The morphology, pigments, metabolic capabilities, and cell wall structure of the eukaryotic microalgae are quite diverse because they represent a multitude of phylogenetically distinct groups of organisms. Recent molecular evidence suggests that the “algae” fit into very different evolutionary lineages including those related to plants, fungi, or animals (Lucentinii 2005) via one or more serial endosymbiotic events (Moestrup 2001a, b). Three classes of primary interest for biofuels pro­duction include the golden-brown algae (Chrysophycea), prymnesiophytes (Prymnesiophyceae), and the eustigmatophytes (Eustigmatophyceae) (Sheehan et al. 1987) and for bioremediation, the Chlorophycea.

6.5.2.1 Chlorophyta

The green algae are a large group estimated to have over 13,000 species, (Guiry

2012) from which the higher plants emerged. Green algae, in common with land plants, have chloroplasts that contain chlorophyll a and b, as well as the accessory pigments P-carotene and xanthophylls, and store carbon as starch and lipids. Their distribution is ubiquitous with marine, freshwater, and terrestrial species and groups adapted to extremes including deserts, arctic zones, hypersaline waters, and deep — sea thermal vents (Lewis and Lewis 2005; De Wever et al. 2009). Several strains are known to have rapid growth rates and high lipid content and have been a focus of biofuel research. Chorella spp., used since the 1950s as a food supplement, have short doubling times and can be cultured to produce between 30 and 55 % lipid content (Becker 1994; Miaoa and Wu 2006). Species of Scenedesmus, Ettlia, Nannochloris, and Monoraphidium grow rapidly (doubling times from 7 to 12 h under nutrient-replete conditions) and have lipid contents ranging from 30 to over 60 % (Griffiths and Harrison 2009). A number of genera in the order of Chloro — coccales, including Actinastrum, Scenedesmus, Chlorella, Closerium, and Gol — enkinia, tend to dominate algae communities in eutrophic waters (Rawson 1956) and algae wastewater ponds (Martinez et al. 2000; Benemann and Oswald 1996). Many are either heterotrophic or mixotrophic.

Kourosh Salehi-Ashtiani, Joseph Koussa, Bushra Saeed Dohai, Amphun Chaiboonchoe, Hong Cai, Kelly A. D. Dougherty, David R. Nelson, Kenan Jijakli and Basel Khraiwesh

Abstract Genomic sequencing is the first step in a systems level study of an algal species, and sequencing studies have grown steadily in recent years. Completed sequences can be tied to algal phenotypes at a systems level through constructing genome-scale metabolic network models. Those models allow the prediction of algal phenotypes and genetic or metabolic modifications, and are constructed by tying the genes to reactions using enzyme databases, then representing those reactions in a concise mathematical form by means of stoichiometric matrices. This is followed by experimental validation using gene deletion or proteomics and metabolomics studies that may result in adding reactions to the model and filling phenotypic gaps. In this chapter, we offer a summary of completed and ongoing algal genomic projects before proceeding to holistically describing the process of constructing genome-scale metabolic models. Relevant examples of algal metabolic models are presented and discussed. The analysis of an alga’s emergent properties from metabolic models is also demonstrated using flux balance analysis (FBA) and related constraint-based approaches to optimize a given metabolic phenotype, or sets of phenotypes such as algal biomass. We also summarize readily available optimization tools rooted in constraint-based modeling that allow for optimizing bioproduction and algal strains. Examples include tools used to develop knockout strategies, identify optimal bioproduction strains, analyze gene deletions, and

Joseph Koussa and Bushra Saeed Dohai contributed equally to this work.

K. Salehi-Ashtiani (H) • J. Koussa • B. S. Dohai • A. Chaiboonchoe • H. Cai K. A.D. Dougherty • D. R. Nelson • B. Khraiwesh

Division of Science and Math, and Center for Genomics and Systems Biology (CGSB), New York University Abu Dhabi, P. O. Box 129188, Abu Dhabi, United Arab Emirates e-mail: ksa3@nyu. edu

K. Jijakli

Division of Engineering, New York University Abu Dhabi, P. O. Box 129188,

Abu Dhabi, United Arab Emirates © 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_10

explore functional relationships within sets in a metabolic model. All in all, this systems level approach can lead to a better understanding and prediction of algal metabolism leading to more robust and cheaper applications.

