Fate Of Pharmaceutical Waste

Published Date: 02 Nov 2017

The inflow of the wastewater was monitored during the period between September 2010 and August 2011. These data showed that the inflow of the wastewater ranged between 271000 and 298400 m3/day. The physico-chemical parameters of the influent during the sampling period exhibit large variations. The wastewater temperature values varied from 23 °C at winter and 35 °C at summer. The pH values varied from 7.0 to 7.2 and the values of nitrogen, phosphorous, chemical oxygen demand (COD), EC and suspended solids (SS) fluctuate during the sample year (see appendix).

5.1.2 Performance of wastewater treatment processes of Sulaibiya

The physico-chemical properties of the effluent are presented in appendix. The total phosphorus content and the total Kjeldahl nitrogen content decreased from 6.3 mg/L to 1.6 mg/L and 43.5 to 2.5 mg/L, respectively. Phosphorus and nitrogen are normally known as limiting nutrients for eutrophication in natural balance of aquatic ecosystems. Therefore, careful manage of their discharge is important to prevent the excessive algal growth (Andersen et al., 2006). The primary and secondary treatments of the wastewater effectively reduced the phosphorus and nitrogen by 75 and 94%, respectively. Significantly, the effluent had a better quality in regard to the nitrogen and organic contents due to the efficiency of the activated sludge process in the WWTP where average COD removals were 93%. The suspended solids in the secondary effluent are much lower than the influent by 95%. During the primary and secondary treatments, the cation's concentrations did not change significantly, which can be seen by the only small change in the electrical conductivity between the influent and effluent with average reduction of 17% where other researcher’s results (7%) as reported by Tchobanoglous et al. (2007).

5.1.3 Occurrence of pharmaceuticals in wastewater influents

The concentrations of the target pharmaceuticals in the influent over the year long sampling period at Sulaibiya are summarized in (Figure 1). Trimethoprim, sulphamethoxazole, paracetamol and ranitidine were found in all influent samples whereas, metronidazole was not detected in October and November. Metronidazole detection was ranging between 4ng/l in December and 58ng/l in April lower than other reported (Rosal et al., 2010). Trimethoprim and sulphamethoxazole were found in the influent in the range of 61-1814 and 11-1669, respectively. The highest concentration of trimethoprim was found in August, and the lowest was found in April whereas the highest concentration of sulphamethoxazole was found in October, and the lowest was in February.

Trimethoprim was reported at 290 ng/L in raw influent water in Switzerland (Goebel et al., 2005) and at relatively high concentrations from2100–7900 ng/L in the US (Batt et al., 2007). On the other hand, sulphamethoxazole was reported at high concentration was 6000ng/l (Batt et al., 2005) and at lower concentration was 70ng/l which was higher than our results. Paracetamol and ranitidine were found in the influent at concentrations significantly higher than other target drugs, which were the top ten pharmaceuticals dispensed in Kuwait. Paracetamol was detected in all the wastewater samples at concentrations ranging from 101-2086 ng/L with the highest concentrations in November 2010 and the lowest concentrations in February 2011. These concentrations were, to some extent, lower than those reported previously (Pham and Proulx, 1997; Ternes 1998; Blanchard et al., 2004). On the other hand, ranitidine was ranging from 365 to 2009 ng/L. These concentrations are consistent with other studies with concentrations of lower detection up to 580 ng/L (Kolpin et al., 2002) and higher detection at 1700 ng/l (Gomez et al., 2006).

The temperature fluctuated during the sampling between summer and winter. This fact might indicate that the concentrations of pharmaceuticals in the influent may be related to higher consumption during the cold periods of the year when more illnesses occur.

Figure 1: Variation of concentration of various target compounds (ng/L) in the influent and each circle in line represents one sample.

5.1.4 Removal of pharmaceuticals during the primary and secondary treatment in Sulaibiya

The removal rates of pharmaceuticals during the sampling period were shown in Figure 2(a,b,c,d,e). Paracetamol was removed efficiently by the secondary treatment, at an average 97.5% with the highest removal reached 99.9%, and the lowest removal was 86.1%. Trimethoprim was removed lower than paracetamol with average removal 86.1% where the highest removal was 96.1%, and the lowest removal was 63%. Removal efficiency of metronidazole in secondary treatment was at average 83.4% with the highest removal was 93.9%, and the lowest removal was 59.4%. Sulphamethoxazole and ranitidine were the lowest removal efficiency with average 77.5% where the highest removal of sulphamethoxazole was 98.7%, and the lowest removal was 31.3% while the highest removal of ranitidine was 99.2%, and the lowest removal was 47.4%.

Figure 2.a: Concentrations of metronidazole (ng/L) in the influent and the removal percentage by the secondary treatment processes of Sulaibiya WWTP during the year.

Figure 2.b: Concentrations of trimethoprim (ng/L) in the influent and the removal percentage by the secondary treatment processes of Sulaibiya WWTP during the year.

Figure 2.c: Concentrations of sulphamethoxazole (ng/L) in the influent and the removal percentage by the secondary treatment processes of Sulaibiya WWTP during the year.

Figure 2.d: Concentrations of paracetamol (ng/L) in the influent and the removal percentage by the secondary treatment processes of Sulaibiya WWTP during the year.

