Photobioreactor controller

Abstract

A controller for a photobioreactor for promoting the growth of an organism therein, the photobioreactor being configured to receive one or more input media flows and contain the organism while it synthesises one or more output media, the controller comprising one or more processors configured to: receive an optical density measurement for a sample comprising the organism; estimate a biomass of the organism based on the optical density measurement; and adjust one or more of the input media flows and / or a condition of the photobioreactor in dependence on the estimated biomass of the organism.

Description

[0001] PHOTOBIOREACTOR CONTROLLER

[0002] FIELD OF THE INVENTION

[0003] This invention relates to photobioreactors, for example for the growth of phototrophic organisms such as microalgae, in particular, to control, apparatus for such a photobioreactor.

[0004] BACKGROUND

[0005] Biofuels are emerging as a promising solution to assist the clean energy transition. In particular, algae can produce lipids that are a viable replacement for conventional fuels such as diesel oil, as well as being a low carbon petrochemical feedstock.

[0006] Algal biofuels can be produced using a photobioreactor. A photobioreactor is a vessel which may be open or closed in which algal cells can be cultured. Photobioreactors can transform raw materials such as carbon dioxide, water and solar energy into cell biomass and complex molecules that can be used as substrates for manufacturing biofuels.

[0007] Microalgae are capable of rapid biomass and bio-oil (for example, triacylglycerides) accumulation when supplied with sufficient light energy, inorganic carbon in the form of CO2 / HCO37CO3, organic carbon, and nutrient supplementation (such as nitrogen, phosphorus and vitamins). Photosynthetic production of glyceraldehyde-3 phosphate feeds biomass and bio-oil accumulation. Microalgae are also capable of heterotrophic and / or mixotrophic accumulation of biomass and bio-oil.

[0008] Microalgal photosynthesis involves two interlinked reactions, as shown below. The light dependent rection shown in Equation (1 ) results in the formation of nicotinamide adenine dinucleotide phosphate (NADPH) and adenosine triphosphate (ATP) and consumes 8 photons (400 nm - 700 nm), which fixes approximately 1.83 kg of CO2 for each 1 .00 kg of biomass generated (See Chisti, 2007). The photosynthetic process shown in Equation (2) uses 8 photons (400 nm - 700 nm light), 9 ATP and 6 NADPH per molecule of CO2.

[0009] The light dependent reaction is given by:

[0010] 2H2O + 2NADP++ 3ADP + 3P, + 8 photons 2NADPH + 2H++ 3ATP+3H2O + O2

[0011] (1 ) The light independent reaction is given by:

[0012] C5O3H8 (PO42)2+ CO2[C6O6H7(PO42’)2] 2 C3O3H4 PO42‘ (2)

[0013] The overall rate of photosynthesis is therefore primarily dependent on available concentration of inorganic carbon, CO2 and 400 nm - 700 nm (photosynthetically available radiation) photon flux. The product of photosynthesis, C3O3H4 PO42' glyceraldehyde- 3 phosphate (G3P), is the metabolic precursor to all organic molecules in the microal- gal cell. Under nitrogen replete conditions, G3P is directed to biomass accumulation, protein, lipid and nucleic acid synthesis for cell division. Under nitrogen depleted conditions, G3P is directed via acyl-CoA and lysophosphatidic acid to phosphatidic acid which is converted to diacyl- and triacyl-glyceride.

[0014] The metabolic switch between photosynthetic cell propagation and carbon energy storage metabolism can be controlled by modulating the carbon:nitrogen:phosphorous ratios in the photobioreactor. Under nitrogen / phosphorous or nitrogen replete conditions, cellular metabolism is directed to cell growth, chromosomal duplication and cell division. Under nitrogen / phosphorous or nitrogen limiting conditions and other stress environmental conditions such as high photon flux density, high temperature and high pH [where inorganic carbon is primarily CO3 ], cell growth and division is repressed and cellular metabolism is directed to triacylglyceride carbon-energy storage in intracellular lipid bodies. Algae undergo another metabolic switch from photosynthesis in the presence of light to respiration in darkness. Respiration utilises photosynthetically generated metabolites such as sugars as energy sources for oxidative phosphorylation and the generation of metabolite pools supporting growth. Respiration reduces cell biomass measurably and represents a metabolic drag on biomass accumulation. The amount of loss is increased by ambient night temperature and exponential growth phase.

[0015] Photobioreactors can also be used to produce other algae-derived products, such as industrial, pharmaceutical and agricultural compounds and lubricants.

[0016] One problem of existing photobioreactor systems is the effective control of culture conditions and inputs to and / or outputs from the photobioreactor to promote optimum production of products.

[0017] In the case of microalgae for biofuel production, to optimise the productivity of the organism, the environment of the photobioreactor should preferably be adjusted to support biomass accumulation and bio-oil production under varying conditions, including under nitrogen repleted conditions and nitrogen-depleted conditions; both conditions requiring optimal inorganic carbon and sunlight.

[0018] SUMMARY OF THE INVENTION

[0019] According to one aspect there is provided a controller for a photobioreactor for promoting the growth of an organism therein, the photobioreactor being configured to receive one or more input media flows and contain the organism while it synthesises one or more output media, the controller comprising one or more processors configured to: receive an optical density measurement for a sample comprising the organism; estimate a biomass of the organism based on the optical density measurement; and adjust one or more of the input media flows and / or a condition of the photobioreactor in dependence on the estimated biomass of the organism. In one particular implementation, the one or more processors may be configured to: receive an optical density measurement for a sample comprising the organism; estimate a biomass of the organism based on the optical density measurement; receive a measured value of an incident light level at the photobioreactor; and adjust one or more of the input media flows in dependence on the estimated biomass of the organism and the light level, wherein the one or more of the input media flows comprise CO2. In a further implementation, the one or more processors may be configured to further receive data indicating a salinity level of contents of the photobioreactor. The one or more processors may be configured to adjust one or more of the input media flows in dependence on the estimated biomass of the organism, the light level and the salinity level, optionally wherein the one or more of the input media flows comprise CO2. In one implementation, the one or more of the input media flows may comprise water (for example, seawater). The one or more processors may be configured to adjust the input flow of water in dependence on the salinity level. The one or more processors may be configured to adjust the input flow of CO2 in dependence on the estimated biomass and the light level.

[0020] The one or more processors may be configured to adjust the one or more input media flows to optimise growth of the organism and / or the production of the one or more output media (such as lipids) based on the estimated biomass.

[0021] The one or more processors may be configured to continuously adjust the one or more input media flows and / or the condition of the photobioreactor by repeatedly performing the following further steps: receive an updated optical density measurement for a sample comprising the organism at a respective time; calculate an updated estimated biomass of the organism at that respective time; and adjust the input media flows and / or the condition of the photobioreactor in dependence on the updated estimated biomass of the organism at that respective time.

[0022] The estimated biomass may be an estimated cell dry weight of the organism. The one or more processors are configured to adjust the one or more of the input media flows and / or the condition of the photobioreactor in dependence on the estimated cell dry weight.

[0023] The received optical density may be measured at between 700 - 800 nm incident radiation. The received optical density may be measured at between 700 - 750 nm incident radiation. The received optical density may be measured at between 730 - 750 nm incident radiation. For example, the light from a light source (such as an LED) with an incident wavelength as above that is blocked when the light passes through the sample can be measured and compared to a reference measurement where the incident radiation is passed through air alone.

[0024] The input media flows may be flows of one or more of a gas, a liquid and a solid.

[0025] The one or more input media flows may comprise one or more of water (such as seawater), air, CO2, nitrogen-, phosphorus- or trace-element containing compounds and vitamins.

[0026] The one or more processors may be configured to adjust a condition of the photobioreactor in dependence on the estimated biomass, and wherein adjusting the condition of the photobioreactor comprises adjusting one or more of: a temperature of the photobioreactor, a pH level of contents of the photobioreactor, a salinity of contents of the photobioreactor and a light level (for example intensity or flux) at the photobioreactor.

[0027] The one or more processors may be further configured to adjust one or more of the input media flows and / or the condition of the photobioreactor in dependence on further received data relating to the state of the photobioreactor and / or the organism.

[0028] The one or more processors may be configured to cause the estimated biomass and / or the further data relating to the state of the photobioreactor and / or the organism to be stored at one or more memories communicatively connectable to the controller. The one or more processors may be configured to adjust one or more of the input media flows and / or a condition of the photobioreactor in dependence on historical data relating to the type of organism.

[0029] The one or more processors may be configured to input the received optical density measurement and / or the estimated biomass to a model, wherein the one or more processors are configured to adjust the one or more media flows and / or the condition of the photobioreactor in dependence on an output of the model.

[0030] The model may be a trained artificial intelligence model. The model may be a machine learning model.

[0031] The model may be specific to the organism. The model may be specific to the type or species of the organism.

[0032] The model may be trained using multiple quantitative culture relationship datasets.

[0033] The quantitative culture relationship datasets may comprise one or more of: incident light level and cell dry weight, temperature and cell dry weight, salinity and cell dry weight, CO2 and cell dry weight, pH and cell dry weight, pH and CO2, nitrogen level and cell dry weight and phosphorus level and cell dry weight. The quantitative culture relationship data / or the trained model may be specific to the organism being cultivated (for example, specific to the type or species of the organism).

[0034] The quantitative culture relationship datasets may comprise data acquired at the photobioreactor and previously received by the one or more processors. The datasets may comprise historical data.

