INTRODUCTION
During the winter/spring of 2017, Plant Chicago Technology Coordinator Eric Weber and Plant Chicago intern Caitlin Hall began investigating the impact of greenhouse gas carbon dioxide (CO2) on the algae in our photobioreactor. A photobioreactor (below) is essentially a clear chamber where a mixture of water, algae and nutrients are exposed to air and light. They are used for things such as biofuel and photosynthesis research. In our case, the bioreactor is for the production of algae as a food source as well as carbon sequestration in the building. In addition to the CO2 us humans breathe out, three of the co-located businesses here at The Plant produce CO2 as part of their fermentation process (Whiner Beer Co., Arize Kombucha and Kombuchade). Rather than let it go out into the atmosphere, we want to see if it can be captured and used.
The first step in this process was seeing how much CO2 a specific amount of our algae culture can process. The two questions at hand were: What amount of CO2 makes the algae grow fastest? and Can too much CO2 kill the algae? Here what we were looking for was the point at which the CO2 would affect the chemistry to a large enough degree that the algae would no longer grow. Injection of CO2 into water is known to create an acidic environment (think greenhouse gasses and ocean acidification), so there theoretically would be a tipping point at which the water chemistry of the bioreactor would be unable to support life. Comparing that point to where we were achieving the maximum growth rate of the algae would give us an idea of how much CO2 the algae can process.
To accomplish this we set up two 10-gallon aquariums and filled them each halfway with water from our main bioreactor. We aerated the systems in a similar manner to the large bioreactor, and lit them with fluorescent lights. One system was left just as such (our control bioreactor, below right), and the other was injected with CO2 from a compressed tank (variable bioreactor, below left). The bioreactors were monitored for three months, periodically increasing the CO2 amount in the variable system while monitoring algae growth and water chemistry in both systems.
WHAT WE FOUND
Our first hurdle in this test was the fact that spirulina likes to clump together, forming what are known as mats. The formation of mats is not necessarily bad, though they make it difficult to calculate the growth rate of the algae. (The addition of circulation pumps to the systems to increase helped slightly, but did not eliminate, mat formation.) We evaluated the algae growth rate by measuring the amount of light that was blocked by the water/algae mixture, and comparing this to the previous measurements. While not ideal for an objective measurement, comparing the variable to the control system gave us a sense of how much faster the algae + CO2 was growing. This method of measuring growth led to our second hurdle: after a few weeks, both tanks were so dense with algae that we reached the limit of what the light meter could measure (in Week 8 for Algae + CO2, Week 9 for Algae only).
Consistently dosing the reactor with compressed CO2 also presented some challenges. The dosing setup was what is typically used for dosing freshwater planted aquariums, and consisted of a 5lb CO2 cylinder, regulator, solenoid, needle valve, and bubble counter. CO2 was then measured in ‘Bubbles per Second,’ or bps. Once the dosing reached 10 bps, we hit the limit of our ability to quickly count. Also, by dosing 24/7, the heat generated by the solenoid caused changes in the clearances of the valves in the system, resulting in the need to periodically adjust the flow to maintain the desired dosing rate.
With a CO2 dosing rate in the variable system of 4 bps and above, it showed no significant algae growth week to week, but did have a measurable amount of CO2 dissolved in the water. Below this rate, there was little to no CO2 dissolved in the water. The control system didn’t contain any dissolved CO2 during entire testing period. This is not to say that the control algae didn’t have any CO2 to take in, but rather the algae was using CO2 at the same rate that it was entering the water from the atmosphere. The variable tank was receiving this same atmospheric CO2 PLUS the CO2 from the dosing system, thus the dissolved CO2 level read higher.
The transfer of a gas across a surface (such as the surface between air and water) is known as diffusion. In this instance if the concentration of CO2 is higher in the atmosphere than in the bioreactor, CO2 molecules will diffuse across the surface. In the case of the bubbles of CO2 produced by the dosing system, the concentration of CO2 there is MUCH higher than that of the water, so more molecules will diffuse into the water at a faster rate. The same is true if the roles were reversed: In the variable system water, CO2 concentrations were higher than that of the room. That system then was constantly LOSING CO2 into the room through diffusion even as we were injecting it with the dosing system.
