Permeability, Seepage and Dewatering

Introduction

Water movement through soil controls the performance and safety of many construction works. It affects settlement, shear strength, excavation stability, retaining-wall pressures, embankment safety, pavement drainage, and the durability of foundations. A technically adequate foundation design therefore requires more than knowing the soil’s strength: the engineer must understand how water flows through its pores and how construction activities modify that flow.

This chapter develops Darcy’s law, laboratory and field permeability testing, seepage analysis using flow nets, uplift and piping assessment, and the selection of drainage, cutoff, and dewatering systems. The discussion assumes saturated soil unless stated otherwise. In the field, however, soils are frequently layered, partly saturated, fissured, anisotropic, and spatially variable; these conditions must be considered when interpreting calculations.

Learning Objectives

By the end of this chapter, a student should be able to:

1. Explain permeability and distinguish it from porosity and hydraulic conductivity.
2. Apply Darcy’s law with consistent units and appropriate assumptions.
3. Select and interpret laboratory and field permeability tests.
4. Construct and use flow nets to estimate seepage, pore pressures, uplift, and exit gradients.
5. Identify the mechanisms of piping and internal erosion.
6. Select practical drainage, cutoff, and construction-dewatering methods.
7. Evaluate limitations, environmental effects, and monitoring requirements in seepage control.

1. Groundwater and Hydraulic Head

Water in soil moves because of differences in total hydraulic head. At a point in a saturated soil mass, total head is

where h is total hydraulic head (m), z is elevation head relative to a chosen datum (m), u is pore-water pressure (Pa or kPa), and γ_w is the unit weight of water (N/m³ or kN/m³). The quantity u/γ_w is pressure head.

Water flows from higher total head toward lower total head. The hydraulic gradient is

where i is dimensionless, Δh is the head loss (m), and L is the length of the seepage path (m). In a general three-dimensional field, the gradient is a vector and flow is governed by the gradient in the direction of greatest head decrease.

Permeability, more precisely hydraulic conductivity, measures the ease with which water flows through soil. It depends on pore size, pore connectivity, fluid viscosity and density, and soil fabric. It is not the same as porosity: a soil may have high porosity but low permeability if its pores are small or poorly connected. Typical orders of magnitude are shown below.

These ranges overlap and are only indicative. Density, gradation, fissuring, compaction, mineralogy, and testing method can change the measured value substantially.

2. Darcy’s Law

For one-dimensional, steady, saturated flow through a homogeneous soil, Darcy’s law is

where q is the discharge rate (m³/s), k is hydraulic conductivity (m/s), i is hydraulic gradient, and A is the cross-sectional area normal to flow (m²). The Darcy flux or discharge velocity is

The Darcy velocity is not the actual average velocity of water in the pores. Because water occupies only the void area, the average linear seepage velocity is approximately

where n_e is the effective porosity. This distinction matters when estimating contaminant travel time or the time required for pore-water pressures to dissipate.

Darcy’s law is an empirical relationship, generally valid for laminar flow in saturated soils at ordinary hydraulic gradients. It may become inaccurate in very coarse gravels, open voids, fractures, or situations involving turbulent flow. For layered soils, the direction of flow relative to the layers is critical.

For layers arranged parallel to flow, the equivalent conductivity is the thickness-weighted arithmetic mean:

For flow perpendicular to layers, the equivalent conductivity is the harmonic mean:

Here k_j and H_j are the conductivity and thickness of layer j. Usually k_h is much greater than k_v in stratified deposits. Assuming isotropy in such soils can seriously misrepresent seepage.

Worked Example 1: Darcy Discharge

A soil specimen has k = 2.0 × 10⁻⁵ m/s. Its length is 0.40 m, cross-sectional area is 0.015 m², and the head loss across it is 0.20 m. The discharge is

Thus the discharge is 0.15 cm³/s. The result is dimensionally consistent because (m/s) × (dimensionless) × m² = m³/s.

