Assessing water use in LCA and related environmental assessments

Special session coordinator: Annette Koehler, ETH Zurich

Note: presentations have not been updated match those given at the meeting

The topic of water use and depletion of freshwater resources is rapidly gaining momentum and nowadays is perceived as important as climate change. While water use has been widely disregarded in LCA in the past, new methodological approaches are being developed both for inventory modeling of water use and for impact assessment describing the impact pathways up to different areas of protection. Here different types of freshwater resources and various environmental mechanisms are being considered. Taking the rapid evolvement of this topic into account, establishing some central guidance for water-use reporting and assessment is recognized as major challenge to be taken already in this early phase. This session welcomes presentations on method development for water-use related LCI schemes, complex LCIA metrics, and simplified assessment approaches for water footprinting. Aspects of spatial differentiation and data availability as well as applications to specific case studies will be discussed. The purpose of this session is to gather researchers and practitioners from business and industry dealing with environmental assessment schemes of freshwater use and depletion.

Session Chairs:

Annette Koehler, ETH Zurich
Emmanuelle Aoustin, Veolia
Cécile Bulle, CIRAIG

Presenters:

Regionalised Assessment of Fresh Water Use in the Swiss Ecological Scarcity Method 2006
Rolf Frischknecht, ESU-services Ltd.
Arthur Braunschweig, E2 management consulting
Norbert Egli, FOEN
Gabi Hildesheimer, oebu
presentation

The Swiss Ecological Scarcity method has first been introduced in 1990 and updated in 1997. The Swiss version of this method was updated and extended in 2006. The update and extension of the method takes into account the recent developments in Swiss and European (as far as it is relevant for Switzerland) legislation and environmental targets.

Furthermore, ISO standard revisions and recent developments in scientific knowledge on environmental effects are also considered where appropriate. The basic principle and main strength of the method, measuring the environmental scarcity with the help of actual pollutants (and resources) flows and maximum allowed (so-called critical) flows, remained untouched. Hence, it is still a distance to political target rather than a damage oriented impact assessment method.

The method allows for a regional assessment of the emission of pollutants and the consumption of resources based on regional scarcities. Being a scarce resource in many areas of the world, regional eco-factors of fresh water consumption were developed. The assessment makes use of the water stress indicator developed by the UN and OECD. It is proposed classify countries (or regions) into six water stress levels from low (using less than 10% of the available fresh water resources) to extreme (using more than 100% of the available fresh water resources).

Several examples show that fresh water consumption matters when assessing water intensive products (e.g. agricultural products). It is of much less relevance for industrial products, even if produced in regions with a moderate water stress.

Operational Characterization Method for Water Use: Case Study Application in the Pulp and Paper Industry
Anne-Marie Boulay, CIRAIG
presentation

Although freshwater resource issues are continuously arising, Life Cycle Assessment (LCA) models are not yet adapted to address this environmental problem. Building on the framework proposed by the project «Assessment of Freshwater Use and Consumption within LCA » of the UNEP/SETAC Life Cycle Initiative, this paper aims to provide an operational method and to illustrate its results using a case study in the pulp and paper industry.

The method presented here introduces a new midpoint impact category Freshwater Deprivation for Human Uses, quantifying the volume of freshwater rendered unavailable because of its use as a function of the type of use (ex: degradative or consumptive), regional water scarcity, water quality, type of resource used (groundwater, surface water) and ability to adapt to freshwater scarcity. The calculated characterization factors provide a country level regional assessment, translating a volume of water used into a drinking water equivalent volume unavailable for human uses (ex: Canada:0; South Africa:0,54; Jordan:3,88).

It was considered that fresh water deprivation does not occur in area where economic resources allow the use of backup technologies, i.e. compensation scenarios. The GDP was chosen to act as an indicator of the adaptation capacity of downstream users to water shortage. For cases where water deprivation does not occur, default backup technologies are identified for several countries considered able to compensate water uses. A case study on the pulp and paper industry modeled and combined impacts from these backup technologies with those allocated to the new impact category “Water deprivation for human uses” resulting in all impacts from freshwater use being included in the LCA results. In addition, data from the pulp and paper industry was used to model the avoided impacts as a consequence of the improvement of the water use efficiency in a plant. These impacts are then compared with the impacts directly avoided from a reduction in water consumption, notably the change in power needed. These results may therefore provide guidance to a decision maker in focusing its environmental efforts.

