Qi Zheng ,Xiaoyin Xu ,Gregory J.O.Martin ,Sandra E.Kentish

1 Peter Cook Centre for CCS Research,Department of Chemical Engineering,The University of Melbourne,Parkville,Victoria 3010,Australia

2 Algal Processing Group,Department of Chemical Engineering,The University of Melbourne,Parkville,Victoria 3010,Australia

Keywords:Bubble Carbon dioxide Flue gas Bioreactors

A B S T R A C T Microalgae have great,yet relatively untapped potential as a highly productive crop for the production of animal and aquaculture feed,biofuels,and nutraceutical products.Compared to conventional terrestrial crops they have a very fast growth rate and can be produced on non-arable land.During microalgae cultivation,carbon dioxide(CO2)is supplied as the carbon source for photosynthesising microalgae.There are a number of potential CO2 supplies including air,flue gas and purified CO2.In addition,several strategies have been applied to the delivery of CO2 to microalgae production systems,including directly bubbling CO2-rich gas,microbubbles,porous membrane spargers and non-porous membrane contactors.This article provides a comparative analysis of the different CO2 supply and delivery strategies and how they relate to each other.

1.Introduction

In 2016,fossil fuel consumption,including oil(34%),natural gas(23%)and coal(28%),accounted for 85%of primary energy consumption in the world,which is the largest contributor to greenhouse gas emissions[1].As these non-renewable fossil fuels are limited resources that impose a large carbon footprint,the world is looking to alternative,renewable sources of energy.Biofuels derived from biomass are renewable and near-carbon neutral.They can compensate for the depleting nonrenewable fossil fuel resources and reduce global carbon emissions.Microalgae are regarded as a feedstock for‘third generation’biofuels due to their ability to be grown at very high productivity on completely non-arable land[2,3].While this means that microalgae are a promising biofuel feedstock,the costs are currently prohibitively high[4].Technologies to reduce the cost of both producing and processing microalgal biomass are required.

One of the major costs relates to the need to provide a carbon source to facilitate microalgae growth.Current CO2sources are atmospheric air,commercial purified CO2or raw flue gas.From a flue gas perspective,at least 13 gigatonnes of CO2is emitted every year from large point sources,such as power plants,cement processing,the steel industry and gas production[5].Effective utilisation of such large point sources of CO2emissions is an opportunity to reduce global greenhouse gas emissions.Every year,around 350 gigatonnes of atmospheric CO2is biologically converted to biomass by photosynthetic organisms[6,7].Consumption of such point sources would thus require only an additional 4%in world biomass production.

In mass microalgae cultivation,simply relying on diffusion of CO2from the atmosphere will mean the cultures are limited by carbon,rather than light availability,reducing productivity and essentially wasting the solar energy resource(the available light falling on that area of land).To date the most common strategy to reduce such carbon limitation has been to bubble gases into the microalgae cultures[8].However,in the process,approximately 50%to 90%of the CO2delivered to the microalgae pond exits to the atmosphere[9].As CO2has been estimated to account for over 50%of the raw material costs in microalgal cultivation[10],more effective delivery mechanisms are vital for both efficient microalgae production and reduction of CO2emissions.This paper provides a critical analysis of different CO2delivery strategies by comparing their efficiency,costs and limitations.

2.The Role of CO2 in Microalgae Growth

2.1.CO2 uptake by microalgae

Carbon is a fundamental compositional element of microalgae,representing approximately 50 wt%of the microalgae dry biomass[11].During photoautotrophic growth,microalgae uptake dissolved inorganic carbon from the surrounding water,and chemically reduce it(via the so-called light-independent reactions)using energy derived from photosynthesis.This reduced carbon goes into central metabolism and from there can be used for production of lipids and biomass.The overall reaction can be summarised as follows[11]:

In eukaryotic algal cells,CO2is fixed in the stroma of the chloroplast by the enzyme Ribulose Bisphosphate Carboxylase Oxygenase(RuBisCO)within the Calvin–Benson–Bassham(CBB)cycle,as shown in Reaction(2)[12].

