What FF20/FL15 Means: Flatness vs. Levelness, Defined

FF20/FL15 is a two-part statistical tolerance for concrete slab profile set by the American Concrete Institute. FF measures flatness, the short-range waviness of a slab surface over 24-inch increments. FL measures levelness, the longer-range slope or tilt of a slab over 10-foot increments. Together they describe how a floor behaves, not just how it looks.

The split matters because a slab can be locally smooth and still slope badly, or it can hold a consistent grade and still ripple underfoot. ACI’s Floor Flatness Report separates the two because they fail differently and get fixed differently: flatness problems show up as bumps and dips that rattle wheels, while levelness problems show up as drainage issues, door-swing clearance problems, or shelving that leans. A slab with an FF20 but a weak FL15 could pass a flatness check and still pool water at a loading dock.

The numbers themselves come from ASTM E1155, the governing test method for determining FF and FL from a defined elevation-measurement plan. Higher numbers mean tighter tolerance. An FF50 slab is dramatically flatter than an FF20 slab under the same statistical method.

Where ACI Places FF20/FL15 in the Classification Scale

ACI organizes random-traffic floor tolerances into five tiers, from conventional to superflat. FF20/FL15 sits at the bottom of that scale, labeled “conventional,” and is intended for noncritical, low-demand applications rather than as a default spec for any commercial slab.

Note where moderate-to-heavy-traffic warehouses actually land: FF35/FL25, two full tiers above conventional. That gap is the source of most of the misapplication problems described below.

Appropriate Uses for Conventional Flatness

FF20/FL15 is appropriate for spaces where surface irregularity has no functional consequence: mechanical and electrical rooms, back-of-house service corridors, nonpublic storage, parking structures, slabs beneath raised-access flooring, and areas scheduled to receive thick-set tile. These uses share one trait: nothing on the floor is sensitive to minor waviness.

ACI’s guidance and corroborating industry sources, including Mecanocaucho’s Recommendations for Floating Floors, place conventional tolerance squarely in this low-demand category. A thick-set tile installation, for example, uses a mortar bed that itself corrects minor substrate variation, so the slab underneath doesn’t need to do that work. Parking structures tolerate conventional tolerance because vehicle suspension absorbs the waviness that would rattle a narrow-aisle forklift mast.

The common thread across every legitimate FF20/FL15 use case is that traffic is light, random, or forgiving, and the finished surface above the slab, or the slab itself, isn’t performing a precision function. Once any of those conditions changes, conventional tolerance stops being safe.

Flooring Systems That Work, and Don’t Work, Over FF20/FL15

Conventional tolerance works under finishes that mask or absorb substrate irregularity: thick-set tile, carpet tile over cushioned backing, and raised-access flooring systems designed with adjustable pedestals that compensate for slab variation. These assemblies have built-in tolerance for a wavy substrate.

It does not work reliably under thin-set flooring, large-format tile, polished concrete, resinous coatings, or adhesive-applied resilient flooring. These finishes telegraph substrate irregularity directly to the walking or rolling surface, because there’s little to no leveling layer between the slab and the finish. A thin-set installation over an FF20 slab will typically show every dip and crown in the concrete once the adhesive cures.

Flooring manufacturers publish written substrate requirements for exactly this reason, and those requirements should control over any generic F-number target. Beyond flatness, thin applied finishes often carry their own moisture-vapor and crack-treatment prerequisites that an F-number alone doesn’t address; readers evaluating substrate readiness for coatings or resilient flooring should also account for moisture testing requirements, a separate but related qualification step covered in IFTI’s comparison of ASTM F2170 and F1869 moisture test methods.

Why FF and FL Matter Separately (and Why “Flat” Isn’t Enough)

A slab specified only as “flat” without a matching FL requirement can still have unacceptable slope, drainage problems, or racking-alignment issues, because flatness and levelness fail independently and a single number can’t capture both.

FF describes curvature over short spans, the kind that affects how smoothly a wheel rolls across the surface. FL describes elevation change over longer spans, the kind that affects whether a floor drains correctly, whether a long run of pallet rack stays plumb, and whether a door at the end of a corridor still clears the frame. A slab can register an acceptable FF number while its FL number reveals a slope problem invisible in the short-range data, and the reverse is just as possible.

