SAMPLING

The value of any product is judged by the characteristics of the sample as determined by laboratory tests.

The sample used for the test(s) must be representative of the bulk material, or data will be produced that are not representative of the material and will, to be blunt, be incorrect no matter how accurate or precise the test method is. In addition, the type and cleanliness of sample containers are important: If the container is contaminated or is made of material that either reacts with the product or is a catalyst, the test results may be wrong.

Thus the importance of the correct sampling of any product destined for analysis should always be overemphasized. Incorrect sampling protocols can lead to erroneous analytical data from which performance of the product in service cannot be accurately deduced. For example, properties such as specific gravity, distillation yield, vapor pressure, hydrogen sulfide content, and octane number of the gasoline are affected by the content of low-boiling hydrocarbons so that suitable cooling or pressure sampling methods must be used and care must be taken during the subsequent handling of the sample to avoid the loss of any volatile components. In addition, adequate records of the circumstances and conditions during sampling must be made; for example, in sampling from storage tanks, the temperatures and pressures of the separation plant and the atmospheric temperature would be noted.

At the other end of the volatility scale, products that contain, or are composed of, high-molecular-weight paraffin hydrocarbons (wax) that are also in a solid state may require judicious heating (to dissolve the wax) and agitation
(homogenization, to ensure thorough mixing) before sampling. If room-temperature sampling is the modus operandi and product cooling causes wax to precipitate, homogenization to ensure correct sampling is also necessary.

Representative samples are prerequisite for the laboratory evaluation of any type of product, and many precautions are required in obtaining and handling representative samples. The precautions depend on the sampling procedure, the characteristics (low-boiling or high-boiling constituents) of the product being sampled, and the storage tank, container, or tank carrier from which the sample is obtained. In addition, the sample container must be clean, and the type to be used depends not only on the product but also on the data to be produced.

The basic objective of each procedure is to obtain a truly representative sample or, more often, a composite of several samples that can be considered to be a representative sample. In some cases, because of the size of the storage tank and the lack of suitable methods of agitation, several samples are taken from large storage tanks in such a manner that the samples represent the properties of the bulk material from different locations in the tank and thus the composite sample will be representative of the entire lot being sampled. This procedure allows for differences in sample that might result from the stratification of the bulk material because of tank size or temperature at the different levels of the storage tank. Solid samples require a different protocol that might involve melting (liquefying) of the bulk material (assuming that thermal decomposition is not induced) followed by homogenization. On the other hand, the protocol used for coal sampling might also be applied to sampling of petroleum products, such as coke, that are solid and for which accurate analysis is required before sale.

Once the sampling procedure is accomplished, the sample container should be labeled immediately to indicate the product, time of sampling, location of the sampling point, and any other information necessary for the sample identification. And, if the samples were taken from different levels of the storage tank, the levels from which the samples were taken and the amounts taken and mixed into the composite should be indicated on the
sample documentation.
Although the above text is focused on the acquisitions of samples from storage tanks, sampling records for any procedure must be complete and should include, but are not restricted to, information such as:

1. The precise (geographic or other) location (or site or refinery or process) from which the sample was obtained.
2. The identification of the location (or site or refinery or process) by name.
3. The character of the bulk material (solid, liquid, or gas) at the time of sampling.
4. The means by which the sample was obtained.
5. The protocols that were used to obtain the sample.
6. The date and the amount of sample that was originally placed into storage.
7. Any chemical analyses (elemental analyses, fractionation by adsorbents or by liquids, functional type analyses) that have been determined to date.
8. Any physical analyses (API gravity, viscosity, distillation profile) that have been determined to date.
9. The date of any such analyses included in items 5 and 6.
10. The methods used for analyses that were employed in items 5 and 6.
11. The analysts who carried out the work in items 5 and 6.
12. A log sheet showing the names of the persons (with the date and the reason for the removal of an aliquot) who removed the samples from storage and the amount of each sample (aliquot) that was removed for testing.

In summary, there must be a means of identifying the sample history as carefully as possible so that each sample is tracked and defined in terms of source and activity.
Thus the accuracy of the data from any subsequent procedures and tests for which the sample is used will be placed beyond a reasonable doubt.