High-temperature steam electrolysis: Technical and economic evaluation of alternative process designs

0360-3199/86 $3.00 + 0.00 Pergamon Journals Ltd. © 1986 International Association for Hydrogen Energy.

Int. J. Hydrogen Energy, Vol. 11. No. 7, pp. 435--442, 1986. Printed in Great Britain.

HIGH-TEMPERATURE STEAM ELECTROLYSIS" TECHNICAL AND ECONOMIC EVALUATION OF ALTERNATIVE PROCESS DESIGNS M. A. LIEPA and A. BORHAN School of Chemical Engineering Practice, Brookhaven Station, Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139, U.S.A.

(Received for publication 26 November 1985) Abstraet--A high-temperature water-vapor electrolysis (HTE) unit operating at an average temperature of 1000°C was integrated into a preliminary process design using electrical and thermal energy derived from coal. Process variations with either steam or water feeds under isothermal or nonisothermal HTE operation were considered. Operating and capital costs were estimated for each process flowsheet, with the lowest costs being obtained for operation with high steam conversions in the electrolyzer. Estimated H2 production costs were compared with estimates obtained from the literature for other H2 production processes. The estimated HTE H2 production costs ranged from $0.17 to $0.22 standard m-3 H2 produced ($13-17 GJ 1 using the higher heating value), assuming $1.90 GJ -1 for thermal energy and $13.90 GJ -1 for electrical energy.

NOMENCLATURE E AGe AHc AHe Q S r/

r/s

= electrical power consumption in HTE process (MW) = free energy change of the water-vapor electrolysis reaction (J mo1-1) = higher heating value of H2 product stream (MW) = enthalpy change of the water-vapor electrolysis reaction (J mol-1) = high-temperature thermal energy consumption in Processes B and C (MW) = low-temperature thermal energy input for offsite steam generation in Processes A and B (MW) = fractional thermodynamic efficiency of HTE process = fractional thermodynamic efficiencyof electrical power generation = fractional thermodynamic efficiency of steam generation.

implemented, a significant amount of process design work must be done. The overall process design for HTE is more complicated than that for conventional electrolysis because of the high operating temperatures. The heat content of the product gases would have to be recovered by heat exchange with the feed steam to reduce the net energy input required. The feed steam could then be heated above 1000°C by a high-temperature heat source which, in principle, can be entirely independent of the electrical power source for the electrolysis cells. Many types of primary energy sources could possibly be coupled with a HTE process. These include fossil sources such as coal or oil shale and inexhaustible energy sources such as fusion. Since fusion energy is not yet available, any current development of a HTE process must rely on an available primary energy source such as coal. Integration of HTE with a heat recovery system and high-temperature heat source must also be done in a way which minimizes H 2 production cost.

INTRODUCTION

Objectives

Background Manufacturing H 2 by electrolyzing water vapor at high temperatures has potential cost advantages over lower temperature processes because a portion of the endothermic heat of reaction can be supplied thermally rather than electrically. As an illustration, supplying any portion of the electrolysis energy directly from coal combustion at a cost of $1.90 GJ -a would be more than seven times less expensive than using electrical energy generated from that same coal. The development of high-temperature steam electrolysis (HTE) cells is currently underway in a number of research groups [1, 2]. These cells have a solid-oxide electrolyte and operate at temperatures above about 900°C. However, before HTE can be commercially 435

The first objective was to develop preliminary process designs which integrated a HTE unit with a thermal energy source and heat recovery equipment. The thermal energy source selected was a coal-fired furnace operating at about 2000°C. Heat exchangers were used to preheat the electrolyzer feed stream while cooling the hot product streams. The second objective was to analyze the economics of the alternative HTE processes we selected. Electrolyzer steam conversions of 15-90% and H 2 product pressures of 0.304-1.01MPa were considered. Operating and equipment costs for the HTE processes were estimated and compared with production costs obtained from the literature for other H2 manufacturing processes [3, 4, 5].

