Mass Production of Microalgae at Optimal Photosynthetic Rates

Microalgae have been studied in the laboratory and in mass outdoor cultures for more than a century and our initial understanding of photosynthesis became unravelled in the laboratories of Otto Warburg. The breakthrough in his laboratories came when he started using Chlorella as model organism [1]. As Grobbelaar [2] pointed out, applied phycology and the mass production of microalgae, became a reality in the 1940’s. Since then, microalgae have been grown for a variety of potential applications, such as the production of lipids for energy using flue-gasses, anti-microbial substances, cheap proteins for human nutrition and the production of various bio-chemicals. At present the focus is on bioenergy [3], however, their only real success has been in wastewater treatment. A major frustration for microalgal biotechnologists has been the realization of much lower yields than what is potentially possible from laboratory measurements. The inability to operate photo-bioreactors including raceway ponds, at maximal photosynthetic efficiencies, impacts directly on the economies of scale. Because of this many large-scale projects have not delivered what was predicted and many investors have lost their investments. Richmond’s [4] observation that “Microalga culture however is yet very far from supplying any basic human needs..” is as true today as then and he concluded that “the major reason for this stems from the failure to develop production systems which utilize solar energy efficiently”. A consequence of low yields is high production costs rendering this technology only suited for exclusive high-priced products.

However, photo-bioreactors are generally considered to be systems in which the culture has no or minimal contact with the atmosphere and they can be of a variety of designs, such as tubes, plates, coils, bags, etc. [6]. Open ponds, on the other hand, have a large area that is in contact with the atmosphere, but they can be enclosed in e.g. plastic covered greenhouse tunnels. Richmond [4] stated that the major weaknesses of open ponds are the absence of temperature control and the long light-path of about 15 cm. The latter results in large culture volumes and consequently low cell mass densities. The question of temperature control is debatable since the temperature fluctuations will be less in large culture volumes, compared to short light-path cultures with small areal volumes.
Grobbelaar [5] analysed the factors governing microalgal growth in "open" and "closed" systems and concluded that the culture depth (optical-depth/light-path) is the single most important factor that determines microalgal growth and photo-bioreactor productivity. Here we analyse the various variables that could impact on microalgal photosynthesis, especially in large commercial scale production systems, with the aim to develop high yielding microalgal production systems that could be scaled-up. Results generated in small high density laboratory cultures have little value when mega ton production plants are required and it is generally agreed that open raceway ponds would be the means of large commercial outdoor cultivation. For this reason, this paper will focus mainly on open raceway production systems for the intensive production of microalgal biomass.

Open raceway ponds
A number of microalgal species have successfully been grown in open raceway ponds, such as Chlorella, Scenedesmus, Spirulina, Heamatococcus and Nannochloropsis. Commercial outdoor cultivation is mostly restricted to warm tropical and sub-tropical areas, preferably with low rainfall and cloud cover. Detailed construction details are not readily available because of commercial sensitivity, however, basic information can be found for open raceway systems in e.g. [7][8][9], centre pivot circular ponds have been used in Japan and Taiwan [10], and sloping cascade ponds in the Czech Republic [11]. Raceway ponds are by far the industrial choice, followed by horizontally stacked tubular systems.
The basic components of open raceway ponds are an oval basin with a central island around which the cultures are stirred. The basins could be excavated or constructed above ground. An impermeable PVC liner is commonly used to seal the ponds. However, open raceway ponds have been constructed using concrete and cement, fibre-glass and even epoxy-coated concrete. Determining factors that dictate the materials used are costs and the requirements of the specific application. For example Spirulina and Dunaliella are grown in high saline growth media that corrodes concrete over time. Another important consideration is the potential toxicity of the materials used, either for the microalgae as such or the quality of the products produced. Since the basins need to be cleaned from time to time, the materials used should be rigid enough to withstand some abrasive actions and repairs should be simple. Culture depth varies from 10 to 50 cm with 15 cm as the most common [7]. Culture depth (optical depth/light-path) is an important factor because it influences pond construction, biomass density, harvesting costs and pond operation. Culture depths of <15 cm becomes an engineering challenge, especially when pond areas are > 500 m 2 . Any variances can influence the areal density and light penetration, as well as flow and mixing (see below). The deeper the 2. Mixing and the resultant turbulence. This differs markedly between open and closed systems, where much higher rates of turbulence can be achieved in closed tubular reactors, while laminar flow is often a problem in open raceway ponds.

