Service guide
Heat pump repair: what fails, and what only looks like a fault
A heat pump runs both ways and about twice the hours of an air conditioner. Three parts carry that difference; one is usually fine.
Get HVAC Options Repair cost estimator
Written by HYRE HVAC Research Desk Primary-source research, data analysis and fact checking
The short version
Everything that fails on an air conditioner can fail on a heat pump. Three parts are different: the reversing valve, the defrost control and the backup electric heat.
They cause most winter complaints. The most common one, cold air for a few minutes while the outdoor unit steams, is a normal defrost cycle, not a fault.
Three things that are not a repair question
Everything below assumes the system is off, or running badly, and nobody is in danger. These three are not diagnostics. Deal with them first and read the rest afterwards.
A heat pump burns nothing, but a great many heat pumps are installed as dual-fuel systems sharing a gas furnace, and the furnace is the half that can produce combustion products.
Leave the building, then call the gas utility’s emergency line or 911 from outside. Do not operate switches on the way out. This is not an HVAC scheduling decision.
A breaker that trips repeatedly is doing its job, and resetting it repeatedly removes the protection without removing the fault. Leave it off and have the system looked at.
Do not open the outdoor disconnect, the air handler cabinet or the panel to investigate.
Heat pump auxiliary heat is a bank of resistance elements drawing tens of amps at 240 volts, and the capacitors inside the equipment hold a charge after the power is off.
Switch the system to emergency or auxiliary heat so the house stays warm, and get it looked at rather than running it.
Do not chip, hammer or pour hot water on the coil — the fins and the refrigerant tubing are soft aluminum and copper, and a punctured coil converts a control fault into a full refrigerant repair.
Running on emergency heat for a few days is expensive; it is much cheaper than the coil.
Change or check the filter. Look at the outdoor unit from a standing position and note whether it is iced, buried, blocked by leaves or sitting under a dripping roof edge or gutter.
Read the thermostat and note whether it is calling for auxiliary or emergency heat. Check that nobody has switched the disconnect off. That is the whole list, and those four observations are worth more to the technician than any part guess.
What should a heat pump diagnosis show?
HyreHVAC analysis: before paying for a repair, ask what measurement shows the part is bad, not just which part is bad.
Watch for two failure modes on this equipment. A diagnosis that names a symptom rather than a cause. Or a proposal to add refrigerant with no leak search. A heat pump can hide a slow leak for two seasons before anybody notices.
What is genuinely different about repairing a heat pump
A heat pump and an air conditioner share a compressor, a condenser coil, an evaporator coil, a metering device, a blower and the controls around them.
Everything that fails on one fails the same way on the other, and a technician who is good at air conditioners is most of the way to being good at heat pumps.
The differences are three components and one duty cycle. A reversing valve — a solenoid-operated valve that redirects hot discharge gas to the indoor coil instead of the outdoor one — is what makes the machine a heat pump at all.
A defrost control watches the outdoor coil and periodically runs the machine in cooling for a few minutes to melt the frost off it.
Auxiliary electric resistance heat covers the house during that, and covers the load when it is too cold outside for the heat pump to keep up on its own.
The duty cycle matters as much as the parts. An air conditioner in a mixed climate works for one season.
A heat pump works in both, so the compressor, the contactor, the fan motor and the capacitor all accumulate roughly twice the annual running hours. HyreHVAC analysis.
That does not make heat pumps unreliable; it means age in years is a poorer proxy for wear than it is on an air conditioner, and a maintenance interval borrowed from a cooling-only system is half of what the machine is actually doing.
It also changes the economics of ignoring a fault. A slow refrigerant leak on an air-conditioning system announces itself in July.
On a heat pump it shows up as a heating bill, blamed on the weather, and the system quietly leans on the resistance strips to make up the difference — which is why the first symptom of a genuine heat pump fault is frequently a bill rather than a noise.
The most common “fault” is the machine working correctly
Steam rising off the outdoor unit, the outdoor fan stopping, a whooshing clunk from the valve, cold air from the registers for a few minutes, then normal service. That is a defrost cycle. It happens on every air-source heat pump ever built, and it is the single most reported non-fault in the trade.
Source fact: DOE’s Building America program monitored inverter-driven heat pumps in cold-climate homes over 667 site-days and described the trigger plainly: a defrost cycle is initiated by the presence of a sufficiently cold coil temperature for a specific length of time.
Defrost terminates once the coil warms to a certain temperature (which occurs very quickly if no frost is present on the coil). The number and duration of defrost cycles, therefore, is generally a function of outdoor air temperature, humidity, and heat pump output.
So the frequency of defrost is set by the weather and by how hard the machine is working — not by a timer you can turn down, and not by a fault.
