Purple Hydrogen

• 

  Bioreactor Design

• 

  Bioreactor Simulation

Inspiration

We all love cars, and we especially like the new wave of environmentally friendly cars that can run at efficiency that hasn't been seen before. We think that hydrogen cars could be the cars of the future, but we wondered why hydrogen fuels weren't taking off, even with hydrogen cars like the Toyota Mirai and Hyundai NEXO already being out in the market along with the recent completion of Rolls Royce's hydrogen plane engine, but when we looked into why we discovered a massive underlying issue in the entire industry of hydrogen production. We fell down a rabbit hole of research into what the hydrogen industry is, and we found out that hydrogen is required for so many manufacturing processes, but is produced in methods unfriendly to the environment. The largest source of hydrogen is literally just from burning oil, this disgusted us, but when we looked into environmentally friendly alternatives, we started to get even more angry. The most popular "environmentally friendly" method of hydrogen production, which has been heralded as the future of hydrogen production, is known as "blue hydrogen"; where the SAME process of burning oil produce hydrogen except now the carbon emissions are just "captured" by carbon capture technology. We already agreed that carbon capture technology, as it is described today, is basically a greenwashing scam, because it simply can not capture enough hydrogen to make a difference. And as it turns out, blue hydrogen plants barely capture any of the carbon they they emit (with average rates being well under 40%). This got us reaching out to experts in the area from Shell, and literally straight-up asking them "why is it so hard to sustainably produce the most abundant element in the universe?" and it turns out that it comes down to dollars and cents. It cheap for facilities to burn oil, and it's also cheap for them to pretend like they are going to capture the carbon they produce with bogus "carbon capture technology". So this gave us the inspiration to apply an actually legit technology to the solution that could make it cheaper, with bioreactors.

What it does

We designed a whole new process of hydrogen production called 'Purple Hydrogen', that utilizes bacteria to produce hydrogen from carbon dioxide and hydrogen sulfide. More specifically, this works by using a bioreactor to precisely control the environment that anaerobic purple sulfur bacteria grown in, to induce a form of anaerobic photosynthesis within the bacteria that produces hydrogen as a byproduct of the hydrogen sulfide. We wrote a article together that goes into detail about how our idea relates to the problem at: https://medium.com/@jacobappleton2/forget-blue-hydrogen-production-make-it-purple-52546e3f5824

How we built it

We built a simulator that can simulate the stoichiometry behind the bacteria's metabolic processes that produce hydrogen, to train an AI off of that can run the bioreactor at peak efficiency. Not only can this produce the most amount of hydrogen possible, from wasted hydrogen sulfide gas, but it can also do it pretty cheaply according to our AI's efficiency, lowering the cost of a kg of hydrogen to about the same price as burning oil. We used python for the computing, and AI of the solution, and then we used pygame to make an interface control board of the simulation, that could theoretically also be used to run the bioreactor. We made a video walkthrough of the AI and simulation at: https://youtu.be/gQKbwfYfhX0 The bioreactor was designed along with feedback and ideas from Clare Francois, the head of hydrogen innovation at Shell. The bioreactor would use IoT to be run remotely and precisely connected to the AI. We wrote a white paper on the solution at: https://docs.google.com/document/d/156HXuMlWVVYHFDfM-lizXADLOQgos0IcbHKuwIVg6Y0/edit?usp=sharing

Challenges we ran into

This was our first time working together as a team, even though we go to the same school. It was also our first devpost hackathon, which meant we had some difficulties navigating the site. When building the simulation, we basically had to learn through failing as none of us have done any frontend before, which was pretty time consuming and challenging for us. Time management, and continuing to work even when we couldn't be meeting in person were a couple other large challenges we faced, but we did our best to overcome it, and learn from the situation. Our biggest challenge actually ended up having to do with time management, because when we were submitting our video with just dozens of minutes left until the deadline, our loom file somehow got half corrupted when we tried to upload it to youtube for devpost, which induced a frantic scramble to re-record and upload the video, which made for a pretty interesting late night team call!

Accomplishments that we're proud of

We are proud that we were able to put ourselves out there to talk to some pretty cool mentors, especially our mentors from Shell who run the company's whole AI and hydrogen branches. We are also proud that we were able to pull together such a project in short amount of time that we can actually show and explain to people. Purple hydrogen is now something that we are all interested in working on into the future to ensure that the idea becomes a reality. And lastly, of course we are proud that we were able to get a cool simulation, with a UI, going with our very minimal front-end experience. You can see our github for the whole ideas simulation and AI at: https://github.com/tiwari-aryan/Purple-Sulfur-Bacteria-Hydrogen-Production-Bioreactor-Simulation

What we learned

We had to learn a lot about purple sulfur bacteria, and the stoichiometry behind anaerobic photosynthesis. But we also learned a lot about how companies like Shell develop projects like this, through talking to our mentors at Shell. We also learned how good a hackathon setting is for getting ideas, even for big companies like Shell. We learned a lot of front end development skills that are needed to communicate an idea, along with the ins and outs of pygame. We learned valuable lessons in time management, while pushing towards our deadlines. We also learned how to work seriously while we were meeting in person, and not just mess around like we would in school. Lastly, we learned a bit about ourselves, and how if we push ourselves, and get enough feedback, we can produce something that we are proud of.

What's next for Purple Hydrogen

Our whole team is pretty interested in the idea of actually trying to build a bioreactor, and grow purple sulfur bacteria to see if our AI can optimize the output of hydrogen. We plan to keep developing the idea along with Shell. We also intend to reach out to a university that is local to us, Ontario Tech, which has Clean Energy Research Facility that specializes in hydrogen, along with a microbiology lab, to see if we can get some lab time to conduct some research to prove if our process of purple hydrogen can work or not. We made a notion webpage with everything on it, that we plan to keep updated at: https://purplehydrogen.notion.site/Purple-Hydrogen-13803900f41b49c78ff7027affdf8e54

Our Team

We are a group of 3 high schoolers from the same school in Ajax Ontario. We have different backgrounds and skill sets, but we were united to do this hackathon through our passion of limiting the effects of climate change. We are also all strongly passionate about preventing the scams and greenwashing that plague the whole area of innovation around climate change, which is why we are a research-first team. We don't want purple hydrogen to end up like blue hydrogen, a decent idea on paper, but a detrimental idea when implemented, which is why we valued reaching out to so many mentors in the area to see if the experts think that the idea actually could be feasible. For now we haven't found any major roadblocks that would prevent our idea from coming to fruition, but as this hackathon is also a learning process for us, we are focused on seeking out the major weaknesses of purple hydrogen to see if it should actually be implemented.

Built With

• julia
• python