Why Relative Humidity Alone Can Mislead a BESS Cooling Review
Condensation depends on the relationship between water vapour in the air and the temperature of the surface it reaches. Relative humidity is only one part of that relationship.
A BESS cooling cabinet may see humid outdoor intake air, a cold coil, a drain pan, insulated pipes, panel seams and locally cooled metal surfaces. A room or ambient relative-humidity reading does not state whether a particular surface is below the incoming air's dew point. ASHRAE guidance therefore treats dew point compared with surface temperature as the more useful moisture-control metric.
The reason is that the two quantities behave differently. Dew point is the temperature to which air must be cooled, at constant moisture content, before it becomes saturated. Relative humidity is the ratio of the moisture actually present to the maximum the air could hold at its current temperature. Warm the air and the relative-humidity reading falls even though not one gram of water has left the enclosure; cool it and the reading climbs. The dew point does not move.
That property is why the recommended envelope in ASHRAE's Thermal Guidelines is written as a dew-point band with a relative-humidity cap attached. Both numbers are needed because they answer different questions, and the lower of the two applies.
This does not make one published dew-point number a universal BESS alarm. The relevant condition is the measured local intake air, the actual coldest exposed surface, air infiltration, insulation and vapour sealing, coil control, drainage, maintenance state and the project's allowed operating modes. A dew point of 19 C is unremarkable in a cabinet whose coldest surface sits at 31 C, and it is a wet-cabinet condition in one whose coolant loop is running at 11 C.
The Moisture Limits Are Written in Dew Point, Not Relative Humidity
The published moisture numbers are screening limits for electronic equipment, not a comfort specification for a battery cabinet.
For the 2021 fifth edition of the Thermal Guidelines, the recommended moisture content for air-cooled classes A1 to A4 is -9 C to 15 C dew point, together with an upper limit of 70% relative humidity where gaseous contamination is low or 50% where it is high, and the standard states that whichever of the two caps represents the minimum moisture content applies. Where chlorine or hydrogen sulphide may be present, the tighter cap and chemical filtration belong in the design until measurements say otherwise.
The relative-humidity cap exists because corrosion does not depend on moisture content alone. ASHRAE research project RP-1755 exposed copper and silver coupons to mixed flowing gas at a reference 21 C and 50% relative humidity, then raised humidity and temperature. Copper corrosion accelerated between 50% and 70% relative humidity in the presence of chlorine, and at 80% significant copper corrosion appeared regardless of which pollutants were present, implying a critical humidity for copper between 70% and 80%. Chlorine dominated copper and hydrogen sulphide dominated silver; the acceptance basis is a thirty-day corrosion thickness below 300 angstroms for copper and 200 angstroms for silver.
The low end of the band deserves the same respect. The reason the older relative-humidity floor was removed is ASHRAE research project RP-1499, which examined electrostatic discharge risk at low humidity and found that a facility already operating with a low rate of ESD-induced damage at 25% relative humidity maintained that low rate at 8%. That is why the recommended band records a dew-point floor of -9 C. None of these figures are condensation limits; they describe the air the equipment is designed to accept.
| Class | Allowable air temperature | Allowable moisture | Maximum dew point |
|---|---|---|---|
| A1 | 15 C to 32 C | -12 C dp and 8% rh up to 17 C dp and 80% rh | 17 C |
| A2 | 10 C to 35 C | -12 C dp and 8% rh up to 21 C dp and 80% rh | 21 C |
| A3 | 5 C to 40 C | -12 C dp and 8% rh up to 24 C dp and 85% rh | 24 C |
| A4 | 5 C to 45 C | -12 C dp and 8% rh up to 24 C dp and 90% rh | 24 C |
Interactive: Screen Dew Point Against a Cold Surface
Adjust the air temperature, relative humidity and cold-surface temperature to see a first-pass condensation screen.
The calculation estimates dew point from air temperature and relative humidity using a standard approximation. It is a teaching and screening aid only: it does not model coil geometry, boundary layers, transient control, drainage, leakage or the BESS safety system. Use calibrated site measurements and the project acceptance plan for a decision.
