Trihalomethane Formation Potential at the Barekese Water Treatment Plant and the Related Cancer Risk to Consumers in the Kumasi Metropolis of Ghana

Abstract

The prevalence of disinfection by-products in drinking water supplies is a global concern due to their carcinogenicity. However, the monitoring of DBPs such as trihalomethanes (THMs) and haloacetic acids (HAAs) in drinking water supplies is non-existent in many developing Asian, South American, and African countries. The formation of THMs during disinfection arises from a reaction between the disinfectant and natural organic matter in the water, particularly, dissolved organic carbon (DOC). This reaction is hastened by increases in temperature, high levels of disinfectant doses or residual, elevated water pH, long disinfection contact times, and high DOC concentrations. However, the inclusion of a granular activated carbon adsorption process in the water treatment process is the most effective method for the removal of the main precursor (DOC) for the formation of THMs in treated water. The Barekese WTP, which disinfects with chlorine, has no adsorption process for DOC removal, and supplies over 80% of pipe-borne water to the city of Kumasi in Ghana, was assessed for the THM formation potential (THMFP). A THM predictive model was used to determine the potential THM concentration in the final water. The THMFP at the Barekese WTP ranged between 22.42 and 38.94 µg/L, which was below the 100 µg/L threshold set by the WHO. The lifetime average daily doses were 3.9494 × 10⁻⁴ µg/Kg/d and 3.9294 × 10⁻⁴ µg/Kg/d for male and female consumers, respectively. The lifetime integrative cancer risks associated with consumption of the water were 1.817 × 10⁻⁵ and 1.808 × 10⁻⁵ for males and females, respectively. The cancer risk posed was acceptably low. However, direct measurement of DBPs is required to corroborate these findings and verify the cancer risk posed to the consumers of treated water from the Barekese WTP to inform policies, regulations, public health interventions, and investment.

1. Introduction

Access to safe drinking water is paramount to human health and well-being. Hence, the provision of wholesome water for domestic use remains a priority for many nations, institutions, and organizations throughout the world, as enshrined in SDG 6 of the United Nations’s Sustainable Development Goals [1]. The consumption of unsafe drinking water is one of the world’s largest causes of morbidity and mortality, particularly among children in developing countries [2]. Children under five years of age in several Sub-Saharan African territories experience diarrhea due to unsafe drinking water and limited access to sanitation facilities [3,4]. It is therefore imperative that water for consumption is treated to make it free from pathogenic microorganisms, improve its physicochemical quality, and sufficiently reduce the disease burden arising from the use of unwholesome water in domestic setting. Conventional treatment processes for drinking water involve coagulation, flocculation, and sedimentation to remove settleable suspended solids (turbidity), filtration to remove fine suspended solids, followed by disinfection to kill or inactivate bacteria and other microorganisms.

The disinfection step in drinking water treatment, which is critical for the deactivation of pathogenic microorganisms that can cause various waterborne diseases [5], is achieved though the addition of disinfectants. These disinfectants are mostly oxidants such as chlorine, chloramine, chlorine dioxide (ClO₂), and ozone [6]. Chlorine is the most widely used disinfectant due to its relatively lower cost and capacity to produce large volumes of ClO₂ for residual protection in the water distribution system [7]. However, microbial disinfectants such as chlorine and chloramine can also unintentionally react with natural organic matter (NOM) that may be present in pre-disinfected water to produce disinfection by-products (DBPs) [8]. The formation of DBPs in drinking water has received lots of attention from researchers and public health scientists worldwide due to their carcinogenic effects on human health and has led to the enactment of regulatory requirements on water treatment companies to monitor and report the concentrations of certain DBPs. The authors of [9] reported that the consumption of water containing DBPs in excess of regulatory thresholds for a long time may increase human health risks, with the most consistent association with bladder and kidney cancer. DBPs can be broadly categorized as aliphatic DBPs (haloacetic acids (HAAs), trihalomethanes (THMs), haloacetonitriles, halonitromethanes, and haloketones), or aromatic DBPs (halophenols, halonitrophenols, halohydroxybenzaldehyde, halohydroxybenzoic acid, halobenzoquinones, and haloanilin) with varying levels of cytotoxicity and carcinogenicity [10]. A meta-analysis for European males showed that bladder cancer was 47% more prevalent among those consuming water with THM concentrations greater than 50 μg/L compared to those consuming water with THM concentrations less than 5 μg/L [11]. Parvez et al. [12] also found a high reproductive health risk to be associated with the consumption of water containing high levels of HAAs.

Due to the grievous harm to human health caused by the consumption of DBPs from chlorine-disinfected water, some water treatment companies have switched from chlorine-based disinfectants to other disinfecting technologies, such as ozonation, ultraviolet radiation, and reverse osmosis to kill or remove pathogens from drinking water, to avoid the formation of DBPs [13]. However, the majority of WTPs include an adsorption process, such as filtration through a granular activated carbon (GAC) bed to eliminate DOC and NOM, which are the main precursors for the formation of DBPs (and contribute to taste, color, and odor), because it the most effective and economically feasible approach to control DBPs by WTPs [14,15,16].

