Carbonate sediments, environments and carbonate porosity

Quick refresher

Carbonates form in the basin itself. The grains are skeletons of organisms and their fragments. The mud is released by algae and microbes. So the factory works where it is warm, bright and shallow, and where little clay and sand come from the land.
Light fades exponentially with depth: each metre takes away the same share of the light that reaches it. The production of reefs and algae is close to its maximum only in the top 10–20 m of clear water. At the bottom of the photic zone (the sunlit layer of water) it falls to a small share of the maximum. In clear water this limit lies at a depth from a few tens of metres to a little over a hundred. In muddy water it is at about 20 m or shallower.
Aragonite and magnesian calcite, from which many organisms build their skeletons, are unstable. Once the sediment is laid down, aragonite dissolves or recrystallises (forms new crystals), and magnesian calcite loses its magnesium. Ancient limestones are made of low-magnesium calcite, and the dolomite in them is mostly secondary (it formed later).
From land to sea, the environments are the sabkha (a salty coastal flat in a hot climate) and tidal flat, the lagoon, the shoals and reefs of the margin, the slope and the basin. On a ramp the sea floor slopes down gently, usually at less than one degree, and the belts (strips of one environment) stretch for kilometres or tens of kilometres. At a rimmed shelf the margin ends in a steep slope, and the belts are narrow.
In the Choquette–Pray classification, pores are fabric selective (interparticle, intraparticle, intercrystal, moldic, fenestral, shelter and growth-framework pores), not fabric selective (fractures, channels, vugs and caverns), or mixed. “Fabric selective” means that the pore follows the structure of the rock: its grains, crystals or the framework built by organisms.
Permeability depends on the size of the particles and on how well the pores connect, not on porosity. Separate moldic pores (voids left where grains dissolved) and vugs add porosity with almost no permeability. Fractures add permeability with almost no porosity.

Key relationships

\dfrac{I(z)}{I_0} = e^{-k z}

fading of light with depth; the lower limit of the photic zone is $z_{1\%} = \ln 100/k$

G(z) = G_m \tanh\dfrac{I_0\,e^{-k z}}{I_k}

production of carbonate, limited by light

V_M = \dfrac{M}{\rho}

molar volume: calcite takes up about 8% more room than the same amount of aragonite

\Delta x = \dfrac{\Delta z}{\tan\alpha}

width of a belt in which the depth changes by $\Delta z$ , on a sea floor with slope $\alpha$

Common pitfalls

Expecting high permeability from a porous carbonate: moldic pores and vugs are often separate.
Judging the porosity of a carbonate from the sonic log alone (a well log that measures how fast sound travels through the rock): it hardly sees vugs.
Judging a layer from core plugs (small cylinders of core): they cannot hold large vugs and rarely cross fractures.
Applying the properties of one facies (one type of rock, laid down in one environment) to the whole buildup: the porosity of a reef margin and of a lagoon can differ several times.
Believing that dolomite is always more porous than limestone: excess dolomite and anhydrite cement close the pores.

After this lesson you can

explain how the carbonate factory works and why carbonates collect where it is warm, bright and shallow and where no sand or clay comes from the land, and estimate how production falls with depth
name the main carbonate grains and minerals (aragonite, magnesian calcite, low-magnesium calcite and dolomite) and explain why the minerals that make up a sediment decide the fate of its pores
describe the environments from the sabkha and tidal flat to the slope and basin, and tell a ramp, a rimmed shelf and an isolated platform apart
classify carbonate pores by the Choquette–Pray scheme, tell pores that follow the rock fabric from pores that do not, and tell separate vugs from touching vugs
estimate how the type of pores affects permeability, log readings and saturation
link the structure of the subsalt carbonate buildups (reef-like mounds that lie below a thick salt layer) of the Pre-Caspian Basin to the places where their porosity is concentrated

The carbonate factory: where carbonates are born

Clastic grains (grains broken off older rock) are born on land, when mountains wear away. Rivers, wind and ice carry them into the basin. Carbonate grains mostly form in the basin itself, at or near the spot where sediment will bury them. Almost all the carbonate of shallow seas is made of skeletons and shells of organisms, their fragments, and limy mud. Limy algae release this mud, and it also settles out of warm water with the help of microbes. So a carbonate sediment is the product of a living community. We speak of it as the product of a carbonate factory.

The factory needs conditions in which its “workers” thrive:

Light. Reef-building corals live in symbiosis (a partnership) with single-celled algae. Limy algae make food from light themselves (photosynthesis). So their production falls with depth, together with the light.
Warmth. Reefs and ooid shoals (banks of small round limy grains) are a sign of tropical water. Reef-building corals of shallow water do not tolerate water colder than about 18 °C. Carbonate also collects in cold seas, but in a different way (see below).
Clear water. Clay and sand brought by rivers make the water muddy, bury the organisms and dilute the carbonate. So carbonate platforms grow far from the mouths of big rivers, and more often in a dry climate.
Normal salinity and water poor in nutrients. When there are too many nutrients, algae and plankton multiply fast, the water turns cloudy, and corals lose the competition. In lagoons with extra salty water, fewer kinds of organisms live. But microbial mats (layers of microbes) spread, and evaporites (salts left when water evaporates) settle out.
Waves and currents. They spread the production around. Mud is carried into lagoons and onto the slope, and sand piles up on shoals.

Light is the most important of these factors, and it is easy to describe in numbers. Each next metre of water absorbs the same share of the light that reaches it. So light fades exponentially with depth.

Fading of light in water

\frac{I(z)}{I_0} = e^{-k z}

where:

$I(z),\ I_0$ light that can drive photosynthesis, at depth $z$ and at the surface, µmol photons/(m²·s)
$k$ attenuation coefficient: how strongly the water absorbs light, 1/m
$z$ depth, m

Units: SI: $k$ in 1/m, $z$ in metres

Holds when

The water is equally clear at all depths. In reality $k$ changes with depth, season and weather.
In the clearest ocean water $k$ is only a few hundredths (about 0.03–0.05 1/m). In muddy coastal water it is 0.2 1/m or more.

