Below-grade foundation waterproofing field guide

Below-grade waterproofing, and why one shot is all you get

Below-grade foundation waterproofing keeps groundwater out of the spaces you put underground: the basement, the below-grade parking, the elevator pit, the mechanical room, the data hall. The wall is concrete, and concrete is not waterproof. A low water-cement ratio mix cuts the concrete's own permeability and is the first line of defense, but the wall still has cracks, cold joints, tie holes, and honeycomb, and water finds every one of them. The membrane and the drainage are what actually hold the water back.

The reason this matters more than almost any other detail on a below-grade structure is access. Once the wall is backfilled, the water side of it is gone. You cannot reach the exterior face again without excavating, and on a tight site or under an occupied building you may never be able to. A leak that would have been a half-day fix during construction becomes a six-figure problem from the inside after the dirt is back. The positive side, done right the first time, is the cheapest waterproofing you will ever buy.

So the order of the work is the whole game. Get the concrete as tight as the mix allows, put the right system on the right side, relieve the water pressure with drainage, detail every joint and penetration, then protect all of it through backfill. Skip a step and the wall finds it for you.

What is the difference between dampproofing and waterproofing?

The difference is hydrostatic pressure, and it decides everything else. Dampproofing resists moisture and water vapor moving through the wall when there is no standing water pushing on it. Waterproofing holds back liquid water under pressure, the kind a water table or saturated soil produces. ACI draws the line the same way: a dampproofing treatment resists water in the absence of hydrostatic pressure, a waterproofing treatment resists water under hydrostatic pressure. Confirm the current ACI reference, but the distinction itself is settled.

You can see it in the materials. Dampproofing is usually a thin asphalt or bituminous coating, often under 10 mils, that slows capillary moisture by plugging the surface pores. It has no ability to bridge a crack or stand up to a head of water. Waterproofing is a membrane, commonly 40 mils and up, elastic enough to bridge a moving crack and continuous enough to hold back standing water.

Here is the mistake that defines the job: dampproofing a wall that has a water table behind it. The sprayed-on asphalt looks like waterproofing to anyone not paying attention, it goes on fast, and it costs a fraction of a real membrane. Then the water table comes up and the basement floods, because asphalt emulsion was never rated for pressure. If there is any chance of a head of water, you waterproof. You do not dampproof and hope.

How does hydrostatic head decide the system?

Hydrostatic head is the pressure standing water puts on the wall, and it grows with depth. Every foot of water above a point on the wall adds roughly 0.43 psi pushing in. A basement wall with the water table near grade can carry several feet of head at the footing, and that pressure works on every crack, joint, and pinhole around the clock.

Where the water table sits, and how high it can rise, is not a guess you make in the field. It comes from the geotechnical report and the soils investigation, and the design responds to it. A seasonal high water table, a perched water table over a clay lens, or a site at the bottom of the grade all change the answer. The geotech tells you what the water does. The engineer of record and the project specification tell you what system answers it.

The blunt version: the water table drives the choice between dampproofing and waterproofing, and it drives whether you need drainage to relieve the head. The building code ties the trigger to exactly this. Where a high water table or severe soil-water conditions are known to exist, the code calls for waterproofing, not dampproofing. Read the geotech before you price the wall.

What is positive-side versus negative-side waterproofing?

Positive-side waterproofing goes on the exterior, the water-bearing face of the wall, and it is the system to use whenever you can. It stops water before it ever enters the concrete, which keeps the wall itself out of the freeze-thaw and the groundwater chemistry. Water never gets into the structure to find a path. That is why it lasts, and why it is worth the access it demands.

Negative-side waterproofing goes on the interior, the dry side, and it is the second-best answer for when you cannot get to the outside. Remedial work on an existing basement, a wall tight against a property line, a structure you cannot excavate: those are negative-side jobs. The catch is real. A negative-side system resists water coming through, but it does not keep water out of the wall mass, so the concrete still saturates and still takes the freeze-thaw and the chemistry. Crystalline coatings are the common negative-side tool because they work inside the concrete itself.

