ESD floor testing to ANSI/ESD S20.20 field guide

What ESD floor testing is

ESD floor testing is the measurement that proves a static-control floor does its job: drains static charge to ground fast enough, and keeps the voltage on a person walking the floor low enough that a discharge into a circuit board never reaches a damaging level. It is two measurements, not one. You measure the electrical resistance of the floor, and you measure the body voltage a person actually builds walking it in their work shoes.

The floor is half of a system, never the whole thing. A floor by itself drains charge to ground through its resistance. A person standing on that floor only drains through it if their footwear connects them to it, so the real protection is the floor and the footwear working together. That is why the program tests the floor alone and then tests the floor with a person on it.

The reading is worthless without the conditions that went with it. Resistance moves with humidity, with temperature, with how clean the surface is, and with the size and weight of the electrode you set on it. A floor that reads in band on a damp morning can read out of band on a dry afternoon. Record the relative humidity and the temperature next to every number, or the number cannot be defended later.

What a static discharge actually does to electronics

Static discharge kills electronics two ways, and only one of them is obvious. A catastrophic failure takes the part out on the spot: the discharge punches through a gate oxide or fuses a junction, and the board is dead at test. You catch that one. The other one is the problem.

A latent failure is the discharge that wounds the part without killing it. The device still works at final test, ships, runs for weeks or months, and then fails in the field for no reason anyone can trace. The energy that caused it left no mark you can find after the fact. Industry estimates have long put a large share of ESD damage in this latent category, which is exactly why you cannot inspect your way out of it on the back end. You control the charge up front or you eat the field returns.

The money is in the trace-back that never happens. A server that resets intermittently, a line card that fails at a customer site, a board that drops out of a burn-in rack: nobody walks that back to the tech who scuffed across a dry floor in the wrong shoes three weeks ago. Data centers, electronics assembly, clean rooms, and defense and aerospace work specify ESD flooring because the discharge that does the damage is the one you never see, and the floor is the cheapest place to stop it.

The two things you actually measure

ESD floor performance comes down to two questions, and a floor has to pass both. First, does the floor drain charge to ground? That is resistance, measured in ohms, and it has to be low enough to bleed charge away but not so low it becomes a shock hazard near energized gear. Second, does a person walking the floor stay at a low voltage? That is walking body voltage, measured in volts, and it is the test of the floor and footwear together with an actual person on it.

Resistance alone does not settle it, and treating it as if it does is the most common shortcut in the trade. A floor can sit comfortably inside its resistance band and still let a person generate hundreds of volts walking across it, because the body voltage depends on the whole system: the floor, the footwear, the person, and how fast charge generates versus how fast it drains. The standard tests both for a reason.

So the program runs resistance per one method and body voltage per another. Resistance per ANSI/ESD STM7.1, on the floor itself. Body voltage per ANSI/ESD STM97.2, with a person in their footwear walking a defined path. Pass the first and fail the second and the floor is not done. The two numbers are not interchangeable, and a clean resistance reading is not permission to skip the walk test.

What floor resistance passes ESD S20.20?

A static-control floor passes when its measured resistance falls inside the band the program specifies, and the widely used ceiling comes from ANSI/ESD S20.20: the flooring system is commonly held below 1.0 x 10^9 ohms, measured per ANSI/ESD STM7.1. Within that ceiling the material splits into two bands. Conductive flooring reads at or below 1.0 x 10^6 ohms. Static-dissipative flooring reads at or above 1.0 x 10^6 ohms and below 1.0 x 10^9 ohms.

Both bands satisfy the S20.20 upper limit, because anything at or below 1.0 x 10^9 ohms will conduct charge away and provide a path to ground. The split between conductive and dissipative is about how fast it drains and about the shock-safety tradeoff, not about whether it qualifies. Some references put the lower edge of the conductive band near 2.5 x 10^4 ohms, since a floor that conducts too freely becomes a personnel-safety concern around energized equipment. Read the actual band off the program.

Verify the band the project's ESD control program requires, because some programs run tighter than the standard ceiling. A program built around very sensitive devices may call for a dissipative target with a hard upper and lower bound, or may require the floor-and-footwear system rather than the floor alone to meet a number. The standard sets the floor of the requirement. The program sets the number you accept against.

