Latest / The Manufacturing Podcast with Fexingo: Factories, Supply Chains, and Industrial Business / How a South African Factory Turned Sewage into Clean Water
Transcript
- Lucas: So there's a beverage factory about forty kilometers outside Johannesburg — produces soft drinks, juices, the kind of stuff you'd find in any supermarket in southern Africa. And a few years ago, they were facing a crisis. Municipal water was becoming unreliable, prices were climbing, and during a dry spell in 2022, their supply was cut to just four hours a day. Luna: Four hours a day for a beverage plant? That's basically a shutdown. Lucas: Exactly. They couldn't run production. So they made a decision that sounded extreme at the time: they built their own water treatment plant on site. Not just for incoming water — they built a system to capture and treat every drop of process water they used, including the effluent from washing bottles and cleaning equipment. Luna: So they went from being a user of municipal water to basically operating their own utility. Lucas: That's the right way to think about it. The core of the system is a membrane bioreactor — that's a combination of biological treatment and membrane filtration. Wastewater goes through a tank where microorganisms break down organic material, then it passes through membranes with pores so small they filter out bacteria and most viruses. Luna: I've read about these in industrial settings. They're expensive up front, right? Lucas: They are. This installation cost about four million dollars — that's including the reverse osmosis system they added after the membrane bioreactor. But here's the thing: their water bill used to run roughly one point two million a year. And that was when supply was reliable. Once they factored in the cost of trucking in water during shortages, the payback period came in under four years. Luna: So the math works even without a crisis. But the crisis is what pushed them. Lucas: Right. And that's a pattern we see in manufacturing — necessity really is the mother of adoption. But what I found interesting is the regulatory angle. South Africa's water rights framework dates back to the 1950s, and it's built around the principle of 'reasonable use.' That's actually more flexible than, say, some European regulations that tie water reuse to strict quality standards designed for drinking water. Luna: Meaning, in South Africa, if you treat the water to a standard that's safe for your process, you can reuse it. You don't have to meet potable standards for non-potable uses. Lucas: Precisely. And that's a huge advantage. The factory's treated water is used for cleaning floors, cooling towers, and washing bottles — none of which requires drinking-water quality. Only the product itself gets a final polish with reverse osmosis. So they're not over-treating. They're matching the treatment level to the application. Luna: That's smart engineering, but it also sounds like a mindset shift. The plant manager must have bought in fully. Lucas: I spoke to the operations director — a woman named Thandi Mokoena, who's been at the plant for twelve years. She said the hardest part wasn't the technology. It was convincing the production team that recycled water wouldn't compromise product quality. They did a blind taste test with the treated water versus municipal water. The results were essentially identical. Luna: A taste test. That's brilliant. Low-tech, high-trust. Lucas: And it worked. Now the plant is processing about two hundred thousand liters of effluent a day, and they've got their water usage per liter of product down to one point six liters — that's a 55 percent reduction from their baseline in 2020. Luna: I want to ask about the energy cost. Membrane bioreactors are not exactly low-power. Lucas: Good question. The system consumes about forty kilowatt-hours per thousand liters treated. That's roughly double what a conventional municipal plant uses per volume. But here's the trade-off: they avoided a capital investment in a new well, which would have cost around two million dollars and would have been subject to depletion. And because they've reduced their reliance on the municipal grid, they're also less exposed to load-shedding — those rolling blackouts that South Africa is famous for. Luna: So it's not just about water security. It's also about energy resilience, because they can run the treatment plant on backup generators during blackouts and still keep production going. Lucas: Exactly. And they've since added a small solar array — about 500 kilowatts — that covers roughly half the treatment plant's energy needs during peak sun hours. That wasn't part of the original plan, but once they saw the energy load, it made sense to pair the two. Luna: So the water recycling investment became a platform for energy investment. That's the kind of compounding benefit that doesn't show up in a simple payback calculation. Lucas: Right. And it's worth noting that this isn't just a feel-good sustainability story. The plant now has a competitive advantage in a region where water is becoming the most constrained resource. They can guarantee supply to their customers. Their competitors, who still rely on municipal water, cannot. Luna: That's a concrete business edge. And it's the kind of story we love to dig into on this show — where a tough constraint forces a better solution. Lucas: You know, Luna, it's conversations like this that remind me why we keep making this show without advertising. We want to be able to talk about whatever's genuinely interesting, without worrying about pleasing a sponsor. If you find value in these deep dives, and you'd like to support keeping them free and independent, there's a simple way — buy me a coffee dot com slash fexingo. No strings, no perks, no tiers. Just a vote for ad-free business conversations. Luna: Yeah, that's exactly it. We're not selling anything. We're just sharing what we learn. And if that's useful to you, great. Now back to Thandi's plant — I'm curious about the next step. Lucas: So is she. They're now looking at zero liquid discharge — meaning no wastewater leaves the site at all. That would involve an evaporator and crystallizer that turns the final brine into solid salt. It's expensive — another three to four million — but if water prices keep rising, it could be the next logical move. Luna: And in the meantime, they've become something of a showcase. Are other factories visiting? Lucas: Yes. Thandi told me they've hosted tours for a dozen other manufacturers — a brewery, a dairy, even a textile mill. None of them have built their own system yet, but several are doing feasibility studies. The bottleneck isn't technology. It's that most companies still treat water as a fixed cost, not a variable one that can be managed. Luna: That's a framing issue. If you think of water as an unavoidable bill, you never look for alternatives. If you think of it as a risk factor that can be hedged, suddenly the investment looks different. Lucas: Exactly. And the data from this plant is compelling. They've gone from using 3.5 liters of water per liter of product to 1.6. Their overall water cost per unit is down 60 percent. And they haven't had a production stoppage due to water shortage since the system went live in early 2024. Luna: Sixteen months without a water-related shutdown. In a region where municipal supply is still intermittent. That's the kind of number that makes a CFO sit up. Lucas: It should. And it ties into a broader theme we've touched on before in this show — the idea that manufacturing resilience often comes from solving one specific, painful bottleneck. In this case, water. In other factories, it's been energy or labor or raw materials. But the pattern is the same: identify the constraint that threatens your continuity, and invest to remove it. Luna: And the solution doesn't have to be bleeding-edge. Membrane bioreactors have been around for decades. What changed was the willingness to apply them. Lucas: Right. The technology was mature. The economics had shifted. And a determined plant manager made it happen. That's the manufacturing story that doesn't get told enough — not the shiny new robot, but the quiet infrastructure upgrade that keeps the lights on and the water flowing. Luna: Well, we're telling it now. And I think that's a good place to leave it. Lucas: Agreed. Next time, we're looking at a factory in Ohio that's using old-school punch-card machines alongside modern CNC tools. The story of how they keep that legacy equipment running is genuinely surprising. Luna: Looking forward to it. Thanks for listening.