Figure 2.e: Concentrations of ranitidine (ng/L) in the influent and the removal percentage by the secondary treatment processes of Sulaibiya WWTP during the year.

In general, the removal efficiencies found in this study were consistent with other WWTP using primary and secondary treatment with activated sludge. For example, 75% removal rate in German (Ternes, 1998; Stumpf et al., 1999), up to 90% in Spain (Santos et al., 2007) and up to 99% in Japan (Nakada et al., 2006) were reported. These removal rates for a single compound can vary greatly from one WWTP to another depending on the type of treatment (e.g. biological and physico-chemical) and the residence time of water in the primary sedimentation tank (Santos et al., 2007).

Removal efficiency of metronidazole has been reported with a large variability range from 65-80% in Spain (Gros et al., 2010). On the other hand, trimethoprim was reported incomplete removal during conventional treatment by several studies ( Gobel et al., 2007; Jelic et al., 2011) while Gros et al., 2010 reports 65 to 80% removal efficiency in the plants with higher hydraulic retention times. Similar observation for the removal efficiency of sulphamethoxazole and ranitidine were found by other researchers where they report removal efficiency 30-92% and 50-98%, respectively (Gros et al., 2010). In Germany, paracetamol was found to be removed efficiently at 95 % due to its biodegradability and was detected in less than 10% of all sewage effluents (Ternes et al., 1998; Kolpin et al., 2004; Roberts and Thomas, 2006).

Concentrations of the target pharmaceuticals detected in the WWTP effluent in a range of 1-1000ng/L are presented in Figure. This is in agreement with Ternes et al (1998), who reported that many pharmaceuticals were detected in the effluents and measured at high concentrations due to incomplete removal in German sewage treatment plants.

The efficiency of modern wastewater treatments have increased the removal of pharmaceuticals from the influent with the introduction of the activated sludge process. Elimination of pharmaceuticals in the activated sludge process occurs due to several reasons adsorption, biological or chemical degradation and biotransformation. Ternes et al., 1998 suggested that activated sludge removes high amounts of pharmaceuticals than other treatments, most likely to the bacterial activity in the activated sludge. The results of this study showed there was not complete elimination of trace pharmaceuticals in the effluent. Therefore, implementing other technologies such as membrane systems would be necessary for complete removal of these traces.

5.1.4.1Effect of temperature on the removal efficiency

Although the total concentrations of target compounds in the influent samples through-out the yearly sampling fluctuated, the removal process in the wastewater, treatment plant worked as efficiently during the summer months as during the winter months. Therefore, effect of temperature was statically analysed using ANOVA. The correlation between the temperature and the removal of COD, BOD, organic nitrogen, TKN, MLVSS, and target pharmaceuticals was highly significant (Table 1). This conclusion contradicts other researchers that found that the removal processes in wastewater treatment plant was higher in summer than in winter (Vieno et al., 2005). They suggested that the reason was the lower biodegradation in the plant because of low temperature in winter.

Table 1: The significance level of the effect of temperature and their interaction on the responses.

Variables

Factor

p-value

Confidence level

Level of Significance

COD removal

Temp

0

100

Highly Significant

BOD removal

Temp

0

100

Highly Significant

Organic nitrogen removal

Temp

0

100

Highly Significant

TKN removal

Temp

0

100

Highly Significant

MLVSS

Temp

0

100

Highly Significant

Metronidazole removal

Temp

0

100

Highly Significant

Trimethoprim removal

Temp

0

100

Highly Significant

Sulphamethoxazole removal

Temp

0

100

Highly Significant

Paracetamol removal

Temp

0

100

Highly Significant

Ranitidine removal

Temp

0

100

Highly Significant

5.1.4.2 Effect of pharmaceuticals concentration on the removal efficiency

The effect of pharmaceutical concentration was highly significant on most responses except sulphamethoxazole, which was a significant effect on COD, BOD, organic nitrogen, and TKN removal, while was highly significant on MLVSS and sulphamethoxazole removal efficiency (Table 2). This highly significant effect of most target pharmaceuticals on the removal COD, BOD, organic nitrogen, TKN, drug, and MLVSS agrees with the previous qualitative analysis discussed earlier (Sections X ), since the removal efficiency and biomass concentrations and characteristics were mainly affected by the dominant factor which was the pharmaceutical concentration. This highly significant effect on removal efficiencies needed to be carefully addressed.

The primary and secondary wastewater treatment gave moderate to high removal efficiencies of pharmaceuticals. However, the effluent still had considerable concentrations of some of these drugs. These concentrations were in the range of 1-1000 ng/L, as most of the pharmaceuticals present in the influent were found in the effluent, which indicate the need for further treatments to remove these pollutants compounds.

