Calculate Rock Density (Mass and Volume)
A rock's density (bulk density ρ = m/V, in g/cm³) depends on its mineral composition and on its porosity: a compact crystalline rock is denser than a porous rock or one rich in light minerals. Two ways to measure it: direct mass and volume if the sample has a simple shape or its volume has already been measured, or hydrostatic weighing (Archimedes' method) for a sample of any shape.
Hydrostatic weighing avoids a geometrically complex volume: the dry sample is weighed, then suspended and immersed in water. The loss of apparent weight is proportional to the volume displaced.
Measurement method
Volume already known (simple geometry, or measured by water displacement in a graduated cylinder).
Irregularly shaped sample: two weighings are enough, no length measurement needed.
A short glossary at the bottom spells out the starred terms.
Density versus specific gravity: same number, two labels
Density* is mass over volume. Written ρ, reported in g/cm³ (or kg/m³), it is an absolute number.
Specific gravity* (SG) is that density divided by the density of water at 4 °C. No unit. Because water at 4 °C is 1.000 g/cm³, a granite at 2.67 g/cm³ has SG 2.67 — the figures look interchangeable, which is why hand-specimen cards, core logs and some lab sheets still say “SG” when they mean bulk density. Mineralogists kept the habit for single crystals (quartz 2.65, olivine ~3.3). A whole-rock value is better labelled bulk density, with dry or saturated stated.
Put another way: 2.67 g/cm³ = SG 2.67 = 2,670 kg/m³. Same rock, three labels. Mix the units on a gravity model and the anomaly is fiction.
Why bulk density moves with mineral mix and pore space
Two knobs, not one. Turn either and ρ moves.
- The mineral mix. Quartz and alkali feldspar sit near 2.56–2.65 g/cm³. Plagioclase climbs a little with anorthite content. Biotite, hornblende, pyroxene and olivine pull the rock up; magnetite and ilmenite pull harder. A modal count on a thin section, weighted by mineral densities, already predicts grain density before a balance is touched.
- The holes. Porosity — pores, vesicles, joints, vugs — is air or water occupying volume that used to be counted as rock. Air is ~0.001 g/cm³; water is ~1.0. Either is a lot lighter than the silicate frame: bulk density falls as soon as the fabric opens.
A tight Cornish granite and a vesicular Icelandic flow, both “igneous”, do not weigh the same in the hand. Composition sets the ceiling; pores knock it down.
Where the Archimedes formula comes from
A body in a fluid is pushed up by a force equal to the weight of the fluid it displaces. That is the whole of Archimedes. On the balance it shows up as a loss of apparent mass.
mair − mwet = ρwater × V
Rearrange and the volume that a ruler could not measure is in hand. Divide the dry mass by that volume and density follows. g cancels: both weighings are already masses.
V = (mair − mwet) / ρwater
ρ = mair / V = mair × ρwater / (mair − mwet)
mwet has to be smaller than mair. If it is not, the sample is floating, still touching the beaker, or the two readings were swapped. A pumice that floats is not a lab error — it is a rock whose bulk density is below 1 g/cm³.
Hydrostatic weighing on the bench: kit and pitfalls
Hydrostatic weighing* is the method for a sample with no useful geometry — a cobble, a core stub, a hammer chip. Two readings on one balance beat a dozen calliper guesses.
- Analytical balance to 0.01 g (0.001 g if the chip is small), with an under-hook or a hanging pan.
- Beaker of distilled or deionised water, deep enough that the sample hangs free of the walls and the free surface.
- Fine nylon or stainless wire; a stand that does not rock.
- Thermometer to ±0.5 °C, because ρwater is not a constant invented at the desk.
- A dry, brushed sample — no mud, no salt crust, no clinging sawdust from the slab saw.
The catch on a porous rock: during the immersed weighing, water starts to enter the pores and the apparent mass creeps upward. For a granite that is a rounding error. For a sandstone it is the whole experiment. Either coat the dry sample (thin paraffin or a spray lacquer) so the envelope volume is what is measured, or switch to a saturation protocol and accept that a dry block is no longer being weighed. Bubbles on the surface act like extra volume; they are tapped off. The wire has mass in water too — tare it, or pick a thread whose immersed weight is negligible next to the sample.