We take the depth where 1% of the surface light is left as the lower limit of the photic zone (the sunlit layer of water): $z_{1\%} = \ln 100/k \approx 4.61/k$ . In the clearest ocean water this is 90–150 m. In clear water at reefs it is usually a few tens of metres, and in muddy coastal water about 20 m or less.

Production does not keep growing with light for ever. In bright light, photosynthesis becomes saturated: it cannot go any faster, and other factors limit growth. We describe this saturation with the hyperbolic tangent, a curve that rises and then levels off. Reef-growth models use it.

Carbonate production limited by light

G(z) = G_m \tanh\frac{I_0\,e^{-k z}}{I_k}

where:

$G(z)$ production: the rate at which carbonate builds up the sea floor at depth $z$ , mm/yr
$G_m$ maximum production when there is more light than needed, mm/yr
$I_k$ saturation light: when $I = I_k$ , production is $\tanh 1 \approx 0.76$ of the maximum, µmol photons/(m²·s)
$I_0,\ k,\ z$ light at the surface, attenuation coefficient and depth, as above, µmol photons/(m²·s), 1/m, m

Units: $G$ and $G_m$ in mm/yr (1 mm/yr is 1 m in a thousand years), $I_0$ and $I_k$ in the same units of light, $k$ in 1/m, $z$ in metres

Holds when

Only light limits production. Temperature, nutrients and salinity are already built into $G_m$ , which holds for one community of organisms in one set of conditions.
The equation describes the build-up of the sea floor in place. Waves and currents carry part of the production away, and organisms destroy another part. So the average rate at which a carbonate sequence builds up over thousands and millions of years is much smaller than $G_m$ .
The parameters of the model are fitted to observations of modern reefs. In the example they are assumed for illustration.

Near the surface there is plenty of light, and production is close to the maximum. Deeper down, light becomes the limit, and production falls fast. The muddier the water (the larger $k$ ), the closer to the surface the “floor” of the factory rises.

Worked exampleHow the production of a reef falls with depth

Given (assumed data)

Maximum production $G_m$: 8 mm/yr (a community of corals and algae)
Light at the surface $I_0$ and saturation light $I_k$: 2000 and 200 µmol photons/(m²·s), $I_0/I_k = 10$
Light attenuation $k$: 0.08 1/m in clear water and 0.2 1/m in muddy water

Solution

Clear water, 10 m: $I/I_0 = e^{-0.08\cdot 10} = 0.449$ and $G = 8\tanh(10\cdot 0.449) = 8.0$ mm/yr. At 20 m, $I/I_0 = 0.202$ and $G = 7.7$ mm/yr.
At 30 m, $I/I_0 = 0.091$ and $G = 8\tanh(10\cdot 0.091) = 5.8$ mm/yr. At 50 m, $I/I_0 = 0.018$ and $G = 1.4$ mm/yr.
Half of the maximum production, 4 mm/yr, occurs where $\tanh x = 0.5$ , that is $x = 0.549$ . So $z = \ln(10/0.549)/0.08 = 2.902/0.08 = 36$ m. The lower limit of the photic zone is $z_{1\%} = 4.61/0.08 = 58$ m.
Muddy water ( $k = 0.2$ 1/m): at 10 m, $G = 7.0$ mm/yr, and at 30 m, $G = 0.2$ mm/yr. Half of the production is at a depth of $2.902/0.2 = 15$ m, and 1% of the light is at $4.61/0.2 = 23$ m.

Answer: In clear water a reef grows at almost full speed in the top twenty or so metres. It grows half as fast at about 36 m. By 58 m, where 1% of the light is left, its growth falls to a tenth of the maximum. In muddy water these marks rise to 15 and 23 m.

What it means. Three main properties of carbonate platforms follow from this. They are flat on top and grow up to sea level, because what lies shallow builds up fastest. They are sensitive to changes in sea level: if the top sinks too deep, it can no longer keep up (see the advanced section). And they avoid muddy water: near river mouths the factory is suppressed. The numbers are of a realistic size. Modern reefs build up the sea floor by millimetres a year. The average rate at which carbonate sequences accumulate over millions of years is only hundredths of a millimetre a year. This is because a long interval includes breaks in deposition and erosion (unconformities, gaps in the rock record). The data in this example are assumed for illustration.

CalculatedProduction of a reef against depth, in the model of the lesson, for clear (

k = 0.08

1/m) and muddy (

k = 0.2

1/m) water. The points mark the depths at which production falls by half. The parameters are assumed for illustration.

Carbonates of cold seas and of the open ocean

Not all carbonates are tropical. In temperate and cold seas, limy sand comes from the shells of molluscs, bryozoans (colonial animals that build lacy crusts), foraminifera, barnacles and red algae. Most of these organisms do not need light. So such sediments also collect at great depths, but more slowly. They have no tropical reefs and no ooids, and usually almost no limy mud.
In the open ocean the main producers are plankton: coccolithophores (single-celled algae with tiny limy plates) and foraminifera. Their skeletons range in size from a few micrometres (the plates of coccoliths) to fractions of a millimetre (the shells of foraminifera). They sink to the bottom as limy mud. Below a certain level, cold water under high pressure dissolves carbonate faster than it arrives. This level is the carbonate compensation depth (CCD), and in today's oceans it is about 4–5 km. Below it, only muds rich in silica and clay muds remain on the sea floor. Compacted limy mud becomes chalk.

Carbonate grains and minerals

The names of the grains (bioclasts, ooids, peloids and intraclasts) and how they are used to name the rock are explained in the lesson on the classification of sedimentary rocks. In short, bioclasts are pieces of shells and skeletons, and ooids are small round grains built up in layers. Peloids are small rounded lumps of limy mud, and intraclasts are chips of half-hard sediment torn up and redeposited nearby. Here something else matters: where each grain is born and which mineral it is made of. The first tells us about the environment. The second tells us what will happen to the grain after burial.