10.1 Introduction

Algal research gained its first round of momentum, beyond the scientific commu­nity, in 1978 with the launch of the Aquatic Species Program to explore alternative transport fuel sources (Sheehan et al. 1998). As many laboratories started focusing their investigations entirely on algal systems, the amount of data available increased at a near exponential pace. In particular, the development of next-generation sequencing platforms has rapidly and dramatically advanced and increased the amount of available data on algal genomes. New high-throughput phenotypic platforms have made metabolic characterizations broader and more rapid. As in other fields of biological research, integration of disparate data types, as well as contextualization of data remains a central challenge that is addressed by the field of systems biology. Systems level understanding of metabolism is needed for pre­diction of biomass and bioproduct optimization strategies, for algae, or for any other organism because of the wide expanse and the high degree of interconnec­tivity in metabolic networks.

Evolutionarily, the term algae describe a polyphyletic group of organisms and is contentious in definite meaning (Proschold and Leliaert 2007). Currently, in some classifications, it includes superphyla in several separate lineages: stramenopiles which include brown, golden, and yellow algae and diatoms; rhodophyta or the red algae; photosynthetic alveolates, such as dinoflagelates; and the viridiplantae which include green algae (Barton et al. 2007). The phylogenetic classification green algae describe the presumed plant predecessors with photosynthetic capabilities, and a characteristic green color (Besche et al. 2009; Harris 2001). Both green algae and diatoms have shown great potentials as sustainable sources of biofuel, biomass, and bioproducts; however, Chlamydomonas reinhardtii, due to its position as a well studied representative green alga (Harris 2001), has received special interest as a model organism for genomic and metabolic studies.

The interest in algal exploration has steadily increased as commercial and large — scale production of lipid-producing algae has provided a practical importance for the research, with particular demand on more integrated goals, i. e., optimizing algae for biofuel production, optimizing growth of strains of interest and achieving economical viability all at the same time (Koussa et al. 2014). The metabolic optimization of algal organisms, which requires extensive characterization of algal metabolic circuitry, calls for an integrative “systems level” approach. This begins with genome sequencing, which then provides the “parts list” for reconstruction of a metabolic network, and finally ends with the ability to make model-based pre­dictions (Fig. 10.1). In this chapter, we first review microalgal sequencing efforts

investigations

Fig. 10.1 Schematic diagram representing the relationships between the reductionist, systems, and synthetic biology approaches. Through reductionism, a “top-down” approach leads to characterization of few or individual target genes. The collective knowledge gained through individual studies provides the framework for developing large-scale methodologies, building knowledgebases, and executing “bottom-up” omics studies, with the results integrated to describe the emergent properties of the system. Investigations of the models, constraints, and simulations provide predictions, which are implemented through an engineering approach, using biological parts or devices, for synthesis of a desired biological outcome that are in progress or have been completed, we then outline metabolic network reconstruction and constraint-based analyses, and to conclude, we briefly describe some computational tools that are used in metabolic modeling, which can aid the design of engineering experiments and optimization of cellular metabolic outputs.

Some microalgae flocculate spontaneously without addition of flocculants or pre­cipitation of minerals, and the phenomenon is generally referred to as biofloccu­lation. Bioflocculation can occur spontaneously in microalgae-based wastewater treatment systems, including high-rate microalgal ponds where it is often used in biomass harvesting (Park et al. 2011b; Benemann et al. 2012).

Bioflocculation is often due to a dominance of microalgae species with high settling rates, often colonial and/or large species of microalgae. For instance, high flocculation efficiencies are observed when the large chlorophyte Pediastrum is dominant (Park et al. 2011a). By recycling part of the harvested biomass, the dominance of this species in the community can be effectively maintained over long periods of time (Park et al. 2013). High flocculation efficiencies are also observed when filamentous cyanobacteria are prominent in the microalgal community (Su et al. 2011). Spontaneously flocculating microalgae can also be used to harvest other, non-flocculating species of microalgae by mixing bioflocculating and non­flocculating microalgae (Salim et al. 2011, 2012). Some bioflocculating species of microalgae such as the diatom Skeletonema appear to produce infochemicals that are capable of inducing flocculation in other microalgal species (Taylor et al. 2012). Cultures of otherwise non-flocculating species microalgae can be ‘trained’ to induce bioflocculation. For instance, Su et al. (2012) repeatedly removed all microalgae that remained in suspension and kept only the rapidly settling microalgae and after one month obtained a culture that flocculated spontaneously. A major advantage of spontaneous flocculation of microalgae is that no chemicals are added during the process and the harvested biomass is free from contaminants. In order to be able to use bioflocculation as a reliable harvesting method though, more research is needed to better understand the underlying mechanisms that cause the phenomenon.

