There's something inherently strange about laying wood floors over concrete. After all, water and wood don't mix, and one of concrete's talents is absorbing and holding water. For contractors in many geographic areas, though, that's what they're faced with everyday on almost every job. First and foremost, they have to learn how to handle moisture. Beyond that, there are other challenges that contractors installing over basements usually don't have to face.
To illustrate the complexities of working over concrete, let's follow the misadventures of Buford, our fictitious wood flooring contractor. After a long, cold, snowy winter, Buford decided he needed a touch of the tropics. He abandoned his business in Iowa to start over in Florida's sunny skies. Without educating himself about the intricacies of working on slabs, however, Buford finds that it's a jungle out there.
Problem #1
Shortly after moving south, Buford lucked out and landed his first Florida job. A builder needed a floor done in a new home immediately, and Buford was available. Buford installed a plywood subfloor over the slab, and then nailed down a solid strip wood floor. One month later, the floor blew up. It cupped throughout the installation and even buckled in some areas. What was Buford's mistake?
It turns out that Buford neglected to test the slab, which was only 50 days old, for moisture before he laid the subfloor. Standard practice calls for doing a moisture test on the slab before any work is started. If the moisture level is found to be unacceptable and the slab is new, waiting for it to cure maybe all that's necessary. If the slab is existing and the moisture level is too high, there may be other moisture sources that have to be eliminated. There are also products on the market that claim to help a slab cure (if you are going to use an adhesive, make sure whatever you put on the slab is compatible with it).
Problem #2
Buford learned from his costly mistake of not testing slabs for moisture before installing wood flooring. A few weeks later, he got a call to do another solid strip job, this time on an addition in an older neighborhood. For this job, he diligently tested the slab, found it to beat acceptable moisture levels, and went ahead with the installation. This time, it was seven months before the floor showed signs of cupping. What happened now?
Buford was mystified by this one. He knew the builder had installed a moisture barrier before he poured the slab (according to code), and he thought he had followed all the right directions. Luckily, Buford had started working with a good distributor, who came out to the job site to help him inspect the floor. After looking the floor over, he went outside and walked around the yard. Landscaping around the house was below the level of the rest of the yard, and to match the existing part of the house, there were no gutters. It was now the rainy season, and from the looks of things, water would drain toward the house. The distributor reminded him that it is the responsibility of the wood flooring contractor to look at the lay of the land to spot such potential problems.
Even with a solid foundation and the protection of moisture barriers, a slab that is a gathering place for runoff will likely cause moisture problems. Slabs are like concrete sponges—they absorb and hold water. Moisture can migrate through concrete block into a slab edge, even without a visible path.
For this job, they had to find someway to divert the moisture. The best solution proved to be installing gutters on the house and a French drain system in the yard. Together they successfully drained the water into the ground far enough away from the house, and eventually, the wood floor flattened.
Problem #3
Although Buford thought he had learned his lesson about moisture and cupped floors, he encountered a similar problem a few weeks later. This time he had glued down a 3/8-inch engineered, factory-finished strip floor over a slab. The slab's moisture level was acceptable and the grading of the yard looked OK, as well. Unfortunately, the customer called the day after installation to say that the edges of the boards looked slightly raised.
As it turns out, Buford gambled and lost when he installed this floor. When he got to the job site to install it, he realized the adhesive his distributor told him to use with that particular flooring was still back in his garage. Rather than drive all the way back home, he grabbed some water-based adhesive he had in his truck from the previous week—all adhesives are pretty much the same, he figured.
Obviously, when Buford chose to save some time by using the wrong adhesive, he ended up with a bigger waste of time. He had to go back and totally reinstall the floor, all at his expense.
Adhesives vary greatly in their chemical composition and their purpose. There are many reasons why contractors use the wrong adhesives—failure to read directions, laziness and trying to save money are the most common—but in the end, none of them are worth the risk of a failed job.
Problem #4
Buford's really wised up now. He's learned about moisture and adhesives, but apparently he's got a ways to go. On his next engineered factory-finished job, he made absolutely sure he had the right adhesive. Yet again, however, he ended up having to speak with an unhappy customer. The home owner called to say the floor was making creaking noises throughout the installation. What did Buford do wrong now?
He forgot one little detail—the trowel.
The recommended trowel, that is. He grabbed a trowel in his truck and figured that as long as it had some notches, it would suffice. Buford may seem a few sandwiches short of a picnic, but contractors make the same mistake every day. What is so important about the way the notches are cut on the trowel? They are critical to achieving the right spread rate for the adhesive. Without the right trowel, there will be too much or too little glue.
