1. 1 Uncertainty in EIO-LCA / Hybrid LCA Models

2. 2 Admin Issues HW #3: Acting TA is Paulina Jaramillo –pjaramil@andrew.cmu.edu (CEE)pjaramil@andrew.cmu.edu –Send email to “both of us” with questions, I will reply as soon as I can. 3 Guest lecturers - their specialty is LCA –Quick intros and what they’ll talk about –Hybrid LCA examples, uncertainty Web page updated.

3. 3 Estimating Direct Environmental Effects in EIO- LCA

4. 4 Economic and Environmental Implications of Online Retailing and Centralized Stock Keeping in the United States

5. 5 E-Commerce Retail Quarterly Volume ($B)

6. 6 Traditional Retail Logistics System Factory to warehouse to warehouse to retailer. Last leg of trip by private vehicle

7. 7 Single Facility Sales LL Bean, Lands End - catalogue sales Amazon (original), MusicOutpost - web based sales from a single facility

8. 8

9. 9 How to Model E-Commerce for LCA?

10. 10 Book Publishing Case Study Traditional System: –logistics: printer > warehouse > warehouse > retailer > home, all by truck/car –unsold returns - roughly 35% for bestsellers E-commerce System: –logistics: printer > warehouse > distribution center >home, by air and truck. –No unsold returns

11. 11 Comparative Analysis: * is EIOLCA Sector Use Traditional: –truck transport (1000 mi)* –Warehousing* –production of returns* –reverse travel of returns* –private automobile transport E-Commerce –air transport (500 mi)* –truck transport (500 mi)* –Warehousing*

12. 12 Comparative Costs ($ 1000s for $ 1 M or 290,000 books)

13. 13 Why are E-Commerce Costs Lower? Higher transportation costs for e- commerce, but: –Returns of unsold copies –Lower retail transactions costs –Lower (private) automobile cost Result is cost advantage for e-Commerce

14. 14

15. 15 Summary Environmental Impacts (per-book basis)

16. 16 Sensitivity Analysis ‘Traditional’ becomes better if: –Local distance to bookstore < 3 miles –Air transport of books > 700 miles –Orders not shipped together Ecommerce better if: –Switch from Air transport –Multiple origin sites –Greater density of sales.

17. 17 Harry Potter Case 250,000 books shipped on release date by Amazon.com –9,000 trucks and 100 airplanes 2.5 lb. book, 0.7 lb. packaging (3.2 lbs.) –Bookstores got 10 per box Shopping trips for books avg. 11 miles –Marginal effects

18. 18 This is Research….

19. 19 Some Analysis Issues What are E-commerce future scenarios? What will happen with local manufacturing technology? What will be impact of new business models for controlling inventory (warehousing), manufacturing and shipping. What is appropriate time scale of analysis?

20. 20 Analysis Boundary Issues (cont.) Buildings - decrease in retail or warehouse space? Shopping - will individuals substitute other travel for reduced shopping travel? Computers - what fraction of personal computer burdens should be allocated to E-commerce?

21. 21 Will E-commerce Improve or Degrade the Environment? Net Effect - hypothesis: depends upon product and processes and upon the analysis boundary. Appropriate Public Policy - –Don’t ignore service industries in environmental policy. –Consider life cycle costs including social costs. –Take advantage of cost savings to create environmental benefits

22. 22 Book LCA Spreadsheet - Advanced EIO-LCA modeling

23. 23 Intro: Uncertainty in LCA Uncertainty exists for all LCA data: mass flows, emissions, impacts, weights and change effects, e.g. –Proprietary data problems –Boundary problems: Lenzen (2000, Journal Industrial Ecology) finds truncation errors on the order of 50% for Australian LCA. Similar to Hocking result. –Measurement, transfer, change, etc.

24. 24 Uncertainty Implications Consistency and reproducibility of results (e.g. are paper or plastic cups superior). Certainty of conclusions and usefulness of LCA. Uncertainty for LCA studies in general obvious - we will focus on EIO-LCA –Data problems, combining data problems, allocation.. Numbers of significant digits – how many digits appropriate for www.eiolca.net result?www.eiolca.net

25. 25 EIO-LCA Uncertainty Sources Survey Errors: sampling and reporting errors – depends on companies and census agencies. Old Data: IO tables are typically 2 to 7 years old. Last US benchmark: 1997 released 12/2002. Incomplete Data: reports from only some sectors or plants (e.g. tri sector and threshold limits, holes in census surveys). Note: similarity to boundary problem in conventional LCA!

26. 26 Uncertainty (Pacca 2003)

27. 27 EIO-LCA Uncertainty (cont) Missing data: Census data missing many topics, such as habitat destruction. Non- monetary inter-sector dependencies also not represented, e.g. congestion effects from truck services. Aggregation: Sectors too large for detailed analysis on specific products. Problem sectors: ?

28. 28 Uncertainty (cont.) Imports: EIO treats imports as similar to domestic production. Model form: Linearity of EIO, lack of substitution as scale economies change. Mapping and Allocation Problems Producer Prices

29. 29 Parameter Stability Over Time Requirements matrix relatively stable over time: –Using 1961 final demand from IO tables of 1939 to 1961 found similar intermediate outputs (Carter, 1970). –Intermediate use relatively constant (Ma, 2003) Environmental impact vectors more dynamic.

30. 30 Intermediate Use 1972-1997 Total Intermediate Use Percentages 1972198719921997 Manufacturing49%41%39%35% Natural Resources11%9%8%9% Trade16%19%20%27% Services22%29%31%28% Miscellaneous2% Infrastructure22%24%25%26%

31. 31 More and Better Data Mixed picture for more and better data. No water use data since 1980s in US. No workfiles for 1997 benchmark released. Better industrial environmental management systems to collect data. More international co-operation and public data – international tri.

32. 32 User Adjustments Many adjustments possible due to known aggregation or emissions problems –Hybrid models including EIO and process models. –Parameter adjustments to reflect non-linearities. –Disaggregating individual EIO sectors. Bayesian methods applicable here – adjusting estimates based on expectations. Multiple approaches: EIO-LCA and Conventional LCA.