Introduction and Executive Summary
This case study supplements a prior report (“Solar-Driven Waves of Water Vapor Within Exterior Wall Cavities”) published in the July 2014 issue of Interface. That article described our team’s large-scale use of humidity/temperature loggers to measure ambient moisture load data within occupiable apartments and exterior wall cavities at multiple two-story buildings at a residential complex near San Francisco. Specifically, this report analyzes the daily spikes of water vapor intrusion into these walls and buildings that occurred when two rain-wetted claddings (traditional three-coat stucco and fiber-cement lap siding) were heated by the sun.
As previously reported, during our 162-day testing period, these dataloggers recorded (every 10 minutes, 144 times per day) ambient relative humidity (RH) and temperature conditions at 24 ground-floor apartments. Concurrent measurements from several exterior loggers supplemented these interior data – in total, more than 4.8 million temperature and RH data points were collected from these various sensors.
This updated report addresses specific datasets recorded (via Loggers 1, 2 and 3) at each of these apartments from the three adjacent spaces (at and adjacent to bedroom closets) identified at Photos 1 and 2.
In summary, this article:
Introduction: Parallel vs. perpendicular walls – key design/construction differences
It is important to note in Photos 1 and 2 that at all exterior walls running parallel to the typical floor/ceiling joists seen in Photo 3 the wood stud framing for the lower apartment terminates at tripled plates at the top of the plenum cavity (as opposed to conventional platform framing where the lower studs would terminate at doubled plates at the bottom of the plenum space).
In contrast, the wood-blocked construction of the opposing perpendicular walls (per Photo 3) is typical for Western platform framing; however, we can see in Photos 3 and 4 that the gypsum wallboard interior envelope provided at the lower unit was not extended up into the plenum space. For example, the 60-minute building paper behind the three-coat stucco is visible in Photo 3.
Photo 1 – Parallel wall: expanded polystyrene foam ‘bellyband’ at rim/plenum transition provides aesthetic relief between lower and upper floors. (Note: building paper is backed only by steel line wires.)
Photo 2 – Parallel wall: at buildings with fiber-cement lap siding, wood board bellybands (not seen in photo) below metal Z-flashings were installed at the rim/plenum transitions.
Photo 3 – Perpendicular wall (stucco): Plenum space above drop ceiling terminates at the building paper. (Note: gypsum wallboard interior envelope was not extended up from the bedroom wall.)
Photo 4 – Perpendicular wall (siding): Plenum space above drop ceiling terminates at inside face of the exterior gypsum sheathing behind the wood board bellyband and the fiber-cement siding.
Photo 5 – Perpendicular wall: Typical EPS foam bellyband at rim/plenum transition provides aesthetic relief between lower and upper floors. (At building corners, stucco is installed to OSB shear panels).
Photo 6 – Perpendicular wall (stucco): No interior envelope was provided at the rim/plenum transition.
Photo 7 – Typical closet ceiling and interior envelope at parallel wall. At all apartments, the drop ceilings are constructed from gypsum board panels hung from steel resilient channels.
A key goal of our testing program was to evaluate potential sources of the excess water vapor and related condensation damage (e.g., Photos 6 and 7) observed at the rim/plenum transitions.
Introduction – the Humidity Ratio
Concurrent measurements of temperature and RH can be converted into the “humidity ratio” – the ratio or percentage of the weight of ambient water vapor molecules per pound of dry air. The software for our loggers was programed to calculate a grains-per-pound (GPP) ratio.
In a manner comparable to a construction professional’s forensic use of pin-style moisture meters to take multiple “wet” vs. “less wet” vs. “dry” readings (i.e., percentages) at wood studs and sheathing throughout a project, our loggers produced 2.4 million humidity ratio (“ambient moisture content”) readings over the 162-day testing period.
