Drinkable - the water podcast

In addition to membrane separations, potable water PFAS treatment in this country runs primarily on two technologies: granular activated carbon and PFAS-selective ion exchange resin. There is no low-cost, readily available alternative sitting between them, which is exactly the gap Juergen Koller is trying to fill at Sequestro, and it is the subject of this week’s episode of Drinkable.

Koller spent about fifteen years in traditional chemical manufacturing before he started Sequestro. He is a trained synthetic inorganic chemist who specialized in process and plant-level projects. That manufacturing background and domain-competence underpins how he talks about his own technology. Most founders in this space lead with sales-stories about the chemistry in ideal conditions. Koller leads with the question of whether the chemistry can survive contact with scaling and whether he will be able to truly deliver what their benchtop testing suggests. This is something we see far too many Cleantech founders either ignore or get horribly wrong.

Founded on textile chemistry, rebuilt from wood

Sequestro’s origin runs through the University of Michigan, though not in the way people usually picture a university spinout. Koller was not a student, postdoc, or faculty member there. He lived nearby, walked into the school’s innovation partnerships office, and asked whether the licensing portfolio had a technology that fit a chemist with his background. That conversation led him to Anne McNeil, an Arthur F. Thurnau Professor of Chemistry at Michigan, whose lab had been developing cellulose modification chemistry originally aimed at cleaning dye contaminants like methylene blue out of textile wastewater. McNeil is Sequestro’s co-founder and Scientific Advisor, and the company licenses the underlying technology directly from her research program.

Koller looked at that chemistry and asked a manufacturing question rather than a scientific one: could the same functional groups be built directly onto a wood-based residual material, skipping the processed cellulose step entirely and going straight from raw sawdust to a finished adsorbent. The first working batch came from sawdust his own table saw produced while he was building a dog bed for a new puppy. Sequestro’s ArborSORB media took shape from there, built on intellectual property that the university has patented.

The underlying chemistry

The published science behind ArborSORB describes what the McNeil lab calls quaternized wood pulp; cellulose fiber grafted with cationic quaternary ammonium groups. Functionally, that makes it behave similarly to an ion exchange resin: a carrier material with functional groups on its surface that grab PFAS molecules electrostatically and hold them. The carrier is what’s interesting here…a conventional PFAS-selective resin uses a polystyrene bead. ArborSORB uses a wood particle instead, which changes both the raw material cost and the potential environmental footprint of the finished product.

The 2022 paper describing this chemistry, published by Koller’s co-founder and her research group, found that quaternized wood pulp removed more than 80 percent of PFOS and PFOA within seconds at environmentally relevant concentrations around 2.5 micrograms per liter, with maximum adsorption capacities of 763 milligrams per gram for PFOS and 605 milligrams per gram for PFOA, figures competitive with both conventional and newer PFAS adsorbents.

The 2022 study found the material was notably less efficient at capturing shorter-chain PFAS (<30% removal), which is not uncommon with organic sorbents. Sequestro’s current marketing describes ArborSORB as offering comprehensive capture of both long-chain and short-chain PFAS. That claim may reflect real, proprietary improvement beyond what was published in 2022, and Koller has said publicly that further functionalization work has shifted the material’s performance across the whole chain-length spectrum.

Technology Readiness Level 4

Sequestro describes its own progress as moving from Technology Readiness Level 4 toward active field pilot coordination. TRL 4, on the standard nine-level scale used across government and industry to describe how mature a technology is, means the material has been validated as a component in a laboratory environment, a meaningfully earlier stage than field-validated or commercially proven. That is a more precise way to describe where things stand than “pilot-scale” or “pre-commercial,” and it is worth keeping in mind against any comparison table that positions ArborSORB as already matching resin performance at carbon pricing. Those are targets the company is building toward, not independently confirmed outcomes.

The federal PFAS drinking water rule sets a maximum contaminant level of four parts per trillion (ppt) for PFOA and PFOS, a limit that pushes water testing and treatment into territory most people have no intuition for. One part per trillion is roughly a drop of food coloring diluted across five Olympic-sized swimming pools. Reliably testing for and treating PFAS compounds in drinking water is a real analytical and engineering achievement, and it asks something of small municipalities that many of them are not equipped to deliver: the testing infrastructure, the treatment technologies, and the funding to install and operate it all have to show up together, and right now they frequently do not.