The lean is not subtle. Positive side if you can dig, negative side if you cannot. The expensive failures come from choosing negative side when positive was available, usually to save the cost of excavation, and then living with a wall that never dries out.

There is a middle ground people forget. On new construction with exterior access, you put the membrane on the positive side and you can still add a crystalline admixture to the concrete for a second line behind it. That belt-and-suspenders approach is standard on the spaces that cannot tolerate a leak, and it costs little when the wall is already open. The decision is rarely positive against negative in the abstract. It is what the access on this site allows, and what the consequence of a leak justifies spending.

Membrane types and where each fits

There is no single best membrane. There is the one that fits the substrate, the access, the water, and the crew you have. Four families cover most below-grade work.

Sheet membranes come in rolls and go on a prepared wall. Self-adhered modified-bitumen sheets, often referenced to ASTM D6135 for below-grade use, peel and stick. Torch-applied and SBS sheets are heat-welded at the laps. They give a factory-controlled thickness, which is their strength, and they live or die at the seams and penetrations, which is their weakness.

Fluid-applied membranes spray or roll on and cure to one continuous layer with no seams. Polyurethane, rubberized asphalt, and cold butyl formulations reach high elongation, which lets them bridge cracks, and they conform to complex shapes a sheet fights. The trade-off is thickness control. A fluid membrane is only as good as the wet-mil and dry-mil readings the applicator actually takes.

Bentonite is sodium clay, in panels or sprayed, that swells when it wets and seals itself around minor voids and penetrations. It is a common blindside and pre-applied choice. Crystalline is the outlier, an admixture or coating that grows crystals inside the concrete to seal the pores. The next section covers it on its own.

What is crystalline waterproofing?

Crystalline waterproofing turns the concrete itself into the barrier. It comes as a surface coating or as an admixture dosed into the mix, and it carries reactive silica chemistry. In the presence of moisture, the soluble silicates penetrate the concrete and react with the calcium hydroxide and lime that hydration leaves behind, growing insoluble crystals that fill and plug the capillary pores. The water path through the concrete closes from the inside.

The feature that sells it is self-sealing. The crystals keep forming wherever water and unhydrated cement are present, so a hairline crack that opens later can re-seal as water reaches fresh material. There are limits. It seals hairline cracks, not structural cracks or moving joints, and the dose, the cure, and the substrate condition decide whether it performs. The manufacturer's data governs the crack width it can close and the conditions it needs.

Crystalline is the common answer on the negative side, where it works because it lives inside the concrete rather than as a film on the dry face that water pressure would push off. As an integral admixture it lowers permeability from the batch, which pairs with a tight water-cement ratio rather than replacing it. It is a system component, not a substitute for drainage where there is a head of water.

Do you need a drainage board?

On a wall with any real head of water, yes. A drainage board, also called a dimple board or dimpled drainage sheet, is what relieves the hydrostatic pressure off the membrane, and relieving the pressure is half the battle. The board is a high-density plastic sheet pressed into dimples, installed dimples toward the wall, so it holds an air gap between the soil and the membrane. Water that reaches the board cannot build pressure against the wall. It falls down the gap.

Where it falls matters. The board carries the water down to the foundation drain at the footing, the perimeter drain, which then runs to daylight on a sloped site or to a sump where it does not. Run the board to the base of the footing so the water is delivered straight to the drain and away from the wall-to-footing cold joint, the single most common leak point on the whole wall. Most boards carry a filter fabric on the soil side that lets water in and keeps the fines out, so the drain does not silt up.

The point that gets lost is that the membrane and the drainage are one system. A membrane with no way to shed the water behind it sits under full head and finds its weakest pinhole. Drainage that dumps water at the footing with no membrane behind it leaks at every joint. You need both, and you need the drain to actually go somewhere.

Grading and the ground around the wall do real work too, and they cost almost nothing. Slope the finished grade away from the building so surface water runs off instead of soaking down against the wall, and keep the downspouts and area drains carrying roof water clear of the foundation. A lot of basements that get blamed on the membrane are really getting fed by a downspout dumping at the corner of the house. Manage the water you can keep away from the wall, and the system behind the wall has less to fight.

