Network refinement (Fig. 10.3c) can be viewed as reconciliation between the content of the model and the available biochemical and genomic data, with the end result of enhancing the reconstructed network. This reconciliation is done based on agreements of model simulations and updated genomics, physiological, and bio­chemical knowledge. A crucial step in the reconstruction of genome-scale meta­bolic models is filling the gaps to decrease the number of dead-end metabolites and improve network connectivity. Metabolic network gaps are filled by the addition of reactions that are missing in the network yet have corroborating evidence for their existence in the system. These may include spontaneous reactions that are not associated with gene products as well as extracellular and intracellular transport reactions and exchange reactions.

Models may not predict the production of biological compounds with existing biochemical evidence if the prerequisite genes have not been added to the model. Manichaikul et al. (2009), using Chlamydomonas as a model, described how genomic data can be used to fill gaps in metabolic models. In their approach, not only genomics and other experimental evidence contributed to the refinement of the network, but also the model itself informed “genomics,” of the presence of missing annotations, justifying the use of more sensitive sequence search and annotation tools to recover the missing genes. One example that can illustrate this is lactate dehydrogenase (LDH), which initially was absent from the Chlamydomonas gene annotation, yet the model reconstruction showed the need for the LDH enzyme in the Chlamydomonas pyruvate metabolism pathway. A PSI-BLAST analysis was carried out to identify the gene encoding LDH; the gene was subsequently added to the model. Additionally, orphan genes, or those biochemically characterized metabolic enzymes lacking sequence data, can be assigned GPRs by reviewing metagenomic sequence data to provide sequences for the missing enzymes. This approach has been experimentally validated (Yamada et al. 2012).

Patrick V. Brady, Mark P. McHenry, M. Carolina Cuello and Navid R. Moheimani

Abstract Industrial-scale microalgae production will likely require large energy­intensive technologies for both culture and biomass recovery; energy-efficient and cost-effective microalgae dewatering and water management are major challenges. Primary dewatering is typically achieved through flocculation followed by sepa­ration via settling or flotation. Flocculants are relatively expensive, and their presence can limit the reuse of de-oiled flocculated microalgae. Natural flocculation of microalgae—autoflocculation—occurs in response to changes in pH and water hardness and, if controlled, might lead to less-expensive “flocculant-free” dewa­tering. A better understanding of autoflocculation should also prompt higher yields by preventing unwanted autoflocculation. Autoflocculation is driven by double­layer coordination between microalgae, Ca+2 and Mg+2, and/or mineral surface precipitates of calcite, Mg(OH)2, and hydroxyapatite that form primarily at pH > 8. Combining surface complexation models that describe the interface of microalgae: water, calcite:water, Mg(OH)2:water, and hydroxyapatite:water allows optimal autoflocculation conditions—for example pH, Mg, Ca, and P levels—to be iden­tified for a given culture medium.

P. V. Brady (H)

Geoscience Research and Applications Group, Sandia National Laboratories,

Albuquerque, USA

e-mail: pvbrady@sandia. gov

M. P. McHenry

School of Engineering and Information Technology, Murdoch University,

Perth, Australia

M. Carolina Cuello • N. R. Moheimani

Algae R&D Centre, School of Veterinary and Life Sciences, Murdoch University, Perth, Australia

© Springer International Publishing Switzerland 2015

N. R. Moheimani et al. (eds.), Biomass and Biofuels from Microalgae,

Biofuel and Biorefinery Technologies 2, DOI 10.1007/978-3-319-16640-7_13

13.1 Introduction

Microalgae achieve a much higher biomass productivity when compared to con­ventional terrestrial biofuel crops (Fon Sing et al. 2011), although several issues that must be resolved include developing cost-effective dewatering and processing technologies, and the associated commercial and environmental challenges (Grif­fiths and Harrison 2009; Moheimani et al. 2011; Vasudevan and Briggs 2008). Cost-effective and energy-efficient dewatering of microalgae, nutrient recycling, and control of effluent wastewater are becoming major challenges to microalgae producers (Borowitzka and Moheimani 2010; Charcosset 2009; Clarens et al. 2010; Wyman and Goodman 1993; Xiong et al. 2008). Open pond microalgae production can become expensive due to variable capital, operational, and downstream pro­cessing costs derived from the low microalgae cell densities (Lee 2001; McHenry

