OWENS CORNING BASEMENT INSULATION SYSTEM
EXPERIMENTAL EVALUATION PROJECT


Principal Investigator:
    Louise F. Goldberg Ph.D (Eng)
    Center for Sustainable Research
    College of Architecture and Landscape Architecture
    University of Minnesota
    Voice:  (612)821-9208 or (612)626-8783
    e-mail:  mailto:goldb001@maroon.tc.umn.edu

Program Manager:
    Traci Aloi
    Owens-Corning Science and Technology Center
    e-mail:  mailto:traci.aloi@owenscorning.com

Revision date: 3/28/02

 

A.    INTRODUCTION

The principal objective of the project was to obtain transient point and weekly system moisture performance data for an interior basement wall finishing system (BWFS) provided and installed by Owens-Corning. 
This system consists of 2.5 in. thick rigid fiberglass panels with a polyeolfin covering on the interior side.  The panels are mounted in a polyvinyl chloride frame that is nailed directly to the wall.  The BWFS has no vapor retarders and all its components have been independently evaluated to show superior resistance to mold/mildew growth compared with the traditional building materials, wood, gypsum board and untreated ceiling tiles (Stukus 1999). 

The following three vapor retarder configurations have been studied:

The insulation system was installed in the west basement test module at the University of Minnesota's Foundation Test Facility by Owens-Corning personnel in the late summer of 1998.  The basement module has a square floor plan with an interior wall length of 19 ft 4 in. and a ceiling height of 92 in.  As shown in the cross-section of Figure A.1, insulation covered the entire interior wall surface to the underside of the ceiling panel leaving a gap above the slab surface produced by the topology of the framing system.


Figure A.1  Basement wall finishing system installation

It is important to note that the walls consist of standard two-core, 12 in. wide hollow masonry blocks without any exterior waterproofing, dampproofing or insulation.  The absence of any exterior moisture control treatments also is intended to yield the worst-case moisture test conditions.  Inclusion of such treatments reduces the severity of the imposed water vapor boundary conditions and would yield improved apparent BWFS performance.  Similarly, the floor is 4 in. thick poured concrete without any insulation or vapor retarder.  The module is surrounded by medium sand that is drained to a water table located at least 20 ft below grade.  This geotechnical configuration provides a well-drained environment eliminating any liquid water incursion into the wall system while simultaneously allowing a significant water vapor source throughout the year.  The overall installation thus maximizes the water vapor load on the BWFS.  It should be noted that the FTF is designed such that the only sources of moisture for the test modules are in the form of water vapor in the surrounding soil and atmospheric air (soil moisture profiles).

The no vapor retarder configurations were installed on the northwest and southeast quadrants, the wall side retarder configuration in the northeast quadrant and the interior retarder configuration in the southwest quadrant.  Each configuration stretched 8 ft along both walls from the corner, so leaving a guard section of about 3 ft  4 in. between test quadrants.  In all cases, the buffer section had no vapor retarders.   Each quadrant consisted of four, 4 ft by 8 ft  panels which left an exposed wall height of about 6" between the top of the insulation system and the ceiling.  This gap was filled with panel insulation as well (as shown in figure A.1).

The transient measurements taken (amongst others) relevant to the objectives were:

The transient data were collected by a computerized data acquisition system with an aggregation period of 20 minutes and a scan interval of about 6 minutes, so producing 72 sets of stored data per day.

The significant weekly measurements (performed manually) were:

The module interior humidity was controlled by a standard room dehumidifier modified to accept direct digital control by computer permitting a relative humidity control bandwidth of 0.5%.  Interior temperatures were controlled by a conventional digital thermostat.

The system was instrumented and commissioned by mid-November allowing transient data collection to start on 11/18/98.  Manual data collection commenced on 11/6/98.

Moisture performance data was gathered over the full project monitoring period extending from 11/18/98 through 3/1/2000.  However, whole module energy performance data only was collected during the 1998/1999 heating season as a means of confirming that the thermal performance of the insulation would agree with well-established data.

