Article pubs.acs.org/IECR
Novel Filtration Photobioreactor for Efficient Biomass Production Di Zhang, Ka Y. Fung, and Ka M. Ng* The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, People’s Republic of China ABSTRACT: A filtration photobioreactor (FPBR) has been designed in this study. It requires a much shorter time to start up than existing biofilm systems, while providing easy biomass harvest. Microalgae are retained on a membrane through filtration so that microalgae growth is spatially separated from nutrient discharge and recharge. Therefore, additional processes for removing viable cells from spent medium are not required, and CO2 transfer efficiency into the medium can be improved without affecting cell growth. Performance of the FPBR for Chlorella sp. cultivation has been studied under various operating conditions, and an operating time and initial cell concentration for achieving optimum productivity has been selected. With a medium recycle rate of 5 mL min−1 and a feed gas containing 7.5% CO2, the biomass productivity was 13.56 g m−2 d−1 as the cell concentration increased from 2.54 to 20 g m−2 at 35 °C, which was higher than that achieved in a biofilm system for Chlorella sp. cultivation. cultivation and biomass productivities of 5.5 g m−2 d−1 at bench scale and 31 g m−2 d−1 at pilot scale were reported. Previous studies have demonstrated the great potential of using biofilm systems for large-scale microalgae production with satisfactory biomass productivity and ease of harvest. However, use of biofilm systems has its disadvantages. Because bacteria play a critical role in biofilm formation, it is difficult to use biofilm systems for axenic microalgae cultivation.13 Also, a biofilm usually takes days to months to develop, which leads to a long start-up period each time after biomass harvest.14 Before the formation of a tight biofilm, microalgae can be easily washed off from the biofilm by the liquid flow. The detached viable cells have to be removed from the water before it can be discharged, as otherwise biological contamination may occur.15 All these add extra costs to microalgae cultivation. To ameliorate these problems, an attached growth system, filtration photobioreactor (FPBR), has been developed. Although membranes have long been used for microalgae harvest from suspended cultures or to facilitate CO2 and O2 transfer into or out of a microalgal culture, they have rarely been utilized as the cultivation surface intended for large-scale microalgae cultivation.16−19 In a FPBR, microalgae are attached to a membrane cultivation surface using the same mechanism as filtration so that biofilm formation is no longer required. Therefore, the FPBR not only provides easy harvest of microalgae but also can be applied for axenic microalgae cultivation and is quick to start up. It is also designed to prevent viable cells from entering the bulk medium so that used medium can be directly discharged. The conceptual design of FPBR is first presented in this paper with the potential advantages summarized. A bench-scale FPBR has been fabricated to evaluate its performance on biomass production, CO2 conversion, and nitrogen and phosphorus removal from the medium. The effect of key operating parameters, including temperature, medium recycle rate, CO2 availability, operating
1. INTRODUCTION Microalgae have attracted much attention due to the increasing interest in their potentials for making biofuels, health care products, and nutritious animal feeds.1−3 The production cost is currently high, partially because biomass harvest is expensive.4 Commercial microalgae products are typically produced using open ponds or enclosed photobioreactors where microalgae are suspended in the culture. Because the biomass contains much oil and has a density similar to the medium, harvest through settling is barely feasible. Although centrifugation, flotation, filtration, and flocculation can be used to harvest microalgae, these methods are very costly due to the low biomass concentration in the suspended culture (0.1−1 g L−1).5,6 Some researchers embedded biomass into a polymer matrix such as a loofa sponge or alginate beads to achieve efficient solid−liquid separation.7,8 However, the total harvest cost was still high, as biomass could only be recovered with additional chemical processes to dissolve the polymer matrix. In recent years, biofilm systems for microalgae cultivation have been developed using wastewater as a nutrient source. In a biofilm system, microalgae attach to a cultivation surface, entrapped by the extracellular polymeric substances excreted by both microalgae and bacteria in the wastewater. The remarkable advantage of such attached systems is easy harvest of biomass as cells are bound together and are readily separated from the medium. One of the most popular designs was the algal turf scrubber, in which a microalgal biofilm was developed on a slightly tilted cultivation surface with wastewater flowing across the biofilm from the higher end to the lower end and being recycled. At the time of harvest, a large proportion of the biomass would be scraped off from the cultivation surface and the remaining cells would initiate another round of biofilm growth. The biomass (microalgae and bacteria) productivity of algal turf scrubbers ranged from 0.71 to 25 g m−2 d−1 in different cases.9−11 Another design was a rotating algal biofilm reactor, where a cotton rope with attached microalgae and bacteria wound around a rotating drum was partially submerged in a wastewater tank.12 As the drum turned, microalgae were intermittently provided with nutrients and light to proliferate. Biomass was continuously squeezed off the rope during © 2014 American Chemical Society
Received: Revised: Accepted: Published: 12927
May 10, 2014 July 28, 2014 July 31, 2014 August 11, 2014 dx.doi.org/10.1021/ie501913k | Ind. Eng. Chem. Res. 2014, 53, 12927−12934
Industrial & Engineering Chemistry Research
Article
Figure 1. Conceptual design of the FPBR.
