Journal of Environmental Management 144 (2014) 118e124
Contents lists available at ScienceDirect
Journal of Environmental Management journal homepage: www.elsevier.com/locate/jenvman
Biomass production and nutrient removal by Chlorella sp. as affected by sludge liquor concentration € m a, *, Leiv M. Mortensen a, Bjørn Rusten b, Hans Ragnar Gislerød a Anette M. Åkerstro a b
Department of Plant Sciences, Norwegian University of Life Sciences, P.O. Box 5003, 1432 Aas, Norway Aquateam COWI AS, P.O. Box 6875, Rodelokka, N-0504 Oslo, Norway
a r t i c l e i n f o
a b s t r a c t
Article history: Received 16 December 2013 Received in revised form 3 April 2014 Accepted 19 May 2014 Available online
The use of microalgae for biomass production and nutrient removal from the reject water produced in the dewatering process of anaerobically digested sludge, sludge liquor, was investigated. The sludge liquor was characterized by a high content of total suspended solids (1590 mg L1), a high nitrogen concentration (1210 mg L1), and a low phosphorus concentration (28 mg L1). Chlorella sp. was grown in sludge liquor diluted with wastewater treatment plant effluent water to different concentrations (12, 25, 40, 50, 70, and 100%) using batch mode. The environmental conditions were 25 C, a continuous lightning of 115 mmol m2 s1, and a CO2 concentration of 3.0%. The highest biomass production (0.42 e0.45 g dry weight L1 Day1) was achieved at 40e50% sludge liquor, which was comparable to the production of the control culture grown with an artificial fertilizer. The biomass production was 0.12 and 0.26 g dry weight L1 Day1 at 12% and 100% sludge liquor, respectively. The percentage of nitrogen in the algal biomass increased from 3.6% in 12% sludge liquor and reached a saturation of ~10% in concentrations with 50% sludge liquor and higher. The phosphorus content in the biomass increased linearly from 0.2 to 1.5% with increasing sludge liquor concentrations. The highest nitrogen removal rates by algal biosynthesis were 33.6e42.6 mg TN L1 Day1 at 40e70% sludge liquor, while the highest phosphorus removal rates were 3.1e4.1 mg TP L1 Day1 at 50e100% sludge liquor. Published by Elsevier Ltd.
Keywords: Microalgae Nutrient removal Sludge liquor Biomass Nitrogen Phosphorus
1. Introduction At a wastewater treatment plant (WWTP), sludge treatment is a costly and complex operation. While many different options exist, anaerobic sludge digestion is the only form of sludge stabilization that produces bioenergy in the form of methane gas. The reject water derived from the dewatering process of anaerobically stabilized sludge, sludge liquor (SL), has a concentration of nitrogen in the order of 1000 mg L1, high chemical oxygen demand (COD), and high total suspended solids (TSS) content (Davis and Masten, 2009). The SL is usually returned to the head end of the WWTP and accounts for 15e25% of the total nitrogen load at the WWTP (Fux et al., 2006; Janus and van der Roest, 1997). Algae could potentially be integrated at a WWTP to treat the side-stream of SL and offers the combined benefits of nutrient removal, energy production, and CO2 sequestration (Rusten and Sahu, 2011; Sahu et al., 2013; Yuan et al., 2012). In this process, algae recover waste nutrients from the SL and transform them into biomass that can be
* Corresponding author. Tel.: þ47 98474400. €m). E-mail address: [email protected] (A.M. Åkerstro http://dx.doi.org/10.1016/j.jenvman.2014.05.015 0301-4797/Published by Elsevier Ltd.
