Jet-loop reactor with cross-flow ultrafiltration membrane system for treatment of olive mill wastewater

The olive oil industry is very important in Mediterranean countries, both in terms of wealth and tradition and it is considered to be as one of the driving sectors of the agricultural economy of the Mediterranean basin (Agapiou et al. 2016). Every year, worldwide olive oil production generates in a short period (from October to March) more than 30 million m³ of olive mill wastewater (OMW) (Belaqziz et al. 2016) causing disposal problems with great negative environmental impact. The major concerns associated with OMW disposal are mainly its high organic content in terms of chemical oxygen demand (80–200 g·L⁻¹, COD), high content of polyphenols (3–12 g·L⁻¹), and acidic pH.

Currently in Jordan, direct OMW disposal is not allowed to avoid contamination of the soil and water resources nor even to discharge it to the municipal wastewater treatment plants (Rusan et al. 2015). Instead, untreated OMW is disposed in dumping sites (Rusan & Malkawi 2016). The reuse of treated OMW should be considered as an additional water source and an important solution from an economical and environmental point of view. The need for appropriate management and treatment of OMW remains the largest single environmental problem in Mediterranean olive oil extraction countries and the development of effective approaches for the treatment of OMW is thus of crucial importance.

Aerobic biological treatment systems could become an interesting alternative due to their fast process kinetics and high removal rates. The use of systems with air jet in the biological treatment of wastewaters produces high turbulent mixing which is responsible for the optimal oxygen and biomass transfer and hence a good biological conversion capacity. Several authors have reported the efficiency of jet aeration systems in the treatment of agro-industrial wastewaters (Bloor et al. 1995; Eusébio et al. 2004; Jail et al. 2010; Wagh et al. 2012). Air is sucked naturally from the environment under the influence of vacuum induced inside nozzle. Proper aeration is one of the most important factors in aerobic wastewater treatment processes.

The aim of this work was to assess the technical feasibility of the aerobic treatment of OMW in the pilot JACTO reactor previously tested (Eusébio et al. 2007) after some modifications: the settler was removed and a cross-flow ultra-filtration (UF) membrane (MBR) unit was incorporated in the system (JACTO.MBR). A JACTO.MBR_1000 L was constructed and installed at JUST for scale-up studies. Moreover, if OMW was properly treated and managed, it can be used as a source of water and nutrients essential to the plants and to the fertility of the soil. Therefore, this study also aimed to evaluate the reuse of treated OMW in irrigation and its impact on seed germination, plant growth and soil fertility and quality parameters through germination and greenhouse phytotoxicity tests.

Raw OMW samples were collected from three-phase olive oil mills, during two olive oil production campaigns during 2012/2013 and 2013/2014, located in the south region of Portugal (Tavira, Algarve) and in the north region of Jordan (Irbid). Samples were pre-filtered with cloth gauze to remove floating solids and were immediately analysed. All samples are kept at 4°C until further use. OMW characterization (Table 1) was performed in the subsequent 48 h after its collection.

JACTO technology is a jet aeration system that combines efficient oxygen transfer with high turbulent mixing to be used for biological treatment of wastewaters. In Portugal, experiments were carried out using the JACTO.MBR, with a working volume of 100 L coupled with a cross-flow UF membrane system, upgraded from a previous reactor (Eusébio et al. 2007). This allows that concentrated biomass recirculates in the system, avoiding biomass wash out, and the treated effluent was collected from UF membrane permeate (cut-off 300 kDa). In Jordan, JACTO.MBR was scaled-up to 1,000 L with vertical configuration and with interior compartments to include PVDF type flat sheet MBR modules. The feed, permeate extraction and the sludge recirculation pumps can operated at 500–2,000 L·d⁻¹ (Figure 1).

The aerobic biological treatment of OMW was carried out under good conditions of aeration rate of 0.33 vvm and turbulent mixing. JACTO.MBR_100 L was tested under batch conditions and different hydraulic retention time (HRT) of 40, 9, 6 and 3 days, with organic loading rates (OLR) of 1, 7, 10, 12, 14 and 18 kg COD·m⁻³·d⁻¹. JACTO.MBR_1000 L was tested under batch conditions at the second milling campaign. Samples of treated effluent (UF permeate) were continuously collected during the OMW treatment processes for monitoring the physicochemical parameters.