CO2 enriched air can be used as a feed gas to provide a carbon source for microalgae. The feed gas is sparged into the medium from the bottom so that CO2 can dissolve into the medium as it rises in the medium preparation unit. Various configurations can be used to design the medium preparation unit, such as a single or a series of bubble columns or packed towers. The selection is mainly based on the desired CO2 transfer rate and unit volume. A bubble column is easy to construct and operate. But the volumetric CO2 transfer rate is also low so that a large column is usually required. Packed towers and pressurized systems may be utilized if a feed gas of low CO2 concentration (e.g., ambient air) is used as the carbon source or if the volume of medium preparation unit needs to be reduced. The feed gas can be supplied either batch-wise or continuously. If the CO2 concentration in the outlet flow needs to be lower than certain limits, the loading rate of a continuous feed gas stream or the CO2 dosage for a batch can be adjusted accordingly. As nutrients in the medium are consumed by the microalgae, the medium has to be refreshed periodically. Because the spent medium has no biomass when it is discharged from the medium preparation unit, there will be little biological contamination caused by the discharged medium. As high solar intensity may cause photoinhibition on microalgae growth, the cover of a cultivation chamber can be designed to shield off part of the sunlight. In addition, the cover of the cultivation chamber also serves as a barrier to airborne contaminants. The medium and feed gas should be filtered before entering the medium preparation unit, as otherwise bacteria will accumulate on the membrane during operation. In cases where the microalgae are cultivated for food or medical industry, the inside walls of all chambers and conduits should be sterilized frequently to meet health and safety standards. To prevent the bioreactor from operating at a temperature outside the suitable range for microalgae growth, temperature control devices can be added to either the cultivation unit or the medium preparation unit. After a certain amount of biomass has grown on the membrane, the biomass is scraped off and harvested from the membrane surface after the medium in the
time, and initial microalgae concentration on biomass production, has also been elucidated.
2. CONCEPTUAL DESIGN OF FILTRATION PHOTOBIOREACTOR The filtration photobioreactor (FPBR) designed to provide easy microalgae harvest and quick start-up is depicted in Figure 1. The system comprises a microalgae cultivation unit and a medium preparation unit. The cultivation unit consists of an enclosed transparent cultivation chamber and a recycle chamber. The two chambers are separated by a membrane on its holder. At the beginning, microalgae are inoculated as a suspended culture in the cultivation chamber. During operation, as medium passes through the membrane, microalgae are intercepted by the membrane to form an evenly distributed biomass layer. The medium is continuously recycled from the bottom of the recycle chamber back to the cultivation chamber after passing through the adjacent medium preparation unit to replenish CO2. The medium recycle to the cultivation chamber should be designed to create minimal disturbance to the even distribution of microalgae on the membrane. As shown in Figure 1, to keep a constant level of medium in the cultivation chamber and the recycle chamber, the medium flow rate out of the recycle chamber (Qr) should be the same as the medium flow rate passing through the membrane (Qm). This requires the pump to have a capacity (Qp) of no less than Qm at all times. Because Qm is dominated by the pore radius, as the cell concentration increases, Qm will decrease. Hence, the maximum medium flow rate across the membrane (Qm0) should be achieved at the beginning of incubation when the cell concentration is at the minimum. Therefore, as long as Qp is set at a value of no less than Qm0, the medium level in both the cultivation chamber and the recycle chamber can be kept constant throughout the operation. The medium preparation unit is designed for CO2 transfer from feed gas to medium in autotrophic microalgae cultivation as well as medium discharge and replenishment. Ambient air or 12928
dx.doi.org/10.1021/ie501913k | Ind. Eng. Chem. Res. 2014, 53, 12927−12934
Industrial & Engineering Chemistry Research
Article
Figure 2. Schematic diagram of the FPBR.
proper initial and final biomass concentration of batch operation trades off productivity with unfavorable mutual shading, and thus, may offer maximum productivity.
