When you build a road or bridge-approach embankment on soft ground, conventional fill keeps settling for decades. EPS (expanded polystyrene) geofoam weighs about 1/100 as much as soil, dramatically reducing long-term settlement. Vary embankment height, EPS density and soft-layer thickness to compare Terzaghi settlement against a conventional fill in real time.
Parameters
Fill material
Density and compressive strength preset
Soft subgrade
Compression index Cc, initial void ratio e0, preconsolidation σ_p
Embankment height H
m
Embankment width B
m
Soft-layer thickness h_soft
m
Existing stress σ_init
kPa
Effective overburden at mid-depth of soft layer
Traffic surcharge q
kPa
Design life
yr
Results
—
Fill self-weight stress (kPa)
—
Final stress on soft layer (kPa)
—
Consolidation settlement (cm)
—
Reduction vs conv. fill (%)
—
Time to 90% consolidation (yr)
—
Weight reduction factor (×)
—
Embankment cross-section and settlement
Blue EPS blocks rest on the soft layer (brown). White dashed outline shows conventional-fill settlement; orange marker shows the current case. Move the sliders to see the response.
H = soft-layer thickness, Cc = compression index, e0 = initial void ratio, σ_i / σ_f = effective stress before/after loading. EPS at 15 kg/m³ is roughly 1/100 the density of soil (1800), reducing imposed stress by two orders of magnitude.
$$t_{90} = \frac{1.131\,H^{2}}{c_v}$$
Time to reach 90% of ultimate consolidation under double drainage. Peat: cv ≈ 1×10⁻⁷ m²/s, soft clay: 10⁻⁷–10⁻⁶ m²/s.
Geofoam EPS Lightweight Embankment Settlement
🙋
"Lightweight embankment" — is it really true that engineers build road embankments out of expanded polystyrene blocks? Don't they just sink under traffic?
🎓
It really is. EPS (Expanded Polystyrene) is the same plastic used in food packaging, but for civil works it is moulded into high-density, high-strength blocks. Density is 15-30 kg/m³ versus around 1800 kg/m³ for soil — about a hundredfold difference. So the stress on the soft layer below a 5 m EPS embankment is about 1% of the conventional case, which is why you can build several meters of road on peat or soft clay. Oslo, Norway, in 1972 was the first highway use (Lakkegata road), and Japanese NEXCO and Shutoko have used it on thousands of projects since the 1980s.
🙋
If I switch "Fill material" to conventional soil, settlement jumps from 28.5 cm to about 135 cm! That's almost five times worse.
🎓
Exactly — that gap is the headline benefit. Terzaghi gives S = H · Cc/(1+e0) · log10(σf/σi). The bigger σf grows compared to σi, the bigger log10(σf/σi) and therefore S. With conventional fill, the embankment alone adds about 90 kPa to a layer that only had 30 kPa to start with. With EPS, the added stress is under 1 kPa. So EPS literally changes the order of magnitude of the stress increment, which then changes the order of magnitude of settlement. Today's default case shows roughly a 79% reduction.
🙋
Settlement is much smaller, but the 90% consolidation time is still 36 years. Don't we have to wait the same time with EPS?
🎓
Good question. Time to 90% consolidation t90 = 1.131 H²/cv depends only on the soft layer thickness and its cv — switching to EPS does not change cv. But the absolute settlement that has to happen is tiny. If 80% of 28 cm consolidates in ten years, only about 6 cm of long-term creep remains. With conventional fill, the same ten years would give 110 cm of settlement and another 20 cm or more after that. EPS wins on the size of the final number, not on the time constant — which is exactly what removes the bumps at bridge approaches.
🙋
So EPS is basically a free lunch? There must be some catch though.
🎓
Three real catches. First, buoyancy. EPS floats; if groundwater rises above the base of the block, you get about 10 kPa of uplift per meter of head. You need enough cover (or pavement + anchors) above. Second, fire and solvents. EPS self-extinguishes but ignites around 360 °C — be careful with welding on site — and gasoline, kerosene and organic solvents dissolve it, so fuel-station vicinities need a barrier sheet. Third, cost. Per cubic meter EPS is 3-5× more expensive than soil, so you justify it against the cost of ground improvement plus long-term settlement repair. Boston's I-93 Big Dig used 250,000 m³ of EPS — the largest in the world — and it paid for itself by reducing loads on the underlying tunnels.
Frequently Asked Questions
Expanded polystyrene has a density of 15-30 kg/m³, roughly 1/100 that of conventional fill (~1800 kg/m³). Because embankment self-weight stress σ = ρ·g·H is proportional to density, the load on the soft layer is reduced about 100-fold for the same height. Terzaghi 1-D consolidation then yields a settlement that is typically 80-95% smaller than that of a conventional fill of the same geometry, making EPS the standard solution for highways and bridge abutments on peat, soft clay and organic soils.
ASTM D6817 and JIS A 9523 define grades from D12 to D35 with different compressive strength and allowable stress. D15 (15 kg/m³, σ_y ≈ 50 kPa) is the lightest and is used for sidewalks and landscaped fills. D20 (≈80 kPa) suits ordinary roads, and D30 (≈150 kPa) is used near the surface of highways and heavy-traffic routes. Design ensures the long-term combined load (traffic plus pavement) stays below ~50% of yield, keeping creep under 2% per 100 years at stress ratios below 0.3.
EPS degrades under direct UV, so it is always covered by pavement or by soil cover plus a geomembrane. If groundwater rises above the EPS base, buoyancy of about 10 kPa per meter of head can lift the fill, requiring counter-weight, anchors or sufficient cover. The ignition point is around 360 °C with self-extinguishing behaviour, but gasoline, kerosene and organic solvents dissolve EPS, so barrier sheets are mandatory near fuel stations. The 1972 Lakkegata road in Oslo and 1990s Kansai Airport access road both demonstrate 30-50 years of service when these precautions are followed.
