Loxone · HVAC
Loxone HVAC — Heating, Cooling and Ventilation Control
How a Loxone control layer sequences chillers, air handling units, fan coils, underfloor heating and radiant circuits, with local logic and open interfaces.
Quick answer
Loxone controls HVAC by running schedules, setpoints, PID and PWM loops and plant sequences on local controllers, while integrating chillers, air handling units, fan coils, underfloor heating and radiant circuits through BACnet, Modbus, 0–10 V signals and dry contacts. One control model then governs the plant and the rooms.
One control model from plant room to room
HVAC in a commercial building is rarely one system. A chiller plant, an air handling unit, fan coils, underfloor loops and radiant panels may come from different suppliers, each with its own controller and its own idea of a setpoint. The control layer has to make them behave as one installation. We map every item to the same occupancy, comfort and energy model, then run the sequences locally on Loxone controllers. The plant keeps working when the wide-area network does not, and the operator sees one status rather than five. The model is documented item by item, so a new floor, a converted tenant area or an added heat pump joins the same structure instead of becoming another island.
- Plant and terminal units share one occupancy model
- Sequences run locally, independent of the cloud
- Every setpoint and command is written to an audit trail
Room and zone strategy
Zoning is the first engineering decision, because it fixes how many sensors, valves and control loops the building will carry. We zone by orientation, thermal behaviour, room use and occupancy pattern rather than by duct runs alone. Perimeter offices, internal meeting rooms, open-plan floors, atria and back-of-house areas usually need different zone sizes. As a typical planning range, a perimeter zone follows one façade and one window-bay grouping, while an open-plan zone is sized so that one temperature sensor represents the air around it; industry reference values commonly sit between 25 and 100 m² per zone, with the final figure set by load, glazing and air distribution. Each zone receives a defined sensor position, setpoint range, occupancy source and terminal device list, so the drawing, the point schedule and the sequence describe the same space. The resulting zone schedule is reviewed with the HVAC designer before it is issued.
- Zones defined by façade, use and thermal behaviour
- Sensor position and setpoint range recorded per zone
- Typical planning ranges quoted, project values confirmed on site
Chillers, heat pumps and air handling units
At plant level the work is protection and coordination. Chillers and heat pumps are staged by load, with setpoint reset, runtime balancing and frost protection. Boilers follow a weather-compensated curve with their own limit and frost logic. Air handling units provide supply-temperature control, damper and valve modulation, filter monitoring and free cooling where the dampers and sensors allow it. Equipment with a factory controller keeps its own safety logic; we read and command it through BACnet/IP, Modbus or a gateway, and the platform adds schedules, alarm routing and trend history around it. The engineering boundary is written down: which controller owns each safety interlock, which one only supervises, and what happens when the link between them is lost. That boundary is agreed before programming starts, not discovered during commissioning. Where the plant is new, the same boundary is agreed with the supplier at order stage.
- Load-based staging, lead-lag and runtime balancing
- Supply temperature and static pressure reset
- Safety ownership documented per plant item
Control sequences and changeover
Every plant item needs a written sequence of operation, not a collection of setpoints. A chiller sequence defines start order, staging thresholds, minimum run and stop times, lead-lag rotation, setpoint reset from return temperature or valve position, and the conditions that stop the plant. An air handling sequence defines damper position before fan start, supply temperature control, economiser changeover, filter alarm and frost protection. Terminal sequences define deadband, valve stroke, fan-speed steps and seasonal changeover. Sequences also cover occupancy transitions: early start, setback, night purge and shutdown. We write them as a control narrative with numbered states, then configure and test each state. A sequence that exists only in the engineer's head is a future fault, because the next person to open the panel will not know why a pump starts before a valve. Each state has an entry condition, an output action and a timeout, which is what makes the sequence testable.
- Numbered states from start-up to shutdown
- Economiser, night purge and seasonal changeover
- Control narrative issued with the point schedule
Terminal units, underfloor heating and radiant surfaces
Terminal units need different treatment. Fan coils respond quickly, so they suit meeting rooms and spaces with variable occupancy; control is a room setpoint with fan speeds, valve modulation and a deadband that prevents heating and cooling from fighting. VAV terminals need duct pressure reset and minimum airflow limits, otherwise a closed damper can starve the rest of the branch. Underfloor heating and radiant or capillary surfaces respond slowly. They need slab or surface temperature sensing, weather compensation and zone valves that open early enough to meet the load without overshooting. Chilled beams add a dew-point limit, because a cold surface in humid air will condense. We keep the fast and slow timescales separate in the logic, and we document which sensor drives which loop, including the fail position of every actuator. Terminal selection and control type are recorded together, because a valve sized for the wrong differential pressure will hunt no matter how well the loop is tuned.
- Room-level setpoints and fan-speed control
- Slab temperature limits for underfloor and radiant zones
- Changeover logic to avoid simultaneous heating and cooling
Device selection and wiring topology
Device selection follows the sequence. A valve actuator is chosen for stroke time, torque and fail position; a sensor for range, accuracy and response time; a controller for the inputs, outputs and spare capacity the zone needs. We reserve spare I/O per panel so a later fit-out does not require a new enclosure, and we keep mains and extra-low-voltage cabling in separate ducts with labelled terminals. Controllers sit in plant rooms and floor risers; Tree or Air devices cover retrofit routes where new cable is impractical; gateways sit close to the equipment they translate. The topology is drawn before installation: controller, extension, field device, network switch and power supply, with addresses reserved. That drawing becomes the as-built and the basis of the point schedule, so an installer and a future maintenance engineer read the same map. Cable schedules record core size, screen and route, and spare cores are terminated rather than left coiled in the ceiling.
