BA-1307: Interior Insulation of Mass Masonry Walls: Joist Monitoring, Material Test Optimization, Salt Effects

Executive Summary

Load-bearing brick masonry buildings are a significant portion of the existing building stock; however, adding insulation to the interior side of walls of such masonry buildings in cold (and particularly cold and wet) climates may cause performance and durability problems in some cases. Previous work has helped disseminate information to North American practitioners on the durability risks associated these retrofits, and measures that can be taken to address these risks. However, there are known knowledge gaps; some topics are covered in the current report:

• Field monitoring of the hygrothermal behavior of moisture-sensitive wood beams embedded in a mass masonry structure retrofitted with interior insulation.
• Analysis of the existing database of masonry material property testing results, to determine whether freeze-thaw resistance generalizations can be made based on easily discernible properties.
• Examination of the effect of dissolved salts on freeze-thaw resistance of masonry materials, and determination whether test methods should be modified due to the presence of salts.
• Optimization of the testing of freeze-thaw resistance of masonry material samples, which included improvements in test repeatability, and reduction in test cycle time.

Field Monitoring of Embedded Wood Members in Insulated Masonry Walls: Wood members embedded in a masonry structure will be colder (and potentially wetter) after an interior insulation retrofit; the potential impact is not as well understood.

This work involves the field monitoring of embedded wood joist ends in a solid brick building which is being retrofitted with interior insulation. Eleven joists scattered throughout the building are being monitored, with a variety of orientations, exposures, and masonry wall types. The joists are being monitored for wood moisture content (low and high at the joist end), and temperature and relative humidity within the joist pocket. In addition, indoor and outdoor conditions are being recorded. Results have been collected for five months (December 2012-May 2013); one limitation of these preliminary results is that construction is still ongoing, and that the interior in much of the building is still at unheated and unoccupied conditions.

One conclusion was that the field monitoring methodology appears to provide valid and relevant data. Current results show that at many orientations (especially the north side of the building), joist moisture contents are high (20-30%), which reflects the long-term "mothballed" (unheated) condition of the building (past two winters). However, visual examination and installation of sensors revealed no signs of existing wood decay. On other, solar-heated orientations (south), the moisture content is in the safe, 10-13% range. The joist pocket relative humidity measurements match the patterns of wood moisture content measurements. The moisture contents at the joist "low" position (near the beam "seat") are consistently higher than the corresponding "high" position. At the portion of the building that is heated, the joist ends dried rapidly after heating, even with the presence of interior insulation. However, air sealing and insulating the joist pockets resulted in increasing relative humidity and moisture contents.

This installation will continue to be monitored, as the building is completed and occupied, which will provide seasonal patterns (e.g., drying in summertime), as well as winter data from occupied conditions after construction completion.

Use of General Masonry Characteristics and Basic Materials Testing: Frost dilatometry (measurement of masonry critical degree of saturation/Scrit) is used to asses freeze-thaw risks to masonry buildings retrofitted with interior thermal insulation.  However, the associated testing and simulation activities add project cost and time requirements.  Some practitioners have questioned whether frost dilatometry testing is required in every case, and whether generalizations could be made based on more easily obtained properties, such as manufacturer, manufacturing method (pressed vs. extruded brick), vintage, density, porosity, etc.

The database of 24 previous projects was analyzed to determine whether generalizations could be made, by plotting variables such as density, porosity, water uptake, and Scrit against vintage or other measurements.  Few useful patterns could be discerned from this database. Further development of test methods and masonry material properties databases is recommended in the hope that future review of larger data sets may identify other opportunities for generalization.

