Category Archives: How to do things

Mineralogy of sandstones: Porosity and permeability


porosity in sandstone

Porosity and permeability – the flow of water and other geofluids

Nearly all geological processes require the presence of water in one form or another. Most sedimentation occurs in water (aeolian deposits are the obvious exception). Sediment burial and compaction involves the expulsion of water. Diagenesis would not take place in the absence of water; hydrocarbons would not migrate to traps and minerals would not be concentrated in ore bodies. Aqueous fluids under pressure reduce cohesion and friction promoting rock deformation.  Metamorphism would be painfully slow, even by geological standards if it were not for the transfer of mass in hot aqueous fluids.

All these processes require not only the presence of water, but its continual movement or flow. Below Earth’s surface, the residence and flow of aqueous fluids requires two fundamental rock-sediment properties:
– voids, commonly in the form of intergranular pores and fractures, and
– connectivity among the voids.
The first of these is referred to as porosity; the second as permeability.

There are two main types of porosity: intergranular porosity that characterizes sands, gravels and mud, and fracture porosity in hard rock. Fracture porosity forms during brittle failure of hard rock or cooling of lava flows. Fracture networks that are connected can provide pathways for fluid flow even when the host rock is impervious (e.g. granite, basalt, indurated sandstone). Highly productive aquifers are not uncommon in fractured bedrock.

Intergranular porosity is the void space between detrital grain contacts and is expressed as a percentage of the total sediment-rock volume. It is a dimensionless number (i.e. it has no units of measure). All sediments begin life with some porosity.  Well sorted beach, river and dune sands have initial porosities ranging from 30% – 40%, muds as high as 70%. These values represent the total void space, namely the large pores plus lots of microporosity in tiny nooks and crannies between grains and crystals. Hydrogeologists have found it useful to define effective porosity as that which permits easy movement of fluid. This excludes microporosity where surface tension forces inhibit flow. Effective porosity is always less than total porosity. Follow this link to a simple experiment designed to measure porosity.

As sediment is buried, the grains settle (i.e. they become more closely packed) as they begin to compact.  The reduction in porosity by mechanical compaction continues during sediment burial, in concert with the precipitation of cements (chemical diagenesis).  This is particularly evident during the compaction of mud. The high initial porosity is due to micro-pores between clay particles that have dimensions measured in microns. Compaction compresses the clays and drives off the interstitial water. Compaction (porosity-depth) curves for mud, like the example shown below, typically show a loss of porosity that at shallow depths is almost exponential, becoming approximately linear at depths where shale forms; total porosities in shale are extremely low.

burial sequence for loss of porosity

The conduits for fluid flow (water, oil, gas) from one pore space to another are the narrow connections adjacent to grain contacts. These connections are commonly referred to as pore throats. Pore throats are susceptible to blockage during sediment compaction (lithic sandstones are prone to this) and cementation, particularly clays.

details of pore filling cements

Porosity can also be enhanced during burial diagenesis. The primary mechanism for formation of secondary porosity is the dissolution, or partial dissolution of framework grains like feldspar and carbonate bioclasts. Many of these secondary pores are larger than the associated intergranular pore spaces; this is an important diagnostic clue to their identification. Likewise, carbonate and clay cements may be prone to dissolution, resulting in enhanced post-depositional porosity.

Burial depths and temperatures where formation of secondary porosity is encountered commonly coincide with chemical reactions involving the break-down of organic matter. By-products of these reactions include carbon dioxide (and carbonic acid) and organic acids like acetic acid. There is a fundamental shift in pH and chemical equilibria, particularly for carbonates, and this promotes dissolution.

Secondary porosity can also form during subaerial exposure of rock and by some bioturbation. However, the secondary porosity seen in most ancient sandstones developed during  burial diagenesis.

Permeability measures the ease with which a fluid flows through sediment or rock. The flow of fluid from one part of a rock to another or to a bore hole, depends on the connections among pores and fractures. It is possible for a rock or sediment to have high porosity but low permeability if the connectivity is low – mud and shale are prime examples. In coarse-grained sediments that are devoid of clay, there is a good correlation between porosity and permeability.  This relationship does not apply where there are significant amounts of clay.