Where carbonate grains come from

Grains: Corals (modern)
Where they form: reefs of warm shallow seas
Mineral: aragonite

Grains: Limy green algae
Where they form: lagoons and shoals; when they break up, they give limy mud
Mineral: aragonite

Grains: Gastropods (snails) and many bivalves (clams)
Where they form: almost any shallow water
Mineral: aragonite; in some bivalves, calcite

Grains: Echinoderms (sea urchins and their relatives), red algae, many bottom-living foraminifera
Where they form: shelf, reefs
Mineral: magnesian calcite

Grains: Brachiopods (lamp shells)
Where they form: open shelf, especially in the Palaeozoic
Mineral: low-magnesium calcite

Grains: Planktonic foraminifera, coccoliths
Where they form: open ocean
Mineral: low-magnesium calcite

Grains: Ooids
Where they form: mobile shoals up to a few metres deep, where the water holds more carbonate than it can keep dissolved
Mineral: modern ones: aragonite

Grains: Peloids, intraclasts
Where they form: lagoons, tidal flats
Mineral: the mineral of the original mud

Grains	Where they form	Mineral
Corals (modern)	reefs of warm shallow seas	aragonite
Limy green algae	lagoons and shoals; when they break up, they give limy mud	aragonite
Gastropods (snails) and many bivalves (clams)	almost any shallow water	aragonite; in some bivalves, calcite
Echinoderms (sea urchins and their relatives), red algae, many bottom-living foraminifera	shelf, reefs	magnesian calcite
Brachiopods (lamp shells)	open shelf, especially in the Palaeozoic	low-magnesium calcite
Planktonic foraminifera, coccoliths	open ocean	low-magnesium calcite
Ooids	mobile shoals up to a few metres deep, where the water holds more carbonate than it can keep dissolved	modern ones: aragonite
Peloids, intraclasts	lagoons, tidal flats	the mineral of the original mud

This is the modern composition. In the past the reef builders were different: in the Devonian, stromatoporoids (extinct sponge-like animals), tabulate corals and rugose corals; in the Carboniferous, bryozoans, sponges, algae and microbial communities. In some epochs ooids were made of calcite.

Carbonate minerals

Mineral: Aragonite
Composition and density: CaCO₃ with an orthorhombic (brick-like) lattice; 2.93 g/cm³
Stability: unstable near the surface: it dissolves first or recrystallises into calcite

Mineral: Magnesian calcite
Composition and density: calcite whose lattice holds more than 4 mol% MgCO₃ (usually 10–20)
Stability: unstable: it loses magnesium and turns into low-magnesium calcite

Mineral: Low-magnesium calcite
Composition and density: less than 4 mol% MgCO₃; 2.71 g/cm³
Stability: stable; ancient limestones are made of it

Mineral: Dolomite
Composition and density: CaMg(CO₃)₂; 2.84–2.87 g/cm³
Stability: stable; in ancient rocks it is mostly secondary, because it replaced limestone

Mineral	Composition and density	Stability
Aragonite	CaCO₃ with an orthorhombic (brick-like) lattice; 2.93 g/cm³	unstable near the surface: it dissolves first or recrystallises into calcite
Magnesian calcite	calcite whose lattice holds more than 4 mol% MgCO₃ (usually 10–20)	unstable: it loses magnesium and turns into low-magnesium calcite
Low-magnesium calcite	less than 4 mol% MgCO₃; 2.71 g/cm³	stable; ancient limestones are made of it
Dolomite	CaMg(CO₃)₂; 2.84–2.87 g/cm³	stable; in ancient rocks it is mostly secondary, because it replaced limestone

The densities are rounded reference values. Aragonite and calcite are the same substance with different crystal lattices (mineral and rock).

The differences in stability decide the fate of the pores. Suppose a shoal or reef is exposed to air when the sea level falls, and fresh groundwater enters the sediment. The aragonite grains dissolve first and leave voids in their own shape. The dissolved carbonate settles out nearby as calcite cement. Magnesian calcite loses magnesium but keeps the shape of the grain. Low-magnesium calcite hardly changes. So two limestones with the same structure, but with grains of different minerals, can have very different pores after diagenesis (the changes in a sediment after burial). There is a numerical example in the section on pores.

The solubility of carbonates has a second consequence. Carbonate sediment is cemented early, often still on the sea floor or soon after it is exposed to air, and it becomes rigid early. So carbonates can hold steep slopes and the cavities of a reef framework. But when they are deformed later, they are brittle: they break instead of bending, and fractures form.

Environments: from the sabkha to the basin

A shallow carbonate sea is usually fringed by belts of environments that run along its edge. From the land to the open sea they follow in this order (the picture is in the next section):