We then used Microsoft Excel spreadsheets to process and graph selected portions of this data to evaluate questions of interest. For example, consider Graph 1, which displays ambient moisture content readings measured concurrently (144 times per day, starting at 12:00 a.m.) over a 17-day period at Logger 3 (within an exterior wall at one of the bedrooms) and Logger 4 (positioned between the kitchen and bathroom). As previously reported, we see at Graph 1 that:
a) The tenants had moved out of the apartment by Day 8, as documented (per the blue line) by the dramatic decline of ambient moisture measured by Logger 4; and
b) More importantly, as indicated by the red line, daily cycles of increased/decreased levels of ambient moisture occurred within the exterior wall (Logger 3) in direct correspondence to solar heating and nighttime cooling.
Graph 1 – Apartment ‘B’: Blue line tracks ambient moisture ratio (GPP) at interior Logger 4, while red line records exterior wall cavity (Logger 3)—144 readings per day for 17 days (starting at 12:00 a.m.).
These conditions certainly are not unique to our testing; at all buildings, some degree of inward solar-driven migration of water vapor (due to increased vapor pressures) always will occur when an exterior wall with the ability to absorb moisture becomes heated by the sun.
“Sun-driven moisture is a phenomenon that occurs when walls are wetted and then heated by solar radiation. Upon solar heating, a large vapour pressure difference may occur between the exterior and interior leading to the inward diffusion of moisture.”
“It is clear that any wet material …that is heated by the sun will generate large inward vapour drives.”
This phenomenon of solar-driven diffusion is further documented by Graph 2, which depicts a remarkably close correspondence between the daily high-low swings in exterior temperature (the dashed black line) measured hundreds of feet away from this particular building and the daily high-low spikes of ambient water vapor (the red line) recorded within an exterior wall cavity at Apartment E. Clearly, a direct relationship exists between daily swings in exterior temperature and the cyclic peaks and valleys of the solar-heated moisture within the wall.
Graph 2 – Dashed black line is exterior temperature (F°) at the site, while red line tracks ambient moisture load (GPP) within exterior wall cavity at apartment E—144 readings per day for 67 days.
The excellent long-term performance across North America of millions of exterior walls clad with stucco or fiber-cement lap siding is evidence that this ubiquitous process of solar-driven diffusion into wall assemblies should not, in and of itself, be considered alarming or problematic.
In general, building envelope professionals know from long experience that building envelopes can safely accommodate such inward and outward drives of water vapor. But what are the key(s) to controlling these spikes in moisture? Our testing and analysis explored this question by quantifying and comparing the sharply increased humidity ratios that occurred when our two rain-wetted cladding systems became heated by the sun.
Introduction – Drought conditions in California
Due to California’s extended drought, weather conditions generally were dry during much of our 162-day testing period; however, as shown at Figure 1, a relatively cool 11-day stretch of intermittent rainfall in March/April 2014 that was followed by five days of hot sun facilitated a close comparison of varying responses of the wetted stucco and fiber-cement lap siding systems to solar heating.
Figure 1 – National Weather Service data from March/April 2014: 11-day period of intermittent rainfall was followed by five days of hot sun.
Note that during the first 11 days of this analysis, a total of 2.23 inches of rain fell at the site, and the average high temperature was only 61°F. In contrast, there was no rainfall during the final five days of sunny weather, and the average high temperature was 80°F.
As documented with the following graphs, these hot, sunny days generated sharply increased humidity ratios within the tested exterior wall cavities. Then, due to the absence of an interior envelope at the perpendicular walls (see Photos 3 and 4), these spikes of moisture readily pushed into the plenum spaces above the drop ceilings. Further, the data from some – but not all – of the apartments revealed that these solar-driven moisture waves also migrated into the occupiable rooms below the drop ceilings.
Graphs 3 and 4 – Stucco (south elevation – parallel wall) at occupied Unit F
As seen at Graph 3 for Unit F, the hot, sunny days generated strong daily increases in ambient moisture at both Logger 2 (plenum space = green line) and Logger 3 (exterior wall cavity = red line). For each logger, we then compared the average (mean) humidity ratio for the first 11 days of testing with the overall average humidity ratio for the final five days.
