Characterization of biomass using instruments – Measurement of forest and crop residues - Biomass as Energy Source: Resources, Systems and Applications - page 61

correct price settled between buyers and sellers, but the cited precision requirement implies a very
large number of samples. The industry seems to have made a tradeoff against what is practically
feasible and, instead of 20-30 samples taken from all parts of a shipment, some 2-3 samples are
typically taken.
A key issue is to differ between random errors and systematic errors. Although the gravimetric
moisture measurement method is quite reliable in itself, the practical implementation deserves
special attention. Representative sampling is also of great importance since the objective is to
obtaining a measurement result that represents the bulk moisture content of a certain shipment.
A common sampling method is to use a shovel to take samples in 2-3 positions in a shipment.
The two main approaches are to collect the samples close to the surface layer of the shipment
or to collect samples from material tipped on the ground. The first alternative might be the only
option when sampling on tipped fuel is either inconvenient or not possible because the biomass is
tipped directly to an entry pocket of a silo or the boiler feed. Another reason for the procedure is
that the reception station might have an inspection bridge adjacent to the lab ovens so that samples
can be taken easily.
In another section, a number of factors that might cause a systematic difference between the
bulk humidity and themoisture close to the surface are discussed. Therefore, amore representative
sampling can be expected if the shipment is unloaded on the ground and sampling can be made in
randomly chosen positions in the bulk.
Another rationale for sampling on tipped material is that the supplier has less potential for
manipulation of the measurement result. One way to manipulate the price would e.g. be to load
wet material in the lower section of the container if sampling is made close to the surface. A key
issue to remember from the above is that the choice of sampling method, and the number of
samples taken, will have an impact on the measurement results.
3.4.1.3 Practical illustration
When the price is based upon a large number of containers, individual errors tend to cancel out.
This is also likely to be the case if similar material is delivered from the same supply. In these
cases, the parties may agree on a lower sampling intensity. In other cases, the number of containers
from a certain supplier might be small. Then, individual errors will have a larger impact on the
price paid to each supplier. This is a reason why high precision is desirable.
Figure 3.4 shows a histogram of the variation in moisture content in a container with chipped
branches and tops sent from a terminal to a Swedish plant in February 2012. The material has been
tipped on the ground. Sampling was made in 20 evenly distributed squares. Then, the moisture
content was determined using the gravimetric method. The y -axis shows the number of samples
that fall within the respective 2%-unit intervals indicated on the x -axis. Assuming that sampling
is made at one point per container it is clear that manual sampling does not give a reliable result
for individual containers. This conclusion is supported in a study by Aulin et al . (2008) which
also indicates that Swedish industry participants are well aware that manual sampling might lead
to erroneous results and that the total error is far larger than the standard deviation for gravimetric
method itself.
3.4.2 Instrumental methods
3.4.2.1 Introduction
Moisture determination through loss of drying has been described earlier in this chapter. It is a
reliable method with a precision of approx. 2%-units of moisture. A clear disadvantage is that the
method is subject to a sampling error and that results usually are obtained after 20-24 hours.
While sawdust or woodchips may be analyzed with, for example, impedance measurement, it is
far more difficult to determine the moisture content of complexmixtures. Today, new technologies
such as NIR spectroscopy, RF spectroscopy and X-ray methods have emerged for more advanced
applications such as process control.