The completed test installation is shown in figure A.2.


Figure A.2

 

B.    EXPERIMENTAL DATA

B.1    Instrumentation layout

B.2    Panel water vapor parameter temporal profiles

B.3    Upper panel baseline humidity ratio correlations

B.4    Upper panel humidity ratio correlations

B.5    Simplified removable section mass temporal profiles

B.6    Removable section mass phenomenology
         This section includes a discussion of the detailed results for the full monitoring period. 

B.7    Energy performance comparison

 

C.    SYSTEM PHYSICAL INSPECTION

On 6/17/1999, a photographic record of the basement system during the summer was made.

At the termination of data collection on 3/1/200, the basement system was dismantled and examined.  The observations and photographic record of the inspection are described for each quadrant as follows:

C.1    Northwest quadrant (no vapor retarder)

C.2    Northeast quadrant (wall side vapor retarder)

C.3    Southeast quadrant (no vapor retarder)

C.4    Southwest quadrant (test cavity side vapor retarder)

 

D.    SUMMARY DISCUSSION

D.1    Simplified physics

D.2    Wetting/drying profiles

 

E.    COMPARISON WITH CURRENT PRACTICE AND PERFORMANCE.

The key issue in comparing the experimental performance of the three vapor retarder strategies tested with current practice is that none of these strategies conforms to the recommended practice in Minnesota.  This practice treats interior foundation wall insulation as a system that is required to provide resistance to both thermal and water vapor flux.  In this context, the current practice is to provide both wall/insulation and insulation/interior vapor retarders in one form or another.  There is still some controversy about the best method of implementing this double vapor retarder strategy, in particular, whether the wall side retarder should extend to the top of the wall or terminate at the grade level.  The theory behind a grade level termination is that it permits drying to the outside in the event that the insulation becomes wet.  There are also issues regarding the appropriate rim joist insulation and vapor retarder configuration, and, in particular, whether the common practice of stuffing the rim joist cavities with batt or blown-in cellulose insulation of some sort is advisable. 

In general terms, the experimental results confirm the efficacy of the dual vapor retarder approach since the negative condensation consequences of single, wall or interior  retarders are apparent.  This is in agreement with the results obtained from the Cloquet Residential Research Facility (CRRF) (publication projected for July, 2000) which focused on determining the moisture performance of various double vapor retarder configurations in combination with a range of rim joist thermal and moisture control strategies.  Hence in terms of the vapor retarding attribute required in Minnesota, the CRRF results confirm that a well-sealed, double vapor retarder configuration extending over the full wall surface is an effective choice provided both retarders essentially are free of penetrations and provision for dealing with condensate that collects on the wall surface is made.  The principal reason for this preference is that a double vapor retarder system offers low net water vapor system permeability and avoids the problems demonstrated by a single retarder on either insulation face with the exception of the incidence of condensation on the exterior side of the wall retarder that needs to be managed by suitable drainage detailing.  Further, when combined with a sub-slab vapor retarder, a dual vapor retarder wall system (or equivalent) yields a significant decrease in net basement envelope vapor ingress.  It should be noted in this context that in the absence of a sub-slab retarder, the benefits of a double vapor retarder over a zero vapor retarder system have been shown experimentally to be not nearly as significant.

Thus while the experimental data show that the zero vapor retarder configuration is an effective thermal insulation system whose performance is not adversely affected by condensation, it cannot be viewed in practice as having any vapor retarding capabilities.  Thus a water vapor management strategy external to the insulation system is required in order to maintain comfortable interior basement conditions and prevent moisture and mold damage to basement contents (furniture, paper, carpets, etc.).  Typically, such a strategy devolves to interior mechanical dehumidification, but other strategies (such as exterior wall treatments) are conceivable.

