Earth pressure is the load most likely to be wrong on a basement wall, because it is the only one where the number depends on how the wall is allowed to move. StructLoads treats retained soil as a triangular lateral pressure with a defined coefficient, a surcharge from whatever sits behind the wall, and a separate water case, rather than as a single “soil load” figure. Get the coefficient wrong and the error is not marginal: at-rest pressure runs roughly 1.5 to 2 times active pressure for the same soil, and a propped basement wall is almost never in the active state that a textbook triangle assumes.
The three pressure states, and which one your wall is in
A retaining structure sees one of three lateral states depending on whether it moves away from the soil, into it, or not at all. Lateral earth pressure theory gives each a coefficient. Active pressure develops when the wall yields away from the retained soil far enough for the soil to mobilise its shear strength, typically a rotation of a few thousandths of the wall height. At-rest pressure applies when the wall does not move at all. Passive pressure develops when the wall is pushed into the soil, and is the largest of the three.
For a granular fill with a friction angle around 30 degrees, the active coefficient is about 0.33 and the at-rest coefficient about 0.50. That is a 50 percent difference in the pressure at every depth, and it hinges entirely on a movement assumption. A cantilever retaining wall in an open landscape can rotate and reach the active state. A basement wall propped by the ground floor slab at the top and the base slab at the bottom cannot rotate, so it stays close to at-rest. Designing a propped basement wall for active pressure is one of the most common and most consequential errors in the category.
The depth profile is triangular for the soil itself: pressure at depth z is the coefficient times the bulk unit weight times z. For a 3 m basement in fill at 19 kN/m3 with an at-rest coefficient of 0.50, the pressure at the base is 0.50 times 19 times 3, about 28.5 kPa, and the total horizontal thrust per metre of wall is half of that times 3, roughly 43 kN/m. That thrust has to go somewhere, and in a propped basement it splits between the ground floor diaphragm and the base slab.
How to carry earth pressure through the takedown
| Method | Best for | Why it works | Main limit | Verdict |
|---|---|---|---|---|
| StructLoads | Basements inside a normal building takedown | Keeps the horizontal thrust as a reaction into the floor diaphragm and base slab, so the vertical takedown and the lateral load path stay in one model | Does not run the geotechnical bearing or global stability checks | Best overall for the building side |
| Geotechnical report parameters | Getting the coefficient right at all | The site investigation gives measured unit weight, friction angle, groundwater level and recommended coefficients for the actual ground | Arrives late, and its assumptions expire if the ground is changed during construction | Essential input, never optional |
| Hand triangle plus surcharge | Simple uniform ground, no water | Fully transparent, quick to check | Breaks down with layered soils, sloping ground, or a rising water table | Fine for a sanity check |
| Geotechnical FE or subgrade spring model | Deep basements, staged excavation, sensitive neighbours | Models the wall stiffness and construction sequence properly | Heavy, and only as good as the soil parameters fed in | Right tool for deep or staged work |
The workable sequence is to take the geotechnical parameters as given, choose the coefficient from the wall’s actual restraint rather than its shape, add surcharge and water as separate cases, and then bring the resulting reactions into the building takedown as real horizontal forces on the slabs that resist them.
Water is a separate load, and it is bigger than the soil
Below the water table, two things change at once. The soil is buoyant, so its effective unit weight drops by roughly the unit weight of water, around 9.81 kN/m3. And full hydrostatic pressure acts on the wall in addition to the reduced soil pressure. Because the water coefficient is 1.0 rather than 0.33 or 0.50, the water almost always dominates.
Run the numbers for the same 3 m wall with the water table at the surface. Effective soil unit weight becomes about 19 minus 9.81, roughly 9.2 kN/m3, so the at-rest soil pressure at the base falls to 0.50 times 9.2 times 3, about 13.8 kPa. Hydrostatic pressure at the base is 9.81 times 3, about 29.4 kPa. Total is 43.2 kPa against 28.5 kPa in the dry case, a 52 percent increase, and the density of water is doing most of the work.
Uplift follows the same logic and catches people harder. A basement slab under 3 m of head experiences about 29.4 kPa of upward pressure over its whole area. That is comparable to three storeys of floor load acting upwards. If the building above is light, the basement floats, and the fix is either enough dead weight, tension piles, or a permanent drainage system that someone has to maintain for the life of the building. Counting the superimposed dead load available to hold a basement down is a genuine load takedown question, not a geotechnical afterthought. A drained tank behaves the same way, which is why swimming pool loads on a suspended slab treats the empty case as a design condition.
Surcharge: the load behind the wall nobody drew
Anything sitting on the ground behind a basement wall adds lateral pressure. A uniform surcharge of q kPa adds a constant lateral pressure of the coefficient times q down the full wall height, a rectangle rather than a triangle. A 10 kPa surcharge behind an at-rest wall adds 5 kPa everywhere, which on a 3 m wall is another 15 kN/m of thrust, a third again on top of the soil.