Bioflocculation may also be caused by other microorganisms in the microalgal culture, such as bacteria or fungi. For example Lee et al. (2012c) showed that bacteria present in cultures of the chlorophyte Chlorella prompted autoflocculation. Bacteria or fungi can be cultured separately and added to a microalgal culture — induced flocculation (Zhou et al. 2012, 2013). Alternatively, bacteria or fungi can be co-cultured with the microalgae, in which case the culture medium should contain a carbon source to support heterotrophic growth of bacteria or fungi (Lee et al. 2008; Gultom and Hu 2013). Bacterial bioflocculation may be a particularly promising flocculation method in wastewater treatment systems, as wastewater often contains a carbon source to sustain bacterial growth. However, an optimal balance between heterotrophic bacterial and autotrophic microalgal production is required for optimal flocculation. This can be controlled by optimizing the ratio of organic carbon over inorganic carbon in the medium (Van Den Hende et al. 2011,

2014) .

Photosynthesis makes use of solar energy to convert CO2 into glucose. This process is vital for life on Earth. Photosynthesis can only use parts of the solar spectrum that are in the photosynthetic active radiation range (PAR) (irradiance between 400 and 700 nm). Based on the measured average solar spectrum at the Earth’s surface, the proportion of total solar energy within PAR is about 48.7 % of the incident solar energy (Zhu et al. 2008). Pigments are responsible for capturing this light. Photo­synthetic organisms contain several pigments. As a matter of fact, pigments are responsible for the names of different divisions and classes of algae. For instance, Cyanophyceae and Rhodophycea contain Chl a and phycobillins, while Haptophy — ceae and Bacilariophycea contain Chl a and c. It is to be noted that all photosynthetic organisms contain Chl a which has the strongest absorption at 430 and 662 nm.

Other accessory pigments have different absorption spectra allowing the organism to more effectively collect different spectra of light (Frigaard et al. 1996). The absorption spectra of some microalgae species are summarized in Fig. 15.2.

Open ponds are the most usual setting for large-scale outdoor microalgae cultiva­tion (Fon Sing et al. 2013; Jeffery and Wright 1999). The major commercial pro­duction of algae is today based on open channels (raceway) which are less expensive, and easier to build and operate compared with closed photobioreactors (Borowitzka 2013b; Tredici and Materassi 1992). In addition, the growth of mic­roalgae meets is less challenging in open than closed cultivation systems; however, just a few species of microalgae (e. g. Chlorella, D. salina, Spirulina. sp., Chlorella sp. and P. carterae) have been successfully grown in open ponds (Moheimani and Borowitzka 2006; Tredici and Materassi 1992). Large-scale outdoor commercial microalgal culture has been methodically developed over the last sixty years (Borowitzka and Moheimani 2013a). Profitable production of microalgae, at present, are limited to a comparatively few small-scale (<10 ha) plants producing high-value health foods, most located in south-east Asia, Australia and the USA (Benemann 1992; Borowitzka and Borowitzka 1990; Richmond 1992). Two major types of large-scale open cultivation systems have been developed and have been used on a commercial basis. These are (a) unstirred ponds and (b) stirred ponds (circular and raceway) (Borowitzka 1993a, b; Borowitzka and Moheimani 2013b). The most common commercial microalgal culture system in use today is the paddlewheel-driven raceway pond (Richmond et al. 1993). The advantages and disadvantages of growing microalgae in open ponds and closed photobioreactors are summarised in Table 1.2. Relatively low cost of construction and operation are the main reasons for culturing algae in open ponds (Tredici and Materassi 1992). However, the high contamination risks and low productivity, induced mainly by poor mixing regime and light penetration, are the main disadvantages of open systems.