Health Damage Assessment Modelling on Agricultural Water Scarcity based on Regression Analysis of Statistical Data
Masaharu Motoshita, AIST
presentation

Population growth and subsequently increasing food demand will collapse the balance of water demand in the world. Thus, high concerns should be paid on the environmental issues related to water consumption and deficiency. In the past studies on environmental problems related to water, qualitative assessment has been mainly conducted and water scarcity has not been given enough importance for the assessment of environmental impacts. Particularly, the shortage of agricultural water may lead to prevention of crop productivity and population growth shall cause multiplier effect through the increase of food demand. The aim of this study was to model the cause-effect chain of undernourishment damage due to agricultural water scarcity and estimate damage factors of each country.

In this study, health damage assessment modeling related to agricultural water scarcity was conducted by applying regression analysis based on statistical data. The cause effect chain of undernourishment damages due to agricultural water scarcity was simply assumed to be composed of two steps, food productivity loss caused by the shortage of agricultural water and subsequent increases of undernourishment damage due to insufficient nutrient conditions in the agricultural water scarce country. At first, food productivity losses in each country caused by the shortage of agricultural water were estimated based on a crop productivity model. Secondly, the estimation model of undernourishment damage was conducted by applying non-linear multiple regression analysis based on several statistical data in each country. Explanatory variables, several social and nutritional indices of each country, seemed to affect on undernourishment damages were selected based on the results of multiple-regression analysis and statistical significant test.

Based on several simplified assumptions in the description model of cause-effect chain, prototype of the assessment model on undernourishment damage due to agricultural water scarcity had developed and damage factors for each country could be preliminary calculated. The estimated average factor weighted by agricultural use of water in each country was 1.44*10^-8 [DALY/m3]. Especially, African areas showed high sensitivity to agricultural water scarcity.

Review of methods addressing water in life cycle assessment
Anna Kounina, EPFL
presentation

Stress on global water resources is recognized as an important issue. Although the impacts related to water use on human life, biotic and abiotic environment can be substantial, such impacts are yet still poorly assessed in Life Cycle Assessments (LCA). Currently, most LCA studies consider water use as simple water inventory and only few methods are available to evaluate the ecological consequences. Impacts related to water use are not directly proportional to the consumed or degraded water amount. Among other parameters, the severity of impacts can depend on the type of water resource used or on water scarcity in the location under study.

This project aims at performing a systematic qualitative review of existing methods (inventory methods, scarcity indexes, and midpoint and endpoint assessment methods) linked to water assessment within a LCA framework. This review is split in different stages: the methods are first briefly described, classified, characterized and finally evaluated. The project will provide as output: (1) key elements to be considered when modelling the cause effect chains in water assessment and provide indications for deriving operational characterization methods and factors to assess water use in LCA to support researchers, (2) interim recommendations on inventory modelling, scarcity indexes, midpoint and impact assessment methods to support practitioners in their short term application.

So far, the following methods and indexes are presented and characterized (further methods might be added to the complete evaluation):

Inventory methods: Classification based on ecoinvent (Frischknecht 2005), Global water tool (WBCSD 2009), Boulay et al. (2009), Water footprint method (Chapagain and Hoekstra 2004), Vince et al. (2007), Bauer et al. (2007).

Scarcity indexes: Gleick et al. (1996, Basic water requirement), Falkenmark et al. (1989, Water resource per capita), Ohlsson et al. (1998, 1999, Social water stress index), Alcamo et al. (1997, 2000, Criticality ratio and index), Seckler et al. (1998, Index of relative water scarcity), Smakhtin et al. (2004, Environmental water requirement), Pfister et al. (2009, Water stress index), Raskin et al. (1997, Water resources vulnerability index), Sullivan et al. (2003, Water poverty index).

