Date: June 25, 2022
Aristotle famously wrote, “The more you know, the more you realize you don’t know.”
In the three months since we wrote our application to Stripe, we have made further strides in refining our understanding and plan.
As a result of our conversations with Stripe, we decided to develop a variation of our Caribbean system that makes 1000+ year durability possible. To share how that will work, and to address additional questions and feedback we received about our application, we have prepared the following addendum in a Q&A format.
We are humbled by how little is known about the deep sea and are grateful for the chance to be published on Stripe’s Github. Our highest hope is that those knowledgeable about the topics herein will read this and get in touch with us to further elucidate any portion of the body of scientific knowledge we should be aware of and cite.
Question 1: Being that carbon gets converted into methane (CH4) by deep sea microbes, how can Pull To Refresh be sure that the carbon in the seaweed you sink won’t be converted into methane? (This is of particular concern being that in the short term methane is a much stronger greenhouse gas than CO2).
Methane is a very potent greenhouse gas composed of one carbon bonded to four hydrogen molecules. Methane absorbs some frequencies of infrared radiation emitted from the Earth’s surface, trapping heat in the atmosphere that would otherwise go out into space.
Answer: In short, the high oxygen content of the deep ocean will prevent the creation of methane from any seaweed suspended in water at the ocean floor. Bohrmann and Torres (2006) show that the high oxygen content in the deep sea, as shown in the images below, would prevent methanogens from existing at the sea floor, and therefore methane creation isn’t expected to occur from sinking seaweed in our proposed locations.
Question 2: Won’t methane be produced under the sediment?
Answer: When seaweed sinks to the ocean floor and is buried under dense materials, such as limestone and silica, the lack of oxygen deep in the sediment has been shown to lead to the production of methane. At our proposed ocean depths, this methane would be frozen as methane hydrate. Frozen methane is of concern because it is less dense than water, which means it would float if there weren’t other forces at play. Floating to the surface is not expected to happen due to the dense materials that make up the sediment. Methane hydrate adheres to the sediment, and the accumulated sediment on top of the hydrate acts as a ‘paperweight’ to keep the hydrate at the bottom of the ocean during our proposed timescales.
This chart shows the density and temperature at which methane becomes its frozen hydrate form.
Question 3: What if global warming causes the ocean to heat up and the frozen methane hydrate thaws and is released as a greenhouse gas?
Answer: Our proposed project does not risk being impacted by the lowering of the hydrate phase boundary. Any sub-sediment hydrates formed from the seaweed we sink will be nowhere near that potential new boundary because of the depth our methodology will require us to sink to in order to be rated at 1000+ year permanence.
Question 4: Would the Air-Sea Flux cause leakage in the 1000-year time scale?
Answer: Wüst (1963) describes the stratified layers in the Caribbean. Over our proposed timescales, mixing happens within distinct layers of ocean rather than between them. Temperature, salinity, and density create these distinct layers. We can think of ocean stratification like a layered cake. You can add more cake to one layer and it doesn’t affect the composition of the other layers. This vertical mixing has been long established, but only recently has the horizontal mixing become better understood. Thus, we referenced in our application Siegel et al. (2021) in which the horizontal mixing by lateral currents up inclines is modeled and underpins the minimum depth of sinking seaweed at which we can assert a given permanence. We expect the seaweed to sink to depths far more than 1000 meters, where vertical mixing to the air can not occur during the first 1000 years.
Question 5: Would it be possible to achieve 1000+ year durability in the Caribbean?
Answer: Yes, we can achieve 1000+ year durability in the Caribbean. There are targeted areas we can sink in the Caribbean in which the literature supports modeling 1000+ year sequestration. These outlined areas have 5000m+ depths. We arrive at this assertion by combining the Siegel et al. model about sequestration timescales with data about bottom current velocities from Shanmugam (2020) and the Arc-Minute global relief data included in the image below.
During our discussions with Stripe we amended our offer to Stripe to the following:
500 Caribbean tonnes with 1000+ year durability
500 Pacific tonnes with 1000+ year durability
The real-time dashboard described in our application and our full reports will show the GPS and current speed at the time of sinking to verify Sargassum gets sunk within the boundaries where it is modeled to hit the ocean floor at 5000m+ depth.
Question 6: Can the vessel you create for the Caribbean system really be used in the Pacific system given the intense and very different conditions of each?
Answer: Due to the need to survive intense weather during hurricane season in the Caribbean that is quantified in Wu et al. (2003), and the larger waves in the Pacific shown in Allan and Komar (2000), we will require equally strong vessels in both locations. The Pacific and the Caribbean will each call for a specific set of carbon removal accessories. These vessels & accessories will ultimately be produced in a location that enables easy access to both oceans.
