LOWERING PEAK TEMPERATURES FOR NUCLEAR THERMOCHEMICAL PRODUCTION OF HYDROGEN Текст научной статьи по специальности «Физика»

LOWERING PEAK TEMPERATURES FOR NUCLEAR THERMOCHEMICAL PRODUCTION OF HYDROGEN

C. W. Forsberg

Oak Ridge National Laboratory* P.O. Box 2008, Oak Ridge, TN 37831-6179, United States of America Tel.: (865) 574-6783; fax: (865) 574-9512; e-mail: forsbergcw@ornl.gov

Dr. Charles Forsberg received his Bachelors of chemical engineering degree from the University of Minnesota and his Doctor of Science degree from the Massachusetts Institute of Technology. He is a professional engineer and a senior scientist at Oak Ridge National Laboratory in the United States of America. Dr. Forsberg has published over 200 articles and reports; holds 9 patents; has served on multiple government, IAEA, and foreign panels; and received multiple awards. In 2002 he received the American Nuclear Society special award for «Advanced Nuclear Power Generation Concepts». This was for development of the Advanced High-Temperature Reactor for hydrogen production.

Forsberg Charles Winfield

The efficient thermochemical production of hydrogen using nuclear heat requires matching the nuclear reactor and the thermochemical processes to convert heat plus water into hydrogen and oxygen. The major challenges are the high temperatures required to produce hydrogen efficiently. Consequently, Oak Ridge National Laboratory, in collaboration with Sandia National Laboratories and the University of California at Berkeley, is investigating nuclear reactor options and thermochemical cycles to minimize those temperatures while efficiently producing hydrogen. We are developing the concept of a molten-salt-cooled Advanced High-Temperature Reactor to produce the heat. The use of a low-pressure liquid coolant minimizes the temperature drops between the hottest fuel elements in the reactor and the thermochemical cycle, thus minimizing peak reactor temperatures. Simultaneously, we are examining the use of inorganic membranes to minimize the temperatures required for the efficient production of hydrogen using the (1) sulfur-iodine, (2) Westinghouse, and (3) Ispra Mark 13 thermo-chemical hydrogen processes.

Introduction

The worldwide demand for hydrogen (H2) is ~50 million tons per year and growing rapidly. Hydrogen is used primarily for production of ammonia for fertilizer and conversion of heavy crude oils into cleaner liquid fuels. An international effort is under way to deliver H2 as a replacement fuel for transport vehicles. Ultimately, the energy required to produce H2 could exceed that for electricity. Consequently, strong incentives exist to develop economic methods to produce H2 using nuclear energy.

Among the leading candidates for low-cost, large-scale H2 production are thermochemical processes. A thermochemical process consists of a set of chemical reactions in which the net result is high-temperature heat plus water yields H2 and O2. Two factors make thermochemical H2 production costs (with nuclear reactors providing the heat) potentially lower than those for electrolysis.

■ Efficiency. Thermochemical processes have potentially greater efficiency because conversion of heat to H2 requires fewer steps than conversion of heat to electricity and electricity to H2.

■ Capital costs. The economics of scale for chemical processes (function of volume) is significantly better than the economics of scale for electrolytic processes (function of area).

If H2 is to be produced economically [1], the nuclear reactor must be matched with the thermo-chemical process. In a recent evaluation [2] of thermochemical cycles, three of the four highest-ranked cycles (Hybrid sulfur, Ispra Mark 13, and sulfur-iodine) were sulfur cycles that have the same high-temperature chemical reactions but different low-temperature chemical reactions. Given these results, we have concentrated our efforts on matching the energy output of the nuclear reactor to the required energy input of these three ther-mochemical processes.

*Oak Ridge National Laboratory, managed by UT-Battelle, LLC, for the U. S. Department of Energy under contract DE-AC05-00OR22725

The three processes require heat input at peak temperatures of ~850 °C. This condition presents a major challenge. If the chemical process requires 850 °C heat, the nuclear reactor must operate at significantly higher temperatures to allow transfer of the heat from the reactor core to the chemical process. Such temperatures are near the limits of practical current materials. To reduce the material challenges, we have initiated a two-part program to better match the nuclear reactor to the thermochemical cycle: (1) develop a nuclear reactor that delivers heat at 850 °C but that is designed to minimize the peak temperatures within the reactor and (2) modify the high-temperature steps within the thermochemical cycles to lower peak temperatures. This paper discusses the status and results of this research.

