023 password removed

7/28/2019 023 Password Removed

1/8

Condensation Risk Assessment

Arnold Janssens, Ph.D. Hugo Hens, Ph.D.Member ASHRAE

ABSTRACTThe methodology afrisk analysis and assessment is reviewedand applied to study the reliability ofcondensation control measuresin lightweight building envelopes. It is generally recognized that airtight construction is an essentialpart ofcondensation control.Nowadays, different air barrier systems are developed and documented to prevent air leakage and moisture accumulation in theenvelope. Butdoes this mean that the condensat ion risk is sufficiently minimizedand that theprotective sys tem is reliable? Considering the high occurrence ofhuman error in the building process, the possibility ofair barrier defects during the service life ofa building envelope may be high.To define the reliability of he condensation control system, the consequences ofair barr ier failure are quantified using a twodimensional numerical control volume model or the calculation ofcombined heat, air, and vapor transfer in multilayered buildingenvelope parts. A set offailure modes and design calculation conditions is definedfor an exemplary woodframe insulated roof,and afailure effect analysis is peiformed in order to predict the condensation risk as a result of air barrier defects. The effectiveness of redundant design measures to improve the reliability of the condensation control system is studied.

RISK ANALYSIS METHODOLOGYRisk analysis, as applied in this work, is a method to study

the effects of uncertain events or unintentional actions on theserviceability of a design in order to prevent or reduce thenegative outcome of these effects (Ang and Tang 1984; Ansellet aI. 1991). In this definition, risk not only refers to the prob-ability offailure but also incorporates a notion of the magni-tude of the failure. The techniques to identify and reduce riskswere developed in the 1960s in the nuclear and chemicalindustries (KIetz 1992). The analysis provides a systematicand structured framework with three main stages:I. Risk identification2. Risk assessment3. Decision making

The first step to be addressed at the design stage of aproject is the recognition and identification of the range ofrisks to which a system is subject. The ultimate aim of riskidentification is to link the possible problems that a systemmight experience directly to external factors and eventsinvolving component failures and human errors. The second

stage implies the use of probabilistic calculation and assess-ment methods in order to predict the system performance andthe probability of failure as a result ofthe unintentional events.In industrial risk analysis, risk is generally determined as theproductof he frequency and the size of the consequence of heevent (Ansell et aI. 1991; Kletz 1992). Risk is accepted whenevents that happen often have no or low consequences or whenevents involving serious problems are rare.

When the predicted risk cannot be justified, decisionsshould be taken to eliminate or minimize the risk, either bypreventing the event or by reducing the consequences in caseof occurrence. These decisions often imply redesign of thesystem by the introduction of protective measures. There aregenerally three types of strategies, with decreasing order ofeffectiveness: inherent, engineered, and procedural safetymeasures. The reliability of protective systems is oftenimproved through some duplication of components. This iscalled "redundancy" if the protective components are the sameor "diversity" if they are different. In a nonredundant system,the failure of the protective component leads inevitably to thefailure of the entire system.

Arnold Janssens is a researcher and Hugo Hens is a professor at the Laboratory of Building Physics, K.U. Leuven, Belgium.

Thermal Envelopes VIIiMoisture Assessments-Principles 199


2/8

Consider the example of fire risk prevention in buildings.A building is constructed with inflammable or fire-retardingmaterials whenever possible and affordable (inherent safety).Recognizing that fire risk is rarely eliminated, a sprinklerinstallation may be designed and installed in the building(engineered safety). Finally, to prevent casualties, a fire alarmand escape may be added (procedural safety). The reliabilityof the fire protection may be increased by installing two ormore fire alarms (redundancy) or adding both sprinkler andalarm systems (diversity).

There are similarities between the design of a hazardprotection and a building envelope system. The building envelope may be regarded as a multicomponent protective systemto eliminate or minimize the discomfort of the environment.The application of the redundancy concept is a commonelement of building envelope design. For instance, the reliability of rain control in the building envelope is oftenimproved by adding redundant components, such as drainedcavities, multilayered membranes, etc. Because of these similarities, the methodology of lisk analysis may well be appliedto the performance analysis of building envelope systems.

