Value intenity of water ued for electrical energy generation in the Wetern U.S.; an application of embedded reource accounting Elizabeth A. Martin and Benjamin L. Ruddell Abtract Thi tudy evaluate the water intenity of thermoelectric power plant in the eleven Wetern tate included within the Wetern Electricity Coordinating Council region. Water intenitie are combined with retail electricity ale uing an embedded reource accounting framework to etimate a value intenity of water embedded in electrical energy production and trade in the Wetern U.S. State with lower water intenitie and higher retail electricity price tend to be net importer of electricity and have the highet value intenitie. A 35% increae in the value intenity of water embedded in electricity traded in export wa hown to have occurred relative to electricity produced for in-tate ue. Thi increae in the value intenity of embedded water ugget that embedded water in the electricity trade ha already emerged a a ubtitute for direct trade in water reource, and that thi trade i organized in a manner that appear to benefit both importing and exporting partie. Index Term embedded reource accounting, water-energy nexu, water in energy, embedded water, electrical grid. A I. INTRODUCTION N analyi of the relationhip between climate and economic change and the energy-water nexu i needed for the purpoe of informing National water and energy policy for the 21 t century. Climate change i expected to caue increaing temperature and evaporation, decreaed rainfall, and more intene drought in the Southwetern U.S. A population and indutry continue to grow, reource demand increae and become more patially concentrated around urban area. Thi i particularly true of demand for electrical energy. Electrical energy production account for the larget percentage of gro water withdrawal in the U.S., placing water reource at the focal point of the energy-water nexu a an important and climate-enitive contraint on electrical energy production. Reallocation of water upplie in addition to reditribution of the production of electrical energy and other reource will be neceary to adapt reduced Thi work wa upported in part by the U.S. Department of Energy Sandia National Laboratory in partnerhip with the Wetern Electricity Coordinating Council. E.A. Martin i a Ph.D. tudent in the School of Sutainable Engineering and the Built Environment, Arizona State Univerity, Tempe, AZ 85283 USA (e-mail: elizabeth.a.adam@au.edu). B.L. Ruddell i an Aitant Profeor with the Department of Engineering, Arizona State Univerity, Mea, AZ 85212 USA. (e-mail: bruddell@au.edu). upplie to meet increaing and patially concentrated reource demand. The re-location of exiting old water reource and acce to low-quality new water reource often involve prohibitive infratructure cot, energy cot, and legal barrier. However, there i a ignificant amount of water embedded in electrical energy production. Therefore, the remote production and virtual tranmiion of water in electricity and other reource provide a powerful management olution for an efficient adaptation to water reource challenge. Embedded (or virtual ) water accounting provide a method for the evaluation of propoed electrical energy production adaptation to water limitation. Embedded Reource Accounting (ERA) i a general input/output framework aeing reource embedded in production and trade. The management of water reource ha played a large role in defining the culture of the Wetern United State. The ettlement of the Wet wa made poible through large government ubidized water project and fixed water right, rather than through market baed economic development of water reource [1]. A a reult, the value of water i difficult to directly oberve or ae. However, a proxy for the value of a reource uch a water, for which there i no functioning market, can be etablihed by oberving water embedded in the proce of production of marketed good and ervice. We perform an analyi of water reource embedded in the electrical energy trade acro the Wetern U.S. to reveal meaningful trend in the value intenity of embedded water in traded electricity. The value intenity for a production proce i the quantity of omething of value created by the proce relative to the quantity of a reource impacted by the proce. Value intenity i not an economic meaure of price becaue many objective, ubjective, economic, ocial, and environmental value are output of a proce, and many input are aociated with a proce. Value intenity i the ratio of the quantity of a ingle value created or produced to a ingle reource conumed or impacted. Thi tudy evaluate the water intenity of power generation plant in the eleven Wetern tate included within the Wetern Electricity Coordinating Council region (Arizona, California, Colorado, Idaho, Montana, New Mexico, Nevada, Oregon, Utah, Wahington, and Wyoming), and combine thi information with retail electricity ale to etimate the value intenity of water embedded in electrical energy production
