ML20099E918
| ML20099E918 | |
| Person / Time | |
|---|---|
| Site: | National Bureau of Standards Reactor |
| Issue date: | 07/31/1992 |
| From: | Rowe J NATIONAL INSTITUTE OF STANDARDS & TECHNOLOGY (FORMERL |
| To: | Michaels T Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 9208120021 | |
| Download: ML20099E918 (24) | |
Text
_ . _ _ _ - _ _ _ _ _ . _.
j .
- get ** c
- g UNITED CTATCO D3PAATMENT OF COMMORCI
- 8 .
g*
{ National institute of Ctandar ds and Technolo0V u w e st a g e. w ,w a e m a
~* . ..:) gq - /yl July 31, 1992 Mr. Theodore S. Michaels Project Manager PDNP Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission
-Washington, D.C. 20555
Dear Mr. Michaels:
NIST respectfully requests review of the enclosed proposal for a new liquid hydrogen cold source to replace the existing heavy water (D2 0) source. Both the NIST staff and the Safety Evaluation Committee have determined that the proposal meets every criterion of 10CFR50.59 and accurdingly does not involve an unreviewed safety question. Because this is a major undertaking of vital national importance that involves extensive effert and high cost, NRC confirniation is requested. The development of the new source represents a major breakthrough for U.S. Science and Technology. When completed, the NIST Cold Neutron Research Facility will not only be the premier and only such facility in the U.S. but will also be among the best in the world.
Please accept our sincere appreciation. I ala enclosing copies of relevant parts of the references to facilitate the revica.
Sincerely, f/
//
CJ.MichaelRowe Ch'ief, Reacter Radiation Division Enclosures cc: T. F. Dragoun Project Scientist U.S. Nuclear Regulatory Commission Region I 475 Allendale Road King of Prussia, PA 19406 9209120021 920731
{'
-PDR ADOCK 05000184 b I p PDR f
i
,~ . _
l ANALYSIS OF SAFETY ISSUES INVOLVED IN A LIQUID HYD SOURCE IN THE NBSR J. M. Rowe, P. Kopotka, and R. E. Williams
- 1. INTRODUCTON The normal spectrum of noutrons emerging from the beam tubos at the NBSR can be well represented by a Maxwelllan distribution of energies with a characteristic temperature of 350 K. For such a spectrum, loss than L5 % of the total neutron flux is comprised of noutrons with energios loss than 5 moV (.005 eV), or, equivatontly, with wavotengths greator than 0.4 nm, which are commonly referred to as cold neutrons. The prosent cold neutron moderator sourco at the NBSR is a block of heavy water (D20) Ice.
This has provided a cold neutron intensity gain of more than five compared to normal beam tubos. For several years, use of this source was restricted to two instruments located insido the confinement building, in recent years, primarily as a result of the work done at the Institut Laue Langevin in Grenoble, France, the great utility of such noutrons has been realized. In fact, the National Academy of Sciences appointed a Committe in 1984 to set priorities for Major Facilities for Materials Research ano delated Disciplinos, which assigned the highest priority to provision of facilities for cold noutron research. In response, NIST and the Department of Commerce approved the construction of the Cold Neutron Research Facility which is now being brought into operation. This major facility, the only one of its kind in the United Statos, includes a large guide hall outside the confinomont building with seven-noutron guide tubos which can accomodate up to 20 experiments simultaneously. The stato of the art instrumentation being installed is fully competitive with any world-wide, and providos an entirely now measurement capability for U.S. researchers.
The initial operation of this- facility has boon a great success, attracting
!ho best rosearchers from industry, universities, and other government laboratories. They have utilized the unique capabilities to perform key measurements on a broad range of materials, with_ applications extending from biology to polymer chemistry to advanced ceramic processing to fundamental neutron physics. This success has given impetus to exploring now technologies that could make the cold source even more competitive.
1
Proliminary proconceptual design Indicated that a considorable improvoment in performanco could be attained by the use of liquid hydrogon or douterium as a moderating matorial, primarily as a result of the moderator romaining liquid down to temporatures below 20 K. Several hydrogen or doutorium based sources have boon operating for the past v: ado or more (at the Institut Laue Langevin in France, at the Saclay reactor near Paris, at the reactor at Julich, Germany, and at the HFBR in Brookhaven), and data from these sources were the basis for the proconceptual work done hore, in addition, a hydrogen based source offors significant simplification in design, construction and operation relative to the curront source. As a result, more extensive design work was undertakon, including Monto Carlo simulations based on the MCNP code '
maintained and distributed by Los Alamos National Laboratory. From those studios, a design that can provide intensity gains in excess of two when compared to the present source over the entire cold noutron enor0y range has omorged.
In this report, the safety implications arising from the use of hydrogen as a moderator are analysed, and shown to be allowable under 10CFR part 50, section 59, " Changes, tests, and experiments", since it does doos not involve changes in the technical specifications or an unroviewed safety question.
- 2.
SUMMARY
DESCRIPTION OF HYDROGEN SOURCE A block diagram of the proposed liquid hydrogen cold source system for the NBSR is shown in Fig.1. The moderator chamber contains the liquid hydrogen used to moderato the noutrons to a lower effective temperature.
