1. Introduction
The oily phase extracted from the HSSB with two outputs contains between 97-99% of oil and 1-3% of solid organic impurities and vegetable water ^2, although these proportions can oscillate. The fraction that including solid impurities is formed by olive fragments, containing proteins, sugars, phospholipids, phenolic compounds 3,4 and microorganisms, as bacteria, yeasts and mold 4-7. This set of organic matter, microorganism and water, can produce fermentations, and then can transmit to the oil some organoleptic defects, as ‘fusty’ or ‘muddy-sediment’ .
Before its storage, these impurities should be removed from the oil by the clarification step. Nowadays, oil clarification is carried out by vertical centrifugation, or/ and by natural settling 1,2,8
Vertical centrifugation is the most used oil clarification method since it is fast, requires little labor and lets to separate impurities effectively. This technology involves to add a quantity of lukewarm water, from 1:1 to 2:1 v/v water/oil, generating a considerable amount of effluents (from 20 to 40 L of aqueous-by-product per 100 kg of olive fruits). Because of the water addition, vertical centrifugation gives a markedly decrease in the phenol content, volatile compounds, induction time and bitterness of the resulting oils, being this effect more important as water is added at larger amount and higher temperature ‘ ‘ .
The natural settling is carried out by conical bottom settling tanks (CBSTs), built in stainless steel, with a cone angle between 45 and 60° and a capacity between 400 and 10000 liters. This method is chosen because do not does not require water addition and can be used both for batch (static settling) and continuous operation (dynamic settling)
Static settling is carried out in individual CBSTs where solids and water contained in the oil are progressively settled out. The oily must from HSSB is led to the bottom of the tank until that is filled, and then it is settled between 24 or 48 hours. Then, the oil is poured to the storage tank.
Other option is the dynamic settling that is carried out using two or more conic bottom tanks connected in series to each other. The oily must is led to the bottom of tank. When it is filled, the oil overflows into other tank through an outlet at the upper part of the tank. The same process occurs in the next settling tanks until the last one. Then, the clarified oil is led to the storage tanks.
For both static and dynamic settling, the settled impurities during the clarification are periodically purged from the bottom of settling tanks, in order to preserve the olive oil quality.
Nowadays, despite of the importance of clarification step in the final quality of VOO, its studies are limited. Previous studies on vertical centrifugation have been carried out using a high l water addition , – . In a recent work, Gila et al. studied the VOO clarification by vertical centrifugation with a minimal water addition, in which also was compared with clarification by settling tank under static conditions. Respect to dynamic settling, up to date, works describing the clarification operation by this system are not available.
Therefore, the aims of this work were: to study the VOO clarification by two different systems, VCS with minimal water addition and CBSTs in dynamic conditions, and their effect on VOO quality, composition and sensory characteristics.
2. Material and methods.
2.1 Plant material.
This work was carried out during three non-consecutive crop years, 2010/2011 (first crop year), 2011/2012 (second crop year) and 2013/2014 (third crop year). For each crop year, ‘Picual’ olive fruits were picked from the experimental orchard of the center ‘IFAPA Venta del Llano’ (Mengíbar, Jaén). Then, olive fruits were immediately transported to the oil mill for processing.
2.2 Olive oil extraction system
The experiments were performed in the experimental oil mill (Pieralisi, Spain), of IFAPA Center «Venta del Llano» in Mengíbar (Jaén, Spain). It consists of a metallic hammer mill with a hole grid diameter of 6 mm, a thermobeater formed by three containers of 600 kg each, a two-three phases horizontal centrifuge Pieralisi SC-90 (working at two phases way). General processing conditions were: 45 min malaxation time, 28°C malaxation temperature, 3.500 rpm centrifugation speed for horizontal decanter and an olive paste flow rate of 1000-1250 kg h-1.
2.3 Clarification systems
The experimental oil mill was also equipped with two different clarification systems working as following:
i. A battery of 3 conical bottom settling tanks (Secovisa, Spain) connected in series with a capacity of 400 liters each and whose dimensions are: 67×122 cm (0xh) for the top cylinder and an angle of the bottom cone of 35°. The oil from horizontal centrifuge was pumped to the first CBTS. When this CBST was filled, the oil overflowed into the second CBST through an outlet at the upper part of the tank. The same process occurs in the second and third CBTS. The time needed to fill each tank was around 100 minutes and the oil temperature during settling was 20 ± 2°C. Purges from the bottom of the settling tanks were carried out after filling each one of them, three purges for the first tank (at 100, 200 and 300 minutes), two purges for the second tank (at 200 and 300 minutes) and one for the third tank (at 300 minutes). These purges were performed manually by butterfly valve until clear oil appeared, with the aim to remove the settled moisture and solid particles.
ii. A vertical centrifugal separator (P1500, Pieralisi, Spain), with a nominal capacity of 1500 L h-1, operating during the experiment at 6500 rpm with a minimal water addition (5% of water regarding to processed oil) at a temperature similar to the oil. Before starting the experiment, a discharge was carried out in VCS.