The Barekese Water Treatment Works (BWTW) is the largest conventional water treatment plant in the Ashanti region of Ghana and it supplies about 80% of public pipe-borne water to residents of the Kumasi Metropolis and its environs [17]. The water treatment processes at the BWTW include coagulation, flocculation, sedimentation filtration, and chlorine-based disinfection. The process has no dedicated technology for the removal of DOC, despite extensive studies outlining the benefits of incorporating a GAC adsorption process in the treatment of water for drinking purposes. The GAC adsorption process in water treatment is effective in removing DOC (the precursor for DBP formation) when compared to a wide range of media types due to its large specific surface area, well-developed pore structure, and exceptional adsorption capabilities [18]. It is a cost-effective technique for removing significant fractions of organic precursors prior to water disinfection [19,20,21] through adsorption and biological activity [22], reduces the chlorine demand and water corrosion potential, and it removes chemicals that produce odors or tastes in water, such as hydrogen sulfide [23]. Nonetheless, significant fractions of NOM (typically involving DOC) may escape removal in these treatment stages at the Barekese WTP and react with the disinfection agent to form sufficiently high levels of disinfection by-products [24].

Despite the overwhelming evidence that disinfection by-products (DPBs) such as trihalomethanes (THMs) are carcinogenic, many developing countries do not monitor their prevalence and concentrations in chlorine-disinfected drinking water. A recent global review of THM concentrations in various countries found no evidence of monitoring or published data by water regulatory agencies, researchers in academia, government, or private research institutions in many African and South American countries [25]. In fact, 48 out of 54 African countries have no data on THM levels in their drinking water, although many of the water treatment plants on the continent employ chlorine gas or hypochlorite compounds as the disinfection agent. In the West African block, Nigeria is the only country out of the seventeen nations to have ever reported THM levels in drinking water supplies, and even then, the levels were greater than the 100 µg/L limit set by the World Health Organization (WHO) [25]. This lack of information on THMs is very concerning and makes it difficult to assess the human health risk they pose to the billions of people drinking chlorine-disinfected water in Africa, including Ghana. The absence of data on DBPs may be due to unavailable research expertise in the analysis of DBPs, nonexistence of instrumentation and equipment for DBP analyses, no financial support to investigate DBPs, or other multiple reasons.

In this study, researchers used measurements of COD as a surrogate for DOC [26], temperature, pH, chlorine residual, and chlorination contact time to compute the THM formation potential [27] in the treated water at the Barekese WTP to provide some data for the assessment of human health risk while innovatively overcoming the numerous barriers of quantifying regulated THMs, namely, trichloromethane (TCM), dibromochloromethane (DBCM), bromo-dichloromethane (BDCM), and tribromomethane (TBM) [28,29]. Although these values may be underestimated or could be misleading due to the lack of validation through direct THMs measurements, it provides a means to conduct health risk assessments, highlights the lack of empirical data, and prompts the need for action in the monitoring and reporting of DBPs in chlorine-disinfected water in developing countries.

2. Materials and Methods

2.1. Study Area

The Barekese WTP stands as a prominent conventional water treatment facility situated in the northern part of the Ashanti region of Ghana (Figure 1). The plant derives its water supply from the Barekese Reservoir, a water source endowed with a remarkable storage capacity capable of meeting nearly 80 percent of the drinking water demands for Kumasi City and its adjoining areas. The reservoir is geographically situated at approximately 19 km northwest of Kumasi, between 6°51′11″ N, 10°42′10″ W and 6°50″ N, 10°39′88″ W, along the Offin River in the region (Figure 1). This important water body spans an expansive surface area of roughly 6.4 km², characterized by an average depth of 33 m, while its catchment area extends over 565 km². Notably, the study area has a semi-humid tropical climate with two distinct rainy seasons: the main season from April to July and the minor season from September to October. The mean annual rainfall in the area is about 1368 mm. The temperature ranges between 21.1 and 31.5 °C.

2.2. Barekese Water Treatment Plant Layout and Processes

The Barekese water treatment process (Figure 2) includes raw water from a 220,000 m³ impounding reservoir [17]. The raw water emerges at the head of the works, where it receives a dose of aluminum chlorohydrate coagulant. At the head of works, the raw water flows down a cascading step where it is aerated to remove volatile organic compounds and odors, while enabling mixing of the raw water and the dosed coagulant. The dosed water arrives at the inner circle of the four sedimentation tanks, where it moves outward to the middle circle for flocculation and to the outer circle for sedimentation. The settled water is combined in the blending chamber and redistributed to sixteen (16) rapid gravity filters (RGFs) containing fine and coarse sand and gravel. The combined filtered water is dosed with chlorine gas or sodium hypochlorite and flows into the partially buried 9000 m³ contact tank for disinfection. The final water is pumped from the contact tank into the Kumasi Metropolitan Area.