[0035] The quantitative culture relationship data may be specific to the organism. The quantitative culture relationship data may be specific to the type or species of the organism. The controller may be configured to control the photobioreactor in dependence on the geographical location of the photobioreactor. The controller may be configured to adjust the one or more media flows and / or condition of the photobioreactor in dependence on the geographical location of the photobioreactor.

[0036] The geographical location may be an input to the model.

[0037] The controller may be communicatively connected with a memory storing a plurality of models. In dependence on the geographical location of the photobioreactor, the controller may be configured to: select a model from a plurality of models; and implement the selected model to control the photobioreactor based on the estimated biomass for the organism. In some implementations, the controller may comprise the memory.

[0038] The controller may have access to a data structure relating values of optical density to corresponding values of biomass for the organism. The data structure may be a lookup table.

[0039] The controller may be configured to adjust one or more of the input media flows in dependence on respective CO2 levels in one or more of (i) gas entering the photobioreactor and (ii) gas leaving the photobioreactor.

[0040] The controller may be further configured to determine, based on the estimated biomass, an optimal time to inoculate the photobioreactor and / or to harvest the one or more output media of the organism.

[0041] The one or more processors may be communicatively connected with a spectrophotometer, wherein the one or more processors are configured to receive the optical density measurement from the spectrophotometer. The one or more processors may be configured to: determine whether the measured light level is above or below a predetermined threshold; and adjust the one or more input media flows in dependence on this determination.

[0042] If the measured light level is below the predetermined threshold, the one or more processors may be configured to stop the flow of CO2 to the photobioreactor.

[0043] The one or more processors may be further configured to: receive data indicating a temperature of the photobioreactor; and adjust the one or more of the input media flows in further dependence on the temperature.

[0044] The one or more processors may be configured to adjust the one or more input media flows in further dependence on one or more of: a pH level of contents of the photobioreactor, the salinity level of contents of the photobioreactor and a culture volume in the photobioreactor media level in the photobioreactor.

[0045] The organism may be a photosynthetic organism. The organism may alternatively or additionally be capable of heterotrophic and / or mixotrophic accumulation of biomass and bio-oil.

[0046] The organism may be a microalgae.

[0047] The one or more processors may be configured to adjust the one or more input media flows in dependence on a measured light level and / or a current time. The one or more processors may be configured to compare the current time with a sunset and / or sunrise time for the geographical location of the photobioreactor. The one or more input media flows may be adjusted in dependence on this comparison. For example, if the current time is between the sunset and sunrise times, one or more of the input media flows may be stopped. For example, the flow of air and / or CO2 to the photobioreactor may be stopped (or continued to be stopped if already stopped). The one or more processors may be configured to determine whether the measured light level is above or below a predetermined threshold. The one or more input media flows may be adjusted in dependence on this determination. For example, if the measured light level is below the predetermined threshold, one or more of the input media flows may be stopped. For example, the flow of air and / or CO2 to the photobioreactor may be stopped (or continued to be stopped if already stopped).

[0048] Adjusting the input media flows may comprise adjusting one or more a quantity, flow rate and a concentration of one or more of the input media flows.

[0049] The controller may be configured to determine, based on the estimated biomass concentration, an optimal time to inoculate the photobioreactor and / or harvest the photobioreactor. When the optimal time to harvest has been reached, the controller may be configured to restrict the flow of air and / or CO2 to the photobioreactor for at least a predetermined dwell period.

[0050] The controller may be configured to adjust oxygen input to minimise night-time respiration of the organism. This may reduce metabolic drag on biomass accumulation providing significant (~7%) saving in cell dry weight biomass lost to respiration.

[0051] The controller may be configured to adjust the one or more input media flows by controlling the operation of one or more valves for controlling the supply of input media to the photobioreactor.

[0052] The controller may be a closed loop controller.

[0053] The controller may comprise a communications unit having a transceiver.

[0054] The controller may be located at the photobioreactor. The controller may be remote from the photobioreactor. The controller may be configured to additionally control multiple other photobioreactors in a network of photobioreactors. In some implementations, sensors at the photobioreactor may communicate with a central controller remote from the photobioreactor. The central controller may receive sensor data from multiple photobioreactors and control those bioreactors.

[0055] The controller may be configured to receive data from the photobioreactor and each of the multiple other photobioreactors and send control signals to respective ones of multiple local controllers of the photobioreactor and multiple other photobioreactors in the network to adjust one or more of (i) one or more input media flows of the respective photobioreactor; and (ii) the condition of the respective photobioreactor in dependence on the respective received data.

[0056] The controller may be a controller for a master photobioreactor in a master-slave network of photobioreactors. The one or more processors may be configured to: receive data relating to a state of the master photobioreactor and / or the organism contained within the master photobioreactor; and adjust one or more of (i) the input media flows to the master photobioreactor and one or more slave photobioreactors; and (ii) a condition of the master photobioreactor and the one or more slave photobioreactors in dependence on the received data.

[0057] The data relating to a state of the photobioreactor and / or the organism may be collected. The data may comprise data indicating one or more of: a culture volume within the photobioreactor (including the volume of the culture media, including water and any nutrients, and the organism), a temperature of the photobioreactor, a pH level of contents of the photobioreactor, a salinity of contents of the photobioreactor, a light intensity on or within the photobioreactor, a colour density of the organism in the photobioreactor (i.e. a property of the colour of the organism), a media level in the photobioreactor, a respective CO2 level in gas entering the photobioreactor and / or in gas leaving the photobioreactor, a respective nitrogen level in gas entering the photobioreactor and / or in gas leaving the photobioreactor, a respective oxygen level in gas entering the photobioreactor and / or in gas leaving the photobioreactor, a nitrogen level of contents of the photobioreactor, a phosphorous level of contents of the photobioreactor, a salinity level of the contents of the photobioreactor, a quantity of a media entering or leaving the photobioreactor and the culture time, and previous datasets on culture performance. The controller may have access to historical culture performance data, which may be used as the quantitative culture datasets described above. For example, one or more of historical light level, temperature, salinity or CO2 supply relative to optical density / biomass and / or triacylglyceride production. The one or more processors may be configured to adjust the one or more input flows and / or condition of the photobioreactor (for example, CO2 sparging) in dependence on the historical data. The one or more processors may be communicatively connected with a memory storing the historical data.

[0058] According to a further aspect, there is provided a system comprising a photobioreactor and a controller as described above. The photobioreactor may be communicatively connected with the controller.

[0059] The controller may be configured to receive the data relating to the state of the photobioreactor and / or the organism from one or more of the following sensors: a salinity sensor, a single turnover active fluorometry sensor, a temperature sensor, a turbidity sensor, a pressure sensor, an optical sensor, a chemical sensor, a conductivity sensor, a fluid flow meter, a light sensor, a moisture sensor, a pH sensor, a liquid sensor, a conductivity sensor, a carbon dioxide concentration sensor, a nitrogen concentration sensor, a phosphorous concentration sensor and an oxygen concentration sensor. The sensor(s) may be located inside the photobioreactor. The sensor(s) may extend into the media contained in the photobioreactor (for example, into the culture medium containing the organism) or may operate without the culture volume such as within the headspace / offgas or input or output piping.

[0060] The one or more processors may be configured to receive the data (optical density measurements, data indicating light level, salinity level, culture volume measurements, etc) at intervals of, for example, between 1-100 ms. The one or more processors may be configured to adjust the input media flows and / or the condition of the photobioreactor at intervals of, for example, between 1 -100 ms. According to a further aspect, there is provided a computer-readable storage medium having stored thereon computer readable instructions that when executed at a computer system comprising one or more processors cause the one or more processors to perform the method above. The computer-readable storage medium may be a non-transitory computer-readable storage medium.

[0061] BRIEF DESCRIPTION OF THE FIGURES

[0062] The present invention will now be described by way of example only, with reference to the accompanying drawings, in which:

[0063] Figure 1 schematically illustrates an exemplary system for a photobioreactor;

[0064] Figure 2 schematically illustrates inputs to and outputs from the photobioreactor system in more detail;

[0065] Figure 3 schematically illustrates the use of local and remote controllers for controlling photobioreactors;

[0066] Figure 4 schematically illustrates a control computing system of an exemplary photobioreactor controller;

[0067] Figure 5 schematically illustrates exemplary apparatus for measuring the optical density of an organism;

[0068] Figure 6 illustrates the relationship between optical density measured using 750 nm incident radiation and cell dry weight;

[0069] Figure 7 schematically illustrates the sensor inputs to the photobioreactor controller and exemplary datasets used by the controller; Figure 8 schematically illustrates how dynamic control models are generated from real-time and historical culture performance, biological and environmental data.

[0070] Figure 9 shows the steps of an exemplary method for controlling a photobioreactor.

[0071] Figure 10 shows a schematic decision flow for night-time shutdown of air and CO2 sparging and recommencing daytime photosynthetic control.