On a damp morning near freezing, which is the worst case for frost, a heat pump may defrost every 30 to 90 minutes and be behaving exactly as designed.
On a dry day at 10°F it may barely defrost at all, because cold air holds very little moisture to deposit.
The question worth asking is what defrost costs, and this is where the field data is more useful than the argument. CARB computed each site’s seasonal coefficient of performance twice: once as measured, and once with every defrost period stripped out of the data.
HyreHVAC calculation: the difference between those two figures is the share of seasonal efficiency that defrost accounted for, and across the seven sites it runs from 0.9% at Site 9 to 5.3% at Site 4, averaging 2.7%.
CARB’s own conclusion was the same one: “While defrost should certainly be minimized for best performance, CARB did not find that excessive defrosting dramatically impacted overall efficiency of any heat pump monitored.”
HyreHVAC analysis: This settles a common upsell. A defrost cycle you can see is not evidence of a problem and is not, on its own, a reason to spend money.
What is worth investigating is defrost that is far too frequent — and the field study found that the usual cause of that is not a component at all.
When defrost really is excessive, look at the installation before the parts
Source fact: The same study recorded a mechanism most repair pages never mention: “If a unit is installed where water or melting snow or ice can drip onto it, ice is very likely to form on the evaporator coil.”
CARB found one unit installed beneath a deck, where snow melting on the deck dripped onto the coil and refroze, and another case where two heat pumps were stacked vertically so that the upper unit’s defrost water fell onto the lower one.
Of that second case: “When testing performance of the lower unit, CARB found the system entered defrost mode every 10 minutes.”
HyreHVAC analysis: A unit defrosting every ten minutes is a dramatic symptom with a boring cause. Nothing in it had failed. The machine was correctly detecting ice that should never have been landing on it, and every component test would have come back clean.
This is the strongest argument on the page for a technician who walks around the outdoor unit before opening it.
Gutters and roof valleys draining onto the coil, a unit set so low it sits in slush, a stacked installation, snow drifted against the fins, and shrubs grown tight around the casing are all cheap to fix. All of them produce symptoms that look like an expensive control fault.
The honest caveat, and CARB states it: “While details of defrost control algorithms are not public”.
Manufacturers do not publish their defrost algorithms, so nobody outside the factory can tell you the exact interval your model should use.
What can be established is whether the coil is icing for a reason that has nothing to do with the coil.
Symptom, candidate causes, and the measurement that separates them
The right-hand column is the useful one. Almost every dispute about a heat pump repair is a dispute about whether a measurement was taken, and you do not need to interpret the reading to ask whether it exists.
| What you observe | What it is usually | What else it can be | The reading that decides |
|---|---|---|---|
| Cold air for a few minutes in winter, steam off the outdoor unit | A normal defrost cycle | Excessive defrost from water dripping onto the coil, or a defrost sensor reading the coil wrongly | How often it happens, against outdoor temperature and humidity. A walk around the unit before any part is opened |
| Heating works, cooling does not (or the reverse) | The reversing valve is not shifting — commonly the solenoid coil, which is the cheap half | A stuck valve body, or a control signal that never reached it | Whether the solenoid is energised and whether the valve actually shifted — a temperature split across the valve body, not a guess from the symptom |
| Auxiliary or emergency heat running constantly in mild weather | A control or thermostat configuration fault, or a heat pump that has lost capacity | Refrigerant undercharge, a failed outdoor unit that the controls are quietly covering for, or a thermostat installed with no heat pump staging | Whether the compressor is running at all when aux heat is on, plus superheat or subcooling. Aux heat masking a dead heat pump is invisible on the thermostat |
| Runs constantly, house never quite warm | Undersizing at the design temperature, or the system operating below its balance point | Low charge, low airflow, or a duct system that cannot deliver what the equipment makes | Static pressure and airflow first, then charge. Low airflow was reported in more than 50% of systems DOE reviewed |
| Outdoor fan spinning, no cooling or heating | A failed run capacitor or contactor — routine, and cheap at any age | A compressor that is not starting, which is a different conversation entirely | Measured capacitance against the rating printed on the part. Ask for both numbers |
| Ice on the indoor coil, weak airflow | Airflow starvation — a blocked filter, a crushed return, a coil that needs cleaning | Low refrigerant charge from a leak | Static pressure across the air handler, then superheat and subcooling. Adding refrigerant to an airflow problem makes it worse |
| Bills up sharply, nothing obviously wrong | A slow refrigerant leak, with the resistance strips silently covering the shortfall | A stuck aux-heat relay, a thermostat setting change, or simply a colder season | Runtime of the auxiliary heat against the compressor, plus a charge check with a leak search. This is the fault a heat pump hides best |
HyreHVAC compilation. The failure modes are standard vapor-compression and heat pump control behavior; the prevalence figures cited in the last column are DOE’s.
