Scholarly article on topic 'Influence of microalgal N and P composition on wastewater nutrient remediation'

Influence of microalgal N and P composition on wastewater nutrient remediation Academic research paper on "Chemical engineering"

CC BY
0
0
Share paper
Academic journal
Water Research
OECD Field of science
Keywords
{Microalgae / "Internal composition" / Species / Biomass / Immobilisation}

Abstract of research paper on Chemical engineering, author of scientific article — Rachel Whitton, Amandine Le Mével, Marc Pidou, Francesco Ometto, Raffaella Villa, et al.

Abstract Microalgae have demonstrated the ability to remediate wastewater nutrients efficiently, with methods to further enhance performance through species selection and biomass concentration. This work evaluates a freshwater species remediation characteristics through analysis of internal biomass N:P (nitrogen:phosphorus) and presents a relationship between composition and nutrient uptake ability to assist in species selection. Findings are then translated to an optimal biomass concentration, achieved through immobilisation enabling biomass intensification by modifying bead concentration, for wastewaters of differing nutrient concentrations at hydraulic retention times (HRT) from 3 h to 10 d. A HRT <20 h was found suitable for the remediation of secondary effluent by immobilised Scenedesmus obliquus and Chlorella vulgaris at bead concentrations as low as 3.2 and 4.4 bead·mL−1. Increasing bead concentrations were required for shorter HRTs with 3 h possible at influent concentrations <5 mgP L−1.

Academic research paper on topic "Influence of microalgal N and P composition on wastewater nutrient remediation"

Accepted Manuscript

WATER RESEARCH

Influence of microalgal N and P composition on wastewater nutrient remediation

Rachel Whitton, Amandine Le Mevel, Marc Pidou, Francesco Ometto, Raffaella Villa, Bruce Jefferson

PII: S0043-1354(15)30452-8

DOI: 10.1016/j.watres.2015.12.054

Reference: WR 11752

To appear in: Water Research

Received Date: 28 July 2015 Revised Date: 21 December 2015 Accepted Date: 31 December 2015

Please cite this article as: Whitton, R., Le Mével, A., Pidou, M., Ometto, F., Villa, R., Jefferson, B., Influence of microalgal N and P composition on wastewater nutrient remediation, Water Research (2016), doi: 10.1016/j.watres.2015.12.054.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Influence of microalgal N and P composition on wastewater nutrient remediation

Rachel Whitton a, Amandine Le Mevel b, Marc Pidou a, Francesco Ometto c, Raffaella Villa a and Bruce Jefferson a'*

a Cranfield University, Cranfield, MK43 0AL, (UK)

b Ecole Nationale Superieure de Chimie de Rennes, 35708 Rennes Cedex 7 (FR) c Scandinavian Biogas Fuels AB, Linkoping University, 58 183, Linkoping (SE)

'Corresponding author: b.jefferson@cranfield.ac.uk Abstract

Microalgae have demonstrated the ability to remediate wastewater nutrients efficiently, with methods to further enhance performance through species selection and biomass concentration. This work evaluates a freshwater species remediation characteristics through analysis of internal biomass N:P (nitrogen:phosphorus) and presents a relationship between composition and nutrient uptake ability to assist in species selection. Findings are then translated to an optimal biomass concentration, achieved through immobilisation enabling biomass intensification by modifying bead concentration, for wastewaters of differing nutrient concentrations at hydraulic retention times (HRT) from 3 h to 10 d. A HRT <20h was found suitable for the remediation of secondary effluent by immobilised Scenedesmus obliquus and Chlorella vulgaris at bead concentrations as low as 3.2 and 4.4 bead-mL"1. Increasing bead concentrations were required for shorter HRTs with 3 h possible at influent concentrations < 5mgP-L-1.

Keywords: microalgae, internal composition, species, biomass, immobilisation 1 Introduction

Microalgae are photosynthetic organisms that assimilate nitrogen (N) and phosphorus (P) during their growth. The subsequent biomass generated can be converted into

25 energy or further raw materials following appropriate processing (Ometto et al., 2014),

26 offering benefits in its use and renewing interest in a microalgae based technology for

27 wastewater nutrient remediation.

28 Nutrient remediation characteristics for N and P have been shown to positively correlate

29 to growth rate (Xin et al., 2010) with growth a function of internal rather than external

30 nutrient concentration (Portielje and Lijklema, 1994). The internal composition of

31 marine phytoplankton has been established as 106:16:1 as a molar ratio for C:N:P,

32 known as the Redfield Ratio (Redfield, 1934). However, in the case of freshwater

33 microalgae, the Redfield Ratio is an exception rather than a rule with N:P molar ratios

34 ranging between 8:1 and 45:1 (Hecky et al., 1993) through a species' specific cellular

35 quota for structural components and storage for growth (Droop, 1968). More

36 importantly, freshwater microalgae have been shown to be able to adjust the N and P

37 concentration in their biomass in relation to the surrounding concentration in the water

38 (Beuckels et al., 2015; Choi and Lee, 2015) with biomass P accumulation influenced by

39 the external P and N supply whereas N accumulation is independent of P (Beuckels et

40 al., 2015). This behaviour is due to the predominate use of nitrogen for protein synthesis

41 with P incorporated into ribosomal RNA. Accordingly, under limited nutrient

42 conditions cell growth is reduced whilst carbon uptake continues (through

43 photosynthesis) resulting in enrichment of carbohydrates or lipids. This is often

44 exploited prior to bioenergy recovery to maximise yield for the microalgae biomass

45 (Craggs et al., 2013). In high nutrient environments, microalgae can also accumulate

46 excess nutrients through luxury uptake pathways (Eixler at el., 2006) enabling

47 adaptation across a wide range of environmental situations. Such flexibility in nutrient

48 compositions enables microalgae to successfully adapt to the local environment and

influences the biochemical composition of the resultant biomass (Loladze and Elser, 2011, Choi and Lee, 2015).

Furthermore, the nutrient remediation characteristics of microalgal species have been correlated to the internal elemental concentration, with P remediation inversely correlated to biochemical composition (Choi and Lee, 2015; Ruiz-Martinez et al., 2015). With the nutrient concentration in microalgal biomass shown to vary significantly from 0.03 - 3% of dry mass for P and between 3 - 12 % for N (Reynolds, 2006), the design of microalgae reactors for wastewater treatment based on fixed stoichiometry (Redfield Ratio) are not likely to be reliable. Studies to date have analysed the impact of varying N:P mediums on cell composition, or evaluated a suitable wastewater nutrient balance for microalgal treatment in relation to internal composition for a specific species (Choi and Lee, 2015). It is posited however, that the efficacy of nutrient remediation can be further enhanced through a targeted selection of a species with a suitable composition following adaptation to a balance of nutrients in a wastewater to be processed, thereby achieving an enhanced level of remediation.

