Constructed wetland-based wastewater treatment of a coffee-washing plant and its impacts: a case study of Kege processing plant, Ethiopia

Coffee is a valuable trading good which is produced in the tropics and mainly consumed in Europe and the United States (Fairtrade 2022). Arabica (Coffea arabica) and Robusta (Coffea canephora) are the two varieties which are internationally traded. Arabica fetches higher prices due to more favorable taste characteristics and makes up 61% of the world's production (Deutsche 2002). Robusta coffee is an important component of commercial coffee blends due to its rich ‘body’ (body is the viscosity, fullness and weight in the mouth of a beverage, ranging from ‘thin, watery’ to ‘thick, heavy’). Coffee is cultivated and exported as a raw, roasted or soluble product to more than 165 countries worldwide providing a livelihood for an estimated 100 million people in the world (ICO 2021). Many countries are involved in coffee production, trade, and communication and it is estimated that about 125 export and re-export coffee products. Ethiopia is the world's third largest coffee exporter after Burundi and El Salvador producing Arabica coffee and an original home of coffee along with the highest diversity in its genetic resource (Haadis & Rani 2008). Coffee plant was founded and cultivated by the Oromo people in the Kaffa province of Ethiopia from which it got its name around 1,000 A.D. Arab people took the coffee seeds from this region and started the first coffee plantation. Then, it spread to the whole of Europe (Adams 1980; Haadis & Rani 2008). Sidama Regional State (SRS) is one of the areas in Ethiopia where coffee is cultivated and processing plants are erected in large numbers. Sidama coffee is a type of Arabica coffee of single origin grown exclusively in the Sidama Province of Ethiopia. Like most African coffees, Ethiopia Sidamo features a small and greyish bean, yet is valued for its deep, spice- and wine- or chocolate-like taste and floral aroma. The most distinctive flavor notes found in all Sidamo coffees are lemon and citrus with bright crisp acidity. Sidama coffee is of very high quality. The SRS, one of Ethiopia's best-known coffee-growing areas, is home to more than 300 wet mills.

After harvesting, coffee can be processed in two ways: dry (natural) processing and wet (washed) processing. Wet processing is done with the help of water, especially to remove the outer red skin and the white fleshy pulp (Wintgens 2004). In addition, the amorphous gel of mucilage around the beans is removed by fermentation. Wet coffee processing can be done in a conventional system, as most of the processing plants do in Ethiopia. The advanced method is currently being practiced in some parts of the country. These technologies are expected to increase the quality of the product and safeguard the environment from pollution. However, the potential of these advanced systems in achieving the required standard is to be evaluated.

The SRS, one of Ethiopia's best-known coffee-growing areas, is home to more than 316CBP wet mills, most of which have not been upgraded to reduce water consumption or treat their waste. For many people in the area, coffee is the sole source of income, providing money for school fees and basic necessities.

Currently, the wastewater frequently released from these agro-industries is impacting the quality and safety of the water for those downstream. More than 1 million people live nearby and rely on the river for their water needs. All wet mill operators in the region lacked the technical knowledge needed to build an adequate treatment system. Better water treatment technologies like constructed wetlands will not only improve water quality for families who depend on the river but will also promote a more sustainable and competitive coffee industry. Therefore, assessing the performance of the newly constructed pilot wetland in the Kege wet coffee-processing plant plays a significant role in this regard for better understanding of the technology.

The overall goal of this research was to compare and establish the potential of natural and constructed wetlands in wastewater treatment of Coffee Berries Processing Agro-industry (CBPA) effluents. The research is conducted under tropical conditions in order to improve the management of agro-industrial liquid effluents in coffee-producing areas of the SRS. The study also evaluated the quality parameters of coffee beans and the soil of the wetland.

In general, the rainfall in the study area is a bimodal type and the main rainy seasons are from June to October and March to May. The maximum rainfall over the northern highlands of the catchment area reaches as high as 180 mm mostly in October and over the western parts it reaches 160 mm in September. The analysis of monthly rainfall indicates that the rainfall pattern over the project area is predominantly bimodal (i.e. the rainfall occurs over a continuous period, but is dominated by two rainfall peaks).

The air temperature of the study area is treated on the basis of observation data obtained from class 3 and class 1 meteorological stations located within and nearby the catchment area of the sub-basin. The monthly temperature data used in this study are the minimum, maximum and the mean temperature. The monthly minimum and maximum temperature values in the Yirgaalem Station (Figure 1), as per the data covering the period from 1986 to 2010, are 14 and 26 °C, respectively.

According to the EPA, humidity above 50% is typically considered too high, while humidity below 30% is usually too low. The average relative humidity (RH) of the study area is 57.5%.

