Life cycle assessment of hydrogen production from a high temperature electrolysis process coupled to a high temperature gas nuclear reactor

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 4 0 1 9 e4 0 3 3

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/he

Life cycle assessment of hydrogen production from a high temperature electrolysis process coupled to a high temperature gas nuclear reactor Mario R. Giraldi, Juan-Luis Franc¸ois*, Cecilia Martin-del-Campo Departamento de Sistemas Energeticos, Facultad de Ingenierı´a, Universidad Nacional Autonoma de Mexico, Paseo huac No. 8532, Col. Progreso, C.P. 62550 Jiutepec, Morelos, Mexico Cuauhna

article info

abstract

Article history:

The life cycle analysis (LCA) is a versatile tool to evaluate process and production systems,

Received 20 April 2014

and is useful to compare environmental burdens. For the purposes of this LCA, a high

Received in revised form

temperature electrolysis process was coupled to a high temperature gas nuclear reactor.

12 January 2015

The system function is the production of hydrogen using electricity and heat from nuclear

Accepted 17 January 2015

power, with a functional unit of 1 kg of hydrogen, at the plant gate. The product system

Available online 10 February 2015

consists of the following steps: (i) the extraction and manufacturing of raw materials (upstream flows), (ii) the electrolytic cell fabrication, (iii) the nuclear fuel cycle, and, (iv) the

Keywords:

hydrogen production plant. Particular attention was paid to those processes where there

Hydrogen production

was limited information available on inventory data, for example mining and processing of

LCA

rare earth metals, and electrolytic cell assembly, which are the primary components of a

GHG

hydrogen generation plant. The environmental impact assessment focuses on the emis-

Nuclear energy chain

sions of greenhouse gases (GHGs), as related to global warming. Additionally, other envi-

High temperature electrolysis

ronmental loads, to complete the environmental profile of the product system, were

High temperature gas reactor

included. The results were low GHGs emissions, with a value of 416 g of CO2eq kg-1H2. As to the process components, the electrolytic cell showed the highest environmental impact. Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

Introduction Hydrogen blended with other elements is very abundant on the Earth; as is the case of oxygen to make up water molecules, of carbon into hydrocarbon molecules; and to a lesser extent as a gas (the lightest of all), mixed with gases in the atmosphere. The water molecule, a natural and massive source of hydrogen, is an abundant resource in sea water;

however, great amounts of energy are required to split its molecule. Hydrogen is a secondary carrier of energy that, unlike electricity, once separated from composed molecules, can be stored in large quantities for long periods of time. One of the benefits is its reversible quality, to return its stored chemical energy back into electrical energy by means of a fuel cell. There are several emerging technologies for hydrogen production, mainly based on renewable sources and nuclear energy. Most of them are still being defined to reach

* Corresponding author. E-mail addresses: [email protected] (M.R. Giraldi), [email protected] (J.-L. Franc¸ois), cecilia.martin.del. [email protected] (C. Martin-del-Campo). http://dx.doi.org/10.1016/j.ijhydene.2015.01.093 0360-3199/Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

4020

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 4 0 1 9 e4 0 3 3

conclusive results. Currently, there are four different conventional ways of producing hydrogen: (i) from natural gas through steam reforming, (ii) from processing oil (catalytic cracking), (iii) from coal gasification and, (iv) from electrolysis using different energy mixes. The latter, the electrolysis technologies, can be categorized into three main types: alkaline electrolysers, proton exchange membrane (PEM) electrolysers, and high temperature electrolysers (HTE). Hydrogen used in vehicles, with fuel cell technology, represents a promising future for the hydrogen economy that mitigates climate change. However, there are challenges to be solved before its intensive use, such as: (i) the cost of the fuel cell, (ii) the method of storing hydrogen on board the vehicle to ensure an adequate cruising range, due to weight, (iii) the creation of hydrogen distribution infrastructure and, (iv) hydrogen production that is energetically efficient. This last point focuses on the possibility of large-scale hydrogen production for the transport sectorwhich is a measure of global warming mitigation. The life cycle analysis (LCA) is a useful tool to quantify environmental burdens in the production chain, defined as “cradle to grave”. This tool facilitates the systematic evaluation of the environmental impacts from new products, processes, and activities. For nuclear power, free of carbon dioxide emissions, hydrogen can potentially be produced on an industrial scale from: (i) high temperature electrolysis (HTE) of water from electricity generated by means of nuclear power; (ii) hightemperature electrolysis of steam (HTES) from a mix of heat and electricity generated by means of nuclear power and, (iii) thermochemical splitting of water from heat produced by means of nuclear power, or by both nuclear heat and electrical power. Literature, reports different LCA studies in the field of hydrogen production. These take into account the impact of the source of energy supply, whose results concerning the electrolytic process are shown in Section 3.4.3. In the case of production from nuclear energy, studies have been performed for different hydrogen production methods:


Utgikar [1] studied a nuclear-high temperature electrolysis plant, whose approach was to supply energy, both heat and electricity, to the hydrogen plant from a nuclear reactor.
Ozbilen [2] analyzed the CueCl thermochemical cycle, which highlighted four scenarios related to the supply of energy for hydrogen production and primary inputs.
Solli [3], Lattin [4] and Giraldi [5] separately assessed the SeI thermochemical cycle, where several scenarios based on the production of hydrogen were analyzed. The results showed variations in the magnitude of GHGs emissions depending on the configuration process technology, and the change of the source of external power supply.
Patyk [6] recently assessed the generation of hydrogen from the HTE process from two energy sources: nuclear and wind; He studied various production scenarios. The following observations are highlighted comparing Patyk's findings with our results: (i) conventional prototype water pressure reactor as a nuclear energy source was selected, (ii) it did not use the Brayton cycle thermal analysis, while on the current study it was included, and (iii) the exchange between electric and thermal energy flows were not considered by Patyk. This work assessed a steam electrolysis process with a nuclear energy supply. The temperature of the steam is heated to temperatures above 1000 K in this process. Hightemperature electrolytic water-splitting using electricity from nuclear to reach the temperature required for hydrogen production was used.; At higher reactor outlet temperatures, the power cycle efficiency increases, thereby increasing hydrogen production efficiency. Gas helium-turbine prototypes with a recuperator and an intermediate cooling have demonstrated thermal efficiency at levels greater than 40%. Although still under development, the use of a direct closed gaseturbine cycle and a modular reactor showed reductions of costs (capital, operation, and maintenance) due to the simplification of the power generation cycle and safety systems [7,8].

