Most HVAC systems do not fail suddenly. Output begins to miss targets. Energy use rises without explanation. Comfort complaints repeat without a clear mechanical fault. The plant continues running, but control accuracy erodes over time.
In commercial buildings, sensors determine how systems behave. They decide when cooling increases, how much air moves through ductwork, and how ventilation responds to occupancy. When those inputs drift or misrepresent real conditions, the system responds correctly to incorrect information. Over time, inefficient operation becomes normal operating behaviour.
Every HVAC control sequence depends on measured input. Zone temperature drives demand. Supply air temperature governs cooling output. Static pressure influences fan speed. CO₂ and airflow readings shape ventilation rates. When any of these signals are inaccurate, the controller interprets the building as being in a different state than it actually is.
Control loops are designed to correct deviation. If the input is wrong, the loop still corrects. It simply corrects the wrong condition. A supply air sensor reading high keeps cooling valves open longer than required. A mixed air sensor reading low drives unnecessary heating. A fluctuating pressure signal causes fans to hunt continuously. System activity increases while stability declines.
This is where HVAC diagnostics often stall. Mechanical components may be operating within specification. The fault sits in the data directing them. Effective diagnostics start by checking whether system response aligns with physical conditions in the building.
In daily operation, BMS troubleshooting depends less on alarms and more on plausibility. Sensor values can remain within expected limits while still being inaccurate enough to drive excess energy use and unstable comfort. This is why teams can spend months adjusting setpoints and schedules without resolving the root cause.
Most sensor faults are gradual rather than obvious.
Drift develops as sensors move away from original calibration due to age, contamination, or thermal stress. Trend lines appear stable and alarms remain inactive. The system responds consistently to values that are no longer correct, driving excess cooling, reheating, and rising baseload energy use.
Bias errors present as fixed offsets. A sensor reads consistently high or low because of scaling errors, reference mismatches, or configuration mistakes during replacement. These issues often follow maintenance when parameters are reused without validation.
Noise and intermittency introduce instability. Loose terminals, degraded insulation, grounding problems, or shared conduits with power cabling distort low-voltage signals. Pressure and airflow sensors are especially vulnerable. Controllers react by constantly adjusting outputs, accelerating wear even when average values appear acceptable.
Placement errors are frequently mistaken for device failure. A return air sensor near a diffuser reads colder than the occupied zone. A mixed air sensor exposed to stratification reports conditions that never reach the coil. A humidity sensor installed in stagnant air reports values unrelated to the space. In these cases, the sensor functions as installed, but the data does not represent reality.
Environmental exposure compounds all of the above. Sunlight, vibration, condensation, and air wash degrade accuracy over time. These effects rarely appear in trend data, which is why software-only diagnostics often miss them.
Sensor faults usually appear as patterns rather than single events.
Persistent overcooling followed by reheating is a common indicator. Zones drop below setpoint, reheat valves open, and energy use rises without changes in occupancy or weather. This behaviour often traces back to temperature sensors reading low.
Unstable zone control is another signal. Spaces swing between hot and cold while adjacent zones remain steady. Trend data shows constant adjustment rather than settling, which often indicates noisy inputs or poor placement.
Ventilation complaints without occupancy changes suggest airflow or CO₂ sensing problems. Dampers move aggressively while indoor air quality feedback does not align with how the space is used. Both under-ventilation and over-ventilation present as comfort and cost issues rather than alarms.
Simultaneous heating and cooling across the same system usually points to conflicting inputs. Different controllers respond logically to different data, creating internal contradiction.
A gradual rise in energy use under similar operating conditions is another warning. When energy intensity increases without mechanical change, input accuracy deserves scrutiny.
Incorrect sensor data affects performance in layers.
Energy impact appears first. Control logic responds to perceived demand. Cooling cycles extend, fans operate at higher speeds, and ventilation rates drift. These decisions accumulate across operating hours without a clear fault event.
Comfort issues follow. Occupants experience drafts, temperature swings, or stale air even when setpoints appear reasonable on the BMS. Complaints are often treated as balancing problems rather than data problems.
Equipment stress builds quietly. Compressors cycle more frequently. Valves fail to settle. Fans operate away from efficient points. Failure is rarely immediate, but service life shortens and maintenance demand increases.
The equipment performs exactly as instructed. The problem lies in the instruction itself.
A Building Management System executes control logic based on the assumption that sensor inputs are valid. Trend logs often show smooth, believable values that remain within limits while the system steadily overcools or over-ventilates.
Most platforms treat values as trustworthy if they fall inside expected ranges. Drift and bias remain invisible because alarms focus on extremes rather than credibility. A sensor can appear healthy while driving inefficient operation for extended periods.