harvested and co-digested with sludge to produce more energy, in the form of biogas, at the WWTP. In addition, flue gas from burning of biogas in the combined heat and power unit (CHP), or CO2 from the digesters may be used as a CO2 source for the algal bioreactor, which can reduce the CO2 footprint of the WWTP (Gronlund et al., 2004; Rusten et al., 2009). The main challenge of using SL for algal cultivation is the high TSS content, which can vary from 790 to 3700 mg L1, with corresponding light transmission values of 16 to 0.1% (670 nm, 1 cm light path) (Rusten and Sahu, 2011). In previous studies of reject water from digested sludge, different pretreatment methods were used to remove the TSS content (Rusten and Sahu, 2011; Udom et al., 2013; Yuan et al., 2012). Increasing the light transmission by diluting with WWTP effluent water may be a preferred method due to its simplicity. A disadvantage of using dilution is that it lowers the nutrient availability and therefore reduces the nutrient uptake rate which in turn may lead to a reduction in both nutrient removal and algal growth rates (Aksnes and Egge, 1991; Hessen et al., 2002; Marschner, 1986). This is the only study, to our knowledge, that uses a high strength reject water from anaerobically digested sludge with a high TSS content (>1 g L1) for autotrophic algae cultivation. The aim was to evaluate the efficiency
€m et al. / Journal of Environmental Management 144 (2014) 118e124 A.M. Åkerstro
of algal biomass production and nutrient removal in different concentrations of raw, unsterilized SL. 2. Methods 2.1. Strain selection An initial experiment was carried out to find a suitable strain of algae for growth in SL. The genus Chlorella was chosen for its high protein content that implies a high nitrogen demand (Kay, 1991; Taub and Dollar, 1965). A screening of 8 strains of Chlorella isolated from locations in Norway was carried out. The strains were received from the Norwegian Institute for Water Research (NIVA) in Oslo, Norway, and were grown in a mixture of 30% SL and 70% effluent water from a rotating biological contactor (RBC) WWTP. The screening was carried out inside a greenhouse compartment using natural light (average PAR were 17 ± 9 mol m2 Day1), a controlled temperature of 25 C and an airflow of 3.0% CO2. After 6 days of cultivation, the strain Chlorella sp. 137 had the highest NTU value and was chosen for this study (Table S1). 2.2. Experimental set-up and cultivation conditions SL and effluent water were collected at Nordre Follo Renseanlegg, a municipal WWTP located in Vinterbro, Norway, with a population equivalent of 41 000. The SL was collected from the dewatering process of the sludge after mesophilic anaerobic digestion: centrifugation with the addition of a cationic polymer and tap water. The SL had a brown color and the following composition: 1590 mg L1 TSS, 3780 mg L1 COD, 906 mg L1 ammonium (NH4eN), and 28 mg L1 total phosphorus (TP). The WWTP uses chemical precipitation for P-removal by in the liquidhandling part of the WWTP and results in a low TP concentration in the SL. Effluent water was collected at the outlet of the dissolved air flotation basin, which was the last step before discharge. The chemical compositions of SL and effluent water are shown in Table S1. The experimental set-up is visualized in Fig. 1. The SL was diluted with effluent water to produce mixtures with 12, 25, 40, 50, and 70% SL. The light transmission of the SL was 2.8% at 100% SL and increased to 85, 71, 20, 15, and 8% for mixtures with 12, 25, 40, 50, and 70% SL, respectively. Growth was compared to a control culture in which Chlorella sp. was grown in a medium composed of a € Oy, Finland), horticultural fertilizer (Superex vegetables, Kekkila phosphate (KPO4), nitrogen (Urea) and dissolved in tap water. The
119
final nitrogen and phosphorus concentrations were 552 and 86 mg L1, respectively (Table S1). Chlorella sp. was cultured in batch cultures, using glass cylinders with an inner diameter of 35 mm and a height of 400 mm. These were placed in a thermostatic water bath (25.0 ± 0.5 C) and continuously illuminated by a panel of white fluorescence tubes (18 W, Narva, Germany), providing a photon flux density of 115 mmol m2 s1. An airflow containing 3.0 ± 0.2% CO2 (from pure liquid CO2) was bubbled through the cultures after passing through a sterile filter (Acro® 37 TF Vent Devices, 0.2 mm). By experience, a concentration of 3% CO2 was needed for avoiding a rise in pH above 8. Before the experiment, the algae were adapted to their specific environmental conditions; they were grown in their respective SL concentrations using a semi-batch mode that included three dilutions. The concentration of algae at the start of the experiment was 0.01 g L1 and corresponded to an optical density (OD) of 0.05 at 680 nm. OD was measured daily for an approximate estimate of growth and TSS. The experiment was carried out until the growth curve of OD leveled out for two days. The nutrient removal due to algal biosynthesis was separated from other mechanical or physical processes by analyzing the algal biomass for the various nutrient elements at the end of the cultivation period. The aluminum content was also measured to test for possible residues from the chemical P-removal process. All cultures were made in triplicates. 