3.
Supply of nutrients and CO 2 , and the prevention of deficit zones.

4.
The biomass concentration or areal density. This determines the in-culture light climate, where a higher biomass will attenuate light energy more and vice versa. This also determines the light acclimated state of the microalgae.

5.
The culture operation being either batch, semi-continuous or continuous.

6.
Temperature within limits as well as the dissolved O 2 build-up in closed reactors.

Light
The production of mega-tons of biomass implicitly means that natural sunlight must be the source of light. Ironically only about 20 to 25 % of photosynthetically active radiation (PAR) from the sun saturates photosynthesis. Furthermore, the action spectrum of photosynthesis has a peak in the blue and red light regions [15], meaning that the green wavelengths have little photosynthetic value. In mass algal cultures light energy and its capture through photosynthesis is complex and it is governed by; 1. The intensity of the light.
2. The quality of the light and the selective absorption of specific wavelengths by the microalgal biomass. 3. The angle of the light impinging on the culture surface.

4.
The condition of the culture surface, e.g. being smooth or rippled.

5.
The optical density of the culture (assuming that only algal mass attenuates light energy).

6.
The movement of individual cells through a light gradient caused by light attenuation in an optically dense culture (turbulence).

7.
The light acclimated state of the culture.  [15]). R d is dark respiration P max the maximum photosynthetic rate, I c the maintenance light energy, α the maximum photosynthetic efficiency, I k the transition light intensity from light limited to light saturated photosynthesis, and I i the light intensity above which photosynthesis is photo-inhibited.
The photosynthetic versus irradiance response curve (P/I) also applies to microalgae and it has been used extensively to describe the response of algae to light energy. Three distinct regions are discernible; i.e. an initial light limited region at low light intensities where photosynthetic rates increase with increasing irradiance, a light saturated region where photosynthetic rates are independent on irradiance, and a region of photo-inhibition in which photosynthetic rates decreases with an increase in irradiance. In the light limited region the rate of photon absorption is correlated with the absorption of light energy and thus the rate of electron transport from water to CO 2 , with the liberation of O 2 . This initial slope or alpha (α) [16] when normalized with chlorophyll a or biomass is the maximum quantum efficiency of photosynthesis. The transition between light limited and light saturated photosynthesis could be gradual or abrupt [17], implying a non-linearity between absorbed light and photosynthetic rates. At light saturation photosynthetic rates reach a maximum (P max , Fig. 1B and 2) irrespective of an increase in light intensity. The transition between light limited and light saturated photosynthesis is donated by I k and it is defined as: At light saturation, the rate of photon absorption exceeds the rate of electron turnover in photosystem II (PS II), without damaging PSII. However, at even higher irradiancies lightinduced depression of photosynthesis occurs commonly referred to as photo-inhibition. Over short term time scales, light induced photo-inactivation of PS II could be viewed as a survival strategy by reducing the number of redundant PS II units. If PS II reaction centres are not repaired through the continuous replacement of the D1 protein, then the damage would become permanent resulting in PS II inactivation [18]. Thus the light-dependent inactivation may be reversible or irreversible. An important finding is that P max was unaffected when PSII reaction centres were reduced by almost 50 % [19]. The reason for this is that the acceptor side of photosynthesis is limited by the capacity of the Calvin-cycle, and the impact of photoinhibition would only become apparent when the numbers of reaction centres are reduced to such a level that the capacity of the Calvin-cycle cannot be met.
Algae (and for that matter all plants) have developed several mechanisms to cope with changes in the quality and intensity of light. In essence the aim of a plant is to acclimate to the prevailing light climate by ensuring that the light reactions exceed the dark photosynthetic reactions. According to Sukenik et al. [20] the Rubisco levels remained relatively constant under varying light regimes, suggesting that the major regulation occurs in the light reactions, especially of PS II. This could be achieved through modulation of the light-harvesting capacity (various photosynthetic pigments) and/or changes in the number and sizes of PS II reaction centres (light harvesting units).
Photo-acclimation affects all components of the P/I curve (Fig. 2), where the open arrows show the direction of the response due to high light and the solid arrows the response due to low light acclimation [15]. The acclimation can take place in time-scales from seconds to hours, depending on the parameters measured [21]. It is important to realize that the microalgae present in a high density algal culture have acclimated to the average light intensity over the entire optical cross section. Furthermore, the microalgae will acclimate during the production cycle in batch cultures, being high light acclimated soon after inoculation of a new culture, to low-light acclimated at the end of the batch process when the cell density is high [22].
There is little that can be done with the above when the scales of mega-mass microalgal production systems are taken into consideration. It may be possible to maintain the biomass in a low light acclimated state by keeping the biomass concentration high. Such acclimated microalgae have high α, low P max at saturating light intensities and low concentration of auxiliary photosynthetic pigments [22]. However, in small-scale production systems, it is possible to utilize the photo-acclimated state of microalgae to achieve very high areal yields. Grobbelaar and Kurano [23] tested a multi-layered flat plate photo-bioreactor, where the microalgae were acclimated to high light in the first layer facing the light source, followed by a layer where the microalgae became progressively more low light acclimated. In this reactor the properties of high and low light acclimated microalgae could simultaneously be exploited and productivities were almost 40 % higher compared to a single layer flat plate reactor of similar optical cross section. Grobbelaar [14] proposed that the key to high productivities lies in turbulence induced short L/D cycles. The three major effects of mixing are:

