Solar Energy Materials 24 (i991) 464-477 North-Holland

Solar Energy Materials

Closed-loop operation of a solar chemical heat pipe at the Weizmann Institute solar furnace

Rachel Levitan, Moshe Levy *, Hadassa Rosin and Rachamim Rubin Materials Research Department, Weizmann Institute of Science, Rehovot, Israel

The performance of a solar chemical heat pipe was studied using CO 2 reforming of methane as the vehicle for storage and transport of solar energy. The endothermic reforming reaction was carried out in an lnconel reactor, packed with a rhodium catalyst. The reactor was suspended in an insulated box receiver which was placed in the focal plane of the Schaeffer Solar Furnace of the Weizmann Institute of Science. The exothermic methanation reaction was run in a tubular reactor filled with the same Rh catalyst and fed with the products from the reformer. Conversions of over 80% were achieved for both reactions.

In the closed-loop mode the products from the reformer and from the ~ethanator were compressed into separate storage tanks. The two reactions were run consecutively and the whole pro~d.~s was repeated for nine cycles. The overall performance of the closed loop was according to expectations.

i . Introduction

A solar chemical heat pipe uses a reversible chemical reactton as a means of storage and transport of solar energy. The reaction is supposed to I:.e operated in a closed loop for many cycles. Therefore, it should be free of any side-reactions that may result in accumulation of undesirable products. The reaction under s tudy is the C O 2 reforming of methane:

CH 4 + C O 2 ~ 2H 2 + 2CO

which is always accompanied by the reverse water gas shift reaction:

CO 2 + H 2 ~ CO + H 2 0

When the initial C O 2 / C H 4 ratio is kept only slightly above unity, the water formed in the reaction amounts to only a few percent and will stay in the gas mixture.

The side-reactions that can be problematic are those leading to carbon forma- tion:

CO + H 2 ~ C + H20

2CO ~- C + C O 2

CH 4 ~ C + 2H 2

These reactions, if not kept under conirol, may eventually lead to blocking of the pipes and the whole process will be interrupted. Therefore, special a t tent ion should be given to any signs indicating the beginning of carbon deposition.

0165-1633/91/$03.50 © 1991 - Elsevier Science Publishers B.V. All rights reserved

R. Levitan et al. / Closed-loop operation of solar chemical heat pipe 465

The reaction was first studied in the laboratory under well controlled conditions. The catalyst chosen was rhodium on alumina pellets. It proved to be quite stable under rather drastic conditions of changes in temperatures and flows of reactants. We studied the kinetics of the reforming and the methanation reactions. We ran the reforming and the methanation separately, then in series, and finally in a dosed loop. We operated the closed loop for up to 1100 cycles with no apparent deterioration in the catalyst activity or accumulation of any side-products [1].

The reforming reaction was then studied under real conditions in the solar furnace. A number of receiver/reactor configurations were tried [2-4]. The directly heated vertical tube configuration proved to be the best for the closed loop. We connected it to a large methanator and operated the loop for a number of cycles. In this presentation, a description of the reformer and the methanator will be given and preliminary results for the performance of both in a closed loop.

2. Reforming

All the reforming experiments were carried out at the Schaeffer Solar Furnace of the Weizmann Institute. A detailed description of the furnace and its performance was given previously [2].

The vertical receiver/reactor used in this work is shown in fig. 1. The receiver is an aluminium box, insulated from the inside by a 5 cm thick alumina blanket. The inside dimensions of the box are 20 × 29 x 60 cm. The aperture is a circle, 10 cm in diameter, in the front panel. The reactor was made of a 24 mm o.d. (20 mm i.d.) Inconel 600 tube, 130 cm long, bent in a U-shape. 115 cm of the tube were filled with catalyst pellets, the rest was filled with alumina pellets of the same size. The catalyst used, was Engelhard (E4823), 0.5% Rh on alumina. The receiver aperture was placed in the focal plane of the solar furnace. The reactcz tube was suspended 18 cm behind the aperture. In order to avoid excessive overheating of the front surface of the reactor tube, two ceramic tubes, 15 mm in diameter, were placed in front of the reactor and thus all the surfaces of the reactor were protected from direct irradiation by the concentrated solar beam. Using this arrangement, the difference in temperature between the front and the back of the reactor, did not exceed 50 o C.

