Research Article | Open Access
Assessment of the Nutrients in the Leachate and the Groundwater Quality for Drinking and Farming around the Nkolfoulou Landfill in Yaoundé, Cameroon
This study focuses on the assessment of the nutrients in the leachate and the groundwater quality around the Nkolfoulou landfill in Yaoundé known in French as “Centre de Traitement de Déchets (CTD).” Landfilling generates leachate that can pollute groundwater. Leachate along with groundwater samples () was collected in January (long dry season) and May (long wet season) 2014 and explored for various parameters including pH, temperature, EC, turbidity, TDS, TA, TSS, TH, , COD, , , , , , , , , , , , and colour using standard methods. In the leachate samples, values of TSS (700.2 and 130.2 mg/L), (140 mg/L), COD (1350 and 1750 mg/L), (82.50 and 39.51 mg/L), (159.32 and 74.82 mg/L), and (702.69 and 345.50 mg/L) exceeded the Cameroonian standards for effluent discharge. All the values of pH and some values of turbidity (4.55 and 4.50 NTU) and (0.51 and 0.73 mg/L) in the groundwater samples violated the Cameroonian standards for drinking water. Based on the water quality index (WQI), an average of 11.53% of groundwater samples was improper for drinking in both seasons. Based on the parameters assessed, all the samples complied with the standard set for irrigation, poultry, and livestock. The hazard quotient (HQ) and the hazard index (HI) of and for children and adults were <1, and hence, the increased non-cancer risks due to these ions through the drinking of groundwater was low. From the statistical analysis, the Nkolfoulou landfill may not be the main source of major ions to the nearby groundwater.
Groundwater is important in potable water supply, livestock, and crops production. Many people in Africa rely on groundwater for drinking . But waste disposal negatively impacts the quality of groundwater . Solid wastes are prevalently managed by landfilling  because it is the simplest and the cheapest technique of waste management . Landfilling generates leachate which is heavily loaded with nutrients such as , , , , , , , , , , and [5–9]. Many scientific reports emphasize that groundwater close to a landfill is often polluted via leachate [10–13]. Nutrients at an elevated concentration in water render it unsuitable for human and animal health and for farming. For example, high levels of and in water are responsible for its hardness. Hard water requires more soap for cleansing and can form scale deposits on domestic appliances, while drinking water containing excessive levels of could result in a purgative effect . Ingestion of nitrate and fluoride beyond the threshold level can cause, respectively, blue-baby syndrome in bottle-fed infants [14, 15] and skeletal fluorosis in human beings in general [14, 16–18].
In Yaoundé, the capital city of Cameroon, municipal solid wastes have been landfilled in Nkolfoulou since 1989. The Nkolfoulou landfill which was then in a bush far away from the city is now surrounded by habitations. The landfill generates an average of 450 m3/day of leachate, out of which only 54 m3 is collected . The uncollected leachate can percolate the soil to pollute nearby groundwater sources. The study area is partly covered by a water distribution network. Besides, the increasing need for water has resulted in frequent pipe-borne water shortage, thus, putting a stress on the population. To tackle this issue more efficiently, many people around the Nkolfoulou landfill have resorted to digging wells or boreholes in their premises. The water abstracted from these wells and boreholes is quite clean and odourless, allowing the people to use it for different purposes without any prior treatment. This practice is unacceptable since water that appears clean may contain harmful microbes and chemicals . Relevant information on the nutrient status of groundwater in the vicinity of landfills in Cameroon is lacking. This is why the present study is aimed at assessing the nutrients in the leachate and the groundwater quality around the Nkolfoulou landfill.
2. Materials and Methods
2.1. Study Site Description
Yaoundé, the capital city of Cameroon, locally called “Ngola,” sprawls over 304 km2. It is ranked as the second most populated town in Cameroon with about 2.5 million inhabitants. It is located at an altitude of about 750 m and lies between longitudes and between latitudes . It has an annual mean rainfall of 1520 ± 268 mm and a mean temperature of 24.4 ± 0.5°C from 1984 to 2008 . Yaoundé has a tropical climate consisting of two rainy seasons (March–June and September–November) and two dry seasons (December–February and July-August). The bed rock of the study area comprises migmatites, migmatitic gneisses, banded gneisses, and mica schists .