The Coldest Surface in a BESS Is Rarely the Room Air
Condensation picks the lowest surface temperature it can reach, and in a battery cabinet that is usually a component rather than the enclosure atmosphere.
Air is a poor conductor and holds little thermal mass, so the air inside a cabinet follows the surfaces around it rather than the other way round. Water therefore appears first on whichever surface is coldest and reachable by the moisture in the air, and that surface is normally one of a small set: the cooling coil and its fins, the wall of a coolant pipe where insulation is missing or damaged, the inside face of the steel enclosure after a rapid ambient drop, the drain pan itself, cold-plate and liquid-loop fittings inside electrical compartments, and any metal penetration such as a gland plate or cable entry that conducts outdoor temperature straight through the envelope.
A coolant pipe is the case that most often gets misread in the field. A liquid-cooled BESS circulates coolant at a temperature set by the battery's own thermal requirement, and in a hot humid cabinet the pipe wall can be well below the dew point of the air around it. Water then appears on the pipe even though the circuit is perfectly sealed, and the first response is usually to look for a leak. Checking loop pressure, coolant level and fitting torque will not find it, because the water never came from inside the pipe.
Intermittent cold surfaces behave differently again. A steel wall exposed to sun by day and to a clear sky at night changes temperature much faster than the air inside, and after a sharp evening drop its inner face can fall below the dew point of air that has not yet cooled. The same applies to a cabinet that has been shut down, a container that has just been transported, and any surface near a door left open for service. These surfaces are not cold because of the cooling system, so a control strategy that only watches the coolant loop will not see them.
- Cooling coil and fins - the coldest reliable surface whenever the coil runs below the local dew point
- Coolant pipe walls and liquid-loop fittings, especially where insulation is compressed, cut or absent at a joint
- Inside face of the steel enclosure after a fast ambient drop, plus ceiling panels and lifting points
- Drain pan, condensate trap and the first metre of the drain route
- Gland plates, cable entries and other metal penetrations that bridge the envelope
- Load-dependent surfaces: power modules, busbars and cold plates that cool down when the load drops
What Surface Moisture Actually Does to a BESS
In a battery cabinet, moisture is not a comfort issue. It is an electrical-insulation issue with a documented path to a safety event.
A grid-scale battery cabinet runs a high-voltage direct-current bus, commonly in the hundreds of volts and increasingly at 800 V or above. Clearances and creepage distances inside the high-voltage box, the combiner, the contactors and the battery management electronics are dimensioned for dry air. Water on a surface defeats that assumption. Insulation resistance falls, leakage current rises, and the battery management system responds to its own insulation-resistance monitoring, which typically trips an insulation fault in the order of one megohm and stops the string rather than waiting to see what happens next.
That trip is the good outcome. A direct-current arc does not extinguish itself at a current zero crossing the way an alternating-current arc does, so surface contamination that progresses from leakage to flashover inside a battery compartment is a genuinely dangerous event rather than a nuisance alarm. The industry framing of the escalation path runs from humidity to condensation, condensation to electrical degradation, degradation to an electrical fault, and the fault to thermal runaway. Moisture management belongs on that chain rather than in a housekeeping checklist.
Corrosion is the slower half of the same problem. Copper busbar joints and terminals exposed to sustained high humidity corrode, the corrosion products increase contact resistance, and the raised resistance produces local heating that further degrades the joint. The corrosion work cited above is directly relevant: the presence of chlorine or hydrogen sulphide lowers the humidity at which damage begins, so a coastal or industrial site has less moisture headroom than a clean inland one for the same cabinet design.
Three Condensation Modes You Have to Test Separately
A cabinet that is dry at full load and mid-day can still be wet at night, at low load, or in the hours after a restart.
The first mode is operational condensation: the cabinet is running and a surface in the normal air path is below the local dew point. The evidence is water on the coil, on the drain line or on an uninsulated joint, and it usually appears together with a slow drift in insulation resistance rather than with any change in fan speed.
