On deep winter nights in northern China, when office buildings, shopping malls and schools sit empty, the HVAC system enters the most dangerous hours of its year. Fresh air units on the roof, pipe shafts along exterior walls and unoccupied end rooms all sit exposed to freezing risk. Tradition offers two extremes: run everything all night — pumps never stop, units hold standby temperature, and the energy bill quietly inflates — or shut everything down and gamble on a mild night, until one cold wave splits a coil and the thaw floods the stairwell. Night anti-freeze mode exists to solve exactly this problem: keep the pipes safe while keeping the energy bill lean.
Many people assume freeze protection means holding indoor temperature. In fact the indoor risk is the smallest — residual heat from occupants, lighting and equipment, plus thermal mass in the walls, makes it hard for occupied-zone temperature to fall below freezing. The real freeze points concentrate in three locations. The first is rooftop equipment: hot-water coils in fresh air units and air handlers sit directly exposed to cold night air, and once fans stop, the still water inside a coil can drop below zero within an hour. The second is pipe shafts and north-facing rooms along exterior walls: stairwells, unheated corridors and north-side washrooms, where pipes cross walls at their thinnest insulation points. The third is ends left unoccupied for long periods: floors closed for holidays, classrooms across winter vacation, vacant shops — no heat is generated there, and nobody notices a leak.
One physical fact matters most: still water freezes at zero degrees, while flowing water tolerates far lower ambient temperatures. Whether a pipe freezes depends not only on air temperature, but also on whether the water moves, how thick the insulation is, and how hard the wind blows across the roof — wind multiplies the convection coefficient several times over, which is exactly why rooftop equipment is far more dangerous than indoor piping. Beyond visible cold, three hidden traps matter more: fresh air dampers that fail to close tight and let cold night air backflow over coils; coil valves with internal leakage, where water seeps slowly and freezes locally; and transoms in smoke vents and elevator machine rooms that create through-draft at night. Designing an anti-freeze mode starts with mapping these three locations and three traps — only then does control logic come into play.
The first principle is zoning. Grade the building by risk instead of applying one strategy everywhere. Occupied core zones only need a normal duty temperature; the real protagonists of the night strategy are the freeze-watch zones — pipe shafts, rooftop equipment floors and unoccupied floors. Zoning pays off by targeting measures precisely: high-risk locations get sensor data scanned every half hour, low-risk ones every hour, aligning monitoring density with risk.
The second principle is temperature setback rather than plain shutdown. At night, setback the heating setpoint of unoccupied zones to eight to twelve degrees and let the system run at low power. Shutting down completely sounds economical, but reheating usually takes longer than the early shift allows, and once walls and pipes go thoroughly cold, local thermal bridges become the most likely freezing spots. Choose the setback value with the envelope and orientation in mind: floors with many exterior walls and curtain walls take the high bound, interior zones the low bound.
The third principle is intermittent circulation. Running circulation pumps continuously through the night is a common waste — well-insulated pipes cool slowly and never need uninterrupted circulation. The correct approach is to set pump intervals from the cooling rate measured by pipe temperature sensors: when pipes cool fast, shorten the interval; when they cool slowly, stretch it. Each pump run lasts ten to twenty minutes, raising the whole-system water temperature by two to three degrees before the pump rests, waiting for the next call. Intermittent logic cuts nighttime pump electricity by roughly half while keeping pipe temperature comfortably far from freezing at every moment.
An anti-freeze mode cannot be alarms alone — it must be a graded response chain: automatic actions by the system first, human actions after, more aggressive at each higher level. The table below gives a reference design of the four levels.
| Level | Trigger | Automatic System Action | Staff Action |
|---|---|---|---|
| Level 1 (Watch) | Outdoor temperature below 5 degrees with forecast nighttime drop | Raise inspection frequency of pipe and equipment temperatures | Aware |
| Level 2 (Warning) | Pipe temperature below 8 degrees | Start pumps at low speed, shorten inspection interval | Monitor the trend |
| Level 3 (Alarm) | Pipe temperature below 4 degrees | Fully open supply and return valves, run pumps continuously, block fresh air unit starts, push alarm | Verify and act on site |
| Level 4 (Emergency) | Pipe temperature below 2 degrees or flow signal lost | Trigger drain-down sequence, phone the duty manager | Act on site and report |
Two points define a sound graded design. First, the thresholds carry hysteresis so the pump does not short-cycle when pipe temperature jitters around a threshold. Second, every level pairs a clear automatic action with a clear human one — an alarm nobody answers equals no alarm at all. The full methodology of dynamic baselines and graded alarms is covered in an earlier article: energy alarm rule design: from fixed thresholds to dynamic baselines in engineering practice.
