Evaluation of the Sodium-Water Reaction in Heat Transfer Systems

intermediate alkali-metal heat-transfer system between the sodium reactor ..... was probably not detected because of a fault in the leak-detection ala...
1 downloads 0 Views 2MB Size
Evaluation of the Sodium-Water Reaction in Heat Transfer Systems

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: January 1, 1957 | doi: 10.1021/ba-1957-0019.ch010

D. D. ADAMS, G. J . BARENBORG, and W. W. KENDALL Knolls Atomic Power Laboratory, General Electric Co. Schenectady, Ν . Y.

In considering sodium for use as a heat-transfer agent, the chemical reaction which results in the event of leakage between sodium and the sec­ ondary fluid is an important factor. The atomic energy industry has acquired considerable exper­ ience in the testing of sodium-heated steam gen­ erators. No instance of violent reaction resulting from leakage in these tests is known. A review of sodium-water reaction experiments forms a basis for estimating the effects of chemical reac­ tion resulting from potential leakage. System de­ sign recommendations to minimize the effects of leakage are also presented.

THE past ten years h a v e seen a g r o w i n g interest i n the use of s o d i u m a n d its alloys w i t h potassium ( N a K ) as heat-transfer fluids. A l t h o u g h the m a i n impetus i n the development of these metals has come f r o m the atomic energy p r o g r a m o n power reactors, the experience gained is also applicable to the use of these metals as h i g h - t e m p e r a t u r e heat-transfer fluids i n other industries. O n e of the problems i n the use of s o d i u m or N a K is the possibility of a v i g o r ­ ous c h e m i c a l reaction r e s u l t i n g f r o m leakage between the l i q u i d m e t a l a n d the fluid w i t h w h i c h it is e x c h a n g i n g heat. T h i s latter fluid is n o r m a l l y water. T h e resulting c h e m i c a l reaction is capable of p r o d u c i n g b o t h m e c h a n i c a l a n d corrosive damage to the system. T h e discussion i n this paper is confined to the m e c h a n i c a l damage aspects (pressure, temperature, a n d flow stoppage effects caused b y the c h e m i c a l reactions). C o r r o s i o n effects m a y nevertheless be important a n d must be considered f o r a n y specific system. F u r t h e r , o n l y s o d i u m - w a t e r reactions are discussed, a l t h o u g h the results are felt to be applicable to other a l k a l i - m e t a l r e ­ actions i n w h i c h the reaction heat a n d amount of gas f o r m e d are s i m i l a r . A n o t h e r p r o b l e m p e c u l i a r to the application of a s o d i u m - t o - w a t e r heat e x ­ changer system i n a n u c l e a r power plant is the reactor safeguards' considerations of a possible leak. I n certain reactors, the i n t r o d u c t i o n of h y d r o g e n o r its c o m ­ pounds c a n cause serious control difficulties. T h e r e f o r e , it is necessary to ensure that a direct leak of w a t e r - t o - r e a c t o r coolant is v i r t u a l l y impossible. T h e a b i l i t y of s o d i u m - c o o l e d reactors to produce steam at c u r r e n t c o m m e r ­ c i a l conditions of superheat a n d pressure has resulted i n considerable design effort to p r e v e n t this direct leakage. A t present, two methods are available to reduce the p r o b a b i l i t y of direct leakage to a safe l e v e l . T h e first of these employs a n intermediate a l k a l i - m e t a l heat-transfer system between the s o d i u m reactor c o o l ­ ant a n d the steam generating system. T h i s m e t h o d is generally applicable o n l y w h e n space limitations are not present. 92

In HANDLING AND USES OF THE ALKALI METALS; Advances in Chemistry; American Chemical Society: Washington, DC, 1957.

ADAMS, BARENBORG A N D KENDALL—SODIUM WATER REACTION IN HEAT TRANSFER

93

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: January 1, 1957 | doi: 10.1021/ba-1957-0019.ch010

T h e second m e t h o d provides a double b a r r i e r w i t h i n the steam generator itself. T h e double b a r r i e r has c o n v e n t i o n a l l y b e e n i n the f o r m of concentric tubes a n d double tube sheets w i t h the intermediate space filled w i t h a static t h i r d fluid to assist i n the transfer of heat. T h e t h i r d - f l u i d system was e q u i p p e d to detect any leaks w h i c h might occur i n either of the single b a r r i e r s . U n d e r the d o u b l e w a l l philosophy as o r i g i n a l l y developed, i f a leak o c c u r r e d i n either b a r r i e r , the heat exchanger was t a k e n out of service for r e p a i r or replacement. F u r t h e r discussion of n u c l e a r reactor c o n t r o l problems r e s u l t i n g f r o m water leakage is b e y o n d the scope of the present paper, a n d the discussion presented hereafter is o n the effects of the c h e m i c a l reaction between s o d i u m a n d water. A s a c o r o l l a r y of this discussion, the necessity for the use of a d o u b l e - b a r r i e r heat exchanger i n systems where n u c l e a r control problems are absent has b e e n e x a m ­ ined. T h e desirability of e l i m i n a t i n g the d o u b l e - b a r r i e r design w h e r e feasible is obvious. T h e d o u b l e - b a r r i e r results i n a m o r e c o m p l e x design w i t h associated fabrication a n d operational problems, requires a d d i t i o n a l heat transfer area due to the increased t h e r m a l resistance of the double b a r r i e r , a n d requires e x t e r n a l e q u i p m e n t to h a n d l e the t h i r d - f l u i d system. A l l these factors increase the size a n d cost of the heat exchanger. I n order to evaluate system b e h a v i o r i n the event of a w a t e r - t o - s o d i u m leak, studies have been conducted b o t h at K A P L ( K n o l l s A t o m i c P o w e r L a b o r a t o r y , operated for the U . S . A t o m i c E n e r g y C o m m i s s i o n b y the G e n e r a l E l e c t r i c C o . , Schenectady, N . Y . ) a n d M S A ( M i n e Safety A p p l i a n c e s C o . , C a l l e r y , P a . ) . A s u m ­ m a r y of data a c q u i r e d i n the K A P L studies a n d i n the earlier w o r k at M S A is i n ­ c l u d e d below. I n addition, three instances of leaks w h i c h actually o c c u r r e d i n operating s o d i u m - w a t e r heat exchangers are presented. T h e i n f o r m a t i o n obtained to date is sufficiently encouraging to justify serious consideration of the use of s o d i u m - w a t e r heat exchangers without d o u b l e - b a r r i e r protection f o r those systems i n w h i c h n u c l e a r c o n t r o l p r o b l e m s are absent. A t least one large project ( D e v e l o p ­ m e n t a l F a s t B r e e d e r P o w e r Reactor, A t o m i c P o w e r D e v e l o p m e n t Associates, Inc.) is p r e p a r i n g to operate a s o d i u m - h e a t e d steam generator of substantial size ( a p ­ p r o x i m a t e l y 1000 k w . ) w i t h a s i n g l e - b a r r i e r design (7). I n the a p p l i c a t i o n of this design to the n u c l e a r power plant, protection f r o m the n u c l e a r h a z a r d is afforded b y a n intermediate a l k a l i m e t a l loop a n d heat exchanger.

