International Journal of Engineering & Technology, 7 (3) (2018) 1762-1767

International Journal of Engineering & Technology

Website: www.sciencepubco.com/index.php/IJET
doi: 10.14419/ijet.v7i3.13593
Research paper

Water quality analysis and model simulation for the second

largest polluted lake in Egypt
Diaa Seif1 1 *, Mahmoud Nasr 1, Mohamed R. Soliman 2, Medhat Moustafa 1, Walid Elbarki 1
1 Sanitary Engineering Department, Faculty of Engineering, Alexandria University, Egypt.
2 Irrigation and Hydraulics Engineering Department Faculty of Engineering, Alexandria University, Egypt
*Corresponding author E-mail: eng-diaa.hamdy@alexu.edu.eg

Abstract

This study investigated the spatial variation in the water quality parameters of Burullus Lake using multivariate analysis and MIKE21
model. The lake was classified into zone-1 at north-east (Z1), zone-2 at south-east (Z2), zone-3 at north-middle (Z3), zone-4 at south-
middle (Z4), zone-5 at north-west (Z5), zone-6 at south-west (Z6), and zone-7 at west (Z7). The obtained parameters were temperature
21.5±5.0 ºC, pH 8.2±0.6, dissolved oxygen (DO) 5.9±1.0 mg/L, biological oxygen demand (BOD) 23.9±5.7 mg/L, NH 3-N 2.5±0.3 mg/L,
NO2-N 1.9±0.3 mg/L, NO3-N 1.2±0.3 mg/L, PO4-P 1.9±0.3 mg/L, SiO4 3.2±0.1 mg/L, Chlorophyll-a (Chl-a) 88.2±10.8 µg/L, and salinity
3.2±1.0 g/L. Principal component analysis showed that agricultural drainage water was the key factor influencing the water quality char-
acteristics of Burullus Lake. Water quality index (WQI) varied between “Bad” to “Medium”, suggesting that the lake wasn’t suitable for
irrigation and fish growth; however, it was appropriate for some aquatic life. A MIKE21 model was developed to provide a recommendation
scenario that could be used to enhance the water quality of Burullus Lake. By improving the water quality of precise drains (namely drains
7 and 8), the WQI at Z4 and Z6 modified from “Bad” to “Medium”. The period required to achieve this self-purification was 5 months.

Keywords: MIKE21 Simulator; Multivariate Statistical Analysis; Spatial Variation; Water Quality Index.

2000–2015 according to collected and measured data, (b) identify

1. Introduction the relations between water quality parameters using multivariate
analysis, (c) estimate the water quality index (WQI), and (d) inves-
Burullus Lake, receiving a total drainage discharge of 3.9×109 m3/y, tigate the possibility of improving the water quality using MIKE21
is the second largest lake in Egypt. The lake has been recognized as modeling.
a consistent water body for the fulfillment of domestic, agricultural,
and industrial requirements [1]. Unfortunately, Burullus Lake has
become a dumping site for huge amounts of agricultural runoff
2. Material and methods
coming from various sectors/lands of the Delta region. Currently,
the lake is facing many challenges, such as eutrophication, high 2.1. Study area
load of contamination, reduction of area, and depletion of fish pro-
duction [2]. Hence, extensive studies should be conducted to deter- Burullus Lake is located in the North-western region of the Nile
mine the water quality of Burullus Lake. Delta; i.e., longitude 30° 33’–31° 07’ E and latitude 31° 22’–31° 26’
Principal component analysis (PCA), and cluster analysis (CA) N (figure 1). The lake has dimensions of 47 km in length, 6–14 km
have been used to identify water pollution sources and assess water in width, and 0.4–2.1 m in depth, with an approximate surface area
quality condition. PCA provides deep information on the most of 410 Km2[5].
meaningful parameters by presenting statistical correlations be-
tween measurements [3]. PCA of the normalized variables is exe-
cuted to extract significant principal components (PCs) by reducing Z1
the data set with minimal loss of original information. CA is a group EB
Z3
of multivariate techniques performed to describe the relationships Z2
MB
between data records. In CA, related parameters are grouped into
certain distinct classes in which the parameters within a group are Z5
Z4
WB Z6
similar to each other (achieve homogeneity). A software package,
Z7
namely MIKE21, has also been used for modelling various aspects
of water quality, complex ecosystem behavior, and ecological ap-
EB= Eastern basin
plications [4]. It is used to simulate several chemical, physical, and WB= Western basin
MB= Middle basin
biological processes in water bodies. The Mike21 software can sim- Z= Zone
Sample point location
ulate water environments such as estuaries, oceans, lagoons, and
coasts.
Hence, this study was conducted to (a) provide an insight into spa- Fig. 1: Study Area of Burullus Lake Showing the Zones and Sampling Point
tial variations of water quality parameters for Burullus Lake during
Copyright © 2018 Diaa Seif1 et. al. This is an open access article distributed under the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
 International Journal of Engineering & Technology 1763

