I Rejected 300 Heat Pump Units. Here’s What “Reliable” Actually Means.

In February 2024, I approved the rejection of 300 heat pump units.

They weren’t broken. In the crate, they ran fine. But when our lab tested their heating output at -10°C—roughly the nighttime low that matters most for cold-climate homes—they delivered 14% less than the vendor’s spec sheet promised.

I’m a quality compliance manager at Growatt, a renewable energy company. My job is to review every product before it reaches customers: roughly 200+ unique items a year across our inverter, battery, and home energy lines. In 2023, I sent back about 7% of first vendor deliveries for specification issues. The heat pump rejection was the largest single batch we’ve ever returned.

This is the story of that rejection—and the lesson it taught me about what “reliable” actually means when you sell home energy equipment.

Why We Started Evaluating Heat Pump Units

Growatt built its reputation on solar inverters and battery storage. But our installers and channel partners kept telling us the same thing: customers don’t want a pile of components anymore. They want a complete home energy system.

On the heating side, a hybrid air source heat pump unit was the obvious starting point. The concept is straightforward: rooftop solar runs the heat pump during the day, a home battery carries it through the night. Homes that used to rely on gas or electric resistance heating get the same comfort with a fraction of the carbon footprint.

In late 2023, we shortlisted three manufacturers of hybrid air source heat pump systems. We picked a vendor with over a decade of experience selling heat pump units in Europe—a mature market with strict performance requirements. Their spec sheets were thorough, their efficiency ratings were competitive, and their installer references checked out.

I assumed their quality standards aligned with ours.

I didn’t verify that assumption. That was my mistake.

The Lab Test Found a 14% Gap

Before we approve any new product, we run independent tests. Not using the vendor’s data sheet. Not at comfortable test conditions. We test at the extremes our customers in North America and Northern Europe actually deal with. For heat pumps, that means down to -15°C.

In March 2024, the pre-production samples arrived. Our lab team ran all three units through a full temperature sweep, from -15°C to +35°C.

The results came back flagged. Across all three units, heating capacity at -10°C averaged 8.2 kW. The vendor’s published specification was 9.5 kW. That’s a 14% gap at exactly the temperature that determines whether a home stays comfortable during the coldest week of the year.

I called the vendor’s engineering manager in Europe. His response was calm, almost rehearsed: “That’s within industry standard tolerance.”

Let me translate that. They were comfortable shipping a heat pump unit that underperformed its own published numbers by 14%, as long as other manufacturers did the same.

I asked whether the numbers were physically measured in a lab or calculated from a performance model. There was a long pause. The low-temperature figures were, in part, calculated—never validated at the conditions we’d tested.

Nobody on their side had flagged this before we found it. That bothered me more than the gap itself.

The Same Story with Home Energy Wind Turbines

While the heat pump issue was unfolding, we were also evaluating three home energy wind turbines for customers who wanted to pair wind with solar. The goal was simple: have real data before recommending anything.

Our test stand results were worse than the heat pumps. One roof mounted wind turbine model claimed 700 kWh per year at a 5 m/s average wind speed. We measured 480 kWh—31% below the claim. Another small wind generator for home use was quieter but generated barely a third of its rated capacity at typical residential wind speeds. (The efficient one was loud. The quiet one underdelivered. Welcome to small wind.)

This is exactly why the Small Wind Certification Council (SWCC) was created (Source: SWCC, smallwindcertification.org). It runs standardized tests on small wind turbines and publishes energy performance labels that buyers can compare directly.

Two different product categories. One identical pattern: advertised performance that didn’t survive contact with real-world testing.

300 Units in a Warehouse: The Decision

By April 2024, our first production batch of 300 heat pump units sat in the vendor’s factory, completed to the original specification. The launch was scheduled for May. Installer training was done. Marketing materials were at the printer.

The upside of accepting the batch: we made our launch date, avoided rework costs, and got revenue flowing. The risk: a family in a cold-climate home would receive less heating output than they paid for on the coldest nights of winter. They wouldn’t know the spec sheet was wrong. They’d just know the system felt underpowered.

I kept asking myself whether a May launch date was worth the goodwill of hundreds of families in their first winter with our product. The answer didn’t change no matter how many times I asked.

We rejected the entire batch.

The vendor pushed back. They pointed at industry norms. They argued that we’d never find a manufacturer who met that standard. Then our procurement team reminded them of the clause we’d put in the contract after a different vendor issue in 2022: any deviation greater than 5% between published specs and independent test results triggers re-manufacturing at the vendor’s cost.

They rebuilt the units and contracted third-party labs to verify them. Our launch moved from May to August. The summer selling season was gone.

Was the Rejection Worth It? The Numbers Say Yes

By January 2025, more than 1,000 corrected units were installed in homes across Canada, the northern US, and Scandinavia. I’ve tracked three metrics since then: heating output complaints, installer reorder rate, and post-winter customer satisfaction.

Heating complaints: 0.3%. Installer reorders: above forecast. Customer satisfaction after the first winter: 4.6 out of 5, from 340+ homeowner responses.

I can’t run a parallel test to show what the original batch would have done. But I can estimate. Based on our lab data, roughly 14% of installations would have underperformed at peak winter demand. If even a quarter of those customers wrote a negative review or told another installer about their experience, the damage to our brand would have far exceeded the cost of the delay.

And that’s the thing about quality: the return on investment doesn’t show up in a ledger. You can’t count the bad reviews that never happened.

What “Reliable Heat Pump” Means to Me Now

If you’re an installer, distributor, or homeowner evaluating heat pumps, wind turbines, or any home energy product, here’s what I now check before putting our name on anything:

• Ask for test data at the conditions customers actually experience. If a manufacturer only provides efficiency at 7°C—the temperature used in most marketing comparisons—ask where the performance drops off. A heat pump unit that can’t hold up at -15°C isn’t reliable. It’s fair-weather equipment.
• Understand whether numbers were measured or calculated. Calculated performance isn’t automatically wrong. But if you base a purchasing decision on it and physical testing doesn’t match, you’ve built your product launch on someone else’s assumptions.
• Put a concrete tolerance clause in your supply contract. Ours is 5% maximum deviation between published specs and independent lab results. It’s the reason this story has a happy ending instead of a costly recall.
• Check for certification. For small wind generators for home use, SWCC certification provides verified performance. For heat pump units, AHRI’s certification directory tells you whether the model’s ratings have actually been certified.
• Treat quality as a brand decision. A product that meets its spec builds trust quietly. A product that doesn’t destroys it loudly.

A reliable heat pump isn’t the one with the best COP on paper. It’s the one that meets its published output when it’s -15°C outside and a family is depending on it.

Quality isn’t something customers see. It’s something they don’t complain about—and that silence is the strongest brand signal there is.