Page 206 - Materials Chemistry, Second Edition

L1644_C05.fm Page 179 Monday, October 20, 2003 12:02 PM

Compilation of damage Classification of data
functions data

Selection of probability Parameter supposed
distribution to be fix
Determination of distribution Distribution for data based
for extensively available data on little information
by literature and experts
Monte Carlo Simulation

Results Sensitivity Analysis
Analysis and discussion of results

FIGURE 5.3 Framework for the assessment of uncertainty and variability in the impact
pathway analysis.

estimations must be considered. All these parameters feed the MC simulation, which
gives the results in the form of a probability distribution around a mean value and
allows a detailed sensitivity analysis to be carried out.

5.5.2 UNCERTAINTY ASSESSMENT IN IMPACT PATHWAY ANALYSIS

Figure 5.3 presents the framework for uncertainty assessment in the IPA. The first
step of the so-called framework is the compilation of damage function data, in which
an exhaustive study must be carried out on all the parameters that have a repercussion
on the final result. Although the model is processing an enormous quantity of data
that are not all relevant, only fundamental facts need really be considered. Thus, a
classification must be made among the most significant parameters, for which prob-
ability distributions should be defined. In addition, the parameters are supposed to
be invariant and are called point estimates. Significant data are further classified into
the above-mentioned two groups for advanced evaluation: extensively available data
and data based on little information. In the same way as in the uncertainty assessment
for LCI, these parameters feed the MC simulation that gives the results in the form
of a probability distribution around a mean value, and allows a detailed sensitivity
analysis to be carried out. The last step of the framework consists of the analysis
and discussion of the achieved results.

5.5.3 RISK CHARACTERIZATION AND UNCERTAINTY ANALYSIS
Risk assessment uses a wide array of information sources and models. Even when
actual exposure-related measurements exist, assumptions or inferences will still be
required. Most likely, data will not be available for all aspects of the exposure
assessment and may be of questionable or unknown quality. In these situations, the
exposure assessor will depend on a combination of professional judgment, inferences
based on analogy with similar chemicals and conditions, estimation techniques and

© 2004 CRC Press LLC