Impact method, at midpoint: Water footprint (Chapagain and Hoekstra 2004), Swiss ecological scarcity method (Frischknecht et al. 2006, 2008), Pfister et al. (2009), Milà i Canals et al. (2008), Boulay and Bayart et al. (2008).

Impact method, at endpoint (or damage) : Pfister et al. (2009), Motoshita et al. (2008), Boulay et al. (2009), Maendly and Humbert (2009), Van Zelm et al. (2009), CExD (Bösch et al. 2007).

These methods are assessed according to defined criteria elaborated based on the project “Recommendation of methods for LCIA” for the International Reference Life Cycle Data System (ILCD), initiated by the European Platform on Life Cycle Assessment (EUPLCA), which builds on the previous work of the Life Cycle Impact Assessment (LCIA) programme of the UNEP-SETAC Life Cycle Initiative (Task force 1).

The scientific criteria address the following areas: 1. Completeness of scope; 2. Environmental relevance; 3. Scientific robustness and certainty; 4. Documentation, transparency and reproducibility; 5. Applicability.

Each of these criteria is further detailed into a number of sub criteria adapted to the specific category (inventory, index or impact), reflecting their most important aspects and focusing on aspects that can support differentiation between the different methods.

In order to systematize the implementation and the interpretation of the water-use assessment in LCA, it is recommended to reach a clear consensus on a simple and comprehensive approach to assess this important issue.

Water Use Impacts from Corn-based Bioethanol Production
Yi-Wen Chiu, University of Minnesota
Annette Koehler, ETH Zurich
presentation

Water is found as a critical requirement in most of the industrial processes and is recognized as an important environmental resource in LCA studies and LCA database. However, only few LCA studies have been conducted so far which consider water input and output information throughout the entire studied product system. The lack of water inputs and outputs in LCA results to neglect impacts induced by water use. With the increasing awareness of the importance of water sustainability, there is an urgent need to integrate water use impact with the existing LCA structure. Therefore, for the proposed special session, we demonstrate the application of a novel water use impact assessment model (WIAM) which incorporates the potentially disappeared fraction of species (PDF) as assessment indicator.

In the past five years, the conflict between water and alternative fuels has attracted numerous research communities to take efforts for determining ethanol’s water implication. Using Minnesota’s corn-based bioethanol production as a study case, we compute the ecosystem damage factor for the 81 watersheds in Minnesota considering each watershed’s hydrological regime. We assumed that ethanol facilities acquire corn for ethanol manufacture from nearby corn fields in order to satisfy the transportation cost efficiency. A cradle-to-gate system has been set-up to take both irrigation water and process water of the ethanol manufacturing into account. The impacts associated with the 22 studied ethanol facilities are assessed by employing WIAM. Our study further shows (a) how water consumption of a product system can be further translated into ecological impacts by employing WIAM, (b) why it is necessary to take adjacent watersheds into account for future ethanol-facility site selection, and (c) what significant information obtained from this approach can support the decision makers in water conservation.

Assessing Water Impacts of Tea and Margarine with a Water Footprint / LCA Approach. Pilot study in Unilever
Llorenç Mila i Canals, Unilever, UK
presentation

Unilever seeks to manage the impact of its water use across the life cycle of a product.

Current methods to measure water use and impact are poor but there is both an increasing internal need for guidance on the impacts of water use and an increasing external level of activity to develop standards for water impact assessment. Water footprinting (WF) is one potential tool to understand water use, and the Water Footprint Network (WFN) provides the forum to work with others to develop the methodology for use with products. In this sense, Unilever is working with other WFN partners (university of Twente) to test the application of the WF method to raw materials and processes of two products (tea and margarine), with explicit distinction of water sources (region and water type). In addition to the accounting (inventory) activity, indicators for impacts related to water stress / scarcity are being developed and applied, at the level of detail where a compromise is reached between practicality and environmental relevance. This pilot builds from the experience within the team in both LCA and WF areas, in order to explore advantages and shortcomings of each tool and the synergies between them.