2.2. Water resources at north-west (Z5), zone-6 at south-west (Z6), and zone-7 at west
(Z7).
Burullus Lake water is of brackish nature. A mixture of drain and
seawater is the two main sources to the lake. Seepage from the cul- 2.5. Multivariate analysis
tivated lands and rainfall in winter are secondary sources [2]. Hu-
man activities wastes are discharged to the lake. According to and Principal component analysis is designed to rotate the original var-
EEA reports since 2001 to 2015, all drainage system provides the iables into new and orthogonal variables (axes), called principal
lake with average four hundred million cubic meters annually in- components (PCs). This step can be achieved by transforming the
creased to be in 2009 about six hundred million cubic meters. original data matrix into a product of two matrices, one of which
contains the information about the samples and the other about the
2.3. Analytical analysis variables [3].
The PCs lie along the directions of maximum variance, and thus,
In this study, the selected water quality parameters were tempera- provide information on the most meaningful parameters. All calcu-
ture, pH, dissolved oxygen (DO), salinity, biological oxygen de- lations were performed using MATLAB R2015a.
mand (BOD), ammonia-nitrogen (NH3-N), nitrite-nitrogen (NO2-
N), nitrate-nitrogen (NO3-N), phosphorus (PO4-P), silicate (SiO4) 2.6. Water quality index calculation
and Chlorophyll-a (Chl-a) as shown in table 1. The measured pa-
rameters were compared with guidelines of the water quality stand- Water quality index (WQI) is basically a mathematical tool used to
ards of the UNEC and WHO [6-7]. calculate a single output (indicator) from multiple water quality pa-
Temperature and pH were monitored on site with a mercury ther- rameters. The calculation of WQI for Burullus Lake was conducted
mometer and a digital portable pH meter, respectively. BOD and using weighted arithmetic index method. This method classified the
DO estimated by using Winkler method titration & membrane elec- water quality according to the degree of purity, using the most com-
trode method. Water salinity was measured using Bechman sali- monly measured water quality variables, such as temperature, pH,
nometer (Model NO. R.S.10). Nutrient salts (N and P species) and turbidity, faecal coliform, dissolved oxygen, biochemical oxygen
Chl-a were measured using a Shimadzu double beam spectropho- demand, total phosphates, nitrates and total solids [8, 9 ,10].
tometer UV-150-02. In this method, each water quality parameter is transferred to a
weighting-curve chart in order to obtain a numerical value (Qi). The
2.4. Data classification (Qi) is multiplied by a weighting factor (Wi), as described in Eq. 1.
Basically, WQI value “water quality status” can be classified as, 0
The comprehensive database elucidated seven zones in the lake, as – 25 “Very bad”, 26 – 50 “Bad”, 51 – 70 “Medium”, 71 – 90
follows: zone 1 (Z1), (Z2), (Z3), (Z4), (Z5), (Z6) and (Z7) (Figure. “Good”, and 91 – 100 “Excellent”[11].
1). This classification was proposed in order to define the best spa-
WQI = i=1 QiWi
n
tial framework for an efficient and reliable water quality assessment
(1)
of the Lake. Winter data during January – March, spring data taken
during April – June, summer data during July – September, while
Where:
fall data samples from October to December.
(Qi) is the sub-index for the water quality parameter; (Wi) is weight
The spatial variation of water quality parameters was classified into
associated with the water quality parameter; (n) is number of water
(Figure 1) zone-1 at north-east (Z1), zone-2 at south-east (Z2),
quality parameters.
zone-3 at north-middle (Z3), zone-4 at south-middle (Z4), zone-5

