Microsoft Word - 48dallavecchia.docx CHEMICAL ENGINEERINGTRANSACTIONS VOL. 49, 2016 A publication of The Italian Association of Chemical Engineering Online at www.aidic.it/cet Guest Editors:Enrico Bardone, Marco Bravi, Tajalli Keshavarz Copyright © 2016, AIDIC Servizi S.r.l., ISBN978-88-95608-40-2; ISSN 2283-9216 Study of the Growth Parameters of the Nannochloropsis Oculata for the Nitrogen and Phosphorus Removal from Wastewater through Design of Experiment Approach Giuliano Saiu*a, Agata Pistisb, Anuta Chindrisc, Massimiliano Grosso a, Maura Barolic, Efisio Antonio Scanoc a Università degli Studi di Cagliari, Dipartimento di Ingegneria Meccanica, Chimica e dei Materiali - Cagliari, Via Marengo 2 - Sardinia - Italy b Sardegna Ricerche, Laboratorio Biocombustibili e Biomasse - Uta, Z.I. Macchiareddu VI strada ovest - Sardinia - Italy c Fondazione IMC- Centro Marino Internazionale Località Sa Mardini, Torregrande, Oristano- Sardinia – Italy g.saiu@dimcm.unica.it Microalgae are a promising renewable energy source because of their capability to produce a large amount of oil, which can be directly used as a fuel or converted into biodiesel through transesterification processes. Microalgae growth needs a certain amount of nutrients such as nitrogen and phosphorus and, with this regard, the wastewater arising from aquaculture and many others agro-industrial processes can be a suitable and cheap source of these elements. Moreover, nitrogen and phosphorus are responsible for the eutrophication of the waters basins, therefore microalgae can be proficiently used to remove these pollutants. In this work Nannochloropsis oculata was tested to verify the nitrogen removal capability from a synthetic wastewater. The medium was composed of marine water added with appropriate macro and micronutrients mixtures. The experiments were carried out by means of lab-scale, completely mixed bubble-columns, photo-bioreactors. A Design of Experiments (DoE) approach was here used in order to assess whether the input factors (i) light intensity, (ii) different nitrogen concentration sources and (iii) carbon dioxide have a statistically significant impact on the microalgae growth. It was demonstrated that the only factors statistically significant are the light intensity and its interaction with the nitrate concentration, at least for the parameter ranges here investigated. As a final remark, it was found that nitrate removal were completed from 3 to 5 days depending on the medium composition, whereas urea and phosphates needed more time. 1. Introduction Algal biomasses represent a promising source of oil for the biofuels production, but microalgae cultivation requires a high amount of water resources and inorganic nutrients. This is a significant cost item for the feedstock production and affects importantly the oil cost. A solution to overcome the high costs is to couple the biodiesel production with waste water treatment based on microalgae, since the secondary effluent of plants contains high amounts of nitrogen and phosphorous (Xin et al., 2010). The treatment of waste water with microalgae to remove the nitrogen and phosphorous is well consolidated in the literature and it was originally proposed 50 years ago by Oswald and Gotaas (Abdel-Raouf et al., 2012). This was followed by several studies on laboratory and pilot scale (see e.g. Shelef et al., 1980, Oswald, 1988, Shi et al., 2007, Zhu et al.,2008). Microalgae growth and nutrient uptake depend on many factors such as pH, light intensity and wavelength, CO2 concentration, temperature and algal density (Dvoretsky, 2015). As CO2 concentration increases, it determines a biomass and total lipid productivity decrease, for some algal species (Chiu et al., 2009).The algal density strongly affects the nutrient removal performance and the photosynthetic efficiency (Lau et al., 1995, Darley, 1982). The nutrient removal efficiency is also affected by the light wavelengths and wavelengths mixing ratio. The latter parameter increases the microalgae production rate and the nitrogen and phosphorous removal rates (Kim et al., 2013). It should be remarked that these results were obtained with the DOI: 10.3303/CET1649093 Please cite this article as: Saiu G., Pistis A., Chindris A., Grosso M., Baroli M., Scano E., 2016, Study of the growth parameters of the nannochloropsis oculata for the nitrogen and phosphorus removal from wastewater through design of experiment approach, Chemical Engineering Transactions, 49, 553-558 DOI: 10.3303/CET1649093 553 single factors that were analyzed one at time. On the other