Article pubs.acs.org/est
Life Cycle Assessment of a Power Tower Concentrating Solar Plant and the Impacts of Key Design Alternatives Michael B. Whitaker,† Garvin A. Heath,*,‡ John J. Burkhardt, III,‡,§ and Craig S. Turchi‡ †
ICF International, 9300 Lee Highway, Fairfax, Virginia 22031, United States National Renewable Energy Laboratory, 15013 Denver West Parkway, Golden, Colorado 80401, United States
‡
S Supporting Information *
ABSTRACT: A hybrid life cycle assessment (LCA) is used to evaluate four sustainability metrics over the life cycle of a power tower concentrating solar power (CSP) facility: greenhouse gas (GHG) emissions, water consumption, cumulative energy demand (CED), and energy payback time (EPBT). The reference design is for a dry-cooled, 106 MWnet power tower facility located near Tucson, AZ that uses a mixture of mined nitrate salts as the heat transfer fluid and storage medium, a two-tank thermal energy storage system designed for six hours of full load-equivalent storage, and receives auxiliary power from the local electric grid. A thermoclinebased storage system, synthetically derived salts, and natural gas auxiliary power are evaluated as design alternatives. Over its life cycle, the reference plant is estimated to have GHG emissions of 37 g CO2eq/kWh, consume 1.4 L/kWh of water and 0.49 MJ/kWh of energy, and have an EPBT of 15 months. Using synthetic salts is estimated to increase GHG emissions by 12%, CED by 7%, and water consumption by 4% compared to mined salts. Natural gas auxiliary power results in greater than 10% decreases in GHG emissions, water consumption, and CED. The thermocline design is most advantageous when coupled with the use of synthetic salts.
■
INTRODUCTION Research and development for concentrating solar power (CSP) is primarily focused on two power plant technologies: parabolic trough (trough) and power tower (tower).1 This study represents the second in a series of life cycle assessments (LCAs) of these CSP technologies led by the National Renewable Energy Laboratory (NREL). The first NREL CSP LCA2 evaluated a hypothetical, reference design for a parabolic trough CSP plant in Daggett, CA along with several design alternatives to estimate four metrics important for environmental sustainability: life cycle greenhouse gas (GHG) emissions, water consumption, cumulative energy demand (CED), and energy payback time (EPBT). Burkhardt et al.2 is hereafter referred to as the trough LCA. The goal of the present study is to evaluate life cycle environmental metrics for a state-of-the-art, reference design of a tower CSP plant hypothetically located in the United States (U.S.). To facilitate comparisons between the CSP technologies, this study uses the same analysis methods and environmental indicators as the trough LCA with similar design alternatives (listed as reference plant design vs alternative design) including (1) mined vs synthetically derived salts, and (2) twotank vs thermocline thermal storage designs, in addition to (3) electricity vs natural gas for auxiliary heating. Hybrid LCA is used to holistically evaluate the environmental impacts of the tower CSP plant. Hybrid LCAs use a combination of two methods: (1) a bottom-up, process-based analysis of component masses and process energy flows and (2) a top-down, economic input− output (EIO)-based assessment of economic costs attributed to © XXXX American Chemical Society
project activity in various industries. Collectively, the two CSP LCAs add to a growing set of literature that enables the comparison of life cycle environmental impacts across a range of renewable and nonrenewable electricity generation technologies.3 Similar to the trough LCA, this is the first LCA of tower CSP in the U.S. in over 15 years and the analysis is distinct in that it both considers a modern design and is the first to quantify life cycle water consumption for this CSP technology. CSP Power Tower Technology. Just over 500 MW of CSP are in operation in the U.S. today.4 While the majority of installed capacity uses trough technology, new tower plants are under construction.4 These new plants are substantially larger than the current operating CSP facilities and many employ dry cooling to reduce water consumption. In particular, the largest tower CSP plants in the world are under construction in southern California (Ivanpah Solar Electric Generating System) and Nevada (Crescent Dunes Solar Energy Project).4 Because of their higher operating temperatures and lower-cost solar collectors, towers can potentially produce power for lower cost than existing parabolic trough designs.5 Utility-scale tower plants generate electricity by using thousands of sun-tracking mirrors called heliostats to focus sunlight onto a receiver mounted on a tall tower. The receiver houses a heat exchanger containing a heat Received: February 23, 2013 Revised: May 9, 2013 Accepted: May 10, 2013
