How to Run an In-Situ RH Test on a Slab

Why this matters

A glue-down floor on a slab fails from underneath. Moisture moving up through the concrete breaks down the adhesive, pushes up plank edges, and grows a smell under resilient flooring, usually months after the crew has been paid. The adhesive and flooring makers both publish a moisture limit, and most of them state it as relative humidity measured inside the slab by ASTM F2170. If the test was not run, or was run in a way the standard does not recognize, the maker's warranty is gone before anyone looks at the floor.

Which test to use, how calcium chloride (ASTM F1869) compares, and what the numbers mean belong to 'Slab Moisture Testing for Flooring'. This card is the procedure for an F2170 test that will stand up when someone asks for it.

What the test measures, and why depth matters

A slab dries from its exposed surface. Near the top it reads drier than the concrete below, and once a floor and adhesive cover it, moisture redistributes and the top comes up toward the deeper reading. F2170 reads RH at a depth chosen to predict that covered-up condition: 40 percent of slab thickness for a slab drying from the top only (on grade, below grade, or on metal deck), and 20 percent for a slab drying from both top and bottom (an elevated slab open to air below). A hole drilled too shallow reads the dry skin and passes a slab that will fail.

  sleeve cap, sealed       room air 70 F / 45 pct RH
        |                  service conditions 48 hr+
 =======[]=================================== top of slab
 |      ||                                  |
 |      ||  sleeve sealed to hole wall,     |  slab
 |      ||  open only at the bottom         |  4.0 in
 |      ||                                  |  thick
 |      \/ sensor reads here, 1.6 in down   |
 |         1.6 in = 40 pct of 4.0 in        |
 |__________________________________________|  bottom
  vapor retarder, then soil: dries from top only

The sleeve isolates the bottom of the hole. If the sleeve is not sealed to the hole wall, the sensor reads the dry concrete near the surface along with the bottom, and the result is low.

Before you start

• Equipment. RH probes or leave-in sensors made for F2170 with calibration verified on the schedule the current edition of the standard requires, sleeves, caps and seals for that system, a rotary hammer with a bit of the diameter the sensor maker specifies and a depth stop, a drill shroud on a HEPA dust extractor, a HEPA vacuum and hole brush, a thermo-hygrometer, and the adhesive and flooring makers' data sheets with their limits.
• Conditions. The slab and the air above it at service conditions, at the temperature and humidity the building will run at, for at least 48 hours before the test. In a new build that means the HVAC permanently running, not a portable heater.
• Slab thickness. From the plans, a core, or a hole drilled at an inconspicuous spot and measured. Without the thickness you cannot set the depth.

Step 1: Confirm and record the conditions

Record the air temperature and RH over the slab, and when the HVAC was set to service conditions. If the space has not been at service conditions for 48 hours, the test does not start. What breaks if you skip it: a slab tested cold, or with the building open, reads a number that describes the construction phase, not the floor's future. That result is neither a pass nor a fail, it is invalid, and the maker will treat it that way.

Step 2: Plan the number and the locations

F2170 calls for three tests in the first 1,000 sq ft and one more for each additional 1,000 sq ft. On a slab on or below grade, the standard also, in substance, asks for a test location near each exterior wall (check the current edition for the distance). Fill the rest of the count where moisture comes from: low spots, near plumbing in the slab, along a wall with poor drainage outside, and the middle of the floor. Mark each on a floor plan with a number.

What breaks if you skip it: a slab tested in three comfortable interior spots averages dry and the one wet edge carries the failure.

Step 3: Clear the slab, then drill dry

Before any hole, rule out what is inside the slab. Look for a post-tensioned slab stamp or notice (common in garages and on the slab edge in some regions); drilling into a stressed tendon can release it violently, and a post-tensioned slab gets no hole until the tendons are located by scanning or the engineer of record clears the location. Ask about radiant tubing and conduit in the slab, and scan or pick locations clear of them.

Drill with a rotary hammer, no water, to the depth the drying condition sets (40 or 20 percent of slab thickness), using the depth stop. Water would wet the concrete you are about to measure. Drilling concrete makes respirable silica: run the drill with a shroud on a HEPA dust extractor and clean the hole with a brush and HEPA vacuum, never compressed air; the controls belong to 'Silica Dust in Flooring Work'. Wear eye protection for the bit and chips.

