IMPACTS OF WATER CONSUMPTION:

Transcription

1 IMPACTS OF WATER CONSUMPTION: UNCERTAINTIES AND THE QUESTION OF SCALE Stephan Pfister, ETH Zurich if h

2 Question of scale No water scarcity on global average Different impacts: Ecosystems Humans Stock/fund resources Especially relevant in agricultural LCA Highly variable inventories 2

3 Uncertainties Uncertainty Inventory Impact assessment Spatial Any LCA RegionalisedLCA Variability Technology If LCA value = f(location) Spatial uncertainty reflects variability Climate Improve by increasing precision Regional Temporal 3

4 Resolution issues Connect to inventory 4

5 Resolution issues Connect to inventory Maximize resolution 35% 65% Use weighted results for features 5

6 Water Impact Assessment (WIA) Midpoint and endpoint factors compatible with Eco indicator 99 (EI99) More than watersheds characterized (global coverage) Publicly available: Google Earth layer: Reference: Pfister, S.; Koehler, A.; Hellweg, S. Assessing the environmental impacts of freshwater consumption in LCA. Environ. Sci. Technol. 2009, 43 (11),

7 Aggregated g damage factors (based on EI99 average weighting) WATERSHED LEVEL 7

8 How to use the regionalized add on Impact factors in Google Earth Download KMZ file: at: h h/ ff/ fi / Impact_factors_LCA_pfister_et_al.kmz 8

9 Application 9

10 Cotton example Analysis based on country level (inventory and impact factors) EI99pt/kg 10

11 Variability of regional impacts Water impacts and inventors highly climate dependant Big countries ti > high h variations iti Example United States: Impact: EI99pt/m3 (~ factor of 1000; average: Inventory: infinite factor Resolution is not enough: Use highest possible resolution 11

12 Simulated irrigation water consumption Neglected: Yield = f(irrigation, temperature, fertilization, machinery, etc.) IWR [mm] 12

13 Impact assessment Damage per kg cotton Damage per m 3 water consumed EI99 pt / kg < >1.5 13

14 Production Statistical data Uncertainty Variability (annual) Calculate production weighted country average Mainly non-irrigated Production 14

15 Difference in results Cotton lint production: global & national average based on country data national average based on local analysis Water Ecosystem consumption Quality [m 3 ] [PDF m 2 yr] Human Health Resources [E 06 DALY] [MJ] Share of total Eco indicator damage Global Average US average data US local assessment % % % 15

16 Aggregate g as late as possible! High spatial resolution: Increased precision Decreased accuracy Combine on country level: Increased precision Increased accuracy 16

17 Conclusions Calculate impact factors on highest resolution Define inventory flows as locally as possible Provide link to inventories (e.g. country/global average) Aggregate factors as late as possible Generic, process specific data (supply chain modeling) Need for consistent uncertainty assessment Impact factors and inventories Trade off accuracy / precision Overalldecreased uncertainties 17

18 LCA IX, Boston, September

19 LCA IX, Boston, September

20 Example: US cotton production Inventory (water consumption) Impact factors Reference: Pfister, S.; Koehler, A.; Hellweg, S. Assessing the environmental impacts of freshwater consumption in LCA. Environ. Sci. Technol. 2009, 43 (11), > SUPP INFO 20

21 Example: US cotton production production per watershed total water impact per watershed Reference: Pfister, S.; Koehler, A.; Hellweg, S. Assessing the environmental impacts of freshwater consumption in LCA. Environ. Sci. Technol. 2009, 43 (11), > SUPP INFO

22 How to use our add-on Click on Information Points i in relevant watershed