2.3. Sampling

Grab samples were collected from the raw water sampling point for color analysis, filtered (pre-disinfection) water was sampled for analysis of the pH, temperature, and COD, whereas final treated water was collected for residual chlorine analysis following established sampling protocols to maintain sample integrity and minimize contamination [30]. Sampling was conducted in the morning (between 8 a.m. and 9 a.m.) and in the afternoon (between 2 and 4 p.m.). Samples were stored in cool boxes and transported to the onsite laboratory for physicochemical analysis. Portions of the pre-disinfection water samples were sent to the Water Quality Laboratory at the Kwame Nkrumah University of Science and Technology, Kumasi, on the same day for further analyses. A total of 36 samples were collected in the first week of January 2023 to capture the coolest possible water temperatures, which occur from November to January in Ghana.

2.4. Analysis of Water Quality Parameters Relevant for THM Formation Potential

For this study to correctly determine the trihalomethane formation potential at the Barekese WTP, disinfection contact time as an operational parameter and water quality parameters such as temperature, DOC concentration, and pH of the pre-disinfected (filtered) water and residual chlorine of the final water were of interest.

2.4.1. pH

Generally, as pH increases from 5.5 to 8.5, THM concentrations increase [31,32]. The formation of DBPs in chloramine-treated water supply systems was also found to have a positive correlation with water pH [33]. pH measurements of the pre-disinfection samples were performed using an HI 2216 pH/ORP/ISE instrument.

2.4.2. Temperature

The THM concentration typically increases as temperature also increases, indicating a stronger temperature dependency of THM formation [34,35]. This is because the rate of chlorine reaction with organic precursors increases with temperature [36]. The temperature of the pre-disinfection samples was measured during sampling using an HI 2216 pH/ORP/ISE instrument (Hanna Instruments, Bangkok, Thailand).

2.4.3. Dissolved Organic Carbon (DOC) Concentration and Color

DOC is the fraction of total organic compounds in a water sample that can pass through a 0.45 μm membrane filter. Generally, these low-molecular-weight aromatic compounds cannot be removed by conventional water treatment processes such as coagulation, sedimentation, and filtration. More than half of the DOC in surface water systems is composed of humic substances, which include humic acids, fulvic acids, phenolic compounds, and other carbon compounds with conjugated double bonds [37]. Humic compounds have an intricate polymeric form and are often generated in water bodies from the decomposition of organisms and various biochemical reactions [38,39]. Both humic and fluvic acids act as precursors for the formation of THMs [40], although humic acids react faster and can produce more DBPs than fluvic acids. Understanding the existence and properties of these precursors is critical for optimizing water treatment operations to reduce the production of potentially dangerous DBPs in drinking water [41]. The DOC concentration in water can have a major influence on THM formation [42]. Evaluation of the effects of DOC concentrations on THM formation [43,44] has shown that THM concentrations increase with increasing DOC concentration in water [45,46,47]. The most frequently used surrogate parameter for assessing NOM levels in water is DOC [42,43,48]. In this study, direct measurement of DOC in the filtered water was impossible due to the lack of an appropriate spectrophotometer. Instead, chemical oxygen demand (COD) of the pre-disinfected water was determined by adding a 2.5 mL aliquot of the sample to the contents of a low range COD cell test kit from HACH, followed by photometric analysis according to the manufacturer’s procedure. The analysis involved the conversion of the total organic carbon in the water sample through oxidation with acidified potassium dichromate (K₂Cr₂O₇) to carbon dioxide, thus allowing COD to be used as a good comparator for the DOC content where direct measurement of dissolved organic carbon content cannot be performed [26,49]. The color of the raw and pre-disinfection water samples, which is also symptomatic of the DOC content, was measured using a HACH LICO 620 colorimeter.

2.4.4. Residual Chlorine Concentration

High residual chlorine levels are linked to increased THM formation [50,51], although it offers protection against microbial resurgence in the distribution network. A colorimetric technique was used to quantify residual chlorine in the treated final water samples following the manufacturer’s protocol. In the method employed, an indicator containing N, N-diethyl-p-phenylenediamine (DPD) was added to the sample, which reacted rapidly with the residual chlorine to generate a pink color for colorimetric determination of the chlorine concentration using a photometer.

2.4.5. Contact Time

According to studies, THMs can continue to form in drinking water over time if precursors and residual chlorine are present [35]. Similarly, the authors of [52] demonstrated that concentrations of THMs and HAAs increased with increasing reaction time.

