Smart Buildings
IoT HVAC Monitoring in Extreme Heat
Published by IOT Phoenix Research & Editorial Team

- What Extreme Heat Does to Cooling Equipment
- The Signals That Matter
- Fault Signatures in Extreme Heat
- Fouled or Blocked Condenser Coil
- Failed or Slow Condenser Fan
- Low Refrigerant Charge
- Hot Air Recirculation
- Economizer Stuck Open
- Short Cycling and Oversized or Mismatched Units
- Chiller Plant Issues
- Practical Rules for Analytics
- Sampling Rates and Storage
- Installation Tips for Roofs
- Limits of Remote Monitoring
- Example: Catching a Failing Condenser Fan on a 113 Degree Day
- What Comes Next
- Building a Baseline Before Summer
- Prioritizing Which Units to Instrument
- Sources
Most articles about HVAC monitoring explain why to monitor. This one explains what to look for. In Phoenix, rooftop units and chillers operate in outdoor air that reached 115 degrees Fahrenheit on August 25, 2026, according to KTAR, and on days like that equipment stops behaving like its brochure. Capacity falls, pressures climb, and small faults that never mattered in April cause shutdowns in July. A monitoring system that collects the right signals can recognize these patterns early. This article is written for facility engineers, controls contractors, and technically minded owners. It describes the signals, the fault signatures that matter in extreme heat, practical sampling and analytics rules, and the limits of what remote data can tell you. For the owner-level overview, see smarter HVAC monitoring for Phoenix commercial properties.
What Extreme Heat Does to Cooling Equipment
Air conditioners reject heat to the outdoors. The hotter the outside air, the higher the condensing temperature must be to reject that heat, which raises compressor discharge pressure and power draw and reduces cooling capacity. A unit that delivers its rated capacity at 95 degrees outdoor temperature delivers noticeably less at 115, while the building load is at its peak. Margin that looked comfortable at design conditions disappears. Any pre-existing weakness, such as a fouled condenser coil, a low refrigerant charge, a failing condenser fan, or a recirculating hot air stream on a crowded roof, is amplified.
The Signals That Matter
| Signal | How to Get It | Why It Matters in Heat |
|---|---|---|
| Outdoor air temperature | Local sensor shaded from sun, or reliable weather feed | Reference for judging everything else |
| Supply air temperature | Duct sensor downstream of the coil | Primary indicator of cooling delivered |
| Return air temperature | Duct or space sensor | Together with supply, gives temperature split across the coil |
| Compressor current | Clamp-on current transformer | Rising current at constant conditions points to high head pressure or mechanical wear |
| Discharge and suction pressure | Refrigerant pressure transducers where installed | Direct view of head pressure and charge |
| Condenser approach | Temperature of condenser air leaving versus outdoor air | Rising approach signals fouled coils or failing fans |
| Fan status and current | Clamp-on sensors | Failed or slowed condenser fans are a common summer killer |
| Run time and starts | Controller or current sensor | Short cycling or continuous run indicates capacity or control problems |
| Filter differential pressure | Pressure sensor across filter | Dust loading reduces airflow and capacity |
Not every unit has pressure transducers, and adding them to hundreds of small rooftop units is often not worthwhile. Many useful rules rely only on temperatures and current, which are cheap to measure.
Fault Signatures in Extreme Heat
Fouled or Blocked Condenser Coil
Dust, cottonwood fluff, and debris cut airflow across the condenser. The unit then runs at a higher condensing temperature. In data, look for compressor current rising above its normal band for a given outdoor temperature, with supply air temperature creeping up and a rising condenser approach if measured. The fix is cleaning, and in dusty seasons it should be scheduled by trend, not by calendar.
Failed or Slow Condenser Fan
Fan current drops or goes to zero while compressor current rises. Units may trip on high pressure and restart, so starts per hour climb. This fault can turn a mild afternoon into an outage during a 112 degree peak.
Low Refrigerant Charge
Cooling capacity declines even on moderate days. Supply air temperature is higher than expected, compressor current is lower than usual, and run time lengthens. With pressure data, suction pressure drops. Low charge often indicates a leak, so topping off without repair is a short-term fix.
Hot Air Recirculation
On crowded roofs, units can ingest warm air from neighboring condensers or from rooftop exhausts. Outdoor air near the unit measures hotter than the reference weather station. Performance degrades on windless afternoons. Fixes include moving units, adding baffles, or installing shade structures.