Figure 1 displays the map of the study area which consists of the neighborhoods of the Nkolfoulou landfill. The Nkolfoulou landfill is situated at 16 km from Yaoundé centre and is surrounded by Nkolfoulou I and II and Nsan villages. It covers a total land area of about 45 ha . The site specification of groundwater samples in the study area is displayed in Table 1.
GW = groundwater; TW = type of well; HDW = hand-dug well; DW = drilled well; DLE = distance from the landfill edge.
2.2. Sampling Procedure
Thirteen (13) groundwater samples and one (01) sample of leachate not stored by the HYSACAM company were collected in 2014 in January, during the long dry season, and in May, during the long wet season. Since the landfill is equipped with leachate collectors, we were more interested in the unstored leachate that flows outside the landfill. For ion analysis, water and leachate were sampled and filtered in situ through a clean 0.45 μm filter paper in 500 mL polyethylene bottles previously washed with deionized water. The samples were kept in a cooler during transportation from the field to the laboratory, stored in a refrigerator at 4°C, and analyzed within 48 hours.
2.3. Physicochemical Parameter Analysis
All the reagents utilized were of analytical (AnalaR) grade. Parameters such as temperature, electrical conductivity (EC), pH, and total dissolved solids (TDS) were measured in situ using a multiparameter HANNA HI 9811-5. Total suspended solids (TSS) were determined following the AFNOR method NF EN 872 , while total alkalinity (TA) and biocarbonate () were determined by titration as outlined in the AFNOR method NF EN ISO 9963-1 . The turbidity was determined using a turbidity meter model ORBECO-HELIGE 966. Chemical oxygen demand (COD) was determined by digestion following the standard method 5220B , while biochemical oxygen demand after 5 days () was assessed as specified by Rodier et al. . The colour was determined in the laboratory by the colorimetric method using the ORCHIDIS standard colorimetric comparator réf.1pco15p220. The last three parameters were assessed only in leachate samples.
The concentration of anions and cations was determined, respectively, employing models ICS1100 and ICS90 liquid-phase ion chromatographic (IC) single-column system at room temperature. All samples and eluents were filtered through a filter paper (0.45 μm) before injection into the IC system. Water samples having higher electrical conductivity were diluted appropriately with distilled water before being analyzed. The chromatograph was computer monitored using a CHROMELEON CM 6.80 SR software.
The anions , , , , and were separated on an AS12A IonPac separating column (4 × 200 mm) with an AG12A guard column (4 × 50 mm) and detected after suppression with an ASRS300 IonPac 4 mm anion electrical self-regenerating suppressor. The anions were eluted using the eluent (2.7 mM + 0.3 mM ) at a flow rate of 1 mL/min. A sample volume of 100 μL was injected and run for 15 minutes. The cycle time was 15 minutes per analysis.
The cations , , , , and were separated on a CS12A IonPac separating column (4 × 250 mm) with a CG-12A IonPac guard column and detected after suppression with a CSRS300 IonPac 4 mm cation electrical self-regenerating suppressor. The cations were eluted with 22 mM as eluent at a flow rate of 1 mL/minute. The injection volume was 50 μL, while the run time was set to 15 minutes, and the cycle time was 15 minutes per analysis.