(All components will bo discussed individually boiow in more detail.) The hoat generated by neutrons and prays in the liquid hydrogen and its aluminum container is removed by boiling of the liquid. The vapor generated oxits from the moderator and cryostat and retums to the hydrogen condensor by the vapor return line, where it is re liquified and returned to the moderator chamber by the liquid supply line. The overall
-vapor-and liquid are maintained at approximately 21 K, and the cooling necessary to remove the heat is supplied by the closed cycle holium gas refrigerator, which can deliver up to 3.5 kW of cooling with a final gas temperature of 18 K. Under normal <porating conditions, the entire
( g c
L_ _ _ _ . _ _ _ _ _ , _ _ _ _ _ _ _ _ . _ _ _ . ~ . _ _ . ~
hydrogen system is maintained at a constant pressure of approximately 22 psi (150 kPa), so that the boiling temperaturo is approximately 21 K. A largo ballast volumo, maintained at room temperature, has sufficient volume to contain the entiro hydrogon inventory with the refrigerator stoppod, and tho system at 300 K, at a system pressure of Mss than 75 psi (500 kPa). The driving furce necessary to establish flow of liquid down to the moderator and vapor back to the condensor is supplied by the density difference betwoon liquid and vapor, acting over the holght difference betwoon the condonsor and the moderator. This is known as a natur circulation loop (sometimos referrod to as a "thermosiphon"), which requires no moving parts to ensuro flow of liquid to the modorator for cooling purposes. The calculated flow for the NBSR installation, based upon both the original calculations of y ray and noutron fluxos, as well as independent Monto Carlo calculations using MCNP, is 3 4 g/s, corresponding to a heat load of 1.2-1.6 kW. Calculations of the loop stability, as well as indopondent friction calculations, indicato that about 20 cm of head is adequato for this circulation, while the actual separation is in excess of 200 cm, leaving amplo margin.
With this cystem, the hydrogen loop and ballast volumo, once leak tight 3 and charged with hydrogon, is entirely closed. This ollminatos further gas handling, and thus minimizes the possibility of inadvertent oxygen contamination, in addition, all hydrogen lines are either within the I biological shiold, encased within heavy stool shields, or run in an existing floor tronch, so as to absoluto!y provent accidental rupturo during operation. All hydrogon containing components are completely surrounded by an atmosphoto of holium Das, maintained at a pressure abovo atmosphoric, so that thoro can be no in loakage of air into the system, and thus no oxygen available to combino chemically with the hydrogon. Tho only contro!!od and monitored system paramotor is the absoluto prossuro l
of hydrogen in the vapor phase, which, in equilibrium, is a measure of the temperature of tho hydrogen in the source (in direct analogy to a vapor bulb ihormomotor). This system has the advantage that no active measuring devico need be installed in a high radiation area. The entire design philosophy is to rely on simplo, passivo safety features that l-minimizo the possibility of a system failure or a procedural problem. With i
i the closed system, gas handling is minimized (the only charging is dono at installation, and after the system is opened for correctivo maintenanco).
3
Thoro is no provision for hydrogen venting in this system, as this would '
again croato more possible falluto modos with no safoty gain. Rathe described below, the capacity of tho ballast tank is relied upon to hold entiro'chargo safely, if the system must 00 emptied of hydrogon to allow
- maintenance on a component, this is done by absorbing the entire inventory into a storage metal hydrido, which then can bo removod from the confinomont building. This rollance on simplo passive systems is, i our view, the best safety philosophy to guarantoo reliable and safe operation.
The components of the system are described in more detail below, it should be noted that all vossols and piping are designed to the ASME c for pressure vessels, using the maximum design stress speciflod fo T6 aluminum with wolded joints (6000 psi). The rupturo strength for suc vossols will be many times the design working pressure. In each caso, th design working. prossure will be specified in the descriptions below.
2.1 MODEIMTOR CHAMBER The reactor vessel, beam port, moderator chamber, cryostat, and neutron guidos are shown in the plan view of cryogonic port F in Fig. 2, and moderator chamber is shown in more detail in Fig. 3. The liquid hydrogon cold moderator is arranged in the form of an annulus formed by two concentric sphorical shells. The outer shell consists of a sphere made from 6061 T6 aluminum which is 32 cm. In diamoter, and 1 mm thick. The liquid hydrogon supply line and vapor return line entor from the top in dome known as a "bubbio cap" (which acts as a liquid vapor phase separator). It should be noted that the liquid ontors via a 0.5 in. OD (
area) innor tubo, while the vapor leaves through the annuius formed b 1.25 in. OD outor tubo, (with 4 cm2 flow area), an arrangement which ensures that the system is isothermal. The innor shell consists of a sphoto mudo from 0.25 mm thick 6061-T6 aluminum, 28 cm in diamo with a tubo penetrating from the bottom. This innor sphere is hold in the contor of the larger sphore by contering and supporting devices, and has a protrusion on the sido facing away from the reactor which matches t outer sphora diamotor. This prevents liquid from entering the area of _th source which is v10wed by the cold neutron beams (see Fig. 2). Thus, the beams view the inner surface of the spherical shell, in
- black-body" l
4
geometry, a design feature which substantially increas@s cold noutron flux. With the reactor operating, the liquid in the shell will be cooling the outor and inner sphoros. The bubbles riso to the top where they food the vapor return lino, leaving a calculated liquid fraction of greater than 80%,
as required for moderation. The total volumo of the annular space is approximately 5 liters, which imp!!os 4 liters of liquid during operation.