A
dx.doi.org/10.1021/es400821x | Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Environmental Science & Technology
Article
transfer fluid (HTF) which is warmed by the concentrated sunlight and is then pumped to the power block containing a standard steam turbine. Power towers can be categorized by the receiver HTF: direct steam towers boil water to produce steam in the receiver, while molten salt towers heat a liquid salt HTF in the receiver. In the latter case, the salt is used to boil water to produce steam via a separate heat exchanger. Direct steam towers eliminate the additional heat exchanger hardware and can run at slightly higher efficiency. Molten salt designs have the advantage that the hot salt can be stored to provide power regardless of solar availability. The energy storage efficiency in molten salt towers has been estimated at greater than 98%.6 The heliostat field, tower height, receiver, and HTF storage tanks are sized to meet power generation and capacity factor requirements. Power generation is utility scale with typical net capacity of 10−100 MW and designed capacity factors ranging between 20% and 70%.4 Tower designs have been the subject of research and development in the U.S. for four decades.7 Power grid analysis suggests that the energy storage capability of CSP systems will gain importance as deployment of intermittent renewable electricity generation technologies such as photovoltaics and wind increases.8,9 Therefore, this study analyzes a state-of-the-art molten salt tower CSP plant that includes thermal energy storage. The system selected is based on the technology designed, built, tested, and operated in the Solar Two research project7 and used in several commercial CSP plants.10,11
materials to Ecoinvent 2.0 material processes with representative manufacturing locations chosen to calculate transportation impacts. The Ecoinvent 2.0 database15 was the primary source for the GHG emissions, CED, and water consumption values assigned to the individual materials analyzed in this study. For the aspects of the plant where detailed material or process energy data were unavailable, the Carnegie Mellon Economic Input Output (EIO) LCA tool16 was used to estimate GHG emission, CED, and water consumption impacts from producer price costs allocated to appropriate industry sectors in 2002 dollar equivalents. EIO LCA was also used to estimate the impacts from highly manufactured components such as pumps, boilers, condensers, heat exchangers, transformers, the turbine generator set, and miscellaneous electrical and mechanical system components. Highly manufactured components were judged to have life cycle impacts that significantly exceed those from the sum of constituent material weights due to high manufacturing energy or consumed materials. The functional unit for the study is 1 kWh of electricity generated at the plant. The four primary metrics evaluated are as follows: • GHGs: using IPCC global warming potentials (GWP) from the fourth assessment report,17 emissions of individual GHGs throughout the tower plant life cycle are weighted and summed to a common metric of grams of CO2 equivalents per unit of electricity generated (g CO2eq/kWh). • Water consumption: consumed water is defined as the amount of water that is removed from the immediate water environment through processes such as evaporation, transpiration, product or crop incorporation, or human or livestock consumption.18 The volume of all surface and groundwater consumed throughout the tower plant life cycle is calculated and reported as liters of water consumed per unit of electricity generated (L/kWh). • CED: the sum of all renewable and nonrenewable energy consumed over the life cycle of the plant reported as megajoule equivalents per unit of electricity generated (MJeq/kWh). For the reference plant and all alternative scenarios, nonrenewable energy sources accounted for 97−98% of CED. • EPBT: the length of time required for the tower plant to generate sufficient energy in the form of electricity to offset the CED over the life cycle of the facility. EPBT is calculated as
■