Fig. 1 e Procedure for the development of the LCA.

4021

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 4 0 1 9 e4 0 3 3

The principle objective of this study is to quantify the GHG emissions associated to the described hydrogen product system, through the following specific objectives: (i) to develop a database with the quantification of the inputs and outputs of the product system, (ii) to identify processes responsible for the generation of GHG emission, (iii) to perform the environmental profile of the product system with additional impact categories and (iv) to compare our results with other LCA studies on hydrogen production using the electrolysis process and nuclear energy.

concluded with the LCI. The scope of the product system can be limited to three boundaries: cradle to grave of production, cradle to gate which concludes when the product is ready for use and, gate to gate which only includes the production process assessed. The product system can be made up of different intermediate flows, co-products and effluents. In order to assign the corresponding environmental impacts, the load allocation procedure was applied. The ISO establishes separate inputs and outputs of the system, between its different functions and products, that reflect the physical causality that links them.

Methodology

Life cycle inventory analysis

The methodology for this LCA was based on ISO standards e 14040 and 14044 [9,10], as well as references given by Guine [11], and the guide from the USEPA [12]. As stated by Giraldi [5] four main steps are involved in the methodology: (i) initial phase, (ii) life cycle inventory (LCI), (iii) life cycle impact assessment (LCIA), and (iv) life cycle interpretation. Fig. 1 shows the procedure applied in this study.

This is an iterative process in which process data, once validated, are linked up to the functional unit for the calculation of the common variables to complete the database. In the data validation, is included the following: data collection and capture, and adequate data with the material and energy balances for each process stage.

Evaluation of the life cycle impact Initial phase: defining the scope and purpose This step establishes what is to be achieved with the LCA in other words the setting of the objectives of the study. Subsequently, three concepts are defined: the functions of the system, the functional unit, and the limits of the system under study with the corresponding criteria that define them. The LCA should account for all inputs to the system, including upstream processing, and the corresponding analysis completed by computing the outputs; they include use and end-of -life of the product. The achievement of this stage is

Characterization or the use of different models proposed for the impact categories result in conversion factors that are applied to the data from LCI. The value obtained is defined as an impact category indicator. The primary environmental impact is on climate change. The global warming potential (GWP) parameter was used to measure the immediate point in the cause-effect chain of the environmental mechanism (midpoint effect), and can be described as the perturbation of the planetary energy balance by a climate change mechanism [13]. The term CO2-equivalent indicates the amount of

Table 1 e Environmental loads assessed in the actual LCA. Impact categories Climate change Human toxicity

Ionising radiation

Ozone depletion

Water depletion Metal depletion Fossil depletion

Cumulative energy demand

Concept

Factor

See description in the text A toxic effect is an alteration of the structure or function of one species, humans in this case, as a result of exposure to a chemical. The factors for determining it can be interpreted in terms of risk equivalents empirical models based on historical data of real exposure in different cases. This category covers the impacts arising from releases of radioactive substances as well as direct exposure to radiation. It takes into account the emissions of ionising radiation to air, water and soil, and calculation of their radiation behavior and burden, based on detailed nuclear-physical knowledge. Stratospheric ozone depletion refers to the thinning of the stratospheric ozone layer as a result of anthropogenic emissions. The Ozone Depletion Potential (ODP) is defined as amount of ozone destroyed by emission of a gas over the entire atmospheric lifetime (i.e. at steady state) relative to the same mass of CFC-11 (trichlorofluoromethane) equivalent emissions. The model uses the concept of natural resources depletion, primarily include non-renewable resources, well as certain renewable resources such as water, which is scarce in certain geographical areas. The calculation considers the ratio of the material used and the total quantity of the resource reserve. The depletion factor is expressed relative to the equivalent of available reserves regarding reference material. It assesses the energy use throughout the life cycle of a good or a service. This includes the direct uses as well as the indirect or grey consumption of energy due to the use of, e.g. construction materials or raw materials (this procedure is not included in the RECIPE).

kg C02 eq kg 1,4 DB eq

kg U235 eq

kg CFC-11 eq

m3 kg Fe eq kg oil eq

MJ eq

4022

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 4 0 1 9 e4 0 3 3

emissions, that when weighted by their respective GWP, lead to an equal climate response [14]. We have added additional environmental impacts to complete the environmental impact profile of the product system based on the RECIPE methodology. This methodology was developed by RIVM (Rijksinstituut voor Volksgezondheid en Milieu), Radboud University, CML (Centrum Milieukunde  Consultants [15]. The purpose was to Leiden) and PRe harmonize previous Dutch methodologies CML 2001 and Ecoindicator 99. However, this study is not intended to focus on these impacts; only a short description of each is presented in Table 1.

Life cycle interpretation Environmental impacts from outputs are analyzed by processes and compared with other LCA studies. As part of the life cycle analysis, three aspects were evaluated: (i) the overall performance of the product system and its characterization, (ii) the system response to changes in energy efficiency, as defined in section 3.4, and (iii) the comparison of the obtained results regarding those collected from existing literature.