Sabkha and tidal flat: A low shore that high tides and storms flood and low tides drain. (A sabkha is the salty part of such a flat above the tide level, on a hot coast.) Fine limy mud builds up in layers on microbial mats. When it dries out, it cracks, and voids are left between the thin layers. These voids are called fenestrae. Tidal channels cut across the flat and have a coarser fill. Above the tide level, on the sabkha, crystals of gypsum and anhydrite grow in the sediment, and the mud is often dolomitized early. Reservoirs here are rare: at best, the fill of channels and porous dolomites. The evaporites act as seals (rock that stops oil and gas from escaping).
Lagoon: Quiet water behind a barrier of reefs or shoals. Mud collects, with peloids, shells and algae, and burrowing animals stir it up. If the exchange of water with the sea is poor, the salinity changes and fewer kinds of animals live there. The usual rocks are mudstones and wackestones (rocks made mostly of mud, with a few grains in it). Their original pores are small, and permeability is low, unless dissolution or dolomitization has created permeability.
Shoals and bars: The water here moves the most. Waves and tidal currents roll ooids and shell fragments about and wash out the mud. What is left is grainstone (a rock of grains with no mud between them) with sloping thin layers (cross-bedding) and pores between the grains. These are the best original reservoirs, if early cement has not closed them. A shoal moves sideways, and in a vertical section its crest comes to lie on its own slope. So from the bottom upwards, mud with grains gives way to cleaner and coarser sand (coarsening upwards).
Reef and platform margin: Organisms build a rigid framework that stands up to waves. Voids are left between its parts and in its cavities, and sediment and early cement partly fill them. The rocks are boundstones (rocks bound together by the organisms that grew in them). Behind the reef lie the clastic sands of the back-reef zone. The reservoirs are complex: framework and intraparticle pores, and dissolution vugs. Right next to them is early cement that has closed part of the pores.
Slope: The slope in front of the margin. Reef debris and sand slide and roll down it. They form breccias and megabreccias (rocks of broken fragments), the deposits of debris flows and grain flows, and turbidites (layers laid down by turbidity currents, which are underwater flows of sediment and water). How steep the slope is depends on the material. Mud slopes are gentle. Slopes of fragments held together by early cement stand at tens of degrees. Aprons of debris are sometimes good reservoirs.
Basin: In deep, quiet water, limy mud from plankton, marls (a mixture of lime and clay) and clays settle out. Where the bottom has no oxygen, dark clays rich in organic matter collect. These are source rocks (rocks that generated oil and gas) of the same age as the platforms on the margins of the basin (Source rocks, maturation and migration).

Carbonate sand obeys the same laws of water flow as quartz sand. Currents build dunes and ripples, waves build wave ripples, and turbidity currents lay down graded beds (coarse at the bottom, fine at the top) on slopes (Primary sedimentary structures). So we read the sedimentary structures in the same way. We must be more careful with grain size. Shell fragments are lighter and flatter than quartz grains of the same size, and water carries them differently. So a more reliable sign of how energetic the water was is whether mud is left between the grains (Dunham classes, the names of limestones by their mix of grains and mud).

Ramp, rimmed shelf and isolated platform

A carbonate platform is a thick body of shallow-water carbonates with a flat top. We tell three main types apart by the way the top passes into the deep sea.

Ramp: The sea floor slopes down smoothly from the shore into the basin, usually at less than one degree, with no sharp break. The waves set the belts. The inner ramp lies above the fair-weather wave base (the depth to which everyday waves stir the sea floor). It has shoals, beaches and the lagoons behind them. The middle ramp lies between the fair-weather and storm wave bases (the storm wave base is the depth to which storm waves stir the floor). It has storm layers among mud. The outer ramp lies below the storm wave base, with mud and marls. The belts are wide, and reefs are separate mounds.
Rimmed shelf: A flat shallow top with a rim of reefs or ooid shoals at the break in slope. Behind the rim lie a lagoon and tidal flats. In front of it is a steep slope down into a basin hundreds of metres deep. The rim breaks the waves, so it is calm behind it.
Isolated platform: A shallow area surrounded by deep water on all sides. It has rims of reefs and shoals on both the windward side (facing the wind) and the leeward side (sheltered from it), and a lagoon inside. Large ring-shaped buildups are called atolls. They are named after modern atolls, which grow on sinking volcanoes, although ancient buildups also grew on other raised parts of the sea floor.

SchematicProfiles of a ramp and of a rimmed shelf (schematic, with the vertical scale greatly exaggerated). On a ramp the belts are set by the wave bases. At a rimmed shelf, a rim of reefs and shoals separates the lagoon from the steep slope.

The gentler the sea floor, the wider each belt of depths. This is simple geometry. A floor with slope $\alpha$ deepens by $\Delta z$ over a distance $\Delta x = \Delta z/\tan\alpha$ .

Worked exampleWidth of the belt of storm layers on a ramp and on the slope of a rimmed shelf

Given (assumed data)

Fair-weather and storm wave bases: 28 and 78 m (the example of the lesson on sedimentary structures: waves with periods of 6 and 10 s)
Slope of the sea floor: ramp 0.1° and 0.5°; slope in front of the margin of a rimmed shelf 25°

Solution

Difference in depth across the belt: $\Delta z = 78 - 28 = 50$ m.
Ramp with a slope of 0.1°: $\Delta x = 50/\tan 0.1° = 50/0.0017453 = 28{,}648$ m, that is about 28.6 km. With a slope of 0.5°: $50/0.0087269 = 5729$ m, about 5.7 km.
Slope of 25°: $\Delta x = 50/\tan 25° = 50/0.4663 = 107$ m.

Answer: The same band of depths takes tens of kilometres on a gentle ramp (28.6 km at a slope of 0.1°), but only about 107 m on the slope of a rimmed shelf.

What it means. On a ramp, a small change in sea level shifts the belts by kilometres, and one facies can be followed over a large area. This is convenient for correlation and field development. At a rimmed shelf, the best reservoirs, the reefs and shoals of the margin, form a narrow strip. It is usually from hundreds of metres to a few kilometres wide, and a well can easily miss it. We look for it in seismic data: mounds, steeply dipping layers of the slope and a change in the pattern of reflections (Seismic interpretation). The wave bases here are assumed values. The real depths depend on wave height and grain size (wave base). The data in this example are assumed for illustration.

A platform changes with time. When more sediment is made than there is room for, the margin advances towards the basin (progradation). When the sea level rises faster, the margin builds upwards or retreats towards the land (accommodation space, the room available for sediment). A ramp can turn into a rimmed shelf, if reefs or shoals grow at the edge of its shallow part. They make a break in the slope of the floor and a steep slope in front of it.

Carbonate pores: the Choquette–Pray classification

In a sandstone almost all pores are gaps between grains, and their size follows the size of the grains. In carbonates, pores occur inside shells, between crystals, where grains have dissolved, in fractures and in caves. Most of the voids have been reshaped after deposition. So we describe them with a scheme of their own, the Choquette–Pray classification. Its main question is this: is the shape of the pore tied to the fabric of the rock, that is, to its grains, its crystals or the framework built by organisms?

