Assessment of Pollution Status in Brunei River Using Water Quality Indices, Brunei Darussalam

Abstract

The Water Quality Index (WQI) is a tool designed to provide a singular figure representing the overall water quality status of a water body. This study applies Malaysia WQI, National Sanitation Foundation WQI (NSFWQI) and statistical analysis to investigate the impact of diverse pollution sources on the Brunei River’s water quality, a critical aquatic ecosystem affected by the rapid escalation of urbanization, industrial activities, and agricultural runoff. Principal component analysis (PCA), expert judgement, and correlation analysis were used to propose parameters for developing Brunei River’s WQI. Eight monitoring stations were selected to analyze 16 water quality parameters (pH, water temperature (T), dissolved oxygen (DO), oxidation-reduction potential (ORP), chemical oxygen demand (COD), the five-day biochemical oxygen demand (BOD₅), salinity, electrical conductivity (EC), total dissolved solids (TDS), turbidity, total suspended solids (TSS), ammoniacal-nitrogen (NH₃-N), fecal coliform (FC), total coliform (TC), phosphate and nitrate (NO₃⁻) in this study. The results showed that NSFWQI classified the Brunei River as moderately polluted, while Malaysia (WQI) status was classified as slightly polluted except for Station Q around the quarry area. Statistical analysis revealed that the primary pollution sources are anthropogenic activities such as quarrying, domestic waste, and agricultural and urban runoff. Other specific areas of concern with low WQI and significant pollution levels are situated at Kampong Ayer Stations (N and J) due to proximal anthropogenic activities. The proposed water quality parameters for developing Brunei River’s WQI are pH, DO, EC, FC, NO₃⁻, BOD₅, T, TSS, turbidity and phosphate. This study addresses the current pollution status of the Brunei River and sets a precedent for future research emphasizing collaborative data-driven strategies for water quality management.

1. Introduction

Rivers are integral components of ecosystems, serving as lifelines for diverse flora and fauna while also meeting the essential needs of the human population. Water quality within the river system is a key determinant of its suitability and the broader impact on surrounding ecosystems. Rivers act as pathways for municipal and industrial wastewater, along with agricultural runoff, which are notably susceptible to pollution across their wide-reaching drainage basins [1]. Human-derived modifications to catchments have altered the water quality of rivers and introduced new water movement pathways—for example, cultural eutrophication from point and non-point source pollution [2]. Point and non-point source pollution affect surface water ecosystems, causing eutrophication, excessive algal growth, and chemical and microbial contamination [3,4].

Numerous countries, including but not limited to the United States, Australia, Indonesia and Malaysia have documented cases of water pollution [5,6,7,8]. Jalilov [9] reported that the lack of household sewerage systems and the contamination of waterbodies with hazardous industrial chemicals are to blame for the poor water quality in many countries in Southeast Asia. The Brunei River is the most polluted major River [10] in Brunei Darussalam, serving multiple roles, including a habitat for many aquatic species. However, like many water bodies, the Brunei River faces threats from human-derived modifications such as the direct discharge of effluents and domestic wastewater into the river [11,12]. Despite extensive efforts to mitigate the pollution of the Brunei River, it continues to be a significant issue. Recently, an initiative by Green Brunei and the Mitsubishi Corporation (Brunei Liaison Office) resulted in the collection of over 400 kg of waste from the river [13]. These challenges motivated this research. Addressing these threats requires effective water quality management, which often involves a variety of chemical, physical and biological water quality indicators as discussed in the works of Gupta et al. [14] and Shah and Joshi [15]. To simplify this process, several researchers proposed a Water Quality Index (WQI) in the form of a simple expression for the representation of the general quality of surface waters [16]. For example, Shah and Joshi [15] evaluated three stations at Sabarmati River in India using WQI, and discovered that the river degradation is impacted by human activities and sewage disposal, with the worst conditions observed around the urban areas.

The WQI enables a straightforward assessment of water conditions by consolidating data from multiple water quality parameters into a single, understandable figure [17]. Uddin et al. [17] explained that the general structure of any WQI consists of several water quality parameters, which are transformed into a common scale called sub-indices. After the sub-indices have been obtained, weights are then assigned which reflect the relative importance of the water quality parameter and/or the appropriate guidelines of water quality [18]. The calculation of the weights of these parameters can vary based on regional priorities such as the importance of protecting aquatic biodiversity or ensuring water safety for human consumption and/or internationally recognized WQI. This flexibility allows each country or region to tailor the WQI to their specific environmental challenges and goals. The final WQI is aggregated from sub-indices and weights values using additive or multiplicative methods, or a mix of both, to avoid issues like eclipsing and ambiguity in the index’s representation of water quality [19]. The index typically ranges from 0 to 100, where higher values indicate better water quality [17]. By providing a single figure, the WQI helps to identify areas where water quality is compromised and allows tracking changes over time, facilitating the evaluation of pollution control measures. The importance of a well-established WQI is particularly evident in regions experiencing polluted rivers due to inadequate waste-management infrastructure and industrial activities, as reported in the studies by Shah and Joshi [15], Sutadian et al. [20] and Liou et al. [21]. There are many WQI calculations tailored towards different regions globally, however Brunei is yet to have its own WQI.

Therefore, our study focuses on the selection of parameters for developing Brunei River’s WQI, based on expert judgement and statistical analysis. An example of such statistical analysis is the principal component analysis (PCA). PCA reduces the dimensionality of a multivariate data set while maintaining its original structure to the maximum extent possible [22]. Thus, some researchers [22,23] have used PCA for parameter selection while dealing with diverse monitoring data for developing WQI in a region. This study addresses the importance of using PCA and experts’ decision for WQI parameter selection.