METHODS A summary of methods for this study is reported here and elaborated in the Supporting Information (SI). Scope. The temporal vintage of the tower plant analyzed here is 2012 with a geographic reference of southwestern Arizona. Weather data for Tucson, AZ are used in the analysis. Annual direct normal insolation (DNI) is approximately 2600 kWh/m2 for Tucson, AZ,12 and the project life is 30 years. Following the guidelines described in the international standard series for LCAs (ISO 14040-44),13 this hybrid LCA evaluates environmental impacts for the following life cycle stages of the hypothetical tower CSP plant with infrastructure impacts amortized over the infrastructure element’s useful life: • Manufacturing: extraction of raw materials, transport of raw materials to the manufacturing facility, component manufacturing processes, and transport of the final product to regional storage. • Construction: transport of materials from regional storage to the project site, site improvement activities and materials, and energy used to assemble the plant. • Operation and maintenance (O&M): manufacture of replacement materials and transport to the project site, plant water consumption for power generation and mirror cleaning, maintenance vehicle fuel consumption, on-site natural gas combustion, propane burning for initial salt melting, and auxiliary grid electricity consumption. • Dismantling: energy consumed to disassemble major tower plant systems. • Disposal: landfilling, incineration, or recycling of demolition waste. No credit is taken for displacing virgin materials through the recycling process. SimaPro v7.214 LCA modeling software is used along with the Ecoinvent 2.0 life cycle inventory database.15 Engineering judgment was used to match specified plant construction
EPBT =
CEDtot αEnet
where CEDtot is the life cycle CED of the plant in MJeq, Enet is the annual net electricity generated by the plant converted to megajoules per year (MJ/yr), and the dimensionless constant α is the average ratio of source energy to site energy for U.S. grid electricity. α is used to ensure that the numerator and denominator are equivalently analyzed as primary energy. The U.S. Environmental Protection Agency uses a value of 3.34 for α for grid-purchased electricity.19 CSP Power Tower Reference Design and Data Sources. The conceptual design of the reference tower CSP plant is based on previous research from Sandia National Laboratories on molten salt tower plant designs7,20 coupled with operational plant modeling work conducted by NREL. The NREL operational modeling was conducted using the publicly available B
dx.doi.org/10.1021/es400821x | Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Environmental Science & Technology
Article
System Advisor Model,21 which is designed as a performance and financial modeling decision support platform for the renewable energy industry. With data from the Sandia National Laboratories research and the NREL operational modeling as the basis, NREL contracted with WorleyParsons Group, Inc. (WPG) to develop a detailed material, process, and cost inventory for the manufacturing and operation of all major systems, excluding the heliostat field, for a hypothetical tower CSP plant.22 The materials and processes associated with the manufacturing and O&M of the heliostat field were separately evaluated by the authors23 with heliostat foundation material estimates adapted from Lechon et al.’s evaluation of solar thermal technologies in Spain.24 The life cycle inventory from mining nitrate salts and producing synthetic nitrate salts is based on analysis reported in the trough LCA. Estimates for construction and dismantling energy consumption between the trough and tower plant designs are scaled based on relative plant capacity using the trough LCA as the base.2 The dry-cooled plant is located near Tucson, AZ as the Southwest U.S. is an optimal location for CSP plants given the high levels of DNI and ability to connect to the western utility grid. The HTF used in the reference 106 MWnet tower design is a molten nitrate salt mixture comprising 60 wt % sodium nitrate salt and 40 wt % potassium nitrate salt, also known as solar salt. In the reference plant’s 2-tank design, solar salt is pumped from a cold storage tank to the receiver where it absorbs sunlight, heats from 288 to 565 °C, and then flows back to a ground-level, hot storage tank. From the hot storage tank, the hot solar salt is pumped to the steam generating system that produces superheated steam for the turbine that generates electricity. The salt is then returned to the cold storage tank for reuse in the receiver system. An auxiliary boiler provides steam during plant startup and overnight to maintain steam turbine seals and condenser vacuum. The auxiliary boiler is not used for supplemental power generation. The reference plant has sufficient storage for six hours of full-capacity power generation with an annual capacity factor of approximately 42%. Six hours of storage is consistent with a tower CSP plant designed to meet afternoon and evening peak grid demand.25 Table 1 summarizes design specifications for the reference tower design.
