Testing HVAC Sequences of Operation
Read and test a written sequence, check economizer and safety logic, and tune or stop control loops that hunt, overshoot or fight each other.
Table of contents
Most control complaints come down to one question: is the system doing what the sequence says, and does the sequence fit the building? A senior controls technician can walk through a sequence point by point, prove each step and explain why a loop is hunting. This is the step between swapping parts and understanding the building. Testing a sequence means commanding real equipment, so it is done by a trained controls technician with the owner's or operator's permission. Some interlocks and fire or smoke functions fall under local code and need the proper authority present.
What You Will Be Able to Do
- Break a sequence down into modes, conditions, setpoints and safeties.
- Test a sequence one step at a time and restore the system afterward.
- Prove economizer operation with temperatures and a percent outdoor air calculation.
- Explain proportional, integral and derivative control in plain terms.
- Tell a tuning problem from a mechanical one.
What a Sequence Contains
A sequence of operation is the written description of how a system should behave. Good ones spell out each of these.
| Part | What it says | Example |
|---|---|---|
| Modes | Which states exist | Occupied, unoccupied, warm-up |
| Enable conditions | What must be true first | Fan proven before heat runs |
| Setpoints | What it controls to | Duct static 1.2 in. w.c. |
| Interlocks | What depends on what | Outdoor damper opens with the fan |
| Safeties | What stops it | Freeze stat stops the fan |
| Failure position | Where it goes on loss of signal | Hot water valve fails open |
Read the sequence, the points list and the drawings together. If the program, the drawing and the sequence disagree, record it. The sequence may be old, or someone may have changed the program.
Walking a Sequence
Work one step at a time.
- Get permission, and tell the operator what you will do and when.
- Back up the program, and record current setpoints and overrides.
- List each sequence step as a test: what you will change, and what must happen.
- Trigger the step, ideally by changing a real condition or a setpoint. If you must override a point, write it down.
- Watch the result at the equipment, not only on the screen. A screen shows a command, and your eyes and meter show what happened.
- Restore the point before moving on, and record pass or fail.
- At the end, confirm that every override has been released.
Warning: Commanding equipment can start fans, pumps and compressors with no warning. People may be working on them. Coordinate with the site, follow lockout rules for any hands-on work, and do not defeat a life-safety or smoke control function during a test unless the correct authority is running that test.
Worked example, a variable air volume air handler:
| Step | Test | Should happen |
|---|---|---|
| Occupied start | Schedule goes occupied | Supply fan starts, status proves |
| Fan failure | Status stays off after command | Alarm, and the sequence stops |
| Static control | Change the setpoint from 1.2 to 1.0 in. w.c. | Fan slows, and static settles at 1.0 |
| Discharge temperature | Raise the setpoint 2°F | Cooling valve closes partway |
| Freeze stat | Trip the test button | Fan stops, outdoor damper closes, valve opens |
Proving an Economizer
An economizer uses outdoor air for free cooling. Economizers fail often, and they fail in ways that do not trip an alarm, so test them on purpose.
- Confirm the enable logic. A common one is that outdoor air is below a high limit and there is a call for cooling. The limit depends on the climate and the energy code.
- Compare outdoor, return and mixed air temperatures with a separate thermometer.
- Command the damper through its range, and watch it move at the actuator.
- Confirm the minimum position for ventilation.
- Confirm what happens when mechanical cooling comes on.
The mixed air temperature must fall between outdoor and return air. If it does not, a sensor is wrong or badly placed. You can calculate outdoor air from those three readings.
Formula: percent outdoor air = (return minus mixed) divided by (return minus outdoor) x 100.
Worked example: return air is 74°F, outdoor air is 54°F and mixed air is 69°F. Percent outdoor air = (74 minus 69) divided by (74 minus 54) x 100 = 25%. If the screen says the damper is at 80% and the math shows 25%, the linkage has slipped, the damper is stuck, or the sensor is wrong. This works best with a large temperature difference between outdoor and return air.
Safeties Come Before Software
Hardwired safeties such as freeze stats, smoke detectors and high-pressure cutouts often act outside the controller, in the fan circuit. A fan that will not start even when the program commands it may have an open safety. Many need a manual reset.
Never bypass a safety to get a system running. Find out why it opened, and make sure the operator knows.
Control Loops in Plain Words
A control loop compares a sensor with a setpoint and moves an output. The difference between them is the error.
Proportional (P) moves the output in step with the error. The range of error over which the output goes from 0 to 100% is the proportional band, also called the throttling range.