The second objective is to address the pollution status of the Brunei River using two established WQI. The two WQI assessment methods (NSFWQI and Malaysia-WQI) were selected after reviewing the existing literature and on the basis of available data. The National Sanitation Foundation WQI (NSFWQI), endorsed by the United States National Sanitation Foundation (USNSF), is the most common method to determine surface water’s WQI globally and has been implemented either in its original form or with modifications [24,25,26,27]. NSFWQI covers a wide range of indicators, including physical, chemical and biological parameters, making its adaptability suitable to address specific local conditions in the Brunei river. NSFWQI evaluates water quality using nine specific parameters, which include dissolved oxygen (DO) saturation, fecal coliform (FC), five-day biochemical oxygen demand (BOD₅), pH, water temperature (T), total phosphate (TP), nitrate (NO₃⁻), total solids (TS) and turbidity. Malaysia WQI was also considered, mainly due to the proximity of the country to Brunei, and it shares the same equatorial tropical climate and geography. Both Malaysia and Brunei face significant river pollution challenges stemming from earthworks, agriculture and sewage—both treated and untreated [11,28]. Malaysia’s WQI [29] utilizes six parameters: DO, chemical oxygen demand (COD), BOD₅, ammoniacal-nitrogen (NH₃-N), suspended solids (SS) and pH.

However, the two models may be less sensitive towards other water quality pollutants, therefore, to increase the reliability of the WQI results, statistical analyses were performed. Several researchers have used different statistical analysis, such as interval plots [30] and correlation [31] to evaluate pollution sources in a river. For example, Pham et al. [32] used WQI and multivariate statistics to determine influence of pollution sources in the Dong Nai River, Vietnam. They attributed the degradation of the river to anthropogenic activities such as domestic, agricultural practices and seasonal change. The use of these various statistical analyses helps in the interpretation of complex datasets to better understand the relationship between parameters, direction and water quality of river bodies. Overall, sixteen water quality parameters (pH, water temperature, oxidation-reduction potential (ORP), turbidity, DO, electrical conductivity (EC), total dissolved solids (TDS), TSS, NH₃-N, salinity, NO₃⁻, phosphate, BOD₅, COD, FC and total coliform (TC)) were assessed in this study.

Prior to this research, WQI had not applied in Brunei, so the findings of this research provide a clear assessment of the pollution level of the Brunei River. Additionally, the research serves as a valuable resource and framework for future studies for developing Brunei’s WQI.

2. Materials and Methods

2.1. Study Area

Brunei Darussalam, occupying 5765 km², is situated on Borneo Island bordered by the South China Sea and surrounded by Malaysia’s Sarawak state [12,33] (Figure 1). Brunei’s climate is characterized as wet–tropical with average monthly temperatures of 26.7 °C to 27.7 °C, and monthly rainfall of 375 mm [33]. Seasonal climatic variations are influenced by the stronger northeast monsoon from October to January, and the weaker southeast monsoon from April to August. The rainfall peaks from October to January and May to July, and there is lower rainfall in February, March and June to August. The country landscape predominantly consists of low undulating terrain, with mountainous areas in the southeast [33]. Its geomorphology was significantly shaped by a major sea-level subsidence event approximately 5000–6000 years ago [33]. The country’s hydrology is dominated by four major rivers: Belait (209 km), Temburong (98 km), Tutong (137 km) and Brunei (41 km), with the latter two being crucial for the nation’s water supply for domestic, industrial and agricultural uses [12,33]. The Brunei River is a tidal estuarine river stretching approximately 41 km in length [34], flowing through Brunei-Muara district and meandering northeastward until it discharges into the Brunei Bay. The river’s catchment primarily consists of alluvial deposits that overlie the Belait Formation [35].

Additionally, the Brunei-Muara area is predominantly covered by recent peat deposits, with nearby hills made up of sandstone and shale. More than sixty percent of the country’s population resides in the Brunei Muara district, where the capital city (Bandar Seri Begawan) is located.

A floating water village (Kampong Ayer) is situated on the river while the capital city stands on its banks. The lower stretches of the river are densely populated by mangroves and Nipah palms, which serve as a vital breeding ground for coastal fisheries. Conversely, the upper reaches of the river are a crucial freshwater source for the western part of the country [36]. Sources of water pollution in the Brunei River have been attributed to wastewater treatment facilities, industrial discharges, agriculture and contaminated streams [36,37,38]. In many areas around the river, monsoon drains are often overlooked and frequently become clogged with silt, leading to anaerobic conditions and problems with odors. During storms, these drains can experience high flow rates, causing effluent from septic tanks and other debris to be rapidly flushed into the river. Domestic wastes and surface runoff constitute about 50 percent and 29 percent of the pollutant load discharged into the Brunei river, respectively [39]. High population density and urban catchments are also primary contributors to the prevalent pollution levels in the river [40].