Furthermore, the majority of the work to date on microalgal wastewater nutrient remediation has considered suspended microalgal biomass operated in relatively passive technologies such as high rate algae ponds (HRAP). HRAPs are typically configured as raceways ponds with shallow depths (20-60 cm) containing dilute biomass concentrations around 0.3 g(DW)-L-1. Biomass concentration is relatively low through the variability of the light source (solar radiation) and associated poor light efficiency, in addition to other external factors related to open systems including temperature, predation and contamination (Park et al., 2011). Consideration and uptake of microalgae based technology for wastewater treatment is restricted in many countries due to the

73 large footprints associated with the required long HRTs and shallow depths (Lundquist

74 et al., 2010). Intensification of the algae biomass (and reduction in footprint) can be

75 achieved through immobilisation where the biomass is encapsulated within an alginate

76 gel affording biomass concentrations of up to 3.3 g(DW)-L-1 (Chevalier and De la

77 Noue, 1985). Whilst the technology is within its infancy, remediation of PO4-P and

78 NH4-N from secondary wastewater effluents from 1.1 mgP-L-1 and 2.6 mgN-L-1 to 0.07

79 mgP-L-1 and 0.02 mgN-L-1 have been demonstrated for immobilised S.obliquus within

80 hydraulic retention times of 6 hours (Whitton et al., 2014). In addition to a concentrated

81 biomass and reduced HRT, immobilisation facilitates the removal of biomass post-

82 treatment through gravity settlement; eliminating costs associated with harvesting

83 technologies which require coagulation and intensive energy requirements i.e.

84 centrifugation. Following these positive attributes, the immobilised technology warrants

85 further research to determine whether the solution can be optimised for adequate

86 treatment within suitable HRTs prior to further development to improve its suitability

87 for application within the wastewater treatment industry e.g. operational costs related to

88 bead longevity and resin material.

89 Immobilisation affords the ability to seed and maintain a chosen species or community

90 with known nutrient removal capacities such that it is posited that appropriate bead

91 concentrations can be tailored to the required loading rates. To date, the work completed

92 to optimise biomass density through bead concentration have pre-selected a bead.mL-1

93 concentration and evaluated remediation performance regardless of the chosen species

94 nutrient uptake characteristics. For example, Abdel Hameed (2007) evaluated the

95 remediation performance of Chlorella vulgaris using bead concentrations of 10.66, 16,

96 32 and 64 bead-mL-1 (1:3 to 2:1 bead:wastewater v/v) at 106 cells-bead-1.

97 Concentrations of 10.66 and 16 beads-mL"1 both achieved 100% NH4+ and 95% PO43"

98 removal efficiency, suggesting a concentration of 10 beads-mL"1 and associated biomass

99 concentration to be suitable for optimal treatment under the conditions tested, with the

100 possibility of a lower concentration performing similarly.

101 With the onset of the water framework directive (WFD) across Europe, the discharge

102 consent for wastewater P will reduce from the current 1 - 2 mgP-L"1 outlined within the

103 Urban Wastewater Treatment Directive (UWWTD) to <0.5 mgP-L"1, with some sites

104 expected to be as low as 0.1 mgP-L"1 (Jarvie, 2006; Mainstone and Parr, 2002).

105 Microalgae can be considered an alternative solution to meet these new stringent

106 targets, providing the solution represents a practical alternative in terms of treatment

107 time (HRT) and footprint, which can be achieved through immobilisation.

108 As such, the objectives of this study are to investigate how the internal composition of

109 microalgae, through their ability to adapt to the external nutrient concentrations, relate

110 to their nutrient uptake. Remediation performance of two of the characterised species,

111 C.vulgaris and S.obliquus, are further analysed within real wastewater effluent. The

112 findings are then translated into the impact on the design of an immobilised reactor for

113 the improved remediation of wastewater nutrients; through the selection of a species for

114 immobilisation and manipulation of biomass concentration through bead concentration

115 to enable a suitable HRT for the integration of a microalgal reactor into a wastewater

116 flow sheet for nutrient polishing.

117 2 Materials and methods

118 2.1 Microalgal biomass culture and immobilisation

119 The freshwater species Chlorella vulgaris (211/11B), Chlorella sorokiniana (211/8K),

120 Microcystis aeruginosa (1450/3), Scenesdesmus obliquus (276/3A) and Stigeoclonium

121 sp (477/24) were obtained from the Culture Collection for Algae and Protozoa (CCAP)

122 (Oban, UK). Mono cultures were cultivated in 100 L reactor containing 50 L of medium

123 as recommended by CCAP for optimal growth with an N:P molar ratio of

124 approximately 2:1 for M.aeruginosa and 6:1 for the remaining species. Cultures were

125 illuminated under a 24 hour light regime with a light intensity of approximately 100 -

-2 -1 -1

126 150 |imol-m- -s- . Constant mixing was through a circulation pump (900 L-h-1) (Hydor

127 Koralia Nano 900), with no external supply of CO2 provided and the temperature

128 maintained at 18oC. Microalgal biomass was harvested prior to the onset of stationary

129 growth phase determined through previous growth experiments to characterise growth

130 under the stated operational conditions and monitored through; cell counts for single

131 celled species using a haemocytometer and light microscope (Olympus, BH Series), or

132 dry weight following standard methods for total suspended solids (TSS) (APHA, 2005)

133 for filamentous species. Knowledge of the chosen species' growth profile enabled

134 biomass harvesting at the latter stages of exponential growth.

135 Microalgal immobilisation and encapsulation within calcium-alginate beads were

136 completed following the method of Ruiz-Marin et al., (2010), with the adsorption

137 capacity of the calcium-alginate resin determined through the method of Gotoh et al.

138 (2004) (see supplementary information, Appendix A).

139 2.1.1 Freshwater species characterisation - nutrient remediation and internal N

140 and P composition

141 Nutrient removal batch trials were completed in 100 L reactors using 50 L modified

142 BG11 medium under the same operational conditions as those for cultivation and

143 supplemented with NH4CI and KH9PO4 for an N:P molar concentration of 2:1, selected

144 as a N:P < 10:1 is associated with enhanced biomass productivity (Choi and Lee, 2015).

145 Biomass was seeded at an approximate concentration of 40 mg(DW)- L~'. Reactors were

146 mixed on a daily basis and pH monitored and corrected to pH 7 using 1 M NaOH and

147 HC1 to prevent alkalisation of the medium and nutrient remediation through the indirect

148 processes of precipitation and volatilisation. Batch trials were run over a period of 10

149 days, with analysis on day 0, 3 5, 7 and 10.