Permeability is expressed in two ways. Henry Darcy’s pivotal experiments with sand-filled tubes (in 1856) established an empirical relationship between hydraulic gradient (that basically is an expression of hydraulic potential energy) and discharge. The proportionality constant in this relationship is called the hydraulic conductivity (K) (a label borrowed from electrical theory), that has units of distance and time (cm/s, feet/s). In mathematical terms, hydraulic conductivity is expressed as a velocity, also known as the Darcy velocity. Hydraulic conductivity is the standard expression of permeability in groundwater studies. Its value depends not only on the connectivity of pores but also on the dynamic viscosity and density of the fluid (viscosity measures the resistance to flow – crude oil is more viscous than water). Thus, for any porous medium the value of K will be different for water and oil, a factor that is important in groundwater remediation.

The hydrocarbon industry deals with fluids of highly variable viscosity (water, oil, gas) and has opted for a standard expression of intrinsic permeability (k) that depends only on the porous medium. The unit is the Darcy that mathematically reduces to units of area (ft2, m2). It is basically a measure of pore size (the oil industry commonly uses the term millidarcy). Frequently used conversions to Darcys are:

1 m2 = 1.013 x 1012 Darcy

1 Darcy = 9.87 x 10-13 m2

Hydraulic conductivity (K) and intrinsic permeability (k) are related by fluid density and dynamic viscosity such that:

k (m2) = K (m/s) x (1.023 x 10-7 m.s) (the time components cancel)

Typical permeability values for unconsolidated sediment and some rock equivalents are shown in the table below.

typical values of permeability

As you can see, the permeability of shale is extremely low. This is the reason why shale beds make good seals to hydrocarbon reservoirs, and aquitards to confined aquifers. Fluid flow in shales and well-cemented sandstones or limestones can be enhanced by hydraulic fracturing. This process (fracing) is front and centre of shale oil production (notwithstanding all the pros and cons of this industrial process). But that is a story for another time.


Here are three excellent texts that detail the theoretical aspects of the above:

P.A. Allen and J.R. Allen, Basin Analysis: Principles and Applications. Blackwell 2005

C.W. Fetter.  Applied Hydrogeology, 2001. PrenticeHall

P.A.Domenico and F.W.Schwartz Physical and Chemical Hydrogeology,1998 John Wiley & Sons


The provenance of detrital zircon


detrital zircon

This post is part of the How To…series – using zircon geochronology to decipher provenance

Zircon is a common accessory mineral in igneous and metamorphic rocks so it’s not surprising that it is also a common constituent of sedimentary heavy mineral suites. Detrital zircon has assumed a remarkable popularity over the last 2-3 decades as a provenance indicator because:

  • crystals contain measurable amounts of uranium (U), lead (Pb) and thorium (Th) isotopes and can therefore be dated radiometrically,
  • zircon is resistant to chemical and mechanical change – crystals can survive multiple sedimentary cycles (i.e. episodes of erosion from source rocks, deposition, burial and uplift, whereupon the whole process begins anew), and
  • they commonly contain multiple stages of crystal growth that record magmatic, metamorphic and depositional episodes.

Continue reading


Provenance and plate tectonics


This post is part of the How To…series – Provenance, sedimentary basins and plate tectonics

Deciphering the history of sedimentary basins is one of the more exciting tasks geologists can undertake; the provenance of sandstones plays an important part in this adventure.

Sedimentary basins are crustal structures. They are regions of long-term subsidence, responding to tectonic and sediment loads, cooling in the crust and upper mantle, and tectonic dislocation along crustal structures like transform faults. The processes that create sedimentary basins and the sediments that fill them are inextricably linked to plate tectonics.