RuBisCO has a very slow turnover and a poor specificity for CO2over oxygen,a product of photosynthesis.To increase efficiency and to favour carbon fixation over competing photorespiration,most microalgae[13,14]have evolved CO2concentrating mechanisms(CCMs)to increase the CO2concentration in the vicinity of RuBisCO.There are a number of different CCMs,including the‘biochemical C4’and the‘CAM’mechanisms of higher plants,and biophysical active transport processes in microalgae,as have been summarised in review papers on this topic[12,15,16].Briefly,the CCMs in microalgae involve a combination of active transport of inorganic carbon from outside the cell into the chloroplast,concentration of CO2at the reaction site of RuBisCO,and the use of carbonic anhydrase enzymes to interconvert between different forms of inorganic carbon[17].

While CO2is the substrate for RuBisCO,most algae species have the ability to uptake and utilise both CO2and[12].This is important as most inorganic carbon is present asat the pH values typical of microalgae cultures.There are a number of known pathways of CO2transport in eukaryotic algal cells[12]involving CO2diffusion and active transport of CO2and(Fig.1).The affinity for the two carbon species varies from species to species[18].The difference is largely related to extracellular carbonic anhydrase(CA),the essential enzyme that catalyzes the reversible conversion ofto CO2.Only algae capable of usinghave been found to express CA[14],which can then maintain the continuous supply of CO2to the active transporters by dehydration ofin the periplasmic space[19,20].These algae show a higher affinity forthan CO2,but the major carbon species entering the cytoplasm is CO2.The existence of external CA in different algae has been reviewed[12,14].In addition to variations between species,the expression of external CA is also largely influenced by pH and CO2concentration[19,21,22].

2.2.Carbon limitation in microalgae cultures

The productivity of microalgae cultures can be limited by a number of factors,of which light and CO2availability are of most practical importance.Fully realising the potential productivity of a microalgae cultivation site requires maximal utilisation of the available light.This means that the cultures must be light-limited,and not limited by CO2availability.Due to the CCMs,the availability of inorganic carbon is often not a limit to photosynthesis in dilute natural systems,in which the overall demand for CO2is not intense.However,CCMs cannot avoid carbon limitation in dense microalgae cultures in which the overall demand for CO2exceeds the supply of CO2that can be provided by diffusion from the atmosphere[24,25].Achieving a typical target biomass productivity of 25 g·m-2·d-1requires an average of 45.8 g CO2·m-2·d-1to be delivered to the growth ponds(assuming a biomass carbon content of 50%w/w).This is a much greater flux than can be provided to algae ponds via atmospheric diffusion,due to the low concentration of CO2in the atmosphere and limited pond surface area.Therefore,to maximise the utilisation of the available light and growth infrastructure,carbon limitation needs to be avoided in mass microalgae production by provision of external CO2[26].However,excessively high CO2concentrations can decrease the medium pH and reduce the activity of external CA[27].This can eventually limit microalgae growth.So it is important to achieve a balance between the CO2demand of microalgae growth and the CO2supply.

3.Global CO2 Supply and Demand

Global anthropogenic CO2emissions were around 33 gigatonnes in 2017,mainly from fossil fuel combustion,industrial processes,residential consumption and transportation[1].Although CO2emissions are numerous,only the large point sources of CO2emissions are amenable for capture using current technology.Understanding the global CO2supply gives us information of potential CO2sources for microalgae cultivation.As the CO2market price depends on the relationship of CO2supply and demand,realising the current and future CO2demand helps to evaluate the economic viability of different CO2sources for microalgae cultivation.

Fig.1.Diagrammatic representation of inorganic carbon uptake processes in eukaryotic algae,reproduced from Raven and Beardall[23].

3.1.CO2 supply

The Intergovernmental Panel on Climate Change(IPCC)reported there was in total 13375 Mt CO2emissions from large point sources(of more than 0.1 Mt CO2per year)in 2002[5].The largest proportion(10540 Mt)was from fossil fuel combustion including coal,natural gas and fuel oil.However,the low CO2concentration(3 vol%to 15 vol%)in these sources may lead to high costs for CO2capture.Other sources of highly pure CO2(above 97 vol%)include the ammonia production process and ethanol fermentation,which could provide 113 Mt and 17.6 Mt CO2in 2002,respectively[5].The United Nations Industrial Development Organization(UNIDO)also reported that 6%of the global industrial CO2emissions was from high purity sources(426 Mt)[28].Ammonia production,gas processing,coal-to-liquid processes and ethylene oxide production provided 240 Mt,160 Mt,20 Mt and 6.3 Mt CO2in 2010[28],respectively.