The water quality parameters which were analyzed in this study are given in Table 1.

For the digestion of coffee beans, a procedure for the digestion of coffee powder as recommended by Dubale (2021) was used. For the digestion of soil samples, the EPA 3050B method was applied.

Next to the division box, there is a constructed sedimentation pond. In this pond, the value of BOD₅ decreased to 385 mg/l, i.e. 34.3% of BOD₅ was reduced . This reduction can be used as a gauge of the effectiveness of constructed wetlands in the treatment of coffee-processing wastewater. The BOD₅ assimilation capacity of this sedimentation pond is 65.7%.

The third point to be checked in the performance of wastewater treatment is the natural existing wetland. It reduces organic matter from the wastewater to 53.4 mg/l against the influent load of 586 mg/l at the division box, where wastewater is discharged immediately from the coffee-processing plant. That is, it reduced 91% of the oxygen requirement for biological oxidation and it shows wastewater is going to be free from the concentration of organic matter and BOD₅ becomes less than the national discharge limit at natural wetlands as well as at the constructed wetland.

pH affects the solubility and availability of micronutrients and how they can be utilized by aquatic organisms. The pH of observed samples ranges between 5.2 and 9 measured at the division box and 9 for the pond (sedimentation pond), respectively. Of the five sampling points, three of them show higher pH than neutrality. The decomposition of basic matter or organic matter that are released from the coffee-processing industries leads to an increase in pH (WHO 1984). The observation (Table 2) showed that the natural as well as the constructed wetlands fulfilled local permissible limits for natural river release, 6–9 (EPA, 2003) as well as international standards for aquatic life 6.5–8.5 (WHO 2006).

In all observation points, the turbidity level is below the existing standards. The consumption of more turbid water may constitute a health risk as excessive turbidity can protect pathogenic microorganisms from the effect of disinfectants, and stimulate the growth of bacteria (Zvikomborero 2005).

Nitrogen can exist in water in four forms, NH₃, NO₃⁻, NO₂⁻ and NH₄⁺, which may cause groundwater and surface water pollution in excessive quantity through leaching, stimulate algal growth in surface water that increases maintenance costs in irrigation practices, is carcinogenic in nature and causes blue-baby diseases in human infants. The concentration of NO₃⁻ in the constructed wetland is below the permissible limit (WHO 2006) for drinking and irrigation water uses. Nitrate is highly soluble and mostly susceptible to leaching, causing groundwater pollution. Its main source is the decomposition of organic matter like coffee waste, fecal matter and municipality waste. The concentration of phosphate in the constructed wetland is lower than the recommended limits for freshwater healthy ecosystems as well (EPA, 2003).

Trace levels of dissolved metals in surface water are essential for proper biological functioning in both plants and animals (CCME 2007). Even though the concentration of some trace metals in the wastewater is below the standard, their source shall be verified to minimize their cumulative effect on the ecosystem. Hence, their presence in coffee beans as well as inside the constructed wetland soil was evaluated (Table 3).

Calcium had the highest concentration of macroelements in all soil samples, followed by K and Na (Table 3). Similarly, Cu was detected in greater abundance among the microelements, followed by Mn and Zn. Co > Cd > Ni was assigned to the remaining trace metals.

Potassium was found to have the greatest content among the macroelements (99.93 ± 0.037 ppm) in coffee beans. The highest amounts of K in coffee beans are likely attributable to the fact that nutrients like K, N, P, S and Mg are highly mobile in plant tissues and can be translocated from old to young plant tissues (Weis & Weis 2004). Another reason for increased K concentrations is that this element is among the most important nutrients for plants. The amount of K reported by the present study is much lower than that found in the Gedeo zone, Ethiopia (Dubale 2021) (Table 3). On the other hand, values reported for K in sample beans from other studies (Omer et al. 2019) are comparable with the results obtained. The concentrations of elements in coffee beans are mostly determined by the ambient conditions in the cultivation area as well as the plant's genetic features (Kelly et al. 2005). As a result, changes in the nature of the soil, climate, agronomic methods, or genetic properties of the coffee beans described in this study could be attributed to low K concentrations in coffee beans.

The findings of the present study revealed that Na and Ca were also detected in significant amounts in coffee bean samples from farms, with concentrations of 22.04 ± 0.042 ppm and 17.23 ± 0.36 ppm, respectively. This result is consistent with that reported in Saudi Arabia (Omer et al. 2019).