Surface preparation before the membrane

The membrane bonds to the wall you give it, not the wall on the drawing. The detail assumes a clean, sound, cured substrate. You will rarely get all three without work. Strip the forms, knock off the fins and the laitance, and get the surface clean of form release, dust, and curing compound, because most membranes will not stick to any of those.

Concrete needs to cure before most membranes go on. Manufacturers commonly call for a cure period, often around 7 days, or a moisture condition they specify, and a self-adhered sheet on green, wet concrete will not bond. Confirm the requirement and the moisture limit for the specific product. Patch the defects that are leak paths waiting to happen: the form tie holes, the honeycomb and bug holes, the cold joints. Tie holes get plugged with a non-shrink grout or the manufacturer's detail material, not smeared over. Honeycomb gets cut out and repaired the way any unsound concrete does, the same patch-bond and low-shrinkage logic that governs spall repair.

Fill and tool a cant or fillet at the wall-to-footing inside corner so the membrane is not asked to bridge a sharp 90 degree angle, which is where sheets tent and fluid coatings thin out. Prep is unglamorous, and it is where the warranty is won.

The details, where below-grade walls actually leak

The field of the wall almost never leaks. The details do. Walk any wet basement and the water is coming in at a joint, a penetration, or a corner, not through the middle of a sound membrane. So the detailing is the job, and the flat field is the easy part.

The cold joint and the construction joint between the footing and the wall, and between pour lifts, is leak point number one. It gets a waterstop cast in and the membrane lapped across it. Penetrations are next. Every pipe, conduit, and sleeve through the wall is a hole in your barrier, and each one needs a boot, a sealant detail, or a hydrophilic collar tied back into the membrane. Inside and outside corners get reinforced, with a fillet at the inside corner and an extra ply or detail strip at both, because that is where a sheet cannot lie flat and a fluid coating runs thin.

The termination at the top, where the membrane ends near grade, has to be sealed and held so water cannot get behind it and run down the dry face. Detail these before you roll the field membrane, not after. The field goes fast. The details are the schedule, and they are the leaks.

Waterstop at construction joints

A waterstop is a sealing strip cast into a concrete joint so water cannot travel through the joint even though the two pours never bonded into one. Below grade, the footing-to-wall joint and every construction joint in the wall is a planned discontinuity, and the waterstop is what closes it from inside the concrete.

Two common types. A PVC waterstop is a profiled strip, often with a center bulb, set so half is in the first pour and half in the second, forming a physical barrier and a long path the water cannot cross. A bentonite waterstop is a strip of swelling clay set on the joint that expands against the concrete when it wets and seals the gap. PVC wants to be centered and supported so it does not fold over during the pour, which is the classic failure, a waterstop laid flat in the bottom of the joint doing nothing. Bentonite wants to stay dry until it is in place, because rain on the strip swells it early.

The waterstop is cast-in, so it is a one-shot detail. There is no fixing it after the second pour closes the joint. Get it located, supported, and continuous around corners and splices before the concrete covers it.

Protecting the membrane through backfill

A membrane that passes inspection and then gets torn during backfill is a membrane that failed. Backfill is rough. Rock, debris, the bucket, and the compaction equipment all want to gouge or puncture the membrane you just installed, and every gouge is a leak you will never see until the wall is wet.

So the membrane gets a protection layer before any dirt goes back. A protection board, a drainage board that doubles as protection, or a specified protection course takes the abuse so the membrane does not. Then the backfill itself is placed with some care: no large rock or construction debris directly against the wall, lifts compacted without slamming the equipment into the protection, and the drainage board kept in place and not crushed.

This is where the schedule pressure does the damage. The waterproofing is done, everyone wants the hole closed, and the backfill goes in fast and dirty. The torn membrane does not announce itself. It shows up as a leak months later, from the one side of the wall you can no longer reach. Protect it like you cannot fix it, because you cannot.

Under-slab membrane and vapor barrier

The wall is only part of the below-grade envelope. The slab on grade sits on the ground too, and water vapor, liquid water, and soil gas all move up through it unless something stops them. Under a slab, the common element is a vapor barrier, a polyethylene or polyolefin sheet laid on the prepared subgrade or granular fill before the slab is poured.