Thermochemical production of hydrogen

To understand the challenges of H2 production, some understanding of the thermochemical cycles is required. As noted earlier, three [2] of the four highest-rated processes (fig. 1) have the same high-temperature chemical step that requires ¡1 heat input at >850 °C. The highly endothermic t (heat-absorbing) gas-phase reaction in each of these f processes is as follows: ^

2H2SO4 ^ 2H2O + 2SO3 ^ |

^ 2SO2 + 2H2O + O2 (850 °C). (1) i

c

The three thermochemical processes have dif- $ ferent lower-temperature chemical reactions. The | sulfur-iodine process [2] has two other chemical q reactions that when combined, (1) yield H2 and

High-Temperature Reactions

Oxygen

Heat

h2SO4—H2O+SO3 ~h2o+so2+1/2o2

Base Case

Heat

I 700°C^>

Membrane Separation

h2so4^h2o+so3 -~h2o+so2+%o2

Inorganic Membrane

Low-Temperature Reactions

Hydrogen

Heat

Sulfur Iodine

Hydrogen Hybrid Sulfur

(Westinghouse, GA-22 and Ispra Mark II)

Hydrogen

I Br2+ S02 + 2H20 Electrolysis ♦ 2HBr+H2S04 2HBr -*H2+Br2

Reject Heat

Fig. 1. Sulfur family of thermochemical cycles

щ>

Reject Heat

Ispra Mark 13

O2 from water and heat and (2) recycle all other chemical reagents:

I2 + SO2 + 2H2O ^ 2HI + H2SO4 (120 °C); (2)

2HI ^ I2 + H2 (450 °C). (3)

< The hybrid sulfur process (also known as

i Westinghouse, GA-22, and Ispra Mark 11) has

£ a single electrochemical step that completes the

S cycle [3]:

| SO2 (aq) + 2H2O (1)^ H2SO4 (aq) + H2 (g). (4)

f (Electrolysis: 80 °C)

I The Ispra Mark 13 process has one chemical g reaction and one electrochemical reaction that com™ pletes the cycle:

Br2 (aq) + SO2 (aq) + 2H2O (l)^ ^ 2HBr(g) + H2SO4 (aq) (77 °C); (5)

2HBr (g H Br2 (l) + H2 (g). (6)

(Electrolysis: 77 °C)

In each of these cycles, the high-temperature sulfur trioxide (SO3) dissociation reaction is an equilibrium chemical reaction that requires a catalyst. High temperatures and low pressures drive the reaction towards completion. Fig. 2 shows this equilibrium as a function of temperature.

Detailed studies have concluded that the required process temperatures need to be very high: ~850 °C. After the high-temperature dissociation reaction, all the chemicals must be cooled to near room temperature, the SO2 separated out and sent to the next chemical reaction, and the unreacted sulfuric acid (formed by recombination of SO3 and H2O at lower temperatures) reheated back to high temperatures. Unless the chemical reactions go almost to completion, the energy losses in separations and the heat exchangers to heat and cool all the unreacted reagents (H2SO4) result in a very inefficient and uneconomical process. This phenomenon is illustrated in fig. 3, in which the overall efficiency of one variant of the sulfur-iodine process is shown as a function of temperature [2]. Efficiency is defined as the higher heating value of H2 divided by the thermal energy into the process. In this flowsheet, the process inefficiencies increase so rapidly with decreasing temperature (incomplete reactions) that the process cannot produce H2 at temperatures below 700 °C. The process thus defines the nuclear reactor requirements.

The advanced high-temperature reactor

There are two approaches to developing a nuclear reactor for H2 production. An existing reactor system can be modified to meet the H2 production requirements, or a new reactor system can be developed. At the current time, only one nuclear reactor system, the gas-cooled (helium) reactor,

0-1-.-1-1-.-

600 650 700 750 800 850 900

Temperature, °C

Fig. 2. Equilibrium concentrations of SO3, SO2, and O2 vs temperature starting with 100 moles of SO3

has the potential high-temperature capabilities to provide the heat at sufficient temperatures to drive a H2 production system. This reactor has historically been considered the reactor that would be used to provide high-temperature heat for H2 production [4]. The gas-cooled (helium) reactor was developed for electricity production and uses a coated-particle fuel (see below) and high-pressure helium as a coolant. Several prototype reactors have been built. Last year, Japan began operation of its 30 MW(t) High-Temperature Test Reactor to develop nuclear heat applications, including H2 production. This specific reactor has a peak exit temperature of 950 °C.

