Methane production from anaerobic pre-treatment of municipal wastewater combined with olive mill wastewater: A demonstration study

One of the most problematic agro-industrial effluents in the Mediterranean basin is olive mill wastewater (OMW) generated by the olive oil extraction process. The annual generation of OMW in the Mediterranean was estimated to be 30 million cubic meters in 2014 (Chiavola et al. 2014). It represents one of the most challenging effluents, demanding special considerations for appropriate management and treatment prior to discharge or recycling, and may also demonstrate characteristics similar to other refractory agro-industrial effluents (e.g., from the processing of vegetables and slaughterhouses of livestock).

OMW poses serious environmental threats due to its high chemical oxygen demand (COD), which ranges from 80 to 200 g/L (100–200 times higher than levels of domestic wastewater (WW)). The high organic load derives from the different organic substances found in OMW, including sugars, tannins, phenolic compounds, organic acids, and lipids (Zbakh & El Abbassi 2012). Specifically, OMW often contains high concentrations of phenolic compounds (total phenolic compounds vary greatly from 0.5 to 24 g/L), and the presence of these inhibitory organic compounds is one of the main challenges in the detoxification and purification of OMW. Additionally, some of these phenolic compounds are responsible for various biological effects, including phytotoxicity, antibiosis, and inhibitory effects (Rodriguez et al. 1988; Capasso et al. 1992; Borja et al. 2006; Barbera et al. 2013; Maragkaki et al. 2017). It is suspected that phenolic compounds contribute to the toxicity and antibacterial activity of agricultural WW in general. Hence, increased concern has been expressed in regard to the effective treatment for these compounds and the safe discharge of OMW into the environment (Silva et al. 2007) and WW treatment systems (Sabbah et al. 2004).

An additional constraint of the treatment of OMW relates to its short and seasonal production (i.e., large quantities of OMW are produced in a relatively short time) (Ramos-Cormenzana et al. 1995; Davies et al. 2004; Gernjak et al. 2004; Sabbah et al. 2004; Paraskeva & Diamadopoulos 2006). Also, OMW composition can vary broadly depending on many parameters, such as olive variety, harvesting time, climatic conditions, and oil extraction process, among others (Zbakh & El Abbassi 2012). Moreover, OMW is typically acidic, with pH values ranging from 4.5 to 5.5, due to the presence of organic acids. The acidic nature of OMW can further exacerbate its toxic effects and hinder biological treatment processes. Thus, the complex composition and inhibitory properties of OMW present additional challenges for its treatment and disposal (Borja et al. 2006; Maragkaki et al. 2017).

In the past, many olive oil-producing countries used to discharge OMW directly into rivers, leading to eutrophication and excessive algal blooms (Galanakis 2016; Manthos et al. 2023). Nowadays, in all EU countries, the direct discharge of OMW into open water bodies and WW treatment systems is strictly prohibited due to its harmful effects on the ecological balance. Despite this, illegal disposal of OMW into nearby aquatic resources has been frequently observed (Koutsos et al. 2018).

Each Mediterranean country has approached the regulation of OMW differently. In Spain, approximately 1,000 evaporation ponds have been constructed to promote evaporation during summer, which guarantees the protection of water bodies from any OMW discharge, but causes odor nuisances (Koutsos et al. 2018), in addition to other risks of water pollution from soil percolation, insect proliferation, and increased methane emissions into the atmosphere (Paraskeva & Diamadopoulos 2006; Jarboui et al. 2010; Messineo et al. 2012). An additional common practice for managing OMW involves using OMW, in small doses, as a soil fertilizer (Batuecas et al. 2019), depending on the soil's hydraulic conductivity (Barbera et al. 2013). Specifically, Italy is the only olive oil-producing country in the world with legislation for the disposal of olive mill waste into the soil. Italian Law (no. 574/96) allows the spreading of up to 50 or 80 m³ ha⁻¹ year⁻¹ OWM generated by the press or continuous centrifugation systems, respectively (Altieri & Esposito 2008; Barbera et al. 2014). Greece, on the other hand, lacks specific regulations (Kapellakis et al. 2006; Tsagaraki et al. 2007). The primary standards for OMW management in Greece require olive mill owners to provide an environmental impact assessment for the generated olive WW (Galanakis 2016; Koutsos et al. 2018).