Mixotrophic cultures are culture systems where light and organic carbons are used as the energy source, while inorganic and organic carbons are used as the carbon source. Although it is sometimes used interchangeably with photoheterotrophic culture, in a strict sense, photoheterotrophic culture involves the use of light as the energy source, while organic carbon is used as the carbon source. In other words, light is required to metabolize the organic carbon source in photoheterotrophic culture. Photoheterotrophic cultivation requires both organic carbons and light at the same time, whereas in mixotrophic culture, both are present, but either can be used without the other. From practical point of view, both mixotrophic and pho­toheterotrophic cultures can be regarded as culture systems where light, organic carbon, and inorganic carbon are present at the same time.

As already discussed, heterotrophic culture has many advantages over photo­autotrophic cultures. However, there are many metabolites whose syntheses are promoted by light, and thus are not efficiently produced in heterotrophic cultures (Chen and Zhang 1997; Lee and Zhang 1999; Cohen 1999; Sukenik et al. 1991). This disadvantage can be overcome by mixotrophic culture which involves simultaneous use of light and organic carbon sources. Mixotrophic cultures have many advantages over other culture systems. For example, inhibition of photo­synthesis by high dissolved oxygen concentration is a major problem in photoau­totrophic cultures, while oxygen limitation is a major problem in heterotrophic cultures. In mixotrophic culture, dissolved oxygen concentration does not increase to inhibitory levels since it is simultaneously used for heterotrophic metabolism of the organic carbon. On the other hand, organic carbon assimilation is hardly limited by dissolved oxygen concentration since oxygen is constantly produced by pho­tosynthetic activities. Furthermore, heterotrophic growth generates carbon dioxide which is used for photoautotrophic growth (photosynthesis).

In mixotrophic cultures, the presence of an organic substrate means that cell growth is not strictly dependent on photosynthesis, and hence, light is not an indispensable growth factor. Read et al. (1989) and Fernandez Sevilla et al. (2004) have reported that mixotrophic growth requires relatively low light intensities and, consequently, can reduce energy costs. In some strains, it has been found that mixotrophic cultures reduced photoinhibition and that the growth rates are higher than those observed in both photoautotrophic and heterotrophic cultures. Further­more, mixotrophic cultivation reduces biomass loss at night and decreases the amount of organic substances utilized during growth (Chojnacka and Noworyta 2004).

In mixotrophic cultures of many strains of microalgae, there are additive or synergistic effects of photoautotrophic and heterotrophic metabolic activities, leading to increases in productivity. Park et al. (2012) reported higher biomass and fatty acid productivities of 14 species of microalgae in mixotrophic culture over photoautotrophic culture. Bhatnagar et al. (2011) found that the mixotrophic growth of Chlamydomonas globosa, Chlorella minutissima, and Scenedesmus bijuga resulted in 3-10 times more biomass production compared to those obtained under photoautotrophic growth conditions. It has also been shown that the addition of glycerol as the carbon source resulted in increased biomass productivity of Phae — odactylum tricornutum (Ceron Garcia et al. 2005, 2006; Moraisa et al. 2009). One of the possible reasons for better growth in mixotrophic cultures is the stability of pH, since carbon dioxide is simultaneously assimilated and released during pho­tosynthesis and respiration. In photoautotrophic cultures, the pH increased to more than 10, but remained stable around 7 in mixotrophic culture (Kong et al. 2011). It is important to note, however, that biomass productivity in mixotrophic cultures depends on many factors such as the strain, type, and concentration of the carbon source, and other medium components, as well as the light intensity. In some strains, for example, the addition of some carbon sources to photoautotrophic cultures inhibits growth, while others stimulate growth (Heredia-Arroyo et al.

2011) . This is because photosynthesis and oxidative phosphorylation of organic carbon substrates seem to function independently in some algae, and growth rates in mixotrophic cultures are the sum of those in photoautotrophic and heterotrophic cultures. This has been reported for Chlorella sp., Spirulina sp., and Haemato — coccus (Ogawa and Aiba 1981; Marquez et al. 1993; Martinez and Orus 1991; Hata et al. 2001). Under certain culture conditions, the presence of organic carbon in some microalgae depresses photosynthetic O2 evolution and inhibits respiration and enzymes of Calvin cycle (Liu et al. 2009). In mixotrophic cultures, photosynthetic fixation of inorganic carbon is influenced by light intensity, while the heterotrophic assimilation of carbon is influenced by the availability of organic carbon. Thus, the ratio of photoautotrophic growth to heterotrophic growth depends on the light intensity, type, and concentration of organic carbon and carbon dioxide concen­tration (Ogbonna et al. 2002a, b). These factors must be controlled to ensure high rates of growth and lipid accumulation.