hand, a thorough analysis of the impact of these factors jointly varied in the analysis is still lacking in literature and a complete understanding of the growth process should require the knowledge of the interactions among them. Nannochloropsis oculata is a marine eustigmatophyte widely used in aquaculture as a food for rotifers. In this sector N. oculata plays a significant role for its robust nature and the ability to accumulate intracellular oil (Kromkampet et al., 2009). In this work the growth of the N. oculata in seawater reinforced with the Guillard medium at different concentration levels of urea and nitrates was studied. The experiments were carried out using bubble column annular photobioreactors with a culture volume of six liters. The light intensity, carbon dioxide, nitrates and urea at two levels of concentration were the parameters that were jointly varied in the Design of Experiment (DoE) (Montgomery, 2013).The aim of this work was to analyze the best condition to improve nitrogen removal from wastewater by microalgae, in the range explored and for the parameters chosen. 2. Materials and methods 2.1 Algae species Nannochloropsis is a genus of algae mostly found in marine environment. This algae is an eukaryotic photosynthetic microorganism that grows rapidly due to its simple structure (Horsman et al., 2008) and because of its small size, ranging from 2 to 4 µm, it is considered to be part of phytoplankton. The microalgae cultures, were permanently kept in Erlenmeyer flasks with a solution of marine water and Guillard F/2 medium (Guillard et al., 1962-1975), under constant illumination and aeration at 22 °C temperature in the laboratories of the IMC Foundation-International Marine Centre, located in Torregrande, Oristano, Sardinia, Italy. Each experiment started from this culture as inoculum, that was previously cultivated to reach a desired concentration and then inoculated into the photobioreactors. 2.2 Medium To simulate a saline aquaculture wastewater, marine water has been modified adding the Guillard medium and different levels of urea and nitrate salts. The Guillard medium composition is reported in Table 1, urea and sodium nitrate used were analytical grade, and all the compounds listed were from AppliChemPanreac. A one ml/l of trace metals and 0.5 ml/l of mix-vitamins were added to complete the nutrients and micronutrients profile of the medium, to ensure the microbial growth. Nitrate and urea were added in the medium when required as further explained in the experimental section. Table1: Modified Guillard medium Compound Conc [mol/l] Compound Conc [mol/l] NaH2PO4 * H2O 3.62 x 10-5 Trace metals: Na2SiO3 * 9H2O 1.06 x 10-4 CoCl2 * 6H2O 4.20 x 10-8 Trace metals: CuSO4 * 5H2O 3.93 x 10-8 FeCl3 * 6H2O 1.17 x 10-5 Na2EDTA * 2H2O 1.17 x 10-5 Vitamins: MnCl2 * 4H2O 9.10 x 10-7 Thiamine HCl 2.96 x 10-7 ZnSO4 * 7H2O Na2MoO4 * 2H2O 7.65 x 10-8 2.60 x 10-8 Biotin Cyanocobalamin 2.05 x 10-9 3.69 x 10-10 2.3 Parameter measurements Monitoring of the algae growth was obtained by means of optical density measurements, which were taken using a dual beam spectrophotometer (UV-VIS Jasco V-530), acquiring both 680 and 750 nm wavelengths. In particular, the first wavelength can be related to the α-chlorophyll and the second to the biomass. Also a portion of the UV-Visible spectrum from 300 to 800 nm was acquired, to allow further studies. Two samples a day were withdrawn from the photobioreactors and frozen, to search the following parameters: dry weight, nitrates and urea. During the experiments pH, temperature and electrical conductivity were monitored and recorded. Different measurements of photosynthetically active radiation (PAR) were also undertaken, to obtain a complete profile of light irradiation in the different sections of the photobioreactors (LI-COR quantum sensor, model LI-190). 2.4 Chemical Analysis In addition to the indirect estimation of the biomass amount, through direct OD measurements, dry weight analysis (Clasceri et al, 1999) were done for each sample withdrawn (twice a day) during the experiments. Forty ml of sample were filtered through 0.45 μm Whatman membranes, under vacuum. After complete filtration, each filter was washed with 20 ml of distilled water under vacuum, to remove the salts from the 554 saline water, and dried at 105 °C. The membranes were then brought to constant weight. The liquid fraction of the sample was used for the nitrate, urea and phosphorous analysis. To obtain the N-NO3 and P-PO4 profile over time and for each experiment, an high performance ion chromatography