Experimental T h e pressure, temperature, a n d p l u g g i n g effects w h i c h result f r o m a s o d i u m water reaction are based o n the f o l l o w i n g c h e m i c a l equations: Na(l)

+ H 0(g) = NaOH(c)

Na(l)

+ N a O H ( c ) = N a * 0 ( c ) + J H ( g ) Δ H ° 291 =

2

+ i H ( g ) Δ H ° 298 = -45.7 k c a l . 2

2

Na(l)

+ è H*(g) = N a H ( c ) Δ Η

β

+1.89 k c a l .

298 = — 1 3 . 7 k c a l .

(1) (2) (3)

R e a c t i o n 1 is e x t r e m e l y r a p i d , generally b e i n g l i m i t e d o n l y b y the speed of m i x i n g . R a p i d m i x i n g a n d reaction are facilitated b y the h i g h s o d i u m flow rates e m p l o y e d i n heat-transfer systems. Less is k n o w n , however, about the rate a n d extent to w h i c h the second a n d t h i r d reactions proceed. U n d e r conditions e n ­ countered i n heat-transfer systems, their rates a r e generally considered to be slow i n c o m p a r i s o n to R e a c t i o n 1. E x a m i n a t i o n of the equations shows that R e a c t i o n 2 w o u l d double the system pressure rise w h i c h w o u l d result w e r e R e a c t i o n 1 to proceed alone. T h e f o r m a t i o n of s o d i u m h y d r i d e i n R e a c t i o n 3 w o u l d reduce the pressure b u i l d - u p . T h e t h e r m a l effects of a s o d i u m - w a t e r reaction are l a r g e l y influenced b y R e ­ action 1 alone. T h e heat p r o d u c e d b y R e a c t i o n 2 is negligible i n comparison, a n d the heat e v o l v e d b y a moderate rate of s o d i u m h y d r i d e f o r m a t i o n w o u l d l a r g e l y be b l a n k e t e d b y the heat exchanger load a n d b y heat losses f r o m the system. P l u g g i n g effects result f r o m the fact that a l l of the solid reaction products f o r m e d are p r a c t i c a l l y insoluble i n s o d i u m . I n considering the pressure effects of a w a t e r - t o - s o d i u m leak, the possibility of shock effects r e s u l t i n g f r o m the c h e m i c a l reaction must be e x a m i n e d . T o this end, experiments w e r e conducted b y N e l s o n (1 ) to measure the shock w a v e e n e r -

In HANDLING AND USES OF THE ALKALI METALS; Advances in Chemistry; American Chemical Society: Washington, DC, 1957.

94

A D V A N C E S IN CHEMISTRY SERIES

gies obtained f r o m the s o d i u m - w a t e r reaction. T h e conditions of the experiments were chosen to obtain m a x i m u m rate of reaction. L i q u i d N a K was substituted for s o d i u m , a n d nine N a K charges (1 to 5 k g . each) were detonated i n water w i t h a n explosive booster charge. Booster charges were detonated without N a K present to serve as a control.

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: January 1, 1957 | doi: 10.1021/ba-1957-0019.ch010

T h e energy c a r r i e d b y the pressure wave i n this experiment was calculated f r o m pressure measurements a n d f o u n d to be v e r y l o w . I n no case was it h i g h e r than 5 cal. p e r g r a m of N a K or m o r e than 0.33% of the available energy. S i m i l a r experiments w i t h trinitrotoluene showed that 25% of the available energy is c a r r i e d i n the shock wave. These tests indicate that it is impossible to m i x s o d i u m a n d water fast enough t h r o u g h a n y conceivable leak i n a heat-transfer system to obtain a reaction p r o d u c i n g the shock waves characteristic of h i g h explosives. T o evaluate the use of p r e s s u r e - r e l i e f devices, a n u m b e r of tests were c o n ducted b y K i n g (3-6). Water was forced into N a K t h r o u g h a à-inch pipe u n d e r a pressure differential of 200 pounds per square inch. T h e pressure generated was vented t h r o u g h a 1-inch relief v a l v e set at 200 pounds p e r square i n c h . These reaction studies are notable, because they demonstrated that the pressure resulting f r o m the reaction c a n be r e l i e v e d b y conventional relief valves. These studies also showed that sharp pressure pulses resulted w h e n the N a K was at r o o m temperature but were r e d u c e d as the N a K temperature was increased. (In tests w i t h the N a K i n i t i a l l y at r o o m temperature, reaction products plugged the relief valve, resulting i n damage to the reaction vessel w h i c h h a d passed a hydrostatic test at 3000 pounds per square inch. In the runs at higher temperature, the relief v a l v e was protected b y a baffle, a n d no m e c h a n i c a l damage occurred.) W i t h the N a K preheated to 600 °F., the reaction proceeded to completion without pressure surges, a n d the m a x i m u m pressure was essentially l i m i t e d to the relief v a l v e setting. It is not expected, therefore, that severe pressure pulses w o u l d r e sult f r o m leaks i n heat-transfer systems, because temperatures i n these systems n o r m a l l y r u n i n the 600 °F. range or higher. These tests f u r t h e r indicated that a higher initial gas pressure over the N a K r e d u c e d the pressure surges. W h e n the i n i t i a l N a K pressure was increased to 200 pounds p e r square i n c h or more, the pressure p u l s i n g effect was m i n i m i z e d . T h e pressure pulses first observed b y K i n g have appeared to some extent i n n e a r l y e v e r y s o d i u m - w a t e r reaction test. It is the authors' o p i n i o n that these pulses are obtained b y n o n u n i f o r m m i x i n g of the reactants. T h e gas f r o m the i n i t i a l reaction separates the reactants, a l l o w i n g a slug of unreacted water to enter the reaction zone. W h e n the gas b u b b l e f r o m the initial reaction disperses, a s u d den m i x i n g of the reactants produces a h i g h local pressure. T h i s pressure peak lasts u n t i l the i n e r t i a of the l i q u i d can be overcome a n d the local pressure e q u a l ized w i t h the pressure i n the r e m a i n d e r of the system. T h e pressure pulses are different f r o m shock waves i n that, although of longer d u r a t i o n ( i n the order of m i l l i s e c o n d s ) , they are r e l a t i v e l y l o w i n pressure. These pulses are generally a t tenuated as they t r a v e l throughout the l i q u i d p i p i n g system a n d do not appear i n the gas phase. T h e tests discussed thus far, w h i l e i m p o r t a n t i n defining the effects expected f r o m a n a l k a l i m e t a l - w a t e r reaction, have i n v o l v e d large quantities of water a n d c o m p a r a t i v e l y little a l k a l i metal. I n the event of water leakage into s o d i u m i n a heat-transfer system, the s o d i u m w o u l d be i n excess. T o investigate the effects of a s o d i u m - w a t e r leak i n heat transfer systems, a s m a l l test system w a s operated at K A P L . T h e system consisted of a water boiler using one d o u b l e - w a l l e d tube w i t h flowing s o d i u m o n the tube side, boiler water o n the shell side, a n d m e r c u r y as the t h i r d fluid. I n each test r u n a hole was fabricated i n the outside tube. A f t e r the b o i l e r reached e q u i l i b r i u m conditions, steam pressure of 100 pounds per square i n c h a n d 475° to 500 °F. s o d i u m temperature, the inner tube was completely p a r t e d b y a tensile l o a d a p p l i e d to a p r e f o r m e d p e r i p h e r a l notch. T o date two test runs have been completed. I n the first, the leak i n the outside tube was a 0.035-inch h o l e ; i n the second, the leak was r e d u c e d to a 0.009-inch hole. It is p r o b a b l e that i n neither r u n was the s m a l l hole t h e l i m i t i n g factor o n the rate of water injection, since water l e a k e d into the m e r c u r y annulus p r i o r to the i n n e r tube f a i l u r e . T h i s reserve of water t h e n was admitted to the s o d i u m