Economizer Stuck Open
Economizers use outdoor air for free cooling when it is cool enough. In Phoenix summers that should be locked out. A stuck damper allows 115 degree air to enter. Mixed air temperature approaches outdoor temperature, supply air rises, and the unit works to exhaustion. Compare damper position commands with actual mixed air readings.
Short Cycling and Oversized or Mismatched Units
Frequent starts with short run times indicate control problems or oversizing, which wastes energy and wears compressors. Heat can aggravate it when high pressure trips cause repeated restarts.
Chiller Plant Issues
For chilled water plants, key signs include rising approach temperature on cooling towers or dry coolers, increasing kilowatts per ton, low delta T across the chilled water loop, and pumps running at maximum while supply temperature rises. These patterns suggest fouling, poor control, or capacity limits.
Practical Rules for Analytics
- Normalize by outdoor temperature. Compare current and supply temperature with the unit's own history at similar outdoor conditions.
- Use time delays. Require a fault condition to persist for 15 to 30 minutes to avoid false alarms from transients.
- Look for trends, not only thresholds. A five percent rise in compressor current over three weeks is more informative than a single high reading.
- Rank by consequence and cost. A unit serving a server room outranks one serving a storage corridor.
- Cross-check with peers. Compare identical units on the same roof. Outliers stand out.
- Flag bad sensors. Flat lines, impossible values, and disagreement with neighbors point to sensor issues, not equipment faults.
Sampling Rates and Storage
For most commercial rooftop units, one-minute to five-minute sampling of temperatures and current is enough for fault detection. Faster sampling, such as seconds, helps diagnose short cycling and startup behavior but creates more data. Keep at least one full year of data so you can compare summer to summer. Store raw data and summaries, and ensure time stamps are accurate and consistent across devices.
Installation Tips for Roofs
- Shade and shield. Mount outdoor sensors on the north side or in radiation shields so they read air, not sun-heated metal.
- Secure wiring. Wind, heat, and UV punish loose cable and tape.
- Label everything. Units, sensors, and gateways should be labeled and mapped.
- Mind safety. Work on electrical equipment by qualified personnel, and use clamp-on devices only as intended.
- Document. Record sensor locations, calibration, and installation dates.
Limits of Remote Monitoring
Data can show that something is wrong and often what kind of problem it is, but not always why. A technician still needs to inspect the unit, measure refrigerant pressures properly, and confirm the diagnosis. Monitoring does not replace maintenance. It targets it. Overconfident automated diagnoses can lead to wasted trips, so choose systems that explain their reasoning and let technicians correct them.
Example: Catching a Failing Condenser Fan on a 113 Degree Day
Suppose a hypothetical retail center has clamp-on current sensors on 25 rooftop units. On a hot afternoon, the system flags that one unit's compressor current has risen 12 percent compared with its normal band at that outdoor temperature, while its condenser fan current has dropped to near zero. Starts per hour have also climbed, indicating high pressure trips. An alert goes to the service contractor with the unit's name, the data, and a suspected cause. A technician arrives within the hour, finds a failed fan motor, and swaps it using a stocked spare. The store never loses cooling. The scenario is illustrative, but it shows the kind of signature-based alert that this article describes.
What Comes Next
Expect fault detection to move closer to the equipment, with controllers that flag issues locally, and more standardized data from manufacturers. Phoenix's heat will keep testing assumptions, and records are likely to keep falling. For related guidance, read smart building sensors: what Phoenix properties need, smart HVAC systems in Phoenix, and our smart buildings section.
Building a Baseline Before Summer
Fault detection works best when each unit has a baseline of normal behavior. Collect data in spring, when outdoor temperatures range from mild to hot, and record compressor current, supply air temperature, and run time against outdoor air temperature. These curves let you see how a unit behaves as it warms, and they provide the reference for summer comparisons. After maintenance, such as coil cleaning or refrigerant repair, record a new baseline. A unit whose behavior does not improve after a repair is a candidate for deeper investigation or replacement.
Prioritizing Which Units to Instrument
Start with units that serve critical spaces, such as server rooms, clinics, and areas with sensitive inventory, followed by the oldest units and those with a history of failures. Add the rest as budget allows. Even partial coverage helps, because patterns on a few units often reveal systemic issues such as poor roof layout or missed filter changes.
Whichever route a team chooses, the practical advice is the same: start with a clear problem, measure a baseline, pilot on a small scale, and publish what you learn internally. Honest review of what worked and what did not builds the trust and skills needed for larger programs, and it surfaces issues with heat, connectivity, and staffing while they are still inexpensive to fix.
