◀ Geology

Engineering geology · hazards

Solid Ground

Every building sits on geology. This is the part where the rock has to hold.

Engineering geology is the question “will it hold?” asked carefully. Read the boring log, work out the factor of safety, know what the ground does when it gets wet — and in Florida, know what’s dissolving underneath you.

Covers ASBOG Domain G — Engineering Geology (12%)

A boring log lands on your desk with a proposed structure. Find the governing hazard — the one that decides whether this design survives.

The call

Loading…

Correct 0Streak 0

The factor of safety is resisting force over driving force. Above 1, the slope stands; below 1, it moves. Watch what happens when you saturate it — this is why landslides come after rain.

25°steepness of the hillside
32°internal strength of the soil
5 kPaclay sticks; clean sand has none
1.00
Factor of safety
Marginal

Florida is a limestone platform with the water table running through it. Dissolve enough rock, remove enough support, and the roof comes down. Learn the three sinkhole types and what tips the odds.

Karst hazard

Loading…

Correct 0Streak 0

A soil is three things at once — solid, water and air — and almost every number in a geotechnical report is a ratio between two of them. Get the phase relationships and the Atterberg limits, and you can classify a soil, predict whether it will swell, and estimate whether water moves through it.

The three-phase model

Weigh the sample wet, dry it, weigh it again. That single pair of numbers — plus the density of the mineral grains — gives you the whole set. These are saturated soils, so every void is assumed full of water.

Wet mass186 g
Dry mass122 g
Grain density ρs2650
Water content w%
Porosity n%
Void ratio e
Bulk density ρkg/m³
Dry density ρdkg/m³

Phase diagram

w = Mw / Ms
Vw = Mw / ρw    Vs = Ms / ρs
n = Vv / V    e = Vv / Vs
n = e/(1+e)    e = n/(1−n)
ρ = M / V    ρd = ρ / (1+w)

Water content is referenced to the dry mass, not the total — which is why a saturated organic clay can honestly report a water content over 100%.

Atterberg limits

A fine-grained soil does not have one strength; it has a sequence of states that depend on how much water it holds. The Atterberg limits are the water contents at which it changes state, and they are the whole basis of fine-grained classification.

Liquid limit (LL)
The water content at which soil starts to flow as a viscous liquid. Measured with the Casagrande cup: a groove is cut in the soil and the cup dropped until the groove closes over 13 mm. Because you cannot hit exactly 25 blows on demand, you run four subsamples at different water contents and read the LL off the flow curve at N = 25.
Plastic limit (PL)
The water content at which the soil stops being mouldable and begins to crumble — determined by rolling a thread until it breaks apart at 3 mm diameter. Four trials, averaged.
Plasticity index (PI = LL − PL)
The width of the plastic range, and the single most useful number in the set. A high PI means the soil stays plastic across a wide range of water contents — which means it holds a lot of water, and means it will shrink and swell as that water comes and goes.

Read the flow curve

Four Casagrande trials. Water content falls as blow count rises; the relationship is linear against log N, so fit a line and read the liquid limit at 25 blows.

Plastic limit24.0 %
Liquid limit%
Plastic limit%
Plasticity index
Flow curve R²

The plasticity chart

Plot PI against LL and the A-line separates clays from silts, while LL = 50 separates low from high plasticity. That single point is the USCS classification for any fine-grained soil.

A-line:  PI = 0.73 (LL − 20)   ·   U-line (upper bound on real soils):  PI = 0.9 (LL − 8)

What the numbers imply about the clay

Plasticity indexSwelling potentialLikely dominant clayWhy
0–15LowKaoliniteLayers hydrogen-bonded tightly together; water cannot get between them. Low surface area, low cation exchange capacity
15–25MediumIllitePotassium ions lock the layers, but less rigidly than kaolinite. Intermediate on every measure
25–40HighIllite–smectite mixturesSome expandable layers present; behaviour scales with the smectite fraction
> 40Very highSmectite (montmorillonite)Water enters between the layers and forces them apart. Enormous surface area and very high cation exchange capacity — the soil that cracks in drought and heaves in wet

The same property that makes smectite a foundation problem makes it good at holding contaminants: a high cation exchange capacity means dissolved metals adsorb onto the clay rather than travelling with the water. The next tab follows that through.

The question a site investigation actually has to answer is rarely “what is this soil”. It is will the contaminant move — and that depends on how fast water can travel through the pores, and how tightly the contaminant is held by the grains it passes.

