![[NF-2000 Database - AIR Program]](../images/Air.gif) |
AIR3-CT94-1987
Development of A Biological Integrated Process for Purifying Olive Oil Waste Water Recovering Energy and Producing Alcohol |
 |
Proposal No: | AIR3-CT94-1987 |
| Date Prepared: | April 1998 | |
| Source: | Final Consolidated Report Summary |
Introduction
The overall objective of the project was to
design, realise and test in a pilot plant, a combined biological process to
purify waste waters that originate from industrial olive milling. This was to
combine a significant reduction in the level of pollutants with energy recovery
in the form of methane and/or ethanol production using an integrated biological
approach.
Olive oil production, one of the oldest agricultural industries, occurs throughout the Mediterranean region and is of economic importance for many countries. The yearly world olive oil production is estimated at around 1.5 to 1.7 million ton/year. Over 95% of the olive trees (714 millions) and over 97% of the agricultural area dedicated to growing olives are in the Mediterranean area. Olive oil production in the EU countries represents 75% of the world production. The main producers are Italy, Spain, Portugal, Greece, and Tunisia. In Greece the mean annual olive oil production is around 300,000 annual tons, in Portugal it varies between 100,000 and 500,000 tons per year and in Italy production of olives vary seasonally from 2.5 to 3.4 million tons, at an average yield of 2.2 to 3.1 tons/hectare. In 1992 olive oil production in Italy was estimated at around 640.000 tons.
In Greece over 70 percent of the olive milling plants have a capacity of more than 1,000 tons per day, while 14% have a capacity between 1,000 tons and 500 tons over an 8 hour shift, with slightly more using modern centrifugal plant than use traditional presses. In Portugal olive oil extraction is still mainly carried out using the traditional discontinuous press process, although over the last few years several units have introduced continuous solid-liquid centrifugation systems. In Italy, over 10,000 olive milling plants are operating with the most common extraction technology still based on simple pressure. This results in large amounts of effluent (olive mill wastewater or OMW).In Greece every year 200.000 to 250.000 m3 of OMW are produced, in Portugal 60,000-350.000 m3 of OMW and in Italy up to 1.5 millions m3 of OMW are produced, depending on the processing cycle, the type of plant and the olive production in a given year.
Summary
For two years research was carried out within the olive
oil sector in three countries (Italy, Greece, Portugal). Work covered olive oil
extraction technologies, OMW production and OMW treatment technology. Analytical
methods (physico chemical, microbiological and toxicological) were developed and
used to characterise OMW. For ethanol fermentation, yeast strains able to
degrade the sugars present in OMW, with a conversion efficiency near to
theoretical values, were selected. However the concentration of sugars in OMW
observed was too low and hence the production of ethanol by fermentation was of
no practical or economic interest. In contrast, anaerobic treatment was
promising.
The results obtained with a 17 litre laboratory reactor gave higher rates of production of methane and COD removal than reported in the literature. This process, when transferred to pilot plant reactors gave good yields during the first phase of the experiment until a dramatic frost stopped the reactor running.
However, the analytical results indicated that using undiluted OMW it was possible to get similar results in the pilot plant as achieved in the laboratory. Hence, the process could be transferred to full scale if the following conditions were fulfilled:
Energy analysis showed that the highest net gas production was obtained using untreated OMWs with intermediate organic loads (5 to 10 g COD/l dig./day) in a reactor in which a heat exchanger was installed. In the case of OMWs with a COD of 30 g/l this corresponds to retention times of 3 to 6 days. The maximum net production of gas were in the order of 0.4 to 0.7 m3/m3 digester per day in unfavourable climate conditions, and 0.85 to 1.15 m3/m3 digester per day in favourable climate conditions. The lowest values were obtained for plants of around 50 m3.
Although there are many small olive mills in the south of Europe, over 90% of the of the OMW is produced by a few very large plants. The size of these mills is much greater than the minimum required to show an economic benefit from applying the anaerobic digestion technology developed. Hence, it is suggested that the technology is restricted to these big mills.
