To achieve the highest productivity, it is necessary to add a source of nu­trients to produce an effective medium. The fertiliser requirements can be calculated using the stoichiometric requirements of the algae. In their LCA, Clarens et al. [87] used to triangular distributions to calculate mini­mum, maximum and most likely dosing rates for nitrogen and phospho­rous. These dosing rates were found to be just under two times the stoi­chiometric requirement providing a surplus but with the excess allowing other reactions to remove the surplus. The fertilisers used were assumed to have been sourced from urea and superphosphate. Stephenson et al. [96] estimated a nitrogen requirements of 59 kg per ton of biodiesel produced which contrasts significantly with the 6 kg per ton estimated by Lardon et al. [77]. Collet et al. [86] in their LCA study assumed a nitrogen dosing of 221 kg per day which equates to 5.74 kg per ton of biomass. The study assumed the biomass if processed to biogas so to compare the studies it is possible to calculate the nitrogen requirement for the energy produced. In which case the requirements in the study by Stephenson et al. [96] is 1.56 kg N/GJ, for the study by Lardon et al. [77] it is 0.16 kg N/GJ and for the study by Collet et al. [86] the value is

0. 65 kg/GJ (assuming the energy content of biodiesel and biogas is 37.8 MJ/kg and 0.036 MJ/L respectively). Clearly each study makes different assumption and thus the fertiliser requirement estimates vary, it’s most likely that higher dosing is required to provide an abundance of nutrients thus avoiding nutrient limitation.

The major drawback to the use of fertilisers is their energy input, cost and environmental impact. Lardon et al. [77] found fertilisers to be one of the major contributors to energy consumption and to the negative en­ergy balance of the whole process system. When the low nitrogen scenario was considered, the energy consumption was far lower. Clarens et al. [87] came to similar conclusions as nutrient-derived energy consumption ac­counted for the greatest energy use in algae production. Similar observa­tions were made by Shirvani et al. [98]. Not only do fertilisers negatively affect the energy balance, but they also provide a significant source of environmental impacts to the system. Fertiliser production requires a high energy input from both electricity and fossil fuels, both of which are high emitters of greenhouse gases. Some studies have ignored the impact of fertilisers however results in the study by Clarens et al. [87] suggested that using alternative sources of nutrients (wastewater) could in fact uptake CO2 and return a positive energy balance in the best case (using source — separated urine as a nutrient source).

The production pathway and experimental methods used in this analysis has been described in detail in previous publications [6,14,19,20]. Further­more, the materials and energy consumption data used in the Experimen­tal Case and the Highly Productive Case are taken from Beal et al. that calculated the second-order energy return on (energy) investment (2nd O EROI) [14]. The term “second-order” refers to the inclusion of direct energy inputs (e. g., electricity consumed for pumping) and indirect en­ergy inputs for consumed materials (e. g., the energy embedded in nitrogen fertilizer that is consumed). Details regarding data collection and uncer­tainty analysis in the Experimental Case and modeling calculations in the

FIGURE 1: The production pathway is represented as three phases: growth, processing, and refining [6]. The data used for each input and output are shown for the Experimental Case (top) and Highly Productive Case (bottom). For the Experimental Case, the material and energy inputs crossing the system boundary were measured for five relatively large scale batches (970-2000 L, each), grown and processed at The University of Texas at Austin, except for the refining inputs (which were modeled from literature data, and are noted in the figure with an asterisk (*) [14]. The Highly Productive Case is an analytical model of a similar production pathway operated more efficiently.

Highly Productive Case can be found in the previous publication [14] and at greater length in a publically available doctoral dissertation (cf. Chapter 4, Appendix 4A, and Appendix 4B of [19]). The work presented herein expands those datasets to incorporate the new analyses mentioned above (water use, nutrients use, FROI, etc.).

Figure 1 shows that the biocrude production process for both of our an­alyzed cases consists of algal cultivation, harvesting (i. e., concentration) with centrifugation or chemical flocculation, cell lysing via electrome­chanical pulsing, and neutral lipid recovery using a microporous hollow — fiber membrane contactor. The Experimental Case is comprised of growth and processing data from five relatively large batches (970-2000 L each), with a total processed volume of about 7600 L. The energy and material inputs that were measured are shown in Figure 1 and the energy outputs are modeled to include bio-oil and biomass fuel (methane) (refining was not conducted during the experiments). The Highly Productive Case mod­els energy-efficient growth and processing methods with higher biomass and lipid productivities.