436

M. A. LIEPA AND A. BORHAN PROCESS D E S I G N A P P R O A C H

Design basis Our process design basis was a H2 production rate of 0.0705 kg s -1, or 0.784 m 3 s -1 (i.e. at standard conditions of 0°C and 0.101 MPa [1 atm]). This rate has a heating value of 10 thermal M W (MWt) if both the heat of H2 combustion and the heat of condensation of the steam combustion product can be extracted to 25°C. This rate was chosen to limit the electrolysis plant electrical power requirements to only about 1% of the peak electrical output of a coal-fired 1000MWe (2600-3000MWt) power plant. This limitation would be appropriate for a H T E demonstration plant built adjacent to that power plant.

Design assumptions The major assumptions made in our preliminary HTE process design are summarized in Table 1. The minimum hydrogen concentration in the high-temperature electrolyzer was 5 mol. % to prevent oxidation of the electrolysis cell cathodes. This requirement led to the inclusion of a hydrogen recycle stream in all process flowsheets. Use of oxidation-resistant cathodes could potentially reduce or eliminate that recycle stream [6]. The average electrolyzer temperature was 1000°C to give good oxygen-ion conductivity in the solid-oxide electrolyte. For nonisothermal electrolyzer operation, a maximum temperature of 1100°C was chosen to minimize materials problems, while a minimum temperature of 900°C was chosen to minimize conductivity loss in the electrolyte. As a result, the maximum temperature drop was 200°C. Because of the high electrolyzer operating temperatures, the feed gases must be preheated to ll00°C in ceramic tubes. As a result, high operating pressures cannot be used. The H2 product pressures considered were 0.304-1.01 MPa. In addition, evaporating 100152°C water in 1000°C ceramic tubes might cause materials problems, so ceramic process equipment was only exposed to gas streams. It was assumed that 153-189°C saturated steam could be supplied by the adjacent power plant and that high-

Table 2. Parameter values for technical and economic evaluation of alternative processes. Parameter Cost of coal Cost of natural gas Cost of electrical energy Cost of residual fuel oil Annual capital recovery charge Thermodynamic efficiency of electrical power generation Thermodynamic efficiency of steam generation Gas-gas overall heat transfer coefficient [8, 11] Gas-water overall heat transfer coefficient [8, 11] Condensing vapor-water overall heat transfer coefficient [8, 11]

2. 3. 4. 5. 6. 7. 8. 9. 10.

$1.90 GJ -1 $3.30 GJ -1 $13.90 GJ -1 $20 bb1-1 20% 35% 100% 60 W m -2 °C-1 230 W m -2 °C-~ 850 W m -2 °C-1

temperature thermal energy could be extracted in ceramic tubes located in the radiant section of a coal-fired furnace in that plant. The construction of a new furnace to provide both low- and high-temperature process heat was therefore not considered. It was also assumed that treated water at 15°C could be purchased from the power plant, so neW cooling towers and water-softening equipment would not have to be built. The cooling water requirement for the electrolysis plant depends on steam conversion in the electrolyzer, but was usually less that 15 kg s -1. A minimum temperature approach of 10°C was used in heat exchangers to avoid pinch points in condensers and boilers. Using a 10°C approach and conservative overall heat transfer coefficient estimates (see Table 2) led to the design of very large gas-gas heat exchangers, however. A larger temperature approach in these exchangers would have reduced our capital cost estimates, but, as shown later, this would only have affected our HTE H2 production cost estimates at electrolyzer steam conversions below about 50%. An average pressure drop of 34.5 kPa was assumed in all heat transfer equipment and in the high-temperature electrolyzer.

Table 1. Major process design assumptions 1.

Value

Feed stream to HTE unit at least 5 mol.% hydrogen, with the balance steam. Average electrolyzer operating temperature of 1000°C. Maximum temperature drop of 200°C in HTE unit. High-temperature ceramic equipment only exposed to gases at less than 1.22 MPa. No heat addition to process below 900°C. Treated cooling and feed water available at 15°C. Minimum temperature approach in heat exchangers of 10°C. Pressure drop of 34.5 kPa in each piece of process equipment. Hydrogen product stream leaves plant at 25°C saturated with water vapor. Hydrogen insoluble in water.