and trade. II. METHODOLOGY Conider three reource: A, B, and C. Here, a "reource" i broadly defined and may be a phyical reource, a good, a ervice, or anything that can be valued, produced, conumed, created, or detroyed a a part of a proce. Reource A i produced through a proce uing a variety of input including reource B. We define the quantity B/A a the intenity of ue of reource B in the production of reource A by uch a proce. There mut exit a market that trade reource A for reource C at a price expreed a C/A. For thi tudy, A = electricity produced by a generation proce and traded for currency (MWh), B = net raw water conumption, i.e. the "blue" water footprint [2] of an electrical energy generation proce (gal), and C = currency paid in exchange for electricity by retail conumer ($ USD). Water, reource B, i a non-market reource for which we can etablih a non-market value metric relative to a markettraded reource, electricity, reource A. Thi non-market value metric i the value intenity (VI) of the embedded water in traded electrical energy that i produced by a pecific proce. In thi cae, VI i found through computation and analyi of the quantity of water (B) embedded in the production and trade of electricity (A) in exchange for valued currency (C) uing the ditribution network of electrical energy trade acro the Wetern U.S. and the reource intenitie, B/A and C/A, dicued above. A. Data The data ued in thi tudy wa obtained from Sandia National Laboratory, the U.S. Environmental Protection Agency and U.S. Energy Information Adminitration online databae. Data ued conit of: Sandia National Laboratory upplied model etimate of megawatt-hour of electricity produced annually (model year 2020) at each power plant within each tate for the eleven wetern U.S. tate elected [3-5], Sandia National Laboratory upplied etimated net water conumption data per day for each of the power plant within each tate [4-9], EIA reported average utility retail price of electricity for each utility within each tate for 2009 [6], and EIA reported total electricity import and export data for each tate for 2009 [10]. The reource intenitie for the network being examined are preented in Table 1. Thee intenitie are the average water intenity of electricity generation, Φ[], and the average price of electricity, f[], in State. The number ued in thi tudy were obtained by computing average for each tate baed on water conumption of energy generation at each plant and average retail price charged by each utility within each tate weighted by total energy produced by each plant and electricity old by each utility. Note that we aggregate electrical production and ditribution together a a ingle proce; thi aumption i neceitated by limited data availability. TABLE 1 WATER INTENSITY OF POWER GENERATION BY STATE, LISTED HIGHEST TO LOWEST, AND PRICE OF ELECTRICITY BY STATE Water Intenity (gal/mwh) Φ[] Price ($/MWh) f[] New Mexico 437.25 $103.56 Utah 411.77 $81.35 Wyoming 384.17 $85.57 Colorado 352.66 $100.26 Nevada 349.23 $80.10 Montana 297.32 $81.57 Arizona 183.81 $86.23 California 129.69 $125.26 Idaho 83.31 $62.91 Oregon 82.04 $67.65 Wahington 52.52 $61.65 California and the Pacific Northwet tate (Oregon, Idaho, and Wahington) have relatively low water intenitie. California low water intenity may be (a i the cae with the Pacific Northwet tate) partly due to the ue of hydroelectric facilitie, but i alo likely a reult of water conervation meaure. Higher water intenitie are aociated with thermoelectric procee [11]. Electricity price in each of the eleven wetern tate reveal that California pay the highet price for electricity in the region, and Pacific Northwet tate pay the leat (Table 1). State with low average retail electricity price are mainly the reult of the preence of low-cot hydro-electric power from federal dam [12]. High average retail electricity pricing in a tate ugget a high energy demand paired with a limited upply and/or high cot of electricity generation. Thee intenitie combined with data for electrical ditribution and trade in the Wetern United State yield the value intenity of water embedded in that electricity. B. Etimating the Ditribution Network The electricity generation and ditribution network in the Wetern United State i compried of power plant, electric utilitie, electrical tranformer, tranmiion and ditribution infratructure, etc. For thi analyi, we conceptualize the ytem imply a a tranportation network with variou reource (electricity, economic currency, and water embedded within electricity) flowing between the node (tate). Conumption of electrical energy mut equal production for the ytem a a whole, after accounting for net export from the network. The EIA production and conumption data for 2009 for the eleven tate under analyi reult in an exce of electrical energy of le than 1% of the total production [10]. To balance the ytem for thi analyi the exporting tate production were reduced by thi amount (Table 2); the exce electricity i preumed to leave the grid for neighboring grid.
