It may not be something you have thought about before, but wastewater treatment plants are highly sensitive to both the weather and human behaviour.
The operation of a wastewater treatment plant can vary dramatically during intense rainfall or a public holiday.
We all know that the Covid-19 pandemic had a profound impact on people’s daily lives, with lockdowns closing schools, businesses and recreational facilities. This not only changed the behaviours and movements of billions of people, these changes impacted wastewater treatment plants.
Accurately predicting the flow-rate of untreated water into wastewater treatment plants (WwTPs) is important for their operation. With the ability to accurately predict the influent flow-rate, the operator can plan for the efficient use of resources and equipment.
“We leveraged new machine-learning techniques to enhance our ability to predict wastewater influent flow-rates within the context of the Covid-19 lockdown situation.”
Now a research team from Canada and China has found that online continuous machine-learning models provide the best information in complex and fast-changing scenarios like pandemics and storms, over and above offline machine-learning.
Data-driven models that predict influent flow-rates are proven to be highly effective tools for operators to use, but most early studies have focused on offline batch-learning. With batch-learning, the data is gathered over time, and the machine-learning model is trained from the data in batches, offline.
These models have proven to be inadequate for wastewater prediction in the era of COVID-19 when the influent pattern changed significantly. In online-learning, the model is trained as new data arrives, which is proven to be more effective.
While batch-learning models typically work faster and require less computational resources, they tend not to be as flexible in handing large and fast-changing datasets as online-learning models. During the pandemic, the research team turned their attention instead to online-learning models, to determine whether they could overcome some of these limitations.
“We leveraged new machine-learning techniques to enhance our ability to predict wastewater influent flow-rates within the context of the Covid-19 lockdown situation,” said Pengxiao Zhou, a civil engineer at McMaster University.
The potential application for this work is that the developed models can be integrated into commercial wastewater modelling software. The online-learning models developed by the team are called Adaptive Random Forest, Adaptive K-Nearest Neighbors and Adaptive Multi-Layer Perception, and are based on conventional batch-learning models.
The team used their newly developed online-models to predict the changing influent flow-rate that resulted because of Covid-19. They developed the models using three to four years of hourly influent flow-rate data and meteorological data collected from two Canadian WwTPs.
They compared the new online-learning models to their respective conventional batch-learning models in predicting the influent flow-rate at the two plants using two different scenarios. In one scenario, there was a 24-hour ahead prediction and in the other there was no lead-time prediction.
The online-learning models produced accurate predictions under changing data patterns and were efficient in dealing with continuous and large influent datastreams. The team found the online-learning models to be superior to batch-learning models.
“The proposed new online learning models can provide more robust decision-support to wastewater operators or managers for coping with changing influent patterns due to emergencies such as Covid-19,” said Zhou.
Looking ahead, the team’s future studies will include more case studies and consider more prediction scenarios to further validate the developed models.
“The ultimate goal is to provide reliable tools for wastewater management and propel the development of wastewater intelligence,” he concludes.
The research team includes Pengxiao Zhou and Zhong Li from McMaster University, Canada; Yimei Zhang from North China Electric Power University; and Spencer Snowling and Jacob Barclay from engineering consultancy Hatch in Canada.