437

HIGH-TEMPERATURE STEAM ELECTROLYSIS This pressure drop is lower than that needed for good heat transfer in gas-gas exchangers but higher than that usually obtained in boilers and condensers [7]. Because of the temperature approach restriction and cooling water temperature chosen, the minimum H2 product temperature was 25°C. The maximum water content of this stream after condensation separation at 25°C was found to be 1.04mo1.% at a total product pressure of 0.304 MPa, so the cost of additional drying equipment was not considered. The solubility of H2 in water was also neglected since its maximum value at the maximum product pressure of 1.01 MPa was about 0.014 mol.% [8].

where, for our study, AHc = 10 MW, r/E = 0.35, and r/s = 1. The values of E, Q, and S were calculated using the ASPEN-PLUS program. These thermodynamic efficiency calculations were not used in comparing HTE processes with other H2 production processes, however. That comparison was based only on production cost estimates. In addition to producing H2 the HTE process produces pure 02 as a byproduct. The 02 byproduct stream was only considered to the extent necessary for heat recovery equipment design. The H2 production cost estimates might be reduced further if the 02 could be sold. However, additional 02 processing equipment would then be necessary.

Technical and economic analysis of alternative HTE processe~ The ASPEN-PLUS process simulator program [9] was used to calculate material and energy balances for each of the processes considered. Thermal and electrical power requirements were calculated with an error of less than 0.01% by this program. Estimates of capital costs for the alternative HTE processes were started with Westinghouse's predicted electrolyzer cost of $50kWt z H2 produced [10]. Total capital cost estimates were then obtained using the techniques recommended by Peters and Timmerhaus [7]. These estimates were only accurate to within about 30%, but this represents a potential H2 production cost error of just $0.01 ms-3 H2 produced ($0.8 GJ -1) because capital cost accounts for only about 20% of the HTE H2 production cost. The cost and efficiency bases used to compare alternative H2 production processes are summarized in Table 2. Both the H2 production cost and thermodynamic efficiency of alternative HTE processes were considered when comparing those processes. Thermodynamic efficiencies were calculated using the formula:

H2/H20 ]..... _ OXYGEN

1-

SELECTION OF A L T E R N A T I V E HTE PROCESSES A flowsheet for the HTE process selected as a base case, Process A, is shown in Fig. 1. In Process A, 153189°C saturated steam is heated to about 1000°C and then electrolyzed nearly isothermally with electricity derived from coal. No ceramic steam-preheating tubes installed in a coal-fired furnace are used, so all the hightemperature energy input must be provided electrically. Process A was selected as the base case because the high electrical power consumption gave an upper bound on H2 production cost. It also gave a lower bound on the thermodynamic efficiency with which the thermal energy of coal can be converted into H2. A flowsheet for the second HTE process considered, Process B, is shown in Fig. 2. In Process B, low-pressure saturated steam is again fed to the HTE plant. The steam is first preheated and then heated to ll00°C in ceramic tubes installed in the radiant sectmn of a coalfired furnace. For our assumed 10 MW t H2 production rate, the required furnace duty is about 1-2 MW t. Finally, the steam is electrolyzed with an accompanying temperature drop of 200°C. In this way, some electrical

q_

HYDROGEN R E C Y C L E

HYDROGEN

RESSOR

]

SUPERHEATED STEAM=I STEAM HEATER

-[

STEAM HEATER

I'~'~ LOW-GRADE STEAM HOT WATER

=

COLD I CONDENSER

F CONDENSATE

Fig. 1. Flowsheet for Process A: steam feed and near-isothermal HTE operation.