LCA of hydrogen production


The LCI was compiled and quantified taking into account the following stages: (i) extraction, mining, milling, and pre-manufacturing of raw materials, (ii) upstream energy flow supplied to the system, (iii) nuclear fuel cycle stages of power generation and, (iv) a hydrogen production plant based on solid-oxide cells technology.
In the hydrogen plant, priority was given to the energy analysis for the evaluation process, since the synergy that supports the electrolysis processes are: (i) the Brayton cycle, (ii) the heat exchange system, and (iii) the energy recovery between processes.
This study was focused on data provided by specialized databases such as Ecoinvent [16] as well as on public databases. Background information is compiled based on information provided in general literature. The software tool used as an interface between the LCI and LCIA was the Simapro® (v.7.3) software. The allocation procedures for multiple streams, coproducts and products, as per the ISO 14044 (2006), were applied in this study. In this case, allocation is based on energy balances from each production stream including intermediate flows. First, calculations of energy, obtained from hydrogen production, were made. Second, their shared energies, based on the energy flow content, that were calculated as explained above, were determined. Then, the emission streams were allotted to each of the intermediate outputs, products and coproducts based on their shared energies.

Initial phase Description of the product system Scope and goal The scope of this study was to quantify the GHG emissions associated to hydrogen production by high-temperature electrolytic water-splitting coupled to a high temperature nuclear reactor, and to compare the results with other LCA studies on hydrogen production technologies. Considering hydrogen as an energy alternative to fossil fuels in the framework of climate change mitigation, the following specific objectives were set:
To develop a database with the quantification of the inputs and outputs of the product system.
To identify processes responsible for the generation of GHG emission.
To perform the environmental profile of the product system with additional impact categories.
To compare our results with other LCA studies on hydrogen production using the electrolysis process and nuclear energy.

Functional unit The functional unit (FU) was defined as 1 kg of hydrogen. The system function was set as the production of hydrogen at the gate of the plant. Use and end-of-life of the product were not included.

The life cycle study is made up of two main systems: (i) the nuclear power plant, including the fuel cycle, and (ii) the hydrogen production plant. Fig. 2 shows a general diagram of the system assessed by this LCA, which is described below.

Nuclear fuel cycle and reactor The evaluation of the nuclear fuel cycle was approached in three phases: (i) the nuclear fuel production, (ii) the nuclear reactor, and (iii) the helium cooling system. The heat from the nuclear reactor is supplied to the electrolyzer by a Brayton cycle, and the helium cooling system will be discussed in section 3.2.2.1 together with the hydrogen plant. The fuel nuclear production included the following stages: (i) mining and production of uranium concentrates, (ii) conversion and enrichment, (iii) Tri-Isotropic (TRISO) coated nuclear fuel particle fabrication. The TRISO is configured on a hexagonal graphite assembly into the reactor. This consists of a High Temperature Gas-Cooled Reactor (HTGR), whose combination of helium cooling and graphite moderator makes the production of a high temperature nuclear heat, capable of supplying enough energy to make steam electrolysis possible. The configuration and quantification of the nuclear fuel cycle and the nuclear plant was based on adaptations and updates from Giraldi et al. [5].

Hydrogen production plant System boundaries and allocation The boundaries for the product system were based on a cradle-to-gate principle according to the following criteria:

The hydrogen process includes: (i) the thermal energy from HTGR coupled to the direct helium recuperated from the Brayton power cycle; (ii) the electrolyzer, as HTE core; (iii) the

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 4 0 1 9 e4 0 3 3

4023

Fig. 2 e Diagram of the product system.

sweep air supply system; and (iv) the hydrogen recuperator. These steps are described in the next sections. Large-scale hydrogen production via HTE from nuclear energy was documented in literature, principally [17e19]. The processes defined in the literature were adapted and replicated according to the procedure outlined in the initial phase of the current LCA; therefore, the data entered into the inventory were supplied by our own calculations.

Brayton power cycle. The HTGR provides both power and heat to drive the HTE process. As mentioned above, thermal energy is transferred to the Brayton power cycle (BPC) through a secondary helium loop that isolates the reactor plant system. The helium coolant leaves the reactor at 1173 K that allows for a split of 80e90% of the flow towards the power cycle. The remainder is routed to the intermediate heat exchanger (IHE) to provide thermal energy to the electrolyzer. This bypassed minor helium flow, once it has been cooled, is returned through a circulator, where the helium is compressed and mixed with the primary coolant returning to the reactor. The helium in the BPC loop flows through the power turbine to generate electric power, where its pressure and temperature drop. To recover the heat contained in the flow, helium is passed through a recuperator and a pre-cooler. Afterwards, the helium flows through a double compression process with an intermediate cooling heat recovery; this process improves the compression efficiencies, and also provides the driving force needed to circulate it back through the recuperator. Here, temperature is raised before it is mixed with another flow that is returned from the IHE. The loop is

completed when the helium flow is returned to the reactor, as mentioned above. The flows of the BPC and the hydrogen plant are shown in Fig. 3; the quantifiable outputs of energy recovered were noted in this diagram.

Electrolyzer. In this study, the electrolyzer is made up of two components: (i) the IHE; and (ii) the planar solid-oxide electrolytic cell (SOEC). Prior to entry at the IHE, a fraction of the gas produced is recycled and mixed with the inlet steam, in order to assure the reducing conditions are maintained in the feed flow to the inlet of the electrolyzer. This mixture is heated up to the electrolysis operating temperature found in the IHE coming from the BPC loop. Typically, high-temperature thermal energy is used to maintain the heat required for isothermal electrolyzer operating conditions; it is supplied from energy recovered in the BPC loop. In this manner, isothermal and adiabatic conditions were used as thermal boundary limits for calculation of the overall energy balance of the system. When the steam/hydrogen flow is fed to the electrolyzer, the water-splitting process occurs. It corresponds to the dissociation or reduction of water, whose overall electrochemical reaction is stated in Eq. (1). H2 O / H2 þ

1 O2 2

(1)

The mixture of steam and hydrogen at ~900  C is supplied to the cathode side of the electrolyte. The half-cell electrochemical reactions occur at the triple-phase boundary near the electrode-electrolyte interface. Oxygen ions are then drawn through the electrolyte by an electrochemical

4024

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 4 0 1 9 e4 0 3 3

Fig. 3 e Diagram of energy streams analyzed within the hydrogen plant.