Aside from increased biomass concentration and productivities (Lodi et al. 2005), mixotrophic cultures can lead to increases in lipid accumulation over the values obtained in photoautotrophic cultures. This has been reported for several species of microalgae such as Chlorella sp. (glucose), P. tricornutum (glycerol) (Fernandez Sevilla et al. 2004), Nannochloropsis sp. (glycerol) (Wood et al. 1999; Liang et al. 2009), and C. vulgaris (Kong et al. 2011). However, the oil contents of the cells in mixotrophic cultures are dependent on the nature of the carbon source. In some cases, the lipid contents of the cells are even lower or the same as those in the photoautotrophic cultures (Park et al. 2012). Nevertheless, because of the higher growth rate, the lipid productivities are, generally, much higher than those in photoautotrophic cultures. Ratha et al. (2013) reported that lipid production by twenty different strains of cyanobacteria and green algae was highest under mixotrophic condition, compared to heterotrophic and photoautotrophic cultures. With either glucose, starch, or acetate, the maximum lipid productivities of Phaeodactylumtricornutum in mixotrophic cultures were several times higher than those obtained in the corresponding photoautotrophic control cultures (Wang et al.

2012) .

Lipid productivity in mixotrophic culture is also dependent on the strain used. For example, the lipid content and lipid productivity were higher under mixotrophic conditions as compared to both photoautotrophic and heterotrophic cultures in all the members of Chlorococcales tested. Yet, the filamentous alga Ulothrix and all the cyanobacterial strains had slightly higher lipid content and lipid productivity in photoautotrophic cultures (Ratha et al. 2013). The increases in fatty acid produc­tivity under mixotrophic conditions can result from the combined increases in biomass productivity and fatty acid content, or from increased biomass productivity at relatively constant fatty acid content. In some strains and under certain culture conditions, there is no positive effect of mixotrophic culture on cell lipid content; thus, the increase in lipid productivity is mainly due to increases in biomass pro­ductivity, shown for C. vulgaris with various carbon sources (Kong et al. 2011). In contrast to photoautotrophic cultures, where conditions that favor lipid accumula­tion often suppress cell growth (Chisti 2007; Hu et al. 2008), in mixotrophic cultures, there can be a linear relationship between biomass and fatty acid pro­ductivities (Griffiths and Harrison 2009; Park et al. 2012).

Mixotrophic cultivation affects the fatty acid profile of microalgae. In 10 out of 14 isolates grown under mixotrophic condition with acetate as the organic carbon source, the percentage of oleic acid content increased significantly (Park et al.

2012) . However, the fatty acid profile was not affected when glycerol was used (Fernandez Sevilla et al. 2004), indicating that high oleic acid content is not a general feature of fatty acids in mixotrophically grown cells and that the carbon source is likely to be an important determinant of the fatty acid profile.

Other advantages of mixotrophic cultures include the feasibility of using open ponds for large-scale cultivation (Perez-Garcia et al. 2011), and the use of waste­waters as sources of organic carbon and other nutrients for reduced production costs (Zhao et al. 2012). When open ponds or non-sterilized bioreactors are used, the addition of the organic carbon sources must be controlled to avoid contamination by fast-growing heterotrophs. In some cases, the organic carbon substrate is only introduced during daylight hours, or alternatively is added only once toward the end of the culture to avoid bacterial contaminants from accumulating to unacceptable levels (Abeliovich and Weisman 1978; Lee 2001).

The main disadvantages of mixotrophic culture, as with heterotrophic culture, are that the cost of carbon source can be high and an excess/uncontrolled addition of organic substrates in an open system is likely to stimulate growth of invasive heterotrophic bacteria, resulting in a low microalgae biomass yield. There is also the problem of photoinhibition of organic carbon metabolism, in some cases, while maintaining an optimum balance of photoautotrophic to heterotrophic metabolic activities can be challenging.