438

M. A. LIEPA AND A. BORHAN

H2/H20 I' I,,1

HTE

F

HYDROGENRECYCLE ~ = ~ . ) R ESSOR HYDROGEN COAL-FIRED FURNACE I-- I

I OXYGENBYPRODUCT SUPERHEATEDSTEAM

=

LOW-GRADE STEAM HEATER

COLD

HOT WArE

t

CONDENSER

C

CONDENSATE

Fig. 2. Flowsheet for Process B: steam feed and nonisothermal HTE operation.

power input is replaced by direct use of thermal energy. As a result of this substitution, Process B has a lower H2 production cost and higher thermodynamic efficiency than Process A. A flowsheet for the third HTE process considered, Process C, is shown in Fig. 3. In Process C, steam is generated from water and preheated by heat exchange with the product gases before again being heated to 1100°C by a coal-fired furnace. The remainder of Process C is the same as Process B. Process C was studied to determine the effect maximizing heat recovery from the product streams would have on H2 production cost and thermodynamic efficiency.

2/H2 o

HTE

.~.OXYGEN

m

COAL-FIRED F URNACE

T E C H N I C A L A N D E C O N O M I C RESULTS FOR A L T E R N A T I V E HTE PROCESSES Process A: steam feed and near-isothermal HTE operation The process flowsheet for our base case, Process A, is shown in Fig. 1. All of the high-temperature energy input to the process is provided electrically. The first part of that electrical energy input drives the watervapor electrolysis reaction, while the second part helps maintain nearly isothermal electrolyzer operation. The first component corresponds to AGe, the second represents the positive difference between AHe and AGe.

HYDROGEN RECYCLE (~

4

HYDROGEN COMPRESSq

I

SUPERHEATEDSTEAM STEAM HEATER STEAM HEATER

lira

,I--

oxY

OXYGEN COOLER "="

STEAM COMPRESSOR

BYPR )DUCT

BOILER/

CONDENSER

CONDENSER~"~--~ CONDENSATE I

HYDROGEN PRODUCT

~"'~TER FEED W ATER~:~ PUMP Fig. 3. Flowsheet for Process C: water feed and nonisothermal HTE operation.

HIGH-TEMPERATURE STEAM ELECTROLYSIS

i

ILl 0.32

r

i

I

028

i

a

rr

.,.¢

%

439

~

0.28

024

I

i

I

I

I

./w'2;°j ZiLs ;7 PRODUCTION COST

% 0.2C

O24 121

i

Z ~ 016

0.2C

~2

PRODUCTIONCOST WITH FREE STEAM

I.- 0 1 6 o9 0 t.)

~ 0.12 O ~ ~ ' ~

ELECTRICALPOWI.~RCOST

Z Q 0.0£

012

,~~ELECTRICAL

- -- - - Process A neer-isothermal HTE operation Process 8 nonisotherrnol

D

8nr 0.04 Q,,. z ,~ O.OC

,', 0 0 8 0 rr o.. z 004 0

rr c3

>- OOC T

I

0,2 STEAM

1

1

l

0.4 0.6 0.8 CONVERSION FRACTION

O I If)

POWER COST

012 STEAM

HTE operation

I I 0.4 0.6 08 CONVERSION FRACTION

I0

Fig. 5. Hydrogen production cost for Process B: steam feed and nonisothermal HTE operation.

Fig. 4. Hydrogen production cost for Process A: steam feed and near-isothermal HTE operation.

In Process A the latter is provided by electrical resistance heating. Design calculations were performed at 5% steam conversion intervals. For a 153°C saturated steam feed and a 0.304MPa H2 product, the total capital cost estimated for Process A decreased smoothly from more than $11 million for a 15% steam conversion in the electrolyzer to less than $4 million at a 90% conversion. The lowest capital cost is obtained at high steam conversions because less feed steam must be processed. Figure 4 shows the H2 production cost estimated for Process A, of which electrical power makes up about 80%, as a function of steam conversion in the HTE unit. In addition, the capital recovery charge only becomes important at low steam conversions. A steam cost of 0.69c kg -1 is an upper bound [7], so the steam cost is also not very important except at low steam conversions. The thermodynamic efficiency of Process A with respect to the primary energy source (thermal energy from coal) increases from less than 30% for a 15% steam conversion in the electrolyzer to 36.9% at a 90% conversion. The efficiency is poor at low steam conversions because almost all of the unconverted steam is condensed, with the accompanying loss of the thermal energy used to generate that steam.