In HANDLING AND USES OF THE ALKALI METALS; Advances in Chemistry; American Chemical Society: Washington, DC, 1957.

ADAMS, BARENBORG A N D

KENDALL—SODIUM WATER REACTION IN

95

HEAT TRANSFER

t h r o u g h the larger f a i l u r e i n the inside tube. In b o t h bases the reaction was r e l a ­ t i v e l y moderate. T h e pressure traces obtained for the first a n d second test r u n s are s h o w n i n F i g u r e s 1 a n d 2, respectively.

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: January 1, 1957 | doi: 10.1021/ba-1957-0019.ch010

In the first test r u n the two pressure traces represent measurements at inlet a n d outlet of the boiler. B o t h traces show the characteristic pressure pulses o b ­ tained i n the w o r k of K i n g . T h e first, second, a n d f o u r t h pulses detected at the s o d i u m inlet appear the strongest, thereby i n d i c a t i n g the reaction p r o d u c i n g these pulses was nearest to this pressure detector. T h e t h i r d pulse is due to a reaction taking place nearer the s o d i u m outlet. These pressure pulses were p r o d u c e d b y n o n u n i f o r m reaction effects s i m i l a r to those p r e v i o u s l y discussed. T h e pressures obtained at the inlet a n d outlet are different, i n d i c a t i n g that appreciable p l u g g i n g of the tube took place. W h e n the unit was disassembled, a m i x t u r e of sodium, m e r c u r y , a n d s o d i u m - w a t e r reaction products was f o u n d i n the tube. In the second, test pressure measurements were also obtained f r o m the system expansion tank. T h i s pressure rose smoothly to a l e v e l corresponding to the p r e s ­ sure relief setting. A g a i n the pressure at the s o d i u m inlet a n d outlet b e h a v e d as though they were independent. T h e pressure at the outlet was moderate a n d is not s h o w n i n F i g u r e 2. T h e negative pressure r e c o r d e d at the inlet is a n absolute

-J0

TIME

IN

i 2

1 4 6 θ SECONDS AFTER LEAK 1

1

1 10

1 12

I 14

I 16

I 18

I 20

I 22

1 24

_J 26

1 28

Figure 1. Pressure traces for a 30-mil leak

pressure of 7 pounds per square i n c h , i n d i c a t i n g that t e m p o r a r y p l u g g i n g of the p i p i n g a n d absorption of the h y d r o g e n b y s o d i u m h y d r i d e f o r m a t i o n m a y have t a k e n place. W h e n the u n i t was disassembled, the tube was f o u n d to be c o m ­ pletely empty. T h i s was caused b y the fact that the leak was a l l o w e d to continue long enough to w a s h out the s o d i u m a n d the reaction products completely. I n b o t h test runs the s o d i u m flow completely stopped i m m e d i a t e l y after the f a i l u r e i n the inside tube. T h e results of the above tests are indicative of a f a i r l y large leak between s o d i u m a n d water. A n actual leak i n a system u s u a l l y takes some time to develop, b e g i n n i n g as a s m a l l crack a n d g r o w i n g larger. In such a system the r a p i d h y ­ d r a u l i c transients, p r o d u c i n g the r a p i d reaction, a n d pressure pulse w o u l d not be expected. It is notable that even i n the test system used here the pressure b u i l d - u p at the expansion tank was u n i f o r m . T o investigate f u r t h e r the effects of water leakage into excess sodium, tests were made at K A P L u s i n g a 1-inch p u m p e d s o d i u m loop. K n o w n amounts of water (6 to 12 grams) were injected into the loop w h i c h contained 20 pounds of s o d i u m . F i g u r e 3 shows a t y p i c a l pressure a n d flow trace obtained after saturated water h a d been injected into flowing s o d i u m at 500°F.(2). T h e gas v o l u m e i n the e x p a n ­ sion t a n k was sized so that the pressure increase w o u l d be 100 pounds per square i n c h if a l l the water were to r e m a i n as steam, or 50 pounds per square i n c h i f a l l the water reacted a c c o r d i n g to R e a c t i o n 1. A s the w a t e r w a s i n j e c t e d s u d d e n l y as saturated water at about 500 °F., the i n i t i a l pressure rise was sharp. Pressure oscillations were noted b y the pressure-sensing element located a p p r o x i ­ m a t e l y 6 feet f r o m the point of injection. T h i s oscillation was p r o b a b l y due to

In HANDLING AND USES OF THE ALKALI METALS; Advances in Chemistry; American Chemical Society: Washington, DC, 1957.