The outer sphere 'with bubblo cap is designed to the ASME pressuro vossol code, for an internal working pressure of 75 psi (500kPa) and an external working pressure of 23 psi (150 kPa).
During norms operation, tho innor sphore will be filled with only hydrogen vapor, as an, liquid will boil off, and the resultant pressure will not allow i
liquid to entor from the bottom. This arrangement allows a largo vlowing l
area for the beams, while reducing the total hydrogen inventory, and l
allowing the use of the spherical shape, the best for stress reduction. The J tube which connects the two parts of the moderator chambor reduces normal liquid vapor interchange to a diffusion limited process, ar.d allows the vapor to have a different ortho / para hydrogen ratio than the liquid.
(Normal hydrogen is 75% ortho, while the equilibrium ratio at 20 K is virtually 100 % para.) For cold neutron officiency, the vapor should approach equit Nm at nearly 100% para, while the liquid should be maintained at no rly 75% ortho, since the ortho form with spin 1 is a strong neutron scattorer while the para form with spin 0 is a weak scatterer. While the spherical shape and other details are now to this source, the concept of a vapor filled domo is not, and has boon shown to work satisfactorily in a hydrogen source at Saclay.
2.2 MODERATOR ORYOSTAT For heat transfer reasons, the moderator chambor and all other low temperature regions of the system will be enclosed in a high vacuum insulation region (<.10 4 mm Hg). In fino with the overall design philosophy, this vacuum must be surrounded by a blanket of holium to provent the possibility of air entering the hydrogon system. The vacuum and helium vessels will be cooled, sinco they will be in a region of radiation heating. All of these requirements are met by the cryostat assembly shown in Fig. 4, which includes the moderator chamber and hydrogen lines. Sinco this assembly is also the barrior betwoon the 5
r
- moderator and the reactor beam tubo, the hollum and heavy water cooling jackots have boon designed with internal working pressures of 139 and 141 psi respectively (945 and 960 kPa), and external working pressuros of 235 and 19 psl respectively (1626 and 130 kPa). Those prossures are for oa h pressuro vossol working indoponc'ently, so that the two together will be correspondingly stronger. The calculated rupture strength of the helium jacket is greator than 1800 psi internal pressure. The water cooling for the aiuminum cryostat assembly is supplied by the auxillary D2 0 cooling system, just as for the existing ice cryostat. The calculated heating rate will requiro 25 gpm of cooling water flow for a temperaturo rise of 18 'F.
In order to ensure adequato cooling of the vacuum shell (which is separated from the cooling water by hollum gas), the space betwoon the two shells will be filled with aluminum " wool", which will not aifoct the hollum, but will increase the thermal conductivity by at least one order of magnitudo over that of helium alone.
The entito cryostat, moderator chambor, and associated linos (hollum, vacuum, liquid and vapor hydrogon) will be attached to a plug similar to tho one used for the existing ico cryostat, with openings for the noutron guidos. This plug and cryostat assembly will be re"od into the beam port from a shielded cask following the same genort' -Ocoduros as were used for insertion of the present source assembly.
2.3 HYDROGEN C')NDENSER AND CONNECTING HYDROGEN UNES The hydrogen condonsor is shown schematically in Fig. 5, along with thu liquid vapor phase separctor at the bottom, the connections to the hydrogon liquid supply anr: vapor return finos, the connection to the helium refrigerator, the connection to the ballast tank, and an auxiliary system for ensuring that the ortho concentration in the liquid remains near 75%
for good n'eutron moderation performanco. The separator itself is a plate.
fin typo boat exchangor obtained commercially, with working pressures for the helium and hydrogen sides of 300 and 150 psi respectively (2040 and 1020 kPa), and a 3.5 kW heat transfer capability. Cold hofium g ;s ontors the heat exchanger at approximately 14K and a pressure greater.
than the hydrogon pressure of 74 psia (warm). As it passes through it is warmed by condensation of hydrogen and returned to the refrigerator at 6
l l
- - - - ..m.,.- 7
approximately 18 K. The liquid drains to the bottom of the heat exchangor into the phase separator region, whkh allows tho incoming vapor stream to expand and any liquid carried over to be separated and returned to the cryostat through the liquid supply lino. This feature ensures against l accidental voiding of the chambor as a result of transient conditions. Tho oporating polni for the condonsor is controlled by the incoming holium temperature and mass flow, which are controlled by the absoluto hydrogen pressure, which is a good measure of the moderator temperature, since the supply and return linos are isothermal by construction.
The auxiliary system shown removos approximately 0.1 g/s of hydrogen vapor from the condonsor system, warms it up to 300 K, converts it to the normal 300 K ortho /p .ra ratlo (3/1), and returns it to the condonsor. This system imposos an additional heat load of loss than 100 W, and is very offectivo in maintaining good neutronic performance for the moderator system. As was discussod' for the moderator, the condonsor is surrounded by an insulating vacuum, which is in turn surrounded by a helium blankot.
The auxiliary system is entirely onclosed by a helium blankot.
2.4 HEllUM REFRK3ERATOR The helium refrigerator is of standard design, obtained commercially with a cooling capacity of 3.5 kW. The gas is circulated in a closed loop from the refrigerator to the heat exchanger /condensor to remove the heat generated in the cold source.
