Improving raceway reactor productivity via vortex induced vibrations for cost effective microalgae production

Abstract

Microalgae research has been becoming more and more common in the last decades due to a number reasons including need for sustainable energy and fuel production, concerns about climate change and orientation of people to biobased products. Microalgae is considered as an excellent feedstock to meet peoples future demands in these fields. Microalgae converts inorganic carbon into sugar using radiative energy in the process so called "photosynthesis". It can grow on non arable land and does not compete with agricultural food products, can assimilate waste products such as flue gas and wastewater and convert them into biomass. Microalgae cultivation is carried out for both remediation of waste streams and commercial purposes. Algal wastewater treatment is a hot topic in environmental engineering and poses several advantages compared to conventional wastewater treatment processes such as reduction of aeration costs. While algal nutrient removal is more common, certain microalgae species can assimilate organic carbon, making it an interesting alternative to activated sludge process. However, algal wastewater treatment is limited to several community scale facilities. Commercial scale algal biomass production is dominated by food and feed industry. While Spirulina and Chlorella are the most commonly cultivated species, cultivation of Haematococcus and Dunaliella is common due to their ability to synthesize high value products such as astaxhanthin and carotenoids. Microalgae based biodiesel is considered as among the main candidates to replace fossil fuels as algae can accumulate lipids up to 70% their dry weight and it can be said that most of the research effort involving microalgae is towards this subject. However, algal biodiesel is not economically feasible yet due to high costs of cultivation, harvesting and other downstream processes. Microalgae cultivation systems are generally classified as open and closed systems. Closed systems offer a more controlled environment with higher light availability; light paths being couple of centimeters to 10 cm. Microalgae growth rate and biomass concentration is higher in this type of systems; however much higher capital and operating costs, as well as upscaling issues strongly limits their utilization. Open systems on the other hand are much easier to build and operate. Raceway ponds is the most common microalgae cultivation system. A raceway can be defined as an oblong channel where culture medium is most commonly circulated with the help of a paddlewheel. A single pond can occupy an area up to 4 hectars. Depth of the pond is kept 20-30 cm to ensure light penetration and flow velocity is typically 0.2-0.3 m/s. While 90% of commercial scale microalgae production is carried out in raceway ponds it has strong disadvantages compared to closed photobioreactors such as limited light availability and vulnerability to environmental and climatic conditions. Among these, light availability is perhaps the most important bottleneck for optimization of low cost microalgae biomass production. Limited light availability results from very limited vertical mixing in long straight channels of raceway ponds. Improving vertical mixing can be achieved by introducing more turbulent to flow by increasing flow velocity, which is energy intensive. Thus, energy efficient systems for improving vertical mixing and creating light dark cycles in raceway ponds is a strong necessity for making algal products more economically attractive. xxii Aim of this thesis to improve vertical mixing in raceway ponds without increasing operational costs. Method to improve vertical mixing is implementation of vortex induced vibrations. Vortex induced vibration is a form of flow induced motion whereby a body becomes excited, with vortices shed from its surface. These vortices, when they shed and leave the surface, exert force on the cylinder. When a vortex separates from the top part, the cylinder feels a downward force. When it separates from the bottom, the direction of the force is then upwards. When the cylinder is allowed to move in the direction perpendicular to the flow, cylinder moves up and down. This is a periodic motion and will last forever as the fluid continues to flow. Vortex induced vibrations make use of the flow energy and convert this power of the fluid to oscillate the cylinder. Within the scope of the thesis vortex induced vibration systems are used to improve vertical mixing in raceway ponds without any additional energy input. Vortex induced cylinder oscillation requires flow uniformity along the width of the channel where the system was implemented. For this, first, flow field of existing raceway pond at the roof of ITU Environmental Engineering Department was numerically investigated using CFD code, to see if it is available for implementation of vortex induced vibration systems. Flow velocity is kept as 0.3 m/s. Paddlewheel was removed from the domain to decrease computational effort and k- ɛ was chosen as turbulence model. In the CFD analyses, the raceway pond was modified with one, two and three flow deflectors and width of the central divider was increased to 5 and 10 cm. It has been seen that by installing 3 semi-circular flow deflectors in the bends of the pond, uniform flow along channel width could be achieved. Existing raceway pond was modified in this way and vortex induced vibration system, which consists of a cylinder with 6 cm diameter and two springs