I Almost Lost a $14K Water Treatment Order Because of a Single Spec Mistake (and Why KSB Pump Curves Saved Me)

It was a Tuesday afternoon in March 2023. I was sitting in our tiny project office—more like a converted storage closet—staring at a spreadsheet of pump specs for a small gold-mine dewatering job. The client was a junior exploration outfit, the kind of company where the CEO answers his own phone. Their order was modest: two submersible pumps and a control panel. Total value, maybe $18,000.

I'd been in this industry for about four years. Long enough to be dangerous, I suppose. I knew enough to pull up the KSB pump curves for the Amarex KRT series. I had the KSB catalogue open on my laptop, the KSB pump data sheet for the KSB-260 model on my tablet. I felt prepared. I was not.

I made a mistake that cost us $14,000 and a three-week delay. But the real lesson wasn't about pump hydraulics—it was about who I was willing to take seriously.

The Setup: A Small Client, A Routine Spec

The client's site was at 4,200 meters above sea level in the Peruvian Andes. That's important—air is thin up there. Pump performance changes dramatically with altitude. I knew this academically. But I'd never actually processed an order for that elevation before.

I specified a KSB Amarex KRT KSB-260 with a 15 kW motor. The pump curve in the KSB catalogue showed it could handle 80 meters of head at the required flow rate. I checked it twice. I was confident.

That confidence is exactly what got me into trouble.

Here's the thing: I was treating this like a standard quote. I'd done dozens of them for large mining companies. But those were for orders worth $200,000+, where the client's engineering team had their own pump specialists and would catch my errors. This small client? They trusted me completely. They didn't have a pump engineer on staff. They had a geologist and a dream.

The Mistake: Reading the Curve Wrong

The KSB-260 pump curve I was looking at was for sea-level conditions. Every pump manufacturer publishes their curves based on ISO 9906 testing standards, which assume an air density of 1.2 kg/m³—essentially sea level. At 4,200 meters, air density is about 60% of that.

The result? The maximum head the pump could deliver dropped from 80 meters to about 48 meters. Not enough for the vertical distance from the mine sump to the surface settling pond.

I said 'Let's go with the standard KSB-260 package.' They heard 'This is the right pump for our application.' What they didn't hear—because I didn't say it—was 'This curve is for sea level, and I haven't applied the altitude correction factor.'

I caught the error myself, thank goodness. But only after the pumps had been manufactured and shipped to the port in Callao. The total at that point: $14,000 in equipment and shipping. Nowhere near salvageable—the pump was already built and sized for a motor that couldn't deliver the required performance at altitude.

The Salvage: How KSB's Support Saved Me (Mostly)

The conventional wisdom is that big suppliers don't care about small orders. Everything I'd read said that if you're a small customer, you're a nuisance. You get standard responses and slow service. In practice, I found the opposite.

I called the KSB regional representative in Lima. I expected a brush-off. Instead, the application engineer—a guy named Ricardo who'd been with them for 17 years—spent an hour with me on a video call, walking through the KSB pump curve re-rating for altitude. He showed me the correction factors in the KSB-260 technical manual that I'd overlooked.

The fix wasn't elegant: replace the 15 kW motor with an 18.5 kW unit, add a VFD to manage the increased starting torque, and modify the impeller trim. Cost to my company: about $4,000 in replacement parts and freight for the new motor. But KSB's factory support team expedited the motor through their system, and Ricardo even helped me prepare a justification document for my client explaining the change.

The original $18,000 order turned into a $22,000 order (and a very educational project). The client? They were understanding, mostly because I was honest about the mistake and had a solution within 48 hours.

That experience taught me something about how to read a pump curve properly—but more importantly, it taught me something about how to treat a small order.

The Lesson: Small Order, Big Standards

When I was starting out in this industry, the vendors who treated my $2,000 trial orders seriously are the ones I still use for $200,000 projects. KSB didn't treat us like a small client. They treated us like a client, period.

Small doesn't mean unimportant—it means potential. That junior mining company? They hit a significant vein six months later. Their next order was for six pumps and a full water treatment package. Total value: $140,000. We got the order because we'd been honest about our mistake on the first one.

Here's what I do now, based on that experience:

1. Never assume a standard pump curve applies. If the site is above 1,000 meters, I recalculate. It takes 10 minutes. It saves tens of thousands in rework.
2. Call the manufacturer's application team before you submit a spec. Every major pump manufacturer—KSB, Grundfos, whoever—has engineers whose job is to keep you from making stupid mistakes like mine. Use them.
3. Treat every order like it's the first order from a future major account. The $14,000 mistake I caught early? That's the kind of error that could have killed a small client's project budget. A small company can't absorb a 3-month delay. A large one can.

Practical Tools: Reading Pump Curves Properly

If you're a small operations engineer or buyer (as I was), here's what I wish someone had told me about KSB pump data sheets:

• The KSB-260 is a specific model in the Amarex KRT range. The curve published in the KSB pump catalogue is for fresh water at 20°C at sea level. If your fluid is anything else—thick, hot, or at elevation—the curve changes.
• Check the NPSHr (Net Positive Suction Head required). On my botched order, the NPSHr was fine. But on other projects, I've seen people miss this and end up with cavitation within weeks.
• Every KSB pump curve has a recommended operating range. Running at the edge of that range—as I was trying to do—is risky. The pump works, but it might not last its intended service life.

As of February 2025, I've caught 17 potential specification errors using the checklist I created after that Peru project. Total savings: roughly $85,000 in avoided rework. Not bad for a lesson that cost me $14,000 and a weekend of panicked phone calls.

The bottom line? Don't let a small order make you lazy with the details. And don't assume a big-name manufacturer won't take your call when you mess up. In my experience, the companies that are serious about engineering don't discriminate by order size.

---

About the author: I've been sourcing industrial pumps for mining and water treatment projects since 2018. My education was in mechanical engineering, but my real education came from a $14,000 re-rating mistake on a KSB-260 order in 2023, which I still refer to as 'the Peru problem' in our team's project pre-checks.