P.5 BALLAST TANK The ballast tank and pipe which connects it to the rest of the system are shown in Fig. 6, which also shows the relativo location of all system components. Tho purpose of the ballast tank is to provido an adequate hydrogen gas reservoir to hold the endro hydrogen inventory at a pressure of loss than 74 psi (500 kPa) when the entiro system is at 300 K. In view of the large ratio of volumos betwoon liquid at 21 K and gas at 300 K (approximately 800 at ono atmosphere pressure), this tank must have a largo volumo it is entirely surrounded by a helium blanket in order to provent air from entering the hydrogen system. The line which connects the ballast volume to the condenser has a check valvo installed to ensure 7
m.%,
that hydrogen can afways return to the tank, but that gas can only ! cave the tank when the solenoid valve (AV 1 in the f!;ure) is activated. This featuro ensures that hydrogen cannot be trapped in the cold system with no expansion volume, while also ensuring that if AV 1 !s closed, gas which enters the tank cannot leave again. The design limits tho volume of gas in the external system when the system is not operating (LL during maintenance), when the system could be vulnerable to accidental rupture, l l
< since shields may be removed.
Tho tank will be located on the first floor of the confinement building
~
wall (C-100, see Fig. 6). The tar.k will be surrounded by a steel frame to protect it against any accidental damage that might occur as a result of any operaCon in C-100 The piping from the tank to the reactor biological shield, where the condenser will be located, will be inside the existing floor trench, where it is protected fwm any possibility of accidental rupture. At the reactor biological shield face, the pipe will pass up through the massive radiation shielding to the condenser, which ill also be surrounded by steel shielding to prevent accidental rupture from any external cause. Since the system will not be operated unless the shielding is in place, the entire hydrogen system will be enclosed by protective shielding which will prevent the possibility of damage as a result of any external occurence while the system is cold, and the inventory of hydrogen available for release is large, When the system is warm, most of the hydrogen will be in the ballast tank, and protected from release by the valve system described above. This design protects against c massive splli of the inventory to the confinement building, with the possibility of delaved ignition and detonation.
it should tw noted that there is no. provision for either automatic or manuai vW' J of the hydrogen inventory, either inside or outside of.the confinemn '. b"ik6ng. This is a conscious decision, based upon analysis of various , anai s that might occur, such as loss of cryogenic cook.ig,
!eaks in ' T system, loss of insulating vacuum, fire and o'hers, in each case, it was concluded that the ballast tank is the safest possible
!ocution for the hydrogen. In any venting system, great care must be taken to maintain high discharge velocities (to prevent back-mixing of ali), to maintain an inert gas atmosphere in the vent lines, to ensure that venting does.aot result in an explosive cloud, to ensure that all of the hydrogen is 8
l pushed out c' the system, and to ensure that the gas used to vent is inert.
A!I of those requirements require complexity in apparatus and procedures, thus introducing new failure modes. The ballast tank is designed to safely withstand any pressures that might be generated by an extemal fire , if the system must be emptied, it will be absorbed into a commercial metal !
hydride storage device, which can then be safely removed from the building. With the system warm, and the ballast tank filled to 75 psi, the remainder of the system will contain less than 10 g of hydrogen (approximately 110 STP liters) out of a total inventory of more than 500
- g. As will be shown below, this amount of hydrogen, even if released and ignited, will not cause structural damage to the confinement building.
- 3. GENERAL SAFEWISSUES For refore.m. the definition of an unroviewed safety question, as defined in 10CFR50.59 is reproduced in whole.
"A proposed change, test, or experiment shall be deemed to involve an unreviewed safety question (i) if the probability of occurrence or the consequences of an accident or malfunction of equipment important to safety previously evaluated in the safety analysis report may be increased; or (ii) if a possibility for an accident or malfunction of a different type than evaluated previously in the safety analysis report may vi created; or (iii) if the margin of safety as defined in the basis for any technical specification is reduced.
The issues raised by the use of hydrogen in the beam' port of the NBSR are limited to damage to the reactor vessel or to the confinement building as a result of a chemical reaction between hydrogen and oxygen with a large energy release. Because the density of hydrogen atoms in the_ liquid is much less than in water or ice, any reactivity offect will be less than that of existing systems.
The primary safety philosophy in the source design consists of preventing oxygen from interaction with the hydrogen by passive design features, such as surrnunding all hydrogen containing volumes with a bfanket of helium gas as a barrier to air in leakage, minimizing external connections which penetrate this helium barrier, and minimizing hydrogen gas handling 9
which could lead _ to inadvertent contamination of tho hydrogen by cir. In addition to these design oloments, all components are fabricated to withstand the reaction which could take place if air were introduced.
Finally, all hydrogen containing pipes are protected by shields designed to prevent accidental rupture of any line. Calculations based upon both theory and experiment are presented which show'that no incident could damage either the reactor vessel or the confinement building.