cultivation chamber is drained. Another batch of suspended culture is to be inoculated into the cultivation chamber to start a new cycle of microalgae growth. Same as the biofilm systems, the FPBR can also have microalgae attached to the cultivation surface. However, a shorter start-up time can be expected in the FPBR, as cell attachment is affected by the medium flow instead of the kinetics of biofilm formation. Without the need of biofilm formation, coexistence of bacteria is also optional, which leads to the possibility of axenic microalgae cultivation using the FPBR. In addition, viable microalgal cells have little chance to enter the medium preparation unit where used medium is discharged. Therefore, subsequent treatment of the discharged medium for the removal of biological contaminant can be saved. Meanwhile, separating the medium preparation unit and the cultivation unit allows designs such as packed tower to be used for enhancing the CO2 transfer rate inside the medium preparation unit. Such designs may not be used in suspended or biofilm systems, as they cause potential loss of viable cells. As identified from the conceptual design of the FPBR, key parameters for controlling the FPBR operation are medium temperature, medium recycle rate, CO2 availability, operating time, and initial microalgae concentration. The medium temperature affects enzymatic activity of microalgae and thus the specific growth rate. Medium recycle rate determines the sufficiency of nutrient supply to the microalgae in the cultivation chamber. A low medium recycle rate may cause nutrient limitation while a high medium recycle rate increases energy consumption. Carbon is the most important component of microalgae. Therefore, CO2 availability directly affects the specific growth rate of microalgae. The operating time between inoculation and harvest determines the final amount of biomass in the cultivation chamber for a specified initial amount. A
3. BENCH-SCALE FILTRATION PHOTOBIOREACTOR AND PERFORMANCE ANALYSIS 3.1. Bench-scale Filtration Photobioreactor (FPBR). According to the conceptual design of the FPBR, a bench-scale unit, as illustrated in Figure 2, was fabricated for evaluating the performance. The cultivation unit was separated by a membrane with a pore size of 5 μm (ISOPORE, USA) into a cultivation chamber and a recycle chamber. The cultivation chamber was illuminated by four 10 W light emitting diodes (LEDs) on the top to provide a photosynthetic photon flux density (PPFD) of 100 ± 5 μmol m−2 s−1 on the culture surface, as measured with a photometer (LI-250a, LI-COR, USA). The illuminated area was 113 cm2. The cultivation unit was kept at a preset temperature by the thermostatic chamber. During start-up, microalgae were inoculated as a suspended culture into the enclosed cultivation chamber. As the medium passed through the membrane, microalgal cells would settle and form an evenly distributed layer. The passing medium was collected in the recycle chamber and was pumped to the adjacent medium preparation unit by a peristaltic pump. The medium preparation unit receiving the recycled medium was a 45 cm tall jacketed bubble column with a maximum volume of 707 mL. The outer jacket was connected to a water bath (Model: 9106, Polyscience, USA) to keep the medium at a preset temperature. The flow rate of CO2 and synthetic air were controlled by separate mass flow controllers (Smart-Trak, SIERRA Instruments, Inc., USA) and were mixed to produce a feed gas of certain volumetric percentage of CO2 (CO2%). The 12929
dx.doi.org/10.1021/ie501913k | Ind. Eng. Chem. Res. 2014, 53, 12927−12934
Industrial & Engineering Chemistry Research
Article
Table 1. 48-h Average Specific Growth Rate (μ) of Chlorella sp. Cultivated in the FPBR under Various Operating Conditions
feed gas was continuously sparged into the medium from the bottom of the bubble column. Part of the CO2 in the feed gas dissolved into the medium, and the remaining gas was released from the headspace of the bubble column where the CO2% in the gas outlet was measured by a CO2 meter (CM-0040, CO2Meter.com, USA). The medium replenished with CO2 would overflow from the bubble column back to the cultivation unit at a flow rate same as the recycled medium. To minimize heat loss, the conduits connecting the cultivation unit and the medium preparation unit were wrapped with insulation tape. The medium flow rate in the whole system equaled to the medium flow rate across the membrane, which was controlled by the peristaltic pump. Increasing the flow rate of the peristaltic pump (Qp) would create a higher vacuum in the recycle chamber to increase the flow rate of the medium across the membrane. As long as the medium was transported from the recycle chamber to the bubble column without accumulation in the recycle chamber, the system would be stabilized and the volume of medium (Vc) in the cultivation chamber would be constant. The hydraulic retention time of medium in the cultivation chamber (HRTc) could be approximated by eq 1.
HRTc =
Vc Qp
temperature (°C) RT 35 35 35
a
medium recycle rate (mL min−1)
feed gasb CO2% (vol./vol.)
μ (h−1)
SD (h−1)
5 5 10 5
7.5 7.5 7.5 25
0.0562 0.0565 0.0569 0.0575
0.00010 0.00006 0.00010 0.00006
a
The experiment was carried out under the ambient temperature of 20 ± 1 °C There was no temperature control for the cultivation chamber or the bubble column. bFeed gas was composed of CO2 and synthetic air, and the feed gas loading rate was 200 mL min−1.