In comparison with other zero vapor retarder systems which necessarily excludes all systems that have inherent vapor retarder qualities (such as those using polystyrene or foil-faced polyisocyanurate insulation or some other diffusive or advective flux retarder), the Owens-Corning zero vapor retarder interior foundation wall system does have better performance attributes.  In particular, the absence of any organic framing material avoids detrimental biological activity common to wood-framed systems, for example.  Similarly, use of a porous cloth surface finish eliminates the moisture problems associated with common relatively impermeable finishing materials such as gypsum board (which acts as a single interior vapor retarder) while meeting the applicable fire code requirements.

In summary, the main issue is that the zero vapor retarder system and current recommended practice for interior foundation insulation systems strictly are not comparable, since current practice has the objective of providing both thermal and water vapor resistance while the zero vapor retarder system provides thermal resistance only.  This is likely to be the primary factor in gaining regulatory sanction for the Owens-Corning foundation system.

AUTHOR'S PUBLICATION NOTE (3/27/2002)

The final version of this report was conveyed to Owens Corning on 6/2/2000.  In the intervening two years, much additional research has been undertaken (such as the completion of the CRRF study referred to above - link) yielding a fuller understanding of the basement envelope moisture transport phenomenology.  Hence it is appropriate to update section E accordingly.

The interior insulation recommendations arising from the CRRF project are given elsewhere on this web site and so the above speculations about the outcome of that project have been resolved.  Briefly these recommendations are that the rim joist cavities not be stuffed with batt insulation and that indeed, a well-sealed, full-wall, double vapor retarder configuration yielded satisfactory moisture performance.  However, the CRRF project also highlighted the requirement that interior dual vapor retarder systems are suitable for use on strictly dry walls only and that wet walls require a different and much more complex treatment (link).

The dual vapor retarder configuration specified in the Minnesota building code for new construction (rule 7672.0600, consisting of a warm-side (interior) full-height vapor retarder and a "moisture barrier" (undefined) from floor to grade on the wall side) is shown by the data in this report to be prone to failure since, in the critical condensation zone at top of the wall, the code configuration functions as a single interior retarder with all the attendant consequences demonstrated in this report.  Therefore (if a minor speculation may be allowed), the presence of the below-grade wall side retarder may actually exacerbate matters by reducing the drying potential to the outside compared with the situation when using a single interior retarder alone.  This supports the anecdotal reports of the service failures of the MN code configuration.

It also needs to be stated that the vapor retarder specification of rule 7672.0600 nominally make the BWFS unsuitable for use in new construction in Minnesota.  However there is nothing in the MN code that precludes its use for retrofit construction (rule 7672.1200).