Fig. 4 e Assessed production steps in the SOEC assembly.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 4 0 1 9 e4 0 3 3

potential. These ions liberate their electrons and recombine to form molecular O2 on the anode.

Sweep air supply system. During the electrolysis process, air is used as sweep gas to remove the oxygen generated on the anode side of the electrolyzer. This inlet air is compressed through an operating system of four-stage compressors; between each of these, the energy is recovered by intercoolers. The air sweep gas (ASG) leaving this system has a temperature and a pressure of about 170  C and 5 MPa respectively. Then, the ASG is heated up to the electrolyzer operating temperature of 800  C by the IHE which supplies nuclear energy directly to the system from the Brayton cycle stage. Hydrogen separation stage. The hydrogen produced and the air-oxygen gas streams are routed through a heat exchanger, also known as an electrolysis heat recuperator (EHR), which recovers energy and reduces the gas temperatures. This heat is then transferred to the feed water flow, which is converted, by boiling and superheating, into steam with output temperatures of about 350  C. It is then routed to the electrolyzer stage. The reduced temperature sweep gas, which can still do the work of compression, is recovered using a turbine situated at the sweep-gas exit. The cooled stream of hydrogen from the EHR, whose composition is 90% hydrogen 10% steam on a mole basis, condenses the water. This flow is routed to a separation tank, where water is collected for recycling, while the hydrogen stream is split into two flows. The first stream is the hydrogen product output. The second stream is hydrogen which is recycled back into the process to blend with the steam flow, in order to maintain reducing conditions at the electrolizer stage. Electrolytic cell manufacture The analysis of the product system of the electrolytic cell (in this study the above-mentioned SOEC) was organized as follows: (i) the cell active materials, (ii) the SOEC assembly, and (iii) their associated energy supplies (see Fig. 4).

4025

This section is based on reports from literature documenting cell manufacturing and component materials [20e29].

Cell active materials. The SOEC are regarded as high temperature electrochemical devices operating at over 800  C, which use non-combustive oxidation methods to split water into hydrogen and oxygen. The SOEC can be made up of two different geometries, tubular and planar; the latter was selected for this study. The cells are comprised of different alloys distributed as follows (i) the anode: consists of a combination of a ceramic (cer) and a metal (met) named nickel-YSZ cermet (yttria stabilized zirconia); (ii) the electrolyte: YSZ dense layer ceramic; and (iii) the cathode: thin porous layer of LSM (lanthanum strontium manganite) ceramic. For producing the three electrodes, materials, such as a nickel, zirconium, lanthanum or manganese, are available in the manufacturing inventory (e.g. Ecoinvent). In the case of yttrium (or its commercial form, yttrium oxide) insufficient documentation is available. Therefore, we documented the production process of yttrium oxide, which was used in the LCI stage; the process is described below. Yttrium oxide. The yttrium is within the group of lanthanides or rare earths. Its chemical form, yttrium oxide is commonly traded for industrial use. The mineral, xenotime, has been one of its sources, as are, some clays found in China. For this study xenotime as an yttrium ore source was selected. This ore is a rare-earth phosphate (YPO4) that contains 60% rare earth oxides, including yttrium oxide (Y2O3) [30]. The ore is processed in a high tension separator, and then in an induced roll magnetic separator, thus a high grade xenotime concentrate is collected. This material is subjected to attrition grinding in a batch mill to form a powder; it is then prepared for chemical processing. The concentrated powder is roasted in a furnace, followed by sulfuric acid digestion, where the YPO4 content is converted into a water soluble sulfate. Phosphoric acid is

Fig. 5 e Production cycle of yttrium oxide.

4026

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 4 0 1 9 e4 0 3 3

generated as by-product of the reaction. The products are allowed to cool with water to form a leach. Oxalic acid is added to the yttrium sulfate solution to precipitate yttrium oxalate. The final stage is the calcination of yttrium oxalate to the oxide [31,32]. In Fig. 5 the process stages are shown, while in Table 2 the main reactions of the processes for obtaining the product were collected. It is important to note that, as a part of the manufacturing process of yttrium oxide, an LCI was performed for oxalic acid and, whose description is out of the scope of this paper.

a more uniform coating until the required thickness of the material is achieved. Solvents are required for plasma spraying suspension, as well as pore-forming precursors for the deposition of electrode layers. The system controller of the torch is accompanied by the corresponding electronic instrumentation and equipment which ensures a correct manufacturing technique, and whose important energy consumption is evaluated. Fig. 6 shows the sequence of a cell assembly for this LCA.

Life cycle inventory SOEC assembly. In the manufacture of planar SOEC, there are different techniques which includes chemical vapor deposition, physical vapor deposition, electrochemical vapor deposition, electrophoresis deposition, and air plasma spraying. For this study, we chose air plasma spraying (APS) of a ceramic coating onto a metallic substrate, whose accurate technique allows for thinner layers to be deposited on the substrate, thus reducing required materials and cell cost. Planar SOECs are fabricated by sequentially adding APS of LSM cathodes, LSGM electrolytes and Ni/YSZ anodes. The deposition process requires a mixture of gases: (i) argon and nitrogen, for the electrodes; and (ii) hydrogen and nitrogen, for the electrolyte. The system is made up of active material reservoirs, a delivery pump, a flow controller, a liquid injector and an atomizing cap. The process consists of the injection of powdered ceramics in a plasma torch along with a carrier gas. The solution is pumped from the reservoirs to the injector and the flow is regulated by a controller. The solution fed into the injector is atomized into a fog stream via the air cap, and then injected into the plasma torch. The ceramic particles are melted within the plasma torch, and are propelled towards the support plate where they flatten and cool to form the electrode and the electrolyte layers. During this process, the torch runs through a support plate, made of steel-chromium alloy, with a constant spray velocity that coats an area of 1 cm2 throughout its path. It repeats the process back and forth in order to achieve

