工業界做生命週期評估的清冊目的大多為1)為產品或製程做熱點分析，作為改進產品的製程的依據 2)提供下游廠商或供應商做為資料庫。目前的工業製程大多複雜並傾向於多產品的產出，因此在進行生命週期清冊建立時(LCI)如何將流程在不影響結果的情況下簡化，以及選定適當的方法分配環境負荷量到各產品上更顯其重要性。本研究所使用之環境負荷量的分配方法有Partitioning 和 System expansion 方法。簡化流程之合併模式分為”branch”，”straight” 以及 “complete”三大類，章節內皆有其定義。
本研究以ㄧ貫煉鋼廠為例子來討論不同分配方法以及合併情況對於生命週期評估結果之再使用性(reusability)與可比較性(comparability)的影響。結果顯示使用System expansion來分配副產品環境負荷量，較能反映同產品在不同地點或使用不同流程生產，其環境衝擊的變化量。
LCA結果顯示，平均每生產一公噸鋼鐵，其二氧化碳排放量約為0.7330公噸。並透過One-at-a-time (OAT)方法，針對所有參數做敏感度分析，得知其對最後LCA結果之影響幅度變化。
但LCA結果只是單單顯示每生產一單位鋼鐵所造成之二氧化碳當量，並無法表現出，當工廠為了配合營運之條件，而實際增加或者減少產量時，溫室氣體排放量之實際改變。本研究設定四種情境分析來探討模擬當減少一噸熱軋鋼鐵產量時，溫室氣體排放量之改變。結果顯示，某些情境下，當減少熱軋鋼鐵產量時，溫室氣體排放量不減反增，其主要原因為，若鋼鐵廠為了保持鍋爐發電系統之蒸氣產量，而添加燃油去補充因鋼鐵減產所短缺之熱量，其因添加燃油所產生之溫室氣體排放量，有時大於減少原物料使用量所節省之溫室氣體排放量。因此直接以stand-alone LCA數值計算溫室氣體減量，有時並無法真的反應實際的減量效果。
當鋼鐵產業(有許多副產品，並其製程與其他工業有互相供需關係)或者發電程序(同ㄧ產品卻有許多種製造方法)在進行LCA研究時，以下兩點須謹慎考量:
1)應配合LCA研究的目的來明確區分，是以利用熱點分析結果來提供工業選擇，或是以工業真實變化為目的，來確實推估結果之改變。當研究目的不同，其選擇之方式也隨之不同。
2)應提供生命週期清冊(LCI)結果和以真實變化為目的之生命週期評估結果，來提供下游廠商做為未來LCA研究之資料庫，以利於未來針對研究目的不同而進行不同之數值選擇。

The objective of construction of an inventory database in a company would be 1) to identify hot-spot for improvement of a product lifecycle, 2) environmental product declaration, and to provide their suppliers and downstream customers with a database for conducting Life-Cycle Assessment (LCA) of their product. Manufacturing procedure is usually complex and tends to have more than one product, therefore, it is not simple to construct an inventory database for LCA. Especially, methodological choices in allocation of environmental loads into multiple products are crucially important because it could affect the final decisions.
In this thesis, first, to what extent and how the flow model in LCA can be simplified, without influencing the result is discussed. There are two major known allocation approaches, namely partitioning and system expansion approaches in life cycle inventory analysis. Three aggregation patterns, namely “branch”, “straight”, and “complete” aggregation patterns were identified in this thesis, and the criteria to allow such aggregation is discussed in conjunction with the two allocation patterns.
Integrated steel manufacturing is taken as a case study to discuss in what cases data aggregation and partitioning could affect the reusability and comparability of LCA studies. And the result shows that it can reflect varied avoided impact by using system expansion for allocation. LCA result shows that emission factor of producing one ton steel is 0.7330 ton-CO2 equivalent and One-at-a-time (OAT) method is used for sensitivity analysis for all parameters in LCA model.
Usually, when the consequences of changes in used amount of steel in a product is evaluated using the amount calculated above. However, above value merely explains how much emission occurred in retrospect to production of one ton of steel, and does not explain how much change will be induced by reduction of or increase in one ton of steel consumption in prospect. To elucidate such difference, four scenarios are set up for a hypothetical integrated steelmaking process, discussing how much change in CO2 emission occurs by decreasing one ton of hot-rolled steel. Interestingly, the result shows that in some scenarios, reduction of use of hot-rolled steel could result in increase of CO2 emission. The case study clearly points out that:
1)Objectives of LCA must be clearly distinguished: whether it is supposed to be a study for identifying the alternatives via hot-spot identification, or it is a change-oriented study that actually evaluate the consequences of applying different alternatives
2)Database for downstream producers should provide both stand-alone LCA results, and change-oriented LCA results, to allow both stand-alone and comparative types of LCA.
Above two points are especially important for products such as steel (lots of co-products, relationships with other industries) and electric power (same product is produced from different production methods), because its application is extremely wide.

[1] Henrikke Baumann & Anne-Marie Tillman, The Hitch Hiker’s Guide to LCA, Studentlitteratur AB, Lund, Sweden (2004)
[2] CHINA STEEL CORPORATION website (2008) http://www.csc.com.tw/index.asp
[3] Bureau of Energy, Ministry of Economic Affairs website (2006) http://www.moeaboe.gov.tw/
[4] Ecoinvent database from Simapro software
[5] CHINESE PETROLEUM CORPORATION website http://www.cpc.com.tw/big5/home/index.asp
[6] Hamby, D.M., A review of techniques for parameter sensitivity analysis of environmental models. Environmental Monitoring and Assessment, (1994), 32(2): P.135-154
[7] Bo Weidema, Avoiding Co-Product Allocation in Life-Cycle Assessment. Journal of Industrial Ecology, Summer (2000), Vol. 4 Issue 3, P.11-33
[8] ISO 14040, (1997), Environmental management - Life cycle assessment - Principles and framework
[9] Tillman, A.-M., Significance of decision making for LCA methodology. Keynote lecture at the 8th Annual Meeting of SETAC-Europe, Bordeaux, April 14-18 (1998), (Submitted for publication in Environmental Impact Assessment Review)
[10] Tillman, A.-M., H.baumann, E. Eriksson, and T. Rydberg. Life cycle analysis of packaging materials. Calculation of environmental load. Goteborg: Chalmers Industriteknik. (1991)
[11] Vigon, B. W., D. A. Tolle, B. W. Cornaby, H. C. Latham, C. L. Harrison, T. L. Boguski, R. G. Hunt, and J. D. Sellers. Life cycle assessment: Inventory guidelines and principles. Washington, DC: United States Environmental Protection Agency, Office of Research and Development (EPA/600/R-92/245). (1993)
[12] Weidema, B. P. New developments in the methodology for life cycle assessment. In Proceedings of the 3rd International Conference on Ecobalance, Tsukuba, November 25-27 (1998), P.47-50