Dependent Indicators for Environmental Evaluations of Desalination Plants
131
LSI (indicators 3, 6, 11, 12 and 14). The procedure used in estimating the values of
these indicators, except free chlorine (indicator 12) is described in Al-Sharrah et al.
(2017). For plant (a), the missing values for EC, Na concentration, total hardness
and LSI were estimated as 72 μS/cm, 16,385 ppm, 734 ppm and 2.5, respectively.
For plant (b), the total hardness was estimated as 45 ppm. Moreover, the LSI values
for plants (d) and (e) were estimated as −0.63 and 0.29, respectively (Al-Sharrah
et al. 2017).
The remaining missing values are the free chlorine concentration (indicator 12)
for all plants except Um-Quwain (plant c). The importance of this indicator and
whether it can be accurately estimated through dependencies will be studied first.
The final decision will be dependent on the ability to proceed with the analysis in
case the indicator has been excluded. As demonstrated in the previous case studies,
the indicators to objects ratio is a reliable measure for exclusion of indicators.
The source of free chlorine in the brine is disinfection of feed seawater, which
is essential to prevent biofouling in desalination processes. Chlorine is toxic to
the aquatic organisms; hence, it is crucial that chlorination is tightly controlled to
achieve the desired task with reduced harm to the marine life. The concentration of
free chlorine is dependent on several factors related to the chemical and physical
conditions of seawater and the flushing ability of the coastal zone (Hamed et al.
2017). Estimation of the free chlorine concentration is not possible with available
dependencies and correlations. In effect, free chlorine can be only determined
by chemical analysis. Hence, the free chlorine indicator cannot be used for this
problem, and the question is whether it can be safely excluded from further analysis.
Going back to the data in Table 5, the number of indicators after excluding the
SiO 2 indicator is 13 and the number of objects is 5, which results in the ratio of
indicators to objects equals to 2.6 prior to excluding the free chlorine indicator
and 2.4 after exclusion. Therefore, exclusion of the free chlorine indicator may be
justified because the ratio is greater than two.
Two scenarios will be presented for ranking the environmental performance
of the five plants by the Copeland method. The first scenario utilizes the 12
remaining indicators (excluding SiO 2 and free Cl 2 ), while the second one considers
the indicators with complete data only (excluding indicators 3, 6, 11, 12, 13 and
14). Hence, the ranking will be performed using 12 × 5 and 8 × 5 datasets. The
ranking results for the two datasets are shown in Table 6. The reported ranks are
normalized in the range from one to zero, where a higher rank indicates more serious
environmental impact.
The ranking results in Table 6 indicate that the ranks for both cases are the same.
The SRCC value calculated for the 8 indicators with respect to 12 indicators is
0.985, which is reasonably high. The results show also that the worst plant in terms
of environmental performance due to brine discharge is Abu-Fintas while the best is
Um-Quwain. These are clear cuts in a sense that identical results have been obtained
from complete and reduced sets of indicators. However, such clarity in decisions
cannot be always achieved with low number of indicators. For instance, the ranks of
Ajman and Um-Quain (plants b and c) can be compared in more confidence when
using 12 indicators. In fact, for the 8 indicators case, it is not realistic to conclude
131
LSI (indicators 3, 6, 11, 12 and 14). The procedure used in estimating the values of
these indicators, except free chlorine (indicator 12) is described in Al-Sharrah et al.
(2017). For plant (a), the missing values for EC, Na concentration, total hardness
and LSI were estimated as 72 μS/cm, 16,385 ppm, 734 ppm and 2.5, respectively.
For plant (b), the total hardness was estimated as 45 ppm. Moreover, the LSI values
for plants (d) and (e) were estimated as −0.63 and 0.29, respectively (Al-Sharrah
et al. 2017).
The remaining missing values are the free chlorine concentration (indicator 12)
for all plants except Um-Quwain (plant c). The importance of this indicator and
whether it can be accurately estimated through dependencies will be studied first.
The final decision will be dependent on the ability to proceed with the analysis in
case the indicator has been excluded. As demonstrated in the previous case studies,
the indicators to objects ratio is a reliable measure for exclusion of indicators.
The source of free chlorine in the brine is disinfection of feed seawater, which
is essential to prevent biofouling in desalination processes. Chlorine is toxic to
the aquatic organisms; hence, it is crucial that chlorination is tightly controlled to
achieve the desired task with reduced harm to the marine life. The concentration of
free chlorine is dependent on several factors related to the chemical and physical
conditions of seawater and the flushing ability of the coastal zone (Hamed et al.
2017). Estimation of the free chlorine concentration is not possible with available
dependencies and correlations. In effect, free chlorine can be only determined
by chemical analysis. Hence, the free chlorine indicator cannot be used for this
problem, and the question is whether it can be safely excluded from further analysis.
Going back to the data in Table 5, the number of indicators after excluding the
SiO 2 indicator is 13 and the number of objects is 5, which results in the ratio of
indicators to objects equals to 2.6 prior to excluding the free chlorine indicator
and 2.4 after exclusion. Therefore, exclusion of the free chlorine indicator may be
justified because the ratio is greater than two.
Two scenarios will be presented for ranking the environmental performance
of the five plants by the Copeland method. The first scenario utilizes the 12
remaining indicators (excluding SiO 2 and free Cl 2 ), while the second one considers
the indicators with complete data only (excluding indicators 3, 6, 11, 12, 13 and
14). Hence, the ranking will be performed using 12 × 5 and 8 × 5 datasets. The
ranking results for the two datasets are shown in Table 6. The reported ranks are
normalized in the range from one to zero, where a higher rank indicates more serious
environmental impact.
The ranking results in Table 6 indicate that the ranks for both cases are the same.
The SRCC value calculated for the 8 indicators with respect to 12 indicators is
0.985, which is reasonably high. The results show also that the worst plant in terms
of environmental performance due to brine discharge is Abu-Fintas while the best is
Um-Quwain. These are clear cuts in a sense that identical results have been obtained
from complete and reduced sets of indicators. However, such clarity in decisions
cannot be always achieved with low number of indicators. For instance, the ranks of
Ajman and Um-Quain (plants b and c) can be compared in more confidence when
using 12 indicators. In fact, for the 8 indicators case, it is not realistic to conclude