Alternatively, a reactor can be designed specifically for H2 production. Given the demanding requirements for H2 production, we are developing a new reactor concept, the Advanced High-Temperature Reactor (AHTR), to match H2 production requirements [5]. This is a joint effort between three organizations in the United States: Oak Ridge National Laboratory (ORNL), Sandia National Laboratories, and the University of California at Berkeley. The AHTR is based on several earlier technological developments:

■ high-temperature, low-pressure molten-salt reactor coolants from the aircraft nuclear propul-

600 700 800 900 1000

Temperature, °C

Fig. 3. Efficiency of the sulfur-iodine process versus temperature

sion program of the 1950s and the molten-salt breeder reactor program of the 1960s;

■ coated-particle graphite-matrix fuel developed in the 1970s for gas-cooled reactors;

■ passive safety systems for gas-cooled and liquid-metal reactors developed in the 1980s;

Concept Description

The AHTR reactor core consists of coated-particle graphite-matrix fuel cooled with a molten fluoride salt. The molten salt (fig. 4) flows through the reactor core to an external heat exchanger (to provide the interface for the H2 production system), dumps the heat load, and returns to the reactor core.

The fuel is essentially the same as that used for the gas-cooled (helium) reactor. The important characteristic of these fuels is that they can operate at very high temperatures with peak fuel operating temperatures of ~1200 °C. Under accident conditions, temperatures can go to 1600 °C for several hundred hours without significant failure. These coated-particle fuels are the only commercially demonstrated nuclear fuels capable of producing heat at temperatures sufficient for H2 production. The fuel consists of small particula-tes of uranium dioxide coated with layers of carbon and silicon carbide. The multiple layers isolate the fuel and fission products (produced by the nuclear reactions) from the coolant. The mi-crospheres are embedded in a compact made of graphite. The fuel compact is embedded in graphite blocks and the hexagonal blocks are then assembled into a reactor core.

Molten fluoride salts are the only high-temperature liquids that have been fully demonstrated to be chemically compatible with graphite fuels.

Several different fluoride salts are being considered. The current leading candidate is a mixture of sodium and zirconium fluorides. The atmospheric boiling points for molten fluoride salts are near 1400 °C. As a consequence, the reactor operates at low pressures.

ji

Matching Reactor Characteristics t

to Those of the Hydrogen Plant |

a

The use of a low-pressure liquid coolant for s

c

production of H2 using nuclear energy has three -§ potential advantages over other systems: mini- ^ mized peak reactor temperatures, operation at low § pressures, and economics. ^

c

c

Minimizing peak reactor temperatures e

The challenge for H2 production is to minimize reactor temperatures while delivering high-temperature heat to the process. This can best be accomplished by using a liquid reactor coolant. Liquid coolants have good heat transfer capabilities and low pumping power costs in comparison with gas coolants, as shown in table 1. The temperature rise across gas-cooled reactors is typically several hundred degrees, whereas that across liquid-cooled reactors is typically under 100 °C. The AHTR, as a liquid-cooled reactor, can deliver its heat with small temperature drops (20-100 °C) with low pumping power.

An example serves to illustrate the benefits of a liquid coolant. If heat is needed at 850 °C, the maximum temperature of the coolant in a gas-cooled reactor may exceed 1100 °C whereas that of the coolant in a liquid-cooled reactor will not exceed 950 °C. While the temperature rise in a gas reactor can be reduced, this requires much

Fig. 4. Advanced High-Temperature Reactor for hydrogen production

Table 1

Temperature drops for different reactor coolants

System (reactor name) AT inlet to outlet (°C) Inlet T (°C) Outlet T (°C) Coolant

GT-MHR 359 491 850 Gas (Helium)

AGR (Hinkely) 355 310 665 Gas (CO2)

PWR (Point Beach) 20 299 319 Liquid (water)

LMR (Super Phoenix) 150 395 545 Liquid (Sodium)

* Abbreviations: GT-MHR — gas-turbine modular helium reactor; AGR — advanced gas reactor; PWR — pressurized-water reactor; LMR — liquid-metal reactor.

a

u

c a

i/i

c c

e

higher gas flow rates with significant additional pumping costs. Liquid coolants minimize materials requirements by lowering the peak reactor and heat-exchanger temperatures.