A dry, uncoated reservoir sandstone dunked and called bulk density is a moving target: water is invading while the balance is read. Coat it, or saturate it on purpose and report which density was measured.
Typical densities of common rocks
Dry bulk ranges, not commandments. A number outside the band is a clue, not a proof.
- Granite, granodiorite — 2.60 to 2.75 g/cm³. Sierra Nevada and Cornish batholith cores cluster near 2.67.
- Basalt, gabbro — 2.85 to 3.10 when massive. Vesicular flow tops drop well below that.
- Sandstone — 2.20 to 2.60, driven more by porosity and cement than by the quartz itself.
- Limestone, dolostone — 2.30 to 2.85. Tight micrite sits high; a vuggy reef sits low.
- Gneiss, schist — 2.60 to 2.85, overlapping granite unless garnet or amphibole pile up.
- Marble — 2.65 to 2.75, close to calcite’s 2.71 when the fabric is tight.
- Peridotite, dunite — 3.20 to 3.40. Felt the moment it is picked up.
- Shale, mudstone — 2.0 to 2.7, a messy band because clay-bound water and compaction both wander.
A 3.3 g/cm³ “basalt” is probably not basalt. Olivine, ore, or a mis-tare is the next place to look.
Bulk density, grain density and the porosity identity
Grain density* (matrix density, sometimes “true” density) is dry mass over the volume of the solids alone. Crush the sample, or use a helium pycnometer* on a plug, and pores drop out of the denominator. Helium sees almost every open pore; the grain volume that remains is the mineral volume. What remains is the mineral mix — the number a petrographic calculation from mineral densities should match. Friable samples, or plugs that would flood during a dunk, belong on the pycnometer; hydrostatic weighing belongs when the envelope of a hand specimen is the question.
Bulk density keeps the pores in the volume. That is what hydrostatic weighing of an intact piece reports, provided water is not sneaking in. Dry bulk uses the dry mass; saturated bulk uses the saturated mass. Same envelope, different numerator.
φ ≈ 1 − (ρb / ρg)
In practice that identity is how many core labs get total porosity without a separate pore-volume run. A jointed granite and a tight granite can share grain density and disagree on bulk density by several percent — the cracks are the difference, and they are also the difference between a dry hole and a small aquifer.
Water temperature and the third decimal
Water is densest at 4 °C (1.0000 g/cm³). At 20 °C it is 0.9982; at 25 °C, 0.9970. Typing 1.000 into the box when the beaker is at room temperature systematically understates volume and overstates ρ.
A 5 °C miss is about 0.1% on density. Fine for a field classification. Not fine if a Bouguer slab or a reservoir-grade grain density is being fed. Read the bath, look up ρwater(T), and type that. Dissolved salt in tap water pushes ρ the other way; distilled water plus a thermometer is cheaper than arguing about the third decimal.
Gravity surveys: density contrast in the subsurface
Gravity does not see lithology. It sees mass. A dense body in lighter country rock makes a positive anomaly; a light body (salt, coal, a thick clastic basin) makes a negative one. Densities are then assigned to units and the map is matched.
Those assigned densities come from this measurement — cores, outcrop chips, hand specimens — not from a textbook average someone remembered. A peridotite pod in granite is an easy contrast (~0.6 g/cm³). A limestone–dolostone contact is not. Ore bodies (massive sulphide, magnetite skarn) shout; a porosity pinch-out whispers. Pair the gravity with seismic velocities and density stops being a free parameter in geology.
Worked example: an irregular chip and a likely lithology
Dry chip, irregular, no coating needed — the rock looks crystalline and tight. Water at 20 °C.
- mair = 248.6 g
- mwet = 155.1 g (hanging, fully immersed, not touching the glass)
- ρwater = 0.9982 g/cm³
V = (248.6 − 155.1) / 0.9982 = 93.67 cm³
ρ = 248.6 / 93.67 = 2.654 g/cm³
2.65 g/cm³ sits in the granite–gneiss window, not in the basalt window. Quartz and feldspar on the broken face would settle it; a thin section would name it. The balance has already ruled out gabbro. Uncertainty on a careful run of this size is around ±0.01 g/cm³ — enough to separate granite from basalt, not enough to separate two granites.