unit from Thermo ScientificTM (DionexTM ICS-2000) was used. The instrument was equipped with a DionexTM Ion Pac TM AS19 Column, operating in the following conditions: constant temperature, 30°C, eluent KOH 10mM from 0 to 10 min, gradient from 10 to 45 [mMol] from 10 to 25 minutes. The sample was previously treated with a DionexTM ON GUARD II Ag 1cc cartridge, to remove the excess of Cl-, Br- and I- ions that could interfere with the nitrate analysis. The calibration curve, with five points, was done every day before the analysis using a multiparametric certified standard mix (F-, Cl-, NO2 -, NO3 -, Br, SO4 2- and PO4 3-). In this working conditions, the retention time for nitrates was 11 min and for phosphates 20.4 minutes. Urea consumption was analyzed by means of an adapted colorimetric method (Mulvenna, 1991). The 0.45 micron filtered samples were treated at acid pH with Diacetylmonoxime and Thiosemicarbazide, incubated at 85 °C for 20 minutes, then refrigerated with a bath of cold water for 5 minutes and, finally, the absorbance at 520 nm was acquired in a Varian Cary®50 UV-VIS spectrophotometer. The calculated detection limit was 1.5 mg/l, uncertainty ±1.5 mg/l (almost constant in the range of calibration curve), according to the UNICHIM statistical procedures, and the calibration points ranged from 0 to 150 mg/l. 2.5 Photobioreactors The experiments were carried out using two lab scale, completely mixed, bubble column, photo-bioreactors of six liter capacity (each). The two reactors are equipped with temperature and pH monitoring and regulation control system. The lighting system is composed of four neon daylight lamp (four fluorescent lamps type cool day light, OSR FQ 24W/865).The gas supply is ensured by a blower and the pure carbon dioxide is fed into the air stream. As known, during the CO2 fixation by photosynthesis, the microalgae, in presence of NO3 - or NH4 +, form OH- ions increasing the pH of the medium. The pH control system operate in order to keep the pH at set point value, opening the CO2 valve when pH exceed 7.60 (optimal pH growing conditions for N.O.). The carbon dioxide was also used as integrated carbon source and to evaluate if its presence enhanced the nitrogen removal. The growth conditions were: Temperature 22°C, pH 7.60 ± 0.2. Air volumetric flow 2 l/min, CO2 volumetric flow 0.2 l/min. 2.6 Experimental design In order to evaluate the behavior of the algae growth due to the different parameters, four factors that may influence the algae growth were chosen and were jointly varied by means of design of experiment approach. Light intensity, carbon dioxide, nitrate and urea were selected to perform the experiments (Spolaore et al., 2006, Faria et al., 2012). A fractional factorial design (indicated as 2 ) consisting of two level, four factors and half fraction of the full factorial, with resolution IV, was used (for further details see Montgomery, 2013) with the aim of reducing the number of experiments. In fact, due to the long term duration of the experiments, the fractional factorial design was an obliged choice to describe the process under investigation in reasonable times. Table 2 reports the levels adopted for each factor varied. Table 2: Levels at low and High concentration/number Factor/Level Low High Units A= Lamps 2 4 number B=NaNO3 60 120 mg/l C=Urea 0 120 mg/l D=CO2 0 1 - The fractional factorial design has been created using the generators and defining relations reported on Table 3, along the complete alias structure, for more details see Montgomery, 2013. As it can be seen in the Table 3 the single effects from the two way interactions can be resolved because of the resolution IV of the chosen fractional factorial design. Table 3: Generators, defining relations and complete alias structure Terms Generators D=ABC Defining relation I=ABCD Alias structure I+ABCD A+BCD B+ACD C+ABD D+ABC AB+CD AC+BD AD+BC 555 2.7 Growth model The algae growth is mathematically modelled by resorting to well-known population models (Seber et al., 2003). Amongst the others, the Gompertz model (eq.1) revealed to be the most proper to fit the experimental results in terms of S (standard deviation of the error, eq.2), compared to the other models (e.g. Logistic equation). The choice was taken by considering the mean square error of the fit. In the equation 1, A is the algae population, µ is the specific growth rate and k is the carrying capacity. μ ∙ log log 1 2 The equation 1 was used to systematically obtain the scalar µ for each experimental run, considering the specific growth rate a representative parameter of the growth process. The specific growth rate was used to evaluate the fractional factorial design. Eq. 2 was also used to compare the goodness of fit (S) among the nonlinear regressions performed for each run. The parameter SSE is the Sum of Squared Errors (∑ ), N is the number of observations and P is the number of parameters. 