2010) , and if not optimised, industrial microalgae production will consume large volumes of water through evaporative loss (Chisti 2007; Clarens et al. 2010), generate effluent and become an energy-intensive process (Borowitzka and Mo- heimani 2010; Charcosset 2009; Clarens et al. 2010; McHenry 2013; Wyman and Goodman 1993; Xiong et al. 2008). It is generally not well understood that mic­roalgae production will require large energy-intensive technologies for both culture and biomass recovery (Chisti 2007), including heat exchangers, scrubbing, pres — surisation, pipeline construction, and pumping of flue gases (as a potential source of CO2) to be intensively managed (McHenry 2010, 2013), and microalgae production sites must be carefully chosen to optimise industrial resources, natural resources, and environmental conditions to facilitate post-harvest processing (Borowitzka 1992).

Using the AM1.5 direct solar spectrum as a baseline, we can model the amount of energy that can be converted into electricity by a solar cell if a portion of the solar spectrum is diverted to algae production and only the remainder is provided to the solar cells. This allows us to determine the viability of a cultivation system based on this concept in terms of generating electivity or increasing the portion of specifically targeted PAR available for cultivation. The proposed lossless system places a filter or device above the algae pond to split the spectrum into the appropriate compo­nents. We do not consider the mechanism used to redirect the light or the specifics of how the system will function. However, one candidate technology would be the luminescent solar concentrator or a variation thereof. Although the exact mecha­nism is not described, the model assumes all the light not provided to the algae is directed to a solar cell. The model assumes there are no losses associated with transmission of light through the filter or reflections from the surfaces of the filter or solar cell. The model also disregards electrical resistance in the transmission of the generated electricity.

The first aspect of this model is to determine the component of the spectra absorbed by the microalgae. As can be seen from Fig. 15.2, on average the main chlorophyll absorption peaks are centered at 434 and 662 nm. The portion of the spectrum transmitted to the algae was varied by changing the threshold around these peaks. For example, full-width half-maximum (50 % threshold) meant the spectra from 400 to 492 nm and 644 to 678 nm was transmitted to the algae, while for a threshold of 80 %, only the spectra from 417 to 458 nm and 656 to 670 nm were transmitted to the algae. Additionally, the light given to the algae is limited to between 400 and 700 nm as this is the region typically considered PAR (photo­synthetically active radiation). All energy not transmitted to the algae is provided to the photovoltaic device for producing electricity. To calculate the power absorbed by the different microalgae species, the AM1.5 solar spectrum is multiplied by the absorbance spectrum (Fig. 15.2). The allocation of the solar spectrum as the bandwidth changes and the power absorbed by the microalgae (nannochloropsis) can be seen in Fig. 15.4.

With this allocation of the solar spectrum, we can calculate the power generated by a solar cell in hypothetical system using the reported spectral response graphs and parameters for crystalline silicon (Beardall et al. 2009) and amorphous silicon (Meier et al. 2004).

The short-circuit current density (Jsc) generated by a solar cell is calculated from:

where EQEQ.) is the external quantum efficiency as a function of wavelength, Ф(к)АМ1 5 is the photon flux density calculated from the AM1.5 (Global Tilted) solar spectrum, PV(k) is function defining the portion of spectrum not transmitted to the microalgae, and q is the charge of an electron.

The open-circuit voltage (VOC) of the solar cell is dependent on the short-circuit current density and will vary with the irradiance that is incident upon the cell. This can be calculated from (Messenger and Ventre 2010):

where Eg is the bandgap of the semiconductor material, k is Boltzmann’s constant, T is the cell temperature in K, and J0 is derived from the published parameters of each device.

The power generated (P) in W. m 2 from the cell is then:

P = FF. Isc ■Voc

where the fill factor (FF) is the value published in the literature for each cell type.

The power generated by the photovoltaic modules from this limited spectrum and as the bandwidth changed can be seen in Fig. 15.4. The power produced from the solar cells can be directed to the powering facilities associated with the growth of microalgae (pumps and monitoring systems) or to provide additional illumina­tion to the microalgae. The former would reduce the costs of running the plant whereas the latter can boost growth productivity. Light emitting diode (LED) arrays can be used to most efficiently provide additional lighting to the microalgae at a specific wavelength. LEDs are highly efficient solid-state devices for converting electricity into light. They can be designed to emit light in a range of wavelengths to match the spectral. The internal quantum efficiency of high quality LEDs can exceed 99 %. This sounds extremely efficient, and however, there are difficulties in extracting the light from the LED which leads to low external quantum efficiencies (EQE) in the order of only a few percent (Schnitzer et al. 1993).

