November 15, 2007

� Adhesive breakdown of adhered finish floorcoverings

� Debonding of coatings� Osmotic blisters of epoxy systems

including coatings and epoxy terrazzo� High pH (alkali) attack of floor finishes

� Microbial growths� Flooring expansion, such as cupping of wood

strips or planks� Reactions between incompatible floor

patching/leveling materialsAlthough the interaction of moisture and

floor finishes (adhered floor coverings andcoatings) has gained much attention in recentyears, moisture issues also affect bare concretefloors through the following modes: � Staining� Efflorescence � Expansion of contaminants and popouts

leading to further cracking� Condensation on slab surfaces making haz-

ardous conditions for traffic� Curling

Despite the development of many ex-cellent, longstanding practices, we con-tinue to see a large number of moisture-related problems and hundreds of millionsof dollars are still spent annually in theU.S. to correct such problems. Some con-tributing factors include: design profes-

By Howard Kanare

MOISTUREAND CONCRETE FLOOR SLABS?

WHY ARE WE STILL HAVING PROBLEMS WITH

G o o d p r a c t i c e s a n d e x c e l l e n t p r o d u c t sa r e o n l y t h e b e g i n n i n g t o t h e s o l u t i o n

W ater is an essential ingredient in

concrete, but uncontrolled excessive

moisture can create a whole host

of problems with concrete floor slabs. Some

of the modes of distress include:

sionals still specify a blotter/cushion/subbase layer over a vapor retarder—sixyears after ACI Committee 302 reinstatedits original recommendation for slabs tobe placed directly on the vapor retarder;vapor retarders are not carefully protectedduring construction; concrete floor slabmixes often are made with smaller thandesirable aggregate and mixes are not op-timized for minimum shrinkage; inappro-priate moisture tests still are specified;moisture testing is not always done cor-rectly; some unreasonably moisture-sen-sitive adhesives are still on the market;there is no performance specification ortest method for moisture resistance of ad-hesives or floor coatings; and failures ofsome moisture-suppression products occurbecause there are no performance specifi-cations for moisture suppression products.Let’s look more closely at some of thesecontinuing causes of moisture problemsand how to avoid them.

VAPOR RETARDERSThis author firmly believes that a floor

slab is part of the building envelope andthat every slab on ground should have avapor retarder meeting ASTM F1745 in-stalled directly beneath the concrete. Justas we do not tolerate water leaks in roofsor walls, we should not accept structures

built so that moisture can infiltrate floorslabs. Highly moisture-resistant floors canbe constructed following the principles inACI 302.1R-06, “Guide for Concrete Floorand Slab Construction,” and PCA EB119,“Concrete Floors and Moisture.” A prop-erly selected and installed vapor retarderis essential for long-term moisture resistance.

Vapor retarder material selection andplacement was the subject of an extraordi-nary amount of debate during the 1990s.The original ACI 302-69, “RecommendedPractice for Concrete Floor and Slab Con-struction,” indicated vapor barriers are usedunder the slab when floor coverings, house-hold goods, or equipment must be protectedfrom damage by moist floor conditions.

In the years following, some field expe-rience suggested that concrete placed di-rectly on a low permeability vapor retarderwas more prone to cracking and curling.

In the subsequent ACI 302.1R-89, “Guidefor Concrete Floor and Slab Construction,”the use of a blotter or cushion layer on topof the vapor retarder was first described.Vapor barriers were said to aggravate theproblem of plastic and drying shrinkagecracking and should be avoided if groundmoisture conditions permit. A granular, self-draining fill was recommended to act as ablotter over the vapor barrier.

In the ensuing decade, many floors builtwith this blotter/cushion layer constructiondetail suffered moisture-related failures. Themany reasons for these failures are nowquite well understood.

1. A 6-mil polyethylene sheet placed overgranular subbase and covered with compactedgranular material often sustains many punc-tures, rendering the vapor retarder ineffective.

2. Dampening the blotter/cushion layer tocompact it adds significant water to the floorsystem. For example, 6% moisture in a 3-inch-thick cushion layer adds roughly 11⁄2pounds of water below every square foot ofconcrete. A concrete mix typically containsabout 3 pounds of water per square foot offloor for a 4-inch-thick slab at 0.5 w/c ratiowith 517 pounds of cementitious material (ne-glecting water in aggregate); thus, the mini-mal water in the cushion layer required forcompaction increases the free water in thefloor system by 50%, providing a long-termsource of unwanted moisture.

