Permeability of Soil — Darcy's Law & Tests
Hydraulic conductivity, constant head and falling head tests, equivalent permeability for stratified soils, and factors affecting permeability — with GATE-level worked examples
Last Updated: March 2026
Key Takeaways
- Darcy’s Law: v = ki, where v = discharge velocity (m/s), k = coefficient of permeability (m/s), i = hydraulic gradient = h/L.
- Seepage velocity vs = v/n (actual velocity through pores); always greater than discharge velocity v.
- Constant head test (IS 2720 Part 36): k = QL/(Aht); used for coarse-grained soils (sands, gravels) with k > 10⁻⁴ m/s.
- Falling head test (IS 2720 Part 36): k = (aL/At) ln(h1/h2); used for fine-grained soils (silts, clays) with k < 10⁻⁴ m/s.
- For stratified soil — flow parallel to layers: kH = (k1H1 + k2H2 + …) / H (weighted average); flow perpendicular: kV = H / (H1/k1 + H2/k2 + …) (harmonic mean).
- kH > kV always for stratified deposits (horizontal permeability always exceeds vertical).
- Typical k values: clean gravel 10⁻² m/s; clean sand 10⁻⁴ to 10⁻⁵ m/s; silt 10⁻⁶ to 10⁻⁸ m/s; clay < 10⁻⁹ m/s.
1. Introduction to Soil Permeability
Permeability (or hydraulic conductivity) is the ability of a soil to allow water to flow through its interconnected void spaces under a hydraulic gradient. It is one of the most fundamental properties in geotechnical engineering — it governs seepage beneath dams and through embankments, controls the rate of consolidation in clay layers, determines the drainage characteristics of foundations, and is central to the design of retaining structures where water pressure must be managed.
Permeability varies by more than ten orders of magnitude across different soil types — from clean gravels (k ≈ 10⁻² m/s) to intact clays (k ≈ 10⁻¹⁰ m/s). This enormous range means that the flow behaviour of gravel and clay are almost incomparable: a gravel allows water to pass freely, while a clay is essentially impermeable on any engineering timescale.
Water flow through soil pores is laminar (not turbulent) under typical hydraulic gradients — this is the fundamental assumption of Darcy’s Law.
2. Darcy’s Law
Henry Darcy (1856), while studying water filtration through sand filters for the town of Dijon, France, discovered that the rate of flow through a porous medium is proportional to the hydraulic gradient. This empirical relationship, now known as Darcy’s Law, forms the foundation of all seepage analysis.
2.1 Darcy’s Law — Equations
v = k i
v = discharge velocity (m/s) — velocity based on total cross-sectional area
k = coefficient of permeability / hydraulic conductivity (m/s)
i = hydraulic gradient = Δh / L (dimensionless)
Q = k i A
Q = flow rate (m³/s), A = total cross-sectional area of flow (m²)
2.2 Discharge Velocity vs Seepage Velocity
Discharge velocity: v = Q / A (A = total area including solids)
Seepage velocity: vs = Q / Av = v / n
Av = area of voids = n × A
Since n < 1 always: vs > v always
The discharge velocity v is a fictitious average velocity — it assumes water flows through the entire cross-section including solid particles. The seepage velocity vs is the actual average velocity of water through the pore channels. In contaminant transport and seepage analysis, the seepage velocity is the physically meaningful quantity.
2.3 Validity of Darcy’s Law
Darcy’s Law applies when flow is laminar — governed by viscous forces rather than inertial forces. The validity criterion is expressed through the Reynolds number for porous media:
Re = v D10 / ν ≤ 1 (laminar; Darcy’s Law valid)
Re > 1–10 → turbulent flow; Darcy’s Law overestimates flow rate
ν = kinematic viscosity of water ≈ 10⁻⁶ m²/s at 20 °C
For most soils (silts through gravels) under normal hydraulic gradients, flow is laminar and Darcy’s Law holds. It breaks down for very coarse soils (coarse gravels, rockfill) under high gradients.
3. Hydraulic Head and Gradient
3.1 Total Head
Total head h = pressure head + elevation head = u/γw + z
(velocity head v²/2g is negligible for flow through soil)
3.2 Hydraulic Gradient
i = Δh / L = (h1 − h2) / L
Δh = head loss between two points (m)
L = seepage path length between those points (m)
i is dimensionless; typical values: 0.2–5 for laboratory tests; 0.01–0.5 in field situations
3.3 Critical Hydraulic Gradient (Quick Condition)
When upward seepage force equals the submerged weight of soil, effective stress becomes zero and the soil loses all strength — this is the quicksand or piping condition.
icr = (Gs − 1) / (1 + e)
For Gs = 2.67, e = 0.67: icr = (2.67−1)/(1+0.67) = 1.67/1.67 = 1.0
Typical icr ≈ 0.9–1.1 for most sandy soils
Factor of safety against piping: FOS = icr / iexit ≥ 3–4
4. Constant Head Test (IS 2720 Part 36)
The constant head test maintains a constant water level difference across a soil specimen throughout the test. Water flows through the soil under this fixed head difference and the flow rate is measured by collecting the outflow over a known time period.