In some recent publications on moisture control, theconcepts of risk analysis are implicitly present. It is recommended that designers incorporate moisture control measuresto ensure the performance of the envelope system in the casethat something goes wrong during construction or operation,for instance, by accidental wetting or by human error. Thissubsystem is called "the second line of defense" (Lstiburekand Bamberg 1996).

UNCERTAINTIES IN CONDENSATIONCONTROLCondensation Control Systems

TenWolde and Rose (1996) presented two majorapproaches to moisture control in the building envelope. Thefirst is to design and construct the building envelope for a hightolerance for moisture. The second is to limit the moisture loadon the envelope. Designing the envelope for a high moisturetolerance implies the use of measures to control the migrationof moisture into the construction, to control moisture accumulation in building materials, or to enhance removal of moisturefrom the building assembly (Lstiburek and Carmody 1993).The limitation of the moisture load often involves controlstrategies on the level of building design and operation, e.g.,ventilation, dehumidification, or depressurization. Table Icategorizes potential condensation control strategies for thebuilding envelope. Table 2 lists measures to restrict the moisture load by proper building design and operation. The individual measures may be combined to create an effectivecondensation control system. The problem is to find out whichcombination of condensation control measures is the mostreliable to reduce the condensation risk.

200

TABLE!Condensation Control Strategiesfor the Building Envelope*Strategy Aim Measure

Control of Eliminate air leakage Air barriermoistureaccess Restrict vapor diffusion Vapor retarder

Control of Raise temperature of Insulation outside ofmoisture condensing surface condensing surface

accumulation Allow harmless High moisture capacityaccumulation at condensing surfaceRemoval of Promote drying Vapor permeable layers

moisture Capillary active layersCavity ventilation

Remove condensation DrainageReduction of Limit conslruction Initially dry materialsmoisture load moisture

*after Lstlburek and Carmody (1993) and TenWokle and Rose (1996)

TABLE 2Reduction of Moisture Load byBuilding Design and Operation*Aim Design Operation

Limit vapor in Natural ventilation Mechanical ventilationinterior air system Dehumidification

Control air pres- Airtight floor separations Depressurizationsure differentials (limit thermal stack height)

*afler Lstlburek and Carmody (199]) and TenWolde and Rose (1996)

Uncertainties in Air Barrier PerformanceIt is generally recognized among building researchers that

airtight construction is an essential part of any condensationcontrol system. Without effective control of air leakage, othercondensation control measures such as vapor retarders arecompletely ineffective (ASHRAE 1997). Nowadays, differentair barrier systems are developed and documented, and material and system requirements are presclibed (Di Lenardo et a1.1995). However, in terms of risk analysis, the air barder is anonredundant protective component. I f airtightness requirements are not met in practice, there is a risk of serious condensation problems. This implies that the uncertainties incondensation control are pdmatily related to uncertainties inair barrier performance.

The authors have conducted a survey of recent cases ofserious condensation problems in lightweight roofs (Janssens1998). They found that most problems were caused by a failure of the air barrier due to human errors. ASHRAE Fundamentals (ASHRAE 1997) lists some of the most frequentlyobserved leakage paths in lightweight wall and roof systems.Thermal Envelopes VII/Moisture Assessments-Principles


3/8

They may be divided into four categories: (1) joints betweendifferent sheets or boards of the air barrier, (2) interfaces withother envelope systems, such as the wall-roof and wallwindow interface, (3) junctions around structural penetrations(columns, joists, trusses, etc.), and (4) electrical and plumbingpenetrations. Air barrier defects at these locations may becaused by a combination of design and construction errors.Different parties in the design team and different trades in theconstruction team may contribute to the defects. Consideringthe high occurrence of human elTor in the bui lding process , itis evident that the likelihood of air banier defects is not negligible. Some redundancy in the design of the condensationcontrol system may become necessary.

To decide in which condition changes in design are necessary and to find out which measures are effective in reducingcondensation risk, it is necessary to quantify the effects of ai rleakage on the moisture performance of building envelopesystems.