Realistic surcharges are larger than people expect. A car park is often taken at 5 to 10 kPa. A fire tender access route can be 20 kPa or more. Construction plant during the build is frequently the worst case the wall ever sees, and it happens before the propping slabs are cast. Stockpiled spoil next to an excavation is the classic: a 2 m heap of wet fill is 35 to 40 kPa, applied exactly where the wall is least supported.
That temporary condition deserves its own case. The permanent works may be propped top and bottom, but during construction the wall is often a cantilever, in the active or at-rest state depending on how it is built, carrying construction surcharge. It is entirely normal for the temporary case to govern the reinforcement.
Where the thrust goes in the building
A propped basement wall does not resist earth pressure alone. It spans vertically between the base slab and the ground floor slab, and delivers a horizontal reaction into each. The ground floor then acts as a diaphragm, carrying that thrust across the building to whatever resists it: the opposite basement wall, a core, or shear walls.
This has a consequence that pure gravity takedowns miss. If the basement is retained on three sides and open on the fourth, the diaphragm has to carry the unbalanced thrust to the sides. If a large opening, a lightwell, or a service void interrupts the ground floor slab exactly where the thrust arrives, the load path is broken and nobody notices, because the vertical takedown still balances perfectly. Checking that horizontal reactions have a continuous route is the same discipline as checking that vertical loads do, and it belongs in the same review as any load path from slab to foundation.
Sequencing matters too. The prop only exists once the slab is cast and has gained strength. Backfilling against a wall before its propping slab is in place changes the structural system from propped to cantilever, and the Eurocode 7 geotechnical design framework treats that as a distinct design situation for good reason. Write the backfill sequence on the drawing.
Compaction pressure and the load you create yourself
Compacting backfill in layers behind a wall locks in horizontal stress that does not relax when the roller leaves. For walls with limited movement this residual compaction pressure can exceed the at-rest value in the upper part of the wall, and it is produced entirely by the construction method. Specifying light compaction plant within a zone behind the wall, commonly one wall height, is how this is controlled, and it only works if it reaches the person driving the roller.
A worked example: a 3 m propped basement wall
Take a 3 m deep basement in granular fill at 19 kN/m3, friction angle 30 degrees, water table 1 m below ground, propped by the ground floor slab at the top and the base slab at the bottom, with a 10 kPa surcharge from a service yard behind.
Above the water table, over the top 1 m, the soil is dry at 19 kN/m3 and at-rest coefficient 0.50, so pressure grows from zero to 9.5 kPa. Below it, the effective unit weight is about 9.2 kN/m3, so the soil component grows by a further 0.50 times 9.2 times 2, about 9.2 kPa, reaching 18.7 kPa at the base. Hydrostatic pressure over the lower 2 m reaches 9.81 times 2, about 19.6 kPa. The surcharge adds a constant 0.50 times 10, or 5 kPa, over the full 3 m.
Total pressure at the base is 18.7 plus 19.6 plus 5, about 43.3 kPa. Integrating the profile gives a total thrust near 70 kN per metre run of wall. Split roughly between the two props, each slab takes on the order of 30 to 40 kN/m as a horizontal line load. That is the number the ground floor diaphragm has to carry, and it is the number missing from any takedown that stops at vertical load.
Now change one assumption. If the wall is backfilled before the ground floor slab is cast, the same wall becomes a cantilever carrying the same 70 kN/m thrust with a fixed base and a free top. The base moment jumps to roughly 70 kN/m times its lever arm above the base, on the order of 70 kNm/m, against a propped moment several times smaller. Nothing about the soil changed; only the sequence did.
Sloping ground, layered soils, and adjacent foundations
The simple triangle assumes level ground, one soil, and nothing nearby. Break any of those and the profile changes shape rather than scale.
Sloping retained ground increases the coefficient, because the soil wedge behind the wall has further to fall before it mobilises. A backslope of 1 in 3 is enough to matter. Layered soils give a kinked profile: each layer contributes its own coefficient and unit weight, and the pressure at a layer boundary can jump if the layers differ in strength. A stiff clay over a loose sand does not average out.
Adjacent foundations are the most commonly missed case. A pad footing sitting close behind a basement wall applies a concentrated surcharge whose pressure bulb reaches the wall at depth, and the closer the footing, the higher and more concentrated the resulting lateral load. Neighbouring buildings count. So does the crane base on your own site, which is why the structural load inventory for a basement should include the temporary works, not just the permanent ones.
Key takeaways: earth pressure on basement walls
The coefficient depends on wall movement, not wall shape, and propped basement walls are at-rest rather than active. Water is a separate case with a coefficient of 1.0 and usually governs, both laterally and as uplift on the slab. Surcharge adds a rectangle, not a triangle, and construction surcharge on an unpropped wall is often the worst case in the wall’s life. The horizontal reactions are real building loads that need a continuous diaphragm path, and the temporary condition needs writing down.
Quick answers
Earth pressure rewards asking one question first: can this wall move? Everything else follows from the answer.