- 4. CHEMICAL REACTIONS When hydrogen reacts with the oxygen in air to form water, it releases a large amount of energy, equal to 1.2 x 10e joules per kilogram (or 24 kg TNT per kg). Further, hydrogen has a wide range of flammable concentrations in air - from 4 to 75 % by volume - and of detonability limits - from 18 to 59 % by volume (the stoichiometric volume fraction is 30 % hydrogen by volume). This wide range of flammable and detonabfe limits is partially compensated by the rapid dispersal of hydrogen in air, as compared to other flammable gases. In this regard, the following definitions are used. A. deflagration (or fire) is a reaction in which the flame front moves through the flammable mixture with a velocity less than the speed of sound. In such a reaction, one can use the usual adiabatic form of the gas law to show that the maximum pressure ratic (gas after combustion / gas before combustion) in a closed vessel filled with a
' combustible mixture is approximately 7. A detonation is a chemical reaction in which the flame front is propagated at velocities exceeding the velocity of sound,' generating a shock wave. For such a reaction, the utio previously defined is less than 16. In an unconfined reaction, a deflagration is of little consequence beyond the damage done by the fire itself, while a detonation can cause damage by the blast loading resulting
.from the shock wave generated.
i As discussed earlier, the primary design goal for the proposed NBSR liquid hydrogen source is to prevem mixing of air or oxygen and hydrogen, by passive design features, and by minimizing hydrogen gas handling. Thus, it is difficult to create a credible scenario that will invofve either a.
deflagration or a detonation of any significant voldme of hydrrgon.
However, in order to show that the system will not involve any unreviewed safety questions, a series of possible occurrences are 10
i I
analyzed, and will be shown not to cause any damage to the cryogenic pI to the reactor, to any safety system, or to the confinement building itself.
4.1 RELEASE OF HYDROGENJO CONFlNEMENT BUILDING As discussed in Section 2 above, all of the hydrogen system external to the biological shield is either surrounded by steel shields, or located in existing floor trenches during cold source operation. The ballast tank is well-protected from damage both by location and by the steel frame wh surrounds it. These features make any massive rupture of the hydrogen lines (which might result from an accidatal contact with any load being moved on the C-100 level) when the hydrogen inventory is primarily located in the moderator not credible. However, during reactor and source shutdown, when maintenance on the guide system might be performed, the shielding at the reactor face (which forms a large part of the protection of the piping) might have to be removed. Although most of the hydrogen will then be in the ballast tank, and held there by the check valve and AV-1, which will be closed, as much as 10 g of hydrogen gas at 74 psi (500 kPa) could be in the moderator chamber, condenser and connecting lines.
'he consequences of the release of that amount of hydrogen to room C-100 have therefore been analysed.
When hydrogen gas is released three possibilities exist - first, the gas can diffuse rapidly until its concentration is below the flammable limit with no other consequences; second, the gas could ignite as it is released, leading to a continuous deflagration in open air (this is highly likely, as an) ident that is of sufficient intensity to rupture pipes will generate spar..., as an instantanous ignition source); and third, the gas could mix with air, followed by delayed ignition and detonation. Of these possibilities, the third is clearly the most severe (although least likely),
and so has been analysed. The theoretical maximum energy release from this is equivalent to 240 g of TNT, if the mixing is 100 % effective at exactly the stoichiometric composition of 30 % hydrogen in air. However, l
this yield cannot be achieved in practice, due to the nature of diffusive mixing, and so the yield will be substantially smaller - 10 % is generally considered cc'servative. Nevertheless, if we do assume the maximam
' possible :teld, then the blast effects can be estimated from standard 1s is distance curves for TNT (see Fig. 7, which was obtained from Ref 1; 11 1
I
e
~
discussed in detail in Ref.1, it is essential to consider both peak pressure and peak Impulse (pressure.x time) in assessing the hazards associated with explosions. This reference also shows, on the basis of extensive analysis.of bomb damage in London during the Second World War, that threshold values of peak pressure and impulse can bo defined Independently for- which no structural damage will occur, in particular, if the impulse generated is less than 120 Pa s, then no structural damage will occur even for residential construction. UsinD Fig. 7, the peak impulse for detonation of 240 g of TNT (10 g hydrogen) will be less than 48 Pa-s for all detonation distances of more than 1.5 m. This impulse can be doubled by reflection from the wall, but even this reflected impulse will be'less than 100 Pa s, well below the limit for structural damage. Any actual hydrogen release would have to occur much further from the confinement building walls 'han 1.5 m, since the only exposed systems which might be damaged, leading to a release, are near the biological shield 10 m from the wall, and air flow is away from the wall. Thus, such a release could not damage the confinement building integrity, although it could have consequences for equipment and personnel in the immediate vicinity. The secondary effects, such as wind driven missiles, present a less hazard to the building.
- Since all of the assumptions in this analysis are conservative (100 %
yield, release of all 10g of hydrogen even though the rate of refease would l
slow very rapidly when the hydrogen pressure reaches one atmosphere, l
! damage criterion based on residential construction standards), it is concluded that such an accident will not lead to damage to the confinement building, and therefore cannot increase the consequences of any accident previously analysed. Every precaution has been taken to minimize the possibility of a gas- release, and procedures will minimize
- this probability even further, since this provides the only possibility of personnel injury associated with the hydrogen source operation. In addition to all passive features described, hydrogen monitors, set to alarm and shut down the refrigerator if the concentration of hydrogen l
exceeds 50 % of the lower flammable limit, will be installed in the area l
L and will give audible and visual warning to any personnel present l
12
4.2 O
_ XYGEN CONTAMINATION OF THE INDROGEN The entire design philosophy of this source (and overy other hydrogen source in operation) is to prevent the possibility of oxygen entering the hydrogen system e.g . all systems are surrounded by helium bl&nkets, gas handling is minimized by use of a closed system with no moving parts.
