Scaling up photoautotrophic production of eicosapentaenoic acid (EPA) using microalgae

Abstract

Food insecurity is one of the major challenges that the world faces today. Improving food security is about not only ending hunger but also ensuring easy, affordable access to nutritious, healthy food for all people. The importance of adequate consumption of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) is well recognized. EPA and DHA are essential omega-3 fatty acids that must obtained from the diet. Producing sufficient amounts of EPA and DHA to meet the global demand from their conventional sources – fish – is becoming increasingly challenging due to population growth and overfishing. To bridge the supply gap of EPA and DHA via sustainable means, the idea of sourcing these nutrients directly from their natural primary producers – microalgae – has emerged. The overarching aim of this thesis was to develop a scalable bioprocess that used photoautotrophic microalgae to produce EPA. Such a strategy would offer the advantages of not requiring arable land or potable water, thus would avoid competition with existing food production techniques for these resources. Currently, commercial production of microalgal EPA is hindered by the process economics. One of the major hurdles is the difficulties of growing EPA-producing microalgae efficiently at large scales, hence the development of scalable, cost-effective production processes is needed. In this research, three broad avenues were taken: (1) identification of suitable strains, (2) optimization of the process operation and (3) optimization of the bioreactor design. Microorganisms used for industrial applications need to have certain characteristics. A key metric is high product accumulation under realistic settings, others include robust growth and consistent performance in industrial systems as well as tolerance to changes in the growth conditions. Systematic evaluations using these industrial metrics have not been done for screening EPA-producing microalgae. To address this gap, a screening procedure was conducted to quantify the EPA productivity for a range of microalgae in both 300 mL flask and 5 L flat-panel photobioreactor cultures. It was found that results from the flask screening offered poor predictions of performance in photobioreactors, suggesting a need for improved screening tools, such as scale-down simulators, for selecting industrial microalgal species. In the photobioreactor cultures Phaeodactylum tricornutum displayed the highest EPA productivity, which was approximately an order of magnitude higher than those of the other species tested. The results demonstrated the potential of P. tricornutum for large scale production of EPA. To further evaluate the suitability of P. tricornutum for scaled-up production, the effects of two key environmental factors, temperature and salinity, on the EPA production of P. tricornutum were examined in photobioreactors. It was found that P. tricornutum could tolerate relatively wide ranges of temperature (13-27 °C) and salinity (35-50 g L-1); within these ranges its EPA content was approximately constant at 5% of the dry biomass weight. These results illustrate the robustness of P. tricornutum which is an obvious advantage from the perspective of industrial production. A comprehensive nutritional analysis was also performed for the biomass of P. tricornutum. In addition to being a good source of EPA, P. tricornutum biomass was also rich in protein (45% of dry weight), iron and vitamin B12. The nutritional information here could serve as a starting point for the formulation of P. tricornutum into food products. With P. tricornutum being identified as a suitable species, improving the EPA productivity was the next step in the process development. The use of reliable computational models could greatly facilitate the identification of optimal operating strategies, with fewer laborious, time-consuming experiments required. This research demonstrated the development of a model for the EPA production of P. tricornutum, which was the first to predict the biomass and EPA productivities along with the EPA concentrations in the biomass and fatty acid fraction for microalgae. The model was built on a kinetic modelling framework where the system behaviour was simulated using a set of ordinary differential equations, with the integrated effect of light and nitrogen availability being accounted for. This mathematically simple model was able to produce satisfactory predictions for different reactor geometries and scales (5 L flat-panel and 50 L cylindrical reactors), light path lengths (5 cm and 19 cm) as well as operating modes (batch and repeated-batch), with the model-data mismatches being typically less than 20%. Using this model, an optimized repeated-batch strategy was developed. Implementation of this strategy over four repeated cycles led to 50% and 20% increases in the biomass and EPA productivities, respectively. The results demonstrated the usefulness of this model as a tool in the scale-up, design and optimization of microalgal EPA production. Another avenue for improving process productivity is developing better reactor designs. A major challenge in improving the performance of photoautotrophic cultures is delivering light into high-cell-density cultures. Using Computational Fluid Dynamics (CFD) with Lagrangian particle tracking, this work examined means of improving the frequency at which the microalgal cells were transported between the light and dark zones, something that was recognized as a way for improving light delivery but has not been sufficiently evaluated under industrially relevant conditions. Different superficial gas velocities (0.6-6.0 cm s-1), reactor diameters (0.14-0.29 m), internal structure designs (split-cylinder airlift, segmental baffles, disc-and-doughnut baffles) and sparger configurations (asymmetrical and oscillatory spargers) were investigated for their effect on the hydrodynamics in 50 L bubble column photobioreactors. It was found that the frequency increased linearly with the superficial gas velocities but did not vary appreciably with the reactor diameter within the tested range. The frequency could be increased by 160% and 50% by the installation of segmental baffles and disc-and-doughnut baffles, respectively; In comparison, using the alternative sparger configurations had smaller effect (within ±30%) on the frequency. The work also developed a modelling approach that predicted the biomass accumulation of P. tricornutum using the simulation results from Lagrangian particle tracking. To the best of the author’s knowledge, this was the first method that could predict the effect of different reactor designs and operating conditions (e.g. superficial gas velocity on algal biomass growth). The model simulation results, together with the reactor hydrodynamics simulated by CFD, could be used to guide the design of more light-efficient photobioreactors. In summary, this thesis presented research in three directions of the development and optimization of a scalable process for producing EPA using photoautotrophic microalgae. Substantial, novel experimental data about the EPA productivity under industrially relevant conditions was generated for a range of species. For the first time a modelling approach was developed which provided accurate predictions of the biomass, EPA and total fatty acid concentrations. This model was found to work for a range of reactor designs, scales and operating conditions, and the approach developed here could be readily applied to other microalgal processes. Finally, the kinetic model was coupled with the results from CFD simulations to develop a novel modelling tool that can be used for the in-silico design of improved photobioreactor designs. Taken together the results from this work are a substantial step towards addressing the challenges in scaling-up microalgal EPA production. Successful scale-up of microalgae EPA production is key in ensuring people have a sustainable, affordable source of EPA.