The total power entering the receiver was regulated by opening and closing of the doors of the building housing the concentrating mirror. The rate of flow of reactants was changed in accorda.ace with the desired conversions, and the reaction pressure was changed by a pressure regulator located at the exit of the reaction system. Due to materials limitations the maximum wall temperature of the Inconel reactor wall was not allowed to exceed 960 ° C. The ratio C O J C H 4 was maintained within the range of 1.1-1.3. The gas flows ranged between 1000-11000 L /h , resulting in Reynolds numbers of 200-1600 and space velocities of 8000-50000 h -~. Under these conditions, maximum conversion of 8470 was reached, The power input was up to 6.4 kW and the power flux on the reactor wall reached 100 k W / m 2.

466 R. Levitan et al. / Closed-loop operation of solar chemical heat pipe

PR(X)UCTS OUT~I

INSUI

~ , ,,t--- FEED IN ® TG

? 'W

Fig. 1. Diagram of the receiver and the reactor. TG: gas mixture temperature. TW: front wall temperature, bTW: back wall temperature.

A typical data acquisition record of one working day is shown in fig. 2. Fig. 2a shows the measured insolation during the day, fluctuating around 700 W / m 2 and dropping down sharply only in the late afternoon. The product gas temperature TG115, measured at the end of the catalyst column, was in the range of 750-800 o C. The wall temperature TWll5, at the same position, was 900-930 o C. These temper- atures were obtained by adjusting the flows of the feed gases shown in fig. 2b. The sharp fluctuatitms were obsereed only when the flow regime was changed and they only lasted for short periods.

A large namber of reforming experiments was carried out under real solar condition,s. P- number of reaction parameters were varied including the reaction pressure (2-8.4 atmospheres)and the rate of flow of reactants (2000 to 10000 L/h). A representative summary of the results is given in table 1.

G 0

0 t- I--

"0

E m

a

R. Levitan et al. / Closed-loop operation of solar chemical heat pipe

' , l Tw°c ~ - - - - ~ ~ ~ ~ . .

o.a. ~TG~$,....,.~.. jr':,. 0.7. f

0.6'

051 0.4 03-', ,

o ,6o ' 2bo ' 3bo time, rain

I

467

28

24-

"~ 20- c: - C H 4 0 u'}

= 1 6 - 0 { -

I - -

1 2 - 0 .

08-

°:t 0 b

r ~

4o 2~o 360

~4me, mln

Fig. 2. (a) TG and TW in centigrade, radiation in W / m 2 as a function of time (rain). (b) Flow rates of CO 2 and CH 4 in L / h as a function of time (min).

A computer model for simulation of the experiments was developed. It is described in detail in another paper [5]. The modeling procedure starta by estimat- ing the total power from the experimentally measured absorbed power and the calculated receiver efficiency. From that, the flux distribution of the solar insolation at lhe focal zone is calculated, followed by its distribution o:', the receiver and the reactor walls. The heat is then transferred through the reactor wall into the gas mixture, raising its temperature and supplying the enthalpy for the chemical reaction taking place, Ei~ergy balance is performed by iterations until convergence is reached for the % conversion of methane, within a given margin of error. The output of the program includes temperature profiles of the wall and gas, composition of the pr,aduct mixture, and energy fluxes.

