Granular fill: why it matters in retaining wall design.

The four jobs the backfill is asked to do

1. Mobilise high friction angle so the active earth pressure on the wall stays manageable. Every degree of φ reduces the active pressure coefficient K_a by about 3 to 4%.
2. Drain freely so water does not accumulate and add hydrostatic pressure on the wall. Saturation can double or triple the lateral force on the wall (see Backfill mechanics).
3. Compact to high density with reasonable site equipment, holding that density under traffic and loading without creep.
4. Interact reliably with the reinforcement (for MSE walls) or transfer load through to the wall facing (for any wall type) without time-dependent loss of stiffness.

Granular fill, properly specified, does all four. Cohesive fill (clay, plastic silt, organic soils) does none of them well.

What "granular fill" actually means

Granular fill is a coarse-grained soil with at least 70% retained on the 75 µm (No. 200) sieve, including:

• Gravel (2 mm to 75 mm particles)
• Sand (0.075 mm to 2 mm particles)
• Crushed rock (quarry-derived angular fragments, mixed gradation)
• Recycled aggregate (concrete or asphalt crushed to gradation)

The opposite is cohesive fill: silt, clay, organic soils, residual soils with significant fines. Cohesive fill derives its strength from electrostatic cohesion between fine particles, not from particle-on-particle friction.

Within granular fill, there are quality tiers:

Costs are 2026 Malaysian indicative ranges. They vary by quarry source and haul distance.

Why fines are the enemy of granular fill

The single most important quality criterion in a granular fill spec is fines content: the percentage of material passing the 75 µm sieve. Low fines = good fill. High fines = problem.

Friction angle

Fines fill the voids between coarse particles. The coarse particles can no longer interlock by dilatancy (they have to push the fines out of the way before they can ride up over each other), and the effective friction angle drops. A backfill with 5% fines might run φ = 37°; the same parent material at 15% fines might run φ = 32°.

Permeability

Permeability scales roughly with the square of the d₁₀ particle size. Adding fines drops d₁₀ dramatically. A clean granular fill at 5% fines has permeability k ≈ 10⁻³ to 10⁻⁵ m/s (effectively free-draining). The same material at 15% fines has k ≈ 10⁻⁶ to 10⁻⁸ m/s (effectively impermeable). The drainage system you designed assumes water flows out; if the permeability is too low, water builds up.

Water sensitivity

Plastic fines (clay) shrink as they dry and swell as they wet. Over a wet-dry cycle, a backfill with 15% plastic fines can experience volumetric changes that crack adjacent facing panels or shear connections. Non-plastic fines (silt) lose cohesion when saturated and flow with seepage water, clogging drains and producing internal piping voids.

Compaction

Granular fill with low fines compacts efficiently with vibratory plate or roller equipment. The same gradation with 15% fines becomes moisture-sensitive: too dry it bulks, too wet it pumps. The compaction window narrows to a tight ±1% moisture range, hard to maintain on a tropical Malaysian site.

Friction angle vs density vs angularity: the design parameters

Density (the easy lever)

The friction angle of a granular fill at maximum modified Proctor density (95 to 100% MDD) is typically 4 to 8 degrees higher than the same material at field-loose density (85 to 90% MDD). Compaction is the most cost-effective way to deliver φ on a project. More on compaction effects →

Angularity (the quarry chooses for you)

Angular crushed rock (φ ≈ 36 to 40°) outperforms rounded river gravel (φ ≈ 30 to 34°) for the same gradation. Why: angular particles interlock more aggressively under shear. Crusher run from Malaysian quarries (granite, limestone, basalt) is naturally angular. Beach sand or river gravel is rounded and is rejected for MSE backfill spec.

Gradation (the spec writer chooses)

Well-graded fills (a continuous distribution of particle sizes from d_max down to fines) achieve higher density and higher φ than gap-graded or single-size fills. The standard gradation envelope cited in BS 8006 Annex A and JKR specs is approximately:

This is a relatively forgiving envelope and most crusher run from Malaysian quarries falls inside it.

Granular vs cohesive backfill: the economic question

Cohesive site-won fill is cheaper to source (often nearly free if it's the excavation spoil from the same site) than imported granular fill. Why don't we use it more?