Step 4: Sleeve, seal and cap

Insert the sleeve to the bottom of the hole, seal it to the hole wall at the top per the sensor maker, and cap it. Record the time each hole was finished. What breaks if you skip it: an unsealed sleeve reads surface concrete, and an uncapped hole equalizes with room air. Both read low.

Step 5: Wait the equilibration time, then read

The concrete around a fresh hole is disturbed by drilling, and the air in the hole has to come to equilibrium with the slab. Read only after the equilibration time the current edition of F2170 sets (24 hours in recent editions; older editions required longer) and after the sensor has conditioned as its maker directs. Record RH and temperature at depth for each hole, and the room air again.

What breaks if you skip it: an early reading is still moving, and depending on what drilling did to the hole it can read high or low. A number taken early is not an F2170 result.

Step 6: Compare each test to the limit

Compare each test to the limit on the adhesive's and the flooring's data sheets, and use the lower of the two. Do not average across holes; the limit is about the worst place on the floor, because that is where the failure starts. Where a maker publishes a higher limit with its own moisture-control system, that is a separate product decision, not a looser reading of the same test.

Step 7: Close the holes and report

Remove sleeves the system does not leave in place, and fill the holes with a cement-based patch the adhesive maker accepts, flush and flat. The report carries: slab thickness and drying condition, hole depth, location plan, the time conditions started, drill and read times, RH and temperature at depth for each hole, room air, the limit compared against and its source, and the result.

Worked example: a ranch slab under glue-down engineered wood

A 3,000 sq ft single-story house on a 4 inch slab on grade over a vapor retarder, so it dries from the top only. Glue-down engineered hardwood is going down; the adhesive's data sheet on this job, read for this product, sets a limit of 85 percent RH by F2170 without its moisture-control system, and the flooring maker's limit is the same or higher, so 85 governs.

Depth. 40 percent of 4.0 in is 0.4 x 4.0 = 1.6 in. Checked the other way: 1.6 / 4.0 = 0.40.

Count. Three for the first 1,000 sq ft plus one for each of the next two thousands: 3 + 2 = 5 tests. The house has four exterior walls, so four tests go within the standard's distance of each exterior wall and the fifth goes in the center of the main room.

Conditions. HVAC at 70 F and about 45 percent RH since Monday at 9 a.m.; holes drilled Wednesday at 9 a.m., which is 48 hours later. Read Thursday at 9 a.m., 24 hours after drilling, after the sensors conditioned.

Reading it. Four of five pass. T4 fails, 88 minus 85 leaves 3 points over. Averaging the five would give (78 + 81 + 79 + 88 + 76) / 5 = 402 / 5 = 80.4, a comfortable pass, which is exactly why the average is not the test. The west wall outside has a downspout discharging at the foundation and grade sloping toward the house.

The decision. The slab as tested fails for this adhesive. Two honest paths: apply the adhesive maker's moisture-control system rated for the reading (in this example, its data sheet covers up to a higher limit, so 88 would be within it), or fix the drainage and re-test after the slab dries, which on a slab can take a long time and is not predictable from the reading. The owner picks with the customer, and the choice and the report go in the job file.

What would change it. If this had been an elevated slab open to air below, the depth would be 20 percent, 0.2 x 4.0 = 0.8 in. On a 6 inch slab on grade, 0.4 x 6.0 = 2.4 in. And if the HVAC had been started Tuesday afternoon, Wednesday morning drilling would have come in under 48 hours, and the whole set would be invalid rather than failed.

How to verify the test will stand up

• The 48-hour service-condition start time is recorded and precedes drilling.
• The hole depth matches the drying condition: 1.6 in on this 4 inch slab drying from the top.
• The test count meets the standard for the area (5 here) and includes the exterior-wall locations.
• Reads came after the equilibration time, with RH and temperature at depth recorded per hole.
• Each test was compared to the governing limit on its own, and the holes are patched.

References

• ASTM F2170, Standard Test Method for Determining Relative Humidity in Concrete Floor Slabs Using in situ Probes, current edition (binds the job through the flooring and adhesive makers' warranty terms and the project specification).
• See related: Slab Moisture Testing for Flooring
• See related: Silica Dust in Flooring Work
• Adhesive and flooring makers' data sheets for moisture limits and moisture-control systems.