Oxilyser and pHlyser II probes (S::CAN, Austria) were used in order to daily monitor the temperature and the dissolved oxygen by fluorescence and the pH, respectively. All data acquisition and registration of signals were performed on-line using a data logger with a control terminal Conlyte4 (S::CAN, Austria).

Membrane washing was performed with a cleaning agent like a solution of ULTRASIL 11 (5 g·L⁻¹) for 30 min, while water was also used (5 L, 4 times), to ensure the membrane cleaning. The washing procedure was carried out at the end of each experimental run or in case of membrane clogging.

Samples from raw and treated OMW were collected under aseptic conditions and used to perform microbial counts as colony forming units (CFU) using the spread plate method, as described in Standard Methods (APHA 9215C, 2001). Bacterial, yeast and fungal isolation was carried out either on Nutrient Agar (NA, Difco), Yeast Malt Agar (YMA, Difco) or Potato Dextrose Agar (PDA, Difco), incubated at 30°C.

A miniaturized growth inhibition test in microtitration plates using a culture of a bacterium, Pseudomonas putida (MIGULA DSM 50026), was performed according to ISO 10712 (1995), but adapted to the microtitration plates. Reliable concentration-response relationships were determined with EC₅₀, which represents the concentration of a compound where 50% of the population exhibits a response, after a specified exposure duration, and with IC₅₀, which is the half maximal inhibitory concentration. The statistical programme used for the calculation of these values (EC₅₀-16 h and IC₅₀-16 h), Maximum Likelihood – Logit method, was the ToxCalcTM V5.0.23F (TIDEPOOL Scientific Software). Samples of raw OMW and UF permeates collected from JACTO.MBR_100 L reactor were tested: permeates ‘a’ and ‘b’ obtained at OLR 10 and 14 kg COD·m⁻³·d⁻¹, respectively.

UF-permeate from JACTO.MBR_100 L (treated OMW) was evaluated as a source of irrigation and its impact on seed germination, plant growth and soil fertility and quality parameters through seed germination and greenhouse phytotoxicity experiments. The UF-permeate composition is shown in Table 2. Germination tests of barley (Hordeum vulgare) seeds were done in an incubator at 20°C. Ten seeds of barley were placed on the filter paper (Whatman #40) and mounted in Petri dishes. Seed germination was determined by counting the number of germinated seeds out of 10, as follows: Germination (%) = [(# of germinated seeds)/10]*100. Greenhouse phytotoxicity tests were conducted in a greenhouse at 23°C, using a calcareous soil with a low organic matter content classified as fine-loamy, mixed, thermic, calcic Paleargid (Khresat et al. 1998) and collected from the Research Center of JUST. Each pot was filled with 5 kg of dried soil. Three maize seeds per pot were seeded. Pots were watered periodically to maintain approximate field capacity water content. The plant heights were recorded once a week during the growing period. Experiments were done comparing different conditions: tap water (W), OMW treated by JACTO reactor (J), raw OMW (100%) and diluted OMW (25–75%).

COD, Total Suspended Solids (TSS) and Electrochemical Conductivity (EC) were measured according to Standard Methods (APHA 2001). The nitrates determination was performed by the method ‘Nitrate Cell Test 1.14542’ (test kits Spectroquant Merck) in a HACH DR/2010 spectrophotometer. A commercial kit for phosphorous determination has been utilised (Also called Orthophosphate) Phosver 3 (Ascorbic Acid) Method using Powder Pillows (HACH DR/2010). Total phenols (TP) content was determined by a modified Folin-Ciocalteau colorimetric method (Singleton & Rossi 1965) and expressed as caffeic acid equivalents.

The technical feasibility of the aerobic treatment of OMW at pilot-scale JACTO reactor with ultrafiltration membrane system (JACTO.MBR_100 L) was assessed. Dissolved oxygen (% of saturation), temperature (°C), and pH were measured on-line and registered daily. Values obtained for each tested parameter are presented in Figure 2.