It was also found that the yield of barley grown in soil irrigated with effluent from the anaerobic plant effluent was higher than that in the control, suggesting that the effluent contained a useful amount of organic matter as well as micro and macronutrients. The plant effluent can therefore be disposed of by spreading on land without harmful environmental effects and with positive effects on crops grown in the treated soil, after one month from application.
These results are of particular interest since a new law (number 574) was introduced in Italy in November 1996 covering the agronomical utilisation of olive oil mill waste water and husks. This law permits the spreading of OMW on the soil at the ratio of 50 m3/ha year if they come from a discontinuous (traditional) extraction process and of 80 m3/ha year if from a modern continuous process.
The results of this work would fit with the official way suggested for the agronomic utilisation of the effluents. The effluent obtained after anaerobic treatment could possibly be spread at higher concentrations than untreated OMW. This would be possible due to the lower phenol content and stabilisation of the treated OMW from anaerobic treatment that has the advantages of methane production and recovery as a source of energy as well as the production of a stabilised and detoxified effluent that can be used as a fertiliser in agriculture.
Results
In order to reach the main objective, preliminary
investigations were carried out to establish physico-chemical and
microbiological methods and to use these to characterise OMW. In addition,
anaerobic cultures able to produce methane from OMW, yeast strains for ethanol
fermentation and aerobic cultures able to degrade phenols were selected. A
laboratory scale anaerobic digester was set up in order to produce the
methanogenic microbial biomass to be used to inoculate the pilot plant. Other
activities included the design, engineering, construction and optimisation of
the pilot plant. Finally the energy balance and economics of the pilot plant
were calculated in order to evaluate the potential for full scale application of
the process.
Analytical methods and results Chemical and microbiological methods were set up for determination of COD, BOD, Total, solids, Dissolved solids, Total phenols, Individual phenols, Total sugars, Reducing sugars, Tannins and Lignins, Total fats, Individual fatty acids, Total phosphorous, Total nitrogen, Ash. Methods for enrichment cultures, microorganisms selection, testing of the biodegradative activity, isolation of pure cultures and identification of microorganisms were also set up.
Particularly attention was paid to analysis of phenolic compounds. This was an important analyte, for which analytical methods are laborious, providing different answers with different methods. Various methods were investigated including the Ciocalteau method modified in various ways. Polyphenols were extracted in ethylacetate, after fat removal using hexane at 80°C, and determined by spectrophotometry. readings. The phenol concentration of OMW using the first methods was in the range of 1.8-2.8 g/l, while the concentrations found with the second method was four or five times higher. The observed phenol concentration did not change after protein precipitation. Extraction of polyphenols in acetonitrile and methanol after removal of the fats and HPLC analysis showed a concentration of phenols of about 0.6 g/l. Phenol determination performed on the same sample by method 1) showed a concentration of about 6 higher than that determined by HPLC. Test for specific phenols, that have been reported by other authors (tirosol, vanillic, protocatechuic, caffeic acid), detected only low amounts in some samples. The level of free fatty acids was quite high (2 to 7 g/l OMW), with C2 to C8 compounds coming from microbial metabolism and C16 to C18 compounds coming from the olive oil.
A large number of chemical analyses were carried in order to characterise OMW and provide a comparison between OMW obtained from the three Countries involved in the BIOWARE project. In addition results were obtained to enable comparison between that obtained by different extraction technologies, between that from two successive (1994-95 and 95-96) harvesting campaigns and OMW from different olive varieties. The results showed great variability and hence it was not possible to establish any relationship between OMW chemical composition and the olive variety, the agronomic and technical conditions, the kind of olive oil processing system or the climatic conditions under which the olives were grown. It was only possible to obtain an idea of the range of values for each parameter, many of which differed by more than an order of magnitude. This variation was seen both between samples from a similar source and between samples from different countries and processes.
