Environ. Sci. Technol. 2010, 44, 9516–9521
Measuring the Embodied Energy in Drinking Water Supply Systems: A Case Study in The Great Lakes Region WEIWEI MO,† FUZHAN NASIRI,‡ M A T T H E W J . E C K E L M A N , ‡,§ QIONG ZHANG,† AND J U L I E B . Z I M M E R M A N * ,‡,§ Department of Civil and Environmental Engineering, University of South Florida, Tampa, Florida, United States, Department of Chemical and Environmental Engineering, Yale University, New Haven, Connecticut, United States, and School of Forestry and Environmental Studies, Yale University, New Haven, Connecticut, United States
Received May 10, 2010. Revised manuscript received October 21, 2010. Accepted October 26, 2010.
A sustainable supply of both energy and water is critical to longterm national security, effective climate policy, natural resource sustainability, and social wellbeing. These two critical resources are inextricably and reciprocally linked; the production of energy requires large volumes of water, while the treatment and distribution of water is also significantly dependent upon energy. In this paper, a hybrid analysis approach is proposed to estimate embodied energy and to perform a structural path analysis of drinking water supply systems. The applicability of this approach is then tested through a case study of a large municipal water utility (city of Kalamazoo) in the Great Lakes region to provide insights on the issues of water-energy pricing and carbon footprints. Kalamazoo drinking water requires approximately 9.2 MJ/m3 of energy to produce, 30% of which is associated with indirect inputs such as system construction and treatment chemicals.
1. Introduction Water systems play a significant role in energy use nationally and globally, demanding not only large amounts of direct energy such as electricity but also a considerable amount of indirect energy embodied in the associated chemicals and materials. It is estimated that 4% of U.S. electricity demand is for the movement and treatment of water and wastewater, both publicly and privately (1), while approximately 2-3% of worldwide energy demand is for supplying water (2). Over the past decade, increasing attention has been placed on energy consumption in water systems as the growth and demand of this service is expected to continue to increase with increasing population and globalization (3). The electricity used directly in water systems, primarily to operate pumps, has been widely studied (1, 4-6). Construction and maintenance of water systems also require vast amounts of * Corresponding author phone: (203)432-9703; fax: (203)432-4837; e-mail:
[email protected]. † University of South Florida. ‡ Department of Chemical and Environmental Engineering, Yale University. § School of Forestry and Environmental Studies, Yale University. 9516
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 24, 2010
energy, both for direct fuel use and as energy embodied in infrastructure materials and treatment chemicals. Water mains and sewers, typically made of energy intensive iron/ steel and concrete, can represent a significant portion of energy consumption (7). Energy is required not only to produce but also to transport these materials. Hence, the total life cycle energy consumption of delivering water, that is the sum of energy inputs for construction, operation, and maintenance phases, is an appropriate metric when considering technology options or when formulating policy. Several researchers have used life cycle assessment (LCA) to evaluate water supply systems or technologies. LCA is an accounting framework for quantifying environmental impacts across the entire life cycle of a product or process. Process LCA, in which individual flows of material and energy are tracked at the process engineering level, has been utilized to analyze the environmental impacts associated with various water treatment options in England (8), South Africa (9), Australia (10), Spain (11), and the United States (12), while analyses of the impacts of entire municipal water and wastewater systems have been undertaken on the basis of a single unit of water (13) and in total (14). Options for water reuse and recycling have also been considered (15). At yet a larger scale, Wilkinson et al. (6) conducted an exploratory life cycle energy assessment of the statewide water supply of California, with spatially explicit results on the embodied energy of water in different areas of the state. Although process LCA allows researchers to examine detailed system characteristics, it can be quite data-intensive and suffers from well-known disadvantages, including user-defined (and therefore somewhat arbitrary) system boundaries, a bias toward direct material and energy inputs, and resulting error from omitting indirect effects that may occur up the supply chain (16). To address these challenges associated with process LCA, researchers have turned to Input-Output (I-O) analysis to estimate environmental impacts, including CO2 emissions and energy consumption across a variety of sectors and services (17-20). I-O analysis is an economic technique that uses a