2.4. Experimental design and sampling.
For the three crop years studied, the VOO from HSSB was passed through a 1 mm vibrational sieve. Then, the oil was clarified simultaneously by the two clarification systems. Half of the oil was pumped to the CBTSs battery and was clarified under dynamic conditions, whereas the rest was clarified by CVS (Figure 1).
During the assays, oil samples were taken at different points. Three oil samples after passing the vibrational sieve (‘HSSB oil’) as reference and three samples more from VCS outlet (‘VCS oil’) were taken after passing enough oil to stabilize the flow rates of the VCS, both coinciding with the beginning of the filling of each CBST. Besides, oil samples from overflow of each CBST (‘overflow 1’, ‘overflow 2’ and ‘overflow 3’) and also samples from the different tank purges carried out during the assay were taken. Both oil and purge samples were taken by triplicate.
For the three crop years, moisture and impurities content were quantified in the oil samples before filtering, and in the purging samples. Then, oil samples were paper- filtered and stored at -24 °C until analysis. Quality parameters, minor components and sensory analysis were analyzed only for the last crop year.
2.5.1 Moisture and solid organic impurities.
To determine water content, samples of olive oil (approximately 10 to 20 g in a capsule with filter paper) were dried in an oven at 105 °C until weight stabilizing. The loss of weight gave the amount (%) of water and volatile matter in the sample15. Solid organic impurities (SOI) content (%) was determined by introducing the dried sample in a SOXHLET system to eliminate the oil16. From these data, the clarification efficiency for the different clarification systems was calculated by Equation 1:
where Ce is the clarification efficiency (%) for each clarification system, cio is the moisture and solid organic impurities (MSOI) content of the initial oil from ‘HSSB’ (%) and cco is the MSOI content of the oil after clarification (%) by VCS or CBSTs under dynamic conditions, either at the overflow of the first, second or third tank.
2.5.2 Quality indexes.
Free acidity (FA), peroxide value (PV), and ultra-violet absorption at 232 and 270 nm (K232 and K270) were measured following the analytical methods described in European Regulation EEC 2568/9117. FA was expressed as percent of oleic acid, PV were expressed as milli-equivalents of active oxygen per kilogram of oil (mEq O2/kg); K232 and K270 extinction coefficients were calculated from absorption at 232 and 270 nm, respectively. Besides, fatty acid ethyl esters (FAEEs) content was analysed 18, result were expressed as mg/kg.
2.5.3 Total phenol content.
Phenolic compounds were extracted from an oil-in-hexane solution with methanol:water (60:40) and their concentration was measured using Folin-Ciocalteau reagent and colorimetric measurement at 725 nm 19. Results are expressed as mg/kg of caffeic acid.
2.5.4 Index K225.
The bitterness index K225 was determined applying the method described by Gutierrez et al. 20.
2.5.5 Tocopherol content.
Tocopherols were analyzed by HPLC, applying the IUPAC method 2432 21. Results are expressed as mg/kg of VOO.
2.5.6 Pigments content.
Carotenoid and chlorophyllic pigments were determined measuring the absorbance of olive oil weighed and dissolved in cyclohexane at 470 and 670 nm as described by Mínguez-Mosquera et al. 22. The results were expressed as mg/kg.
2.5.7 HPLC analysis of phenolic compounds.
The individual phenolic compounds were determined. The Phenolic Compounds were extracted with methanol/water and the extracts were analyzed by RP-HPLC 23. Phenolic compounds were quantified at 280 nm using syringic acid as internal standard and the response factors determined by Mateos et al. 24. The results were expressed as mg/kg.
2.5.8 Sensory analysis.
Sensory characterization was performed by the virgin olive oil analytical sensory panel of Fundation Citoliva formed by trained tasters as described by EU Regulation25. The results were expressed as the median of the intensity of the sensory attributes.