Table 1: Average Measurement Parameters of Burullus Lake during Seasons

zone Tem ºC pH Salinity g/L DO mg/L NH3 mg/L NO2 mg/L NO3 mg/L Po4 mg/L Chlorophyll g/L Sio4 mg/L BOD mg/L
Z1 win 16.63 7.65 12.14 8.44 1.55 0.36 1.06 0.59 32.33 2.73 10.04
Z1 spring 24.87 8.77 10.07 7.70 1.38 0.28 0.94 0.45 20.33 2.60 13.62
Z1 summer 27.79 8.87 8.24 7.45 1.32 0.21 0.70 0.38 28.00 2.48 13.66
Z1 Fall 18.55 8.15 10.63 7.94 1.46 0.36 1.03 0.56 32.55 2.60 14.14
Z2 win 16.63 7.71 9.39 7.35 3.92 2.95 1.86 2.76 67.83 4.67 17.78
Z2 spring 24.03 8.83 5.27 5.44 3.69 2.66 1.74 2.10 88.57 4.54 26.78
Z2 summer 27.79 8.93 3.31 4.71 3.52 2.51 1.50 1.76 93.50 4.43 34.17
Z2 Fall 18.29 8.44 8.10 6.44 3.44 2.29 1.57 2.28 68.22 4.55 19.99
Z3 win 15.47 7.51 3.89 7.82 1.08 0.54 0.76 0.81 54.67 3.83 16.08
Z3 spring 23.84 8.63 2.08 7.28 1.02 0.42 0.64 0.69 75.40 3.72 22.60
Z3 summer 27.19 8.73 1.04 7.02 0.97 0.36 0.45 0.52 82.80 3.63 29.81
Z3 Fall 17.24 7.84 2.49 7.49 0.92 0.49 0.70 0.71 62.89 3.72 20.69
Z4 win 16.40 7.16 2.26 6.05 4.00 2.26 1.87 4.28 99.00 4.82 16.78
Z4 spring 22.02 8.28 1.14 5.04 3.80 1.56 1.75 3.43 119.73 4.69 28.65
Z4 summer 27.87 8.38 0.68 4.43 3.69 1.41 1.51 2.43 127.13 4.57 32.76
Z4 Fall 18.08 7.99 1.83 6.25 3.66 1.77 1.17 3.25 93.59 4.67 21.99
Z5 win 14.10 7.34 0.88 6.77 1.60 0.68 1.28 1.59 52.33 2.05 13.58
Z5 spring 20.82 8.46 0.56 5.96 1.45 0.27 1.16 1.15 73.07 1.92 17.90
Z5 summer 26.19 8.56 0.36 5.70 1.38 0.17 0.26 0.90 80.47 1.81 25.51
Z5 Fall 18.53 7.78 0.75 6.60 1.32 0.46 0.77 1.32 63.12 1.90 18.99
Z6 win 17.40 7.26 0.79 6.84 4.90 1.18 2.32 5.10 106.00 2.94 30.17
Z6 spring 22.19 8.38 0.46 3.41 4.67 0.65 2.20 4.80 124.40 2.81 50.33
Z6 summer 28.25 8.48 0.29 2.94 2.21 0.50 1.35 4.54 131.80 2.70 51.50
Z6 Fall 21.15 8.04 0.79 5.32 3.06 0.87 1.64 4.83 114.40 2.73 36.13
Z7 win 18.40 7.51 0.74 6.04 2.35 0.79 1.20 0.32 140.17 2.31 19.08
Z7 spring 24.02 8.63 0.42 2.96 2.34 0.59 1.08 0.20 141.73 2.18 22.80
Z7 summer 28.33 8.73 0.24 2.09 2.15 0.44 0.41 0.17 149.13 2.06 23.51
Z7 Fall 21.28 8.17 0.62 5.11 2.19 0.70 0.88 0.29 145.90 2.15 19.99

used to cover the irregular domain of Burullus Lake with 19509 el-
2.7. MIKE21 application ements and 9000000 nodes (Figure 2).
Bathymetric charts were presented using an average water depth for
The numerical model of Mike21 is established according to the so- a unit area of 150 m×150 m, and the model domain was discretized
lution of the two or three-dimensional incompressible Reynolds av- into non-overlapping elements/cells.
eraged Navier-Stokes equations [4]. A finite-element model was The Courant number was selected as 0.8 based on the mesh resolu-
tion, field velocity, and 100 sec time interval (2.7×106 time steps).
 1764 International Journal of Engineering & Technology

