Analyzing Riyadh Treated Wastewater Parameters for Irrigation Suitability Through Multivariate Statistical Analysis and Water Quality Indices

Abstract

An alternative irrigation water supply that prioritizes quality standards, promotes sustainable water resource management, and uses ecologically friendly approaches is still being researched. The purpose of this study is to evaluate the thirteen physicochemical properties of Riyadh wastewater treatment plants (WWTPs) over eight years for their potential use in irrigation. Wastewater quality was assessed using the Comprehensive Water Pollution Index (CPI) and the Canadian Wastewater Quality Index (CWQI). Principal component analysis and heatmaps were also used to identify trustworthy parameters. The CWQI results, ranging from 72.95 to 95.55%, showed acceptable variations over eight years, indicating adequate quality. The CPI values varied from 0.19 to 0.77. However, the average CPI was determined to be 0.6, indicating that there had been some slight contamination throughout the study. The first and second components (PC1 and PC2) represented 32.6% of the data, revealing a dominant pattern for a better understanding of the effluent characteristics. The effluent parameters loaded onto PC1 were EC, Ca²⁺+Mg²⁺, NO₃, and COD, whereas NH₄, DO, and turbidity were loaded onto PC2. The effluent from the Riyadh WWTPs is appropriate for irrigation, highlighting the necessity of TWW for agriculture and supporting Saudi Arabia’s Green Riyadh Initiative.

1. Introduction

The arid regions have undergone natural hazards caused by climate change, such as droughts and floods, harsh weather, and a faster hydrological cycle. According to the average climate change prediction, the Middle East and North Africa (MENA) region’s total water demand will rise to 393 km³ per year by 2050, while the region’s overall water scarcity will climb to 199 km³ per year. A 50% rise in water demand and a 12% decrease in water supply are the reasons for this worsening of the crisis [1]. The Gulf Cooperation Council (GCC) countries produced 2853 million m³/year of wastewater in 2015, with Saudi Arabia contributing 54% and treating 69% [2]. In addition, Saudi Arabia’s national population is expected to increase by 77% by 2050, from 32 million to more than 56 million. Furthermore, the difference between water supply and water demand in Saudi Arabia is 11.5 billion m³. This gap can be replenished through depleting groundwater resources, desalinating water, and reusing treated wastewater (TWW) as non-traditional water sources [3]. The use of TWW in agriculture is not only an effluent disposal solution but also a solution for sustainable agriculture, particularly in areas where freshwater is scarce. According to the 2030 Saudi Vision, Saudi Arabia intends by 2025 to reach 100% TWW utilization [3]. In 2022, Saudi Arabia’s TWW utilized in agriculture was 412 million m³, accounting for 22.6%. However, the TWW amount in Riyadh was 596 million m³, but only 37 million m³, or 9%, was used for agriculture [4]. TWW is available all year and contains nutrients required for agricultural growth [5]. Thus, it can be utilized as a source all year. The total annual capacity of the 133 wastewater treatment plants (WWTPs) is 1.93 billion m³. In the future, greater reliance will be placed on available TWW supplies, especially in the Green Riyadh project [6].

Environmental issues due to declining natural water supplies, urbanization, population growth, poor water quality, and climate change necessitate exploring alternative water sources to prevent depletion [7]. Water reuse is proven to be a reliable alternative for sustainable water management [8]. Reusing TWW can enhance global water resource management, especially in semi-arid and arid regions where water supplies are becoming scarce in quantity and quality [9]. TWW is now an essential resource that is utilized to irrigate pastures, crops, vegetables, and fruit in countries including Mexico, Peru, Colombia, Chile, Argentina, and Bolivia [10]. Remarkably, these regions have witnessed the adoption of TWW for irrigation on a substantial scale, potentially spanning over two million hectares. This extensive utilization encompasses both indirect and direct methods of irrigation [11]. However, Shafiee, et al. [12] reported that wastewater had a higher social risk than the other sources in each sector. The explanation might be an absence of confidence in the WWTP’s ability to maintain the appropriate and consistent quality, as well as concerns about pathogens and decreased productivity [13]. The current treatment techniques at a Saudi Arabian water treatment plant (WWTP) fulfil the quality requirements for restricted irrigation but not for unrestricted irrigation, according to a one-year monitoring survey via microbiological analysis [14]. This suggests that safer water usage practices are necessary for irrigation of food crops. The analysis of TWW used for irrigation in Riyadh reveals high salinity, low sodium content, and higher-than-allowed nitrate levels [15]. In desert regions like Abu Dhabi and Oman, irrigation water displays high levels of salinity and sodicity. All samples in Oman had a high salinity and a high sodium concentration. Researchers are investigating the reuse of TWW to alleviate water scarcity and salinity hazards, as well as to boost agricultural output [16,17,18,19,20,21]. Historical data analysis from WWTPs can be utilized to analyze plant performance, recognize patterns, and evaluate the effectiveness of adjustments. This approach has the potential to enhance treatment processes and strengthen environmental protection measures [22].

The dynamic nature of wastewater poses a challenge for maintaining water sources and developing effective management strategies. Wastewater quality fluctuates due to various conditions, making it challenging to assess and improve. This variability occurs over time, seasons, and operational changes, complicating the evaluation and optimization of wastewater processes [23]. The conventional methods encompass geo-physicochemical analysis, experimental sampling, and regular monitoring of water quality. However, because there are so many different physical, chemical, and microbial components, temporal evaluation is challenging [24]. Furthermore, it is challenging to identify trends and choose the best options because of the intricate interactions between the components. Moreover, it is challenging to assess wastewater quality using a single evaluation due to the complex interactions between attributes and treatment methods. It is necessary to create comprehensive strategies for efficient water quality monitoring programs in order to decrease the quantity of quality indicators that are regularly evaluated. Understanding the relationships and identifying the parameters that enhance wastewater reuse can be facilitated by multivariate analysis [23]. Water quality indices are methods that reduce data volume to a great degree and simplify the expression of the state of water quality [25]. Numerous physicochemical and bacteriological factors are used to calculate the water quality index [26]. The Canadian Council of Ministers of the Environment (CCME) has developed Canadian Water Quality Index (CWQI), simplifying the monitoring of complex and technical data on water quality [27,28]. The CWQI is a numeric term used to measure the quality of a water body and summarize vast quantities of the tracked parameters into one [23,29]. The CWQI is a unitless number that defines the efficiency of the determined chemical, physical, and microbiological parameters using an analytical method [30]. CWQI ranges from zero to 100% based on measured parameters and established standards [31]. This index has developed to assess spatial and temporal changes in water quality [32,33]. It also helps water managers clarify the overall quality of water in a more clear and continuous manner [34]. This index provides a concise overview of comprehensive data on water quality that the general public, water providers, planners, managers, and policymakers can easily understand [30]. The CWQI has been used in many studies of drinking water, groundwater, and surface water [35,36,37,38,39,40,41,42], but few of wastewater [43,44,45,46]. Additionally, the CWQI method was used in watersheds [47]. Important obstacles to widespread adoption exist due to certain possible pollutants that may have detrimental effects on soil quality and/or public health [48]. The sodium adsorption ratio (SAR) was calculated by the salinity laboratory staff of the USDA [49]. This assesses the cation content expressed in milliequivalents per liter [50]. It is a valuable measurement for determining the acceptability of irrigated water based on salt risk [51,52]. It examines the interaction between the soluble forms of sodium, calcium, and magnesium. SAR has become an indicator of feasible soil seepage problems because of its capacity to monitor variations in calcium in soil water. The ability of the soil to absorb Na ions and the penetration of dissolved cations into its cation exchange zones are both determined by the SAR [53]. The intricate relationships between the many aspects of TWW quality make it challenging to define TWW quality efficiently using a single measure or combination of parameters. Furthermore, interpreting variable variance without comprehensive analysis would not result in a full assessment of treatment system efficiency [23,54]. Therefore, multivariate analysis, such as principal component analysis and heatmap clustering, might be useful in comprehending this intricate connection and identifying the main factors that could improve TWW reuse [55].