2.6 Statistical analysis
ANOVA analysis was performed in order to compare the effect on the oil of the two different clarification systems studied. When a significant F value was found, Tukey’s HSD was used to test differences between means (p < 0.05). These determinations were carried out using software Statistix, version 8.0.
3. Results and discussion
3.1 Process efficiency
Table 1 shows the moisture and SOI content of the VOO from HSSB before and after clarification by CBSTs under dynamic conditions and VCS. As can be observed, HSSB oils showed differences in their moisture and SOI content between the crop years analysed. The moisture ranged from 2.68% to 7.72%, whereas the SOI content varied between 0.41% and 1.28%. This variability was obtained changing the regulation of HSSB.
Table 1. Moisture and solid organic impurities content of the VOO from HSSB and clarified oils by CBSTs in dynamic conditions and VCS for the three crop years.
In general, clarified oils by VCS showed lower moisture content than those clarified by CBSTs. For solid organic impurities, CBTS oils showed low content (0.05-0.15%), while for the oils clarified by VCS they were not detected.
Clarification efficiency was calculated by Eq.1, using the sum of moisture and organic solid impurities. In general, the oil clarification by both systems was influenced by the MSOI content of the HSSB oil. Higher clarification efficiency was achieved when initial HSSB oil had higher MSOI content. VCS showed higher clarification efficiency (higher than 85%) achieving similar values between years. The dynamic settling by CBSTs showed clarification values lower than 85%.
Each CBST showed different clarification rates, due to the dynamic conditions of transition of the oil among tanks. In general, the second tank showed the highest clarification efficiency (higher than 45%). In the first tank less than 45% of MSOI were removed, being more efficient for those HSSB oils with higher MSOI content. The oils that arrived to the last settling tank were cleaner, therefore lower partial clarification percentages were obtained (lower than 35%) at this tank, being for the second crop year lower than 15%. These differences between efficiency of CBTS can be explained because in the first tank moisture and larger solid particles are settled, whereas in the next tanks lower amount of water and smaller solid particles are present.
Concerning the clarification times, CBTSs needed 300 minutes to obtain clarified oil, while VCS was faster and clarification was carried out immediately.
The removal efficiency of the settled impurities was also evaluated by measuring the MSOI and oil content of the tank purges and their weights (Table 2). In general, considering the MSOI content of the HSSB oils (Table 1) and CBST volume (0.4 m3 equivalent to 364 kg for oil density of 909 kg/m3 26), the total MSOI weight estimated for the full battery of three CBSTs was 63.3, 33.7 and 98.3 kg for the three crop years. Only 5.52, 4.48 and 54.46 kg of these MSOI contents were removed by purging the tanks, thus most of the MSOI remained in the CBSTs after purging.
The first CBST was the receptor of the HSSB oil, therefore, the most of settled MSOI was accumulated in the bottom of this tank. In general, higher MSOI amount was removed in the first tank (1.84-46.89 kg). Besides, a higher MSOI amount was removed in the third purge (300 min), since the most of MSOI remained in suspension for the two previous purges (100 min and 200 min), as observed for the MSOI content of the oil from ‘overflow 1’ in Table 1 (2.55-5.24 %).
For the second CBST, a lower total MSOI amount was removed (1.56-7.39 kg), since the oil from ‘overflow 1’ showed lower MSOI content compared to HSSB oil. The two purges carried out for this tank showed similar MSOI percentages for all crop years.
The last CBST, for the 2011-12 and 2013-14 crop years, showed the lower MSOI content in the purges, since the oil feeding this tank was previously clarified in the other two CBSTs. The first crop year showed different behavior, observing higher MSOI content (2.12 kg) than the purges performed for the first and second CBST, 1.84 and 1.56 kg, respectively.
Regarding to the oil losses in the purges, the use of HSSB oils with high MSOI content, as for the last crop year (9%), showed a greater oil loss (68.48 kg), than those HSSB oil with low MSOI content, for the first and second crop years, respectively.
Table 2. MSOI (moisture and solid organic impurities) and oil content of the different purges carried out during clarification process in the three CBSTs for the three crop years.
Therefore, these results showed that the purges performed were not efficient since a considerable amount of the MSOI remained in the tanks and a significant oil volume was lost14. The low efficiency of this purges system is due to this operation depend to subjective parameters: manual open and close of valve, visual appreciation of the clear oil and conical bottom characteristics. By contrast, VCS was a clarification system faster, more constant, with higher clarification efficiency and with minimal losses of oil.