2.5. Data Analysis

The theoretical contact time of the 9000 m³ contact tank was calculated using the daily flow measured at the intake, since there was no other upstream flowmeter. Due to the varied intake pump speed settings between 1220 and 1341 rpm, the flow rate arising from the operations at the abstracted point exhibited daily fluctuations. The contact time was calculated using Equation (1) and assuming that the hydraulic efficiency of the tank was 1.

$$\mathbf{Contact}\mathbf{time}(\mathbf{t})(\mathbf{mins})={\displaystyle \frac{\mathbf{V}\mathbf{o}\mathbf{l}.\mathbf{o}\mathbf{f}\mathbf{h}\mathbf{o}\mathbf{l}\mathbf{d}\mathbf{i}\mathbf{n}\mathbf{g}\mathbf{t}\mathbf{a}\mathbf{n}\mathbf{k}\left(\mathbf{L}\right)\times \mathbf{h}\mathbf{y}\mathbf{d}\mathbf{r}\mathbf{a}\mathbf{u}\mathbf{l}\mathbf{i}\mathbf{c}\mathbf{e}\mathbf{f}\mathbf{f}\mathbf{i}\mathbf{c}\mathbf{i}\mathbf{e}\mathbf{n}\mathbf{c}\mathbf{y}}{\mathbf{F}\mathbf{l}\mathbf{o}\mathbf{w}\mathbf{r}\mathbf{a}\mathbf{t}\mathbf{e}\mathbf{o}\mathbf{f}\mathbf{w}\mathbf{a}\mathbf{t}\mathbf{e}\mathbf{r}\mathbf{i}\mathbf{n}\mathbf{t}\mathbf{o}\mathbf{h}\mathbf{o}\mathbf{l}\mathbf{d}\mathbf{i}\mathbf{n}\mathbf{g}\mathbf{t}\mathbf{a}\mathbf{n}\mathbf{k}\left(\mathbf{L}/\mathbf{m}\mathbf{i}\mathbf{n}\right)}}$$

(1)

Simple linear regression analysis was also performed using Microsoft Excel 2019 to find the correlations between the final WQPs and THM formation potential. Principal component analysis (PCA) was employed to explore complex relationships between THMFP and the WQPs using XLSTAT version 2022. To elucidate the relationships between physicochemical parameters and THM formation, Pearson’s correlation matrix analysis was applied.

2.5.1. Conversion of Chemical Oxygen Demand to Dissolved Organic Carbon Concentration

In this study, researchers lacked the instrumentation to directly determine the dissolved organic carbon (DOC) concentration in the pre-disinfection water samples. Instead, the chemical oxygen demand (COD) was measured using COD cell test kit from VWR (UK). This study used an indirect approach to estimate the DOC content in the sample by calculation using the COD-to-DOC conversion equation derived by [26], as shown in Equation (2):

$$\mathbf{D}\mathbf{O}\mathbf{C}={\displaystyle \frac{\mathbf{C}\mathbf{O}\mathbf{D}+\mathbf{3.338}}{\mathbf{5.292}}}$$

(2)

2.5.2. Determination of Trihalomethane Formation Potential (THMFP)

The study also applied a DBP predictive model developed by [27] to estimate the DBP formation potential depending on the final WQPs described in the sample analysis. The model allows a good estimation of the total trihalomethane (THM) concentration (µg/L) in drinking water using Equation (3), with an R² value of 0.90:

$$\mathbf{T}\mathbf{H}\mathbf{M}\mathbf{s}=\mathbf{4.4}\times {\mathbf{10}}^{-\mathbf{2}}\times {\mathbf{D}\mathbf{O}\mathbf{C}}^{\mathbf{1.030}}\times {\mathbf{C}\mathbf{t}}^{\mathbf{0.262}}\times {\mathbf{p}\mathbf{H}}^{\mathbf{1.149}}\times {{\mathbf{R}}_{\mathbf{C}\mathbf{l}}}^{\mathbf{0.277}}\times {\mathbf{T}}^{\mathbf{0.968}}$$

(3)

where DOC and R_Cl are the concentrations of dissolved organic carbon (mg/L) of the pre-disinfected water and the residual chlorine (mg/L) in the final water, respectively; T is the temperature of the final water in °C; and Ct is the disinfection contact time in hours.

2.5.3. Human Health Risk Analysis

The consumers of chlorine-disinfected water are exposed to THMs through oral ingestion, dermal contact, inhalation during drinking of the water, bathing, swimming, washing, and showering [25]. In this study, the predicted concentration of THMs in the drinking water produced at the Barekese WTP was used to estimate the lifetime average daily dose (LADD) in mg/kg-day via ingestion only, since the inhalation and dermal absorption of THMs is significant when consumers take warm baths from pipe-fed boilers, which is not the case in Ghana. Equation (4) was used to calculate the LADD via ingestion [53], as follows:

$${\mathbf{L}\mathbf{A}\mathbf{D}\mathbf{D}}_{\mathbf{i}\mathbf{n}\mathbf{g}\mathbf{e}\mathbf{s}\mathbf{t}\mathbf{i}\mathbf{o}\mathbf{n}}={\displaystyle \frac{{\mathbf{T}\mathbf{H}\mathbf{M}}_{\mathbf{s}}\times \mathbf{I}\mathbf{R}\times \mathbf{E}\mathbf{F}\times \mathbf{E}\mathbf{D}\times \mathbf{C}\mathbf{F}}{\mathbf{A}\mathbf{B}\mathbf{M}\times \mathbf{A}\mathbf{T}}}$$