Multivariate statistical approaches are useful for comprehending the relationships between various datasets. Correlation analysis, factor/principal component analysis (FA/PCA), and cluster analysis (CA) are a few examples of multivariate statistical methods that are useful for environmental investigations. PCA is a robust dimensionality reduction approach that extracts the essential information from associated variables by using diagonal transformations to split them up into independent principal components. By reducing complex chemical datasets to a few factors and uncovering previously undetected relationships, these techniques help to improve understanding of water quality [56,57,58]. Using measurement data, especially the large databases accumulated in WWTPs, PCA improves the quality and performance of processes by detecting faults and diagnosing them and treating them with changing operating conditions by updating the covariance structure recursively [59]. The raw process data often contain serious correlations between measured variables and are of high dimensionality; PCA provides independence and reduces dimensionality [60,61]. The correlated parameters are transformed into the component factors, which are uncorrelated with each other [62]. Lefkir et al. [62] used PCA and partial least squares (PLS) to examine seasonal influences and compositional changes in sewage and industrial waste. CWQI and statistical techniques constitute a powerful approach that provides a clear narrative of a region’s wastewater quality, transforming complex data into a comprehensive understanding. Principal component analysis (PCA) is a linear technique that combines variables to provide a low-dimensional representation of data in order to capture maximum variation, although it may sometimes compromise the quality of the results [63]. A heatmap is a two-dimensional visual representation of microarray data, allowing easy understanding and analysis of different TWW parameters and months [64]. The combination of CWQI, PCA, and heatmaps can be a useful technique for assessing TWW characteristics and discovering the relationships among the TWW parameters. Notwithstanding the fact that TWW characteristics have been the subject of several investigations, this study emphasizes the complications in assessing and enhancing TWW quality efficiency because of its dynamic nature and variable elements, underscoring the need for more research. Conventional techniques like physicochemical analysis and monitoring are insufficient due to complex variable relationships. Furthermore, evaluating TWW quality with one type of evaluation is challenging. The current understanding of TWW quality is limited due to limitations in laboratory measurements. To improve the accuracy and reliability of quality assessments, it is crucial to assess critical parameters and examine their relationships, as certain variables are essential for understanding TWW quality factors. There has been little research recently on the suitability of TWW from the Riyadh WWTPs for irrigation. The research hypothesis is based on the possibility of utilizing all of the unexploited TWW, which accounts for 90.5% of the effluents, through TWW quality acceptability evaluation for irrigation in Riyadh, while considering the parameters governing quality.

The objectives of this study were: (i) to determine the temporal fluctuations of TWW quality characteristics and determine the appropriateness of TWW for irrigation purposes by CWQI, and (ii) to integrate PCA and heatmap clustering analysis in order to comprehend the relationships between TWW quality parameters and to understand the trustworthiness of those parameters.

2. Materials and Methods

2.1. Study Area

Riyadh Province spans 406,291 km², representing approximately 18.9% of Saudi Arabia (longitudes 45°59′12″ to 47°20′29″ E, latitudes 24°13′51″ to 25°10′30″ N, and altitude 620 m), with a dry tropical climate that receives erratic rainfall, high rates of evaporation, and low relative humidity. The city was built on the Wadi Batha alluvial plain, which receives 120 mm of rain per year on average [65].

2.2. Riyadh Municipal Wastewater Treatments Plants (WWTPs)

There are 7.5 million people (25% of the total population) living in Riyadh, the capital and largest city of Saudi Arabia [66]. Riyadh WWTPs are divided into four populated areas: Manfouha, Heet-Alkharj, Alhayer, and Refinery, as illustrated in the details shown in

Table 1. Major wastewater treatments plants in Riyadh city [69].

Municipal Wastewater Plant	Design Capacity (m3/day)	Technology	Treatment Type	Purpose
Manfouha	North 200,000 South 200,000 East 200,000	Trickling filter-activated sludge	Tertiary	Agricultural irrigation
Heet-Alkharj	Phase I 100,000 Phase II 100,000 Phase III 200,000	Activated sludge	Tertiary	Groundwater recharge
Alhayer	Phase I 400,000	Activated sludge	Tertiary	Irrigation and groundwater recharge
Refinery	20,000	Clarification and filtration	Tertiary	Agricultural irrigation

, and the locations of these municipal centralized plants are shown in Figure 1. The municipal WWTPs in Riyadh adhered to the activated sludge process. After being sieved, the sewage effluent was sent to a grift chamber, a primary sedimentation tank, an aeration tank to activate the sludge, a secondary sedimentation tank, a tertiary filter, and finally chlorine disinfection to obtain treated sewage [67]. These municipal plants use tertiary treatment, which offers an efficient and environmentally benign method of treating wastewater [68].

2.3. Treated Wastewater (TWW) Quality Parameters

The historical TWW parameters data were collected for eight years from WWTPs in the Riyadh region. The TWW quality parameters included thirteen parameters: chemical oxygen demand (COD) (mg/L), dissolved oxygen (DO) (mg/L), free chlorine (free Cl) (mg/L), Na⁺ (meq/L), Ca²⁺ (meq/L), Mg²⁺ (meq/L), NH₄-N (mg/L), NO₃-N (mg/L), total dissolved solids (TDS) (mg/L), EC (dS/m), pH, turbidity, and Escherichia coli (E. coli) (Cell/100 mL). A total of 37,552 data for the thirteen quality parameters were obtained from 3134 samples, sometimes taken twice a day, over an eight-year period. However, E. coli was tested twice to three times per week. All samples and analytical methods used were carried out in accordance with APHA [70].

The traditional standard plate count technique (MPN) per 100 mL was employed for the microbiological tests [71]. A flame photometer was used to detect the soluble concentrations of Na⁺ and K⁺, while the versenate titration technique (EDTA) was utilized to calculate the soluble concentrations of Ca²⁺ and Mg²⁺ [72]. H₂SO₄ was used to calibrate the NH₄ measurement using a Kjeldahl [73]. A spectrometer in a nephelometric turbidity unit (NTU) was used to measure turbidity [74]. The gravimetric technique was employed to quantify the total dissolved solids (TDS), as stated by Hussein and Magram [75]. An EC meter was used to measure the EC. The appropriateness of TWW for irrigation was also evaluated using EC, TDS, and sodium adsorption ratio (SAR). SAR was calculated from the ratio of sodium to calcium and magnesium as follows:

$$\mathrm{S}\mathrm{A}\mathrm{R}={\displaystyle \frac{{\mathrm{N}\mathrm{a}}^{+}}{\sqrt{({\mathrm{C}\mathrm{a}}^{2+}+{\mathrm{M}\mathrm{g}}^{2+})/2}}}$$

(1)

where Na⁺, Ca²⁺, and Mg²⁺ are in meq/L.

2.4. Calculation of Canadian Water Quality Index (CWQI)

The CWQI was adapted from Saffran et al. [28] and Ebrahimi et al. [76] as follows:

$$\mathrm{C}\mathrm{W}\mathrm{Q}\mathrm{I}=100-{\displaystyle \frac{1}{1.732}}·\sqrt{{\displaystyle \frac{{\mathrm{F}}_{\mathrm{P}}}{{\mathrm{T}}_{\mathrm{P}}}}+{\displaystyle \frac{{\mathrm{F}}_{\mathrm{t}}}{{\mathrm{T}}_{\mathrm{t}}}}+{\displaystyle \frac{\left({\sum }_{i=1}^{\mathrm{n}}\left(\frac{{{\mathrm{V}}_{\mathrm{F}\mathrm{t}}}_{\mathrm{i}}}{{\mathrm{L}}_{\mathrm{i}}}-1\right)\right)/{\mathrm{T}}_{\mathrm{t}}}{\left(0.01\left({\sum }_{\mathrm{i}=1}^{\mathrm{n}}\left(\frac{{{\mathrm{V}}_{\mathrm{F}\mathrm{t}}}_{\mathrm{i}}}{{\mathrm{L}}_{\mathrm{i}}}-1\right)\right)/{\mathrm{T}}_{\mathrm{t}}\right)+0.01}}}$$

(2)

where ${\mathrm{F}}_{\mathrm{P}}$ is the number of failed parameters, ${\mathrm{T}}_{\mathrm{P}}$ is the total number of parameters, ${\mathrm{F}}_{\mathrm{t}}$ is the number of failed tests for all parameters, ${\mathrm{T}}_{\mathrm{t}}$ is the total number of tests for all parameters, ${{\mathrm{V}}_{\mathrm{F}\mathrm{t}}}_{\mathrm{i}}$ is the value of failed test i for all parameters, ${\mathrm{L}}_{\mathrm{i}}$ is the limitation standard of the test i, and n is the total number of failed tests for all parameters. The wastewater quality was then ranked in different categories, as described in

Table 2. Treated wastewater quality category based on CWQI.