1.
Prevention of the cells settling to the bottom of the pond. Should this happen, so-called "dead zones" occur where the accumulated of organic material leads to anaerobic decomposition and the release of unwanted substances and metabolites.

2.
To prevent the formation of nutritional and gaseous gradients. Over and above the obvious lowering of oxygen super saturation, mixing would decrease the boundary layer around the cells [23][24]. This would increase the mass transfer rates between the cells and the culture medium for both nutrient uptake and exudation of metabolites. Grobbelaar [26] went on to show that a synergistic enhancement of productivity takes place between decreasing the boundary layer and L/D cycles, with increasing turbulence (see next point).

3.
To move the cells through an optically dense gradient, with variations in the quantity and quality of light energy. In outdoor mass cultures the algae are subjected to the dynamic natural environment, with its variations in the quantity and quality of the light, both diurnally and seasonally, as well as through the mechanical means of mixing.
Kok [27] showed that photosynthesis is influenced by the intensity of light, L/D fluctuations and the ratio of dark time to light time. He used the term 'flashing light' to indicate the light time and it has often been confused with single turnover flashes. He concluded that the light time should be less than 4 ms in order to achieve 'full' efficiency and that the dark time should be at least ten times as long as the light time. For many decades, the enhancement of photosynthesis in intermittent light was interpreted as some residual light energy that was captured during the light time and then utilised in the dark until the next light flash is received. Today we know that this is not possible, because the time-scales for electron turnover in PSII and PSI range from femto-to milli-seconds [14].
Kok [27] also found that the pattern of intermittent light required, for high yields of Chlorella was dependent on turbulence manifested in a flowing culture. The increased production of biomass as a result of increased turbulence was eloquently demonstrated by Laws et al. [28] who placed aerofoil type of devices in the channels with flowing cultures. These caused vortices of 0.5-1 Hz at a flow rate of 30 cm/s, which resulted in photosynthetic conversion efficiencies increasing from an average of 3.7 to up to 10%.
Increasing photosynthetic rates and efficiencies in intermittent light has been shown by several authors, e.g. [21,26] and Grobbelaar [2] found that the increase in specific production rates were consistently exponential with increasing L/D frequencies. The observed increases in productivities with increasing L/D frequencies are directly linked to electron turnover rates in the electron transport chains of photosynthesis. As the L/D frequencies approach the turnover rates, productivities and photosynthetic efficiencies increase until the L/D frequencies match the turnover rates, where production and photosynthetic efficiencies would be at their highest. These rates are in the millisecond range and in practical terms this should form part of the operational considerations when designing photo-bioreactors.
Grobbelaar [14] showed that the photosynthetic rates increased on average 2.1 times at equal L/D cycles when the frequencies were increased from 0.1 to 10 Hz and that the enhancement of photosynthetic rates only becomes significant at cycles > 0.1 Hz. In practical terms this should be achievable in SLP (short light path) reactors [14]. However, special devices would be needed in open raceway ponds, such as aerofoils [28], rippled floors and sides, and curvilinear end geometries.