3. The blotter/cushion layer can be ex-posed to precipitation and become exceed-ingly wet prior to concrete placement.

4. Compacted granular subbase fill typi-cally has 10% to 20% void space. This spaceon top of the vapor retarder can act as a“plenum” for unwanted moisture intrusionover long distances. For example, water infil-tration at the perimeter of the structure cantravel with little resistance in this layer.

5. If the vapor retarder is installed intact,then it functions as a “bathtub” beneath thegranular blotter/cushion layer. Liquid wateror moisture vapor that gets into this layer afterconstruction (e.g., pipe leaks, perimeter infil-tration) cannot easily drain outward throughthe vapor retarder and therefore infiltrates andrecharges the slab from below.

The commonly cited advantages of ablotter/cushion layer (less bleeding, shortertime to trowel, less plastic shrinkage crack-ing, reduced curling) can be achieved just aswell by using well-designed mixes with prop-erly graded aggregates and following ACI rec-ommended hot weather concreting practiceswhen necessary.

In April 2001, ACI Committee 302 pub-lished an update to this guideline, revertingto early ACI and PCA practice, recommend-ing that concrete slabs be placed directly onvapor retarders if the slab will receive vaporsensitive floor covering. It is this author’s opin-ion that all floor slabs should be constructeddirectly on an appropriate vapor retarder forthe following reasons:

Modern electronic thumb-size relative humidity probes now are available that can be placed below the sur-face of the concrete slab to monitor drying without interfering with construction or use of the floor. Thedevices provide real-time relative humidity and temperature at the push of a button. Photo: Howard Kanare

2 � Concrete Construction � Floors 2007 www.concreteconstruction.net

Concrete Construction � Floors 2007 � 3www.concreteconstruction.net

1. All finish flooring systems (wood,textile, resilient, hard surfaces, and poly-mer coatings) have some degree of mois-ture sensitivity.

2. Bare concrete floors can suffer fromefflorescence and moisture accumulationthat at least requires unanticipated main-tenance and at worst may create slippery,hazardous conditions.

3. Omission of a vapor retarder permitscontinual movement of moisture into theoccupied space of a building, increasing rel-ative humidity.

4. Floor finishes are routinely changedduring the life of a building. Even if theinitial intention is to have a highly mois-ture tolerant floor finish in some areas (e.g.,ceramic tile in cementitious thinset withcement grout), at some point in the fu-ture a more moisture-sensitive flooringproduct may be installed in that area.

5. Adaptive reuse of buildings, whichwere initially designed not to receive floorcoverings (e.g., warehouses, big box retailstores), often requires a completely differ-ent and more moisture-sensitive floor cov-ering to be installed for the new use.

6. Tenant buildouts of areas not orig-inally intended for occupancy with finishflooring do receive moisture-sensitive fin-ishes.

When concrete is placed directly on avapor retarder, intimate contact developsbetween the top of the vapor retarder andthe underside of the concrete slab. Mois-ture does not move horizontally betweenthe concrete and the vapor retarder. Minorpunctures or gaps in the vapor retardermay produce only localized effects insteadof widespread moisture movement into ablotter/cushion layer. The vapor retarderdirectly under the concrete also acts to re-duce friction at the underside of the slab,reducing the occurrence of random cracksdue to restraint.

Vapor retarders should be specified tomeet ASTM E1745-97 (2004), “StandardSpecification for Water Vapor RetardersUsed in Contact with Soil or Granular FillUnder Concrete Slabs.” The PerformanceCriteria and Test Method table shows thethree classes of vapor retarder recognizedin the E1745. An ordinary 10-mil poly-ethylene sheet will meet the permeancerequirements of the standard but may ormay not meet the tensile strength andpuncture resistance requirements. Mostordinary low-density polyethylene sheets

aren’t certified to the standard. Designprofessionals should include a require-ment in project documents that vapor re-tarders meet this standard. A number ofproducts have been introduced in recentyears to meet even the most stringent re-quirements of this standard.