The acute toxicity of OMW was investigated using various bioassays to determine EC50. These included Thamnotoxkit F (24h), Daphnia (48h), Microtox (5 min), Microtox (15 min) which confirmed the high toxicity of the OMW.
A high microbial diversity was also recorded for OMW. Among the strains identified were several species of Acinetobacter, Pseudomonas and Enterobacter. However, much of the microbial activity was represented by 71 strains, showing different metabolic patterns. In addition the pathogenic Klebsiella pneumoniae ss pneumoniae was also isolated from untreated and treated OMW.
Phenol degrading cultures A high number of resistant microorganisms (1000 to 10000 cfu/g) were detected in the soil used for enrichment cultures used to prepare a column to treat OMW. The effects of phenolics depended on the nature of the compound used. Caffeic acid (the only phenol amongst the standards used detected by HPLC in the OMW), was the most difficult to degrade and the most toxic. The most active degrading organisms were actinomycetes of the Nocardioforms group, with the best one being Gordona sp., which was isolated from soil, not OMW. Strains of Pseudomonas able to grow in OMW were also isolated.
Yeasts, both isolated from OMW and obtained from culture collections, were able to metabolise the sugars in OMW at a high rate in laboratory tests. This work was not transferred to pilot plant as the concentrations of sugars in the OMW were too low to make this worth while. Some of the yeasts showed a capacity to degrade phenols. This included a strain (Y101) able to degrade model compounds at concentrations as high as 5 g/l. This strain was tested for ability to degrade polyphenols in OMW in a laboratory reactor as was the fungus Pleurotus ostreatus. On 100 percent OMW, this organism reduced the phenol content by 78 percent, mainly due to laccase activity oxidising phenols to quinones.
Laboratory work on methanogenesis Various methanogenic cultures were tested with OMW, the most efficient were derived from a lab-scale filter, adapted to phenolic rich wastewaters. The efficiency of methanogenic biomass selected in a 15 litre upflow anaerobic filter was tested over a 10 month trial, using various operating conditions and loading rates. This was adapted to OMW over a period of three months using stored OMW derived from the 1994 olive harvest. This was diluted to obtain the required organic load of 2.5 g COD per litre reactor per day and operated with a hydraulic retention time of 48 hours, giving 85 percent COD removal, biogas of 75 percent methane and a daily methane production of 11 litres. Using fresh OMW from the harvest of 1995 the COD removal was 90 percent, with a specific biogas production of 0.45 l/g COD removed, containing 74 percent methane.
The volumetric organic load was progressively increased up to 20g COD/ litre reactor per day. Under all conditions the performance of the reactor was better than that previously reported in the literature. The removal of soluble COD was around 90% with an average value of 0.25 litre methane per g COD removed. The effectiveness of the process was indicated by the low volatile fatty acid content of the effluent, by the high cell count of the main microbial groups in the biofilm and the low content of F-420 in the effluent. The latter being a method used to monitor loss of methanogenic bacteria in the effluent, and hence an indicator of the stability of the process.
The populations of methanogens which grew under various conditions with selected phenols and other toxic compounds present in the OMW were studied using polymerase chain reaction (PCR) on samples of the biofilm covering the wood chip support used. This involved a simple and rapid boiling lysis for DNA release, without any purification or precipitation, followed by PCR. The quantitative detection of different groups of methanogens in the sample was accomplished in less than eight hours. Methanobrevibacter-like cells were more numerous than Methanobacterium. Among the acetoclastic methanogens Methanosaeta and Methanosarcina were high. The levels of Methanococcales and Methanomocrobiaceae were under the detection limit of the method. Of the hydrogenotrophic methanogens, during OMW digestion (which differed from other wastewaters), Methanobacteriaceae were more numerous than Methanomicrobiaceae and Methanococcaceae.
