
Written and maintained by the PhDino author · Last reviewed 21 September 2026 · Every calculator used here is tested against independent reference values · how PhDino checks its numbers
Check the ground under and around a retaining wall before sizing the wall itself, so you're not designing a wall on soil that can't hold it.
The most common way a DIY retaining wall fails isn't the wall itself. It is the ground underneath it, the slope it sits on, or the water building up behind it. Before sizing the wall, it's worth checking that the foundation soil can carry it and, if the wall sits on or above a slope, that the slope itself has an adequate safety factor. Only then does the wall's own lateral force and base width mean something.
This guide runs those two ground checks first, then finishes with the wall's loading. As with the deck guide, treat it as a sizing sanity-check alongside your local code, not a replacement for it. Retaining walls above a few feet tall are also where many jurisdictions require a permit and an engineer's stamp, so check that threshold locally before you build.
A retaining wall carries a load that never goes away, pushed sideways by soil that behaves like neither a liquid nor a solid and differs from one yard to the next. A wall can fail four ways. It can slide forward along its base, tip over about its toe, sink or rotate because the soil beneath it is too weak, or move together with the whole slope it stands on. The strength of the wall material, which is what most people worry about, comes after all four.
That is why this project checks the ground before the wall. The bearing check decides whether the base can sit where you plan to put it. The slope check decides whether the hillside is going anywhere with or without your wall. Only then does the sideways push on the wall mean something, because a wall that resists its load perfectly still fails if the slope beneath it goes.
Each of the three calculators answers one narrow question and is blind to the rest, so know its blind spots before you trust its number.
A 4 ft high wall is planned along the downhill edge of a level terrace in a sloping yard. The soil is a silty sand with a friction angle of 30°, a small cohesion of 100 lb/ft² and a unit weight of 120 lb/ft³. Those are the kind of values a soil test returns; without one, use more cautious numbers. The wall's base will be 2 ft wide and buried 1 ft, and below the terrace the ground falls away as a slope 8 ft high at 25°.
| You enter | Value |
|---|---|
| Cohesion (c) | 100 psf |
| Friction Angle (φ) | 30 degrees |
| Soil Unit Weight (γ) | 120 pcf |
| Footing Width | 2 ft |
| Footing Depth | 1 ft |
| The calculator returns | Value |
|---|---|
| Ultimate Bearing Capacity | 7,030.5 psf |
| Allowable (FS=3) | 2,343.5 psf |
| Bearing Factor Nc | 30.14 |
| Bearing Factor Nq | 18.40 |
| Bearing Factor Nγ | 15.07 |
For a 30° soil the bearing factors are Nc = 30.1, Nq = 18.4 and Nγ = 15.1. The equation adds three terms: cohesion, the surcharge of the soil above the base, and the width of the footing. That is 3,014 + 2,208 + 1,808 = 7,030 lb/ft² ultimate, which the calculator divides by a safety factor of 3 to give 2,343 lb/ft², about 2.3 kips per square foot.
That is more generous than the presumptive values most codes give for ordinary soils, which are often in the range of 1,500 to 2,000 lb/ft². When the two disagree, use the lower one unless a soil report supports the higher. Remember, too, that this figure has no water table and no settlement check in it.
| You enter | Value |
|---|---|
| Slope Height | 8 ft |
| Slope Angle | 25 degrees |
| Cohesion | 100 psf |
| Friction Angle | 30 degrees |
| The calculator returns | Value |
|---|---|
| Factor of Safety | 1.48 |
| Stable | 0 yes/no |
The calculator adds a friction term, tan φ ÷ tan β = 1.24, and a cohesion term, 0.25, for a factor of safety of 1.48. It flags the slope as not stable because that is just under the 1.5 it looks for.
Read that as too close to call, not as a small shortfall. The calculator cannot see the weight of the wall and the terrace fill you plan to add at the top, and any water in the slope lowers the real number. The honest choices are to flatten the slope, move the wall back from the edge, or get a geotechnical opinion before you build.
| You enter | Value |
|---|---|
| Slope Height | 8 ft |
| Slope Angle | 20 degrees |
| Cohesion | 100 psf |
| Friction Angle | 30 degrees |
| The calculator returns | Value |
|---|---|
| Factor of Safety | 1.89 |
| Stable | 1 yes/no |
Cutting the slope back from 25° to 20° raises the factor of safety from 1.48 to 1.89. Both terms grow as the slope flattens: the friction term to 1.59 and the cohesion term to 0.30. A margin of that size can absorb the things the calculator cannot see, which is why regrading is often cheaper than any wall detail.
| You enter | Value |
|---|---|
| Wall Height | 4 ft |
| Soil Friction Angle | 30 degrees |
| Soil Unit Weight | 120 pcf |
| The calculator returns | Value |
|---|---|
| Total Lateral Force | 320.0 lbs/ft |
| Overturning Moment | 426.7 ft-lbs/ft |
| Base Width (H/2 rule of thumb) | 2.0 ft |
Rankine's active pressure coefficient for 30° is Ka = 0.333, so the total push is ½ × 0.333 × 120 × 4² = 320 lb per foot of wall. It acts a third of the way up, which makes an overturning moment of 427 ft-lb per foot. The base width from the H/2 rule of thumb is 2 ft, the same base we assumed in step 1.