2.2. Sampling Points

Eight monitoring stations—B, D, E, G, J, N, P and Q—are observed along the river, as shown in Figure 1. These points were selected due to their proximity to areas affected by human-induced modifications causing point and non-point source pollution. The co-ordinates extend from 4°52′ N to 4°56′ N, and from 114°54′ E to 115°0′ E. Station Q is located on Sungai Damuan, a tributary flowing into Sungai Brunei, approximately 5 km upstream from the commercial hub and the capital city, Bandar Seri Begawan. This area, rich in mangrove forests, is monitored to assess pollution impacts from a nearby quarry along the riverbank. Downstream, approximately 2 km from Station Q, lies Station P at the merging point of Sungai Brunei and Sungai Damuan within a largely forested area surrounded by agricultural activities. Further down, 2.5 km from Station P, Station N is found in the initial segment of Kampong Ayer within the Pengiran Tajuddin Hitam residential area, a location concerned with potential untreated sewage discharge. Station J, situated about 1 km downstream from Station N, lies at the interface of Kampong Ayer’s latter section and Bandar Seri Begawan, bordered by cement walkways and subjected to domestic and industrial waste, including runoff from the city center and the Sungai Kianggeh tributary. Approximately 1.5 km downstream from Station J, Station G marks the transition from urban walkways to natural riverbanks. Station E is positioned about 1.3 km downstream from Station G, close to the Pintu Malim Sewage Treatment Plant. A distance of 1.3 km downstream from Station E, Station D is near the Brunei Darussalam Maritime Museum. The final station, Station B, is found roughly 3.4 km downstream from Station D and 3.6 km from the Brunei Bay mouth, in an area blending forested and residential settings.

2.3. Data Collection

Measurements were conducted once or twice a month from August 2022 to December 2023, with COD, nitrate and phosphate measured from September to December 2023. The Aqua Troll 600 Multiparameter Sonde was used for in situ water quality monitoring in the Brunei River. Recognized for its advanced capabilities in assessing aquatic environmental health, the Aqua Troll 600 facilitates the simultaneous measurement of multiple water quality parameters including, but not limited to, pH, DO, water temperature, ORP, TDS, NH₃-N, EC and salinity. Prior to deployment, a calibration protocol was implemented to ensure the precision and reliability of the data collection. Each calibration was done on the sensor to a known calibration standard [41]. For instance, pH sensor was calibrated using standard buffers at pH 4.00, 7.00 and 10.00 every 10 weeks. Factory calibration of Aqua troll 600 instruments was done annually at In-Situ Incorporated, Fort Collins CO, USA, or when significant data drift was observed. This process included cleaning and complete functionality checks of all sensors.

Three data points were frequently obtained at each monitoring station to boost accuracy due to potential instability in readings caused by river current, boat movement and the removal of any potential outliers due to measurement errors. Water samples were collected at depths of approximately 2 to 4 m using a Van Dorn sampler, then immediately transferred to sterile bottles, ensuring no air bubbles were present, and stored in a chilled container throughout the sampling process. Turbidity was measured using the HF scientific M100þ laboratory turbidimeter. Monthly readings of TSS concentration readings were collected from the Department of Water Services Brunei, Jabatan Kerja Raya (JKR), Brunei Darussalam. A HI83399 Water and Wastewater Multiparameter (with COD) Photometer was used to measure COD and phosphate, while H1782 nitrate checker was used for analyzing the nitrate concentration. The BOD was measured over 5 days [42]. In contrast, coliform bacteria were quantified using a multiple tube fermentation approach [43] at the wastewater laboratory of the Department of Drainage and Sewerage (DDS), JKR, Brunei Darussalam.

2.4. Statistical Analysis

Interval plots were used to visually compare the central tendency and variability of the 16 (pH, water temperature, DO, ORP, COD, BOD₅, salinity, EC, TDS, turbidity, TSS, NH₃-N, FC, TC, phosphate and NO₃⁻) water quality parameters across the eight sampling sites along the river. These plots displayed the mean values and the associated 95% confidence mean intervals for each measured parameter.

Principal component analysis (PCA) was used to simplify and select the proposed Brunei river WQI parameters by reducing the dimensionality of the data of thirteen water quality parameters [22], excluding COD, NO₃⁻ and phosphate due to low data availability. Prior to applying PCA, the parameters were standardized due to their differing units. This standardization was done by calculating the z-scores for each parameter, setting the mean to zero and the standard deviation to one [44]. This step ensured that no single parameter with a larger scale and variance dominated the PCA results. The suitability of the data for the PCA analysis was assessed using the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy and Bartlett’s test of sphericity [11]. The KMO measure evaluates the proportion of variance among variables that might be common variance. A KMO value closer to 1 indicates that the data are suitable for PCA [22]. Bartlett’s test of sphericity tests the null hypothesis that the correlation matrix is an identity matrix, which would indicate that variables are unrelated and unsuitable for structure detection. A significant Bartlett’s test (p < 0.05) suggests that the correlations in the data set are appropriate for PCA [11,22]. Following the normalization and suitability tests, PCA transformed the original parameters into a new set of variables known as principal components (PC) based on eigenvalues greater than one. Furthermore, PC loadings were categorized as strong (greater than 0.75), moderate (0.75–0.50) and weak (0.50–0.30) [45]. Pearson correlation was also used to measure the strength and direction of the linear relationship between the water quality parameters.