The major systems that comprise the reference tower CSP plant design (Figure 1) are described briefly below, with details reported in the SI and ref 22: • Site Improvement: activities that prepare the land for plant construction. Site improvement activities include clearing and grading the site, installing the water supply infrastructure, and establishing the road system, parking facilities, and fences required for the power plant complex. • Collector System: the collector system consists of 6682 two-axis tracking Advanced Thermal System (ATS) heliostats and the control systems that operate the heliostat field.23 Each heliostat has an aperture area of 148 m2. • Tower/Receiver System (Receiver): the receiver system consists of a cylindrical tube wall heat exchanger (the “receiver”), pumps to circulate the solar salt, and inlet and outlet vessels, as well as the tower itself, and the insulated salt piping running up and down the tower. • Steam Generation System: the thermal energy from the hot solar salt is used by the steam generation system to produce superheated steam (with a reheater) for the turbine-generator to generate power. • Thermal Energy Storage System: the two-tank thermal energy storage system stores the high temperature solar salt from the receiver and the low temperature solar salt from the steam generation system in separate tanks connected by insulated piping and pumps. • Electric Power Generation System: the electric power generation system consists of the turbine-generator, feedwater systems, steam condenser, and pumps required to convert the energy in the steam to electric power for grid delivery. It also includes balance of plant buildings and mechanical and electrical systems that are associated with the power block and not included in other major systems. Design Alternatives. In addition to the reference plant, this study also evaluates three primary design alternatives that can be incorporated individually or in combination. The design alternatives include the following: (1) Synthetic salts: mined nitrate salts from the reference design are replaced with synthetically produced nitrate salts. With this alternative, salt sourcing and production processes change but the amount of salts used remains the same. This alternative was also explored in the trough LCA.2 (2) Natural gas-fired auxiliary boiler: the electric-powered auxiliary boiler of the reference design is replaced with a natural gas-fired boiler resulting in increased natural gas consumption and decreased grid electricity consumption. (3) Thermocline: the two-tank system for hot and cold nitrate salt storage from the reference design is replaced with a thermocline system that maintains a heat gradient in one tank. The main modifications are that the cold salt tank is eliminated and the amount of nitrate salt media is reduced by approximately two-thirds. This alternative was presented in the trough LCA2 which reports further details. The Supporting Information contains the authors’ visual interpretation of the plant and key systems and subsystems (Figures S1−S4) and reports additional information regarding the inputs used in the LCA model (Tables S1−S14).
Table 1. Specifications of Reference Power Tower Concentrating Solar Power Plant Design
a
parameter
value
units
gross capacity parasitics (at design point) net capacity net annual generation capacity factor availability grid electricity consumption natural gas consumptiona annual water consumption number of heliostats total aperture area HTF mass TES storage capacity tower height total land area
115 9 106 378,463,439 41.7 96 7,920 0 71,934 6,682 964,712 17,418 6 172 6,345,471
MW MW MW kWh % % MWh/yr MMBtu/yr m3 heliostats m2 tonnes hours m m2
The reference case does not use natural gas. C
dx.doi.org/10.1021/es400821x | Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Environmental Science & Technology
Article
Figure 1. Schematic of the major systems and components of the reference CSP tower design evaluated here (illustration adapted from ref 7).