A good aeration ability of the reactor was observed, since high values of dissolved oxygen (up to 70% of the saturation) were achieved. Temperature self-maintained around 40°C and pH increases to values above 6.0, which is probably due to organic acids degradation.

HRT is the most important parameter to evaluate the performance of JACTO.MBR. The main objective was to decrease HRT and simultaneously maintaining the performance of the reactor, mainly reaching high removal efficiencies of COD and TP. The results obtained during the 1^st and 2^nd milling campaigns are shown in Table 3.

In batch conditions high percentage of TP removal was observed, but COD removal was lower than expected. This might be due to a still deficient adaptation of the microorganisms to the medium. For HRT 40 d the COD removal increased while TP remained constant, revealing that the aerobic microorganisms were possibly well adapted. Based on this assumption, the organic load was increased and the results showed that the lower HRT tested (9 d) and the higher OLR (7 kg COD·m⁻³d⁻¹) resulted in the maximum removal rates for COD (72%) and TP (85%). In the 2^nd milling campaign, a decrease in the HRT (6 d) allowed to increase gradually OLR up to 14 kg COD·m⁻³d⁻¹ achieving COD and TP removals of 88% and 79%, respectively. This is a very good result, since the objective was to decrease HRT, and increasing OLR with high COD and TP removal rates. However, when decreasing to HRT 3 d, the performance of the reactor was negatively affected. This fact was attributed to low F/M ratio: high biomass content inside the reactor and low substrate availability. This situation could be avoided by purging the reactor periodically and guaranteeing that the concentration of biomass was around 10 g L⁻¹. The evaluation of pH is another important parameter for irrigation purposes. Raw OMW is an acidic effluent with pH value less than 5. During the OMW treatment with the reactor JACTO.MBR, pH in UF-permeate increased to values higher than 6. According to Portuguese legislation, this permeate can be discharged to WWTPs or used for irrigation purposes.

Preliminary results obtained for the scale-up of the JACTO.MBR_1000 L reactor have shown that COD and TP removal rates were around 85% and 92%, respectively, under batch conditions. These results seem to be very promising for continuous feeding regimens.

Viable microorganism populations were evaluated as microbial counts in samples collected at the beginning and at the end of OMW treatment from the pilot reactors (Table 4). At the beginning of the OMW treatment, it seems that high diversity of microorganisms is found in OMW samples used to feed the reactor in terms of aerobic total bacteria, yeasts and moulds plate counting. During the aerobic treatment process, the conditions developed favour the growth of aerobic bacteria and simultaneously yeasts are slightly affected. However, the hydraulic shear stress imposed by the reactor aeration system seems to induce a selective pressure in the composition of the microbiota and may be responsible for the elimination of fungi and filamentous bacteria, since moulds were not detected, as also observed by Eusébio et al. (2007). During the OMW treatment in pilot JACTO.MBR_100 L, only gram-negative bacteria were detected. The presence of isolates belonged to the genus Pseudomonas, a genus well known for the ability of many of its species to degrade complex molecules (e.g. polyphenols).

This microbial community is according with the microbiota active of the OMW and so in the organic matter decaying and they can play a role in the organic components degradation leading to a natural transformation of organic components to minerals.

The results of the ecotoxicological evaluation of the OMW samples, before and after the treatment by JACTO.MBR (UF-permeate samples), using the Pseudomonas putida (MIGULA DSM 50026) growth inhibition test (ISO 10712 1995 but adapted to microtitration plates) are summarized in Table 5. The Pseudomonas putida was selected as the test organism because it is a ubiquitous organism in the environment (soil and water ecosystems). The IC₅₀s values showed that the toxicity levels are dependent on the samples tested and increased one order of magnitude in the treated samples (UF-permeates), indicating thus an effective toxicity reduction. In fact, the OMW treatment by JACTO.MBR was able to reduce the raw effluent toxicity by 69%, producing UF-permeate samples (a and b) with a similar behaviour. These results highlight that UF-permeate from JACTO.MBR technology can be safely used for soil irrigation purposes, in slightly controlled conditions.