The calculator stops there, and the wall's own weight is the next input. Take a solid concrete wall at 150 lb/ft³, 2 ft thick and 4 ft high: 1,200 lb per foot of wall. Three quick hand checks then need only that weight and the calculator's outputs.
Overturning: the weight acts about 1 ft from the toe, so it resists 1,200 × 1 = 1,200 ft-lb against the 427 ft-lb pushing the wall over, a safety factor of 2.8. Bearing: the weight spread over the base averages 600 lb/ft², and because the push moves the resultant toward the toe the peak there is about 1,241 lb/ft². Both are far below the 2,343 lb/ft² allowable from step 1, so the ground is not what limits this wall.
Sliding: friction under the base resists the weight times tan of two thirds of the friction angle, a common assumption for concrete cast on granular soil. That is 1,200 × 0.364 = 437 lb against a push of 320 lb, a factor of only 1.36, short of the 1.5 normally wanted. This is the check the H/2 rule hides: it fixes the footprint, not the weight.
To reach 1.5 the wall needs about 1,319 lb per foot, which is 2.20 ft of solid concrete instead of 2 ft. Cheaper fixes are a wider base with backfill sitting over the heel, a shear key cast below the base, or bedding the wall on compacted gravel, which grips better than silty sand.
The chain told us three different things. The soil under the wall is not the problem: 2,343 lb/ft² allowable against a peak near 1,241. The hillside needs regrading before anything is built, since 1.48 at 25° becomes 1.89 at 20°. And the wall's real design question is sliding, a check that appears nowhere on the calculator's results panel.
Had we gone straight to the wall's numbers, we would have designed a tidy footprint on a marginal slope and never met the sliding problem. That ordering, ground first and wall second, is the whole point of running the calculators as a chain.
Every result above assumes a dry, free-draining backfill. Fill that same backfill with water and the wall must resist the soil's thrust and the water's pressure together. For a fully saturated backfill of this 4 ft wall, using the buoyant soil weight (120 − 62.4 = 57.6 lb/ft³) plus water at 62.4 lb/ft³, the total push becomes 653 lb per foot instead of 320, 2.04 times as much, and the overturning moment rises from 427 to 870 ft-lb per foot.
The sliding factor that was 1.36 would fall to 0.67, which means the wall moves, and that ignores any uplift under the base. This is why walls are built with a free-draining stone layer directly behind them, a perforated pipe at the bottom that runs out to daylight, and filter fabric to keep soil from clogging the stone. The calculators' numbers are only valid if that drainage exists and keeps working.
Many jurisdictions require a permit and engineered drawings for walls above about 4 ft, measured from the bottom of the footing to the top of the wall, and some require them at any height when there is a slope above, a driveway or building nearby, or the wall sits at the edge of a slope. The wall in this example is right at that threshold, and the terrace it protects sits on a slope, so a local engineer would be a sensible second opinion.
Also bring in a professional when the soil is soft clay or fill, when groundwater is close to the surface, when the wall is tiered (one wall standing on the backfill of another), or when failure would endanger a building, a road or a neighbor. The figures in this guide are a sizing check on paper, not a design.
Because it is designed for free-draining granular backfill, where cohesion is essentially zero. Cohesion in a clay backfill can vanish as the soil dries, cracks and then wets again, so wall design ignores it on the loading side. That conservative choice is also why the guide tells you to backfill with clean stone rather than native clay.
Then the calculator understates the push. Rankine's coefficient for a backfill sloping at angle β is Ka = cos β × (cos β − √(cos²β − cos²φ)) ÷ (cos β + √(cos²β − cos²φ)). For a 25° backfill slope with φ = 30° that is 0.494 instead of 0.333, so the total push is about 48 percent higher, and the horizontal part of it, the part that slides and tips the wall, about 34 percent higher.
No. The calculator divides its computed capacity by 3. Codes publish presumptive values for each broad soil type, usually lower, that you can use without a soil test. Use whichever is lower unless a geotechnical report supports the higher number, and keep in mind that neither accounts for settlement.
It is convention, not a law of nature. Bearing capacity rests on soil strength, the most uncertain input in the whole job, so a large factor of 3 is customary. Sliding and overturning are resisted mostly by the wall's own weight, which is known far better, so the customary factors are about 1.5 for sliding and 2 for overturning. Your code names the factors to use.
As a starting footprint for a gravity wall, roughly. It is a rule of thumb, not a result: the base width that actually works depends on the wall's weight, the soil, water, and how the wall is built, and the example above shows a wall on an H/2 base that still fails the sliding check.
PhDino earns a commission on qualifying purchases made through this link, at no extra cost to you.
→ The full PhDino bookshelf on Bookshop.org (UK delivery only)