4.1 Test Setup
A cylindrical specimen of length L and cross-sectional area A is placed in a permeameter. Water enters at the top under a constant head h (maintained by an overflow arrangement) and exits at the bottom. The volume Q of water collected in time t is measured.
4.2 Formula
From Darcy’s Law: Q = k i A t = k (h/L) A t
k = Q L / (A h t)
Q = volume of water collected (cm³ or m³)
L = length of specimen (cm or m)
A = cross-sectional area of specimen (cm² or m²)
h = constant head difference (cm or m)
t = time of collection (s)
4.3 Application
Used for coarse-grained soils (clean sands and gravels) where k > 10⁻⁴ m/s. Flow rates are measurable within a reasonable collection time. For fine-grained soils, the flow rate is too small to measure accurately by this method.
5. Falling Head Test (IS 2720 Part 36)
In the falling head test, water is allowed to flow through the soil specimen from a standpipe of small cross-sectional area. As water flows out, the head in the standpipe falls. The head at the start and end of a time interval is recorded and k is back-calculated.
5.1 Formula Derivation
At time t, head in standpipe = h. In small time dt, head falls by dh:
Flow out of standpipe = −a dh (negative because h decreases)
Flow through specimen = k (h/L) A dt
Setting equal: −a dh = k (h/L) A dt
Integrating from h1 at t = 0 to h2 at t = t:
k = (aL / At) ln(h1 / h2)
Or: k = 2.303 (aL / At) log(h1 / h2)
a = cross-sectional area of standpipe (cm² or m²)
A = cross-sectional area of specimen (cm² or m²)
L = length of specimen (cm or m)
h1 = initial head (cm or m)
h2 = final head after time t (cm or m)
5.2 Application
Used for fine-grained soils (silts and clays) where k < 10⁻⁴ m/s. The falling head method is sensitive enough to measure very small flow rates. For very low permeability clays, even this method may be impractical and consolidation test back-calculation of cv is used to infer k.
6. Equivalent Permeability — Stratified Soils
Natural soil deposits are usually layered (stratified). The equivalent permeability of the layered system depends on the direction of flow relative to the layers.
6.1 Flow Parallel to Layers (Horizontal Flow)
Each layer carries flow proportional to its k and thickness. Total flow = sum of flows in each layer.
kH = (k1H1 + k2H2 + … + knHn) / H
H = H1 + H2 + … + Hn = total thickness
kH is a weighted arithmetic mean — dominated by the most permeable layer
6.2 Flow Perpendicular to Layers (Vertical Flow)
Head loss is distributed across layers; flow rate is the same through all layers.
kV = H / (H1/k1 + H2/k2 + … + Hn/kn)
kV is a weighted harmonic mean — dominated by the least permeable layer
6.3 Relationship Between kH and kV
kH ≥ kV always (arithmetic mean ≥ harmonic mean)
Equality only when all layers have the same k
Typical field ratio: kH/kV = 2–10 for alluvial deposits
This inequality has profound practical implications: horizontal drainage (e.g., sand drains, horizontal drains) is far more effective than vertical drainage through a layered soil profile. Even a thin clay layer in an otherwise permeable sand deposit can dramatically reduce vertical permeability.