Table 2: The significance level of the effect of pharmaceutical concentration and their interaction on the responses

Variables

Factor Concentration

p-value

Confidence level

Level of Significance

COD removal

Metronidazole

0

100

Highly Significant

Trimethoprim

0.002

99.8

Highly Significant

Sulphamethoxazole

0.06

94

Significant

Paracetamol

0.001

99.9

Highly Significant

Ranitidine

0

100

Highly Significant

BOD removal

Metronidazole

0

100

Highly Significant

Trimethoprim

0.003

99.7

Highly Significant

Sulphamethoxazole

0.064

93.6

Significant

Paracetamol

0.001

99.9

Highly Significant

Ranitidine

0

100

Highly Significant

Organic nitrogen

removal

Metronidazole

0

100

Highly Significant

Trimethoprim

0.002

99.8

Highly Significant

Sulphamethoxazole

0.053

94.7

Significant

Paracetamol

0.001

99.9

Highly Significant

Ranitidine

0

100

Highly Significant

TKN removal

Metronidazole

0

100

Highly Significant

Trimethoprim

0.003

99.7

Highly Significant

Sulphamethoxazole

0.063

93.7

Significant

Paracetamol

0.001

99.9

Highly Significant

Ranitidine

0

100

Highly Significant

MLVSS

Metronidazole

0

100

Highly Significant

Trimethoprim

0

100

Highly Significant

Sulphamethoxazole

0

100

Highly Significant

Paracetamol

0

100

Highly Significant

Ranitidine

0

100

Highly Significant

Drug removal

Metronidazole

0

100

Highly Significant

Trimethoprim

0.002

99.8

Highly Significant

Sulphamethoxazole

0.05

95

Highly Significant

Paracetamol

0.001

99.9

Highly Significant

Ranitidine

0

100

Highly Significant

5.1.5 Physical removal of pharmaceuticals during the WWTP

5.1.5.1 Physicochemical Characteristics

The physicochemical characteristics of the feed and permeate for the ultra filtration process during the sampling year are presented in Appendix. The average value of the physicochemical characteristics of the inlet of ultra filtration was pH (7.04), TSS (8.68mg/l), TDS (437.1mg/l), COD (24mg/l), BOD (4.07mg/l), total iron (1.38mg/l), and total coliforms (426261.2 colonies/100ml).

The average removal efficiencies of the ultra filtration process were 98%, 95% and 99% for the TSS, total iron and total coliforms measurements, while there were no significant changes in the TDS measurements. The concentrations of COD were measured for ultra filtration feed and permeate. Results did not show high removal of trace organic contaminants through the filtration processes where the average removal of COD was 42% while the BOD was 72%. Thus, the ultra filtration process provides an essential pre-treatment for the RO by removing particulate and colloidal material from the feed, but the removal is limited to particles larger than the membrane pore size (Van der Bruggen et al., 2003).

The average removal efficiencies of the RO process for the TSS and total coliforms measurements were 68% and 99%, while there was a highly significant removal in the TDS measurements with average removal 96%. Furthermore, the concentration measurements of BOD for RO feed and the permeate show high removal of trace organic contaminants through the filtration processes with average removal 90%.

Trace Organic Compounds may be completely or partially degraded in wastewater treatment a plant that takes place mostly in the activated biological sludge process. Pharmaceuticals fluctuate in their degradation in various wastewater treatment processes. The remaining pharmaceuticals are passed into the ultra filtration then to RO systems.

Pharmaceutical compounds were detected in the RO feed that was derived from the WWTP. These variations were formerly a result of annual fluctuations of compounds in raw wastewater in addition to other processes involved in wastewater treatment. Most of the pharmaceutical, namely metronidazole, Trimethoprim, sulphamethoxazole, paracetamol, and ranitidine were found in all samples from RO inlets during the sampling year. The average concentrations of these compounds found in the RO inlets were 4ng/l, 61ng/l, 47ng/l, 8ng/l, and 210ng/l for metronidazole, trimethoprim, sulphamethoxazole, paracetamol, ranitidine respectively (Figure 3a,b,c,d,e). The highest removal efficiency of these compounds was 97% for ranitidine then 92% for sulphamethoxazole and paracetamol as for Trimethoprim was 86%. Lastly, the lowest removal efficiency between these compounds was 56% for metronidazole due to low concentration found in RO inlets.

Figure 3a: Concentrations of metronidazole (ng/L) detected in the RO inlet and the removal percentage by RO process in the Sulaibiya WWTP during the year.

Figure 3b: Concentrations of trimethoprim (ng/L) detected in the RO inlet and the removal percentage by RO process in the Sulaibiya WWTP during the year.

Figure 3c: Concentrations of sulphamethoxazole (ng/L) detected in the RO inlet and the removal percentage by RO process in the Sulaibiya WWTP during the year.

Figure 3d: Concentrations of paracetamol (ng/L) detected in the RO inlet and the removal percentage by RO process in the Sulaibiya WWTP during the year.

Figure 3e: Concentrations of ranitidine (ng/L) detected in the RO inlet and the removal percentage by RO process in the Sulaibiya WWTP during the year.

All pharmaceuticals were found in all RO feeds, permeate and brine (Figure 4). The average concentrations of these pharmaceuticals found in the RO brine were up to 6, 22, 109, 5, and 35ng/L for metronidazole, trimethoprim, sulphamethoxazole, paracetamol, and ranitidine, respectively. In general, concentrations of these compounds in the brine were 16 times than concentration in feed for metronidazole; twice for trimethoprim, sulphamethoxazole and paracetamol; once for ranitidine.

Figure 4: Concentration of metronidazole, trimethoprim, sulphamethoxazole, paracetamol and ranitidine (ng/L) in RO feeds, permeate and brine for December 2010.