The normalized inorganic charge balance (NICB) was computed to check the accuracy of the results using the following equation , and the calculated NICB was within ±5%:
The quality rating scale for each parameter was computed as follows:where is the estimated concentration of the parameter in the analyzed water, is the ideal value of this parameter in pure water ( = 0, except for pH = 7.0), and is the maximum limit of the parameter. The unit weight, , for each parameter is computed as follows:where = constant of proportionality and can be computed as
The increased non-carcinogenic risks due to nitrate and fluorite were evaluated by calculating the hazard quotient (HQ) and the resulting hazard index (HI). HQ was computed from equation (7), while ADD was computed using equation (8) :where ADD represents the mean daily dose of a pollutant ingested from drinking water (mg/kg/day) and RfD is the reference dose for non-carcinogenic risk. The RfD for nitrate is 1.6 mg/kg/day while that for fluorite is 0.06 mg/kg/day. AAD is computed using the following equation:where represents the concentration of a particular pollutant in groundwater (mg/L), IR is the water ingestion rate (L/day), BW is the body weight (kg), EF is the exposure frequency (days/year) (EF = 365 days per year), ED is the exposure duration (year) (ED = 70 years), and AT is the average exposure time (days) (AT = days)). Thus, from equation (8),and equation (8) becomes
In this study, the IR for adults and children is 2 and 1 L/day, respectively, whereas the BW for adults and children is 60 and 18 kg.
The resulting hazard index (HI) due to exposure to many non-carcinogenic substances was calculated from the following equation [29, 30]:where are the non-carcinogenic substances and is the hazard quotient for the substance.
Sodium adsorption ratio (SAR) was computed using the following equation with and in meq/L:
The Stiff and Piper plots were generated using DIAGRAMME 5.1 software. The geographical coordinates of the selected sampling points were recorded using a Geographical Positioning System (GPS) Magellan Triton-300. These coordinates were loaded in ArcGIS 10 software in order to draw the map of the study area and to evaluate the distances between the border of the landfill and the monitoring sites.
2.4. Data Analysis
Statistical processing of the data for t-test and correlation matrix was carried out using the Statistical Package for Social Sciences (SPSS) 23.0.
3. Results and Discussion
3.1. Physicochemical Quality of Leachate
The physicochemical quality of the leachate is reported in Table 2. From the long dry season to the long wet season, the pH and temperature fell slightly from 8.9 to 8.2 and 26.7 to 26.2°C, respectively. The pH remained within the safe range of , while the temperature dropped below the safe limit of 30°C set by MINEP  for effluent discharge. A similar pH value of 8.1 was found in leachate from Taiwan . The pH values obtained indicate that the leachate samples were from an old cell since old leachate is characterized by a pH > 7.5 .
ND = not detected; SD = standard deviation.
During the long dry and long wet seasons, respectively, the values of (100; 140 mg·/L), COD (1,350; 1,750 mg·/L), and TSS (70.2; 130.2 mg/L) were all found to be higher than the regulatory limits for the effluent discharge set by MINEP  (Table 2). During this same period, leachate had a biodegradability ratio (/COD) of 0.07 and 0.08, respectively. This implies that the leachate was from an old cell since a ratio of /COD < 0.1 is typical of leachate from an old landfill (>10 years) [7, 33]. Again, during this same period, the concentration of (82.5; 39.51 mg/L), (159.32; 74.82 mg/L), and (702.69; 345.50 mg/L) was all superior to the safe limits for effluent discharge stated by MINEP  (Table 2). Ammonia-nitrogen in leachate may originate from the catabolism of protein-rich compounds [7, 34, 35]. is produced during ammonification  and can thereafter be partially transformed into by the anaerobic ammonium oxidation process (ANAMMOX) as shown in equation (13). It can also be transformed into under aerobic conditions through nitrification , while leachate stagnates or flows on the ground: was not detected in any of the seasons, while concentration dropped from 50.24 to 10.60 mg/L. found in leachate may be the result of the oxidation of , while the absence of might be due to its coprecipitation by and to form magnesium ammonium phosphate (MAP) hexahydrate known as struvite as shown in the following equation :
In both seasons, leachate had a dark-brownish colour ranging from 5208.33 to 4654.75 mg/L Pt-Co with an offensive pungent smell. In a similar study , the values were in the range of 955–15.142 mg/L Pt-Co for an old leachate. The dark brown colour of the leachate may be due to the presence of FeS and PbS formed, respectively, by the reaction of iron and lead with hydrogen sulfide. It may also be due to the presence of MnO resulting from the oxidation of manganese in air. The pungent odour of the leachate is attributable to the presence of either , , , and volatile fatty acids (VFAs) or putrescine and cadaverine from the putrefaction of wastes from the abattoirs.