We found that Logger 2 had recorded an average 40.0% increase within the plenum space, and Logger 3 had measured a 43.3% average increase within the exterior wall cavity.
In contrast, data from both Logger 1 (closet = purple line = 7.3%) and kitchen Logger 4 (3.3% – not graphed) suggested that solar heating of the wetted stucco-clad walls below the rim/plenum transition had not greatly impacted humidity levels within the occupied rooms.
Graph 3 – Unit F (stucco – parallel wall): Purple line records GPP ratio in closet; green line records plenum; and red line records exterior wall – 144 readings per day for 16 days (starting at 12:00 a.m.).
To better understand the substantial differences in vapor-retarding performance between the rim/plenum transition (behind the bellyband) and the lower wall below, which share the same solar-heated stud bay (see Photo 1), we considered Graph 4, which focuses only on the data from hot, sunny Days 14 and 15 of this testing. Upon review, we determined:
a. The closely similar, cyclic shapes of the red line (exterior wall – Logger 3) and green line (plenum space – Logger 2) again prove that the daily cycles of increased/decreased moisture within the ceiling cavity correspond to solar heating and nighttime cooling of the rain-wetted stucco walls.
b. However, because the greatly dissimilar shape of the purple line (bedroom closet – Logger 1) is not consistent with direct solar heating, the interior envelope – typical painted gypsum wallboard – at this apartment has, to a great degree, resisted and tempered the daily pushes of solar-driven moisture.
c. Further, because the daily rise of the heating phases at the green line occurs approximately one to two hours later than at the red line (even though these two data loggers are only inches apart) the unpainted gypsum wallboard (see Photo 7) above the drop ceiling also served to satisfactorily retard the daily drives of solar-driven moisture…
d. …Therefore, the most likely source (and route) of these daily increases in ambient moisture above the ceiling was the opposite perpendicular wall (Photo 3), where the absence of an interior envelope had created the path of least resistance.
Graph 4 – Unit F (stucco – parallel wall): Purple line records GPP ratio in closet; green line records plenum; and red line records exterior wall (144 readings each for days 14 and 15, starting at 12:00 a.m.). Every vertical line on this graph corresponds to one hour.
To evaluate this seemingly obvious finding, we then considered comparable data from another parallel wall – at an apartment clad with fiber-cement siding.
Graphs 5 and 6 – Siding (north elevation at parallel wall) at occupied Unit G
At Graph 5 for Unit G, clad with fiber-cement lap siding, we see that the hot, sunny weather beginning on Day 12 (again reference the weather matrix at Figure 1) similarly produced strong increases in average moisture content at Logger 2 (plenum = green line = 24.2%) and logger 3 (exterior wall cavity = red line = 31.8%).
And again, the data from Logger 1 (closet = purple line = 6.1% increase) and kitchen logger 4 (6.7% – not graphed) indicated that solar heating (from the same stud bay) of the wetted siding below the rim had not greatly impacted humidity levels within the occupied rooms. Then, to better confirm this analysis, we reviewed data (per Graph 6) from the hottest Days 14 and 15 of this testing period.
Graph 5 – Unit G (siding – parallel wall): Purple line records GPP ratio in closet; green line records plenum; and red line records exterior wall—144 readings per day for 16 days (starting at 12:00 a.m.).
Graph 6 – Unit G (siding – parallel wall): Purple line records GPP ratio in closet; green line records plenum; and red line records exterior wall (144 readings each for days 14 and 15, starting at 12:00 a.m.) Every vertical line on this graph corresponds to one hour.