Wind friction was determined from the observation of Egyptian several drains such as Nasser drain and El-Garbiya drain. The de-
Meteorological Authority according to the wind speed; e.g., wind crease in NH3-N concentration at Z3 could be because the zone is
friction of 0.001255 at 7 m/s. not connected to any drain. The drains 7, 8, and 9 carry agriculture
contaminants to the lake with a total flowrate of 1.7×109 m3/year;
and thus, the NH3-N concentrations at zones 4 and 6 increased.
As shown in Figure 3c, a low concentration of NO2-N (0.5±0.1
mg/L) was observed at Z1 (a zone not connected to drains). How-
ever, the concentration of NO2-N increased at Z2 (2.2±0.4 mg/L),
El-Boughaz Z4 (3.3±0.8 mg/L), and Z6 (4.8±0.2 mg/L). NO2-N is not a stable
nitrogen species and can appear in the water mainly as a result of
Zone 1 biochemical oxidation of ammonia (nitrification) or the reduction
of nitrate (denitrification) [13].
Zone 3
Zone 2
The spatial variation of NO3-N showed the minimum and maximum
Nasser values of 0.6±0.1 mg/L at Z3 and 1.9±0.5 mg/L at Z6 (p < 0.01),
Drain

Zone 5 Zone4
610 million
m3/year
respectively (Figure 3d). According to Ayers and Westcot [9], the
Drain 7
degree of restriction for NO3-N in Burullus Lake on reuse was
Zone 6
478 million
m3/year
“None”. However, the increase in NO3-N concentration over 0.3
Zone 7
Brinbal
Canal Drain 8
mg/L suggests the occurrence of eutrophication.
200
million Drain 9
426 million
m3/year The increased concentration of PO4-P at Z2 (2.2±0.4 mg/L), Z4
m3/year Drain 11 732 million m3/year 780 million m3/year
(3.3±0.8 mg/L), and Z6 (4.8±0.2 mg/L) could result from the dis-
charges of drains [7-9] (Figure 3d). Similarly, El-Zeiny & El-
Fig. 2: (2D) Visualization of a Computational Mesh and Drainage Sources
of Burullus Lake. Kafrawy [5] indicated that the increase in the percentage of agricul-
tural and urban lands contributed to an increment of nutrient con-
centrations in lake ecosystems.
3. Results and discussion Silicate
The spatial variation of SiO4 indicated the minimum and maximum
3.1. Water quality variation values of 1.9±0.1 mg/L at Z5 and 4.7±0.1 mg/L at Z4, respectively
(p < 0.01) (Figure 3e). Zone-5 was not directly connected to drains,
Temperature and thus observed low SiO4 concentrations (Figure 1). However,
The maximum temperature was found at Z7 reaching 23.0±4.2°C, the high concentration of SiO4 at Z2 and Z4 could be related to the
which decreased to 19.9±5.0°C at Z5 (Figure 3a). The non-signifi- discharge of drainage water coming from several drains such as
cant (p: 0.984) variation in temperature could be linked to the small drain-7, Nasser drain, El-Gharbya drain, and Burullus drain.
region of latitude (between 31° 22’ and 31° 26’ N). Chlorophyll-a
Hydrogen ion concentration The concentration of Chl-a significantly (p < 0.01) fluctuated from
As shown in Figure 3a, the maximum pH-value was observed at Z2 28.3±5.7 µg/L at Z1 (directly connected to the Mediterranean Sea)
(8.5±0.6), while the minimum pH of 7.2–8.4 was detected at Z4 (p > to 144.2±4.1 µg/L at Z7 located in the western region (Figure 3e).
0.01). Based on WHO [6], Z2 was considered strongly alkaline (pH The highest Chl-a concentration was recorded at Z7 because this
8.5–9.0), whereas most zones represented a neutral condition (pH zone is connected to Rosetta Branch containing several floating and
6.6–7.3). The pH-standards indicated that most data were moder- submerged aquatic plants [14]. Moreover, the high concentration of
ately alkaline in nature, suggesting that the lake was suitable for Chl-a at Z7 was possibly associated with the use of phosphorus-
irrigation, and different kinds of aquatic life [6]. According to based fertilizers within the watersheds of drain-11 and Burullus
UNECE [7], the water quality could be classified as Class I (pH drain (Figure 1). Chl-a concentration is considered as a good indi-
6.5–9.0) for the maintenance of aquatic life. Based on the pH range cator of the phytoplankton biomass, and its increase can reflect a
of 6.5–8.4, the lake could be used for irrigation purposes [11]. sign of eutrophication. According to UNECE [7], the average Chl-
Dissolved oxygen a during the year of 88.0±39.0 µg/L is appropriate for the mainte-
The horizontal distribution of water quality parameters revealed nance of aquatic life, following “Class V”.
that DO concentration increased from the western part to the eastern Salinity
zones (Figure 3b). Zone-1 showed a significantly (p < 0.01) high The lowest salinity (0.5±0.2 g/L) was recorded at Z7, whereas Z1
DO level of 7.9±0.4 mg/L, suggesting a low discharge of irrigation demonstrated the highest salinity of 10.3±1.6 g/L, with a p-value <
water to the lake. On the contrary, the minimum DO value was ob- 0.01 (Figure 3e). The salinity at Z7 was not affected by the rate of
served at Z7 (4.0±1.8 mg/L), which could be due to its consumption exchange with the sea water because this zone is located far from
by microorganisms for the decay of organic pollutants [12]. Ac- adjoining open of Mediterranean Sea. Thus, the significant spatial
cording to UNECE [8], the DO standard for the maintenance of variation in water salinity could be explained by the mixing of Z1
aquatic life is Class I (DO > 7) at Z1 and Class III (DO 4–6 mg/L) with salty seawater; i.e., Burullus Lake is connected with the Med-
and Class IV (DO 3–4 mg/L) at Z7. Other DO values could be clas- iterranean Sea through El-Boughaz opening at Z1 (Figure 1). Based
sified as Class II (6–7 mg/L). on Ayers and Westcot [11], the salinity concentration at Z1 ex-
Biological oxygen demand ceeded the severe degree of restriction on reuse (> 2.0 g/L). Accord-
The highest BOD concentration of 42.0±10.6 mg/L was reported at ing to FAO/WHO [15], salinity conditions in Z7 and Z1 can be clas-
Z6 (Figure 3b), which could be due to the negative impacts of drains sified as “Non-saline” and “Very saline”, respectively. However,
8 and 9. These drains carry a high amount of agriculture and mu- the average salinity of the lake was medium saline (2.0–4.0 g/L) to
nicipal drainage water to the lake with a total flow rate of 1.2×109 highly saline (> 4.0 g/L).
m3/year (Figure 1). According to UNECE [7], the concentration of
total BOD in water bodies for fisheries and aquaculture should not
exceed 3 mg/L at 20°C. This finding suggested that several zones
of Burullus Lake could suffer from high rates of pollution.
Nutrients
The regional variation of NH3-N was fluctuating from 1.0±0.1
mg/L at Z3 to 3.8±0.2 mg/L at Z4 (Figure 3c). The significant (p <
0.01) spatial variation of NH3-N could be due to the discharge of
large amounts of drainage water from agricultural sectors through
 International Journal of Engineering & Technology 1765