There is a significant amount of research effort into increasing the external quantum efficiencies of LEDs. As a result, LEDs are produced from a range of materials and use a variety of technologies. Some of the resulting LEDs include blue emitting InGaN-GaN LED’s with a EQE of 40 % (Gardner et al. 2007), thin — film GaAs LEDs with a 30 % EQE, (Schnitzer et al. 1993), and organic LEDs with an EQE of 30 % (Kim et al. 2013). In some cases, careful texturing can improve the light extraction efficiency which yields LEDs with an EQE greater than 50 % and in some cases up to 60.9 % (Krames et al. 1999a).

The most useful LED for adding targeted illumination to a microalgae pond would be those LEDs with high external quantum efficiencies at particular wave­lengths, such as those discussed by Krames et al. with efficiencies of 60.9 % (Krames et al. 1999a).

The additional power (P) in Wm 2 that can be produced using the power gen­erated (Pin) using the system modeled above can be calculated from:

P = EQELed .Pin

where EQELed is the external quantum efficiency of the LED.

The additional irradiance from these LEDs is assumed to be tailored to the peak absorbance of the respective microalgae species. The total power absorbed by the algae is thus the irradiance absorbed directly by the microalgae and the peak absorbance multiplied by P.

Algenol (Florida, USA) has been continually refining a process for continuous growth and harvest of ethanol excreted by modified cyanobacteria. In this process, ethanol is released by the organism in the vapour phase and then captured for extraction using a novel distillation process, eventually the spent microalgae bio­mass is converted to fuel using a variant of the HTL process. Algenol claims the following figures for their pilot plant (http://www. algenol. com/):

• Yield: 8000 gallons per acre of total liquid fuel production (80,000 L/ha) of which 85 % is ethanol and the remaining 15 % is hydrocarbons

• Cost: $1.27 per gallon

Two other companies pursuing milking-based projects are Joule unlimited and Proterro who are focused on chemicals/fuels and sugars as feedstock to traditional biofuel processes, respectively. In all of these approaches, the process is funda­mentally different as the milking process extracts the oils, ethanols or other chemicals of interest from the growth medium without killing the microalgae. As a comparison, the traditional microalgae production systems ‘kill’ the ‘cow’ (microalgae) to extract the ‘milk’ (oil) rather than keeping the cow (microalgae) productive and continually harvesting the milk (oil) (Moheimani et al. 2013a). Milking addresses the shortfalls of the existing production systems in two major ways:

• Nutrients—Only the products of interest are removed (which typically contain very low N and P), and as a result, there is a limited requirement for fertilisers. Only water, CO2 and sunlight are required to continually produce the compounds

• Dewatering—Microalgae are typically not removed from the culture to be milked to limit the need for dewatering (Moheimani et al. 2013b).

These systems are currently at various stages of early development. Despite this, the potential of these novel approaches to address the major issues with traditional methods warrants their continued investigation.

Many industries and human activities generate wastewater, and as a consequence, there are many different types of wastewaters, each with a different chemical composition and volumetric production over time. Table 5.1 gives an overview of different types of wastewaters and their content of N and P. The two major sources of wastewater are domestic wastewater and wastewater derived from animal manure. Each person supplies about 3 kg N per year through domestic wastewater, which translates to about 21 million ton of N per year for a global population of 7 billion (similar to the estimate of Smil 2002; Van Harmelen and Oonk 2006). The major livestock animals that produce manure are pigs, cattle, and chickens.

