One of the big social problems that I have really become passionate about over the last year is sewer overflows. A sewer overflow is an event where the capacity of a pipe is exceeded and the remaining flow (sanitary sewage) is dumped into the river. My work at the sewer district has very rapidly accelerated my interest in this topic. Many of the professionals at MSD are very passionate about not putting raw sewage in the creeks, which to me makes a whole lot of sense. What I’ve come to realize is the sheer magnitude of possible solutions to this problem. There are thousands of methods of coming to the same conclusion and many ways of being effective for various dollar amounts. Over the previous one and a half years, I had primarily been focused on essentially building bigger pipes. While this is very effective, it also costs a lot of money.
This year, I have been involved in the remote monitoring department at MSD. What this department does is strategically place flow monitoring equipment throughout the sewer network to obtain data that can then be used for other purposes. These other purposes range from educating us for future projects to making decisions on how to operate our assets in the current time. The latter method is one of the cheapest methods we have to provide measurable reductions in overflows. We can use real-time data gathered by the 650+ sensors we have throughout Hamilton County to directly inform us of the conditions in our system. We can then use that data to hold water, operate gates, and activate remote wet weather facilities. This can all be done at the cost of a penny per million gallons. Compare that to the near 40 cents required to increase pipe capacity (size) and you have a good idea about just how cheap and effective it is.
As an environmental organization, we have a responsibility to be stewards to the environment. We also have the Ohio EPA harping on us to report our overflows within 24 hours. Part of the issue that we face is that sensors fail and conditional alarms can never be complex enough to capture every scenario. What I have done at MSD is set up a system where a team of employees ranging from unskilled help to highly skilled principal engineers can look at a potential overflow event utilizing a system where the lower skill levels (read: paid a lot less) can filter through all the false alarms and move the more complex ones to a more skilled and experienced engineer to handle. This system has been particularly effective at reporting this overflows and is something I am particularly proud of.
In the future, I would like to attempt to implement machine learning to predict the future state of the MSD sewer system. I am currently taking a course in machine learning and believe it could be an incredibly useful tool in overflow management. I will also be working on the EPA Campus Rainworks project in the fall with UC Society of Environmental Engineers, which is a challenge to build green stormwater infrastructure to help reduce stormwater entering our sewer system. I am definitely looking forward to a productive year!