THE ROLE OF TRANSPORTATION ON THE GHG EMISSIONS OF WINE

Transcription

1 THE ROLE OF TRANSPORTATION ON THE GHG EMISSIONS OF WINE C. Reich-Weiser 1,3, P. Paster 2, C. Erickson 1, D. Dornfeld 3 1 Climate Earth Inc., ClimateEarth.com 2 Ask Pablo, AskPablo.org 3 UC Berkeley, Laboratory for Manufacturing and Sustainability, lmas.com Abstract: This paper follows on earlier studies demonstrating the importance of transportation to the total greenhouse gas (GHG) footprint of wine and discusses the variability of transportation emissions depending on the specific type of transportation utilized. Variations in freight transportation emissions by truck, ship, train, and airplane are discussed along with how these values affect the lifecycle wine GHG emissions of wine. Ultimately, this work represents the opportunity for the wine industry to reduce their GHG emissions by influencing the GHG efficiency of their distribution networks, reducing packaging weights, and selecting the appropriate freight mode. It is also demonstrated that emissions tradeoffs may exist between transportation and regional GHG vineyard emissions. 1. Introduction Wine is an internationally traded commodity with continued growth across international markets because of the wine consumer s desire to enjoy regionally specific products. Green wines are becoming an increasingly desired commodity as evidenced by numerous popular articles on the subject [1, 2, 3, 4]. Based on existing studies, regional variations in wine production methods, varietals, rainfall, and fertilizer requirements influence the environmental impact of each bottle of wine. Additionally, the transportation of wine from the producer to the consumer can be a significant portion of the total greenhouse gas (GHG) emissions associated with a bottle of wine [5]. While production, packaging, and transportation related impacts are all worthy of additional exploration, this paper focuses only on transportation and explores the opportunity for transportation greenhouse gas (GHG) reductions through decisions on transportation mode. Specifically, we review previous work on regional GHG emissions variations and explore the variability in the transportation GHG emissions associated with wine. Regional variations in wine production GHG emissions are not explored here; however this paper will further the ability of future studies to make regional production versus transportation tradeoffs. As a case study, several possible transportation emissions scenarios are used to explore the emissions breakeven point of purchasing wine from Bordeaux versus purchasing wine from California. 2. Literature Review Previous research on the GHG emissions associated with wine production is divided on the importance of transportation emissions. Colman et al. [5] found that transportation is the most

2 significant part of a wine s lifecycle GHG emissions, given the large distances that many wines travel to their final destination. Whereas, Fearne et al. [6] found trellising and viticulture practices along with bottling and packaging to account for over half of the lifecycle GHG emissions. Additionally, Gonzalez et al. [7] found that wine production, glass production, and transportation had similar GHG emissions; and then concluded that because transportation depends on distance it is not in the wineries power to control and that improvements should focus on wine production. Ardente [10] studied Italian wine production and found packaging to be 50% of the footprint, followed by transportation, electricity, and fertilizers. Point [11] focused on the emissions in the vineyard and production of wine, but acknowledged that transportation to the consumer may be a large part of emissions. Gazula et al. [8] concluded that the viticulture and bottle making emissions are the most important contributors to lifecycle GHG emissions, but also said that transportation could be up to 30% of total emissions. The discrepancy in whether a researcher determines transportation is important or not depends on whether they have included wine distribution and how far they have assumed the wine must travel; however in the studies where transportation is included, and multi-country distances considered, transportation is generally considered an important component of wine greenhouse gas emissions. Previous researcher s results are summarized in Table 1, not including distribution to retail and/or to the consumer. There are multiple reasons for the variations seen in Table 1, including: regional differences in soil quality and fertilizer requirements; grape and wine sugar content variations; different boundaries set on the life-cycle assessment. TABLE 1: Wine GHG Emissions before Distribution Wine GHG Before Distribution (kg-co 2 eq) Bottle Volume (ml) Source [9] [10] [7] [11] [8] 3. Transportation GHG Overview Transportation GHG emissions are often given relative to the weight of the goods transported and the distance transported. While there are theoretical problems with this approach (such as it assumes a vehicle containing no load has no emissions), it provides a reasonable way to estimate emissions when all that is known is weight, mode of transportation, and distance. This method also assumes that packing efficiencies have been optimized so that weight is the driving factor rather than volume. Researchers of wine emissions have generally utilized a single GHG value to estimate transportation emissions relative to total LCA emissions. This is a reasonable way to come up with an estimate; however, trusted emission factors vary and use of these estimates for conclusive results is risky. Tables 2 through 5 demonstrate a range of emissions factor values depending on source. The variations seen in these values are likely from the assumptions of the study, regional variability