Electromagnetic bioeffects from relatively weak signals are often due to a time-varying electric field, induced by a time-varying magnetic field [1]. However, the ability of weak oscillating EMF fields to interact with liv­ing cells has been a source of controversy since thermal and other noise poses restrictions to the detection of weak signals by a cell. Activation of signal pathways by external stimuli connects the physical interactions of the applied EMF to the biological response [70]. In nonlinear systems such as biological sensory apparatus, presence of noise can actually en­hance the detection of weak signals, called stochastic resonance [84]. Very small changes in the underlying non-linear kinetics caused by very weak coherent signals and noise can lead to strong, but reversible alterations in the internal nonlinear processes and associated biological function such as ELF influences on G-protein activation dynamic, magnetic field influence on radical pair recombination reactions and weak signal amplification by stochastic resonance incorporated within the Ca2+ signal pathway models [70]. The mechanism of stochastic resonance has shown an amplification factor that may exceed a factor of 1,000. This is because in a nonlinear system, the reaction to an external signal may be much greater when act­ing as a whole than the response of the system’s individual elements. This resonance manifests itself by the appearance of sharp peaks in the power spectrum of the system at the driving frequency and in some of the higher harmonics. Currently, the cell membrane is considered the most likely cel­lular site for interactions with EMF’s and the possible role of ionic chan­nels of the membrane in the amplification process. The potential well-like structure of an ionic channel makes it the ideal system for stochastic reso­nance amplification [85].

SAMPLING AND ISOLATION OF PURE CULTURES

Microalgae grow in most of the natural environments including water, rocks and soil, but interestingly also grow on and in other organisms. Their main habitats are freshwater, brackish and marine ecosystems. Microalgae can be found and collected not only in general aquatic ecosystems such as lakes, rivers and the oceans, but also in extreme environments such as vol­canic waters and salt waters. Local microalgae species should be collected because it can be expected that they have a competitive advantage under the local geographical, climatic and ecological conditions. Our experience has shown that water and sediment samples from aquatic environments that undergo fluctuating and/or occasional adverse conditions provide a higher chance of isolating high lipid accumulating microalgae. Most likely these conditions would favor robust and opportunistic (fast-growing) al­gae with superior survival skills (e. g., by accumulation of storage lipids). Examples of these environments are tidal rock pools, estuaries and rivers.

Isolation is a necessary process to obtain pure cultures and presents the first step towards the selection of microalgae strains with potential for biodiesel production. Traditional isolation techniques include the use of a micropipette for isolation under a microscope or cell dilution followed by cultivation in liquid media or agar plates. Single cell isolation, based on traditional methods from the original sample is time-consuming and requires sterilized cultivation media and equipment, but the result of this elaborate process is always a pure culture that is usually easily identifi­able. Another approach in the laboratory includes the enrichment of some microalgae strains by adding nutrients for algal growth. The most impor­tant nutrient sources for algal growth are nitrogen and phosphate. Some particular algae species may require trace minerals for their growth (e. g., silicon for diatoms). Soil water extract is an excellent source of nutrients for algae growth at this stage because this medium is easy to produce and satisfies nutrient intake of many algae strains. Although automatic isola­tion techniques have offered some advantages towards traditional methods (see below), single cell isolation by a micropipette (e. g., a glass capillary) is still a very effective method that can be used for a wide range of samples and is very cost-effective. An automated single cell isolation method that has been developed and widely used for cell sorting is flow cytometry [17]. This technique has been successfully used for microalgae cell sort­ing from water with many different algae strains [18], primarily based on properties of chlorophyll autofluorescence (CAF) and green autofluores­cence (GAF) to distinguish algae species such as diatoms, dinoflagellates or prokaryotic phytoplankton.