Comparatively, the average physicochemical characteristics of the leachate samples collected in this work from the Nkolfoulou landfill differ from those observed elsewhere (Table 3). This may be due to the differences in waste composition, meteorological conditions, waste age, and landfilling technology [6, 7]. As observed in Table 3, all comparable parameters were in the typical range stated by Christensen et al.  except for TSS which was much lower, more likely as a result of the soil type on which the leachate leaks or runs.
ND = not detected.
Seasonal variation of the levels of nutrients as well as the pH, EC, TDS, TA, TH, and colour exhibited a decreasing trend with percent decrease ranging from % (EC) to −78.90% () in going from the long dry season to the long wet season. This decrease is attributable to the dilution of leachate by runoff or rain water. In contrast, the increase in TSS and turbidity as well as and COD in the long wet season could be due to runoff that brings solid particles and organic matter to the leachate. Seasonal variation in major ion concentrations is illustrated by the Stiff diagram (Figure 2) in which the shapes of polygons differ from season to season implying that the leachate composition varies according to the season. In the leachate samples, the trends in the cation concentrations were in the order > > > > in the long dry season and > > > > in the long wet season while that of the anions was > > > > > in both seasons.
3.2. Groundwater Quality
The physicochemical quality of groundwater in both seasons is displayed in Tables 4 and 5. During the long dry season, the groundwater temperatures were in the range °C (mean 23.5°C), while during the long wet season, the range was °C (mean 23.3°C). All the temperature values for both seasons were lower than 25°C, which is the standard limit according to Cameroonian regulations  for drinking water.
ND = not detected; SD = standard deviation.
ND = not detected; SD = standard deviation.
The pH of groundwater samples varied from 5.2 to 6.1 (mean 5.6) during the long dry season and 5.8 to 6.4 (mean 6.1) during the long wet season. These pH ranges are lower than recorded by Mor et al.  for groundwater around a landfill in Delhi (India). For both seasons, the pH values were outside the safe ranges of 6.5–9.5, 6.5–8.5, and 6.5–9.0 set, respectively, by EU , EPA , and Cameroonian standards  (Table 6) for drinking water.
Disinfection effectiveness; taste consideration.
During the long dry and long wet seasons, the EC values were, respectively, in the range 30–130 μS/cm (mean 76 μS/cm) and 20–150 μS/cm (mean 49 μS/cm), while the TDS values were in the range 10–60 mg/L (mean 32 mg/L) and 10–70 mg/L (mean 21 mg/L). As can be observed, the EC and TDS values are several folds lower than 12,745 μS/cm and 9,895 mg/L, respectively, as reported by El-Salam and Abu-Zuid  for groundwater near a landfill in Alexandria (Egypt). With respect to drinking water, the EC and TDS values complied with the standards set by regulatory bodies [14, 18, 42, 43] (Table 6).
During the long dry and long wet seasons, the turbidity values registered were, respectively, between 0.70 and 12.60 NTU (mean 4.55 NTU) and 0.70 and 14.5 NTU (mean 4.50 NTU), whereas the TSS values were between 5.00 and 90.00 mg/L (average 32.44 mg/L) and 5.00 and 103.79 mg/L (average 31.89 mg/L). Comparatively, Han et al.  reported high values in the range NTU in groundwater near a landfill at Zhoukou (China). All the recorded TSS values were lower than the recommended maximum limit set by MINDIC for drinking water  (Table 6). In the long dry season, , , , and wells representing 30.76% of the samples and , , and wells in the long wet season representing 23.07% of the samples had turbidity less than 1 NTU which is within the standard limit for drinking water disinfection effectiveness laid down by WHO . Higher turbidity can seriously interfere with the efficiency of disinfection by providing protection for organisms .
During the long dry and long wet seasons, the TH values increased, respectively, from 16.02 to 49.16 mg/L (average 26.14 mg/L) and 11.14 to 44.90 mg/L (average 25.93 mg/L), while the TA values increased from 19.02 to 90.28 mg/L (mean 34.66 mg/L) and 14.27 to 53.27 mg/L (mean 29.39 mg/L). These TH and TA values were, respectively, lower than those of mg/L and mg/L recorded by Mor et al.  for groundwater near the Gazipur landfill (India).