3. Factors Affecting Hydraulic Conductivity

Hydraulic conductivity increases with larger, better-connected pores. In sands and gravels, grain size and gradation are important; in clays, mineral type, diffuse double-layer effects, fabric, fissures, and stress history are often more important. Increasing void ratio generally increases conductivity, although the relationship depends on soil structure.

Temperature affects viscosity. A test conducted at one temperature may be corrected to a reference temperature using the ratio of water viscosities, provided the soil structure remains unchanged. Compaction generally decreases conductivity by reducing void ratio and pore connectivity. Construction vibration, cracking, desiccation, or piping may instead create preferential flow paths.

The hydraulic conductivity of a soil is a property of the soil–fluid system, not an absolute constant independent of conditions. For groundwater containing unusual contaminants or high concentrations of dissolved substances, fluid viscosity and density should be considered.

4. Laboratory Permeability Tests

4.1 Constant-Head Test

The constant-head test is suited mainly to relatively permeable soils such as clean sands and gravels. A constant head difference is maintained across a specimen, and the collected volume is measured over a known time.

where Q is collected water volume (m³), L is specimen length (m), A is cross-sectional area (m²), h is constant head loss (m), and t is elapsed time (s). The test assumes saturated flow, steady discharge, representative specimen preparation, and negligible leakage or sidewall flow.

4.2 Falling-Head Test

The falling-head test is commonly used for silts and less permeable sands. Water falls in a standpipe connected to the specimen. If the head decreases from h₁ to h₂ over time t, then

where a is the standpipe area (m²), A is the specimen area (m²), and ln denotes the natural logarithm. The heads must be measured relative to the same reference pressure condition. The formula assumes saturated, one-dimensional flow and constant specimen geometry during the test.

4.3 Testing Errors and Interpretation

Important sources of error include incomplete saturation, trapped air, disturbance during trimming, leakage around the specimen, incorrect head measurement, evaporation, sidewall friction, and nonrepresentative layering. The specimen should be de-aired or saturated using an appropriate back-pressure procedure where necessary. Results should be reported with specimen density, water temperature, saturation procedure, stress condition, and test direction.

A laboratory result should not automatically be used as the design value. Field deposits may contain lenses, fissures, root channels, and construction defects that dominate seepage. Multiple tests and conservative interpretation are usually preferable to reliance on a single number.

5. Field Permeability Tests

Field methods are valuable when scale, fabric, or fissuring makes small laboratory specimens unreliable.

• Pumping tests: Water is pumped from a well while drawdown is observed in the pumping and observation wells. They estimate aquifer-scale transmissivity and storage behavior, but require appropriate boundary and steady-state assumptions.
• Slug tests: A sudden change in water level is introduced and recovery is recorded. They are relatively rapid but represent a smaller volume of ground.
• Borehole or packer tests: Water is injected into an isolated borehole interval. They are useful in fractured rock and grouted ground, although results depend strongly on fracture orientation and test pressure.
• Infiltrometer and double-ring tests: These are primarily used for near-surface infiltration and drainage studies rather than deep foundation seepage.

For a confined, homogeneous aquifer under ideal steady radial flow, a simplified pumping relationship may be written as

where Q is pumping rate, b is aquifer thickness, s is drawdown, r_w is well radius, and r_e is an effective radius of influence. This expression is a specialized idealization; real pumping-test interpretation commonly requires transient solutions, well-loss corrections, boundaries, leakage, and anisotropy. It should not be applied indiscriminately to an unconfined or layered aquifer.

6. Seepage Through Soil and Flow Nets

A flow net is a graphical representation of seepage through an isotropic, homogeneous soil. It consists of:

• Flow lines, which indicate the paths followed by water; and
• Equipotential lines, which join points of equal total head.

Flow lines and equipotential lines intersect approximately at right angles in a properly constructed two-dimensional isotropic flow net. The spaces between adjacent flow lines are flow channels, while the spaces between adjacent equipotential lines represent equal head losses.