468 R. Levitan eta!. / Closed-loop operation of solar chemical heat pipe

Table 1 CO 2 reforming of methane

CH 4 CO 2 GHSV Re P Rad Door TG0 TGll5 TWll5 bTWll5 (L/h) /CH 4 th-l) (atm) (W/m 2) (m) (o C) (o C) (°C) (o C)

1514 1.22 8976 272 2.0 579 5.0 449 746 884 838 2050 1.24 12219 386 2.0 635 5.5 402 ~17 878 820 2850 1.10 15963 508 2.6 673 6.0 365 700 901 828 3440 1.17 19899 662 2.9 664 6.0 324 672 892 811

1350 1.02 7264 206 4.8 705 5.5 471 788 899 857 1916 1.22 11325 347 3.9 569 5.0 436 714 836 807 2916 1.15 16699 533 3.5 768 6.0 349 719 859 847 4246 1.17 24555 822 4.2 788 6.0 297 686 835 826

1435 1.18 8328 237 5.7 740 4.5 492 811 891 875 1894 1.24 11315 339 7~ 682 5.0 441 767 878 857 3744 i.25 22464 754 7.0 703 6.0 312 679 844 810 4532 1.20 26624 925 7.4 745 6.0 283 633 787 763

CH 4 H 2 CO CH 4 CO 2 H20 COnV. QE Qr Qt Rate (L/h) (%) (%) (%) (%) (%) (%) (W) (W) (W) (mol/g. h)

1514 42.0 42.1 7.7 8.2 0.0 73.2 875 3135 40i0 3.0 2050 36.7 39.2 11.3 11.5 1.2 62.7 1132 3683 4816 3.4 2850 33.1 35.3 16.6 13.9 1.1 50.7 1532 4140 5673 3 9 3440 27.0 31.0 21.4 18.6 2.0 40.3 1850 4028 5878 3.7

1350 38.7 44.5 3.6 10.3 2.9 85.4 723 3328 4051 3.1 1916 29.8 38.8 9.9 17.0 4.5 63.4 921 3573 4495 3.2 2916 24.5 32.1 19.7 19.9 3.8 41.8 1644 3612 5255 3.3 4246 18.4 26.0 25.2 26.6 3.8 30.5 2376 3883 6259 3.5

1435 34.3 44.5 4.3 11.9 5.1 82.2 810 3452 4272 3.1 1894 28.7 39.9 9.4 16.4 5.6 64.7 1056 3640 4695 3.3 3744 20.2 25.4 22.5 29.3 2.6 33.6 2067 3691 5758 3.4 4542 17.4 23.5 27.3 28.8 3.0 27.2 2275 3680 5955 3.3

Qh - power absorbed as sensible heat. Qr - power absorbed in the reaction. Qt - total power absorbed.

A representa t ive example of the ca lcu la ted t e m p e r a t u r e prol~le is s h o w n in fig. 3.

I t can be seen that the ag reemen t wi th the experimet/~al po in t s is qu i te good .

However , it shou ld be stressed tha t no t all the expe r imen t s s h o w e d such a g o o d

agreement .

In o rder to c o m p a r e the expe r imen ta l and the ca lcu la ted results, we selected 45

exper iments where the average wall t em pe ra tu r e at the e n d of the ca ta lys t bed , T W ,

was cons tan t , 860 _+ 1 5 ° C . A p lo t o f these values versus the rate o f the m e t h a n e

feed shows that the calcula ted T W was also cons t an t , a l t h o u g h a b o u t 20 ° C h igher

t han the exper imenta l T W (fig. 4). The solid line represen t s the l eas t - square fit for

R. Lecitan et aL / Closed-loop opereiion ,~1 ~olar chemical heat pipe 469

9ooi-

~ 8oo|..-- - ' ~ - ~ ' - - ' ~ - ~ '" / "~ 700]_ ° ~ T""~

5 0 0 1 ~ "

400 t- I I I I I I 0 20 40 60 80 I00 120

Catalyst bed length (cm)

Fig. 3. Calculated temperature profiles of TG and average TW as a function of the catalyst bed fraction length. The symbols are the experimentally measured values.

the experimental points and the dotted line, for the corresponding calculated values. The TG decreases with increasing flow, as expected, and the agreement between calculation and experiment is very good. The total power absorbed, Qt, h~creases with the rate of flow of the reactants (fig. 5) flow, in spite of the fact that the % conversion decreases. This indicates that the total heat transfer is higher at higher flows because of improved turbulence. The agreement between calculated and experimental values of Qt is good at low flows, but at higher flows the deviation becomes rather large.