3 in fuels, and regional variability in vehicles. For comparison, the sources utilized here include: a PhD thesis on freight emissions [12]; the World Resources Institute [13]; the UK Department for Environment, Food and Rural Affairs [14]; EcoInvent (a European LCA database) [15]; and CE Delft [17]. TABLE 2: Air Freight Emissions (grouped by source) Transport Type Value (mg CO 2 /kg-km) Source Air Transportation - Direct 597 [12] Air Transportation - Total 870 [12] Short haul (<452 km) 570 [13] Medium haul (452 to 1600 km) 800 [13] Long haul (>1600 km) 1580 [13] Long-Haul Flights 610 [14] Short-Haul Flights 1320 [14] Domestic Flights 1900 [14] Transport, aircraft, freight, intercontinental, RER 1068 [15] Transport, aircraft, freight, RER 1100 [15] Transport, aircraft, freight, Europe, RER 1669 [15] TABLE 3: Water Freight Emissions (grouped by source)* Transport Type Value (mg CO 2 /kg-km) Source Large container vessel (20,000 tonnes) 13 [14] Small container vessel (2,500 tonnes) 15 [14] Transport, transoceanic freight ship, OCE 11 [15] Transport, liquefied natural gas, freight ship, OCE 52 [15] Container 15 [17] Refrigerated Cargo 77 [17] RoRo Cargo 59 [17] * Tankers and barges are not included here TABLE 4: Trucking Emissions (grouped by source) Transport Type Value (mg CO 2 /kg-km) Source Road Transportation - Direct 87 [12] Road Transportation - Total 118 [12] Road Freight 72 [13] All articulated - UK average 86 [14] ALL HGVs - UK average 132 [14] All rigid - UK average 276 [14] Transport, lorry >16t, fleet average, RER 126 [15] Transport, lorry t, fleet average, RER 334 [15]

4 TABLE 5: Train Emissions (grouped by source) Transport Type Value (mg CO 2 /kg-km) Source Rail Transportation - Direct 15 [12] Rail Transportation - Total 17 [12] Diesel Locomotive 20 [13] Electric Locomotive 26 [13] Coal Locomotive 40 [13] Rail Freight 21 [14] Transport, freight, rail, CH 15 [15] Transport, freight, rail, RER 40 [15] Transport, coal freight, rail, CN 44 [15] Transport, freight, rail, diesel, US 50 [15] 4. Transportation s Influence: Napa versus Bordeaux Case Study In the work by Coleman et al. [5] the authors presented the concept of the wine line, an imaginary line representing the greenhouse gas emission breakeven point between the transportation of wine from Napa and Bordeaux. In the working paper on the Journal of Wine Economists website this line passed through Ohio [16]. Due to an error in a cited report [17], the emission factor for shipping was adjusted, resulting in a revised line, now through Kansas, in the final published paper [5]. The concept of the wine line builds upon the work of Andrea Paxton, Tim Lang, and others on the concept of food miles [18, 19]. Rather than calculating the absolute food miles traveled by wine, the wine line takes into account the relative greenhouse gas emissions not only from distance, but also based on weight and transportation mode. Here, we take into account multiple possible emissions factors for transportation to calculate a range of possible wine line alternatives. 4a. Breakeven Distance between Napa and Bordeaux To assess the influence of transportation, an emissions breakeven point where transportation emissions from Bordeaux exceed emissions from Napa can be calculated using the values from Tables 2 through 5. Equations 1 through 4 setup the breakeven calculation, where d is distance, f is a GHG factor from Tables 2 through 5, and w is the weight of a full bottle of wine. The goal is to solve for the breakeven distance from Napa (d Napa.to.breakeven ) across the range of possible transportation emissions values. Note that in reality, the breakeven between Napa and Bordeaux cannot be a straight line, because road distances vary; however the lines calculated in this section are simply intended to demonstrate the variability in results depending on chosen transportation mode and therefore the power distributors have to affect this breakeven point through smart transportation decision making. This study is intended to show the variability in GHG emissions depending on transportation choices made by wine distributors. Therefore, the following simplifying assumptions are made: 1) Emissions up until the point of distribution are the same for both wines 2) Packaging weight is negligible 3) Packaging efficiencies are maximized

