Setting the System Boundaries of “Energy for Water” for Integrated Modeling

Many studies over the last two decades have addressed the “water-energy nexus,” generally defined as the interdependency between water and energy in their supply, processing, distribution, and use. The research community currently disaggregates the water-energy nexus into two components: “water for energy” and “energy for water.” While there seems to be clear consensus on the definition of “water for energy”—that is, water required for the extraction, processing, and transformation of energy as well as the irrigation of bioenergy—there has been less agreement on the definition and system boundaries of “energy for water.” We represent six integrated assessment modeling teams presently incorporating the hydrologic system and water demands into existing global models of energy, agriculture, land use, and climate. In this article, we propose system boundaries of “energy for water” that are appropriate for integrated energy and water modeling, and introduce a third category of processes that are relevant for the water-energy nexus, but that are not logically classified as either “water for energy” or “energy for water.”

In the literature, some studies have estimated “energy for water” as the energy used for water abstraction, treatment, distribution, and postuse wastewater treatment.(1) Others have also included water-related energy consumption in the residential, commercial, and industrial sectors (e.g., for water heating and cooling). When included, these “end-use” processes typically account for more than two-thirds of total “energy for water.”(2) Using even broader system boundaries that consider all processes where energy is applied to water, including all primary energy used at thermoelectric power plants, Sanders and Webber classified 47% of total primary energy in the United States as “energy for water.”(3)

While interesting as an accounting exercise, classifying all energy applied to water as “energy for water” mischaracterizes the purpose of the energy consumption in processes that take energy and water inputs in order to produce a nonwater output or service. This classification is especially problematic in the context of modeling the water-energy nexus; integrated whole-economy models of the water and energy systems explicitly represent inputs (e.g., labor, capital, energy, and/or water) to modeled processes whose outputs are either: (a) commodities that are in turn inputs to other modeled processes; or (b) final end-use demands. In this context, no input to a modeled process should be classified as being for another input to the same process. While the primary energy used for thermoelectric power generation does heat water, this is an intermediate step within a larger process whose output is (secondary) energy, not water. In place of the currently adopted definition, we propose that “energy for water” should refer only to the energy used for processes whose main output is water (see Figure 1), such as water abstraction, treatment, distribution, and postuse wastewater treatment.

Figure 1

High Resolution Image

Download MS PowerPoint Slide

Our proposed system boundaries of “energy for water” exclude energy used for end-use technologies (e.g., dishwashers) that use both energy and water as inputs to produce services (e.g., clean dishes). However, we recognize the importance of this type of process in quantifying and understanding the interlinkages between water and energy systems, and therefore propose a third category of processes within the water-energy nexus: “water and energy for other purposes.” Processes within this category take inputs of water and energy, among others, and provide desired services or commodities that are neither water nor energy. The specific representations of this type of process in any model will depend on the model’s level of technological detail. As most global models currently do not have appliance-level representations of end-use sector energy and/or water consumption, improving the representation of end-use technologies is recognized as a priority for future model development.

Fortunately, for modeling of any process considered as “water and energy for other purposes,” it is not necessary to estimate the fraction of the energy input that is physically applied to the water input. Adopting this approach more broadly would significantly mitigate the data burden currently facing the water-energy nexus research community, as the presently adopted definition of “energy for water” (as in Sanders and Webber)(3) often requires information at a finer level of detail than is available even in bottom-up surveys. For instance, cooking energy consumption surveys do not disaggregate energy applied to water from energy applied to nonwater substances, in part because this information has not been considered worthwhile by relevant stakeholders, and also because there is no clear delineation between water and nonwater substances in this process.

Nevertheless, even with our proposed definition, the data challenges for estimating global “energy for water” remain substantial. In most nations, the quantities of energy used for water abstraction, treatment, distribution, and postuse wastewater treatment are not known. In detailed national energy statistics(4) and international energy balances,(5) the energy used by municipal water systems is typically not disaggregated from other commercial sector energy use. For several other water-related uses of energy, such as irrigation systems and seawater desalination, it is unclear in which sector energy use is reported.

In this context, the lack of understanding of how much energy is currently used globally for water abstraction, treatment, distribution, and wastewater treatment limits assessment of the future feedbacks between the energy and water systems. Quantifying the energy consumed by these processes whose primary output is water is the first step toward integrated modeling of future energy for water. This is the endeavor to which the research article by Liu et al.(6) contributes. Despite our proposed revisions to existing definitions, we underscore that our suggested “energy for water” system boundaries do not trivialize the energy and water linkages in end-user processes that consume both energy and water. Rather, these processes remain important to track in the context of the future evolution of the water-energy nexus. In the longer term, as the models’ characterizations of end-use energy and water consumption become more detailed, the capability to quantify the interdependencies between energy and water will likewise be enhanced.