The idea that the composition of sedimentary rocks was related to large-scale crustal processes was acknowledged by Charles Darwin, Charles Lyell, Henry Sorby and others, but it wasn’t until the 1950s – 60s that sedimentologists began to develop empirical models of sandstone provenance. Robert Folk, Francis Pettijohn, Robert Dott and contemporaries observed links between sediment composition and tectonic domains (or provinces), such as stable cratons and geosynclines – this was the era before plate tectonics.

For geoscientists, the discovery and development of plate tectonic theory changed everything. This is where William R. Dickinson comes into the picture. Dickinson recognized the fundamental link between sedimentary basins and plate tectonics, particularly at plate boundaries. He developed models that relate the modal composition of sandstones to plate tectonic provinces such as collision orogens, magmatic arcs, forearc basins and stable cratons. It is important to remember that these models are based on empirical evidence – analysis of 1000s of thin sections that he and many others had recorded from diverse locations.

The models are based on ternary plots like those used by Dott to classify sandstones. Dickinson and his co-workers used different combinations of the quartz, feldspar and lithics end-members to emphasize certain characteristics of the sediment and the source rocks. Both models shown here use the full suite of minerals. Other plots used only the lithic components, or polycrystalline quartz and lithics.

The Qt-F-L plot combines all varieties of quartz (mono- and polycrystalline quartz, including chert) as a single category and as such emphasizes the maturity of the sediment. Deposits with greater volumes of quartz are generally considered more mature, where mechanical and chemical weathering during sediment transport and deposition have removed less- stable components like feldspar and lithics.

In the Qm-F-L plot, polycrystalline quartz is shifted to the lithic field and in so doing emphasizes the source rocks and production of rock fragments (the quartz component consists only of monocrystalline varieties). Lithics here are key indicators of reworked orogenic provinces along continent-continent and continent-magmatic arc collision provinces. Here, erosion of sedimentary cover and volcanic rocks tends to produce greater proportions of rock fragments.

The second set of diagrams shows typical plate tectonic configurations that correspond to the various QFL fields. The diagrams are highly simplified. In addition to pigeonholing sandstone compositions, the plots provide a useful means of documenting systematic changes as uplift and erosion expose deeper crustal rocks.



For example, unroofing a fold and thrust belt along a collision margin will yield an initial rush of lithics derived from the deformed sedimentary cover. Gradual exposure of a metamorphic core will yield increasing volumes of quartz (mostly polycrystalline) and a new suite of heavy minerals. Likewise, unroofing a magmatic arc complex (the “dissected arc” field in these plots) will provide abundant volcanic lithics followed by more felsic sediment from the deeper intrusive rocks. This is shown schematically in the cartoon below.

The Dickinson plots, like any scientific model, are highly simplified versions of the real world. No two collisional orogens are alike, no two magmatic arcs are exact duplicates. Finding exceptions to any of the models does not indicate their failure – quite the opposite. Their value lies in providing a direction for investigation.

Dickinson’s models have been through several iterations, but their basic structure has survived 40 years of intense scrutiny by geoscientists. They are still useful starting points for unravelling the links among sediment composition, sedimentary basins and plate tectonics.

Check out the companion article – Provenance of sandstones


Some useful texts and papers:

Petrology of Sedimentary Rocks, Sam Boggs Jr. 2012

Dickinson, W. R. and C. A. Suczek, 1979, Plate tectonics and sandstone compositions: American Association of Petroleum Geologists Bulletin, v. 63, p. 2164–2182.

W. R. Dickinson, 1988. Provenance and Sediment Dispersal in Relation to Paleotectonics and Paleogeography of Sedimentary Basins. In New Perspectives in Basin Analysis, Editors, Karen L. Kleinspehn & Chris Paola,  Springer-Verlag, pp 3-25.