2.3. Analytical methods The following parameters of the wastewater were analyzed, using Hach-Lange kits (Hach Lange, Germany): chemical oxygen demand (COD), total nitrogen (TN), ammonium nitrogen (NH4eN), nitrate nitrogen (NO3eN), total phosphorus (TP), and phosphate (PO4eP). The elements S, K, Ca, Fe, Mg, Mn, Zn, Cu, and Al were analyzed with an Inductively Coupled Plasma Optical Emission Spectrometer (ICP) (Optima 5300 DV, Perkin Elmer, USA) after the addition of HNO3 to 10% v/v and were decomposed by UltraClave (UltraClave III, MLS, Leutkirch) at 250 C for 1.5 h. A calibrated pHmeter, Orion Aplus TM (Thermo Electron Corporation, USA), was used to measure pH. Turbidity (NTU) was measured with a HACH 2100AN turbidimeter (HACH Company, USA). Light transmission (%) was measured at 670 nm (Rusten and Sahu, 2011) and was compared with that of de-ionized water (using water for the zerobase measurements) (Unicam Helios, UVevis spectrometry) in 1 cm light path. The OD of the algae culture was measured at 680 nm and was compared with the respective SL dilution, without algae (L. Wang et al., 2010).
Fig. 1. The experimental set-up for evaluating the nutrient removal and biomass productivity of Chlorella sp. as affected by sludge liquor concentration (12, 25 40, 50, 70 and 100%) diluted with WWTP effluent water and compared to a control culture composed of a horticultural fertilizer.
€m et al. / Journal of Environmental Management 144 (2014) 118e124 A.M. Åkerstro
120
TSS were measured using pre-weighed and pre-washed GF/F Whatman glass microfiber filters and were dried at 105 C for 3 h. The filters were weighed on a Mettler Toledo XP6 microbalance (Mettler-Toledo Inc., USA). The algae's dry weight was determined by subtracting the TSS value of the SL without algae. The volumetric biomass production (g dry weight L1 Day1) was determined by the method of least square regression, after removing the lag and stationary phase. The algal biomass was analyzed by harvesting 10 mL of the culture by centrifugation (Eppendorf Centrifuge 5810) at 4000 rpm for 10 min. The cell pellet was washed three times by re-suspending it in Milli-Q water (Milli-Q Synthesis A10, Millipore Corporation, USA), and it was then centrifuged at 4000 rpm for 5 min. The washed algae pellets were re-suspended in 10 mL of Milli-Q water and frozen (at 21 C) until they were analyzed. The Norwegian standard NS 4743 method was used for measuring TN. The nutrient elements P, K, S, Ca, Fe, Cu, Zn, Mg, and Mn were measured by ICP using the same method as for the wastewater, but after automatically diluting the samples 50e200 times. As a standard, NutrientsWP (RTC) was used in concentrations of 0.5 and 5.5 mg L1. The cellular content of the various nutrients was calculated as a percentage of dry weight: Cellular content (%) ¼ 100*nutrient element (gL1)/dry weight (gL1). The nutrient removal due to algal biosynthesis was calculated by using the values of cellular content of nutrients and multiplying these values by the overall biomass production: Nutrient removal (mgL1) ¼ (cellular content (%)/100) *biomass (mgL1). The nutrient removal rates were calculated as follows: Nutrient removal rates (mg L1 Day1) ¼ (cellular content (%)/100)*volumetric productivity (mg L1 Day1). Bacteria were present in the SL, but their proportion of biomass in the samples was assumed negligible, as most were rinsed off when washing the cell pellet. The method of least square regression was executed using Excel. Differences between the cultures were tested with the ANOVA and Tuckey's test in Minitab 16 software, using a 95% confidence level. The graphs were made using SigmaPlot 10.0. 3. Results and discussion 3.1. Biomass productivity The biomass was harvested when the OD measurements leveled out, which occurred after 5 days for the control, after 7 days for 12% SL dilution, and after 8 days for the other dilutions. At the time of harvest, the dry weights of the different SL concentrations varied from 0.96 g L1 for 12% SL to 2.11 g L1 for 40% SL (Table 1). In general, linear biomass growth started at a dry weight concentration of 0.2 g L1 and lasted for at least 3 days. Linear biomass growth indicates the productivity in a continuous system and is a comparable parameter, as it does not include lag or stationary phases. The results showed that the optimal light and nutrient conditions for biomass production were achieved at 40e50% SL that reached the same productivity (0.45e0.42 g L1 Day1) as the
