Grain-size analysis sits at the foundation of nearly every sediment characterisation study, and the choice of method shapes the data you get long before any calculation begins. Laser diffraction instruments like the Malvern Mastersizer have become the default in many laboratories over the past two decades, largely because they are fast, reproducible, and require small sample masses. But the hydrometer method, standardised under ASTM D422 and ISO 17892-4, has not been displaced — it remains the legally recognised procedure in many geotechnical and environmental regulatory contexts, and it behaves differently enough from optical diffraction that treating the two as interchangeable leads to real interpretive errors. This article works through the physical principles behind each technique, the sediment types and project conditions where one outperforms the other, and the practical steps needed when you must reconcile data from both methods in a single dataset.

The physics each method actually measures

The hydrometer method derives particle size from Stokes' Law: a sphere settling through a viscous fluid at a known temperature reaches a terminal velocity that depends on its diameter and the density contrast between the particle and the fluid. You measure the density of the suspension at timed intervals using a calibrated hydrometer, and from that density profile you reconstruct a cumulative size distribution. The particle is treated as a sphere of equivalent settling velocity, which is why the result is often called the equivalent spherical diameter or ESD. Laser diffraction works on a completely different physical principle. A collimated laser beam passes through a dilute suspension, and the forward-scattered light pattern is recorded on a detector array. Mie scattering theory, or Fraunhofer approximation for coarser fractions, relates the angular intensity pattern to a volume-weighted particle size distribution. Crucially, the instrument still assumes spherical particles, but it detects size from optical cross-section rather than settling behaviour. For equidimensional grains these agree reasonably well. For platy phyllosilicates, elongated biogenic fragments, or aggregated clay flocs, they can diverge by a full phi unit or more.

Where the methods agree — and the grain-size window where they do not

For well-sorted medium to coarse silt (roughly 10 to 50 micrometres) composed of quartz or feldspar, laser diffraction and hydrometer analysis typically return D50 values within 5 to 10 percent of each other, which for most interpretive purposes is negligible. The agreement tightens further when samples are properly dispersed with sodium hexametaphosphate and when the laser instrument's optical model is set to a refractive index appropriate for quartz (approximately 1.544 real, 0.001 imaginary). Divergence becomes practically significant in two windows. In the clay fraction below 2 micrometres, laser diffraction consistently under-reports clay content relative to the hydrometer, because platy kaolinite and smectite particles scatter light as if they were larger spheres oriented randomly in flow. Published inter-laboratory comparisons on reference kaolinite suspensions have shown laser diffraction reporting clay fractions 8 to 20 percentage points lower than hydrometer results on the same material. At the coarse silt and very fine sand boundary, around 50 to 100 micrometres, organic coatings and biofilms can make particles behave as lower-density aggregates in settling, inflating the apparent size in hydrometer results while laser diffraction sees through to the optical grain boundary.

Sample preparation: where most errors originate

Both methods are sensitive to dispersion quality, but they fail differently when dispersion is incomplete. In a hydrometer test, undispersed flocs settle faster than the constituent particles, so the distribution shifts coarse and the clay fraction is under-reported. You catch this by running a blank and checking that the initial hydrometer reading after 40 seconds corresponds to total solids in suspension. In laser diffraction, undispersed flocs register as large particles in the volume distribution and produce a bimodal or right-skewed output that looks real but is not. For fine lacustrine or estuarine sediments, the standard preparation sequence at Silt Lab Zone uses 4 percent sodium hexametaphosphate solution at 40 grams per litre, 12 hours of overnight soaking, followed by one minute of ultrasonication at 40 watts directly before measurement. For carbonate-rich samples, a 10 percent hydrogen peroxide treatment removes organic matter before dispersion, because organic binding agents are the primary cause of persistent aggregation in calcareous muds. Skipping the peroxide step on an organic-rich sample and simply increasing dispersant concentration rarely achieves full deflocculation and can introduce surfactant interference in the optical detector.