Table 2 e Quantification of the production stages of yttrium oxide (coefficients were calculated by mass). Materials

Input

Output

2YPO4 þ 3H2SO4 / Y2(SO4)3 þ 2H3PO4 2 YPO4 1.63 1.30 3 H2SO4 1 Y2(SO4)3 2 H3PO4 Y2(SO4)3 þ 3H2C2O4 þ 4H2O / Y2(C2O4)3$4H2O þ 3H2SO4 1 Y2(SO4)3 2.06 1.20 3 H2C2O4 4 H2O 0.32 1 Y2(C2O4)3 4 H2O 3 H2SO4 Y2(C2O4)3$4H2O / Y2O3 þ 3CO2 þ 3CO þ 4H2O 1.96 1 Y2(C2O4)3 0.32 4 H2O 1 Y2O3 3 CO2 3 CO 4 H2O

2.06 0.86

1.96 0.32 1.30

1.00 0.58 0.37 0.32

For the LCI in this study, the most important processes were selected according to the following two criteria: (i) the processes which involve significant changes in the current LCA, and (ii) the processes with limited information in literature. We were able to provide a source of inventory data. Construction materials reported for building the HTGR reactor were collected from literature; however, other equipment and ancillary installations were based on data from conventional nuclear facilities. This same approach was used for the capital goods that make up the hydrogen plant. Table 3 provides the baseline characteristics of the system and is described below. Nuclear fuel production data inventory was based on similar research in literature [5].

Hydrogen plant The recompilation of the inventory of the hydrogen plant focused on two aspects: (i) the manufacturing of SOEC as described above, and (ii) the quantification of material and energy balances of the hydrogen production. Based on the information described in section 3.2.3, Table 4 presents the inventory data of the raw materials and the energy consumption for the cell fabrication. These aggregated data include quantification of the yttrium oxide production. For the hydrogen plant, mass and energy inventory data was obtained from Refs. [17,18,34] on the system flows in the HTE process. Data from the performance of the hydrogen production plant coupled to a HTGR, collected from selected literature, was used for the UniSim process analysis software. For the current data inventory, the model was replicated and modified based on mass and energy balances according to the process flow diagrams from hydrogen production at realistic conditions. Table 5 provides the main inventory values of input and output from the energy balance of the plant (see Fig. 3). This table does not show all the intermediate values and the exhaustive information which was compiled into spreadsheets, and was used for assembling the system. Since the plant is considered a provider of energy of hydrogen fuel, in this LCA, the energy allocations in the processes were prioritized. The hydrogen plant inventory data was computed by Simapro® software, following these stages: (i) the water provision process, (ii) the air supply system, (iii) the SOEC device, (iv) the thermal energy produced by the nuclear reactor, and (v) the capital goods.

Life cycle interpretation: results and discussion The results are analyzed as follows: (i) the environmental impact from LCA; (ii) the comparison of the contribution of

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 4 0 1 9 e4 0 3 3

4027

Fig. 6 e Outline of the electrolytic cell assembly [19,24,33]. environmental burdens by process type; (iii) the system sensitivity with respect to energy conversion efficiency of hydrogen production and the environmental loads assessed; and (iv) the comparison of electrolysis hydrogen production from different energy sources. In order to evaluate hydrogen production efficiency, including the variable of environmental impact, Eq. (2) is used, where:

Table 3 e The baseline characteristics of the product system.


hC is the overall efficiency of the hydrogen production cycle;
EH is the energy from the produced hydrogen, it was calculated with the lower heating value (LHV) of 120 MJ kg1;
ER is the energy, thermal and the electric power, which can be potentially recovered from the hydrogen plant;
EP is the overall energy supplied to the hydrogen plant by the HTGR; and
CED is the cumulative energy demand of the production chain, this includes the whole of the energy supplied to the upstream process before arriving at the hydrogen plant.

System technical specifications Reactor power Burn-up fuel Cell electric potential Electric current density Effective area of cell Area specific resistance Reaction yield Effective area-yield Turbine/Reactor energy ratio Steam/sweep gas ratio Water feed Air supply Hydrogen production

600 175 1.14 0.22 100 0.40 0.97 1.96 0.88 15.16 21.53 11.12 2.41

MW GWd Mg U1 Volts A cm2 cm2

hC ¼

H2 H2O1 (molar) kg H2 cm2

Environmental impact from LCA

kg s1 kg s1 kg s1

EH þ ER EP þ CED

(2)

Table 6 shows the results of GHG emissions from the climate change category. The results show that nuclear energy and cell fabrication, both together, represent the largest contribution of emissions (~73%). This is related to the intensive use of energy in the upstream of: (i) the nuclear fuel cycle,

4028

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 4 0 1 9 e4 0 3 3

Table 4 e Inventory of materials for the electrolytic cell assembly. Material Polyvinyl butyral Ethanol Trichloroethylene Polyethylene glycol Dibutyl phthalate Electricity [kWh m2 cell; kWh FU1] La-cathode Sr-cathode MnO3/Mn2O3-cathode ZrO2-anode Y2O3-anode NiO/Ni-anode ZrO2-electrolyte Y2O3-electrolyte Argon Nitrogen Hydrogen Nitrogen CuO Carbon Pore Former 2-Butoxyethanol Inconel 600 Al2O3 substrate Cr-Steel support plate Electricity [kWh m2 cell; kWh FU1]

Production stage

kg m2 cell

kg FU1

Powder preparation

4.20E-01 1.50Eþ00 3.14Eþ00 3.80E-01 3.40E-01 1.91Eþ01 3.81E-02 6.01E-03 5.18E-02 9.85E-01 1.57E-01 1.48Eþ00 3.34E-01 5.32E-02 1.01Eþ00 2.36E-01 5.65E-02 3.15Eþ00 1.13Eþ00 2.47E-01 2.24Eþ01 1.29Eþ00 6.00E-01 1.41Eþ01 5.65Eþ02

