HVAC · Control basics
HVAC Control Basics — Loops, Sequences and Setpoints
How HVAC control works: sensors, controllers and actuators, the sequences that link them, and the strategies that keep plant and terminals coordinated.
Quick answer
HVAC control holds temperature, humidity, air quality and pressure for each space by measuring conditions, comparing them with setpoints, and modulating valves, dampers and fans. A control loop does the moment-to-moment work; a sequence coordinates equipment across modes, schedules and faults.
The control loop
Every control loop has the same parts: a sensor, a setpoint, a controller, an actuator and a disturbance. The controller compares the measured value with the setpoint and produces an output. On/off control is simple but cycles; proportional control reduces cycling but leaves an offset; integral action removes the offset; derivative action damps the response. Most building sequences use PI control and reserve PID for difficult loops. Tuning is not a one-time ceremony. It should be checked at part load, because a loop that is stable at design conditions can hunt at low output. A loop that is stable on a mild day can oscillate on a cold morning, when the coil is near full output.
The equipment it coordinates
A sequence coordinates equipment rather than a single loop. At plant level it stages chillers and boilers, enables pumps and sets supply temperatures. At air-handling level it starts fans, positions dampers and modulates coils. At terminal level it controls fan coils, VAV boxes or radiant circuits. The sequence decides which equipment leads, which follows and when a mode changes from heating to cooling. That mode logic matters more than any single setpoint, because simultaneous heating and cooling is usually a coordination failure, not a tuning failure. A written mode table prevents one engineer from assuming heating while another assumes cooling. It also makes the changeover visible in the graphics.
Strategies that do the real work
Most energy and comfort outcomes come from a small set of strategies. Scheduling keeps equipment off when the building is empty. Optimum start brings conditions into range before occupancy without running all night. Setback and setpoint limits reduce demand outside core hours. Economiser operation uses outdoor air when it is genuinely useful. Demand-controlled ventilation matches fresh air to occupancy. Static-pressure and supply-temperature resets reduce fan and coil energy at part load. Each strategy needs a clear precondition, a defined exit and an alarm when it fails. A strategy without a precondition is a guess, and a strategy without an exit is a trap. Document how the strategies interact in operation, because a schedule and a temperature reset can quietly undo each other without either being wrong.
Writing a sequence that can be tested
A sequence is only useful if a commissioning engineer can test it. Write it as states, transitions, setpoints, timers and fail positions. State what happens on sensor failure, fan failure, network loss and manual override. Define which alarms are generated and who acknowledges them. Separate safety functions from efficiency functions so a fault cannot quietly disable both. Then write the acceptance test next to each requirement, not in a separate document that nobody reads. A sequence that cannot be tested will be interpreted differently by every engineer who touches it. Handover should include the test results, not only the sequence document.
- State: what the equipment is doing now
- Transition: what causes the next state
- Setpoint: the value and its allowed range
- Failure: the safe position and the alarm
Core HVAC control strategies
6 rows| Strategy | Mechanism | Precondition |
|---|---|---|
| Scheduling | Runs equipment only when needed | Reliable occupancy pattern |
| Optimum start | Pre-conditions the space before occupancy | Known thermal response |
| Setback | Relaxes setpoints outside occupied hours | Space can tolerate drift |
| Economiser | Uses outdoor air for free cooling | Suitable climate and dampers |
| Demand-controlled ventilation | Matches fresh air to occupancy | CO₂ or occupancy sensing |
| Pressure and temperature resets | Reduces fan and coil output at part load | Variable-speed fans and valves |
Project images
3 images
Frequently asked
01What is the difference between a loop and a sequence?
A loop controls one variable, such as room temperature. A sequence coordinates several loops and equipment states over time, including startup, changeover and fault handling.
02Do I need PID for every loop?
No. Many building loops work well with PI control. PID is useful where the response is difficult or the load changes quickly, but it is not a default requirement.
03What is a deadband?
A deliberate gap between heating and cooling setpoints. It prevents the system from heating and cooling at the same time and reduces actuator wear.
04Why do setpoints drift?
Overrides, sensor error, changed occupancy and unclear ownership. A periodic setpoint audit is part of routine commissioning, not a one-off task.
05Can the BMS run the control sequences?
It can hold schedules and setpoint limits, but fast control loops should run in controllers so they survive a network outage.
Next step
Bring us the building.
Tell us the building type, systems in scope and the outcome you need. We will map the shortest path to a working system.