Types of carbonate pores in the Choquette–Pray classification

Group: fabric selective
Pore type: interparticle (between grains)
How it forms: gaps between grains

Group: fabric selective
Pore type: intraparticle
How it forms: cavities of shells, chambers of foraminifera, pores of skeletons

Group: fabric selective
Pore type: intercrystal
How it forms: gaps between crystals, most often of dolomite

Group: fabric selective
Pore type: moldic
How it forms: voids where grains have dissolved, often aragonite grains

Group: fabric selective
Pore type: fenestral
How it forms: voids between thin layers of mud on tidal flats

Group: fabric selective
Pore type: shelter
How it forms: under flat shells, where mud did not get in

Group: fabric selective
Pore type: growth-framework
How it forms: between the parts of the growing framework of a reef

Group: not fabric selective
Pore type: fracture
How it forms: splitting of rigid rock

Group: not fabric selective
Pore type: channel, vug, cavern
How it forms: dissolution that cuts across grains and cement; caverns are as big as a person or bigger

Group: mixed
Pore type: breccia, boring, burrow, shrinkage
How it forms: follow the fabric or cut across it, depending on the case (borings are holes bored by organisms, burrows are tunnels dug by animals)

Group	Pore type	How it forms
fabric selective	interparticle (between grains)	gaps between grains
fabric selective	intraparticle	cavities of shells, chambers of foraminifera, pores of skeletons
fabric selective	intercrystal	gaps between crystals, most often of dolomite
fabric selective	moldic	voids where grains have dissolved, often aragonite grains
fabric selective	fenestral	voids between thin layers of mud on tidal flats
fabric selective	shelter	under flat shells, where mud did not get in
fabric selective	growth-framework	between the parts of the growing framework of a reef
not fabric selective	fracture	splitting of rigid rock
not fabric selective	channel, vug, cavern	dissolution that cuts across grains and cement; caverns are as big as a person or bigger
mixed	breccia, boring, burrow, shrinkage	follow the fabric or cut across it, depending on the case (borings are holes bored by organisms, burrows are tunnels dug by animals)

The groups are fabric selective, not fabric selective and mixed. By size, pores are divided into micropores (smaller than 1/16 mm), mesopores (1/16–4 mm) and megapores (4–256 mm). By time, they are primary, formed before and during deposition, and secondary.

SchematicEight common types of carbonate pores (schematic): six are fabric selective and two are not. The pores are shown in light blue.

The difference is practical. Fabric-selective pores follow the original structure, and so they follow the facies. Interparticle pores are in the grainstones of shoals, framework and intraparticle pores are in reefs, fenestrae are in tidal-flat deposits, and intercrystal pores are in dolomitized muds. We predict their distribution from a facies map. Pores that are not fabric selective follow later events: the bending of layers and faults (fractures, Fractures and naturally fractured reservoirs), and exposure to air and dissolution under unconformity surfaces (vugs, channels, caverns). We predict them from the structural and diagenetic history.

Dolomitization and karst. Dolomite in ancient rocks is mostly secondary. Brines (very salty waters) of lagoons and sabkhas, or formation waters (the water in the rock pores), bring in magnesium, and dolomite replaces calcite. Suppose the replacement goes “mole for mole”, so that all the carbonate of the dolomite comes from the replaced calcite. Then the volume of the solid would fall by about 13% (molar volume). But this is the limiting case. Often dolomite replaces limestone almost “volume for volume”, each volume of calcite giving a volume of dolomite. This uses extra carbonate that is brought in, and the porosity of the dolomite is largely inherited from the original rock. Dolomites are often more porous than limestones at the same depth. One reason is that they resist compaction and pressure solution (the dissolving of grains where they press on each other) better. The size of their intercrystal pores grows with the size of the crystals. But excess dolomite that grows into the pores, and anhydrite cement, destroy porosity. Karst is the dissolving of carbonates by fresh water when the platform is exposed to air. It makes vugs, channels, caverns, collapse breccias and fractures widened by dissolution, concentrated under unconformity surfaces. Karst voids give very high permeability. In drilling, they cause losses of drilling mud (it drains into the voids) and sudden drops of the drilling tool. More on both is in the lesson on diagenesis.

Worked examplePorosity without permeability: dissolution of aragonite in an oolite

Given (assumed data)

Oolitic grainstone before diagenesis: intergrain porosity 42%; grains are 58% of the rock volume
Composition of the grains: half of the volume is aragonite ooids, half is calcite fragments
What happens: all the aragonite dissolves; all the dissolved carbonate settles out nearby as calcite cement in the intergrain pores (a closed system)
Reference data: density of aragonite 2.93 and of calcite 2.71 g/cm³; $M(\mathrm{CaCO_3}) = 100.09$ g/mol

Solution

Molar volumes: aragonite $100.09/2.93 = 34.16$ cm³/mol, calcite $100.09/2.71 = 36.93$ cm³/mol. From 1 m³ of aragonite we get $36.93/34.16 = 1.081$ m³ of calcite.
Voids where the ooids were: $0.5\cdot 0.58 = 0.29$ of the rock volume.
Cement: $0.29\cdot 1.081 = 0.313$ . The intergrain porosity that remains is $0.42 - 0.313 = 0.107$ .
Total porosity: $0.29 + 0.107 = 0.397$ , that is 39.7% instead of 42%. About 73% of it is separate moldic pores, which communicate only through the remaining intergrain pores.
Permeability is set by the network of intergrain pores. If the grain size stays the same, the Kozeny–Carman model (a standard formula that links permeability to porosity), $k \propto \phi^3/(1-\phi)^2$ , gives $\dfrac{0.107^3/(1 - 0.107)^2}{0.42^3/0.58^2} = 0.0070$ . Permeability falls about 140 times.

Answer: Total porosity has hardly changed (39.7% instead of 42%). But the connected intergrain porosity has fallen from 42 to 10.7%, and by the Kozeny–Carman estimate permeability has fallen by about two orders of magnitude, that is, by more than a hundred times.