was installed to pond. Continuous cylinder oscillation was achieved with 6.5 cm vertical amplitude and 1.24 s-1 oscillation frequency while water level was 0.3 m. Impact of this cylinder motion on vertical mixing was numerically analyzed using CFD code. To simulate the cylinder oscillation, governing equations of vortex induced vibration was implemented to model as user defined function. Flow velocity was kept as 0.3 m/s as in the experiments and k- Ω SST was chosen as turbulence model. Model was run under steady conditions until the dynamic equilibrium was reached. After this, model was run for 10 seconds to investigate VIV motion. Model output revealed that vertical motion of flow covered 2/3 of pond depth. Cylinder oscillation directs flow upwards with a magnitude of 0.3 m/s and creates high frequency light dark cycles to effectively utilize so called flashing light effect. Light to dark cut off point was assumed as 3 cm below culture surface and average frequency of L/D cycles in the first 60 cm downstream of the cylinder for uppermost, neutral and lowermost cylinder positions were calculated as 21.17 s-1, 5.28 s-1 and 2.33 s-1, respectively. Pure culture of Chlorella vulgaris was grown comparatively to assess the effect of VIV on biomass production capacity. Culture was first grown under laboratory conditions in 10 L plastic bottles. Temperature was kept constant at 28 oC and Bald's Basal Medium was used as growth medium. Culture was grown for 1 week in laboratory and after that transferred to open ponds at the roof of ITU Environmental Engineering department. Culture was further grown for 1 week to acclimate outdoor conditions and diurnal cycle. VIV system was removed from the pond and culture was grown in two identical ponds for additional one week two make sure identical ponds demonstrated the same performance in terms of biomass production capacity. VIV system was implemented to one of the ponds at the end of this week and comparative cultivation xxiii with and without VIV system was carried out for one week. Biomass growth was monitored by optical density measurement under wavelength of 540, 690 and 750 nm. Experiments revealed that VIV increased biomass production capacity in the pilot scale raceway pond with 3 m channel length and 1 m total width by over 20%. Amplitude response of cylinder achieved in the pilot scale raceway pond for 0.3 m/s flow velocity was lower compared to literature. To investigate the reason of this and to investigate effect of VIV on vertical mixing and light dark cycles in the raceway pond in detail, flow visualization technique was applied. Particle image velocimetry using LED illumination was applied under experimental conditions mentioned above. Frame rate of PIV camera was 165 FPS and focal length was 35 mm. Flow visualization experiments without the VIV system revealed that flow velocity decreases through pond depth in the paddlewheel driven system with a 4.5 cm bottom clearance. Distribution of horizontal flow velocity could be modeled with 2nd order polynomial. This uneven distribution of flow velocity through the depth suppresses cylinder motion which resulted in lower amplitude response compared to literature. Several other equipment such as archimedes pumps, centrifugal pumps, airlift pumps and propellers are proposed to replace the paddlewheel for further exploitation of effect of VIV motion on vertical mixing and thus light availability and biomass production capacity. Indeed, it has been reported that propellers and airlift pumps are more energy efficient than paddlewheels. Furthermore, paddle induced circulation would become more disadvantageous when culture depth was increased due to mechanical reasons. Flow field in the raceway pond when vortex induced vibration system was implemented was analyzed using particle imaginary technique. Vertical component of flow was in accordance with CFD analyzes in general. 75 cells were selected at the downstream of VIV cylinder and were tracked for 20 cm in horizontal direction, until they disappeared from the other side of cameras projection area. Initial cell positions were set as three equidistant planes through the depth of the raceway channel to represent the average situation. Flow visualization experiments revealed that 33% of selected cells entered high frequency light dark cycles with the help of VIV. Average frequency of light dark cycle was found to be 35.69 s-1 with a light fraction of 0.49. 44% of cells entered light limited zone from dark zone as a result of VIV motion. Pilot scale RWP has 3 m channel length, which means, compared to full scale facilities, cells pass through paddlewheel, where vertical mixing happens, more frequently. In other words, pilot scale RWP is more effectively mixed compared to full scale systems. By installing one VIV cylinder, it can be said that a 2nd vertical mixing point was created in the pond. On the other hand, real scale RWPs have much higher channel lengths, thus effect of paddle induced vertical mixing in these systems would be less pronounced. In these long channel sections, cells near the surface will become "over- charged" after a certain period of time. On the other hand, cells at lower parts of pond depth will reside in photobiologically inactive parts of the pond for a prolonged period. As indicated above, VIVs can cycle cells between photobiologically active and inactive parts of channel and increase number of cells that perform photosynthesis in one circulation around pond. Thus, it is believed that the effect of VIV could be more pronounced in larger ponds.