Process B: steam feed and nonisothermal HTE operation The second process flowsheet considered, Process B, is shown in Fig. 2. Process B is similar to A except that a coal-fired furnace has been fitted with ceramic tubes to preheat the electrolyzer feed stream to 1100°C. This

extra heat supplied by the furnace permits a temperature drop of 200°C from the inlet to the outlet of the HTE unit, while still maintaining a 1000°C average operating temperature. In this way, direct use of less expensive thermal energy ($1.90 GJ -~) is substituted for electrical resistance heating in the electrolyzer ($13.90 GJ-1). At a steam conversion of 30% in the electrolyzer, for example, the total electrical power consumption is reduced by about 15 % (from 11.47 MJ m q3 H2 produced in Process A to 9.76 MJ m-~3 in Process B). In addition, the enthalpy of the 02 product stream is not recovered, eliminating one steam-preheating exchanger. Eliminating the 02 heat exchanger increased the heat duty required in the furnace by up to about 0.2 MWt, but also reduced capital cost. This elimination would be less economically attractive in Process A, where the lost heat recovery must be replaced by electrical resistance heating in the electrolyzer. The total capital cost estimated for Process B was lower than that for A at all steam conversions, but the difference may not be significant because our capital cost estimates were only accurate to within about 30%. Figure 5 shows the H2 production cost estimated for Process B as a function of steam conversion in the HTE unit. The electrical power cost for Process B is lower than that for A because some thermal energy has been substituted for electrical power. The fraction of thermal energy substituted is greater at lower steam conversions, where the electrolyzer feed stream has a higher initial enthalpy content. The total H2 production cost was lower for Process B than for A because of the reduced electrical power cost, as was the thermodynamic efficiency, as shown in Fig. 6, because less electrical power (generated at a 35% conversion efficiency) is used.

440

M. A. LIEPA AND A. BORHAN 50

I

I

I

Table 4. Comparison of H2 production costs for HTE and competing processes

I

40

Process Z LJ

t 3O

Hydrogen production cost* ($ mq3 H2 produced)

t

Catalytic steam reforming of natural gas

f

0.07-0.10 0.08--0.11 0.10--0.12 0.11-0.14 O. 12-0.16 0.14-0.17 O. 17-0.20 0.17-0.22

s/

Partial oxidation of residual fuel oil

-- - - - - Process A~ neor-isothermol HTE operolion

a 0 w ..r F-

Process B: nonisothermol HTE

operotion

Koppers-Totzek coal gasification

I0

High-temperature steam electrolysis I Q2 STEAM

I O4

/ 06

CONVERSION

I 0.8

I0

FRACTION

Fig. 6. Comparison of thermodynamic efficiencies for Processes A and B: steam feed.

[3] [4] [12] [4] [3] [3] [4] [13]

Advanced solid-polymer electrolysis

0.22-0.24 [5]

Conventional electrolysis

0.34--0.41 [4]

* Cost bases are given in Table 2.

Process C: water feed and nonisothermal HTE operation The third process considered, Process C, is shown in Fig. 3. Rather than feeding saturated steam to the system, the enthalpy of the product streams is used to boil liquid water within the process. As in Process B, the electrolyzer operates with a 200°C temperature drop and coal-fired heat is used to preheat the electrolyzer feed stream to ll00°C. In order to generate steam internally, the pressure of the steam-hydrogen product stream must be higher than that of the feed water so that its condensing temperature range will also be higher than the boiling temperature of the feed water. The pressure difference must be increased in order to operate at high electrolyzer steam conversions because the dew point of the steam-hydrogen product mixture decreases with decreasing steam mole fraction unless the total product pressure is increased. When boiling water at 0.101 MPa, steam conversions up to about 35% can be achieved for a H2 product pressure of 0.304MPa. This upper limit increases to about 55% for a H2 product pressure of