The second is night or swing condensation, driven by the enclosure rather than by the cooling system: a rapid ambient drop cools the steel envelope faster than the internal air, the inner face crosses the dew point, and water forms on walls and lifting points before it forms anywhere the coolant loop can see. The third, most often missed in commissioning, is restart condensation. After an outage or a service visit with the doors open, the thermal mass of the cabinet is cold while the air arriving with the next shift of humid weather is warm and wet. The moisture does not need a running cooling system to condense; it only needs a cold surface. Each mode has its own trigger, timing and evidence, so the acceptance plan should name which ones are being demonstrated.
| Mode | Cause | What you see | What to verify |
|---|---|---|---|
| Operational | Cabinet running with the coil or a pipe below the local dew point | Water on the coil, drain line or an uninsulated joint; slow insulation-resistance drift | Coil set point against the measured dew point; insulation and vapour barrier on cold lines; drain route and trap |
| Night or swing | Fast ambient drop cooling the envelope faster than the internal air | Water on inner wall faces and lifting points; drips onto the top of modules | Wall insulation and thermal breaks; sealed penetrations; heater and fan interlock; panel seals |
| Restart or service | Warm humid air arriving at a cabinet whose thermal mass is still cold | Water appearing shortly after restart rather than during steady state | A defined dry-out step before closing up; restart ramp; door-open log |
Follow the Water Path, Not Just the Fan Command
A fan can alter air distribution and coil duty, but it does not replace a complete condensate-management design.
A practical commissioning record aligns measurements in time: local inlet dry-bulb and humidity, calculated or measured dew point, selected surface temperatures, fan command and FG feedback, coil or cabinet mode, drain observation and named equipment temperatures. Record panels and louvers fitted, because an open service door can invalidate the pressure and infiltration condition. Time alignment is what makes the record usable: a humidity reading at 14:00 and a surface temperature at 16:00 describe two different cabinets as far as moisture is concerned.
If a wet condition appears, do not conclude that the fan is defective from RPM alone. Inspect the sequence of operation, air leaks, drainage, coil condition, sensor placement and the installed pressure path before deciding whether the fan duty or air-path design needs changing. Tachometer feedback reports that a rotor is turning; it says nothing about whether the coil is colder than the air, whether the drain is carrying water away, or whether humid air is entering through a gland plate that was never sealed.
| Check | Why it matters | Evidence to retain |
|---|---|---|
| Local intake condition | Weather-station readings can differ from air entering a louver or cabinet | Air temperature, RH or dew point, probe location, time and cabinet state |
| Cold surfaces | A coil, drain line, uninsulated joint or leakage path can be colder than the room air | Surface-temperature locations, insulation and vapour-seal details |
| Drainage | Water that forms must leave without pooling or reaching sensitive interfaces | Drain-pan slope, trap/drain route, blockage inspection and wet-condition check |
| Air leakage and pressure | Unplanned humid-air ingress can reach cold components outside the intended path | Panel, gland and service-door condition; fan mode and measured pressure |
| Control response | Low load, fan speed and coil control can produce a different wet condition than peak thermal load | Mode sequence, commanded speed, temperatures and observation window |
Where a Fan Helps, Where It Cannot, and What to Alarm
Air movement is one input to condensation control. It is not a moisture-removal device, and treating it as one produces wet cabinets with healthy airflow readings.
A fan changes where air goes, how fast it crosses a surface, and how much of the coil's capacity is actually used. Moving more of the humid intake air across a coil that is below the dew point condenses more water out of it, which is how dehumidification works in the first place. The catch is that the water has to leave. If the drain is blocked or its trap has dried out, increasing airflow across a cold coil moves more water into the cabinet rather than out of it.
The second trap is the stopped fan with a cold coil. When cooling stops but the coil is still below the dew point, a stagnant pocket of humid air sits against a cold surface and condenses there, with no airflow to carry dried air away. Idle periods at night are exactly when ambient humidity is highest in many climates, which makes this a design condition rather than a hypothetical one. The third trap is dry-out after service: airflow alone only redistributes moisture that has been absorbed by insulation, dust on the coil or an unsealed crevice.