Results for the two case studies will be presented, including water volumes accounting and preliminary results for the impact assessment. The discussion will focus on how the WF and the LCA approaches can build from each other, and at what level they need to be kept separate in order to make the most out of each.

Direct and Indirect Water Withdrawals for US Industrial Sectors
Michael Blackhurst, Carnegie Mellon University
presentation

Effective water management is critical for social welfare and ecosystem health. Nevertheless, information necessary to meaningfully assess sustainable water use is incomplete. Using publicly available data, we estimate a vector of water withdrawals for all 428 sectors in the 2002 US economic input-output table. The vector was applied using economic input-output life cycle assessment (EIO-LCA) techniques to estimate direct and indirect water withdrawals for each sector’s production, both in terms of total water use and per dollar of economic output. We estimate that the US economy used 500 trillion gallons of water in 2002, excluded household water use. Given that domestic total non-household water use is approximately 140 trillion gallons, results indicate that the US imports 3.5 times more water than is used domestically. We estimate that 60% of water is used indirectly, ie, is embodied in goods and services. Agricultural activities and power generation constitute a majority of direct water use (>60%). However, these sectors constitute only 30% of total water use when accounting for indirect supply chain effects. For 85% of the sectors, indirect or supply chain water withdrawals are larger than direct water use. Food and beverage processors are the largest indirect water users (20%). Power supply and grain farming are the most frequent sources of indirect water use. Several industrial sectors have relatively high water use per dollar of output, namely paint manufacturers, paperboard mills, and pesticide manufacturers. These results should be useful for environmental life cycle assessment of US production and other studies, but we conclude that better information on water use is essential for effective water management.

Poster Spotlight 1: Testing of Operational Methods for Impact Assessment of Freshwater Use on Midpoint and Endpoint level: Insights and Conclusions
Stephan Pfister, ETH Zurich, IFU
Annette Koehler, ETH Zurich
Stefanie Hellweg, ETH Zurich
presentation

We developed regionalized characterization factors for the environmental impacts from freshwater consumption on the midpoint and endpoint level. Midpoint factors are developed for the existing midpoints abiotic resource depletion and land use (by calculating land-use equivalents). These are further assessed for their impacts on the areas of protection resources and ecosystem quality. An additional midpoint category “water deprivation” is introduced, as a basis for modeling related effects to human health. Our method can be used within most existing LCIA methods, either on midpoint or endpoint level. Relative importance of water consumption is analyzed by integrating it into the fully aggregating Eco-indicator 99 method (EI99). Both, importance of regionalization in impact assessment and relative relevance of water consumption are illustrated by the case of several agricultural products produced worldwide. The results show that in arid regions, water consumption might dominate the aggregated overall impact whereas in humid regions, impacts from water use are usually not relevant. Consequently, water-consumption assessment considering a regionalized modeling approach is a crucial development in LCA. This is particularly important when products traded on the global market shall be compared and relevant environmental information shall be provided to decision makers in food supply chains or for consumers interested in sustainable consumption.

Based on the agricultural case studies, we compared the method as add-on of the conventional EI99 to the Swiss Ecological Scarcity Method 2006 (UBP06), which includes regional impact factors for water use. In general, the relative impact of water use compared to other impact categories has been smaller in EI99 than UBP06. We analyzed the reasons for this difference and reflected also further indicators for water scarcity in order to evaluate the limitations of the methods presented and give recommendations for the interpretation of the results.

Poster Spotlight 2: Characterization Factors for Damage to Aquatic Biodiversity caused by Water Use
Sebastien Humbert, Quantis
presentation

The way society uses water can significantly affect ecosystems and aquatic biodiversity. However, due to the complexity of this topic, these water use impacts are poorly addressed by current life cycle impact assessment methodologies. This work proposes a framework to calculate characterization factors to quantify damage to aquatic biodiversity caused by water use. Damage can be expressed as the potentially damaged fraction of species (PDF) that disappear over a certain area affected (in PDF*m2) because of a certain amount of water used per year for a specific purpose.