1.01MPa. A centrifugal compressor was used to increase the pressure of the saturated steam leaving the boiler. The electrical power cost is higher for Process C than for B because of the addition of a steam compressor. This effect is greatest at low steam conversions, where the steam generation rate is higher. The total production cost is about 10% greater for Process C than for B, primarily because of the steam compression cost. However, the thermodynamic efficiency is higher for the same steam conversion because of greater heat recovery. For example, the thermodynamic efficiency of Process B is 34.7% for a 30% steam conversion in the electrolyzer, while that of Process C is 38.5%. Use of higher product pressures reduces efficiency because more electrical power is needed for steam compression. Table 3 summarizes the results we obtained through technical and economic analysis of the three alternative H T E processes. Process B had the lowest total electrical power consumption and lowest H2 production cost, while Process C had the highest thermodynamic efficiency.

Table 3. Technical and economic comparison of alternative HTE processes HTE process A B C

Steam conversion fraction

Electrical power consumption (MJ m-] H2)

0.30 0.90 0.30 0.90 0.30

11.47 11.25 9.76 10.66 11.03

* Cost and efficiency bases are given in Table 2. t Includes steam cost of 0.69c kg -~.

Hydrogen production cost* ($ mq3H2) 0.22 0.19 0.18 0.17 0.21

(0.24t) (0.19t) (0.20t) (0.18t)

Thermodynamic efficiencyt

(%)

31.9 36.9 34.7 38.2 38.5

HIGH-TEMPERATURE STEAM ELECTROLYSIS C O M P A R I S O N O F H T E W I T H O T H E R H2 P R O D U C T I O N PROCESSES Typical H2 production costs for several H2 producing processes are presented in Table 4 (in 1983 U.S. dollars). Hydrogen production costs obtained from two of the references were recalculated so that cost comparisons could be made on a common raw material cost basis [3, 4]. As can be seen from this table, steam reforming and partial oxidation of hydrocarbons are currently more economical than the HTE process. On the other hand, low-temperature water electrolysis processes are less economical than high-temperature steam electrolysis because of their higher electrical power consumption. Even the most efficient process under development, General Electric's solid-polymer electrolysis process, has a higher H2 production cost. The production costs shown in Table 4 for the Koppers-Totzek coal gasification process are similar to our estimates for the H T E process. This is somewhat misleading because these coal gasification production costs were based on a 399 M W t H2 plant capacity. The coal gasification, steam reforming, and partial oxidation processes all have significant reductions in production cost with increased plant size due to economics of scale [4]. A t a 399 M W t H2 plant capacity, about 75% of the Koppers-Totzek H2 production cost of $0.16--0.18 ms 3 H2 produced ($13-14 GJ -~) is accounted for by the 20% annual capital recovery charge. Decreasing the plant capacity to 10 MW t would cause the KoppersTotzek H2 production cost to jump to about $0.40-0.46 ms 3 produced ($31-36GJ-1). Hydrogen production costs for the low- and high-temperature electrolysis processes are not strongly affected by plant size because electrolyzer cost is almost directly proportional to plant capacity. Two of the three catalytic steam reforming H2 production cost estimates given in Table 4 were also based on a 399 M W t H2 plant capacity [3, 4]. In these two cases, only about 33% of the H2 production cost of $0.08--0.11 m~-1 H2 produced ($6-9 G J-1) was accounted for by the capital recovery charge. As a result, decreasing the plant capacity to 10 MW t H 2 only increases the steam reforming H2 production cost to about $0.140.18 m~-3 produced ($11-14 GJ-1). More significantly, about 66% of the catalytic steam reforming H 2 production cost is accounted for by the cost of the natural gas feedstock. An increase in the cost of natural gas from $3.30 to $7.60 GJ -1 would increase the steam reforming H2 production cost from the value given in Table 4 ($0.08-0.11 ms 3 H2 produced) to about $0.16-0.20 m~-3 H2 produced ($13-16 GJ-1). This production cost is comparable to the authors' HTE H 2 production cost estimates of $0.17-0.22 m~-3 H2 produced. CONCLUSIONS Based on performance and cost figures available in 1983, high-temperature electrolysis coupled with thermal and electrical energy derived from coal is not yet