What it means. The rock has rearranged its voids. The gaps between the grains have filled with cement, and pores have appeared where the grains were. These pores hardly connect with each other. The permeability estimate rests on an assumption: that permeability follows the intergrain porosity by the Kozeny–Carman model at the same grain size. The estimate is optimistic. Cement narrows the pore throats (the narrow passages between pores) first, so in reality permeability usually falls even more. On the density and neutron logs (two other kinds of porosity log) such a rock looks like a good reservoir. But the sonic log hardly sees such voids (secondary porosity). And the cementation exponent $m$ in Archie's equation (the equation that gives water saturation from resistivity logs) is clearly larger than 2 for it (formation resistivity factor). If groundwater carries the dissolved carbonate away (an open system), there is less cement and the porosity grows. The data in this example are assumed for illustration.

CalculatedVolumes of grains, pores and cement in the oolite of the example, before and after the aragonite ooids dissolve (fractions of the rock volume).

What the pore type means for permeability, logs and saturation

Permeability is set by the size of the pore throats and by how well the pores connect, not by their total volume. For the petrophysics of carbonates (the study of the physical properties of rocks) it helps to separate two groups. Interparticle pores lie between grains or crystals and are not much bigger than them. Vuggy pores are all the others: inside grains, or clearly bigger than them. Vugs, in turn, are of two kinds. Separate vugs are connected to each other only through interparticle pores (moldic pores, shell cavities, isolated vugs). Touching vugs form a network of their own (fractures, channels and dissolution vugs, caverns, breccias). The petrophysical classification of carbonates, the Lucia classification, is built on this division.

In rocks without vugs, permeability depends on the interparticle porosity and on the size of the particles. The bigger the grains or crystals, the wider the throats (Kozeny–Carman model: at the same porosity, $k \propto d^2$ ). So with the same interparticle porosity, grainstones and coarse-crystalline dolomites (crystals larger than about 100 µm) are more permeable than packstones, which have mud between the grains, and medium-crystalline dolomites (20–100 µm). These in turn are more permeable than rocks with a muddy framework and fine-crystalline dolomites (smaller than 20 µm). Inside each group the link between porosity and permeability is fairly close. Between the groups, at the same porosity, permeability differs by orders of magnitude (factors of 10, 100 or more).

What the pore type means

Pores: Interparticle: between grains and crystals
Permeability: grows with the size of particles and with interparticle porosity
Logs and saturation: porosity methods agree; $m$ about 2

Pores: Separate vugs: moldic pores, shell cavities
Permeability: add porosity and hardly add permeability
Logs and saturation: the sonic log hardly sees them; $m$ above 2

Pores: Touching vugs: fractures, channels, dissolution vugs
Permeability: very high, with low porosity
Logs and saturation: $m$ nearer 1; early breakthrough of water and gas, mud losses in drilling

Pores: Micropores in mud and chalk
Permeability: low, with high porosity
Logs and saturation: a lot of residual water: high water saturation on logs, yet the wells produce no water

Pores	Permeability	Logs and saturation
Interparticle: between grains and crystals	grows with the size of particles and with interparticle porosity	porosity methods agree; $m$ about 2
Separate vugs: moldic pores, shell cavities	add porosity and hardly add permeability	the sonic log hardly sees them; $m$ above 2
Touching vugs: fractures, channels, dissolution vugs	very high, with low porosity	$m$ nearer 1; early breakthrough of water and gas, mud losses in drilling
Micropores in mud and chalk	low, with high porosity	a lot of residual water: high water saturation on logs, yet the wells produce no water

For the values of the cementation exponent $m$ see the lesson on the formation resistivity factor. For an estimate of the volume of vugs see the lesson on sonic porosity.

Scale matters too. A standard core plug 2.5–3.8 cm across cannot hold large vugs and rarely crosses fractures. So we study vuggy and fractured intervals on whole (full-diameter) core. We check the permeability of the layer as a whole by well tests (what core gives). Carbonates are more often oil-wet (water-repelling) or of mixed wettability than sandstones. (Mixed wettability means that part of the rock surface prefers oil and part prefers water.) This affects how water pushes oil out (Wettability and interfacial phenomena). And hydrochloric acid dissolves the rock itself. Acid treatments of carbonate reservoirs are based on this (Acid treatments: principle, economics and candidate selection).

Subsalt carbonate buildups of the Pre-Caspian Basin

The largest fields of Kazakhstan, Tengiz, Kashagan and Karachaganak, lie in the carbonates of the subsalt tier of the Pre-Caspian Basin. From the Late Devonian to the Early Permian, carbonate platforms and buildups grew on the margins of the deep basin. In its centre, thin deep-water black clays collected: the main source rocks. Later, the salt of the Kungurian stage covered it all (how the basin formed, the salt and the two tiers). According to published data summarised in the early 2000s, the buildups are built differently on different margins.

Northern and western margins. There are three thick carbonate sequences, from the Upper Devonian to the Lower Permian. Each has a belt of barrier reefs along its edge, and these pass towards the basin into thin deep-water black clays. We know both the advance of these sequences into the basin and the retreat of the younger reefs. Beyond the barrier reefs, only separate buildups have been found. One of them is Karachaganak, a gas-condensate field (the gas carries light liquid hydrocarbons) with oil. The field lies in an atoll-type buildup of the Upper Devonian–Carboniferous and in a Lower Permian pinnacle reef (a single tall reef) that grew on top of it. The field was discovered in 1979, and its area is more than 280 km².
Southern margin, the Karaton–Tengiz zone. Atolls many hundreds of metres high, made of carbonates from the Upper Devonian to the Bashkirian stage of the Middle Carboniferous. Beyond their edges the carbonates pass into deep-water deposits of the basin. The tops of the atolls are covered, across an unconformity, by thin Lower Permian black clays. These were laid down on sunken uplifts, below the depth at which carbonate could still form. By that time the buildups had stopped growing (unconformities). Tengiz was discovered in 1979. The first well flowed oil from subsalt deposits at a depth of 4045–4095 m. The accumulation measures about 20 by 21 km in plan view.
Kashagan is a buildup of the same type under the shallow water of the North Caspian, about 80 km from Atyrau, where the sea is 3–4 m deep. The reservoir lies at a depth of about 4.2 km and is under high pressure. Oil flowed in the first well in 2000. By an estimate of the early 2000s, the buildup is almost three times larger in area than Tengiz.