Net trade in electricity, T NET [], i taken a production minu conumption within each tate. State with a poitive net trade in electricity will export that amount, and tate with a negative net trade will import needed electricity from exporting tate via the wetern power grid network. Uncontrained electrical tranmiion between all tate i aumed. Nonzero value for Gro Export, T O [], and Gro Import, T I [], for the 11 tate are hown in Table 2. Only nonzero value are hown. By definition, the um of all T O [] over all tate i equal to the um of all T I [] over all tate. Net Intertate Trade, T NET [], (MWh) Arizona 31,685,245.00 Montana 5,775,543.00 New Mexico 15,700,958.00 Nevada 1,655,392.00 Oregon 5,079,110.00 Utah 12,389,184.00 Wahington 2,117,039.00 Wyoming 26,882,529.00 California (84,137,000.00) Colorado (4,815,000.00) Idaho (12,333,000.00) TABLE 2 INTERSTATE ELECTRICITY TRADE Gro Export, T O [], (MWh) Gro Export Percentage, C O [], (%) 31,685,245.00 31.3% 5,775,543.00 5.7% 15,700,958.00 15.5% 1,655,392.00 1.6% 5,079,110.00 5.0% 12,389,184.00 12.2% 2,117,039.00 2.1% 26,882,529.00 26.5% Gro Import, Gro Import T I [], (MWh) Percentage, C I [], (%) 84,137,000.00 83.1% 4,815,000.00 4.8% 12,333,000.00 12.2% Nonzero value of Gro Export/Import Percentage, C O [] and C I [], are preented a well. Gro import and export percentage give the percentage of the total electrical energy on the power grid that wa upplied by each exporting tate and the percentage of the total purchaed by each importing tate. Note that the um of all C O [] over all tate i equal to 100%, and the ame i true for C I []. The amount exported from each tate to each importing tate i etimated a the product of the Gro Export, T O [], of the exporting tate and the Gro Import Percentage, C I [], of the importing tate. i.e., California C I [] = 83.1%; therefore 83.1% of the gro export from each exporting tate i allocated to California. The reulting electricity tranfer are hown in Fig. 1. A 2005 tudy by Marriott and Matthew [13] on electricity generation and conumption mixe in the Wetern U.S. utilized a imilar tranportation network approach with a linear programming optimization model to etimate intertate electricity trading for year 2000 which reulted in energy tranfer very imilar to thoe found in our tudy, with California dominating import and Arizona the larget exporter with correponding order of magnitude for all intertate electricity tranfer. Fig. 1. Electricity trade network for the Wetern United State (TWh). Electricity tranfer quantitie are hown for all exporting tate to all importing tate, indicated by node and arrow. Internally produced and conumed electricity for each tate i hown in blue. California dominate import conuming 83.1% of the traded electricity. C. Etimating Value Intenitie The average retail price (f[]) paid per megawatt-hour of electricity from Table 1 wa applied to the electricity tranfer acro the Wetern U.S. hown in Fig. 1 to obtain the average retail price of each tranmiion. Mot tate were able to upply their electricity demand with dometically produced electricity; however, in the cae where electricity wa exported the price realized by the producer i the average retail price (f[]) of the receiving (or importing) tate. f repreent the weighted average market price of available electricity on the grid (of all import). A hown in (1), f i the average f[] weighted by volume of gro import per tate [] in $USD per megawatt-hour. I f T = (1) f I T f repreent the average local price of available electricity on the grid (of all export). A hown in (2), f i the average f[] weighted by volume of gro export per tate [] in $USD per megawatt-hour. O f T f = O (2) T The volume of water conumed per megawatt-hour of electricity produced, Φ[] (Table 1), wa applied to the electricity ditribution network acro the Wetern U.S. ( Fig. 1) in order to determine the amount of embedded water contained within each tranfer. Internal water intenitie (Φ[]) were applied for electricity demand met with dometically produced electricity; however, for tranfer of electricity from exporting to importing tate the Φ[] applied wa that of the upplying (or exporting) tate. Φ repreent the weighted average embedded water content of the available electricity on the grid (total of all export). Φ i found according to (3) a the average Φ[]