Regulatory and geotechnical contexts where hydrometer remains mandatory

Many national geotechnical standards still specify the hydrometer or pipette sedimentation method explicitly, and regulatory submissions prepared under those standards cannot substitute laser diffraction data without a formal equivalence study or explicit approval from the supervising authority. In Australia, for instance, AS 1289.3.6.3 governs hydrometer analysis for engineering soil classification, and laboratory accreditation under NATA (National Association of Testing Authorities) requires method traceability to that standard for geotechnical reports used in infrastructure assessments. For contaminated land investigations under the National Environment Protection (Assessment of Site Contamination) Measure, grain-size data used to support contaminant mobility modelling is typically required to follow a sedimentation-based method, because the transport physics of fine-grained contaminant carriers in groundwater is more directly analogous to Stokes settling than to optical cross-section. Environmental consultants working on sites subject to EPA Victoria or EPA NSW oversight should confirm the accepted method with the relevant authority before generating a laser diffraction dataset they expect to use in a regulatory context.

When laser diffraction is the better tool

Speed and sample throughput are the obvious advantages: a Malvern Mastersizer 3000 completes a measurement in under two minutes per replicate, with three replicates taking around ten minutes including rinse cycles. A full hydrometer run requires a technician present at 40 seconds, 2 minutes, 5 minutes, 15 minutes, 30 minutes, 1 hour, 2 hours, and then at 24 hours. For a research project involving 200 core sub-samples, the labour difference is not marginal. Laser diffraction also handles mixed mineralogy datasets more consistently when the goal is relative comparison rather than absolute size. If you are tracking downcore changes in grain-size distribution through a sediment sequence to reconstruct palaeoenvironmental energy conditions, what matters is that every sample is measured the same way. Laser diffraction's reproducibility, with coefficients of variation on D50 typically below 2 percent when sample preparation is standardised, makes it well suited to studies where pattern and trend matter more than exact agreement with a settling-velocity definition. It is also the practical choice when sample mass is limited: hydrometer analysis requires 15 to 50 grams of dry sediment, whereas the Mastersizer works reliably with 0.1 to 0.5 grams.

Converting between methods: what the correction factors actually look like

Several published conversion approaches exist for translating laser diffraction outputs to hydrometer-equivalent values, most based on empirical regression on paired datasets from the same sediment type. Konert and Vandenberghe (1997) proposed multiplying the laser diffraction D50 by a factor of approximately 0.54 to obtain the settling-equivalent diameter for fine silts and clays, derived from comparisons on loess and riverine mud. That factor has since been refined by multiple groups working on marine sediments, lacustrine carbonates, and glaciogenic tills, and it varies between 0.4 and 0.7 depending on mineralogy and particle shape. The practical implication is that there is no universal correction factor. If your project requires laser diffraction data to be reported in hydrometer-equivalent terms, you need to run a method-comparison sub-study on a representative subset of your own samples, ideally 15 to 20 pairs spanning the full textural range of your material, and derive a site-specific regression. Linear regression on paired D10, D50, and D90 values is the minimum; for datasets with complex bimodal distributions, a component-by-component comparison of the sand, silt, and clay fractions produces more defensible conversion parameters. Document the regression statistics and report the uncertainty explicitly — a conversion derived from a ten-sample subset applied to 300 core samples carries propagated uncertainty that should appear in your methods section.

Practical decision framework for method selection

The decision reduces to four considerations examined in sequence. First, is the output destined for a regulatory submission or a geotechnical classification report under a named national standard? If yes, use the specified sedimentation method and do not substitute. Second, is the primary goal relative comparison across a large sample set, where throughput and reproducibility outweigh absolute accuracy against a settling-velocity standard? Laser diffraction is appropriate. Third, does the sediment contain more than roughly 20 percent clay-sized material, and does that clay fraction matter for the interpretive conclusions? If so, be aware that laser diffraction will under-report the clay fraction and plan either to correct for it or to supplement with a pipette or hydrometer measurement on a subset. Fourth, is sample mass limited to below 5 grams per specimen? Hydrometer analysis is not feasible at that mass; laser diffraction is your only option. For mixed projects, the most defensible approach is a combined workflow: laser diffraction for the full sample set to capture high-resolution volume distributions efficiently, with hydrometer analysis run in parallel on 10 to 15 percent of samples selected to span the textural extremes. The parallel dataset anchors the laser diffraction results to a physically interpretable settling-velocity framework and provides the comparison data needed for any regulatory review.

Method choice in grain-size analysis is a technical decision with downstream consequences for interpretation, regulatory acceptance, and inter-study comparability. The hydrometer and laser diffraction methods measure different physical proxies of size, and understanding that distinction precisely is what allows you to select, apply, and report either technique with confidence. The Silt Lab Zone team runs both methods in-house and can advise on method selection, parallel calibration studies, and data reconciliation for projects at any stage.