For a two-dimensional flow net, the discharge per unit width is

where q is discharge per unit width (m²/s), H is total head loss (m), N_f is the number of flow channels, and N_d is the number of potential drops. The expression assumes a well-developed net with approximately curvilinear squares, homogeneous isotropic soil, steady seepage, and negligible three-dimensional effects.

The head loss per potential drop is

At any point, pore-water pressure can be calculated from

The pressure head h − z must be positive for positive gauge pore pressure. If the point lies above the phreatic surface in an unsaturated zone, the simple saturated flow-net interpretation is not sufficient.

Flow-Net Construction

A practical construction procedure is:

1. Define the soil boundaries, impervious boundaries, upstream and downstream water levels, and drainage boundaries.
2. Draw likely flow lines using boundary conditions: an impervious boundary is a flow line, while a constant-head boundary is an equipotential line.
3. Add equipotential lines approximately normal to flow lines.
4. Adjust the network so that the fields are close to curvilinear squares.
5. Count flow channels and potential drops, then calculate discharge and local head.
6. Refine the net near entrances, exits, corners, and drains, where gradients change rapidly.

For anisotropic soil, a transformed-section method may be used. If the principal conductivities are k_x and k_y, the geometry can be transformed using a scale factor related to √(k_x/k_y), depending on the selected coordinate transformation. The resulting flow-net discharge must then be interpreted using the appropriate equivalent conductivity. Numerical methods are often preferable for strongly anisotropic or irregular conditions.

Worked Example 2: Flow-Net Discharge

A sheet-pile cutoff retains water across a homogeneous sand layer. A flow net gives H = 6 m, k = 4 × 10⁻⁶ m/s, N_f = 4, and N_d = 12. Discharge per metre width is

For a 20 m long wall, the estimated total discharge is

This estimate represents seepage through the idealized soil domain. Leakage around the ends of the wall, through joints, or through defects may increase actual flow.

7. Uplift, Exit Gradient, Piping and Internal Erosion

7.1 Uplift Pressure

Seepage beneath a floor slab, dam, weir, or basement can create upward water pressure. If pressure varies linearly between u₁ and u₂ over area A, the resultant uplift is approximately

For a nonuniform pressure distribution, uplift is the integral of pressure over the area:

The effective stress relationship is

where σ′ is effective normal stress, σ is total normal stress, and u is pore-water pressure. Increased uplift reduces effective stress and may reduce the available resistance against sliding, bearing failure, or heave.

7.2 Critical Hydraulic Gradient

For saturated cohesionless soil, the approximate critical upward gradient at which submerged particles become suspended is

where γ′ is submerged unit weight, G_s is specific gravity of solids, and e is void ratio. This is an idealized quick condition for boiling or heave. Natural soil structure, seepage direction, confinement, gradation, and particle interlock influence actual behavior.

The factor of safety against heave is often expressed as

where i_exit is the actual exit gradient. A design must use a suitable margin and account for uncertainty rather than treating the critical value as an acceptable operating gradient.

7.3 Piping and Internal Erosion

Piping is progressive removal and transport of soil particles by seepage. It commonly begins at a downstream exit where gradients are high and confinement is low. Internal erosion can also occur within embankments, around conduits, through cracks, at material interfaces, or through poorly compacted zones.

Common warning signs include muddy discharge, increasing seepage, sinkholes, sand boils, turbid water, settlement, concentrated leaks, and unexpected changes in piezometric levels. A high gradient is a warning indicator, but piping resistance also depends on particle-size distribution, filter compatibility, fabric, stress state, and the presence of cracks or concentrated defects.

Filters should satisfy two competing requirements: retain the base soil while allowing water to pass. Filter design is normally based on validated particle-size criteria and project standards, not on an isolated rule of thumb. Graded filters, drainage blankets, toe drains, relief wells, and properly designed transition zones can control seepage and prevent migration of fines.