Effect of Salts on the Durability of Masonry Materials: This research examined how salts affect the durability of masonry materials—and specifically, how they affect assessments of freeze thaw degradation risk when adding interior thermal insulation to masonry buildings.  A literature search demonstrated that dissolved salts exacerbate freeze-thaw issues in concrete, and similar behavior is expected in masonry.  These dissolved salts might influence frost dilatometry testing; this effect warrants further research.  In addition, salt damage can occur by sub- efflorescence, or subfluorescence, where dissolved salts recrystallize within the pores below the masonry surface, and cause spalling due to the expansive force of crystal growth.  Freeze-thaw and subfluorescence damage may be confused; determining salt content at damaged areas in the field is useful to understand the underlying mechanism.  Furthermore, it is unclear if interior insulation retrofits can cause or exacerbate significant salt decay for walls, due to changes in temperature regime and/or drying magnitude or direction.  Additional exploration is warranted.

Optimization of Freeze Thaw Dilatometry Testing: The frost dilatometry (freeze-thaw test) process often has to be repeated, and has significant time and cost requirements. Minimizing these time and cost obstacles would allow more thermal insulation retrofit projects to benefit from this testing. This work examined optimizations to testing methods, determining whether they had any detrimental effects on measurements.

Reducing the physical sample size allowed more samples to be fit into the chilled bath, doubling throughput. Sample length measurement was fine-tuned, switching from metal screw “targets” to a jig that ensured consistent placement of the sample and micrometer (without the use of targets). This reduced sample preparation time, and increased repeatability of measurements.  Issues were found with length measurements of softer brick samples (sample erosion), and wet vs. dry measurements; protocols were improved to address these issues.

Freeze-thaw cycle time was reduced by almost half, by increasing the rate of cooling, decreasing the target thaw (upper) temperature, and decreasing the hold time from an hour to 30 minutes.  These modifications could be made without any loss in accuracy in the test.

1 Introduction

Load-bearing brick masonry buildings are a significant portion of the existing building stock in the East Coast and Midwest regions of the United States. However, adding insulation to the interior side of walls of such masonry buildings in cold (and particularly cold and wet) climates may cause performance and durability problems in some cases. Exterior insulation provides the ideal conditions for building durability; however, many buildings cannot be retrofitted with insulation on the exterior for reasons such as historic preservation, cost, zoning or space restrictions, or aesthetics.

Previous Building America and other publications (Maurenbrecher and Shirtliffe 1998; Gonçalves 2003; Straube and Schumacher 2002, 2004, and 2007; BSC 2011; BSC 2012) have helped disseminate information to North American practitioners on the durability risks associated with interior insulation of masonry walls, and measures that can be taken to address these risks.  However, there are known knowledge gaps in the body of literature, some of which are addressed in the current document. These research topics include the following items, and are covered in more detail under “Research Topics and Research Questions.”

• Field monitoring of the hygrothermal behavior of moisture-sensitive wood beams embedded in a mass masonry structure retrofitted with interior insulation.
• Analysis of the existing database of masonry material property testing results, to determine whether freeze-thaw resistance generalizations can be made based on easily discernible properties, such as vintage, geographic location, brick type (pressed vs. extruded), porosity, etc.
• Examination of the effect of dissolved salts on freeze-thaw resistance of masonry materials, and determination whether test methods should be modified due to the presence of salts.
• Optimization of the current regiment of masonry material property testing: in particular, determination of freeze-thaw resistance of samples. This optimization included improvements in test repeatability, and reduction in test cycle time requirements.

Background and Previous Work

It is well established that exterior insulation provides the ideal conditions for building durability and performance (Hutcheon 1964); this approach is further described by Lstiburek (2007a). However, many buildings cannot be retrofitted with insulation on the exterior for reasons such as historic preservation, cost, zoning or space restrictions, or aesthetics.

Numerous obstacles to more wide-scale deployment of interior retrofits include concerns about freeze-thaw damage due to reduced outward heat flow and reduced inward drying, and the potential for decay of wood structural framing members (typically floor joists) that are embedded in mass assemblies. The problems and some case studies of interior retrofits are outlined by practitioners such as Maurenbrecher et al. (1998), Gonçalves (2003), and Straube and Schumacher (2002, 2004).