Alarm fatigue increases risk. Slow deviation rarely triggers urgency and blends into background noise. By the time performance issues are acknowledged, the system has already adapted around inaccurate data.
Highly responsive control logic amplifies the problem. Fast correction improves stability when inputs are accurate. When they are not, the same responsiveness produces instability.
Effective BMS troubleshooting treats the system as a relay, not a source of truth. The key question is whether the data reflects real conditions in the building.
Validation checks whether sensor readings make sense relative to each other and to physical behaviour.
Cross-checking is the starting point. Supply air temperature should align logically with mixed and return air. Duct pressure should move with fan speed. Zone temperatures should follow load patterns rather than remain static while outputs change.
Baseline comparison adds context. Most systems have periods of stable operation after commissioning or major service. Those periods define reference behaviour. If current readings drive different responses under similar loads, drift or configuration errors are likely.
Rate-of-change checks expose noise and intermittency. Real thermal loads change gradually. Sensor values that spike or oscillate faster than physical conditions allow warrant investigation.
Rule-based checks rely on system intent. A fully open cooling valve paired with rising supply air temperature points to either actuator failure or faulty sensing. Outside air dampers remaining open while mixed air mirrors return air indicate sensing chain issues.
Effective diagnostics avoid adjusting setpoints or replacing components until the data itself has been tested.
Maintenance begins with verification. Sensor readings should be confirmed against physical conditions using handheld instruments, temporary reference sensors, or spot measurements.
Isolation follows. Disconnecting a suspect sensor and observing system response helps confirm whether the input drives the behaviour. Substituting a known-good sensor separates device failure from wiring or configuration problems.
Calibration should only occur once placement and signal integrity are confirmed. Calibration cannot correct poor location or environmental interference.
Documentation closes the loop. Control systems often adapt through tuning changes and overrides. When a sensor fault is corrected, those adjustments must be reviewed. Leaving compensations in place after fixing inputs creates new imbalance.
Repeated sensor replacement without cause analysis usually points to deeper issues in placement, wiring, or system design.
Recurring sensor problems are rarely random. They usually trace back to earlier design, installation, or retrofit decisions.
Placement remains the leading cause. Sensors installed for convenience rather than representativeness report conditions that do not reflect the occupied zone or airstream.
Control zoning errors compound the issue. A single sensor may represent spaces with different loads or usage patterns. The controller responds to an average that suits none of them.
Wiring design also matters. Shared conduits with power, long unshielded runs, and poor grounding introduce interference that calibration cannot correct. These problems often surface after retrofits that reuse legacy infrastructure.
Access limitations lock faults in place. Sensors hidden above ceilings or deep within ductwork are rarely checked after commissioning. Dirt, moisture, and damage degrade accuracy while the system compensates silently.
Systems that perform consistently are designed with verification and maintenance in mind, not only initial operation.
Automated fault detection highlights behaviour that no longer aligns with expected operation.
It is effective at identifying slow degradation. Sensor drift, increasing valve runtimes, or widening temperature spread across similar zones become visible when tracked over time.
Automation does not replace physical verification. It cannot confirm poor placement, condensation, sunlight exposure, or physical damage.
Its value lies in timing. Early identification prevents months of unnecessary energy use and unstable control, reducing maintenance driven by symptoms rather than causes.
In South Africa, HVAC performance intersects with energy and workplace standards. SANS 204 and SANS 10400-XA require reasonable energy efficiency relative to building use and climate. Inaccurate sensor data undermines energy reporting even when equipment selection is correct.
Green Star SA and similar rating tools place increasing emphasis on measured performance. Indoor environmental quality depends on stable temperature and ventilation control. Faulty sensing creates gaps between certified intent and actual operation.
Local operating conditions add pressure. Coastal humidity accelerates corrosion. Inland temperature swings stress exposed sensors. Power interruptions affect low-voltage signals. These factors reduce tolerance for sensing errors.
Under the Occupational Health and Safety Act, employers must provide working environments without risk to health. While sensor accuracy is not specified directly, ventilation adequacy and thermal comfort form part of broader duty-of-care obligations.
HVAC performance depends on more than installed capacity. It depends on whether control decisions are based on accurate, representative data throughout the system’s operating life.
As a company that designs and manufactures commercial and modular HVAC systems, Air Options considers long-term operation during system design and build. Sensor placement, access, signal integrity, and control compatibility are addressed to ensure that performance can be verified and maintained well beyond handover.
Early decisions around sensing and control architecture reduce downstream energy waste, instability, and maintenance burden in complex commercial environments.
Get the latest updates in your email box automatically.
Your nickname:
Email address:
Subscribe
Request A Quote