Unlike for many agricultural crops, a targeted selection and domes­tication of microalgae strains is still in its infancy, while technology to economically grow microalgae with high lipid content is still be­ing developed [4]. Each microalgae strain requires careful selection and optimization in order to increase lipid productivity with the aim to provide a cropping system with improved biofuel production and per­formance properties [19]. Each micro-environment may provide algae strains with very different properties. As opposed to only manipulating a few individual cultured algae strains in the laboratory for high lipid productivity, a more efficient and useful strategy to identify oleagi­nous microalgae would be an in-depth and systematic investigation of a whole taxonomic group of microalgae over a wide geographical and ecological distribution [20]. By correlating this with algal oil contents and optimal environmental growth conditions, a predictive tool for se­lecting optimal microalgae strains for biofuel production maybe devel­oped. A bioinformatics approach could assist with the discovery of new algae isolates capable of biodiesel production and their phylogenetic grouping may suggest that potentially many more species have this ability. Typically, the steps involved in obtaining data for phylogenetic analysis include primers design (Table 1), DNA and/or RNA extrac­tion, PCR amplification, denaturizing gradient gel electrophoresis and/ or sequencing.

In 2010, seven microalgae genomes had been completed [27] and cur­rent efforts to obtain many other microalgal genome sequences will en­hance gene-based biofuel feedstock optimization studies (e. g., by meta­bolic engineering). The accumulation of storage lipid precursors and the discovery of genes associated with their biosynthesis and metabolism is a promising topic for investigation. For example, genes encoding key en­zymes involved in biosynthesis and catabolism of fatty acids/TAG and their regulation are currently not well understood (for a review of the lipid biosynthesis pathway in microalgae see Schuhmann et al. [28]). By pro­viding insight into the mechanisms underpinning the relevant metabolic processes, efforts can be made to identify molecular markers for selection or to genetically manipulate microalgae strains to enhance the production of feedstock for commercial microalgal biofuels. To date, genetic engi­neering approaches have been successfully used to improve biofuel phe­notypes only in Chlamydomonas reinhardtii, Nannochloropsis gaditana and Phaeodactylum tricornutum [29].

Although the potential of microalgae lipid production is tremendous, no commercial development of microalgae based fuels has been achieved so far because of lack of price competitiveness versus petroleum diesel. The bottleneck that limits the development of microalgae biomass energy is that we do not have technologies to produce large quantities of low-cost and high lipid content microalgae biomass. In order to achieve large-scale and low-cost cultivation of microalgae, the following bottleneck issues should be addressed.

1.6.1 DEVELOPMENT OF HIGH PERFORMANCE MICROALGAE STRAINS

Breeding of high-quality microalgae with characteristics of high lipid pro­ductivity and strong adaptability is the key to realize both of high lipid content and great biomass. Excellent microalgae strains can be obtained by screening of a wide range of naturally available isolates, and the ef­ficiency of those can be improved by selection and transformation. Con­sidering the good results of mixotrophic cultivation, we should do some domestication making microalgae adapt to different organic matters such as sucrose, glycerol, xylan, organic acids in slurry. This will hopefully greatly reduce the cost of cultivation.

Much of the research reported so far is based on laboratory work and spec­ulation while testing a full system would allow realistic life-cycle assess­ments to be carried out investigating similar systems in a number of indus­trial scenarios. Data related to energy consumption and yield would prove the viability of the concept. Setting up pilot scale infrastructure within most industries with suitable wastewater would be a simple undertaking with great research benefits. The ponds would need to be inoculated with a mix of local strains and the dominance of those strains monitored. Suitable harvesting techniques would need to be tested for the algal mix, cultivated within the ponds to identify the most effective and sustainable method for each case. Further research is required to optimize energy recovery from conversion techniques which provide the maximum energy yield, most likely anaerobic digestion or combustion.

2.4 CONCLUSIONS

The review of the current state of knowledge and technology suggests that it is unlikely that there is one solution to biofuel recovery from algal bio­mass. Production of energy from algae is most likely to be successful on a case by case basis based on applicability to the particular industry and the site under question. The majority of wastewaters from common industries have shown capacity to support the cultivation of various strains of algae. Allowing natural domination of algal strains means that algae which are most effective for that particular situation should develop. If a preferred algal strain is required, the pond could be seeded with the algae and re­cycled continuously to promote growth.

