Comprehensiveness, "basic data" and "individual data"

IDEA contains LCI datasets of non-manufacturing sectors (agriculture, forestry and fisheries, mining, construction and civil engineering) as well as manufacturing sectors (food and beverage, textile, chemical industry, ceramics and building materials, metal and machinery) and also sectors such as electricity, gas, water and sewerage. It covers all products that are classified within the scope of the Japan Standard Commodity Classification, so the comprehensiveness is guaranteed. This means that a situation where some background data is missing will never occur. On the other hand, there are some datasets for service and processing (which are not products), where the comprehensiveness is not guaranteed.

Datasets that ensure comprehensiveness are defined as "basic data". "Basic data" is mainly created utilizing statistical data, and reflects the average production process of Japan. Further, datasets created to meet high needs, but for which the comprehensiveness is not guaranteed, is defined as "individual data". "Individual data" is created from interviews, surveys and utilizes past literature or existing data, and has a more limited applicability compared to "basic data".

IDEA classification code

In order to develop a comprehensive inventory database, a product classification that defines names and target scopes was developed in IDEA. This classification code is called "IDEA classification code". "IDEA classification code" is created by referring to "Japan Standard Industrial Classification" (Ministry of Internal Affairs, 2002) and other statistics. There are four types of classification codes, "Major groups <2 digits>", "Groups <3 digits>", "Details <4 digits>" and "Subclass <6 digits>". The "Japan Standard Industrial Classification" covers all economic activities conducted by businesses in Japan, and is very useful in constructing a classification for a comprehensive database such as IDEA. Correspondence between classification numbers of "Japan Standard Industrial Classifications" and IDEA classification code are shown bellow.

In IDEA, "Major groups <2 digits>" and "Groups <3 digits>" are "industry classification" and unit process data does not exist. The most detailed classification in "Japan Standard Industrial Classification" is "Details", which describes "industries" and is displayed as a 4-digit code. However, 4 digits in the IDEA classification is a product classification, and is considered as a product family that represents an aggregation of all products that the industry produces.An example of the hierarchical structure of IDEA classification code (until "Details") is shown bellow.

In order to set a further detailed classification other than "Details <4 digits>", an IDEA individual "Subclass <6 digits>" is created under "Details <4 digits>", based on literature and statistics.

IDEA classification code "Subclass <6 digits>" represents "products". An example of the hierarchical IDEA classification structure (until the "Subclass" level) is shown bellow.

This industry classification method is useful when similar products are produced in multiple industries. Each product generated as the output product flow of a unit process dataset is assigned a unique code number . This code is based on the IDEA classification code and is used to manage products.


The geographical scope of IDEA is set to the whole of Japan. The temporal scope (database reference year) is 2010. However, it is difficult to create all inventory data perfectly matching the predefined conditions, and in such case, processes were modeled using the most recent data available. Some datasets are from the 1990s, and this means that the quality of the temporal scope is judged as poor when assessing the inventory data quality (temporal data quality is ranked according to difference from the reference year). The technical scope is the current average technology used in facilities existing in Japan, and is basically Gate to Gate.

Data quality

IDEA raw data is collected from a wide variety of sources, and the quality varies. For example, some datasets have their quality assured such as those created by industry associations, while some data partly use assumptions or approximations to cover the lack of input and output inventory data. Performing LCA using background datasets without a thorough understanding of its quality may have significant impacts on the results. Therefore, quality evaluation of background datasets is important. Also, both "basic data" datasets and "individual data" datasets coexist in IDEA, and sometimes there are situations where multiple datasets represent the same product. If there are multiple datasets that represent the same product, the LCA practitioner will have to judge which dataset is appropriate, and in those circumstances, quality evaluation of background datasets will play an important role in assisting the LCA practitioner to adopt a higher-quality dataset.

Copyright (C) AIST All Rights Reserved.