Chickens produce a relatively dry type of manure that is suitable for application as fertilizer on agricultural land and we do not consider the N supply from chicken manure for algae production. Pigs produce about 16 kg N animal 1 year 1, cattle 35 kg N animal-1 year-1, and dairy cattle 75 kg animal-1 year-1. With global population sizes of 1 billion pigs, 0.9 billion cattle and 0.25 billion dairy cattle, wastewater from pig and cattle manure can theoretically provide 65 million tons of N, or about 3 times more than domestic wastewater (similar to the estimate of Van Harmelen and Oonk 2006). If we assume a “N” content of microalgal biomass of 7 %, the total human, pig, and cattle wastewater N nutrient is enough to produce about 778 million ton dry microalgal biomass per year. This is in the same order of magnitude as the global production of wheat or of corn. Although this is a lot of biomass, it can produce only about 233 million ton of oil (assuming a 30 % lipid content in microalgae). This corresponds to 1800 million barrels of oil, or only about 5 % of the global oil consumption. Thus, wastewater alone cannot supply sufficient nutrients for microalgal biomass to make a large contribution to the world’s energy demand. This conclusion is in line with Peccia et al. (2013) or Chisti (2013), who estimated that nutrients from domestic wastewater of a typical large city can only produce enough microalgal biofuel to supply 3 % of the fuel demand of that city. If microalgal biofuels are ever to make a larger contribution to the global fuel demand, it will be essential to recycle the nutrients during conversion of microalgal biomass to biofuels (Venteris et al. 2014). Although animal manure is a potentially significant source of nutrients for microalgae production, it should be noted that a significant proportion of animal manure is already used as a fertilizer in conventional agricultural production (Bouwman and Van Der Hoek 1997). Because synthetic fertilizer prices are increasing, the value of nutrients in animal manure also increases. Therefore, the use of animal manure as fertilizer in conventional agriculture is likely to increase in the future (Shilton et al. 2012), and microalgae and conventional agriculture may compete for animal manure nutrients. However, the main limitation of the use of raw animal manure in conventional agriculture is the high transport cost resulting from the high water content of animal manure. In areas where livestock numbers are high and agricultural crop production is nutrient — limited, it will unlikely be economically feasible to transport animal manure to the field, and microalgae may become a more attractive option to process large volumes of manure on a relatively small land area.

In addition to manure, there are many other sources of wastewater that could be used for microalgae biomass production, such as wastewater from olive mills, wineries, breweries, vegetable processing, tanneries, or the paper industry (Cai et al.

2013) . Some emerging technologies also generate a lot of wastewater. The use of anaerobic digestion to convert organic waste streams into methane is growing worldwide and generates a nutrient-rich effluent that could be processed with microalgae (Uggetti et al. 2014). Aquaculture is also increasing worldwide and generates a similar nutrient-rich wastewater that may be suitable for treatment using microalgae (e. g., Van Den Hende et al. 2014). The volumes that are produced by these industries are relatively small compared to the volumes of domestic and animal manure wastewater. Nevertheless, microalgae may be a solution to treat some of these wastewaters as conventional water treatment technologies may be too expensive or ineffective.

The RNAi pathway has been studied in the unicellular green alga C. reinhardtii, and used as a reverse genetics tool in different algal species. Complex sets of endogenous small RNAs, including candidate microRNAs and small interfering RNAs, have been identified in four algal species, C. reinhardtii (Molnar et al. 2007; Zhao et al. 2007), Porphyra yezoensis (Liang et al. 2010), Phaeodactylum tricornutum (De Riso et al. 2009), and Ectocarpus siliculosus (Cock et al. 2010). However, RNAi mechanisms and their applications remain largely uncharacterized in most algae. RNAi against specific genes can be induced by the introduction of exogenously synthesized dsRNAs or siRNAs into cells or whole organisms (Cerutti et al. 2011; Moellering and Benning 2010; Molnar et al. 2009). Within algal species, in vitro — synthesized long dsRNAs have been electroporated into Euglena gracilis cells and shown to silence two endogenous genes homologous to the introduced dsRNAs (Iseki et al. 2002; Ishikawa et al. 2008). Recently, an RNAi triple knockdown of the three most abundant LHCII proteins (LHCBM1, 2 and 3) has been reported in Chlamydomonas with the aim of increasing the efficiency of photobiological H2 production (Oey et al. 2013). Artificial microRNA (amiRNA) expression success­fully exploits endogenous miRNA precursors to generate small RNAs that direct gene silencing in C. reinhardtii (Molnar et al. 2009; Schmollinger et al. 2010; Zhao et al. 2008). Zhao et al. (2009) developed an artificial amiRNA-based strategy to

knock down gene expression in Chlamydomonas using an endogenous Chlamydo — monas miRNA precursor (pre-miR1162) as the backbone. Other studies show that amiRNAs can be used as a highly specific, high-throughput silencing system, and they propose that they will become the system of choice for analysis of gene function in Chlamydomonas and related organisms (Molnar et al. 2009; Schmollinger et al.