A D V A N C E S IN CHEMISTRY SERIES

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: January 1, 1957 | doi: 10.1021/ba-1957-0019.ch010

96

expansion a n d contraction of the gas b u b b l e f o r m e d , before the b u b b l e pressure equalized w i t h the expansion tank pressure. T h e pressure at the expansion tank d i d not show a n y oscillation. T h e system pressure increased b y 20 pounds p e r square i n c h i m m e d i a t e l y a n d thereafter d r o p p e d b y about 8 pounds per square i n c h i n 20 seconds. If no s o d i u m h y d r i d e were f o r m e d , the m i n i m u m pressure increase at the expansion tank should have been 50 pounds per square i n c h . A s the m a x i m u m increase was o n l y 20 pounds per square i n c h , it appears that a considerable fraction of the h y d r o g e n e v o l v e d b y R e a c t i o n 1 must have reacted f a i r l y r a p i d l y b y R e a c t i o n 3. T h e pressure d e ­ crease i n the e x p a n s i o n tank is f u r t h e r i n d i c a t i o n of h y d r o g e n absorption b y s o ­ d i u m formation. T h e trace of the s o d i u m flow shows a response p e c u l i a r to this system. T h e s u d d e n flow d r o p after injection was caused b y v a p o r i z a t i o n of the water w h i c h pushed s o d i u m t h r o u g h the flowmeter i n the reverse d i r e c t i o n to the expansion tank. A f t e r this transient, the s o d i u m flow r e t u r n e d to n o r m a l , u n t i l the reaction products r e a c h e d a v e r t i c a l l e g . T h e gaseous reaction products t h e n r e d u c e d the h e a d o n the p u m p discharge, thereby increasing the flow. T h e final decrease i n flow w a s p r o b a b l y caused b y the reaction products' p a r t i a l l y p l u g g i n g the system a n d thereby increasing the system pressure d r o p .

-J SODIUM

SURGE TANK PRESSURE

SODIUM

PRESSURE

I

1-

I S £ί 5 1

$

25 Ο 125

NEAR INJECTION

POINT

Fi Β

25 0

2 TIME

IN

4

6

10

12

14

20

22

24

SECONDS

Figure 3. Pressure a n d sodium flow traces after water injection

In HANDLING AND USES OF THE ALKALI METALS; Advances in Chemistry; American Chemical Society: Washington, DC, 1957.

26

ADAMS, BARENBORG A N D

KENDALL—SODIUM WATER REACTION IN

HEAT TRANSFER

97

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: January 1, 1957 | doi: 10.1021/ba-1957-0019.ch010

F u r t h e r runs were m a d e o n the 1-inch system i n order to determine the p l u g p i n g effects of the reaction products on restricted flow passages, i n this case s i m u lating the close h y d r a u l i c clearances of a n u c l e a r reactor. F o r this purpose, a n orifice plate h a v i n g f o r t y - f o u r 5 0 - m i l holes was inserted i n the loop, and r u n s were made i n w h i c h the s o d i u m temperature was v a r i e d f r o m 400° to 625°F. A t the lowest temperature, complete p l u g g i n g of the orifice plate o c c u r r e d , w h i l e at the highest temperature there was no immediate p l u g g i n g effect. A t intermediate temperatures, p a r t i a l p l u g g i n g was noticed (2). T h e lack of p l u g g i n g at 625°F. is attributed to s o d i u m h y d r o x i d e h a v i n g a m e l t i n g point of 605 °F. T h e results of this test show that the presence of excess s o d i u m significantly reduces the pressure rise obtained because of the r a p i d f o r m a t i o n of N a H ( R e a c tion 3). A l t h o u g h the test was p e r f o r m e d on a b a t c h - i n j e c t i o n basis, the data indicate that w i t h a continuous leak o n l y m i n i m u m - p r e s s u r e oscillations w o u l d occur, a n d the o v e r - a l l pressure w o u l d be less t h a n predicted f r o m Reaction 1. T o explore f u r t h e r the system effects w h i c h result f r o m the continous l e a k age of water into a c i r c u l a t i n g s o d i u m stream, K A P L is c u r r e n t l y m o d i f y i n g a n 8i n c h pipe diameter system for l a r g e - s c a l e water injection studies. T h e system to be used for the large-scale tests w i l l contain a p p r o x i m a t e l y 4000 pounds of s o d i u m a n d have a surge v o l u m e of 25 cubic feet. W a t e r - l e a k rates i n a range f r o m 0.1 to 100 pounds per m i n u t e are to be tested, a n d pressure, temperature, a n d p l u g g i n g p h e n o m e n a w i l l be studied. T h e results of these tests, complemented b y the tests discussed above a n d b y experience gained f r o m operating test-heat exchangers, w i l l help f o r m a sound e x p e r i m e n t a l foundation for any future decision to e l i m inate the d o u b l e - b a r r i e r philosophy.

Experience from Operating Systems D u r i n g the p e r i o d K A P L has been testing steam generators, instances of l e a k age between s o d i u m a n d water have been r a r e . T h e r e have been, however, three h e a t - e x c h a n g e r failures that resulted i n the m i x i n g of water a n d s o d i u m . T h e data f r o m these failures give a n i n d i c a t i o n of what m a y be expected f r o m a s o d i u m water leak. O f the failures w h i c h definitely resulted i n s o d i u m - w a t e r reactions, two of these failures o c c u r r e d i n one steam generator b e i n g used i n n a t u r a l c i r culation heat-transfer studies at K A P L . T h i s unit, s h o w n i n F i g u r e 4, was a s h e l l and tube evaporator w i t h s o d i u m on the tube side a n d water on the s h e l l side. Because it was an e x p e r i m e n t a l unit, o n l y a single b a r r i e r was used i n separating the s o d i u m a n d the water.

Figure 4.

Natural circulation evaporator

In HANDLING AND USES OF THE ALKALI METALS; Advances in Chemistry; American Chemical Society: Washington, DC, 1957.

A D V A N C E S IN

98

CHEMISTRY SERIES

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: January 1, 1957 | doi: 10.1021/ba-1957-0019.ch010

