work [62] focused on the role of CW design (SF, HSSF, VF and hybrid systems) and
respective operational factors, making use of correlation analysis to establish links
with pharmaceuticals removal efficiency. It should be mentioned that these latest
review papers [28, 62] pre-filtered the 25–30 compounds for which the number of
data points was sufficient to provide statistically significant results. Therefore, one
cannot find updated information for the more than 100 different pharmaceuticals that
have been reported at least once in a paper about CWs [27, 50]. To find metadata on
specific compounds, the review papers from 2014 are a better reference.
Ilyas and van Hullebusch [28] compared the performance of the different CWs,
SF, HSSF, VF and hybrid systems (Table 2). In addition, besides looking into the
removal of the 29 more common pharmaceuticals, the authors included as well
information about 19 transformation products (for the specific name of the compounds see Table 2). This provided the first overview on the dynamics of transformation products, including formation and removal, in CWs. Hybrid designs were
considered to perform better, followed by VF, HSSF and SF by Ilyas and van
Hullebusch [28]. Results provide an interesting overview of the differences among
designs. However, one should bear in mind the large standard deviation (Table 2)
due to the compromise of including for each CW design, results from lab to full-scale
systems and from primary to tertiary treatment.
2.3 Performance
A careful look at Table 2 shows that for 17 of the 29 common compounds,
differences among designs were not significant. VF seems to perform better for
caffeine, ibuprofen and naproxen, while hybrid designs have shown better removal
for diclofenac and trimethoprim. Overall, the differences pinpoint that removal can
be compound-specific and that part of the efficiency can be explained by the removal
mechanisms associated with each design. A very good example is the different usage
of VF (for oxidation of pollutants) and HSSF (explored as anaerobic of facultative
systems) that can result in different removal efficiencies for the same compound
(e.g. ibuprofen and diclofenac). Thus, compounds removed more efficiently by VF
tend to be mainly driven by aerobic degradation, while anaerobic degradation prevails for HSSF. Hybrid systems, and the coexistence of aerobic and anaerobic
conditions, seem to favour other compounds, such as diclofenac [63, 64]. However,
other factors such as the retention time can play a role. More on processes is detailed
in the next section.
Regarding specifically CWs for secondary wastewater treatment and tertiary
treatment, a wide range of variability for several compounds is consistently seen in
all 2014 revisions. Further reading [27, 49] is recommended to ascertain detailed
comparisons between compounds, designs and treatment stages. Here the best
performances from Table 2 and major conclusions from the last-mentioned reviews
are listed per removal range:
Constructed Wetlands and Phytoremediation as a Tool for Pharmaceutical Removal
391
respective operational factors, making use of correlation analysis to establish links
with pharmaceuticals removal efficiency. It should be mentioned that these latest
review papers [28, 62] pre-filtered the 25–30 compounds for which the number of
data points was sufficient to provide statistically significant results. Therefore, one
cannot find updated information for the more than 100 different pharmaceuticals that
have been reported at least once in a paper about CWs [27, 50]. To find metadata on
specific compounds, the review papers from 2014 are a better reference.
Ilyas and van Hullebusch [28] compared the performance of the different CWs,
SF, HSSF, VF and hybrid systems (Table 2). In addition, besides looking into the
removal of the 29 more common pharmaceuticals, the authors included as well
information about 19 transformation products (for the specific name of the compounds see Table 2). This provided the first overview on the dynamics of transformation products, including formation and removal, in CWs. Hybrid designs were
considered to perform better, followed by VF, HSSF and SF by Ilyas and van
Hullebusch [28]. Results provide an interesting overview of the differences among
designs. However, one should bear in mind the large standard deviation (Table 2)
due to the compromise of including for each CW design, results from lab to full-scale
systems and from primary to tertiary treatment.
2.3 Performance
A careful look at Table 2 shows that for 17 of the 29 common compounds,
differences among designs were not significant. VF seems to perform better for
caffeine, ibuprofen and naproxen, while hybrid designs have shown better removal
for diclofenac and trimethoprim. Overall, the differences pinpoint that removal can
be compound-specific and that part of the efficiency can be explained by the removal
mechanisms associated with each design. A very good example is the different usage
of VF (for oxidation of pollutants) and HSSF (explored as anaerobic of facultative
systems) that can result in different removal efficiencies for the same compound
(e.g. ibuprofen and diclofenac). Thus, compounds removed more efficiently by VF
tend to be mainly driven by aerobic degradation, while anaerobic degradation prevails for HSSF. Hybrid systems, and the coexistence of aerobic and anaerobic
conditions, seem to favour other compounds, such as diclofenac [63, 64]. However,
other factors such as the retention time can play a role. More on processes is detailed
in the next section.
Regarding specifically CWs for secondary wastewater treatment and tertiary
treatment, a wide range of variability for several compounds is consistently seen in
all 2014 revisions. Further reading [27, 49] is recommended to ascertain detailed
comparisons between compounds, designs and treatment stages. Here the best
performances from Table 2 and major conclusions from the last-mentioned reviews
are listed per removal range:
Constructed Wetlands and Phytoremediation as a Tool for Pharmaceutical Removal
391