Quality Range	CWQI	Water Category
Excellent	95–100	Very near to perfect levels
Good	80–94	Rarely departs from desirable levels
Fair	65–79	Sometimes departs from desirable levels
Marginal	45–64	Often departs from desirable levels
Poor	0–44	Quality is almost always threatened

[33].

2.5. Calculation of Comprehensive Pollution Index (CPI)

CPI is calculated by comparing the measured concentration of a parameter to its limitation as prescribed by Saudi standards [77], as follows:

$$\mathrm{C}\mathrm{P}\mathrm{I}={\displaystyle \frac{1}{N}}{\sum }_{\mathrm{i}=1}^{\mathrm{N}}{\displaystyle \frac{{\mathrm{P}}_{\mathrm{i}}}{{\mathrm{S}}_t}}$$

(3)

where P_i is the measured concentration of a TWW parameter, S_t is the standard permissible concentration of the parameter, and N is the number of tests for each parameter.

CPI ranges are defined between 0 and 2. The CPI classifications for TWW quality are five scores (clean, sub-clean, slightly polluted, moderately polluted, and severally polluted), as illustrated in

Table 3. TWW quality classification based on CPI.

Score	Criteria
0.0 ≤ CPI ≤ 0.2	Clean
021 ≤ CPI ≤ 0.4	Sub-clean
0.41 ≤ CPI ≤ 1	Slightly polluted
1.01 ≤ CPI ≤ 2.0	Moderately polluted
CPI > 2.01	Severely polluted

.

2.6. Multivariate Statistical Analysis Approach

Temporal variability in TWW parameters was investigated through the use of principal component analysis (PCA) as multivariate analysis. PCA was performed using XLSTAT software (Version 2019.2.2) in the following order [78]: (i) Data were normalized, which involves detracting the mean and dividing by the standard deviation of each parameter. The standardization process ensures that each variable makes an identical contribution to the study, preventing any bias in principal component analysis favoring variables with higher variation. (ii) A covariance matrix was created. This matrix represents the relationships between variables. (iii) Eigenvectors and eigenvalues were estimated. (iv) The major components were determined (PC1 matches the largest eigenvalues). (v) The dataset dimensions were reduced to improve representation quality. The highest level of data variability is between observed parameters estimated via PCA. It identifies the parameters that are the cause of the discrepancy. Thus, we may evaluate the effectiveness of wastewater treatment by using these parameters. The results of this analysis identify whether the water is suitable for agricultural irrigation and its environmental effects according to the Saudi standard.

PCA served as a powerful tool to unravel the intricate relationships between the numerous parameters influencing TWW quality. Based on Kaiser’s rule, eigenvalues of 1.0 or more are regarded as significant, and the greatest eigenvalues represent the components that are most important [79,80]. The statistical analysis was carried out using the XLSTAT statistical package utilizing the xlstat program (Version 2019.2.2; Excel Add-ins Soft SARL, New York, NY, USA).

2.7. Heatmap Analysis

A hierarchical cluster analysis for each month based on multiple parameters was carried out to categorize TWW quality parameters according to temporal change. To find the cluster groups, a dendrogram of clusters was created. Over all eight years, a heatmap clustering analysis (HCA) was conducted with 13 parameters and 12 months. The HCA divided the 12 months into two groups. The 13 parameters were grouped together and resulted in the highest values for all parameters. Utilizing R Studio software (version 2022.12.0+353, R Core Team, 2022), a heatmap clustering analysis was created to illustrate the integration of parameters throughout various months. The heatmaps were used to illustrate monthly variations in key TWW quality measurements, with clusters indicating similarity across parameters and months and significant monthly fluctuations in positive and negative values.

3. Results and Discussion

3.1. Treated Wastewater (TWW) Quality

The TWW’s COD, NO₃, and NH₄ parameters showed comparable tendencies. COD is a biochemical parameter that measures the organic matter in wastewater by examining the oxygen released after organic matter is oxidized by powerful chemical oxidants [81,82]. Nearly every COD data result was under the maximum permitted level for restricted irrigation (RI), as shown in Table 4 and Figure 2a. The average COD contents in the TWW were found to comply with Saudi standards, falling below the permitted limit of 80 mg/L. This result serves as evidence for the efficient operation of the aeration units and biological reactors, staying within the parameters of their design. This finding was in agreement with those of international plant treatments [83,84]. Regarding NO₃, the average was 7.67 mg/L, with a range of 0.98 to 40.98 mg/L (Figure 2a). About 19.1% of NO₃ data values are above the maximum permitted level for RI, as revealed in Table 4 and Figure 2a. The established limitation for NO₃ is 10 mg/L, and exceeding this limit may be hazardous owing to its effects on the environment and human health [85,86]. Furthermore, the average NH₄ was 3.8 mg/L, with a range of 0.85 to 13.37 mg/L, as shown in Figure 2a. During the study period, 71.3% of NH₄ data were under the maximum permitted, which is 5 mg/L for RI, as shown in Table 4 and Figure 2a. This result was consistent with Badr et al. [87] in Alhassa, Saudi Arabia, who discovered that NH₄ levels rose in September 2017 compared to October 2016, indicating a considerable rise in TWW. NH₄ toxicity is more harmful in alkaline water than acidic water, and its toxicity can be influenced by elevated pH levels, affecting pollutant toxicity [85]. However, post-treatment of effluent using photocatalytic material–microorganism systems may assist in reducing effluent levels of ammonia and nitrate [88]. The TDS values fluctuated between 1115.02 and 1435.39 mg/L, as revealed in Figure 2d. No test for TDS and EC has ever yielded results beyond the recommended limit. The average EC content was 1.98 ds/m, with a range of 1.74 to 2.42 mg/L, shown in Figure 2b. TDS and EC are related to the ion content, which reflects the clarity and cleanliness of the water body [89]. Moreover, soil salinity is largely dependent on the agricultural water quality [90]. TWW discharged without proper treatment has the highest EC and hence obtains the highest grade. TWW effluents with a high salt content have the potential to negatively impact freshwater aquatic environments [91]. In the context of pH values, there are no tests that exceeded the suitable limit. The pH values ranged from 6.61 to 7.59 with an average of 7.23. pH, as related to the measurement of the acidic status of water, has a significant influence on soil microbial activity, crop growth, and metabolism [89]. The mean pH values of Kuwait’s WWTPs were between 6.5 and 7.5 [92], which are similar to our findings in Riyadh. However, a minor difference in pH values, ranging between 7.1 and 7.3, was noted at the same plants in Riyadh in 2011 [93].