(4)

In Equation (4), IR is the ingestion rate of treated water in liters per day, EF is the exposure frequency in days per year, ED is the exposure duration in years, ABM is average body mass in Kg, and AT is an average lifetime in days (

Table 1. Values used in the calculation of lifetime average daily dose and lifetime integrative cancer risk of THMs in Ghanaians.

Variables	Notation	Value	Reference
Average body mass of an adult in Ghana (Kg)	ABM	Male: 72.43	[54]
		Female: 68.12
Ingestion Rate (L/day)	IR	2	[55]
Exposure frequency (days/year)	EF	365
Exposure duration (days/year)	ED	365	[56]
Average lifetime in Ghana (days)	AT	Male: 23,170.2 (63.48 years)	[57]
		Female: 24,761.6 (67.84 years)
Age-dependent adjustment factor	ADAF	<2 year = 10	[58]
		2 to 16 years = 3
		>16 years = 1
Average carcinogenic slope factor for THM4 (TCM: 0.031; BDCM: 0.062; TBM: 0.0079; CDBM: 0.084) via oral/dermal ingestion	SF	0.046	[59]

). The value obtained is multiplied by a conversion factor (CF) of 0.001 to change the units of measurement from μg/Kg to mg/Kg [53].

The lifestage integrative cancer risk (LICR) from the total amount of LADD through ingestion was calculated using Equation (5), as follows:

$$\mathbf{L}\mathbf{I}\mathbf{C}\mathbf{R}={\sum }_{\mathbf{n}=\mathbf{1}}^{\mathbf{\infty }}\left({\mathbf{L}\mathbf{A}\mathbf{D}\mathbf{D}}_{\mathbf{i}\mathbf{j}}\times {\mathbf{S}\mathbf{F}}_{\mathbf{i}\mathbf{j}}\times \mathbf{A}\mathbf{D}\mathbf{A}\mathbf{F}\right)$$

(5)

where i represents different THM species, and j represents different exposure routes. SF is the cancer slope factor of each THM species for a special route (mg/kg-day), and ADAF is the age-dependent adjustment factor to the SF for a certain lifestage (Table 1).

2.6. Limitations of the Study

This study did not incorporate the direct measurement of THMs concentration to cor-roborate the predicted possible concentrations under the given treatment conditions and water parameters at the Barekese Water Treatment Plant. This lack of validation of the results could be misleading to a higher or lower health risk. The predicted concen-trations of THMs although useful, must be used with caution until actual measure-ments are published. The second limitation of this research is the determination of DOC concentration, which was also not obtained from a direct measurement, but from the COD concentration. Although COD relates directly to DOC as demonstrated in Section 2.5.1, it is worth noting that this conversion could lead to underestimation of DOC which will in turn diminish the predicted THMs potential.

3. Results and Discussion

The Barekese WTP daily flow rates into the contact tank showed significant variability, with values between 70,974 and 79,404 L/min in the six days of sampling. These fluctuations in flow rate were due to intake pump adjustments to meet consumer demand. Consequently, the water contact time, which reflects the duration of exposure between chlorine and microorganisms (including the organic precursors for THMs), also varied over the study period, ranging from 95 to 101 min. The water quality parameters of raw, filtered, and final water samples obtained from the BWTP during the research period are shown in

Table 2. Water quality parameters and THM formation potential at the Barekese WTP.

	Raw Water	Filtered (Pre-Disinfection) Water				Final (Post-Disinfection) Treated Water
Day	Color (PCU)	pH	Temp. (°C)	COD (mg/L)	DOC (mg/L)	R. Cl2 (mg/L)	Contact Time (t) (minutes)	Flow Rate (L/min)	THMs (µg/L)
1	240	7.03	25.80	5.00	1.58	2.50	95.34	79,403	22.42
2	300	7.13	26.10	10	2.52	2.60	100.73	75,150	38.23
3	260	7.08	25.70	8.00	2.14	1.90	106.66	70,974	29.38
4	250	7.00	25.20	6.0	1.76	2.00	100.68	75,189	23.24
5	300	7.10	26.10	11.00	2.71	2.10	103.61	73,062	38.94
6	300	7.20	25.00	8.00	2.14	2.05	101.26	74,756	29.38
Mean	275	7.09	25.65	8.00	2.142	2.192	101.38	74,755.7	30.265
Min.	240	7.00	25.00	5	1.58	1.9	95.34	70,974	22.42
Max.	300	7.2	26.1	11	2.71	2.6	106.66	79,403	38.94
STDev	25.66	0.072	0.459	2.28	0.431	0.287	3.749	2553.1	7.087

Notes: STDev = standard deviation; R. Cl₂ = residual chlorine.