As mentioned above, the criterion used for convergence in the calculations was the % conversion of methane. Using the kinetic equations derived previously, we could also calculate the concentrations of all the gases in the product (figs. 6a and 6b). The calculations show higher values for H20 and CO and lower values for H2

1 0 0 0

9 0 0

soo ! - -

L9

700

o v ~ Expt

x • .... CoIc

: ..... .-. :.,,,..~ ........ o........=..-...~..-...,,... ~w y,-'_ _ v,~t~...~ ~, t _v ~, g~v_~,," .......... -.

o

- ~ , ~ , ~ x^ o x Z ~ " ~ - ~ ~ ,, o o

6 0 0 t 1 . , _ _ i I i

1 0 0 0 2 0 0 0 3 0 0 0 4 0 0 0

CH4 ( I / h ~ Fig. 4. Comparison of experimental and calculated wall and product gas temperatures, o C, as a function

of methane flow, L/h.

470 R. Levitan et al. / Closed-loop operatio:~ of solar chemical heat pipe

5

o 3

7

8

2

1

[ ] x '~x_x ~ , . l ~ , . j ~ l ; l ~

• . . . . . . . . . . . . . . . . . .

XX

~ ^ ~ x

D x

a ~ E~pt

x . . . . Co Ic

0 i I , i i

t000 2000 aooo 40oo

CI-I 4 (I/h) Fig. 5. Comparison of experimental and calculated total power absorbed, kW, as a function of methane

flow, L /h .

and CO2. It indicates that the rate equations used for the reverse water gas shift reaction should be modified.

From these experiments we concluded that the reforming reaction is reasonably well under control and it is possible to use this reactor for the closed-loop operation.

3. Methanation

The methanator was build for the sole purpose of closing the loop and was not intended, at this stage, for energy measurements and calculations. A diagram of the methanation system is shown in fig. 7. Its main components are the reactor tube, the gas preheater, two steam generators, two flow controllers for the gas, and a metering pump for water. The methanator (M) is made of a 1" OD SS tube, packed with catalyst - 0.5% Rh on alumina pellets. The length of the catalyst bed is 90 cm. The reactor is insulated with alumina wool. The temperature is measured along the catalyst bed at 15 cm intervals. Gas samples are withdrawn from positions 15, 45, 75 and 90 cm along the catalyst bed, and pass through the condensers (C) and the automatic sampling valves to the analysis system (GC). The feed gas for methana.- tion (a pure CO + H E mixture from a cylinder or the product gas from the solar reformer) enters into the system through a valve (V) and the mass flow control!or (MFC 1). Two additional gas lines are combined with the feed line: (a) Argon through MFC 2, to purge the system or dilute the reactants; (b) CO2 which flows through the system during start up, until steady temperatures are reached.

R. Levitan et al. / Closed-loop operation of solar chemical heat pipe 471

N 0

,¢ -r-

a

5 0

40

3 0

2 0

tO

C02 CH4

V Q

x Q

- - Expl

. . . . Colc

v ,j~ w ~. ~ ..............

c . ,

I I i I i

2000 3000

CH4 (I/h)

0 1000 4000

5O H20 H~ CO

~... ...... • e ; o a v ~EKpt

4 0 " ~ ...... . * . o x , - Calc

X xX X

2 0 Hz I

10

. . . . . . . . . . . . . . .

1000 2000 3000 4000

b CHn(I/h) Fig. 6. (a) Compar i son of experimental and calculated % CH 4 and C O 2 in product gases, as a funct ion of methane flow, L / h . (b) Compar ison of exper imental and calculated% H20 , H2 and C O in p roduc t gases,

as a funct ion of methane flow, L / h .

