Battery Enclosure Condensation: Why Sealed Packs and BESS Cabinets Still Get Wet

Part of our iHaaS deep-dive series. New to HaaS? Start here.

Battery systems are built to keep water out. Then a teardown after a humid season finds water marks on the cold plate and corrosion on a busbar. Nobody spilled anything. The water came in as vapor and turned to liquid on the coldest surface it could find.

At ATMOSIScience®, we make anti-condensation packs for enclosed spaces, and new-energy batteries are one of the applications we design them for. We also build the HaaS DEW condensation sensors that show what's happening inside the cavity. This guide covers why battery packs and battery energy storage system (BESS) cabinets sweat, the math for the two numbers that matter most, and how to instrument and validate a fix. If you want the general dew point basics first, start with our enclosure condensation guide.

Why battery systems invite condensation

Most enclosures only have one cold surface: the wall facing the night sky. A battery system often has a second one built in. Cells get hot under load, so most traction packs and many BESS racks use liquid cooling. That cold plate is now the coldest thing in the box, sitting right under the cells and close to sensing wires and connectors.

A study of traction battery systems from Technische Hochschule Ingolstadt, published in Energies, puts it plainly: liquid or evaporative cooling can result in water condensing inside the battery system, and condensation occurs when the cooling plate is below the dew point. The authors list what that water can do: damage electrical components, cause corrosion, reduce insulation resistance, and possibly accelerate cell aging (Kim et al., Energies, 2019).

BESS cabinets and containers face the same physics at room scale. Cooling coils run cold, doors open for service on humid days, and systems sit idle during commissioning while the weather cycles. The rule doesn't change: any surface below the dew point of the air touching it gets wet.

Three ways water gets into a "sealed" pack

The same Ingolstadt team built a physical battery simulator with a liquid-cooled plate and tested three housing designs: fully sealed, open, and fitted with a pressure compensation element (PCE), the vent that lets a housing equalize pressure. They identified three routes for water (Kim et al.):

1. It's already inside. Water vapor in the air at assembly condenses on cold parts as they cool. This part is reversible: it evaporates again when things warm up, and it doesn't keep accumulating on its own.
2. It's pulled in when the pack cools. Cooling lowers the internal pressure, so outside air is drawn in through the openings. Roughly the same amount leaves when the pack warms. Bigger temperature swings, such as cold coolant during heavy loads, mean bigger pressure changes.
3. It diffuses in toward the cold plate. Even with equal pressure inside and out, vapor moves along partial pressure gradients. When vapor condenses on the plate, the local vapor pressure drops, and more moisture moves in to replace it.

Two findings stand out. First, PCEs are designed to balance pressure while blocking liquids, dust and larger particles, so they have little effect on vapor getting in. Second, condensation formed on the internal surface in all three designs. In their normal-operation test, condensation on the cold plate peaked at about 16 mg in the sealed system and was at least 26 mg in the open one. Under a worst-case faulty-cooling profile, those figures rose to about 98 mg for the sealed system and at least 140 mg for the open one.

The lesson for designers: sealing changes how much condenses, not whether it does. You still need a plan for the water that gets there.

Worked example: the water you seal in at assembly

Route 1 is the one you control at the factory, so let's put numbers on it. You need two formulas. The Magnus formula gives saturation vapor pressure from temperature, using the constants in the World Meteorological Organization's guide (WMO-No. 8, Annex 4.B):

e_s(T) = 6.112 × exp(17.62 × T / (243.12 + T)), in hPa

Then absolute humidity (grams of water vapor per cubic meter of air) follows from the ideal gas law, as used in the Ingolstadt study:

H = 2.167 × e / (273.15 + T), with e in Pa and H in g/m³

Take a pack with 40 L of free air inside, closed up on a mild day at 25 °C and 50% RH:

• e_s(25 °C) = 31.6 hPa, so the actual vapor pressure is 50% of that: 15.8 hPa, or 1,580 Pa.
• H = 2.167 × 1,580 / 298.15 = 11.5 g/m³.
• Water in 40 L (0.040 m³): 11.5 × 0.040 = 0.46 g.
• Dew point of that air: 13.9 °C. Any surface colder than that gets wet.

Now run the coolant so the plate sits at 10 °C. Saturated air at 10 °C holds only 9.4 g/m³, or 0.38 g in 40 L. If all the air reached plate temperature, up to 0.08 g (about 84 mg) would have to come out as liquid. In practice only the air near the plate gets that cold, so treat it as an upper bound for the first cool-down.

Here's how the assembly room changes the picture:

Same pack, same coolant. A humid summer shift on the assembly line can put several times more liquid water on the cold plate than a dry one. Compare that with the tens of milligrams measured in the Ingolstadt simulator, which had a much smaller free volume of about 1.5 L, and you can see why the starting air matters. Closing packs in a dry, controlled space is one of the cheapest fixes you have.

Cold plates, coolant lines and the dew point margin

For routes 2 and 3, the number to manage is the dew point margin: surface temperature minus the dew point of the air touching it. Positive margin, dry surface. Zero or below, water forms.