. The pH of the pre-disinfected water ranged from 7.0 to 7.2, while the temperature ranged from 25.0 to 26.1 °C. This narrow range of pH values suggested that the raw water quality did not change significantly during the week of sampling and, therefore, the chemical doses that may influence the pH were relatively constant. The residual chlorine levels measured in the final water were considerably high, ranging between 1.90 and 2.60 mg/L, as compared to the 0.2 to 0.7 mg/L recommended limits set by the Ghana Standard Authority [60]. Such high residual chlorine in the distribution network will provide a considerable level of protection against the resurgence of pathogenic microorganisms in the treated water but increases the risk of THM formation in the water supply network. The COD values ranged between 5 and 11 mg/L and the corresponding DOC concentrations were between 1.58 to 2.52 mg/L. The color of pre-disinfected (filtered) water, which is an indicator of the DOC content, particularly the humic acid concentration, was very low at 5 PCU. The COD and DOC levels in the Barekese filtered water were significantly low. However, these DOC levels were significant for interaction with the dosed chlorine to form DBPs [43].

3.1. Trihalomethane Formation Potential (THMFP)

The total trihalomethane formation potential, expressed as microgram per liter of treated final water, at the Barekese WTP ranged from 22.42 to 38.94 µg/L (Table 2), with a mean concentration of 30.27 µg/L. This range of predicted THM concentrations was lower than the maximum contaminant level (MCL) recommended by the World Health Organization [28] and United States Environmental Protection Agency (USEPA) for drinking water quality (100 µg/L and 80 µg/L, respectively [25]) and findings of studies in other tropical climates [61].

Although these findings are not alarming, it is essential that direct measurement of the concentrations of THM compounds is performed in drinking water; thus, TCM, DBCM, DCBM, and TBM should be assessed for their individual concentrations at the Barekese WTP and all other WTPs in Ghana to determine the actual levels of THMs in treated drinking water across the country. The actual THM concentrations will provide the data required for public health risk assessment, policy decisions, and regulatory oversight of the drinking water sector, with regard to the formation of these carcinogenic compounds in drinking water treatment processes. The low predicted THM concentrations could stem from the Barekese water treatment plant’s efficiency in removing the precursors for THM formation, leading to relatively low DOC concentrations in the water, and other factors such as the category of natural organic compounds that might be in the water and the unknown initial dose of chlorine added for disinfection [7,62,63].

3.2. Correlations Between THM Formation Potential and Water Quality Parameters

The WQPs with the most influence on THM formation potential were DOC > pH > temperature > residual chlorine > contact time, as seen by the R² values from the linear correlations (Figure 3, panels A–E), the aligning of DOC with THMs in the PCA (Figure 3, panel F and

Table 3. Principal component analysis of the water quality parameters.

	PC1	PC2	PC3	PC4	PC5
Eigenvalue	3.681	2.092	1.014	0.154	0.058
Variability (%)	52.589	29.881	14.491	2.204	0.836
Cumulative %	52.589	82.470	96.961	99.164	100.000

), and the correlation coefficient between THMs and DOC (

Table 4. Correlation analysis of water quality parameters.

Variables	pH	Temp	Contact Time	R. Cl2	COD	DOC
pH
Temperature	−0.110
Contact time	0.265	0.054
R. Cl2	0.015	0.535	−0.692
COD	0.576	0.516	0.583	0.000
DOC	0.576	0.520	0.579	0.005	1.000
THMs	0.560	0.615	0.466	0.174	0.984	0.985

Note: Values in bold are different from 0 with a significance level at alpha = 0.05; DOC = dissolved organic compounds; COD = chemical oxygen demand; R.Cl₂ = residual chlorine.

). This comparison of the influence of different WQPs on THM formation potential could be used to prioritize investments into improving processes at WTPs to mitigate the formation of THMs. The magnitude of the each WQP’s contribution to THM formation could also be used as a guide for decision making and trade-offs in water treatment operations. The data revealed that pH has a moderate influence on THM formation, as observed from the linear relationship between pH and THFP (Figure 3, panel A). For instance, the THM formation potential in the final water was higher at pH 7.20 (29.38 µg/L). The residual chlorine levels ranged from 1.9 to 2.6 mg/L and exhibited a discernible impact on THM concentrations such that the water sample with a relatively higher residual chlorine content (2.6 mg/L) correspondingly showed an elevated THM concentration (38.23 µg/L), and the one with a lower residual chlorine level (1.9 mg/L) showed a comparatively lower THM concentration (29.38 μg/L). A time-dependent formation trend of THMFP for chlorinated water samples of the Barekese WTP was also seen. The concentration of THMs gradually increased from 22.42 to 38.94 µg/L when the contact time changed from 95.34 min to 103. 61 min. The critical temperature at which THMFP was more pronounced was 25.0 to 26.1 °C.