In this case, Buford was left without enough glue. The thinner spread not only shorted the adhesive, it also shortened the working time. The result was movement in the floor and objectionable creaking noises. Even if he never noticed that he had the wrong trowel, there are a couple of things that Buford could have done to help him notice a problem. First, he could have periodically stopped, lifted up a piece of flooring, and checked the transfer of the glue. Another would have been to read the adhesive can for the recommended spread rate. If the directions say a gallon of adhesive should cover 60 to 65 square feet and you're starting a second bedroom with the same can, that's a warning flag that something's wrong.
Problem #5
Just when Buford thinks he's got the noisy floor problem fixed, it happens again. It's the same type of job—a glue down engineered floor—but this time the customer calls a few weeks later and says that there are just a few areas of the floor that are noisy, but they are driving her crazy. This time Buford used the right glue and the recommended trowel. What was the problem?
Inspection of the floor proved that Buford didn't compensate for some unacceptable variations in the slab. That left a few "hollow" spots in the wood floor that were causing the noises.
Slabs (like any subfloor) should be checked for flatness before installation begins. NWFA Installation Guidelines say that any variation greater than 3/16 inch in 10 feet or 1 in 6 feet is unacceptable. If larger variations are found, there are several ways to flatten the floor. One of the easiest is to use a self-leveling underlayment. Such products can be mixed in a bucket on the job site, spread on the floor and they find their own level. Another common way to flatten the floor is to use an old buffer to grind off high spots.
Problem #6
It doesn't take Buford long before he's in trouble yet again. It's another solid strip floor. He checked the slab for moisture, installed a vapor retarder and a layer of plywood over the concrete, then went ahead and nailed down the floor. Now this floor's making noises, too. Buford thinks that if he has one more noisy floor, he's getting out of the business.
A look underneath the wood flooring reveals Buford's problem: There were two steel prehung door units in place, and1-inch plywood was the maximum thickness he could use to allow the doors to swing open. Additionally, since the 1/2-inch plywood laid flat, he used only nine fasteners per piece.
Using the wrong thickness of subfloor can make it difficult for the wood flooring fasteners to hold well, and using too few fasteners on the subfloor can allow too much movement. Both problems can contribute to squeaks or other noises in the floor.
For a nailed-down subfloor over concrete, NWFA Installation Guidelines call for a minimum of 4-by-8-foot sheets of 5/8-inch CDX-grade plywood, or, preferably, 4-by-8-foot sheets of 3/4-inch CDX-grade plywood. They should be spaced 1/8 to 1/4 inch apart with staggered joints and fastened every 12 inches on the border and at 12 inches on center per sheet (a total of 32 shots per sheet; higher humidity areas may require more shots). An expansion space of 3/4 inch should be left at vertical obstructions.
Problem #7
On the next job, Buford had to gluedown finger block parquet in a new house. He checked the slab for moisture, filled in a few low spots and glued down the floor using the recommended trowel and adhesive. When he went back to sand, while he was edging he noticed that some of the areas around the perimeter seemed loose. He knew he had the correct spread rate and good transfer. What did Buford do?
Pulling up a few of the loose pieces revealed that Buford had misunderstood the concept of a "clean" slab. He had swept it clean, but he had neglected to deal with the areas of latex paint left by the painters, who had sprayed the room. The same problem can occur with joint compound. The adhesive may bond well with the paint or drywall, but the paint or joint compound pulls up off the slab along with the flooring. Any contaminated areas of the slab need to either be scraped clean or ground off.
Problem #8
By this time, word is getting around that Buford isn't the brightest bulb, and jobs don't seem to be coming his way. To bring in some money, he has to take a few jobs with a local wood flooring contractor who does commercial work. Unfortunately, it isn't too many weeks before Buford gets sent out to start a glue-down job alone. The moisture level was OK and the slab was flat enough, so he went ahead and started laying flooring. By the time his boss got to the job later that morning and saw the slab, Buford already had a good start, but his boss told him they would have to rip it out and start all over again. Why did he say that?
Buford's boss recognized that the slab was slick and shiny. A couple of drops of water on the slab beaded up. In order for the wood flooring contractor to guarantee the job, it was the company's standard practice to mechanically abrade such floors to make sure the adhesive would stick to the slab.
Buford's boss also reminded him of a common peril with commercial slabs—incompatible sealers. On commercial jobs, raw slabs are frequently treated with sealers, which may cause adhesion problems. To test for an incompatible product on the slab, an easy method is to put a few drops of water on the slab. Because concrete likes water, the drops should absorb into the slab. If they bead up on top of the slab, it is probably contaminated. The slab then has to be mechanically abraded to allow for proper adhesion when using adhesive.
Hopefully most contractors take a little more initiative than Buford to educate themselves about the types of jobs they're doing and the risks involved, especially when working over concrete. Those who don't may get away with taking shortcuts job after job, but sooner or later, their laziness or ignorance will catch up to them, and they'll be no better off than the pathetic and ill-fated Buford, who might be found now flipping pancakes at the Waffle House.