Long term, we are likely to evolve our technology and design as we continue to iterate and optimize. But initially we have designed accessories that allow the same vessel to be used in both the Caribbean and the Pacific, and it’s the accessorization that allows them to perform very different carbon removal tasks in each area of ocean.
Question 7: What measures are you taking to ensure your project doesn’t harm sea life?
Answer: Since writing our application, we have made progress on our machine vision system that is used to detect whether our collection receptacles have Sargassum in them or not. We can also apply this same technology to detect anomalies, such as the presence of life in the receptacles, and use the local processing available on the vessel to initiate reversing the vessel to empty whatever would be in the receptacles that shouldn’t be. This is one of the ways we are taking measures to protect sea life. Another factor that impacts sea life is sound. Our brushless motors produce far less noise than combustion engines, and this supports marine life’s ability to sense prey with echolocation and communicate with their communities. Additionally, we have identified light spectrums that we can use for flash photography that our machine vision depends on so that we won’t disturb sea life. We are legally required to place a 360 degree light at the rear of the vessel and can use the narrow wavelength shown below that can provide the needed human safety without impacting the sea life on the surface.
Question 8: How will your project affect deep sea ecosystems?
Answer: To address the need to further understand marine life on the ocean floor, in our application we proposed using observation equipment to monitor interactions between sunk seaweed and other life. We don’t anticipate any marine life at the depths we describe in our application having the ability to either produce methane or in any other way transfer the CO2 back to the surface. However, the exact impact of additional seaweed on ocean floor ecosystems isn’t known. The data we can gather can inform the industry as we all seek to better understand the deep ocean.
Question 9: Will the operation really be running 365 days of the year?
Answer: In our application we stated 365 day a year operation. We wish to clarify that we don’t propose that all vessels will be at max performance every day of the year but that we will optimize for 7 day and 14 day averages of our continuously deployed vessels that will only be pulled in if they need maintenance. So they don’t dock but are simply charging in the sun whenever there is down time. Certain times of the year will have more Sargassum so that is also factored into our calculations. We expect to have a daily average that is our annual yield divided by 365.
Question 10: Why have you not specified any use of CO2 level monitoring equipment?
Answer: The reason is that we can sample the seaweed, (be it Sargassum or Macrocystis,) take it to a lab and analyze it for carbon content. We are considering adding sensors to our vessels to contribute more CO2 data to the scientific community, but this would not be part of our monitoring and verification and instead would be in pursuit of better global models of the climate. Our ability to measure the weight of what we sink can be calibrated to provide accurate monitoring and verification of how much carbon we remove from the carbon cycle. The measurements we would get from CO2 sensors would provide far less meaningful data, containing mostly noise from other biological factors. It’s important to note the boundary of our system is the Earth’s carbon cycle over a timescale of 1000 years, not simply an area of ocean, and the goal is to reach gigatonne-scale removal in order to meaningfully reduce the parts per million in the atmosphere and ultimately stop the planet from warming.
Question 11: Why does your application refer to CO2 removed from the atmosphere when seaweed grows in the water?
Answer: In order to be comparable to direct air capture, tree planting, biochar, and mineralization, we use the weight that the carbon would have if it had two oxygen molecules attached and was in the air. This number will be the amount of carbon our system has measured that we are sinking times 3.66666. To our knowledge, this is how the industry measures its impact and sells tonnes of carbon removal. Although called “carbon removal,” it is actually tonnes of CO2, which weigh more since the oxygen also has weight. This metric makes sense for accounting purposes because emissions are measured in tonnes of CO2. Durably storing CO2 with the oxygen still attached is very difficult and most other projects also turn it into carbon with no oxygen and then multiply its weight by 3.66666 to find out how much atmospheric CO2 was removed. What matters is that the carbon in the seaweed we sink will not be a part of the surface ocean, air, or any surface biology for more than 1000 years.
Question 12: Why doesn’t your application account for the ships needed to transport the seaweed that is grown back to land for monitoring and verification?
Answer: It’s possible to use digital scales and cameras connected via satellite to effectively monitor and verify locally on the vessel and broadcast back to servers for us and any stakeholders to remotely monitor. An essential calibration process creates the underpinning math and will provide a margin of error that will be known. This is possible because each species has a fairly consistent buoyancy-to-mass ratio and carbon content. As variations are recorded in carbon content with frequent sampling we can recalibrate the system to remain accurate. We will need to bring samples back to land a few times a year and we have budgeted for those in our system maintenance but did not detail them in our proposal. Far less than one percent of the seaweed will be brought to land to be analyzed for carbon content.
Question 13: Isn’t kelp a net producer of CO2 in some locations?
Answer: Yes, and so we aren’t sinking in any of those locations. It’s important to note we will only gather or cultivate and sink seaweed in very deep areas of the ocean, not along the coast.