Mn was the most accumulating microelement in coffee bean samples, followed by Ni, Cu and Zn. Higher manganese levels in coffee beans might be attributed to the availability of this micronutrient in relatively acidic soils of the farmland. The availability of Mn in soil is known to be influenced by the pH of the soil (Wood 1985). Because Mn has a high solubility at low pH, its concentration in acidic soil is likely to be high (Wood 1985). The high concentration of trace metals Ni, Cu and Zn in coffee beans may be due to the fact that these ions are easily transmitted from soil to plants and accumulate in them (Wood 1985; Tsegaye & Struik 2001; Saadia & Nabila 2013).

In the coffee beans of the selected study site, the toxic heavy metals (Pb and Cd) were not found in the coffee beans. This finding is consistent with the previous study reported by Getachew & Worku (2014) on raw and roasted coffee beans, by Abera Gure et al. (2017) in raw coffee beans and Dubale in coffee bean samples from farms. These toxic metals were not detected in the present study area, because there is no environmental degradation owing to industrial operations in Dale Woreda, toxic metals (Pb and Cd) concentrations in coffee beans and soils are expected to be quite low. Furthermore, the absence of commercial fertilizers and pesticides for coffee plantations in the woreda may be evidenced by the low amounts of hazardous metals. Furthermore, Pb and Cd have no nutritional value for humans, despite their very low concentration with no health risks as a result of these hazardous substances in Dale Woreda coffee beans.

In coffee beans from Kege farmers' farms, Mn was determined to be the most abundant minor element, followed by Cu, Zn and Ni, of which Mn (0.927 ± 0.004 ppm) was present in higher concentration than the other microelements (Table 3). Likewise, the levels of Cu (0.22 ± 0.0026 ppm) and Zn (0.073 ± 0.003 ppm) were higher than that of Ni (0.055 ± 0.004 ppm) (Table 3) except for Pb, Cd, Co and Cr, which were not detected in coffee bean.

Soil sample analysis shows that calcium (1,355 ± 18.02 ppm) was the largest of all the detected metals. K and Na were found in larger amounts next to calcium with values of 673 ± 2.65 ppm and 111.63 ± 0.35 ppm, respectively. Out of the analyzed microelements, Cu (42.66 ± 2.52) was found to be in larger amounts followed by Co (5.33 ± 0.305 ppm) and Mn (0.62 ± 0.238 ppm). Likewise, the concentrations of Cd, Ni and Zn were found to be 3.1 ± 0.1 ppm, 0.191 ± 0.009 ppm and 0.161 ± 0.007 ppm, respectively. The average metal concentrations in the constructed wetland soil were determined to be in the following order: Ca > K > Na > Cu > Co > Cd > Mn > Ni > Zn.

Even though sample preparation and analysis methodologies differed, the results obtained in this study were compared to those reported by other studies (Table 4). The concentration of essential and non-essential metals in raw and roasted coffee beans cultivated in various parts of the world, including Ethiopia, has been documented by a number of writers. In Table 4, the concentrations of metals in coffee beans obtained in the current investigation are compared to reported levels.

The results show that the metal content found is more or less comparable to levels published in the literature. In this investigation, however, the concentration of K was found to be lower than previously reported values. This may indicate that Ethiopian coffee is grown without the use of fertilizers. Pb and Cd concentrations were much below the technique detection limits, exactly as they had been in previous studies (Ashu & Chandravanshi 2011; Gure et al. 2017; Dubale 2021).

The findings of this study are more or less compatible with those reported by different researchers from various nations. Cu and Mn contents, on the other hand, are lower than the previous studies (Gure et al. 2017; Silva et al. 2017; Dubale 2021). The current study's findings are in good agreement with the majority of published values. Furthermore, for macroelements K > Na > Ca, the general trend of metal concentration is in good agreement. Mostly trace microelements trend follows Mn > Cu > Ni > Zn in coffee beans from different farms.

Maximum allowable limits of heavy metals in coffee beans were not found in the literature. As a result, standards for other foods and herbal plants were used to compare (Table 5). Trace heavy metal concentrations in plants were critical for animal and human health, but the metals should be kept within allowed ranges, as established by FAO/WHO and other standard-setting authorities. Health is harmed by concentrations that are higher or lower than the prescribed limits (FAO/WHO 1993; WHO 2005).

Nickel concentrations in coffee beans from farms (CBFF) and coffee from the washing industry varied from 0.05 to 0.08 ppm (CBWI). FAO/WHO (1993) defined a tolerable level of 1.63 ppm in edible plants. When the metal limits in the researched medicinal plants are compared to those proposed by FAO/WHO (1993), it is discovered that all plants collect Ni below this level. The WHO (2005) limits for Ni have not yet been defined for medicinal plants. Because Ni absorption by the body is quite low, Ni poisoning in humans is not very common (Onianwa et al. 2000).