[0072] DETAILED DESCRIPTION

[0073] A photobioreactor refers to any cultivation system designed for growing an autotropic or heterotrophic photosynthetic unicellular or multicellular organism. A photobioreactor can be used in the production of biofuels, such as biodiesel. A photobioreactor may define a volume or reactor chamber in which organisms can grow, reproduce and / or synthesise one or more output media. Any suitable organism may be grown in the photobioreactor. The organism may be an algae. The organism may be a microorganism such as a microalgae. The photobioreactor may be configured to contain the organism while it is propagating or otherwise living. The photobioreactor may be open (i.e. in communication with the atmosphere, for example a pond or raceway, or open to the atmosphere via a line or valve) or closed (i.e. not in communication with the atmosphere). The photobioreactor contains matter comprising the organism. The organism may be contained in a culture medium. The culture medium may comprise, for example, water and nutrients. The culture medium may comprise non-sterile seawater. In some examples, no freshwater may be used in the process. The nutrients may comprise one or more fertilisers, which may contain biologically accessible sources of nitrogen (N) and phosphorus (P). Other fertilisers may include potassium (K) and other trace elements. While contained in the photobioreactor, microalgae may synthesise one or more output media. The one or more output media may comprise one or more output molecular products. The output media may comprise a biofuel product. The output media may be a lipid, such as a triacylglyceride. In some examples, the photobioreactor may comprise an enclosure. The enclosure can contain the organism being cultivated. The photobioreactor may have walls that enclose the reactor chamber. Some or all of the walls may be translucent so that external light can pass through the walls and enter the reactor chamber. The external light is preferably sunlight. However, artificial light may also be used. In some examples, the walls may be transparent. The light may promote photosynthesis of organisms within the chamber.

[0074] The photobioreactor may have one or more openings. The opening(s) may permit input media such as water, growth media, gases and nutrients to be introduced into the photobioreactor. The input media can be supplied as flows of solid, liquid or gaseous media. One or more of the input media flows may comprise a solid dissolved in a liquid. The input media flows may be flows or one or more gases, which may be a mixture of gases. The opening(s) may also permit organisms that are to be cultivated in the photobioreactor to be introduced into it. The opening(s) may be located at the top or the side of the photobioreactor. The opening(s) may also permit culture media and / or organisms to be removed from the photobioreactor.

[0075] Empty space above the culture medium biofuel product synthesised by an organism inside the photobioreactor, which may be referred to as the headspace, can act as a venting space through an opening in the photobioreactor to the atmosphere and so prevent levels of oxygen in the system from rising to levels at which the photosynthetic activity of the organism is impeded through inhibition of RuBisCo. Such oxygen in the system may be both dissolved in the culture medium and present within gas inside the photobioreactor.

[0076] Gases supplied to the photobioreactor may encourage mixing of the contents of the photobioreactor and / or encourage one or more reaction products to float to the surface of the liquid contained within the photobioreactor.

[0077] Atmospheric, recovered or flue gas-derived CO2 dissolved in seawater spontaneously forms HCOs' which is transported into the cells of the organism where, using chemical energy derived from light in the chloroplast, it can be converted into biomass and triacylgylceride bio-oil. Uptake of CO2 from industrial flue gases confers a carbon negative footprint on the fuel.

[0078] Figure 1 shows a schematic diagram of an exemplary system 100 showing the inputs and outputs to the system. The photobioreactor 106 may comprise a single photobioreactor unit or an assembly of individual photobioreactor units. The inputs and outputs to the photobioreactor 106 may be controlled by a controller (described in more detail below). In this example, the photobioreactor 106 receives the following inputs: CO2 101 , water 107, nitrogen 114, phosphorous 115, trace element (TE) supplements 116, air 119 and photosynthetic organism inoculum 123. The water 107, nitrogen 114, phosphorous 115 and TE supplements 116 make up the culture medium. The culture medium can be mixed and stored at a tank 110 before injection into the photobioreactor, or the components of the culture medium can be individually injected into the photobioreactor via their individual supply lines.

[0079] CO: 101 can be input from the atmosphere, from direct air capture, flue gas from biological, fossil fuel and / or other industrial sources (e.g. a gas separation plant, petrochemical plant CO2 emissions, concrete manufacture emissions or other CO2 / HCO37CO2' containing industrial effluent). The input CO2 can be stored at a storage tank 102 and then supplied via a supply line such as CO2 pipe 103. In some cases, for example where input CO2 is derived from hot flue gases or similar emissions, a pump 104 and heat exchanger 105 can be used to pressurise and cool the CO2 gas respectively.

[0080] The water 107 may be one or more of seawater, brackish water and freshwater. The type of water may in some cases be selected based on the type of organism. In this example, the water is pumped from the water source 107 using a water pump 108 and passed through an optional filtration unit 109. The nitrogen 114, phosphorus 115 and TE supplements 116, which may be stored at respective tanks, can be pumped to the culture medium storage tank 110 using respective pumps 111 , 112, 113.

[0081] The culture medium is fed from tank 110 via a supply line 117 to the photobioreactor system 106. The supply line 117 may comprise a pump 118. Alternatively, the photobioreactor may be gravity-fed.

[0082] The air 119 may be injected to the photobioreactor system 106 to assist mass transport. The air is fed to the photobioreactor 106 via a supply line 112. The supply line 112 may have a gas pump 120 and / or a filtration unit 121.

[0083] The photosynthetic organism inoculum 123 can be injected into the photobioreactor 106 via a supply line 125. A pump 124 may be used to assist the supply of the photosynthetic organism inoculum 123 to the photobioreactor 106.

[0084] In this example, the photobioreactor 106 generates two outputs: oxygen rich off-gas 126 produced from photosynthetic cleavage of H2O to H2 + >2 O2; and biomass in culture medium. Output biomass in culture medium can be output from the photobioreactor 106 via harvest line 127, with an optional harvest pump 128 and may be stored for subsequent downstream processing at a tank 129.

[0085] As described above, the photobioreactor array is supplied with a culture medium which may be comprised of water and nitrogen-containing, phosphate-containing and / or trace element-containing supplements. The photosynthetic organism 123 can be inoculated via inoculate pump 124 and an inoculum pipe 125. Under suitable light flux (for example, photosynthetically available radiation in the range 400 nm - 700 nm) and in the presence of CO2 and water, photosynthesis can generate an oxygen-rich off-gas 126 which can be released from the photobioreactor via a pipe and optionally stored. Cultivation of photosynthetic organisms in the photobioreactor array can capture input CO2 and convert it to biomass and biomass-related products, which in this example includes triacylglycerides. The culture can be harvested via harvest pipe 127 and harvest pump 128 to biomass / culture storage tank 129.

[0086] In this example, the photosynthetic organism may be a freshwater / brackish water / seawater microalga and the products may be protein-rich biomass, sugar-rich biomass and lipids including triacylglycerides.

[0087] Figure 1 schematically illustrates the control of some inputs to and outputs from a photobioreactor 106, as well as the monitoring of conditions in the photobioreactor, in an exemplary system 200. Cultivation initiates with inoculation of the photosynthetic organism 230 via inoculation pipe 231 and inoculation valve 232, which introduces the photosynthetic organism into the photobioreactor culture medium 233. This is supplied via inoculum pipe 234 and valve 235. A source of CO2 236 is supplied via CO2 pipe 237 and valve 238. Mass transfer can be assisted by means of air-lift. Filtered air 239 can be supplied via air pipe 240 and valve 241 . Culture volume is controlled by addition of water 242, which may be fresh water, brackish water or seawater, via water pipe 243 and valve 244. The values 232, 235, 238, 241 , 244 may be motorised and controlled by the controller to control the input flows to the photobioreactor.

[0088] In this example, culture performance is monitored by an optical density (OD) sensor

[0089] 245, as will be described in more detail below, and the following further sensors at the photobioreactor: a light sensor 247, a thermal (temperature) sensor 249, a conductivity sensor 251 , a pressure sensor 253 and a CO2 sensor 255 which can measure the CO2 level in the headspace of the photobioreactor above the culture medium / biomass. The outputs of these sensors are OD (for example, measured at 700-800nm incident light)

[0090] 246, light flux (klux or photons. s-1. nr2) 248, culture medium temperature (°C) 250, culture medium salinity 252, culture medium level 254 and CO2 level 256 respectively.

[0091] Control of photobioreactor culture may be based on one or more of the measured outputs 246, 248, 250, 252, 254, 256. Generally, the sensors may sense the state of the photobioreactor and / or the state of an organism in the photobioreactor and exemplary sensors may include a temperature sensor, a sunlight sensor, a fluid flow meter, a conductivity sensor, a moisture sensor, a liquid sensor, a turbidity sensor, a CO2 concentration sensor, a single turnover active fluorometry (STAF) sensor, an optical sensor for sensing the colour of organisms in the photobioreactor, a salinity sensor (which may measure salinity using electrical conductance or other suitable techniques), a sensor for measuring optical density (OD) of the organism (such as a spectrophotometer) and a light sensor for measuring photon flux (for example in lux / pM photons. s’1.rrr2). The sensors may extend into the media contained within the photobioreactor or may operate without the culture volume such as within the headspace / offgas or input or output piping.