The solubility of these pharmaceuticals varies; some are moderately soluble such as sulphamethoxazole and Trimethoprim where the solubility was 281mg/l, 400 mg/L, respectively; some are highly soluble like ranitidine, paracetamol and metronidazole where the solubilities were 24.7g/l, 14g/l and 10g/l, respectively. Log Kow values of these pharmaceuticals ranged between -0.02 and 0.92. The plot of log Kow versus removal efficiency showed a weak positive (Figure 5). On the other hand, the solubility and log Kow did not correlate with the behaviour of these pharmaceuticals in the RO. According to Tolls (2001) found that log Kow may not be a good indicator of the behaviour of pharmaceuticals in the environment. It has been reported that the removal efficiency of solute by ultra filtration and RO is affected by different parameters such as pH, solute charge, molecular weight and geometry, polarity and hydrophobicity, as well as the membrane surface charge (Van der Bruggen et al., 1998; Van der Bruggen et al., 1999; Kiso et al., 2000; Kiso et al., 2001; Ozaki and Li, 2002; Kimura et al., 2003b; Kimura et al., 2004).

Figure 5: Plot of the water coefficient (log Kow) of the pharmaceutical against removal efficiency.

Investigations have been done by previous studies on the removal efficiency of RO compared to other types of membranes and low pressure reverse osmosis where they found a great advantage of using RO in producing a high quality of recycled water. According to Lopez-Ramirez et al., 2006 who found that the reclaimed wastewater for widely exceed the RO membrane the drinking water standards by. However, the removal results of the RO membrane represent highly reduced pollutants in permeate. Furthermore, micro-organisms were removed from the RO permeate, which would allow safe reuse of water.

5.1.5.2 Effect of temperature and pH on the removal efficiency

Similarly, to biological treatment, the temperature was affecting the removal processes in the RO system. The concentrations of target pharmaceuticals in the RO inlet samples during the sampling year was fluctuated during the summer and winter months. Therefore, effect of temperature and pH was also statically analysed using ANOVA. The correlation between the temperature and pH with the removal of BOD, TSS, TDS, total coliforms and all target pharmaceuticals was highly significant (Table 3).

Table 3: The significance level of the effect of temperature and pH, and their interaction on the responses

Variables

Factor

p-value

Confidence level

Level of Significance

BOD removal

Temp

0

100

Highly Significant

pH

0

100

Highly Significant

TSS removal

Temp

0

100

Highly Significant

pH

0

100

Highly Significant

TDS removal

Temp

0

100

Highly Significant

pH

0

100

Highly Significant

Total coliforms removal

Temp

0

100

Highly Significant

pH

0

100

Highly Significant

Metronidazole removal

Temp

0.008

99.2

Highly Significant

pH

0

100

Highly Significant

Trimethoprim removal

Temp

0

100

Highly Significant

pH

0

100

Highly Significant

Sulphamethoxazole removal

Temp

0

100

Highly Significant

pH

0

100

Highly Significant

Paracetamol removal

Temp

0

100

Highly Significant

pH

0

100

Highly Significant

Ranitidine removal

Temp

0

100

Highly Significant

pH

0

100

Highly Significant

In this study, there was no relationship between the removal of pharmaceuticals and the other removal parameter tested such as TSS, TDS, BOD, and total coliforms distributed in the RO streams. The regression analysis shows no correlation between the removal efficiency of pharmaceuticals and TSS, TDS, BOD, and total coliforms. This might be due to the complexity of RO feed in treatment plants and to a broad range of rejection. Therefore, it's very difficult to associate these operating parameters with these removal rates.

Although to the wide range of variability and limitation of data, there was no possible to determine the relationship between the removal and molecular weight or molecular size. According to Kimura et al., 2003 there was a linear relationship between molecular weight of the non-charged compounds and removal. However, in this study, the relationship between the molecular weight and the removal of metronidazole, trimethoprim and ranitidine was observed a linear regression while sulphamethoxazole and paracetamol were not on the regression line (Figure 6). The physico-chemical characteristics of tested drugs in this study differ from each other. Thus, a relationship between any of these removals could be described by different physico-chemical characteristics such the charge, shape and polarity of compounds. Steric hindrance is the main removal mechanisms in RO membranes referring to the electrostatic interaction and hydrophobic interaction between compounds and the membrane (Bellona et al., 2004). Removals by RO were investigated by many researchers suggesting that the removal may be influenced by dipole moment of compounds, hydrophobicity of compounds represented by Kow and molecular size (Ozaki and Li, 2002; Van der Bruggen et al., 2003). Positive correlation between hydrophobicity of non-phenolic compounds (log Kow) and their removal by nanofiltration was reported by Kiso et al., 2000. On the other hand, hydrophilic compounds do not adsorb to the membrane polymeric matrix Alturki et al., 2010. According to Snyder and co-workers, 2007 found that some compounds pass the RO, thus no clear relationship between molecular structure and membrane could be established. Penetration of the RO could be the result of diffusion into and through the membrane, short-circuiting of the membrane or supporting media failure. The removals of micropollutants by RO were determined by complex interactions of electrostatic and other physical forces acting between specific solute, the solution and the membrane. Furthermore, electrostatic attraction or repulsion forces can affect the removal of some micropollutant in RO membrane due to their charge (e.g. negative charge of sulphamethoxazole).