Initial cell concentration covers the range of 1-6.76 g L-1 which is summarized in Table 7.5. For example, two mutants of Chlorella sp. (MT-7 and MT-15) was investigated at different initial cell concentrations on CO2 fixation rates from 1 to 3 g L 1 (Ong et al. 2010). The CO2 fixation rate increased from 0.0124 to 0.0168 g L-1 d-1 and from 0.0109 to 0.0177 g L-1 d-1 for Chlorella sp. MT-7 and MT-15, respectively (Table 7.5), indicating a significantly higher CO2 fixation rate at higher biomass concentrations (Ong et al. 2010). Furthermore, Table 7.5 shows the effect of initial cell concentration on CO2 removal and biomass concentration of cyanobacterium Synechococcus sp. (Takano et al. 1992). When the initial micro­algal cell concentration increased from 1.4 to 6.8 g L-1, the CO2 fixation rate increased from 1.06 to 2.22 g L-1 and biomass concentration increased from 1.92 to 7.76 g L-1 (Takano et al. 1992). Additionally, CO2 retention times will be increased in higher biomass concentration as a result of higher culture medium viscosity, resulting in an enhanced CO2 removal rate and fixation efficiency (Ong et al. 2010).

Genome-scale metabolic networks are often computationally explored to charac­terize functional relationships between their reactions. For instance, identification of correlated reaction sets (cosets), i. e., reactions that are always “on” or “off” con­currently (Papin et al. 2004) can define functional relationships between reactions that are not necessarily in the same pathway or obvious. The significance of these reactions is evident from the observation that mutations in correlated reactions may lead to a manifestation of the same aberrant (or disease) phenotype (Jamshidi and Palsson 2006). Because, the solution space of genome-scale networks can be

enormous, uniform sampling of the space is often carried out using the Monte Carlo method to identify the cosets and the overall shape and size of the steady state flux space (Becker et al. 2007). This sampling method, which is implemented in COBRA, identifies a set of randomly distributed solutions to serve as a proxy for the entire space. In Monte Carlo sampling, points are picked randomly from the space and the fraction inside the defined constraints is counted. This sampling method allows a uniform exploration of the metabolic network space while reducing the computational power demand required for the analysis.

F. C. Thomas Allnutt and Ben A. Kessler

Abstract Harvesting of dilute cultures of algae from large volumes of culture needed for production of biofuels and bioproducts is a substantial hurdle to the economic viability of algal biofuels. While centrifugation and sedimentation are already scaled to volumes that would allow direct application to algal biofuel production, their economics to the production of biofuel are not favorable. The industry has reevaluated the existing technologies and continues to innovate around the harvesting of microalgae for biofuels and bioproducts. This review discusses the historical approaches and recent advances while comparing and contrasting the different methods. An engineering estimate of comparative costs is also provided.

14.1 Introduction

A major challenge facing the microalgae industry is how to economically harvest microalgal biomass from millions of gallons of culture medium containing biomass at densities of less than 1 % total solids. Typically, open pond photoautotrophic production reaches biomass densities ranging from 0.01 to 1 gdw L 1, while cell densities in enclosed photobioreactors (PBRs) range from 4 to 10 gdw L-1 (Chisti 2007; Stephens et al. 2010). The higher levels of biomass reported for PBRs that provided high-intensity artificial light are still low relative to cell densities that can be reached in heterotrophic cultures (which can exceed 100 gdw L 🙂 and also introduce a different hurdle to commercial viability, the operating expense (OpEx) of the electricity for the lights, and availability of inexpensive, durable, and highly efficient lights (Chen et al. 2011). The contribution that harvesting the algal biomass makes toward the overall cost for renewable biofuel production has been estimated

F. C.T. Allnutt (H)

BrioBiotech LLC, P. O. Box 26, Glenelg, MD 21737, USA e-mail: fct. allnutt@gmail. com

B. A. Kessler

Phycal Inc, 51 Alpha Park, Highland Heights, OH 44143, USA © 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_14 to be between 20 and 30 % and remains a bottleneck for the industry (Brennan and Owende 2010; Dismukes et al. 2008; Gudin and Thepenier 1986).

Microalgae have been harvested by centrifugation in industrial algal applications that, to this point, have mostly focused on higher value nutritional products, such as P-carotene, astaxanthin, edible algal biomass, or higher value nutritional oils— where the high energy costs of centrifugation can be borne by a high-value product (Spolaore et al. 2006; Wijffels et al. 2013). In the case of wastewater treatment applications, where the purpose is to reduce the biological oxygen demand (BOD), algae have mostly been harvested using flocculation and either settling or flotation tanks in order to lower the overall cost. While such methods have already been operated at scales relevant to commercial biofuel production, they do not provide for preservation of the biomass for downstream processing into fuels and co­products.