In addition to specifications for the ma-terials and placement of a vapor retarder,design details must be provided to ensurethe integrity of the vapor retarder fromwall to wall and around structural elements,such as columns and footings, and at allother penetrations.

MOISTURE TESTINGIn recent years there has been a renewed

interest and emphasis on moisture testingfor concrete floor slabs. However, a minorproportion of all floors that should be testedactually are being tested before installationor application of floor finishes. Hopefully,new technologies will make floor moisturetesting more economical and attractive tocontractors and facility owners.

Electronic moisture meters have beenavailable for several decades. They oper-ate on principles of electronic capacitanceor resistance and do not provide an ab-solutely accurate measure of the moisturecondition of a concrete floor slab. Theyare useful as survey tools to evaluate com-parative differences in moisture distribu-tion across the floor. Most of these in-struments are designed to read the upperapproximately 1 inch of a concrete slab,avoiding interference from steel mesh orreinforcement within the slab. Therefore,these meters cannot indicate the moisturecondition deeper in the slab.

The old plastic sheet test, ASTM D4263,has been found to be unreliable becauseit depends upon temperature and dew pointat the concrete’s surface. If the test is pos-itive (dampness under the plastic sheet),then the floor is likely too damp to pro-ceed with flooring or coating installation.However, a negative result under the plas-tic sheet does not necessarily indicate thatthe floor is sufficiently dry.

A major change is taking place in theway that the floor covering industry meas-ures moisture in concrete floors. This changeaffects concrete floor contractors, generalcontractors, construction managers, own-ers, architects, and floor covering installers.Just as the concrete floor construction in-dustry made the shift in flatness tolerances

from straightedge to the more appropriate,statistically based FF/FL system, now thefloor covering industry is facing a majorshift from measuring moisture vapor emis-sion rates (MVER) to measuring relativehumidity (RH).

The anhydrous calcium chloride test forMVER was developed in the 1940s as aqualitative evaluation of floor moisture con-dition. Without any documented scientificbasis, it became a quantitative test in the1960s. Now, nearly 500,000 MVER testsare performed each year in the U.S. In thepast decade, we have learned that the testcan be unreliable, capable of producingboth false high and low results, and de-pendent on ambient temperature and hu-midity, water-cement ratio, use of light-weight aggregate, the presence of curingcompound, how hard a floor was trow-eled, and how the test site is prepared.

Over the past 10 years, investigations ofthe MVER test method in the field and in thelab have found that it suffers from several se-rious deficiencies:� The test has no pedigree: there are no

published or existing data from the 1940sto 1960s that were used to establish thetest kit dimensions, time of exposure,choice and mass of desiccant, or basisfor calculations. There are no flooringperformance data supporting the widelyused 3-pound limit publicized in the 1960s(actually 2 pounds in the earliest printedversions), or the 5-pound limit for someproducts published by flooring manufac-turers in the late 1990s.

� There is no practical way to calibrateMVER test kits. There are no standardreference concretes available with con-trolled MVER levels. Kit dimensionshave been arbitrarily standardized toprovide reproducibility between brands,but they are not absolutely “accurate.”

� The test determines a portion of the freemoisture near the surface of a slab, gen-erally the upper 12 to 20 mm (1⁄2 to 3⁄4inches), providing no information aboutmoisture conditions deeper in the slab.

� The test does not accurately determinethe true MVER of concrete; it overes-timates low moisture emission levels andunderestimates high emission levels.

� Ambient conditions interfere with test re-sults—warmer, more humid room air canyield higher MVER results even if the in-ternal concrete moisture condition is un-changed. Floor surface preparation for

Posted with permission from the November 15, 2007 issue of CONCRETE CONSTRUCTION magazine. © Hanley Wood.#1-23091613 Managed by The YGS Group, 717.399.1900. For more information visit www.theYGSgroup.com/reprints.

For more information on relative humidity testing go to www.rapidrh.com

testing, such as gentle grinding, can changemeasured MVER drastically.