[0092] Generally, the sensors are configured to acquire data relating to the state of the bioreactor and / or the state of the organism. The data relating to a state of the photobioreactor and / or the organism may be data indicating one or more of: a temperature of the photobioreactor, such as a temperature of the culture medium for the organism in the photobioreactor, a pH level of contents of the photobioreactor, such as a pH level of the culture medium, a salinity of contents of the photobioreactor, such as a salinity of the culture medium, a light intensity on or within the photobioreactor, such as a light intensity on or within the culture medium, a colour density of the organism being cultured in the photobioreactor (i.e. a property of the colour of the organism), a media level (such as liquid level, such as a level of the culture medium) in the photobioreactor, a respective CO2 level in gas entering the photobioreactor and / or in gas leaving the photobioreactor, a respective nitrogen level in gas entering the photobioreactor and / or in gas leaving the photobioreactor, a respective oxygen level in gas entering the photobioreactor and / or in gas leaving the photobioreactor and a quantity of a media entering or leaving the photobioreactor, an incident light level at the photobioreactor, a level or volume of matter contained in the photobioreactor (for example, in comparison to a maximum volume of the photobioreactor), a property of a media or product contained within the photobioreactor and a liquid level in the photobioreactor. The data relating to a state of the organism may be data relating to one or more of the following: an optical density (OD) of the organism, a biomass of the organism, a colour of the organism, a pH level of the organism or its surrounding medium in the photobioreactor, a concentration of the organism, a colour density of the organism (i.e. a property of the colour of the organism) and the incident light level on the organism. The data relating to the state of the organism may include data relating to an output product of the organism, such as biomass-related products such as triacyglycerides / lipids.

[0093] As shown at 210, the photobioreactor system 200 is communicatively connectable with a controller 210. The controller 210 comprises one or more processors, such as processor 211 , and a memory 212 storing in non-volatile form program code executable by the processor(s). The memory may also store data collected by the sensors, which may be used in the data sets used to train control models, as will be described in more detail later. The controller may comprise a solar panel and battery system to allow it to operate in the absence of an electrical grid connection. Alternatively or additionally, the controller may be coupled to an electric grid supply.

[0094] In addition, the controller can implement algorithms that can access locally or remotely stored sensor output data (shown at 257) that correlate culture biomass and output product (for example, triacylglyceride) production performance with controller control inputs to control the input flows 230, 233, 236, 239, 242. The data set 257 may be stored at memory 212 or another memory of the controller.

[0095] The controller 210 may be located at the photobioreactor 106 or at a remote location. The photobioreactor may in some cases comprise the controller.

[0096] Figure 3 schematically illustrates the use of local 259 and remote 258 controller systems. In the local system, the controller 259 may be co-located with the photobioreactor and communicate with the photobioreactor sensors (for example sensors 245, 247, 249, 251 , 253 and 255). Alternatively, the controller 258 may be remote from the photobioreactor and may communicate with distal sensors at the photoreactor, for example via standard communications protocols. The communication between the controller and the photobioreactor may be wired or wireless. For both local and remote controllers, the controller is communicatively connectable with the photobioreactor and its associated components (such as valves and sensors). The controller receives inputs from the sensors (245, 247, 249, 251 , 253, 255) and based on control algorithms and on current and historical data 257 can operate valves (for example, 232, 235, 238, 241 , 244) to adjust inputs (for example, 230, 233, 236, 239, 242) into the photobioreactor 106.

[0097] Thus, in dependence on the data 257, the controller may control one or more valves of the control unit to regulate a flow of media into or out of the photobioreactor. The memory may also store data collected by sensors, which may be used in the data sets used to train control models, as will be described in more detail later.

[0098] Figure 2 schematically illustrates the controller 210 (which may be a local 259 or remote 258 controller) in more detail.

[0099] In this example, the controller 210 comprises a central processor 211 , a communications processor 213, a memory storing the historical data set 257 and CPU memory 212.

[0100] Sensor inputs from the photobioreactor sensors (such as sensors 245, 247, 249, 251 , 253, 255) can be used to control the operation of valves (such as valves 232, 235, 238, 241 , 244) to adjust the flow of input media to the photobioreactor.

[0101] The controller may control parameters that are influential to the growth of the organism in the photobioreactor and / or its synthesisation of one or more output media, such as a biomass, triacylglyceride / lipids and other metabolic products and molecules. The controller may control the conditions within the photobioreactor. The controller may alternatively or additionally monitor conditions within the photobioreactor. The controller may receive data from one or more sensors such as sensors 245, 247, 249, 251 , 253, 255. As described above, the photobioreactor may comprise the sensors or may be communicatively connectable with them. The data collected from these sensors may be used to control the photobioreactor, as will be described in more detail below.

[0102] Returning to Figure 2, one or more supply lines, such as supply lines 231 , 234, 237, 240, 243, may be configured to supply one or more input media to the photobioreactor 106. The one or more input media preferably comprise one or more liquids or gases. The one or more input media may comprise one or more substrates. The one or more input media may comprise solid components, for example, a solid fertiliser. The solid components may optionally be suspended in a liquid. Each supply line may comprise a conduit. The input media can be supplied as one or more input media flows. As described above, each supply line may comprise a valve 232, 235, 238, 241 , 244. The valves control the flow of the respective media into the photobioreactor 106. The valves may be controlled from the controller in dependence on data derived from sensors in the photobioreactor 106. The supply lines deliver their respective input media to the photobioreactor 106 in quantities or rates controller by the controller, and / or during time periods controlled by the controller. The input media supplied to the photobioreactorl 06 may be controlled by the controller. The controller may control the quantities and / or rates of the one or more input media supplied to the photobioreactor 106.

[0103] In this example, the photobioreactor 106 is configured to receive one or more gases as input media. The one or more gases may be received as one or more input media flows. In other implementations, one or more of the input media may not be gaseous. In some implementations, the photobioreactor may be configured to receive CO2 as dissolved CO2 in a liquid input flow. Where the input flows comprise air and CO2, the air and CO2 may be supplied via different conduits (as in Figure 2) or via the same conduit. In some cases, the photobioreactor may receive air without additional CO2. Although atmospheric air has a CO2 content (~420 ppm), the addition of supplemental CO2 has been found to increase the growth rate of photosynthetic organisms such as algae. For example, the growth rate of some algae has been seen to double on the addition of 5% CO2 (vol / vol per unit time, 1 minute). In other embodiments, up to 100% CO2 may be used. The gas input to the photobioreactor can agitate its contents, in particular the culture medium containing the organism. This can help to keep the organism cells suspended and expose them all approximately equally to the light energy that drives photosynthesis. In a photobioreactor used to cultivate microalgae, the microalgal cells can be mixed during culture via the air and\or CO2 bubbling into the culture media. The gas can also provide carbon to the cells of the organism which allows them to grow.

[0104] Gaseous products or excess supplied gases in the photobioreactor can be exhausted outside out the photobioreactor chamber through a vent line. The vent line may comprise a CO2 sensor configured to measure the CO2 concentration of the gas being exhausted from the photobioreactor. The vent line may also comprise sensors for measuring the concentration of other gases such as nitrogen and oxygen in the gas exhausted from the photobioreactor. There may also be sensors located at the exit of one or more of the supply lines such as lines 231 , 234, 237, 240, 243 to measure the concentration of gases being supplied to the photobioreactor 106.

[0105] An array of multiple photobioreactors may be arranged around a central pump station. Each photobioreactor may be connected with the pump station for the supply of the input media. The harvesting and exchange of the microalgae and culture medium can be effected via tubes between the photobioreactors and the pump station, and from the pump station to a processing plant. Multiple photobioreactors may be coordinated via a wireless communications infrastructure.

[0106] In the exemplary implementation shown in Figure 2, a pressure sensor 253 is configured to measure the level of matter in the photobioreactor 106. The pressure sensor may measure the total level of matter in the photobioreactor (for example water, the organism, and any product of the organism). The controller may determine how full the photobioreactor is in dependence on data from the pressure sensor 253 and a known maximum volume of the photobioreactor. The controller may adjust the flow of additional input media to the photobioreactor in dependence on the determined level, which can prevent the photobioreactor from becoming too full and / or overflowing.

[0107] An outlet control valve may be configured to control the flow of media exiting the photobioreactor 106. The valve can be controlled to direct the flow of media exiting the photobioreactor to multiple lines. Each line may comprise a conduit. In some cases, the valve may control the flow of media to either a harvest line (such as line 127 in Figure 1 ) or a dump line. The dump line may be used in the event that a catastrophic event occurs, such as an infection, attack from predation, or poisoning of the system due to contamination. In this event, the controller may connect the biomass line to the dump channel so that the infected culture medium can be disposed of.

[0108] The controller may control the valves 232, 235, 238, 241 , 244 for controlling the flow of input media into the photobioreactor. Those valves may be controlled from the monitoring station or automatically by the controller itself in dependence on data derived from the sensors 245, 247, 249, 251 , 253, 255.

[0109] The controller may also adjust one or more conditions of the photobioreactor. Adjusting the condition of the photobioreactor may comprise adjusting one or more of: a temperature of the photobioreactor, a pH level of contents of the photobioreactor (such as the pH level of the culture medium), a salinity of contents of the photobioreactor (such as the salinity of the culture medium) and a light intensity / light path at the photobioreactor.

[0110] In one example, the controller may be configured to adjust one or more of the input media flows in dependence on respective CO2 levels in one or more of (i) gas or media entering the photobioreactor and (ii) gas or media leaving the photobioreactor. The controller may be configured to adjust the input flow of CO2 (for example in gas and / or dissolved CO2 in a gaseous or liquid flow) to the photobioreactor in dependence on a determined level of CO2 in gas leaving the photobioreactor. As described above, the photobioreactor may comprise one or more exhaust conduits via which gas may be output from the photobioreactor. The photobioreactor may comprise a sensor 255 configured to determine the CO2 level in the gas leaving (i.e. being output from) the photobioreactor via the exhaust conduit. The controller may be configured to receive data from the sensor 255 indicating the CO2 level in the gas leaving the photobioreactor. In response to detecting that the level of CO2 in the gas leaving the photobioreactor is greater than a predetermined value, the controller may be configured to reduce the amount of CO2 supplied to the photobioreactor as an input flow.