DO concentration showed a uniform trend from 2013 to 2014. However, from 2015 to 2020, it showed a non-uniform trend. DO fluctuation may be due to the change in the wastewater flow rate. The DO values ranged between 5.07 and 8.07 mg/L, with an average of 6.24 mg/L, as shown in Figure 2b. DO was a crucial parameter in aeration tank operation, affected by both increasing and decreasing biomass consumption. Long-term wastewater retention increases DO concentration, while short-term retention decreases DO [94]. There were no tests that exceeded the standard limit. Moreover, the average free Cl value was 0.23, within a range of 0.02 to 1.04 mg/L (Figure 2b). In line with our findings, Al-Jasser [93] discovered that the average free Cl content of TWW in Riyadh plants, Saudi Arabia, was 0.26 mg/L. However, in Al-Turki [95]’s investigation, it was discovered that TWW samples from Buraidah City in KSA had a mean free Cl value of 0.52 mg/L. Excessive organic matter effluents and sewage pollution cause the low DO levels in streams, where microorganisms require it for physiological processes [96]. Yin et al. [97] highlighted the decrease in chlorine percentage in the analyses as a sign of improved water quality and protection against potential risks associated with chlorine presence.

The range of sodium concentration was 8.13 to 45.31 mg/L, and the Na average was 11.89 mg/L; 97.9% of samples had less than the recommended limit (40 mg/L), as shown in Table 4 and Figure 2c. The levels of metals in wastewater were influenced by regional factors such as industries, social standards, and environmental awareness regarding the negative effects of inadequate waste disposal [98,99,100]. The average content of calcium and magnesium was 11.74 meq/L, which was lower than the recommended limit (25 meq/L), within the range of 6.44 to 29.78 meq/L, as shown in Figure 2c. SAR was found to be between 2.46 and 8.19 (meq/L)^1/2, which was lower than the recommended limit of 11.3 (meq/L)^1/2 for RI. No SAR or Na test has ever exceeded the recommended limit for RI. SAR may identify potential infiltration problems in soil because of its impact on water permeability [93]. The turbidity concentration ranged from 2.56 to 26.29 NTU, falling over the required level, as illustrated in Figure 2a. Approximately 50% of turbidity data were over the maximum permitted level, which was 5.8 NTU for RI.

Table 4. The maximum value, percentage over the limit, minimum value, percentage below the limit, and standard allowed value of the parameters for unrestricted and restricted irrigation in Saudi Arabia, including FAO standards and EPA standards.

Parameters	Unit	Maximum Value	Over the Limit (%)	Minimum Value	Below the Limit (%)	Unrestricted Irrigation in Saudi Arabia	Restricted Irrigation (RI) in Saudi Arabia	FAO Standards [101]	EPA Standards [102]
Free Cl	mg/L	1.04	7.4	0.02	92.6	0.5 (<0.2)	0.5	-	1
DO	mg/L	8.07	0.0	5.07	100.0	>4	>4
SAR	(meq/L)1/2	8.19	0.0	2.46	100.0	-	11.3	15	-
Na	mg/L	45.31	0.0	8.13	100.0	-	40	40
Ca2++Mg2+	mg/L	29.78	2.1	6.44	97.9	-	50
E. coli	cell/100 mL	2416.90	6.4	0	93.6	2.2	1000 (monthly)	-	200
COD	mg/L	60.35	0.0	5.95	100.0	20 (monthly)	80 (monthly)		40
NO3-N	mg/L	40.98	19.1	0.98	80.9	10	10	10	50
NH4-N	mg/L	13.37	28.7	0.85	71.3	5	5	5–30
Turbidity	NTU	26.29	50.0	2.56	50.0	5.8	5.8	10	<2
TDS	mg/L	1435.39	0.0	1115.02	100.0	2500	2500	0–2000	450–2000
EC	dS/m	2.24	0.0	1.74	100.0	-	3	0–3.0	0.7–3.0
pH		7.59	0.0	6.61	100.0	6–8.4	6–8.4	6.5–8.4	6.0–9.0
CWQI	%	95.5	13.8 *	72.95	86.2 *	-	-	-	-

* CWQI was ranked based on class.

Table 4. The maximum value, percentage over the limit, minimum value, percentage below the limit, and standard allowed value of the parameters for unrestricted and restricted irrigation in Saudi Arabia, including FAO standards and EPA standards.

Parameters	Unit	Maximum Value	Over the Limit (%)	Minimum Value	Below the Limit (%)	Unrestricted Irrigation in Saudi Arabia	Restricted Irrigation (RI) in Saudi Arabia	FAO Standards [101]	EPA Standards [102]
Free Cl	mg/L	1.04	7.4	0.02	92.6	0.5 (<0.2)	0.5	-	1
DO	mg/L	8.07	0.0	5.07	100.0	>4	>4
SAR	(meq/L)1/2	8.19	0.0	2.46	100.0	-	11.3	15	-
Na	mg/L	45.31	0.0	8.13	100.0	-	40	40
Ca2++Mg2+	mg/L	29.78	2.1	6.44	97.9	-	50
E. coli	cell/100 mL	2416.90	6.4	0	93.6	2.2	1000 (monthly)	-	200
COD	mg/L	60.35	0.0	5.95	100.0	20 (monthly)	80 (monthly)		40
NO3-N	mg/L	40.98	19.1	0.98	80.9	10	10	10	50
NH4-N	mg/L	13.37	28.7	0.85	71.3	5	5	5–30
Turbidity	NTU	26.29	50.0	2.56	50.0	5.8	5.8	10	<2
TDS	mg/L	1435.39	0.0	1115.02	100.0	2500	2500	0–2000	450–2000
EC	dS/m	2.24	0.0	1.74	100.0	-	3	0–3.0	0.7–3.0
pH		7.59	0.0	6.61	100.0	6–8.4	6–8.4	6.5–8.4	6.0–9.0
CWQI	%	95.5	13.8 *	72.95	86.2 *	-	-	-	-

* CWQI was ranked based on class.

WWTPs have excessive turbidity owing to insufficient effluent treatment, design capacity overflow, technical challenges, and sedimentation tank issues, which reduce chlorine efficiency and necessitate continual maintenance and monitoring. The increase wastewater turbidity can be attributed to poor sludge settling and organic adsorption due to the microplastics, which remain suspended and difficult to settle, potentially reducing the efficiency of clarification and filtration processes [103].

The quantity of dissolved oxygen decreases as turbidity increases, which might affect the water’s quality and safety. As a result, photosynthesis decreased and aquatic life diminished. Furthermore, the rise in turbidity raised the concentration of microbes and nutrients, which might lead to the regrowth of harmful bacteria. Increased turbidity or insufficient influent treatment may be the cause of the decline in chlorine efficiency observed in water with high turbidity [104,105]. This finding agreed with Al-Jasser [93], who demonstrated that NO₃ and turbidity surpassed the maximum permitted levels. This was likely due to a combination of factors, including a flow rate exceeding the design capacity, technical issues, and poor settling characteristics in the sedimentation tanks. NO₃ in TWW can stress aquatic environments by generating eutrophication, promoting the development of algae, and lowering DO levels [106,107], and reduced DO levels in TWW indicate higher toxicity.

The E. coli concentration fluctuated from 0 to 2400 cells/100 mL, with an average of 159.5 cells/100 mL, falling below the required standard limit except in September and October 2014, June 2016, September 2018, and August and November 2019. The reduced effluent quality from the plants in summer is likely caused by operating at flow rates exceeding their design capacity. As a result, the treated wastewater did not meet the required standards. To ensure compliance, increasing the contact time in the chlorine contact tank will be essential for further reducing E. coli levels in the wastewater [93]. About 95.8% of E. coli data were under the maximum permitted level, which is 1000 cells/100 mL for RI, as shown in Table 4. Microplastics increase E. coli levels, disrupt microbial communities, and reduce sludge settling efficiency. Their biofilm growth, removal blockage, and disinfection interference lead to waterborne illnesses and pollution [103,108]. There was an obvious relationship between turbidity and E. coli from October 2016 to December 2016 and from September 2013 to March 2014. But in the subsequent months, no pattern stands out. Turbidity was linked to E. coli reduction, with sand filters and sedimentation tanks contributing to turbidity reduction [109]. The allowed standards of unrestricted and restricted irrigation are shown in Table 4.

