Two subfloor moisture tests, and the floors they rank in opposite orders

A concrete slab can pass one manufacturer’s moisture limit and fail another’s, without changing, because the two makers are not measuring the same property of it.

Two tests are in general use on a slab. One drives a probe into a hole drilled to a specified depth and reports the relative humidity inside the concrete. The other seals a dish of anhydrous calcium chloride to the surface and weighs how much moisture the slab gives up in three days. The first describes the state of the slab through its depth; the second describes what is currently leaving its surface.

The two tests do not convert into one another. Not approximately, and not with a correction factor: a slab that is drying from the top can be quiet at the surface while the interior is still wet, and one under a vapor retarder can read the opposite way. Which one a manufacturer names decides which slabs its floor may be laid on.

15 of the 16 products here declare a relative humidity ceiling for the slab, and nine products declare a calcium chloride ceiling. Nine products declare both, which is what makes the comparison below possible at all.

Three calcium chloride cells in this release are held back rather than published: two independent readings of the same document disagreed about the limit, and nothing goes out until a person settles it. Those products are absent from the comparison, and absent for a reason that is ours rather than the maker's.

Of the 36 pairs of products that can be compared on both tests, 14 are ranked in opposite orders by the two. Mannington ADURA Rigid 6x48 sets the stricter humidity ceiling of the two at 85 percent against 95, and Bruce LifeSeal Classic sets the stricter calcium chloride ceiling at 5 pounds against 8. A slab that is acceptable to one of them on the test the installer happens to run is not acceptable to the other, and swapping the test swaps the answer.

The reversal means that "which floor has the stricter moisture requirement" is not a question with an answer until the test is named. An installer who runs the probe because that is the equipment in the van, and reads the result against a manufacturer that publishes a calcium chloride ceiling, has checked a number against a limit that does not govern it. The slab was not measured wrongly; it was measured for a different thing.

It also means that a specification reading "subfloor moisture within manufacturer limits" has deferred the decision rather than made it. The limit is only a limit once the method is attached, which is why the method gets its own column here beside every figure, and marks the cell where a maker names a limit without naming a method.

This note does not say which test is better. That argument belongs to the standards bodies that maintain them, and both are current. What it says is narrower and harder to argue with: the two produce different orderings over the same set of products, so a comparison that quietly mixes them is not a comparison.

The comparison also leaves out wood subfloors, which are governed by moisture content rather than by either of these tests, and the ambient conditions the same documents set for the room. Those are separate fields, read per product, and they sit on the product pages.