Nevertheless, two possible accident scenarios have been analysed, one of
' which is considered possible, arising from residual oxygen in the supply gas, and one of which is not considered credible, arising from an unknown cause and leading to a stoichiometric hydrogen air mixture in the moderator.
All hydrogen gas contains residual oxygen at levels measured in ppm by volume. Since this is far outside the limits of flammability, it is normally of no concern. However, in a system containing liquid hydrogen, solid oxygen might concentrate in the liquid. Again, this is normally of no concern, since there is no ignition source, but in a radiation environment, the creation of ozone, followed by conversion back to oxygen can provide such an ignition source. The NBSR source will be filled with high grade hydrogen of less than 10 ppm oxygen, and will be tested before being brought into the confinement building. For conservatism, we assumed that in fact the oxygen contamination is 100 ppm by volume, and that all of this will end up in the' moderator as solid (this is so unlikely as to be incredible - most of the oxygen will remain in the condenser). Since the maximum hydrogen inventory will be less than 1 kg, this implies 1.6 g oxygen in the moderator chamber. Assuming maximum yield, this corresponds to the reaction of 0.2 g hydrogen, with an energy release of 2.4 x 104 joules. This energy will show up as a temperature (and therefore pressuro) rise of the hydrogen and/or as steam. In either case, the resultant pressure will be less than the elastic limit for the moderator chamber, and will therefore result in no damage to any component of the j
source. Therefore, this accident scenario presents no hazard to any part of l
the reactor, its safety systems, or to the confinement building.
L Although no credible scenario that will produce the appropriate conditions has been identified, a case in which the moderator vessel is filled with a stoichiometric mixture of hydrogen and air at a pressure of 14.7 psi and 300 K (clearly, the pressure cannot exceed one atmosphere, or else there 13 i
i
p would be no air present) has also been analyzed, in this case, the resultant pressure from a deflagration would not excoed 7 times the initial pressuie, or 103 psi. This is below the rupture strength of the moderator vessel, and so would present no problem for the reactor or any component of the source.- A detonation, on the other hand, could lead to pressures of 16 times the initial pressure, or 235 psi, which should not rupture the moderator vessel for which the calculated rupture strength is greater than 500 psi. However, the working pressure of the helium containment chamber is 139 psi, and the volume of this vessel is six times that of the moderator. Therefore, even if the moderator vessel did rupture, the outer vacuum and helium vessels would contain the resultant prrissures easily, and the chect of such an event would be entirely contained in the cryostat assembly. It is therefore concluded that neither scenario could result in damage to the reactor, its safety systems, or to the confinement building.
4.3 _NTRODUCTION OF AIR INTO THE INSULATING VACUUM S I
Although all insulating vacuum spaces surrounding the cryogenic portions of the system are surrounded by helium, the consequences of air entering these regions, and being cryopumped onto the cold surfaces, has been analyzed. This would require both that a leak to the vacuum space exists and that the helium gas is contaminated by air (the helium containment systems will be continuously monitored for oxygen during operation, and l
maintained above atmospheric pressure). The possibility of cryopumping
' air into the existing cryostat has been previously analysed in NBSR-13 (Ref. 3), and shown not to be a problem in and of itself. We therefore go beyond this scenario, and further postulate that after prolonged, undetecteo cryopumping, the moderator vessel ruptures, and a strong detonation source is present (all of these postulates are necessary -
otherwise, the reactor would be shut down, the source would be warmed up, and the problem repaired). In order to estimate the magnitude of such an event, some level of air contamination in the helium containment I system must be assumed. A level of 100 % was used in spite of the fact
' that it is' not cred;ble. A leak rate that would go undetected, for example as a degradation of the insulating vacuum, was also assumed. This has been analysed in detail in NBSR-13 (Ref. 3) for the present D2 0 ice source, where it was concluded that no more than 1 g/ year could result from a leak with 10 % contamination of air. Therefore, for the postulated 14
scenario of 100 % air,10 g of oxygen would be the absolute maximum, in a series of tests with a strong detonation source (Ref. 2), the peak pressure ever recorded for a deliberate detonation of hydrogen with 80 g of solid oxygen was 1120 psi, well below the rupture pressure of the cryostat helium and vacuum jackets. Thus, the entire effect would be contained in the cryostat system, and would have no effect on the reactor or any of its systems. This scenario, which goes well beyond the maximum credible accident, in that it assumes failure of many independent systems simultaneously, and violation of many operating procedures, serves as the design basis accident for the NBSR hydrogen source. The results of this postulated accident are entirely contained within the cryostat, and do not affect the reactor or any safety system.
5.0 STARTUP TESTING AND PROCEDURES Prior to insertion of the new source into the beam port, a set of startup testing and operation procedures will be developed and reviewed by the Safety Evaluation Committee. The startup tests will be designed to verify the design calculations of heating rates, thermal stability, shutdown mechanisms and other critical parameters. The initial tests will be performed at low reactor power, with limits on observed parameters which must be met before proceeding to higher power tests. During thess tests, the neutron performance will also be measured, using the already installed instruments in' the guide hall.
l All existing reactor operation procedures relating to operation of the
! existing source will be reviewed for applicability to the new source, modified as required, reviewed by the Safety Evaluation Committee, and updated prior to operation of the new source.