3. Results and discussion 3.1 Nitrate and Urea consumption As described in Table 2, the NaNO3 concentration at t0 ranged from 60 to 120 mg/l, and its degradation was accomplished in 3-5 days (e.g. Figure 1), depending on the initial concentration and on the experimental conditions. On the other hand, degradation of urea revealed to require more time, probably because the microalgae started degrading first the nitrate, more easily degradable than the urea. However, the presence of urea, that ranged from 0 to 120 mg/l, slightly improved the microbial growth in terms of specific growth rate µ. Figure 1 shows the complete depletion of nitrate (almost in all the runs) and the partial degradation of urea and phosphorous reached in ten days. Table 4: Growth rates, removal rates (R-xx) and per- centage consumption (%C) for Nitrate, Urea (when present) and Phosphorus. Run R-NO3 [mg/l/Die] %C NO3 R-Urea [mg/l/Die] %C Urea R-PO4 [mg/l/Die] %C PO4 1 14.9 99.3 - - 0.27 30.8 2 20.0 100.0 - - 1.01 35.0 3 30.0 100.0 - - 0.19 26.6 4 29.9 99.7 - - 0.39 36.2 5 12.0 100.0 4.8 57.7 0.57 48.4 6 10.5 100.0 9.0 72.6 0.53 59.5 7 19.4 96.9 5.2 44.7 0.23 52.1 8 24.0 100.0 3.8 47.8 0.21 40.6 Figure 1: Dry Weight Vs. Urea, N-NO3 and P-PO4 consumption (Run 6 – t0: Urea & NO3 120 mg/l) Table 4 reports the removal rates (R-xx) for nitrates, urea and phosphorus, and the amount of the nutrients removed during the experiment period (10 days) on a percentage basis, for each experiment performed. The algae biomass, in terms of dry weight, ranged from 0.366 to 1.097 g/l, depending on the experimental conditions. 3.2 Statistical analysis As already introduced in Section 2.7, the impact of the four factors chosen has been evaluated by means of the representative scalar µ. By use of the Yates algorithm (Nelson et al., 2003) we were able to assign an effect to each factor in order to quantify its impact on the system. In figure 2, the Pareto chart shows the absolute value for the effects computed. The dashed line reported in figure 2 represents the threshold value for the statistically significant terms (light and light-N-NO3 interaction on the right) from the not significant, on the left. As described by Montgomery, the not significant factors could be neglected (threshold value: α=10%), and the experimental campaign could be projected to a 22 full factorial design, with two replicates (urea and 556 CO2 discarded). This procedure allowed to re-process the data obtained from the experimental campaign, without losing significant information, and to perform the analysis of variance (ANOVA), that is summarized into Table 5. The Light, as expected, is confirmed as significant factor, p-value 0.01, and interaction light- NaNO3 is statistically significant as well, p-value 0.028. The nitrates have a mild negative impact on the system, but are not statistically significant (p-value 0.762). D AC B AD C AB A 0.140.120.100.080.060.040.020.00 Te rm Effect 0.0777 A Light B NaNO3 C Urea D CO2 Factor Nam e Lenth's PSE = 0.0292723 Figure 2: Pareto chart for computed effects [dimensionless] Table 5: Effects and coefficients for each significant factor (coded unit) – ANOVA table (dimensionless) Term Effect Coef SE Coef DoF SSE MSE F-ratio p-value Constant 0.31642 0.01438 2 0.0362 0.0181 10.97 0.024 Light 0.13440 0.06720 0.01438 1 0.0361 0.0361 21.83 0.010 NaNO3 -0.00931 -0.00466 0.01438 1 0.0001 0.0001 0.10 0.762 Light/NaNO3 -0.04855 -0.04855 0.01438 1 0.0189 0.0189 11.40 0.028 In figure 3 is reported the growth curve for the run 6 (4 lamps, 60 mg/l N-NO3, 120 mg/l Urea), in terms of dry weight, and the non-linear Gompertz equation, that well represents the data, along with the confidence intervals (MSE=0.00154 and S=0.03929). Picture 4 clearly shows the interactions plot for light and N-NO3. Indeed, a marked curvature on the interactions lines is evident. The fastest growth was found for high levels of light end low levels of N-NO3. 