CONSEQUENCES OF AIR BARRIERDEFECTSAnalytical Study of Condensation Dueto Air Leakage

Let us first consider the simple case of one-dimensionalheat, air, and vapor flow. In steady-state conditions, this problem may be solved analytically (Hens 1996). The solution mayhelp in understanding the conditions when ai r leakage causescondensation problems. It may also give a first idea of thepotential measures to prevent problems. The condensationrate at a material interface in air permeable envelope systemsis calculated according to Equation I:

(I)

wheregc = condensation flow rate (kg/mz/s),

= airflow rate (negative sign in case of exfLltration) (kg!

PiPe

mZ/s),= vapor capacity of air = 6.2xIO-6 kg/kglPa,= intetior vapor pressure (Pa),= exterior vapor pressure (Pa),psatcec) = saturation vapor pressure at the condensation plane(Pa),

Zj = sum of the vapor diffusion resistances betweeninterior and condensation plane,

= sum of the vapor diffusion resistances betweencondensation plane and exterior.

With even moderate air exfiltration rates (negative sign),the denominators with the exponential terms become unityand infinity, respectively. Equation I reduces to Equation 2:

Thermal Envelopes VII/Moisture Assessments-Principles

(2)The condensation flow rate per unit flow of air is inde

pendent of the vapor transfer properties of the envelopesystem. Condensation will occur whenever the indoor vaporpressure exceeds the saturation vapor pressure of the condensation plane or when

(3)where/1p = vapor supply in the building (Pa),Pi = Pc +/lp.

The left-hand side of Equation 3 characterizes the indoorclimate, the right-hand side, the exterior climatic conditions.

Equation 2 shows that when exfiltration through the envelope is not stopped. there is only one way to prevent condensation: by increasing the temperature of the critical intelfaceso as to lower the difference between the indoor vapor pressure and the saturation vapor pressure at the interface. This ispractically done by applying insulated sheathing or by changing the envelope design with the principles of a warm deckconstruction.

Going back to Equation 1, itis possible to define a secondtype of potential protective measure. Suppose that the saturation vapor pressure at the condensation plane is above theexterior vapor pressure an d that the vapor diffusion resistanceto the outside Z / is at least one order of magnitude smallerthan the diffusion resistance to the inside Z/o For a certainrange of air exfiltration rates, Equation I reduces to

(4)This shows that measures to enhance moisture drying to

the outside may be capable of preventing condensation problems, even when a perfect airtightness is not achieved. On theother hand, the use of vapor retarders is ineffective when airleakage is not stopped.Failure Effect Analysis:Insulated Wood Frame Roof

Calculation Model. The authors developed a twodimensional numerical control volume model for the calculation of combined heat, air, and vapor transport in multilayeredbuilding envelope parts (Janssens 1998). The model considersairflow in porous material and laminar flow in cracks and airlayers only. As a consequence, the equations for the conservation of momentum reduce to the linear Darcy and Hagen-Poiseuille flow equations. Airflow is driven by external pressures or thermal stack. The equations for the conservation ofmass, momentum, and energy are iteratively solved. Underrelaxation techniques are applied to improve and acceleratethe convergence of he solution. The heat and vapor flow equa-

201


4/8

tions are integrated using the exponential differential scheme.Condensation flows are calculated iteratively by changingvariables whenever the calculated vapor pressure rises abovethe saturation vapor pressure. Eventually, in every calculationnode, the airflow rates, temperatures, relative humidities, andcondensation flow rates are defined. Thus, the model makes itpossible to assess the effect of air leakage on the thermal andmoisture performance of the building envelope.

Description of Boundary Conditions, EnvelopeSystem, and Failure Modes. A preceding comparisonbetween the results of transient and steady-state calculationsshowed that steady-state calculations are a good check for themoisture performance of lightweight envelope systems withnonhygroscopic layers (Janssens 1998). Unlike moistureaccumulation as a result of vapor diffusion, condensationlinked to air leakage is a fast and short-term phenomenon. Theone-dimensional transient calculation was perronned using aclimate file of Manchester, U.K.. with hourly averaged data.The results showed typical condensation periods of a couple ofdays to a week. A common characteristic of the boundaryconditions during these periods was the negative net radiationto the exterior surrace, in addition to the conditions describedabove. From this analysis. a set of weekly average designboundary conditions was composed for the two-dimensionalsteady-state calculation of condensation problems linked toair leakage. They are listed in Table 3. The air pressure isimposed at the lowest point of the boundary.