5 4) The distance from Bordeaux to New York is 3170 nautical miles (based on shipping distances provided by searates.com [20]). 5) The distance from Napa to New York is 3000 miles (based on google.com driving distances [21]). 6) To get to the breakeven-point between Napa and Bordeaux, wine must travel through New York. The breakeven point between New York and Bordeaux can be calculated using the following equations (and as illustrated in Figure 1): d Bordeaux.to.NY 3170 nautical miles 5870 km (1) d Napa.to.NY 3000 miles 4800 km (2) 1) If the breakeven point is west of New York (breakeven < d Napa.to.NY ) A B C d Napa.to.breakeven (f truck.napa ) = d NY.to.breakeven (f truck.ny ) + d Bordeaux.to.NY (f ship.bordeaux ) (3) d Napa.to.breakeven = d Napa.to.NY - d NY.to.breakeven (4) 2) If the breakeven point is east of New York (breakeven > d Napa.to.NY ): A B C d Napa.to.NY (f truck.napa ) + d NY.to.Breakeven (f ship.ny ) = d Bordeaux.to.Breakeven (f ship.bordeaux ) (5) d NY.to.breakeven = d Bordeaux.to.NY d Bordeaux.to.breakeven (6) d Napa.to.Breakeven d Napa.to.NY + d NY.to.breakeven (7) FIGURE 1: Illustration of breakeven calculation segments as used in equations 3 and 5.

6 Assuming the same type of truck is used for wine coming from Napa and wine coming from New York, and that the bottles are the same weight regardless of origin, equations 1-7 simplify to 1) If the breakeven is west of New York (breakeven < d Napa.to.NY ) d Napa.to.breakeven [km] ½ (d Napa.to.NY + d Bordeaux.to.NY (f ship / f truck )) (8) d Napa.to.breakeven [km] (f ship / f truck )) (9) 2) If the breakeven is east of New York: If breakeven > d Napa.to.NY (i.e. breakeven is east of NY) d Napa.to.breakeven [km] d Napa.to.NY + ½ (d Bordeaux.to.NY d Napa.to.NY (f truck /f ship )) (10) d Napa.to.breakeven [km] (f truck /f ship ) (11) These equations are then used to determine the breakeven line for every possible combination of water freight and trucking emissions from Tables 3 and 4. Results are shown in Figure 2. Table 6 demonstrates three example combinations of trucking and shipping emission combinations along with their breakeven distance. Note that the western-most line on Figure 2 is approximately equivalent to the wine-line originally presented by Coleman et al. [5]. Alternatively, breakeven lines can be approximated by assuming that rail freight is used on land instead of trucking. In this case, the breakeven results are shown in Figure 3. Rail emissions are much lower than trucking emissions, therefore the breakeven lines move to the east in Figure 3. Table 7 demonstrates three example combinations of train and shipping emission combinations along with their breakeven distance These results indicate that the breakeven point for wine from Bordeaux versus wine from Napa can be anywhere from midway across the United States to somewhere in the Atlantic Ocean. Note that, in reality, the breakeven between Napa and Bordeaux cannot be a straight line because road distances vary; however Figures 1 and 2 are sufficient to demonstrate variability. TABLE 6: Example Breakeven Lines Shown in Figure 2. f ship f truck d Napa.to.breakeven [km] [mg-co 2 /kg-km] [mg-co 2 /kg-km] A ,497 B ,491 C ,761 TABLE 7: Example Breakeven Lines Shown in Figure 3. f ship f train d Napa.to.breakeven [km] [mg-co 2 /kg-km] [mg-co 2 /kg-km] A ,046 B ,268 C ,273