[0111] As described above, sensor 245 is configured to measure the OD of a sample comprising the organism. The OD measures the ability of the organism to slow or delay or block or absorb the transmission of light. OD measurements describe a logarithmic measure of the percentage transmission of light through a sample. The OD is calculated as the logarithmic ratio of the radiation incident on the sample to the radiation transmitted through the sample. The radiation incident on the medium may be determined as a reference value. The OD is therefore a logarithmic measurement of the percent transmission (%T) and it can be represented by the equation, A = Iog10 100 I %T. A sample with 1 OD allows 10% of light to be transmitted through the sample.

[0112] This value can be used to infer the concentration of microalgal cells in the sample, which can be used to infer the cell dry weight (CDW) of the organism. The OD of the organism can be measured using a spectrophotometer. A spectrophotometer measures the number of photons absorbed or blocked when the light passes through a sample. This can be determined using an ultra-violet (UV), visible or infrared (IR) spectroscopy. The OD can be measured at a particular wavelength of incident light. For example, the OD of a sample for incident light of 750nm is referred to as OD750. In the examples described herein, measurements of the OD may be OD 700-750, more preferably 730-750 (OD730-OD750).

[0113] The sample comprises cells of the organism. In a preferred implementation, the sample comprising the organism comprises culture medium. The sample may have a smaller volume than the total volume of matter in the photobioreactor. The sample may be extracted from the photobioreactor to take the OD measurement. The sample may be a sample of culture broth comprising cells of the organism and culture medium. The sample comprising the organism may comprise a cell suspension (i.e. a suspension of cells of the organism in a liquid medium, such as the culture medium).

[0114] A schematic diagram of the OD measurement process using a spectrophotometer 500 is shown in Figure 5. Light sources are shown at 564 and 566. The light source may emit light in the visible, UV or IR spectra or a part thereof. The light source may be, for example, a light-emitting diode (LED). The light beam 571 emitted from reference light source 564 passes through air or some other medium devoid of the sample. The amount of light transmitted through the air or other medium is detected by the reference light sensor 565. The light beam 572 emitted from sample light source 566 passes through a sample holder 550 which contains the sample. In this example, the sample comprises cells of the organism in a medium, which may be the culture medium. This may allow the measurement to be made by extracting a small amount of culture medium and biomass from the photobioreactor. Extraction may be intermittent or continuous. The sample holder may be a transparent tube. The amount of light transmitted through the sample is detected by the sample light sensor 568. Light flux can be measured over a set time and linear regression analysis of reference and sample can be used to determine biomass against a standard curve of light flux against standard known cell dry weight cell cultures.

[0115] The OD measurements therefore measure how much of the incident light is blocked by the culture medium / biomass as it passes through the sample as compared to a reference light path of the same wavelength passing through air or medium without the sample analyte.

[0116] The arrangement shown in Figure 5 is exemplary and the OD may alternatively be determined using different setups or instruments.

[0117] The controller receives the value of the determined OD for the sample. This can be correlated to the microalgae biomass measured as cell dry weight (CDW) in g.L-1to infer a value of instantaneous biomass for the organism. The OD measurement can be made continuously. The OD measurements may be made at time intervals of, for example, 100 milliseconds.

[0118] The correlation between the values of OD and dry cell weight may be based on a previously collected dataset. The cell dry weight may be determined from a sample of cell suspension. For example, cells can be recovered from the suspension by centrifugation at 5,000 g for 10 minutes at 30°C, in pre-weighed centrifuge tubes. Following overnight drying in a 40°C hot air oven, the dry biomass is determined by a balance (for example, to 4 decimal places). This process can be time consuming (24 hrs). Therefore, the measurement of OD, which is much quicker, can be used to infer the biomass based on previously obtained data.

[0119] Figure 6 shows an example of microalgal CDW vs photobioreactor culture optical density at 750 nm. The plot shown in Figure 6 illustrates that this relationship is stable and can be used as a reliable index of biomass. Linear regression analysis can be performed on the dataset. In the example shown in Figure 6, linear regression analysis was performed on the OD measured at 750 nm (x) plotted against the biomass concentration in the form of the CDW of the organism (y) measured in g.L-1to give a relationship of the form f(x)=m*x+b. In this example, m and b were determined as follows: m = 4.98502 + / - 0.02264 (0.4542% asymptopic standard error); b = 0.0100369 + / - 0.0149 (148.4% asymptoptic standard error).

[0120] Therefore, based on measurements of OD of the organism in the photobioreactor, the instantaneous cell biomass and growth stage of the organism can be inferred. The spectrophotometer 500 can measure the optical density of the sample against a reference light path. The suspended cells in the sample block light leading to a lower light flux in the sample as compared to the reference (see Figure 5). The CDW of the sample can be determined from a dataset of light flux values versus known cell dry weight cell culture suspensions. Amongst other parameters, the production of biomass and biomass-related products including triacylglycerides / lipids is driven by photon flux, inorganic carbon and N and P levels. Cultivation of the organism can be optimised through one or more of the following: maintenance of a constant light path and / or photon flux; control of inorganic carbon (CO2 / HCO37CO3) which may be supplied through a metered sparging system or through a metered inorganic carbon saturated medium; controlled mass transfer by variable air lift sparging rates; and adjusting the N and P levels in the culture medium, for example through the addition of N and P compounds. The input flows to the photobioreactor and / or the conditions in the photobioreactor can be optimized based on the inferred cell biomass concentration.

[0121] Values of OD measured by the spectrophotometer or other suitable apparatus can be periodically sent to the controller. As discussed above, this value can be used to infer the cell biomass concentration in the form of CDW. The control system can also receive measured values of other parameters such as the incident light level at the photobioreactor, the temperature of the culture medium containing the organism and the salinity of the culture medium containing the organism.

[0122] In one implementation, as illustrated with reference to Figure 7, culture performance is periodically and / or continually monitored using OD sensor 245, light sensor 247, a temperature sensor 249, conductivity sensor 251 , pressure sensor 253 and CO2 sensor 255. In other implementations, different sensor types may be used.

[0123] Data 257 may also be collected by the sensors at the photobioreactor and stored in a data set of one or more quantitative culture relationships. The historic data can be used to predict future relationships. Data sets collected from the bioreactor (or other bioreactors) may be used to train a model to predict the optimal supply for media to the photobioreactor.

[0124] For example, as shown in Figure 7, the data sets 257 of quantitative culture relationships may comprise any of the following: OD-CDW (770), light-CDW (771 ), tempera- ture-CDW (772), salinity-CDW (773), CO2-CDW (774), pH-CDW (775), pH-CO2(776), N-CDW (777), P-CDW (778), N-Bio-oil (779), P-Bio-oil (780), CO2-bio-oil (781 ), CDW- bio-oil (782), salinity-bio-oil (783), light-bio-oil (784), temperature-bio-oil (785), pH-bio- oil (786), time-CDW (787), time-bio-oil (788) and geographic data (789). The field ‘biooil’ may be the bio-oil quantity (for example of triacylglyceride) in g or kg as a fraction of total CDW. The geographic data may comprise data indicating parameters such as sunrise time and sunset time, latitude, longitude and altitude above sea level for specific geographical locations.

[0125] Control routines for the photobioreactor may be derived based on one or more of the data sets 770-789. In some implementations, machine learning can be used to identify optimal culture control scenarios. In one example, to effect automated closed loop control, the controller can use a programmable decision tree based on vector machine- optimised strategies derived from the data sets of quantitative culture relationships, such as one or more of the data sets 601-611 .

[0126] As mentioned above, the memory of the controller may store one or more models. Alternatively or additionally, the memory may store one or more look up tables containing optimal control parameters given current measured sensor values. The models may be trained artificial intelligence models. The models may be trained machine learning models, such as neural networks. The models may be trained based on historical data collected by the sensors, at the same photobioreactor or one or more other photobioreactors. The models can be trained using one or more of the data sets 770-789. The machine learning step can generate the trained modes and / or look up tables. Different models and / or look up tables may also be generated for different geographical locations. In dependence on the geographical location of the photobioreactor, the controller may be configured to select a model from a plurality of models (which may be stored at a memory communicatively connected with the processor of the controller) and implement the selected model to control the photobioreactor based on the estimated biomass for the organism determined by OD measurement. As mentioned above, the controller may comprise a communications module 213. The communications module 213 may comprise a wired or wireless transmitter for transmitting data to a remote monitoring station and / or central controller unit for an array of photobioreactors. The controller may receive new or updated control models from an external source via the communications module 213. For example, models may be trained in the cloud and sent to the controller.

[0127] The one or more models may be implemented by the one or more processors of the controller to modulate processes such as photobioreactor mass-transfer (e.g. air sparging) and CO2 injection with environmental conditions (such as light, temperature, salinity, N / P) and / or to give an indication of biological parameters (such as biomass and culture phase) to manage the photobioreactor production run from inoculation to harvest.

[0128] Data measured by the sensors (such as sensors 245, 247, 249, 251 , 253 and 255) can be input to the model, which may output adjusted values of one or more input flows to the photobioreactor and / or conditions of the photobioreactor. The primary input to the model may be the measured OD, which can be used to infer CDW, as discussed above. This can give an indication of the growth stage of the organism. The conditions in the photobioreactor can be adjusted accordingly in dependence on the growth stage of the organism.