150 Nutrient remediation was determined by measuring the residual concentration of NH4-N

151 and total phosphorus (TP) within the medium in triplicate using Spectroquant test

152 kitsl. 14752.0001 (NH4-N) and 1.14543.0001 (PO4-P) (Merck Millipore), following the

153 manufacturer's instructions and read via a Spectroquant Nova 60 spectrophotometer.

154 Biomass growth was monitored as previously described and specific growth rate

155 calculated using Equation 1, where 11 = specific growth rate (d"1), xl and x2 the

156 biomass concentration in cells-mL"1 or mg(DW)-L_1 at time tl (d) and t2 (d).

_ Equation 1

^ tl - tl

157 Characterisation of the internal total nitrogen (TN) and C content of the microalgal

158 biomass was analysed following freeze drying (ModulyoD Freeze Dryer, USA) and

159 analysis using a TCN Vario III Elemental Analyser (Isoprime, DE) according to

160 standard method ISO 10694:1995. Phosphorus content of the digested biomass sample

was measured by UV/VIS spectrophotometry, following calibration with P standards of 0 - 7 mg-L-1 with a ± 0.005 accuracy, according to standard methods (USEPA, 1995).

2.1.2 Wastewater nutrient remediation trials for S.obliquus and C.vulgaris

Secondary wastewater effluent was delivered weekly from a wastewater treatment works located in the Midlands, UK and stored at 4oC until use. The 32,000 population equivalence (PE) wastewater treatment plant (WWTP) comprises of an oxidation ditch operated for biological nutrient removal in addition to iron salt precipitation prior to the secondary clarifier. Effluent was selected from this site as the WWTP is located in a catchment designated as a Site of Special Scientific Interest (SSSI) which will be required to meet the stricter P consents prescribed within the WFD. As such, the WWTP has been selected as a trial site to evaluate the performance of multiple alternative technologies for the purpose of P polishing to meet the upcoming change in consent.

The average characteristics of the effluent collected were 0.3 mg-L-1 PO4-P and 0.1 mg-L-1 NH4-N. Effluent was supplemented with NH4Cl and KH2PO4 to compensate for the current dosing strategy (which ensures appropriate nutrient discharge to the SSSI) and maintain a set NH4-N concentration of 5 mg-L-1 and a range of PO4-P concentrations between 0.5 and 10 mg-L-1. These concentrations represent a possible range of secondary effluent characteristics that could be encountered by a tertiary microalgal system without advanced upstream treatment (A. Brookes & P. Vale, 2011, pers. Comms., 20 October), with an N:P molar ratio between approximately 22.1 to 1.1 suitable for microalgal activity with ratios < 22 indicating sufficient phosphorus (Hecky et al., 1993). The wastewater also contained a non-supplemented and variable NO3-N concentration of a maximum of 2 mgN-L-1, lower than the concentration of NH4-N,

which was not analysed through the preference of microalgae to assimilate NH4-N over NO3-N (Lau et al., 1995); and results from previous trials which found no accumulation of NO3-N associated with the nitrification of NH4 but rather a decrease of NO3-N in parallel to NH4-N remediation.

Conical flasks with 250 mL modified effluent were seeded with 104 and 105 cells.mL'1 of S.obliquus and C.vulgaris respectively to ensure a sufficient initial biomass concentration for growth. Batch trials were run over a period of 7 days under the same operational conditions as those for cultivation with sample analysis on days 3, 5 and 7. Analysis included pH, NH4-N, P04-P and cell concentration. Residual nutrient concentrations and cell concentration were analysed as previously described.

The NH4-N and P04-P cell uptake rate for S.obliquus and C.vulgar is was estimated through the analysis of the residual concentration according to Equation 2, where V is the cell uptake rate (mg-cell^-d"1), N the cell concentration (cells-mL"1) at time t (d"1) and Ci and Cf the initial and final residual concentrations (mg-L"1) respectively.

C- — C

V = —-- Equation 2

2.2 Calculations for optimal biomass concentration for wastewater treatment by S.obliquus and C.vulgaris - suspended and immobilised cultures

Following determination of the cell uptake rate for S.obliquus and C.vulgaris, calculations were completed to determine initial biomass concentrations required for 'optimal remediation' of phosphorus (residual <0.1 mgP-L"1) in line with changes to P discharge consent within the forthcoming WFD. Results were translated to an immobilised culture assuming a fixed cell stocking of 106 cells-bead"1 as recommended by Abdel Hameed, (2007). Calculations estimated the biomass concentration required for the remediation of phosphate at concentrations of 1, 5, and 10 mgP-L"1 operating at a

208 range of HRTs varying from 3 h to 10 days. The range of HRTs chosen complement the

209 work by Whitton et al., (2014) which characterised remediation of an immobilised

210 system at HRTs of up to 20 h and compared to the typical retention time of a HRAP (4211 10 days).

212 3 Results

213 3.1 Freshwater species characterisation - nutrient remediation and internal N and

214 P composition

215 3.1.1 N:P composition changes with change in external N:P

216 Following cultivation, the internal molar N:P composition of the tested algal species

217 ranged from 7.8 to 20.3 (Table 1) despite the similar N:P molar concentration (6:1) of

218 the growth medium (excluding the medium forM.aeruginosa, (2:1)), demonstrating the

219 potential significance of algal selection.

220 Transferring the algal species from the growth medium to a more N limited test medium

221 (N:P 2:1) for the nutrient remediation trials, reduced the difference in the biomass

222 nutrient molar ratio to between 11.3 and 16.3 (average N:P 13.7) (Table 1). The molar

223 N:P composition of M.aeruginosa was found to change the least, with a decrease of 1.5

224 from a molar N:P of 16.6 to 15:1 due to the similarities in the N:P characteristics of the

225 cyanobacteria growth medium and test medium (both N:P 2:1) (Table 1). The difference

226 observed with the change in N:P medium is congruent with the microalgae adapting the

227 nutrient concentration within its biomass to the new environment (Beuckels et al.,

228 2015), with the nitrogen content varying between 6.5 and 9.0% by weight (0.065 and

229 0.09 mgN-mg biomass-1) with comparable variations in biochemical N composition of

230 Chlorella sp of 3.6 to 10% previously demonstrated (Ákerstróm et al., 2014).

All microalgae were found to adjust their internal N:P content within the first 3 days of the trial and remained at the their new N:P composition for the remainder of the trial (Table 1). For example, Stigeoclonium sp. demonstrated a change from an initial N:P of 18.9 (± 0.2) to an average of 13.5 (±0.6) for days 3 - 10 of the experiment, concluding with an N:P of 13.9 (±1.3) by day 10 with mass balances estimating 94.0 and 93.9% of N and P removal through microalgal growth. The final internal N:P values exhibited by all the microalgae analysed generally decreased (excluding S.obliquus) through the adjustment of the microalgae.