TABLE 3Design Boundary Conditions for CondensationRisk Assessment (Weekly Averages)

Vapor Surface Net AirBoundary Temp. Pressure Coefficient Radiation PressureInterior 18'C 950Pa 8 W/(m2K) OW/m' 2 PaExterior -2.5'C 450Pa 19 W/(m'X) -30W/m2 oPa

As an exemplary case, a failure effect analysis isperformed on an insulated wood frame roof that has, frominside to outside, the interior finish, a glasswool insulationlayer, an air cavity, the underlay, and tiles. The interior finishalso includes the air/vapor barrier. It is assumed that the tilelayer, due to an intensive mixing of air around the tiles, affectsthe temperature distribution in the roof only. Its thermal resistance is included in the underlay. Table 4 lists the materialproperties. The length of the roofis 3 m, and the slope is 45.The U-value is 0.22 W/(m2K). The moisture performance ofthe roof design is acceptable according to existing standards(e.g., DIN 4108,1981) . Under the above conditions, a Glasercalculation predicts a condensation amount of 0.01 Llm2 perweek.

The calculation domain is discretized into a rectangulargrid with dimensions of 19 x 75 in the x and y direction,respectively. The first calculation assumes that no defects arepresent in either of the envelope layers. The underlay and

202

LayersInternal lining

FiberglassCavity

Underlay

TABLE 4Material Properties

d A.j}...y Rxm W/m2/K m 2.K/W0.01 0.1 0.10.14 0.035/0.040 4.00.01 0.120.01 0.14 0.07

/1fl KjKym m25.0 le-120.2 4e-9/6e-9

1.5 le-12intemal lining are essentially airtight, and airflow maydevelop in the air permeable fiberglass insulation only, due tothermal stack.

Then a set of failure modes is defined, and all possiblecombinations are included in the envelope geometry (FigureI). Table 5 lists the set of failure modes. One particular modemay consist of two or more defects that are supposed to occurtogether, for instance, air layers on either side of the thermalinsulation or overlaps in the underlay. Some of them are notnecessarily defects but may be inherent in the envelopedesign, for instance, overlaps in the underlay or cavity vents.

In total, 256 combinations (= 28) are calculated and theresulting heat flow, air flow, and condensation flow rates arestatistically processed.

7 5 1

6

8

4 4 3 0 mV82

7 5 3

~ 017 m ~Figure 1 Fai/ure modes.

Thermal Envelopes VWMoisture Assessments-Principles


5/8

TABLESSet of Failure ModesLayer Failure Mode Width (mm) Position from Bottom Simulated Defect

Intemallining 1 1 3.0m Crack-unsealed junction2 1 0.5m Penetration (equivalent 08 mm)3 1 O.Om Crack-unsealed junction

Thermal insulation 4 5 - Air layers5 5 0.0/3.0 m Junction with eaves and ridge6 5 1.212.4 m Joints between boards

Underlay 7 2 0.0/3.0m Junction-vent8 2 1.0/2.0 m Overlap---joint

Moisture Performance. The calculation results arerepresented by means of the cumulated distribution function(CD F). This is a common statistical function to show the variation of a collection of different data. The function defines thefraction of calculation results falling below a certain value.Since the distribution of calculation results characterizes thepotential variation in system properties, the CDF may beconsidered a response property of the system to a se t ofdefects. Figure 2 shows the cumulated distribution function(CDF) of the average condensation flow rate of all 256 calculated cases. In this case, the condensation rate in a roof withoutdefects coincides with the 10% percentile of the CDF (seeTable 6). Thus, the consequences of defects in the envelopelayers are serious and may jeopardize the roof performance.Experiments have shown that condensate starts to run offwhen it amounts to approximately 0.1 Llm2 (Janssens 1998).In more than 70% of the failure comhinations, the weeklycondensation mass is higher. The two curves in the graphrepresent the effect of the thermal radiation from the outsideboundary (net radiation q = 0 vs. q = -3 0 W/m2).

-1----- .0.2 ..................... . ......................._.!,oL -__________L-__________ __________

o o.s 1.5condensation rate (Ifw eeklm2)

Figure 2 CDF of condensation rate (Um2/week).