Pressure

For H2, a lower-pressure reactor is preferred. The H2 production facility will contain significant inventories of hazardous chemicals. A low-pressure, non-chemically reactive coolant minimizes safety risks by minimizing the consequence of heat-exchanger failures between the chemical and nuclear facilities. High-pressure reactor coolants create the potential for pressurization of the chemical plant and releases of toxic gases. At high temperatures, high-pressure coolants also place much greater stresses on the materials of construction. Low-pressure molten-fluoride coolants can match the low pressures of the hydrogen production systems. Molten fluoride salts have boiling points near 1400 °C (and thus avoid the potential for chemical plant pressurization), do not react with air, and react only slowly with water.

Economics

Economics ultimately determines whether a particular approach to H2 production will be viable. Nuclear power production of H2 faces two main competitors: fossil production of H2 by steam reforming of natural gas (or coal) and nuclear production of electricity for electrolysis of water. Hydrogen from fossil fuels is a potential long-term option if environmentally-acceptable methods for sequestration of carbon dioxide can be developed to avoid the potential consequences of greenhouse gases. The newest world-class H2 production plant (that is under construction and will be fueled with natural gas) will have a H2 production capacity of 8.5 million cubic meters per day (300 million cubic feet per day). An equivalent nuclear H2 plant would require an energy output of 2400 MW(t), assuming 50 % efficiency, to produce an equivalent quantity of H2. Current nucle-

ar power plants for electricity production are of similar size.

To match the economics of these H2 plants, the AHTR is a large (2400 MW(t))] reactor with passive safety systems. Passive safety systems do not require operators for functioning and have no moving parts (motors, pumps, etc.). Such systems offer major advantages in terms of safety and also have the potential to reduce costs. Although, historically, these systems could be used only on smaller reactors, the use of a high-temperature, low-pressure coolant may allow their use in large reactors. If this can be demonstrated, it has major economic advantages.

If a reactor shuts down, heat continues to be generated from the decay of short-lived radionu-clides in the fuel. The decay heat decreases with time. If a method to remove decay heat is not provided, the reactor core will overheat with damage to the reactor core. Several types of passive decay heat removal systems have been developed for modular reactors, all of which are similar. Decay heat from the reactor core is conducted through the reactor vessel to some type of passive cooling system outside the reactor vessel. The decay heat option shown in fig. 4 is similar to that proposed for the General Electric S-PRISM liquid-metal-cooled modular reactor. In this pooltype reactor, decay heat is conducted through the reactor vessel wall, transferred across an argon gap by radiation to a guard vessel, conducted through the guard vessel, and then removed from the second wall by natural circulation of air. The radiation heat transfer from the reactor vessel to the guard vessel increases by T4; thus, a small rise in the reactor vessel temperature greatly increases heat transfer out of the system. The argon gap acts as a thermal switch to limit heat losses during normal operation but allows radiation heat transfer to increase heat losses if the reactor vessel heats up.

The reactor size is limited by the ability to transfer decay heat from the nuclear fuel to the outside of the reactor vessel (fig. 5) in an emergency. The use of a molten salt coolant and a high-temperature

Fig. 5. Evolution of passive decay heat removal systems in similar size reactor vessels to allow larger reactor power outputs

fuel allows much higher reactor power ratings than those found in other reactors with similar passive safety systems in the same size reactor vessel. Reducing plant size per unit output reduces plant costs. There has been an evolution in the design of passive safety systems that allows reactors of larger size to use passive safety systems.

■ Gas-cooled reactors. In an emergency in which the other cooling systems have failed, decay heat must be moved from the center of the reactor to the vessel boundary by conduction and radiation. This process requires a large temperature drop to transfer heat through the graphite fuel, the graphite reflector, and a thick-wall pressure vessel. To ensure that the fuel in the center of the reactor does not fail, the power production of the reactor is limited to 600 MW(t). Conduction from the center of the reactor to the outside of the pressure vessel limits the ultimate size of the reactor.