T h e i n i t i a l f a i l u r e i n this unit was first indicated b y a drop i n the s o d i u m flow to zero d u r i n g a n a t u r a l c i r c u l a t i o n heat transfer r u n . T h e u n i t at this time was operating w i t h a s o d i u m - i n l e t temperature of about 700°F., a s o d i u m pressure of about 5 pounds per square i n c h gage, and a b o i l i n g water pressure of 485 pounds. F l o w was restored i n about 0.5 h o u r b y a p p l y i n g a d d i t i o n a l heat to the s o d i u m a n d b y activating the loop electromagnetic p u m p . O p e r a t i o n of the loop was then continued for about 9 hours i n order to p r o v e conclusively the existence of a s o d i u m - w a t e r leak. T h e second f a i l u r e i n this u n i t o c c u r r e d sometime after the unit h a d been put b a c k into operation. A g a i n the first i n d i c a t i o n of a leak was a l o w s o d i u m flow rate d u r i n g a n a t u r a l c i r c u l a t i o n heat-transfer r u n . F r o m experience gained w i t h the first leak, the flow trend was noticed a n d p u m p power applied early enough to prevent complete loss of flow. T h e gas i n the s o d i u m surge tank was sampled a n d f o u n d to contain h y d r o g e n , thus p r o v i d i n g further evidence that a leak existed. A f t e r the discovery of the leak, the system was continued i n operation for about 12 days. D u r i n g this time three more n a t u r a l c i r c u l a t i o n heat-transfer runs were made. L e a k a g e a p p a r e n t l y continued d u r i n g these runs, as the i m p u r i t y content of the s o d i u m increased d u r i n g the 12-day p e r i o d . [ I m p u r i t y content was measured b y use of a " p l u g g e d i n d i c a t o r " (1).] W h i l e the t h i r d r u n was u n d e r w a y on the twelfth day, a sudden pressure rise was experienced i n the s o d i u m surge tank f r o m 5 to 30 pounds per square i n c h gage i n less t h a n 2 minutes. A t this time the r u n was t e r m i n a t e d a n d no f u r t h e r heat-transfer runs were made on the loop. D a t a available on the first f a i l u r e d i d not p e r m i t an estimate of the water l e a k rate. F o r the second leak, however, sufficient data were available to p e r m i t an estimate of the leakage rate d u r i n g the final heat-transfer r u n . Estimates of this leakage rate were made f r o m data both on the pressure rise and o n the increase i n s o d i u m i m p u r i t y content. These estimates indicated an average leak rate of a p p r o x i m a t e l y 0.01 p o u n d per m i n u t e . M e t a l l u r g i c a l e x a m i n a t i o n of the failure showed that the leak resulted f r o m a crack i n a t u b e - t o - t u b e sheet w e l d . A section t h r o u g h this crack, w h i c h e x t e n d ed a r o u n d about two thirds of the tube p e r i p h e r y , is s h o w n i n F i g u r e 5. A c a l c u lation was made of the water leak rate t h r o u g h the failure. T h i s calculated result was a p p r o x i m a t e l y 10,000 times h i g h e r than the 0.01 p o u n d per m i n u t e noted above, a leak rate too large for the effects observed. A l t h o u g h appreciable error is expected i n such a calculation, other factors aside f r o m calculation limitations

Figure 5.

Failure in natural circulation evaporator

In HANDLING AND USES OF THE ALKALI METALS; Advances in Chemistry; American Chemical Society: Washington, DC, 1957.

ADAMS, BARENBORG A N D

KENDALL—SODIUM WATER REACTION IN HEAT

TRANSFER

99

caused this d i s c r e p a n c y : p a r t i a l p l u g g i n g of the c r a c k b y the reaction products of s o d i u m a n d water, a n d the presence of steam w i t h the water. T h i s steam c o u l d have come either f r o m flashing of the saturated water or f r o m the steam n o r m a l l y present i n a b o i l i n g evaporator. T h e presence of these two factors w o u l d a p p r e c i ­ ably reduce the water leak rate for a crack of g i v e n size. T h e second unit i n w h i c h a f a i l u r e resulted i n the m i x i n g of water a n d s o d i u m o c c u r r e d i n a superheater unit w h i c h was part of a steam generator w i t h a p e r h o u r capacity of about 10 χ 10 B . t . u . b e i n g tested at the M S A (8). T h i s s u p e r ­ heater, s h o w n i n F i g u r e 6, was a tube a n d shell heat exchanger w i t h s o d i u m o n the s h e l l side a n d steam o n the tube side. T h e unit was of d o u b l e - b a r r i e r design w i t h m e r c u r y as the t h i r d fluid. T h e t h i r d fluid was n o r m a l l y m a i n t a i n e d at a pressure intermediate between the s o d i u m a n d steam system pressures.

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: January 1, 1957 | doi: 10.1021/ba-1957-0019.ch010

e

T h e first evidence of the s o d i u m - t o - w a t e r leak was a drop i n s o d i u m flow f r o m 40 to 10% i n 30 seconds w i t h a subsequent decrease to zero i n the next 15 minutes. T h e plugs w h i c h caused the drop i n s o d i u m flow were p r i n c i p a l l y i n the s o d i u m tubes of the evaporator unit of the generator w h i c h was d o w n s t r e a m of, and i n series w i t h , the superheater. A b o u t 75% of the tubes i n this u n i t were cleared without cutting open the evaporator. T h e r e m a i n i n g 2 5 % , however, h a d to be opened b y treating each tube i n d i v i d u a l l y w i t h a steam lance. E x a m i n a t i o n p r i o r to cleaning of the plugs i n the latter tubes showed the plugs to be p r i m a r i l y composed of s o d i u m h y d r o x i d e a n d / o r s o d i u m oxide. A s the steam generator was b e i n g operated i s o t h e r m a l l y at 700 °F. without water w h e n the flow stoppage occurred, the s t e a m - t o - m e r c u r y b a r r i e r must h a v e failed i n i t i a l l y without detection, d u r i n g previous steaming operations. T h i s leak w o u l d then have a l l o w e d steam to condense i n the m e r c u r y system a n d f o r m a n undetected water l a y e r i n the m e r c u r y - s y s t e m expansion tank. T h e i n i t i a l f a i l u r e was p r o b a b l y not detected because of a fault i n the l e a k - d e t e c t i o n a l a r m system. T h e subsequent leak i n the s o d i u m - m e r c u r y b a r r i e r then a l l o w e d the water c o l ­ lected i n the m e r c u r y system to enter the s o d i u m . T h e only m e t h o d available to estimate the rate of leakage for this f a i l u r e is based on the size of the m e r c u r y - t o - s o d i u m crack s h o w n i n F i g u r e 7. A s i n the

STEAM

Figure 6. Superheater unit tested at MSA

In HANDLING AND USES OF THE ALKALI METALS; Advances in Chemistry; American Chemical Society: Washington, DC, 1957.