3.2.CO2 demand

Parsons Brinckerhoff[29]investigated the demand for purified CO2in 2011 and found that the total CO2demand was 80 Mt,in which 50 Mt was for enhanced oil recovery(EOR)and the remaining 30 Mt for beverage carbonation,the food industry,precipitated calcium carbonate and other uses[29].

The CO2demand ultimately determines the CO2price.In 2010 this was quoted as anywhere from 3 USD to 55 USD per ton CO2depending upon the source and the location of this source[5,29].Although CO2emissions from fossil fuel power plants provide a large potential supply,it is important to notice that there is a gap between the CO2market price and the costs for CO2capture.Rubin et al.[30]estimated the CO2capture costs of different technologies for fossil fuel combustion power plants to be between 36 USD and 111 USD per ton CO2.The implementation of a carbon tax or other incentives to reduce emissions can partly offset this difference.However,new strategies must be proposed to completely eliminate the gap between the CO2market price and carbon capture costs.

4.Types of CO2 Supplies Available for Microalgae Cultivation

As discussed above,CO2must be provided to improve microalgae growth such that the available light is fully utilised.But contrary to EOR or other chemical industry processes,it is not necessary to use 100 vol%CO2for microalgae cultivation,so atmospheric air and raw flue gas can also be regarded as potential CO2sources.Secondly,some impurities(e.g.NOxand SOx)present in flue gas,even at low concentration,may detrimentally affect microalgae growth.The following paragraphs compare the different types of CO2supplies for microalgae cultivation,including air,raw flue gas,commercial purified CO2,purified CO2from carbon capture facilities and CO2-containing solvents,in relation to CO2delivery.The advantages and disadvantages of each type of CO2supply is discussed,with a summary provided in Table 2.

4.1.Air

The atmosphere is the most readily available CO2source for microalgal growth,and is the source for growth of these organisms in natural systems.In this case,there is no need for gas transportation from a CO2production plant.However,atmospheric air has only about 0.035 vol%CO2.This low concentration means that there is a low driving force for mass transfer into a microalgal growth medium.This fact,combined with the low gas–liquid interfacial area on the surface of ponds,means that relying on passive atmospheric diffusion is not sufficient to maintain high biomass concentrations or productivities.A gas compressor and a pump on site can be used to bubble air through the cultures to improve carbon delivery for microalgae and enhance the agitation of the culture,however the low CO2concentration still limits mass transfer efficiency due to the low concentration driving force.Jiang et al.[31]showed that with air aeration,Nannochloropsis sp.only achieved one third of the biomass concentration of cultures grown using gas containing 15%CO2.Even more importantly,the low concentration of CO2means that an enormous volume of gas has to be pumped to provide the required carbon for intensive cultivation(Table 1).These issues mean that air is not a practical source of CO2despite being free and highly available.

Table 1 Volume of gas required①(t·m-2·d-1)to provide carbon②for an areal microalgal biomass productivity of 25 g·m-2·d-1 as a function of CO2 concentration and CO2 utilisation.CO2 utilisation is the proportion of CO2 put into the ponds that is converted to microalgal biomass,and will increase with increasing CO2 concentration and decreasing bubble size

4.2.Raw flue gas from fossil fuel combustion

Raw flue gas from fossil fuel combustion accounts for 78%of anthropogenic CO2emissions[5].The high capture costs associated with the low CO2concentrations in this flue gas restrict its usage for EOR and other industrial applications,but raw flue gas can be utilised as the carbon source for microalgae cultivation.The CO2concentration ranges widely,from about 4%for natural gas combustion up to about 33 vol%for cement or steel production[32].The above-ambient concentrations increase the rate of mass transfer compared to air,and reduce the required volume of gas by orders of magnitude(Table 1).

However,in addition to CO2,raw flue gas usually contains several other components,such as N2,O2,CO,NOx,SOx,CxHy,heavy metals and particulate matter[33,34].The concentration of these components varies depending on the flue gas source[35];NOxcan be in a range from 2 to 1500 μl·L-1and SOxcan be in a range from 0 to 1400 μl·L-1[34].Components,such as NOxand SOx,have been reported to have a negative effect on microalgae growth[32,36].SOxfrom flue gases may hydrolyse in the microalgae medium to release H+,causing a pH decrease and inhibit the microalgae growth[35].NOxhas two forms NO and NO2,that may be oxidized by the oxygen produced by microalgae photosynthesis to form nitrate or nitric acid in the culture medium[37].Kao et al.[35]showed that flue gas containing less than 100 μL·L-1NOxwith 0.2 L·L-1·min-1aeration rate can be used as an additional nitrogen source for Chlorella sp.MTF-15,which could be used to increase the maximum biomass concentration reached in the cultures.