2.5. National Sanitation Foundation WQI (NSFWQI)

The NSFWQI is calculated based on the evaluation of several fixed water quality parameters (n = 9), which include DO saturation, FC, pH, BOD₅, water temperature, TP, NO₃⁻, turbidity and TS. The NSFWQI is segmented into five classes of water quality categorized as very poor (0–25), poor (25–50), medium (50–70), good (70–90) and excellent (90–100). The fixed parameters have different weight coefficients (w_i), with DO saturation (%) and TS having the highest and lowest weights (0.17 and 0.07, respectively) in the calculation of NSFWQI [24]. The weights assigned to each parameter were based on a statistical survey using the Delphi technique. This method involved 142 experts who contributed to the formulation of weights for the index. NSFWQI is defined in Equation (1) as follows [24]:

$$\mathrm{NSFWQI}={\sum }_{i=1}^n{w}_i{q}_i$$

(1)

where q_i is the sub-index for ith water quality parameters, w_i is the weight (in terms of importance) associated with the ith water quality parameter, and n is the number of water quality parameters. Other relative weights for the parameters (w_i in Equation (1)) are as follows: FC (0.16), pH (0.11), BOD₅ (0.11), water temperature (0.10), NO₃⁻ (0.10), TP (0.10) and turbidity (0.08). Colony-forming units (CFU/100 mL) are required for FC calculations [24]. However, in our study, FC was measured in MPN/100 mL. Recent research [46] has demonstrated that utilizing MPN/100 mL instead of CFU/100 mL is acceptable and does not significantly impact the results of the WQI. Furthermore, the parameter subindex used in this study (q_i) was derived from the respective rating curves [47,48]. DO concentration (mg/L) at a given water temperature was converted into DO saturation (%) to define the subindex for parameter DO. After the q-value (q_i) was obtained, it was multiplied by a weighting factor (w_i) to give the overall water quality.

2.6. Malaysia WQI

The Malaysia WQI fixed system [29] was also utilized to gauge the health and quality of water bodies on a regional scale. The Malaysia WQI formula (Equation (2)) assesses water quality based on the measurements of six parameters: DO, BOD₅, COD, NH₃-N, SS and pH. DO is converted to DO saturation at a given temperature. The parameter concentrations are converted into subindices by rating curve functions [49] based on water quality standards. The index combines these parameters into a single value that ranks water quality on a scale from 0 to 100, categorizing it into three main classes: clean (90–100), slightly polluted (70–90) and polluted (0–70).

WQI = (0.22 ∗ SIDO) + (0.19 ∗ SIBOD) + (0.16 ∗ SICOD) + (0.15 ∗ SIAN) + (0.16 ∗ SISS) + (0.12 ∗ SIpH)

(2)

where, SI relates to subindices of those parameters (n = 6) that were obtained from a series of equations.

Fit equations for the estimation of various subindex values for the Malaysia WQI:

Subindex for BOD

SIBOD = 100.4 − 4.23x          for x ≤ 5

(3)

SIBOD = 108 ∗ exp(−0.055x) − 0.1x    for x > 5

(4)

Subindex for DO (in % saturation)

SIDO = 0                   for x ≤ 8

(5)

SIDO = 100                   for x ≥ 92

(6)

SIDO = −0.395 + 0.030x² − 0.00020x³    for 8 < x < 92

(7)

Subindex for COD

SICOD = −1.33x + 99.1           for x ≤ 20

(8)

SICOD = 103 ∗ exp(−0.0157x) − 0.04x    for x > 20

(9)

Subindex for NH3-N

SIAN = 100.5 − 105x              for x ≤ 0.3

(10)

SIAN = 94 ∗ exp(−0.573x) − 5 ∗ [x − 2]    for 0.3 < x < 4

(11)

SIAN = 0                   for x ≥ 4

(12)

Subindex for SS

SISS = 97.5 ∗ exp(−0.00676x) + 0.05x       for x ≤ 100

(13)

SISS = 71 ∗ exp(−0.0061x) − 0.015x    for 100 < x < 1000

(14)

SISS = 0                  for x ≥ 1000

(15)

Subindex for pH

SIpH = 17.2 − 17.2x + 5.02x²          for x < 5.5

(16)

SIpH = −242 + 95.5x − 6.67x²       for 5.5 ≤ x < 7

(17)

SIpH = −181 + 82.4x − 6.05x²       for 7 ≤ x < 8.75

(18)

SIpH = 536 − 77.0x + 2.76x²         for x ≥ 8.75

(19)

The sub-indices were calculated using the parameter mean of each station based on Equations (3)–(19). For example, the BOD mean in Station B was 2.54 (

Table 1. Descriptive statistics of the 16 water quality parameters.

