Comparison of sampling and grain-size analysis methods 
Scope 

Current large-scale infra-structural plans in the Dutch nearshore area will require , if carried out , substantial amounts of the adequate type of sand. As a result, high quality surveys of potential extraction areas will be needed as well as state-of-the-art soil sample analysis to estimate (sand) grain-size distributions. 

As a first step, an appraisal of existing archive data is imperative. Archive data on sand from the southern North Sea originate from a variety of sources, largely collected over the last 35 years. Sand for grain-size estimates and/or analysis was sampled by different means: from sea bed samples, cores and bore holes. In most cases only the D50 was estimated. In some cases grain-size distributions were constructed. 

In the past this was done by various kinds of sieving analysis. Nowadays, grain size distributions are largely analysed by laser counter. Another, recently introduced, method is based on quantities of automatic grain measurements. It is important for the future that there is a good assessment for inter-comparison of these sampling and analytical methods, as well as for the relationship between single D50 values and full grain size distributions. 

In this context a short, preliminary study was carried out very recently using sand from a few existing bore holes and recent cores from the same general location, a sand wave area in Dutch offshore hydrocarbon concession block S2, that was or had been analysed by various means.

Methods

Sampling methods

Counter flush drilling system

The Netherlands Institute of Applied Geoscience-TNO  (TNO-NITG) uses two types of counterflush drilling systems. For this study the Geodoff counterflush/airlift drilling device was used, which penetrates down to 12 metres below seabed. The resulting

mixture of water and sediment particles, per metre of penetration, is led into a separate sampling basin with an overflow. Care is taken to collect a representative sample from the sampling basin. Samples from the uppermost 5 metres have been taken into account for the present study.

Vibrocoring system

Various vibrocoring systems are operated by TNO-NITG. For this study a powerful hydraulic vibrocorer called "Trilflip", capable of taking undisturbed 5-6 metre cores in a matter of seconds, was deployed. Averaged samples from between 2-3 and 3-4 metres have been used for grain-size analysis.

Grain-size analysis methods

Sieving analysis

Two hundred grams of dry sand were sieved for 15 minutes using a Retsch sieving machine. Sieves used followed the whole phi Wentworth system i.e. 63, 125, 250, 500, 1000 and 2000µm. The sand that remains in each sieve was weighed and the resulting values were used for the construction of cumulative curves.

Laser counter

The sediment sample was put through a 2000µm sieve and several times through a 1 : 1 sample splitter. The resulting 5g was introduced into a Malvern laser counter Z, version 1.2a, equipped with a MSX15 automatic sampler. The sample was  brought into suspension using an ultrasonic bath with a stirring device and subsequently pumped through the measurement cell. A laser bundle traverses the cell in such a way that the diffraction pattern caused by the passing grains can be measured. These grain surface measurements were being processed by Malvern software so that grain volumes are calculated based on a globular grain shape.

Go to page 4