(a) 30 Temperature pH 9.0 (b) 9 DO BOD 50

(b) 4
Temperature (°C) 25 8.8 8
7
45
40
(a) (b)
8.6 35

BOD (mg/L)
20 6

DO (mg/L)
8.4 30 3
5

pH
15 25
8.2 4
20
10 3 15
8.0
2 10 2
5 7.8 1 5
0 7.6 0 0
Z1 Z2 Z3 Z4 Z5 Z6 Z7 Z1 Z2 Z3 Z4 Z5 Z6 Z7
1

(29%)%)
(c) 5.0 NH₃-N NO₂-N 6 (d) 2.5 NO₃-N PO₄-P 6

PC2(29.09
4.5
4.0 5 2.0 5
0
NH3-N (mg/L)

NO2-N (mg/L)
3.5

NO3-N (mg/L)

PO₄-N (mg/L)
4 4

PC2
3.0 1.5
2.5 3 3
2.0 1.0 -1
1.5 2 2
1.0 1 0.5 1
0.5 -2
0.0 0 0.0 0
Z1 Z2 Z3 Z4 Z5 Z6 Z7 Z1 Z2 Z3 Z4 Z5 Z6 Z7

(e) 12 SiO₄ Salinity Chl-a 160 -3

140
10
120
Salinity (g/L)

Chl-a (µg/L)
SiO₄ (mg/L)

8 -4
100
6 80 -5
60
4
40
2
20
0 0
(b) 4
(a)Z1 Z2 Z3 Z4 Z5 Z6 Z7 (b) Z7
Fig. 3: Spatial Variations of Water Quality Parameters for Burullus Lake:
3
(A) Temperature and Ph, (B) DO and BOD, (C) NH3-N and NO2-N, (D) Z6
Z5 Z7
NO3-N and PO4-P, and (E) Sio4, Salinity, and Chl-a.
2
Z3
Z7
Z6
3.2. Principal component analysis 1 Z5

