Known as the Rubber Capital of the World, Akron is Ohio’s fifth largest city and has a long tradition of innovation and civic engagement. The City of Akron Water Supply Bureau has a similar vaunted history, and it produces and delivers roughly 35 million gallons of water per day (MGD) to roughly 300,000 customers in the city and surrounding areas.
Akron continually invests in its water system to consistently meet or exceed Federal and State drinking water standards. Its plant draws water from the protected watershed of the Upper Cuyahoga River and is a “conventional” surface water treatment plant with coagulation/flocculation/filtration for particle removal and chlorine addition for disinfection. Specific chemicals used in various combinations to meet their treatment goals include potassium permanganate, chlorine dioxide, powder activated carbon (PAC), alum, and sodium hypochlorite.
Matt Ebie, utilities chief operator with the Water Supply Bureau, has led the team that produces the water, and has kept his eye out for innovations to improve plant performance. “One of the tools we use to reduce organics and control disinfection by-products is powdered activated carbon (PAC),” he says. PAC is dosed ahead of the filters to remove organic compounds or DBP precursor material prior to disinfection.
Beginning in 2015, a partnership with Dr. Chris Miller, a professor from the University of Akron’s Department of Civil Engineering, alerted the Water Supply Bureau that it may need to look at its PAC in a new light.
Like hypo [sodium hypochlorite] and caustic, Akron Water treated PAC as a commodity. But University of Akron research showed that different PAC products had different efficiencies depending on the specific composition of the organic matter in the water. “No two carbons are the same,” notes Miller.
On the advice of Miller, Akron purchased a fluorescence spectrophotometer to precisely measure the type of organics in the water, and Miller’s research team conducted experiments to verify that the different PAC products had different treatment efficiencies.
Miller’s analyses were immediately convincing. One PAC product stood out from the rest as having the highest efficiency with Akron’s raw water. But Akron had a problem – its procurement of PAC was based on price alone. Miller and his researchers helped Akron Water produce a detailed analysis that calculated the value of each PAC product based on both its cost and its treatment efficiency.
“Our PAC selection process is now performance and cost based,” explains Ebie. “Bidders get an equal opportunity – they provide samples and we run them through a test procedure using the fluorescence spectrophotometer and the testing software developed by Miller. We can normalize the price to performance. I now have a nice, documented process for the purchase decision I make.”
Developing a Treatment Strategy
Having identified the best value in PAC options for Akron, Ebie and his team turned to improving plant performance with respect to disinfection byproducts (DBPs). “DBPs are our greatest challenge, so we wanted to optimize treatment for that first,” says Ebie.
By 2017, Miller had launched Fontus Blue, Inc., with the goal of developing software tools to help operators improve treatment plant performance and efficiency. One early software product was called Virtual Jar. Virtual Jar takes in water chemistry and chemical use data from the treatment plant. Using advanced mathematics based on artificial neural networks (ANNs), it can accurately forecast treatment process performance for turbidity and organics removal.
Up to that time, Akron Water operators based their chemical dosing on the performance of their filters. Instead, Virtual Jar utilized other parameters including settled turbidity and fluorescence measures. Virtual Jar allowed Ebie and his team to “experiment” with different combinations of alum and other chemicals on a computer and arrive at an optimal treatment strategy to simultaneously remove turbidity for filter performance and minimize DBPs. “The software tool predicted settled turbidity, DBPs and cost. That really impressed us,” says Ebie.
The implementation team for Virtual Jar consisted, at first, of Ebie and two other process control team members who met with Fontus Blue monthly to explore treatment scenarios on the computer and come up with recommendations to the operations supervisor. With time, as the operators grew more confident in the predictions from Virtual Jar, more people joined the planning calls with Fontus Blue, and the calls became more frequent.
“Many of our operators started with a ‘we’ve already done it this way’ mentality,” notes Ebie, adding that there was some concern at the outset that a software model would simply be used to shave alum dosing to save money. “There was a bit of fear that you’d just look at cost and sacrifice water quality. But at our treatment facility, quality comes first.”
But sometimes, the best treatment scenarios use less chemical.
As the Akron team continued to refine its treatment strategies, they discovered that the optimal treatment resulted in less chemical use than they had previously relied upon. Furthermore, filter performance was substantially improved. Filter run times in the summer often dropped significantly relative to those in winter months.
“It used to be that an operator would see filter runs drop and turbidity rise, and they’d test an alum adjustment on two basins,” recalls Ebie. This manual approach to alum optimization usually took several days and a lot of work. “But once we had Virtual Jar, there was no guesswork anymore. Once we saw the water quality improvement, and also the improved filter runs, adoption took off across the organization,” he says.
Optimizing for Lead & Disinfection Byproducts
The City of Akron Water Supply Bureau has utilized a corrosion control strategy in place for a long time. Akron began to remove lead service lines 60 years ago that, at the outset, numbered as high as 40,000. “Akron was way ahead of the curve,” Ebie says of lead service line replacement. Today, the city has fewer than 4,000 lead service lines in the ground and more are removed every year.
Akron’s corrosion control consisted of a constant dose of zinc orthophosphate and maintaining a system pH of around 7.3. But with the success using Virtual Jar to reduce DBPs and improve plant performance, Ebie and the team were curious: Why is the ortho set at this level? Was there anything they could do to improve lead performance in the distribution system further?
“Ortho tends to be ‘set-and-forget’ it,” says Ebie. “But, would more be better? Would pH adjustment improve things? We just didn’t know.”
The Fontus Blue team returned with a new tool, the Corrosion Sentry. Corrosion Sentry utilizes advanced chemical models and proprietary insights to optimize orthophosphate addition and distribution system pH. It can be applied to systemwide analyses as well as predicting lead risk in premise plumbing.
It is well known that higher pH tends to reduce corrosivity and lower rates of lead release in the distribution system. But higher pH also increases the production of trihalomethanes, the most common disinfection byproduct. Systems attempting to reduce lead levels while simultaneously managing DBPs can be caught in a Catch-22: as one makes improvements in one water quality variable, you lose ground in the other.
Ebie and the Akron team then discovered the power of simultaneous optimization. By combining the software tools for DBP management with the tools for optimizing corrosion control, the team discovered they could avoid the Catch-22 and develop an overall optimum treatment plant that delivered improvements in both DBPs and lead. “It’s the best results we’ve ever had for lead,” says Ebie.
Human-Assisted AI Can Optimize Treatment & Reduce Costs
The best news came in 2020. In an otherwise challenging year, the Akron water team continued to refine its treatment operations using Fontus Blue’s Decision Blue toolbox. In 2020, Akron not only reduced lead to historically low levels and reduced DBPs, but the optimized use of chemicals resulted in a savings of $1.2 million for the city and a 25 percent overall reduction in chemical cost of treatment per million gallons.
“This is a fantastic tool,” says Ebie. “It helps operators arrive at a decision – not dictating one.”
The City of Akron is now allowing water treatment professionals from other cities in Ohio to sit in on their weekly calls with the Fontus Blue team. By giving others a glimpse of how the Akron team has integrated this powerful new tool into their daily operations, they hope to promote “best water practices” for their colleagues serving other communities.
This article was compiled by WF&M staff with input from Dr. Chris Miller, founder of Fontus Blue. Miller is an associate professor in the civil engineering department at the University of Akron and a registered professional engineer.