This is also why ACI publishes FF and FL as a pair in its classification table rather than a single composite score. Specifying only one number, or using a straightedge check like “1/8 inch in 10 feet” as a substitute, leaves half the slab’s behavior unverified. Steel King’s technical guidance on concrete floor flatness makes this point directly: a straightedge tolerance does not replace a complete FF/FL requirement, because it doesn’t produce the statistical data needed to evaluate the whole floor.

When an Owner Should Specify Higher Than FF20/FL15

An owner should move up from conventional tolerance whenever the space involves narrow-aisle lift trucks, robotics or automated guided vehicles, thin-applied flooring finishes, repeated forklift traffic, or any equipment with tight dimensional clearance requirements. In most of these cases, FF25/FL20 or FF35/FL25 is the realistic minimum, not FF20/FL15.

Wide-aisle warehouse floors with standard counterbalance forklifts commonly land in the FF25 to FF35 / FL20 to FL25 range, according to IFTI’s Floor Flatness Requirements by Industry Guide and the Concrete Floor Construction Association’s F Number Tolerances. Narrow-aisle and automated operations frequently need tolerances well above that range, or a different measurement approach entirely (covered in the next section). The decision isn’t about building type alone; it’s about equipment, aisle geometry, and finish, assessed together before the slab is poured.

Getting this decision wrong is expensive to reverse. A slab poured to FF20/FL15 that turns out to need FF35/FL25 for its actual rack and lift-truck configuration typically means grinding, shimming, or a leveling overlay after the fact, costs and schedule delays that a correct spec avoids entirely. Readers setting up a new specification from scratch can review IFTI’s guide on how to specify FF/FL floor flatness and levelness correctly before finalizing contract language.

Random-Traffic FF/FL vs. Defined-Traffic Fmin: Why the Distinction Matters

FF/FL is a statistical, random-traffic measurement that describes overall slab profile across a sampled area. Fmin is a defined-traffic measurement that evaluates the floor along the actual travel paths of specific equipment. They answer different questions and aren’t interchangeable.

Random-traffic FF/FL assumes traffic crosses the floor in unpredictable patterns, so the statistics describe the slab as a whole. That assumption breaks down for very-narrow-aisle trucks, wire-guided or rail-guided vehicles, and some automated systems, where equipment travels the same few inches of floor thousands of times a day. The Concrete Floor Construction Association’s F Number Tolerances makes the point plainly: a slab can pass its overall FF/FL requirement and still contain a localized defect sitting directly in a critical forklift wheelpath, because the statistical survey averages across the whole sample area rather than isolating that one path.

For VNA, high-bay, wire-guided, rail-guided, robotic, or automated facilities, the Concrete Floor Construction Association and MHI’s Applied Industrial Flooring both recommend establishing requirements jointly with the rack supplier, vehicle manufacturer, automation integrator, flooring installer, and concrete contractor before bidding. Depending on the equipment, solutions may include a higher FF/FL target, an Fmin specification, narrow strip pours, specialized screeding, or post-placement grinding. Deciding this after the slab is poured is far more expensive than deciding it during design, a sequencing problem addressed in IFTI’s FF/FL Floor Flatness Specs: Owner’s Pre-Construction Control Guide.

Measurement, Timing, and Acceptance: Getting FF20/FL15 Verified Correctly

FF and FL values are determined using ASTM E1155, which defines the elevation-measurement plan, including test locations, sample panel size, overall and local requirements, the testing window, exclusions near joints and penetrations, and responsibility for surveying and any corrective work. Verification should happen on a defined schedule, not whenever it’s convenient.

Conspectus’ Floor Slab Flatness & Levelness, Tech Tips A4010, recommends measuring within 72 hours of slab placement when the goal is to evaluate the original as-built profile. Beyond that window, curling, shrinkage, ongoing construction traffic, and environmental exposure start changing the slab’s surface, so a later measurement reflects conditions that have already shifted rather than the as-placed profile the contract is actually judging.

F-numbers are statistical measurements of slab profile, not a guarantee that every point on the floor meets a uniform deviation from a theoretical plane. That distinction matters contractually: specifications should separately define overall and local tolerances, the test method, the testing timeline, and the acceptance area, rather than relying on a single F-number pair to carry the entire requirement, as Conspectus’ guidance emphasizes. Readers who want the underlying math behind FF and FL calculations can review IFTI’s explainer on how FF/FL numbers are calculated for the full measurement logic.