Test of barley seed germination under different conditions is shown in Figure 3. It can be seen that the raw OMW has a phytotoxic effect and prevents germination. No germination was observed with raw OMW even at dilutions higher than 25%. It has been reported that 25% dilution of raw OMW significantly improved plant growth grown in the soil (Rusan & Malkawi 2016). However, such dilution prevented seed germination as shown in current study. On the other hand, it was observed that the phytotoxicity of OMW was eliminated using treated OMW (UF-permeate) from JACTO.MBR and seeds were able to grow. Plant growth under different conditions is shown in Figure 4.

Unlike in the germination test, raw OMW did not prevent plant growth. Treated OMW (UF-permeate 1) resulted in positive effect on plant growth by eliminating the phytotoxicity effect of OMW, similar to the results obtained for the non-stressed conditions, W and W + F. Results obtained for UF-permeate 2 show that the operating conditions for JACTO.MBR_1000 L must be further optimized. This technology shows promising results and seems to be suitable also for germination and irrigation purposes.

Table 6 shows the limit values legislated in Portugal and Jordan for irrigation on soils and for effluents discharge to municipal wastewater treatment plants (WWTP). UF-permeate complies with the Portuguese legislation (Dec-Law 236/98) for irrigation purposes. To fulfil the legislation for irrigation in Jordan or for discharging to WWTP (e.g. Portugal), an additional membrane system, nanofiltration or reverse osmose should be suggested after the biological treatment process, in order to decrease COD values. This facilitates obtaining concentrates rich in polyphenols for valorisation and a clean water to be used for irrigation or other purposes.

The overall maintenance and operating costs of the JACTO.MBR technology for the treatment of 50 m³ of OMW per day were calculated for a period of 10 years (Ioannou-Ttofa et al. 2017). The estimated value obtained for the reactor reached on an average annual level of 9700 € for Portugal and 7250 € for Jordan, based on personnel salaries, electrical energy, chemicals and consumables, which is different for each country. The JACTO.MBR process was found to be highly efficient in removing the organic content of the raw OMW, with a quite low operational cost.

The technical feasibility of the aerobic treatment of OMW at a pilot-scale JACTO reactor with UF membrane system (JACTO.MBR) was assessed. The main operational parameter HRT that is related with the OLR to feed the reactor was studied. Under continuous feeding, the best efficiency of the OMW treatment process was 88% and 79% for COD and TP removal, respectively, achieved at tested HRT 6 d and OLR up to 14 kg COD·m⁻³d⁻¹.

In terms of ecotoxicity tests, IC50 s values showed an effective toxicity reduction in UF-permeate, reducing the raw effluent toxicity by 69%.

Treated OMW (UF-permeate 1) resulted in positive effect on plant growth by eliminating the phytotoxicity effect of OMW, similar to the results obtained for the non-stressed conditions, W and W + F. Results obtained for UF-permeate 2 show that the operating conditions for JACTO.MBR_1000 L must be further optimized.

Experience shows that using a biological treatment, such as JACTO.MBR, there is no need for initial or further dilutions of OMW, which means water saving. Another advantage is that the neutral pH obtained in the UF-permeate after the treatment saves chemicals. Moreover, the chemical composition of the UF-permeate, allows further applications, such as soil irrigation or discharge to WWTPs. Furthermore, the experimental data obtained during this study implies that JACTO.MBR technology can be applied to small-scale olive mills.

This work was prepared in the framework of the project ‘Mediterranean Cooperation in the Treatment and Valorisation of Olive Mill Wastewater (MEDOLICO)’ which is funded by the European Union under the “ENPI Cross-Border Cooperation Mediterranean Sea Basin Programme″. MEDOLICO total budget is 1.9 million Euro and it is co-financed through the European Neighbourhood and Partnership Instrument (90%) and national funds of the countries participating in the project (10%). The authors wish to thank Céu Penedo for technical laboratory assistance.