Upon review of both graphs, we concluded:
a. The generally similar shapes of the red line (Logger 3) and green line (Logger 2) again demonstrate that the daily cycles of increased/decreased moisture within the ceiling cavity corresponded to solar heating and nighttime cooling of the wetted siding assembly.
b. Again, because the generally dissimilar shape of the purple line (Logger 1) is inconsistent with direct solar heating of the occupied spaces, the interior envelope (painted wallboard) had resisted and tempered the spikes of solar-driven moisture.
c. Similarly, because the heating phases at the green line occurred three to five hours later than the red line, the unpainted gypsum wallboard (Photo 7) between Loggers 2 and 3 again had retarded these waves of water vapor…
d. …Thereby demonstrating again that the source (and route) of the increased moisture above the ceiling was solar-driven migration through the perpendicular wall (Photo 4) where the absence of an interior envelope had created the path of least resistance.
Note also that comparing Graphs 3 and 4 (stucco: south-facing parallel wall at Unit F) with Graphs 5 and 6 (lap siding: north-facing parallel wall at Unit G) reveals highly different shapes for the moisture cycles being recorded at the two red lines. At Graphs 3 and 4, we see high, steep, and narrow spikes of increased vapor on hot, sunny days, while the corresponding red line cycles at Graphs 5 and 6 tend to be flatter and wider.
• Upon review: the primary factor in determining the cyclic shapes of these red lines appears to be solar exposure; wetted walls that receive greatest exposure to direct sunlight during the hottest portion of the day simply will, in general, encounter higher, steeper spikes of solar-driven moisture.
We then evaluated data collected from loggers installed at typical perpendicular walls clad with stucco (Graph 7) and fiber-cement lap siding (Graphs 8 and 9).
Graph 7 – Stucco (south elevation – perpendicular wall) at occupied Unit J
Graph 7, at stucco-clad Unit J, demonstrates that at all four loggers recorded large increases in average moisture content during the final five days of hot weather: Logger 1 (bedroom closet = purple line = 19.5%), Logger 2 (plenum space = green line = 35.2%), Logger 3 (exterior wall cavity = red line = 58.6%), and also at kitchen/bathroom Logger 4 (12.8% – not graphed).
• Further, close correspondence between the shapes and timing of the purple and green lines indicated that solar heating of the rain-wetted stucco had concurrently driven water vapor directly into both the occupied bedroom and the plenum space above.
Graph 8 – Siding (east elevation – perpendicular wall) at occupied Unit K
Similarly, Graph 8 for Unit K (clad with fiber-cement), reveals strong increases in average moisture content at all four loggers: Logger 1 (bedroom closet = purple line = 23.7%), Logger 2 (plenum = green line = 30.4%) and Logger 3 (exterior wall cavity = red line = 55.1%), and also kitchen/bathroom Logger 4 (16.4% – not graphed).
• And again, the remarkably close correspondence between the shapes and timing of the purple and green lines during the final five days demonstrated that solar heating of the rain-wetted wall had pushed moisture directly into both the occupied bedroom and the plenum space.
Graph 7 – Unit J (stucco – perpendicular wall): Purple line records GPP ratio in closet, green line is plenum space, and red line is exterior wall—144 readings per day for 16 days (starting at 12:00 a.m.).
Graph 8 – Unit K (siding – perpendicular wall): Purple line records GPP ratio in closet, green line is plenum space, and red line is exterior wall—144 readings per day for 16 days (starting at 12:00 a.m.).
Graph 9 – Siding (south elevation – perpendicular wall) at occupied Unit H
Finally, Graph 9 for the south-facing perpendicular wall at Unit H (also clad with fiber-cement) again showed large increases in average moisture content at all four loggers: Logger 1 (bedroom closet = purple line = 31.2%), Logger 2 (plenum = green line = 39.4%), Logger 3 (exterior wall cavity = red line = 49.6%), and also kitchen/bathroom Logger 4 (25.4% – not graphed).
And yet again, close correspondence between the shapes and timing of the purple and green lines during the final five days demonstrated that solar heating of the rain-wetted perpendicular walls (all of which lacked an interior envelope at rim/plenum transitions, per Photos 3 and 4) had driven moisture directly into the occupied bedroom and the plenum cavity above.