Figure 6: The relationship between the molecular weight and the removal efficiencies of metronidazole, trimethoprim, sulphamethoxazole, paracetamol and ranitidine.

The concentrations of pharmaceutical were reduced as these compounds passed through the ultra filtration systems in the water reclamation plant. Metronidazole, trimethoprim, sulphamethoxazole, paracetamol and ranitidine were detected in ultra filtration effluent with a maximum concentration at 13, 190, 73, 15, and 236ng/l, respectively.

Other researchers have found high levels of pharmaceuticals in the effluent of WWTP (Nakada et al., 2006; Roberts and Thomas, 2006; Gomez et al., 2007; Santos et al., 2007). Therefore, the removals of pharmaceuticals are more effective in the advanced treatment plant using RO systems than other conventional treatment plants (Snyder et al., 2007).

However, removal rates with RO membrane were high, which are in agreement with results obtained by other researchers (Alturki et al., 2010; Radjenovic et al 2008; Reznik et al 2011; Snyder et al., 2003; Dolar et al., 2012). Dolar et al., 2012 observed in pilot scale experiment using MBR with RO membranes that the majority of compounds studied in the influent were completely removed. Joss et al., 2011 reports that most organic micropollutants are removed or retained to below the detection limit by RO. Carbamazepine, sulphamethoxazole, metoprolol and sotalol were removed with high removal rates (>98%) using RO by researchers Radenovic et al., 2008 and Gur-Reznik et al., 2011.

5.1.6 Chlorination effect on the pharmaceuticals during the treatment

The effluents of the RO were treated further by chlorine before discharge. Most of the pharmaceutical escape from the RO system and the trace were detected in the RO effluent. The maximum concentration detected was 19ng/l and 15ng/l for Trimethoprim and ranitidine where the other pharmaceuticals were detected at 7ng/l, 5ng/l, and 4ng/l, respectively. On the other hand, the lowest concentration detected in the RO effluents for metronidazole, trimethoprim, sulphamethoxazole, paracetamol, and ranitidine were 0.2ng/l, 1ng/l, 0.2ng/l, 1ng/l, and 1ng/l, respectively. The incomplete removal of these pharmaceuticals at wastewater treatment plants have permitted their spread through surface waters (Boyd et al., 2003; Carballa et al., 2004; Kim et al., 2007; Metcalfe et al., 2003; Okuda et al., 2008; Paxeus, 2004; Reemtsma et al., 2006; Tauxe-Wuersch et al., 2005; Ternes et al., 1998), which is mainly the source of drinking water.

Sulaibiya wastewater treatment was design to treat the product water with chlorine before discharged. Samples were taken to follow the fate of these pharmaceuticals when treated with chlorine. The effect of chlorine was varying between 0 to 100% removal or transformation of these pharmaceuticals (Figure 7). However, most of the detected pharmaceuticals in the final effluent after chlorination were below 4ng/l or completely removed indicating a complete oxidation of the investigated pharmaceuticals in the presence of chlorine residue. Transformations of the pharmaceuticals may occur in the presence of chlorine residues during treatment.

Figure 7.a: Concentrations of metronidazole (ng/L) detected in the RO outlet and the removal percentage by chlorination process in the Sulaibiya WWTP during the year.Figure 7.b: Concentrations of trimethoprim (ng/L) detected in the RO outlet and the removal percentage by chlorination process in the Sulaibiya WWTP during the year.

Figure 7.c: Concentrations of sulphamethoxazole (ng/L) detected in the RO outlet and the removal percentage by chlorination process in the Sulaibiya WWTP during the year.

Figure 7.d: Concentrations of paracetamol (ng/L) detected in the RO outlet and the removal percentage by chlorination process in the Sulaibiya WWTP during the year.

Figure 7.e: Concentrations of ranitidine (ng/L) detected in the RO outlet and the removal percentage by chlorination process in the Sulaibiya WWTP during the year.

The complete removal of the pharmaceuticals after chlorination was expected since 90% removals have been previously reported for most sulfonamides and for trimethoprim in river water at a chlorine dose of 1 mg/l (Adams et al., 2002). Some of the pharmaceutical still remained at a noticeable level after RO (from 1 to 19ng/l) in the product water, the discharging of this water could cause potential ecological risks and / or the proliferation of bacterial resistance in downstream environments. However, chlorination, showed its ability to completely remove pharmaceuticals in recycled water but the complete removal of the toxicity of these pharmaceuticals has not yet been confirmed since the formation of chlorinated byproducts may be more harmful than their parent antibiotics (Von Gunten et al., 2006).

To confirm the removal of pharmaceuticals by chlorine a study was done on the chlorination of drugs in lab scale. Other treatments found in water reclamation plant such as ozonation. Therefore, all treatments was tested to compare the removal efficiency of these different treatments in different sample media such as synthetic wastewater, real wastewater and tab water to evaluate the effect of chlorine and ozone concentration in each drug. Researchers found that the residual of chlorine in tap water can react with organic pollutants and producing by-products (Canosa et al., 2006; Negreira et al., 2008). Previous studies have found that pharmaceuticals can be degraded under chlorination treatment (Pinkston and Sedlak, 2004; Westerhoff et al., 2005).