There have been many other studies into the effect of flue gas concentration on microalgae productivity[38–43],as well as some recent investigations into the effects of flue gas on microalgae physiology and biochemistry[44–48].Broadly,there are two options for directly using raw flue gas containing NOxand SOxto enhance microalgal growth at large scale.One approach is to remove the SOxand NOxfrom the flue gas prior to microalgae use by adding desulfurization and denitrification units[35].However,to avoid the cost of gas processing,another method is to isolate NOx-and SOx-tolerant microalgae strains[39].Chlorella fusca LEB 111,which was isolated by Duarte et al.[26]from the ash settling ponds of a thermoelectric plant in Brazil,can tolerate up to 400 μl·L-1SO2and 400 μl·L-1NO,respectively.Radmann et al.[49]isolated Chlorella vulgaris from the sewage treatment pond of a thermoelectric plant in Brazil,which can grow with 12%CO2,60 μl·L-1SO2and 100 μl·L-1NO.Chlorella sp.,Dunaliella tertiolecta and Scenedesmus obliquus are the most commonly reported species that can tolerate SOxand NOx[32].

Usually the microalgae pond needs to be located near to the flue gas source to reduce the CO2delivery costs,but this also limits the potential scale of algae production.While there are many point sources of flue gas,only a small proportion of these will be suitably located for algae production(e.g.suitable climate,proximity to cheap flat land and water)[8].The distance between the CO2point source and algae cultivation site influences the cost and therefore the potential utilisation of this form of CO2supply.For example,Quinn et al.[50]estimated that with a baseline maximum distance between the CO2source and the algae cultivation site of 4.8 km,only 44 million barrels of oil equivalent could be produced from microalgae in the USA.Increasing the CO2transportation distance to 80 km would allow the production of 1.8 billion barrels of oil,however the costs of CO2transportation would not be economically feasible.Rubin et al.[30]estimated the CO2transportation costs through onshore pipelines as 1.3 to 10.9 USD per ton CO2per 250 km depending on the different capacities.The feasibility of CO2transportation is therefore dependent on both the CO2concentration of the gas and the required distance.A life cycle analysis completed by Stephenson et al.[51]indicated that an increase in CO2concentration in the feed gas from 5 vol%to 12.5 vol%could reduce the energy required for microalgal biodiesel production from 23.7 to 6.5 GJ·t-1of biodiesel.For direct flue gas utilisation,higher volumes of gas need to be delivered than that provided by a pure CO2stream(Table 1).As gas compression and transportation are energy intensive[52],purified CO2is therefore more preferable for microalgae plants located further from point sources of CO2.

4.3.Purified CO2

For microalgae ponds that do not have flue gas sources nearby,the more economic choice may be commercial purified CO2.As mentioned above,ammonia production,gas processing,coal-to-liquid processes and ethylene oxide production can provide commercial sources of purified CO2[28].Research by Kadam[52]compared the supply costs of CO2from raw flue gas and purified CO2extracted from the flue gas using monoethanolamine(MEA),with a 100 km pipeline distance between the flue gas source and the microalgal cultivation site.The results showed that the direct flue gas supply(containing 14%CO2)incurred expenses in compression,drying(46.6 USD per ton CO2)and transportation(10.6 USD per ton CO2).Including MEA extraction added a cost of 28.72 USD per ton CO2,but decreased the costs of compression,drying(8.48 USD per ton CO2)and transportation(3.30 USD per ton CO2)due to the smaller volume of the purified stream.For this scenario,the total cost of directly pumping flue gas was thus found to be 40%more expensive than supplying purified CO2by MEA extraction.

4.4.Bicarbonate addition

As discussed,a large amount of energy is required for gas compression and transportation over the vast areas required for large scale microalgae cultivation.An alternative is to directly add sodium bicarbonate to the medium to enhance microalgae growth[32].This avoids any pumping of gas,however the price of CO2via sodium bicarbonate is higher than even purified CO2(Table 2).