The porosity of these buildups is uneven. It was made mainly by dissolution, which is best developed in the reef rocks. At Tengiz, by the same data, porosity ranges from a few per cent up to 20%, and the average over the atoll is about 6%. Most of it is vugs, and primary pores give no more than 2–3%. Porosity above 10% is found in a ring-shaped zone of the reef core along the edge of the atoll. On the slopes and in the central lagoon it is clearly lower. Permeability is controlled mainly by fractures. On core plugs it ranges from a few to hundreds of millidarcies (mD, the unit of permeability). At Karachaganak, vuggy porous reef limestones and dolomites have porosity of 10–14% on average. The neighbouring lagoon facies usually have 6–10% or less.

Everything we have said above comes together here. The environment (margin against lagoon) decides where the best reservoirs lie. Dissolution decides how many pores they have. Fractures decide how fast fluid flows to the well. So models of such accumulations are built from facies, pore types and fracturing, not from average porosity. The fluids of these fields contain a lot of hydrogen sulfide (Composition of reservoir fluids). The initial pressure in the subsalt reservoirs of Tengiz and Kashagan is about twice the hydrostatic pressure (the pressure of a column of water) (Abnormal formation pressures: origin, prediction and recognition).

Terms and symbols

Carbonate factory: The community of organisms, and the conditions in which a carbonate sediment forms. Its production depends on light, depth, temperature and the clarity of the water.
Photic zone: The layer of water that receives at least 1% of the surface light, down to $z_{1\%} = \ln 100/k$ .
Aragonite, magnesian calcite and low-magnesium calcite: Minerals of skeletons and sediments. The first two are unstable, and after deposition they dissolve or recrystallise. The third is stable.
Carbonate buildup, reef, atoll: A body that grew from organisms. A reef is a buildup with a rigid framework, and an atoll is a ring-shaped buildup around a lagoon.
Ramp: A platform with a gentle slope of the floor, usually less than a degree, and no sharp break. The belts are set by the wave bases.
Rimmed shelf: A platform with a rim of reefs or shoals at the break in slope, a lagoon behind it and a steep slope in front of it.
Isolated platform: A shallow platform surrounded by deep water on all sides.
Sabkha: A salty coastal plain of a hot climate above the tide level, where gypsum and anhydrite grow in the sediment.
Fenestrae: Voids between thin layers of mud on tidal flats. They are fabric-selective pores.
Choquette–Pray classification: A division of carbonate pores into fabric-selective pores, pores that are not fabric selective, and mixed pores.
Moldic pores: Voids where dissolved grains used to be, most often aragonite grains.
Separate and touching vugs: Separate vugs are connected to each other only through interparticle pores. Touching vugs form a network of their own.
Carbonate compensation depth (CCD): The depth in the ocean below which carbonate dissolves faster than it arrives. Today it is about 4–5 km.
Dolomitization, karst: The replacement of calcite by dolomite, and the dissolving of carbonates by fresh water when the platform is exposed to air.

Common mistakes and misconceptions

Wrong

“Limestones settle out of sea water like salt from brine.”

In fact

Almost all the carbonate of shallow water is made of skeletons of organisms, their fragments, and mud released by algae and microbes. So the carbonate factory depends on light, warmth and clear water, and not only on chemistry.

Wrong

“The porosity of the carbonate is high, so its permeability is high too.”

In fact

Moldic pores and separate vugs add porosity with almost no permeability. The micropores of mud give high porosity with low permeability. Permeability is set by the size of the particles, the interparticle pores and how well the voids connect.

Wrong

“Dolomite is always better than limestone.”

In fact

Dolomitization can create intercrystal pores. But excess dolomite that has grown into the pores, and anhydrite cement, destroy them. The quality of a dolomite depends on the size of its crystals and on its history.

Wrong

“The core is dense, so the layer is dense.”

In fact

Core plugs cannot hold large vugs and rarely cross fractures, yet these often control the inflow. We check the permeability of the layer on whole core and by well tests.

Wrong

“A reef is the whole platform.”

In fact

A reef is a narrow margin or a set of separate buildups. Behind it lie the lagoon and tidal flats, with very different rocks and pores. A well in the middle of a platform may meet no reef reservoir at all.

In depth: advanced detail

Can the buildup keep up with the sea level

Let the relative sea level rise at a rate $\dot A$ . This is the rate at which room for sediment is created (sinking of the sea floor plus rise of the ocean level). Then the depth of water over the top of the buildup changes as the difference between the rise of the level and the build-up of the top:

Water depth over a growing top

\frac{\dd z}{\dd t} = \dot A - G(z)

where:

$z$ depth of water over the top of the buildup, m
$t$ time, years
$\dot A$ rate of relative rise of the sea level, mm/yr
$G(z)$ production of carbonate at depth $z$ (the equation above), mm/yr

Units: $\dot A$ , $G$ and $\dd z/\dd t$ in mm/yr (1 mm/yr = 1000 m per million years), $z$ in metres

Holds when

All the production stays on the top. Removal of sediment and destruction reduce $G$ .
$\dot A$ is constant. In reality the sea level oscillates with various periods.

While $G(z) > \dot A$ , the top catches up with the sea level and in the end stays near the surface. The depth $z^*$ at which $G(z^*) = \dot A$ is a point of unstable equilibrium (a balance that the slightest push destroys). A top that ends up deeper than this falls further and further behind, production drops, and the buildup “drowns”. If $\dot A$ is larger than the maximum production, any buildup will drown. English-language literature calls platforms that keep pace with the sea level “keep-up”, those that catch up with it “catch-up”, and those that give up “give-up”.