A D V A N C E S IN

100

CHEMISTRY SERIES

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: January 1, 1957 | doi: 10.1021/ba-1957-0019.ch010

calculation for the n a t u r a l c i r c u l a t i o n evaporator, it was assumed that the crack acted as an orifice a n d that the fluid flowing t h r o u g h the crack was l i q u i d water. T h e m a x i m u m leak rate was calculated i n this m a n n e r to be about 200 pounds per m i n u t e . T h e average leak rate w o u l d be expected to be less t h a n this v a l u e . In the absence of definite t h i r d - f l u i d system pressure data, it was assumed i n the calculation that the t h i r d - f l u i d system was at its n o r m a l pressure of 300 pounds per square i n c h gage. T h e authors estimated that the total amount of water w h i c h entered the s o d i u m system was about 6 pounds. A g a i n the value of m a x i m u m leakage rate is p r o b a b l y h i g h , a l t h o u g h this calculation is felt to be more accurate t h a n the orifice flow c a l c u l a t i o n for the n a t u r a l c i r c u l a t i o n evaporator, since g e o m etry of the crack was simpler, thus p e r m i t t i n g a more accurate estimate of the flow area. T o s u m m a r i z e the experience gained, it was f o u n d in the case of the n a t u r a l c i r c u l a t i o n evaporator that a s m a l l leak d i d not cause the system to p l u g w h e n

Figure 7.

Failure in superheater unit

In HANDLING AND USES OF THE ALKALI METALS; Advances in Chemistry; American Chemical Society: Washington, DC, 1957.

101

ADAMS, BARENBORG A N D KENDALL—SODIUM WATER REACTION IN HEAT TRANSFER

p u m p i n g p o w e r was available a n d that c o n t i n u e d operation was possible i n the presence of the leak. A l s o , the leak rate was a p p a r e n t l y r e d u c e d over that possible for l i q u i d water either b y the presence of steam w i t h the w a t e r o r b y p a r t i a l p l u g ­ ging of the c r a c k . F o r a leak of moderate size, as i n the superheater u n i t , loss of s o d i u m flow d i d occur a n d c o n t i n u e d operation i n the presence of the leak was not possible. H o w e v e r , i n none of the leaks observed i n operating e q u i p m e n t to date has there b e e n evidence of a c h e m i c a l reaction of sufficient violence to p r e ­ sent a danger to operating p e r s o n n e l or adjacent equipment. I n the case of a l l three leaks, the s o d i u m - w a t e r reaction of itself was not vigorous enough to i n ­ dicate the leak, b u t r a t h e r i n each case the l e a k was detected i n d i r e c t l y b y the p l u g g i n g effects of the insoluble reaction products.

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: January 1, 1957 | doi: 10.1021/ba-1957-0019.ch010

Proposed Design Concepts O n the basis of the reaction studies completed a n d the operating experience a c ­ q u i r e d to date, it is n o w possible to develop a n e w philosophy i n the design a n d operation of s o d i u m - w a t e r heat-transfer systems. R e a c t i o n studies have i n d i c a t e d that pressures i n a c i r c u l a t i n g system for a p a r t i c u l a r leak rate do not n o r m a l l y exceed those p r e d i c t e d b y R e a c t i o n 1. W h e r e pressure peaks have o c c u r r e d i n c i r c u l a t i n g systems, they have been of v e r y short d u r a t i o n a n d not of serious c o n ­ sequence. N e i t h e r pressure rises n o r temperature rises have caused m e c h a n i c a l damage i n those cases w h e r e leaks have o c c u r r e d i n operating heat exchangers. A l t h o u g h substantiation i n a large scale demonstration is still r e q u i r e d , it w o u l d n o w appear that the designer c a n , w i t h reasonable assurance, e m p l o y elementary tools i n designing a s o d i u m - w a t e r heat-transfer system to p r e v e n t serious m e c h a n ­ i c a l damage s h o u l d leakage occur. A s noted above, the amount of gas e v o l v e d f r o m a t y p i c a l s o d i u m - w a t e r leak w i l l n o r m a l l y be somewhat less t h a n that i n d i c a t e d b y R e a c t i o n 1. B y c o m b i n i n g the stoichiometry of this reaction w i t h the i d e a l gas l a w , a conservative v a l u e of the r e s u l t i n g pressure rise c a n be a p p r o x i m a t e d b y Δ Ρ =

kl

(

) RT

2 χ 18 χ 60 where

(4)

~y

Δ Ρ = system pressure rise, pounds p e r square i n c h L = average leak rate, pounds p e r m i n u t e (V

2

t = time,

mole H / m o l e H 0 ) 2

8

seconds

R = gas constant, 10.73 ( l b . / s q . inch)

(cu. f t . ) / l b . m o l e ° R .

Τ = temperature, ° R . V = surge v o l u m e , c u . feet A s e m p l o y e d i n these derivations, L represents a n average constant leak rate. I n actual practice, the leak rate w i l l v a r y w i t h the differential pressure across the leak. N o r m a l l y the use of a constant v a l u e of L s h o u l d p r o v e adequate. I n a p a r t i c ­ u l a r application, however, the designer m a y w i s h to consider the decrease i n L w i t h rise i n s o d i u m pressure. E q u a t i o n 4 assumes a constant gas v o l u m e i n the s o d i u m system a n d the instantaneous transmission of pressure throughout the system. A s s u m i n g a n average system temperature of 700°F., E q u a t i o n 4 reduces to: L A Ρ — 5.76 — t (5) V E q u a t i o n 5 is plotted for several values of L/V i n F i g u r e 8. I n system design, a plot of this type c a n serve to determine the surge v o l u m e r e q u i r e d to protect the s o d i u m system f r o m exceeding design pressures. T o use the plot, first select the m a x i m u m allowable system pressure a n d the system pressure w h i c h w o u l d

In HANDLING AND USES OF THE ALKALI METALS; Advances in Chemistry; American Chemical Society: Washington, DC, 1957.

A D V A N C E S IN CHEMISTRY SERIES

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: January 1, 1957 | doi: 10.1021/ba-1957-0019.ch010

102

0

10

20

30

40

50

60

70

80

90

WO HO

120 130 140 150

TIME * SEC.

Figure 8. Rate of system pressure rising resulting from leakage of water into sodium