- Actuator, sensor and controller selection by duty
- Spare I/O reserved per panel
- Mains and extra-low-voltage separation with labelled terminals
Interfaces, alarms, commissioning and handover
Interfaces are agreed per item: BACnet/IP or Modbus where the equipment exposes a network port, 0–10 V or 4–20 mA for analogue commands, dry contacts for status and safety, and a gateway where the protocol is proprietary. Alarms are classified by consequence — safety, plant protection, comfort, maintenance — with a response time and an owner for each class. During commissioning every point is proven at its address, each sequence is run through its states, and fault behaviour is tested by simulating sensor loss, valve failure and network interruption. Trends for supply and return temperature, valve position, runtime and room conditions are reviewed against the design intent. Handover includes the point schedule, control narrative, alarm list, tested backup and operator training on views, schedules and permitted overrides. Alarm floods are suppressed by grouping related points under one parent alarm, and every suppression has an expiry so a real fault cannot hide indefinitely.
- Interface and command rights confirmed in the audit
- Alarm classes with response time and owner
- Point schedule, narrative, backup and training at handover
HVAC scope, interfaces and control approach
14 rows| Equipment or system | Typical interface | Control sequence | Engineering note |
|---|---|---|---|
| Chiller plant | BACnet/IP, Modbus, dry contacts | Load staging, lead-lag rotation, setpoint reset, minimum run | Vendor safety logic stays active; platform supervises and schedules |
| Boilers | Modbus, 0–10 V, dry contacts | Weather compensation, staging, limit and frost protection | Flue, gas and safety interlocks remain with the boiler controls |
| Heat pumps | Modbus, BACnet/IP, dry contacts | Bivalent staging, defrost status, supply temperature limit | Source-side limits and defrost logic confirmed with the manufacturer |
| Air handling units | BACnet/IP, Modbus, 0–10 V | Damper before fan, supply temperature, economiser, filter alarm | Fan and damper interlock proven during functional test |
| Fan coil units | Modbus, 0–10 V, relays | Room setpoint, fan steps, valve modulation, deadband | Seasonal changeover configured by zone, not by unit |
| VAV terminals | BACnet, Modbus, 0–10 V | Airflow setpoint, reheat, minimum flow, duct pressure reset | Damper position used to reset fan pressure |
| Chilled beams | 0–10 V, zone valves, sensors | Dew-point limit, changeover, zone temperature | Condensation protection has priority over comfort |
| Underfloor heating | 0–10 V, relays, Modbus | Slab temperature limit, weather compensation, zone valves | Slow response written into the schedule and early start |
| Radiant and capillary surfaces | 0–10 V, zone valves, surface sensors | Supply temperature limit, dew-point protection, zone control | Surface sensor placement agreed with the ceiling or wall build-up |
| Ventilation and DX units | Modbus, dry contacts | CO₂ and humidity demand control, boost, fault status | Air-quality trends used to review the ventilation schedule |
| Valves and actuators | 0–10 V, three-point, PWM, feedback | Stroke time, fail position, position feedback | Selected for torque and close-off pressure, not price alone |
| Room sensors | Tree, Air, 0–10 V, Modbus | Temperature, humidity, CO₂, presence | Position away from sun, draughts and supply jets |
| Windows and occupancy | Tree, Air, dry contacts | Setback, mode change, window-open alarm | Local override expires and is logged |
| Meters and gateways | Modbus RTU/TCP, pulse, IP | Metering per plant item, protocol translation | Address list and spare capacity included in the as-built |
Project images
3 images
Frequently asked
01Can Loxone control an existing chiller or AHU?
Usually through BACnet/IP, Modbus, 0–10 V signals or dry contacts, provided the existing controller exposes a usable interface. The audit confirms what is readable and what is commandable.
02Does this replace our existing BMS?
Not necessarily. Loxone can act as the local control layer, or supervise an existing BMS through an open protocol. Replacement is a separate commercial decision.
03How is underfloor heating controlled without overshoot?
Slab or surface temperature sensing, weather compensation and zone valves, with a lower limit on the supply temperature. The slow response is built into the schedule rather than corrected by frequent setpoint changes.
04Can a room heat and cool at the same time?
Only by design error. Changeover logic, deadbands and valve interlocks prevent simultaneous heating and cooling; the sequence is documented and tested.
05Is PID control available?
Yes. PID loops are configured for temperature and pressure, and PWM outputs are used for thermal actuators that cannot accept a continuous signal. Loops are tuned against the installed equipment.
06What happens during a network outage?
Local controllers keep running the schedules, sequences and interlocks. Central supervision and remote access may be unavailable, but the plant does not stop.
07How do you prove the strategy works?
Through trend data: setpoints, valve position, supply and return temperatures, runtime and room conditions. These are reviewed at commissioning and can be exported for review later.
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.