The effect of harvesting on biomass production and nutrient removal in phototrophic biofilm reactors for effluent polishing

Abstract

An increasing number of wastewater treatment plants require post-treatment to remove residual nitrogen and phosphorus. This study investigated various harvesting regimes that would achieve consistent low effluent concentrations of nitrogen and phosphorus in a phototrophic biofilm reactor. Experiments were performed in a vertical biofilm reactor under continuous artificial lighting and employing artificial wastewater. Under similar conditions, experiments were performed in near-horizontal flow lanes with biofilms of variable thickness. It was possible to maintain low nitrogen and phosphorus concentrations in the effluent of the vertical biofilm reactor by regularly harvesting half of the biofilm. The average areal biomass production rate achieved a 7 g dry weight m⁻² day⁻¹ for all different harvesting frequencies tested (every 2, 4, or 7 days), corresponding to the different biofilm thicknesses. Apparently, the biomass productivity is similar for a wide range of biofilm thicknesses. The biofilm could not be maintained for more than 2 weeks as, after this period, it spontaneously detached from the carrier material. Contrary to the expectations, the biomass production doubled when the biofilm thickness was increased from 130 μm to 2 mm. This increased production was explained by the lower density and looser structure of the 2 mm biofilm. It was concluded that, concerning biomass production and labor requirement, the optimum harvesting frequency is once per week.

Appendices

Appendix

$$ {{\mathrm{Penetration}\ \mathrm{of}\ \mathrm{NO}}_3}^{\hbox{--} }{{,\mathrm{PO}}_4}^{3\hbox{--} }{{,\mathrm{and}\ \mathrm{HCO}}_3}^{\hbox{--} } $$

The algal growth on NO₃ ^-, PO₄ ^3-, and CO₂ can be described by a stoichiometrical reaction equation. With the measured elemental composition of the biomass, this equation is as follows:

$$ 1\ {\mathrm{CO}}_2+0.15\ {{\mathrm{N}\mathrm{O}}_3}^{\hbox{--} }+0.008\ {{\mathrm{P}\mathrm{O}}_4}^{3\hbox{--} }+0.0036\ {{\mathrm{S}\mathrm{O}}_4}^{2\hbox{--} }+0.98\ {\mathrm{H}}_2\mathrm{O}\to {\mathrm{CH}}_{1.77}{\mathrm{O}}_{0.59}{\mathrm{N}}_{0.15}{\mathrm{P}}_{0.008}{\mathrm{S}}_{0.0036}+1.35\ {\mathrm{O}}_2+0.18\ {\mathrm{O}\mathrm{H}}^{\hbox{--} } $$

From this stoichiometrical growth equation, the yield of the different components can be calculated. The biomass content and the yields are shown in Table 6.

Table 6 Parameters for calculating the penetration depth of NO₃ ⁻, PO4³⁻, and HCO₃ ⁻ in the biofilm and parameters for calculating the light intensity at the penetration depths

Fig. 10

The normalized spectral distribution of the PAR photons (E_n,PAR,λ 400-700 nm) of the Phillips Master PL-L

Fig. 11

The specific absorption coefficient (a_λ) for Chlorella sorokiniana

The penetration depths of NO₃ ^–, PO₄ ^3–, and HCO₃ ^– are calculated according to the following formula as described for instance in Perez et al. (2005):

$$ {L}_{p,i}=\sqrt{\frac{2\cdot {D}_i\cdot {C}_{i,l/b}}{\frac{\mu_{\max }}{Y_i}\cdot {C}_x}}\left[\mathrm{m}\right] $$

with L _pi the penetration depth of nutrient i (meters), D _i the diffusion coefficient of nutrient i (square meters per second), C _i,l / _b the concentration of nutrient i at the liquid–biofilm interface (grams per cubic meter), μ_max the maximum specific growth rate (per second), Y _i the yield of biomass on nutrient i (gram biomass per gram nutrient i), and C _x the algae concentration (gram per cubic meter).

Table 6 shows the parameters that were used for the calculation. For the concentrations at the biofilm surface, it was assumed no mass transfer limitation occurred at the liquid–biofilm interface.

Penetration of light

The following formula was used to calculate the light intensity at depth z inside the biofilm:

$$ \mathrm{PFD}(z)={\displaystyle \sum_{\lambda =400}^{\lambda =700}\left({\mathrm{PFD}}_{\mathrm{in}}\cdot {E}_{n,\mathrm{PAR},\lambda}\cdot {e}^{-{a}_{\lambda}\cdot {C}_x\cdot z\cdot d}\cdot \varDelta \lambda \right)\left[\upmu \mathrm{mol}{\mathrm{m}}^{\hbox{-} 2}{\mathrm{s}}^{\hbox{-} 1}\right]} $$

with PFD_in the photon flux density of the incoming light (picomoles per square meter per second), E _n,_PARλ the normalized spectral distribution of the PAR photons (400–700 nm nm^–1), a _λ the specific absorption coefficient (square meters per gram), C _x the algae concentration (grams per cubic meter), z the biofilm depth (meters), d a light-path enhancement factor (−), and ∆λ the wavelength interval (nanometers).

Table 6 shows the parameters that were used for this calculation. To obtain the E _n,PAR,λ for the Phillips Master PL-L lamps for the experiments, the spectral photon flux density was measured using a fiber-optic CCD-based spectroradiometer (AvaSpec-2048 detector, Fiber FC-IR100-1-ME, Avantes, Eerbeek, The Netherlands) at 1-nm intervals (for details on the measurement protocol, see Vejrazka et al. (2011)). This measurement was normalized for the PAR range to obtain the normalized emission spectrum according to the following equation:

$$ {E}_{n,\mathrm{PAR},\lambda }=\frac{{\mathrm{PFD}}_{\lambda }}{\mathrm{PFD}}\left[{\mathrm{nm}}^{-1}\right] $$

with PFD_λ the spectral photon flux density (micromoles per square meter per second per nanometer) and PFD, the photon flux density in the PAR range(400–700 nm, micromoles per square meter per second ). Figure 10 shows E _n,PAR,λ of the Phillips Master PL-L lamps. The a_λ used for phototrophs adapted to high light conditions (top layer of the biofilm) was the a_λ measured for Chlorella sorokiniana shown in Fig. 11 (for details on the cultivation see Kliphuis et al. (2010) and for details of the measurement protocol see Vejrazka et al. (2011)). In order to simulate low light adapted phototrophs this a_λ was multiplied with a factor 2.