Electro Active Membranes: Cut Fouling and Desalination Costs

Antoine: Hi Arian, welcome to the show.

Arian: Thank you for the opportunity and good to be here.

Antoine: I almost wanted to say welcome back because we already had a recorded conversation — it was on the Water Show when you were just winning the Tech Idol at the Global Water Summit in Berlin. We’ll come back to that because I’d like to understand how important that was on your path. I think I discussed that with Praash Gobindidan from Gradient because he’s one of the alumni of that prize. But before that, let’s really start with what you’re doing specifically.

Reverse osmosis membranes tend to foul and scale and you have a differentiated take on that. So first: is that the number one problem you’re aiming at? If so, what’s your differentiated solution?

Arian: To answer that a little better, I have to give you some of my background. Since late 2003, when I finished my PhD, I have had my entire career in desalination and membrane-based processes — 22 years now. I spent the first half of my career operating desalination plants. Back then it was a newer technology; not many people understood scaling and fouling, so I had a front-row seat to those phenomena and how they impact operations.

People say “we just clean it and it’s all good,” but as an operator I can tell you it’s a pest, a nuisance; it impacts operations. If you’re outside in 50°C heat and the desalination plant feeding a community goes down, every hour that plant is out because of a corrective action for fouling or scaling is an hour that community doesn’t have water. It’s the tail that wags the dog. Everything in a desalination plant is designed one way or another around mitigating scaling and fouling — from pre-treatment to hydraulics to the recovery you try to control. Every nut and bolt is related to controlling and mitigating scaling and fouling.

Now let’s take a step back. Why does it happen? It happens because membranes — whether from one company or another, whatever the material — operate as passive separators. You pass water through the membrane, you clean what you clean, a concentrate exits. The membrane itself is essentially a dumb plastic; it has no control over what it sees, what it separates, or how it fouls. So it relies on all the upstream processes to mitigate fouling and scaling.

Early in my career I used to imagine: if I could talk to these membranes each morning — “Hi, how are you? How’s the gypsum treating you?” — that was the seed idea behind what years later became Active Membranes.

Like every venture, Active Membranes grew from personal experience, circumstance and serendipity. In 2019 I started a company with colleagues Eric Hook and Rick Wolfen; Rick seeded the business and asked me to run it. That company, Pacifica Water Solution, still operates as a technology services provider to organizations like SWA and NEOM.

In 2022 we debated whether to reinvest in the consultancy or pursue a venture. Serendipitously, the work behind Active Membranes came out of UCLA for licensing. I know the inventor David Jassby and the research well — I’d followed it for a decade and always thought it could change the fouling and scaling game. So I took a licence.

What Active Membranes does is change passive separation into active separation — hence the name. If you give the membrane a way to control fouling and scaling without relying on upstream processes, you change the phenomena of scaling and fouling, and you change desalination and water separation. I don’t like passive processes; I like processes that take control. We’re giving the membrane theoretical control to do what it’s supposed to do.

We do that with electricity: we make the membrane electrically conductive. When the membrane is conductive you can apply electricity — in our case a low-potential AC (alternating current). That gives the membrane the power to repel charged species in the feed channel, which changes nucleation behavior of contaminants on the membrane surface, as long as those contaminants are charged. It changes separation, nucleation and related processes.

The bigger picture: when you send electrons back and forth to a medium you establish a language. You then learn that language by operating the system. Think of archaeologists who dig up texts — the language is there; you just have to learn how it works. I’m indulging the analogy a bit, but that imagination is the point. I want smart membranes that we can talk to, and that can talk back and tell us what’s going on and how to run at best efficiency. That is the key difference between Active Membranes and others: active versus passive separation.

Antoine: Let me unpack that. First, the story itself is interesting. You built a company advising on tech and became an expert in desalination. You studied and lived the problem, then the solution appeared next door and fit what you identified. If I try to visualise your tech, it’s like an electromagnet: apply electricity, and it attracts or repels charged stuff. Depending on the current changes, you can develop a “language” with the membrane.

What is the key message you get from the membrane? Is it “clean faster,” “perform this,” “I’m getting dirty — backwash me,” or something else?

Arian: The membrane can tell you when it’s getting dirty: “I’m getting dirty.” The voltage, current and frequency you apply might not be enough — the membrane signals that. Then you can change conditions to mitigate the problem, or over time collect enough data so the membrane can make that decision on its own. That’s the bigger picture.

For example, take a notorious scalant like gypsum (calcium sulfate). We’ve found that for certain voltages, frequencies and application patterns, the membrane shows anti-scaling behavior; for other conditions the effect worsens. We learn this because the membrane tells us — you see near-immediate changes in performance. Often with scaling and fouling you find out after the fact; here you get early warning and the potential for corrective action. The tools are the waveform, the frequency and the current. If we optimise those three parameters for each type of scaling, fouling and water quality, we can reach an optimal process. That is the basis of a smart desalination system: leaner, with less pre-treatment, cheaper, more sustainable and quicker to bring online.

Why doesn’t desalination scale everywhere today? Because of footprint, cost and complexity. Places like Dubai or Jeddah have the political will and money to make it work; smaller communities do not. We rely on desalination because 97% of our water resources are saline. To serve smaller communities — towns of 5,000–10,000 — desalination must be more cost-effective at small scales. My goal is to democratise desalination so the market can change and more communities gain access. You can’t do that without changing the separation phenomenon — and that’s what Active Membranes aims to do: make membranes smart.