2010) . The synthesized miRNAs provide a convenient tool for reverse genetic studies in Chlamydomonas. More recently, epitope-tagged protein-based amiRNA (ETPamir) screens were developed, in which target mRNAs encoding epitope — tagged proteins were constitutively or inducibly co-expressed in protoplasts with amiRNA candidates targeting single or multiple genes (Li et al. 2013). This design allowed parallel quantification of target proteins and mRNAs to define amiRNA efficacy and mechanism of action, circumventing unpredictable amiRNA

Fig. 8.2 Genome engineering tools. a miRNA pathway. MIR genes are transcribed by RNA polymerase II into pri-miRNA transcripts that are further processed into pre-miRNAs harboring a characteristic hairpin structure. From the stem of the pre-miRNA the miRNA/miRNA* duplex is excised by DCL1 and can be assisted by HYL and SE proteins. miRNA-guided AGO-containing RNA-induced silencing complex (RISC) directs mRNA cleavage or translation inhibition of the target transcript. b Summary of Transcription activator-like effectors (TALEs) nuclease. Custom — designed nucleases introduce double-strand breaks with high precision at predetermined genomic loci. Double-strand breaks are either repaired by error-prone non-homologous end-joining (NHEJ) or high fidelity homologous recombination (HR). NHEJ repair causes random insertions and/or deletions of nucleotides around the target site and some of these mutations will knockout gene function. Gene replacement, tagging, or correction can be achieved by HR-mediated targeted integration of a donor construct that is provided together with a nuclease pair. c CRISPR/Cas9 mediated target DNA cleavage. The CRISPR loci include Cas genes, a leader sequence, and several spacer sequences derived from engineered or foreign DNA that are separated by short direct repeat sequences. Cleavage occurs on both strands, 3 bp upstream of the NGG proto-spacer adjacent motif (PAM) sequence on the 3′ end of the target sequence and is followed by DNA repair by the endogenous cellular repair machinery expression/processing and antibody unavailability. These screens could improve algal biofuel engineering research by making amiRNA a more predictable and manageable genetic and functional genomic technology. From a practical perspec­tive, RNAi is becoming a customary method for directed gene silencing in algae. As the necessary molecular tools are developed, RNAi approaches are expected to contribute to the functional characterization of novel genes, as well as to the strain engineering of algae (Fig. 8.2a). Ultimately, RNAi technology may provide much-needed insights into gene function, metabolic pathways, and regulatory networks allowing us to comprehend the role of algal species in nature, as well as to engineer these organisms for the synthesis of valuable bioproducts.

Fusion Proteins Specific regions within the coding regions of chloroplast genes enhance efficient expression of foreign genes in plants and algae (Anthonisen et al. 2002; Gray et al. 2009; Kuroda and Maliga 2001b). Genetic fusion of endogenous regions to an exogenous protein of interest may represent another effective strategy for high-level transgene expression (Gray et al. 2011; Kasai et al. 2003). A dis­advantage of the fusion protein approach may be reduced industrial or clinical values or increased cost of purifying the protein of interest from its fusion partner.

Codon Optimisation Codon optimisation has been shown to be an important factor of heterologous gene expression in algae (Heitzer et al. 2007). It has been dem­onstrated that protein expression levels can be improved in the chloroplast by up to * 80-fold (Franklin et al. 2002a) by adjusting the codon bias of the transgene to the AT-rich chloroplast codon bias. Despite this, the effects of codon bias on expression levels are largely heuristic and still not well understood at a theoretical level.

Replacement of Highly Expressed Photosynthesis Genes The highest protein expression levels in Chlamydomonas have so far been demonstrated in transfor­mants carrying the psbA promoter and 5′ UTR for transcription and translation initiation in a psbA knockout background, which leads to a non-photosynthetic strain (Manuell et al. 2007; Surzycki et al. 2009). PsbA encodes for the D1 protein in the photosystem II reaction centre and is the most rapidly synthesised protein at high light intensity in higher plants and algal cells (Trebitsh et al. 2000). Although photosynthesis was restored by introducing the psbA gene at a different location, the presence of the psbA protein decreased the yield of foreign protein production (Manuell et al. 2007). This may reflect competition between the two genes, but could also be due to primary energetic or biosynthetic limitations.