8. Drainage and Cutoff Systems

Seepage control generally combines reduction of inflow, safe collection of water, and monitoring.

Cutoffs

Cutoffs lengthen the seepage path and reduce gradients. Common systems include sheet piles, slurry walls, secant or contiguous pile walls, grout curtains, geomembranes, and compacted clay cores. A cutoff may be vertical, inclined, partial, or keyed into a low-permeability stratum. Its effectiveness depends on continuity, penetration, joints, defects, and the hydraulic properties of the underlying layer.

Drainage Systems

Drainage systems collect and convey water without allowing soil particles to migrate. Examples include:

• Drainage blankets and chimney drains in embankments;
• Toe drains and collector pipes;
• Under-slab drainage layers and sump systems;
• Weep holes behind retaining walls;
• Horizontal drains in slopes;
• Relief wells beneath slabs, dams, or hydraulic structures; and
• Prefabricated vertical drains for accelerating consolidation, where their purpose is pore-pressure dissipation rather than excavation pumping.

A drain must have adequate hydraulic capacity, a stable filter or geotextile interface, access for inspection, and an outlet that remains functional. A low-permeability liner without a drainage layer can trap water and produce damaging pressure if not detailed correctly.

9. Construction Dewatering

Dewatering lowers or controls groundwater during excavation and construction. The system should be designed around the required excavation level, soil permeability, aquifer boundaries, allowable drawdown, settlement sensitivity, discharge quality, and construction duration.

Common Methods

The selection should begin with a conceptual groundwater model. A system that merely pumps harder may worsen settlement outside the site, mobilize contaminants, or destabilize the excavation.

Dewatering Design Considerations

The engineer should establish baseline groundwater levels and seasonal variation, characterize soil and aquifer layers, identify nearby basements and utilities, estimate inflow, check drawdown-induced settlement, and plan treatment and discharge. Pumping rates should be increased in stages while piezometers and settlement points are observed. Backup pumps, standby power, alarms, and emergency response are essential where flooding could endanger people or damage adjacent property.

Water discharged from a construction site may contain sediment, hydrocarbons, metals, nutrients, or contaminated groundwater. Discharge permits, turbidity controls, treatment systems, and lawful disposal routes must be addressed before pumping begins. Dewatering is therefore both a geotechnical and an environmental activity.

10. Illustrative Case: Excavation Beside an Existing Building

Background

A contractor must construct a two-level basement 8 m below ground beside an occupied masonry building. A sand layer extends from the ground surface to 14 m depth and overlies a stiff clay aquitard. The groundwater table is 2 m below ground. The excavation is 40 m by 25 m, and the neighboring building is only 5 m from the proposed wall.

Problem

Open sump pumping would likely draw water through the sand toward the excavation. The resulting external drawdown could cause consolidation and settlement beneath the neighboring building. Uncontrolled seepage could also loosen sand at the excavation base.

Analysis and Approach

The project team installs monitoring wells and piezometers inside and outside the site, performs borehole permeability tests, and maps utilities and building foundations. A continuous secant-pile wall is selected as a cutoff and retaining system, extending several metres into the clay aquitard. The wall is not assumed to be perfectly watertight; joints are inspected and pressure-grouted where necessary.

Inside the wall, a staged system of wellpoints and localized sumps controls residual leakage. A drainage layer and collection trench are placed below the basement slab. Pumping is started before excavation and increased in stages. Trigger levels are established for external drawdown, wall movement, building settlement, turbidity, and piezometric response.

Outcome

The cutoff substantially reduces inflow and limits drawdown outside the site. Internal pumping handles leakage and rainfall without producing high exit gradients. Monitoring shows small, recoverable movement, and pumping is adjusted to maintain the required internal water level rather than lowering the surrounding groundwater unnecessarily.

Learning Points

1. A cutoff and a pumping system perform different functions and are often used together.
2. Groundwater control must consider neighboring structures, not only excavation inflow.
3. Field monitoring converts a design assumption into an observational-control process.
4. Filtered drainage and controlled discharge are essential to prevent soil loss and environmental harm.
5. Temporary works require the same level of hydraulic and geotechnical risk management as permanent works.