Pilot plant A pilot plant was established during the second year of the project. and the layout of the reactor and tanks decided. This could be used to investigate various treatment possibilities. This included the use of the mycetes to degrade phenols. However, the laboratory results indicated that while Pleurotus grew well in batch culture on OMW, this requires addition of specific nutrients the addition of which would entail costs and complications in a full-scale plant that could not be justified. Hence, it was decided not to use this organism. The yeasts, that in the laboratory studies showed some efficiency in phenol degradation, were tested with the aim of developing this in the pilot plant. One possibility was to add a preliminary treatment, in an integrated system, to recover ethanol which reducing the phenol concentration prior to anaerobic treatment. However, as a result of considerations of high energy consumption and instability it was concluded that a preliminary aerobic stage was not useful.
Both diluted (7-20 g/l COD) and undiluted (25-35 g/l COD) OMW were used in pilot plant anaerobic tests, and monitored for suspended solids, soluble solids, COD, total N, lipids, total phenols, total viable anaerobic bacteria and methanogens as well as gas yield. Significant operational problems were encountered during testing of diluted OMW. However, the results indicated that while inoculation problems are encountered when passing from the laboratory to the pilot plant scale, the anaerobic filter technology could be transferred to the pilot plant, without the need for a sophisticated system for stabilising the input stream. Realistically it would be difficult to establish as process that required the dilution of the input stream, while the input loads (1.5 to 2 g COD per l per day) were too low, while the specific gas production was also very low. The efficiency of gas production was 66% of that achieved in the laboratory under comparable load conditions. With undiluted OMW the work was interrupted by frost. This resulted in a need, when operation was resumed, to speed up the incremental increase in loading rate (due to the short oil extraction season), leading to a potential overload situation. However, specific organic loads were attained comparable to those achieved in the laboratory reactor. However, the COD removal rates were lower, since the transformation took place at 25 degrees C, rather than 35 degrees as in the laboratory, while the need for rapid inoculation probably did not permit the formation of a stable biofilm. There was also evidence of the formation of channels that conveyed the OMW directly to the digester output, while obstruction occurred due to the suspended solids. In general it was concluded that the process was viable without a need for pre-treatment, other than some mechanical filtering to reduce suspended solids and adjustment of pH.
The best performance in terms of gross methane production do not necessarily lead to the highest net energy production. The limiting parameters for optimising performance are temperature, concentration of OMW and the hydraulic retention time. The highest net productions were found with untreated OMWs with intermediate organic loads (5 to 10 g COD per l digester per day). For OMWs with a COD of 30 g/l corresponds to a retention time of 3 to 6 days. The maximum net gas productions were 0.4 to 0.7 m3/m3 digester/day in unfavourable climate conditions and 0.85 to 1.15 m3/m3 digester/day in favourable conditions. These can be increased using a heat exchanger to recover heat from the effluent to heat the incoming OMW, with a maximum value of 1.6 m3/m3. This should make plant of greater than 350 to 500 m3 economically viable, where the plant operate within a climate of high energy prices. With smaller anaerobic plant it would be very difficult to reach a break-even point.
Use of anaerobic effluent The simplest way to use the effluent from the anaerobic digestion of OMW id to spread it on land. In order to evaluate the effect of any residual phenolic substances on the soil, lysimeter tests were carried out and the percolating water was analysed. It was concluded that no appreciable contamination of ground water would occur and hence the anaerobic plant OMW effluent can be applied to soil without risk of contamination. The water soluble phenols are reduced due to absorption into the humic acid fraction and by break down by soil bacteria and fungi. Yields of barley grown in soil percolated with anaerobic plant effluent were higher than in the control, suggesting that the effluent supplied nutrients.
Conclusions
It was possible to transfer from laboratory to pilot plant an anaerobic process for treatment of OMW, which it should be possible to scale up if the following conditions are fulfilled:
It is also concluded that the effluents can be spread on soil without further treatment.
Of the other alternatives investigated it is concluded that:
© Copyright 2006 Policy Statements
Updated
by CPL Press:
03/07/2007
- biomatnet@biomatnet.org
 |
 |
 |
News |
 |
Events |
|
|