■ Sodium-cooled reactors. In an emergency in which the other cooling systems have failed, decay heat is transferred from the center of the reactor to the vessel wall by natural circulation of sodium. (Natural circulation of a liquid is an efficient way to transfer heat.) If the fuel in the center of the reactor is not to fail in an accident, the power production must be limited to ~1000 MW(t). The limitation in this reactor is that the peak temperature must be significantly below the boiling point of sodium.

■ AHTR. Decay heat removal in the AHTR is similar to that in a sodium-cooled reactor. However, for the AHTR, the coolant boiling point is 1400 °C and the fuel failure temperature is above this level. Thus the only limitation is the reactor vessel. With current vessel materials, the vessel temperature can be as high as 750 °C. This may allow a reactor power level of ~2400 MW(t). Because the coolant and fuel can go to such extreme temperatures, the vessel has an internal insulation layer (core graphite reflector that also reduc-

es neutron damage to the reactor vessel) to reduce heat losses during normal operation. This allows the molten salt to operate at higher temperatures than the reactor vessel.

Status of AHTR Development

The AHTR is a new reactor concept (~2 years old). The basis for a preconceptual design has been developed. These preliminary results have been highly favorable; however, significant work is required before a major commitment can be made to the large-scale development of the technology.

Lower-temperature sulfur thermochemical cycles

The high temperatures of the efficient ther-mochemical cycles present a major engineering challenge. Therefore, ORNL has initiated a parallel effort to reduce the peak temperatures required for these thermochemical cycles [6]. This is a new effort based on the use of inorganic membranes. For over 50 years, ORNL has been developing various inorganic membranes for other applications such as the separation of uranium isotopes by gaseous diffusion.

Membrane Reactor Concept

An inorganic membrane process is proposed to reduce the peak temperature of the SO3 dissociation step by several hundred degrees to 700 °C. This is accomplished by the separation of SO2, H2O, and O2 from the SO3. If these reaction product gases are removed, the remaining SO3 (with a catalyst and heat) will disassociate into its equilibrium concentrations as shown in fig. 2. If the reaction gases can be selectively removed, the process can be driven to completion. The membrane operates with high pressure on one side and a lower pressure on the other side. This pressure difference drives the separation process.

The operating temperature of the membrane is limited by two considerations. First, significantly lower temperatures are not allowed because the membrane processes separate gases, not liquids. As the temperature decreases, condensation of various sulfur compounds will occur. Membrane operating temperatures need to be a reasonable margin above the temperatures at which condensation of any species under any condition might occur. From a thermodynamic perspective, lower temperatures would be expected to reduce the process efficiency; thus, there is an incentive to operate at higher temperatures. It requires mechanical work to provide the pressure difference across the inorganic membrane. However, the irreversible (non-thermodynamic) losses in heat exchangers to heat and cool reagents are the primary source of inefficiencies between an ideal process and the real process. Inorganic membranes reduce these inefficiencies. As a consequence, lowering temperatures is not expected to result in major loses in efficiency. The thermodynamic efficiency is less but the irreversible losses are also reduced. Studies are being initiated to define the optimum membrane temperature.

Figure 6 shows a schematic of two ideal high-temperature reactors with inorganic separation membranes. Each alternative option consists of two zones:

■ oxygen separation. The top membrane reactor shows the operation of a perfect membrane that allows H2O and O2 through the membrane but blocks all other chemical species. At the high temperatures, the H2SO4 dissociates into H2O and SO3. When these reagents contact the catalyst, the SO3 partly disassociates into SO2 and O2 (2). This is a highly endothermic reaction; thus, heat

must be added to enable this reaction. The dissociation is limited by its equilibrium. As the gas mixture flows to the right past the membrane, O2 and H2O go through the membrane. The reaction is driven to the right with the resultant greater

concentrations of SO2. A

mixture of SO2, SO3, exits the reactor. Re-

Fig. 6. Membrane reactor systems

and small quantities of O2 moval of oxygen alone can not drive the reaction to completion (see next section);

■ oxygen and SO2 separation. The membrane reactor is similar to the first case, except that the membrane selectively allows H2O, O2 and SO2 to pass through the membrane. In this case, a perfect membrane would drive the reaction to completion (see next section).