continuously measured to determine the consumption rate of CO2. Liquid samples were collected from the bubble column at the beginning and the end of the experiment. The chemical oxygen demand (COD), nitrogen concentration, and phosphorus concentration of samples were determined. The samples were mixed with COD digestion reagent vial, low range (Product #: 2125815, HACH, USA), NitraVer nitrogennitrate reagent set, high range (Product #: 2605345, HACH, USA), and phosphorus (total) TNT reagent set (Product #: 2742645, HACH, USA) for measurement of COD, nitrate, and phosphate concentration using a colorimeter (DR/870, HACH, USA), respectively. Experimental procedures followed the corresponding instructions provided in the HACH DR/870 Datalogging Colorimeter Handbook. At the end of the experiment, Chlorella sp. was harvested from the membrane and was resuspended in 200 mL of fresh BBM. The OD680 of the resuspended harvest was measured with the UV−vis spectrometer to determine the average specific growth rate, μ, of the microalgae, defined below
(1)
3.2. Inoculum Preparation. A freshwater Chlorella sp. was employed in this study. The Chlorella sp. was cultivated in the modified Bold’s basal medium (BBM) with the following composition (per liter): 250 mg of NaNO3, 75 mg of K2HPO4, 175 mg of KH2PO4, 25 mg of CaCl2·2H2O, 25 mg of NaCl, 75 mg of MgSO4·7H2O, 500 mg of EDTA anhydrous, 31 mg of KOH, 4.98 mg of FeSO4·7H2O, 1 mL of H2SO4, 11.42 mg of H3BO3, 8.82 mg of ZnSO4·7H2O, 1.44 mg of MnCl2·4H2O, 1.19 mg of NaMoO4, 1.57 mg of CuSO4·5H2O, and 0.40 mg of CoCl2·6H2O. The inorganic carbon source was provided by feed gas comprising CO2 and synthetic air. In this study, all batches were started with an equal amount of inoculated cells. The seeding culture with an optical density at 680 nm (OD680) of 0.270 ± 0.005, as measured with a UV− vis spectrophotometer (Ultrospec 4300 pro, USA), was prepared. For each batch, 200 mL seeding culture was inoculated into the cultivation chamber. This provided an initial biomass concentration of 1.28 g m−2 on the membrane. 3.3. Analysis of the Filtration Photobioreactor Performance. Effect of Key Operating Parameters on Microalgae Growth. To study the effect of temperature, medium recycle rate, and CO2 availability on microalgae growth in the FPBR, experiments were conducted with the bench-scale FPBR under the different operating conditions listed in Table 1. All experiments were repeated three times. The experiments were conducted for 48 h under the photosynthetic photon flux density (PPFD) of 100 ± 5 μmol m−2 s−1. The cultivation chamber and the bubble column were either exposed to room temperature (20 ± 1 °C) or controlled at 35 °C by the thermostatic chamber and the water bath. While the volume of medium in the cultivation chamber was kept at 200 mL (1.77 cm above the membrane), the peristaltic pump controlled the medium recycle flow rate at 5 or 10 mL min−1 to give an HRTc of 40 or 20 min. Regarding the supply of the inorganic carbon source, a feed gas with 7.5% or 25% CO2 was continuously sparged into the bubble column at a flow rate of 200 mL min−1. The CO2% in the gas outlet from the headspace was
μ=
Δ(ln OD680 ) Δt
(2)
where Δt is the time between two measurements of OD680. Selection of Operating Time and Initial Microalgae Concentration for Batch Operation. After the appropriate operating conditions were selected, including temperature, medium recycle rate, and CO2 concentration in the feed gas for cultivating Chlorella sp., the same cultivation experiment was conducted in the FPBR under the selected conditions. Instead of 48 h, a longer period of 96 h was conducted, to select an optimum operating time and initial microalgae concentration for batch operation. During the experiment, biomass was harvested every 24 h and resuspended in the culture medium. The OD680 of the resuspended harvest was measured and the average specific growth rate for each harvest period could be calculated using eq 2. Biomass productivity (P), as defined by eq 3, was also calculated for each period. P=
ΔN Δt
(3)
where N is the amount of biomass per unit area and can be obtained from the OD680 of the resuspended harvest according to eq 4. k·OD680 ·V (4) A where k is the correlation constant between OD680 and dry weight density, and equals 0.267 g L−1 for Chlorella sp., V is the N=
12930
dx.doi.org/10.1021/ie501913k | Ind. Eng. Chem. Res. 2014, 53, 12927−12934
Industrial & Engineering Chemistry Research
Article
volume of the resuspended harvest (200 mL), and A is the illuminated area (0.0113 m2).
