Nobody Has Measured What an HVAC Fault Costs You in Years

There is a sentence on nearly every HVAC monitoring product page, and it was once on ours: catch problems early and your equipment lasts longer. It is intuitive, it is almost certainly true in direction, and when we went looking for the research that puts a number on it, there was none to find.

We keep a corpus of several hundred documents for exactly this purpose — DOE and national-laboratory reports, NIST technical notes, peer-reviewed HVAC&R papers. We ran five separate extraction passes over it looking for one specific thing: a published figure connecting a named fault to a quantity of equipment life lost. We found direction for nearly every fault. We found magnitude for none of them. Every source below is linked so you can check that yourself.

The Department of Energy says the same thing

This is not a gap we discovered. DOE's Building Technologies Office named it in two separate reports — one on residential HVAC and one on commercial HVAC — listing among the outstanding research needs:

understanding which system problems lead to decreased equipment lifetimes

U.S. DOE Building Technologies Office, Energy Savings Potential and RD&D Opportunities (2011, 2012)

More than a decade on, no published work has closed it. On compressor cycling specifically — the fault most often blamed for shortened life — Huang et al. state plainly that no direct quantitative relationship between the number of starts per hour and compressor lifetime is specified in the literature.

The one measured number, and why it is not enough

Across the whole corpus there is exactly one passage connecting a specific fault to a specific time-to-failure. It comes from an Oak Ridge National Laboratory report on a prototype heat pump running with liquid returning to the compressor:

This liquid floodback caused the compressor to fail after about one year of operation.

Baxter, Oak Ridge National Laboratory, ORNL/CON-50

One unit. One prototype. That is an anecdote, not a rate, and it would be dishonest to build a customer-facing number on it. We mention it because it is genuinely the strongest thing anybody has, and knowing that is useful in itself.

What the research does support

Direction, repeatedly and clearly. NREL, on undercharge: “Undercharging a system reduces its cooling capacity, which results in low evaporating pressures, evaporator coil icing, and reduced system lifetime.” DOE's 2018 review of installation practices reaches the same place by way of Kim & Braun: undercharge or overcharge can reduce air conditioner life, capacity and efficiency.

And on heat specifically, Hu, Yuill et al. (2021) make a point worth sitting with:

The increased operating temperature will reduce the lifespan of the system, potentially affecting operating cost more significantly than the energy penalties.

Hu, Yuill et al., Energy and Buildings vol. 242

Read that carefully. The authors expect the thermal damage from running out of spec to matter more than the extra electricity — and nobody has sized it. The energy penalty is the part that is easy to measure, so it is the part that gets measured and quoted. The larger effect goes unquantified.

Why the number would be wrong even if we invented one

There is a structural reason this figure cannot exist in the form people want it. How much life a fault costs depends on how long the fault is left running — and that is precisely the variable monitoring changes.

A refrigerant leak caught in its second week and the same leak caught in its third year are not the same event. Any single percentage would have to assume some fixed duration of neglect, at which point the number describes the assumption rather than the equipment. This is why “extends equipment life by 40%” is not merely unsupported. It is not a well-formed claim.

So what can honestly be said

Three things, and we hold ourselves to them:

• The mechanism. Liquid reaching a compressor washes oil off the bearings. Suction gas returning too hot stops cooling the motor windings. These are described in the literature and we can name them.
• The running cost. NIST Technical Note 1848 measured a 30% undercharge at roughly 12% lower COP and 15% lower capacity. That is a real number from a real test stand, and it prices the electricity — not the lifespan.
• The rate, while the fault lasts. A fault that removes capacity makes the unit run more hours for the same work, and field data does show run hours predicting service life. That supports saying a system is wearing faster right now. It does not support converting that into years.

What we will not do is give you a number of years, or a percentage of lifespan. When you see one, it is worth asking which study it came from. We looked for that study for a long time.