weighted by volume of gro export per tate [] in gallon per megawatt-hour. Φ O T O T Φ = (3) The VI of water embedded in locally produced and conumed electricity for each tate (VI l []) wa found according to (4), comparing local water intenity to local electricity pricing: l f VI = Φ (4) The VI of water embedded in exported electricity (VI O []) wa found for each exporting tate according to (5), comparing the average market price of electricity available on the grid with the water intenity of each exporting tate: O f = (5) VI Φ The VI of water embedded in imported electricity (VI I []) wa found for each importing tate according to (6), comparing the price of electricity of each importing tate with the average market water intenity of electricity available on the grid: I f = (6) VI Φ An average market VI for embedded water in electricity purchaed from the grid by conumer (import) wa found uing (7), comparing the average market price of electricity available on the grid with the average market water intenity of electricity available on the grid: f VI = (7) Φ Similarly, an average producer VI for embedded water exported wa found uing (8), comparing the average local price of electricity exported to the grid with the average market water intenity of that electricity: VI = (8) f Φ Comparing VI obtained uing (7) and (8), the percent change in VI due to trade can be determined. See (9). VI VI VI = (9) III. RESULTS Scott and Paqualetti (2010) reported the reult of a thorough multi-year tudy of the Energy-Water Nexu within Arizona and Sonora. The reult of our analyi for export of water embedded in electricity from Arizona to other tate are comparable to thoe of Scott and Paqualetti, though our analyi i at a larger cale and lower reolution. Reulting embedded water tranfer for the entire network are not preented here, only thoe for Arizona for comparion purpoe. Table 3 compare reult from thi tudy with thoe found previouly by other [12]. TABLE 3 NET EXPORT OF EMBEDDED WATER IN ELECTRICITY Net Export from Arizona to: Martin and Ruddell (2012) Scott and Paqualetti (2010) California 4,838 Mgal 7,984 Mgal Northwet (Idaho) 709 Mgal 1,932 Mgal Colorado 277 Mgal -1,100 Mgal The value intenity reult are preented in Fig. 2. The reult how that higher value intenitie are generally aociated with tate that have lower power generation water intenitie and which import electricity (ee alo Table 1). The percent increae or decreae in value intenity due to trade i hown for each tate. A trend i that exporter generally increae the value intenity of water embedded in export compared to the value intenity of water embedded in locally conumed electricity, and importer generally decreae in the value intenity of water embedded in imported electricity compared to that of water embedded in dometically produced electricity. Comparion of the reulting VI and VI, which are weighted average of the previouly dicued value intenitie, how that the overall percent change,, i a 35% increae in value intenity of embedded water through trade in electricity on the Wetern power grid. IV. CONCLUSIONS Undertanding the value intenity of water that i embedded in traded product and ervice provide ueful information and perpective. The Embedded Reource Accounting framework wa ued to expoe the value intenity of water embedded in electricity traded acro the Wetern U.S. Thi analyi how the urpriing reult that an implicit trade in embedded water appear to exit in the Wetern U.S., and that thi trade appear to benefit both importing and exporting tate a exporter increae and importer decreae the value intenity of water embedded in electricity trade. In general, State with higher water intenitie are net exporter of electricity and embedded water and have lower internal retail price for electricity. The trade in embedded water, and the value intenity of the water, i dominated in the Wetern U.S. by the import of thee reource by California. Further work will focu on applying thi analyi on a maller cale (at county, utility, and waterhed level) and performing a imilar analyi on the value intenity of water embedded in agricultural product produced and traded in the Wetern U.S.
