Environmental Engineering Reference
In-Depth Information
practical in a typical field-sampling situation. First, it is not possible to analyze a very
large number of samples for each and every area of interest. Second, the size of the area
to be treated may be controlled by other than pure sampling or statistical considerations.
For instance, when soil plant nutrient levels are determined as a basis for the commercial
application of agriculture chemicals, the size of the area sampled is determined by
equipment size, capacity, and economics.
For contaminated fields, we are generally interested in differentiating between
contaminated and uncontaminated areas. It is important to know when the level of
contaminant is at or below some specified target level. When three samples from an area
have analytical results showing that the contaminant is below the target level at the 95%
confidence level, then the desired information has been obtained and further sampling
may not be needed. (Confidence levels will be discussed in Chapter 6.)
For these reasons, a more systematic sampling scheme is often used. An imaginary grid
can be set up to cover the field. At this point each grid can be randomly sampled, or a
selected number of grids can be sampled on a random basis. Even fewer random samples
can be obtained by specifying positions within each grid area in which samples are to be
taken. In this case, these positions can be repeatedly sampled to determine the changes in
contaminant or component concentration over time.
In the case of agriculture, grids large enough to be treated individually are usually
sampled. Here the desire is to have an estimate of the amount of fertilizer to be added to
reach a certain yield goal, but not to cause environmental harm. Also, if an adjacent grid
requires more or less fertilizer that adjustment can be made as needed. In this type of
sampling, the average amount of available plant nutrient present in each grid is important
for making fertilizer application. It should be kept in mind that very precise and accurate
applications rate cannot be obtained, and because of this it is not necessary to be able to
distinguish very small differences.
When contaminated areas are sampled the desired sampling outcomes are different.
Where in situ bioremediation is to be carried out repeated sampling of the area will be
necessary. Here we wish to make sure that the contamination is not spreading and to
know when the contaminant has fallen below a specified level in a specific area. In other
cases, the soil in the field sample is to be removed and remediated in another location. In
this case, sampling will only be done once, and so consideration of repeated sampling is
not in question [15].
5.16. QUALITY CONTROL
There is always a great deal of concern regarding quality control and quality assurance.
Does the analysis and reporting of the results accurately and faithfully represent the
situation in the field or in the environment? Often samples are spiked; a precisely known
quantity of contaminant is added to a control sample. Spiking should be done in the field
after the sample is in its container. If this precisely known quantity has changed during
transportation, storage, or handling, the results of analysis of samples are suspect.
A problem arises when the contaminant of concern is either highly toxic or volatile. It
is undesirable to increase the amount of toxic material in a sample, both from the
 
Search WWH ::




Custom Search