[0129] In dependence on the output of the model, the controller can adjust conditions such as mass transfer (air-sparging), CO2 injection and N / P content of the culture in response to changes in environment and culture parameters and control the culture from inoculation to harvest.

[0130] An overview of this process in schematically illustrated in Figure 8. Dynamic control models are generated from real-time and / or historical culture performance, biological and environmental data. The historical datasets 770-789 are used to train a model 800 which is implemented by one or more processors of the photobioreactor controller 210. Based on current values measured by one or more of the sensors 245, 247, 249, 251 , 253 and 255, which are provided as input to the model, the controller can control one or more the valves 232, 235, 238, 241 , 244 to adjust one or more of the input media flows to the photobioreactor 106. The one or more valves may be controlled based on the output of the model. One or more conditions of the photobioreactor may also be adjusted in dependence on the output of the model. For example, if the temperature in the photobioreactor is found to be below its optimal level for growth given the current measured parameters, an artificial heat source may be activated.

[0131] One output of the model may also provide an indication of the culture phase. For example, the output of the model may indicate that the organism is in the inoculate phase or the harvest phase. Appropriate action may then be taken to inoculate the bioreactor or harvest the product respectively at an appropriate time.

[0132] When the output of the model, in dependence on the sensor data, indicates that the photobioreactor should be inoculated (i.e. is in the inoculate culture phase), water (such as filtered seawater) can be added to the photobioreactor via the water line 243 to a predetermined level. This level may be, for example approximately initially 15- 25% of the full depth of the photobioreactor. As the culture progresses, the culture medium level will be increased to full capacity. The pressure sensor 253 may measure the level of the media currently in the photobioreactor and send data indicating the current level to the controller. The controller may adjust the quantity or flow rate of water supplied to the photobioreactor, for example by adjusting the valve 244 on the water line 243. A flow of nutrients to the photobioreactor may also be adjusted.

[0133] The organism may have different requirements for CO2 and other inputs depending on its growth stage. As mentioned above, by measuring the OD of the organism, which can be used to infer the CDW of the organism and hence the cell biomass concentration and / or culture phase, the input media flows to the bioreactor can be controlled accordingly to optimise the growth of the organism.

[0134] For example, microalgae may propagate, undergoing cell division. During this phase, the cell biomass density can increase rapidly. The liquid level in the photobioreactor may decrease due to evaporation. The salinity may also increase. Once the photobioreactor is full of culture medium to the maximum level (which may be determined by the pressure sensor 253) and the OD measurements indicate that the organism has reached its target density, the controller may control the input flows to deprive the organism of nutrition. The supply of N and P may be switched off. CO2 and air input flows may remain to flow at a rate determined by the model as optimal. The organism under these conditions enters a state where it no longer divides, but lays down a store of triacy lglyceride / 1 ipid (bio-oil).

[0135] CO2 injection may be optimized in dependence on the inferred CDW based on Sterner’s stoichiometric equation for microalgal composition (see Sterner, R. Warner., & Elser, J. J. (2002), “Ecological stoichiometry: The Biology of Elements from Molecules to the Biosphere”, Princeton, NJ, Princeton University Press):

[0136] 106 CO2+ 16 HNO3+ H3PO4 + 122 H2O -> (CH2O)42(CH2)64(NH3)i6H3PO4+ 170 O2

[0137] (3) and / or optimized based on the Redfield index (see Redfield, A. C. (1958), “The biological control of chemical factors in the environment”, American Scientist (Vol. 46, p. 230), Sigma Xi, The Scientific Research Honor Society, https: / / doi.Org / 10.2307 / 27827150).

[0138] All of the 106 carbons are derived from atmospheric CO2, which can be dissolved into the water medium (for example, seawater) provided to the photobioreactor. Therefore, as the cells grow and divide, the amount of CO2 required for algal growth also increases. CO2 dissolved in a seawater medium, which can be used in the photobioreactor, can occur in three chemical forms depending on pH: CO2, HCOs' and CO32-. The biological substrate for microalgal photosynthesis is HCO3; which reaches a maximum concentration at pH 8.2. CO2 gas is readily soluble at low pH and at high pH bicarbonate predominates. It is desirable to maintain pH in the photobioreactor to be in the range 7-8.5. The CO2 requirement of the algal cells may also vary in dependence on the light incident on the photobioreactor and hence on the organism. CO2 (HCO3) growth substrate can be supplied when the organism is photosynthetically active (for example, during sunlight hours when the measured light level is greater than a predetermined threshold), as well as with culture phase as indicated by the inferred dry cell weight. Therefore, the proportions of CO2 and / or air can be changed during the course of the cycle as prescribed by the model.

[0139] The proportion of CO2 in the exhaust (output) from the photobioreactor may change much less over the course of the growth and harvest cycle. As CO2 is one of the key costs in running the system, it is desirable that every mole of CO2 supplied to the photobioreactor is taken up by the cells and made into biomass (algae) and lipid. It is desirable to supply the CO2 sparging to just satisfy photosynthetic demand and minimise waste. If the amount of CO2 detected in the exhaust gas output from the photobioreactor is higher than a predetermined level (for example, greater than 400ppm CO2), then the controller can determine that some of the incoming CO2 is not being taken up by the cells of the organism. The model may take as input the value of the level of CO2 in the output exhaust gas from the photobioreactor, and the controller can reduce the amount of CO2 supplied to the photobioreactor as an input flow.

[0140] In dependence on the CDW (or other measure of biomass) estimated from the OD measurements, the controller may also be configured to limit the amount of N in the photobioreactor so as to limit or prevent the organism from dividing. N may be supplied as soluble urea or nitrate, or a mixture with phosphate compounds, and may be supplied at levels required to support a given end state of cell biomass of the organism.

[0141] In general, the controller may be configured to adjust one or more of the input media flows and / or a condition of the photobioreactor in dependence on the inferred biomass (e.g. CDW). Adjusting the input media flows may comprise adjusting one or more a quantity, flow rate and a concentration of one or more of the input media flows. Adjusting a condition of the bioreactor may comprise adjusting one or more of: a temperature of the photobioreactor, a pH level of contents of the photobioreactor (such as the pH level of the culture medium), a salinity of contents of the photobioreactor (such as the salinity of the culture medium) and a light intensity at the photobioreactor. In some examples, the controller may adjust a condition of the photobioreactor by adjusting one or more of the input media flows.

[0142] Therefore, the relationship between OD and CDW, as illustrated in Figure 6, can be used to determine the estimated biomass as CDW from the measured OD. Using the collected datasets, the other parameters (e.g. input media flow(s) and / or condition(s) of the photobioreactor) can be adjusted accordingly to optimise growth of the organism at the instantaneous biomass (e.g. CDW).

[0143] The datasets 770-789 can provide fine-tuning for photobioreactor productivity. These datasets can be continually populated from sensor data streams to the controller. Correlation of conditions with biomass concentration and bio-oil productivity measurements can be used to generate a learning and testing data sets for machine learning training, to train models to predict input flows and / or conditions with optimal productivity outcomes for a particular instantaneous biomass.

[0144] The controller may also control the photobioreactor in dependence on the geographical location of the photobioreactor. The geographical location may be an input to the model, along with the measured sensor data. Alternatively, based on the geographical data, a specific control model may be chosen for that geographical location. The geographical location may in some examples be determined from a location sensor, such as a GPS sensor. Alternatively, the location may be input to the control system by an operator of the system. The geographic data set 789 may provide data indicating parameters such as sunrise time, sunset time, latitude, longitude and altitude for particular geographical locations. This may allow the model to be trained to account for the location of a photobioreactor and for control strategies to be further adapted based on the geographical location of the photobioreactor. The controller program objective is to control culture inputs to achieve maximum productivity, in terms of organism growth and / or bio-oil production, under the variable environmental conditions experienced in the field. This may be achieved using lookup tables to store the optimal culture parameters, in terms of input flows of e.g. air and CO2 and conditions of photobioreactor such as temperature. On-line data capture of culture performance can be used to enhance the controller lookup table accuracy and reliability.

[0145] In one particular example, the controller may be configured to, if the optical density / biomass concentration is measured to be between 0.2 and 0.5, control the CO2 flow rate to be between 1 % v / v and 5% v / v.

[0146] The model may be configured with boundary conditions that cause one or more of the following conditions of the photobioreactor to be maintained between the following exemplary limits. The model may cause the temperature at the photobioreactor to be maintained at between about 20°C to about 50°C, preferably between 25°C - 35°C. The one or more input flows and / or conditions in the photobioreactor may be adjusted to maintain the pH in the photobioreactor between 6 and 11 , preferably at approximately 8.2 to maximise solubility of HCOs'. The one or more input flows and / or conditions in the photobioreactor may be adjusted to maintain a salinity of between 15 to 60 PSU, preferably 30 to 40 PSU. The photobioreactor may be controlled to maintain a photonic intensity of between 100 and 2500 pmol photons nr2.s’7400 nm - 700 nm Photosynthetically Available Radiation.

[0147] As discussed above, the controller may determine the culture phase time based on the estimated CDW. When the output of the model indicates that the product should be harvested (i.e. culture phase time = harvest), the input flow lines to the photobioreactor may be switched off. The culture medium containing the biomass and the product may be left to settle for a predetermined dwell time (for example, minutes or hours) during which cells with a lower percentage of oil generally settle in the lower levels of the photobioreactor and the fattest cells preferentially float to the upper levels of the photobioreactor. Following the dwell time, an outlet valve may be opened. The outlet valve may be controlled to select the ‘harvest’ path 275 in Figure 2 and the biomass from the photobioreactor can be pumped out of an outlet conduit to a tank for further processing.