Dry bulk versus saturated bulk
Same piece of rock, two masses, one envelope volume. Dry bulk is the core-analysis default and the number gravity wants for the unsaturated zone. Saturated bulk is what sits below the water table, and what a saturated-surface-dry (SSD) protocol reports.
ρsat ≈ ρdry + φ × ρfluid
A sandstone at φ = 22% and ρdry = 2.08 g/cm³ is about 2.30 g/cm³ when the pores are full of water — a jump of 0.22, which is a large Bouguer contrast if forgotten. Oil in the pores instead of water lands somewhere in between. Which bulk density is meant has to be written; “density of the sandstone” is not a sentence.
North Sea reservoir sandstone: Brent and Forties plugs
North Sea core labs quote dry bulk density because it pairs with grain density to give porosity without a separate Boyle’s-law run. A clean Brent (Middle Jurassic) or Forties (Paleocene) plug at 22–28% porosity, quartz framework near 2.65 g/cm³, lands around 1.90–2.10 g/cm³ dry bulk. Tight, cemented intervals climb toward 2.40. That is not a granite number and it is not a basalt number: it is a pore-volume number.
BGS memoirs and operator core reports for the UKCS treat 2.05 g/cm³ as a perfectly ordinary sandstone. Feeding 2.67 into a Bouguer slab over a thick Forties fairway leaves a hole in the residual that is just the reservoir. The same plug, saturated with brine, sits near 2.25–2.30 — the figure a seismic modeller wants below the oil–water contact, not the figure a wellsite geologist writes on a dry-core ticket.
Icelandic basalt versus granite in the rucksack
A massive Icelandic tholeiite, vesicles gone, sits at about 2.90 to 3.00 g/cm³, pyroxene and plagioclase, little or no quartz, a color index deep into the melanocratic band. A Cornish or Sierra granite at ~2.67 g/cm³, modest biotite, is a different material in the bag — because it is.
The flow top of that same Icelandic basalt, scoriaceous, can fall below 2.5. Composition did not change; porosity did. Mafic minerals raise grain density, vesicles lower bulk density, and a single lithologic name covers both if the notebook is careless. USGS and BGS map units for flood basalt both need the qualifier “massive” or “vesicular” before density is assigned.
Bouguer reduction at 2.67 g/cm³
A Bouguer* correction replaces the rock between the station and datum with an infinite slab of chosen density. Continental work still defaults to 2.67 g/cm³ — a granite number — because so much of the upper crust behaves like that. Use 2.67 on a Columbia River or Icelandic plateau and a slab of extra mass is left in the residual. Use 2.90 on a granite terrain and a hole is carved that is not there.
Nettleton profiles (trial densities until terrain and Bouguer stop correlating) are how the guess is checked. A 0.1 g/cm³ error on a 1 km slab is ~4 mGal. That is a real anomaly, or a fake one, depending on whether the rocks being modelled were measured. This calculator is the unglamorous source of that 2.67.
Glossary
- Density (ρ)
- Mass divided by volume. For rocks it is usually reported in g/cm³. 2.67 g/cm³ = 2,670 kg/m³.
- Specific gravity (SG)
- Density of the sample divided by the density of water at 4 °C. Dimensionless. Numerically equal to density in g/cm³ when water is taken as 1.000 g/cm³.
- Hydrostatic weighing
- Archimedes’ method: the dry sample is weighed in air, then weighed while fully immersed. The apparent-mass loss equals the mass of water displaced, hence the volume.
- Grain density (ρg)
- Dry mass divided by the volume of the solids only — pores excluded. Also called matrix density. The number a mineral-mode calculation should match.
- Pycnometer
- A device that measures the volume a fluid (liquid, or helium gas) can occupy. Helium pycnometry on a plug gives grain volume, hence grain density.
- Bouguer correction
- In gravity surveying, the replacement of the rock between station and datum by an infinite slab of chosen density. Continental default is often 2.67 g/cm³.