In Israel, the Environmental Protection Ministry adapted this land spreading method according to the Italian bylaw (olive plantations in particular), but under specific controlled conditions. The recommended application guides OMW owners to spread up to 6 m³ per dunam in the same area every season (Ministry of Environmental Protection 2014). However, some of the OMW is still illegally discharged into the open environment and nearby sewerage facilities, causing irreversible damage to the biological WW treatment systems. The Karmiel municipal WWTP (10 Mio. m³/year) in Israel is an example of a medium-size WWTP that faces severe problems due to shock loads of COD and fat, oil and grease (FOG) because of illegal discharges of OMW during the olive harvest period and from other sources (i.e., slaughterhouses) in the area. This causes substantial operational difficulties for the WWTP, leading to financial losses and non-compliance with regulations.

Within the ULTIMATE project, the Karmiel case study aimed to close the water and energy loops. The objective was to offer a special management and treatment paradigm to mix OMW with municipal WW under controlled conditions, with the advantage of exploiting the high organic load to increase the production of biogas. This can be applied through the advanced anaerobic technology (AAT) that we developed (Sabbah et al. 2016) as pretreatment of the municipal WW mixed with agro-industrial effluents (i.e. WW from olive mills), while protecting the downstream treatment by the typical activated sludge system of the Karmiel WWTP. The application of AAT was proposed as a first barrier against the fluctuations caused by infused olive oil mills' residues in order to improve the removal of dissolved organic material prior to the aerobic stage. That step would lead to a reduction in both the energy consumption required for the aeration pond and the quantity of sludge generated during the aerobic treatment stage.

The AAT treatment is based on a unique patented process for preparing ‘pre-treated’ biomass, immobilized in a polymeric matrix that has a high handling capability and fast acclimatization to WW streams containing high organic loads (Massalha et al. 2015a; Sabbah et al. 2016). These WW streams are treated mainly by anaerobic biological treatment that produces biogas. This unique technology was tested for increasing process stability and enhancing the efficiency of methane production within the Karmiel municipal WWTP treating agro-industrial WW mixed with municipal WW (Massalha et al. 2015a). The novelty of this study lies in the treatment of OMW combined with municipal WW using stable anaerobic processes at a demonstration-scale plant, marking the first demonstration of the long-term performance of the high-rate AAT system.

Three water partners are involved in the Karmiel case study under the ULTIMATE project: an SME company, AgRobics Ltd (AGB); a research center, the Galilee Society – Institute of Applied Research (GSR); and the national water company of Israel, Mekorot (MEK), as a big industrial partner and end user. The goal of this symbiosis is to demonstrate the treatment of agro-industrial WW using an immobilized high-rate anaerobic system (AgRobics' AAT; see AAT specifications and dimensions in Table 1, prior to the conventional aerobic biological process [activated sludge]).

The Karmiel municipal WWTP (northern Israel) includes pretreatment, physical settling, biological treatment (activated sludge-based), and tertiary treatment (sand filtration). The plant faces problems due to shock loads caused by the illegal discharge of OMW during the harvest period, as well as WW from nearby slaughterhouses.

This immobilized high-rate AAT system protects the next downstream process of high-load organic matter (high shock COD) as a result of the discharge of agro-industrial WW into the municipal WW mainstream.