7. Factors Affecting Permeability
| Factor | Effect on k | Reason |
|---|---|---|
| Grain size | k ∝ D10² (Hazen’s formula) | Larger grains → larger pores → less resistance to flow |
| Void ratio (e) | k increases with e | More void space → easier flow paths; k ∝ e³/(1+e) (Kozeny-Carman) |
| Degree of saturation | k decreases as S decreases below 100 % | Air bubbles block pore channels |
| Soil structure / fabric | Flocculated > dispersed for same e | Flocculated structure has larger inter-aggregate pores |
| Fluid viscosity μ | k ∝ 1/μ | Viscosity decreases with temperature → k increases at higher temperature |
| Organic content | k decreases | Organic matter fills pores and swells when wet |
| Clay content / mineralogy | k decreases sharply | Clay particles clog pores; swelling further reduces pore size |
7.1 Hazen’s Empirical Formula
k = C D10²
k in cm/s, D10 in mm
C = 100 for loose uniform sand (range: 100–150)
Valid for: 0.1 mm ≤ D10 ≤ 3 mm, Cu ≤ 5
7.2 Kozeny-Carman Equation
k = CKZ × e³/(1+e) (CKZ depends on grain shape and specific surface)
More rigorous than Hazen; accounts for void ratio change during consolidation
Used to estimate change in k during consolidation: k2/k1 = [e2³/(1+e2)] / [e1³/(1+e1)]
8. Typical k Values and Drainage Classification
| Soil Type | k (m/s) | Drainage Classification |
|---|---|---|
| Clean gravel | 10⁻² to 10⁻¹ | Free draining |
| Coarse sand | 10⁻³ to 10⁻² | Free draining |
| Medium sand | 10⁻⁴ to 10⁻³ | Free draining |
| Fine sand / loose silt | 10⁻⁵ to 10⁻⁴ | Poorly draining |
| Dense silt / silty clay | 10⁻⁷ to 10⁻⁵ | Practically impermeable |
| Intact clay | 10⁻¹⁰ to 10⁻⁸ | Impermeable |
| Fissured clay | 10⁻⁷ to 10⁻⁵ | Depends on fissure spacing |
9. Field Permeability Tests
| Test | Principle | Best For |
|---|---|---|
| Pumping test | Pump water from a well at constant rate; measure drawdown in observation wells; k from Dupuit-Thiem equation | Large-scale in-situ k of aquifer; coarse soils |
| Borehole tests (slug test) | Instantaneous injection or withdrawal; monitor head recovery; k from recovery curve | Small-scale in-situ k; any soil |
| Piezocone (CPTU) | Dissipation of excess pore pressure after cone penetration; k inferred from dissipation rate | Fine-grained soils; quick assessment |
Field permeability values are typically higher than lab values because undisturbed samples capture the effect of natural fissures, root channels, and macro-pores that are not reproduced in remoulded lab specimens.
10. Worked Examples
Example 1 — Constant Head Test
Problem: A constant head permeability test is performed on a sand sample. Length of sample L = 300 mm, diameter = 100 mm. Under a head of h = 500 mm, 450 cm³ of water is collected in 5 minutes. Find k.
Solution
Q = 450 cm³, t = 5 × 60 = 300 s, L = 30 cm, h = 50 cm
k = QL / (Aht) = (450 × 30) / (78.54 × 50 × 300)
k = 13500 / 1178100 = 1.146 × 10⁻² cm/s = 1.15 × 10⁻⁴ m/s
This value is consistent with a medium-coarse sand. ✓
Example 2 — Falling Head Test
Problem: A falling head test is conducted on a silt sample. Sample: L = 150 mm, A = 50 cm². Standpipe diameter = 4 mm (a = π/4 × 0.4² = 0.1257 cm²). Initial head h1 = 1000 mm, final head h2 = 400 mm after t = 900 s. Find k.
Solution
= (0.1257 × 15) / (50 × 900) × 2.303 × log(100/40)
= 1.886 / 45000 × 2.303 × log(2.5)
= 4.191 × 10⁻⁵ × 2.303 × 0.3979
= 4.191 × 10⁻⁵ × 0.9164
k = 3.84 × 10⁻⁵ cm/s = 3.84 × 10⁻⁷ m/s
Value consistent with a silty soil. ✓
Example 3 — Stratified Soil: Equivalent Permeability
Problem: A soil deposit has three horizontal layers:
| Layer | Thickness H (m) | k (m/s) |
|---|---|---|
| 1 (Sand) | 2.0 | 3 × 10⁻⁴ |
| 2 (Silt) | 1.0 | 5 × 10⁻⁷ |
| 3 (Sand) | 3.0 | 2 × 10⁻⁴ |
Find kH (horizontal) and kV (vertical). Total H = 6.0 m.
Horizontal Permeability (parallel flow)
= (3×10⁻⁴×2 + 5×10⁻⁷×1 + 2×10⁻⁴×3) / 6
= (6×10⁻⁴ + 5×10⁻⁷ + 6×10⁻⁴) / 6
= (1.2×10⁻³ + 5×10⁻⁷) / 6 ≈ 1.2005×10⁻³ / 6
kH = 2.0 × 10⁻⁴ m/s
Vertical Permeability (perpendicular flow)
= 6 / (2/3×10⁻⁴ + 1/5×10⁻⁷ + 3/2×10⁻⁴)
= 6 / (6667 + 2×10⁶ + 15000)
= 6 / (2×10⁶ + 21667) ≈ 6 / 2.022×10⁶
kV = 2.97 × 10⁻⁶ m/s
kH/kV = 2.0×10⁻⁴ / 2.97×10⁻⁶ ≈ 67 — the thin silt layer dominates vertical permeability.
Example 4 — GATE-Style: Critical Hydraulic Gradient
Problem (GATE CE type): A sand deposit has Gs = 2.65 and void ratio e = 0.65. Compute the critical hydraulic gradient. If the actual upward hydraulic gradient is 0.6, find the FOS against piping.
