Optimisation of Anaerobic Digestion

Optimisation of Anaerobic Digestion

Author : Steve Oxtoby, EngD student. His host university is Imperial College London and his project, 'Improvement, validation and implementation of predictive models on full scale anaerobic digestion assets' is sponsored by Thames Water and Severn Trent Water.

In my copious (2 week) experience of the STREAM programme, I’ve discovered that what makes it more exciting and (arguably) better than other doctorate options is the community feel between us streamers (the collective noun for people who are doing a stream EngD or PhD) as well as the wide variety of backgrounds and projects we are tackling. One 2017 streamer is a forestry student, others are geographers, chemical engineers (like myself), microbiologists, and assorted other science and maths disciplines. The projects are varied as well, mine is titled ‘Improvement, validation, and implementation of predictive models on full scale anaerobic digestion assets’ and is sponsored by Thames Water. I’ll be building on a previous project by improving a model and using it in order to improve industrial operation of anaerobic digesters, but more on that later. Other projects include using flow cytometry (… I know) in the water sector as well as such brilliant titles as “Rooting out Resilience” which I believe is a study into problems created during the transport of potable (drinking) water and lots of studies into various methods for wastewater treatment.

So that’s a fountain of wisdom about STREAM, now for a splurge regarding my project. In most (all) modern wastewater treatment plants, there is an initial settling/ dewatering stage. This splits into a liquid stream which will often be treated by an activated sludge process in which a tank is aerated to digest solids in the liquid, and a more solid rich stream, known as sludge. This sludge is where my project comes in. This sludge is anaerobically (without air) digested in order to produce methane (CH4), a gas most famously produced by cows. This can be burnt to produce energy and is generally valuable. This digestion takes place throughout several stages, acidogenesis and methanogenesis being a couple. The problem with this is that these stages make it quite a complex system to solve. When that is coupled with the competing constraints including that the sludge must be digested to produce a solid that can be used as fertiliser, as much methane as possible, and water that can be put back into the activated sludge process, it becomes even more complex. In short, it’s not easy to model but there’s obviously a lot to be gained from doing so. For the past 4 years a previous EngD has tackled that, my mission is to complete and capitalise on his work. Bet you wish you were reading the Geographer’s project now eh.

By now it should’ve become clear that this isn’t quite a traditional PhD, partially because its not. What made me want this project so much is the ability to make a difference in the short term. Potentially during the project but if not at the end of the 4 years I should have some tangible results and suggestions on how the process should be ran. That or a big fat failure of a model that oversimplifies or worse, overcomplicates, the problem such that it is beyond solving. A traditional PhD, a friend of mine is researching graphene, can end with neither progress nor failure, I don’t think I’ll have that luxury.

This is all for blog Episode 1, Episode 2 is set to arrive whenever it feels pertinent.