Enhancement of algae growth using bicarbonate has been demonstrated.Pancha et al.[55]found a 23%and 21%increase in biomass and lipid productivities when Scenedesmus sp.CCNM 1077 was supplied with 0.6 g·L-1sodium bicarbonate instead of normal BG11 medium without sodium bicarbonate addition.Nunez and Quigg[56]showed that the growth of Nannochloropsis salina was about 3-fold higher with 5 g·L-1sodium bicarbonate,than without sodium bicarbonate addition.It has also been shown that delivery of gaseous CO2can be supplemented by the addition of sodium bicarbonate to increase productivity[57].

Usually seawater species have a higher tolerance to bicarbonate addition than fresh water species,as seawater species are already conditioned to high ionic strengths.As has been mentioned above,some microalgae strains(for instance,Nannochloris atomus and Nannochloris maculata[14])do not have external carbonic anhydrase,so the major,or only inorganic carbon species taken up is CO2.This limits the potential application of this approach to species that can utilise bicarbonate.Sodium bicarbonate addition is typically in a range from 0.1 to 5 g·L-1[55,56,58,59],with the highest reported concentration of 0.60 mol·L-1NaHCO3being applied to the highly salt-tolerant Dunaliella salina[59].An issue with this approach is that while the bicarbonate is consumed,the sodium concentration within the medium will grow over time,increasing the total ionic strength and the cost of production.Nonetheless,it has been shown that there are certain algae present in alkaline natural soda lakes which can be highly productive at very high ionic strengths and high pH(＞10)[60].The high pH increases the rate of CO2mass transfer[61]and the high alkalinity provides good buffering capacity.The potential to grow algae suited to cultivation under these extreme conditions is therefore of growing interest[62–65].

4.5.CO2-containing solvents

Using CO2-containing liquid solvents for microalgae growth could also reduce the energy required to transport the CO2as liquid does not require compression and is much more efficient to pump.This could provide a cheaper source of bicarbonate and reduce the energy associated with carbonate regeneration.It would also avoid energy and losses associated with pumping a gas and decouple CO2capture and utilisation[53,54].Noel et al.[66]used a sodium carbonate solution,loaded with CO2to form sodium bicarbonate,as a carbon source.The CO2was transferred into seawater through contact of the two liquids either side of a non-porous poly dimethyl siloxane(PDMS)membrane.The CO2loaded seawater was used as a microalgal medium,while the sodium carbonate solution was recycled.

An alternative approach developed by our own group is to capture the carbon dioxide into a solvent such as MEA and to circulate this solvent directly throughout the algal bed within similar PDMS non-porous membranes[67,68].The CO2desorbs from the solvent into the algal media,allowing this liquid to be directly recycled to the capture process to absorb more CO2.Assuming a pressure drop of 300 kPa across a 100 km delivery distance,and a pump efficiency of 75%,the energy penalty of pumping is only 0.02 GJ·t-1CO2,significantly reducing the energy penalty for CO2delivery.Aside from the reduction in compression costs,this approach also significantly reduces the cost of the CO2capture step.When purified CO2is obtained from a typical chemical absorption process in a power plant,the energy penalty for CO2regeneration is 2.4–4.2 GJ·t-1CO2[69].By contrast,when the membrane system is implemented to deliver a CO2-loaded chemical solvent,the energy penalties for CO2regeneration are avoided.Further,with this approach,CO2might be temporarily stored in the solvents when the microalgae stop photosynthesis at night,allowing the capture process to continue operating on a continuous basis and avoiding the need for any gas storage.Conversely,a large membrane area is required within the ponds and this can add significantly to the capital costs(see Table 1).We estimate that a membrane cost of under 10 USD per square meter is needed to ensure a competitive process.

4.6.Comparison of types of CO2 supplies

A comparative summary of information relating to different types of CO2supplies is shown in Table 2.While air is free,practically unlimited in volume,and available on-site without requiring transportation,its low CO2concentration means that it is not a feasible option due to enormous volumes of air that have to be injected into the microalgae cultivation ponds.