Thermal Envelopes VllIMoisture Assessments-Principles

Figure 3 gives more information about the conditionsleading to condensation. The condensation flow rate is plottedas a function of the total air exfiItration rate. There is no unamhiguous relationship between both variables. Roughly, threezones may be distinguished. The first one, with no air exfiltration, contains all failure combinations without a continuousairllow path between the boundaries of the construction. Thecondensation amounts are limited but may still be 30 timesmore than in a roof without defects. This is the case when atleast two defects occur in the internal lining.

When air leaks exist in both the internal and externallining, humid indoor air exfiltrates through the envelope andthe condensation amounts are more substantial. The relationship between condensation and exfiltration rate, however,depends strongly on the minimum length of the airllow path inthe construction. The second zone in the graph, with the highest condensation amounts, contains all combinations withdefects in the internal lining and central overlaps in the underlay (failure mode 8). The third zone, with the highest exfiltration rates but clearly lower condensation amounts, containsthe comhinations with defects in the internal lining and at least

5 r-----'.!.---;.--.TI------r-----'l------T"I-----,..,1I .m.!

I I "i ! 1 .... ..+....... .. + . - . ~ .. ..........__ ! ,,1.. ...... 1 ,i i ~ i- - = r . - ~ l : : F ~ ~ ; . ~

o L -_ _ _ _ _ __L____ ~ __L-___ J______12 ." .. ., oThousands

Air exfillration rate (lIhfm)Figure 3 Condensation vs. air ex iltration rate.

203


6/8

TABLE 6Distribution ofRoof Performancesas a Result of Roof DefectsPercentile UAPPARENT Va,ROOF g,

% W/m 2/K ELEAK1Eo m 31h1m kg/m2/week10 0.15 1.01 0.0 om50 0.20 1.09 2.4 0.3890 0.26 1.27 7.3 0.84

the outer overlaps in the underlay (failure mode 7). In this case,the airflow path is shorter, the exfiltrating humid air is cooleddown less effectively, and, consequently, less moisture isdeposited in the cavity.

Thermal Performance, The effects of air leakage on thethermal performance of the roof is represented in two ways.First, the apparent V-factor of the roof is calculated by dividing the total conductive heat loss at the internal boundary bythe roof area and the temperature difference between insideand outside (case with no radiation to the exterior surface).Table 6 lists the statistical distribution of the calculated apparentV-factors. The V-factor of a roofwithout defects coincideswith the 60% percentile of the CDF.

The analysis of the consequences of defects on the energyefficiencyof the roof is incomplete without considering the airleakage rate through the roof and the associated enthalpy loss.For a correct interpretation, these extra heat losses should bestudied on the scale of the building. Let's consider, for example, a dwelling with a vertical section as in Figure 4. The V-factor of the roofUo is 0.22W/m2/K, and the average V-factorof the other envelope parts U j is 0.82 W/m2/K. This way, thedwelling fulfills the Flemish insulation requirements (insulation level K55). Suppose that the ventilation air change rate nis 0.5 h-1; then the energy consumption of the dwelling is Eo:

EO = (2AUo +A j U,+0.34nV)(8,-8,) (5)

A Uo

4.50

4.25 mFigure 4 Dwelling section.

204

In the case with defects, air leakage through the roof mayincrease the air change rate of the building and contribute tothe ventilation heat losses. As a worst estimate, let's add the airleakage rate Va,ROOF to the ventilation air change rate n. Theenergy consumption of the dwelling becomes ELEAK:

(6)

The increase in energy loss in the dwelling due to defectsin the roof is estimated from the ratio ELEAK/Eo. Figure 5shows the CDP of the energy loss increase. The two curvesrepresent the effect of radiation from the outside boundary. Inall calculation cases, the roof failure modes have a negativeeffecton the energy demand of he dwelling. In the worst case,it may be increased with a factor of 1.45.REDUNDANT PROTECTIVEMEASURES INCONDENSATION CONTROLRedundancy