Table 2. Life Cycle GHG Emissions, Water Consumption, CED, and EPBT by Life Cycle Phase for the Reference Plant and Alternative Design Scenariosa Reference Guide scenario scenario description
ref. plant reference plant design
alt. 1 synthetic salt
alt. 2
alt. 3
alt. 4
natural gas boiler
thermocline
synthetic salt + natural gas boiler
alt. 5 thermocline + synthetic salt
alt. 6
alt. 7
thermocline + natural gas boiler
thermocline + synthetic salt + natural gas boiler
GHG Emissions (g CO2eq/kWh)
a
life cycle phase
ref. plant
alt. 1
alt. 2
manufacturing construction O&M dismantling disposal grand total
14 2.4 17 0.38 4.0 37
18 2.3 17 0.38 4.0 42
14 14 2.4 2.2 13 17 0.38 0.38 4.0 3.4 33 36 Water Consumption (L/kWh)
alt. 3
life cycle phase
ref. plant
alt. 1
alt. 2
manufacturing construction O&M dismantling disposal grand total
0.43 0.0239 1.0 0.0016 0.008 1.4
0.49 0.0224 1.0 0.0016 0.008 1.5
0.43 0.0239 0.7 0.0016 0.008 1.2
life cycle phase
ref. plant
alt. 1
alt. 2
alt. 3
alt. 4
alt. 5
alt. 6
alt. 7
manufacturing construction O&M dismantling disposal grand total EPBT (months)
0.17 0.018 0.29 0.0056 0.0066 0.49 15
0.20 0.018 0.29 0.0056 0.0066 0.52 16
0.17 0.018 0.23 0.0056 0.0066 0.42 13
0.17 0.018 0.29 0.0056 0.0069 0.49 15
0.20 0.018 0.23 0.0056 0.0066 0.46 14
0.18 0.018 0.29 0.0056 0.0069 0.50 15
0.17 0.018 0.23 0.0056 0.0069 0.42 13
0.18 0.018 0.23 0.0056 0.0069 0.43 13
alt. 3 0.42 0.0221 1.0 0.0016 0.007 1.4 CED (MJeq/kWh)
alt. 4
alt. 5
alt. 6
alt. 7
18 2.3 13 0.38 4.0 38
15 2.2 17 0.38 3.4 38
14 2.2 13 0.38 3.4 32
15 2.2 13 0.38 3.4 34
alt. 4
alt. 5
alt. 6
alt. 7
0.49 0.0224 0.7 0.0016 0.008 1.3
0.44 0.0217 1.0 0.0016 0.007 1.4
0.42 0.0221 0.7 0.0016 0.007 1.2
0.44 0.0217 0.7 0.0016 0.007 1.2
GHG = greenhouse gas, CED = cumulative energy demand, EPBT = energy payback time.
■
Reference Plant. The reference tower plant is estimated to generate normalized life cycle GHG emissions of 37 g CO2eq/ kWh. Carbon dioxide accounts for 94% of CO2-equivalent GHG emissions with methane contributing 5% and nitrous oxide and other gases contributing less than 1% (Table S18). The O&M
RESULTS AND DISCUSSION
Table 2 reports the GHG emissions, water consumption, CED, and EPBT for the reference plant and each alternative configuration disaggregated by life cycle phase, with additional details reported in Tables S15−18. D
dx.doi.org/10.1021/es400821x | Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Environmental Science & Technology
Article
reference plant value. Alternative 1 reports the change in impacts due solely to switching from mined solar salt to synthetic solar salt without changing the thermal energy storage system or auxiliary boiler fuel. Substituting synthetic solar salt increases life cycle GHG emissions by 12% from 37 to 42 g CO2eq/kWh. Water consumption also increases from the switch to synthetic solar salt, but the increase is only 4% (1.5 vs 1.4 L/kWh). The switch to synthetic salt increases CED by 7% from 0.49 to 0.52 MJeq/kWh. The 7% increase in CED extends the EPBT by approximately 1 month to 16 months. Alternative Design: Auxiliary Boiler Power Source. Auxiliary grid electricity is the largest contributor to O&M GHG emissions, water consumption, and CED for the reference plant design. As the reference plant’s auxiliary electric boiler consumed 40% of the total auxiliary electricity, an alternative design was evaluated to test the impacts of switching the boiler fuel from electricity to natural gas. Table S19 outlines the auxiliary energy consumption by fuel source for both the base case with an electric boiler and the alternative cases that assume a natural gas boiler. The switch to natural gas decreases life cycle GHG emissions by 11% to 33 g CO2eq/kWh, water consumption by 16% to 1.2 L/ kWh, CED by 13% to 0.42 MJeq/kWh, and EPBT by 2 months to 13 months (Table 2, alternative 2). Based on the life cycle inventory inputs used in this analysis, a switch from U.S. average grid electricity to natural gas for auxiliary boiler operation provides environmental benefits for all metrics evaluated. It should be noted, however, that upstream environmental impacts of natural gas are a topic of significant current interest,26 with the potential for the understanding of burdens to change as the results of new research are published. Alternative Design: Thermal Energy Storage System Configuration. Design alternatives for configuration of the thermal energy storage system are two-tank (reference) or thermocline.27 In the thermocline design, the molten salt is stored in a single tank that is filled with a porous, solid medium. As salt flows through the low-cost media, thermal energy is transferred to or from the solid. For this analysis, the solid medium is assumed to be silica sand. High-temperature