efficiency is depicted in Fig. 5. Secondary effluents have been irra-
Fig. 4. Effect of sonication on particle size distribution.
335
Fig. 5. Effect of ultrasonic pre-treatment density (sonication time: 5 s) on reduction of fecal coliforms and on number of particle counts >50 lm.
diated by UV light for 5 s so that the concentration of fecal coliforms (FC) was reduced by 2.5 log units. Prior sonication for 5 s resulted in improved UV disinfection efficiency: at a low sonication level, disinfection rate could be improved by 0.8 log units and for the higher ultrasound level efficiency could be enhanced by 1.2 orders of magnitude (compared with the not pre-treated sample). These results coincides with the data we obtained from particle size analysis: dependant on the ultrasound dose applied, particles bigger than 50 lm could be reduced by 25% and 60%, respectively (volume related) by ultrasonic pre-treatment. We did not only focus on fecal coliforms, but also analysed the effect on the less vulnerable group of streptococci, a group of microorganisms that is more resistant to environmental stress than commonly monitored coliforms [10]. Fig. 6 depicts disinfection efficiencies for a secondary effluent (TSS: 5.2 mg/l, mean diameter: 68 lm) that was irradiated with sole UV light and with a combination of US and UV. Thirty seconds of UV irradiation were needed to bring down the number of E. coli to less than 1000 and fecal streptococci
Fig. 6. Reduction of E. coli and fecal streptococci by ultrasonic pretreatment for 10 s at different densities followed by UV irradiation for 30 s.
336
T. Blume, U. Neis / Ultrasonics Sonochemistry 11 (2004) 333–336
ml, an exclusive UV irradiation of as much as 30 s is not capable of achieving this goal and is even more ‘‘costly’’ as the consumed energy of 1500 W s/l is about 66.7% higher.
References
Fig. 7. Improved UV disinfection (density: 50 W/l) by ultrasonic pretreatment (sonication time: 5 s, density: 80 W/l) to meet the EU bathing water guide value.
to less than 100 CFU/100 ml, respectively. Ultrasonic pre-treatment for only 10 s at densities of 170 W/l brought down mean particle size to 35 lm, with an increased US density of 310 W/l the samplesÕ mean diameter could even be lowered to 20 lm. For both microbiological parameters observed, the ultrasonic pre-treatment has a clear beneficial effect; disinfection efficiency is by more than one order of magnitude higher. As expected, the thicker-walled streptococcus species seems to be less vulnerable than E. coli. In order to demonstrate that the combination of a short ultrasonic and a subsequent ultraviolet treatment is even cost-efficient in our lab-scale experiments, Fig. 7 demonstrates that this combination is very meaningful although the specific energy consumption of ultrasound (80 W/l) is higher than the UV lampÕs (50 W/l). Whereas 5 s of ultrasonic pre-treatment and 10 s of UV disinfection consume 900 W s/l to reduce fecal coliforms to a level beneath the critical concentration of 100 FC/100
[1] EEC, Council directive concerning the quality of bathing water (76/160/EEC), Brussels, 1976. [2] T. Nelle, Abwasserdesinfektion mit UV-Licht und Ozon unter Einbeziehung der vorgeschalteten Reinigungsverfahren, Inaugural-Dissertation an der Naturwissenschaftlichen Fakult€at der Universit€at Witten/Herdecke, 1994. [3] N. Narkis, R. Armon, R. Offer, F. Orshansky, E. Friedland, Effect of suspended solids on wastewater disinfection efficiency by chlorine dioxide, Wat. Res. 29 (1) (1995) 227–236. [4] J.L. Darby, K.E. Snider, G. Tchobanoglous, Ultraviolet disinfection for wastewater reclamation and reuse subjects to restrictive standards, Wat. Environ. Res. 65 (2) (1993) 169–180. [5] M.W. LeChevallier, C.D. Cawthon, R.G. Lee, Factors promoting survival of bacteria in chlorinated water supplies, Appl. Environ. Microbiol. 54 (1988) 649–654. [6] G. Sakamoto, C. Zimmer, UV disinfection for wastewater reclamation: the effect of particle size and suspended solids, in: 1997 PNPCA Annual Conference, 26–29 October 1997, Seattle, Washington. [7] American Public Health Association, American Water Works Association, Water Environment Federation, Standard Methods for the Examination of Water and Wastewater, 19th ed., Washington, 1995. [8] V. Herwig, A. Tiehm, U. Neis, Particle size analysis for improved deep bed filtration in waste water treatment, in: World Filtration Congress 8, Brighton, UK, 3–7 April 2000, Proceedings Volume 2, 2000, pp. 827–830. [9] I. Hua, J.E. Thomson, Inactivation of Escherichia coli by sonication at discrete ultrasonic frequencies, Wat. Res. 34 (15) (2000) 3888–3893. [10] M.J. Figueras, F. Polo, I. Inza, J. Guarro, Past, present and future perspectives of the EU bathing water directive, Mar. Pollut. Bull. 34 (3) (1997) 148–156.