The reference most specs point to is ASTM E1745, which classes these sheets A, B, and C by puncture and tensile strength, with the heavier classes for rougher subgrades and harder use. The number that matters for a flooring-sensitive slab is the permeance. A low-perm sheet, with the tightest specs calling for very low permeance, keeps vapor from driving up into the slab and wrecking the floor covering bonded on top. Where soil gas is the concern, the same under-slab membrane is part of the radon or methane control, sealed at the laps and the penetrations so the gas has no path up.

On a true below-grade slab with a head of water under it, this becomes waterproofing, not just a vapor barrier, and the slab membrane ties into the wall system so the two are continuous. A gap between the wall membrane and the slab membrane is a leak at the worst possible joint.

Blindside and pre-applied waterproofing

Sometimes there is no exterior to waterproof. On a zero-lot-line building, against soldier-pile-and-lagging shoring, or anywhere the excavation is cut to the property line, you cannot place a membrane on the outside face of the wall after the wall exists, because that face is poured against the earth retention. The answer is blindside, also called pre-applied, waterproofing.

The membrane goes in first, against the shoring or the lagging on the walls and against the mud mat under the slab, before any reinforcing or concrete. Then the concrete is cast directly against it. The defining feature of a pre-applied system is that it bonds to the concrete as it cures, so the membrane and the structure become integral and water cannot travel between them. That intimate bond is what matters, because if water can migrate in the gap between a loose membrane and the wall, a leak at any point can show up anywhere, and you can never trace it.

Blindside is unforgiving. It is installed before the concrete, so it is buried before it is ever tested under head, and there is no access to repair it. HDPE pre-applied sheets and bentonite systems are the common choices. The detailing at the slab-to-wall transition and at every penetration is the whole job, because that is where blindside systems leak and where you cannot get back to fix them.

High-stakes below-grade: pits, vaults, and data halls

Some below-grade spaces forgive a little water. A slightly damp utility tunnel is a nuisance. Others forgive nothing. An elevator pit with standing water, a vault full of switchgear, a below-grade data hall with millions of dollars of equipment on the slab: those are spaces where a leak is not a callback, it is a catastrophe, and the waterproofing gets designed accordingly.

On these the approach changes from one system to redundancy. You may see a positive-side membrane and a crystalline admixture in the concrete and a full drainage system, because no single layer is trusted alone. You see more testing, tighter joint detailing, and leak detection designed in. The cost of the waterproofing is trivial against the cost of what it protects, so the design stops optimizing for cost and starts optimizing for never leaking.

The principle carries over to any below-grade job: match the system to the consequence of failure. A storage basement and a data hall do not get the same wall, because the cost of being wrong is not the same. Let the consequence, the water table, and the project specification set the level, and on the high-stakes spaces, build in the redundancy before you need it.

Testing and quality control

A below-grade membrane gets tested before it is buried, because after backfill there is no testing it and no fixing it. The two tools you will see most are the flood test and electronic leak detection.

A flood test floods a horizontal membrane, a plaza, a pit floor, a slab, with water to a set depth for a set time, commonly on the order of a day, and you watch for leaks below. It is direct and convincing on flat work, and it has a real risk: if the membrane leaks, the substrate underneath can saturate, so you flood-test where you can drain and dry afterward. Electric field vector mapping, EFVM, and similar electronic leak detection put a low-voltage potential between the membrane, which does not conduct, and the grounded concrete beneath it, then trace the current to find any breach. It finds a pinhole as easily as a torn seam, it works on sloped and vertical surfaces a flood test cannot, and it can be run after overburden is placed.

Either way, the inspection that matters happens before the dirt goes back. Walk the details, not just the field: every penetration, every corner, every termination, the laps, the waterstops. The field of the membrane rarely fails the test. The details do, and the test is your last look at them.

The maintenance the owner takes on

Waterproofing is not install-and-forget, and the part that needs attention is the part nobody thinks about until the basement is wet: the drainage. The membrane itself is buried and maintenance-free, but the foundation drain, the sump, and the discharge are mechanical, and they fail the way mechanical things fail.