culture. Previous data showed that optimum specific growth rate of Chlorella sp. was achieved under an ambient temperature of 30 to 35 °C. Under an ambient temperature of about 20 °C, the specific growth rate was only one-third to that achieved under 30 to 35 °C. A possible reason for the different observations was that the medium temperature would be similar to the ambient temperature in the suspended growth system but could be much different in the FPBR. In fact, the shallow layer of culture medium in the cultivation chamber of the FPBR was found to be heated up to 27 °C under light illumination while the ambient temperature was only about 20 °C. Thus, there was no surprise that the specific growth rate would be similar to that under the medium temperature of 35 °C. For this reason, temperature control might not be necessary for the FPBR, even in a cool environment, as long as the medium in the cultivation chamber can be heated up to around 30 °C by light illumination. However, temperature control would be necessary if the medium could not reach the optimum temperature or was heated up to over 35 °C under strong illumination. The cultivated cells directly absorb nutrients from the medium in the cultivation chamber. Therefore, the flow rate of medium recycle should be selected to ensure that sufficient nutrients were provided to the microalgae in the cultivation chamber. Medium recycle rates of 5 and 10 mL min−1 were studied and the 48 h average specific growth rate was 0.0565 and 0.0569 h−1, respectively. If a medium recycle rate of 5 mL min−1 was too low to provide sufficient nutrients for microalgae, a significantly higher growth rate would be expected under a medium recycle rate of 10 mL min−1. Because the specific growth rate exhibited little increase when the medium recycle rate was doubled, the microalgae must have already been provided with sufficient nutrients at the medium recycle rate of 5 mL min−1. A lower recycle flow rate was also beneficial, as it reduced the energy cost for pumping the medium. The effect of CO2 availability on microalgae growth was also studied at two concentrations, 7.5% and 25%. This corresponded to 210 and 650 ppm equilibrium dissolved inorganic carbon species concentrations in the medium, represented as CaCO3, respectively, as measured by a carbon dioxide ion selective electrode (9502BNWP, Orion, USA). The average specific growth rates measured in the experiments were 0.0565 and 0.0575 h−1 for 7.5% and 25% CO2 in the feed gas, respectively. These experimental results corroborated with that from our previous experiments by cultivating Chlorella sp. in a bubble column. The specific growth rate was also kept at optimum when the equilibrium dissolved inorganic carbon species concentration in the medium ranged between 210 and 650 ppm, whereas the specific growth rate was lower if the concentration was below or above this range. This was because CO2 deficiency would directly constrain cell growth. Once it was increased to a range where all growth related enzymes were saturated, the specific growth rate could reach its optimum. However, extremely high concentrations of CO2 would cause unfavorably low medium pH and hamper cell growth instead. A similar trend was also observed by Widjaja et al.20 The amount of CO2 consumed by the FPBR at 7.5% and 25% CO2 in the feed gas were also determined, as illustrated in Figure 4. The CO2% in the outlet gas of the FPBR for control experiment without microalgae was stabilized at 4.2% and 24% for inlet gas with 7.5% and 25% CO2, respectively. This was probably because CO2 desorbed from the medium in the
4. RESULTS AND DISCUSSION 4.1. FPBR Start-up, Biomass Harvest, and Medium Discharge. The Chlorella sp. cells of 8−10 μm in diameter were added to the cultivation chamber. Within 20 min after inoculation, cells settled on the membrane with a pore size of 5 μm and left a clear layer of culture medium in the cultivation chamber. Compared with biofilm systems, which required a start-up period of up to 20 days,14 start-up of the FPBR was shown to be much faster. It was also noticed that the cells settled on the membrane did not form tight attachment to the membrane until about 48 h after inoculation. Before that, the green cells settled on the membrane could be easily resuspended when the cultivation chamber was tilted. Only after about 48 h of incubation would the biomass stay on the membrane when the supernatant was gently disturbed. The biomass could be easily scraped off from the membrane surface (Figure 3) at the end of 48 h after the medium was drained.
Figure 3. Scrape to harvest the biomass.
Samples were taken from the bubble column at the end of incubation and no microalgae were observed under a microscope. This indicated that microalgal cells were completely retained in the cultivation chamber and did not pass through the membrane so that the spent medium could be discharged from the medium preparation unit without the need of removing viable cells. 4.2. Effect of Key Operating Parameters on Microalgae Growth. The effect of temperature, CO2 availability, and medium recycle rate on microalgae growth were studied in the FPBR. Similar 48 h average specific growth rates of 0.0565 and 0.0562 h−1 were obtained for experiments conducted with heating (35 °C) and without heating (20 ± 1 °C), respectively, showing small influence of low ambient temperature on the growth of Chlorella sp. in the FPBR. These experimental results were very much different from our previous observation when Chlorella sp. was grown in a bubble column as a suspended 12931
dx.doi.org/10.1021/ie501913k | Ind. Eng. Chem. Res. 2014, 53, 12927−12934
Industrial & Engineering Chemistry Research
Article
Figure 5. COD, nitrogen (NO3−N), and phosphorus (PO4−P) concentration before and after 48 h of incubation using feed gas with 7.5% and 25% CO2. (The cultivation process was conducted under 100 ± 5 μmol m−2 s−1 and 35 °C. Medium recycle rate was 5 mL min−1.)