Open Reading Frame Orientation In the chloroplast, 3′ UTR of the RNA often shows the potential for stem-loop formation, which serves for transcript stabilisa­tion rather than transcription termination (Rott et al. 1998). Therefore, a degree of ‘read through’ from the upstream gene is possible (Oey et al. 2009), which in parallel can increase the amount of translatable RNA and thus potentially increase the amount of protein (Stern and Gruissem 1987).

Heterologous and Hybrid Regulatory Systems Rasala et al. (2011) have recently shown that an increase in protein production can be achieved by the fusion of the 16S ribosomal promoter, which does not contain translation initiation signals such as Shine-Dalgarno sequences, to the endogenous atpA 5′ UTR containing the translation initiation signal. This, and the demonstration that heterologous regula­tory elements can significantly induce the expression rate (Kuroda and Maliga 2001a; Oey et al. 2009; Ruhlman et al. 2010), suggests that designed regulatory elements could serve to improve expression.

Inducible Systems Environmental changes or developmental factors naturally influence the up — and down-regulation of genes. Understanding those regulatory mechanisms provides a valuable genetic tool, for example to switch protein expression automatically or under the control of specific circumstances. Examples of inducible algal promoters include light responsive genes (Falciatore et al. 1999), nitrogen starvation (Poulsen et al. 2006; Poulsen and Kroger 2005), a sulphur — regulated arylsulfatase gene and an ISG glycoprotein. Tightly controlled expression of toxic proteins (e. g. the growth factor DILP-2) is also often desirable (Surzycki et al. 2007). The expression of psbD (D2 component of PSII) is dependent on the Nac2 gene fused to the copper-sensitive cytochrome c6 promoter (cyc6) and induced in copper deficiency and repressed in the presence of copper.

Riboswitches Riboswitches that can be used to regulate protein expression at the translational level have been shown to be functional in C. reinhardtii and Volvox carteri (Croft et al. 2007) suggesting that it may be a useful technique for other algae species as well.

Chemical flocculation adds flocculating and/or coagulating agents to the culture medium to speed cell aggregation. This process is often used with microalgae as a pretreatment in combination with other processes such as dissolved air flotation (DAF). While this improves the speed at which the cells are collected, it has the added complication of dosing chemicals at a specific desired concentration to achieve this rate. Generally, coagulants are used to neutralize the charges of the particles in the solution, and flocculants are the chemicals used to aggregate the particles. These chemicals (flocculants and coagulants) complicate the overall process in that they add additional cost, often add metals or other compounds that need to be disposed of in the resulting biomass, and complicate any downstream processing of the materials into primary and co-products. Negative impacts have been countered in a number of ways, including the use of degradable biopolymers as flocculants (e. g., polyacrylamide and starches) and electroflocculation where no flocculant is directly added.

Inorganic flocculants and coagulants are typically iron or aluminum based and are used to neutralize the surface charge. This method requires a significant input of the inorganic flocculant which adds to the sludge; this adds to the OpEx both in the inputs and in processing (to remove the chemicals). Additionally, the process is sensitive to pH, usually working best at higher pH but varies with strain and culture condition. All algal strains do not respond the same to a particular chemical, so tailoring will be required to fit the organism being harvested (Chen et al. 2011). The chemical flocculants can also be a problem for downstream use of the biomass for feeds, feedstock for anaerobic digesters, and residual ions can be a problem for use of digestates for land application as a soil amendment (Christenson and Sims 2011).

An example is the use of alum (hydrated potassium aluminum sulfate) for flocculation of Scenedesmus and Chlorella cultures (Molina Grima et al. 2003). Knuckey (2006) used ferric-induced flocculation at 0.5 mg L 1 to concentrate algae that had been pH-adjusted to about pH 10 by 200-800-fold. The floccs needed to be neutralized after concentration. They successfully flocculated Chaetoceros calci — trans, C. muelleri, Thalassiosira pseudonana, Attheya septentrionalis, Nitzschia colesterium, Skeletonema sp., Tetraselmis suecica, and Rhodomonas salina all with >80 % efficiency. All were marine algae useful for aquaculture feeds (Knuckey et al. 2006). As an example of strain differences, it has been reported that B. braunii flocculated better at pH 11 (Lee et al. 1998).