The perimeter drain silts up if the filter fabric was wrong or missing, and a clogged drain puts the full head of water back on a membrane that was never sized to take it alone. The sump pump has a float, a check valve, and a power supply, and any of the three can quit, which is why critical pits get a backup pump and an alarm. The discharge has to keep flowing to daylight or the storm system and not freeze or get buried. Hand the owner a system they can actually maintain: a drain they can clean out, a sump they can test, and an alarm that tells them before the water is on the floor.

When a previously dry basement starts leaking years later, the cause is usually not the membrane. It is a drain that stopped draining or a sump that stopped pumping, and the head came back.

What to document

The waterproofing gets covered up, so the record is the only proof of what is behind the dirt. When a leak shows up later, the documentation is what tells you whether the system was right and where to look.

Record it by area, because a below-grade structure rarely gets one system everywhere. The walls above the water table may be dampproofed while the deep walls and the slab are waterproofed, and the record has to capture which is which and where the line was. Capture the system on each surface, the membrane product and thickness, the drainage, the waterstops, the test results, and the manufacturer's details used.

Common mistakes

• Dampproofing a wall that has a water table behind it, when the conditions called for waterproofing.
• Installing a membrane with no drainage board, leaving the full hydrostatic head on the membrane.
• Leaving joints, penetrations, and corners undetailed and trusting the field membrane to cover them.
• Tearing the membrane during backfill because no protection board went on first.
• Choosing negative-side waterproofing to skip excavation when positive side was available.
• Omitting the waterstop at the footing-to-wall and construction joints, or laying it flat so it does nothing.
• Applying the membrane to green, dirty, or uncured concrete that it will not bond to.
• Stopping the drain nowhere, so it has no daylight or sump to discharge to.

Field checklist

Standards and references

The building code sets the trigger. The residential code addresses foundation dampproofing and waterproofing, commonly in the Chapter 4 foundation provisions, with the waterproofing-versus-dampproofing decision tied to whether a high water table or severe soil-water conditions are known to exist. It also sets where the treatment runs, from finished grade down to the footing or near the basement floor. Confirm the section against the code edition the jurisdiction has actually adopted, and the local amendments, before you cite it.

ACI gives the definitions and the guidance, and the ACI document on below-grade waterproofing is where the dampproof-versus-waterproof distinction and the system guidance live. ASTM covers the materials and the tests, including the self-adhered sheet membrane specification and the under-slab vapor retarder specification with its Class A, B, and C ratings. Verify the current designations rather than trusting a number from memory.

Above all of it, two things govern in the field. The manufacturer's instructions and data sheets control the product: the cure requirement, the thickness, the primer, the detail, and the crack width a crystalline system can seal. The geotechnical report and the engineer of record control the design: the water table, the system, and the redundancy. When any of these conflicts with a rule of thumb, the rule of thumb loses.

Units and terms

Below-grade waterproofing borrows terms from a few trades, and the same idea shows up under different names across the geotech report, the spec, and the product data sheet.

Hydrostatic head is given as a height of water in feet or as a pressure in psi or kPa, where every foot of water is about 0.43 psi. Membrane thickness is in mils, thousandths of an inch, or in millimeters on metric data sheets, where 40 mils is about 1 mm. Permeance for vapor barriers is in perms. Positive side means the exterior water-bearing face. Negative side means the interior dry face.

Hydrostatic head

The pressure standing water exerts on the wall, growing about 0.43 psi per foot of depth

Dampproofing

A treatment that resists moisture with no hydrostatic pressure, typically a thin coating

Waterproofing

A treatment that resists water under hydrostatic pressure, typically a membrane

Positive side

The exterior, water-bearing face of the wall; the preferred side to waterproof

Negative side

The interior, dry face; used when the exterior cannot be reached

Waterstop

A strip cast into a concrete joint to block water from passing through the joint

Blindside / pre-applied

Membrane installed before the pour, against shoring or a mud mat, bonding to the concrete

Mil

One thousandth of an inch, the unit for membrane thickness

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What is a concrete mix design?
• Why does spalled concrete happen?
• How do you drain a patio?
• What is a footing?
• What is a waterstop?