Latest / The Operator's Manual with Fexingo: Daily Decisions, Tactical Wins, and Business Execution / How One Brewery Cut Water Use 35 Percent With a Single Sensor
Transcript
- Lucas: There's a craft brewery in Portland — we'll call them River Bend Brewing — that was spending about six hundred thousand dollars a year on water and wastewater combined. That's pretty typical for a brewery their size: about seven barrels of water for every barrel of beer. Luna: Seven to one — that's the industry benchmark, right? Most people don't realize how much water goes into beer. Lucas: Exactly. And the brewery had already done the obvious stuff: low-flow nozzles on the hoses, fixing leaks, training staff to not leave the hose running. They'd plateaued. The head brewer figured they were as efficient as they could get without major capital — new equipment, a wastewater pretreatment system. Luna: But that's not the story, I assume. Lucas: No. What they did instead was buy one $400 inline flow sensor and clamp it onto the main supply line coming into the brewhouse. Just one sensor, logging flow every 15 seconds to a simple dashboard. Luna: One sensor? That's the whole intervention? Lucas: That's the whole intervention. They didn't change a single pipe or buy a new machine. They just started measuring — and looking at the data in real time. Luna: What did the data show? Lucas: Within a week, they noticed a spike every morning around 9:30. A two-minute burst of flow at about 40 gallons per minute. It happened like clockwork, seven days a week — even days they weren't brewing. Luna: So something was cycling on automatically. A wash cycle? A CIP cycle? Lucas: That's exactly what it was. One of their clean-in-place systems — the automated tank washer — had a valve that was failing intermittently. Every morning during its self-test cycle, it would open and dump about 80 gallons of hot water straight to the drain before the test failed and the valve closed. Nobody had noticed because the system's error log only recorded a generic 'test failure' — no indication that water was involved. Luna: So they were literally flushing money down the drain, every single day. 80 gallons a day, 365 days a year — that's almost 30,000 gallons. Lucas: 29,200 gallons. Plus the energy to heat it, plus the sewer charges on that volume. They replaced a $30 valve seat, reprogrammed the test cycle to start with a compressed-air check instead of a water flow check, and the spike disappeared overnight. Luna: But that's a one-time fix. Where does the 35 percent figure come from? Lucas: The valve fix alone got them about 5 percent. The rest came from a cascade of similar discoveries over the next three months. Once they could see flow data at 15-second intervals, they found three more behavioral leaks: a keg washer that was running a rinse cycle thirty minutes longer than necessary because the timer was set wrong, a floor drain that constantly had water running — from a hose valve that was only partially closed — and a heat exchanger that was over-pumping cooling water because the flow control had drifted. Luna: Each one small. But cumulative. Lucas: Exactly. And none of them would have been found without granular measurement. The old monthly utility bill only showed total consumption. You couldn't see the spikes. The head brewer described it as 'trying to find a leak in your house by looking at your water bill once a month.' Luna: So the sensor paid for itself in what — a day? Lucas: The $400 sensor paid for itself in less than a week, just from the valve fix. Over the first year, the brewery cut water consumption by 35 percent and saved $210,000 total — $140,000 on water, $70,000 on wastewater treatment surcharges. Their water to beer ratio dropped from 7:1 to about 4.5:1. Luna: That's not just an environmental win. That's a 35 percent reduction in a direct operating expense, with no capital outlay. That's the kind of thing that goes straight to the bottom line. Lucas: It is. And the interesting part is what happened next. The brewery shared their data informally at a regional craft brewers' meetup — just a slide in a presentation — and within six months, five other breweries in the Pacific Northwest had installed similar sensors. One of them found a stuck valve in their glycol chiller that was wasting 50 gallons of water an hour, 24 hours a day. That's 438,000 gallons a year from one valve. Luna: That's crazy. And all because someone was willing to look at a 15-second flow log. Lucas: The principle here is straightforward but surprisingly rare: you can't manage what you don't measure at the right resolution. Most companies measure at the monthly or quarterly level, which is fine for financial reporting but useless for operational improvement. The signal is buried in the noise of aggregation. Luna: So the lesson isn't 'buy a flow sensor.' The lesson is 'measure at the frequency of the problem.' Lucas: Exactly. The problem — a valve cycling for two minutes — happens at a 15-second resolution. A monthly reading could never detect it. The right measurement frequency is the one that matches the rhythm of the waste. Luna: That's such a good framing. And it applies everywhere, not just water. Energy, labor, material yield — any input you're consuming consistently. Lucas: Totally. And the barrier is almost never the cost of the sensor. It's the discipline to look at the data. River Bend didn't even have an automated alert set up. The head brewer just glanced at the dashboard every morning over coffee. He noticed the spike on day three. Luna: That's a good reminder too — the tool only works if someone's paying attention. But the cost of attention is also low: thirty seconds of looking at a graph. Lucas: Right. And that's the kind of thing that scales. You don't need a data science team or a digital twin. You need a $400 sensor and someone who's curious. Luna: You know, it's funny — we talk a lot on this show about the power of a single small change. And this episode is a perfect example. A $30 valve seat and a $400 sensor, and you save over $200,000 a year. That's not a moonshot. That's just paying attention. Lucas: And if conversations like this one have made your work just a little sharper — or saved you from chasing the wrong metric — then that's exactly what we're here for. The Operator's Manual is ad-free and listener-supported, and the way that stays true is through people who find value in it chipping in at buy me a coffee dot com slash fexingo. That's it. No tiers, no perks. Just keeping the show independent and focused on real operations, not ad reads. Luna: Yeah. And honestly, it's because of that support that we can spend an entire episode on a $400 sensor without having to sell you a mattress or a meal kit. So thank you. Lucas: Alright. Back to the brewery — one more detail I think is worth noting. After the first year of savings, River Bend used the data to negotiate a better water rate with the city. They went to the utility with their consumption history and said, 'Look, we've reduced our demand permanently. Can we get a volume-based discount?' And the city agreed to a 15 percent rate reduction because the brewery's lower usage meant less strain on the municipal treatment plant during peak hours. Luna: That's brilliant. So the data didn't just save them water — it gave them leverage in a negotiation they wouldn't have had otherwise. Lucas: Exactly. The sensor produced operational savings, and the operational data produced strategic savings. That's a compounding effect that most companies leave on the table because they don't bother to collect the data in the first place. Luna: So the single sensor — it's really a gateway to a whole culture of measurement and continuous improvement. That's the Toyota Production System applied to water. Lucas: It is. And it's worth asking: where else in a business is waste hiding at a frequency you're not measuring? It might not be water. It might be idle time on a machine, or rework cycles, or the time between a customer inquiry and a response. The principle is the same. Luna: Right. Find the right resolution, look at the data, find the spike, ask why. Lucas: That's the operator's manual, right there.