1086420 1.2 1.0 0.8 0.6 0.4 0.2 0.0 Days D ry W ei gh t [ g/ l] Gompertz equation 95% Confidence Interval Figure 3: growth curve in terms of Dry Weight and fitted Gompertz equation Light (n. lamps) N - N O 3 (m g/ l) 4.03.53.02.52.0 120 110 100 90 80 70 60 > – – – – < 0.20 0.20 0.25 0.25 0.30 0.30 0.35 0.35 0.40 0.40 µ Gomp Figure 4: Interaction plot for Light and Nitrates, computed with µ [day-1] as output 4. Conclusions The DoE analysis applied to the experimental campaign showed that the light was the most statistically significant parameter for the algae growth. It was also found that the interaction between light and N-NO3 was 557 statistically significant as well. Indeed, the main value for the specific growth rate was found for the experimental run 6 (4 lamps, 60 mg/l N-NO3, 120 mg/l Urea), whereas the lower value was found in the run 3 (2 lamps, 120 mg/l N-NO3, 0 mg/l Urea). This confirms that, as found with the DoE analysis, the light has the main impact on the system. The total N-NO3 degradation was accomplished in 3-5 days, while the Urea consumption reached removal rates ranging from 44.7 to 72.6% in 10 days. The P-PO4 removal, in the same operating conditions, achieved values between 27 to 60 %. Urea and CO2, in the range here explored, were found not statistically significant. References Abdel-Raouf N., Al-Homaidan A.A., Ibraheem I. B. M. (2012) Microalgae and watewater treatment, Saudi Journal of Biological Sciences, 19, p.257-275. Chiu S. et al. (2009) Lipid accumulation and CO2 utilization of Nannochloropsis oculata in response to CO2 aeration, Bioresource Technology, 100, p.833-838. Clasceri L., Greenberg A., Eaton A. (1999) Standard Methods for the Examination of Water and Wastewater, 20th ed. Amer Public Health Assn, p.1931. Darley W. M. (1982) A phisiological Approach. Algal Biology. Basic Microbiology, 9, Blackwell Scientific Publications Oxford. Down, BT22 1PF, Northern Ireland - Estuarine, Coastal and Shelf Science 34, p.429-438. Dvoretsky D. et al. (2014). Optimization of the Process of Cultivation of Microalgae Chlorella Vulgaris Biomass with High Lipid Content for Biofuel Production. Chemical Engineering Transaction, Vol.43. Faria G. R. et al. (2012) Effects of the availability of CO2 on growth, nutrient uptake, and chemical composition of the marine microalgae Chlorella sp. and Nannochloropsis oculata, two potential useful strains for biofuels production, International Research Journal of Biotechnology, 3,5, p.65-75. Guillard R.R.L., Ryther J.H. (1962) Studies of marine planktonic diatoms I. Cyclotella nana (Hustedt) and Detonula confervacea (cleve), Can. J. Microbiol. 8, 229-239. Guillard R.R.L. (1975) Culture of phytoplankton for feeding marine invertebrates in W.L.Smith, M.H. Chanley (Eds.), Culture of Marine Invertebrate Animals, Plenum Press, New York, p.26-60. Lau P. S., Tam N. F. Y., Wang Y. S. (1995) Effect of algal density on nutrient removal from primary settled waste water, Environmental Pollution, 89, p.56-66. Li Y. et al. (2008) Biofuels from microalgae, Biotechnol. Progr.24 815-820. Montgomery, D. C. (2013) Design and analysis of experiments, Hoboken, NJ: John Wiley & Sons, Inc. Mulvenna P.F. & Savidge G.(1991) The Queen's University of Belfast, Marine Biology Station, Portaferry, Co. Natrella M.G. (2013) Experimental Statistics, Dover Pubblications. Mineola, NY. Nelson R. et al. (2003) Introductory Statistics for Engineering Experimentation. USA, Elsevier Science, p.354. Oswald W. J. (1988) Microalgae and wastewater treatment. In Borowitzka M. A., Borowitzka L. J. (Eds), Microalgal Biotechnology, Cambridge University Press, p.305-328. Seber G.A.F., Wild C.J. (2003) Nonlinear regression. Hoboken, NJ: John Wiley & Sons, Inc. Chapter 7. Shelef G. et al (1980) Algal mass production as an integral part of waste water treatment and reclamation system in algal biomass . Elsevier, p.163-190. Shi J., Podola B., Melkonian M. (2007) Removal of nitrogen and phosphorous from waste water using microalgae immobilized on twin layers: an experimental study, Journal of applied Phycology, 19, 5, 417- 423,. Spolaore p., Joannis-Cassan c., Duran e., Isambert a., Optimization of Nannochloropsis oculata growth using the response surface method (2006) Journal of Chemical Technology and Biotechnology, 81, p.1049- 1056. Tae-Hyeong Kim et al. (2013) The effects of wavelength and wavelength mixing ratios on microalgae growth and nitrogen phosphorous removal using Scenedesmus sp. For waste water treatment, Bioresource Technology, 130, p.75-80 Zhu G. et al. (2008) Biological removal of nitrogen from wastewater, Reviews of Environmental contamination toxicology, 192, p.159-195. Xin L., Hong-ying H., Jia Y. (2010) Lipid accumulation and nutrient removal properties of a newly isolated freshwater microalga, Scenedesmus sp.LX1, growing in secondary effluent, New Biotechnology, 27, 1, p.59-63. 558