experiment was conducted for 4 days and biomass was harvested every 24 h to determine the amount of biomass per unit area, as summarized in Figure 6. According to eq 4, with OD680 measured for samples collected every 24 h, biomass productivity was calculated to be 4.81 g m−2 d−1 in Day I and increased to 13.56 g m−2 d−1 in Day II. At the end of Day II, the microalgal cells were found to be tightly attached to the membrane by observation. After that, lower productivities of 5.56 and 5.46 g m−2 d−1 were observed for Day III and Day IV, respectively. The daily biomass productivity varied as more biomass grew on the membrane. Throughout the experiment, three scenarios of membrane coverage by microalgae (Figure 7) were observed as the biomass concentration increased. At the beginning of the experiment, the membrane was only partially covered and cells were well illuminated (Figure 7a). After a period of time, the membrane was densely covered as cells proliferated. However, these cells were only loosely packed on the membrane as they could be easily resuspended with gentle disturbance of the supernatant. Under this scenario, although mutual shading appeared, light could still penetrate the cell layers so that most cells were illuminated (Figure 7b). As the amount of biomass on the membrane continued to increase, the membrane would be covered by layers of tightly packed cells (Figure 7c). In this case, only the top layer of cells could be illuminated while the bottom layers could not get access to light at all. Before the membrane was tightly covered by biomass (Figures 7a,b), all microalgal cells were expected to grow exponentially and the productivity could be expressed by eq 5 obtained from substituting eqs 2 and 4 into eq 3:
Figure 4. CO2 consumption by Chlorella sp. cultivated with a feed gas of (a) 7.5% CO2 and (b) 25% CO2. (The cultivation process was conducted under 100 ± 5 μmol m−2 s−1 and 35 °C. Medium recycle rate was 5 mL min−1.)
cultivation chamber and recycle chamber so that the medium in the preparation unit was capable of absorbing CO2 from the feed gas continuously. In the experiments with microalgae grown in the FPBR, the CO2% in the gas outlet was even lower and reached 3.3% and 22.5% for inlet gas with 7.5% and 25% CO2, respectively. With the control of the inlet flow rate at 200 mL/min, the molar flow rate of the inlet gas could be calculated using the ideal gas law. Assuming the molar flow rate of the gas outlet was equal to that of the inlet, we determined the maximum CO2 consumption rate was approximately 3.5 and 5.9 mg min−1 for inlet gas with 7.5% and 25% CO2, respectively. Figure 5 also shows the COD, nitrogen, and phosphorus concentrations before and after 48 h of microalgae cultivation. Metabolites were secreted into the medium during microalgae growth so that the COD of the culture supernatant was increased after 48 h of incubation. Nitrogen and phosphorus in the medium were also consumed as cells proliferated. Similar to the CO2 consumption rate, both nitrogen and phosphorus consumption rate by Chlorella sp. were higher under 25% CO2 than that under 7.5% CO2. This is probably because the cell metabolites such as lipids and enzymes were different in the cells cultivated under the two carbon availability levels.21 4.3. Selection of Operating Time and Initial Microalgae Concentration for Batch Operation. Experiments with operating conditions concluded in the above session were conducted to select the operating time and initial microalgae concentration for batch operation of the FPBR. The system was provided with 200 mL min−1 feed gas with 7.5% CO2 and surface illumination of 100 ± 5 μmol m−2 s−1, while medium recycle rate was kept at 5 mL min−1 (HRTc = 40 min) and medium temperature at 35 °C. The microalgae growth
Ne μΔt − N (5) Δt After the membrane was tightly covered, the productivity could not be expressed by eq 5 as, instead of all microalgal cells, only the top layer of cells was illuminated and could proliferate (Figure 7c). Productivity was determined by how many times the top layer of cells could be doubled, where the doubling time was determined by the specific growth rate of the top layer (μ) and was ln 2/μ. For an incubation period of Δt, there should be m1 complete cycles of doubling time and a period remained after these complete cycles (m2). For each complete cycle, the top layer of cells with a concentration of N0 would be reproduced to form a new layer of cells. Because the new layer P=
12932
dx.doi.org/10.1021/ie501913k | Ind. Eng. Chem. Res. 2014, 53, 12927−12934
Industrial & Engineering Chemistry Research
Article
Figure 6. Growth of biomass in the FPBR. (The cultivation process was conducted under 100 ± 5 μmol m−2 s−1 and 35 °C. Medium recycle rate was 5 mL min−1. Feed gas of 7.5% CO2 was supplied.)