Worked example: the setpoint is 55°F, and a cooling valve runs from 0% at 55°F to 100% at 65°F, so the band is 10°F. At 58°F, the valve is at 30%. If the load needs the valve at 30%, the temperature stays at 58°F. It holds 3°F above setpoint for good. That leftover error is called offset, and proportional control alone cannot remove it.
Integral (I) adds output over time while any error remains, so it removes offset. Too much of it causes slow, rolling swings. If an output cannot move any farther while error stays, the integral term can build up and cause a large overshoot later. This is called windup, and it often shows up after a system restart.
Derivative (D) reacts to how fast the error is changing. HVAC loops rarely need it, and it can amplify sensor noise. Many loops run with it set to zero.
Hunting or Something Else
Trend the sensor, the setpoint and the output together. The shape tells you a lot.
| Trend shape | Likely cause | First step |
|---|---|---|
| Steady swing around setpoint | Gain too high, or too much integral | Reduce gain in small steps |
| Sits off setpoint, steady | Proportional only, or a limit reached | Check integral and the device's range |
| Slow to recover | Gain too low | Raise gain in small steps |
| Output steps in jumps, swings stay | Sticking valve or slack linkage | Check the actuator |
| Output pinned at 100%, setpoint missed | Not enough capacity, or a failed device | Check the equipment |
| Sensor jumps with no output change | Noise or a bad sensor | Check wiring and location |
Before you tune, rule out mechanical problems. No tuning fixes a sticking damper, an oversized valve or a sensor in the wrong place. Change one term at a time, wait long enough to see the result, and record the old value. Slow systems, like space temperature, need a long time between changes.
When Loops Fight
Two loops can work against each other while each looks fine by itself.
- Heating and cooling valves both open, because their setpoints overlap.
- Supply and return fans hunting together around a building pressure setpoint.
- Reheat running on a box that is getting air colder than it needs.
- A static pressure loop and zone dampers chasing each other.
Trend both outputs on one graph. Simultaneous heating and cooling wastes energy and often has no complaint attached to it, so look for it on purpose.
Worked Diagnosis
A floor complains of being hot every afternoon. The air handler shows a 55°F discharge temperature, right at setpoint, with the cooling valve at 40% and the economizer at minimum position. Outdoor air is 58°F, so free cooling should help.
- A separate thermometer in the duct, past the sensor, reads 67°F. The sensor reading is wrong, so the loop is holding a false 55°F.
- The sensor reads low because it sits right behind the chilled water coil, in a stream of cold air leaking past a poor seal.
- Percent outdoor air calculates near 10%, matching the minimum position.
- Outdoor air at 58°F is well below return air, but the economizer stays at minimum. The high limit in the program is set below 58°F, which is too low for this climate.
You move the sensor, confirm the discharge temperature, and report the high limit setting. Changing a sequence value belongs to the person with the authority to approve it. Record that, then retest the next day.
Common Mistakes
| Mistake | What it costs |
|---|---|
| Trusting the screen without checking the device | A stuck damper reads as open |
| Leaving a point overridden after a test | A hidden fault later |
| Tuning a mechanical problem | Hours lost, and the swing stays |
| Changing two loop settings at once | No idea which one helped |
| Calling a working sequence a fault | A correct system gets changed |
Check Your Understanding
The program, the drawing and the written sequence disagree about a damper's failure position. What do you do? Record it. The sequence may be old, or someone may have changed the program.
A cooling valve runs from 0% at a 60°F setpoint to 100% at 70°F. The temperature holds at 62°F. Where is the valve, and what is the 2°F called? The valve is at 20%, and the 2°F is offset.
You lower a loop's gain to stop a swing, and the swing does not change. What next? Put the old value back and check the hardware, because tuning cannot fix a sticking valve or slack linkage.
The screen shows a fan commanded on. Does that prove the fan is running? No. A screen shows a command, so check the status and the fan itself.
Economizers rarely trip an alarm when they fail. How do you find one that has? Test it on purpose: confirm the enable logic, check temperatures with a separate thermometer and watch the damper move at the actuator.
Key Takeaways
- A sequence spells out modes, conditions, setpoints, interlocks, safeties and failure positions.
- Test one step at a time, watch the device itself and restore every override.
- Prove an economizer with real temperatures and the percent outdoor air formula.
- Proportional control leaves offset, and integral removes it.
- Rule out mechanical faults before you tune.
- Trend paired outputs to catch loops that fight each other.