A first chlorination and ozonation test of the five drugs were performed in order to assess their degradability. Thus, they were treated for 30 min with a 5 and 10 mg/L chlorine and 5, 10 and 15 mg/l ozone. The removal percentage of each drug at the end of this experiment is shown in Figure 8(a,b,c,d,e). The removals of those pharmaceuticals were varied where metronidazole was the lowest and ranitidine was the highest by the oxidation process of chlorination and ozonation.

Metronidazole was affected by the doses of chlorine and ozone in different media such as synthetic wastewater, tab water and real wastewater. The removal was gradually increased with the increase of doses. The highest removal was 96% in tab water 10mg/l of chlorine while, in synthetic wastewater and real wastewater with the same doses of chlorine were 59 and 58% respectively. On the other hand, the effects of ozone doses were almost comparable as the highest removal of metronidazole in 15mg/l of ozone was in synthetic wastewater with rate 55% while was 51% in tab water and real wastewater.

Trimethoprim was almost disappeared in tab water and real wastewater at 10mg/l doses of chlorine with removal 99% while, the removal rate at same doses in synthetic wastewater was 92%. On the other hand, the removal rates of trimethoprim by ozonation were similar synthetic wastewater, tab water and real water with rate >84% at 15mg/l ozone dose.

Sulphamethoxazole was almost removed in synthetic wastewater, tab water and real wastewater at all doses of chlorine, whereas, the highest removal was at 15mg/l dose of ozone. The removal of sulphamethoxazole in synthetic wastewater, tab water and real wastewater was comparable and the removal rate was 90%, 92% and 95% respectively.

Paracetamol was removed in synthetic wastewater, tab water and real wastewater at different doses of chlorine and ozone with similar rate. The maximum removal found at 15mg/l chlorine dose with rate >96%, while, the average removal rate at 15mg/l ozone dose was 91%.

Ranitidine was highest removed compound between the other pharmaceuticals which almost completely removed at different doses of chlorine and ozone except 5mg/l ozone dose where the removal rate was 71%, 73% and 75% for synthetic wastewater, tab water and real wastewater respectively.

Figure 8.a: Effect of the initial chlorine and ozone dose on the metronidazole removal or transformation in experiments performed in different doses (5mgL-1 O3, 10mgL-1 O3, 15mgL-1 O3, 5mgL-1 cl2, and 10mgL-1 cl2) with synthetics wastewater, tap water and real wastewater.

Figure 8.b: Effect of the initial chlorine and ozone dose on the trimethoprim removal or transformation in experiments performed in different doses (5mgL-1 O3, 10mgL-1 O3, 15mgL-1 O3, 5mgL-1 cl2, and 10mgL-1 cl2) with synthetics wastewater, tap water and real wastewater.

Figure 8.c: Effect of the initial chlorine and ozone dose on the sulphamethoxazole removal or transformation in experiments performed in different doses (5mgL-1 O3, 10mgL-1 O3, 15mgL-1 O3, 5mgL-1 cl2, and 10mgL-1 cl2) with synthetics wastewater, tap water and real wastewater.

Figure 8.d: Effect of the initial chlorine and ozone dose on the paracetamol removal or transformation in experiments performed in different doses (5mgL-1 O3, 10mgL-1 O3, 15mgL-1 O3, 5mgL-1 cl2, and 10mgL-1 cl2) with synthetics wastewater, tap water and real wastewater.

Figure 8.e: Effect of the initial chlorine and ozone dose on the ranitidine removal or transformation in experiments performed in different doses (5mgL-1 O3, 10mgL-1 O3, 15mgL-1 O3, 5mgL-1 cl2, and 10mgL-1 cl2) with synthetics wastewater, tap water and real wastewater.

5.1.6.1 Chlorination

The effective removal of pharmaceuticals by chlorination from water requires sufficient free chlorine concentration and contact time. Chlorination can degrade or transformation the chemical compounds via one of two pathways; one by chorine substitution or addition reactions, which may alter active functional groups; second chlorine radicals may oxidizing (break down) the target compound such as pharmaceutical into smaller molecules, which may or may not possess the active properties (Crain and Gottlieb, 1935).

Adams et al., 2002 has been reported 90% removal for most sulfonamides and trimethoprim when use free chlorine at 1.0 mg/l with contact times greater than 16 min and greater than 40 min in river water respectively.

A study by Gibs et al., 2007 on the effect of presence free chlorine on transformation of some pharmaceutical compounds in drinking water during distribution found 50%- 80% for sulphonamides and 42% for trimethoprim were removed after one day and completely removed after 10 days. At concentration 3.5- 3.8 mg/l of free chlorine, 90% to 99% removals of were achieved for sulphamethoxazole and trimethoprim in river water after 24 h contact time (Westerhoff et al., 2005). HClO and ClO2 oxidises sulphamethoxazole at specific functional groups with high electron densities, such as neutral tertiary amines and aniline (Huber et al., 2005). On the other hand, rapid and substantial transformation of trimethoprim to a wide range of chlorinated and hydroxylated products is expected to occur under typical conditions of wastewater and drinking water chlorination (Dodd and Huang, 2007).