Zn levels in CBFF varied from 0.054–0.076 ppm to 0.051–0.09 ppm in CBWI. FAO/WHO (1993) defined an acceptable level of 27.4 ppm in edible plants. Following a comparison of the metal limit in the investigated coffee beans with those specified by FAO/WHO (1993), it was discovered that all samples fall within this range. The WHO (2005) limits for Zn have not yet been defined for medicinal plants.

Cu levels in CBFF and CBWI were 0.14–0.29 ppm and 0.14–0.28 ppm, respectively. FAO/WHO (1993) defined an acceptable level of 3.00 ppm in edible plants. After comparing the metal limit in the investigated coffee beans to the FAO/WHO (1993) recommendations, it was discovered that all coffee beans collect Cu below this level. However, the WHO (2005) limits for Cu have not yet been determined for medicinal plants. Cu acceptable limits in medicinal plants were determined by China and Singapore at 20 and 150 ppm, respectively (WHO 2005). Cu levels in agricultural products should be between 4 and 15 parts per million, according to Allaway (1968).

Mn concentrations ranged from 0.08 to 0.11 ppm in CBFF and 0.07 to 0.10 ppm in CBWI. FAO/WHO (1993) defined an acceptable level of 2 ppm in edible plants. After comparing the metal limit in the investigated coffee bean to the FAO/WHO (1993) recommendations, it was discovered that all plants collect Mn below this level. The WHO (2005) limits for Mn have not yet been defined for medicinal plants. Dubale (2021) found that Mn levels in coffee beans ranged from 17.3–23.6 ppm in CBFF and 17.2–22.6 ppm in CBWI.

Pb and Cd contents in the coffee beans tested were well below the detection limits of the technique. The FAO/WHO (1993) permitted limit for Cd and Pb in edible plants was 0.21 and 0.43 ppm, respectively. However, the WHO (2005) set a permissible limit for Cd of 0.3 ppm and a maximum for Pb of 10 ppm for medicinal plants.

According to international heavy metal guidelines, the quantities of metals identified in coffee bean samples were below the maximum allowed limit. As a result, it may be concluded that there is now no health risk linked with the use of Dale Woreda coffee beans.

Coffee is one of the world's most popular and frequently consumed beverages, with significant commercial and social significance. In international trade, it is also one of the most important agricultural goods. As a result, the current study attempted to investigate the performance capacity of a constructed wetland in Dale Woreda of the SRS. The source of the contaminants was traced through the analysis of the concentrations of K, Na, Ca, Mn, Cu, Zn, Ni, Cr, Co, Cd, and Pb in coffee beans from farms and from the washing industry. The constructed wetland showed a superior performance in treating the wet coffee effluent to national and international standards. It reduced BOD₅ to 65.69% at the sedimentation pond and 3.4% at the outlet of the constructed wetland against 9.1% at the outlet of the natural wetland. The constructed wetland performs on par with the natural wetland in treating the effluent from the coffee-washing station. The levels of essential and non-essential metals in coffee beans (from farms and washing industries) and soil samples were determined in this study. Metal levels in coffee bean samples from farmers' farms are in the following order: K > Na > Ca > Mn > Cu > Ni > Zn. Metal levels were found to be K > Na > Ca > Mn > Cu > Zn > Ni in coffee beans from the washing industry. However, the average metal concentrations in the soil were determined to be in the following order: Ca > K > Na > Cu > Co > Cd > Mn > Ni > Zn. In both cases, the levels of toxic metals (Pb and Cd) were not detected, and trace heavy metal levels were below the FAO/WHO maximum permissible limits. As a result, there is no health risk linked with the use of coffee beans due to harmful and trace heavy metals.

According to the findings of this study, there are permitted levels of macro and trace elements in coffee beans from farms and washing plants. With the intervention of a constructed wetland at the wet coffee-processing plants, the organic contamination of rivers can easily be mitigated. As a result, organic as well as metal pollutants have no effect on the coffee grown in Dale Woreda.

Based on the results of this research, it is recommended to implement the constructed wetland for treating effluent from small coffee-washing stations in a cost-effective manner in the tropics. Policymakers are advised to enforce this technology in wet coffee-processing stations as an affordable intervention in the Gidabo Watershed of Southern Ethiopia. This study needs to be continued at least for 3 years to verify the sustainability of the treatment system. It is recommended to go for further research to optimize the effluent flow rate with the size of the constructed wetland.

The authors would like to offer particular thanks to the Directorate of Research, Hawassa University Institute of Technology for funding this research project.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.