Betcha won’t drink it! The natural chemicals hiding in water, and how a new technique hopes to remove them using an unlikely ally- bacteria

Citation: Fu, Jie, et al. “Pilot Investigation of Two-Stage Biofiltration for Removal of Natural Organic Matter in Drinking Water Treatment.” Chemosphere, vol. 166, 2017, pp. 311–322., doi:10.1016/j.chemosphere.2016.09.101.
Om Nom Nom? Maybe don’t drink it.
Take a look at any of the photos of the recent flooding in Houston, Puerto Rico, Mumbai, Florida, and beyond and you’ll notice muddy brown waters. This color comes in large part from floating materials, aka dirt, but there’s an invisible story here: the heavy mixing of water and solids extracts other chemicals into the water, much like your washer extracts out chemicals from your dirty clothes.
These chemicals aren’t necessarily bad- in fact, they’re perfectly natural. We group all the many kinds of these chemicals into a category called natural organic matter (NOM). NOM can vary in composition depending on where you find it: NOM that comes from soils tends to have bigger, more complex compounds present as a result of decaying plants, than NOM found out in the ocean or in a lake, which largely comes from the breakdown of algae and bacteria.
If you take another look at this picture below of water from Texas after Harvey, another thought might cross your mind- I certainly don’t want to drink that! Which is fair- it doesn’t meet our basic standards for drinking water: clear/free from solids, no taste, no smells, and free from disease-causing pathogens.

NOM causes problems with our quest for high drinking water quality. In addition to contributing to taste, odor, and color in water, NOM is also food for certain kinds of bacteria- bacteria that the we try to kill during the drinking water treatment process itself. NOM also can contribute to pipe corrosion, and when it mixes with common disinfectants like chlorine, it can produce a variety of toxic and carcinogenic chemicals we call disinfection byproducts (DBPs), which we regulate to protect our health.
In addition to requiring extra procedures to remove these DBPs, NOM makes cleaning water more expensive in other ways. It requires higher doses of chemicals to treat drinking water, because NOM might interfere with the removal of some contaminants. If there is bacterial buildup (because, remember, NOM is food for microbes), you might have to replace filters and other parts more frequently.
Looking at the past two decades of data, we expect the problem of NOM to not only stick around, but to get worse: as temperatures rise, rainfall increases, atmospheric carbon dioxide increases, and acid deposition decreases, the amount of NOM in surface waters, where we get most of our drinking water, increases.
What to do about NOM
Finding more efficient and inexpensive ways of eliminating NOM from our water supply is a high priority for research. The current best practice is to add a step that filters water through granular activated carbon (GAC), which absorbs these NOM onto/into its pores. GAC is what you find in your Brita filter- like your Brita filter, it helps remove taste and odor compounds as well as DBPs precursors.

Also like your Brita filter, GAC filters don’t last forever. Depending on the flow rate and how NOM-filled the water is coming in, GAC has to be replaced frequently. That’s because over time, these pore sites get full of NOM and can’t absorb any more.
Fu et. al. piloted a design to get the GAC to auto-generate new sites using an unlikely ally: bacteria, growing within the GAC itself. As bacteria grow, they consume NOM on the surface or in the pores of the GAC, making those sites available to new NOM chemicals to stick.
A Pilot
The pilot plant, located at a water treatment plant outside of Atlanta, GA, consisted of a two-stage process to remove NOM and regenerate the GAC. This process was placed in the middle of standard drinking water treatment, so the water was treated to remove large solids first. Following this process, the water was further treated to complete disinfection (UV and chlorine additions).
The first filter was a simple sand/anthracite filter column to remove any remaining solids; and the second was the GAC/biofiltration filter column. These columns were added to the end of the standard treatment process.

As you can see from the above graphic, the group tested four columns, or two sets of this two-step process. As an environmental engineering research group, the engineers wanted to optimize the column, so they chose to test two designs: a passive biofiltration column, labeled Filter-1 and GAC-1, and an engineered column, labeled Filter-2 and GAC-2, where nutrients for the bacteria were added to the incoming water just before Filter-2. Adding nutrients would act like multivitamins to the bacteria, helping them eat even more NOM than their potentially nutrient-starved counterparts in Set Up 1. To see if this actually worked, the scientists varied the nutrients over the months of this study.
To evaluate water quality, many parameters were measured, but we only focus on turbidity and dissolved organic carbon. Turbidity reflects the number and size of particulates suspended in the water, while dissolved organic carbon describes the amount of NOM that is fully dissolved in the water. Both the particulates and the dissolved chemicals can contribute to NOM.
And… IT WORKED! But not like you’d think.
What Fu et al. report is that both the passive and engineered biofiltration paths resulted in high turbidity and NOM removal efficiencies. By sampling at many points throughout the two column set ups, the scientists were able to determine that the sand/anthracite filters were responsible for most of the turbidity removal, and GAC+bacteria columns played the dominant role in NOM removal.
Surprisingly, they found that the addition of nutrients (nitrogen and phosphorous) only had mild effects on improving water quality. While adding nutrients contributed to lowering dissolved organic carbon, it didn’t affect the overall levels of NOM.
It may be the case that the microbes degrade as much of the NOM as possible, hitting a ceiling; there are some compounds of NOM that bacteria just can’t digest, and adding more nutrients doesn’t make these compounds any more edible for bacteria.

Future Recommendations
To see what the bacteria can really do, Fu et al. recommended relatively strong pre-oxidation step prior to this two-phase biofiltration process- the drinking water equivalent of adding oxiclean to your water when doing laundry. This step would convert more of the NOM into biodegradable compounds, allowing the bacteria to have an even greater impact on NOM removal.
This is all to say…
To many, bacteria seem like the bad guys, playing the role of “germs.” We’ve all heard stories of uh… GI distress caused from drinking bacteria-filled water, to put it delicately.
But bacteria don’t always play this role in water, at least in wastewater; there, bacteria have helped break down wastes from many sources for decades, making dirty water clean enough to return to the environment or use in other processes. This journal paper expands on this role in the context of drinking water treatment. As water becomes an increasingly precious resource, designing efficient, economically feasible processes will mean cleaner, safer water, more accessible for all.