Question 14: Isn’t there more CO2 in the water than in the air, so aren’t you removing CO2 from the wrong place and therefore discount the carbon removed from the water since some CO2 will be replenished from below?
Answer: All carbon removal of any kind is followed by an equilibrium process. Our method removes carbon from the air/water/biomass system as a whole given time for equilibration to take place. When CO2 is turned into carbon through photosynthesis of seaweed in the water, chemical forces chase new balance and excess CO2 in the air replenishes the levels in the ocean. If replenishment happens from CO2 that is lower in the water column rather than from the air, then this would be due to momentarily lower CO2 levels in that region of the atmosphere.
That momentary upwelling of CO2 would only be chemically possible within the upper 1000 meters, and then readjusts by sinking back down to balance out, pulling with it CO2 back out of the air when that occurs. In other words, in a very short-term timescale circulation of CO2 can be detected, but if you wait a longer period of time, (such as several months or a couple of years,) we expect those fleeting perturbations to not be relevant to our accounting. Over all it will average itself out.
We don’t need to discount the tonnes of CO2 removal due to them coming from the water, just as direct air capture facilities that remove large amounts of CO2 will ultimately have some of their work undone as the oceans put some CO2 back into the air to take its place and restore the equilibrium. Since the industry does not currently discount direct air capture facilities, it doesn’t need to discount seaweed operations . Biology does not appear to be constrained by the availability of CO2 in the air / water and so there appears to be as much biology on Earth at any given point as the other constraining forces will permit.
That biology threshold is expected to be met whether or not we sink any seaweed. There is no evidence we have found in the literature that suggests sinking seaweed gives rise to more growth than otherwise would have happened anyway. So just as direct air capture or mineralization are simple subtractions from the carbon cycle, sinking seaweed to sufficient depths is also a simple subtraction from the carbon cycle.
Capacity Calculations Clarification
According to data from Florida State University, for every square meter of Caribbean water, there are about 10g of Sargassum on average. If a gathering accessory is being driven around on an exploration path that is constantly covering new waters, then statistically every square meter covered should equal 10g of Sargassum collected. In reality, the Sargassum is not spread out, but concentrated into mats. Therefore, on some days some vessels will collect nothing, and on more prosperous days they will collect an abundance.
For the purposes of estimating the effectiveness of a fleet, we can multiply the cross-section of the collector, which is three meters, by the distance we can travel each day, which is one meter per second.
The tonnes of seaweed we sink will contain carbon that came from CO2 that was photosynthesized. The tonnes we sell will be labeled as tonnes of CO2, which has a weight 3.66x higher than pure carbon due to the two oxygen molecules attached. The dry weight analysis of carbon content in seaweed is listed as pure carbon. In order to arrive at the industry standard for CO2 removal, we multiply by 3.66 to arrive at tonnes of CO2 removal.
With a 50% reduction to the Florida State estimates because of possible margin of error and dry to wet weight ratios that are linked in our references, we calculate the following annual tonnes of CO2 removal for each vessel:
((WTSC)/R)M = V
W (weight ) = 5 grams per square meter of ocean
T (time) = 31,536,000 seconds in one year
S (speed) = 1 meter per second annual average
C (cross-section) = 3.2 meters
R (dry to wet weight ratio) = 0.03
M (ratio of CO2 molecular weight per carbon molecule) = 3.66
V (volume) = 55.402 tonnes of CO2 removal per year per vessel
Allan, J., & Komar, P. (2000). Are ocean wave heights increasing in the eastern North Pacific?. Eos, Transactions American Geophysical Union, 81(47), 561-567.
Bohrmann, Gerhard & Torres, Marta. (2006). Gas Hydrates in Marine Sediments. 10.1007/3-540-32144-6_14.
Shanmugam, G. (2020). Mass transport, gravity flows, and bottom currents. Elsevier. https://doi.org/10.1016/B978-0-12-822576-9.00008-4.
Siegel, D. A., DeVries, T., Doney, S., & Bell, T. (2021). Assessing the sequestration time scales of some ocean-based carbon dioxide reduction strategies. Environmental Research Letters. https://doi.org/10.1088/1748-9326/ac0be0
Tillotson, K., & Komar, P. D. (1997). The Wave Climate of the Pacific Northwest (Oregon and Washington): A comparison of data sources. Journal of Coastal Research, 440-452.
Wu, C. S., Taylor, A. A., Chen, J., & Shaffer, W. A. (2003). 3.5 Tropical Cyclone Forcing of Ocean Surface Waves. NOAA.
Wüst, G. (1963, July). On the stratification and the circulation in the cold water sphere of the Antillean-Caribbean basins An abstract. In Deep Sea Research and Oceanographic Abstracts (Vol. 10, No. 3, pp. 165-187). Elsevier.