Whereas species with higher initial N:P composition (>16) reduced to between 12.0 and 16.3, S.obliquus with the lowest initial N:P of 7.9 increased to 11.6 (Table 1) suggesting S.obliquus can tolerate a greater N concentration than supplied by the growth medium (N limited) for incorporation and conversion into new biomass, supporting previous work of tailoring growth mediums to species' biochemical composition for growth optimisation (Mandalam and Palsson, 1998).

An average phosphorus content of 0.01 mgP-mg biomass-1 (ranging 0.8 - 2.1% by weight) (Table 1) was found for all species, characterising growth in non-limiting P conditions (Hessen et al., 2002). The narrow variation in P within the biomass of the different algae is consistent with previous work that has shown that P level vary less when the N content within the biomass is relatively low (Beuckels et al., 2015). The N content therefore dictated the overall N:P composition of the species (varying from 0.07 - 0.09 mgN-mg biomass-1), and remained within the N:P ranges of 8.5 - 42 and 4.1 - 32 previously reported for C.vulgaris and S.obliquus (Rhee, 1974; Oh-Hama and Miyachi, 1988, Beuckels et al., 2015) when grown within varying N:P concentrations of differing retention times and growth conditions.

255 3.1.2 Nutrient remediation performance and internal N:P composition

256 Analysis of the remediation through a reduction in liquid phase nutrient concentration

257 revealed > 99% removal of NH4-N during the experimental period with species with an

258 initial N:P >18 compared to between 24.7 and 60.8% for species with an N:P <18. The

259 increase in uptake by species characterised with a greater N composition is associated

260 with the greater nitrogen content required per cell supporting previous work

261 demonstrating a relationship between cell growth rate (Droop, 1974) and N remediation

262 characteristics (Choi and Lee, 2015) in relation to internal concentration of the algal

263 cell. The remediation efficiency of phosphate was lower than ammonium, at between

264 12.5 and 19.6% and unlike N, species characterised as more P limited were found to

265 remediate at the higher end of this range supporting the inverse relationship

266 demonstrated by Ruiz-Martinez et al., (2015) for P uptake and biomass composition.

267 Mass balances considering biomass growth and N:P composition estimates between

268 82.6 - 94.5% and 83.7 - 98.1% of N and P remediation is attributed to incorporation

269 into new biomass for all species analysed. Remediation performance attributed to

270 abiotic processes such as precipitation and volatilisation were considered negligible

271 through the attainment of average pH values (prior to pH adjustment) during the trials

272 of 6.9 (±0.09) for C. vulgaris, 6.2 (±0.32) for C. sorokiniana, 7.7 (±0.30) for

273 M.aeruginosa, 7.7 (±0.38) for S.obliquus and 7.7 (±0.15) for Stiegoclonium sp

274 respectively. As such, when considering the pKa value for ammonium at 20oC an

275 estimated 2 - 4% (equivalent to 0.2 - 0.4 mg-L-1)of removal can be contributed to

276 volatised free ammonia at the peak pH value of 7.7 (±0.38); and minimal P precipitation

277 through pH values lower than the required 8 - 9 required for precipitation with metal

278 ions such as calcium (Ca) (Montastruc et al., 2003).

Species with a lower N composition were found to have an enhanced specific growth rate (Figure 1) through the reduced N requirements for growth. This is illustrated by a 1.5x increase in the final biomass concentration of 2.0 g(DW)-L"1 compared to 1.3 g(DW)-L"1 at the end of the trials by species with the lowest (0.06 mgN-mg(DW)"1) compared to the highest (0.09 mgN-mg.(DW)"1) N composition respectively. However, the increase in biomass concentration of those species with a lower N:P could not outcompete the remediation performance of those species with a greater N:P composition at a lower biomass volume. Overall, species with a greater N:P composition (high N and low P), where found to remediate ammonium and phosphate more efficiently than those species with a lower N:P under the specified conditions even when considering total biomass concentration.

3.2 Wastewater nutrient remediation by S.obliquus and C.vulgaris

Two commonly used singled celled species initially characterised by a low and high

N:P composition during cultivation; S.obliquus and C.vulgaris, were selected for trials with secondary effluent from a municipal wastewater treatment works. Ammonium concentration was fixed at 5 mgN-L"1 and the phosphorus concentration varied between 0.5 and 10 mgP-L"1, varying the medium N:P molar ratio between approximately 22.1 to 1.1 and encompassing a range of concentrations possible within secondary wastewater effluent (A. Brookes & P. Vale, 2011, pers. Comms., 20 October) . Increasing the P concentration and hence reducing the N:P ratio resulted in an increase in cell uptake for both species in terms of P (Figure 2a) and a decrease in cell uptake for N (Figure 2b) reflecting nutrient availability within the supplemented effluent.

The phosphorus removal rate increased from 0.1 to 1.6 pgP-cell"1-d"1 and 0.2 to 2.7 pgP-cell"1-d"1 for S.obliquus and C.vulgaris respectively as the initial concentration

303 increased from 0.5 - 10 mgP-L-1. A minimum P remediation efficiency through the

304 incorporation into new biomass of 47 - 82% and 18 - 77% for S.obliquus and

305 C.vulgaris is estimated, prior to an increase in the effluent pH to values indicative of

306 remediation through abiotic processes, with the initial biomass incorporation rates

307 decreasing with increasing P concentration.

308 The removal of phosphate is associated with N removal through their respective roles in

309 cellular metabolism (Loladze and Elser, 2011). In microalgae, N is mainly integrated

310 into proteins that in turn links to the production of ribosomes and ribosomal RNA.

311 Phosphate uptake is predominately associated with storage into the ribosomal RNA

312 such that the observed function between P concentration and uptake rate requires

313 sufficient N to ensure no restriction of protein synthesis. Previous work has shown that

314 in low N environments, the uptake of P into the biomass remains low irrespective of the

315 P concentration in the solution (Beuckels et al., 2015). In such cases the uptake rate also

316 relates to the ability of microalgae to store available phosphate in time of surplus,

317 through a luxurious uptake pathway where polyphosphate accumulates within the cells

318 (Wang et al., 2010).