Conductive flooring

Reads at or below 1.0 x 10^6 ohms; drains charge fastest, with a lower bound for personnel safety near energized gear

Static-dissipative flooring

Reads above 1.0 x 10^6 up to below 1.0 x 10^9 ohms; drains charge in a controlled, slower way

Ohm

The unit of electrical resistance; ESD floor values run high and are written in scientific notation

Point-to-point and resistance-to-ground

The resistance test reports two values that answer different questions, and the spec usually calls for both. Point-to-point resistance, written Rtt or Rp, is measured between two electrodes set apart on the floor surface. It tells you the floor is uniform, that one tile or one pour drains the same as the one next to it. Resistance-to-ground, written Rtg or Rg, is measured from one electrode on the floor to the floor's bonded ground point. It tells you the floor actually connects to ground.

You need both because they fail independently. A floor can read fine point-to-point, meaning the surface conducts evenly across itself, and still read open to ground because the bond to the grounding system was never made or has corroded loose. A floor that conducts beautifully but is not grounded is a charged plate waiting to discharge. Point-to-point proves the material. Resistance-to-ground proves the connection.

When resistance-to-ground reads high in one zone and fine everywhere else, that is almost never a bad floor. It is a bonding break under that zone, a ground tie that was missed or backed loose. Retest the spot, then chase the bond, not the flooring. Logging both values by coordinate is what lets you tell a material problem from a connection problem six months later.

What is walking body voltage?

Walking body voltage is the peak electrical charge a person builds on their body while walking across the floor in their work footwear, and the common acceptance limit is under 100 V peak, measured per ANSI/ESD STM97.2. A person generates charge with every step through contact and separation between shoe and floor, and the floor-and-footwear system has to bleed that charge away faster than the walking generates it. If it does, the body voltage stays low. If it does not, the charge stacks up step by step.

This is the test that catches what resistance misses. A floor can pass its resistance band and still let a person reach several hundred volts walking it, because the body voltage depends on the footwear and the person as much as the floor. STM97.1 measures the resistance of the floor-and-footwear system with a person standing on it, commonly held below 1.0 x 10^9 ohms. STM97.2 measures the voltage that same person generates walking. The pair is how you prove the system, not just the slab.

Run it with the operators' actual footwear, not a clean reference shoe from the kit. The whole point is the real condition: the heel grounders people forget to replace, the soles that have glazed over, the floor finish that has built up. A walking test passed with pristine lab footwear and failed by the night-shift crew in their own shoes is a test that proved nothing. Test what walks the floor.

Walking body voltage

Peak voltage a person builds walking the floor in their footwear; commonly held under 100 V peak per STM97.2

System resistance (STM97.1)

Resistance of person plus footwear plus floor in series, commonly held below 1.0 x 10^9 ohms

Charge generation

Static built by contact and separation of shoe and floor with each step

How to run the resistance test

Run resistance with the defined electrodes, the right meter, the right voltage, and a grid of points across the whole room, not a spot check at the door. ANSI/ESD STM7.1 specifies a 5 lb (about 2.27 kg) cylindrical electrode. Point-to-point uses two of them set a defined distance apart, commonly 36 in for the floor-uniformity check. Resistance-to-ground uses one electrode on the floor and a lead to the floor's bonded ground point. The weight matters because it sets the contact pressure, and a lighter probe reads higher than the standard one.

Use a compliant megohmmeter that puts out between 10 and 100 VDC and reads across the wide range these floors cover. The convention is to start at 10 V. If the reading comes back at or below about 1.0 x 10^6 ohms, record it at 10 V. If it reads higher, switch to 100 V and record at 100 V. Reading a high-resistance dissipative floor at the wrong voltage is a classic way to get a number that does not match the standard's method, so note the test voltage with every reading.

Lay a grid over the room and test it like a grid. Take readings on a defined pattern of points so coverage is even and so any point can be retested at the same spot later. Record the temperature and the relative humidity at the time of each set of readings, because both move the number. A reading without its RH is a reading you cannot reproduce or defend.

• Confirm the floor is bonded to ground before any resistance-to-ground reading; an ungrounded floor makes Rg meaningless.
• Use the 5 lb (2.27 kg) cylindrical electrodes the method specifies; do not substitute a lighter probe.
• Set point-to-point electrodes the specified distance apart, commonly 36 in, and use one electrode plus the ground lead for resistance-to-ground.
• Start at 10 V; if the reading exceeds about 1.0 x 10^6 ohms, retest and record at 100 V.
• Take readings on a defined grid across the room, not a single spot, and key each to a coordinate.
• Record temperature and relative humidity with every set of readings.

Grounding the floor before the reading means anything

Resistance-to-ground is a measurement to a ground point, so if the floor is not actually bonded to ground, the number is fiction. The floor has to tie into the equipment grounding and bonding system through a real connection: a grounding tab on a conductive tile system, a copper strip or grid embedded in the pour, a bond from the access-floor understructure to the building ground. Confirm that connection exists and is tight before you read Rg, not after the reading comes back wrong.