The failure effect analysis clearly demonstrates that evensmall defects in the air barrier or in one of the other layers ofthe envelope system may seriously affect the thermal andmoisture performance of the envelope part. The most problematic for the serviceability of the envelope and the buildingis the risk of condensation problems. This risk cannot be justified, as we have shown, because of the high possibility of airbarrier defects and because of the potentially important effectsof air barrier failure. Therefore, i t is necessary to look for someredundancy in the envelope design, directed toward preventing or minimizing the condensation risk. Performance requirements could be established in order to define whether thecondensation risk of a design solution is acceptable or not. Forinstance, the performance check could exist of a failure effectanalysis with a design set of failure modes and design calculation conditions. The condensation risk of an envelope design

0 .

i0. 6 ...........j.... ....1' ..........._" ...... .---Ii0.4 1 . j.2

!~ ____ ____-L____ ______L-_ _ _ _1.2 1.3 1.4 1.5

Energy demand increase in dwelffngFigure 5 CDF ofenergy demand increase.



7/8

could be acceptable if the 90th percentile of the calculatedcondensation amounts would be smaller than a critical level,i.e., 0.1 kg/m'.

An obvious way to improve the reliability of the airbarrier system and to reduce the condensation risk is the dupli-cation of the air banier, for instance, by applying the underlayor the sheathing according to air banier requirements.However, a second air barrier also is susceptible for damage.Once a continuous airflow path in the envelope develops, thecondensation problems are essentially as problematic as in thecase with a single air barrier. Protective measures to reducecondensation amounts even though air barrier defects occurwould be more effective.Diversity

In the introductory calculation example, two potentialmeasures were defined to reduce the condensation rates inleaky envelope parts: the use of thermally insulating or vaporpermeable materials at the cold side of the structural cavity.The effectiveness of these measures in reducing condensationrisk is studied more in detail here by failure effect analysis.

The analysis is done for a roofwith the same properties aslisted in Table 5 but with different underlay properties. Figure6 shows the CDF of the condensation rate in a roof with avapor permeable underlay with diffusion thickness 0.02 m,together with the CDF for the original roof design. Thecondensation risk is reduced, as expected, butis not acceptableyet. In the same graph, the results are shown for a calculationthat does not account for thermal radiation to the outsidesurface (q = 0 instead of -3 0W1m'). The influence of radiationon the condensation rate is more important in the case with avapor permeable underlay. The results with no radiationlargely underestimate the condensation risk. This demonstrates the importance of including radiation at the outsidesurface in condensation calculations. The failure effect anal-ysis also confirms that a vapor barrier is completely ineffec-

oL -________ L__________L-_________o

Figure 60.5 1.5

Condensation rate (Ifw eeklm2)Effect of vaporpermeable underlay, parameter= internal lining / underlay lin mi.

Thermal Envelopes VII/Moisture Assessments-Princip les

tive in reducing the condensation risk: the CDF is essentiallythe same no matter whether the diffusion thickness of theinternal lining is 0.5 m, 5 m, or 50 m.

Figures 7 and 8 show the effect of exterior insulation onthe condensation risk. The thermal resistance of the layers atthe cold side ofthe structural cavity is the parameter in the failure effect analysis. The analysis is performed for the originalroof design and for the design with vapor permeable underlay.In the original roof design, an exterior insulation withhigh thermal resistance is needed in addition to the cavity insu-lation in order to reduce the condensation risk to the desiredlevel. The 90% percentile of the condensation rate CDFbecomes lower than 0.1 kg/m2/week when a thermal resis-tance of 2 m'KJW is added to the structural framing, theequivalent of 6 em of extruded polystyrene (Table 7).

15"

15"

0.8

0 .

OA

0.2

00 O.S

Condensation rale (I/w eeklm2)Figure 7 Effect of exterior insulation = 1.5 mY,

parameter ~ thennal resistance of layers atcold side of structural cavity (in m2KIW).

0.8

0 .

0.4

0.2

00 0.5

Condensation rate (t.W eeklm2)1.5

Figure 8 Effect of insulating and vapor permeablesheathing = 0.02 mY, parameter = thennalresistance of layers at cold side of structuralcavity lin m2-KIW).