control culture. In 70% SL, productivity was 0.38 g L1 Day1, and in 100% SL, the productivity was further reduced to 0.26 g L1 Day1. In a study by Osendeko and Pittman (2014) that used SL from activated sludge, prior to digestion, growth was reduced already at 40% SL and was assumed to be due to oxidative stress resulting from a high concentration of Na and metals such as Fe and Zn. In this study, the high TSS contents and their resultant low light transmissions assumingly reduced the productivity in the highest SL concentrations (L.A. Wang et al., 2010). At 100% SL the light transmission was 2.8% compared to 40 and 50% SL that had a transmission of 20 and 15%, respectively. Although growth was reduced at 100% SL the results imply that algae may be used for bioremediation of other types of wastewater with a limited light transmission, such as industrial wastewater with dyestuff (Naim and El Abd, 2002). Another growth reducing factor may have been the high ammonium content. Tam and Wong (1996) showed that the growth of Chlorella was reduced at >700 mg L1 ammonium. Although they refer to ammonia (NH3eN) in their study the pH levels they listed were consistently below 7 and the main form must then be as ammonium (NH4eN) nitrogen. The growth of Oscillatoria sp. was reduced already at a concentration of 100 mg L1 NH4eN but with pH levels listed as being on the alkaline side (Hashimoto and Furukawa, 1989). The dissociation of ammonium to ammonia occurs at pH levels above 8. Ammonia is highly membrane permeable and dissipates the pH gradients within the cell (Taiz and Zeiger, 2006). Insertion of 3% CO2 was sufficient for keeping the pH below 8 and thereby avoiding the toxic effects of ammonia. The pH was stable in all SL concentrations except at 70% and 100% SL, in which the pH dropped from 8.1 to 6.1 (70% SL) and 6.6 (100% SL) (Table 2) and seemed to be linked to a rapid growth since it coincided with the linear biomass growth. In SL concentrations
Contents lists available at ScienceDirect
Journal of Environmental Management journal homepage: www.elsevier.com/locate/jenvman
Biomass production and nutrient removal by Chlorella sp. as affected by sludge liquor concentration € m a, *, Leiv M. Mortensen a, Bjørn Rusten b, Hans Ragnar Gislerød a Anette M. Åkerstro a b
Department of Plant Sciences, Norwegian University of Life Sciences, P.O. Box 5003, 1432 Aas, Norway Aquateam COWI AS, P.O. Box 6875, Rodelokka, N-0504 Oslo, Norway
a r t i c l e i n f o
a b s t r a c t
Article history: Received 16 December 2013 Received in revised form 3 April 2014 Accepted 19 May 2014 Available online
The use of microalgae for biomass production and nutrient removal from the reject water produced in the dewatering process of anaerobically digested sludge, sludge liquor, was investigated. The sludge liquor was characterized by a high content of total suspended solids (1590 mg L1), a high nitrogen concentration (1210 mg L1), and a low phosphorus concentration (28 mg L1). Chlorella sp. was grown in sludge liquor diluted with wastewater treatment plant effluent water to different concentrations (12, 25, 40, 50, 70, and 100%) using batch mode. The environmental conditions were 25 C, a continuous lightning of 115 mmol m2 s1, and a CO2 concentration of 3.0%. The highest biomass production (0.42 e0.45 g dry weight L1 Day1) was achieved at 40e50% sludge liquor, which was comparable to the production of the control culture grown with an artificial fertilizer. The biomass production was 0.12 and 0.26 g dry weight L1 Day1 at 12% and 100% sludge liquor, respectively. The percentage of nitrogen in the algal biomass increased from 3.6% in 12% sludge liquor and reached a saturation of ~10% in concentrations with 50% sludge liquor and higher. The phosphorus content in the biomass increased linearly from 0.2 to 1.5% with increasing sludge liquor concentrations. The highest nitrogen removal rates by algal biosynthesis were 33.6e42.6 mg TN L1 Day1 at 40e70% sludge liquor, while the highest phosphorus removal rates were 3.1e4.1 mg TP L1 Day1 at 50e100% sludge liquor. Published by Elsevier Ltd.