Stations		pH	T	DO	ORP	COD	BOD5	Salinity	EC	TDS	Turbidity	TSS	NH3-N	FC	TC	Phosphate	NO3−
Unit			°C	mg/L	mV	mg/L	mg/L	PPT	μs/cm	mg/L	NTU	mg/L	mg/L	MPN /100 mL	MPN /100 mL	mg/L	mg/L
B	min	6.47	24.6	3.5	84.41	218	0.9	14.1	19,612	11.66	2.7	18	0.07	400	2300	0.03	0.04
	max	7.7	30.9	8.4	256.4	696	3.9	39.5	43,459	25.94	10.3	68	0.18	5000	90,000	0.04	0.04
	mean	6.99	28.85	6.2	168.1	546	2.54	21.6	29,204	17.39	6.3	36.4	0.0017	1733	18,333	0.03	0.04
D	min	6.41	24.2	3.9	95.98	215	0.4	11.9	17,886	10.66	2.9	21	0.047	700	2300	0.02	0.08
	max	7.46	30.8	7.9	234.6	738	4.4	33.6	40,323	24.01	16.9	88	0.103	5000	90,000	0.04	0.78
	mean	6.88	28.23	5.9	163.7	518	2.39	18.2	26,687	15.99	8.01	41.3	0.0024	1800	14,183	0.02	0.43
E	min	6.31	23.9	4.1	78.88	201	0.3	4.2	20,788	12.4	2.5	21	0.05	200	2300	0.01	0.22
	max	7.41	30.9	6.8	230.1	639	4.8	39.4	39,467	23.51	20.45	70	0.09	28,000	160,000	0.02	0.9
	mean	6.84	28.31	5.5	152	344	2.7	17.5	25,290	15.25	9.88	40.2	0.0045	5258	36,025	0.08	0.56
G	min	6.37	22.5	3.9	81.3	169	7.9	10.6	18,430	11.33	2.6	4	0.035	400	2300	0.01	0.13
	max	7.14	30.9	6.7	202.6	565	2.9	22.4	32,823	19.52	20.6	58	0.13	13,000	90,000	0.17	0.57
	mean	6.8	28.13	5.4	141.9	453	2.89	16.1	22,939	13.92	9.44	40	0.0087	2967	18,158	0.07	0.35
J	min	6.4	23.8	3.6	81.04	198	0.9	10.6	19,023	11.35	3.2	16	0.02	900	3000	0.01	0.05
	max	7	30.7	6.1	178.1	645	3.6	30.2	37,130	22.07	59.3	98.8	0.046	35,000	160,000	0.03	0.88
	mean	6.78	28.2	4.9	131.5	457	2.59	15.5	23,282	13.9	11.79	40.9	0.009	5292	75,583	0.02	0.47
N	min	6.35	24	3.5	79.29	191	1.2	4.4	17,330	10.42	2.2	11	0.022	200	300	0.01	0.11
	max	6.99	30.6	6.4	179.6	583	6.6	26.4	32,959	19.65	16.25	76	0.035	11,000	50,000	0.2	0.5
	mean	6.74	28.18	4.9	129.7	467	2.67	13.7	22,157	13.16	7.39	38.9	0.0015	2275	14,350	0.07	0.305
P	min	6.34	23.6	3.3	97.35	192	1.1	9.72	16,590	10.18	2.3	13	0.026	200	2300	0.01	0.03
	max	6.91	30.9	6.6	208.7	640	3.7	15.6	32,506	19.39	18.6	61	0.075	11,000	16,000	0.82	0.12
	mean	6.67	28.15	4.6	149.3	456	2.22	12.5	21,079	12.5	6.46	36.8	0.0013	2825	20,192	0.22	0.075
Q	min	6.32	23.2	2.9	97.34	192	1	6.8	80,43	5.08	3.2	6	0.034	800	3000	0.05	0.53
	max	7.02	30.6	4.5	185	544	4.2	13.1	29,890	17.8	40.7	57	0.07	17,000	160,000	0.12	0.53
	mean	6.65	28.1	3.8	156.3	411	2.48	9.52	16,290	9.76	14.06	34.1	0.001	4067	27,083	0.05	0.53

) and the appropriate equation to use is Equation (3), which is for x ≤ 5. The x stands for the value of the parameter. The subindices result of each parameter are then multiplied with Equation (2) to give each station its own WQI.

3. Results

3.1. Descriptive Statistics

Results of the 16 minimum (min), maximum (max) and mean of water quality parameters measured in the eight monitoring stations—B, D, E, G, J, N, P and Q—are presented in Table 1, Figure 2. The parameters assessed included pH, water temperature, DO, ORP, COD, BOD₅, salinity, EC, TDS, turbidity, TSS, NH₃-N, FC, TC, phosphate and NO₃⁻, each crucial for evaluating the health and quality of water bodies. Table 1 showed that observed pH values ranged from a slightly acidic 6.31 to a slightly alkaline 7.7. Water temperature measurements varied between 22.5 °C and 30.9 °C, and DO levels spanned from a low of 2.9 mg/L to a high of 8.4 mg/L. Significant variability was also noted in the ORP, with values ranging from 78.88 mg/L to 256.4 mg/L; COD and BOD₅, which are crucial indicators of organic pollution, showed considerable ranges from 169 mg/L to 696 mg/L, and from 0.3 mg/L to 7.9 mg/L, respectively. Salinity and EC values ranged from 4.2 to 39.5 PPT, and from 8043 to 43,459 μS/cm, respectively. Similarly, TDS results varied from 5.08 to 25.94 mg/L. Turbidity and TSS ranged from 2.2 NTU to 59.3 NTU, and from 4 mg/L to 98.8 mg/L, respectively. Nutrient levels, including NH₃-N, phosphate and NO₃⁻, had ranges of 0.02–0.18 mg/L, 0.01–0.82 mg/L and 0.03–0.9 mg/L, accordingly. The microbial contamination concentrations in this study indicated a wide range in FC (200 to 35,000 MPN/100 mL) and TC counts (300 to 160,000 MPN/100 mL).

3.2. Proposed Parameter Selection for Building Brunei River’s WQI

The suitability of the data for PCA was evaluated using the KMO measure of sampling adequacy and Bartlett’s test of sphericity. The KMO value was 0.713, while Bartlett’s test of sphericity was significant at p < 0.05, confirming that the 13 parameters were appropriate for PCA. Given the practical challenges and costs associated with incorporating all water quality parameters, PCA reduced the 13 selected parameters into five principal components (PC), accounting for 12 parameters using an eigenvalue greater than one (

Table 2. Presents the eigenvalues, total variance and cumulative variance of the first five principal components.

Total Variance Explained
Component	Initial Eigenvalues			Extraction Sums of Squared Loadings
	Total	% of Variance	Cumulative %	Total	% of Variance	Cumulative %
1	4.431	34.082	34.082	4.431	34.082	34.082
2	1.75	13.458	47.54	1.75	13.458	47.54
3	1.33	10.229	57.77	1.33	10.229	57.77
4	1.236	9.507	67.277	1.236	9.507	67.277
5	1.031	7.931	75.207	1.031	7.931	75.207
6	0.92	7.08	82.288
7	0.81	6.227	88.515
8	0.534	4.105	92.62
9	0.359	2.759	95.379
10	0.294	2.265	97.644
11	0.273	2.101	99.745
12	0.023	0.179	99.924
13	0.01	0.076	100
Extraction method: principal component analysis.