Rooftop fresh air units deserve separate treatment within the anti-freeze mode, and three actions must be in place on a winter night. The first is the coil water circuit: during downtime, either keep the coil valve open so water keeps moving, or drain the coil completely — pick one, and the worst state is the ambiguous half-open. The second is dampers: fresh air and return dampers must close tight with their position confirmed — a leaking damper leaves the coil naked in the wind. ZYH-DA damper actuators perform the anti-freeze damper interlock and confirm tight closure; the end-position switches built into the actuator feed the control system directly, and the system refuses to confirm the anti-freeze state until the dampers are proven closed. The third is fan interlock: when coil temperature falls below the threshold, the fan must be forbidden to start — manual override included. This interlock belongs in hardwired control logic, not in the duty operator memory.
Compare night anti-freeze mode with running everything all night, and the savings come from four sources. Stopping fresh air units and closing dampers tight eliminates fan electricity and outdoor-air heating energy — typically the bulk of a building nighttime consumption. Intermittent circulation replaces continuous pumping and cuts pump electricity by about half. Setback in unoccupied zones reduces the night load on the boiler plant or heat exchange station. On-demand inspection replaces full-scale monitoring, and even the resource footprint of the control system shrinks. Taken together, a well-executed night anti-freeze mode reduces building nighttime heat consumption by twenty to thirty percent and pump electricity by forty to sixty percent compared with running all night — while freezing risk actually falls, because the system shifts from blind operation to data-watched operation.
Field experience from a tech park office building (desensitized) illustrates the point. The building runs fourteen fresh air units, all of which previously held standby operation through winter nights with pumps running continuously. After retrofitting the anti-freeze mode — the unit shutdown trio, pumps circulating intermittently by cooling rate, and a ten-degree setback in unoccupied zones — one heating season of comparison showed: nighttime pump electricity down about fifty-five percent, nighttime fresh air system energy down about sixty percent, and whole-building nighttime heat consumption down about twenty-eight percent; meanwhile, anti-freeze alerts rose from eleven freeze damage incidents in the previous season to thirty-seven effective early warnings, with zero actual freeze damage. The savings comfortably cover all sensor and control retrofit investment within two years.
Every line of anti-freeze logic stands on real-time data. The sensing layer has three critical mounting spots: coils and supply sections of fresh air units, the top and bottom of heating risers, and wind-exposed elbows of rooftop piping. Duct temperature sensors mounted on fresh air unit coils for anti-freeze monitoring sit on the coil surface itself and catch a freezing trend earlier than supply-air measurement — by the time supply air still reads normal, parts of the coil may already be icing up. The protection bottom line must stay independent of software: the ZYH-100 freeze protection thermostat provides hard-wired low-temperature cut-off on heating pipes and coils, breaking the fan circuit or opening the heating valve through a mechanical low-temperature contact — even if the whole control system collapses, this layer still stands. The verification layer answers whether water really moves: the ZYH-WFS water flow switch verifies real flow in the pipe during night intermittent circulation — a running pump signal does not mean flow is established; a clogged strainer or a half-open valve gives the signal without the water, and the flow switch exposes that illusion. At the control execution layer, building controllers carry the anti-freeze logic locally: sequenced control of supply and return valves, pump start-stop interlocks, and fan start permission all execute on ECC200 or WEB-8000 series controllers in local closed loop, so local logic keeps running even when the network goes down.
Five pitfalls recur. First, staring at fixed thresholds while ignoring the weather forecast — freezing incidents happen on cold wave nights, and finely tuned daytime thresholds are useless then. Feeding the next-day minimum temperature into the night strategy is exactly what separates an anti-freeze mode from an ordinary alarm. Second, freeze stats that alarm without interlock: the alarm rings, nobody knows; somebody knows, nobody acts; and the device ends up as a decoration remembered only at the annual inspection. Third, setting pump intervals by gut feeling: the pipe cooling rate data should do the talking — a well-insulated system and a poorly insulated one can justify intervals twice as far apart. Fourth, doing only part of the fresh air unit trio: a loose damper, a leaking coil valve or an incomplete drain can zero out everything else. Fifth, holiday mode fighting the anti-freeze mode: when the whole-building vacation routine shuts down all non-essential systems, it sometimes shuts freeze monitoring down along with them — the single most common scenario in winter freeze damage accidents.
The implementation checklist condenses to six steps: map the three freeze locations and three traps; define zones and setback values; fit temperature sensors at coils, risers and elbows; configure the four-level response chain with hysteresis; keep the freeze stat hard-wired protection independent of software; and run a full-chain drill before the first cold wave. Finish these six, and the winter night stays as cold as ever — but the pipes sleep in warm circulating water, and the energy bill sleeps easier too.