319 The pattern of remediation for NH4-N were similar for both species and decreased as the

320 concentration approached 2 mgP-L-1 prior to stabilising. Remediation rates of 0.7 - 1.5

321 pgN-cell-1-d-1 and 1.1 - 2.3 pgN-cell-1-d-1 for S.obliquus and C.vulgaris were

322 demonstrated (Figure 2b), with remediation through incorporation into new biomass

323 estimated at a minimum of 54 - 99% and 38 - 93% for S.obliquus and C.vulgaris

324 respectively, prior to an increase in the effluent pH and the contribution of volatilisation

325 to total remediation.

326 Remediation performance was furthermore reflected in the cell uptake rate for both

327 species remediating the equivalent of 18.5 - 82.1 and 36.3 - 95.6 fmolN-cell"1-h"1 for

328 S.obliquus and C.vulgaris respectively; greater than that reported for ammonia-

329 oxidising bacteria (AOB) in wastewater of 0.03 - 53 fmolNcell-1-h-1 (Lydmark, 2006).

330 When considering the difference in mass of AOB and microalgae, AOB concentrations

331 of 1010 cells-gVSS-1 (Hallin et al., 2005) in comparison to approximately 5 x109 cells-g-

332 1.VSS for Chlorella (considering 7.7x109 cells-gCOD-1 and 1.43 gCOD-gVSS-1 (Ras et

333 al., 2011)) are reported. The associated mass uptake rates based on the higher ranges of

334 5.3x10-9 and 1.8x10-8 fmolNgVSS-1-h-1 for AOB and Chlorella respectively further

335 demonstrate the effectiveness of microalgal cells for ammonium remediation.

336 Comparison of the two microalgal species in terms of cell uptake revealed similar levels

337 of 0.4 (+ 0.07) pgP-cell-1-d-1 in low phosphate wastewater (< 2.5 mg-L-1) (Figure 2a). In

338 contrast, at higher initial phosphate concentrations of 5 mgP-L-1, uptakes rates of 0.5

339 (±0.2) pgP-cell-1-d-1 and 0.9 (± 0.3) pgP-cell-1-d-1 were observed for S.obliquus and

340 C.vulgaris respectively (Figure 2a). Nitrogen uptake was also slightly greater for

341 C.vulgaris at higher P concentration consistent with C.vulgaris' higher N:P cell content

342 but in contrast to previous work that showed that N concentration in biomass was

343 independent of P supply (Beuckels et al., 2015) indicating other mechanisms. Notable

344 differences were observed between species in terms of cell growth and associated

345 alkalisation of the surrounding medium. The growth rate of C.vulgaris was lower and

346 more consistent across all concentrations with a range of specific growth rates between

347 0.16 and 0.29 d-1 (in comparison to 0.51 and 0.71 d-1 for S.obliquus) suggesting the

348 greater P and N uptake observed at higher P concentration was not due to cell growth.

349 Luxury phosphate uptake has been demonstrated to take effect for Scenedesmus beyond

350 a critical growth concentration of 1.5 mgP-L-1 (Azad and Borchardt, 1970). At

351 concentrations beyond this level uptake through luxury consumption has been observed,

352 with no impact on growth (Azad and Borchardt, 1970). Similar observations were found

353 within this study, and supports the improved remediation performance at the higher

354 concentrations despite the narrow range of growth rates observed for the different PO4-P

355 concentrations. The lower growth rate observed for C.vulgaris resulted in a reduced

356 degree of alkalinisation as evidenced by an average final pH of 10.9 for S.obliquus in

357 comparison to 9.7 for C.vulgaris.

358 4 Discussion: implications for an immobilised microalgal reactor for tertiary

359 wastewater nutrient remediation

360 The aim of the study was to examine the impact of variation in nutrient content of algal

361 biomass on the associated nutrient uptake rates and understand the importance of algal

362 species selection when operating a tertiary treatment system. Overall, the nutrient

363 concentration in algal biomass is not fixed and so does not map to predictions based

364 around the Redfield ratio (Hecky et al., 1993). Furthermore microalgae display

365 flexibility in the nutrient concentrations in the biomass enabling adaptation to the local

366 environment with nutrient uptake limited by the species' specific cellular quota for

367 structural components and storage for future growth (Droop, 1968). Accordingly, design

368 of microalgae reactors for wastewater treatment need to consider species selection and

369 nutrient concentrations in the biomass and ability to adapt to external concentrations as

370 it impacts on the maximum treatable loading rate and associated footprint.

371 Experimental results characterising nutrient uptake for the species S.obliquus and

372 C.vulgaris were used to calculate cell concentrations required for 'optimal remediation'

373 (<0.1 mgP-L-1) from initial concentrations similar to those found within tertiary

effluent. Required concentrations were predominately calculated through PO4 -remediation as phosphate is targeted for further reductions in consent through the WFD. Required cell concentrations for both S.obliquus and C.vulgaris were found to increase to > 3x106 cells-mL-1 for HRTs less than 24 hours, with maximum cell concentrations

of 8.5x10' and 4.7x10' cells-mL-1 at a treatment time of 3 h for S.obliquus and C.vulgaris respectively for the higher P concentrations (Figure 3).

The required cell concentration for C.vulgaris was found to be less variable than S.obliquus with similar biomass concentrations necessary for the remediation of 1, 5 and 10 mg-L-1. C.vulgaris is therefore considered a better option for the treatment of wastewater with a varying influent concentration. However, for both species, greater biomass concentrations were necessary for treatment at 5 mgP-L-1 than 10 mgP-L-1due to a cell uptake rate three times greater when treating 10 mgP-L-1 highlighting the ability of microalgae to adjust their performance to suit a changing environment, with enhanced remediation through luxury uptake within nutrient rich environments (Eixler at el., 2006) thus the reduced biomass concentration at the higher P concentration.