molten salt flows in and out of the top of the tank with low-temperature molten salt flowing in and out of the bottom. The design advantages of the thermocline system include the material and cost savings from the elimination of one storage tank and an approximate two-thirds reduction in the mass of salt media required for the system. However, a thermocline thermal energy storage system for a tower CSP plant has yet to be deployed and tested at utility scale. Designs must overcome the increased stresses from the thermal variation in the tank. In addition, the problem of thermal ratcheting resulting from settling of the fill material may present excessive stress on the tank wall.28 Alternative 3 in Table 2 displays results for GHG emissions, water consumption, CED, and EPBT for switching the reference design to use a thermocline. The switch to a thermocline system increases the total mass of storage media (solar salt plus silica sand) by approximately 47%. The net effect of switching to a thermocline system while maintaining the use of mined salts is insignificant (less than 2% decrease) across all environmental parameters as the impacts from the increased total storage media and subsequent disposal requirements mitigate the benefits associated with reducing the mass of mined salt and eliminating the cold storage tank. Reductions in GHG emissions, water consumption, and CED were more significant for the trough LCA system (−7%, −2%, and −7%, respectively)2 due to the
phase is the largest contributor to reference plant GHG emissions at 45% followed by the manufacturing phase at 38% and disposal at 11%. Within O&M, approximately 97% of GHG emissions are from auxiliary electricity consumption supplied by the grid, i.e., electricity required at times of no power generation. (Table S19 outlines the distribution of auxiliary electricity and natural gas to major plant systems.) Total O&M grid electricity consumption is estimated at 7920 MWh/yr with the electric auxiliary boiler consuming 40%. Note that this amount of electrical demand depends on the reasonable assumption that the temperature of the salt is allowed to decay overnight, despite implication to the contrary in the initial design (see appendix to ref 18). The analysis uses the Ecoinvent 2.015 U.S. average grid mix for the electricity supply with a life cycle GHG emission factor of 775 g CO2eq/kWh. Deviations in assumed GHG emission intensity of the plant’s auxiliary electricity supply will result in approximately proportional adjustments to the tower plant’s O&M GHG emissions. For the manufacturing GHG emissions, the collector system is the primary contributor representing 54%, in line with its 50% contribution to total plant material mass. The electric power generation system represents only 8% of the plant material mass but contributes 18% of manufacturing GHG emissions due to the presence of several highly manufactured components such as turbines and transformers. Refer to the SI Tables S16 and S17 for additional breakdowns of percent contribution to manufacturing and O&M phase GHG emissions, water consumption, and CED by major plant system. Normalized life cycle water consumption for the reference plant is estimated at 1.4 L/kWh. The two phases that primarily contribute to water consumption are O&M (67%) and manufacturing (30%). Within O&M, 59% of water consumption is attributed to the supply of auxiliary grid electricity, 22% to ongoing steam generation system water requirements, and 18% to heliostat mirror washing. The life cycle water consumption intensity for the U.S. electricity grid mix used in this analysis is 27 L/kWh.15 For the manufacturing phase, the collector system, owing to high material requirements, and the electric power generation system, owing to many highly manufactured components, are the primary contributors to water consumption at 51% and 20%, respectively. Normalized CED for the reference plant is estimated at 0.49 MJeq/kWh with an EPBT of 15 months. The two phases that primarily contribute to CED are O&M at 59% and manufacturing at 35%. Within O&M, approximately 94% of CED is attributed to the auxiliary electricity supply with 5% attributed to the operation of heliostat water wash trucks and the provision of water required to maintain the collector system. The life cycle CED intensity of the auxiliary U.S. grid electricity supply analyzed in this study is approximately 13 MJeq/kWh.15 Alternative Design: Salt Type. The reference plant design assumes the use of mined sodium nitrate and potassium nitrate salts. The SI of the trough LCA2 discusses the role of mined sodium nitrate and potassium nitrate salts in the marketplace along with production processes used to generate synthetic salt as an alternative design. This study uses the same source data and calculation methods as the trough LCA2 to estimate life cycle GHG emissions, water consumption, and CED impacts of mined and synthetic solar salt on a per metric ton basis (Table S7). For O&M calculations, salt replacement is assumed to occur at a rate of 0.1% per year.22 The life cycle environmental impacts of the synthetic salt alternative designs are reported in Table 2 along with the E
