Video 4 – Calcium Chloride Shown to Give False Readings (pt.1)
ASTM F1869, often called the moisture vapor emission kit or chloride kit, it has been around for a long time, but what we’ve learned is that it’s actually not necessarily fundamentally accurate or useful in the way we had thought it had been for many years.
Over the last 10 years in our laboratories and in the field and other folks have looked at this test method, and we have to come learn that there are a lot of things wrong with it. Let’s take a look at some of those in some detail.
This first slide talks about how the test was invented back in, first published in 1941. I had the opportunity to look through many old books and magazines and newspapers and interview people who were involved in moisture testing and flooring going back all the way to the 1940s.
The very first reference that we found to this test method is what was called a “dampness test” published in 1941 by Armstrong in their installation book on linoleum. The way they did the test was to drill a small 1/2 inch diameter hole, about an inch deep into the concrete and over the top of it place a small glass dish and put crystals of calcium chloride into that dish. They put a ring of putty around that and over the top a big glass and covered it up and let it sit overnight.
You come back and look at and if the crystals of chloride look wet, then the floor is wet. It was strictly a qualitative go or no go sort of test. There was no numerical results associated with it. It was not a quantitative test. It’s what we would simply call a qualitative type of assessment. That started in 1941.
This next slide, you see a page from Kentile’s magazine, published in 1960, so roughly 20 years later. Kentile in Brooklyn, New York, who manufactured vinyl asbestos tile, began to actually manufacture and sell kits to do the calcium chloride test. The way it was done was totally unknown. There’s no pedigree to this test. There is nothing that we can find in the literature from the 1940s, ’50s, or ’60s as to why they chose the particular size and shape of the dish and of the dome.
If you look at the slide, you can see it’s a rectangular dome about an inch and a half high and roughly half a foot square. No one knows why we’re working with calcium chloride instead of some other sort of desiccant. So the test really has no scientific pedigree. There’s no research that we can find that indicates why the original test conditions were established.
In 1960, as you see on this next slide, in Kentile’s instructions, they said a couple of interesting things. One was that there has to be a moisture barrier underneath the slab, and that makes sense because if you’re testing the moisture condition in the concrete slab, that test is only going to be as useful as it is at the time you take the test because you don’t know if the moisture picture will change. If there’s a vapor retarder underneath the slab, then there will be no communication or ingress of moisture from below the slab. So that was a requirement for the test right from the beginning that there be a vapor barrier or moisture retarder under the slab.
The second thing you see at the bottom of this slide is that the original number put to the test result was 2 pounds of moisture per 1,000 square feet for 24 hours. Where that number came from is anybody’s guess. We think we have some idea that if you test a floor that’s old and as dry as it can be at 50% relative humidity, you typically will get a number somewhere around 2 or 3 pounds. But again, there was nothing we could find in the scientific literature indicating any testing had been done to demonstrate that adhesives or floor coverings work at 2 pounds very specifically. This was simply something that the manufacturer of the test kit first published back about, almost 50 years ago.
On this next slide, you see some laboratory tests that we did here at our facility, CTLGroup in Skokie, Illinois. We asked the question a few years ago about how accurate the test actually is. Because there was nothing published in the scientific literature, we thought that it would be a good idea to try to determine the absolute accuracy of the test method.
So, as you see in this picture, we took a series of one inch thick concrete slabs made at a water/cement ratio that permitted a high rate of moisture passage. We put them over stainless steel pans as you see in the picture. We allowed them to sit for many weeks or even months until they reached a steady state. That is, if we weigh them day after day, they’re losing a constant rate of moisture from the water in the pans underneath the slab.
So, you’ve got pans filled with water, a concrete slab over the top, and every day some of that water is evaporating through the concrete and out the top. If you weigh them every day, you can calculate the true moisture vapor emission rate. We made these as standards and then put chloride kits on the surface to see if the chloride kits measured the true actual amount of moisture coming through accurately.
On the next slide, you see those results. The vertical axis on the left- hand side of the graph shows the apparent moisture vapor emission rate as measured by the kit. The horizontal axis down at the bottom of the page shows the true vapor emission rate as measured by the ASTM E96 test procedure. On the left-hand vertical axis, you’ve got 0, 2, 4, 6, 8 pounds of moisture, and the same scale down at the bottom.
You see the points plotted in that graph generally tend from the lower left-hand corner to the upper right. So it seems like that the chloride kits are generally giving the results similar to what’s coming out of the pan. That is at a low emission level, a true low emission level, the chloride kits are giving us a low number. If you look over at the right-hand side, a higher emission level is giving us a higher number with the chloride kits. That’s at first glance.
If we look at it a little more closely though, and I’m bringing up a diagonal line here. This is the line the points should fall directly on, if the test were actually accurate. The reason is that line represents the one-to-one ratio, where the chloride kit is actually measuring the moisture coming out of the slab. That line represents the line of accuracy. Now you can see that these points really fall considerably far off of the line. In fact, the points down in the lower left-hand corner of the graph all represent an overestimation.
The chloride kits tell us there’s more moisture coming out of this slab than there is at low levels. At the upper end of the graph, the chloride kits underestimate the moisture. So if you have a really wet floor, the chloride kit simply cannot grab or pick up all of that moisture coming out of the concrete. So the test fundamentally doesn’t seem to be terribly accurate.
Now, we could, perhaps, correct the graph and say that there’s a calibration factor to use with the kits. But there’s another thing we saw which was a little more disturbing. If you look at the points in the lower left corner, there are four X’s indicated on the graph, and those are right above the zero on the horizontal axis.
Those were four concrete slabs that actually sat in the room at 50% humidity for several years. When we weighed them day after day, after several years, they were neither emitting moisture or gaining moisture. They were simply at a constant steady state. Very happy, very dry, 50% relative humidity. You could probably put flooring on there with Elmer’s glue and it would work just fine. Those floors have very low moisture and no PH issue.
But when we put the chloride kits on them and measured according to ASTM F1869, we found results ranging from 2.5 to over 4 pounds, and the reason for that is because the chloride kit, the desiccant in the calcium chloride actually sucks out more moisture than is coming out of the concrete. So we had these slabs which were not emitting anything, but when we put the chloride kit on top, it actually sucked out moisture. That just goes to show that if you had a limit of let’s say 3 pounds with some very dry concrete, you’d actually measure 4 and you would think the floor was too wet and you would not install. The reason is simply because the test gives you a false positive result, a high result when there’s really no emission.
Again, going back and looking at all this data, we see the test is fundamentally just not very accurate. Secondly, it actually sucks moisture out of the concrete that’s not really coming out of the concrete. I think, therefore, we’ve got a test which now we know has no scientific pedigree and doesn’t provide fundamentally accurate results. So we begin to question whether or not the test is really useful in understanding how adhesives and floor coverings will perform.