3.2. Canadian Water Quality Index (CWQI)

The study used CWQI scores to group TWW into adequate irrigation categories, integrating data on physiochemical and biological TWW quality parameters to construct a weighted and standardized score. The CWQI serves as an indicator of the parameters allowed for RI, as shown in Table 4. The CWQI classifies TWW into five categories (excellent, good, marginal, fair, and poor). Figure 3 depicts the fluctuation patterns over a period of eight years, falling within the classifications of acceptable and adequate quality.

The CWQI results ranged from 72.95 to 95.55%, with an average of 84.81%. The findings of the CWQI illustrated that 14.1% of all months fell into the fair category in the study period. About 81.9% and 3.2% of the sample tests were in the good class and excellent class, respectively (Table 2). This finding demonstrated that reuse of TWW would not pose a harm to public health, especially for users and irrigated crops. Saudi Arabia’s TWW quality was compared with that of other Middle Eastern countries, such as the data of Al-Mafraq in Jordan, which fell into the fair category [110], while in Baghdad in Iraq, Obiad and Al-Sultan [46] proved that the CWQI of TWW was 80.19%, indicating good quality. They recommended that TWW was not suitable for food crops because of its soluble organic content, but it was ideal for irrigating a wide range of other crops, such as ornamental plants and grasses.

3.3. TWW Heatmap Clustering

Figure 4 illustrates how hierarchical clustering applies to both rows (months) and columns (TWW parameters), forming a clustering tree or dendrogram. The dendrogram shows clustering parameters based on similar patterns over time, seasonal trends, and correlations between parameters. The heatmaps used dendrogram clustering, which was applied to understand the trend of data means and the monthly high and low parameter values from 2013 to 2020 (Figure 4a–f). The color scale shows values ranging from blue (the lowest value) to red (the highest value), with intermediate or moderately high levels indicated by yellow and orange, respectively.

In 2013, various winter months, such as December and November, were grouped together, and the heatmap revealed that the majority of the measured parameters had low levels. Although the heatmap shows that free Cl reached maximum values, it did not exceed the permissible limits, as shown in Figure 2b. A considerable increase in NO₃ was seen in January 2013, indicating a high NO₃ concentration that exceeded the maximum allowable limit for restricted irrigation (10 mg/L), as shown in Figure 4a. This may be harmful owing to its impact on the environment and human health [85,86]. However, NO₃ usage during cultivation may be reduced as a result of this rise in effluent, which may save money, but constant monitoring is needed to adjust fertilizer quantities accordingly [111]. Furthermore, DO and pH exhibit relatively low values in August and September 2013. This may be attributed to nitrate-containing TWW, which can lead to eutrophicated environments by generating eutrophication, algae growth, and decreasing DO levels [107]. DO levels incline around the same time as pH and other parameters, indicating lower oxygen content in TWW during the later months of 2013, which did not fall under the acceptable limit, as shown in Table 4. The pH here ranges from 7.13 to 7.59; therefore, it will not turn the soil into an alkaline, irrigating the soil without degrading its quality. TDS and EC show relatively low values from September to December, but high values in March, without exceeding the allowable salinity for sensitive crops [112]. The parameters such as free Cl, E. coli, and SAR show relatively intermediate values over 2013, with only moderate fluctuations. Elevated levels of Ca²⁺+Mg²⁺ were seen in August 2013, which did not exceed the standard limit, possibly indicating mineral-rich water during that period. CWQI was in the good class and excellent class, except for January 2013, when it was in the fair class.

The heatmap of 2014 showed that all of the parameters (Na, Ca²⁺+Mg²⁺, EC, and TDS) had low amounts except for free Cl, which had the highest value in December. In addition, NO₃, pH, and COD all rose in July 2014, while E. coli and NH₄ levels were elevated in October. The residuals of the months of the year were combined in the other group, because the values of the different parameters were comparable. The heatmap of 2014 (Figure 4b) reveals that the dendrogram clusters parameters like “pH” and “NO₃” together, which fluctuate over months. NO₃ levels rose substantially in March, as observed in Figure 2a and Figure 4a, exceeding the maximum permissible limit for RI. The behaviour of “E. coli” and “DO” suggests that bacterial contamination influences water oxygen levels, indicating a correlation between these variables. The average E. coli concentration fell below the regulatory limit in all months with the exception of September and October, because of the low concentration of DO in these months. The fluctuation in turbidity levels in water can cause decreasing dissolved oxygen concentration, cloudiness, and potential impacts on aquatic life. Oxygen is crucial for aerobic bacteria, microorganisms that break down organic waste and promote the aerobic decomposition of the organic wastes [113]. Without adequate levels of DO, these bacteria cannot effectively break down waste, leading to inefficient treatment and potential pollution issues. The level of DO in wastewater can also influence the types of organisms present, with high levels supporting complex life forms like fish and invertebrates and low levels causing the growth of anaerobic bacteria, producing harmful by-products like hydrogen sulphide. According to Al-Jasser [93], exceeding the design capacity, technical problems, and inadequate sedimentation tank settling characteristics are some of the factors affecting the effluent quality of the Riyadh WWTPs. The warmer months, such as June, July, and August, display more stable or slightly positive values, suggesting higher treatment activity or natural environmental changes, while cooler months, like December, show more neutral or negative values, possibly indicating reduced treatment activity or less environmental impact. CWQI was in the good class and excellent class, except for January 2013, when it was in the fair class.

In 2015, there was no clear trend in the two groups for the temporal changes in all months. The values of Na, EC, and TDS increased in December, but did not exceed the limit, as in 2013. The summer months in 2015 (June, July, and August) show neutral values, suggesting stable TWW quality as in 2014, as shown in Figure 4c; however, January and April show slightly lower values, indicating better TWW quality. The winter months (December and January) display both negative and positive outcomes, whereas May has a rise in NO₃, which may potentially reduce fertilizer consumption due to higher concentration in TWW [111]. Parameters like F chlorine and E. coli are clustered closely, suggesting similar behaviour over time, similar to the 2014 trend, as illustrated in Figure 4c. F chlorine affects water clarity, while increases in E. coli could indicate events where higher chlorine usage is required. Furthermore, turbidity, COD, and DO are clustered together, suggesting that high bacterial contamination reduces oxygen levels, impacting aquatic ecosystem health. The clustering of SAR and Na also suggests potential impacts on water saline levels, potentially affecting TWW and ecosystem health. In February 2015, E. coli showed a red spike, indicating a contamination event affecting public health and aquatic life, but did not exceed the standard limit. The rise in chlorine may indicate increased disinfection measures due to contamination or other TWW quality issues. Blatchley III et al. [114] found that disinfectant exposure reduced bacterial populations in undisinfected wastewater, while disinfected samples showed significant recovery under similar conditions. CWQI was in the good class and excellent class, except for May, June, July, and December 2015, when it was in the fair class.

In 2016, the heatmap indicated that all of the parameters had low levels, with the exception of SAR, which had its greatest value in October but did not reach the maximum permitted level for RI. Elevated findings, especially in SAR, could point to seasonal factors affecting water salinity. In the second group, in June 2016, elevated values were noted for E. coli and turbidity, which might indicate potential pollution, decreased treatment activity, or changes in the natural environment. About 50% of turbidity data exceeded the permissible limit, as shown in Figure 4a and Table 4, affecting TWW quality and safety and potentially regrowing harmful bacteria because of insufficient treatment. This may cause a decline in chlorine efficiency in high-turbidity water [105]. Elevated turbidity levels often accompany spikes in E. coli and NO₃, as contaminated water tends to be more turbid due to particles and bacteria. The hydraulic conductivity of irrigation soil is significantly reduced by algal and bacterial development, and these reductions are linearly correlated with an increase in the quantity of bacteria [115]. However, the turbidity problem can be overcome by sand media filters, which are recommended for TWW in micro-irrigation [93]. Monitoring and remediation of soil after irrigation is crucial for sustainable use and maintaining agricultural quality [116]. Leal et al. [116] discovered that TWW irrigation boosted soil sodicity and salinity, clay dispersion rates, and electrical conductivity. However, electrical conductivity in the topsoil diminished after a short period of discontinuance. The other parameters, such as pH, TDS, EC, COD, and F chlorine, maintain more stable or lower values, as shown in Figure 4d. The concentration of these parameters was discovered to be lower than the recommended levels for RI. Some parameters exhibit distinct seasonal trends. For example, the spike in NH₄ in May might be linked to increased influent discharge during that period. As shown in Table 4, 71.3% of NH₄ data did not exceed the maximum permitted level for NH₄. The NH₄ recovery from wastewater reduces costs, energy, and environmental footprint and can be used for fertilizer production [117]. CWQI was in the good class and excellent class, except for June 2016, when it was in the fair class.

