Page 123 - Materials Chemistry, Second Edition

106 LIFE CYCLE ASSESSMENT HANDBOOK

GaBi or SimaPro. Software, however, is the subject of another chapter and
will not be covered here. Another option is to build the LCI in a tailor-made
fashion, directly from data sources. This approach is covered in the following
sections.
In many instances, creating an LCI begins with the collection of raw data
which are data extracted from various sources, such as bookkeeping of a plant,
national statistics, technical journals, etc., but not yet related to the process
for which the dataset is being developed. Typically, a number of sources are
needed to be called upon to collect a sufficient amount of data. Other examples
of data sources that may be drawn from or utilized include the following:

• Meter readings from equipment
• Equipment operating logs/journals
• Industry data reports, databases, or consultants
• Laboratory test results
• Government documents, reports, databases, and clearinghouses
• Other publicly available databases or clearinghouses
• Journals, papers, books, and patents
• Reference books
• Trade associations
• Related /previous life cycle inventory studies
• Equipment and process specifications
• Best engineering judgment (EPA 2006)

Once raw data are collected, following a pre-determined data collec-
tion approach, unit process datasets can be created by defining mathemati-
cal relationships between the raw data and various flows associated with
the dataset in a defined reference flow. Data modeling requirements, with
desired quality attributes and adequate documentation, are specified to accu-
rately transform raw data into unit process datasets, and incorporate proper
review and documentation to address verification and transparency issues
(Consoli, Allen et al. 1993; Curran 2011). Therefore, understanding how data
flow from raw data providers to LCI data users (shown in Figure 5.1) is
important because data move from the raw state to and through datasets
and databases.
Recycling provides an example of some of the strengths and limitations
encountered in gathering data. For some products, economic-driven recycling
has been practiced for many years, and infrastructure and markets for these
materials already exist. Data are typically available for these products, includ-
ing recycling rates, the consumers of the reclaimed materials, and the resource
requirements and environmental releases from the recycling activities (col-
lection and reprocessing). Data for materials currently at low recycling rates
with newly forming recycling infrastructures are more difficult to obtain. In
either case, often the best source for data on resource requirements and envi-
ronmental releases is the processors themselves. For data on recycling rates
and recycled material, consumers and processors may be helpful, but trade