This is where ESD floor work meets the wider grounding and bonding system of the room, and the two get verified the same way: by measurement, not by sight. A bond that looks made can still read open through a painted contact surface, a corroded lug, or a connection that was snugged but never torqued. Where the spec calls for it, confirm the bond with a low-resistance ohmmeter, the same instrument you would use to prove a ground-resistance and bonding connection anywhere else in the room.

The failure mode is specific and common: the floor material is perfect, the ground bond is open, and the floor reads high to ground in the zone the open bond serves. Crews chase the flooring when the problem is a missing tie. Prove the ground first. Then any high resistance-to-ground reading is the floor's problem, not the bond's, and you are troubleshooting the right thing.

Humidity, wax, and contamination move the reading

Floor resistance is not a fixed property of the material. It moves with the air around it and with whatever is on the surface, which is why a floor can pass clean at install and fail months later with nothing changed but maintenance and weather. Humidity is the big one. A film of moisture lowers surface resistance, so a floor reads lower on a humid day and higher when the air dries out. That is why STM7.1 conditions and tests at controlled low humidity, commonly around 12 percent RH, and often at a higher point near 50 percent RH as well. Testing dry proves the floor works on its own conductivity, not on help from the weather.

Surface contamination is the field killer. The most common own-goal in the trade is waxing or finishing a static-control floor with the wrong product. A standard floor wax or finish is an insulator, and a coat of it lays an insulating skin over a conductive or dissipative floor and drives the reading straight out of band. The floor underneath is fine. The finish on top has disabled it. Dirt, residue, and the wrong cleaning chemical do the same thing more slowly.

So a floor that passed at acceptance and fails in service is usually a maintenance story, not a material defect. Stripping the wrong finish and recleaning with the manufacturer's approved product often brings the reading back. The lesson for the program is to lock the cleaning and maintenance products to what the floor manufacturer approves, because the cheapest way to kill an expensive ESD floor is one pass with the wrong mop bucket.

Qualification, acceptance, and ongoing monitoring

There are three testing moments, and they answer different questions. Qualification testing happens before the floor goes in, on a sample of the material, usually in a lab at controlled humidity, to prove the product can meet the band at all. STM7.1 covers this preinstallation qualification as well as the field methods. It is how a floor gets specified in the first place, and the manufacturer's qualification data is what you check the submittal against.

Acceptance testing happens on the installed floor at turnover. This is the field test: resistance on a grid, resistance-to-ground at the bonds, and the walking body-voltage test where the program requires it, all on the real floor in the real room. This is the reading that says the floor as built meets the program before anyone signs for it, and it is the set of numbers that goes in the acceptance packet by coordinate.

Monitoring, or compliance verification, is the ongoing part, and it is the one programs let slide. An ESD floor is not a one-time pass. The program sets a recurring cadence, often periodic resistance checks on a sample of points, because the floor degrades through wear, finish buildup, and bond corrosion. The exact interval is a program decision, but the principle is fixed: a floor that was qualified and accepted still has to be re-verified on schedule, or you are running on a number that expired.

The floor is half of the system

A floor drains charge to ground through its own resistance whether anyone is on it or not. A person only drains through the floor if their footwear ties them to it. So the protection an operator actually gets is the floor and the footwear in series, and either one out of spec breaks the path. The floor can be perfect and the person can still be a charged hazard in ordinary shoes.

That is why the standard has the floor-and-footwear methods at all. STM97.1 measures the resistance of the whole series path, person plus footwear plus floor, commonly held below 1.0 x 10^9 ohms. STM97.2 measures the body voltage that combination generates while walking. Conductive or dissipative ESD footwear, heel grounders worn correctly on both feet, and an approved floor are what make the system work. One ESD shoe and one regular shoe, or a heel grounder tucked inside the sock instead of under the heel, breaks it.

On the floor, the practical failures are footwear failures more often than flooring failures. Soles glaze and stop conducting. Heel grounders wear through and get pulled instead of replaced. People wear their own shoes on the floor for one task and forget. The floor test proves the slab. The system test proves the slab plus what is standing on it, and that is the number that reflects what actually happens when a tech walks a row.

Conductive vs dissipative flooring: which do I need?

Both meet ANSI/ESD S20.20, so the choice is about how fast the floor drains charge and about safety near energized equipment, not about which one qualifies. Conductive flooring reads at or below 1.0 x 10^6 ohms and drains charge fastest. Static-dissipative flooring reads above 1.0 x 10^6 up to below 1.0 x 10^9 ohms and drains it in a slower, more controlled way. The program picks the band from the sensitivity of the devices handled and the electrical environment of the room.

