Extreme Heat and What It Does to Equipment
Why this matters
On the third day of a heat wave the phones carry one sentence: "it's running but it can't keep up." Some of those calls are real failures. Many are equipment doing exactly what physics allows at an outdoor temperature above what the system was sized for, and a tech who cannot tell the two apart either adds refrigerant to a correctly charged system or sells a replacement nobody needed. The other half of the heat-wave workload is heat killing the parts that were already marginal: capacitors, contactors, boards in a 140 F attic, and motors running on sagging grid voltage.
This card owns the equipment side. Crew heat illness belongs to the FieldHealth cards, surge scheduling to Operations, and finding your area's design temperature to the Place group. See related: Design Events: What the Code Assumes Will Happen to the House.
The design day, and why "can't keep up" is often the answer
A residential cooling system is sized to a design outdoor temperature, usually the 1 percent cooling design dry-bulb for the location (the temperature exceeded about 1 percent of the hours in a typical year), taken from the climate tables a load calculation such as ACCA Manual J uses. Equipment is rated at a fixed test point: the AHRI 210/240 rating condition for residential air conditioners is 95 F outdoor.
Two consequences follow, and both are physics, not faults.
- Capacity falls as outdoor temperature rises. The condenser rejects heat to outdoor air across a temperature difference. Raise the outdoor air temperature and the refrigerant has to condense hotter to keep rejecting heat, the compressor works against a higher pressure ratio, and delivered cooling drops. The amount is in the manufacturer's expanded performance table, which lists capacity by outdoor temperature and indoor wet-bulb. Read it for the model in front of you rather than using a rule of thumb, because the slope differs by compressor type and coil size.
- Load rises at the same time. A house at 108 F outdoor gains more heat through walls, glass and attic than the same house at the 98 F design day.
So a correctly sized, correctly charged system can run continuously and lose ground on a day hotter than design. The tell is that it is running continuously with normal readings for that ambient, the indoor temperature drifts up in the afternoon and recovers after sundown. A failure looks different: abnormal readings, short cycling, or no recovery at night.
Head pressure: what is normal at 108 F
Head pressure (the high-side or discharge pressure) tracks condensing temperature, and condensing temperature tracks outdoor temperature plus the condenser's approach (condensing saturation temperature minus outdoor air temperature entering the coil). The approach is set by the coil and fan design and is published by many manufacturers; for a given unit it stays roughly constant across outdoor temperatures when the coil is clean and airflow is right.
That is the useful fact in a heat wave. If a unit's normal approach is, say, 20 F (illustrative), then at 108 F outdoor the condensing temperature should be near 128 F, and the pressure is whatever the refrigerant's pressure-temperature chart says for 128 F. A reading that converts to 145 F condensing at the same ambient means an approach of 37 F, and the extra 17 F is the fault: a dirty or blocked coil, a failing fan motor, recirculating discharge air, or non-condensables. A high pressure that matches the ambient is not a fault, and it is not a reason to recover charge.
Charge on a hot day. Subcooling on a fixed-metering or thermostatic expansion valve system is still a valid check at high ambient, but only with the system at steady state (15 minutes running) and with the indoor airflow verified first. Adjusting charge to bring the head pressure "down to normal" on a 108 F day undercharges the system for every cooler day after it.
What heat does to the parts
| Part | What heat does | What you see |
|---|---|---|
| Run capacitor | Life falls steeply with temperature; a weak capacitor fails under the hottest, highest-load hours | Reading below the rated tolerance band on the label (commonly plus or minus 5 or 6 percent), bulged top, compressor or fan humming on start |
| Contactor | Higher current and longer run time pit the contacts | Pitted or welded contacts, chatter at low voltage, scorched coil |
| Compressor overload | An internal or external overload opens on high winding temperature, which high ambient and low voltage both push | Compressor off, fan running, shell too hot to touch; it resets after cooling down, often an hour or more |
| Control board in attic | Board components are rated for an ambient range on the spec sheet; attics in heat waves often reach 130 to 150 F | Intermittent lockouts in the afternoon, normal in the morning |
| ECM blower motor | Electronics derate with temperature; motor modules in hot attics fail in clusters during heat waves | Blower stops and restarts, or runs at the wrong speed |
The low-voltage multiplier. During peak demand, utility voltage at the house can sag. A motor asked to deliver the same power on lower voltage draws more current, which heats windings, which trips overloads. Measure voltage at the contactor's line side with the system running, and again at compressor start. The tolerance is on the nameplate, typically plus or minus 10 percent of rated voltage for single-phase residential units.
Grid stress and demand response
Utilities manage peak load with demand response programs: a load control switch on the condenser, or an enrolled smart thermostat that the utility adjusts during an event. During an event the condenser may be cycled off for part of each half hour, or the setpoint raised a few degrees. The customer often forgot they enrolled.
Find out before you diagnose. A load control device usually sits near the condenser disconnect and carries the utility's marking; an enrolled thermostat shows an event message. A compressor that stops for a fixed period with the fan also off, on a hot afternoon, with a signal on the device, is the program working. Do not bypass a utility device; it belongs to the utility, and tampering with it is usually a breach of the program terms. Tell the customer how to opt out of an event through the program, if the program allows it.