For one-dimensional consolidation with double drainage, t90 = 1.131 H²/cv where H is the half-thickness of the soft layer and cv is the coefficient of consolidation. Peat and organic soils have cv on the order of 1×10⁻⁷ m²/s, so a 10 m layer takes about 30-40 years. Soft clays span 1×10⁻⁷ to 1×10⁻⁶ m²/s, giving years to a decade. With EPS the final settlement itself is small, so absolute settlement rate matters less than the fact that very little residual movement remains for differential settlement at bridge approaches.
Real-world applications
Highways over soft-ground sections: Japanese NEXCO, Shutoko and Hanshin expressways use EPS extensively on embankments and bridge-approach backfills over alluvial clay under Kanto loam and reclaimed coastal soft layers. Conventional fill produces 10-20 cm of differential settlement at bridge abutments within 10-20 years of opening, harming ride quality. EPS makes the embankment settle uniformly with the abutment, almost eliminating the bump. The Sanriku coast post-earthquake reconstruction roads (after Tohoku 2011) also used EPS extensively on soft clay subgrades.
Road embankments above underground structures: Building a road on top of a subway or utility tunnel imposes heavy earth pressure on the underground structure. Lightening the fill with EPS slashes the reinforcement cost. Boston's I-93 Big Dig used a record 250,000 m³ of EPS above the underground highway tunnels. In Japan, EPS protects underground structures at the Toyosu market embankment and Shutoko's Ohashi Junction.
Bridge abutment and retaining-wall backfill: Backfilling against an abutment or wall with conventional soil produces large active earth pressure. EPS at 1/100 the density cuts lateral pressure by the same factor, allowing slimmer abutments and walls. Under seismic design, EPS also has tiny inertial mass, lowering earthquake-induced abutment damage. After the Hanshin earthquake EPS was widely used in lightweight repair backfills.
Cold-region, slope and special-ground applications: EPS doubles as thermal insulation, suppressing frost heave on northern roads (extensive use in Hokkaido and Tohoku). On steep terrain, EPS forms the valley-side fill of a widened road, giving the slope stability needed without overloading the toe. Norway's Statens Vegvesen Lakkegata project of 1972 was the first installation worldwide and is still in service.
Common misconceptions and pitfalls
The biggest trap is to assume "EPS is light, so anything goes." The weakness of EPS is exactly that it floats. With a specific gravity of 0.015–0.030 versus water at 1.0, the base of the block submerging under flood, heavy rain or a high water table can float the whole embankment off if the cover and anchorage are insufficient. Design must specify pavement plus concrete slab plus anchors (or sufficient soil cover) that exceeds the maximum credible water level. Riverside, low-lying and high-water-table sites need the most attention.
Next, treating consolidation alone as the whole settlement story. The Terzaghi equation used here gives the long-term consolidation of the soft layer only. Real embankment construction also produces (1) immediate elastic settlement of the granular bedding, (2) lateral squeeze of the soft layer outward, and (3) punching into deeper firm strata. EPS dramatically reduces the consolidation term but the bedding-layer compression and shear of thin soft seams still need separate checking. In practice this tool is the quick screen, followed by PLAXIS or FLAC for detailed analysis.
Finally, overconfidence in EPS durability. EPS has 50-year track records, but only when protected. UV exposure, prolonged saturation, contact with hydrocarbons, and mechanical damage from excavator buckets all degrade the surface within a few years if they coincide. The design must include a geomembrane plus sand layer, barrier sheets near fuel stations, and inspection access, with planned coring at 10 and 20 years of service. The Oslo Lakkegata road still works today because exactly these long-term controls have been respected for half a century.
How to Use
Enter embankment height (m) and width (m) to define geometry; typical road embankments range 2–6 m high.
Input soft subgrade thickness (m) and pre-existing vertical stress (kPa) from field boring data; soft clay typically shows 40–80 kPa initial stress.
Simulator calculates EPS Geofoam self-weight stress (density ~15 kPa/m), final stress on subgrade, primary consolidation settlement using Cc (compression index), and compares settlement reduction versus conventional granular fill (density ~18 kPa/m).
Review weight reduction factor (typically 0.8–0.95× conventional fill) and time to 90% consolidation using vertical strain rate coefficients for the soft layer.
Worked Example
Bridge approach embankment: 4 m high EPS Geofoam, 12 m wide, over 3 m soft clay (Cc=0.28, cv=2.5×10⁻⁸ m²/s, preexisting stress 50 kPa). Self-weight stress from EPS = 60 kPa; final stress = 110 kPa. Consolidation settlement ≈ 8.5 cm versus 34 cm for gravel fill. Settlement reduction = 75%. Weight reduction factor = 0.83×. Time to 90% consolidation ≈ 4.2 years versus 16.8 years for conventional fill, allowing faster project completion.
Practical Notes
EPS density varies by grade (EPS-15 to EPS-40); confirm manufacturer certification to ensure compression index and long-term creep assumptions (typically 1–3% additional settlement over 20 years).
Soft subgrades with OCR <1.5 and Cc >0.35 yield maximum benefit; saturated clay with low permeability extends consolidation time—install vertical drains (PVD spacing 1–1.5 m) to accelerate.
Account for EPS lateral extrusion under embankment load; geotextile wrapping and slope stability analysis (FSslip ≥1.3) are mandatory design steps not simulated here.
For marine/coastal sites, confirm EPS moisture absorption limits (<1% by volume) and UV degradation protection with polyethylene facing.