Keywords: Microalgae Nutrient removal Sludge liquor Biomass Nitrogen Phosphorus
1. Introduction At a wastewater treatment plant (WWTP), sludge treatment is a costly and complex operation. While many different options exist, anaerobic sludge digestion is the only form of sludge stabilization that produces bioenergy in the form of methane gas. The reject water derived from the dewatering process of anaerobically stabilized sludge, sludge liquor (SL), has a concentration of nitrogen in the order of 1000 mg L1, high chemical oxygen demand (COD), and high total suspended solids (TSS) content (Davis and Masten, 2009). The SL is usually returned to the head end of the WWTP and accounts for 15e25% of the total nitrogen load at the WWTP (Fux et al., 2006; Janus and van der Roest, 1997). Algae could potentially be integrated at a WWTP to treat the side-stream of SL and offers the combined benefits of nutrient removal, energy production, and CO2 sequestration (Rusten and Sahu, 2011; Sahu et al., 2013; Yuan et al., 2012). In this process, algae recover waste nutrients from the SL and transform them into biomass that can be
* Corresponding author. Tel.: þ47 98474400. €m). E-mail address: [email protected] (A.M. Åkerstro http://dx.doi.org/10.1016/j.jenvman.2014.05.015 0301-4797/Published by Elsevier Ltd.
harvested and co-digested with sludge to produce more energy, in the form of biogas, at the WWTP. In addition, flue gas from burning of biogas in the combined heat and power unit (CHP), or CO2 from the digesters may be used as a CO2 source for the algal bioreactor, which can reduce the CO2 footprint of the WWTP (Gronlund et al., 2004; Rusten et al., 2009). The main challenge of using SL for algal cultivation is the high TSS content, which can vary from 790 to 3700 mg L1, with corresponding light transmission values of 16 to 0.1% (670 nm, 1 cm light path) (Rusten and Sahu, 2011). In previous studies of reject water from digested sludge, different pretreatment methods were used to remove the TSS content (Rusten and Sahu, 2011; Udom et al., 2013; Yuan et al., 2012). Increasing the light transmission by diluting with WWTP effluent water may be a preferred method due to its simplicity. A disadvantage of using dilution is that it lowers the nutrient availability and therefore reduces the nutrient uptake rate which in turn may lead to a reduction in both nutrient removal and algal growth rates (Aksnes and Egge, 1991; Hessen et al., 2002; Marschner, 1986). This is the only study, to our knowledge, that uses a high strength reject water from anaerobically digested sludge with a high TSS content (>1 g L1) for autotrophic algae cultivation. The aim was to evaluate the efficiency
€m et al. / Journal of Environmental Management 144 (2014) 118e124 A.M. Åkerstro
of algal biomass production and nutrient removal in different concentrations of raw, unsterilized SL. 2. Methods 2.1. Strain selection An initial experiment was carried out to find a suitable strain of algae for growth in SL. The genus Chlorella was chosen for its high protein content that implies a high nitrogen demand (Kay, 1991; Taub and Dollar, 1965). A screening of 8 strains of Chlorella isolated from locations in Norway was carried out. The strains were received from the Norwegian Institute for Water Research (NIVA) in Oslo, Norway, and were grown in a mixture of 30% SL and 70% effluent water from a rotating biological contactor (RBC) WWTP. The screening was carried out inside a greenhouse compartment using natural light (average PAR were 17 ± 9 mol m2 Day1), a controlled temperature of 25 C and an airflow of 3.0% CO2. After 6 days of cultivation, the strain Chlorella sp. 137 had the highest NTU value and was chosen for this study (Table S1). 