R.V. Ingersoll, T. F. Lawton, and S.A. Graham, 2018. Tectonics, Sedimentary Basins, and Provenance: A Celebration of the Career of William R. Dickinson. Geological Society of America, Special Paper v.540, 757 pages


The provenance of sandstones


Provenance – the origin of sand

This post is part of the How To… series

See the companion posts – Provenance and plate tectonics

The provenance of zircon

Provenance is an investigative process that attempts to find out where things originated; a piece of art, an ancient manuscript, a sandstone. In geology, provenance usually applies to sediments and sedimentary rocks – where did they come from? At its most basic, provenance determines the kind of source rock that supplied loose sediment. But provenance can also provide information about:

  • Paleoclimate, particularly weathering of rock where the physical agents of erosion and chemical agents of mineral alteration, determine the volumes of sediment produced and its composition. For example, warm humid climates favour the formation of deep soil profiles where minerals like feldspar, and ferromagnesians like amphiboles and pyroxenes are susceptible to dissolution and alteration.
  • Paleotopography – was there mountainous relief (e.g. orogenic belts or volcanic arcs), or low relief across an ancient craton?
  • Ancient sediment transportation corridors such as terrestrial drainage systems or sediment distribution across a marine shelf or platform to the deep ocean,
  • Possible tectonic dislocation between source areas and sites of sedimentation (sedimentary basins), for example shortening across a mountain thrust belt, or lateral displacement along strike slip or transcurrent faults? Such dislocations may even involve tectonic plates or terranes that have moved 100s of kilometres, separating source rocks from the sedimentary basin.

  • Changes in mineralogy resulting from unroofing. For example, the upper crustal levels of magmatic arcs (including volcanoes) will produce abundant lithics and plagioclase, and lesser amounts of quartz. As uplift and erosion exposes (unroofs) deeper intrusive rocks, there may be a change to lithic-poor sediment with concomitant increases in quartz and or feldspar. Likewise, uplift and erosion of an orogenic belt may produce an initial pulse of sedimentary lithics and recycled quartz, that with gradual unroofing of a metamorphic core produces sediment laden with polycrystalline quartz and a new suite of heavy minerals, particularly micas.

It is instructive to look at suites of sedimentary rocks. This is important because there may be an opportunity to gauge regional trends in composition and texture in relation to changes in the source area (e.g. uplift), drainage, and depositional paleoslope (i.e. the regional dip of sedimentary basins). Your first task is to identify the mineralogy and textural properties like grain sorting and angularity.

Some of the problems associated with the determination of provenance are nicely illustrated using quartz arenites as an example. Quartz arenites contain 95% and more detrital quartz grains. They tend to be well sorted; grains are well rounded. Whence all that quartz? Some useful questions and considerations to begin your determination might be (the questions apply to sedimentary rocks of any composition):

  • Have the distribution and age of sedimentary units and potential source rocks been mapped in sufficient detail?
  • Stratigraphic-sedimentologic evidence may help pin-point potential sediment sources? Particularly useful are any mineralogical changes that coincide with down-dip facies changes or vertical stratigraphy.
  • Can regional paleocurrent trends be linked to drainage of potential source rock areas?
  • Do the potential source rocks contain sufficient amounts of quartz; do they contain the right kinds of quartz?
  • What is the mineralogy of the quartz? You will need to determine the proportions of monocrystalline and polycrystalline varieties, and whether the quartz has strained (i.e. deformed crystal lattice) or unstrained extinction. Is there any volcanic quartz?
  • The suite of heavy minerals can be very instructive. Common minerals include the ferromagnesian groups of like pyroxenes, amphiboles, micas, and olivine, plus iron oxides such as magnetite and ilmenite, and tourmalines, garnets and zircon. Most of these minerals occur in several different rock types (intrusive, metamorphic, volcanic), and on their own may not be diagnostic. Ferromagnesians are prone to alteration by weathering and during diagenesis. But with careful observation you should be able to tease useful provenance information from the suite. Of the micas, muscovite is probably the most useful in that it does not occur in extrusive volcanics but is common in metamorphic and intrusive rocks. Minerals like sillimanite and kyanite are almost exclusively metamorphic. Olivine (although uncommon as a heavy mineral), is particularly prone to alteration and in most cases will be a first cycle product that pinpoints mafic volcanic and intrusive source rocks.
  • Zircon is extremely durable and can survive two or more cycles of deposition, burial, metamorphism and uplift. Over the last 2 decades, Zircon has assumed a position of importance because technology now allows the determination of radiometric dates not just from whole single crystals, but from different parts of a crystal. Knowing zircon ages helps  pinpoint source candidates.