Hydrophilic polyurethane pre-polymer, ‘Hypol FHP 2002™’, was purchased from Dow Chemicals Company (‘Dow’). ‘Hypol FHP 2002™’ is a toluene diisocyanate pre-polymer. Technical grade chemicals were purchased from CARLO ERBA Reagents S.A.S. The preparation of the immobilized composite polyfoam containing wet anaerobic sludge was done according to the AgRobics© patent (Sabbah et al. 2016). The anaerobic granular biomass was collected from a well-operated UASB bioreactor treating WW of a citrus-based soft drink factory (PRIGAT) in Kibbutz Givat Haim, Israel. The total solids (TS) measurement of the sludge was around 100 g/kg and that of the volatile solids (VS) was around 70 g/kg. To produce hydrophilic polyurethane foam, the pre-polymer was heated to 40°C in a water bath before mixing it with the aqueous solutions. The amounts of pre-polymer, water, and dry sludge were 15, 24, and 18 kg, respectively. The solution was vigorously mixed for 20 s using a mechanical stirrer and then poured into a rectangular chamber. The foam was left for 40 min to reach its maximal strength before the mold was removed. The foam was cut into ropes and then inserted into the pilot scale AAT (21 m³ of net anaerobic volume).

The biogas flow rate was measured continually by the gas flow meter of VORTAB (ST75 Series Mass Flow Meters). Samples of the produced biogas were collected four times a week into 1 L RESTEK^® (USA) gas sampling bags connected to the upper outlet of the AAT reactor by a gas tube. Methane (CH₄) content was analyzed in the lab using a gas chromatography GC system (Agilent 7890B Gas Chromatograph), combined with a thermal conductivity detector, equipped with Agilent DB-1ht GC Column, part number: 122–1,111, 15 m length, 0.25 mm inner diameter, 0.1 μm film thickness, with nitrogen used as the carrier gas. The column temperature was fixed at 150°C, while the injector and detector were set at 200°C, and the volume of the injected biogas sample was 250 μL.

Grab samples of 2 L were collected manually every 3–4 days from the WW influent, the OMW influent, the effluent of the mixing tank, and the effluent of the AAT and were analyzed for COD, TS, and VS according to Standard Methods for the examination of water and WW (APHA, AWWA & WEF 2005). The concentration of volatile fatty acids (VFAs) was analyzed using a combination of the potentiometric titration methods for acidity (Method 2310B) and alkalinity (Method 2320B), according to the Standard Methods for the examination of water and WW (APHA, AWWA & WEF 2005). The total alkalinity was calculated using the amount of acid needed to titrate the sample from the starting pH down to pH 4, and the volatile acids were calculated using the amount of hydroxide needed to titrate the sample from pH 4 back to pH 7 (APHA, AWWA & WEF 2005). pH was measured using a combo water tester (MRC Labs IP67, Israel).

Total polyphenols content was analyzed colourmetrically following the procedure of Davies et al. (2004) and Box (1983). In brief, a 1 mL water sample and 0.5 mL of the Folin–Ciocalteu reagent were placed in a 10 mL centrifuge tube. After 5 min, 3 mL of a saturated sodium carbonate solution and 5.5 mL of water were added. The resulting solution was centrifuged for 10 min at 3,000 rpm, and the supernatant's absorbance was measured at 725 nm using an ultraviolet (UV)–vis spectrophotometer (Thermo Scientific Evolution 260 Bio). Quantification of the total phenol concentration from obtained absorbance was done using a calibration curve for tannic acid (at concentrations of 0–25 mg/L).

The methane yield was determined based on the mass balance of total COD and solid removal. It accounted for the total methane observed in the daily collected biogas and the dissolved methane, which averaged 31% of the total methane produced during the entire operation of the AAT system, as measured using the method described by Souza et al. (2011). Considering a total evacuation of 3–4 m³/week of sludge with an average TSS of 13,250 mg/L leads to 7.57 kg/day of COD as excess daily sludge produced by the system. Additionally, 14% of the removed COD was assumed to be attributed to biomass growth. This mass balance indicated that an average of 78.5% of the total removed COD could be considered as net COD available for methane production.

No effective treatment or management solution has yet been developed for upstream, on-site treatment of OMW that meets the criteria of technical feasibility, economic viability, and social acceptability. As a result, this challenging WW is often discharged without proper treatment in the worst cases or spread on land, where feasible, as a relatively better alternative.