Many winter and autumn months were grouped together in 2017, such as November, December, January, February, and March. The heatmap indicated that all of the parameters had low levels, with the exception of Na and turbidity, which had their greatest values in January 2017; they exceeded the maximum permitted level. Moreover, Ca²⁺+Mg²⁺ had its greatest value in November 2017 but did not reach the maximum permitted level. Elevated findings, especially in turbidity, suggest increased colloidal material resulting in cloudy water. The residuals of the months of the year were grouped together in the second group, because the values of the different parameters were intermediate or moderately high levels that did not reach the maximum limit, with the exception of NH4, which exceeded the 5 mg/L limit in June, July, August, and October, as shown in Figure 2a and Figure 4e. Chlorine levels were elevated in September and April 2017, but did not exceed the standard limit of 0.5 mg/L. NO₃ levels were elevated in April 2017, which can lead to environmental degradation and health risks [85]. E. coli spikes in August 2017 indicate bacterial contamination, requiring immediate disinfection efforts; however, its value did not exceed the limit, as shown in Figure 2d. In February, turbidity slightly increased, indicating a decline in water clarity due to suspended particles like silt, algae, or organic material. This increase may be linked to runoff or organic material entering the water, as contaminated water tends to be more turbid due to particles and bacteria. Increases in turbidity can lower the concentration of DO, which may affect aquatic life by influencing aerobic bacteria that aid in the breakdown and decomposition of organic waste [113]. This ensures a lower DO value, which can be found in January and February. However, DO levels show an increase in November 2017, which could suggest improved TWW quality, potentially following water treatment efforts or natural aeration processes. Higher dissolved oxygen levels suggest that the water was well-oxygenated, possibly due to lower organic contamination or the implementation of measures to improve aeration, which is beneficial for aquatic ecosystems, as shown in Figure 4e. The neutral levels suggest fewer contamination events or reduced disinfection. CWQI was in the good class and excellent class, except for January 2017, when it was in the fair class. The groups of Na, EC, TDS, Ca²⁺+Mg²⁺, pH, turbidity, and E. coli indicate consistent behavior over January, February, March, April, and May, suggesting low values in 2018. However, there were greater concentrations of NO₃, NH₄, F chlorine, COD, and DO with low values in the residual months from June to December. NO₃ and NH4 levels were increased in February and March, although they fell short of the acceptable limit. Turbidity and E. coli had higher values that exceeded the standard limit, which reduced the dissolved oxygen. Low-dissolved oxygen discharge effluents with high turbidity pose a threat to surface water by increasing the toxicity of certain elements when combined with toxic substances [118]. However, NO₃ and NH₄ had greater values but did not reach the standard limit, as shown in Figure 4f. CWQI was in the good class and excellent class, except for January 2018, when it was in the fair class.

In 2019, all months were grouped together except December, with low levels of the parameters except in January, February, and March. Na, NH₄, EC, and TDS parameters had the highest value in January, but did not reach the maximum limit. Residuals were grouped together due to intermediate or moderately high values, except for NH₄, which exceeded the 5 mg/L limit in June, July, August, and October. The clustering of Na, NH₄, TDS, and EC indicates consistent behavior over time, suggesting a shared source or related processes affecting these measurements. There were greater concentrations of NO₃ and Ca²⁺+Mg²⁺ in December 2019, which may have been caused by human runoff and colder weather. The rising NO₃ content in wastewater can reduce cultivation costs, but regular monitoring is necessary to adjust the amount of additional fertilizers [111]. Furthermore, EC, TDS, Na, and NH₄ peaked towards the beginning of the year in January and February 2019, which may have been caused by agricultural activities. The clusters of DO, E. coli, F chlorine, and NO₃ indicate possible seasonal or environmental variations brought on by agricultural runoff or rainfall.

There were higher concentrations of E. coli in August 2019, possibly indicating reduced contamination. However, DO (dissolved oxygen) shows a drop, possibly indicating a decrease in water aeration or other ecological factors affecting oxygen levels. From June to July 2019, the data show that most metrics have consistent values, suggesting a balance in TWW quality. The observed seasonal variations suggest that water management strategies should be adaptive, focusing on peak contamination periods like late summer and winter to mitigate health risks and environmental impacts (shown in Figure 4g). CWQI was in the good class and excellent class, except for September 2019, when it was in the fair class. The clustering of EC, TDS, COD, and SAR indicates common variability due to agricultural runoff or wastewater inputs, suggesting that managing one parameter may positively impact others (as shown in Figure 4h).

The clustering of DO, NO₃, Ca²⁺+Mg²⁺, and pH suggests that these parameters respond similarly to seasonal and environmental factors, with high inorganic contamination, often from sewage or fertilizer runoff. January 2020 shows significantly decremented levels, indicating low organic pollutants during this period. High NO₃, Ca²⁺+Mg²⁺, pH, and DO levels are typically associated with increased organic matter or pollutants. Increased NO₃ levels in water bodies can result from agricultural runoff, which can deplete oxygen and cause eutrophication and toxic algal blooms. A rise in E. coli was observed in February 2020. High E. coli levels, typically indicative of fecal contamination, pose risks to human health and the ecosystem due to potential stormwater runoff, flooding, or untreated wastewater.

Furthermore, EC and TDS were higher in June. The increases in evaporation in the summer lead to higher concentrations of salts and minerals. Increased water usage, seasonal inputs from surface runoff, and soil accumulation also contribute to these effects. COD and SAR show a mix of negative and positive values, suggesting stable TWW quality. March and April showed relatively stable TWW quality, possibly due to favorable weather conditions or reduced human activity. The elevated levels of turbidity and Na are clustered closely with July and September, which also display moderately high values in several parameters, suggesting possible late-summer contamination events, potentially linked to rainfall and runoff. CWQI was in the good class and excellent class, except for April 2020, when it was in the fair class.

3.4. Comprehensive Pollution Index (CPI)

Based on the 2013 data (Figure 5a), the CPI results varied from 0.04 to 1.7, from 0.05 to 1.62, from 0.05 to 2.02, and 0.05 to 2.22 in the winter, spring, summer, and autumn, respectively. TWW was divided into four different groups: clean, sub-clean, slightly contaminated, and heavily polluted. The CPI data for 2014 (Figure 5b) showed fluctuations of 0.01–1.73, 0.02–1.35, 0.04–1.17, and 0.2–1.33 in the winter, spring, summer, and autumn, respectively. The increase in turbidity in the autumn of 2013 caused the quality of TWW to deteriorate, whereas the elevated E. coli caused TWW to worsen in the autumn of 2014. The TWW in 2014 was classified as being slightly polluted by E. coli. The CPI values fluctuated in 2015 between 0.01 and 0.78, 0.03 and 2.04, 0.04 and 1.93, and 0.01 and 1.57 in the winter, spring, summer, and autumn, respectively. The TWW was classified as severely polluted in the winter because of the increase in turbidity, as depicted in Figure 5c. However, the quality of TWW in the autumn worsened because of increased ammonia (NH₄). Complete ammonia oxidation can result from microplastics’ inhibition of nitrifying bacteria’s development and activity, including those of Nitrosomonas. Ammonia build-up and increased toxicity in effluents might result from this disturbance [103]. In most seasons, rising turbidity was the primary cause of TWW contamination. The CPI values fluctuated in 2016 between 0.01 and 1.57, 0.01 and 1.05, 0.25 and 1.85, and 0.03 and 1.53 in the winter, spring, summer, and autumn, respectively. Rising turbidity was the main factor contributing to TWW pollution across all seasons, as Figure 5d illustrates. Concerning 2017 data (Figure 5e), the CPI results varied from 0.1 to 2.22, from 0.03 to 0.98, from 0.14 to 1.18, and 0.05 to 0.87 in the winter, spring, summer, and autumn, respectively. The CPI data for 2018 (Figure 5f) showed fluctuations of 0.03–0.96, 0.14–0.6, 0.09–0.93, and 0.26–1.4 in the winter, spring, summer, and autumn, respectively. The increase in turbidity and ammonia in the autumn and summer of 2017, respectively, caused the quality of TWW to deteriorate, whereas the elevated turbidity caused TWW to worsen in the autumn of 2018. The TWW in 2014 was classified as being slightly polluted by turbidity.