Thermodynamics

A thermodynamic analysis of the separation process was undertaken to understand the ideal theoretical performance of this system. The classical thermodynamic equation for this equilibrium reaction is

K(T, P) = [SO2][O2]/[SO3], (7)

where K (T, P) — equilibrium constant (a constant at any temperature but increases with temperature); [SO2] = gas-phase concentration of SO2, typically in moles per liter; [O2] = gas-phase concentration of O2; [SO3] = gas-phase concentration of SO3.

As can be seen from (7), as SO2 and O2 are removed from the catalyst bed, more of the SO3 must dissociate to maintain the required equilibrium until all of the SO3 is disassociated. However, if only the O2 is removed, the concentration of SO2 increases as the SO3 decreases. With the removal of only one reaction product, the reaction can go far toward, but not all the way to, completion.

A parametric study was conducted to determine the potential benefit that the removal of have on the to SO2. Using the FactSage computer program, the equilibrium conversion as a function of temperature was calculated (fig. 2).

Next, the effect of the removal of O2 was studied. Calculations were made by first assuming that the reaction reached equilibrium in the first (theoretical) stage. At that stage, all of the O2 was assumed to be removed and the remaining SO3 and SO2 were allowed to come to equilibrium again (stage 2). The O2 was again removed and this process was repeated through six stages. As shown in table 2, the residual SO3 at 700 °C (21.6%) using inorganic membranes is ap-

O2 and SO2 could conversion of SO

Table 2

Effect of removal of oxygen and sulfur dioxide from sulfuric acid decomposition reactor

using an ideal inorganic membrane*

Stage # Removal of O2, T = 850 °C Removal of O2, T = 700 °C Removal of O2 and SO2, T = 700 °C

O2 SO2 SO3 O2 SO2 SO3 O2 SO2 SO3

0 0 0 100 0 0 100 0 0 100

1 39.42 78.87 21.13 23.78 47.55 52.45 23.78 47.55 52.45

2 5.43 89.74 10.26 6.8 61.16 38.85 12.47 24.94 27.51

3 1.91 93.55 6.45 3.54 68.24 31.76 6.54 13.08 14.43

4 0.92 95.4 4.6 2.26 72.76 27.24 3.38 6.86 7.57

5 0.53 96.49 3.54 1.6 75.97 24.03 1.8 3.6 3.97

6 0.34 97.14 2.86 1.21 78.4 21.6 0.94 1.89 2.08

* Initial value for SO3 = 100 moles. Table shows moles of various components remaining in the reaction chamber after each stage.

proximately equal to the residual SO3 at equilibrium at 850 °C (21.13 %) with no membrane separation. For the chemical reactor configuration shown in fig. 4, lengthening the tubes increases the number of theoretical stages (the stages do not represent physical stages of this equipment).

Lastly, the effect of the removal of both O2 and SO2 was studied. Calculations were made by first assuming that the reaction reached equilibrium in the first stage. At that stage, all of the O2 and SO2 were assumed to be removed and the remaining SO3 was allowed to dissociate and come to equilibrium again (stage 2). The O2 and SO2 were again removed and this process was repeated through six stages. After six stages, only 2.08 % of the SO3 remained.

Although the analysis indicates that an ideal membrane that separates only O2 can effectively lower the peak dissociation temperature 150 °C and reduce the unreacted SO3 to 21.6 % at 700 °C, there are strong incentives to remove both SO2 and O2. An idealized membrane can reduce the unreacted SO2 to 2.06 % with six ideal states of separations.

Characteristics of Inorganic Membranes

The relative rates of transport of different molecules through the membrane determine the capability of the membrane to separate different gases, and multiple gas-transport mechanisms are involved [7]. The precise transport mechanism that is dominant for each gas depends upon a variety of physical factors including temperature (T), pressure (P), molecular mass (m), pore diameter (dp), molecular size and shape, pore surface composition, pore morphology, and mutual interactions between molecules traversing the membrane.

The performance of a membrane is measured by two parameters: permeance and selectivity. The permeance, defined as flow of the pure gas in question per unit membrane area per unit time per unit pressure, is expressed in moles per square meter per second per pascal (mol/(m2/s • Pa)). The selectivity is defined as the ratio of the permeances of two pure gases. The separation factor for a mixture of two gases is defined as [y/(1 - y)][(1 - x)/x]. Here, y is the concentration of the fastest-permeating component on the permeate side of the membrane and x is the concentration of the fastest-permeating component on the feed side.The product of the separation factor and permeance is often taken as the figure-of-merit by which to judge a particular membrane-gas mixture combination.

