During the Fall of my senior year at Tufts, I designed, manufactured, and tested a kelp farm proof-of-concept as a capstone project. In this project, three teammates and I worked to design a new method for farming kelp which could be more scalable than existing farming techniques.
As one of the fastest growing plants, kelp is capable of sequestering roughly 20x more CO2 per acre than trees. Compared to other sequestration methods, kelp can be less expensive, more scalable, and does not compete for land with other industries. This makes kelp an attractive climate solution.
In the Summer of 2021, I obsessed over climate change. In my free time, I found myself reading books, listening to podcasts, researching ongoing efforts, and even reading academic journals on the subject. I felt extremely motivated to do something to help our planet, but I didn't know how to channel this feeling into something productive.
When I heard my Senior Design class would be given the opportunity to pitch project ideas, I decided to present an idea for a climate-related project. I spent some time researching areas that could benefit from new technologies and that were feasible to tackle in a 3 month timeline. After looking into many fields ranging from renewable energy to carbon-free transportation, I decided to present a concept for using kelp to sequester CO2 and deacidify the ocean.
I pitched this idea to my class in the first week of the semester, and received interest from several of my peers. We decided to form a team called "Kelp is on the Way", and got to work researching our design space with the ultimate goal of helping to make the kelp aquaculture process more scalable.
Slides from an early presentation on the project.
Through speaking with several industry experts and researching the current state of the kelp industry, we reached several important findings. The most pressing result of our research was finding substantial controversy surrounding the use of kelp for carbon sequestration. Kelp sequesters a huge amount of carbon, but it is hard to reliably quantify how much. This is important if it is to be used in exchange for carbon credits in the future. Using kelp for sequestration also requires sinking the kelp to the bottom of the ocean, something that critics claim is a waste of a valuable resource and could harm ocean ecosystems.
Using kelp strictly for carbon sequestration is an extremely young industry, with more research needed to establish a scientific foundation and define engineering problems in the design space. Because of this, we decided to target our designs towards the use of kelp as a crop, meaning we would be harvesting our kelp rather than sinking it. Growing kelp as a crop provides a huge number of ecological benefits — most notably ocean deacidification and habitat restoration, making it widely considered one of the most sustainable sources of food.
After thoroughly researching problems and bottlenecks in the kelp aquaculture process, we began the concept generation phase of the project. We first came up with a "Set A" list of low-fidelity concepts in three areas: remote kelp monitoring, data readout (user interface), and actuation for automatic farm maintainence.
The core belief in our ideation was that by increasing the automation of kelp farming, farmers can save time on things like checking on their farm, performing maintenance, and harvesting the kelp. With less active time required, farmers are then free to grow more kelp, scaling up the entire aquaculture process.
During this initial phase of ideation, I came up with a concept which would allow for kelp to be grown more space-efficiently and harvested easier. This idea was centered around growing the kelp off of a textile rather than a rope, which expands the growing into the third dimension to make use of all the space available to a farmer. To enable easier harvesting, the textile has ballasted buoys at each corner, allowing for the kelp to be raised out of the ocean at harvest time. The kelp then dries out in the sun, and can be picked up by the farmer after it has dried. Dry kelp weighs roughly 10% as much as wet kelp, so this change would make havesting much easier and more efficient for farmers.
My initial concept drawing of the "Textile Matrix".
With about 50 concepts in our Set A group, we used post-it notes to split our ideas into categories. We then paired down and combined ideas in these categories to arrive at a "Set B" of about 10 designs. For each of these designs, we then made more detailed design and manufacturing decisions, and sketched the design in higher fidelity.
Three sketches from our Set B designs.
From our 10 Set B designs, we used decision matrices to select the best 4. These "Set C" designs were then fleshed out in much higher detail. We selected specific parts for purchase, and manufacturing techniques to use for certain mechanisms. Each member of the team then 3D modeled and performed a computer simulation on one of the designs. Since I had proposed the Textile Matrix concept, I took ownership of this design.
My Set C CAD model of the Textile Matrix.
My CAD model consisted of an aluminum extrusion frame and four ballasted buoys. Each buoy was made of a cast acrylic tube, a ballast tank, two water pumps, space for sensors and electronics, and two lids fit with O-ring seals.
My teammates' Set C CAD models: a remote sensing buoy, a cheap clip-on sensor suite, and an autonomous underwater vehicle.
With our set of four high-fidelity designs, it was time to choose a final concept. We met with our sponsor and discussed the pros and cons of each concept. Ultimately, our sponsor recommended the Textile Matrix as our final design, due to its originality and manufacturability.
With just one week between our sponsor meeting and the deadline for purchasing components, we got to work critically reviewing and revising the Textile Matrix design. Rather than pumping water for our ballasting system, we decided to switch to compressed air, which would reduce costs. Additionally, we changed the frame to an "X" shape to remove mass. We selected components, remodeled the design as a team, and purchased our parts. The entire system cost roughly $300 — within our $320 budget.
A sketch of the new design.
The final CAD model of the Textile Matrix.
The final design consisted of a central mounting plate with four aluminum extrusions, four buoys, and a control box which housed electronics, sensors, the compressed air, and valves to control the flow of air to the buoys. The buoys were to be made of an aluminum ring, a thin polymer sheet, an acrylic lid, and a 3D-printed mount. Pneumatic tubing is run from the control box to each buoy. A 1-square-meter textile is placed on top, and seeded with kelp.
We got to work fabricating custom components. I got trained on the waterjet at Tufts and manufactured some of the aluminum structures. My teammate lasercut the lids for the buoys, and together we assembled a test buoy.
My waterjet components, and my teammate assembling a buoy.
Testing the buoy with our compressed air system.
We attached adapters to our compressed air tank and tested the system with our solenoid valves. We were able to successfully inflate the buoys from the tank, but due to safety concerns, we decided to halt work with the air tanks and use lower pressure shop air as a proof-of-concept instead.
Next, we assembled the frame, finished the rest of the buoys, attached all of the pneumatic tubing, and mounted the textile to the frame. A teammate wired up a circuit that allowed us to open and close the valves, and wrote a Python script to control them. At this point, we were ready to test an inflation of the system!
Assembling the frame.
Inflating the four buoys by sending a command to open the air valve.
The final step of the build was mounting the electronics into the waterproof control box, and adding a water depth sensor to the box. I wired up an ADC (analog to digital converter) to the Raspberry Pi, and wrote a Python script to map voltage readings from our sensor to water depth. I then created a server that gave us the ability to inflate the farm over the internet, and read the live depth of the farm.
With the hardware and software complete, we took the Textile Matrix to the beach to test the buoys' ability to support the farm. The farm successfully floated on the surface, demonstrating its ability to hold kelp above the water while it dries.
Testing the kelp farm in the ocean.
To wrap up our Senior Capstone class, each team demonstrated and pitched their project. This was an opportunity to showcase our work throughout the semester, and answer any questions about the project.
Explaining operation of the farm to a professor and peers at Tufts.
The team at the product demo.
Overall, this project was pivotal in my engineering education, and I believe in the future I will point to this work as the foundation of many skills I use. I was able to see a vision through, from a seed of an idea all the way through manufacturing a prototype. This process was crucial in gaining a large number of technical skills, but also subtler knowledge such as perspective and wisdom on the engineering design process. Working closely with my teammates helped me improve at managing team dynamics, communicating design ideas, and having empathy for all perspectives. I am grateful to have had this experience, and to have the opportunity to work with three other excellent engineers.