7 FIGURE 2: Trucking vs. Water Freight Breakeven Lines. FIGURE 3: Rail vs. Water Freight Breakeven Lines

8 4b. Transportation emissions relative to total emissions It is useful to determine the total one-way emissions per bottle attributable to transportation so this can be compared with previous assessments of emissions up until distribution (Table 1). The Liquor Control Board of Ontario [22] reports that a 750 ml wine bottle weighs anywhere from 270 to 800 grams, with a mean weight of 510 grams and a median weight of 500 grams. Additionally, the weight of the wine is roughly 750 grams (assuming it is similar to water). Therefore, the total transported weight per bottle is between 1020 and 1550 grams (median of 1250 grams). Given this range of possible wine values, the transportation emissions may range from 0.21 to 1.6 kg-co 2 eq for trucking and water freight; or 0.07 to 0.54 kg-co 2 eq for rail and water freight. Given that wine emits somewhere around 1 to 2.5 kg-co 2 eq per 750 ml bottle up until the point of transportation (Table 1), the added emissions from transportation may or may not be significant depending on the weight of the bottle and the choice of transportation (rail and water freight favored over truck). The relative similarity between transportation emissions and the variations in wine manufacturing emissions indicates that in certain cases it may be wise to consider the tradeoffs possible between the GHG emissions associated with production of wine at a certain site and the transportation of that wine to the consumer. For example, in particularly poor climates or regions with poor soil conditions, the additional energy and fertilizer requirements of local wine production may mean that imported wine from a more favorable region has lower total GHG emissions. This area will require further research to understand how to minimize global wine GHG emissions. Conclusions: 1. This work demonstrates a methodology to quantify the GHG impact of shipping and distribution systems for any commodity produced and consumed worldwide, and offers some clarification on how to more reliably determine the impacts given uncertainty in transportation mode. 2. Emissions values for transportation vary depending on the type of transportation and the methodology used to calculate the emissions factor. While the most accurate way to measure emissions is to examine a known quantity of fuel type combusted, researchers often need to use estimated emissions per tonne-km. Researchers should be aware of the variability, potential error, and uncertainty behind these calculations before coming to over-simplified conclusions. 3. Assuming the same bottle of wine, with the same emissions up until the point of distribution, the breakeven point for Napa versus Bordeaux wine is somewhere between the middle of the United States and the middle of the Atlantic Ocean. In particular, the decision to transport wine by train or by truck has a large influence on the breakeven point because train emissions are generally lower than trucking emissions. 4. Transportation is an important part of wine GHG emissions; therefore, careful selection of efficient modes of transportation for wine transportation is an effective method to reduce GHG emissions from the global wine trade.