3.3. Principal Component Analysis (PCA) of Parameters Influencing THMFP

Principal component analysis (PCA) was utilized in this study to understand the relationships and dependencies between water quality indicators and THM formation potential (THMFP). Eigenvalues, scree plots, and ordination plots were used to identify the dominant principal components and to interpret the contribution of each parameter to these components. The explained variance and patterns observed within the reduced dimensional space provided insight into the parameter interactions and their influence on THM formation.

3.4. Correlations Between THMFP and Water Quality Parameters

The Pearson correlation matrix, as shown in Table 4, elucidated the strength of the correlations between THMFP and the physicochemical parameters. The correlation between THMFP and residual chlorine in the present study was marginally positive but statistically significant (r = 0.17). This was attributable to the minor differences in residual chlorine concentrations among water samples obtained from the BWTP. Studies have revealed that other variables, like chlorine dose and chlorine decay, may have an impact on the relationship between residual chlorine and THM formation [7,30,63]. However, there were moderate positive correlations between THMFP and pH (r = 0.60), temperature (r = 0.62), and contact time (r = 0.50). Thus, the study demonstrated that these physical water quality factors can enhance THM formation at the BWTP. This was supported in the literature by other studies reporting that such operational and environmental factors can affect THM formation [30,64,65]. Additionally, a correlation coefficient of 0.99 was obtained for THMFP and DOC, which indicated a strong positive relationship between these parameters. As a result, the current study found that increasing DOC concentration resulted in an increase in total THMs. This indicated that the organic matter present in the water contributed significantly to THM formation.

3.5. Predicted Impact of Climate Change on THM Formation Potential at the Barekese WTP

The potential THM concentration in the warmer season was 31.37 µg/L. It is important to acknowledge that the current study was conducted during a season when water temperatures were at their lowest. Notably, during the warmer season, average water temperatures can reach 26.70 °C. Consequently, this elevated temperature profile corresponded to a projected potential THM concentration of 31.37 µg/L, which is considerably lower than the maximum level of contamination established in the WHO recommendations for drinking water quality. This dynamic underlines the intricate relationship between seasonal variations, temperature influences, and THM formation. However, the projected temperature rise due to global warming may be expected to increase the THM formation potential at the Barekese WTP above than the levels found in this study, unless the water treatment process is improved.

3.6. Estimated Health Risk from the Predicted THM Concentration at the Barekese WTP

The predicted LADD via ingestion was 3.95 × 10⁻⁴ and 3.93 × 10⁻⁴ mg/Kg/d for males and females, respectively. These values were significantly lower (approx. 58% less) than the findings of [61] in an Ethiopian study. The low LADD_ing was because the method used in this study to quantify the concentration of total THMs was a conservative approach, which could lead to an underestimation. The LICR posed to consumers of the Barekese WTP through the ingestion of treated water was 1.816 × 10⁻⁵ and 1.808 × 10⁻⁵ for males and females, respectively (

Table 5. Lifetime average daily dose and lifestage integrative cancer risk of THMs in water from the Barekese WTP.

Estimates	Notation	Value	Reference
Mean THM concentration (μg/L)	THMs	30.265	This study
		76.31	[61]
Lifetime Average Daily Dose (ingestion) (mg/Kg/d)	LADDingestion	Male: 3.949 × 10−4	This study
		Female: 3.929 × 10−4
Lifestage Integrative Cancer Risk (for over 16-year-old)	LICR	Male: 1.816 × 10−5	This study
		Female: 1.808 × 10−5

), which were rated as acceptable low risk. According to [61], cancer risk (Rc) is defined as negligible if Rc < 10⁻⁶; acceptable low risk if 1 × 10⁻⁶ ≤ Rc < 5.1 × 10⁻⁵; acceptable high risk if 5.1 × 10⁻⁵ ≤ Rc < 10⁻⁴; and unacceptable risk if Rc ≥ 10⁻⁴. The level of cancer risk found from oral ingestion of the Barekese treated water may be low, but the cumulative risk from inhalation and dermal absorption may significantly higher. In this study, LADD via inhalation and dermal absorption was excluded because THMs may enter the body through these channels during showering with hot water from a boiler that is fed by pipe-borne water. As this scenario does not apply to most of the consumers in the Kumasi Metropolis, the cancer risks due to inhalation and dermal absorption were not calculated. It is also worth noting that, although the THM formation potential is an estimate of the collective concentration of individual compounds, each of these halo-organic compounds have varying carcinogenicity [61]; hence, the WHO sets the limits for each of these compounds to reflect its potency. For instance, the recommended limits for chloroform, dibromochloromethane, dichloroacetate, and monochloroacetate in drinking water are 300, 100, 50, and 20 µg/L, respectively [28]. To holistically tackle this unknown public health risk posed by DBPs in drinking water, researchers, medical and health professionals, government, and private institutions must work together to formulate a drinking water safety plan, detailing the levels of DBPs arising from each WTP in the country, the deficit in reduction of THMs required to meet the current WHO standards, the needed infrastructure and process improvement at each WTP to meet the standards, the financial burden associated, and the funding sources that could be harnessed to bring WTPs in developing countries like Ghana into the 21st century.