In this study, we simulated the scenario of illegal discharge of agro-industrial WW and demonstrated the capability of the AAT system to treat such WW, characterized by high organic loads, fluctuations in organic loading rates (OLR), and the presence of inhibitory substances, under controlled conditions. The anaerobic high-rate reactor utilizes a polymer-based foam to maintain stable microbial activity by preventing microorganism washout, ensuring a robust biogas production process. The AAT pilot system processes 100 m³/day of WW. During operation, 0.5% OMW is mixed with 99.5% municipal WW in an equalization tank, and the resulting mixture is directed to the AAT reactor (for experiments labeled ‘with OMW’). The mixed stream undergoes high-rate anaerobic treatment in the AAT pilot with an HRT of approximately 5 h, leading to biogas production.

Figure 2(a) shows that the influent total (COD_T) increased when OMWs combined with municipal WW; however, the effluent total (COD_T) is stable at a value of about 900 mg/L. It is clearly shown that the high organic peaks observed as a result of the mixed OMW were shaved and eliminated; this demonstrates a future potential for protecting WWTPs from sudden agro-industrial discharges. Moreover, during the summer period (hot weather with an average temperature of around 30 °C), a COD removal efficiency of 46% was achieved (Figure 2(b)) for an inlet COD_in of 2,280 mg/L (WW + OMW), while a removal efficiency of 49% was achieved (Figure 2(b)) for a COD_in of 1,430 mg/L (without OMW). In the winter period (warm period with average temperatures of 19–20°C), a COD removal efficiency of 47% was obtained for a COD_in of 1,770 mg/L (WW + OMW), while the removal efficiency was slightly decreased to about 36% (Figure 2(b)) for a COD_in of 1,088 mg/L (without OMW). In general, it can be seen from Figure 2(b) that, during the summer period, the removal efficiencies' range was 32–67%, while it was 16–67% in the winter period. The soluble COD_S was lower than the total COD_T, 33% of COD_T for WW, and 43% of COD_T for WW + OMW (Figure 2(c)). The removal efficiencies were lower as well (Figure 2(d)).

A similar trend was also observed for the soluble COD, where the effluent COD_s was at an average value of about 460 mg/L (results not shown). These results are in line with the results of other mesophilic UASB pilot systems treating diluted olive pomace leachates working at an HRT of 3 days, and OLR between 0.33 and 1.67 g COD/(L day) (Katsoni et al. 2014). It is important to emphasize that these OLR levels are much lower than the much more challenging OLRs applied in this current study, as will be shown later on, where the COD removal efficiencies ranged between 35 and 70%.

Figure 3 shows the AAT performance efficiency of total COD removal as a function of the OLR during the performance period. As can be seen from this figure, the OLR with added OMW was higher than without OMW in the feed. The addition of OMW at 0.5% has increased the OLR to 9–11 kg COD m⁻³ per day, where the corresponding COD removal was between 40 and 55%. It is important to mention that these applied OLR values are very high for municipal WW systems, where they are comparable to operational conditions of the challenging high-rate anaerobic treatment of agro-industrial WW. The main outcome of this long operation period (four seasons) shows the removal efficiency was relatively stable throughout the operation of AAT. A slight decrease in the removal efficiency was observed in days 133–183, which may be mainly due to relatively low temperatures (around 20 °C).

Anaerobic biodegradation results in biogas production, which is an important aspect of energy recovery. Figure 4 presents biogas production (m³/day) throughout the AAT's operation time; the average OLR of each operational period is also presented in the upper part of the figure. From the figure, it is clear that the biogas production was highly dependent on both temperature and OLR. In all cases, the addition of OMW resulted in a 10% increase in biogas production. In particular, the first two operational periods were the best for comparison, as they were operated under almost the same temperature (30 °C). In these periods, the addition of OMW resulted in an average increase of about 10%; from 8.7 m³/day (municipal WW without OMW) to 9.6 m³/day when OMW was added. This increase corresponds to the change in OLR, where the high OLR of 9.3 kg/m³ day (municipal WW mixed with OMW) resulted in a higher biogas production rate than the lower OLR of 6.57 kg/m³ day (without OMW). A similar increase in the biogas production rate was also observed for the period at a low temperature of 20 °C (days 133–233), where a biogas production rate of 3.7 m³/day was obtained with the addition of OMW (OLR = 8.2 kg/m³ day) compared to 3.4 m³/day without the addition of OMW (OLR = 5 kg/m³ day). Comparison of the biogas production at the different temperature ranges indicates the strong effect of temperature on biogas production (Tsagaraki et al. 2007), as biogas production was two and a half times higher at 30 °C than at 19.5 °C despite comparable OLR of 9.3 kg/m³ day and 8.2 kg/m³ day. Similar results were obtained in the study by Maragkaki et al. (2017), in which a pilot-scale anaerobic co-digestion process treated sewage sludge with 5% (v/v) OMW under mesophilic conditions (35 °C) at an HRT of 24 days (low rate). Their results demonstrated a significant increase in biogas production, nearly 220%.