The use of purified CO2,flue gas,and CO2-loaded solvents is all dependent on the availability of point sources of CO2and their proximity to land suitable for large scale microalgae production.To further explore the topic of CO2supply,the reader is referred to a number of detailed analyses of the availability,location and overall potential of point CO2sources for microalgae cultivation in the USA[50,70,71].Briefly,flue gas is widely available and has limited commercial value.However,it also contains impurities that may inhibit microalgae.Purified CO2has the advantage of higher mass transfer,utilisation and transportation efficiencies,but has a commercial value and is more limited in availability than flue gas.

While the gaseous CO2supplies require considerable energy for compression and transportation,the energy needed to deliver bicarbonates or CO2-loaded solvents is much lower.The price of bicarbonates is higher than even purified CO2,and is more limited in availability than flue gas.In addition,not all species of algae can uptake bicarbonates.The use of CO2-loaded solvents would require co-investment with a CO2capture plant,but in doing so would reduce the cost and energy required for solvent regeneration.While this approach appears to have potential,it is still an emerging technology.Ultimately,the selection between flue gas,purified CO2and bicarbonate addition will be made on a site-by-site basis according to the relative economics.

5.Comparison of CO2 Delivery Technologies

5.1.CO2 delivery requirements

At typical microalgal growth conditions,only about 0.8 mmol·L-1of inorganic carbon can be stored in seawater[8].For an areal biomass productivity of 25 g·m-2·d-1this CO2would be consumed in an average of about 5.5 h.As such,inorganic carbon must be frequently replenished throughout the cultures.It is important to realise that areal productivity targets(e.g.25 g·m-2·d-1)are averaged over a year.CO2supply during peak photosynthetic periods(e.g.the middle of a sunny day in summer)will be many times higher than the average,while night time demand will be zero.CO2delivery strategies must be able to cope with fluctuating demands,both seasonally and diurnally,and ensure that inorganic carbon is available throughout the cultures.For very large cultivation ponds,multiple CO2injection points will be required,the spacing of which will be determined by the rate at which CO2supplies are depleted from the cultures[8].

5.2.Conventional CO2 delivery by sparging

Microalgae can be grown in artificial systems such as open raceway ponds and closed photobioreactors[73–75].The conventional method for delivering CO2to microalgae ponds or photobioreactors is to use a sparger or diffuser[76–78].For large scale production,microalgae are typically cultivated in open raceway ponds[8].A raceway pond is a shallow pond configured as a loop,around which the algal cultures are circulated,typically using a paddle wheel(Fig.2)[79].The shallow 0.2–0.5 m pond depth allows the microalgae to best utilise the available light,by minimizing the effects of self-shading.Nutrients(such as nitrate,phosphorus,trace metals)are introduced into the pond after the paddle wheel.As mass transfer of CO2from the atmosphere is insufficient to avoid C-limitation[15],purified CO2or air is usually injected into the pond through a sparger at the bottom of the pond[80].The paddlewheel is operated to circulate and agitate CO2,the microalgae broth and nutrients[81].

5.3.CO2 utilisation efficiency

Fig.2.A typical raceway pond configuration for algae production[79].

As a carbon source,CO2has been estimated to account for around 66%of the raw material costs(including CO2,fertilizers and water)for microalgae production in a large scale plant using sparged tubular photobioreactors[10].Increasing the efficiency of utilisation of CO2can therefore significantly reduce operating costs.CO2utilisation efficiency is defined as the proportion of CO2injected into the ponds that is converted into microalgal biomass,rather than lost to the atmosphere.As shown in Table 1,the rate of gas delivery is directly dependent on the CO2utilisation efficiency.For gaseous forms of CO2supplies(i.e.air,flue gas,purified CO2),the efficiency of CO2utilisation in microalgae cultures is dependent on the rate of CO2mass transfer from the gas to the liquid relative to the retention times of the gas in the liquid.

The mass transfer of CO2from a gas bubble to an algae cell involves a number of processes(Fig.3),which are described below.In an ideal system with a CO2-containing gas over water,there will be an equilibrium state according to the following:

Fig.3.Schematic of CO2 mass transfer pathways from a gas bubble to an algae cell.The important steps are identified by numbers in parentheses and are explained in the text.