  • Are there any trends in textural properties or mineralogy along depositional dip or stratigraphically? Are there down-stream changes in grain size and angularity with distance from source? Has the proportion of less stable minerals been reduced? For example, if the primary source rock is granite there will be an initial mix of monocrystalline quartz, potassium feldspar and plagioclase, and heavy minerals like biotite and muscovite. Mechanical abrasion and chemical weathering will preferentially reduce the feldspar population.  The micas, even though they are denser than quartz or feldspar, will be preferentially removed because they behave like hydraulically light minerals. There will be a tendency for quartz to become dominant even though it originally was volumetrically subordinate to the feldspars.From your thin section observations and having considered some of the above, what are the potential source rock types? In answering this question, you need to keep in mind the likelihood that the sedimentary rock under the microscope may look nothing like its progenitor.

Having decided on likely source rocks and source areas, it is now time to consider provenance in relation to tectono-stratigraphic domains associated with plate tectonics – are the source rocks-source terrain and adjacent sedimentary basin part of an orogenic belt and foreland basin, an accretionary prism, a magmatic arc-trench complex?  A companion post examines the rationale behind these questions.



Some useful texts and papers:

Petrology of Sedimentary Rocks, Sam Boggs Jr. 2012

R.V. Ingersoll, 1978, Petrofacies and Petrologic Evolution of the Late Cretaceous Fore-Arc Basin, Northern and Central California. The Journal of Geology Vol. 86, pp. 335-352


The mineralogy of sandstones – matrix & cement


SEM of quartz overgrowths

This post is part of the How To… series – How to identify matrix and cement in arenites

The primary architectural element of a sandstone is its framework of sand grains. Spaces between grains will be filled, to varying degrees by much finer sediment – or matrix. The amount of matrix initially deposited is strongly dependent on the energy of the depositional system and the degree of sediment reworking. Wind-blown dune sands and beach sands will have very little matrix at the time of deposition; those deposited farther out to sea will tend to accumulate more.

The proportion of matrix to framework is an important determinant of sandstone classification.  Arenites have less than 15% matrix, wackes have more than 15%. Sandstones with little matrix (commonly referred to as clean sands) have the potential to preserve some of their original porosity.

Detrital matrix generally consists of clay minerals mixed with silt-sized quartz and feldspar.  At the point of deposition there is a significant volume of interstitial water, most commonly seawater or freshwater.  The mix of solid and liquid phases means that matrix is highly reactive in both mechanical and chemical contexts. Compaction that begins soon after deposition and continues during burial, will physically compress the matrix and at the same time drive off some of the interstitial fluid. As burial proceeds, the increase in temperature and accompanying fluid flow will promote chemical reactions involving dissolution of some detrital minerals (particularly clays and feldspar) and precipitation of new minerals.  Thus, the matrix gradually changes from purely detrital to a mix of detrital components and diagenetic products. At some point in this transformation there may be no recognisable detrital matrix. Continue reading


The mineralogy of sandstones: lithic fragments


Identifying detrital lithic fragments

This post is part of the How To… series

Lithic fragments are the bits of eroded or broken rock that can’t be easily slotted into either the quartz or feldspar classification end-members. They are the fragments that are not broken down into single minerals. They tend to be fine-grained and rather dirty looking in shades of brown and grey. In thin section they are nowhere near as exciting to look at as other framework grains. But If we want to know something about sediment source rocks (provenance) or the longevity and survival of granular sediment during transport and deposition, then lithics are no less valuable than quartz, feldspar – perhaps more so.