11. Limitations, Technology and Professional Practice

Flow nets remain useful for developing physical intuition and checking numerical models, but they are limited by idealized geometry, isotropy, two-dimensional assumptions, and subjective drawing. Finite-element and finite-difference seepage models can represent layered, anisotropic, transient, and coupled conditions, but their results are only as reliable as the conceptual model, boundary conditions, material parameters, and mesh.

Modern practice uses vibrating-wire piezometers, automatic water-level loggers, distributed settlement sensors, remote telemetry, and cloud-based dashboards. Artificial intelligence can assist with anomaly detection in monitoring data, but it should not replace engineering judgment, physical verification, or responsibility for safety-critical decisions. Models should be calibrated against observations and subjected to sensitivity analysis.

Good practice includes independent checks of units and water levels, explicit identification of boundary conditions, conservative treatment of uncertain conductivity, staged construction analysis, inspection of cutoffs and filters, and a documented response plan for unexpected seepage. Designers must also comply with applicable building, environmental, occupational-safety, water-discharge, and dam or excavation regulations. Ethical practice requires clear communication of uncertainty, prompt reporting of warning signs, and refusal to conceal unsafe seepage or monitoring results.

Chapter Summary

Permeability describes the ease with which water flows through soil, while hydraulic gradient describes the driving force. Darcy’s law, q = kiA, provides the fundamental relationship for steady laminar saturated flow. Constant-head and falling-head laboratory tests measure conductivity under controlled conditions; field pumping, slug, borehole, and packer tests provide information at larger scales.

Flow nets estimate seepage discharge, head distribution, pore-water pressure, uplift, and exit gradients. Uplift reduces effective stress, while excessive upward gradients may produce heave, boiling, piping, or internal erosion. Filters and drainage layers must retain soil while allowing water to pass. Cutoffs lengthen seepage paths and reduce inflow, whereas wells, wellpoints, sumps, drains, and recharge systems manage water during construction. Effective dewatering requires hydrogeological characterization, staged pumping, monitoring, environmental controls, emergency planning, and professional judgment about uncertainty.

Key Terms

Hydraulic conductivity: The proportionality coefficient relating Darcy flux to hydraulic gradient.

Permeability: Common engineering term for hydraulic conductivity; technically, intrinsic permeability is a soil–fluid-independent property.

Hydraulic head: Elevation head plus pressure head.

Pressure head: Pore-water pressure divided by the unit weight of water.

Hydraulic gradient: Head loss per unit seepage length.

Darcy flux: Discharge rate divided by gross cross-sectional area.

Seepage velocity: Approximate average water velocity through the void space.

Effective porosity: Connected pore volume available for flow divided by total volume.

Constant-head test: Laboratory test for relatively permeable soils using a maintained head difference.

Falling-head test: Laboratory test using the decline of water level in a standpipe.

Flow line: A line tangent to the direction of seepage.

Equipotential line: A line joining points of equal total head.

Flow net: Network of flow lines and equipotential lines used to analyze two-dimensional seepage.

Uplift: Upward water force acting on a structure or soil boundary.

Exit gradient: Hydraulic gradient where seepage leaves the soil mass.

Piping: Progressive removal and transport of soil particles by seepage.

Filter: Graded or manufactured drainage material that retains base soil while passing water.

Cutoff: Low-permeability barrier that lengthens or restricts seepage paths.

Dewatering: Temporary removal or control of groundwater during construction.

Piezometer: Instrument used to measure pore-water pressure or hydraulic head.

Review Questions

Short Answer Questions

1. Distinguish hydraulic conductivity from porosity.
2. State Darcy’s law and define each variable.
3. Why is the falling-head test generally preferred for silts and clays?
4. What are flow lines and equipotential lines?
5. Why can dewatering cause settlement outside an excavation?