Converting bubbles per second to a more standard measurement, our 10 bps upper measurement equates to roughly 0.018 cubic feet per hour. (This is a very small number by industrial standards, though this is also for 5 gallons of algae. Stepping up to our large photobioreactor, which holds roughly 250 gallons, equals a capacity of 4.48 cubic feet per hour of CO2.) At 10 bps, the dissolved CO2 level in the variable tank got as high as 50 ppm, while at 4 bps, dissolved CO2 was around 25 ppm. More data points will need to be collected to verify the relationship between the CO2 added to the reactor and the resulting level of dissolved CO2, though the test we were using has a maximum range of 50 ppm, which may make achieving this difficult.
By the 12th week of testing we decided to increase the CO2 dosing rate to a near stream of bubbles, since incremental increases were not realizing significant changes in water chemistry and algae growth could no longer be accurately measured. Our goal was to measure the chemistry the following morning to see what effect this large increase of CO2 would have. Unfortunately the compressed CO2 tank ran out sometime overnight and the following morning’s tests resulted in 0 ppm dissolved CO2 and no change in the other parameters. So ended our first stage of this experiment.
WHAT’S HAPPENING HERE?
When CO2 enters into water, some of it will remain as dissolved CO2; the rest will either be utilized by organisms (such as the algae), be taken up as part of chemical reactions, or exit the water through diffusion. The primary chemical reaction taking place is known as hydrolysis, where carbon dioxide reacts with the water itself to form carbonic acid. This acid is responsible for lowering the pH of the reactor (and the world’s oceans). The test we do to measure CO2 only accounts for the CO2 that stays in the dissolved state, so additional tests (such as pH) need to be done to track where CO2 goes. The pH of the variable system consistently measured 0.5-1.0 lower than the control tank, verifying this occurrence.
An example of the CO2 exiting the system before getting a chance to react with anything is seen when measuring the air around the bioreactors. While the background atmospheric CO2 level in the aquaponic farm averages around 460 ppm, a sensor placed above the variable system can read well above 2000 ppm! This indicates that potentially most of the CO2 entering the bioreactor simply comes right back out again. A more efficient method for dissolving CO2 into the water is needed. The method used in our experiment simply involves air line and a small ceramic airstone similar to what would be used in an aquarium. Ideally we would make use of a CO2 reactor (below), which is basically a chamber filled with a plastic media or set of baffles to increase the contact time between the water and the CO2 bubbles, thereby increasing the rate of dissolution of CO2 in the water.
The next stage of this experiment will be aimed at additional data gathering around CO22, both within the bioreactor and the building at large. We will incorporate a CO2 reactor to our bench scale test to improve the efficiency of the dosing equipment. The removal of the solenoid from the compressed CO2 tank as well as the sourcing of a low-range airflow meter will improve the accuracy and consistency of dosing. Decreasing the volume of the bioreactors will increase CO2’s effectiveness. Acquiring a more sensitive light meter will improve our ability to gauge algae growth. (The only sure-fire way to measure growth is to measure the dry mass of the algae, which requires dehydrating, weighing and rehydrating the entire amount algae. This would be difficult to accomplish repeatedly.)
We will also measure the outputs of CO2 from the various sources in the building in order to get a sense of flow rates and frequency. If these measurements match up well with the calculated CO2 capacity of the large bioreactor, then discussions could be had regarding the transfer of CO2 from its sources to the bioreactor.
An attempt will be made to identify the species of algae present in the large photobioreactor. When the system was first started, a culture of spirulina was introduced. Since then our main source of nutrients has been the aquaponic wastewater, which contains any number of different species of algae. It is conceivable that, once introduced, another species could take over for the spirulina. This new species might then end up the primary or even sole driver of nutrient consumption within the system. In the end, since the main purpose of the bioreactor is to absorb CO2 and process wastewater while growing algae that can be harvested and fed to fish, it is not critical that the algae culture be of a particular species. Knowing WHICH species, however, can help us adjust other water parameters more to its liking. Stay tuned to this blog, or come visit us at The Plant to check up on how this project is going!