Grundfos Pump Drift: What It Is and Why Your System's 'Efficiency' Might Be a Lie
Your pump's 'peak efficiency' is a marketing number, not a guarantee. The real metric is drift.
I'll say it straight: if you're buying a Grundfos pump based solely on its published efficiency curve, you're almost certainly going to be disappointed. Not because the pump is bad — it's not — but because the gap between lab-tested performance and real-world operation is a chasm, and that gap is called drift.
I've spent the last six years on the front lines of pump emergencies. In my role coordinating emergency replacements for industrial and municipal clients, I've seen more 'high-efficiency' installations underperform than I care to count. The culprit? Almost always drift — the slow, invisible degradation of pump performance that no curve chart accounts for.
What is drift?
Drift is the difference between the pump's theoretical performance (the shiny curve on the datasheet) and its actual performance at a given point in its life. It's not a bug. It's a feature of reality. Pumps wear. Impellers get eroded or fouled. Wear rings open up. VFDs (Variable Frequency Drives) lose calibration. The best pump in the world — and Grundfos makes some of the best — cannot escape physics.
For instance, a Grundfos CR vertical multistage pump handling slightly abrasive water might lose 5-8% of its head within the first 18 months. The curve says it should deliver 60 meters at 10 m³/h. In reality, you're getting 55 meters. The pump is still running. It's not 'broken' in the sense of a fault code. But it's drifting. And if you sized the system based on that 60-meter curve, you're now underperforming.
Why standard efficiency ratings mislead
Industry standards like the EU's MEI (Minimum Efficiency Index) or the US's DOE regulations are useful baselines, but they test pumps under ideal conditions: clean water, new components, stable flow. Your water might have silt. Your duty cycle might be pulsed. Your pipework might have a few extra bends. Suddenly, that 'MEI ≥ 0.7' pump is operating at 65% of its rated efficiency.
According to the Hydraulic Institute, real-world pump efficiency is often 10-25% lower than published curves due to installation effects, fluid properties, and wear (Source: HI Pump Standards & Guidelines, 2023).
That's not a criticism of Grundfos. It's a criticism of the entire industry's habit of selling curves instead of systems. The published curve is the promise. Drift is the reality check.
Three types of drift I see most often on site
1. Hydraulic drift
This is the classic: wear. Impeller gaps grow, surface roughness increases. It's slow, cumulative, and often misdiagnosed as a control problem. I've seen a Grundfos SP submersible pump that was 'running fine' for years, but its flow had dropped 15% because the rubber bearings had worn. The owner only noticed when his irrigation schedule stopped working.
2. Control drift
This one's sneaky. Grundfos pumps with CUE drives or Magna3 circulators use sophisticated PID loops. Over time, sensor offsets accumulate. A pressure transmitter drifts by 0.2 bar. The pump compensates, but now it's running at a different setpoint than you think. You'd be surprised how often I arrive at a call where the display shows 'System pressure: 4.0 bar' and a physical gauge shows 3.6 bar. That's drift, and it costs energy.
3. Application drift
The system changes, but the pump doesn't. A factory installs a new heat exchanger with a different pressure drop. A building adds a floor. A municipality extends a water main. The original Grundfos selection was perfect — five years ago. The drift here isn't in the pump; it's in the mismatch between the pump's curve and the new system curve. And the pump can't tell you. It just runs, inefficiently.
How to catch drift before it becomes an emergency
In my years handling rush repairs, I've learned one thing: the emergency call is always preceded by months of undetected drift. The pump didn't suddenly fail. It slowly degraded until it couldn't meet demand, then someone panicked.
Here's a cheap, practical check you can do on any Grundfos pump with a pressure gauge and a data sheet:
- Isolate the pump from the system (close the discharge valve, recirculate if possible).
- Run it at full speed.
- Measure the shut-off head. Compare to the curve.
If it's more than 5% below the published shut-off head, you have drift. (Note to self: I should check this on our own test rig more often — it's too easy to ignore routine checks.)
For VFD-driven pumps like those with CUE controllers, log the frequency required to maintain a fixed setpoint over time. If the frequency keeps climbing (ugh, that's a bad sign), your system resistance is rising or your pump's performance is dropping. Either way, it's drift.
The boundary condition: not all drift matters
I don't want to scare you. Some drift is harmless. If your pump was oversized with a 20% margin (common in conservative engineering), a 5% drift might not affect anything. The pump still meets demand. The motor isn't overloading. The process works. Don't chase perfection if the system is stable.
But if you're running a pump at 95% of its curve, and it drifts 5%, you're now at the edge. That's when the phone rings at 2 a.m. (trust me on this one). Or, if you have automated monitoring through Grundfos' Grundfos GO portal, you might catch it before 2 a.m. — that's a luxury I wish I'd had more of in my early years. Give me a pump with a data logger over a manual gauge any day. The difference is way bigger than I expected.
Bottom line: Drift is real. It's not a defect. It's a metric you need to track. Ignore it, and you'll be paying for a rush replacement — and maybe a $10,000 penalty clause — instead of a planned overhaul. I still kick myself for not tracking drift data on a municipal booster set in 2022. If I'd had the logs, we'd have caught the bearing wear in February instead of April. That was a lesson learned the hard way.
So check the numbers. Seriously. Your pump's curve is a starting point, not a promise.