130
6 Risk Assessment and Management of Chemical Products
Generally, the distribution of a chemical between compartments is described by
partition coefficients, and degradation processes are described by degradation rate
constants (see Appendix B).
Multi-compartment models as described here can often be run with a relatively
small set of chemical property data, e.g., vapor pressure, water solubility, K ow ,
and k transformation . However, it is important to keep in mind that these property
data describe the environmental fate of nonionizing organic chemicals. The environmental fate of acids and bases, salts, or other types of materials, for example,
nanoparticles, cannot be described by this set of chemical property data. Multicompartment models can also be set up for these other types of substances, but then
the models have to be fed with property data that reflect the chemical and physical
behavior of these substances (Praetorius et al., 2014).
Another fundamental assumption of multi-compartment models is that within
each compartment, a chemical is always distributed homogeneously. Therefore,
concentration gradients, for example, for a chemical in a lake or river, can only
be described by several separate compartments, for example, for the layers of a
stratified lake (epilimnion and hypolimnion) or different stretches of a river.
Depending on the problem investigated, multi-compartment environmental fate
models can be built with lower or higher spatial and temporal resolution. Models
with few compartments (e.g., air, water, soil, sediment) at steady state are generic
and relatively coarse. They can be used to understand the basic features of a
chemical’s circulation in the multi-compartment system (Scheringer and MacLeod,
2021). But it is also possible to set up multi-compartment models with a high spatial
resolution (many grid cells or spatial domains) (see MacLeod et al., 2011 (BETRGlobal) and Glüge et al., 2016), and high temporal resolution (level IV models with
short time steps) (Camenzuli et al., 2012). Of course, with increasing spatial and
temporal resolution of the models, also more data on environmental conditions are
needed.
Many models have been developed that operate with different amounts of
required inputs and different adjustable settings and within different programming
environments. Reviews of some of them are provided by Pistocchi et al. (2010)
and Di Guardo et al. (2018). Some examples of applying these models to various
settings from local to global include the investigation of polychlorinated biphenyls
(PCBs) in an urban setting (Gasic et al., 2009), the global fate of the organochlorine
insecticide endosulfan (Becker et al., 2011), and the global emissions and fate of
perfluorooctanesulfonic acid (PFOS) (Wang et al., 2017).
Finally, for developers and users of any kind of environmental fate model, it is
important to properly assess the validity of the model results. To help with this,
Buser et al. (2012) developed a set of guidelines for good modeling practice.
6 Risk Assessment and Management of Chemical Products
Generally, the distribution of a chemical between compartments is described by
partition coefficients, and degradation processes are described by degradation rate
constants (see Appendix B).
Multi-compartment models as described here can often be run with a relatively
small set of chemical property data, e.g., vapor pressure, water solubility, K ow ,
and k transformation . However, it is important to keep in mind that these property
data describe the environmental fate of nonionizing organic chemicals. The environmental fate of acids and bases, salts, or other types of materials, for example,
nanoparticles, cannot be described by this set of chemical property data. Multicompartment models can also be set up for these other types of substances, but then
the models have to be fed with property data that reflect the chemical and physical
behavior of these substances (Praetorius et al., 2014).
Another fundamental assumption of multi-compartment models is that within
each compartment, a chemical is always distributed homogeneously. Therefore,
concentration gradients, for example, for a chemical in a lake or river, can only
be described by several separate compartments, for example, for the layers of a
stratified lake (epilimnion and hypolimnion) or different stretches of a river.
Depending on the problem investigated, multi-compartment environmental fate
models can be built with lower or higher spatial and temporal resolution. Models
with few compartments (e.g., air, water, soil, sediment) at steady state are generic
and relatively coarse. They can be used to understand the basic features of a
chemical’s circulation in the multi-compartment system (Scheringer and MacLeod,
2021). But it is also possible to set up multi-compartment models with a high spatial
resolution (many grid cells or spatial domains) (see MacLeod et al., 2011 (BETRGlobal) and Glüge et al., 2016), and high temporal resolution (level IV models with
short time steps) (Camenzuli et al., 2012). Of course, with increasing spatial and
temporal resolution of the models, also more data on environmental conditions are
needed.
Many models have been developed that operate with different amounts of
required inputs and different adjustable settings and within different programming
environments. Reviews of some of them are provided by Pistocchi et al. (2010)
and Di Guardo et al. (2018). Some examples of applying these models to various
settings from local to global include the investigation of polychlorinated biphenyls
(PCBs) in an urban setting (Gasic et al., 2009), the global fate of the organochlorine
insecticide endosulfan (Becker et al., 2011), and the global emissions and fate of
perfluorooctanesulfonic acid (PFOS) (Wang et al., 2017).
Finally, for developers and users of any kind of environmental fate model, it is
important to properly assess the validity of the model results. To help with this,
Buser et al. (2012) developed a set of guidelines for good modeling practice.