%)
Z4

(29%)
Z1
Z3 Z2
PC2(29.09
Z5 Z7 win Z6
The first two principal components (i.e., PC1, and PC2) affected the 0
Z1
quality of the lake with a total variability of 68%. The first principal Z4
PC2
Z3
component (PC1: 39% of the variance) represented high loadings -1 Z2
on BOD: 0.36, NH3-N: 0.43, NO3-N: 0.38, PO4-P: 0.40, and Chl-a: Z1 Z5 Z4
Z3
0.34 (Figure 4a). These positive loadings indicated that PC1 in- -2
Z2
Z6
Z1
creased with nutrients concentrations coming from agricultural land.
-3
In addition, DO was distributed on the negative side of PC1, sug- Z4

gesting that the discharge of drainage water into the lake tended to Z2
-4
reduce the oxygen content. This result indicated that agricultural -5 -4 -3 -2 -1 0 1 2 3 4 5
drainage water was the key variable influencing the water quality PC1
PC1(39.05
(39%)%)
characteristics of Burullus Lake. The second principal component Fig. 4: Multivariate Analysis of Water Quality Parameters for Burullus Lake:
(PC2: 29% of the variance) was associated with high positive load- (A) Principal Component Analysis, and (B) Cluster Analysis.
ings on temperature: 0.44 and pH: 0.36, and a negative loading on
DO: -0.38 (Figure 4a). This finding indicated an opposite correla- 3.3. Water quality index
tion between temperature and DO, suggesting that DO concentra-
tions tended to decrease in summer. BREABĂN [16] reported that The results indicate that is no seasonal difference in WQI values
the peak levels of DO in winter could be because the solubility of can be obviously witnessed because the lake temperature could be
oxygen increases as water temperature decreases. linked to the small region of latitude between 31º 22’ and 31º 26’
Cluster analysis N. As might be expected, the WQI decreased during the summer
According to Figure 4b, the samples were distributed into two season (Figure 5a). The WQIs of zones 2, 4, and 6 were found be-
groups: tween 30 and 40, referring to “Bad” water quality type (Figure 5b).
Cluster 1: positioned at the right part of PC1 and contained mainly The low WQI in these zones could potentially increase the risk for
the zones of Z2, Z4, and Z6. Since organic substances and nitrogen health-related problems . Zone-2 suffered from organic and nutri-
and phosphorus species had strong positive loadings in PC1, these ents loads coming from several drains. These drains include (a) El-
zones were mainly influenced by nutrients concentrations coming Burullus drain that carries average waste of 70×106 m3/year of un-
from agricultural lands and catchment run-off processes. Zone-2 is treated and industrial wastewater, (b) El-Gharbya drain, Nasser
located near a large population community, discharging domestic drain, and drain-7 carrying about 1.6×109 m3/year of sewage and
sewage to El-Gharbya drain and Burullus drain. Zone-4 is near agricultural drainage water. In addition, Z4 and Z6 were affected by
drain 7 that carries huge amounts of agricultural drainage water the discharge of untreated wastewater coming from many small vil-
(478×106 m3/year). Zone-6 receives discharges coming from lages and communities distributed along drain side areas, reaching
drain-8 (426×106 m3/year) and drain-9 (780×106 m3/year). over 1.0×109 m3/year.
Cluster 2: containing the zones Z1, Z3, and Z5, was approximately The zones Z1, Z3, Z5, and Z7 were categorized as “Medium” water
distributed on the negative direction of PC1. Zone-1 is located in quality type. The highest salinity was reported in Z1 due to the mix-
the Eastern basin, indicating that it was affected by high salinity due ing of Burullus Lake and Mediterranean Sea [2]. Zone-3 could be
to the influence of the exchanges with the Mediterranean Sea (Fig- affected by the pollution coming from drain-7 (478×106 m3/year).
ure 1). The zones Z3 and Z5 are located far from the drainage water, Additionally, zone-3 is connected to Khalig canal and Nasser drain.
and thus they were characterized by low nutrient concentrations and Zones 5 and 7 are located near Brimbal canal and drain-11, dis-
an increase in DO levels. charging approximately 932×106 m3/year of agricultural discharges
Zone-7 was distributed on the positive direction of PC2, indicating [5].
that this zone suffered from increased chlorophyll-a concentrations,
and low DO levels. This result might be due to the decrease of flow
rate coming from Rosetta Branch, leading to poor water circulation
[12]. Moreover, the lack of hydrodynamic water circulation in the
lake contributed to the appearance of dead zones, such as zone 7
[17].
 1766 International Journal of Engineering & Technology