HyreHVAC analysis: Notice how often the deciding measurement is airflow rather than refrigerant. That ordering is not a stylistic preference — it is what the field literature reports, and it is the ordering most commonly reversed in practice.
Why “it is low on refrigerant” is a diagnosis of a leak
A sealed refrigeration circuit does not consume refrigerant. If it is low, it leaked, and the refrigerant went somewhere. Topping it up without finding where is not a repair; it is a scheduled payment.
Source fact: DOE’s review of 44 field studies reported undercharge in 29-78% of the systems tested and overcharge in 4-50%. Both figures are ranges across studies and we are not going to average them into one number — the spread is the finding, and it says the population of installed systems is charged very inconsistently.
HyreHVAC analysis: On a heat pump that spread matters more than on an air conditioner, for two reasons. The first is exposure: twice the running hours means twice the opportunity for a vibration leak at a flare or a brazed joint.
The second is concealment — an undercharged air conditioner stops cooling and you notice, while an undercharged heat pump keeps the house warm by running the resistance strips and you notice in February when the bill arrives.
On an R-410A system there is a further consideration, and it is a price story rather than a deadline story.
EPA’s Technology Transitions program restricts new equipment to refrigerants below a global warming potential of 700, with a manufacture and import cutoff of January 1, 2025.
It does not require anyone to replace working equipment and it does not make it unlawful to service what you own — EPA’s current sector table states there is no installation compliance date for residential and light commercial systems where all specified components were manufactured or imported before January 1, 2025.
What is true is that the production phasedown tightens R-410A supply over time, so refrigerant to recharge an older system is a rising cost, and that belongs in the arithmetic of a large leak repair as a price rather than as a countdown.
The useful next step: if a quote proposes refrigerant, ask two questions. What was the measured superheat or subcooling before the charge was adjusted, and what leak search was performed. A technician who did the work will answer both without pausing.
The repair cost estimator covers who is legally permitted to open the circuit under EPA Section 608, and our study of the refrigerant transition covers the rule itself, including the widely reported installation deadline that is no longer in force as reported.
Reading the scope of work
A repair quote is a short document and most of its content is in what it does not say. These are the distinctions that separate a scope you can hold someone to from one you cannot.
“Not heating” is the complaint you telephoned in. “Reversing valve solenoid coil open, measured infinite resistance, valve did not shift on a call for cooling” is a diagnosis. The second one can be checked afterwards; the first cannot.
Capacitance against the rated microfarads. Superheat or subcooling where charge is claimed. Static pressure where airflow is claimed. These are by-products of doing the work, not extras, and a technician who took them will read them out on request.
A location, or an explicit statement that the leak search was inconclusive and what happens next. “Added 2 lbs” with no leak line is the single most common padded item on a heat pump invoice.
Perfectly legitimate either way. What is not legitimate is finding out afterwards. Ask before the visit, not after.
A short list of what was ruled out is the cheapest evidence that a diagnosis happened. Its absence does not prove anything, but combined with a missing cause it is a pattern.
There is no rule requiring you to replace working equipment and no date by which you must act. If urgency is being generated by a compliance deadline, the stated reason is wrong, which is a fact about the quote worth noticing on its own.
A compressor is the most expensive component in the system and it usually fails because of something else — chronic low charge, chronic low airflow, or a failed capacitor left in service.
Replacing it without establishing what killed it buys the same failure again. Check the manufacturer’s parts warranty on your model and serial number before agreeing to anything.
Replacement may well be right. It is not something that can be established before the fault is. A quote for new equipment arriving in place of a diagnosis is a sales visit that was booked as a service call.
What drives the price, and why there are no part prices on this page
HyreHVAC does not install, service or sell HVAC equipment, and holds no data on quotes, invoices or finished jobs. Nothing here is a price we observed. That is also why these tools can tell you to keep the system you have.
What we can set out is what moves a heat pump repair up or down the range, because those drivers are structural rather than regional.
The part class: a capacitor, contactor or defrost sensor is an inexpensive component and a short visit; a reversing valve or a coil means recovering the refrigerant, brazing, evacuating and recharging, which is most of a day and a technician who is certified to handle refrigerant.
The refrigerant quantity: any repair that opens the sealed circuit means a full charge afterwards, and on an R-410A system that quantity is a rising cost.
Access: an air handler in a conditioned closet and one in a 3-foot crawlspace are the same repair and not the same job.
Timing: the first hard freeze and the first heatwave are when after-hours rates apply and when parts availability is worst.
HyreHVAC analysis: The number that actually decides the repair-or-replace question is not the repair price in dollars; it is the repair price as a fraction of what replacement would cost, set against how much service life the equipment has left.