In an algae culture the transfer of CO2from a gas bubble represents a non-equilibrium state in which CO2is pulled through by photosynthetic consumption of the CO2(aq)and.The overall driving force for mass transfer is the difference between the concentration of CO2in the gas bubble and the concentration of CO2(aq)andthat remains in solution despite uptake by the algae cells.CO2mass transfer is a multi-step process(Fig.3).(1)CO2(g)must first be solubilised across the liquid film layer to form CO2(aq),the equilibrium concentration of which is dependent on the CO2(g)concentration in the gas bubble according to Henry's law(Table 3).The rate of CO2mass transfer is dependent on the resistance to mass transfer in the liquid film layer and the concentration driving force.The concentration driving force is the difference between the equilibrium CO2(aq)concentration as determined by Henry's law,and the CO2(aq)concentration in the bulk liquid.A sufficiently high CO2(aq)concentration has to be maintained in the bulk liquid in order to drive mass transfer through to the algae either directly(2)or via(3),(4),(5),and(6).This means a high equilibrium CO2(aq)from a concentrated CO2supply is needed to maintain a sufficient concentration driving force for efficient CO2mass transfer.

5.4.Gas delivery sumps

To achieve high efficiencies of CO2utilisation,CO2must be rapidly transferred out of the gas bubbles and into solution before the gas bubbles reach the surface of the ponds.The retention time of gas bubbles has a cubic dependency on the bubble diameter and linear dependency on the pond depth.As microalgae ponds are typically only about 0.2–0.5 m in depth,large bubbles can rise so quickly to the pond surface that most of the CO2is lost to the atmosphere.To increase the retention time of the bubbles,Lundquist et al.[8]introduced a sump(1 m depth)in the raceway pond at the point of gas injection.In addition to increasing the depth for bubble rise,by placing the diffuser below the downward stream the effective contact time can be further increased.For bubbles with a rise velocity of 0.30 m·s-1,and a downward water velocity of 0.25 m·s-1,the contact time of gas and water could be increased from 1 s to around 20 s(Fig.4),and the CO2loss reduced.Nonetheless,it was assumed that CO2utilisation would still be limited to about 50%for a flue gas using this design[8].

Fig.4.Schematic diagram of a sump for gas delivery in microalgae ponds,modified from Lundquist et al.[8].

5.5.Microbubbles

Reducing the bubble size is another approach to improving CO2utilisation efficiency.Compared to conventional bubbles(with adiameter of 1–2 mm[32])microbubbles with a diameter of less than 100 μm have a much higher surface to volume ratio and slower rise velocity in the microalgae medium,so more CO2can dissolve in the medium[32,82](Table 3).

Table 3 Influence of bubble size on key factors affecting CO2 utilisation efficiency:residence time and specific surface area

Compared with the rapid rise of larger bubbles,microbubbles can reside for much longer periods(Table 3).The difference in residence times between large bubbles and microbubbles can be even greater than predicted by Stokes'law owing to the likely coalescence of large bubbles and the ability of micro-turbulence to entrain microbubbles.Further,microbubbles of purified CO2can actually decrease gradually in size as they are dissolved in the water(Fig.5)[83].Several methods have been applied to create microbubbles,such as fluidic oscillation,ultrasound or pressurized gas liquid circulation system[82,84].Such microbubbles have been applied for O2injection for wastewater treatment or oyster cultivation[82,84,85].Zimmerman et al.[78,86]first introduced microbubbles into microalgae cultivation by using a fluidic oscillation diffuser.The bubble size was around 550 to 750 μm,larger than the usual definition of microbubbles.Later research by Al-Mashhadani et al.[87]showed that with a fluidic oscillation diffuser,a 29%increase in the CO2overall mass transfer coefficient was obtained with 550 μm microbubbles compared with 1.3 mm bubbles(see Table 2).

Although microbubbles can greatly increase CO2utilisation efficiency,there is also a much greater head pressure loss associated with generating microbubbles that leads to an increased energy consumption compared to conventional sparging.The energy associated with forming bubbles is given by the need to create new surface:

where R is the bubble radius and σ is the surface tension,which for air–water systems at ambient temperature is 72 mJ·m-2.This energy clearly increases as the bubble radius(R)decreases(Table 3).To minimize energy loss through microbubble generation,purified CO2supplies are preferable over flue gas as much lower volumes of gas need to be pumped(Table 1).An additional limitation to this approach is that the strong shear stress effect created by microbubble generation may damage less robust microalgae cells[88,89].