The methane production rate over OLR as a function of time and temperature is presented in Figure 5. In this figure, it is clear that the normalized methane per OLR was lower when applying OMW relative to the scenario without OMW addition at a similar temperature. This can be attributed to the inhibiting effect of the OMW constituents (i.e., polyphenols and tannins) (Dhouib et al. 2006). Regardless of the different operational conditions, it can be clearly seen that methane was produced throughout the AAT operation. The concept of combining the treatment of municipal WW and agro-industrial effluents (i.e., olive mill effluents) has been previously proposed elsewhere (Boari & Mancini 1990; Croce et al. 1994). However, in this study, a demonstration-scale plant was conducted for methane production with the anaerobic pretreatment of municipal WW combined with olive mill WW.

Figure 6 illustrates that the methane yield ranged from 0.07 to 0.22 m³ CH₄/(kg COD_T) over a total operational period of more than 400 days under varying temperatures and OMW addition conditions. During the initial period, under a high OLR of 9.3 kg COD/(m³ day) and a temperature of 30 °C, the methane yield was 0.17 ± 0.04 m³ CH₄/(kg COD_T) when municipal WW was mixed with fresh OMW, compared to 0.22 ± 0.06 m³ CH₄/(kg COD_T) without OMW addition. Both operations in the first 50 days occurred in summer at an average temperature of 30 °C but under different OLRs. The reduced methane yield with OMW addition suggests an inhibitory effect caused by either the high OLR (Guo et al. 2022) or the intrinsic characteristics of OMW (Pluschke et al. 2023).

This trend was also observed during the colder period. In run #2, between days 133 and 175 (WW only), the methane yield was 0.14 ± 0.06 m³ CH₄/(kg COD_T), while during days 193–225 (WW mixed with OMW), it dropped to 0.07 ± 0.02 m³ CH₄/(kg COD_T), despite similar average temperatures of 18–20 °C. These findings support the hypothesis of a relative inhibitory effect during colder temperatures (Bodik et al. 2000). The reduced methane yield in colder conditions can be attributed to lower anaerobic activity at 20 °C compared to 30 °C, where COD removal was likely dominated by physical sedimentation rather than biological processes. The study of de Oliveira et al. (2024) shows that temperature fluctuations significantly affect UASB reactor performance for swine WW treatment. Winter temperatures below 15 °C reduce methanogenic capacity, while mesophilic temperatures (∼25.6 °C) enhance organic matter degradation and methane production (de Oliveira et al. 2024). Another study by Kumar et al. (2020) investigated the co-digestion of waste microalgal biomass with cattle dung in a pilot-scale reactor and reported higher volumetric and specific biogas yields during summer (0.45–0.65 m³/day and 0.72–1.04 m³/kg VS fed/day, respectively) compared to winter (0.06–0.21 m³/day and 0.096–0.336 m³/kg VS fed/day). The methane content of the biogas remained between 55 and 65% across seasons (Kumar et al. 2020).

In the final two summer operational periods, insufficient sampling points limited the ability to determine a clear effect of temperature on methane yield.

Figure 7 presents the methane production rate as a function of OLR and temperature. This 3D presentation visually illustrates the relative combined impact of OLR and temperature on methane production. As can be seen in the figure, higher methane yields were achieved at higher temperatures regardless of the changes in OLR. As anticipated, OLR has a much lower effect on methane production than the temperature in an anaerobic process. It is clear that, for the same temperatures, higher OLRs did not lead to increased methane production.