During the long dry season, the concentration of , , , and fluctuated, respectively, between 0.48 and 22.83 mg/L (mean 4.07 mg/L), 0.70 and 5.71 mg/L (mean 2.46 mg/L), 0.20 and 6.05 mg/L (mean 2.47 mg/L), and 2.54 and 13.75 mg/L (mean 7.61 mg/L) while in the long wet season, they ranged from 0.03 to 2.31 mg/L (mean 1.18 mg/L), 1.20 to 8.44 mg/L (mean 2.30 mg/L), 0.55 to 4.58 mg/L (mean 1.98 mg/L), and 1.45 to 15.74 mg/L (mean 7.11 mg/L), respectively. The average values of , , , and obtained in this work are lower than 23.75, 11.17, 84.6, and 360 mg/L, respectively, reported by Singh et al.  for groundwater around the Pirana landfill in Ahmedabad (India). In this study, all the concentrations of obtained were many folds lower than the maximum limits stated by the regulatory bodies [14, 42, 43] for drinking water (Table 6). In both seasons, the levels of and were lower than the permissible limits laid down by MINDIC  for drinking water (Table 6).
In the long dry and long wet seasons, the levels were observed, respectively, between ND (non detected) and 0.51 mg/L (mean 0.14 mg/L) and ND and 0.73 mg/L (mean 0.23 mg/L), while the values varied from 0.63 to 11.76 mg/L (mean 3.76 mg/L) and 0.07 to 7.97 mg/L (mean 2.29 mg/L). Comparatively, high values of in the range ND–4.3 mg/L were reported by Mor et al.  for groundwater near the Gazipur landfill in India, while, for , a mean value of 49.90 mg/L higher than those found in this study was recorded by Singh et al.  for groundwater at the vicinity of the Pirana landfill also in India. In the study area, the in the groundwater might have originated from the leachate or leaked from the septic tank. The values registered in the long dry season at (0.51 mg/L) and in the long wet season at (0.73 mg/L) exceeded both the EU  and MINDIC  standards for drinking water (Table 6). However, it has been recently substantiated that the additional exposure to from water in the concentration range of mg/L is negligible and thus does not pose a risk to human health . in drinking water may have resulted from the protonation of given the acidic nature (pH < 6.5) of the water. The values of in both seasons were lower than the threshold set by EU , EPA , and MINDIC  for drinking water (Table 6).
During the long dry season, the , and values were, respectively, between 0.02 and 0.08 mg/L (average 0.04 mg/L), 0.22 and 3.72 mg/L (average 1.02 mg/L), and 0.17 and 3.26 mg/L (average 1.28 mg/L), while in the long wet season, they were, respectively, between 0.01 and 0.10 mg/L (mean 0.04 mg/L), 0.15 and 0.91 mg/L (mean 0.49 mg/L), and 0.03 and 3.60 (mean 0.78 mg/L). The mean average values of , , and are, respectively, several orders of magnitude lower than 0.67, 253.94, and 96.94 mg/L recorded by Pujari and Deshpande  for groundwater around a landfill in Nagpur (India). As far as , , and were concerned, groundwater sources were fit for drinking for both seasons since their concentrations were lower than the prescribed limits set by EU , EPA , MINDIC , and WHO  (Table 6).
was barely detected in the three wells (, , and ) during the long dry season and in six wells (, , , , , and ) during the long wet season with concentrations ranging slightly from 0.01 to 0.03 mg/L in both seasons. The concentrations of increased from 23.20 to 110.10 (mean 42.26 mg/L) in the long dry season and from 17.40 to 64.96 mg/L (mean 35.84 mg/L) in the long wet season. The mean values of and are, respectively, lower than 0.29 mg/L and 391.8 mg/L reported for groundwater around the Pirana landfill in Ahmedabad (India) . was not detected in the leachate samples, so its occurrence in some groundwater samples in the study area may be attributed to anthropogenic activities such as domestic waste discharge and agriculture. In groundwater, may have originated from the dissolution of atmospheric and carbonate minerals  or from sulfate reduction of organic compounds in the aquifer as represented in the following equation :
According to Piper’s diagram  (Figure 3), the hydrochemical facies of groundwater in the study area were of the (84.62%) and (15.38%) types in the long dry season, while in the long wet season, all the groundwater samples were of the type (100%).