The feed is preheated in H, which is a 0.5" SS tube, packed with ceramic rings, and electrically heated. The hot feed gas is combined in an insulated and heated line with the superheated steam. The superheated steam at 600 °C is combined with the feed gas at the inlet into the methanator, at this point the mixture is heated to 350-400 ° C, which is the "ignition temperature" for the methanation reaction. The methanation products pass through the water cooled condenser C, where the water

472 R, Levitan et al. / Closed-loop operation of solar chemical heat pipe

• GA~ SAMPLING

UlSV~ATEO LINE

METI-IANATOR SYSTFM O l l E M P E f l A I ~ ( IOIEASI, WIE I~E N l

~1 I EIMpEr#A~IJI"IE CONTIIOL L ,r t."

"31 ti ~,.vts ; I ;2 - '

I I l'lK_

,,,,,,En r.--I ,!

f, -t ~.1"

INUt] °I~S N

I'UIAP

L 0

0

t;

P

I /,/) PIICDLICT OAS OUT

7 h . - - " - IIILHT GAS IN

,¢ t ~ ~ C t l ~

INFC

Fig. 7. Diagram of the methanation system.

is condensed. The dry gas is either vented or compressed into storage cyl inders and

returned into the reformer for closed-loop operat ion. The pressure in the system is controlled by the Grove pressure regulator (G), which is opera ted by an external gas

pressure. The pressure in the me thana to r was 1 -5 ~tm. The feed ranged f rom 1000 to 1500 L / h .

The first series of experiments were run with a mix ture of C O / H 2 = 0.9 f rom a cylinder. A short summary of the results is presented in table 2. The gas composi -

t ion is based only on the carbon containing gases in o rde r to be able to compa re

Table 2 Methanation of a carbon monoxide/hydrogen mixture

Run TG Feed CO CH 4 C O 2 C O 0 2 / H 4 no. (kg/h) (%) (%) (%) cony.

(%)

* 100.0 0.0 0.0 0.90 1 447 0.36 0.6 47.4 52.0 99.4 1.00 2 486 0.36 3.1 45.3 51.6 96.9 0.94 3 496 0.36 3.8 42.4 53.7 96.2 1.04 4 449 0.36 0.7 48.1 51.2 99.3 0.93 5 563 0.51 16.0 35.6 48.4 84.0 0.93

563 0.51 10.1 39.4 50.5 89.9 1.00 7 583 0.51 19.4 35.0 45.7 80.6 0.95 8 561 0.51 14.8 37.8 47.4 85.2 0,98

* Feed composition based on C-containing gases.

R. Levitan et at.. / Closed-loop operation of solar chemical heat pipe

Table 3 Methanation of the reformer products

473

Run TG Feed CO CH4 CO 2 CO no. (kg/h) (%) (~) (t~) conv.

(~)

O2//H 4

* 1.20 49.7 25.0 25.3 1.02 1 ~ 501 9.6 41.1 49.3 80.6 1.11

1.22 50.8 24.4 24.8 1.06 2 520 10.9 41.4 47.7 78.6 1.12

1.24 52.7 306 !6.7 0.79 3 542 11.2 42.0 46.7 78.7 1 A2

1.13 67.6 16.1 16.2 1.06 4 519 13.6 41.2 45.2 79.9 1.02

1.22 58.2 17.4 24.4 1.32 5 528 11.8 43.2 45.0 79.8 1.07

1.23 68,5 16.7 14.8 1,05 6 512 10.9 43.7 45.4 84.1 1.06

1.28 65.8 19.0 15.2 1.03 7 419 5.4 46.0 48.6 91.7 1.06

1.29 61.6 21.4 17.0 1.02 8 358 9.5 44.6 45.8 84.6 1.10

1.27 60.1 23.6 16.3 0.96 9 368 8.3 47.0 44.7 86.2 1.01

1.27 64.9 20.6 14.5 0.99 10 431 4.0 46.4 49.7 93.9 1.05

1.71 65.4 18.6 16.1 1.10 11 569 6.3 44.3 49.5 90.4 1.06

1.73 68.2 18, ° 13.0 1.02 12 561 6.3 44.1 49.7 90.8 1.08

1.78 66.8 18.1 15.1 1.05 13 477 14.5 42.8 42.7 78,3 1.13

* First lines feed compositions. a~ Second lines product compositions.

methanation and reforming experiments. In the experiments where the feed flow was 0.36 kg/h, the CO conversions were higher than 95%. In the others they were over 80%. In the latter experiments the exit gas temperatures were fairly high and therefore conversion was not complete. The O2/H4 ratio (i.e. O2/2H2) which was used for monitoring the overall composition remained fairly steady within experi- mental error.