Fig. 2. Value intenitie of embedded water within electricity traded throughout the Wetern United State ($USD/gal). Higher value intenitie are generally aociated with State that have lower local water intenitie per MWh (i.e. le water ued per MWh, or more water-efficient generation). Blue bar how local value intenitie for water embedded in locally produced and conumed electricity (VI l ). Red bar how value intenitie of water embedded in exported electricity (VI O ), and green bar how value intenity of water embedded in imported electricity (VI I ). Exporter generally ee an increae in value intenity compared with local value intenity, and importer generally ee a decreae in value intenity compared with local value intenity. The percent change in value intenity due to trade for each tate i hown. Comparion of reulting VI and VI the value intenity of water embedded in electricity production and trade increae an average of 35% overall. ACKNOWLEDGMENT We would like to thank Vincent Tidwell of the Sandia National Laboratory and hi Water and Energy cience team for providing data and valuable ongoing collaboration on the development of thee finding, and Mike Paqualetti of ASU for review of ongoing work. REFERENCES [1] Reiner M. (1993). Cadillac Deert: The American wet and it diappearing water. New York: Penguin Book. [2] Hoektra, A.Y..; Chapagain, A.K.; Aldaya, M.M.; & Mekonnen, M.M.. (2011). The Water Footprint Aement Manual: Setting the Global Standard. Earthcan Publihing. Water Footprint Network. Available online. http://www.waterfootprint.org/download/thewaterfootprintaemen tmanual.pdf. [3] EPA (U.S. Environmental Protection Agency), 2010. Emiion & Generation Reource Integrated Databae (egrid), verion 1.0 [4] EIA (U.S. Energy Information Adminitration), 2005. Steam-Electric Plant Operation and Deign Report (EIA-767). [5] Sandia National Lab Energy/Water Nexu Group (2011): Etimate of water withdrawal and conumption at exiting plant were developed from a variety of ource. [6] EIA (U.S. Energy Information Adminitration), 2011a. Online. Sale data table. Electricity. Acceed October 2011. http://www.eia.gov/electricity/data.cfm#ale. [7] Kenny, J., Barber, N., Huton, S, Liney, K., Lovelace, J., & Maupin, M. (2009). Etimated ue of water in the United State in 2005. U.S. Geological Survey Circular 1344. [8] Macknick, J., Newmark, R., Heath, G., & Hallett, K.C. (2011). A Review of Operational Water Conumption and Withdrawal Factor for Electricity Generating Technologie, NREL/TP-6A20-50900, National Renewable Energy Laboratory, Golden, CO, pp. 29. [9] Solley, W., Pierce, R., & Perlman, H. (1995). Etimated Ue of Water in the United State in 1995, U.S. Geological Survey Circular 1200. [10] EIA (U.S. Energy Information Adminitration), 2011b. Online. Supply and Dipoition of Electricity. State electricity profile. Electricity. Acceed October 2011. http://www.eia.gov/electricity/tate/. [11] Scott, C., & Paqualetti, M. (Fall 2010). Energy and water reource carcity: Critical infratructure for growth and economic development in Arizona and Sonora. Natural Reource Journal, 50, 645-682. [12] EIA (U.S. Energy Information Adminitration), 2011c. Online. Factor affecting electricity price. Energy Explained. Acceed December 2011. http://www.eia.gov/energyexplained/index.cfm?page=electricity_factor _affecting_price. [13] Marriott, J.,& Matthew, S. (2005). Environmental effect of intertate power trading on electricity conumption mixe. Environmental Science and Technology, 39, 8584-8590.