Long Answer Questions

1. Explain the construction and interpretation of a flow net, including its boundary conditions and principal assumptions.
2. Compare constant-head, falling-head, pumping, slug, and packer permeability tests.
3. Explain the mechanisms of uplift, heave, piping, and internal erosion.
4. Discuss how cutoffs, filters, drainage blankets, relief wells, and collector drains control seepage.
5. Develop a dewatering strategy for a deep excavation in a layered sand–clay deposit beside existing buildings.

Analytical / Application Questions

1. A soil has k = 3 × 10⁻⁵ m/s, cross-sectional area 0.02 m², length 0.50 m, and head loss 0.15 m. Calculate the Darcy discharge.
2. A flow net has H = 5 m, k = 2 × 10⁻⁶ m/s, N_f = 3, and N_d = 10. Calculate discharge per metre width and explain two reasons actual discharge might differ.
3. An excavation shows increasing turbidity and sand boils at its downstream toe. Diagnose the likely mechanism and propose immediate and permanent control measures.

Multiple Choice Questions with Answers

Question 1

For saturated one-dimensional laminar flow, Darcy’s law is represented by which expression?

A. q = k i A

B. q = k A/i

C. q = i/(kA)

D. q = k + i + A

Correct Answer: A. q = k i A

Question 2

A constant-head permeability test is most suitable for which material?

A. Highly plastic clay

B. Clean sand or gravel

C. Unsaturated expansive clay

D. Intact concrete

Correct Answer: B. Clean sand or gravel

Question 3

In a falling-head test, which quantity is measured as a function of time?

A. The specimen grain specific gravity

B. The decline in standpipe water head

C. The soil total unit weight only

D. The excavation wall movement

Correct Answer: B. The decline in standpipe water head

Question 4

For a two-dimensional flow net, seepage discharge per unit width is calculated as:

A. q = kH(N_d/N_f)

B. q = kH(N_f/N_d)

C. q = H/(kN_fN_d)

D. q = kN_d/(HN_f)

Correct Answer: B. q = kH(N_f/N_d)

Question 5

What is the usual relationship between flow lines and equipotential lines in an isotropic flow net?

A. They are parallel everywhere

B. They intersect approximately at right angles

C. They never intersect

D. They are always at 45 degrees

Correct Answer: B. They intersect approximately at right angles

Question 6

If k = 1 × 10⁻⁵ m/s, i = 0.4, and A = 0.5 m², what is q?

A. 2 × 10⁻⁶ m³/s

B. 2 × 10⁻⁵ m³/s

C. 1.25 × 10⁻⁴ m³/s

D. 2 × 10⁻⁴ m³/s

Correct Answer: D. 2 × 10⁻⁴ m³/s

Question 7

Which condition most directly increases the risk of piping at an excavation exit?

A. A low exit gradient with a stable filter

B. A high exit gradient and unfiltered sandy soil

C. A fully grouted impervious boundary

D. A thick intact clay aquitard

Correct Answer: B. A high exit gradient and unfiltered sandy soil

Question 8

The critical hydraulic gradient for a cohesionless soil depends principally on which properties?

A. Specific gravity and void ratio

B. Plastic limit and liquid limit only

C. Wall height and footing width

D. Water temperature and pipe diameter only

Correct Answer: A. Specific gravity and void ratio

Question 9

Which system is commonly appropriate for deep drawdown in a relatively permeable aquifer?

A. Deep wells

B. A single shallow sump only

C. A surface paint coating

D. A soil nail without drainage

Correct Answer: A. Deep wells

Question 10

Why may a cutoff wall and internal pumping be used together?

A. The cutoff reduces inflow while pumping removes residual water

B. The cutoff increases external drawdown and pumping prevents settlement

C. Both systems serve only to measure pore pressure

D. Pumping makes the cutoff hydraulically unnecessary in every soil

Correct Answer: A. The cutoff reduces inflow while pumping removes residual water