- 6. CONCUJS]QN Since none of the accidents analysed here, or identified in the overall safety analysis, have any effect on the reactor or its safety systems, nor on the confinement building, it is concluded that the liquid hydrogen l
l source for the NBSR does not involve any unreviewed safety questions.
15
o
~ .. .
-7. REFERENCES 1 '. - Explosion' Hazards and Evaluation, W E. Baker, P. A. Cox, P. S. Westin, J. J. Kulesz, and R. A. Strehtow,- Elsevier Publishing Company, New York,' NY,1983.
4
-2.- D. L. Ward,- D. G. Pearce, and D.' J. Merret, Advances Cryo. Eng. 2, 390 -
(1964)
- 3. NBSR-13, Safety Analysis Report on the D20 Cold Neutron Source for the National Bureau of Standards Reactor, August 1984 (Revised May, 1987).-
F Si
+
16
' ~
r-~~-"'***~~~~7; *
- CONDENSE 9' I' (21K) 'l I
1 *
.I. .
1 CLOSED LOOP
$
- HEUUM REFRIGERATOR (COOUNG CAPACITY TO 3.SKW)
I 1 I l
- /
Hg GAS BALLAST s l
u TANK f.
(300K) _,,
___ _ _ __j i
l' M
H, YAPOR RETURN UNE C
\
)_ Q Ha UQUID SUPPLY UNE SPHERICAL SHELL OF ,
UQUID HYDROGEN .
MODERATOR CHAMBER ,
MODERATOR CRYO.WAT i
Figure 1
t +
_;.~-
e
. . ? l S.
~ ,, 29 I E
$j
$ $ o- h5- d #$! i -
f -
A l./ ./ ~
- \ ~). - - -
b
\ 7:::.- j jl,! U
\1
\ : -
I!//
/
\. l lj
\
\.
}$? ll /)
/ ,\ , i, ,
y/
1 I. 1
[ .
II 6ziazzA C, ~
gj%4,[y #
! [
n y a i a.
il,,,
,3 t y )- .I ag f
t
- l
\./ //i- lMj
\;
r
$0
=
2 ~% ' - \ f l 4
\
jj tx !
- f. ,I
~
Oc :a DanOg e-0 0/
~
~,m ' ! ' .
[ .
- ,~
" R
.. E .
B
.M A _
H .
C R
O .
.T A
R -
,_. vE D O F L
M A D H E E L R B E M
E H
P S S S
A R 4 E T E m U S i
c O A B -
2 3 R O
T A
R E
E D R O E M H
P S
R. R O E T N A N R N I
3 A
- r e
^
P. T_E D UI~
E,,
i
,,:I:I:I:I S OU Q _ ~
. ~
u O RU- g, P P F i R_ O~
A V
/,
D A
V .
F I
- .\
U
.\'
Q U
.N' l/
~
{.p l; ];
i:1:l
. k' i. .
m c
ke _-
8 2
~
s E
R E
H PF
- SL A
RH E
T
- U
' O
?
vL .
4 a S ,
4 7 T A
T S
3 O Y
R C
- m T'
A W
m R
s O
C W jwE M D
X 8
.M E
S S
A m,
W w -
E I
E i
V v L t A A 4 TN T A N AO TI T0 e -
ST 0C Sn Oc r
" YE RS YE RS u
C- L C- g S E S i RS S RS F OO S OO TR E T R AC V _
AC R T R
EC N EN D0 DIR E O T ME M q
- O Mn M C
r N .
7 A v A M I A
J O T E S L G I N 7, E N
\.
O C
M Qjft
- ~ g O
V _- C 2
1 E R E
- H __
T A
W vUUC M
VL GS E
RM E c NS x m.
DH OC M
A
._ t >
IE TV A
L U
S N
I f
/!
~
\
/
/[t\
- '.' : - ,t , , '
F.
,,A . T
.O Sr.
AN
's s
_ LA
- 0) . TLT
- i A
B D. 7 MN E
T- Gc EN SNOi
!i:
YiRT
'- +
, 3
- l, I
- \:i.
-l:l,I:
I i__SADA T YR L E
_-{ !
_YN
_RIHN T S S
i -
AE E
- lA MOC
. T' -
L V
. SO YH IOHN X TO T
._ S LT AR UTRC N E
A TON A U
. G A 0 M NP C1AS . N u.\
I s
. I EM &Rw A A T wl GU OP R
D^
I RP v E
TRo A OC h A_U L
E S
S A
O C .
.. Y EF _
. E M
_ H .H _
V U
_ M U E
. U 1,
- !i:l!
_ U H
- 1:l:1
-j:l .1 C .
c.-} ti A
Y g K 6= y$
g 5
r,l i< g g e g i u
g i
~
F i) s " d C
N),/ - n 1
.M R
/R .WS O P
E 6= A V
'L'NR A
YNAE CIHS G '
_, N>" u R FCN E XE % Wh MEED T
C MATN 6 0 OLA O CPEC H
6 3 1
1
\
~ d C
_w 1
)
~ o
- c. c 4
v - :
),,M
,g
)5%
e)'-
m, s a w!W i!Q
,.jW '
i j:
~
~ }'
N s _
~
u !- I ,L
. ! : l i
g 1 E N y - V g V L E i
"o g A A V
EC DI f 7 IV RE
- K- DD N j fC g
g 5,
z t
g g
C E
H C
Y HE LA AR G
TO u
/
j s ET MS _
N '
i q ic .