1.27E-04 4.53E-04 9.49E-04 1.15E-04 1.03E-04 5.76E-03 1.15E-05 1.82E-06 1.57E-05 2.98E-04 4.74E-05 4.48E-04 1.01E-04 1.61E-05 3.05E-04 7.13E-05 1.71E-05 9.51E-04 3.43E-04 7.47E-05 6.77E-03 3.89E-04 1.81E-04 4.25E-03 1.71E-01

Cell active materials

Cell assembly

including the HTGR energy analysis, and (ii) the processing of rare earths regarding the manufacturing of the electrolytic cell. Fig. 7 shows the GHG emissions for the intermediate products which were added to the LCI to complete the overall LCA. It was noted, that as products move from being simple to more complex, or as in the case of materials that are less abundant, the emissions rise considerably. Although GHG emissions seem to be high for nuclear fuel and the electrolytic cell, its magnitude, per functional unit produced, is low. The LCI shows consumption of TRISO fuel as low as 50 mg to produce 1 kg of hydrogen; while just over 0.5 cm2 active cell area was used by the FU.

Table 5 e Energy flows data from hydrogen plant. Process

Stage plant

MJ FU1

Energy inflows (demand and supply) Reactor heat Nuclear plant 248.40 Compressors (High & Low) Brayton cycle 100.72 Recuperator (In) Brayton cycle 268.63 IHE (Air & Steam) Brayton cycle 20.03 Air compression system Air supply system 4.85 Pumping and water steaming Steam supply system 55.81 SOEC (heat & power) Electrolyzer 123.34 Product recovery stage Separation module 2.17 Energy outflows (production and recovery) Intercooler & Precooler Brayton cycle 98.17 Recuperator (Out) Brayton cycle 268.92 Turbine Brayton cycle 185.48 Intermediate Heat Exchanger (IHE) Brayton cycle 20.04 Air compressors recovery energy Air supplied system 2.66 Electrolysis Heat Recuperator Separation module 38.53 Sweep Gas Turbine Separation module 1.38

Comparison of the contribution of the environmental burdens by process stage The environmental impact is shown in Fig. 8. The component related to the energy from the HTGR has the largest environmental impact contribution in five of the eight evaluated categories. The most important categories are the ionizing radiation and the CED, which are closely related, since 95% of the CED comes from nuclear energy. The main source of the ionizing radiation impact comes from the nuclear fuel cycle upstream. Milling produces waste referred to as mill tailings, which are a mix of ground rock, liquid wastes, radioactive products from the natural radioactive decay process, and other chemical toxic wastes. The volume of these tailings can be large when mining low grade uranium ore, which maximizes its environmental impact. The fate and exposure model was used to evaluate the impact category of ionizing radiation. The effect factors are based on disease statistics resulting from relatively high work-related or accident-related exposure. However, in the LCA approach, the exposure doses are generally very low since were corrected due to difference in incidences per exposure dose. The models use a time horizon of 100,000 years in order to consider significant impacts of the different pathway, and a stabilized world population was assumed for this period [15,43,44]. Therefore, the uncertainty in measuring the impact can be considerable, since there is not enough experimental information. The electrolytic cell components showed the highest environmental impact on ozone depletion, due to the use of solvents from laser spraying in the electrolytic layer preparation, which causes 65% of the impact in this category. In general, as is shown in Fig. 8, the highest share of environmental burdens comes from the nuclear energy

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 4 0 1 9 e4 0 3 3

Table 6 e GHGs emissions by stage process. [kg CO2eq FU1]

Climate change impact Total emissions

0.416

Energy from HTGR Electrolytic cell fabrication Facilities Inputs supplied to the H2 plant Energy recovery surplus

0.157 0.148 0.073 0.037 0.001

4029

supply, followed by the manufacturing of the electrolytic cell. Moreover, the hydrogen plant showed the lowest environmental impact, which takes into account all the facilities and the supply of intermediate inputs as well, such as the process water and its conditioning.

Comparison with other hydrogen production systems In this study, the comparisons were carried out regarding two GHG emission issues: (i) the electrolysis methods which were

Fig. 7 e Results of the product system components.

Fig. 8 e Contribution of environmental loads from product system stages.

4030

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 4 0 1 9 e4 0 3 3

Fig. 9 e Comparison with other studies for hydrogen production. supplied from alternative energy sources (non-fossil), and (ii) the production of hydrogen from a nuclear energy supply. The results in Fig. 9 show a range corresponding to the highest, intermediate and the lowest values collected from literature data [1,3e5,35e42]. This information shows the variability of the values, which demonstrates the sensitivity caused by the management system limits and data robustness. The actual

LCA is comparable to the lower range of: (i) the electrolytic process supplied by biomass energy, and (ii) the SeI thermochemical process energized by nuclear energy. The variations shown for the HTE case are due to the boundary delimitation of the product system, the characteristics and the parameters that are subject to change, as can be seen in the next section 3.4.4.

Fig. 10 e The sensitivity of the system product from transport and nuclear fuel burn-up.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 4 0 1 9 e4 0 3 3

4031

Fig. 11 e Analysis of three categories of impact regarding the supply and energy efficiency of the product system.