). PC 1 accounted for 34.082%, PC 2 (13.458%), PC 3 (10.229%), PC 4 (9.507%) and PC 5 (7.931%). These components capture most of the data’s variance (

Table 3. Principal component loadings of the 13 water quality parameters.

Parameter	1	2	3	4	5
pH	0.665	0.285	0.426	0.16	0.158
EC	0.971	0.048	0.052	−0.032	0.066
Salinity	0.95	0.048	0.003	−0.024	0.094
TDS	0.972	0.033	0.033	0.017	0.048
ORP	0.161	0.636	−0.346	0.033	−0.475
DO	−0.624	0.511	−0.024	−0.119	0.36
Turbidity	−0.129	0.008	0.44	0.647	−0.345
BOD5	−0.174	0.203	0.623	−0.196	0.26
TSS	0.205	−0.039	0.228	−0.707	−0.035
TC	0.009	0.458	0.394	0.188	−0.01
FC	0.205	0.094	−0.411	0.419	0.646
NH3-N	0.751	−0.401	−0.076	0.007	−0.124
T	0.297	0.758	−0.246	−0.157	−0.122

Note: factor loadings greater than ±0.5 are in bold.

).

Each PC represents a different aspect of the water quality parameters with the loadings (positive or negative) of each parameter’s contribution to the Brunei River. In this study, we focused on parameters with strong-to-moderate factor loadings greater than ±0.5. The retained parameters (Table 3) are: pH, EC, TDS, salinity, ORP, turbidity, BOD₅, TSS, FC, NH₃-N, T and DO. However, including all 12 parameters in the Brunei River’s WQI remains costly and impractical, especially considering the correlations among some parameters. For instance, TDS and salinity are highly correlated (

Table 4. Correlation matrix of the 12 retained water quality parameters.

	pH	EC	Salinity	TDS	ORP	DO	Turbidity	BOD5	FC	NH3-N	T	TSS
pH	1.00
EC	0.58	1.00
Salinity	0.63	0.98	1.00
TDS	0.54	0.98	0.98	1.00
ORP	0.04	0.06	0.04	0.05	1.00
DO	−0.23	−0.44	−0.42	−0.47	0.03	1.00
Turbidity	0.00	−0.22	−0.24	−0.18	0.07	−0.10	1.00
BOD5	0.04	−0.11	−0.13	−0.12	−0.11	0.24	0.03	1.00
FC	0.10	0.11	0.12	0.11	0.03	0.02	−0.04	−0.15	1.00
NH3-N	0.23	0.66	0.64	0.67	−0.12	−0.55	−0.03	−0.17	0.08	1.00
T	0.17	0.29	0.29	0.30	0.52	0.10	−0.08	0.00	0.05	0.00	1.00
TSS	0.15	0.18	0.12	0.10	−0.06	0.03	−0.18	0.09	−0.09	0.09	0.03	1.00

Note: Significant correlations in bold.

) and TDS can be derived from EC measurements. Thus, to avoid redundancy, the highly correlated parameters were not retained. Furthermore, despite the correlations between some parameters, such as pH and EC, they were included in the parameters selected for the Brunei River’s WQI due to their critical importance in assessing the river’s water quality.

The PCA biplot (Figure 3) shows the first two principal components (PC1 and PC2), which explains 47.54% of the total variance. The biplot shows that the further away the water quality parameters are from a PC origin, the more influence they have on that PC. For example, salinity, EC, TDS, T, ORP, DO and NH₃-N have strong influences on the first two components. The spread of red dots (scores) along the principal components illustrates significant variability among the samples.

3.3. Assessment of Water Quality across the Eight Stations Using Malaysia WQI and NSFWQI

The summarized results (

Table 5. Malaysia WQI values of the eight monitoring stations.

Station	Parameter Sub-Indices						WQI	Status
	DO	BOD5	COD	NH3-N	SS	pH
	(0.22)	(0.19)	(0.16)	(0.15)	(0.16)	(0.12)
B	89.21	89.66	21.82	99.82	78.05	99.37	79.54	Slightly Polluted
D	84.77	90.29	20.72	100.25	75.78	99.32	79.42	Slightly Polluted
E	78.86	96.17	13.3	100.28	76.31	99.16	76.85	Slightly Polluted
G	76.95	88.18	18.04	99.59	76.39	98.98	75.61	Slightly Polluted
J	68.46	79.67	18.19	99.56	76.05	93.82	71.47	Slightly Polluted
N	68.46	89.11	18.61	83.96	76.93	98.67	71.23	Slightly Polluted
P	63.31	91.01	18.16	100.36	77.79	98	73.4	Slightly Polluted
Q	47.58	84.33	16.28	100.39	79.12	98.11	68.58	Polluted

Note: WQI values are derived from the product of parameter weights and subindices, which are aggregated to represent overall water quality.

) from the Malaysia Water Quality Index indicate that most of the monitoring stations, specifically Stations B, D, E, G, J, N and P, were classified as slightly polluted, with WQI values ranging from 71.23 to 79.54. These stations showed a moderate presence of pollutants that could be addressed through targeted water quality management practices. Station Q, however, was classified as polluted with a WQI of 68.58, suggesting a higher level of contamination that may pose significant risks to the ecosystem.