3.4. Simulation by MIKE21 model

Model calibration and validation

The model was subjected to a calibration procedure to estimate its
applicability for fitting the collected data. The model parameters of
both hydrodynamic and ecological models were iteratively adjusted
until the model results agreed with the measured data. It was found
that the coefficients of determination (R2-values) of water quality
parameters were 0.95 for BOD (Figure 6a), 0.74 for NH3-N (Figure
6b), 0.62 for NO3-N (Figure 6c), and 0.93 for PO4-P (Figure 6d).
The calibrated parameters included nitrification rate: 2.3/day, NH3-
N yield by plants 0.078 g-N/g-DO, denitrification rate 1.03/day,
BOD decay rate: 0.47–2.24/day, and half-saturation oxygen con-
centration: 2.1 mg/L.
The validation procedure was conducted over a period of one year
using new data that were not involved in calibration. The progress
of validation was confirmed by the determination of R2: 0.92–0.95,
and adj-R2: 0.87–0.91. This result ensured the model reliability and
usability.
Water quality improvement
The developed model was used to provide a recommendation sce-
nario that could be used to enhance the water quality of Burullus
Lake. The values of physicochemical parameters for drains 7 and 8
were assumed to be allowable according to UNECE [7]. The pollu-
tants in drains can be minimized by (a) reducing the agricultural
discharge that contains pesticides and washed and leached fertiliz-
ers, (b) using small communities with separate treatment plants that
allow for wastewater reuse in irrigation purposes, (c) restricting the
over-application of fertilizers by using biological methods to resist
the agricultural pests. Under this scenario, the water quality param-
eters of Burullus Lake improved as follows: BOD from 23.9 to 9.2
mg/L (Figure 6e), NH3-N from 2.5 to 0.8 mg/L (Figure 6f), NO3-N
from 1.2 to 0.4 mg/L (Figure 6g), and PO4-P from 1.9 to 0.5 mg/L
(Figure 6h). In addition, the WQI at Z4 and Z6 improved from “Bad”
to “Medium”. It was found that the period required to achieve this
Fig. 5: Water Quality Index for Lake Burullus: (A) At Seasonal, (B) At Spa- self-purification could be 5 months.
tial.

Fig. 6: Simulation of Water Quality Parameters by MIKE21 Software: Calibration Procedure for (A) BOD, (B) NH3-N, (C) NO3-N, and (D) PO4-P, As Well
as, Enhancement Scenario for (E) BOD, (F) NH3-N, (G) NO3-N, and (H) PO4-P.
 International Journal of Engineering & Technology 1767