matrix formulation of the entire economy to specify how demand for goods or services in one sector creates activity in other sectors (21). This technique can be extended to quantify environmental impacts, by identifying how demand for a particular good or service from one sector results in environmental impacts, such as energy use, throughout the supply chain (19). I-O analysis provides a comprehensive method for accounting for the indirect impacts of supplying water that occur in other sectors of the economy, in producing piping materials or water treatment chemicals, for example. These indirect effects can be highly significant; in the case of greenhouse gas emissions, indirect effects were shown to dominate the carbon footprints of most sectors of the U.S. economy (22). Because basic I-O models are based on national average data, the method may be inappropriate to apply directly to an individual product or water system (16). In this paper, a hybrid approach is utilized, combining the comprehensive coverage of I-O analysis with the specificity of process assessment, to estimate energy embodiment and to perform an energy path assessment in drinking water supply systems. Hybrid assessment is routinely used in LCA and can take various forms (16, 23-25) however, it has rarely been applied to energy analysis of water systems. Filion et al. (26) constructed a hybrid model to evaluate net energy use over the life cycle of an urban pipe network, using New York City as a case study. Stokes and Horvath (27) used 10.1021/es1015845
2010 American Chemical Society
Published on Web 11/24/2010
a hybrid approach to evaluate various water supply options for California on the basis of various environmental parameters, including energy use. While these models are notable for their complexity and breadth, they were created for specific contexts and are thus difficult to adapt to other water supply systems in the U.S. and elsewhere. The objective of the present study is to develop a flexible model for estimating both the direct energy used to construct and operate the system, and the indirect (or upstream) energy embodied in the production and transport of materials and energy used by the system. The applicability of this approach is tested through a case study of a large public water utility in Kalmazoo, Michigan, a city in the Great Lakes region to provide insights on the issues of water-energy pricing and greenhouse gas emissions.
2. Methodology 2.1. Calculation of Initial Embodied Energy Intensities. The I-O model used here is based on the latest 2002 Use table (U) and Make table (M) published by U.S. Bureau of Economic Analysis (28), which were used to create a commodity-by-commodity direct coefficient table (CC) as follows CC ) B × D;B ) Ug-1 ;D ) Mg-1
(1)
where CC ) commodity-by-commodity direct coefficient table; B ) commodity-by-industry direct coefficient matrix; D ) industry-by-commodity direct coefficient matrix; U ) use table; M ) make table; g ) a column vector showing the total $ output of each industry; and q ) a column vector showing the total $ output of each commodity. The coefficients in the commodity-by-commodity direct coefficient matrix (28) show the monetary amount of different commodities (in columns) directly needed to produce one dollar of output of a certain commodity (in rows). There are, in total, 424 commodity sectors in the BEA tables. Of these, two target sectors were selected to represent water systems (r ) 1, 2): the water, sewage, and other systems sector (WSOS) and the other nonresidential structures sector (NS). The former is used to represent the operation and maintenance of water supply systems and the latter is used as a proxy for the construction of water supply systems. These sectors clearly include economic activities not directly related to the provision of water, and the necessary and appropriate adjustments made to CC table are described in later sections. These adjustments require the use of the direct requirements table used in this study, as opposed to the more aggregated total material requirements table used in other I-O formulations (16). The direct and total (direct + indirect) embodied energy per unit output of the adjusted WSOS and NS sectors are calculated by examining the inputs to both sectors of study from other energy-related sectors. Direct energy for WSOS is the energy used for operating and maintaining the water system, such as electricity for pumping; indirect energy for WSOS is the energy used to manufacture and deliver nonenergy inputs used for operation or maintenance, such as treatment chemicals. Direct energy for NS is that used for constructing the water system, such as diesel fuel for mixing the concrete, while indirect energy for NS is that used to manufacture and deliver the cement to the site, for example. Five energy supply sectors are included here: oil and gas extraction, coal mining, power generation and supply, natural gas distribution, and petroleum refineries. The distribution of coal and petroleum is not included in these energy supply sectors as they represent a small fraction (