Product system sensitivity Sensitivity analysis was performed based on two approaches. The first focused on the magnitude of impacts of GHG emissions regarding two issues: (i) the transport of materials and supplies, that aims to assess the direct impact with the use of fossil fuels in the production cycle; and (ii) the burn-up of the fuel discharged, which relates to the efficiency of energy exploitation from the nuclear reactor These results are shown in Fig. 10. In the case of the fuel burn-up, an increase in GHG emissions was noted when the fuel burn-up decreases and fuel utilization is at its worst. It showed that a reduction of 50% in the fuel burn-up ratio produces an increase of about 13% in GHG emissions. Although the reactor energy source is nuclear, if fuel burn-up efficiency is improved, then inputs supplied are reduced which in turn decrease overall GHG emissions. While, in the case of transportation of raw materials, a direct impact on GHG emissions was noted because of the fossil fuel consumption. Nevertheless, a smaller proportional increase of GHG emissions was obtained when the transportation distance of raw materials increases. For example, a 50% increase in the distance of transportation caused an increase of about 5% in emissions. The second approach was focused on the relationship between impact categories and the system product, for which the following relevant categories were selected: (i) climate change, (ii) ionizing radiation, and (iii) human toxicity potential. As the main function of the product system is the energy supply, two parameters related to it were selected: (i) fuel burn-up, and (ii) energy efficiency (hC), as was defined in Eq. (2). The results are shown in Fig. 11. In the three impact categories there was a reduction in the level of measurement, while the efficiency of the system was increasing. As expected, the parameter of fuel burn-up has a greater influence in the category of ionizing radiation, where a greater energy yield from nuclear fuel gives rise to a reduction of the environmental burden. For the other two impact categories (i and iii), their correlations with burn-up fuel and hC have different implications in other parameters in the product life cycle. The processes of the hydrogen plant should remain within a certain range of thermodynamic stability, notwithstanding that changes are promoted in the overall efficiency of the product system or fuel burn-up. In response to any change, even if minimal, an alteration is triggered in the flow of supply of both energy and raw materials through the processes upstream that ends up modifying the values of the impact categories. For example, a gradual decrease in the ratio of energy which supplies the Brayton cycle system (energy split in Fig. 3) causes an increase in GHG emissions, due to the replacement of a nuclear energy deficit by an external power grid. Finally, it is important to mention that because of data availability and the complexity of impact mechanisms the assessment of ionizing radiation and also human toxicity is much more uncertain than that of the climate change impact.

Conclusions According to the results obtained in this LCA, the HTE coupled to a HTGR showed an environmentally competitive result

4032

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 4 0 1 9 e4 0 3 3

compared with other hydrogen production experimental technologies, including others supplied by nuclear energy, specifically in the case of GHG emissions. The conclusions of the results are as follows:
The reduced value of GHG emissions in the HTE technology, compared to other technologies, was achieved due to the internal settings of the energy flows which involve considerable energy recovery from: (i) the Brayton cycle with four heat exchangers, (ii) the air supply system with three heat recovery units, (iii) the two principal interchangers: IHE & EHR, in the electrolytic process, and (iv) the two turbines which supply the required electric power.
Result of the above, 40 MJ kg1 of hydrogen produced, of energy recovery were the output of the product system (see LCI phase, Table 5); this energy comes from the coolers in the Brayton cycle system, the compression stages of the air supply flow, and surplus heat outflows from the electrolytic cell.
From the viewpoint of GHGs mitigation, HTE technology could be a viable option for the future, as a massive hydrogen supplier for replacing fossil fuels in the transport sector. However, in the product system other environmental loads are generated that require further investigation. For example, the human toxicity impact category is a relevant consequence of the production chain of the electrolytic cells.
The ionizing radiation impact category is an important issue to be managed. It must be noted that HTE coupled to a HTGR has characteristics that improve its environmental performance. HTGRs produce less heavy metal radioactive waste per unit of energy produced than current light water nuclear reactors, because of their high thermal efficiency and high fuel discharge burn-up.
The ionizing radiation impact category is an important difference between nuclear and non-nuclear hydrogen production technologies.

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

Acknowledgments Special thanks to the National Council for Science and Technology (CONACYT) for the financial support provided to Mario Giraldi, and to the National Autonomous University of Mexico for the support provided through the research project PAPIIT IN106310.

[18]

references

[20]

[19]

[21] [1] Utgikar V, Thiesen T. Life cycle assessment of high temperature electrolysis for hydrogen production via nuclear energy. Int J Hydrogen Energy 2006;31:939e44. [2] Ozbilen A, Dincer I, Rosen MA. Life cycle assessment of hydrogen production via thermochemical water splitting using multi-step Cu-Cl cycles. J Clean Prod 2012:1e15. [3] Solli C. Fission or Fossil?: a comparative hybrid life cycle assessment of two different hydrogen production methods

[22]

[23]