Worked exampleCritical depth for the reef of the example

Given (assumed data)

Reef from the factory example: $G_m$ = 8 mm/yr, $I_0/I_k$ = 10
Rise of the sea level $\dot A$: 2 mm/yr; for comparison, 10 mm/yr
Water: clear ( $k$ = 0.08 1/m) and muddy (0.2 1/m)

Solution

The condition $G(z^*) = \dot A$ gives $\tanh x = 2/8 = 0.25$ . So $x = 0.255$ and $z^* = \ln(10/0.255)/k = \ln 39.2/k$ .
Clear water: $z^* = 3.669/0.08 = 46$ m. Muddy water: $z^* = 3.669/0.2 = 18$ m.
At a rise of 10 mm/yr, production is smaller than the rise at every depth ( $G \le 8$ mm/yr). There is no depth of balance.

Answer: At a sea-level rise of 2 mm/yr, a reef survives if its top is not deeper than about 46 m in clear water, or than 18 m in muddy water. It cannot keep up with a rise of 10 mm/yr at any depth.

What it means. A fast rise of the sea level, or muddier water that lowers the production curve, can “drown” a platform. This can happen even though, on average, carbonate factories make sediment much faster than the sea floor usually sinks. A drowned top stops growing, and deep-water sediments collect on it. This is how the tops of the atolls of the Karaton–Tengiz zone, for example, were covered. The model does not say why a particular buildup stopped. Cooling, too many nutrients or clastic material can also stop growth. The data in this example are assumed for illustration.

Connections to other topics

Sedimentary rocks and their classification — names of carbonate rocks by Dunham and Folk.
Minerals and rocks — calcite, dolomite and molar volume when one mineral replaces another.
Facies, facies successions and sedimentary logs — facies successions, including shallowing-upward cycles.
Porosity — total and effective porosity, primary and secondary pores.
Rock permeability — why permeability grows as the square of the particle size.
Porosity from the sonic log — how the volume of vugs is estimated by comparing methods.
Archie's equation and saturation of clean reservoirs — the cementation exponent of rocks with vugs and fractures.
Fractures and naturally fractured reservoirs — natural fractures and fractured reservoirs.
Folds, salt diapirs and their relationship to faults — Kungurian salt and the subsalt tier of the Pre-Caspian Basin.
Plate tectonics and sedimentary basins — how the Pre-Caspian Basin formed.
Acid treatments: principle, economics and candidate selection — acid treatments of carbonate reservoirs.
Diagenesis: compaction, cementation and dissolution — dissolution, cementation and dolomitization after burial.

Check yourself

1
A reef grows in water with a light attenuation coefficient of 0.05 1/m. At what depth does the light fall to 1% of the surface value?
1. A.About 460 m
2. B.About 92 m
3. C.About 9 m
4. D.About 46 m
Show the answer
Correct answer: B. About 92 m
$z_{1\%} = \ln 100/k = 4.61/0.05 \approx 92$ m. At 46 m, 10% of the light is left ( $\ln 10/0.05$ ). In muddy water with $k = 0.2$ 1/m the same limit would rise to 23 m.
2
In a thin section, there are rounded voids where ooids used to be, and the gaps between them are filled with clear calcite. What do we call these pores, and what do we expect for permeability?
1. A.Fenestrae; medium permeability
2. B.Intergrain pores; high permeability
3. C.Vugs, not fabric selective; high permeability
4. D.Moldic pores, fabric selective; low permeability with noticeable porosity
Show the answer
Correct answer: D. Moldic pores, fabric selective; low permeability with noticeable porosity
The voids repeat the shape of the dissolved grains. These are moldic pores, which are fabric selective. They connect only through the intergrain pores, and those are filled with cement. So porosity can be high and permeability low, as in the example of the lesson.
3
A band of depths from 10 to 50 m lies on a ramp with a slope of 0.2° and on a slope of 20°. Estimate the width of the band in both cases.
Show the answer
$\Delta z = 40$ m. Ramp: $40/\tan 0.2° = 40/0.00349 \approx 11.5$ km. Slope: $40/\tan 20° = 40/0.364 \approx 110$ m, which is about a hundred times narrower. So the facies of a ramp can be followed for kilometres, while a well can easily miss the facies of the margin of a rimmed shelf.
4
A shoal of fresh carbonate sand has dried out (emerged from the sea), and fresh water has entered it. Which grains dissolve first?
1. A.Aragonite ooids and shells of gastropods
2. B.Plates of echinoderms (magnesian calcite)
3. C.All at the same time
4. D.Fragments of brachiopods (low-magnesium calcite)
Show the answer
Correct answer: A. Aragonite ooids and shells of gastropods
Aragonite is unstable near the surface and dissolves first, leaving moldic pores. Magnesian calcite is also unstable, but it usually loses its magnesium and keeps the shape of the grain. Low-magnesium calcite is stable.
5
In a limestone, density and neutron logs give a porosity of 17%, and the sonic log gives 8%. In thin sections we see moldic pores left by dissolved shells. What does this tell us about permeability and about the calculation of saturation?
Show the answer
About 9 porosity units (p.u.) belong to separate moldic pores, which the sonic log hardly sees. Permeability is set by the interparticle pores (about 8%) and the size of the particles, not by the 17%. The cementation exponent of such a rock is larger than 2. If we take $m = 2$ , the water saturation from Archie's equation will be too low, and the oil saturation too high.
6
Why, on an atoll of the Tengiz type, is porosity higher in the ring-shaped zone along the edge of the buildup than in its centre and on its slopes?
Show the answer
The edge of the atoll is the reef core: a framework of organisms with primary voids, and coarse clastic rocks. Dissolution then developed best in them and made vugs. In the centre is a lagoon with muddy rocks and small pores. On the slopes is redeposited material. So the best reservoirs form a ring along the margin, and fractures largely control permeability.

Updated: 15 October 2026