F.    PROJECT CONCLUSIONS

  1. A simplified phenomenological model describing moisture transport through foundation walls is well supported by the experimental data.  This model defines a neutral, time varying, horizontal plane around the foundation perimeter below which vapor transport is always directed from the ground into the basement interior.  Above the neutral plane, vapor transport is from the basement interior through the wall system during the winter and reverses during the summer.  The two annual reversal points are dependent on the local climate.  In terms of this model, the condensation source is always in the upper section of the wall and hence, the relative performance merits of the 3 vapor retarder systems tested are apparent in the upper portion of the wall with particular localization to the above-grade section.  Thus the following is based on the phenomenology of the upper half of the basement wall.
  2. The basement insulation system with a polyethylene vapor retarder located between the wall and the insulation collects condensation on the insulation side of the vapor retarder during the winter and on the wall side of the retarder during the summer.  During the winter, the condensate volume is large enough to run down the face of the vapor retarder and collect on the floor.  While the experimental data does not suggest that the insulation itself absorbs any significant amount of condensate, the large volume of condensate deposited on the vapor retarder and its eventual collection on the floor renders this vapor retarder configuration an unacceptable choice, particularly in a permanent installation.
  3. The upper removable panel in the basement insulation system with a polyethylene vapor retarder covering the interior surface absorbed condensate from 3/26/99 through 9/8/99 reaching a maximum gravimetric moisture content of 39%. From 9/8/99 through 3/1/00, the moisture content declined to 15% suggesting that at the same weekly drying rate, the moisture content would reach about 13%  by 3/26/00, before moisture absorption is likely to recommence.  This suggests an unstable, annual positive feedback loop in which the insulation becomes progressively wetter as the years go by.  Thus, the interior vapor retarder configuration clearly is unacceptable.
  4. In the worst-case northwest quadrant, while the basement was maintained at a 56 oF setpoint temperature without dehumidication (approximating an unheated condition), the upper removable panel in the basement insulation system with no vapor retarder absorbed moisture from 8/5/99 through 1/31/00 reaching a maximum gravimetric moisture content of 2.4%  before beginning to dry out.  Upon increasing the basement setpoint temperature to 68 oF on 2/7/00 after a 1.3 g/week decline in water mass was measured, the drying process was accelerated producing a gravimetric moisture content of 2% at the end of the experiment on 3/1/00.  Under heated basement conditions (68 oF setpoint and a 45% maximum relative humidity), the maximum gravimetric moisture content reached 1% and dried out completely within 80 days.  Although the wall surface is wet during the heating season, there is no evidence of any gross wetting of the insulation or the wall, neither is there any evidence of condensate running down the wall surface.  In a heated basement (even if initially maintained at low temperatures for a significant part of the heating season), the drying rate of the insulation is sufficient to confidently project that the small moisture absorption will dissipate by the onset of the next moisture uptake period beginning in August.  Hence, it may be concluded strictly that the zero vapor retarder configuration has a stable moisture wetting/drying annual cycle in a heated basement and thus is a configuration suitable for use as an interior basement thermal insulation system in a heated basement.  The data also permit the conservative estimate that the zero vapor retarder configuration has a stable wetting/drying annual cycle in an approximately unheated (minimum interior temperature of 56 oF) basement as well.
  5. The energy performance of the insulation system was in agreement with expectations yielding the same performance as all the full-wall, interior thermal insulation systems tested in the module within the FTF experimental error of 3%.  In particular, the energy performance was numerically identical to conventional R-11 fiberglass batts in a 24 in. on-center stud frame.

 

G.    COMMERCIAL INSTALLATION RECOMMENDATIONS

In the light of the experimental data and the CRRF data, there are few significant installation recommendations since, by its nature, the zero vapor retarder Owens-Corning system is quite forgiving of low installation quality.  The significant recommendations are:

  1. The insulation system should cover the entire structural wall system without leaving a gap between the top of the wall and the top of the insulation system.  This ensures that the experimentally measured thermal performance will be achieved in practice since the largest heat loss occurs in the top foot or so of the wall.
  2. The insulation system should not extend over the rim joist region as this will produce condensation on the interior band joist or sheathing surface.  Following the recommendations arising out of the CRRF basement insulation / rim joist project, foil-faced polyisocyanurate insulation should be firmly attached to the interior band joist or sheathing surface and the rim joist cavity should not be filled with insulation (batts or similar), allowing a free circulation of interior air.  In practice, this may mean that the basement ceiling system needs to be positioned in the vicinity of the top of the foundation wall.
  3. Electrical wiring conduits or other hardware should not be placed more than 1/3 of the slab/grade height above the slab.

 

H.    APPENDICES

H.1    Experiment log

H.2    Weekly manual data readings

 

I.    PROJECT REVIEWS

        I.1    John Carmody

 

J.    ACKNOWLEDGEMENT AND CERTIFICATION

The research described herein has been performed under the auspices of a research contract with the Owens Corning Corporation which provided all the project funding.  While this financial support is gratefully acknowledged, the Principal Investigator assumes complete responsibility for the contents herein with the exception of section I (Project Reviews).

K.    REFERENCES

Stukus, P. 1999.  Comparative Analysis of Owens Corning Basement Wall Finishing System and Other Building Materials, Report prepared for Owens Corning by Denison University, February, 1999.

back to CONTENTS