Kozeny–Carman: permeability from texture

You can estimate intrinsic permeability from three things you already measured: porosity, median grain size, and grain shape. Permeability scales with the square of grain size, which is why a small change in texture produces an enormous change in how water moves.

k =  ψ2 · n3 · Dp2
180 (1 − n)2
Porosity n0.65
D504 µm
Shape ψ0.65
Permeability k
In millidarciesmd
Hydraulic conductivity Km/s

The shape factor

The Corey shape factor compares the short axis with the two longer ones:

ψ = c / √(a · b)

with a long, b intermediate and c short. A sphere gives 1.0; well-rounded sand runs about 0.7–0.8; angular silt around 0.6–0.7; platy clay flakes can fall below 0.5. Flatter grains pack with more tortuous pore throats, so water has further to travel — which is what the factor is standing in for.

Also note k and K are different things. Intrinsic permeability k (m²) is a property of the rock alone. Hydraulic conductivity K (m/s) folds in the fluid: K = kρg/µ, so the same soil conducts warm water faster than cold, and water far faster than oil.

Know where this equation fails. Kozeny–Carman is derived for a bundle of capillary tubes through uniform granular material. It reproduces measured values well for sands and silts, but for clay-rich soils it commonly overestimates permeability by three orders of magnitude or more, because it takes no account of electrostatically bound water, the platy shape of clay particles, or the way the fabric itself changes as the clay consolidates. For a clay, treat the answer as an upper bound, and get a falling-head or consolidation test before anything depends on the number.

Will the contaminant travel?

Two properties decide it, and they usually point the same way.

1 · Can the water move?
A clean sand at 10−4 m/s moves metres a day; a fat clay at 10−9 m/s moves centimetres a year. Five orders of magnitude decide whether a plume reaches a well in a season or in a human lifetime — and it is why clay liners are what landfills are built with.
2 · Does the contaminant stay with the water?
Dissolved metals such as lead carry a positive charge, and clay surfaces carry a negative one. So they adsorb — and how strongly depends on cation exchange capacity, which scales with surface area and therefore with clay mineralogy. Smectite has a CEC an order of magnitude above kaolinite. Lead in particular binds strongly, and a high-PI clay holds onto it hard.
The two together
A high-PI, smectite-rich, low-permeability soil is doubly retentive: the water barely moves, and what the water carries is stripped onto the grains as it goes. The contamination stays where it is. That is genuinely good news for the aquifer — and it also means the material is a long-lived reservoir, so disturbing it is the real hazard. Conversely, a sandy, low-PI, high-permeability soil retains almost nothing, and a plume there is already moving.
Retardation, put as a number
The retardation factor R compares how fast the contaminant travels with how fast the water does: R = 1 + ρdKd/n, where Kd is the partition coefficient. R = 1 means no adsorption — the contaminant moves at the speed of the groundwater. R = 50 means it moves at a fiftieth of that speed. For lead on a clay-rich soil R is routinely in the hundreds or thousands, which is why lead contamination tends to stay near where it entered.

Dig it out, or leave it?

Excavation is expensive and it is not automatically the right answer. Work the cost, then weigh it against what the numbers above actually say about risk.

Area100
Depth1.00 m
Truck volume27
Cost per load$2000
Volume
Truckloads
Disposal cost

What the cost figure leaves out

Disposal is the easy number, and it is usually the smaller one. A real estimate also carries excavation and haulage, dewatering a saturated marsh soil, characterising the waste before a landfill will accept it, disposal at a hazardous rather than a clean-fill rate if lead exceeds threshold, replacement backfill, and site restoration.

And excavation itself creates exposure: digging a wetland mobilises material that was stable, and the truck route matters.

Monitored natural attenuation — leaving contamination in place and watching it — is a legitimate remedy where the soil demonstrably immobilises the contaminant. Making that case is exactly what the plasticity and permeability numbers are for. Round up truckloads: half a load still needs a whole truck.

The reference bench: soils by behaviour, foundations by problem, the tests that tell you what’s down there.

Soil by particle size & behaviour

Grain size drives everything — drainage, strength, and how the soil behaves when it gets wet.

Unified Soil Classification System (USCS)

Two letters. The first says what the soil is, the second says how it is graded or how plastic it is. Learn the letters and you can read any geotechnical log.

Foundations & ground treatment

Site investigation methods

Hazards worth knowing cold