It is worth separating heating from dehumidification for the same reason. An anti-condensation heater raises a surface above the dew point, which does prevent condensation on that surface. It does not remove water from the enclosure: in a sealed cabinet the moisture content is unchanged, so the moment the enclosure cools again the margin is gone. Heating is a surface-temperature measure and dehumidification is a moisture-content measure.
Where a fan does earn its place: keeping the air mixed so that no stagnant pocket sits under the dew point, distributing dried air after a dehumidification cycle, and providing modulated rather than on-off flow that would re-entrain water from a drain pan.
The control decision follows from the same physics. A condensation control that alarms on a room humidity percentage is watching the wrong variable, because the same reading corresponds to a dry cabinet and a wet one depending on what the coolant is doing. The variable worth alarming is the margin: the coldest instrumented surface temperature minus the dew point of the air reaching it. The dew-point sensor belongs in the intake air path and surface sensors belong on the named cold surfaces. Interlocks are where the design is enforced: the heater, fan and coolant loop should not be able to enter a state that leaves a cold surface exposed to humid air with no airflow.
- More air over a coil below the dew point condenses more water - useful only if the drain and trap can carry it
- A stopped fan with a cold coil strands moisture against the coldest surface in the cabinet
- Alarm on the margin between the coldest instrumented surface and the local dew point, not on a room humidity percentage
- Calibrate humidity sensors on a schedule, and treat a suspiciously stable reading as a fault
- Interlock heater, fan and coolant states, and put a dry-out step in the restart sequence
Specify the Fan Around the Installed Duty
The custom-fan requirement belongs to the complete normal-cooling air path, including the humid or service state that the cabinet is expected to see.
MEGA Tech can review a custom DC cooling fan or a cost-competitive replacement using the actual frame and mounting, target flow at stated pressure, voltage, PWM/FG/RD interface, local inlet condition, noise target and expected filter, louver and coil configuration. For a humidity-sensitive cabinet, add the intended fan control modes, drain and sealing arrangement, condensation scenario and sample acceptance measurements.
Do not approve a replacement from dimensions or maximum free-air CFM alone. The specification is mostly about the part of the duty that is easy to leave out: the fan has to remain controllable across the low-load and idle states where humidity is highest, its feedback has to be truthful at those low speeds, and its airflow has to be characterised at the pressure the cabinet presents with the coil and filter installed. A fan that meets its numbers only at maximum command does not cover the state in which a cabinet actually gets wet.
- Record dew point and cold-surface temperatures in the same time window.
- Inspect drains, seals, cable entries and panels with the fan in its normal installed mode.
- Separate normal cooling requirements from emergency or battery-safety functions.
- Define the measurement locations and acceptance criteria before a wet-condition test.
- Document the service state rather than relying on a one-time clean, dry demonstration.
Engineer FAQ: BESS Dew Point and Condensation
These answers keep humidity discussions attached to measurable installed conditions.
Q: Is relative humidity enough to set a BESS condensation alarm? A: Usually no. Relative humidity changes with air temperature; compare local dew point with relevant cold-surface temperatures and the installed moisture path.
Q: Does a normal fan RPM prove that the cabinet is free from condensation risk? A: No. Tach feedback indicates reported rotation, not coil temperature, water drainage, panel leakage or local surface conditions.
Q: Can a fan solve a drainage or sealing problem? A: No. Fan selection and control must work with the drain path, enclosure seals, insulation, coil control and project sequence of operation.
Q: What should an evaluation sample include? A: Supply the fan or drawing, pressure-flow duty, control interface, enclosure layout, local temperature/humidity conditions and the agreed wet-condition acceptance test.
Q: Why does the ASHRAE envelope use a dew point and a relative-humidity cap together? A: Because the two answer different questions. The dew point describes how much moisture is in the air and does not change when the air temperature changes; the relative-humidity cap exists because corrosion accelerates at high humidity regardless of absolute moisture content. The lower of the two applies.