Organic flocculants and coagulants can also be used; these are high molecular weight bridging polymers (e. g., chitosan and starch) that react with cells in the culture to make large aggregates and help speed up flocculation (Edzwald 1993). It is reported that these biodegradable polymers do not contaminate the biomass as much as inorganic coagulants and that cationic polymers are superior to anionic and neutral ones. Cationic polyelectrolytes (e. g., Dow C-31) induced algal cells to flocculate while anionic and nonionic polymers were shown to be ineffective (Tenney et al. 1969). Changing the pH was necessary to optimize the flocculation with these cationic polymers, with essentially no flocculation at high pH (above 8) and maximal flocculation between pH 2 and pH 4. The flocculation was attributed to a bridging by the polymers between algal cells to form a matrix. The chemicals necessary for pH adjustment (and neutralization) significantly add to the overall OpEx of flocculation processes. The flocc formation is inhibited by high salt concentrations, forms better at high biomass density, and is subject to shear dis­ruption, which slows the flow rates possible.

Growth stage has also been reported to have an impact on the flocculation procedure. For example, Botryococcus braunii cultures flocculated easiest after two weeks of culture regardless of harvesting method (Lee et al. 1998).

Current development of new anaerobic digester designs and configurations can play an important role in the future of hybrid microalgae production and anaerobic digestion facilities. New digester reactor designs that decouple the hydraulic and solid retention times can be extremely beneficial for anaerobic digestion of micro­algae (Ward et al. 2014) and can be achieved by upflow anaerobic sludge blanket reactors, anaerobic membrane reactors, anaerobic filters and bed reactors, in-pond fermentation pits and also by two-stage anaerobic digestion (Gao et al. 2007; Goodwin et al. 2001; Green et al. 1995a, b; Haridas et al. 2005; Inglesby and Fisher 2012; Shin et al. 2010; Varel et al. 1988; Vergara-Fernandez et al. 2008; Zamalloa 2012; Zhou et al. 2009). New reactor designs allow better control, breakdown and conversion of organic matter within the digester (Vergara-Fernandez et al. 2008). In particular, two-stage anaerobic digestion physically separates stages of digestion into separate reactors, allowing much better control over the anaerobic digestion process (Dinsdale et al. 1996; Dugba and Zhang 1999; Shin et al. 2010; Varel et al. 1988; Vergara-Fernandez et al. 2008; Yu et al. 2002). Furthermore, molecular research underway is providing new insights into the microbial communities associated with the anaerobic digestion process and aim to understand the impact of internal envi­ronmental conditions and change within the digester community (Keyser et al. 2006; Shin et al. 2010; Supaphol et al. 2011). For example, new molecular methods in the anaerobic digestion process (such as “poly chain reaction denaturing gel gradient electrophoresis” and “real time poly chain reaction”) are under investigation to improve metabolic pathways within the digester and the abundance and species composition of bacterial populations (Patil et al. 2010a, b; Shin et al. 2010; Skilman et al. 2009; Supaphol et al. 2011; Ward et al. 2015; Zhang et al. 2012; Ziganshin et al.

2011) . A greater understanding of the anaerobic digestion processes at the molecular level can optimise the anaerobic digestion process to specific microalgal biomass and co-digestate substrates (Ward et al. 2014).

16.2 Conclusions

The integration of anaerobic digestion with microalgae-based biofuels production is able to attain higher efficiency and improve sustainability of the production of microalgae-based biofuels. Several of the technical issues including the low con­centration of biodegradable (digestible) microalgae substrates, cell wall disruption and high lipid concentrations can be overcome by the pre-treatment methods used to process microalgae for liquid and gaseous biofuels. Gas produced by the anaerobic digestion of residual microalgae biomass can be used for electrical or thermal energy within the microalgae biofuels bio-refinery, while the high density microalgae cultures can provide efficient biogas purification. The resulting digestate has been shown to be an ideal nutrient source for the continued growth of additional microalgae biomass and help to close the nutrient loop associated with large-scale microalgae biomass production. With a greater understanding of the different algae species and their characteristics, the anaerobic digestion of microalgae and their residues can be optimised to play an essential role in the sustainable future of clean energy derived from microalgae biomass.