FOS = icr / iactual = 1.0 / 0.6 = 1.67
FOS = 1.67 < 3 → insufficient safety against piping; design intervention required (sheet pile cutoff, drainage blanket, or filter protection).
11. Common Mistakes
Mistake 1: Confusing Discharge Velocity with Seepage Velocity
What happens: The seepage velocity vs = v/n is used in Darcy’s Law (Q = vs A n = v A) — but substituting vs directly into Q = vA gives a flow rate that is n times too high.
Fix: Darcy’s Law Q = kiA uses the discharge velocity v = ki. The seepage velocity vs = v/n is the actual pore velocity — use it only when computing travel times of contaminants or tracers, not for flow rates.
Mistake 2: Using Arithmetic Mean for Vertical (Perpendicular) Permeability
What happens: kV = (k1H1 + k2H2)/H is used for both horizontal and vertical flow, giving kV = kH — physically impossible for a stratified deposit.
Fix: Horizontal (parallel) flow = arithmetic mean. Vertical (perpendicular) flow = harmonic mean: kV = H / (H1/k1 + H2/k2 + …). Always kH ≥ kV.
Mistake 3: Using the Constant Head Formula for Fine-Grained Soils
What happens: k = QL/(Aht) is applied to a silt or clay, but the collected volume Q is negligibly small over any reasonable test time, giving an unreliable k.
Fix: Constant head test → coarse soils (k > 10⁻⁴ m/s). Falling head test → fine soils (k < 10⁻⁴ m/s). For very low permeability clays, back-calculate k from the coefficient of consolidation cv.
Mistake 4: Forgetting Units — Mixing cm/s and m/s
What happens: All lengths in the formula are converted to cm except the flow rate Q which is left in m³/s, giving k in incorrect units. This is a very common source of error in GATE numerical problems.
Fix: Before substituting into any permeability formula, convert all quantities to a consistent unit system (all cm and cm³, or all m and m³). State units explicitly in every step.
Mistake 5: Applying Hazen’s Formula Outside its Valid Range
What happens: Hazen’s formula k = CD10² is used for a well-graded sand with Cu = 12 or for a clay with D10 = 0.001 mm. Hazen’s formula is only valid for clean uniform sands with D10 = 0.1–3 mm and Cu ≤ 5.
Fix: Check the validity limits before applying Hazen’s formula. Outside this range, use laboratory permeability tests or Kozeny-Carman for a more reliable estimate.
12. Frequently Asked Questions
Q1. Why is permeability temperature-dependent, and how is it corrected?
Darcy’s Law is fundamentally a viscous flow equation — the resistance to flow through pore channels depends on the viscosity of water. Viscosity decreases significantly with temperature: water at 4 °C is about twice as viscous as at 40 °C. Since k = kintrinsic × γw/μ, where kintrinsic depends only on soil structure, k measured in the lab varies with the water temperature during the test. IS 2720 Part 36 requires permeability to be reported at 27 °C (standard temperature for India). The correction is: k27 = kT × (μT/μ27), where μT is the viscosity at test temperature T. For most practical purposes in GATE problems this correction is not required unless specifically asked.
Q2. Why is the field permeability of clay usually higher than lab-measured values?
Laboratory permeability tests are performed on small, reconstituted or trimmed specimens that do not capture the macrostructure of natural clay deposits. Natural clays often contain fissures, root channels, sand lenses, and organic inclusions that create preferential flow paths with permeability several orders of magnitude higher than the intact clay matrix. When the test specimen is trimmed, these features are either absent or closed. Field permeability tests (pumping tests, piezometer dissipation tests) measure the bulk permeability of the deposit including all its structural features, which is the relevant value for seepage analysis under dams, through embankments, and around sheet pile walls.
Q3. What is the relationship between the coefficient of permeability k and the coefficient of consolidation cv?
In Terzaghi’s consolidation theory, the rate at which pore water drains from a compressing clay layer is governed by both its permeability (k) and its compressibility (mv): cv = k / (γw mv), where mv is the coefficient of volume compressibility. A clay with high permeability consolidates faster (higher cv); a highly compressible clay consolidates slower (lower cv for the same k). This relationship allows k to be back-calculated from consolidation test data — a more reliable method for very low permeability clays than direct permeability testing.
Q4. What happens to the permeability of a soil after compaction?
Compaction significantly reduces permeability, particularly for fine-grained soils and on the wet side of OMC. At OMC, compaction reduces void ratio and creates a more uniform, dispersed soil structure with smaller, less interconnected pores. Compaction wet of OMC produces a highly dispersed, low-permeability structure (k can be 10–100 times lower than dry-side compaction). Dry-side compaction produces a more open, flocculated structure with higher inter-aggregate pores and higher permeability. For dam cores and liners where low permeability is required, specifications always require compaction at or slightly wet of OMC. For subgrades and fills where strength is the priority, OMC compaction balances strength and permeability requirements.