9 5. Reduction of weight by using innovative new packaging materials (TetraPak, PET wine bottles, bag-in-box, etc.), lighter glass [23], or by shipping in bulk and packaging nearer to the consumer [9] reduces transportation emissions; although tradeoffs relative to packaging manufacturing emissions and regional differences in wine production emissions must be considered as well. 6. There is the potential for future work to assess if it makes sense to consider the tradeoffs possible between the GHG emissions associated with production of wine at a certain site and the transportation of that wine to the consumer. This is possible given that the variations in wine emissions shown in Table 1 are of the same magnitude as transportation emissions; and could mean that in some cases the lowest GHG wine alternative is non-local wine. For example, imagine that wine from location A has emissions of 1.2 kg-co 2 eq/bottle and wine from Location B has emissions of 0.8 kg- CO 2 eq/bottle. Furthermore, consider that the emissions to transport wine from location A to location B are 0.1 kg-co 2 eq/bottle. Then, a consumer in location A has lower total emissions by purchasing wine from location B rather than local wine. Acknowledgments: This work is supported by the industrial affiliates of the Laboratory for Manufacturing and Sustainability (LMAS) and Climate Earth. Thank you to the reviewers for their excellent comments and insights. References: [1] Arnold, E. (2009) Green wine: eight environmentally friendly wines worth trying Forbes Magazine Online, accessed online: < [2] Asimov, E. (2007) When the Wine is Green The New York Times, accessed online: < [3] Root, J. (2008) The World s 5 Most Wicked Green Wineries. Tree Hugger, accessed online: < [4] O Rourke, C. (2009) Italian Vintner Aims for Carbon Neutrality. New York Times blog, accessed online: < [5] Colman, T. and P. Paster (2009) Red, White, and Green : The cost of greenhouse gas emissions in the global wine trade. Journal of Wine Research, 20(1), pp [6] Fearne, A., C. Soosay, R. Stringer, W. Umberger, B. Dent, C. Camilleri, D. Henderson, A. Mugford (2009) Sustainable value chain analysis: a case study of South Australian wine. Government of South Australia

10 [7] Gonzalez, A., A. Klimchuk and M Martin (2006) Life Cycle Assessment of Wine Production Process: finding relevant process efficiency and comparison to eco-wine production. [8] Gazulla C, Raugei M, Fullana P (2010) Taking a life cycle look at crianza wine production in Spain: where are the bottlenecks? International Journal of Life Cycle Assessment, 15, pp [9] Paster, P. (2009) Wine In A Box: The Greenest Way To Drink? accessed online: < [10] Ardente, F., G. Beccali, M. Cellura, A. Marvuglia (2006) POEMS: A Case Study of an Italian Wine-Producing Firm. Environmental Management, 38(3), pp [11] Point, E. (2008) Life Cycle Environmental Impacts of Wine Production and Consumption in Nova Scotia, Canada. Dalhousie University, Halifax, Nova Scotia, Master of Environmental Studies Thesis. [12] Facanha C. (2006) Life-cycle air emissions inventory of freight transportation in the United States. Ph.D. thesis, University of California at Berkeley. [13] World Resources Institute (2004) The greenhouse gas protocol initiative. World Resources Institute and World Business Council for Sustainable Development, accessed online: < [14] Defra (2008) 2008 Guidelines to Defra s GHG Conversion Factors: Methodology Paper for Transport Emission Factors. Department for Environment, Food and Rural Affairs, < [15] Ecoinvent v2.1, Product Ecology Consultants, Swiss Center for LCA Inventories, [16] Colman, T., P. Paster (2007) Red, white and green : the cost of carbon in the global wine trade. American Associates of Wine Economists, AAWE Working Paper, No.9, accessed online: < [17] CE Delft (2006) Greenhouse gas emissions from shipping and implementation guidance for the marine fuel sulphur directive. Publication number: [18] Waye, V. (2008). Carbon Footprints, Food Miles and the Australian Wine Industry. Melbourne Journal of International Law, 9, [19] Weber, C., H. Matthews (2008) Food-Miles and the Relative Climate Impacts of Food Choices in the United States. Environmental Science & Technology, 42(10), pp [20] SeaRates.com Global Shipping Platform, July 27, 2009, accessed online: <searates.com> [21] Google Maps, July 27, 2009, accessed online: <maps.google.com> [22] Ford, P. (2009) Wine Council of Ontario Environmental Seminar, Pillitteri Estates WineryPatrick Ford, Senior Director, Policy & Government Relations, LCBO, July 28. [23] Fetzer (2008) Fetzer lightweight bottle carbon savings analysis. accessed online: <