It is important to note that the performance of the algal mass culturing system and the properties and value of the product portfolio being produced have direct impacts on how much the harvesting or dewatering step can cost. The reasons for this are the following: (1) In most cases, the size of the equipment is based on volumetric throughput, not dry mass throughput, and (2) in many cases, the operating costs (OpEx) are also based on volumetric throughput. Consequently, culturing systems that produce higher densities of algae and higher densities of product per unit volume of culture medium will require smaller equipment and have lower OpEx.

Additionally, the microalgal feedstock being produced cannot be degraded during the harvesting such that it cannot be used for the production of algal biofuels and any co-products necessary for commercial viability. The industry has, to this point, focused on the production of biomass that is directly converted to energy by a number of different processes (e. g., combustion, hydrothermal liquefaction, and catalytic gasification), secretion into the medium (e. g., ethanol), or biomass that is high in lipid as a feedstock for production of biofuels. But it should be noted that many of the value propositions being put forward for commercial production rely on value-added co-products (e. g., animal feeds, single-celled protein) to meet profit targets. Because of the different requirements for the growth systems, productivi­ties, and extraction criteria, it can be difficult to directly compare harvesting tech­nologies and their associated economics. This chapter will present qualitative considerations for the various harvesting options as well as present ranges for the associated economics.

In 1965, a study completed by Golueke and Oswald compared all of the available harvesting methods and concluded that only centrifugation and chemical flocculation were economically feasible (Golueke and Oswald 1965). However, both of these technologies have characteristics that make them less than ideal for algal biofuel production, and their commercial relevance has been reexamined. Centrifugation is energy intensive and requires expensive equipment to carry out at scale. The addition of flocculants adds OpEx in the form of chemicals and addi­tional bulk of harvested material, and the flocculant itself can negatively impact the final product and valuable co-products (e. g., residual metals). In order to reduce the cost impact of harvesting on production of biofuels, a number of technologies or modifications of old technologies have been evaluated and new technologies have been and are currently being evaluated and developed to reduce harvesting as a hurdle to the commercial viability of algal biofuels. This review will briefly provide an historical backdrop on algal harvesting technologies, describe the existing technologies, compare and contrast the developed technologies, and provide a description of new technologies that have begun the process of crossing into scaled use by the industry.

16.3.1 Low Concentration of Biomass

One of the major factors hampering anaerobic digestion of microalgae biomass is the low concentration of algal biomass under culture conditions. Most outdoor microalgae cultures are very dilute and may only contain 1 g L-1 of solid biomass (Golueke et al. 1957; Stephans et al. 2010); too dilute for anaerobic digesters, potentially causing bacterial washout due to biomass and excessive water addition to achieve the required VS solid loading rates (De Schamphelaire and Verstraete 2009; Golueke et al. 1957; Parkin and Owen 1986). This problem is overcome by harvesting, concentrating and dewatering microalgae cultures to concentrate the biomass. This is a relatively expensive and time consuming production requirement for most microalgae biofuel production methods (Harun et al. 2010; Pragya et al. 2013; Stephans et al. 2013; Ward et al. 2014). However, when anaerobic digestion is integrated with a microalgae biofuel production system, the energy requirement for harvesting and dewatering processes is offset by resultant methane production from the anaerobic digestion process (Sialve et al. 2009). Many different methods are used for the harvesting, concentration and dewatering microalgae, including settling of biomass (Collet et al. 2010), use of chemical flocculants (Golueke et al. 1964; Kalyuzhnyi et al. 1998), centrifugation (Benemann et al. 1977), alum (Golueke and Oswald 1963) and electro-flocculation (Pragya et al. 2013). The above listed harvesting methods have been shown to be nontoxic to anaerobic digestion and may improve digester performance due to better retention of solids (Callander and Barford 1983; Campos et al. 2008; Golueke and Oswald 1963; Kalyuzhnyi et al. 1998; Krishnan et al. 2006). However, with the high cost asso­ciated with some of these harvest methods, many new laboratory and pilot scale methods are under development, the resulting impact to anaerobic digestion is yet to be assessed for toxicity effects causing inhibition of anaerobic digestion processes (Ward et al. 2014).