That gap between a federal standard and a local municipality’s actual capacity is one reason the case for decentralized treatment keeps getting stronger.

Municipal systems can realistically aim to make water broadly safe at the city level, and Point of Entry (POE) or Point of Use (POU) treatment address the specific concern/s that matters most, which for most PFAS exposure is oral consumption rather than every gallon that leaves the tap. No single technology is going to solve PFAS everywhere at every scale. The honest answer, and the one Koller gives when asked why his technology can compete against established competitors, is that the scale of the contamination problem needs multiple approaches working at once, not a single winner.

The credential still matters, regardless of the media

Whichever adsorbent ends up in a given filter vessel or cartridge housing, the decision about what technology fits a specific water chemistry, a specific regulatory requirement, and a specific budget still belongs with a trained professional. A WQA Certified Water Specialist or Master Water Specialist is trained to evaluate the entire scope of the project, rather than just reaching for whatever product is easiest to sell.

The full conversation with Juergen Koller, including the story behind his first employee, the parts of being a CEO he finds harder than the chemistry, and his advice to his own younger self, is available now on Drinkable, wherever you get your podcasts, and on YouTube.

What a great time to be in the water business!

Sources and further reading

Harris, J. T., de la Garza, G. D., Devlin, A. M., and McNeil, A. J. (2022). Rapid Removal of Poly- and Perfluoroalkyl Substances with Quaternized Wood Pulp. ACS ES&T Water 2(2):349-356. DOI: 10.1021/acsestwater.1c00396.

McNeil, A. J. et al. Ionic-Functionalized Wood Pulp and Related Methods for Water Treatment. U.S. Patent Application Publication US 2022/0370982 A1, priority date May 21, 2021.

Ann Arbor SPARK (2026). Ann Arbor SPARK Invests $150,000 in Sequestro to Advance PFAS Water Treatment Technology. annarborusa.org.

Sequestro, Inc. ArborSORB Technical Specifications and Performance Advantages. sequestro.co.

Walk into almost any plumbing supply house in the country and you will hear some version of this: “soft water kills water heaters”. It gets repeated at kitchen tables, in online forums, and occasionally even in the trade press. The problem is that it is a significant oversimplification of a real electrochemical phenomenon, and oversimplifications lead to bad treatment decisions. Silly phrases like “You’re over-softening the water” get used, but they are also not fact-based.

The most common driver of accelerated anode consumption is elevated electrical conductivity in the water, not the mere presence of sodium ions (or the absence of Calcium).

I have been solving corrosion problems for three decades, from residential systems in Utah to mission-critical industrial and institutional systems on multiple continents. The corrosion-mechanism is not mysterious once you understand what is happening at the electrode interface inside your water heater. Once understood, the path to a permanent, low-maintenance solution becomes rather obvious.

How a sacrificial anode works

Modern steel water heater tanks are lined with a glass (porcelain enamel) coating, but microscopic defects, weld seams, and fitting penetrations inevitably expose bare steel to the water column. In the presence of water carrying dissolved ions (electrolytes), those exposed steel surfaces form galvanic cells: anodic sites where metal oxidizes and releases electrons, and cathodic sites where those electrons are consumed in reduction reactions. Without intervention, the tank will eventually corrode from the inside…this is natural and normal.

A sacrificial anode rod addresses this by using a metal that sits higher in the galvanic series than steel: magnesium, aluminum, or an aluminum-zinc alloy. The anode material is electrochemically more active, so it oxidizes first (preferentially), releasing electrons that suppress corrosion at the steel surfaces. The rod sacrifices itself so the tank does not have to. The rate at which that sacrifice occurs depends on three primary variables: temperature, the electrochemical potential difference between the anode and the tank steel, and the electrical conductivity of the water acting as the electrolyte. That last variable is the one most relevant to the softener conversation.