The next series of experiments involved solar reforming followed directly by methanation, table 3. Because the capacity of the reformer was higher than that of the methanator, only a fraction of the refermer products were directed into the methanator. The rest was vented off. In these experiments the flow of reactants into the methanator was higher and accordingly the CO conversions were somewh~ lower, ranging between 79 and 94%. The feed composition in this case was not constant and varied according to the conditions of the reforming. However, except for a couple of exceptions the O2/H., ratios remained constant within 10%.

474 R. Levitan etal. / Closed-loop operation of solar chemical beat pipe

SOLAR ENERGY

STORAGE CH4÷ CO z

~ - -~CO~H2

Fig. 8. Diagram of the dosed loop.

4. Closed loop

A schematic representation of the closed-loop system is shown in fig. 8. The reformer and the methanator are connected via a compressor to two separate storage tanks. The two reactions were run consecutively. The working procedure was the following: The reformer was first heated by the sun, to the appropriate temperature, under a flow of CO 2. At this stage methane was introduced from a storage tank and the flow controllers kept the ratio between the two gases constant. We usually worked at a CO2/CH 4 ratio of 1.1 to 1.3. The reforming was done under conditions where the product gases, at the exit of the reactor tube, were at a temperature close to 850 °C and therefore the conversions should be 80~ or above. The product gases were compressed into a battery of 20 cylinders, connected in parallel, up to a maximal pressure of 25 atm. After the reforming was finished, the reformer was allowed to cool down, under a flow of carbon dioxide, and the methanation was started. The methanator tube was heated under a flow of CO2 and steam, and when stationary conditions were reached, the CO 2 flow was turned off and a constant flow of reactants was introduced from the C O / H 2 storage. The reaction mixture was preheated to a temperature that will allow the initiation of the methanation, namely about 400 ° C. The temperature along the methanation tube was monitored. It raised fairly quickly to 700 °C and higher and then the tempera- ture dropped down to 3 5 0 ° C - 4 0 0 ° C at the end of the tube. At these exit temperatures the CO conversions were usually over 80%. The methanation products, which consist mainly of a mixture of CH4 and CO 2, were compressed into a battery of 10 cylinders. This concluded the first loop. The second and other loops started from the methanation products and followed the exact same procedure. Altogether,

R. Levitan et al. / Closed-loop operation of solar chemical heat pipe

Table 4 Closed-loop operation

Reforming

475

Loop TG Feed CO CH 4 CO 2 CH 4 O2 /H 4 no. (kg/h) (70) (70) (%) cony.

(70) 1 ~ 4.02 0.0 46.2 53.8 1.17 * 865 57.2 15.6 27.3 66.3 1.26 2 3.84 9.1 40.2 50.7 1.27

872 60.7 9.6 29.7 76.1 1.52 3 4.10 8.9 39.3 51.8 1.30

900 61.4 8.2 30.4 79.1 1.59 4 3.36 10.9 39.2 49.9 1.28

847 60.5 8.9 30.6 77.2 1.60 5 3,26 10.0 40.8 49.2 1.21

893 65.1 8.2 26.7 79.9 1.49 6 1.85 9,4 38.6 52.0 1.32

888 65.3 7,0 27.7 81.9 1.54 7 1.25 11.9 37.2 50,9 1.35

903 71.1 6.2 22.7 83.3 1.39 8 1,25 11.5 36.0 52.5 1.41

807 64.7 11.4 23.9 68.4 1.31 9 1.29 13.3 36.6 50.1 1.33

883 64.8 10.2 25.1 72.2 1.36

Methanation

Loop TG Feed CO CH 4 CO 2 CO O2 /H 4 no. (kg/h) (~) (70) (70) conv~

(70) 1 ~ 1.26 57.2 15.6 27.3 1.25 • 448 9.1 40.2 50.7 84.1 1.27 2 1.39 60.74 9.6 29.7 1.52