That is arithmetic you can only do with your own quote in front of you, which is why it lives in a calculator rather than in a paragraph here.
The useful next step: put your diagnostic and repair quotes into the repair cost estimator, which expresses them as a share of replacement and does not invent a part price.
Check remaining life with the lifespan estimator, which is built on the one sourced horizon we could find and reports a range rather than a year; and if the answer is close, run repair or replace.
When repair stops being the right answer
There is no threshold that settles this, and anyone who gives you one has substituted a rule of thumb for your circumstances. There are, however, four situations where the arithmetic usually tips.
A compressor failure on an out-of-warranty system: The most expensive single component, in a machine whose remaining life is the question. If the parts warranty has expired, this is the point at which the replacement conversation is legitimate rather than opportunistic.
A leaking evaporator or condenser coil on an R-410A system: The repair is major, the recharge is substantial, and the refrigerant is on a long-term price rise. It can still be the right call on a young system; on an old one it rarely is.
The third significant repair in two years: Not because of a rule, but because a system producing recurrent unrelated faults is telling you something about its condition that no single diagnosis captures.
The house was never comfortable: If the complaint predates the fault — one room that has never been warm, a system that has always run constantly — then repairing a component fixes a different problem from the one you have.
That is a design and duct question, and no repair addresses it. Ductwork and its replacement is where that conversation belongs.
HyreHVAC analysis: Note what is not on that list: age on its own. A twelve-year-old heat pump with a failed capacitor should have the capacitor replaced. Age becomes decisive only when it is multiplied by the cost of the specific repair in front of you.
Get connected
Tell us about the work
Three short steps: the job, your home, and how to reach you. Nothing goes to an HVAC company unless you agree.
A straight answer first. Matching is still being built, so this is not an instant quote. If you smell gas or a CO alarm sounds, leave the house and call 911 or your gas utility first.
Get HVAC Options
Tell us about the system and where you are. We will follow up about connecting you with heating and cooling contractors serving your area.
Questions
My heat pump is blowing cold air in winter. Is it broken?
How do I know if the reversing valve has failed?
Why is my auxiliary or emergency heat running so much?
Does a heat pump need servicing more often than an air conditioner?
Can I add refrigerant to my heat pump myself?
Is my heat pump repair urgent because of the R-410A rules?
What should a heat pump repair quote contain?
Written and audited by
HYRE HVAC Research Desk
Primary-source research, data analysis and fact checking
We are a research desk, not a sales floor. We read the federal microdata file, the statute or the manufacturer data sheet ourselves, and we publish the figure with the document it came from and the date we retrieved it.
Where a number cannot be traced to a primary source, we publish the shorter page and say what we could not verify.
The counts below are generated from the published pages themselves, last counted September 28, 2026, and they are what we have actually published rather than what we intend to.
- 13
- studies published
- 12
- federal sources read and cited
- 8
- studies published with their full dataset as CSV
- 51
- jurisdictions reproduced against EIA’s own tables
How this desk works
- Primary sources only. Federal data comes from the agency that collects it, in the file that agency publishes. We do not cite an article that cites a source; we download the source and compute the figure ourselves.
- We validate against the agency before we publish. First, we use each federal microdata file to reproduce the agency’s own published tables. Our cooling research reproduces EIA’s state estimates and standard errors for all 51 jurisdictions. That check caught a variance formula that was off by a factor of four.
- Every estimate carries its uncertainty. These are survey figures, not counts. Standard errors are computed from the replicate weights the federal file supplies and printed beside the estimate. An estimate too imprecise to publish is reported as such rather than printed.
- Nothing is typed by hand. Prose, tables and charts all read from one dataset built by script, so a number in a sentence and the same number in the table below it cannot disagree.
- We publish the data, not just the conclusion. 8 of our 13 studies offer the full computed table as a CSV download on the page, so you can check the analysis or disagree with it. A study without a row-level dataset gets no download link and claims none in its structured data.
- We correct in public. Where we have published a figure wrongly we fix the figure, rewrite any analysis that rested on it rather than patching the number underneath it, and leave a dated correction note on the page.
- We do not install or sell HVAC equipment, and we take no payment for placement, ranking or a favorable mention. Nobody buys a position on this site.
Data as of DOE Building America field monitoring (2015); DOE BTO literature review (2018); EPA rules read September 4, 2026. Authorship on this site is organizational: the analysis belongs to the desk rather than to a named individual, and we do not publish credentials we do not hold.
Our editorial policy sets out how we source, date and correct what we publish.
Sources & retrieval dates
HyreHVAC does not install, service or sell HVAC equipment and does not perform electrical work. This page is general information about how heat pump faults are diagnosed and priced; it is not a diagnosis of your system and it is not a substitute for a qualified technician inspecting it.