[0148] As mentioned above, multiple photobioreactors may be arranged in an array. Each photobioreactor in the array can be controlled by a central controller. The central controller may be remote to each of the photobioreactors. The central controller may be configured to receive data indicating the state of the photobioreactor and / or the organism it is cultivating from each of the multiple photobioreactors. The central controller may process the data using the model as described above and then send control signals to respective ones of multiple local controllers of the multiple photobioreactors for adjusting the one or more input flows to a respective photobioreactor or a condition of the respective photobioreactor. The controller may be wirelessly connected to local controllers of each of the multiple photobioreactors. The local controllers may control valves of the respective photobioreactor to adjust one or more input media flows and / or a condition of the photobioreactor.

[0149] In one embodiment, multiple photobioreactors may be configured in a master-slave network comprising a master photobioreactor which is configured to control a plurality of slave photobioreactors. The slave photobioreactors may be controlled based on data collected from the master photobioreactor and based on the model output for that data. This may reduce the number of sensors in the array and / or the number of controller valves. The master photobioreactor may comprise a controller configured to receive data relating to a state of the master photobioreactor and / or the organism contained within the master photobioreactor, as described above. The slave photobioreactors may not comprise a controller as described above. The master photobioreactor and the one or more slave photobioreactors are each configured to receive one or more input media flows and to contain a respective organism while it synthesises one or more output media. The master photobioreactor may comprise one or more sensors configured to acquire the data relating to a state of the master photobioreactor and / or the organism in the master photobioreactor, as described above. The one or more slave photobioreactors may not comprise sensors. The flow of the input media that is supplied to each photobioreactor in the network and / or the conditions of each photobioreactor may be controlled by the master controller. The controller of the master photobioreactor may receive the data as described above and adjust one or more of (i) the input media flows to the master photobioreactor and the one or more slave photobioreactors; and (ii) a condition of the master photobioreactor and the one or more slave photobioreactors in dependence on the received data. The input media flows to the master photobioreactor and the one or more slave photobioreactors may be controlled at a master valve for each respective input media flow. The master valve may be controlled by the master controller. The master valve may control a respective one of the input media flows to each photobioreactor in the master-slave network (i.e. to the master photobioreactor and the one or more slave photobioreactors). Each photobioreactor in the master-slave network may receive the same quantities of input flows and / or be controlled with the same conditions (temperature, salinity, pH etc), based on the received data from the master photobioreactor.

[0150] Figure 9 shows the steps of an exemplary method for controlling a photobioreactor. The photobioreactor can promote the growth of an organism therein. As described above, the photobioreactor is configured to receive one or more input media flows (such as CO2, air, nutrients and water) and contain the organism while it synthesises one or more output media (such as a product for a biofuel). At step 901 , the method comprises receiving an optical density measurement for a sample comprising the organism. At step 902, the method comprises estimating a biomass of the organism based on the optical density measurement. At step 903, the method comprises adjusting one or more of the input media flows and / or a condition of the photobioreactor in dependence on the estimated biomass of the organism.

[0151] Figure 10 shows a schematic decision flow 1000 for night-time shutdown of air and CO2 sparging and recommencing daytime photosynthetic control. In this example, as shown at 1001 , the controller receives a measurement from the light sensor 247. The controller may compare this received measurement with a predetermined threshold light level. As shown at 1002, the controller may alternatively or additionally compare the current time with a known (for example, historical) sunset time. As shown at 1003, the CO2 and / or air supply to the photobioreactor may be stopped when the light level, as measured by sensor 247, drops below a predetermined threshold or may be stopped when the measured time is at or beyond the known sunset time for the geographical location of the photobioreactor. The air valve 241 and / or the CO2 valve 238 can be turned off so that air and / or CC^ are not supplied to the photobioreactor overnight. The controller may continually monitor the light level and the photobioreactor and / or the time. As shown at 1004, the controller receives a measurement from the light sensor 247. The controller may compare this received measurement with a predetermined threshold light level. As shown at 1005, the controller may alternatively or additionally compare the current time with a known (for example, historical) sunset time. As shown at 1006, the CO2 and / or air supply to the photobioreactor may be resumed when the light level, as measured by sensor 247, increases above a predetermined threshold or may be stopped when the measured time is at or beyond the known sunrise time for the geographical location of the photobioreactor. The air valve 241 and / or the CO2 valve 238 can be turned on so that air and / or CC^ are supplied to the photobioreactor during the day.

[0152] This may help to reduce metabolic drag from respiration on biomass accumulation and may provide a significant saving in cell dry weight biomass lost to respiration. For example, experiments have shown an average ~7% loss in cell dry weight overnight as respiration consumes sugar in oxidative phosphorylation.

[0153] As well as being turned off overnight, the supplies may generally be turned off when the light level, as measured by sensor 247, drops below a predetermined threshold. Features of any of described embodiments may be applied to features of any of the other described embodiments.

[0154] The algae produced by the photobioreactor system can, for example, be used for producing biofuel or related bio-derived products. As described above, the photobioreactor may conveniently have inbuilt sensing, control and / or communications systems. This can enable organisms and / or associated products to be produced in an economically viable form at scale.

[0155] To optimise the productivity of the microalgae, the environment of the photobioreactor is monitored and can be constantly adjusted to support biomass accumulation under a range of conditions throughout the growth cycle, including nitrogen repleted conditions, and to support bio-oil production under nitrogen-depleted conditions.

[0156] In particular, nitrogen limitation may be an important environmental trigger for triacylglyceride accumulation. In one particular example, for microalgae cells cultured at up to 130 Klux PAR, 26 - 35 PSU, 20°C - 45°C with atmospheric CO2, reducing available nitrate to below 2.7 mg.L’1can induce neutral triacyglyceride synthesis.

[0157] Optimum productivity in terms of both biomass concentration and bio-oil (triacylglycerides) is achieved when light, inorganic carbon and N / P supplements are titrated to the dynamic requirements of the cultured microalga. The closed loop control system described herein for photobioreactors leverages datasets of the culture and environmental parameters to facilitate a scalable, effective and low-maintenance control, sensor and database system using OD as the primary determinant of biomass concentration. This information, in combination with other parameters such as the culture stage, photon flux data, temperature profile data, salinity profile data, and final biomass and bio-oil productivity datasets combined within a lookup table which may be generated by vector machine analyses, define control inputs for maximum productivity. Data capture can be recorded and integrated into the datasets for continued analysis and optimisation of the controller function.

[0158] The controller and associated models / algorithms can measure parameters such as incident light, culture OD, salinity and culture volume in real-time to optimise CO2 and / or seawater input flow supplementation in response to measurements of light intensity, culture biomass (inferred from OD measurements) and culture salinity. Short-term light capture adaptations may drive CO2 demand. Marine microalgae have adapted to a highly mobile environment in which photosynthetically available radiation 400 nm - 700 nm photon flux changes rapidly with time, whether it is day or night, water depth and water opacity. Therefore, their light capture and conversion systems are highly dynamic so as to optimise energy capture and biomass growth, limit photodamage and photo inhibition and maintain ATP / NADPH redox balance. The short-term adaptive mechanisms which the controller engages may include: (1 ) rapid Non-Photochemical Quenching (NPQ) which operates over seconds in which excess photons are dissipated as heat protecting photosystem II (PSII), while simultaneously absorbing photons; (2) State Transitions which operate over minutes in which light harvesting complexes (LHC) associate with PSII or photosystem I (PSI) and so distribute electron flow between PSII and PSI; (3) Cyclic electron flow around PSI which operates in seconds to control electron flow between PSII and PSI. For optimum growth, these short-term adaptations to incident light drive RuBisCo modulated CO2 demand, which can be modulated in, for example, a millisecond(s) cycle of light flux measurement and CO2 release. This may be based on an algal-specific model / algorithm in the controller.

[0159] Total photosynthetic biomass is a predominant long-term factor in setting CO2 supplementation rate. Therefore, the controller model / algorithm, in addition to modulating CO2 supplementation in response to incident light, for example at intervals of milliseconds, can also increase the CO2 supplementation baseline in response to the measured cell culture OD, which has a linear relationship with cell growth, as illustrated in Figure 6.

[0160] Photobioreactor systems can be used with the controller described herein that can cultivate organisms in arid and semi-arid regions using saline waters / seawater, without using freshwater in the cultivation process. In traditional systems, because of evaporative water loss from photobioreactor headspace venting and through film plastics, the inevitable rise in culture media salinity is controlled by adding freshwater. This freshwater demand may be avoided through the use of high-salinity organisms, such as that described in WO 2024 / 132882 A1. Microalgae may exhibit preferred ranges of salinity. For example, Chlorella vulgaris requires freshwater for cultivation, Scenedesmus species have a preference for brackish PSU 10 - 15 waters found in estuaries and littoral areas, while others require higher salinities PSU 25 - PSU 32 for optimum function, such as Nannochloropsis oceanica. Other organisms may require extremely high salinity conditions for growth, such as Dunaliella salina.

[0161] The one or more processors may be configured to adjust the input flow of water (for example, seawater) in dependence on the salinity level of the contents of the photobioreactor and / or the culture volume, and / or other parameters such as estimated biomass and light level. The present controller can therefore measure salinity and / or culture volume and additional seawater and / or culture medium input flows can be added to control salinity and / or culture volume. The monitoring of salinity can thus optimize growth of the organism.