Open pond systems (i.e. HRAPs) are reported to maintain an approximate biomass concentration of 0.4 g(DW)-L-1, equivalent to a cell concentration of approximately 106 cells.mL-1 for S.obliquus (based on laboratory growth data). This concentration is sustained through the variation in light, temperature and biotic factors including zooplankton grazers and pathogens (Park et al., 2011). As such, HRTs >4 days are necessary to achieve an optimal level of treatment when using a HRAP through the biomass concentration achievable. To illustrate the consequence of this, a wastewater treatment works with population equivalence (PE) of 2,000 treating the standard 0.18

3 -1 -1 2

m -pe- • d- of effluent would require a HRAP with a surface footprint of 7,200 m at a

depth of 0.2 m and minimum HRT of 4 days. This footprint is considerably larger than that of conventional tertiary treatments such as rotating biological contactor (RBC) unit or trickling filter with land footprints of 40 - 50 m (Butterworth et al., 2013). As such, a HRAP would be an unlikely solution for retrofitting to a site of 2,000 PE. To overcome limitations around treatable load and footprint an intensification of the algal biomass is required. Immobilisation enables biomass concentrations beyond 107

cells-mL-1 through either an increase in the cells per bead or the number of beads per

unit volume, with example levels of up to 10 cells-mL" (Abdel Hameed, 2007) and biomass concentration of up to 3.3 g(DW)-L-1 (Chevalier and De la Noue, 1985) reported, equivalent to the typical biomass concentration found within the activated sludge (AS) process (Metcalf & Eddy et al., 2003)

To illustrate the impact of immobilisation on intensification of microalgae based wastewater treatment, the cell concentration required for each influent concentration (Figure 3) was used to determine the required bead concentration (based on an initial internal bead concentration of 106 cells-bead"1) (Figure 4). Calculations for PO4-P removal considering 106 cells-bead-1 found bead concentrations of 3.2 to 85.1 beads-mL-1 (1:12.5 to 2.1:1 bead:wastewater v/v) for S.obliquus and 4.4 to 47.1 beads-mL-1 (1:9 to 1.2:1) for C.vulgaris (Figure 4) for remediation of influent concentrations of 1, 5 and 10 mgP-L-1.

The required bead-mL-1 concentration was found to increase with the reduction in HRT due to the increased loading rate and required increase in biomass concentration. Overall, a lower bead-mL-1 concentration was found for S.obliquus at the lower PO4-P influent concentrations (1 mgP-L-1) and C.vulgaris for the higher concentrations (>5mgP-L-1) (Figure 4) due to the increased cell uptake rate demonstrated by C.vulgaris

422 during batch trial characterisation (0.13 - 1.6 and 0.12 - 2.7 pgP-cell-1-d-1 for 0.5 - 10

423 mgP-L-1 for S.obliquus and C.vulgaris respectively) (Figure 2a) and the corresponding

424 performance associated to the characterised internal N:P composition.

425 These calculated bead concentrations can be compared to observed experimental

426 performance of an immobilised algae reactor of S.obliquus treating a PO4-P

427 concentration of 0.7 mg-L-1 at fixed bead concentration of 10 beads-mL-1 and variable

428 HRTs of 3, 6, 12 and 20 hours (Whitton et al., 2014). Similar residual concentrations,

429 following the initial start-up period of 0.10, 0.17 and 0.11 mgP-L-1 were observed for 6,

430 12 and 20 h respectively confirming a suitable bead/biomass concentration. However, a

431 reduction in residual performance at a 3 h HRT of 0.43 mgP-L-1 was observed

432 suggesting the biomass concentration to be inadequate. Based on the calculated bead

433 concentrations presented (Figure 4a), a concentration of approximately 13 beads-mL-1

434 (equivalent of an additional 106 cell-L-1 and approx. 0.2 g(DW)-L-1) would have

435 provided the additional biomass necessary to remediate within the shortened retention

436 time and as such, the predicted biomass concentrations presented in Figure 4 can be

437 used to inform cell concentration through bead volume for S.obliquus and C.vulgaris.

438 Extending this to higher loading rates needs to consider other practical aspects which

439 limit the applicable bead concentration to 1:1 v/v in order to minimise practical issues.

440 These include sinking and crushing of beads under their own weight (Abdel Hameed,

441 2007) and self-shading restricting light penetration (Lau et al., 1995) which can

442 contribute to a significant reduction in NH4+ remediation performance (Abdel Hameed,

443 2007) as well as improving irradiation efficiency.

When applying the maximum bead concentration (1:1 v/v) to the range of influent concentrations, a treatment period of 1.5 - 2.5 h for an effluent with a concentration < 1mgP-L-1 is achievable by immobilised S.obliquus and C.vulgaris (Figure 5). Treatment periods >3 h are then necessary for the remediation of effluents >2.5 mgP-L-1 by both S.obliquus and C.vulgaris with required HRTs of 6.3 h and 3.5 h for S.obliquus and C.vulgaris at 5mgP-L-1 respectively (Figure 5). In situations where immobilised algae are used as a tertiary treatment the solution is unlikely to encounter influent concentrations greater than 5 mgP-L-1. As such the required HRT is less than 3 hours indicating the potential for effective use of microalgae without the need for large footprint technology.

Immobilisation also introduces an additional component in the form of the calcium-alginate beads that contain the microalgae and offers an additional uptake pathway. Adsorption trials with blank Ca-alginate beads found PO4-P uptake by the resin material to be negligible across the tested PO4-P concentrations (see Appendices, Figure B. 1), confirming previous trials with blank alginate beads in sterile conditions (Cruz et al., 2013). However, within non-sterile wastewater Cruz et al. (2013) demonstrated a capacity of > 15 ^gP-g-1 over a 48 hour period with removal contributed to the formation of a concentrated biofilm layer supported by the bead's surface area and not directly through the adsorption capacity of the resin material. In contrast, uptake of NH4-N resulted in removal efficiencies of 9.1, 20.6, 25.4 and 23.4% for NH4-N at starting concentrations of 0.5, 2.5, 5 and 10 mgN-L-1 respectively, with an adsorption capacity of 6 |igN-g-1 determined through fitting the data to a Freundlich isotherm model (see Appendices, Figure B.2) and providing an additional pathway for nutrient removal when using immobilised systems. As such, this study provides a conservative

estimate on the ability of immobilised microalgae to remediate wastewater nutrient through species selection and biomass concentration as additional mechanisms, including the Ca-alginate resin and indirect methods of volatilisation and precipitation, would further enhance the overall remediation performance.

5 Conclusions

• A relationship between internal N:P composition and nutrient remediation is evident and can be considered when selecting a species for remediation. Species with a high N and low P internal composition remediate ammonium and phosphate more efficiently.

• Required biomass concentrations varied with wastewater characteristics and nutrient uptake abilities. When translated into immobilised beads, concentrations as low as 3.2 beads-mL-1 is possible for S.obliquus at HRT of 20 h.

• A HRT <3 h is impractical for an immobilised microalgal solution for concentration > 5 mgP-L-1, due to the volume of beads required to achieve maximum remediation.

Acknowledgements

The authors gratefully acknowledge financial support from the Engineering and Physical Sciences Research Council (EPSRC) through their funding of the STREAM Industrial Doctorate Centre, and from the project sponsors Anglian Water, Severn Trent Water and Scottish Water. Also a special thank you to Maria Biskupska and Jan Bingham for assisting with the internal biomass composition analysis.