Graph 9 – Unit H (siding – perpendicular wall): Purple line records GPP ratio in closet, green line is plenum space, and red line is exterior wall—144 readings per day for 16 days (starting at 12:00 a.m.).
Interestingly, Graph 9 also demonstrates that these solar-induced waves of water vapor had commenced on Day 9 (even though the high temperature was only 61°F, per the Figure 1 weather matrix) and continued through Day 11 (cloudy all day, with minor precipitation and a high temperature of only 59°F).
After review, we concluded that these seemingly “early” spikes of moisture at this south-facing perpendicular wall at Unit H represented the expected effects of direct heating of atypically high amounts of trapped moisture within the wall assembly–very likely due to rainwater leakage behind the wood bellyband boards.
Per Photos 8 and 9, destructive investigation confirmed areas of rainwater infiltration adjacent to typical locations being tested with Loggers 1, 2, and 3 at these perpendicular walls, below wood bellyband boards below poorly installed Z-flashings. (Note the “burned” building paper.)
Photo 8 – At wood bellyband boards, rainwater infiltration at non-sealed joints and transitions at metal Z flashings has led to increased levels of solar-driven moisture through some exterior walls (Graph 9).
Photo 9 – Rainwater infiltration at non-sealed joints and transitions at metal Z-flashings has led to increased levels of solar-driven moisture migration through the exterior walls (see Graph 9).
Virtually all exterior walls experience some degree of inward vapor drive when heated by the sun; however, after a period of rainy weather, even greater quantities of water can be driven into the wall cavities from wetted claddings. While this ubiquitous process is not problematic, in and of itself, providing a continuous interior envelope (or solid wood blocking) at these walls will keep these daily cycles of solar-heated moisture safely contained within the wall cavities. Our testing suggests that in the moderate climate zones near San Francisco, an interior envelope constructed only of unpainted gypsum wallboard will satisfactorily retard these inward pushes of moisture – even on the sunny days that immediately follow a rainy period.
Conversely, our data clearly demonstrate that the general failure to provide any form of interior envelope at certain rim/plenum transitions at this residential complex created pathways for large amounts of water vapor to be pushed directly into first-floor building interiors, including occupiable spaces below drop ceilings. The size and extent of such solar-driven moisture (vapor) waves of course were exacerbated wherever rainwater (liquid) already had leaked into the wall systems due to construction defects at the exterior envelopes.
This suggests that whenever forensic investigators find condensation damage and fungal growth at interior walls within residential spaces, they should closely consider the possibility that primary origin(s) of this problematic moisture might be solar heating of trapped rainwater at potentially-distant defects in the exterior walls.
Further, this investigation and similar recent datalogger testing by our firm indicate, as evidenced by Graph 1, that diffusion (due to vapor concentration differentials) into exterior walls of occupant-produced ‘lifestyle’ moisture generally is far less likely to cause damage than is convection (via air pressure differentials) of humid interior air into exterior walls due to unsealed gaps and voids. As noted within ASTM International’s Manual 18, Moisture Control in Buildings: The Key Factor in Mold Prevention: “Requiring careful moisture management by the occupant is no substitute or excuse for inadequate moisture-resistant design.”
In cases where a wide gamut of opinions and theories are being offered by the investigating consultants, the data collected from well-positioned loggers can be used to better allocate to the parties whose deficient work caused or allowed this excess vapor moisture intrusion their fair shares of financial responsibility for interior remediation. In particular, our firm’s data analyses center on short- and long-term changes in ambient moisture content (‘humidity ratio’) readings because most parties involved in the typical construction defects litigation process intuitively understand the probative values of “wet” vs. “less wet” vs. “dry” moisture content readings.
We recommend that most data-centric evaluations of interior humidity levels and/or building envelope performance similarly should be focused on tracking changes in actual moisture content over time, as opposed to changes in relative humidity (which, of course, is a ‘relative’ value that can change greatly even when the actual moisture content ratio remains the same).