[0162] The present system can also implement optimised geo-located environmental models. The models can be generated from (for example, trained using) data collected during operation of the controller and can improve control and output.

[0163] A central component of the controller model / algorithm is data storage and analysis correlating culture performance with controller command inputs in response to changes in measured environmental conditions. The model / algorithm may be initialised with initial conditions set in dependence on the geographical location, for example by geographical location environmental data (for example, diurnal light ranges, hours daylight, seasonal changes, diurnal temperature ranges, diurnal temperature ranges with season, local salinity of seawater etc.). The working database of measured and / or inferred data may be subsequently populated with more processspecific, fine-detailed information derived from operations (for example, using historical data measured by sensors in communication with the controller) to provide a more effective control strategy tailored to the specific deployment site conditions. The geographical location of the bioreactor may be used to give the overall predicted pattern of use of CO2, or other parameters such as salinity of seawater, for that location, with the monitoring allowing fine tuning to the specific situation. The monitoring may therefore be a calibration of the expected conditions needed to optimise growth.

[0164] The controller described herein can aid the efficient production of bio-oil, which when produced from cultures supplied with flue gas, constitutes a carbon negative drop-in biofuel for marine and heavy transport as well as a carbon negative SAF feedstock and other petrochemical products. Using this approach, the delivery of sustainable low carbon biofuel, or other biomass or metabolic products at scale with minimal environmental impact can be achieved using marine microalgae cultured in photobioreactors in non-agricultural land, as well as, by capitalising on the capabilities of microalgae strains, in semi-arid and arid areas. The latter not only provides a route to biofuel production without compromising agricultural land and freshwater resources, but also for semi-arid and arid areas with sparse of no vegetation. This may increase global photosynthetic capacity, allowing for a reduction in atmospheric CO2 reservoir.

[0165] The applicant hereby discloses in isolation each individual feature described herein and any combination of two or more such features, to the extent that such features or combinations are capable of being carried out based on the present specification as a whole in the light of the common general knowledge of a person skilled in the art, irrespective of whether such features or combinations of features solve any problems disclosed herein, and without limitation to the scope of the claims. The applicant indicates that aspects of the present invention may consist of any such individual feature or combination of features. In view of the foregoing description it will be evident to a person skilled in the art that various modifications may be made within the scope of the invention.

Claims

CLAIMS1 . A controller for a photobioreactor for promoting the growth of an organism therein, the photobioreactor being configured to receive one or more input media flows and contain the organism while it synthesises one or more output media, the controller comprising one or more processors configured to: receive an optical density measurement for a sample comprising the organism; estimate a biomass of the organism based on the optical density measurement; receive a measured value of an incident light level at the photobioreactor; receive data indicating a salinity level of contents of the photobioreactor; and adjust one or more of the input media flows in dependence on the estimated biomass of the organism and the light level, wherein the one or more of the input media flows comprise CO2.

2. The controller as claimed in claim 1 , wherein the one or more processors are configured to adjust the one or more input media flows to optimise growth of the organism and / or the production of the one or more output media based on the estimated biomass.

3. The controller as claimed in any preceding claim, wherein the one or more processors are configured to continuously adjust the one or more input media flows by repeatedly performing the following further steps: receive an updated optical density measurement for a sample comprising the organism at a respective time; calculate an updated estimated biomass of the organism at that respective time; and adjust the input media flows in dependence on the updated estimated biomass of the organism at that respective time.

4. The controller as claimed in any preceding claim, wherein the estimated biomass is an estimated cell dry weight of the organism and wherein the one or more processorsare configured to adjust the one or more of the input media flows in dependence on the estimated cell dry weight.

5. The controller as claimed in any preceding claim, wherein the received optical density is measured at between 700 - 800 nm incident radiation.

6. The controller as claimed in any preceding claim, wherein the input media flows are flows of one or more of a gas, a liquid and a solid.

7. The controller as claimed in any preceding claim, wherein the one or more input media flows comprise one or more of water, air and nitrogen-, phosphorus- or traceelement containing compounds and vitamins.

8. The controller as claimed in any preceding claim, wherein the one or more processors are configured to adjust a condition of the photobioreactor in dependence on the estimated biomass, and wherein adjusting the condition of the photobioreactor comprises adjusting one or more of: a temperature of the photobioreactor, a pH level of contents of the photobioreactor, a salinity of contents of the photobioreactor and a light level at the photobioreactor.

9. The controller as claimed in any preceding claim, wherein the one or more processors are further configured to adjust one or more of the input media flows and / or the condition of the photobioreactor in dependence on further received data relating to the state of the photobioreactor and / or the organism.

10. The controller as claimed in claim 9, wherein the one or more processors are configured to cause the estimated biomass and / or the further data relating to the state of the photobioreactor and / or the organism to be stored at one or more memories communicatively connectable to the controller.

11. The controller as claimed in any preceding claim, wherein the one or more processors are configured to adjust one or more of the input media flows in dependence on historical data relating to the type of organism.

12. The controller as claimed in any preceding claim, wherein the one or more processors are configured to input the received optical density measurement and / or the estimated biomass to a model, wherein the one or more processors are configured to adjust the one or more media flows in dependence on an output of the model.

13. The controller as claimed in claim 12, wherein the model is a trained artificial intelligence model.

14. The controller as claimed in claim 12 or claim 13, wherein the model is specific to the organism.

15. The controller as claimed in claim 14, wherein the model is specific to the type or species of the organism.

16. The controller as claimed in claim 13, or in claim 14 or 15 as dependent on claim 13, wherein the model is trained using multiple quantitative culture relationship datasets.

17. The controller as claimed in claim 16, wherein the quantitative culture relationship datasets comprise one or more of: incident light level and cell dry weight, temperature and cell dry weight, salinity and cell dry weight, CO2 and cell dry weight, pH and cell dry weight, pH and CO2, nitrogen level and cell dry weight and phosphorus level and cell dry weight.

18. The controller as claimed in claim 16 or claim 17, wherein the quantitative culture relationship datasets comprise data acquired at the photobioreactor and previously received by the one or more processors.

19. The controller as claimed in any of claims 16 to 18, wherein the quantitative culture relationship data is specific to the organism.

20. The controller as claimed in claim 19, wherein the quantitative culture relationship data is specific to the type or species of the organism.

21. The controller as claimed in any preceding claim, wherein the controller is configured to control the photobioreactor in dependence on the geographical location of the photobioreactor.

22. The controller as claimed in claim 21 as dependent on any of claims 12 to 20, wherein the geographical location is an input to the model.

23. The controller as claimed in claim 21 as dependent on any of claims 12 to 20, wherein the controller is communicatively connected with a memory storing a plurality of models and wherein, in dependence on the geographical location of the photobioreactor, the controller is configured to: select a model from a plurality of models; and implement the selected model to control the photobioreactor based on the estimated biomass for the organism.

24. The controller as claimed in any preceding claim, wherein the controller has access to a data structure relating values of optical density to corresponding values of biomass for the organism.

25. The controller as claimed in any preceding claim, wherein the controller is configured to adjust one or more of the input media flows in dependence on respective CO2 levels in one or more of (i) gas entering the photobioreactor and (ii) gas leaving the photobioreactor.

26. The controller as claimed in any preceding claim, wherein the controller is further configured to determine, based on the estimated biomass, an optimal time to inoculate the photobioreactor and / or to harvest the one or more output media of the organism.

27. The controller as claimed in any preceding claim, wherein the one or more processors are communicatively connected with a spectrophotometer, wherein the one or more processors are configured to receive the optical density measurement from the spectrophotometer.

28. The controller as claimed in any preceding claim, wherein the one or more processors are configured to: determine whether the measured light level is above or below a predetermined threshold; and adjust the one or more input media flows in dependence on this determination.

29. The controller as claimed in claim 28, wherein if the measured light level is below the predetermined threshold, the one or more processors are configured to stop the flow of CO2 to the photobioreactor.

30. The controller as claimed in any preceding claim, wherein the organism is a photosynthetic organism.31 . The controller as claimed in claim 30, wherein the organism is a microalgae.

32. The controller as claimed in any preceding claim, wherein the one or more output media comprises a bio-oil.

33. The controller as claimed in any preceding claim, wherein the one or more processors are further configured to: receive data indicating a temperature of the photobioreactor; and adjust the one or more of the input media flows in further dependence on the temperature.

34. The controller as claimed in any preceding claim, wherein the one or more processors are configured to adjust the one or more input media flows in further dependence on one or more of: a pH level of contents of the photobioreactor, the salinity level of contents of the photobioreactor and a culture volume in the photobioreactor media level in the photobioreactor.

35. A method for controlling a photobioreactor for promoting the growth of an organism therein, the photobioreactor being configured to receive one or more input media flows and contain the organism while it synthesises one or more output media, the method comprising:receiving an optical density measurement for a sample comprising the organism; estimating a biomass of the organism based on the optical density measurement; receiving a measured value of an incident light level at the photobioreactor; receiving data indicating a salinity level of contents of the photobioreactor; and adjusting one or more of the input media flows in dependence on the estimated biomass of the organism and the light level, wherein the one or more of the input media flows comprise CO2.

36. A computer-readable storage medium having stored thereon computer readable instructions that when executed at a computer system comprising one or more processors cause the one or more processors to perform the method of claim 35.

Images