The masonry freeze-thaw issue has been examined by (among others) Litvan (1975a), Mensinga et al. (2010), and Lstiburek (2010). These practitioners propose the use of material property testing as an input to hygrothermal simulations, using a limit states design approach, where load (climate exposure) is compared with the capacity (freeze-thaw resistance of the brick). The embedded floor joist decay issue has been studied by some practitioners (Dumont et al. 2005, Scheffler 2009, Morelli 2010, Morelli and Svendsen 2012, Ueno 2012), but many issues remain unresolved.

Straube and Schumacher (2007) reviewed the moisture control principles that must be followed for a successful insulated retrofit of a solid load-bearing masonry wall.  Interior insulation reduces heat flow through the assembly, thus changing the existing moisture balance of wetting and drying. Given this reduced drying, the retrofit design should reduce wetting in a commensurate manner. The winter temperatures of parts of the inner layers of masonry are also significantly lowered, thereby raising the risk of freeze-thaw damage. This document was intended for architectural practitioners.

Previous Building America work includes BSC TO2 Task 7.3: “Internal Insulation of Masonry Walls: Final Measure Guideline” (BSC 2012). This Measure Guideline presents the current knowledge in terms of moisture-safe retrofits of mass masonry walls, expanding on Straube and Schumacher (2007). It also includes a checklist intended as a brief outline of key points and takeaways, which can be used by a field practitioner when assessing a mass masonry building for interior insulation.  The intention was that the Measure Guideline would provide more detailed explanations and visual examples of items covered in the checklist.

In addition, BSC led a Building America Experts Meeting on this topic, with contributions from multiple practitioners, and discussions on gaps in the current knowledge and recommendations for future work. A full report is provided in BSC Task 1.3: “Recommended Approaches to the Retrofit of Masonry Wall Assemblies: Final Expert Meeting Report” (BSC 2011). Relevance to Building America’s Goals Given the Building America goals of reducing home energy use by 30%-50% (compared to 2009 energy codes for new homes and pre-retrofit energy use for existing homes), insulation and air sealing of mass masonry walls will need to be a component of the work if mass masonry residential buildings are to be addressed. Potentially millions of housing units could benefit from a better understanding of the moisture risks associated with interior retrofits, and the means of reducing these risks.

The majority of construction that can benefit from this research is in locations with older building stock (i.e., mass masonry). The greatest concentrations are likely on the East Coast and in the Midwest (i.e., cold climates), although these types of buildings are definitely present throughout the country.

One belief that appears to be common among less technical practitioners is that insulation of mass masonry structures is not possible (due to potential for damage) and/or unnecessary (as thermal mass effects provide sufficient benefits).  Previous work discussed above covers risk management for interior insulation of mass masonry walls. Basic energy models can be used to show that uninsulated thermal mass does not result in significant improvement to energy performance for buildings located in heating-dominated (cold) climates. Thermal mass is of greater benefit in locations with high diurnal swings around the interior setpoint, as commonly found in the United States Southwest. This is covered in more detail by BSC (2012). Note, however, that thermal mass is not dismissed as a concept: energy savings can be obtained by combining insulation with thermal mass (e.g., charging mass with passive solar gains). Temporal load shifting effects can be another benefit, such as pre-cooling of a meeting room to reduce peak load effects and minimize installed equipment size.

The research topics proposed here are specifically relevant based on the current needs of practitioners.  The Building America Standing Technical Committee on Enclosures has identified both of these as important topics for additional research work (see Enclosures STC Strategic Plan: #30 Walls: Interior Masonry Retrofits; and #39 Walls: Embedded Wood Beams and Joists.)

Cost-Effectiveness

Load bearing masonry has a wide range of thermal properties.  However, even a ‘thick’ multi- wythe load-bearing masonry wall is likely to have an R-value in the range of R-3.2 to R-6.8, with an average R-value of around R-5. Surface heat transfer coefficients (“air films”) of another R-1 may result in thermal performance comparable to that of a high-end (i.e. triple glazed) modern window; however this level of insulation is too low for most practical purposes (considering current energy costs, building durability, health and thermal comfort issues). Hence insulation is often added during retrofits, and is critical to achieving high performance (as per Building America program goals) in any climate with significant heating loads.

The R-values of uninsulated masonry walls are also substantially below modern code requirements for cold climates: for Zones 5 and 6, the typical opaque-wall “true” R-value requirements are in the range of R-12 to R-17 (as calculated from U values given in the 2009 International Energy Conservation Code/IECC, Table 402.1.3; ICC 2009).

Thermal insulation follows the law of diminishing returns, with decreasing return on investment with increasing insulation thickness. Given that these wall assemblies are being changed from uninsulated (base case) to insulated (final) assemblies, it is likely that the initial inch or two of insulation should be highly cost-effective. Optimization would be a function of insulation cost, energy costs, and climate zone.

BSC’s current retrofit recommendations include the use of closed-cell air-impermeable spray foam as an interior insulation material for mass masonry walls. Closed-cell spray foam (ccSPF) has a typical installed price of roughly $1.00 per board foot (although prices appear to be falling below this level); when normalized by R value (instead of volume) this is equivalent to approximately $0.16/sf·R. In comparison, typical loose-fill fibrous insulations are sold at $0.02- 0.04/sf·R; note that this is a material cost, not an installed cost. If a 1:1 material-to-installation cost ratio is assumed for this estimate, the use of closed-cell spray foam is still roughly two to four times as expensive as the lowest-cost loose fill materials. However, as discussed by Straube and Schumacher (2007) and BSC (2012), the use of air permeable, moisture-sensitive fibrous insulations in this application results in assemblies with higher risk of moisture-related failures.

Tradeoffs and Other Benefits

Basic benefits of the retrofit insulation of uninsulated masonry walls include energy savings and thermal comfort improvements for occupants (due to radiant surface temperature effects and air leakage reduction).  The assemblies under discussion could meet requirements for modern energy codes, as discussed earlier, unlike the non-retrofitted assemblies.

Research Topics and Research Questions

The research topics and associated questions are broken down here by section; the answers to these research questions are provided in the conclusion of this report.

Field Monitoring of Embedded Wood Members in Insulated Masonry Walls

Previous simulation work on embedded wood members in insulated masonry walls left many open questions, which could be addressed—at least in part—by in-situ measurement of embedded member moisture content.  This field testing was done in collaboration with Merrimack Valley Habitat for Humanity (MVHfH) in Lawrence, MA. This organization is currently in the process of renovating a mass masonry multi-family building into 10 condominium units; the monitoring equipment is installed at joists at this site. The associated research questions were as follows:

• What are the temperature conditions at the joist pockets or beam pockets with and without retrofitted interior insulation?
• What are the measured seasonal moisture contents of embedded beams in insulated and uninsulated cases, and can these be mapped to durability risks?
• Does orientation have a specific effect of temperature and moisture conditions?  This will include both the effects of solar gain and wind-driven rain.
• Does distance from grade (and thus “rising damp,” via capillary activity) have a specific effect on temperature and moisture conditions?
• Is there a significant difference in the performance of an insulated beam without an air seal around the beam pocket, compared to one that is rigorously air sealed?
• Are there measurable airflows occurring at joist pockets in wintertime or summertime conditions?  Given the limits of hot wire anemometers, it is unlikely that airflows are within the measurement range, but will be checked for reference.
• How do the monitored data correspond to previous three-dimensional and one- dimensional simulations?

Use of General Masonry Characteristics and Basic Materials Testing

Frost dilatometry is a costly and time-consuming test, which has probably limited its application to high profile and/or high value projects. If generalizations could be made on freeze-thaw resistance from variables such as brick vintage, type (pressed vs. extruded), manufacturer, porosity, or other easily recognizable physical characteristics, it would simplify risk assessments for interior insulation. BSC has collected a database of material properties from previous projects, which can be compared with the variables described above. However, there is a great deal of variability in measurements; a larger sample set would be required to obtain robust and reliable results. The relevant research question is:

• Based on analysis of the existing database of masonry material property testing results, can freeze-thaw resistance generalizations be made based on easily discernible properties, such as vintage, geographic location, brick type (pressed vs. extruded), porosity, etc.?

Effect of Salts on the Durability of Masonry Materials

The issue of effect of salts (dissolved in mass masonry) on freeze-thaw behavior was raised at BSC’s July 2011 Building America Experts Meeting (BSC 2011). Salts, acting alone, can cause damage to masonry materials (via subfluorescence); it is possible that visible damage may be due to this salt migration.  In addition, the presence of dissolved salts can exacerbate freeze-thaw damage in masonry materials. The associated research questions were as follows:

• What is the theoretical impact of salts in masonry materials?
• How might dissolved salts affect the mechanics of freeze-thaw?
• What range of osmotic pressures (pressure drives due to different salt concentrations) might be expected?  Could these pressures result in damage?
• What are the mechanics of subfluorescence?
• Do practitioners confuse freeze-thaw damage and salt damage in the field?

Optimization of Freeze Thaw Dilatometry Testing

As described previously, masonry material property testing (frost dilatometry) is a costly and time-consuming procedure. Further development is warranted to make it a more available, repeatable, and commonplace tool for practitioners. The research team has done work to optimize testing, such as improving the repeatability of dilatometry measurements, decreasing freeze-thaw cycle time, increasing precision in sample length measurement, and increased accuracy in interpretation of results. The associated research questions were as follows; they examined improving repeatability (in measuring the length of masonry samples), and improving test throughput (reducing cycle time from the current two day cycle).

• Can targets (metal reference points attached to the sample, similar to those used in structural testing) be used to improve repeatability of dimension measurements?
• Is there an alternative to measuring dimensions; for example, can freeze-thaw damage be identified via acoustic methods such as measurements of attenuation?
• Can the freeze-thaw cycle be modified in order to decrease test cycle time, and therefore increase throughput?
• Alternately, could the cycle time be compressed to allow more freeze-thaw cycles and a greater dilation to be achieved in a shorter time period?
• Are the temperatures & time steps currently used sufficient to bring the entire masonry specimen (including the core) to target temperatures?

2 Field Monitoring of Embedded Wood Members in Insulated Masonry Walls

Wood members embedded in a masonry structure will be colder (and potentially wetter) after an interior insulation retrofit; the potential impact is not as well understood. The work in this section involved the field monitoring of embedded wood joist ends in a solid brick building which is being retrofitted with interior insulation.

Background

There are durability risks and concerns when applying interior insulation to mass masonry walls. Some of them have been (or are being) addressed, including interstitial condensation and brick freeze-thaw damage. However, another durability risk that has received less investigation is the hygrothermal behavior of moisture-sensitive wood beams embedded in the load-bearing masonry. With the retrofit of interior insulation, the embedded beam ends will spend longer periods at colder temperatures than their pre-retrofit condition. Therefore, these wood members will have reduced drying potential due to reduced heat or energy flow (as described by Lstiburek 2008). The wood will also be subjected to higher relative humidity (RH) conditions in the beam pocket, and therefore remain at a higher moisture content (MC): both of these factors increase the risk to the beam’s durability.

Some solutions that have been proposed and/or used to protect embedded members include borate preservative injections into the wood; metal plates next to the member (to provide passive heat flow); active heating at the beam ends; or construction of a load bearing structure inside the masonry, and cutting off the end of the beam.

Literature Review

Several practitioners have examined the problem of moisture risks at embedded members, using both in-situ monitoring and simulations.

Dumont et al. (2005) monitored moisture content of wood structural members embedded in masonry in two low-rise residences that were retrofitted with insulation in Wolseley, SK (DOE Zone 7, “dry” climate) and Kincardine, ON (DOE Zone 6A, “moist” climate). The Wolseley house was insulated with mineral wool, with a polyethylene vapor barrier; the Kincardine house was insulated with spray polyurethane foam. The foam insulation encased the wood members where they were seated in the masonry wall.

Data showed that the wood members of the Wolseley house remained at safe moisture content levels (10-15%) throughout the monitoring period. This is not meant as an endorsement of the use of air-permeable insulation with an impermeable interior vapor barrier: this is actually a high risk assembly. However, the lack of moisture load (dry climate) is likely the dominant factor. In contrast, the Kincardine house showed consistently elevated moisture contents (20%+) at several locations.  It was suspected that the moisture source was capillary uptake from the wet foundation, but rainwater absorption through the face of the masonry (due to surface detailing) was not eliminated as a possible source. The limited drying to the interior available through spray polyurethane foam was also likely a contributing factor.

Scheffler (2009) used simulations to examine the problem of moisture accumulation at wooden beam ends. He used DELPHIN two-dimensional hygrothermal simulation software, running a wooden beam embedded in brick masonry under steady state conditions (23° F/-5° C/80% RH exterior; 68° F/20° C/50% RH interior; 90 days). These simulations indicated the moisture risks associated with insufficient control of airflow or moisture vapor flow (diffusion) from interior sources. This was followed by one-dimensional and two-dimensional simulations using transient weather data (Bremen; mild, maritime climate with high rain and humidity), indicating increases in relative humidity and liquid water (condensation) at the beam ends due to the addition of insulation.

Scheffler also described the historic methods to increase embedded beam longevity, such as charring the beam end to increase moisture resistance, and the addition of exterior-to-interior ventilation at the beam pocket.  He discussed current methods to ameliorate these moisture issues due to insulation retrofits, including replacement of wood floor/ceiling assemblies with non- moisture sensitive materials (e.g., concrete), and possibly the addition of heat and/or ventilation at the wood beam end.

Morelli et al. (2010) collaborated with Scheffler, continuing examinations of this issue. They proposed the solution of leaving a gap in the insulation of 12” (300mm) above and below the floor, resulting in a 30” (770 mm) gap (12” gap×2 plus floor depth).  Two- and three- dimensional heat transfer simulation showed that the heat flow was reduced by 60% going from the uninsulated to insulated cases, while the “gap” case was only a 45% reduction. This work was followed by two-dimensional DELPHIN hygrothermal simulations of the embedded beam (in a Bremen climate).  Relative humidity levels in a corner of the beam pocket (and equilibrium wood moisture content) were compared between cases. The existing, uninsulated wall showed a drying trend; the fully insulated wall showed seasonal increases in RH; and the gapped insulation wall showed performance between the two previous cases (but with increasing moisture levels).  However, these results assumed a relatively high wind-driven rain loading factor. Switching to a lower loading factor resulted in the gapped insulation assembly showing a general drying trend.

Morelli and Svendsen (2012) continued the previous work of Morelli (2010), showing further simulations with continuous interior insulation, and insulation installed with a gap at the floor beams. Another result of this research was a methodology for assessing retrofit measures on brick masonry walls based on a failure mode and effect analysis.

Morelli and Svendsen also performed an extensive literature review from the 1980s to current day. Several in-situ field studies were examined; the typical findings were that driving rain did not result in moisture problems at beam ends.  Some researchers found that cracks in the façade could result in problems, but crack-free façades had acceptable performance. Another researcher examined the effect of interior air leakage at beam ends: at high air leakage rates, temperatures in the beam pocket were increased, reducing condensation risks. At low air leakage rates, air- transported moisture was negligible.  However, at intermediate flows, condensation did occur. That researcher recommended the addition of localized exterior wall insulation at beam ends, or insulating the beam end cavity. Other researchers examined techniques such as adding heat to the beam ends to avoid moisture problems. The results kept wood moisture contents low, but this is an expensive (and therefore unlikely) retrofit. . .

Download the complete report here.