[Thesis of Master]. Norwegian University of Science and Technology; 2004. Lattin WC, Utgikar VP. Global warming potential of the sulfur-iodine process using life cycle assessment methodology. Int J Hydrogen Energy 2009;34:737e44. Giraldi MR, Franc¸ois JL, Castro-Uriegas D. Life cycle greenhouse gases emission analysis of hydrogen production from SeI thermochemical process coupled to a high temperature nuclear reactor. Int J Hydrogen Energy 2012;39:13933e42. Patyk A, Bachmann TM, Brisse A. Life cycle assessment of H2 generation with high temperature electrolysis. Int J Hydrogen Energy 2013;38:3865e80. O'Brien JE, Stoots CM, Herring JS, McKellar MG, Harvego EA, Sohal MS, et al. High temperature electrolysis for hydrogen production from nuclear EnergyeTechnologySummary (No. INL/EXT-09-16140). Idaho National Laboratory (INL); 2010. OECD (Organisation for Economic Co-operation and Development). Nuclear production of hydrogen. Fourth Information Exchange Meeting. 2009. ISO 14040. Environmental managementelife cycle assessmenteprinciples and framework. International Organization for Standardization; 2006. ISO 14044. Environmental managemente life cycle assessmenteRequirements and guidelines. International Organization for Standardization; 2006. e JB, de Bruijn H, Van Duin R, Huijbregts MAJ, Guine Lindeijer E, Roorda AAH, et al. Handbook on life cycle assessment: operational guide to the ISO standards.. Dordrecht: Kluwer Academic Publishers; 2004. USEPA (U.S. Environmental Protection Agency). Life cycle assessment: principles and practice. 2006. Report EPA/600/R06/060. UNFCCC (United Nations Framework Convention on Climate Change). Information on global warming potentials. Bonn: Technical Paper: FCCC/TP/2004/3; 2004. Fuglestvedt JS, Berntsen TK, Godal O, Sausen R, Shine KP, Skodvin T. Metrics of climate change: assessing radiative forcing and emission indices. Clim Change 2003;58:267e331. Goedkoop MJ, Heijungs R, Huijbregts M, De Schryver A, Struijs J, Van Zelm R. ReCiPe 2008; Ministry of Housing, spatial planning and environment of Netherlands. 2012. Ecoinvent, data v2.2. 2010 development. In: Proceedings of a Technical Committee Meeting. Johannesburg: Swiss Centre for Life Cycle Inventories; 1996. www.ecoinvent.org. McKellar MG, O'Brien JE, Harvego EA, Herring JS. Optimized flow-sheet for a reference commercial-scale nuclear-driven high temperature electrolysis hydrogen production plant. Idaho National Laboratory, U.S. Department of Energy; 2007. Harvego EA, McKellar MG, O'Brien JE, Herring JS. Parametric evaluation of large-scale high-temperature electrolysis hydrogen production using different advanced nuclear reactor heat sources. Nucl Eng Des 2009;239(9):1571e80. O'Brien JE. Large scale hydrogen production from nuclear energy using high temperature electrolysis. In: Keynote Lecture and proceedings, 14th international heat transfer conference; 2010. Karakoussis V, Brandon NP, Leach M, Van der Vorst R. The environmental impact of manufacturing planar and tubular solid oxide fuel cells,. J Power Sources 2001;101(1):10e26. Baratto F, Diwekar UM. Life cycle assessment of fuel cellbased APUs. J Power Sources 2005;139(1):188e96. Wincewicz KC, Cooper JS. Taxonomies of SOFC material and manufacturing alternatives. J Power Sources 2005;140(2):280e96. Eccleston KL. Solid oxide steam electrolysis for high temperature hydrogen production. Doctoral dissertation. University of St Andrews; 2007.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 4 0 1 9 e4 0 3 3

[24] Schiller G, Ansar A, Lang M, Patz O. High temperature water electrolysis using metal supported solid oxide electrolyser cells (SOEC). J Appl Electrochem 2009;39(2):293e301. [25] Sohal MS. Degradation in solid oxide cells during high temperature electrolysis. Idaho National Laboratory Report INL/EXT-09e15617, U.S. Department of Energy; 2009. [26] Bailey L, Khan M, Maheandiran J, Raza A. Life cycle costing analysis of solid oxide fuel cells [Bachelor degree thesis]. University of Toronto; 2009. [27] Kenney BA. An experimental and modeling study of oxygen reduction in porous LSM/YSZ solid oxide fuel cell cathodes. Doctoral dissertation. Queen's University; 2009. [28] Liang M, Yu B, Wen M, Chen J, Xu J, Zhai Y. Preparation of NiOeYSZ composite powder by a combustion method and its application for cathode of SOEC. Int J Hydrogen Energy 2010;35(7):2852e7. [29] Zhao L, Brouwer J. Life cycle analysis of ceramic anodesupported SOFC system manufacturing processes. ECS Trans 2012;42(1):247e63. [30] Long KR, Van Gosen BS, Foley NK, Cordier D. The principal rare earth elements deposits of the United States. A summary of domestic deposits and a global perspective. U.S. Geological Survey Scientific Investigations Report; 2012. [31] Vijayalakshmi R, Mishra SL, Singh H, Gupta CK. Processing of xenotime concentrate by sulphuric acid digestion and selective thorium precipitation for separation of rare earths. Hydrometallurgy 2001;61:75e80. [32] Gupta CK, Krishnamurthy N. Extractive metallurgy of rare earths. CRC Press; 2004. [33] Hino R, Yamada K, Kasai S. Steam electrolysis. In: Nuclear hydrogen production handbook. Taylor & Francis Group; 2011. [34] O'Brien JE, Stoots CM, Herring JS, Hawkes GL. Hydrogen production from nuclear energy via high temperature electrolysis. In: The 1st energy center hydrogen initiative symposium. Purdue University; 2006.

4033

[35] Koroneos C, Dompros A, Roumbas G, Moussiopoulos N. Life cycle assessment of hydrogen fuel production processes. Int J Hydrogen Energy 2004;29:1443e50. [36] Spath PL, Mann MK. Life cycle assessment of renewable hydrogen production via wind/electrolysis. Golden: National Renewable Energy Laboratory; 2004. [37] Granovskii M, Dincer I, Rosen MA. Exergetic life cycle assessment of hydrogen production from renewable. J Power Sources 2007;167:461e71. [38] Manish S, Banerjee R. Comparison of biohydrogen production processes. Int J Hydrogen Energy 2008;33:279e86. [39] Ozbilen A, Dincer I, Rosen MA. A comparative life cycle analysis of hydrogen production via thermochemical water splitting using a Cu-Cl cycle. Int J Hydrogen Energy 2010;36:11321e7. [40] Djomo SN, Blumberga D. Comparative life cycle assessment of three biohydrogen pathways. Bioresour Technol 2011;102:2684e94. [41] Ozbilen A, Dincer I, Rosen MA. Comparative environmental impact and efficiency assessment of selected hydrogen production methods. Environ Impact Assess Rev 2013;42:1e9. [42] Ozbilen A, Dincer I, Rosen MA. Environmental impact assessment of nuclear assisted hydrogen production via CueCl thermochemical cycles. Sustain Cities Soc 2013;7:16e24. [43] Frischknecht R, Braunschweig A, Hofstetter P, Suter P. Human health damages due to ionising radiation in life cycle impact assessment. Environ Impact Assess Rev 2000;20(2):159e89. [44] De Schryver AM, Van Zelm R, Humbert S, Pfister S, McKone TE, et al. Value choices in life cycle impact assessment of stressors causing human health damage. J Industrial Ecol 2011;15(5):796e815.