R.H. Dott’s classification divides the lithic end-member into sedimentary, volcanic, and metamorphic clasts; R.L. Folk takes each of these a step further (I’m not convinced Folk’s subdivision is practical). Part of the problem with too fine a subdivision is the diagenetic and mechanical alteration that lithics are prone to, rendering them indeterminate. Lithic fragments are much softer and chemically more reactive than their quartz-feldspar counterparts during burial diagenesis. Continue reading


The mineralogy of sandstones: feldspar grains


Identifying detrital feldspar.

This post is part of the How To… series

Quartz may be the most common mineral in sandstone, but feldspar is the most abundant mineral in pretty well every other rock type; in fact, it is the most abundant mineral in the Earth’s crust. Unlike quartz, feldspar is an essential ingredient in nearly all igneous rocks, felsic through ultrabasic. It begins to crystallize in magmas at temperatures about 1000oC – 200o warmer than quartz crystallization. Feldspars are also common in metamorphic rocks. As such, feldspar is an important (usually subordinate) component of most terrigenous clastics, reflected in its inclusion in QFL classification schemes.

The two major groups of feldspar are potassium feldspars and plagioclases. All have low relief in plain polarized light (similar to quartz). Both groups have two good cleavage planes at 90o to each other such that broken crystal fragments tend to be blocky. Twinning is common. Continue reading


The mineralogy of sandstones: Quartz grains


This post is part of the How To… series – quartz mineralogy in sandstones

Classification of terrigenous sandstones depends on the identification of two main components: framework grains and matrix. Frameworks are represented by a QFL triad – quartz, feldspar and lithic fragments, where the proportion of each grain type is determined from thin section.  Most classification schemes aggregate all types of quartz, feldspar and lithics into each end-member. This approach is sensible and easy to use.

But simply naming a sandstone (or any rock type for that matter) is not enough. We also want to know about its provenance, the sediment source or sources – was it a stable continent or active mountain belt, volcanic arc or ocean basin, perhaps a far-travelled terrane or tectonic sliver for which the only evidence is the collection of grains that have survived multiple cycles of attrition.

Teasing this information from the rocks requires us to delve into the mineralogy in greater detail. The simplest and cheapest way to do this is with thin sections and a polarizing microscope. We begin with the most common terrigenous component – quartz. Continue reading


Classification of sandstones


sandstone classification

This post is part of the How To… series – how to classify sandstones using QFL plots

In science, classification of things is one of those tasks we readily identify as a crucial component of knowledge but prefer that someone else does it. Classification schemes don’t just name things, they organize them according to their properties, appearance, structure, composition.  If I wish to talk about a particular rock or fossil, then the people who are interested in such things will have a frame of reference to understand and contribute to the discussion, based on whatever classification scheme applies.

The classification of sandstones matured in the 1940s through 1960s; many publications were devoted to the subject; some of the key players were Robert Folk, Harvey Blatt, Francis Pettijohn, Raymond Siever, P. Krynine, E. McBride, H. William, F Turner, C Gilbert, Robert Dott, and R. Fisher. Several schemes were proposed and debated; few were accepted. One of the central topics of discussion was the relative importance of sandstone texture versus sandstone composition. A classification based on texture alone was deemed inadequate; if the rock or sediment had >50% sand, then it was a sandstone, or arenite. Qualifications such as pebbly, silty or muddy might be applied, but this said nothing about the variability of mineral types. However, textural properties such as the percentage of matrix (clay plus silt) did provide grounds for distinguishing between ‘clean’ sandstones (i.e. those lacking significant matrix) and wackes – those rocks containing significant matrix. Continue reading


Some controls on grain size distributions


poorly sorted conglomerate

This post is part of the How To… series

What processes determine the size distribution of clasts in clastic sediments? Why is it that dune sands tend to be very well sorted, river deposits less so, and at the other end of the spectrum, turbidites on a submarine fan are left with no sorting at all? In this article we will look briefly at three determinants of grain size distributions: inheritance, depositional hydraulics, and post-depositional changes.  The three determinants are discussed in greater detail by R.L. Folk under the heading Textural Maturity (a PDF of the 1980 issue can be downloaded here)