Table 6. NSFWQI values of the eight monitoring stations.

Station				Q-Curve Sub-Indices						WQI	Category
	DO	T	BOD5	FC	TS	Turbidity	pH	NO3-	TP		Pollution
	(0.17)	(0.1)	(0.11)	(0.16)	(0.07)	(0.08)	(0.11)	(0.1)	(0.1)
B	87	12	70	19	83	82	88	98	90	67.58	Medium
D	75	12	80	18	83	78	87	96	90	65.85	Medium
E	62	12	67	14	82	75	87	95	90	63.8	Medium
G	60	13	64	16	82	80	80	97	94	64.14	Medium
J	55	12	68	14	81	72	78	96	90	61.83	Medium
N	50	13	65	18	82	85	78	97	94	62.55	Medium
P	50	12	90	16	81	83	75	99	94	64.52	Medium
Q	45	11	70	15	80	75	73	95	90	59.23	Medium

presents the NSFWQI values calculated for the eight monitoring stations, alongside the categorized pollution level. These values were derived from assessing the mean of nine water quality parameters, with each parameter assigned a specific weight reflecting its relative importance to overall water quality. The parameter subindices were calculated using the Q curves. The analysis revealed that all eight monitoring stations fell within the medium category concerning pollution levels, indicating a moderate degree of water quality pollution.

4. Discussion

4.1. Evaluation of the 16 Water Quality Parameters for the Eight Monitoring Stations

The observed, slightly acidic pH across the river’s monitoring stations was likely influenced by the oxidation of acid sulfate soils (ASS) surrounding the Brunei River [40,50]. Disturbances such as excavation or changes in water levels expose ASS to oxygen, causing the production of sulfuric acid and subsequent acidification of adjacent Brunei water bodies [51]. Such disturbances are noticeable around quarry operations at Station Q. Mean water temperature readings were relatively consistent, with minor variations possibly attributed to microclimatic conditions or differential solar insolation [52] across the sampling sites. Both temperature and pH were within the permissible USEPA limits for surface water [53].

Station P was mainly affected by nutrient pollution (phosphate) due to agricultural runoff from small paddy fields around the monitoring station [54]. It is well known that high levels of nutrient losses from agricultural fields, such as paddy fields, from which nutrients flow downstream, can cause eutrophication, including excessive algal blooms in the adjacent receiving water [54]. Prambudy [55] reported that the existence of waste-disposal activities added to the COD pollution load on the Cipager river, Cirebon, Indonesia. Similarly, the wastewater constantly being discharged into the Brunei River from Kampong Ayer could have increased the COD concentration, especially around Stations N, J and G. This is undesirable since the continual discharge of effluent [11] has impacted the receiving Brunei River waterbody to some extent, which may have negative effects on the quality of freshwater and ultimately cause harm to aquatic life, notably fish, if not controlled. A higher BOD₅ found at Station G was possibly from sewage discharges or agricultural runoff [56,57].

Station G pollution was attributed to urban and agricultural runoff from the capital city. Urban runoff is a common problem in densely populated areas like Brunei where the landscape is dominated by impervious surfaces that do not allow water to infiltrate into the ground [58]. Downstream stations B, D and E had higher EC, TDS and salinity concentrations, which decreased upstream due to dilution by freshwater inflows and lesser tidal effects [59,60]. The observed variations in EC may also be influenced by adjacent geological formations within the watershed [61]. For example, sandstone is made up of more inert elements that do not dissolve into ionic components when washed into the water, therefore, water that travels through places with sandstone tends to have lower electrical conductivity. Water that flows through clayey soils, on the other hand, typically has a higher EC due to ionized components. The concentration of ORP describes the health of the river, with higher values indicating greater health. Wetzel [62] stated that, in healthy waters, ORP should read between 300 to 500 mV, but in this study ORP read between 84–230 mV. A sufficient supply of DO is vital for all aquatic life. Most aquatic plants and animals need oxygen to live. For instance, fish typically cannot survive in water that has a dissolved oxygen content of less than 5 mg/L for an extended period [63]. The DO is generally low in the upstream stations due to anthropogenic activities such as sewage waste and agricultural and urban runoff, especially around Stations N, J, P and Q.

Turbidity and TSS are measures of water clarity impacted by particulate matter that could be of allochthonous or autochthonous origin, including soil erosion, algal growth, or resuspended sediments [64,65,66]. Moreover, the increased turbidity in Stations Q and J could impede light-penetration, thereby affecting the primary productivity of aquatic organisms and geochemical processes within the photic zone around these stations [11]. Stations D, E and G experienced raised TSS, possibly from erosion and urban runoff [67] from the capital city, Bandar Sei Begawan. The detection of NH₃-N, phosphate and NO₃⁻ in the Brunei River shows that the organic pollution could be attributed to agricultural runoff, wastewater treatment effluent, or other sources of organic waste [67,68,69,70,71]. Stations E, J, D and Q showed signs of moderate-to-high levels of nutrients which could lead to eutrophication [70].

Generally, TC and FC are relatively high at all stations and above the permissible Malaysia limit [29]. However, noticeable levels of FC and TC at Station E, near the Pintu Malim Sewage Treatment Plant (PMSTP), imply that inefficiencies in sewage processing or overflow incidents may be contributing to microbial pollution. In contrast, increased contamination at Station J in the Kampong Ayer area shows that the direct discharge of waste from residential areas is a primary source of the river’s water quality degradation. The situation is further exacerbated by Kampong Ayer’s population and lack of proper waste-management infrastructure. Meanwhile, despite its closeness to a quarry, Station Q exhibited higher FC and TC levels, which are more indicative of urban runoff.