The table below shows how the safe floor for any cold surface moves with the climate the system lives in. We've used a 2 °C margin as a working example. It covers sensor uncertainty, which comes to roughly half a degree at typical readings (we show that math in the enclosure guide), with some room to react.

Read it this way: a coolant setpoint that's comfortably dry in a temperate spring can sit several degrees below the dew point in a humid summer. If your thermal management can take a dew point signal, the coolant floor can follow it. If it can't, add a material inside the enclosure that absorbs moisture first and blocks condensation, plus a sensor that tells you when the margin closes.

Instrumenting a battery pack or BESS cabinet

Our DEW condensation sensors read air temperature, humidity and surface temperature at the same time, and condensation is calculated directly from those three readings. Electrical cabinets, energy storage and automotive are the uses we designed them for. Each one has a surface temperature probe plus a temperature and humidity probe behind a waterproof breathable membrane, is rated IP65, and mounts with double-sided tape.

Where you put them matters more than how many you buy:

Three practical notes:

• Plan the data path. Each sensor stores 260,000 records at a settable interval and reports to a phone or through a gateway. As a general rule of radio engineering, a closed metal housing can weaken wireless signals, so decide up front how data gets out of yours. Tell us the housing and we'll confirm the gateway position for your test. For sizing, 260,000 records at 1-minute sampling is 260,000 / 1,440 = about 180 days of data.
• Check the temperature range. Our DEW sensors are rated from -10 °C. If your pack or cabinet sees colder ambient, tell us the site and we'll confirm options.
• Alert on margin, not RH. Alarms go out by SMS, voice call or email through our HaasLink platform, and the HaaSMind Condensation Analysis Platform analyzes any time window you pick (see our HaaSMind guide).

Prevention: absorb first, then prove it

Our anti-condensation packs are designed for enclosed spaces, with new-energy batteries, cameras and radar, automotive headlamps and 5G base stations among the applications. They use reversible absorption and desorption materials that absorb moisture first and block condensation. They're easy to install, and we co-develop each application with your engineering team. We've been working on protecting new-energy vehicle electronic equipment since 2017 and supplying anti-condensation technology for electric vehicles since 2022.

Then we prove it under your conditions, not generic ones. Our climate room is about 20 m² with a temperature range of -20 to 60 °C. Above 30 °C it can hold 30 to 90% RH, and at 0 °C it can hold 50 to 90% RH, with refrigeration, heating, humidification and dehumidification on a programmed schedule. That lets us simulate a real duty cycle: a hot, humid soak, then a cold phase with the coolant running, repeated over many cycles.

A useful validation compares a baseline pack with a pack fitted with anti-condensation material, through the same cycles. Judge it on three numbers per cycle at the cold plate:

• Minimum dew point margin. How close the surface came to wetting.
• Minutes below zero margin. How long water was actually forming.
• The trend across cycles. A margin that shrinks cycle after cycle means water is accumulating somewhere.

Ready to size packs for a specific enclosure? Start with our Anti-Condensation Fiber Pack, or see how the shipping side of the problem plays out in our container sweat testing guide.

A checklist for battery and BESS teams

1. Close packs in dry air, and record temperature and RH at closing, just as you'd record torque values.
2. Find the coldest surface, usually the cold plate near the coolant inlet.
3. Tie the coolant floor to dew point, not to a number that only works in one season.
4. Don't count on the vent for vapor.
5. Measure surface and air together, and alert on margin.
6. Add a material that absorbs first, and validate it against a baseline under your real duty cycle.

Frequently asked questions

Why does a sealed battery pack still get condensation?

Because the water is already inside, or it gets in as vapor. Air sealed in at assembly carries moisture, cooling pulls outside air in through any opening, and vapor diffuses toward a cold plate even when pressures are equal. In the Ingolstadt study, condensation formed in sealed, open and vented designs alike.

Do pressure compensation vents stop moisture?

They stop liquid water, dust and particles while letting the housing equalize pressure. They have little effect on water vapor. In the study's tests, a vent reduced moisture transport compared with an open housing, but condensation still formed.

How do I calculate the dew point inside a battery enclosure?

Measure air temperature and RH, then use the Magnus formula with the WMO constants: γ = ln(RH/100) + 17.62 × T / (243.12 + T), and dew point = 243.12 × γ / (17.62 - γ). At 25 °C and 50% RH, that's 13.9 °C. Any surface colder than that will collect water.

Where should condensation sensors go in a battery pack?

Put the surface probe on the coldest surface, usually the cold plate near the coolant inlet, and sample the air near the BMS and high-voltage connectors. Add positions under the lid and just inside the vent if you want to see where moisture comes from.

Can anti-condensation packs be used in battery enclosures?

Yes. New-energy batteries are one of the applications our anti-condensation packs are designed for. They absorb moisture first in the enclosed space and block condensation. Tell us the pack design, cooling method and climate, and we'll suggest a pack and a validation plan.

Sources: Kim, Steger, Kotak, Knudsen, Girgsdies and Schweiger, Water Condensation in Traction Battery Systems, Energies 12(6), 1171 (2019); World Meteorological Organization, Calibration of humidity instruments, Part 1 (citing WMO-No. 8, Annex 4.B). Worked examples are our own calculations.