c l e a r l y indicate a n i n - l e a k a g e of water. T h e latter pressure, c a l l e d the a l a r m p r e s ­ sure, s h o u l d be safely above a n y transient pressures w h i c h m i g h t occur i n n o r m a l system operation. T h e difference between these two pressures determines the total time available for relief of the l e a k - i n d u c e d pressure b u i l d - u p . W h i l e relief c a n be accomplished i n a n u m b e r of ways, perhaps the most s t r a i g h t f o r w a r d is the r e l i e f of pressure f r o m the water side ( d o w n to the s o d i u m side p r e s s u r e ) , essentially stopping f u r t h e r leakage. A n alternative method w i t h some advantage involves direct r e l i e f of the s o d i u m system. U s i n g the difference i n the two pressures d e t e r m i n e d above, as w e l l as the total time r e q u i r e d for the control system to f u n c t i o n a n d for the water side p r e s ­ sures to be r e l i e v e d , the designer c a n refer to a plot of the type of F i g u r e 8 a n d determine the m a x i m u m v a l u e of L/V to be designed into his system. T h e v a l u e of L selected w i l l d e p e n d p r i n c i p a l l y o n the design of the heat exchanger, a n t i c i ­ pated operating conditions, the f r e q u e n c y of inspection, a n d , if available, e x p e r i ­ ence w i t h s i m i l a r equipment. O n c e a conservative v a l u e of L is chosen, the n e c ­ essary surge v o l u m e , V , is selected to give the r e q u i r e d L/V ratio. T h e above process c a n be reversed, first selecting Δ Ρ a n d L/V a n d then designing the control system a n d relief system to f u n c t i o n w i t h i n the r e q u i r e d time. T h e significance to be attached to use of the above type of analysis is twofold. F i r s t , suffiicient data have n o w been collected to indicate that explosive reactions w i t h energetic shock waves do not occur i n t y p i c a l heat exchanger failures. S e c ­ o n d l y , experience a c q u i r e d i n recent years n o w provides a basis for estimating the rate of leakage w h i c h m i g h t be expected i n a p a r t i c u l a r design of s o d i u m - w a ­ ter heat exchanger. It is expected this basis w i l l be extended b y future e x p e r i ­ ence. T h e temperature rise r e s u l t i n g f r o m water leakage into a c i r c u l a t i n g s o d i u m system m a y be considered b y c o m p a r i n g the temperature effect w i t h the pressure effect. A s s u m i n g that a l l heat generated b y leakage f r o m water into s o d i u m is that due to R e a c t i o n 1, the rise i n temperature of the s o d i u m system is a p p r o x i m a t e d b y A

T

=

QL/60t Wc

where

Δ Τ = Q = L = t = W = c =

(6)

s o d i u m temperature rise, ° F . heat of reaction, 4,550 B . t . u . / l b . P L O leak rate, pounds p e r m i n u t e time, seconds weight of s o d i u m charge, pounds specific heat of s o d i u m , 0.3 B . t . u . / l b . - ° F .

E q u a t i o n 6 reduces to: (7)

In HANDLING AND USES OF THE ALKALI METALS; Advances in Chemistry; American Chemical Society: Washington, DC, 1957.

ADAMS, BARENBORG A N D KENDALL—SODIUM WATER REACTION IN HEAT TRANSFER

103

T h i s a p p r o x i m a t i o n is based o n the assumption that s o d i u m c i r c u l a t i o n rates are sufficiently h i g h to distribute generated heat u n i f o r m l y throughout the s y s ­ tem. E x c e p t i n the case of n o n c i r c u l a t i n g systems, this assumption is m o r e t h a n counterbalanced b y the fact that system heat losses a n d the heat capacity of sys­ tem p i p i n g a n d components have not been considered. W h i l e s h o r t - l i v e d , localized temperature excursions i n excess of that calculated m a y occur a n d m a y be d e t r i ­ m e n t a l to the extended operation of a system, immediate failure w o u l d not be expected. C o m b i n i n g E q u a t i o n s 5 a n d 7, w e find that V = 43.1— Ρ W

ΔΤ

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: January 1, 1957 | doi: 10.1021/ba-1957-0019.ch010

A

(8)

A n average system operating temperature of 700 °F. was assumed i n f o r m u l a t i n g E q u a t i o n 5. R i g o r o u s l y , E q u a t i o n 8 takes the f o r m , Δ Τ / Δ Ρ = 49,900/Tav. W W , w h e r e T . = operating system plus one h a l f the temperature rise d u e to leakage. A s V/W w i l l generally range f r o m a v a l u e of 0.002 to 0.01 cubic foot p e r p o u n d (10 to 5 0 % surge v o l u m e ) , the ratio of Δ Τ / Δ Ρ w i l l range between 0.0862 a n d 0.431. A relationship such as E q u a t i o n 8 c a n be used to estimate the temperature effects w h i c h w i l l result w i t h a corresponding pressure incident. I n most cases a system adequately protected f r o m overpressure w i l l also be protected f r o m o v e r t e m p e r a ture. a v

W h i l e temperature a n d pressure effects have generally been considered to be the most severe results of a w a t e r - t o - s o d i u m leak, recent experience indicates that the most substantial problems w h i c h result are i n v o l v e d w i t h system p l u g g i n g . Because of the difficulty i n differentiating between the various reaction products, it is not k n o w n whether p l u g g i n g is caused b y one of the reaction products or b y some combination of them. C o n s i d e r a b l e i n f o r m a t i o n is available on the p l u g g i n g characteristics of s o d i u m oxide i n sodium. Less is k n o w n , however, about the p l u g g i n g characteristics of s o d i u m h y d r o x i d e a n d s o d i u m h y d r i d e . F o r this reason a n d also because the solubilities of these three reaction products i n s o d i u m are s i m i l a r f r o m 400° to 800 °F., p l u g g i n g effects are discussed i n terms of s o d i u m oxide concentration. T h e concentration of o x y g e n - b e a r i n g components i n s o d i u m is n o r m a l l y e x ­ pressed as weight per cent o x y g e n . T h e concentration of o x y g e n i n s o d i u m r e s u l t ­ i n g f r o m water leakage c a n be expressed b y

C = where

C L t W

= = = =

-

8

J ° - 100% w

(9)

weight % o x y g e n leak rate, pounds p e r m i n u t e time, seconds weight of s o d i u m charge, pounds

Setting C = 0.016%, the s o l u b i l i t y of s o d i u m oxide at 600°F., w e get W L = 0.0108 — t

(10)

E q u a t i o n 10 relates water leak rate to the time to reach s o d i u m oxide saturation at 600°F. T h i s relationship is plotted i n F i g u r e 9 for two values of W . It is signifi­ cant to note that a n y large leak w i l l cause p l u g g i n g conditions i n just a matter of seconds, e v e n i n a large system. C o m b i n i n g E q u a t i o n s 4 a n d 10, a n d assuming a n average system temperature of 700 °F., we find W A P = 0.0622 — (11) V Inspection of this relationship, w h i c h holds for a n y size leak, indicates that a system w i t h 10% surge v o l u m e w o u l d become saturated w i t h o x y g e n w h e n the pressure rise r e s u l t i n g f r o m a water leak reached a little over 30 pounds p e r square

In HANDLING AND USES OF THE ALKALI METALS; Advances in Chemistry; American Chemical Society: Washington, DC, 1957.

A D V A N C E S IN CHEMISTRY SERIES

104 'ΟΟ-ϋ

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: January 1, 1957 | doi: 10.1021/ba-1957-0019.ch010

90-fl

; 0

ι 20

1 40

1 60

1 80

1 100

1 120

1 140

TIME - SEC.