� Limits set for MVER based on prod-uct type—one level for sheet vinyl, car-pet, or rubber, and a different level forVCT and felt-back resilient flooring—neglect the fundamental fact that ad-hesives play a major role in flooringperformance.Relative humidity for floor moisture

measurement is not new—it was first usedto measure moisture in concrete floors aspart of Portland Cement Association’s(PCA) applied research programs in CTL-Group’s laboratories beginning in the1950s. RH instruments can be independ-ently calibrated directly traceable to na-tional standards. There are a variety ofcommercially available RH instrumentsspecifically for measuring moisture in con-crete. Although RH instruments costmore than a calcium chloride MVER kit,large savings in testing time and laborare making RH the method of choice.More importantly, RH testing gives amuch more useful picture of the actualmoisture condition within the concreteregardless of mix, aggregate types, floorthickness, or surface conditions. Prop-erly conducted RH testing can preventpremature flooring installation that canlead to costly repairs and litigation.

An all-too-common case in point: Sev-eral months after it was installed, sheetvinyl in a large, new hospital in Chicagowas badly rutted and bubbled. Floors weretested before installation using calciumchloride kits and found to have about 4lb/1000sq ft/24 hr MVER, below the floor-ing manufacturer’s 5-pound limit for floorcovering installation. However, after theflooring failure, RH tests indicated greaterthan 95% RH, well above the generallyaccepted 75% RH specified in ASTMF710. If RH testing had been done priorto installation, remedial action could havebeen taken to avoid the failure and the

expensive, disruptive repairs that must bedone throughout this facility. However,the story is more complicated—some typesof flooring in the building are perform-ing fine while others are not. Two typesof sheet vinyl exhibit failure while vinyltiles and a third type of sheet vinyl areperforming well. These observations in-dicate that different adhesives have vary-ing moisture sensitivity and are critical tofloor covering performance. Acceptablemoisture levels should not be specifiedby flooring type (carpet, tile, or sheet)but should be based on actual systemperformance. CTLGroup has developedand implemented methods to evaluateperformance of flooring system compo-nents at various, precisely reproduciblemoisture conditions.

RH has been the preferred method forassessing concrete floor moisture condi-tions in several countries for many years.Limits for installation range from 60% RHfor direct glue-down wood parquet to 90%RH for some vinyl tile products. Flooringmanufacturers in the U.S. must establishrealistic RH limits for acceptable perform-ance of their products through rigorousscientific testing, taking into account thevarious components of their systems, suchas patching/leveling compounds, primer,adhesive, and floor coverings. Manufac-turers will probably develop “tiered sys-tems” that will allow design professionalsand contractors to select flooring productsfor various moisture levels to produce en-during, successful outcomes.

REMEDIATIONIf a floor is not dry enough when it is

time to apply an adhered floor covering orcoating, the owner and contractor haveseveral options ranging from considerationof alternate finishes all the way to removaland replacement of the floor slab. Whathas become a fairly common approach isto isolate the moisture (and accompanying

high pH) of the concrete floor from the ap-plied floor finish system. Two types of mois-ture barrier remediation systems are gen-erally available: preformed membranes andliquid applied membranes.

Preformed membranes are available fromseveral manufacturers designed for use undervarious types of finish floor systems suchas carpet, wood flooring, resilient sheet,and hard surface tile. These products havevery low moisture vapor transmission andtherefore cause moisture to accumulatewithin the concrete floor slab below themembrane. This accumulation of moistureshould not present a problem as long asthe membrane is intact and sealed. Thereare no performance standards or prescrip-tion specifications for these products, sousers must look carefully at their physicalproperties and compatibility with the in-tended finish floor system.

Liquid applied topical moisture vaporsuppression systems are available in a widevariety of products including silicates,acrylics, epoxies, cementitious, and othermaterials. The products that provide thebest resistance to water vapor transmissiongenerally are high cross-link density, 100%solids two-part epoxies; some of these prod-ucts when applied at 12 to 15 mils havewater vapor transmission rates as low as10 mg·hr1·m2, roughly as good as a 6-milpolyethylene vapor retarder sheet belowthe slab. Silicates may create a water-repel-lent surface, but they do little to suppressmoisture vapor transmission. At this time,there are no performance standards or pre-scription specification requirements for theseproducts. Manufacturers of some of theseproducts also provide moisture-resistantadhesives to use in combination with theliquid membranes.

— Howard Kanare is a senior principal scientist at Construction Technology Laboratories (CTL) Group,Skokie, Ill., 847-522-2285 or hkanare@ctlgroup.com.