2.2. Experimental set-up and cultivation conditions SL and effluent water were collected at Nordre Follo Renseanlegg, a municipal WWTP located in Vinterbro, Norway, with a population equivalent of 41 000. The SL was collected from the dewatering process of the sludge after mesophilic anaerobic digestion: centrifugation with the addition of a cationic polymer and tap water. The SL had a brown color and the following composition: 1590 mg L1 TSS, 3780 mg L1 COD, 906 mg L1 ammonium (NH4eN), and 28 mg L1 total phosphorus (TP). The WWTP uses chemical precipitation for P-removal by in the liquidhandling part of the WWTP and results in a low TP concentration in the SL. Effluent water was collected at the outlet of the dissolved air flotation basin, which was the last step before discharge. The chemical compositions of SL and effluent water are shown in Table S1. The experimental set-up is visualized in Fig. 1. The SL was diluted with effluent water to produce mixtures with 12, 25, 40, 50, and 70% SL. The light transmission of the SL was 2.8% at 100% SL and increased to 85, 71, 20, 15, and 8% for mixtures with 12, 25, 40, 50, and 70% SL, respectively. Growth was compared to a control culture in which Chlorella sp. was grown in a medium composed of a € Oy, Finland), horticultural fertilizer (Superex vegetables, Kekkila phosphate (KPO4), nitrogen (Urea) and dissolved in tap water. The
119
final nitrogen and phosphorus concentrations were 552 and 86 mg L1, respectively (Table S1). Chlorella sp. was cultured in batch cultures, using glass cylinders with an inner diameter of 35 mm and a height of 400 mm. These were placed in a thermostatic water bath (25.0 ± 0.5 C) and continuously illuminated by a panel of white fluorescence tubes (18 W, Narva, Germany), providing a photon flux density of 115 mmol m2 s1. An airflow containing 3.0 ± 0.2% CO2 (from pure liquid CO2) was bubbled through the cultures after passing through a sterile filter (Acro® 37 TF Vent Devices, 0.2 mm). By experience, a concentration of 3% CO2 was needed for avoiding a rise in pH above 8. Before the experiment, the algae were adapted to their specific environmental conditions; they were grown in their respective SL concentrations using a semi-batch mode that included three dilutions. The concentration of algae at the start of the experiment was 0.01 g L1 and corresponded to an optical density (OD) of 0.05 at 680 nm. OD was measured daily for an approximate estimate of growth and TSS. The experiment was carried out until the growth curve of OD leveled out for two days. The nutrient removal due to algal biosynthesis was separated from other mechanical or physical processes by analyzing the algal biomass for the various nutrient elements at the end of the cultivation period. The aluminum content was also measured to test for possible residues from the chemical P-removal process. All cultures were made in triplicates. 2.3. Analytical methods The following parameters of the wastewater were analyzed, using Hach-Lange kits (Hach Lange, Germany): chemical oxygen demand (COD), total nitrogen (TN), ammonium nitrogen (NH4eN), nitrate nitrogen (NO3eN), total phosphorus (TP), and phosphate (PO4eP). The elements S, K, Ca, Fe, Mg, Mn, Zn, Cu, and Al were analyzed with an Inductively Coupled Plasma Optical Emission Spectrometer (ICP) (Optima 5300 DV, Perkin Elmer, USA) after the addition of HNO3 to 10% v/v and were decomposed by UltraClave (UltraClave III, MLS, Leutkirch) at 250 C for 1.5 h. A calibrated pHmeter, Orion Aplus TM (Thermo Electron Corporation, USA), was used to measure pH. Turbidity (NTU) was measured with a HACH 2100AN turbidimeter (HACH Company, USA). Light transmission (%) was measured at 670 nm (Rusten and Sahu, 2011) and was compared with that of de-ionized water (using water for the zerobase measurements) (Unicam Helios, UVevis spectrometry) in 1 cm light path. The OD of the algae culture was measured at 680 nm and was compared with the respective SL dilution, without algae (L. Wang et al., 2010).
Fig. 1. The experimental set-up for evaluating the nutrient removal and biomass productivity of Chlorella sp. as affected by sludge liquor concentration (12, 25 40, 50, 70 and 100%) diluted with WWTP effluent water and compared to a control culture composed of a horticultural fertilizer.