The CPI values fluctuated in 2019 between 0.01 and 0.82, 0.23 and 2.19, 0.21 and 2.66, and 0.17 and 2.66 in the winter, spring, summer, and autumn, respectively.

The TWW was classified as severely polluted in the winter and the autumn because of the increase in turbidity, as depicted in Figure 5g. In most seasons, rising turbidity was the primary cause of TWW contamination. The CPI values fluctuated in 2020 between 0.06 and 1.22, 0.01 and 0.94, 0.01 and 1.35, and 0.01 and 1.17 in the winter, spring, summer, and autumn, respectively. TWW quality was classified as clean, sub clean, slightly polluted, and moderately polluted. Rising turbidity was the main factor contributing to TWW pollution across all seasons, as Figure 5h illustrates.

The CPI average values of all years ranged from 0.19 to 0.77, but the average CPI was determined to be 0.6, indicating slight pollution throughout the entire period. An identical finding was reported by Shakir et al. [104], who discovered an average CPI of 0.69, indicating light contamination. Figure 5 shows the changes in the CPI value for each parameter taken into consideration. The study found that 50% of turbidity data exceeded the maximum permitted level for RI. This might be because of inadequate influent treatment or elevated turbidity, which lowers chlorine effectiveness in water with high COD and turbidity. A similar result was found by Al-Hammad et al. [119], who mentioned that 66% of samples exhibited turbidity levels between 6.0 and 8.2 NTU, above the 5.0 NTU limit. The turbidity measurements of 33% of the samples fell within the recommended limit. However, Al-A’ama and Nakhla [120] and Al-Turki [90] recorded lower findings, with mean values of 1.6 and 2.3 NTU, respectively, from studies of municipal WWTPs in the KSA (Jubail and Buraidah). Although WWTPs lessen the number of harmful microorganisms, they do not get rid of them entirely. Many crops that are watered with wastewater that has been treated present different health risks, while some of these crops may include pathogenic microorganisms that are not thought to be detrimental to humans. These include industrial crops like cotton or fodder, fruits that are dried for at least 60 days after irrigation, watermelons produced for edible grains or seeds, and crops that are inaccessible to humans. There might not be any health risks associated with these crops [121,122].

3.5. TDS Hazard

TDS levels ranged from 1115.02 to 1435.39 mg/L (mean value = 1282.16 mg/L) in research samples, demonstrating a high level of acceptance for the usage of this TWW in irrigation. The permissible TDS level for TWW was 2500 mg/L. It is clear that TWW irrigation does not raise the soil salt concentration, because the soil salt concentration was below the permissible threshold (3 dS/m). Irrigation water with a conductivity of 1.74 to 2.24 dS/m is allowed for irrigation [93]. TWW can therefore be utilized for agricultural irrigation.

3.6. Sodium Adsorption Ratio (SAR)

According to the data enquiry, the SAR average was found to be 4.81, with a range of 2.75 to 8.13. SAR falls into class II (about 90% of the data), which denotes low sodium. Class I comprises 2% of the data, and Class III includes 8% of the data with a value of less than 8. Most of the data have few problems for irrigation usage, except with sodium-sensitive crops. SAR values are described in

Table 5. Classification of irrigation water based on SAR according to [123].

Class	Sodicity Class	SAR (%)
Class I	No sodium problem	<3
Class II	Low sodium, few problems except with sodium-sensitive crops	3–6
Class III	Medium sodium, increasing problems	6–8
Class IV	High sodium, not generally recommended	8–14
Class V	Very high sodium, unsuitable	>14

based on the categorization of irrigation water.

3.7. Principle Component Analysis (PCA) of TWW Parameters

The PCA analysis produced four main factors, with the most significant having the greatest eigenvalues.

Table 6. The PCA’s squared cosines with factor loadings from 2013 to 2016 and from 2017 to 2020.

	From 2013 to 2016				From 2017 to 2020
	PC1	PC2	PC3	PC4	PC1	PC2	PC3	PC4	PC1
Free Cl				0.251
DO						0.289
Na			0.303				0.328
Ca2++Mg2+	0.532				0.466				0.466
E. coli				0.433
COD		0.502			0.541				0.541
NO3	0.336				0.576				0.576
NH4		0.425				0.540
Turbidity		0.269
EC	0.569						0.623
pH			0.576					0.491
Eigenvalue	2.018	1.520	1.234	1.167	2.153	1.484	1.290	1.032	2.153
Variability (%)	18.341	13.821	11.221	10.609	19.569	13.489	11.724	9.379	19.569
Cumulative %	18.341	32.162	43.383	53.992	19.569	33.059	44.783	54.162	19.569

show the eigenvalues of the five components representing the entire variation in TWW quality across the study period and explaining around 53.9% and 54.2% of the total data variability for 2013–2016 and for 2017–2020, respectively, with loss of information of about 46.1% and 45.8%. The dimension of the data was reduced from 13 variables to 4 components. The PCA explained the difference in datasets by identifying the TWW quality characteristics. The total fluctuation of all parameters was described by the first and second PC over 2013–2016, accounting for 32.1% (18.3% and 13.8%, respectively) of the total variance across all parameters, as shown in Table 6. Nevertheless, during 2017–2020, it accounted for 33.1% (19.6% and 13.5%, respectively), which is also explained by the first and second PCAs, as shown in Table 6. All parameters of WWTP effluent in all years were accurately represented by the four components.

The parameters of 2013–2016 effluent, including Ca²⁺+Mg²⁺, NO_3, and EC, were loaded onto PC1. However, COD, NH₄, and turbidity were loaded onto PC2. The remaining traits (Na and pH, and free Cl and E. coli) were loaded onto PC3 and PC4, respectively, as shown in Table 6. While PC1 has effluent parameters including COD, NO₃, and Ca²⁺+Mg²⁺ in 2017–2020, PC2 was laden with NH₄ and DO in 2017–2020. PC3 and PC4 were loaded with the remaining features (Na, pH, and EC), as shown in Table 6. The highest loading in 2013–2016 was attributed to the first main component (PC1) in the dataset, which makes up 18.3% of the total information and is the most significant component, based on PCA findings, whereas the first component (PC1) in the dataset makes up 19.5% of the total information in 2017–2020. This component can offer a dominant pattern of the data to help better comprehend TWW characteristics. It represented the loading of the parameters (Ca²⁺+Mg²⁺, NO₃, and EC) in 2013–2016, while it indicated the loading of the NO₃, COD, and Ca²⁺+Mg²⁺ parameters in 2017–2020. They have large positive loadings on component 1. The second component (PC2) accounts for 13.8% of the initial data variance, for COD, NH₄, and turbidity, of TWW parameters for 2013–2016, as shown in Table 6. On the other hand, PC2 accounts for 13.5% of the initial data variance of TWW parameters (NH₄ and DO) throughout 2017–2020, as shown in Table 6. The third component, loaded by Na and pH, contributes to 11.2% of the variability in 2013–2016, as shown in Table 6. However, in 2017–2020, about 11.8% of the variability is attributed to the third component, which is loaded by Na and EC, as shown in Table 6. The fourth component (PC4) demonstrates a notable positive correlation with E. coli and free Cl parameters, explaining 10.6% of the variability, as shown in Table 6. PC4 exhibits a positive relationship with the pH parameter, accounting for 9.4% of the variability, as illustrated in Table 6. These elements give rise to a dominant pattern that helps comprehend the TWW characteristics. The resulting components are shown in Table 6. The squared cosines of a PCA measure the degree of correlation between variables and principal components, taking values between 0 and 1, with a close 1 indicating a high correlation. The components with higher loadings better describe the data [56].

