Big Data and Water Safety

Marina Gutierrez (ESR11)


Some weeks ago, I started to work with some data from a wastewater treatment plant (WWTP), as part of my PhD thesis. I was working with this huge Excel file full of data on flow regimes, quality parameters, and operational conditions. The data was from the last five years, and it was incredibly accurate. As an example, I could check the wastewater flow entering the plant each minute of every day from 2018.

A few days later, I read a news article that I thought was fake. Some hackers almost poison a city in Florida by entering the control system of the drinking water treatment plant and increasing the concentration of sodium hydroxide1. A chemical used for increasing the pH level of the water and thus avoiding the release of heavy metals and corrosion, but very dangerous at high concentrations. Luckily, an operator passing by noticed the 100-fold increase of this parameter and reduced it on time before any harm to the city’s water supply was produced.

How can it be possible? Is it so easy to hack this kind of software? Were there examples of hacking of WWTP instead of drinking water treatment plants?

Many developed cities have created very creative and efficient ways to overcome the challenges related to the urban water cycle: Upgraded sensors and pumping systems to improve the operation of many automated systems; restructuration of the sewage system to better deal with natural events such as storms; reuse of water for groundwater injection or irrigation of recreational areas. All of them seem cutting-edge solutions for the optimization of the urban cycle. Still, it surprises me how fragile it may be and how it can still be improved in many ways. In the example of the drinking water treatment plant, it was enough to find out the password for the software used in the facilities to enter the system.

Before the internet era, much of the data obtained in WWTPs was gathered without a purpose or analyzed afterward. In the era of big data, I have the opportunity to work with a great amount of data online that allows me to understand very well what is going on inside the WWTP.

The point is that big data can benefit water issues on a local scale. New technologies with big data offer solutions to solve imminent issues (and don’t let hackers enter into the system!) or contribute to scientific research. Information collected allows WWTP operators (and researchers like me) to understand, manage and optimize the reliability and performance of the plant. Information becomes meaningful and useful to make better decisions and be more proactive during plant operations. No matter what specific tools or wastewater utilities are used. It is essential to have a management plan that pulls all the important data together and identifies trends or their absence, so potential problems are identified before they happen.

We are currently working on how to dosage activated carbon according to the necessities of the plant. It has been a very nice experience to learn about how the panel control and the security system works in this WWTP. Now, we can adapt to the fluctuations of the load influent and look for the best conditions to remove micropollutants on the real scale (and they let me push a lot of buttons on the control panel).

Picture source: