Building preservationists take precautions. That’s good. Precautions are actions taken to help assure good outcomes, and are predicated on a belief linking the action with the outcome.

There is a big concern for freeze-thaw damage in North American buildings. If you ask the preservationists why they take measures to prevent freeze-thaw damage, they are likely to tell you that water can get in masonry, and as ice, it has a greater volume than water. So it can expand and damage the masonry, they say. That is the commonly-held link between an action—for instance, avoiding applying insulation at the interior—and avoiding freeze-thaw damage.

How strong is the evidence linking an energy measure such as avoiding thermal insulation and avoidance of freeze-thaw damage? It’s weak, to say the least.

1. Definition. We do not have a good definition of “freeze-thaw damage”. There are many types of damage to masonry, and we might wish to cross off seismic cracking, efflorescence and biological discoloration from our list of candidates. Spalling looks like a good candidate. But there are many cases of spalling that are independent of icing. Sideways brownstone. Soft brick. Rising damp. Sky-exposed materials. We can identify damage, say from spalling, but there is no means of attributing that damage to icing. It’s difficult to say the least to try to solve a problem where we cannot clearly identify instances of the problem.
2. Direct experience. Preservationists who have seen defects which they attribute to freeze-thaw for reasons in their judgment have a basis for taking precautions for avoidance. I’m not in that group. I have not met a preservationist in that group. That does not undermine the importance of direct experience, however those with direct experience do not number among my acquaintances.
3. Indirect experience. There have been surveys of historic buildings for damage. “Practices to Improve the Thermal Energy Performance of Heritage Buildings: A Literature Review”. Sahar Sahyoun and Hua Ge, 2020. I have their permission to cite from the search, though it is not published. They reviewed 52 reports. Here are the first two conclusions, which appear internally at odds:
   • Most of the guidelines focus on cultural preservation in general. As for the building envelope thermal performance upgrade, most guides do not suggest or are very cautious about internally insulating the exterior walls.
   • Although analyses and modeling showed a decrease of temperature and an increase of condensation potential within the wall retrofitted with interior insulation, there was no obviously observed adverse effect of adding interior insulation on historical masonry walls based on field assessments by CMHC projects and others. (my emphasis)
4. Insulate half the building. Wouldn’t it be nice if we could do an experiment where only half the building is insulated and the other half is not? Actually, every building provides this experience. Outside corners are never heated–their geometry does not permit heat from indoors to get to the corner. Any infrared scan will tell you this is true. So simply by looking at the condition of the corner, you can tell what the condition in an uninsulated clear wall area will be.
5. Mechanical model. Ice is a crystal which is presumed to grow in size with prolonged sub-freezing temperatures. All crystals grow at a solid-fluid interface, where fluids feed the crystal lattice. There is no crystal growth at a solid-solid interface. Therefore ice growth does not apply direct force against stone or brick surfaces. The only forces present are fluid pressures. For fluid to apply pressure it must be 1) fully, perfectly confined with not even pinhole leaks, and 2) essentially free of air pockets which can accommodate fluid expansion with only low resulting pressures. The likelihood of achieving these two conditions in porous masonry elements is vanishingly small.
6. Field experiment. In the last few decades the space conditioning systems in many buildings have been converted from hydronic to forced air, involving yanking maybe a million hot radiators from beneath windows. The area beneath a window is often wet. Has there been an outcry, with at-risk areas of masonry suddenly subject to lowered temperature and greater swings? Not to my knowledge.
7. Benchtop experiment. Fagerlund [1] developed the critical degree of saturation test, which appears to be the only repeatable way of causing dimension strain in a masonry unit through icing. Air in the sample is first evacuated using vacuum process and the pores are water-filled. Then the sample is bagged and immersed in a bath that cycles above and below freezing. Dimension change is measured with calipers. Notice that these experimental conditions match the mechanical model requirements in #4 above, and do not represent field conditions.
8. Data distribution. If spalling is the purported freeze-thaw effect, is there a distribution of this defect weighted to the north? Are there any forms of damage that show a north climate weighting? To my knowledge, no.
9. Temporal distribution. Do we hear reports of masonry popping during or after a deep freeze event? To my knowledge, no.

In my estimation, a link between an energy measure such as avoiding the use of internal insulation and “freeze-thaw damage” is weak to non-existent. If the term “precaution” describes an action taken to help assure a welcome outcome, where a link between the action and outcome is strong, what do we call it when the link is weak to non-existent? Superstition. Precaution is good. Superstition is not good.

There is much more to say on this subject. Masonry exposed to the sky is more at risk of damage than vertical wall installation. Salts play a role which is more chemical than mechanical. Good rainwater management is always a good idea. Thermal expansion plays an important role in most instances of masonry defects. Diurnal temperature swings of surfaces oriented to the sky are about double the temperature swings in wall surfaces. Projecting parts of building are the most at risk for weathering, and projecting parts of buildings are never warmed with absence of insulation. Avoiding use of interior insulation in order to avoid “freeze-thaw damage” is far too random a measure to be seen as an engineering solution. Avoiding wall insulation is not at the top of the list of energy measures in a building; other measures provide a more clear and immediate payback. Concerns for freeze-thaw damage are rare in the European literature—its focus seems to be North America. The standard method for assessing risk from freeze-thaw (critical degree of saturation testing plus transient hygrothermal modeling) predicts strain that is similar in magnitude to thermal strain in a normal temperature range. Those points will be taken up in future posts. For now, we note that the claim of precaution is undermined by the weakness of the link between avoiding interior insulation and avoidance of damage claimed to be due to freeze-thaw.

[1] Fagerlund, Göran 1973. Significance of critical degrees of saturation at freezing of porous and brittle materials.