The occurrence of NOM, a heterogeneous mixture of different organic compounds that can be divided into humic (hydrophobic) and non-humic (hydrophilic) substances [66], is expected to increase as the land use and land cover in the catchment of the Barekese Reservoir also changes. According to [67], the Barekese watershed has lost over 70% of its natural vegetation, farmland has increased by 5%, and human habitation has increased by 4%. These changes will alter the nature and concentration of organic compounds that may enter the reservoir and lead to the formation of DBPs in the water treatment process.

The formation potential of THMs in drinking water will also be negatively impacted by the current trend in increasing global temperatures. Although, an insignificant temperature increase has been observed in the Barekese catchment over the past decade [68], the abstracted water temperature will rise significantly through the treatment process, which will increase the THM reaction kinetics and lead to high THM concentrations in the treated water as global warming ravages on. In the decade preceding 2020, the surface water quality in the Barekese catchment also worsened by about 20% [68], which undoubtedly influences the nature and concentration of NOM in the raw water and alters the THM formation potential. Fortunately, the amount of rainfall in the watershed has risen each year, resulting in a marginal decrease in raw water pH. This trend would be antagonistic to the formation of THMs in the disinfection process, although it increases the cost of pH correction at the end of the water treatment process. The generation of THMs in drinking water treatment will also negatively impact ecosystem plants, animals, and microorganisms in the receiving water bodies of untreated wastewater emerging from domestic and industrial premises where the treated water is used [5,69].

The trend in DBP formation in chlorine-disinfected water in tropical climates is set to rise amidst the increasing global temperature and the rapidly changing land use and land cover in most developing countries. It is therefore essential to control the types and concentrations of DOC from anthropogenic sources into water resources and to invest in water treatment technologies with high performance in DOC removal to minimize the formation potential of DBPs and to ensure drinking water safety. Among the technologies available for reducing the occurrence and concentration of THMs in drinking water, nanofiltration and Fenton-based processes are the most efficient techniques for removing DOC, whilst GAC-mediated microbial degradation and air stripping are the most effective removal methods for THMs and HAAs [70].

4. Conclusions and Recommendations

4.1. Conclusions

This research provides a novel approach to estimating THM concentrations in treated drinking water supplies and generating some data in places where direct measurements are not possible due to lack of infrastructure and equipment. The THM formation potential in drinking water produced at Barekese WTP in Ghana ranged from 22.42 to 38.94 µg/L, with a mean concentration of 30.27 µg/L. The resultant lifetime average daily dose was 3.95 × 10⁻⁴ µg/Kg/d for males and 3.93 × 10⁻⁴ µg/Kg/d for female consumers. The lifetime integrative cancer risks resulting from these potential concentrations were calculated as 1.82 × 10⁻⁵ and 1.81 × 10⁻⁵ for males and females, respectively. These potential THM concentrations were lower than the specified maximum standards of 100 µg/L outlined by the USEPA and WHO in drinking water. Many developing countries lack the equipment and instrumentation to conduct direct measurements of disinfection by-products in drinking water; hence, data on the concentrations of various DBPs in chlorine-disinfected waters are non-existent. This study has provided a novel blueprint for an indirect protocol for the computation of THM formation potential. Such data can be used to inform water treatment companies on asset planning and investment for the future to reduce the potential formation of THMs. This data will also help regulators in developing countries to assess the potential risk to consumers and formulate policies to guide water utility companies toward better monitoring of THMs in their final treated water and improved disinfection procedures to reduce the formation potential. However, due to the lack of validation of this study findings through direct measure-ment of THMs concentration in final treated water, the use of this approach could be misleading or result in underestimation of potential THMs formation in drinking wa-ter and hence finding must be applied cautiously.

4.2. Recommendations

It recommended that research institutions and regulatory agencies of the water industry in developing countries dedicate resources to monitoring the prevalence of all DBPs in chlorine-disinfected final water to contribute to the data generation on THM concentrations in public water supplies. Where DBP concentration data are available, the human health risk posed by these THMs and HAAs should be investigated further to identify the risk factors that may lead to disease outcomes based on exposure to DBPs. Water treatment companies should also invest in technologies such GAC adsorption processes prior to disinfection to reduce the fraction of NOM that may persist in the filtered water to be disinfected. Research should be conducted into the development of point-of-use water filtration systems that may reduce the formation of DBPs in treated water and minimize the exposure and the consequential cancer risk posed to consumers.