The Deck Stain That Turned White Overnight

Why this matters

A deck stained on a hot, dry afternoon looks like the easiest weather call of the year, and that is why this failure catches good crews. Nothing went wrong while anybody was watching. The finish failed after the crew had gone home, under a clear sky, with no rain anywhere in the forecast, and the shop had to rebuild what happened from records it had not meant to keep. The case is worth following for two reasons: the cause was printed on the hourly forecast available to the crew that morning, and the reconstruction shows that the reading a careful crew would have taken at the last stroke would have passed.

What the customer saw, and what the job record held

A two-person crew applied a water-based semi-transparent stain to a cedar deck on a late-August afternoon, about 3:00 to 6:30 p.m. The deck had been cleaned and brightened two days earlier and had dried in hot weather since.

At 7:30 the next morning the homeowner sent a photo: the open part of the deck had a milky white haze over it, and the finish was tacky to bare feet. Three days later it was still hazy and still slightly tacky. The strip of deck along the house wall, under the roof overhang, and the patch under the patio table looked normal.

The office pulled the job record. It held:

• Clock-in at 2:40 p.m. and clock-out at 6:55 p.m. for both techs.
• Three photos, the last stamped 6:41 p.m., showing the finished deck in the shade of the house.
• The work order naming the stain line, but not the lot.
• No air temperature, no humidity, no dew point and no surface reading at any point.

The crew remembered it as a hot, perfect stain day, which it was, at 3 p.m.

Rebuilding the evening

The lead found the stain cans in the truck and wrote the lot number onto the job. The product's data sheet set an application range for air and surface temperature, said not to apply when rain or heavy dew was expected before the film dried, and gave a dry time at moderate temperature and humidity that it said would run longer in cool, humid air. Every one of those limits is about the hours after the last stroke, and the job record held nothing about them.

For the hours themselves he pulled the history from the nearest public weather station, a little over two miles away on similar open ground, and laid it against the photo and clock times. The dew point column is computed from the station's air temperature and humidity.

 Rebuilt from station history, job photos, clock times
 Time    Air(F)  RH(%)  Dew pt(F)  On the job
 3:00p   86      45     62         crew starts, deck in sun
 6:00p   81      52     62
 6:30p   -       -      -          last stroke
 6:41p   -       -      -          last photo, deck in shade
 8:00p   75      66     63
 8:05p   -       -      -          sunset
 10:00p  69      82     63
 2:00a   64      96     63         customer: dew on the car

Two things in that table are the whole case.

The dew point never moved. It sat at 62 to 63 F from mid-afternoon to the small hours. The air was drying out and cooling, and the moisture in it stayed put. At 3 p.m. the margin between air and dew point was 24 F, which is why the afternoon felt dry. By 10 p.m. it was 6 F, and by 2 a.m. it was 1 F.

The deck is not the air. A station reports air temperature at about head height in a ventilated shelter. Horizontal deck boards under a clear, calm night sky radiate heat straight up and commonly run several degrees colder than the air around them. A 6 F margin in the air at 10 p.m. can be zero at the surface of an open board, which means condensation was forming on the stain film somewhere between about 9 p.m. and midnight, roughly two and a half to five and a half hours after the last stroke.

A water-based stain at that point has lost much of its water but has not finished forming a film. Water landing on it from outside slows the drying, and the surfactants and other water-soluble parts of the coating can be drawn to the surface, which reads as a white or milky haze and can leave the film soft. That is a widely seen failure in water-based exterior coatings dewed or rained on before they had their dry time, and it fit what the homeowner photographed.

The tell that confirmed it

A reconstruction from a station two miles away is a hypothesis, not proof. The deck itself supplied the confirmation, because the haze had a shape.

• Under the roof overhang, the finish was normal. Boards under a roof cannot radiate to the open sky; they radiate to the underside of the overhang, which is near air temperature. They stay warmer, and they did not dew.
• Under the patio table, the same. The tabletop shaded those boards from the sky the same way. The outline of the table was visible on the deck as a rectangle of normal finish.
• The open boards, and the open steps worst of all, carried the haze. The steps stood farthest from the house and fully open to the sky.

A problem with the product, the cleaning, or moisture coming up out of the wood would not draw a rectangle where a table stood. Radiational cooling to the sky does exactly that.