Grid stress has a second form, rolling outages: an older unit without a restart delay can try to start its compressor against unequalized pressures when power returns. See related: How to Prepare a House for a Utility Power Shutoff.
What a tech checks first in a heat wave
Order matters because heat-wave days are short on time. This is the order that separates "physics" from "fault" fastest.
- Is it the program? Look at the thermostat and the load control device before touching anything. Two minutes, and it removes a whole class of calls.
- Is the condenser moving air through a clean coil? Fan running at full speed, discharge air leaving the top, nothing within the manufacturer's clearance, coil face clear of cottonwood, grass clippings or a hedge. Lock out the condenser disconnect before reaching into the fan or coil area; a fan that is stopped on a high-pressure or overload trip can restart.
- Voltage under load. At the contactor, running and at start.
- Capacitor. With power off and the capacitor discharged through a resistor, measure microfarads against the label.
- Indoor airflow. Filter, blower speed, evaporator coil. A frozen evaporator on a hot day is usually low airflow or low charge.
- Approach and subcooling, only now, with the system at steady state.
- Attic equipment. Check board fault codes and the condensate safety. Limit attic time in a heat wave; the attic is often the hottest workspace the tech enters all summer, and the FieldHealth rules on work-rest and water apply.
Do not hose a running condenser to "help it." Spraying cools it for minutes and, with hard water, leaves scale that insulates the fins for good. Washing a dirty coil with power locked out is maintenance.
Worked example: a 108 F afternoon
A 3-ton split system, 12 years old, on a house whose load calculation used a 98 F design day. Outdoor temperature at the condenser is 108 F, 10 F above design. Setpoint 74 F, indoor 82 F at 4 pm. The customer says it was fine yesterday morning.
Readings, all illustrative:
- No demand response event on the thermostat; no load control device on site.
- Condenser fan running. Coil face about half covered by cottonwood fluff on the side facing the yard.
- Line voltage at the contactor 228 V running against a 230 V nameplate, 214 V for an instant at compressor start. Nameplate range is plus or minus 10 percent, so 207 to 253 V. Both readings inside.
- Run capacitor 44 microfarads on a label of 45 plus or minus 6 percent (42.3 to 47.7). Inside.
- Supply-to-return temperature difference indoors 15 F, filter clean.
- Manufacturer's published approach for this condenser: 18 F. Measured condensing saturation from the high-side pressure on the chart: 134 F. Approach 134 minus 108, 26 F. That is 8 F over published.
The reasoning. Voltage and capacitor clear the electrical side. The 8 F excess approach and the visibly blocked coil point the same way. With power locked out at the disconnect and the fan blade still, the tech washes the coil from the inside out at low pressure. After 15 minutes running, condensing saturation reads 126 F, an approach of 18 F, matching published. Indoor delta stays near 15 F.
What the tech tells the customer. The coil was the fault and it is fixed. The house will still drift above 74 F this afternoon, because 108 F is 10 F hotter than the system was designed for, and the expanded performance table for this unit shows it delivering less than its rated capacity at that temperature. It should recover to setpoint after sundown. If it does not recover by morning, call back.
What would have changed the answer. If the approach had been normal with the coil clean, the next suspects are indoor airflow and charge. If the voltage at start had dropped below 207 V and the compressor tripped on its overload, the call goes to the utility for a voltage complaint, with the readings and times written down, rather than to a parts swap. If the indoor temperature had kept rising overnight, the system is not keeping up even at lower load, which is a capacity or a duct problem, not the heat wave.
The failure mode. A tech who reads 134 F condensing, decides the head pressure is "high," and recovers refrigerant to bring it down will undercharge the system. On the next 85 F day it runs with low subcooling, the evaporator may frost, and the callback comes three weeks later on a day when nothing was wrong with the weather.
Beyond air conditioning
Pool: chlorine demand rises with water temperature and sunlight, so readings drift faster between visits. Refrigeration: a refrigerator in a hot garage may be outside the ambient range in its manual; read that range before diagnosing. Garage openers: backup batteries age faster in a hot garage. Roofing: shingles soften and scuff underfoot, and the manufacturer's instructions give handling temperatures.
How to verify you got this right
- Every "can't keep up" ticket records outdoor temperature, published approach, measured approach and indoor delta-T, so a reviewer can tell physics from fault afterward.
- No heat-wave ticket shows refrigerant removed to lower head pressure without a subcooling reading that justified it.
- Review heat-wave callbacks two weeks later. A cluster of capacitor failures is normal; charge adjustments followed by low-charge callbacks are the error above.
References
- AHRI Standard 210/240, rating conditions for unitary air conditioners (95 F outdoor rating point)
- ACCA Manual J, residential load calculation and outdoor design conditions
- Manufacturer installation and expanded performance data for the specific unit
- See related: Design Events: What the Code Assumes Will Happen to the House; How to Prepare a House for a Utility Power Shutoff; How Service Work Quietly Removes a House's Hazard Resistance