For high-temperature separations, the mechanisms of nanopore diffusion provide the best performance. The term nanopore diffusion encompasses several distinct mechanisms that take place in nanometer-diameter pores. For larger molecules, the membrane may function effectively as a molecular sieve, eliminating the transport of such molecules through the membrane and giving high separation factors. For smaller molecules, the transport exhibits thermally activated behavior: as the temperature is increased, the permeance increases exponentially, rather than decreasing as in Knud-sen diffusion. This characteristic of improved performance with increases in temperature is a requirement for an efficient high-temperature membrane. Typically, the membrane pore size is no more than three times the diameter of the molecule. One thermally activated mechanism that has been described in the literature is termed gas trans-lational diffusion (also referred to as thermally activated Knudsen diffusion), in which molecules jump between pore walls but with an activation

barrier that must be overcome in order to make a diffusion jump. This thermally activated characteristic is similar to the diffusion of defects or atoms in the solid state in the presence of traps, with an activation energy (£a). Physically, this is plausible, since the lower limit on the size of a pore must correspond to interatomic spacing in the solid state. In the dp, ~1 nm regime, separation factors >100 are possible. For example, report [8] that a separation factor >200 has been measured for a mixture of H2 and C3H6 gases using a supported amorphous silica membrane with a pore diameter of ~1 nm.

Status of Inorganic Membrane Development

ORNL has developed and fabricated a wide variety of inorganic membranes and has several test loops. The development of an inorganic membrane for this particular separation has just begun and one of the existing test loops is being modified for these gas mixtures. The operation of inorganic membranes is not fully understood. Consequently, several membranes from our inventories will be chosen and tested using O2, SO2 and SO3 as a function of temperature. Based on the experimental results and theory, custom membranes for this specific application will be fabricated and tested. This is an iterative procedure. In parallel, studies have been initiated to understand the performance requirements for such membranes.

Conclusions

Thermochemical hydrogen production using nuclear energy has the potential to be an economic, efficient, and environmentally friendly source of H2 for the world. However, major engineering challenges remain — particularly the high temperatures required. To address these barriers, two different technologies are being investigated: an improved nuclear reactor to produce the high-temperature heat and an improved chemical reactor using inorganic membranes to reduce peak thermochemical

process temperatures. Although the research is still in an early stage of development, both approaches appear to be potentially attractive.

References

1. Forsberg C. W. Hydrogen, Nuclear Energy, and the Advanced High-Temperature Reactor // Int. J. of Hydrogen Energy. 2003. Vol. 28, No. 10. P.1073-1081.

2. Brown L. C., Besenbruch G. E., Lentsch R. D. et al. High Efficiency Generation of Hydrogen Fuels Using Nuclear Power // GA-A24285, General Atomics Corporation, La Jolla, California. June 2003.

3. The Conceptual Design of an Integrated Nuclear Hydrogen Production Plant Using the Sulfur Cycle Water Decomposition System// NASA CR-134976, Contract No. NAS-3-18934, April 1976.

4. LaBar M. P. The Gas Turbine-Modular Helium Reactor: A Promising Option for Near Term Deployment // Int. Congress on Advanced Nuclear Power Plants, Hollywood, Florida, June 9-13, 2002. GA-A23952, San Diego, California: General Atomics Corp., 2002.

5. Forsberg C. W., Pickard P. S., Peterson P. F. Molten-Salt-Cooled Advanced High-Temperature Reactor for Production of Hydrogen and Electricity // Nuclear Technology. December 2003.

6. Forsberg C. W., Bischoff B., Mansur L. K., Trowbridge L., Tortorelli P. A Lower-Temperature Iodine-Westinghouse-Ispra Sulfur Process for Ther-mochemical Production of Hydrogen // Proc. Glo-bal-2003, American Nuclear Society, La Grange Park, Illinois.

7. Burggraaf A. J., Cot L., Eds. Fundamentals of Inorganic Membrane Science and Technology // Membrane Science and Technology Series, No. 4. Elsevier, Amsterdam, 1996.

8. Uhlhorn R. J. R., Keizer K., Burggraaf A. J. Gas Transport and Separation with Ceramic Membranes // J. of Membrane Science. 1992. Vol. 66. P. 259-287.