The lead then measured the mechanism directly. The following week, on another clear, calm evening, he went back with a hygrometer and an infrared thermometer at 10 p.m. The air read 70 F with a dew point of 64 F. An open board on the deck read 63 F, already under the dew point. A board under the overhang read 68 F. That is a different night, and it is one set of readings, but it showed the same 5 F gap between open and sheltered boards that the haze pattern implied.

What it was not

Two other causes produce a white or soft deck finish, and both were checked before the strip.

Moisture left in the wood from cleaning. Stain over wet wood can fail to penetrate and can haze. But the deck had dried two days in hot weather, and when the lead read the damaged boards and the protected boards with a pin moisture meter before stripping, both read about the same and inside the range the stain sheet asked for. Wet wood would have hazed under the overhang too.

Over-application. Stain puddled thick in a few spots can stay tacky for days. The haze here ran evenly across thin areas as well as heavy ones, and the thick spots under the overhang had cured. Over-application would not stop at the roofline.

Strip and recoat

There was no repairing the film in place, since a coat over a soft, hazed film inherits the failure.

1. Strip. The crew used a deck stain stripper of the type the stain maker recommends for its own product. Most are strongly alkaline, so the crew wore chemical splash goggles, chemical-resistant gloves and long sleeves, pre-wet and covered the plantings at the deck edge, and set out the rinse so the runoff went to the lawn rather than the driveway and street drain, since wash water from stripping is not the storm drain's to take. See related: Wash Water and the Storm Drain Rule.
2. Neutralize and brighten. Stripper leaves the wood dark and alkaline, and the follow-up brightener is typically acid based. The two were applied separately with a full rinse between, never mixed in one sprayer, and with the same eye and skin protection.
3. Dry. Two days of dry weather, then moisture meter readings across open and sheltered boards, all inside the stain sheet's range before any stain was opened.
4. Recoat inside a window that ends before dark. The recoat started at 10:00 a.m. with a full set of readings logged at the coldest board, the north end steps, and finished by 1:45 p.m. That left the film more than six hours of afternoon before sunset and the evening's dew.

The original application took about 7 crew-hours. Stripping, brightening and rinsing took about 8, and the recoat about 7, so the redo ran 15 crew-hours, roughly 2.1 times the original job, plus two days of the customer's deck being out of service.

The record the shop keeps now

The fix the shop adopted is not a new rule about stain. It is a record that would have made this decision visible before the cans were opened. Every exterior coating job now carries a conditions log: air temperature, humidity, dew point and surface temperature at the start, at midday and before the last application, with the product lot and the tech's initials, under The Coating and Sealant Conditions Log SOP.

On this job the log would have caught it at the planning line, not at the readings. Run the 6:00 p.m. station figures through the last-application check: air 81 F, dew point 62 F, and a deck surface in the shade of the house that would have read somewhere in the high 70s. That is a spread of roughly 15 F or more, far past the usual 5 F margin, and the line would have been marked PASS. A crew that took the readings and stopped there would have produced a clean log for a failed deck. The readings describe the surface at the moment; only the forecast describes the surface at 10 p.m., which is when this finish failed. That is why the log's planning line, the "last application by" time worked back from the forecast, comes before the readings and not after them. The rule the lead added for horizontal surfaces is: when the hourly shows the overnight air reaching within 5 F of the dew point, the last coat goes on early enough to have the sheet's full dry time before sunset. On this August day the hourly showed exactly that, and with the sheet's dry time at moderate conditions rounded up to 4 hours, the last stroke would have been due by about 4:00 p.m., not 6:30. The crew would have finished the open deck first and left the steps for the next morning.

What changes that answer is the sky and the air. A cloudy or breezy night holds deck boards closer to air temperature, so the same 6 F air margin may never close at the surface. A dry climate where evening dew points sit 20 F or more under the overnight low makes this failure rare. A deck fully under a roof barely needs the rule at all. The log is how the crew finds out which night it is, instead of finding out from a photo at 7:30 in the morning.

References

• Stain manufacturer's product data sheet and stripper and brightener directions for the system in use.
• See related: What Dew Point Predicts
• See related: The Surface Five Degrees Above Dew Point Rule for Coatings
• See related: Radiational Frost: Why a Surface Freezes Above 32
• See related: The Coating and Sealant Conditions Log SOP