5.1.6.2 Ozonation

Other researchers have reported the effective treatment of ozonation for removal of pharmaceuticals in water and wastewater effluents (Adams et al., 2002; Ternes et al., 2003; Huber et al., 2005). Study of Adams et al., 2002 found that ozonation removed more than 95% of several sulphonamides and trimethoprim from river water within 1.3 min contact time at an ozone dose of 7.1 mg/l. Huber et al., 2005 also observed that at doses >2mg/l of ozone oxidised 90% - >99% of sulphonamides in secondary wastewater effluents.

Oxidative degradation of organic chemicals by ozone treatment can occur either by direct reaction with molecular ozone (O3) or indirectly via hydroxyl radicals (Staehelin and Hoigne, 1985). During wastewater ozonation, many antibiotics, including sulphonamides have been shown to be predominantly transformed via direct reaction with ozone (Dodd et al., 2006. The oxidation pathway will depend on the ratio of molecular ozone and hydroxyl radicals, the corresponding reaction kinetics, and presence of organic matter (Elovitz et al., 2000, von Gunten, 2003). Ozone and/or hydroxyl radicals deactivate bactericidal properties of antibiotics by attacking or modulating their pharmaceutically active functional groups, such as aniline moieties of sulphonamides (Huber et al., 2005) and the phenol ring of trimethoprim (Dodd et al., 2009). The good removal (>90%) by ozonation was observed for those compounds with electron-rich aromatic systems, such as hydroxyl, amino (e.g. sulphamethoxazoles), acylamino, alkoxy and alkyl aromatic compounds, as well as those compounds with deprotonated amine (e.g. trimethoprim) and nonaromatic alkene groups since this key structural moieties are highly amendable to oxidative attack (Dickenson et al., 2009).

Adams et al. (2002) found more than 95% conversion of trimethoprim by ozonation in a pre-filtered river water sample spiked with these antibiotics at an initial concentration of 50 mg/l. similar reactivity of trimethoprim and sulphamethoxazole by ozonation in wastewater (Ternes et al., 2003). Ternes et al., 2003 confirmed that 5 mg/L of applied ozone could completely remove 0.62 mg/L sulphamethoxazole present in a biologically treated municipal wastewater. Similar results were also reported elsewhere (Huber et al., 2003, 2005). Paracetamol was effectively degraded by ozone, which could degrade 0.8 g/L paracetamol in 30 min with an ozone flow rate of about 72 g/h (Andreozzi et al., 2003). A number of degradations intermediates were found during the ozone treatment follows typical phenol ozonation pathways, such as hydroxylation of phenol ring, anomalous ozonation to cleave aromatic ring of hydroquinone, and decarboxylation by hydroxyl radicals (Andreozzi et al., 2003).

5.1.7 Overall removal of pharmaceuticals during wastewater treatment

Modern WWTP can effectively remove the micro pollutants, as well as microbial pollution, which also receives a large number of different traces organic polluting compounds, among them pharmaceuticals. Conventional treatments have not been specifically designed to remove these pharmaceuticals (Suárez et al., 2008). Therefore, pharmaceuticals may occur in effluents, because either they do not have the tendency to adsorb onto activated sludge or their biodegradation was not fast enough to be completed within the hydraulic retention time. However, using further treatment such as filtration (ultrafiltration, nanofiltration and RO) and chemical used for disinfection (chlorine and ozone) beside the conventional treatment increased the efficiency of WWTP.

Sulaibiya WWTP consists of screen and grit removal in Ardiya then pumped it to an activated sludge process (biological treatment). Secondary effluent was then sent to ultra-filtration and reverse osmosis; the product water was chlorination before discharge it. Pharmaceuticals were followed throw the overall treatment to evaluate the removal process. Box plots figure to indicate levels of pharmaceuticals found in influent and the effluent of each process in the Sulaibiya WWTP are shown in Figure 9. Box plot summaries the removals process of drugs during full-scale treatment of municipal wastewater. These graphics were built from 12 sampling period measurement for all sample locations.

5.1.7.1 Biological removal

Metronidazole concentration in influent was low compare to other drugs, and at all sampling time was highly variable, with a median concentration of 15ng/L and a mean of 22.33ng/L. Statistical analysis of the data from Table 2 showed that the biological removal of metronidazole was highly significant where the median and mean concentration of the biological effluent was 3ng/L and 6.28ng/L respectively. On the other hand, trimethoprim concentration in the influent was high and also highly variable from one point outside the box plot, and the median concentration was 470ng/L while the mean concentration was 622ng/L. The biological removal of trimethoprim was also statistically highly significant where the median and mean concentration of the effluent was similar 107.5ng/L. Similarly, sulphamethoxazole concentration in the influent was high and moderate variable with two points outside the box plot, and the median concentration was 241ng/L while the mean concentration was 365ng/L. Highly significant biological removal of sulphamethoxazole to reduce the concentration, where the median and mean concentration of the effluent was 66.5ng/L and 134.1ng/L respectively, with two points outside the box plot. However, paracetamol concentration in the influent was also high and highly variable, with a median concentration of 740ng/L and a mean of 881ng/L. Paracetamol mostly is effectively removed in activated sludge wastewater treatment plant with near 100% efficiency and statistically highly significant; the median and mean concentration of the effluent was 10.5ng/L and 15.25ng/L respectively. Lastly, ranitidine concentration in the influent was high and also highly variable from one point outside the box plot, and the median concentration was 624ng/L while the mean concentration was 812ng/L. The biological removal of ranitidine was also statistically highly significant where the median and mean concentration of the effluent was 199ng/L and 320ng/L, respectively, with one points outside the box plot.