What happens if you back-fill an MSE wall with cohesive soil

1. Effective friction angle is too low. Cohesive Malaysian residual soils often have effective φ = 22 to 28°, vs the 34 to 40° the design assumed. Active earth pressure coefficient jumps from 0.27 (φ = 35°) to 0.40 (φ = 25°), a 50% increase in lateral force on the wall.
2. Drainage fails. Permeability is too low for water to drain out. The wet-season pore pressure builds up and the effective stress holding the reinforcement in place collapses.
3. Creep occurs. Saturated plastic fines continue to deform under sustained load over months and years, with the wall continuing to displace until the active wedge has fully extended.
4. Reinforcement interaction degrades. The interface friction between steel reinforcement and clay-rich backfill is much lower than steel-on-granular, with α (interaction coefficient) often dropping to 0.3 to 0.5 vs 0.8 to 1.2 for granular.

The result: a wall that looks fine on day one but fails progressively over the first wet season. JKR's mandatory granular fill for federal projects exists because the agency has seen this exact failure mode multiple times on contractor-substituted cohesive backfill.

The hybrid approach

Many real Malaysian projects use a hybrid backfill scheme: granular fill in the active reinforced zone (the zone within 0.7 H of the wall face), cohesive site-won fill in the random fill zone behind. This recovers the cost savings from site-won material without compromising the design-critical zone near the wall. AnchorSOL® routinely specifies this approach when the contractor has clean site-won fill to dispose of.

Why AnchorSOL® can run on crusher run

The reason AnchorSOL® specifies crusher run at ≥34° friction angle rather than premium granular fill at ≥36° comes down to the anchored mechanism. In a friction-based MSE wall, the reinforcement pullout resistance is distributed along the strip length and depends on α · σ_v · tan(φ). Every degree of friction angle matters. Drop φ by 2° and the reinforcement length must grow proportionally, eating into project cost and footprint.

In an anchored MSE wall, the pullout resistance is dominated by passive earth pressure on a discrete deadman block: P_r ≈ (K_p − K_a) · γ · z · A_block. K_p at φ = 34° is 3.5; at φ = 36° it's 3.85. A small change in φ produces a small change in passive resistance, and the design is naturally robust to it.

The result: AnchorSOL® can use crusher run from local Malaysian quarries (RM 40 to 70 per m³) instead of premium granular fill (RM 80 to 140 per m³), saving 30 to 50% on backfill cost for a typical project. On a 10,000 m² wall (≈ 30,000 m³ of fill), that's RM 1.2 to 2.1 million in backfill savings alone. See Crusher run as MSE backfill for the full case.

Frequently asked questions

Can I use recycled aggregate (crushed concrete) as MSE backfill?

Yes, with caveats. Recycled aggregate from concrete-crushing typically achieves φ = 34 to 38° and satisfies the gradation envelope. It can have slightly higher fines content (10 to 15%) and a higher water absorption. Acceptable for AnchorSOL® projects after a material-characterisation test programme. Cost is typically 20 to 30% below virgin crusher run on projects near major demolition sites.

What about sand alone?

Clean uniform sand (river sand or beach sand) has φ = 30 to 33° (loose) to 35 to 38° (well-compacted). Acceptable for MSE backfill if well-graded, but rounded sand particles produce lower α (interaction coefficient) with reinforcement than angular crushed rock. Premium granular spec usually requires a minimum sand-gravel ratio.

Does laterite work as backfill?

Laterite is the iron-rich residual soil typical of Malaysian uplands. It is cohesive when intact, with significant plastic fines. Untreated laterite is not acceptable as MSE backfill. Modified laterite (broken to a granular gradation, with fines controlled by screening) can sometimes meet the spec but requires a project-specific test programme.

How is field acceptance verified?

Per BS 1377 Part 9 or ASTM D6938: nuclear density gauge readings on every 200 mm compacted lift, minimum 1 test per 500 m² of lift area. Density must achieve 95% modified Proctor MDD at OMC ± 2%. Plus material acceptance per supply: sieve analysis on every 500 m³ delivered, Atterberg limits where fines exceed 10%.