What ion exchange does to your water chemistry

Ion exchange softening is a substitution reaction. The cation resin preferentially strips divalent calcium (Ca²⁺) and magnesium (Mg²⁺) ions from the water and releases sodium (Na⁺) ions in their place. Every divalent hardness ion removed is replaced by two monovalent sodium ions to maintain electrical neutrality. That increased ion count drives measurable changes in water conductivity after softening.

A cation exchanger targets only positively charged ions and has no affinity for anions, so anions like carbonate, chloride and sulfate that are present in the source water pass through the softener unreacted. Those anions remain in the softened water at essentially the same concentrations as in the feed water. Softening does not add chloride or any other anion to the product water.

Physical total dissolved solids (measurable with a dry-down test), do not change dramatically and the mass balance is roughly equivalent because two sodium ions weigh roughly the same as one calcium or magnesium ion at comparable concentration.

The conductivity increase after softening is real…in my experience, the post-softening conductivity increase averages around 10%, though that figure can vary considerably with source water composition and should not be treated as a universal constant.

Proper softener programming and adequate rinse cycles are also critical. Older systems or poorly programmed units that allow regeneration byproducts to carry through will produce brief conductivity spikes that field technicians have historically blamed on softeners as a category, rather than on the specific operational failure causing the slug of salty regeneration byproduct that can have TDS levels >10,000 mg/L. This is very important to explore and diagnose properly.

Bradford White’s published technical guidance on anode rod selection confirms the conductivity framing: “…in conditions where water is more conductive, including softened water, the sacrificial process in magnesium anode rods is accelerated, and aluminum rods become the preferred material choice for those applications”.

Their guidance goes further, suggesting that installers target 10 to 12 grains per gallon of residual hardness rather than treating to zero, on the basis that lower conductivity will extend sacrificial anode life. I disagree with this assertion. It is a simplistic view, since residual hardness does not completely protect against corrosion, especially when the natural conductivity of the water is elevated.

As this article documents in detail, scale is non-uniform, and the LSI is an unreliable corrosion predictor in softened water. The WQA’s technical position is that hardness neither causes nor prevents tank corrosion. Bradford White’s 10-12 gpg target reflects a warranty-protection posture for its sacrificial anode products, not a scientifically defensible corrosion management strategy. Under-softening the water to coax another year out of a rod is engineering around the wrong challenge.

Any source of elevated conductivity has an impact

Ion exchange Softened water is not uniquely aggressive. Conductivity-driven anode acceleration occurs whenever the dissolved ionic load is elevated, regardless of which ions are responsible or where they came from. A raw water source carrying 500 mg/L of TDS in calcium, magnesium, sulfates, and bicarbonates with zero sodium will consume an anode faster than ion exchange softened municipal water with 250 mg/L of TDS.

The WQA Corrosion Educational Kit, authored by Joseph F. Harrison, PE, CWS-VI, and Duane Nowlin, PhD, identifies high mineral concentration as a primary corrosion accelerator because it increases electrical conductivity and therefore the current flow in electrolytic cells. That same document states the conclusion directly: ion exchange softening neither causes nor controls corrosion.

The reasoning follows directly from the mechanism itself. Ion Exchange Softening does not change pH, dissolved carbon dioxide, dissolved oxygen, or total mineral concentration. Those are the actual corrosion drivers. Ion exchange changes which cations are in solution; it does not alter the other factors that govern corrosion rate. This is WQA’s official technical position, written by credentialed professionals, and it aligns precisely with the conductivity-based analysis here.

The Langelier Saturation Index (LSI) provides a useful but limited professional tool. The LSI calculates whether a given water is likely to precipitate calcium carbonate (scale-forming, positive LSI) or dissolve it (corrosive, negative LSI), using pH, temperature, TDS, alkalinity, and calcium hardness as inputs. Softened water, with calcium hardness removed, shifts the LSI in the negative direction: the water becomes undersaturated with calcium carbonate and therefore (according to Langlier) potentially more prone to dissolving carbonate scale.