loosely packed in Day I and II and the productivity would follow eq 5. As microalgae were tightly packed in Day III and Day IV, the productivity would follow eq 6. As can be deducted from eq 5, the daily productivity would increase as the amount of biomass on the membrane increased to fully utilize the illuminated area. However, once the packing of settled cells became tight and light penetration was severely hampered, the productivity of the FPBR would be low and determined only by the top layer of cells (N0). Biomass production per day should become stable after microalgae were tightly packed, as can be deducted from eq 6. The laboratory results validated these deductions, as the productivity increased from Day I to Day II, but decreased and was stable in Day III and Day IV. Hence, an optimum operating time and initial concentration of cells for batch operation should allow full utilization of the illuminated area while preventing cells from forming tight packing on the cultivation surface. With the productivity of Day III obtained from the experiment and assuming the specific growth rate of the top layer in Day III was similar to that in Day I (0.065 h−1), the biomass concentration of a fully illuminated layer (N0) in the FPBR could be calculated from eq 6 and was 2.54 g m−2. This indicated that in order to fully utilize the illuminated area, the initial concentration of biomass should be at least 2.54 g m−2. Meanwhile, it was observed that tight packing of cells would appear toward the end of Day II when the cell concentration approached 20 g m−2. Therefore, the optimum final concentration of biomass before harvest should be 20 g m−2. The optimum operating time for Chlorella sp. cultivation in the FPBR was thus selected to be the period for biomass concentration to grow from 2.54 g m−2 (inoculation) to 20 g m−2 (harvest), which was about 40 h.
Figure 7. Scenarios of membrane coverage. (a) Partially covered membrane, (b) loosely packed cells on the membrane, (c) tightly packed cells on the membrane.
of cells would completely cover the original one, the amount of biomass produced per unit area for m1 complete cycles of doubling time should be m1N0. For the time remained after m1 complete cycles (m2), the amount of cells produced per unit area from the top layer could be determined from eqs 2−4 and expressed as N0(eμm2 − 1). Therefore, the productivity of the FPBR after the membrane was tightly covered could be expressed by eq 6. P = N0
m1 + e μm2 − 1 Δt
5. CONCLUSIONS A novel attached growth system, filtration photobioreactor, for microalgae cultivation with easy harvest has been designed and studied. Without the need of biofilm formation, coexisting bacteria is not mandatory in microalgae cultivation and a shorter start-up time can be expected. As the microalgae are retained by the membrane, the spent medium is free of viable microalgal cells and will not cause contamination to the
(6)
From the observations of the 4 day experiment, microalgae did not tightly attach to the membrane until toward the end of Day II. Therefore, the microalgae could be considered as 12933
dx.doi.org/10.1021/ie501913k | Ind. Eng. Chem. Res. 2014, 53, 12927−12934
Industrial & Engineering Chemistry Research
Article
(6) Johnson, M. B.; Wen, Z. Development of an attached microalgal growth system for biofuel production. Appl. Microbiol. Biotechnol. 2010, 85, 525−534. (7) Akhtar, N.; Iqbal, J.; Iqbal, M. Removal and recovery of nickel(II) from aqueous solution by loofa sponge-immobilized biomass of Chlorella sorokiniana: Characterization studies. J. Hazard. Mater. 2004, 108, 85−94. (8) Lam, M. K.; Lee, K. T. Immobilization as a feasible method to simplify the separation of microalgae from water for biodiesel production. Chem. Eng. J. 2012, 191, 263−268. (9) Mulbry, W.; Kondrad, S.; Pizarro, C.; Kebede-Westhead, E. Treatment of dairy manure effluent using freshwater algae: Algal productivity and recovery of manure nutrients using pilot-scale algal turf scrubbers. Bioresour. Technol. 2008, 99, 8137−8142. (10) Ozkan, A.; Kinney, K.; Katz, L.; Berberoglu, H. Reduction of water and energy requirement of algae cultivation using an algae biofilm photobioreactor. Bioresour. Technol. 2012, 114, 542−548. (11) Schnurr, P. J.; Espie, G. S.; Allen, D. G. Algae biofilm growth and the potential to stimulate lipid accumulation through nutrient starvation. Bioresour. Technol. 2003, 136, 337−344. (12) Christenson, L. B.; Sims, R. C. Rotating algal biofilm reactor and spool harvester for wastewater treatment with biofuels by-products. Biotechnol. Bioeng. 2012, 109, 1674−1684. (13) Irving, T. E.; Allen, D. G. Species and material considerations in the formation and development of microalgal biofilms. Appl. Microbiol. Biotechnol. 2011, 92, 283−294. (14) Posadas, E.; García-Encina, P.; Soltau, A.; Domínguez, A.; Díaz, I.; Muñoz, R. Carbon and nutrient removal from centrates and domestic wastewater using algal-bacterial biofilm bioreactors. Bioresour. Technol. 