The biomass processing stages can use existing technologies which are tested for many strains of algae. Harvesting can be optimised for each individual scenario. Optimal recovery of energy for maximum efficiency is likely to be similar for each industry. Literature suggests that recovery through anaerobic digestion or combustion provides the highest energy return for mixed strains. As the strains are likely to be mixed and varying there is little point in designing systems for specialised biofuels (biodiesel or bioethanol) which require specific biomass characteristics. Therefore a system which is flexible for numerous industries is possible. Pilot scale tests of such systems will be essential for implementation to optimise sys­tems individually.

While re-use of CO2 would be desirable and some water requirements could be met with wastewater or saline water, the increased demand for fertilizer and electricity could have negative economic impacts. Depend­ing on the scale of production, this electricity input requirement could impact electricity prices and yield a significant, unintended increase in carbon emissions.

5.3.6.1 CARBON DIOXIDE

Under ideal conditions, algae require roughly 2 kg of CO2 for each kg of algal biomass produced [2,15,17]. However, in the experiments, most of the CO2 delivered to the growth volumes was not retained in biomass (and released as outgas). As a result, 9.35 g of CO2 were consumed per liter of pond water processed, which only contained 0.26 g of algae, on average. Based on this consumption, 3.7 Mg of CO2 were consumed per L of bio-oil. For the Highly Productive Case (with 1 kg algal biomass/kL of processed volume, 8 kg of CO2 per kL of processed volume, and 0.26 L of bio-oil per kL of processed volume) 31 kg of CO2 would be required for each L of bio-oil produced. For 19 GL/yr of bio-oil (5 Bgal/yr), this equates to 5.8 x 1011 kg of CO2 consumed per year, which is ~11% of the total CO2 emissions from the U. S. [42].

Lactic acid (2-hydroxypropanoic acid) is the most widely occurring car­boxylic acid in nature. Annual production reaches 120,000 tons per year, 90% of which is produced by bacterial fermentation of biomass sugars [25], including pentoses [26]. The bacterial route affords lactic acid in high yields (e. g. 90%) although it possesses important drawbacks namely low reaction rates and troublesome separation/purification from the re­action broth of the lactic acid product. Lactic acid is obtained in form of calcium salt, and subsequent neutralization generates large amounts of residual CaSO4 (1 kg per kg of lactic acid) which raises production costs and produces a waste disposal problem. Recently, a non-biological route for the conversion of aqueous sugars into lactic acid based on eas­ily separable and recyclable solid zeolites has been developed [27]. This promising technology opens the possibility of producing lactic acid from biomass sugars at more competitive prices in near future thereby consider­ably increasing the platform potential of lactic acid. The classical market for lactic acid involves food and food-related applications; however, the development and commercialization of new applications in the field of polymers and chemicals has caused steady expansion of the lactic acid market since the early 1990s [28].

Lactic acid possesses a rich chemistry based on its two functional groups (e. g. single bondOH and single bondCOOH). Thus, a variety of transformations to useful compounds such as acetaldehyde [29] (via de — carbonylation/decarboxylation), acrylic acid [30] (via dehydration), pro­panoic acid [31] (via reduction), 2,3-pentanedione [32] (via condensation) and polylactic acid (PLA) [33] (via self-esterification to dilactide and subsequent polymerization) has been described. All these transformations convert lactic acid in an attractive feedstock for the renewable chemicals industry [34].