References

Abdel Hameed, M. S. (2007), "Effect of algal density in bead, bead size and bead concentration on wastewater nutrient removal", African Journal of Biotechnology, vol. 6, no. 10, pp. 1185-1191.

492 Âkerstrom, A. M., Mortensen, L. M., Rusten, B. and Gislerad, H. R. (2014), "Biomass

493 production and nutrient removal by Chlorella sp. as affected by sludge liquor

494 concentration", Journal of Environmental Management, vol. 144, pp. 118-124.

495 APHA (2005) Standard methods for the examination of water and wastewater.

496 Washington, D.C. : American Public Health Association.

497 Azad, H. S. and Borchardt, J. A. (1970), "Variations in phosphorus uptake by algae",

498 Environmental Science and Technology, vol. 4, no. 9, pp. 737-743.

499 Beuckels, A., Smolders, E. and Muylaert, K. (2015), "Nitrogen availability influences

500 phosphorus removal in microalgae-based wastewater treatment", Water Research, vol.

501 77, no. 0, pp. 98-106.

502 Butterworth, E., Dotro, G., Jones, M., Richards, A., Onunkwo, P., Narroway, Y. and

503 Jefferson, B. (2013), "Effect of artificial aeration on tertiary nitrification in a full-scale

504 subsurface horizontal flow constructed wetland", Ecological Engineering, vol. 54, no. 0,

505 pp. 236-244.

506 Chevalier, P. and De la Noue, J. (1985), "Efficiency of immobilized hyperconcentrated

507 algae for ammonium and orthophosphate removal from wastewaters", Biotechnology

508 Letters, vol. 7, no. 6, pp. 395-400.

509 Choi, H. J. and Lee, S. M. (2015), "Effect of the N/P ratio on biomass productivity and

510 nutrient removal from municipal wastewater", Bioprocess andBiosystems Engineering,

511 vol. 38, no. 4, pp. 761-766.

512 Craggs, R., Lundquist, T., Benemann, J., (2013). Wastewater treatment and algal

513 biofuel production. In: Borowitzka, M.A., Moheimani, N.R. (Eds.), Algae for Biofuels

514 and Energy. Springer, Dordrecht, pp. 153-163.

515 Cruz, I., Bashan, Y., Hernàndez-Carmona, G. and de-Bashan, L. E. (2013), "Biological

516 deterioration of alginate beads containing immobilized microalgae and bacteria during

517 tertiary wastewater treatment", Applied Microbiology and Biotechnology, vol. 97, no.

518 22, pp. 9847-9858.

519 Droop, M. (1968), "Vitamin B12 and marine ecology. IV. The kinetics of uptake,

520 growth and inhibition in Monochrysis lutheri", Journal of the Marine Biological

521 Association of the United Kingdom, vol. 48, no. 3, pp. 689-733.

522 Droop, M. (1974), "The nutrient status of algal cells in continuous culture", Journal of

523 the Marine Biological Association of the United Kingdom, vol. 54, no. 4, pp. 825-855.

524 Eixler, S., Karsten, U. and Selig, U. (2006), "Phosphorus storage in Chlorella vulgaris

525 (Trebouxiophyceae, Chlorophyta) cells and its dependence on phosphate supply",

526 Phycologia, vol. 45, no. 1, pp. 53-60.

527 Gotoh, T., Matsushima, K. and Kikuchi, K. (2004), "Preparation of alginate-chitosan

528 hybrid gel beads and adsorption of divalent metal ions", Chemosphere, vol. 55, no. 1,

529 pp. 135-140.

530 Hallin, S., Lydmark, P., Kokalj, S., Hermansson, M., Sörensson, F., Jarvis, Ä. and

531 Lindgren, P. (2005), "Community survey of ammonia-oxidizing bacteria in full-scale

532 activated sludge processes with different solids retention time", Journal of Applied

533 Microbiology, vol. 99, no. 3, pp. 629-640.

534 Hecky, R. E., Campbell, P. and Hendzel, L. L. (1993), "The stoichiometry of carbon,

535 nitrogen, and phosphorus in particulate matter of lakes and oceans", Limnology and

536 Oceanography, vol. 38, no. 4, pp. 709-724.

537 Hessen, D. O., F^rovig, P. J. and Andersen, T. (2002), "Light, nutrients, and P: C ratios

538 in algae: grazer performance related to food quality and quantity", Ecology, vol. 83, no.

539 7, pp. 1886-1898.

540 Jarvie, H.P. et al. (2006) 'Sewage-effluent phosphorus: a greater risk to river

541 eutrophication than agricultural phosphorus?', The Science of the total environment,

542 360(1-3), pp. 246-53.

543 Lau, P. S., Tam, N. F. Y. and Wong, Y. S. (1995), "Effect of algal density on nutrient

544 removal from primary settled wastewater", Environmental Pollution, vol. 89, no. 1, pp.

545 59-66.

546 Loladze, I. and Elser, J. J. (2011), "The origins of the Redfield nitrogen-to-phosphorus

547 ratio are in a homoeostatic protein-to-rRNA ratio", Ecology Letters, vol. 14, no. 3, pp.

548 244-250.

549 Lundquist, T. J., Woertz, I. C., Quinn, N. and Benemann, J. R. (2010), "A realistic

550 technology and engineering assessment of algae biofuel production", Energy

551 Biosciences Institute, , pp. 1.

552 Lydmark, P. (2006), "Population dynamics of nitrifying bacteria in biological

553 wastewater treatment" (PhD thesis), Department of Cell and Molecular Biology,

554 Microbiology, Göteborg University.

555 Mainstone, C.P. and Parr, W. (2002) 'Phosphorus in rivers — ecology and

556 management', Science of The Total Environment, 282-283, pp. 25-47

557 Mandalam, R. K. and Palsson, B. 0. (1998), "Elemental balancing of biomass and

558 medium composition enhances growth capacity in high-density Chlorella vulgaris

559 cultures", Biotechnology and Bioengineering., vol. 59, pp. 605-611.

560 Metcalf & Eddy, I. et al. (2003) Wastewater engineering: treatment and reuse. Boston:

561 McGraw-Hill.

562 Montastruc, L. et al. (2003) "A thermochemical approach for calcium phosphate

563 precipitation modeling in a pellet reactor", Chemical Engineering Journal, vol 94, no 1,

564 pp. 41-50.