Antoine: Thanks — that helps. To close the arc: if I use a simple analogy, it’s like a shower head that starts to scale. You can rub it off often and keep it working, but if you don’t it eventually dies. You’re saying the membrane can give early indicators and then you can use potential, waveform and current to “brush it off.” That raises a question: where do you put the intelligence — membrane, module, or skid? I imagine not fiber by fiber.

Arian: The medium is the module. We use spiral-wound modules — the form factor is standard and widely used. Our product is a spiral-wound module that’s easily retrofittable into existing pressure vessels without altering them.

We also made a small box we call the Active Box — essentially a function generator. You can hook it to one pressure vessel, an entire skid, or multiple vessels. It supplies the signal to the membrane (for example, four volts at one hertz as a sine wave) and you then observe plant performance to see whether fouling or scaling is mitigated.

In our pilots we’ve run head-to-head Active Membrane modules versus standard modules; monitoring shows whether the active device makes a difference. So the work happens at module level plus the Active Box that sends and receives the signal between the pressure vessel and the Active Membranes.

Antoine: You’re tempting me — do you run side-by-side comparisons? Your website mentions up to 45% higher recovery and up to 50% cost reduction. Can we qualify those claims? What would you see on brackish inland water, on an island, or on very salty produced water? Give some ballpark figures.

Arian: Absolutely. We’re looking at standard desalination applications: brackish and seawater. In brackish applications, 45% recovery is a bit high as a general claim — it’s more about footprint than recovery. We can operate at higher recovery with much less chemical use.

For example, in a pilot in Yuma, Arizona, an existing surface brackish system ran at ~74–75% recovery without chemistry. We ran an active train at ~84% recovery and then used that brine to run another stage at ~65% recovery, achieving about 92% recovery overall for the plant. The impacts: lifecycle cost reductions from lower chemical costs, lower waste disposal costs because there’s less brine, reduced membrane replacement and fewer cleanings. In one case the passive train had nine cleaning events over nine months; the active train had only three. Putting it together, lifecycle OPEX savings ranged roughly from ~32–60%, depending on the operating recovery.

In seawater, we see two clear advantages: eliminating chemistry for scaling mitigation and reducing biofouling. Because salinity is higher, the absolute benefit on scaling is less than brackish, but still meaningful — on the order of 20–30% cost savings. More importantly, we’ve identified and filed a patent on enhanced boron rejection. Boron is an issue in desalination — for example, in Israel seawater desalination is used for agriculture and boron in permeate is toxic to crops. Traditionally you raise pH (to ~10.2) to convert boric acid to borate and improve rejection — that requires significant NaOH dosing and subsequent acid dosing to neutralise, which operators dislike. We can achieve comparable boron rejection electrically at much lower cost and without handling large quantities of chemicals. That’s a big OPEX win.

Another value is footprint reduction. On greenfield produced-water projects we’ve eliminated chemical clarification steps, reducing footprint and cost by at least ~50% in one application. That cuts trucks hauling chemicals, reduces waste, and lowers environmental impact. Those are the core value propositions: higher recovery with less chemistry, lower OPEX, smaller footprint and better environmental performance.

Antoine: Just to be clear — can you apply current to any membrane type?

Arian: In principle yes, but we apply our coating. The coating — our secret sauce — is something we developed over the past 2½ years and iterated through two generations. Our current generation makes polyamide membranes far more conductive. To give a quick physics note: a polyamide surface might start at ~10^7 ohm per square (very resistive). After our coating, we bring that resistivity down to ~100–200 ohm per square — many orders of magnitude more conductive. That change is what enables the active phenomena.

Antoine: You were so interesting I broke my usual order and went deep into the history. But I want to understand how you arrived at that coating — it’s a lot of moving parts. Where did you start and how did you assemble them? Did the coating come first or the idea to apply current?

Arian: When UCLA decided to license the tech in 2020, David Jassby had a lab method; we had to transfer it into a field-ready process. Crazy as it sounds, there was a USBR competition called “More Water, Less Concentrate.” It was me, one membrane scientist and another colleague. We said: let’s make an electrically conducting membrane and take it to the competition.

To transfer the lab method we scaled up from lab sheets. Our first rudimentary setup was a Home Depot drum and a handheld paint sprayer — under $200, but it worked. From there we asked: how do we scale to larger, longer sheets for bigger modules? Long story short, we developed a coating machine we call the Spitfire. Spitfire uses spraying and heat to apply the coating at scale with low cost. Over the past year the development team focused on making the coating homogeneous and fast; we tried ~20 spraying technologies and settled on a robust method that gives the consistency required.

The Spitfire is inspired by rocking/weaving frames; we’re now on the second generation. The first generation got us through pilots; the second is more automated and efficient. As a membrane company you can either invest huge CAPEX in membrane manufacturing, or be creative. We deliberately chose a low-CAPEX, scalable approach so we can ramp capacity as demand grows without huge upfront investments. That’s how we moved from lab to commercial coating for 4-inch and 8-inch elements.