All those grains of sand, pebbles and cobbles come from somewhere, from other sediments or source rocks. Erosion and weathering of rock generally results in a wide range of clast sizes – from large blocks to silt and finer. During sediment transport, particularly in fluvial, alluvial and high energy coastal systems, the larger fragments are whittled to smaller fragments. However, there are situations where the source contains a limited range of clast sizes which means that sediment sourced from these deposits will also have similar grain size limitations. The example of modern beach and dunes sands illustrated in the previous post is a case in point. The siliciclastic component of the modern sands is derived from consolidated and weakly lithified Pleistocene deposits that have similar, if not identical grain size distributions. The grain size characteristics of the modern sands are inherited from the older deposits.

Environments of deposition are strong determinants of grain size and grain size distributions in clastic sediments. The primary control here is hydraulics – the strength and longevity of water-wind currents and waves acting on a sediment bed. The strength of a flowing medium determines the size of clasts being moved; in general the stronger the flow, the coarser or heavier the clasts that will be transported (note that size here depends on mass and density).

However, the strength of the flowing medium alone is not the only determinant; the longevity and/or repetition of flow is also critical. For example, turbidity currents and debris flows are basically single depositional events. For contrast, compare sediments in river channels where flow is continuous (albeit fluctuating). Sediment movement along the length of a river commonly produces a down-stream increase in finer sediment fractions, where much of the coarser material (especially gravel) remains upstream. Wave washing along coasts results in repetitive sediment movement, such that grains may travel many kilometres in the swash and backwash and yet never leave the beach.  In both these examples the potential for episodic sediment movement is high. We refer to this process as reworking.

There is a direct correlation between sediment sorting and the degree of reworking; well sorted sands have generally been subjected to high degrees of reworking (notwithstanding the possibility of inheritance). For any given flow velocity there is a maximum grain size that can be moved across a sediment surface; smaller or lighter grains will move across the bed (or in suspension), coarser or heavier grains will not move. Thus, sediment is sorted according to size and mass; lighter or smaller grains are separated or winnowed from coarser-heavier grains. The more frequently this process occurs, the greater the degree of grain sorting.

Movement of sediment also results in mechanical wear and tear of clasts (abrasion). Prolonged abrasion during reworking will ultimately reduce the size of clasts. This process depends primarily on the mechanical strength of clasts. Quartz and feldspar (the two most common components of terrigenous clastic rocks and sediments) react differently during prolonged reworking; quartz is mechanically stable and although grain sizes may become smaller over time, grains survive several cycles of deposition and reworking. Feldspars on the other hand tend to break along crystal cleavage planes. Thus, a sediment that that originally contains equal amounts of quartz and feldspar can, following prolonged reworking, become a well sorted quartz sand with little or no feldspar.

Post-depositional changes
Diagenetic changes that specifically effect the size distribution of grain populations can involve either dissolution (size reduction) or precipitation (enlargement) of the more common rock-forming minerals such as quartz, feldspar and carbonate.

Pleistocene shallow marine deposits in northernmost New Zealand contain bivalve moulds and casts but no calcium carbonate. Post-depositional leaching of the original coarse carbonate size fraction has created deposits that now are as well-sorted as the associated ancient dune deposits.

Calcium carbonate is not the only mineral component affected by leaching in these Pleistocene deposits; feldspar grains too show a remarkable degree of dissolution, resulting in size reduction and even complete removal. Thus, diagenetic changes have skewed the overall grain size distribution towards that of the surviving quartz grain population.

Grain size enlargement also occurs in more lithified deposits.  This commonly takes the form of crystal overgrowths on quartz and feldspar grains.  Mineral overgrowth not only changes the size of clasts, but also their textural properties such as shape and angularity. Care needs to be taken when observing disaggregated sands to distinguish these post-depositional changes from original depositional textural attributes. The most reliable way to do this is using thin-sections and a polarizing microscope.

Some other useful links

Describing sedimentary rocks; some basics

Analysis of sediment grain size distributions

The hydraulics of sedimentation: Flow Regime