5.6.Porous membrane sparger

Porous membranes are another technology that can be used to generate small bubbles in order to maximise CO2utilisation efficiency.The porous membrane can provide smaller and more uniform CO2bubbles(0.2–0.9 mm)[90]than conventional sparging,but is not able to generate microbubbles.Fan et al.[89,91]tested the performance of PVDF hollow fibre membranes through sparging 1%CO2into a helical tubular photobioreactor and achieved a tenfold higher mass transfer coefficient compared to tube bubbling.Cheng et al.[92]compared the performance of a polypropylene hollow fibre membrane contactor with a bubble photobioreactor,and showed that the CO2fixation rate of C.vulgaris was enhanced over threefold.Mortezaeikia et al.[90]also used a polypropylene microporous membrane contactor to supply CO2bubbles for Synechococcus elongatus growth.Similar to microbubble generation,porous membrane sparging results in greater pressure losses and therefore energy consumption than conventional sparging.In addition,the use of porous membranes imposes additional capital and maintenance costs that have to be factored into any design decisions.

5.7.Non-porous membrane contactor

Fig.5.Effect of CO2 delivery technology and supply type on bubble size and mass transfer and resulting CO2 utilisation efficiency for microalgae cultivation.Depth of shading indicates the concentration of CO2 in the gas.Not drawn to scale.

The delivery of molecular CO2directly rather than via bubbles has been investigated as a means to ensure complete CO2utilisation in microalgae cultures.A membrane contactor fabricated with a dense membrane can minimize CO2loss by producing bubbleless CO2in the aqueous phase[93–95].In this case,the CO2crosses the membrane as individual molecules in a vapour state through the solution diffusion mechanism.This eliminates the surface energy required to form a bubble(Eq.4)and thus significantly reduces the compressor demand.Kumar et al.[94]used a composite hollow fibre membrane composed of a thin,dense polyurethane layer sandwiched between two microporous polyolefin layers to transfer gas containing 2%–15%CO2to Spirulina platensis.A similar three layer composite membrane was used by Kim et al.[93,95](Fig.6).

Fig.6.Non-porous membrane for bubbleless CO2,modified from Kim et al.[93].

As described above,both Noel et al.[66]and Zheng et al.[67]have used a similar thin film composite hollow fibre membrane composed of a thin,dense PDMS layer and a microporous support layer to deliver CO2from CO2-loaded solvents. With solutions containing 0.5 mol CO2·mol-1of solvent,a 16 to 20-fold increase in volumetric productivity was achieved compared with simple air diffusion[68].This approach not only eliminates the surface energy requirement for bubble formation,but also eliminates the need for gas compression and transport,replacing this with the pumping of a liquid phase,which is less energy intensive.

5.8.Comparative summary

A comparative summary of the various CO2delivery technologies is provided in Table 4.Each of the CO2delivery technologies differ in their inherent capabilities,limitations,and energy requirements.They also differ in terms of their suitability to different CO2supplies.In particular,sparging technologies will result in incomplete utilisation of the CO2supply and are therefore less suited to purified CO2sources that cost money to purchase.Conversely,microbubble generation and porous membrane sparging are less suited to flue gas due to the high energy requirements for pumping large volumes of gas.The selectionof CO2delivery technology is complex and site specific,depending on the type and price of CO2supply.

Table 4 Qualitative comparative assessment of the different strategies for delivering CO2 to microalgae

6.Conclusions

This paper has described a range of CO2supplies and CO2delivery strategies for microalgae systems to enhance microalgae growth.As the largest proportion of CO2emissions is from large point sources,flue gas is a potential supplement to commercial purified CO2.While flue gas is cheap,the use of purified CO2can greatly reduce gas volumes and pumping energy and increase mass transfer efficiency.However,the gap between CO2market price and carbon capture costs has to be narrowed.The energy costs of CO2delivery and the loss of CO2to the atmosphere are important contributors to the viability of gaseous delivery approaches.The use of concentrated CO2supplies and the generation of smaller bubbles can improve utilisation efficiency,however they come at a cost.The use of microbubble generators or the delivery of CO2through porous hollow fibre membrane spargers can achieve higher CO2utilisation than sparging,but require considerable energy for bubble formation.The use of CO2containing solvents,delivered through non-porous membranes offers another approach that ensures low CO2loss and avoids gas compression and transportation energy.However,the capital costs associated with the membrane installation can be high.

Acknowledgements

The authors thank the Pratten Foundation for their valuable financial contribution to this work.