8/8

TABLE 7Distributiou of CondensationRate as a Result of Roof Defects (kglm2/week)

Properties of Layers at Cold Side of Structural CavityDiffusion Thic kness J1d = 1.50 m

Percentile Thermal Resistance (m2.K/W)

(% ) 0.001 0.07 1.010 0.01 0.01 050 0.42 0.38 0.1390 0.95 0.84 0.24

A vapor permeable underlay or sheathing is unable toreduce the condensation risk without additional thermal resistance. However, the combination of a high vapor permeancewith a thermal resistance even as small as that of an air cavityis very effective in minimizing condensation risk. Theseresults indicate that very simple and inexpensive designmeasures are capable of creating reliable condensation controlsystems in lightweight building envelopes.

CONCLUSIONSThe methodology of risk analysis and assessment has

been reviewed and applied to study the reliability of condensation control measures in lightweight building envelopes. Itis generally recognized that lack of airtightness is the moreimportant cause of serious condensation problems, but intenns of risk analysis, the air barrier protective system is anonredundant condensation control system. Considering thehigh occurrence of human error in the building process, it hasbeen shown that the possibilityof air barrier defects during theservice life of the building envelope is high. To define the reliability of the condensation control system, the consequencesof air barrier failure have been quantified using a two-dimensional numerical control volume model for the calculation ofcombined heat, air, and vapor transfer in multilayered buildingenvelope parts.

A set of failure modes and design calculation conditionshas been defined for an exemplary wood frame insulated roof,and a failure effect analysis has been performed to predict thecondensation risk as a result of air barrier defects. The analysishas demonstrated that the condensation risk cannot be justifiedbecause of the potentially important effects of air barrier failure. Therefore, the effectiveness of redundant protectivemeasures to reduce the condensation risk has been studied.The results have shown that the use of thermally insulating orvapor permeable materials at the cold side of the structuralcavity are capable of creating reliable condensation controlsystems in lightweight building envelopes. The use of a vaporbarrier, on the other hand, is completely ineffective in reducing the condensation risk.206

Diffusion Thic kness Jld = 0.02 mThermal Resistance (m2.K1W)

2.0 0.001 0.07 0.17 0.330 0 0 0 0om 0.34 0.20 0.08 00.07 0.78 0.51 0.23 0.005

REFERENCESAng. A.H.-S., and W.H. Tang. 1984. Probability concepts in

engineering planning and design, Volume II-Decis ion ,risk and reliability. John Wiley & Sons, Inc.

Ansell, J., F. Wharton, and e. Wiley (eds.). 1991. Risk:Analysis, assessment and management.ASHRAE. 1997. i997 ASHRAE Handbook-Fundamen

tals. Atlanta: American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.

Di Lenardo, B., w.e. Brown, W.A. Dalgleish. K. Kumaran,and G.F. Poirier. 1995. Technical guide for air barriersystems for exterior walls of ow-rise buildings. Ottawa:Canadian Construction Materials Centre, NationalResearch Council Canada.

DIN 4108. 1981. Wiirmeschutz im Hochbau. DeutschesInstitut fiir Normung e.Y.

Hens, H. 1996. Heat, air and moisture transfer in insulatedenvelope parts (HAMTIE): Modelling. InternationalEnergy Agency Annex 24. Final report, Volume I.

Janssens, A. 1998. Reliable control of interstitial condensation in lightweight roof systems: Calculation and assessment methods. Ph.D. thesis, K.U.Leuven, LaboratoriumBouwfysica, Heverlee.

Kletz, T. 1992. Hazop and Hazan, ident ifying and assessingprocess industry hazards. Institution of Chemical Engineers. Bristol, Pa.: Hemisphere Publishing Corporation.

Lstiburek, J.W., and J. Carmody. 1993. Moisture controlhandbook: Principles and practices for residential andsmall commercial buildings. New York: Van NostrandReinhold.

Lstiburek, J.W., and M.T. Bamberg. 1996. The performancelinkage approach to the environmental control of buildings, Part II: Construction tomorrow. 1. Thermal. Insul.and Bldg. Envs. Vol. 19, pp. 386-403.

Ten Walde, A., and W.B. Rose. 1996. Moisture control strategies for the building envelope. J. Thermal. Insul. andBldg. Envs. Vol. 19, pp. 206-214.


023 password removed

Documents