3.3. Statistical Correlation
The correlation matrices for 19 measured variables during the long dry and long wet seasons are illustrated in Table 7. The perfect positive correlation between TSS and turbidity (r = 1, ) in both seasons means that they have exactly the same contributor which could be mud brought in by the infiltrating rain water. The strong positive correlation between EC and TDS (r = 0.98, ) in the dry season signifies that they have nearly the same contributors (the dissolved ions). At the 0.05 level, a significant negative correlation was observed only between the temperature and (r = ) in the long dry season, while for the long wet season, it was registered between pH and (r = ), between pH and (r = ), between pH and (r = ), and between pH and (r = ) indicating the opposing distribution of these pair variables. In both seasons, the depth of the wells exhibited no significant correlation with any of the variables in the matrices except for with a significant positive correlation (r = 0.78, ) in the long dry season. In the long dry season, there was a significant positive correlation between pH and (r = 0.68), and (r = 0.64), and (r = 0.68), and and (r = 0.66) at the 0.05 level, while at the 0.01 level, there was a significant positive correlation between and (r = 0.89), and (r = 0.87), and (r = 0.91), and (r = 0.78), and (r = 0.83), and (r = 0.84), and and (r = 0.77).
LDS = long dry season; LWS = long wet season. Correlation is significant at the 0.05 level; correlation is significant at the 0.01 level.
In the long wet season, there was a significant positive correlation between and (r = 0.61), and (r = 0.59), and (r = 0.56), and (r = 0.63), and (r = 0.58), and (r = 0.66), and and (r = 0.68) at the 0.05 level, whereas at the 0.01 level, there was a significant positive correlation between and TSS (r = 0.75), and turbidity (r = 0.74), and (r = 0.82), and (r = 0.83), and (r = 0.85), and (r = 0.76), and (r = 0.88), and (r = 0.75), and (r = 0.90), and (r = 0.74), and (r = 0.74), and and (r = 0.73).
The distance between the wells and the landfill barely exhibited significant negative correlation with pH (r = , ) and temperature (r = , ), implying that the landfill may not be the main source of the major ions in the nearby groundwater. The strong positive correlations observed between the various ions in both seasons imply that they may have common sources (weathering of rocks, septic tanks, pit latrines, domestic effluent, and agricultural activities), which contribute to the ion input in the groundwater in the area.
3.4. Seasonal Variation of Groundwater Quality
Although the t-test is usually applied on larger sample sizes, it can be used for small sample sizes (), as long as the effect sample size is large . The t-test was performed to check the significance of the seasonal variation of the mean value of the parameters assessed. Results of this test are summarized in Table 8. The resulting values were compared with the significance level, = 0.05. The seasonal variation of parameters was statistically significant only for pH (), T (temperature) (), and EC () (Table 8), whereas it was insignificant for the remaining data pairs representing 84.21% as . It can therefore be inferred that the change in groundwater quality from the dry season to the wet season was generally insignificant within a 95% confidence limit.
d = dry season; w = wet season; turb = turbidity. Bolded figures indicate significant difference.
In the groundwater samples, the order of abundance of cations was > > > > in the long dry season and > > > > in the long wet season while that of anions was > > > > > in both seasons.
3.5. Drinking Water Suitability Based on TDS, TH, and WQI
Following the classification of van der Aa  of drinking water based on TDS (Table 9), apart from the and wells in the long dry season and the and wells in the long wet season, the remaining wells, representing 84.61% of the samples in both seasons, had very low mineral concentrations. According to the same classification, 15.38% of the samples in the long dry season (