Photo-inhibition
As discussed above, photo-inhibition is the light-induced depression of photosynthesis when the rate of photon absorption exceeds the rate of electron turnover in PS II. Over the short term, light-induced photo-inactivation of PS II is a survival strategy, whereby the number of redundant PS II units is reduced. Prolonged exposure to high light intensities eventually results in PS II reaction centres not being repaired, through the continuous replacement of the D1 protein. The damage then becomes permanent resulting in PS II inactivation [18]. Whether photo-inhibition occurs in dense mass algal cultures is open for debate, especially in turbulent tubular and plate photo-bioreactors. Congming and Vonshak [29] reported a midday maximum quantum efficiency (Φ max ) depression of dark adapted Spirulina platensis as a result of reaction centre inactivation, which they ascribed to photo-inhibition. Grobbelaar [30] measured a similar midday depression of Φ max on open outdoor raceway Spirulina platensis cultures but concluded that it was as a result of light energy being lost as heat dissipation. This was particularly evident in low density cultures where more than 60 % of the reaction centres became "silent", meaning that they neither reduced Q A , nor returned their excitation energy to the antenna.
Laminar flow is common in large open raceway ponds and Laws et al. [28] found that they had to space their foil arrays 1.2 m apart, because of the rapid dissipation of mixing in the raceway channels. Since raceway channels can be >100 m, it is reasonable to expect photoinhibition to be a factor in the surface layers of the cultures. Grobbelaar et al. [31] modelled microalgal productivity in large outdoor raceway ponds. The model was calibrated against two years of data collected from five raceway ponds differing in surface area from 71 -263 m 2 . In a generalized form the model is written as: Where PROD = net production, PRD = productivity, RES = respiration, and INB = photoinhibition.
PRD is calculated from inputs of the biomass concentration in the culture, culture temperature and light impinging on the surface of the culture. The equation for PRD is: where A 1 -A 3 are constants, X 1 is the biomass concentration in mg (dw) L -1 , I z the irradiance in mol quanta m -2 h -1 at depth z in meters, I s the light half saturation constant and T a temperature factor. Equation (3) has temperature/biomass and temperature/light energy terms. A 1 is the light utilization efficiency, A 2 the Q 10 of photosynthesis and A 3 the Q 10 for the light half saturation.
The component RES includes all losses due to respiration and exudation of organic compounds from the cells and is calculated from the following equation: Photo-inhibition (INB) was included in the model and the equation was constructed such that it only took new production (PRD) into account, that it increases linearly with an increase in irradiance above a threshold irradiance of 1 mol quanta m -2 h -1 , and that temperature affected the overall rate. Photo-inhibition was calculated from:  3D). The exponential decrease in irradiance with increasing depth is seen where all the light energy is absorbed at different depths, depending on the biomass concentration and the resultant attenuation of light energy. Grobbelaar et al. [31] defined the condition where all the light energy is absorbed over the depth profile (optical cross section) as the optimal areal density for maximum productivity (Fig. 2 B). Net productivity typically showed photoinhibition at the surface, where the depth of P max depended on the areal biomass density, ranging from 6 cm at an areal density of 30 g m -2 to 2 cm at an areal density of 120 g m -2 . Below P max productivity decreased as the light energy was attenuated. The impact of photo-inhibition is seen at optical depths deeper than that at P max when the plots of net production (PROD) are compared to the plots where photo-inhibition is excluded (PRD -RES, the dot dot dash lines in Fig. 3). The significant impact that photo-inhibition has on the overall productivity is evident in all four examples (Figs. 3A -3D), when the depth integrated rates are compared (Table 1). Losses due to photo-inhibition are clearly dependent on areal density and light attenuation where it is the lowest at the optimal areal density. At an areal density of 120 g m -2 losses due to photoinhibition can be as high as 66 % (Table 1). Respiratory losses are also dependent on areal density and increases linearly with an increase in areal density. However, as the "dark" zone increases with an increase in areal density a larger percentage of the biomass is lost due to RES ( Fig. 3 and Table 1). For example the total RES loss at a biomass concentration of 200 mg L -1 is 9.6 %, whereas at 800 mg L -1 it is 28 % (Table 1).
With such losses, it is surprising that very little research has gone into eliminating or limiting the losses due to photo-inhibition. Several authors, e.g. [25, 27 and 31], have already reported significant increases in volumetric and areal productivities where mixing was increased. As mentioned above, Laws et al. [28] attributed an increase of up to 10 % in photosynthetic efficiencies in their systems where foils created vortices with rotation rates of about 0.5-1 Hz. However, Grobbelaar et al. [22] showed minimal enhancement at light/dark cycles ≤1 Hz. This was acknowledged by Laws and Berning [33] and they attributed the increased productivity in flumes with foils, rather to the re-suspension of settled cells and breaking down of nutrient gradients. As stated before, Grobbelaar [26] clearly showed how increased turbulence enhanced the exchange rates of nutrients and metabolites between the cells and the growth medium, and this, together with increased light/dark frequencies, synergistically increases productivity and photosynthetic efficiency.