4. Conclusion Vol.12, No. 2, 283-289, 2013.

https://doi.org/10.30638/eemj.2013.034.
[10] Iticescu, C., Georgescu, L.P., Topa, C. M. - "Assessing the Danube
This study succeeded to develop a large-scale picture of water qual- water quality index in the city of Galati, Romania," Carpathian Jour-
ity in Burullus Lake and it can be concluded that: nal of Earth and Environmental Sciences, Vol 8, No 4, 2014.
Seasonally, significant (p < 0.05) variations were observed in tem- [11] R. Ayers and D. Westcot, "Water quality for agriculture. Rome: FAO
perature and PH, however significant (p < 0.05) spatial differences Irrigation and Drainage," Paper 29. Revision. 1. pp. 1-130, 1994.
were determined in salinity, DO, NH3-N, NO2-N, NO3-N, PO4-P, [12] H. Hossen and A. Negm, "Change detection in the water bodies of
Burullus Lake, Northern Nile Delta, Egypt, using RS/GIS," Procedia
Chl-a, and SiO4. Engineering, Vol. 154, pp. 951-958, 2016.
According to PCA results, PC1: 39.01% represented high loadings https://doi.org/10.1016/j.proeng.2016.07.529.
on NH3-N: 0.43, NO3-N: 0.38, BOD: 0.36 and PO4-P: 0.40, whereas [13] J. Lai, F. Jiang, K. Ke, M. Xu, F. Lei and B. Chen, "Nutrients distri-
PC2: 29.19% was associated with high positive loadings on salinity: bution and trophic status assessment in the northern Beibu Gulf,
0.42 and pH: 0.57, and with a negative loading on Chl-a: -0.22. China," Chinese Journal of Oceanology and Limnology, Vol. 32, No.
The results from cluster analysis revealed that zones 2, 4 and 6 (lo- 5, pp. 1128-1144, 2014. https://doi.org/10.1007/s00343-014-3199-y.
cated in the southern part of the lake) were mainly influenced by [14] M. Nassar and S. Gharib, "Spatial and temporal patterns of phyto-
nutrients concentrations coming from agricultural drainage water, plankton composition in Burullus Lagoon, Southern Mediterranean
Coast, Egypt," The Egyptian Journal of Aquatic Research, Vol. 40,
while zone 7 (located in the western part of the lake) suffered from No. 2, pp. 133-142, 2014. https://doi.org/10.1016/j.ejar.2014.06.004.
increased Chl-a concentrations, and lower DO levels. [15] FAO/WHO, "Contaminants," In: Codex Alimentarius, Vol. XVII,
Zone-1 was affected by high salinity due to the influence of the ex- and Rome: FAO/WHO, Codex Alimentarius Commision, 1984.
changes with the Mediterranean Sea, while zones 3 and 5 (situated [16] I. BREABĂN, D. GHEŢEU and M. PAIU, "Determination of Water
in the northern part of the lake) were subject to lower pollution. Quality Index of Jijia and Miletin Ponds," Bulletin of the University
A simulation model using MIKE21 software provided an enhance- of Agricultural Sciences & Veterinary, Vol. 69, No. 2, p. 160, 2012.
ment scenario to modify the WQI of Z4 and Z6 from “Bad” to “Me- [17] M. Okbah and N. Hussein, "Impact of Environmental Conditions on
dium” within 5 months. the Phytoplankton Structure in Mediterranean Sea Lagoon, Lake Bu-
rullus, Egypt," Water, Air, and Soil Pollution, Vol. 172, No. 1-4, pp.
In conclusion, the lake water was not adequate for drinking as a 129-150, 2006. https://doi.org/10.1007/s11270-005-9066-x.
surface water sources or irrigation purposes or even as a fish farm.
However, it can be suitable for some aquatic habitats and plants.
Thus, it is recommended to:
1) Control the discharge of drainage water and sewage dis-
charged into the lake.
2) Reduce the usage of fertilizers in agricultural sectors.
3) Improve the quality of agricultural drainage water coming to
the lake by providing an appropriate treatment method.
Further investigations should be conducted to studying the possibil-
ity of dividing the lake into separate parts and providing mechanical
aeration system to improve water quality to be suitable as a fish
farms.

References
[1] K. Shaltout, "Reed Products from Lake Burullus, Egypt" The Wet-
land Book, Netherlands: Springer, 2017.
https://doi.org/10.1007/978-94-007-6172-8_211-8.
[2] H. El-Asmar, M. Hereher and S. El Kafrawy, "Surface area change
detection of the Burullus Lagoon, North of the Nile Delta, Egypt,
using water indices: A remote sensing approach," The Egyptian Jour-
nal of Remote Sensing and Space Science, Vol. 16, No. 1, pp. 119-
123, 2013. https://doi.org/10.1016/j.ejrs.2013.04.004.
[3] M. Nasr and H. Zahran, "Performance evaluation of agricultural
drainage water using modeling and statistical approaches," The
Egyptian Journal of Aquatic Research, Vol. 42, No. 2, pp. 141-148,
2016. https://doi.org/10.1016/j.ejar.2016.04.006.
[4] DHI, MIKE 11-A "Microcomputer Based Modelling System for Riv-
ers and Channels," Technical Reference. Version 2008, Denmark:
Danish Hydraulic Institute, 2008.
[5] A. El-Zeiny and S. El-Kafrawy, "Assessment of water pollution in-
duced by human activities in Burullus Lake using Landsat 8 opera-
tional land imager and GIS," The Egyptian Journal of Remote Sens-
ing and Space Science, Volume 20, Pages S49-S56, 2017.
https://doi.org/10.1016/j.ejrs.2016.10.002.
[6] WHO, "Guidelines for drinking-water quality," Second ed, Vol. 2.
Health criteria and other supporting information, Geneva: World
Health Organization (WHO), 1996.
[7] UNECE, "Protection and Sustainable Use of Water Resources and
Aquatic Ecosystems," New York: United Nations Economic Com-
mission for Europe (UNECE), 1995.
[8] Roşu, C., Piştea, I., Călugăr, M., Martonoş, I., Ozunu, A., "Assess-
ment of ground water quality status by using water quality index
(WQI) method in Tureni village," Cluj county, Aerul siapa compo-
nente ale mediului, pg 111-118, 2013 Cluj-Napoca, Romania.
[9] Piştea, I., Roşu, C., Martonoş, I., Ozunu, A., "Romanian surface wa-
ter quality: Tarnava Mare river between Medias and Copsa Mica case
study," Environmental Engineering and Management Journal,