dx.doi.org/10.1021/es400821x | Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Environmental Science & Technology
Article
greater initial solar salt requirements of trough as compared to tower CSP designs. Combined Alternative Designs. In addition to evaluating the alternative designs in isolation, alternative design combinations were also evaluated. The results are reported in Table 2 under alternative 4 (synthetic solar salt and natural gas boiler), alternative 5 (thermocline and synthetic solar salt), alternative 6 (thermocline and natural gas boiler), and alternative 7 (thermocline, synthetic solar salt, and natural gas boiler). Compared to the reference plant, alternative 4 is estimated to increase life cycle GHG emissions by 2%, decrease water consumption by 12%, decrease CED by 6%, and shorten EPBT by 1 month as the increased impacts of synthetic solar salt production are mitigated by the environmental benefits of the boiler fuel switch. Alternative 5 demonstrates how the thermocline design, by reducing the solar salt requirement, helps to mitigate the impacts that would otherwise occur from the switch to synthetic solar salt. Compared with the two-tank synthetic salt case (alternative 1), GHG emissions are decreased by 9% with water consumption decreased by 4%, CED decreased by 5%, and EPBT shorted by 2 months. Alternatives 6 and 7 both result in significant environmental benefits compared to the reference plant due to the fuel switching to natural gas and mitigation of solar salt impacts with the thermocline design. For alternative 6, life cycle GHG emissions compared to the reference case are decreased by 13%, water consumption by 17%, CED by 13%, and EPBT by 2 months. Similarly, for alternative 7, evaluated impacts are lower than the reference plant for all parameters, with life cycle GHG emissions decreased by 9%, water consumption by 16%, CED by 11%, and EPBT by 2 months. The design that simultaneously minimizes GHG emissions, water consumption, CED, and EPBT is alternative 6 (thermocline thermal energy storage system, mined solar salt, and natural gas boiler). However, all alternatives yield results across the analyzed parameters that are in a relatively tight range with less than 20% variability from the reference plant. Comparison of Reference Plant Results to Other Technologies. It is informative for future energy deployment decisions to benchmark the scale of impacts estimated in this study against other categories of power generation technologies. This section makes comparisons of the tower reference plant design to other power generation technologies at a high level; interested readers can use the references cited herein for more detailed comparisons. Table 3 reports life cycle GHG emission and on-site operational water consumption estimates for various power generation technologies based on systematic review and meta-analysis. The GHG emission estimates reported in Table 3 represent median values from meta-analysis studies that considered utility-scale systems of various designs and locations within broad technology categories: CSP technologies,30 crystalline silicon photovoltaics31 (PV), wind power,32 nuclear power,33 coal,29 and natural gas.26 The meta-analysis studies harmonized the GHG emission results to account for differences in system boundary assumptions, site specific data, and system design alternatives. Detailed discussions of the impacts of each key parameter on the results and the range of published and harmonized results across the literature for each technology are available in those references. Life cycle GHG emissions for the reference tower plant and design alternatives evaluated herein are of the same order of magnitude as those for PV, wind, and nuclear as well as those reported in the broader CSP literature with all reporting median harmonized GHG emissions of less than 50 g CO2eq/kWh. In
Table 3. Comparison of Central Estimates of Life Cycle GHG Emissions and On-Site Operational Water Consumption for Various Electricity Generation Technologiesa life cycle GHG emissionsb (g CO2eq/kWh)
technology this study benchmark CSP tower - dry cooled CSP trough - dry cooled crystalline silicon photovoltaics wind nuclear coal natural gas combined cycle
on-site operational water consumptionc (L/kWh)
37 38
0.34 0.10
22
0.30
44