441

competitive with processes which make H2 from hydrocarbons. If natural gas costs were to double, however, catalytic steam reforming and H T E would have comparable H2 production costs. Electrical power cost is the major component of H T E H2 production cost. For electrolyzer steam conversions greater than about 50%, electrical power cost makes up about 80% of the H2 production cost. In addition, replacement of electrical power input with thermal energy reduces HTE H2 production cost. Production cost reductions of 10-15% can be obtained in this way. The capital cost for a H T E process is only an important component of H2 production cost for low electrolyzer steam conversions. This gives H T E an advantage over coal gasification, where capital cost is the major component of production cost. Finally, high electrolyzer steam conversions give the best thermodynamic efficiency in converting coal-fired heat into H2. A thermodynamic efficiency of 30.6% for a 15% steam conversion can be increased to 38.2% at a 90% conversion. Acknowledgements--The authors thank T. A. Hatton, A. D. Richards, J. W. Tester, H. P. Meissner and other members of the MIT staff for their guidance and support, and for their critical review of our work. Thanks are also due to G. T. Skaperdas, A. Mezzina and F. J. Salzano of Brookhaven National Laboratory for the constructive criticism and suggestions provided during our study. This paper is a summary of work performed at Brookhaven National Laboratory under the auspices of the U.S. Department of Energy. A more detailed description of our study is available elsewhere [13].

REFERENCES 1. W. Doenitz, R. Schmidberger, E. Steinheil and R. Streicher, Hydrogen production by high temperature electrolysis of water vapour, Int. J. Hydrogen Energy 5, 55-63 (1980). 2. A. O. Isenberg, Energy conversion via solid oxide electrolyte electrochemical cells at high temperatures, Solid State lonics 3/4, 431-----437(1981). 3. H. G. Comell and F. J. Heinzelmann, Hydrogen in oil refinery operations. In W. N. Smith and J. G. Santangelo (eds), Hydrogen: Production and Marketing, ACS Symposium Series No. 116, pp. 67-94. American Chemical Society, Washington, D.C. (1980). 4. D. P. Gregory, C. L. Tsaros, J. L. Arora and P. Nevrekar, The economics of hydrogen production. In W. N. Smith and J. G. Santangelo (eds), Hydrogen: Production and Marketing, ACS Symposium Series No. 116, pp. 3-26. American Chemical Society, Washington, D.C. (1980). 5. D. K. Gupta and J. H. Russell, Solid polymer electrolysis economic and design study, U.S. Department of Energy Contract No. DE-AC02-78ET26202, General Electric Company, Wilmington, Massachusetts (1981). 6. A. Mezzina and M. Bonner, Visit to Westinghouse R&DAugust 3--4, Memorandum, Brookhaven National Laboratory, Upton, New York (1983). 7. M. S. Peters and K. D. Timmerhaus, Plant Design and Economics for Chemical Engineers, 3rd edn. McGrawHill, New York (1980).

442

M. A. LIEPA AND A. BORHAN

8. R. H. Perry and C. H. Chilton, Chemical Engineers' Handbook, 5th edn. McGraw-Hill, New York (1973). 9. Aspen Technology, Inc., ASPEN PLUS Introductory Manual. Cambridge, Massachusetts (1983). 10. J. A. Fillo, HYFIRE: Fusion/high temperature electrolysis conceptual design study--annual report, Report No. 33701, Brookhaven National Laboratory, Upton, New York (1983).

11. J. P. Holman, Heat Transfer, 5th edn. McGraw-Hill, New York (1981). 12. M. Belier, Personal communication, Brookhaven National Laboratory, Upton, New York (1983). 13. M. A. Liepa and A. Borhan, High-temperature steam electrolysis: technical and economic evaluation of alternative process designs, Report No. 51798, Brookhaven National Laboratory, Upton, New York (1983).