Internet of Things

How we used IoT data to optimize wastewater treatment

Share this post:

Tweaking wastewater treatment operations allowed us to dramatically lower costs and reduce waste

A city with 2 million people will likely have a wastewater treatment plant (WWTP) that handles hundreds of millions of gallons of sewage in a day. This is about the amount that would fit into all 2 million of those residents’ bathtubs! Treating this amount of sewage uses tens of millions of dollars in electricity. By just saving 10 to 20 percent of this energy, we can save millions of dollars per year. This is just one of the things my team at IBM Research-Haifa accomplished in an Internet of Things pilot to improve efficiency at a treatment plant in Lleida, Spain.

Wastewater treatment plan in Lleida Spain

Wastewater treatment plant in Lleida Spain

We joined forces with Aqualia Spain, the world’s third largest private water management company, and developed a technology for operations optimization using sensors and analytics. Up until now, the plant planned their settings and adjustments of resources on a seasonal basis. This was done based on experience and intuition. Our goal was to use mathematics and analytics to improve the process, reducing the energy and resources used, while improving the compliance with standards for water treatment.

Starting in November 2014, our system was in use every day of the week, around the clock for one calendar year, with recommendations updated every 2 hours. The system use showed a dramatic 13.5 percent general reduction in the plant’s electricity consumption. The plant also uses its resources more effectively, with a 14 percent reduction in the amount of chemicals needed to remove phosphorus from the water, and a 17 percent reduction in sludge production. Other benefits include significant improvement in total nitrogen removal, especially in low temperature conditions. In addition to these measurements, the plant observed a significant improvement in the control of effluent parameters and satisfaction with regulations.

Creating electricity from bio-waste

Wastewater treatment is the process of cleaning sewage that enters the WWTP from various sources. The WWTP takes in sewage, which is often wet waste from processes from restaurants or factories. After the main solids settle out of the sewage, the general treatment process uses bacteria that ‘eat’ the sewage and separate it into clean water and solid sludge. To carry out this treatment process we need energy, oxygen, and chemicals. The energy, turned into electricity, runs the plant; the oxygen aerates the sewage and activates the bacteria; and the chemicals remove the pollutants from the water.

We used sensors, simulation models, and optimization algorithms to better understand the plant’s state and how it impacts the plant’s need for oxygen, energy, and chemicals. Then we could produce and plan for the operations and how to adjust the use of these resources to be most efficient and have the least amount of waste.

The system use showed a dramatic 13.5 percent general reduction in the plant’s electricity consumption. The plant also uses its resources more effectively, with a 14 percent reduction in the amount of chemicals needed to remove phosphorus from the water, and a 17 percent reduction in sludge production.

Wastewater treatment plant operations

Wastewater treatment plant operations

Plants consists of three main parts: a liquid line, a sludge line, and a gas line. First, the sewage is cleaned in the liquid line. The clean water is diverted to the nearest lake or river, and the solid matter is delivered to the sludge line for additional treatment. The sludge line produces water that is sent to the liquid line again for additional treatment, sludge that is loaded on the tracks and taken away from the plant, and methane gas that is used in the gas line as a source for self-produced electricity. The plant wants to treat water to the level required by the regulator, while keeping expenses as low as possible. This is done primarily by controlling the concentration of oxygen in the reactor, adjusting various pump rates, and adding chemicals.

Using IoT to conserve resources

For WWTPs such as Aqualia’s, there are challenges due to the dynamic nature of the operations. For example, different volumes of water come in at different times of day or night. Moreover, in cold weather, the bacteria used to break down the wastewater behave differently. Electricity rates can also vary at different times of day or night. But using IoT to optimize various systems has the potential to change how we deal a WWTP, along with many aspects of daily life, such as healthcare, transportation, insurance, and many other industries. With information coming in from so many sensors everywhere, our vision is to provide more insight into how we use our world’s resources and how we can conserve them in a way that makes sense.

More Internet of Things stories

IBM Researchers Design Winning AI Algorithms for Drones

For the second year in a row, researchers at the IBM-Illinois Center for Cognitive Computing Systems Research(C3SR) won a competition challenging experts worldwide to design low-power embedded systems for Internet-of-Things (IoT) applications. The 2019 Design Automation Conference (DAC) System Design Contest’s objective: create algorithms that can accurately detect and locate objects from images taken by […]

Continue reading

Helping to Improve Medical Image Analysis with Deep Learning

Medical imaging creates tremendous amounts of data: many emergency room radiologists must examine as many as 200 cases each day, and some medical studies contain up to 3,000 images. Each patient’s image collection can contain 250GB of data, ultimately creating collections across organizations that are petabytes in size. Within IBM Research, we see potential in […]

Continue reading

Video Scene Detection Using Optimal Sequential Grouping

Our team at IBM Research recently developed a new approach for automated video scene detection. Videos are used today for everything from entertainment and marketing to knowledge-sharing, news, and social journaling. Automated scene detection can help consumers and enterprises utilize this video content in new ways. Video scene detection is the task of temporally dividing […]

Continue reading