456 8.9 39.3 51.8 85.3 1.30 3 1.26 61.4 8.2 30.4 1.59

437 10.9 39.2 49.9 82.2 1.28 4 !.17 60.5 8.9 30.6 1.60

400 10.0 40.8 49.2 83.4 1.21 5 1.24 65.1 8.2 26.7 1.49

421 9.4 38.6 52.0 85.5 1.32 6 1.24 65.3 7.0 27.7 1.54

464 1 !.9 37.2 50.9 81.7 1.35 7 1.21 71.1 6.2 22.7 1.39

410 11.5 36.0 52.5 83.9 1.41 8 1.30 64.7 11.4 23.9 1.31

366 13.3 36.6 50.1 79.4 1.33 9 1.15 64.8 10.2 25.1 1.36

356 9.2 36.2 54.6 85.8 1.44

a~ First lines feed compositions. * Second lines product compositions.

476 R. Levitan et al. / Closed-loop operation of solar chemical heat pipe

I ' [ ' I J I ' I I 2 .0 • reforming feed "l/

• rnefhanation feed -~ /

1.8~- I

J L4 J

I l

to[ , I . = , i , I 7 I 3 5 7 9

loop number

Fig. 9. O z / H 4 ratio for the different loops.

nine loops were performed with the same initial gases. The results are given in table 4, where the reforming and the methanation data are given separately in spite of the fact that except for the first reforming and the last methanation the compositions are the same. It should be noticed that analysis was performed at the entrance, as well as at the exit of both the reformer and the methanator. However, in the table only the appropriate entrance results are given as they represent an average composition after being allowed to mix in the cylinders, and not the momentary composition that varies according to the conditions prevailing at the time of the analysis. In the reforming experiments the C H 4 conversions were kept high enough by adjusting the feed flow and the door opening to give a product gas temperature higher than 850°C, in most cases. In the methanation experiments, the CO conversions were also fairly high. It should be noticed that the maximal flow of feed gases into the methanator did not exceed 1.4 kg/h due to the size of the methanator tube. The O2/H 4 ratio remained quite steady (fig. 9). The moderate increase in the ratio is probably due to the CO 2 added during the initiation procedure of every run.

5. Condusions

Closed-loop operation of a chemical heat pipe, based o n C O 2 reforming of methane, was performed previously in the laboratory. However, this is the first time that this cycle was closed under real solar conditions. The special conditions of working with solar energy create problems that do not occur under well controlled electric heating. There are still a number of factors that have to be improved, however, the results to date are quite satisfactory.

A c k n o w l e d g e m e n t

This work was supported by a grant from the Israel Ministry of Energy and Infrastructure.

R. Levitan et al. / Closed-loop operation of solar chemical heat pipe 477

References

[I] M. Levy, R. Levitan, H. Rosin, G. Adusei and R. 7~ubin, Proceedings of the 4th Int. Symposium on Solar Thermal Energy, Santa Fe, USA (June 1988).

[2] R. Levitan, H. Rosin and M. Levy, Solar Energy 42 (1989) 267. [3] R.B. Diver, J.D. Fish, R. Levitan, M. Levy, H. Rosin and J.T. Richardson, Solar Energy, submhted. [4] M. Levy, H. R~sin and K. Levitan, J. Solar Energy Engineering 111 (1989) 96. [5] A. Segal and M. Levy, Proceedings of the 5th Symposium on Solar High-Temperature Technologies,

Davos, Switzerland (August 1990) Sol. Energy Mater. 24 (1991) 725.