When this framework is applied to the case of hydroelectric dams, the impacts can be expressed either per amount of water flowing through the dam per year or relative to the amount of electricity produced per year. This framework expresses results using PDF*m2*yr per unit of water or electricity. This unit is advantageous in that it is consistent with other damage-oriented impact categories describing damage to ecosystems and thus allows comparison between different types of impacts. Suggested characterizations factors are calculated for dams.

Results show that for run-of-river hydroelectric dams, damage to aquatic biodiversity is on the order of 0.01 (0.002-0.1) PDF*m2*yr per m3 of water turbined or 0.04 (0.01-0.2) PDF*m2*yr per kWh produced. For alpine dams, damage is about 0.001 PDF*m2*yr per m3 of water turbined or 0.0008 PDF*m2*yr per kWh produced.

Thus, alpine dams appear to have damages that are approximately 50 times lower than run-of-rive dams (on a per kWh produced basis), partly explained by the higher efficiency from alpine dams (higher drop), thus less m3 of water is needed to produce 1 kWh. Overall, these values are between 1 and 3 orders of magnitude higher than the total damage to ecosystems currently evaluated for electricity production, indicating that the damage to aquatic biodiversity by hydroelectricity is significantly underestimated and should be considered when comparing the environmental impacts of competing electricity production technologies. These characterization factors can be directly combined with appropriate inventory databases and used in life cycle assessments related to water use. This is especially useful for processes using electricity from hydroelectric utilities, but also for processes using water for irrigation, power plant cooling or public use. Thus the inclusion of water use damage to aquatic biodiversity within overall damage to ecosystems quality increases the reliability of final results reported from LCA containing processes using water.

Poster 1: Relevance of Water Use in LCAs of Biofuel Production
Mireille Faist Emmenegger, Empa

Many countries promote biofuels with the main aim to reduce their greenhouse gas emissions. However, the potential of biofuels to reduce both greenhouse gas emissions as well as overall environmental impacts is highly controversial. Switzerland is one of the first countries to set requirements in its legislation on tax exemption for biofuels to ensure a reduction of greenhouse gas emissions while avoiding additional overall environmental impacts.

The assessment of environmental impacts of biofuels has until now mainly focused on greenhouse gas emissions and fossil energy use. Only few studies (e.g. Zah 2007) have adopted a more comprehensive approach which includes overall environmental impacts within the full life cycle of biofuels. Moreover, water use in biofuel production has only been considered by very recent papers, without, however, relating their results to overall environmental impacts. This step is nevertheless very important for a comprehensive assessment of biofuels. Indeed, one recent publication showed that the assessment of water consumption in an LCA of cotton production strongly influences the results, especially in water scarce regions (Pfister 2009).

Our paper analyses water use in biofuel production. It shows a case study of rape seed methyl ester production in Argentina based on irrigated and non-irrigated cultivation, and compares the results with values for other biofuels from the ecoinvent database. We use regionalized factors for the assessment of the environmental impacts of water use and show the importance of taking into account water scarcity in the respective region. Furthermore, we relate the results to the Swiss legislation, showing that water use should be included in future assessments of biofuels.

Poster 2: Water Footprint – Principles, Requirements and Guidance
Sebastien Humbert, Quantis

Freshwater is a very important natural resource, which everyday becomes scarcer and therefore urgently requires appropriate management tools to be used in an internationally consistent fashion by different stakeholders. The objective is to propose and ISO Standard on principles, requirements & guidelines for a water footprint metric of products, processes and organizations, as well as its communication. It aims at being:

1. Coherent with ISO14000 series,
2. Coherent with environmental metrics such as carbon footprint, LCA (ISO14040), GHG accounting (ISO14064),
3. Coherent with environmental communication (ISO14020, ISO 14067), and
4. Consistent with existing and ongoing work of GHG protocol (same boundaries, scope, etc.).

This poster will outline the foreseen process, starting in 2009 and planned to last 3 years, aiming at creating such an ISO Standard.