The treatment of secondary effluent with sand filtration is intended to remove suspended solids and turbidity that persist after clarification. Pharmaceutical degradation can also occur in these systems through, further biological degradation within the biofilms on the filter media (Gobel et al., 2007), adsorption to the solids is also possible.

Metronidazole was removed in the sand filtration during the sampling period by 50% where the mean and median concentrations in the sand filtration effluent were 3.56ng/L and 1.5ng/L, respectively with one point outside the box plot 13ng/L. Similarly, trimethoprim was removed by 53% where the mean concentration was 61.3ng/L and the median concentration was 50ng/L with one point outside the box plot 253ng/L. In contrast, sulphamethoxazole was removed with nearly constant concentrations in the effluent with mean 47.2ng/L and median 46ng/L. Paracetamol was removed by 57% where the mean and median concentrations in the sand filtration effluent were 8.58ng/L and 4.5ng/L, respectively. However, ranitidine was the lowest removal by sand filtration with 16% where the mean and median concentration in the effluent were 210.1ng/L and 166.5ng/L, respectively, with one point outside the box plot 1057ng/L, which high compare to other removed drugs.

Metronidazole, trimethoprim and paracetamol reveal removal efficiencies (≥50%), while sulphamethoxazole and ranitidine were 31% and 16%, respectively, when sand filtration is employed. These removals by sand filters are attributable to biological activity or adsorption, so it could predict from structural and physical properties of those pharmaceuticals what will be more susceptible to treatment. Although to the observations with sorption tendencies (i.e., correlation to Kow values), the highest removal efficiencies were obtained for pharmaceuticals were previously identified as highly removed during the active sludge process. Moreover, there are some influences of operational variables on pharmaceutical removal during sand filtration such as hydraulic residence time, hydraulic loading rate and bulk water quality characteristics. Gobel et al., 2007 found significant differences in the removal of trimethoprim (15% versus 74%) in two sand filters comparable hydraulic retention times and hydraulic loading rates per biofilm surface area in each case. Furthermore, Nakada et al., 2007 suggest that the removal of 24 different pharmaceutical during sand filtration. Similarly, Gobel et al., 2007 observed that some of the pharmaceutical eliminated to the greatest extent during sand filtration.

5.1.7.2 Pharmaceuticals removal by membrane filtration

The efficiency of pharmaceutical removal by membrane filtration depends on different parameters such as molecular weight cut-off, hydrophobicity, surface roughness, and charge (Bellona et al., 2004). Furthermore, depend on different physio-chemical parameters of the pharmaceutical compounds which can influence their retention such as the molecular weight and size, acid dissociation constant (pKa), octanol-water partitioning coefficient (Kow), polarity and aqueous diffusion coefficient (Bellona et al., 2004 and Bellona et al., 2005).

Pharmaceutical removal in UF and RO process in wastewater treatment is sufficient compare to the previous treatment. The concentration of pharmaceutical was low, where median concentrations was in the UF influent 1.5, 50, 46, 4.5 and 166.5 ng/L for metronidazole, trimethoprim, sulphamethoxazole, paracetamol and ranitidine, respectively. The median removal by UF was 0%, 44%, 64%, 77% and 74% for metronidazole, trimethoprim, sulphamethoxazole, paracetamol and ranitidine, respectively. In contrast, pharmaceutical removal in RO process is high compare to UF process where the removal rate was 33%, 82%, 94%, 100% and 97% for metronidazole, trimethoprim, sulphamethoxazole, paracetamol and ranitidine, respectively. RO is characterized to be the very high pharmaceutical removal efficiencies. However, some pharmaceuticals have been detected in RO permeate, and their breakthrough cannot be rationalized by their physical-chemical properties.

5.1.7.3 Pharmaceuticals removal by chlorination

Box plot analysis reveals that most of the pharmaceutical escape from the RO removal process, and the detected median concentrations of pharmaceuticals were ranging between 0.667 to 5ng/L. Chlorination process in all effluent were achieved pharmaceutical removals greater than 86%, while the highest concentrations were 3, 3, 1, and 4 for metronidazole, trimethoprim, sulphamethoxazole and ranitidine, respectively. Laboratory investigations with model systems have convincingly demonstrated that chlorination of common pharmaceutical can lead to the formation of known toxicants and probable carcinogens. Dodd and Huang 2007 found that chlorination reacted with trimethoprim, and the products were predominantly multi-chlorinated and hydroxylated. Bedner and MacCrehan 2006 demonstrated that free chlorine doses typically used in water treatment could react with paracetamol and led to the production of several products.

Figure 9: Box plot showing pharmaceuticals removal efficiencies in Sulaibiya wastewater treatment plant. Plots show the distribution of removal efficiencies, expressed in terms of fraction of pharmaceuticals remaining after each treatment approach. The solid line in each box represents the median.