Lactic acid, with its two adjacent functional groups concentrated in a small molecule of three carbon atoms, can be considered as a proto­type of an over-functionalized biomass-derived molecule. This chemical structure determines its high reactivity as well as its natural tendency to decompose with temperature [35]. As indicated in the Introduction, an ef­fective approach for the conversion of biomass derivatives into advanced biofuels involves a requisite oxygen removal step that helps to reduce re­activity leaving molecule more amenable for subsequent Csingle bondC coupling upgrading processes. Following this approach, lactic acid can be converted into hydrophobic C4-C7 alcohols suitable as high energy density gasoline-compatible liquid fuels for the transportation sector (Fig. 2) [31]. In this scheme, lactic acid is first deoxygenated to generate two reactive intermediates, namely propanoic acid and acetaldehyde, by means of dehydration-hydrogenation and decarbonylation/decarboxyl — ation processes, respectively. These intermediates were detected at low lactic acid conversions indicating that they are primary products in the synthesis [36]. Importantly, these intermediates are less reactive than lactic acid but still preserve oxygen functionality for subsequent Csingle bondC coupling upgrading. Thus, acetaldehyde can undergo self-coupling by aldol-condensation to generate butanal (after hydrogenation of the cor­responding C4 unsaturated aldol-adduct), whereas propanoic acid is self­coupled into 3-pentanone via ketonization. As shown in Fig. 2, successive aldol condensations between acetaldehyde (which is present in the reactor in high amounts) and butanal and 3-pentanone products generate C6 and C7 ketones. There are several important aspects of this process: (i) lactic acid is processed solved in water which the classical medium in which this molecule is obtained after microbial fermentation of sugars; (ii) the num­ber of reactions leading from lactic acid to C4-C7 carbonyl compounds can be carried out in a single reactor by employing a multifunctional and water-stable Pt/Nb2O5 catalyst in which niobic support plays a crucial role catalyzing dehydration, decarboxylation/decarbonylation and Csingle bondC coupling reactions; and (iii) C4-C7 carbonyl compounds (precur­sors of the corresponding alcohols by simple hydrogenation) are stored in a spontaneously separating from water organic layer accounting for 50% of the carbon in the lactic acid feed.

2.2.5.1 HARVESTING

As the biomass cannot be utilised efficiently at low concentrations in me­dia, the first step in the biomass processing stage is harvesting the algae for subsequent processing. The method of harvesting used depends very much upon the type of algae which is under cultivation. Microalgae re­quire more intensive harvesting methods in comparison to macroalgae, because of their cell size. Depending upon circumstances, often a series of harvesting methods is required to produce a final biomass below a desired moisture content. Common methods of harvesting of algae are: microfil­tration, flocculation, sedimentation, flotation and centrifugation [58].

One of the most effective methods of harvesting is filtration using mi­cro-filters. This method of filtration generally uses a rotary drum covered with a filter to capture the biomass as the influent passes through from the centre outwards [59]. Initial harvesting tests in the 1960s tested micro­filters but found that the majority of algal cells simply passed through most of the filter types [60]. It was later suggested that micro-filtration was suitable for strains of algae with a cell size greater than around 70 pm and was not suitable for those species with cell sizes lower than 30 pm

[61] . The size of the opening in the filter mesh dictates what percentage of biomass is captured likewise with the size of the biomass cells. The pore size also affects how much pressure is required to facilitate the flow of water through the filter which will in turn affect the energy consumption

[62] . The concentration of the algae in suspension also influences the ef­ficiency of removal as highly concentrated biomass will foul the filter very quickly causing reduced performance and a requirement for backwashing and thus further energy consumption. If filtration is to be used it is essen­tial that the method suits the species of algae which is being harvested, otherwise the filtration will be ineffective and provide low yields of bio­mass. If the cultivated algal species allows for filtration (e. g., Spirulina, Spirogyra, Coelastrum), the filtration method can prove very efficient and cost effective method of harvesting. Mohn [63] for example found that gravity filtration using a microstainer and vibrating screen both provided good initial harvesting of Coelastrum up to a total suspended solid of 6% with low energy consumption (0.4 kWh/m3). Mohn [63] also investigated pressure filtration of Coelastrum which provided even higher total solids of concentrate up to 27%, although requiring more than twice the energy. Clearly inexpensive and low energy harvesting of biomass is possible with filtration, providing the dominant algae being harvested is of a suitable cell size and optimal concentration level.

Sedimentation and flotation have also been proven as viable options for harvesting algal biomass with no requirement for specific cell size. Both sedimentation and flotation rely on biomass density to facilitate the process, both processes are aided by flocculation and flotation is aided additionally with bubbling. As a method of biomass removal, sedimentation was consid­ered a viable process in the 1960s due to its prominence in wastewater treat­ment and its low energy requirement [60]. Due to the low specific gravity of algae, the settlement process is, however, slow but, under certain conditions, the self-flocculation of some strains of algae is possible. Nutrient and carbon limitation and pH adjustment appear to be methods of auto-flocculation of algae which may provide a low-cost solution to the initial harvesting process [64]. Recent studies have focussed upon bio-flocculation which occurs as a result of using several bacteria or algal strains to flocculate with the desired algal biomass to allow settlement. Gutzeit et al. [65] found that gravity sedi­mentation was possible using bacterial-algae flocs developed in wastewater for the removal of nutrients, and reported that the flocs of Chlorella vulgaris were stable and settled quickly. Other approaches investigated the combined use of autoflocculating microalgae (A. falcatus, Scenedesmus obliquus and T. suecica) to allow for flocculation of non-flocculating oil-accumulating algae (Chlorella vulgaris and Neochloris oleoabundans) [66], which re­sulted in a faster sedimentation as well as a higher percentage of biomass harvested. This method of harvesting appears viable due to its low energetic inputs but also because it does not rely on chemicals, thus allowing the water to be discharged or recycled without further treatment. However, it should be noted that this method of flocculation may not be suitable for all types of algae, and thus further research is required in this area.