Figure 9. Time required to reach oxygen saturation vs. leakage of water into sodium

inch. A c o r r e s p o n d i n g l y smaller pressure rise w o u l d occur i n systems w i t h l a r g e r surge volumes. T h e equation is suited to use f o r systems w i t h a n average t e m p e r a ­ ture of 700°F. a n d a c o l d l e g temperature of 600°F. S i m i l a r expressions c a n be developed f o r a n y other temperature conditions. T h e r a p i d rise i n o x y g e n concentration w h i c h results f r o m water leakage i m ­ mediately suggests the use of a p l u g g i n g indicator as a means of leak detection i n h i g h - t e m p e r a t u r e systems. A c t u a l l y , most of the leaks encountered i n the past were first signaled b y flow restrictions. I n designing specifically f o r leak detection, however, a special p l u g g i n g indicator should be installed i n a s m a l l , continuously operating bypass d o w n s t r e a m f r o m the heat exchanger. T h e p l u g g i n g disk s h o u l d be m a i n t a i n e d at system temperature where it w i l l not be affected b y n o r m a l c o n ­ centrations of s o d i u m oxide, p r o v i d i n g the latter is generally c o n t r o l l e d w e l l b e l o w saturation. B a s e d o n present knowledge, a n operating temperature for this type indicator above the m e l t i n g point of s o d i u m h y d r o x i d e c o u l d not be r e c o m m e n d e d . D a t a presented above f o r the 1-inch p u m p e d loop test p r o v e the effectiveness of an indicator of this type at temperatures of 500 °F. a n d below.

Conclusions T h e reaction between s o d i u m a n d water i n p r a c t i c a l heat-transfer systems w i l l not generally result i n m e c h a n i c a l damage to the system. T h e l a c k of damage is p r o b a b l y due i n large part to two things: the absence of shock waves f r o m the s o d i u m - w a t e r reaction p e r se, a n d the relative smoothness of the reaction at n o r ­ m a l h e a t - e x c h a n g e r temperatures. T h e m a x i m u m pressure e x p e r i e n c e d f o r a n y significant d u r a t i o n of time w o u l d n o r m a l l y be no greater t h a n the water p r e s ­ sure a n d the temperature effects w o u l d be r e l a t i v e l y m i n o r . S y s t e m p l u g g i n g a n d subsequent problems i n the r e m o v a l of the reaction products m a y occur, however, if the leak rate is sufficiently large. These conclusions are supported b y the r e ­ sults of test w o r k at K A P L a n d other laboratories o n the injection of water into s o d i u m a n d b y experience gained f r o m failures i n operating heat transfer systems. C r i t e r i a presented i n E q u a t i o n s 4 a n d 8 c a n be used to design the system to w i t h s t a n d the pressure a n d temperature effects of the s o d i u m - w a t e r reaction. T h e use of these criteria requires a reasonable estimate of the m a x i m u m leak rate expected. T h e experience gained w i t h the s o d i u m - w a t e r reaction has also indicated the desirability of some design considerations i n order to m i n i m i z e the effects of l e a k ­ age i n a s o d i u m h e a t - t r a n s f e r system. T h e considerations a r e : P r o v i d e the system w i t h adequate v e n t i n g capacity. P r e f e r a b l y this v e n t ­ i n g capacity s h o u l d be p r o v i d e d o n the h i g h - p r e s s u r e side ( n o r m a l l y the water side) to p r o v i d e for r a p i d equalization of the pressure across the leak, thus essen­ t i a l l y stopping the leakage.

In HANDLING AND USES OF THE ALKALI METALS; Advances in Chemistry; American Chemical Society: Washington, DC, 1957.

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: January 1, 1957 | doi: 10.1021/ba-1957-0019.ch010

ADAMS, BARENBORG A N D KENDALL—SODIUM

WATER REACTION

IN HEAT TRANSFER

105

W h e r e feasible, design the low-pressure s i d e (usually t h e s o d i u m s i d e ) of the heat exchange system for n o n r u p t u r e at the m a x i m u m pressure of the h i g h pressure side. N o r m a l l y this n o n r u p t u r e design should not be based o n the same factor of safety used i n designing for n o r m a l operating conditions. O f t e n this n o n ­ r u p t u r e condition w i l l be automatically achieved t h r o u g h code design of the s y s ­ t e m for n o r m a l operating conditions or b y the use of standard components w h i c h are overdesigned for n o r m a l operating conditions. T h e i n f o r m a t i o n discussed above warrants serious consideration of the use of s i n g l e - w a l l heat exchangers i n s o d i u m heat-transfer systems. C o n s i d e r a t i o n of the use of s i n g l e - w a l l exchangers should include a t h o r o u g h e x a m i n a t i o n of a l l possible consequences to the system a n d its components i n the event of a h e a t exchanger leak. It m a y be desirable to use several s m a l l heat exchangers r a t h e r t h a n a single unit, so that a leak i n one u n i t w o u l d not r e q u i r e complete s h u t d o w n of the entire heat-transfer system. If it is d e t e r m i n e d that s i n g l e - w a l l heat e x ­ changers are permissible, their use w i l l result i n a r e l a t i v e l y simple, s m a l l , a n d l o w - c o s t unit. T h i s simplified design m a y result i n a h i g h l y r e l i a b l e piece of e q u i p ­ ment.

Literature Cited (1) (2) (3) (4) (5) (6) (7) (8)

J a c k s o n , C . B . , "Liquid M e t a l H a n d b o o k , S o d i u m - N a K S u p p l e m e n t , " 3rd ed., U. S. G o v ­ ernment P r i n t i n g Office, Washington 25, D . C . , 1955. K a r n e s , H. F., private c o m m u n i c a t i o n . K i n g , E. C., " R e a c t i o n of N a K a n d H2O," M i n e Safety A p p l i a n c e s C o . , TR-XI (Sept. 7, 1951). Ibid., T R - X I I ( J a n . 30, 1952). K i n g , E. C., Wedge, C . Α . , Jr., " R e a c t i o n of NaK a n d H2O," M i n e Safety A p p l i a n c e s C o . , T R - I I I (Feb. 1, 1950). Ibid., T R - V I I ( N o v . 1, 1950). M o r a b i t o , J. J., S h a n n o n , R. H., P a p e r 55-A-189, ASME A n n u a l M e e t i n g , N o v . 13-18, 1955. T i d b a l l , R . Α . , Progress Reports 25 a n d 26, M i n e Safety A p p l i a n c e s C o . , October 1954 to J a n u a r y 1955.

In HANDLING AND USES OF THE ALKALI METALS; Advances in Chemistry; American Chemical Society: Washington, DC, 1957.