4.2. Proposed Parameter Selection for Building Brunei WQI

In this study, PCA was employed to identify and select key water quality parameters for developing the WQI for the Brunei River. PCA effectively reduced the complexity of the dataset by transforming the original 13 water quality parameters into five principal components which collectively explained 75.207% of the total variance. Initially, 12 parameters (pH, EC, salinity, TDS, ORP, DO, turbidity, BOD₅, FC, NH₃-N, water temperature and TSS) were retained from the PCA. These parameters were further processed through correlation analysis and authors’ judgement to identify and exclude redundant parameters, thereby simplifying the dataset without significant loss of information.

EC, salinity and TDS were found to be correlated as they collectively represent the ionic content of water bodies. Although previous research indicated that TDS and EC correlation are not always linear, the concentration of TDS from EC can be used to give an overview of water quality [72]. Consequently, only EC was retained to represent the ionic content of the Brunei River, while TDS and SAL were excluded to avoid redundancy. ORP showed weak correlations with most parameters except water temperature, and is often measured alongside DO in polluted rivers [62]. However, in this study, the authors decided to exclude ORP and retain DO and water temperature as they are more vital for assessing the water quality of Brunei rivers [11].

The observed levels of NH₃-N in this study were not significantly high (Table 1), leading to its exclusion from the parameter list. Instead, using the Delphi method, we included nitrate and phosphate because they are key indicators of nutrient pollution and are particularly relevant for assessing eutrophication, a major concern for the Brunei River’s water quality. pH was retained due to its fundamental role in assessing the acidity or alkalinity of water bodies, considering the surrounding acid sulfate soils near the river [73]. Prior research [74] noted a linear correlation between turbidity and TSS, but our study showed a minimal correlation between them. Therefore, the authors decided to keep TSS and turbidity. BOD₅ and FC showed low correlations with other parameters, revealing their independent roles in assessing the Brunei River’s water quality. BOD is crucial for determining the level of organic pollution, while FC measures microbial contamination.

Based on our analysis and for future research and development of the Brunei River WQI, we recommend the following 10 parameters for inclusion: pH, EC, DO, turbidity, BOD₅, FC, water temperature, phosphate, NO₃⁻ and TSS.

4.3. Monitoring Stations’ Water Quality Using NSFWQI and Malaysia WQI

Overall, Stations B and D had the best WQI values for both Malaysia and NSFWQI assessments due to these locations being situated in areas with relatively low anthropogenic activity. The Malaysia WQI categorized Station Q as polluted and other Stations B, D, E, G, J, N and P as slightly polluted. Conversely, the NSFWQI categorized all stations as having medium pollution levels, though Station Q had the lowest WQI. In the Malaysia WQI and NSFWQI indices, Station Q had the lowest WQI, particularly due to the station’s lowest DO saturation, which carried the highest weights of 0.22 and 0.17, respectively. However, these indices do come with inherent limitations. The selection and weighting of parameters in these indices might not fully represent the specific water quality challenges of the Brunei River, potentially overlooking pollutants that are locally relevant but not included in the indices. NSFWQI offered a broader and better perspective on the Brunei River’s water quality by incorporating additional parameters such as temperature, fecal coliform and nutrients (nitrate and phosphorus) compared to the Malaysia WQI. Incorporating these additional parameters clearly influenced the WQI, moving it from slightly polluted to a moderately polluted status. The comparative analysis of WQI and statistical assessments showed that upstream stations are more polluted than downstream stations, given that low WQI scores were mainly found in the upstream stations, especially stations around Kampong Ayer (N and J) and the quarry area (Q). The Kampong Ayer Water Village, with its dense population and likely direct discharge of domestic wastes into the water and the quarry area, with potential sediment and pollutant runoff, are environments where the water quality is severely compromised.

5. Conclusions

This research assessed the pollution level and proposed parameters for building Brunei River’s Water Quality Index. Sixteen water quality parameters in eight monitoring stations along the Brunei River were analyzed using both statistics and two established water quality indices. We used a combination of descriptive statistics, interval plots, NSFWQI and Malaysia WQI to assess pollution levels in the river. The NSFWQI classified the Brunei River as medium polluted, whereas the Malaysia WQI rated most of the monitored stations as slightly polluted except for Station Q, which was deemed polluted. NSFWQI and Malaysia WQI identified pollution hotspots as Station Q (Quarry area) and Stations N and J (Kampong Ayer). Addressing these hotspots requires waste-management strategies, regulatory frameworks and continuous monitoring to safeguard the water quality of the Brunei River. Furthermore, the implementation of stringent pollution control measures, coupled with community awareness programs for Kampong Ayer residents and urban planning that prioritizes green infrastructures, could substantially mitigate the identified risks.

Statistical analysis revealed that sources of pollution of the Brunei River are mainly attributed to anthropogenic activities such as mining, domestic waste and agricultural and urban runoff around the river.

The combination of principal component analysis, correlation analysis and expert judgment ensures that the WQI parameters selected remains comprehensive and relevant to local water quality management needs. This research enhances the understanding of the current state of the Brunei River by objectively selecting key parameters and reducing subjectivity in the index development. The selected 10 parameters: dissolved oxygen, water temperature, electrical conductivity, phosphate, nitrate, pH, total suspended solids, turbidity, biological oxygen demand and fecal coliform form a reliable basis for assessing the water quality of the river. The findings of this research support effective water quality management in Brunei Darussalam.