The WQA Corrosion Educational Kit cautions, however, that the LSI should not be applied directly to ion-exchange-softened water as though it were equivalent to naturally-soft acidic water, because the two are chemically different. Softening does not change pH, alkalinity, or CO₂ content to resemble natural soft waters. Stumm and colleagues, cited in that same WQA document, put the broader limitation plainly: corrosion in natural waters depends on so many interdependent variables that no simple index adequately describes corrosive potential. The LSI is a scale-tendency tool, not a complete corrosion predictor. Use it as one data point among several, which is exactly why a full water analysis by a qualified professional is the non-negotiable first step in any corrosion manag…

A jury sat through seven days of testimony. The suit was brought by Brian Miller, a Kansas dairy farmer with decades of tractor experience. It heard from his experts, from CNH’s engineers, and from Miller himself about the moment he jumped off a moving tractor to clear an object from its path and was crushed under the seed drill it was pulling. The jury found for Miller and put sixty percent of the fault on CNH Industrial America. Then the Tenth Circuit reversed the whole thing.

That reversal, handed down July 21, 2026 in Miller v. CNH Industrial America, is worth your attention whether you build tractors, water softeners, reverse osmosis systems, or chemical dosing equipment. We operate in an era of record setting jury verdicts, an era where a plaintiff’s lawyer can win seven days of trial, sixty percent fault, and still lose everything on appeal. Understanding exactly where the law still protects manufacturers and dealers from claims dressed up as warning failures, when the real issue is a person ignoring a danger everybody already understands, is not optional anymore. It is survival.

The facts

Miller had operated tractors his whole working life. On the day of his injury he slowed his tractor to a crawl, between four and five miles an hour, and stepped off while it was still moving to retrieve something in his path. The tractor kept rolling. Its thirty foot seed drill ran over him.

CNH’s five hundred page manual told him more than once not to dismount a moving tractor. It told him to park on level ground, set the parking brake, lower the implements, and stop the engine before leaving the seat. The cab had an alarm that sounded if an operator left the seat without setting the brake. Miller argued none of that mattered because one line in the manual, stating the park brake engages automatically if the operator leaves the seat for more than five seconds, should have told him more clearly that this automatic engagement would not actually stop a tractor already in motion.

The Tenth Circuit was not persuaded. Kansas law, like the law in most states, excuses manufacturers from warning about dangers that are patent, open, or obvious to a reasonable user. The court’s own words carry the point better than any paraphrase could: a reasonable person should know that leaving a moving vehicle and stopping in front of it is likely to cause injury. Miller’s own expert witness testified that stepping off a moving tractor would require someone to be, in his words, suicidal.

The procedural twist

Whether a danger is open and obvious is what a jury decides after hearing arguments from both sides. That is precisely why Miller’s win at trial seemed final. The Tenth Circuit explained that this general rule has an exception…when the evidentiary record points in only one direction, the fact question becomes a legal one a court can resolve, even after a jury has already ruled the other way (Miller v. CNH Industrial America, 2026).

That is not a small technicality. It means a defendant who loses at trial is not necessarily out of options if the underlying record never actually supported a contrary conclusion. CNH renewed its motion for judgment as a matter of law after the verdict, preserving exactly this argument. The lesson for anyone facing litigation in our industry is straightforward. Document the record during discovery and trial as though the appeal is already being written, because it often is.

The Restatement (Third) of Torts frames it plainly: a warning is not required when the product’s own obvious danger is already the warning. The Tenth Circuit has applied this logic before, holding that a reasonable tractor operator would already understand the risk of falling off a tractor without a seatbelt (Ahrens v. Ford Motor Co., 2003). The Kansas Supreme Court reached a similar conclusion when it held that manufacturers can assume ordinary users already know that alcohol burns and knives cut, without a label reminding them (Delaney v. Deere & Co., 2000).

Contrast this with Kincaid v. Kubota Tractor Corp. (2020), a case Miller’s counsel leaned on and lost. In Kincaid, a passenger riding on the front of a skid steer was injured when its grapple attachment trapped him against the cab. The Kansas Court of Appeals let that claim proceed because the plaintiff had not read the operator’s manual and did not understand the exterior warning sticker as a general prohibition against riding on the machine. The difference is not academic. Kincaid turned on what one specific person actually knew about an unusual practice. Miller turned on what any reasonable person already understands about tractors in general. Subjective ignorance of a specific warning is a different animal from an objectively obvious hazard, and courts treat them differently.