2013, 139, 50−58. (15) Gressel, J.; van der Vlugt, C. J. B.; Bergmans, H. E. N. Environmental risks of large scale cultivation of microalgae: Mitigation of spills. Algal Res. 2013, 2, 286−298. (16) Zhang, X.; Hu, Q.; Sommerfeld, M.; Puruhito, E.; Chen, Y. Harvesting algal biomass for biofuels using ultrafiltration membranes. Bioresour. Technol. 2010, 101, 5297−5304. (17) Rossignol, N.; Vandanjon, L.; Jaouen, P.; Quéméneur, F. Membrane technology for the continuous separation microalgae/ culture medium: Compared performances of cross-flow microfiltration and ultrafiltration. Aquacult. Eng. 1999, 20, 191−208. (18) Fan, L. H.; Zhang, Y. T.; Zhang, L.; Chen, H. L. Evaluation of a membrane-sparged helical tubular photobioreactor for carbon dioxide biofixation by Chlorella vulgaris. J. Membr. Sci. 2008, 325, 336−345. (19) Liu, T.; Wang, J.; Hu, Q.; Cheng, P.; Ji, B.; Liu, J.; Chen, Y.; Zhang, W.; Chen, X.; Chen, J.; Gao, L.; Ji, C.; Wang, H. Attached cultivation technology of microalgae for efficient biomass feedstock production. Bioresour. Technol. 2013, 127, 216−222. (20) Widjaja, A.; Chien, C.; Ju, Y. Study of increasing lipid production from fresh water microalgae Chlorella vulgaris. J. Taiwan Inst. Chem. Eng. 2009, 40, 13−20. (21) Tsuzuki, M.; Ohnuma, E.; Sato, N.; Takaku, T.; Kawaguchi, A. Effects of CO2 concentration during growth on fatty acid composition in microalgae. Plant Physiol. 1990, 93, 851−856. (22) Moheimani, N. R. Long-term outdoor growth and lipid productivity of Tetraselmis suecica, Dunaliella tertiolecta and Chlorella sp (Chlorophyta) in bag photobioreactors. J. Appl. Phycol. 2013, 25, 167−176.
receiving water bodies during discharge. As all microalgae will attach to the membrane after the medium is drained in the cultivation chamber, they can be easily scraped off from the membrane and easy harvest can be realized in the FPBR. Also, by separating the medium preparation unit and the cultivation unit, designs such as packed towers can be used to improve the gas transfer rate inside the medium preparation unit. The effect of major operating parameters for Chlorella sp. cultivation was evaluated. Results showed that the FPBR can be readily warmed up by light illumination so that microalgae growth rate would be less affected if the ambient temperature is lower (20 °C) than the optimum cultivation temperature (30− 35 °C). The medium recycle rate of 5 mL min−1 was sufficient to supply enough nutrients to the microalgae. Experimental results also showed that similar growth rate was obtained for a feed gas of 7.5% and 25% CO2, corresponding to 210 and 650 ppm of dissolved inorganic carbon species in the medium. However, the CO2, nitrogen, and phosphorus consumption rates were all higher for inlet gas of 25% CO2 than that for inlet gas of 7.5% CO2. Biomass productivity of Chlorella sp. was also determined in this study. The highest biomass productivity was achieved before microalgal cells tightly attached to the membrane and was 13.5 g m−2 d−1, whereas the stable productivity with cells tightly packed on the membrane was 5.5 g m−2 d−1. The initial amount of biomass should be no less than 2.54 g m−2 to fully utilize the illuminated area, whereas the final concentration of biomass before harvest should be no more than 20 g m−2, to prevent tight packing of cells. This led to an operating time of about 40 h. Compared with a biofilm system for Chlorella sp. cultivation whose productivity was 2.57 g m−2 d−1, the productivity in the FPBR was much higher.6 However, the biomass productivity of Chlorella sp. (13.5 g m−2 d−1) was still lower than that in the suspended systems such as open ponds and enclosed photobioreactors, where the biomass productivity ranged from 20 to 40 g m−2 d−1.19,22 This was mainly because certain growth-related operating parameters such as mixing and light penetration were compromised for the sake of easy harvest of biomass in the FPBR.
■
AUTHOR INFORMATION
Corresponding Author
*K. M. Ng. E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
REFERENCES
(1) Pulz, O.; Gross, W. Valuable products from biotechnology of microalgae. Appl. Microbiol. Biotechnol. 2004, 65, 635−648. (2) Spolaore, P.; Joannis-Cassan, C.; Duran, E.; Isambert, A. Commercial applications of microalgae. J. Biosci. Bioeng. 2006, 101, 87−96. (3) Milledge, J. J. Commercial application of microalgae other than as biofuels: A brief review. Rev. Environ. Sci. Biotechnol. 2011, 10, 31−41. (4) Pires, J. C. M.; Alvim-Ferraz, M. C. M.; Martins, F. G.; Simões, M. Carbon dioxide capture from flue gases using microalgae: Engineering aspects and biorefinery concept. Renewable Sustainable Energy Rev. 2012, 16, 3043−3053. (5) Cao, J.; Yuan, W.; Pei, Z.; Davis, T.; Cui, Y.; Beltran, M. A preliminary study of the effect of surface texture on algae cell attachment for a mechanical−biological energy manufacturing system. J. Manuf. Sci. Eng. 2009, 131, 064505. 12934
dx.doi.org/10.1021/ie501913k | Ind. Eng. Chem. Res. 2014, 53, 12927−12934