Faster is not automatically better. A very conductive floor near energized equipment can be a personnel-safety concern, because a low-resistance path to ground under a person standing near live gear is exactly what you do not want in a fault. That is why the conductive band has a practical lower bound and why many electronics and data-center rooms specify dissipative. The dissipative floor drains charge well enough to protect the devices while keeping enough resistance for personnel safety.

The decision is the program's, made against the device sensitivity, the equipment in the room, and any personnel-safety requirement, then written into the spec as a band with an upper and lower limit. Do not pick the metal-fast floor because it sounds safer for the electronics. Pick the band the program calls for and accept the floor against it.

Field example: ESD acceptance in a data hall

Take a data hall with a static-dissipative access floor, bonded to the signal reference grid, with an ESD control program that calls for resistance-to-ground and point-to-point below 1.0 x 10^9 ohms and a walking body-voltage check under 100 V peak. The room is gridded the same way the rest of the acceptance packet is gridded, letters one way and numbers the other, so every reading carries a coordinate.

The crew runs the megohmmeter with the 5 lb electrodes on a grid of tiles, starting at 10 V and switching to 100 V where the reading climbs above 1.0 x 10^6 ohms, recording resistance-to-ground at the bonded points and point-to-point across the field. Temperature and RH go on the sheet with every set: 22 degrees C, 38 percent RH that morning. One zone reads high to ground while its neighbors read mid-band, which points to a bond, so that pedestal tie gets chased and retested rather than the tile replaced. Then a tech in the site's actual ESD footwear walks the defined path for the body-voltage test.

The packet closes with the resistance grid in band, the one high-to-ground zone traced to a loose bond and corrected and retested, the walking body voltage at 60 V peak against the 100 V limit, and the conditions recorded next to every number. The table shows the spine of that record for a handful of representative points.

Acceptance criteria

The acceptance numbers below are the commonly cited S20.20 and STM values, but they are a starting point, not the contract. The project's ESD control program is what you accept against, and a program for sensitive devices can be tighter than the standard ceiling or can require the system tests in addition to the floor tests. Read the band, the methods, and the body-voltage requirement off the program, then hold the floor to those.

Carry the limits as a set, not as a single number. A floor passes when point-to-point is in band, resistance-to-ground is in band, and, where the program calls for it, the walking body voltage is under the limit with real footwear. A clean reading on one of the three is not a pass on the floor.

IEC 61340-5-1 and the international program

Outside North America the equivalent program is IEC 61340-5-1, and by topic it lines up closely with ANSI/ESD S20.20: an ESD control program standard, with floor resistance commonly held below 1.0 x 10^9 ohms and a person-plus-footwear-plus-floor system limit and a body-voltage requirement. The two standards are recognized as counterparts, S20.20 the American document and EN 61340-5-1 the European one, so a program written to one maps to the other on the major requirements.

The details are not identical, and that is the trap. The IEC document phrases some limits differently and includes a charge-decay style requirement in places, such as dissipating a defined test charge to a low residual within a couple of seconds, and it carries its own system-resistance figure for the operator path. If a project is specified to IEC rather than S20.20, do not assume the S20.20 numbers transfer one for one. Pull the limits from the standard the project actually names.

The practical move on an international job is to confirm which standard governs before you write the acceptance sheet, then test and record to that one's methods and limits. Mixing an S20.20 acceptance sheet onto an IEC-specified room is how a floor gets accepted against the wrong number.

Why did my ESD floor fail the test?

When an installed floor reads out of band, the cause is almost always one of a short list, and they rank by how often they actually bite. Run them in order before you condemn the floor.

The bond to ground is open or high, so resistance-to-ground reads high in a zone while point-to-point reads fine. That is a connection, not a material, and it is the most common finding. The surface has the wrong finish or contamination on it: a coat of standard wax, a residue from the wrong cleaner, or built-up dirt laying an insulating skin over a good floor. The humidity is low and the floor is being read dry without accounting for it, or the reading was taken at the wrong test voltage for the range. And the electrode is wrong, lighter than the 5 lb the method specifies, which reads high because the contact pressure is low.

The body-voltage failure has its own short list, and it is usually footwear. Glazed soles that have stopped conducting, heel grounders worn through or worn wrong, or operators in their own shoes for a task. Fix the system, not the slab, when the floor resistance passes and only the walk test fails. Chasing the flooring when the problem is a worn heel grounder wastes a day and replaces nothing that was broken.