When warnings become theater

Here is where I want to editorialize a little, because three decades in this trade has taught me that the fear behind Miller is not new, and manufacturers have been reacting to it for decades in ways that would be funny if they were not also so expensive and time-consuming. In 2009, the Foundation for Fair Civil Justice ran its annual Wacky Warning Label contest and gave the top prize to a small tractor manufactured by New Holland. The label read, in full: avoid death (Forbes, 2011). New Holland, for anyone who does not follow the machinery world closely, is a brand under the same corporate umbrella as the defendant in Miller. Same industry, same instinct, the same fear of being the next headline.

That fear did not appear out of nowhere, it came from decades of manufacturers watching juries punish companies for warnings that were genuinely inadequate, then overcorrecting into warnings that insult the reader’s intelligence. Laundry machine tags that tell you not to put a person inside. Stroller tags that tell you to remove the child before folding it up, a label that probably exists because someone, somewhere, actually folded a stroller with a child still in it and then sued over the lack of a label telling them not to. Chainsaw labels warning against grabbing the spinning blade with your hands, prompted by a real injury and a real lawsuit that a jury found in the plaintiff’s favor.

I want to be fair here, because it is easy to wave all of this away as frivolous litigation run amok. The case most people cite as the poster child for lawsuit abuse, the McDonald’s hot coffee case, gets treated as a punchline more often than it gets treated honestly. Stella Liebeck suffered third degree burns requiring skin grafts after coffee served at 180 to 190 degrees Fahrenheit spilled in her lap, and internal McDonald’s documents showed the company already knew of hundreds of prior burn complaints before her injury. That is not a story about a woman blaming a company for the ordinary fact that coffee is hot. It is a story about a company that had already been warned, repeatedly, and chose not to change anything until a jury made the decision expensive enough to matter. Genuinely dangerous products deserve genuinely serious warnings, and companies that know about a hazard and sit on it deserve to the consequences.

The problem is not that our legal system holds companies accountable for real, non-obvious hazards. The problem is that the same system gives plaintiffs’ counsel every incentive to argue that any hazard, however obvious, was actually a documentation failure in disguise. Miller’s team tried exactly that argument, built around one ambiguous sentence buried in a five hundred page manual, and it took a full trial and an appeal to sort out the difference between a real warning gap and a manufactured one. That process is expensive, even when you win. Ask any manufacturer who has sat through a seven day trial defending a case they eventually won on appeal whether that felt like justice moving efficiently. Either way, it cost you money and time.

Prop 65

Nowhere is this warning fatigue more visible than California’s Proposition 65. Our industry sh…

There is a meme that captures something I have watched play out in boardrooms and jobsites for thirty years. A golden retriever sits at a restaurant table, ignored, looking miserable. The label on the dog: “Engineering team, knowing the solution is not technically possible.” At the table, happy and chatting: “Sales team presenting the solution in PowerPoint” and “Excited customer.”It is funny. It is also a slow-motion organizational disaster, and most water treatment companies treat it as an unavoidable feature of doing business rather than a correctable failure of management.In this episode of Drinkable, I go after the full picture: the handoffs nobody manages well, the compensation structures generating the problem, what Frank Blau, Jim Abrams, and Nexstar Network figured out the hard way, the private equity consolidation reshaping our channel, and what independent dealers can do that no PE-backed platform ever will.

Companion episode to this article:

About this episode

Every service call in the water treatment industry is a trust problem before it is a technical problem. In this episode, Greg draws on behavioral science research from a former FBI special agent to explain why homeowners resist the very professionals who can help them most, what that resistance actually signals, and what skilled technicians can do about it in the first five minutes of a service call.

The conversation covers the psychology of threat response, the difference between sympathy and empathy in a customer interaction, why water quality problems create a uniquely high vulnerability gap compared to other home service trades, and how WQA certification gives both technicians and homeowners a concrete bridge across that gap.

If you work in home service, manage technicians, or train water treatment professionals, this episode is for you.

Reference

Schafer, J. R., & Karlins, M. (2015). The Like Switch: An Ex-FBI Agent's Guide to Influencing, Attracting, and Winning People Over. Touchstone / Simon & Schuster.

ISBN 978-1-4767-5448-2. Available wherever good books are sold.