Conventional methods of flocculation using flocculants common to wastewater treatment such as alum, ferric chloride, ferric sulphide, chi — tosan among other commercial products are likely to provide a more consistent and effective solution to flocculation. Much research has been conducted upon the removal of algae using flocculants with vary­ing degrees of success (Table 4). For example, a complete removal of freshwater microalgae, Chlorella and Scenedesmus, using 10 mg/L of polyelectrolytes while 95% removal using 3 mg/L of polyelectro — lites has been reported [60]. A comparative study where alum and fer­ric chloride were use as flocculants for three species of algal biomass (Chlorella vulgaris, I. galbana and C. stigmatophora) indicated the low dosages of alum (25 mg/L) and ferric chloride (11 mg/L) were sufficient for optimal removal of Chlorella vulgaris, while higher dos­ages of alum and ferric chloride were required for the removal of ma­rine cultures I. galbana (225 mg/L alum; 120 mg/L ferric chloride) and C. stigmatophora (140 mg/L alum; 55 mg/L ferric chloride) [67]. Additionally it has been reported that the combined use of chitosan at low concentrations (2.5 mg/L) and ferric chloride provided much quicker flocculation of the algal cells, Chlorella vulgaris, I. galbana and C. stigmatophora, and reduced the requirement of ferric chloride [43]. The use of chitosan as a flocculant for the removal of freshwater algae (Spirulina, Oscillatoria and Chlorella) and brackish algae (Syn — echocystis) has been investigated [43], and chitosan has been found to be a very effective flocculant, at maximum concentrations of 15 mg/L removing about 90% of algal biomass at pH 7.0. The use of conven­tional and polymeric flocculants for the removal of algal biomass in piggery wastewater has been recently investigated [43]: ferric chloride and ferric sulphate were found to be effective flocculants at high doses (150-250 mg/L) providing removal rates greater than 90%; polymeric flocculants required less dosing (5-50 mg/L), although provided lower biomass recoveries; chitosan performed poorly at both low and high dosages for each of the algal species types with a maximum removal of 58% at a dose of 25 mg/L for a consortium of Chlorella.

We performed a literature study on residues and waste potentials in 2050 for five main categories:

• Oil and fat residues and waste

• Forestry residues and wood waste

• Agricultural residues

• Wet waste and residues

• Dry waste

In our study we analysed the potential for the most important sub­categories of each main category. After obtaining the literature values for the potential of each residue and waste (sub)category, we performed three additional analyses to arrive at the final residue and waste potential figures:

1. We adapted literature projections for 2050 for manure and waste animal fat potential to reflect the meat consumption level described above.

2. We altered the dry waste potential from municipal solid waste (MSW) to reflect the fact that not all MSW is renewable and that some MSW is wet and some is dry.

3. We updated the recoverable fraction, the share of residues and waste that is available for bioenergy production, for some of the categories because they were inconsistent with other developments such as im­proved future economic feasibility of residue collection or our study’s framework of closed nutrient loops. The recoverable fractions used always take into account other uses, e. g. the use of wood residues for production of fibre board, and sustainability considerations.

The results of our analysis are displayed in Table 4. Table 4 also in­cludes data on the used recoverable fractions per subcategory. In some cases, the recoverable fraction was implicitly included in the analysis of the literature sources we consulted; in those cases the value cannot be reported. It should be noted that the reported recoverable fractions do not include the residue-to-product ratio, which is a measure of how much resi­due is produced per quantity of main product. This ratio was implicitly included in the analyses of the consulted literature sources.