Field checklist

What to document

An ESD floor reading is only as good as the conditions logged beside it, because without the temperature and humidity at the time no one can reproduce the number or trust it. The record answers the question a year out when a board failure gets traced back and someone asks whether the floor was ever right.

Capture the location or coordinate, the point-to-point resistance, the resistance-to-ground, the test voltage each was taken at, the walking body voltage where it was run, the temperature and relative humidity at the time, the instrument and its calibration, the program limit each was held to, and pass or fail. Note the footwear used for the body-voltage test, because the system result is only as meaningful as the shoes that walked it. The table below is the minimum spine.

Common mistakes

• Reading resistance-to-ground on a floor whose ground bond was never verified, so the number measures nothing.
• Treating a clean resistance reading as full S20.20 compliance when the program also required the walking body-voltage test.
• Running the walk test with pristine reference footwear instead of the operators' actual shoes and heel grounders.
• Using a lighter probe than the 5 lb electrode the method specifies, which reads high on low contact pressure.
• Reading a high-resistance dissipative floor at the wrong test voltage and getting a number that does not match the method.
• Recording resistance with no temperature or relative humidity, so the reading cannot be reproduced or defended.
• Waxing or finishing a static-control floor with the wrong product and laying an insulating skin over a good floor.
• Spot-checking at the door instead of testing a grid across the room.
• Chasing the flooring when one zone reads high to ground, instead of finding the open or corroded bond under it.
• Accepting a floor against the standard ceiling when the project's ESD program specified a tighter band.

Standards and references

ANSI/ESD S20.20 is the ESD control program standard, the document a facility's program is built to, and it sets the flooring requirement commonly held below 1.0 x 10^9 ohms. The floor resistance itself is measured per ANSI/ESD STM7.1, the test method for resistive characterization of flooring systems, which covers preinstallation qualification and the field point-to-point and resistance-to-ground methods, the 5 lb electrode, and the 10 to 100 V test convention.

The floor-and-footwear system is covered by the STM97 pair: ANSI/ESD STM97.1 measures the resistance of the person-plus-footwear-plus-floor path, commonly below 1.0 x 10^9 ohms, and ANSI/ESD STM97.2 measures the walking body voltage that system generates, commonly held under 100 V peak. Footwear on its own is characterized per ANSI/ESD STM9.1. The standards are revised on a cycle, so confirm the edition the program references rather than assuming the numbers carried forward unchanged.

Internationally, IEC 61340-5-1 is the counterpart ESD control program standard, close to S20.20 by topic but not identical in its limits and decay requirements, so test and accept to whichever standard the project names. This guide covers the ESD floor measurement in detail; the broader access-floor acceptance, the signal reference grid, and the grid-coordinate packet that the ESD readings live inside are covered in the raised-floor acceptance work, and the bond-verification side ties to ground-resistance and bonding testing. Verify every limit against the project's ESD control program and the adopted standard editions before you accept a floor against them.

Units, terms, and conversions

ESD floor work runs on ohms, volts, and percent relative humidity, with the resistance values written in scientific notation because they span many orders of magnitude. The same idea shows up under a few names across a spec, a data sheet, and a meter readout, so it helps to keep the synonyms straight.

Resistance is in ohms, written as a number times a power of ten, so 1.0 x 10^9 ohms is one gigohm and 1.0 x 10^6 ohms is one megohm. Point-to-point resistance is also called Rtt or Rp; resistance-to-ground is also called Rtg or Rg. Body voltage is in volts, reported as a peak value. Relative humidity is in percent, and electrode weight in the standard is 5 lb, about 2.27 kg. Temperature is in degrees C in the test conditions and is recorded with every reading because resistance moves with it.

Rtt / Rp

Point-to-point resistance, measured between two electrodes on the floor surface

Rtg / Rg

Resistance-to-ground, measured from one floor electrode to the bonded ground point

Ohm and scientific notation

The resistance unit; ESD values run high, so 1.0 x 10^9 ohms is one gigohm and 1.0 x 10^6 ohms is one megohm

Walking body voltage

Peak voltage a person builds walking the floor, in volts, per STM97.2

Relative humidity (RH)

Air moisture in percent; lower RH raises resistance, so it is recorded with every reading

5 lb electrode (2.27 kg)

The defined-weight cylindrical probe STM7.1 uses; its contact pressure is part of the method

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What load rating does a data center raised floor need?
• What is the 62 percent rule in fall-of-potential testing?
• What is an acceptable duct leakage class?
• What is a concentrated load rating on an access floor?
• What megger voltage do I use on busway?