€m et al. / Journal of Environmental Management 144 (2014) 118e124 A.M. Åkerstro
120
TSS were measured using pre-weighed and pre-washed GF/F Whatman glass microfiber filters and were dried at 105 C for 3 h. The filters were weighed on a Mettler Toledo XP6 microbalance (Mettler-Toledo Inc., USA). The algae's dry weight was determined by subtracting the TSS value of the SL without algae. The volumetric biomass production (g dry weight L1 Day1) was determined by the method of least square regression, after removing the lag and stationary phase. The algal biomass was analyzed by harvesting 10 mL of the culture by centrifugation (Eppendorf Centrifuge 5810) at 4000 rpm for 10 min. The cell pellet was washed three times by re-suspending it in Milli-Q water (Milli-Q Synthesis A10, Millipore Corporation, USA), and it was then centrifuged at 4000 rpm for 5 min. The washed algae pellets were re-suspended in 10 mL of Milli-Q water and frozen (at 21 C) until they were analyzed. The Norwegian standard NS 4743 method was used for measuring TN. The nutrient elements P, K, S, Ca, Fe, Cu, Zn, Mg, and Mn were measured by ICP using the same method as for the wastewater, but after automatically diluting the samples 50e200 times. As a standard, NutrientsWP (RTC) was used in concentrations of 0.5 and 5.5 mg L1. The cellular content of the various nutrients was calculated as a percentage of dry weight: Cellular content (%) ¼ 100*nutrient element (gL1)/dry weight (gL1). The nutrient removal due to algal biosynthesis was calculated by using the values of cellular content of nutrients and multiplying these values by the overall biomass production: Nutrient removal (mgL1) ¼ (cellular content (%)/100) *biomass (mgL1). The nutrient removal rates were calculated as follows: Nutrient removal rates (mg L1 Day1) ¼ (cellular content (%)/100)*volumetric productivity (mg L1 Day1). Bacteria were present in the SL, but their proportion of biomass in the samples was assumed negligible, as most were rinsed off when washing the cell pellet. The method of least square regression was executed using Excel. Differences between the cultures were tested with the ANOVA and Tuckey's test in Minitab 16 software, using a 95% confidence level. The graphs were made using SigmaPlot 10.0. 3. Results and discussion 3.1. Biomass productivity The biomass was harvested when the OD measurements leveled out, which occurred after 5 days for the control, after 7 days for 12% SL dilution, and after 8 days for the other dilutions. At the time of harvest, the dry weights of the different SL concentrations varied from 0.96 g L1 for 12% SL to 2.11 g L1 for 40% SL (Table 1). In general, linear biomass growth started at a dry weight concentration of 0.2 g L1 and lasted for at least 3 days. Linear biomass growth indicates the productivity in a continuous system and is a comparable parameter, as it does not include lag or stationary phases. The results showed that the optimal light and nutrient conditions for biomass production were achieved at 40e50% SL that reached the same productivity (0.45e0.42 g L1 Day1) as the
control culture. In 70% SL, productivity was 0.38 g L1 Day1, and in 100% SL, the productivity was further reduced to 0.26 g L1 Day1. In a study by Osendeko and Pittman (2014) that used SL from activated sludge, prior to digestion, growth was reduced already at 40% SL and was assumed to be due to oxidative stress resulting from a high concentration of Na and metals such as Fe and Zn. In this study, the high TSS contents and their resultant low light transmissions assumingly reduced the productivity in the highest SL concentrations (L.A. Wang et al., 2010). At 100% SL the light transmission was 2.8% compared to 40 and 50% SL that had a transmission of 20 and 15%, respectively. Although growth was reduced at 100% SL the results imply that algae may be used for bioremediation of other types of wastewater with a limited light transmission, such as industrial wastewater with dyestuff (Naim and El Abd, 2002). Another growth reducing factor may have been the high ammonium content. Tam and Wong (1996) showed that the growth of Chlorella was reduced at >700 mg L1 ammonium. Although they refer to ammonia (NH3eN) in their study the pH levels they listed were consistently below 7 and the main form must then be as ammonium (NH4eN) nitrogen. The growth of Oscillatoria sp. was reduced already at a concentration of 100 mg L1 NH4eN but with pH levels listed as being on the alkaline side (Hashimoto and Furukawa, 1989). The dissociation of ammonium to ammonia occurs at pH levels above 8. Ammonia is highly membrane permeable and dissipates the pH gradients within the cell (Taiz and Zeiger, 2006). Insertion of 3% CO2 was sufficient for keeping the pH below 8 and thereby avoiding the toxic effects of ammonia. The pH was stable in all SL concentrations except at 70% and 100% SL, in which the pH dropped from 8.1 to 6.1 (70% SL) and 6.6 (100% SL) (Table 2) and seemed to be linked to a rapid growth since it coincided with the linear biomass growth. In SL concentrations