A PCA biplot is a visual illustration that combines loadings and scores from a PCA and shows variables as arrows or vectors, illustrating the relations between 11 variables in PC1 and PC2 in a reduced-dimensional space, as shown in Figure 6a,b. Longer arrows indicate stronger correlations and impacts on main components, while pointing in the same direction implies positive variables. In PC1, there is a strong positive relationship in 2013–2016 between PC1 and Ca²⁺+Mg²⁺, EC, NO₃, and Na, but a small negative relationship between PC1 and free Cl, E. coli, DO, and NH₄. On the other hand, there is a strong positive relationship in 2017–2020 between PC1 and Ca²⁺+Mg²⁺, pH, and NO₃, but a small negative relationship between PC1 and turbidity, E. coli, and NH₄, and a large negative relationship between PC1 and Na and COD. In the perpendicular axes (PC2), NH₄, COD, and turbidity have a large positive relationship with PC2, while NO₃ has a minor negative one with Na and EC in 2013–2016, whereas there is a small negative correlation between PC2 and E. coli and EC and a large negative correlation between PC2 and DO and turbidity in 2017–2020. There is a large positive correlation between PC2 and free Cl and NH₄; however, there is a minor positive correlation between PC2 and Ca²⁺+Mg²⁺, NO₃, Na, and pH. Figure 6a,b illustrate a biplot of correlations between the variables and components.

In a PCA biplot (Figure 6a,b), the distribution of the initial variables in factor space is depicted on a map called the correlation circle. It shows whether there is a substantial positive correlation (r near to 1), an orthogonal correlation (r close to 0), or a negative correlation (r close to −1). The angle between variables’ arrows indicates their relationship or similarity, with small angles indicating strong positive correlation and larger ones indicating weaker correlation. The relative angle between arrows is crucial for assessing their relationship, with the length of arrows representing the variables’ importance and direction indicating their relationship with principal components. A small angle implies positive correlation, a large one suggests negative correlation, and a 90° angle indicates no correlation between two characteristics.

The arrows’ length represents the strength or importance of the variables in explaining the variation in the data, while the arrows’ direction indicates the direction of the relationship with the principal components. The vectors of the pH, COD, and turbidity parameters formed an acute angle (less than 90°) with PC1 in 2013–2016, indicating a positive correlation. Conversely, the vectors of the Na, Ca²⁺+Mg²⁺, EC, and NO₃ parameters formed an angle greater than 90° with PC1. However, there was a negative relationship between PC1 and the free Cl, DO, E. coli, and NH₄ parameters.

On the other hand, in 2017–2020, the vectors of the free Cl, Ca²⁺+Mg²⁺, COD, NO₃, and turbidity parameters formed an acute angle (less than 90°) with PC1, indicating a positive correlation; conversely, the vectors of the Na, NH₄, and COD parameters formed an angle greater than 90° with PC1. However, there was a negative relationship between PC1 and the EC, E. coli, and turbidity parameters. The vector of the DO parameter formed an angle greater than 90° with PC1.

4. The Significance and Practical Application of This Study

In 2022, Saudi Arabia’s use of TWW in agriculture was just 22.6%. The 2030 Saudi Vision aims to achieve 100% TWW utilization by 2025. This study’s findings demonstrate the appropriateness of TWW for restricted irrigation in support of the Saudi Green Initiative, which aims to plant 10 billion native trees across Saudi Arabia. PCA and heatmap clustering analysis are usually examined independently. However, this study combines the two approaches. This investigation may be essential to comprehending the relation between TWW variables in evaluating TWW quality, particularly when there is a drought in the agricultural sectors, which consume a large amount of water in such arid areas. In order to provide a useful technique for assessing TWW quality, the study suggests integrating CWQI, PCA, and heatmap clustering analysis. This would assist in determining how to evaluate these factors most effectively. Planners and decision-makers can refer to the results, which pinpointed the crucial variables, while deciding whether or not to reuse these TWWs. The study examines the reuse of TWW in the Riyadh region, supporting Saudi Arabia’s Green Riyadh Initiative, and promoting sustainable water resource management and environmental conservation. The analysis helps identify and enhance key elements influencing TWW quality, contributing to sustainable urban development and aligning with regional environmental goals.

5. Conclusions

A potential solution for farmers in Riyadh faced with Saudi Arabia’s water deficit is to use TWW for irrigation; however, environmental and health hazards must be considered before implementation. Using historical data from 13 parameters collected over an eight-year period, this study was able to determine the temporal fluctuations in TWW quality. The data acquired for TDS, EC, pH, free Cl, SAR, Na, and COD parameter values were less than the maximum permitted levels for restricted irrigation (RI). However, 6.4%, 19.1%, 28.7%, and 50% of E. coli, NO₃, NH₄, and turbidity data were over the permitted level, respectively. The CWQI results showed a range of 72.95 to 95.55%, with an average of 84.81%. The CWQI revealed that 14.1% of months were fair, 81.9% of months good, and 3.2% excellent, indicating that reuse of TWW does not harm public health. The average CPI for all years was 0.6, indicating light pollution. According to the findings of the principle component analysis (PCA), the characteristics of 2013–2016 effluent, such as Ca²⁺+Mg²⁺, NO₃, and EC, were loaded onto PC1, which accounted for 18.3% of data fluctuation. Furthermore, COD, NH₄, and turbidity were placed onto PC2. PC1 had effluent parameters such as COD, NO₃, magnesium, and calcium in 2017–2020, which account for 19.5% of the overall information, while PC2 was loaded with NH4 and DO. These two components can provide a prominent pattern of the data to assist in better understanding TWW features.

The heatmap of the WWTPs in Riyadh showed increases in NO₃ in January 2013, exceeding the maximum allowable limit for restricted irrigation. Constant monitoring is needed to adjust fertilizer quantities accordingly. In 2014, the behavior of “E. coli” and “DO” suggested that bacterial contamination influences water oxygen levels, indicating a correlation between these variables. In 2015, the clustering of F chlorine and E. coli suggested similar behavior over time, with high bacterial contamination reducing oxygen levels and impacting aquatic ecosystem health. The summer months in 2015 (June, July, and August) showed neutral values, suggesting stable TWW quality, as was the trend in 2014. In 2016, elevated values in June were noted for E. coli and turbidity, which might indicate potential pollution, decreased treatment activity, or changes in the natural environment. About 50% of turbidity data exceeded the permissible limit in general, affecting TWW quality and safety and potentially regrowing harmful bacteria due to insufficient treatment. This may cause a decline in chlorine efficiency in high-turbidity water. However, the turbidity problem can be overcome by sand media filters, which are recommended for TWW in micro-irrigation. Other parameters, such as pH, TDS, EC, COD, and free Cl, maintained more stable or lower values in 2016 and other years.

In conclusion, TWW is suitable for meeting RI requirements using these approaches, though not for unrestricted irrigation. It offers potential resources in dry and semi-arid regions, enhances resource recovery, reduces fertilizer and agricultural input needs, and contributes to sustainable irrigated agriculture, climate change mitigation, and reduced greenhouse gas emissions. These approaches are accurate and resource-saving, as they reduce the parameters utilized to measure TWW quality. This approach may be used in different areas to evaluate the TWW quality and provide information for managing water resource decision-making.