565 Oh-Hama, T. and Miyachi, S. (1988), "Chlorella", in Borowitzka, M. A. and

566 Borowitzka, L. J. (eds.) Microalgal Biotechnology, Cambridge University Press,

567 Cambridge, pp. 3-26.

568 Ometto, F., Quiroga, G., Psenicka, P., Whitton, R., Jefferson, B. and Villa, R. (2014),

569 "Impacts of microalgae pre-treatments for improved anaerobic digestion: Thermal

570 treatment, thermal hydrolysis, ultrasound and enzymatic hydrolysis", Water Research,

571 vol. 65, no. 0, pp. 350-361.

572 Park, J. B. K., Craggs, R. J. and Shilton, A. N. (2011), "Wastewater treatment high rate

573 algal ponds for biofuel production", Bioresource Technology, vol. 102, no. 1, pp. 35-42.

574 Portielje, R. and Lijklema, L. (1994), "Kinetics of luxury uptake of phosphate by algae-

575 dominated benthic communities", Hydrobiologia, vol. 275, no. 1, pp. 349-358.

576 Ras, M., Lardon, L., Bruno, S., Bernet, N. and Steyer, J. (2011), "Experimental study on

577 a coupled process of production and anaerobic digestion of Chlorella vulgaris",

578 Bioresource Technology, vol. 102, no. 1, pp. 200-206.

579 Redfield, A. C. (1934), "On the proportions of organic derivatives in sea water and their

580 relation to the composition of plankton", James JohnstonesMemorial, , pp. 176-192.

581 Reynolds, C., (2006). In: Ecology of Phytoplankton. Cambridge University Press,

582 Cambridge, p. 550.

583 Rhee, G. (1974), "Phosphate uptake under nitrate limitation by Scenedesmus sp. and its

584 ecological implications", Journal of Phycology, vol. 10, no. 4, pp. 470-475.

585 Ruiz-Marin, A., Mendoza-Espinosa, L. G. and Stephenson, T. (2010), "Growth and

586 nutrient removal in free and immobilized green algae in batch and semi-continuous

587 cultures treating real wastewater", Bioresource Technology, vol. 101, no. 1, pp. 58-64.

588 Ruiz-Martinez, A., Serralta, J., Romero, I., Seco, A. and Ferrer, J. (2015), "Effect of

589 intracellular P content on phosphate removal in Scenedesmus sp. Experimental study

590 and kinetic expression", Bioresource Technology, vol. 175, pp. 325-332.

591 USEPA, E. (1995), "Method 3051: Microwave assisted acid digestion of sediments,

592 sludges, soils, and oils", Test Methods for Evaluating Solid Waste.

593 Wang, L., Min, M., Li, Y., Chen, P., Chen, Y., Liu, Y., Wang, Y. and Ruan, R. (2010),

594 "Cultivation of green algae Chlorella sp. in different wastewaters from municipal

595 wastewater treatment plant", Applied Biochemistry and Biotechnology, vol. 162, no. 4,

596 pp. 1174-1186.

597 Whitton, R., Santinelli, M., Chazaux, M., Pidou, M., Ometto, F., Villa, R. and Jefferson,

598 B. (2014), "Immobilised microalgae for phosphorus remediation", IWA World Water

599 Congress and Exhibition, 21st - 26th September 2014, Lisbon, Portugal, .

600 Xin, L., Hong-ying, H., Ke, G. and Ying-xue, S. (2010), "Effects of different nitrogen

601 and phosphorus concentrations on the growth, nutrient uptake, and lipid accumulation

602 of a freshwater microalga Scenedesmus sp.", Bioresource Technology, vol. 101, no. 14,

603 pp. 5494-5500.

Table 1 Specific growth rate and internal N:P composition throughout the nutrient remediation trials, mean (±standard error).

Species

Specific growth rate

Start (Day 0) Day 3 - 10 End (Day 10)

N P (MgM (Mg'Mg g-1) 1)

N:P (molar)

N:P (molar)

(Mg'Mg-1)

N:P (molar)

66.0 7.72 46.8 7.3

Stigeoclonium sp. 0.19 (0.03) (0.6) (0.1) 13.5 (0.6) (0.3) (0.1)

18.9 (0.2) 13.9 (1.3)

0.17 (0.04) 69.8 9.2 75.6 24.0

C.vulgaris (0.4) (12) 10.5 (0.9) (2.9) (118)

16.8 (1.3) 12.0 (12)

0.10 (0.03) 79.5 22.6 81.5 16.0

S. obliquus (2.1) (2.8) 9.6 (0.5) ^ (7.8) (3.0)

7.8 (0.8) 11.3 (2.2)

0.12 (0.05) 82.4 9.8 75.4 24.5

C.sorokiniana (17.8) (0.8) 15.0 (1.0) (2.7) (110)

20.3 (1.1) 16.3 (0.8)

0.06 (0.03) 87.0 11.6 91.0 13.3

M.aeruginosa (nd) (nd) 16.6 (1.0) (nd) (nd)

16.6 (nd) 15.1 (nd)

nd = undetermined

<u S1 S3 S

H3 '<3

6 7 8 9 10

N biomass composition % (mgN-mg biomass-1)

Figure 1 N remediation and specific growth in relation to species' internal N composition. N remediation (■) and specific growth rate (o). Uptake rates calculated using TSS data when available (mean ± standard error).

Initial PO4-P concentration (mgP-L"1) Initial PO4-P concentration (mgP-L-1)

Figure 2 Cell uptake rate for a) PO4-P and b) NH4-N (mean ± standard error) by suspended S.obliquus (□) and C.vulgaris (•) in secondary wastewater effluent with varying initial PO4-P concentration.

HRT (d) HRT (d)

Figure 3 Optimal cell concentration for a) S.obliquus and b) C.vulgaris with HRT for influent PO4-P concentrations of (■) 1, (o) 5 and (▲) 10 mgP-L-1 (mean ± standard error), (-- ) denotes approximate equivalent biomass concentration for a HRAP.

HRT (d)

23 HRT (d)

Figure 4 Corresponding bead.mL-1 concentration for a) S.obliquus and b) C.vulgaris with HRT up to 5 days for influent PO4-P concentrations of (■) 1, (o) 5 and (▲) 10 mgP-L"1 (mean ± standard error), (--) denotes 1:1 (bead:wastewater v/v) and maximum bead-mL"1 concentration possible.

2 4 6 8

Influent PO4-P concentration (mg-L-1)

Figure 5 Theoretical minimum HRT at 1:1 (v/v) bead concentration for PO4-P

remediation by S.obliquus (□) and C.vulgaris (•).

Highlights

• Relationship between species' N:P composition and nutrient remediation evidenced.

• Species selection can be informed through internal composition characterisation.

• A composition of high N and low P demonstrate enhanced removal.

• Remediation in <3 h through biomass and species selection for effluents < 5mgP-L-1.