The polyphenol concentrations during the AAT operation are shown in Figure 8(a). It is clear that the OMW discharge increased the polyphenols concentration in the inlet (45 mg/L on average). However, the AAT was able to tolerate these concentrations, achieving removal efficiencies ranging between 18 and 45%. Anaerobic treatment of olive oil mill effluent in UASB reactors resulted in a reduction of simple phenolic compounds, according to the study by Angelidaki et al. (2002).

Another important aspect of achieving a ‘healthy’ anaerobic digestion process is pH. The ideal pH for anaerobic digestion is 6.8–7.2, while the process can be functional without inhibition at pH values of 6.5–8 (Ward et al. 2008; Yu & Wensel 2013). As shown in Figure 8(b), the AAT effluent pH generally remained within these permissible values. Specifically, the pH dropped to 6.6 when OMW was added to the system. Therefore, only 0.5 m³/day of OMW per 100 m³/day of municipal WW was applied to avoid rapid acidification. It is important to indicate that the VFA levels remained very low in the AAT outlet. This demonstrates that the methanogenic stage was not disturbed, and the methane was easily formed from these intermediates.

No appropriate treatment or management solution has yet been developed for upstream and on-site olive mill WW, which can be considered technically feasible, economically viable, and socially acceptable. Thus, most of this challenging WW is discharged without adequate treatment, in the worst-case scenarios, or through land spreading, if available, in the relatively better scenarios.

In this current study, we simulated the scenario of the illegal discharge of agro-industrial WW and showed that OMW mixed with municipal WW could be successfully treated and partially biodegraded without harming the downstream process in any WWTP. However, this can be achieved under limited and controlled conditions. More specifically, we have shown in this study that, when our developed AAT technology is applied and operates well, the system can tolerate the addition of up to 0.5 m³/day of OMW per 100 m³/day of municipal WW (0.5% OMW). The system shows a COD reduction of about 50%, a polyphenol reduction of 18–47% with a stable performance, as well as a decent increase in biogas production.

Moreover, AAT enabled treating high organic loads (9.3 kg COD m⁻³ day⁻¹) resulting from OMW discharge by shaving the high peaks of the organic content, leading to protection of the subsequent activated sludge process. The AAT system allows the recovery of a significant part of the organic load through the production of biogas, the main product of anaerobic biodegradation, which was shown to be highly dependent on temperature and partly on the OLR. Additionally, it was also shown that the addition of OMW reduces methane yield; The methane yield dropped to 0.17 ± 0.04 m³ CH₄/(kg COD_T) with OMW addition during warm conditions (30 °C), compared to 0.22 ± 0.06 m³ CH₄/(kg COD_T) without OMW. During cold conditions (18–20 °C), the yield decreased to 0.07 ± 0.02 m³ CH₄/(kg COD_T) with OMW, compared to 0.14 ± 0.06 m³ CH₄/(kg COD_T) without OMW.

This work shows that the AAT system has the potential for pretreating municipal WW, increasing the energy efficiency of the plant, and protecting small-medium WWTPs from sudden agro-industrial discharges.

This paper has been prepared and the research was carried out with the financial assistance of the European Union's Horizon 2020 research and innovation programme under grant agreement No. 869318 as part of the circular economy called CE-SC5-04-2019. The authors would also like to thank Khalid Farah for reviewing the manuscript and for contributing his valuable comments.

This research was funded by the European Union's Horizon 2020 research and innovation programme under grant agreement No. 869318 as part of the circular economy called CE-SC5-04-2019.

I.S. and K.B.K. conceptualized and acquired funding for the study, wrote, reviewed, and edited the article. M.H. rendered support in data curation, data analyzing and validation. H.R.K. wrote, reviewed, and edited the article, validated the project, developed the methodology. N.M. reviewed and edited the article. All authors have read and agreed to the published version of the manuscript.

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

The authors declare there is no conflict.