1tj j 6 3dpj= T ,,r :'
Q /
/
e r
u
/ g z
- i F _
i;hj p .,
/ '
j
$ /
h~ '
R /
.^
- h f
E S
N E !,
/
y D N
O C
/ -
~
j '. -
- : ,l !1 .*l1 -
A
.- W
=..
h.
g, '
T _
., - ] _
z _
?.
~
].
' i :u
.. 10 10d,-
8 .,r x . .m. ,.c.y,;f,3!L,yV c{y. ,. . . . .@ .mm;. 3r, L;m. J t
d, r....m g_._ g in.un. g. r;m.P, uRs g y .a
_g. ,._ i, 10 '[ ,p d !Ow.rps:vW, a o. , %!
.e, . . v. g , . . _
a
. . s.- ,. ....,... ,
m _..%.s.e. g4 , ... . . a . _. . .
.a :.. :_ ,s.a F.. .: n. _:tw;p3+g.,
% i 5, w .y -
- 2 Lm. A . ..
_+
, i i i 4< . p. .,n a 'r
- w a nm:A p* a .. > .
n ,
- m.7,wm-r
. c. i u f_.; 1 4w r mE 7 rrc L.75- ..w J._- ,: T .c 1{_ .L .i yn L ..A . g; qi: '
5 .
c r_ 5 g<. 7.
m i 4.Q.; . h. . ". .,-
- U .w
. t.
ao n. ..4. 4 r w.m1.F t.i . F.
m m T.yri pg W .e rmgw.a q;;t4.s.
n " m .- . .
- l . P. vr#y_
. s. . . r. - J. Y.;.
e-- ,s-: n,2 -- ...,r..r t ,1/3
"- : .M .: ....- -- -
4 .
- : r-x -
r ,.7 ,33 P, a_- .- r..t p+ ;- .
,.:.r ys..a t.r.T .... ,,;. . 5 .w._p.,L.
l ..,. .i..
+
..-:.q...-t , . . ,.. :5_,. : : e. .,- _r, ' .m,, . ,
m i gw i r b. .
1,: , .,.. .
343 .#
.g..
av r ,, y
. _ _., ,.- 6 2 7 ,
4,j,4 ,_. ,
.q . .,,
.A..,..-
. . _w ,g #
- f. ,wo ,
- . 4,.7__9
.e.,. .,
4 10 ,....3.,,....,. . .. . .
- 4. .
._ g ,
L' s 10 ,
- e. _ __ . , ss .L. , . . .- . . a+ - a. ,e -
' ~.u, .-
i T' gM' .4 # *g y,: t"gj 'fg }
+
g'
~"4-w u + ~i%~1: s'
~+~-
,w' p "r-- g~ g p'-y' y7 1/3 -
2 4
p Up'"gmy.g
+TmW. : 1 :qq f g rz M..g]@q t3IW e 191rF W q 7: s m%
- .__ eg
.m y ;
~.
3.7_ _
q i m~g# e
- a. ,
+ -- 7 , g4 + j.j 1
.+s.
+ +-
g
.,J -4
- t_ -
i u.
as
- v. . . ...
M- , _0_ - ![F ' P w.".
gg1 /3 gnd a
E.. ,.4. 9
-i gr ajr
% .z _M .:E--m. :
u.. r. . , .
. ,. . .ig-0r h-3 ' - k. . ..t ---= :,.r-3.,.
l
.i ,- ,-
-l' . = p .rm.2,.J 10 q
y.
6 .Tll: , .
- u; ; _; 4 _rv m pup .t,.m:pr r . _ _n_i Lw ha r. +u_.
3 10 10 :
rg. r:m.v..,-.. r_
3 ,.-- __ '
A +- pd+H H _.m-,-w %a
. ._ _ ._ _ ._ ._m . _
_ p. ._ . . .r _
p_ - , .
1/3 37 W w- e ;. -
r, . m _,,A . ..
r -._4a, s
- 4. n -
. .. . 1 ..
i 5IW _.
! 9 4- r .
_3 m .
-_ ..t2
- . , 2-
,. q . .7. ._y- t iT.
m2 (t, n. .. r. . . -g--
e
,- _,.- - i s
_u _ .
5 , p-_[_ -
.. j_..._ _f T / W f3 :n _.
.g _ g :w_.g .m. _. p _. y_g3
.n w _n 93 .p A n
u-x3 r, - '
~,
,as m 5 - -=-* = :
.w mm =am -
--..L1--
kg ii@ n.m.4 i -u? , ==
^
u=
". . .DN .' -. h.-4 -: ".m F .b 3' ~ ~ ': z'y w~F--.: f; m7 ". %, R_.' +" *.. n ~~.~
-4 5
- . . ... ="P
. %=--
p p .o9,,,. p$i u x n. m; . .. .- q
- +r 4, m'sd autm ,] p .,
L q yi d
1(
- t=W.n--:
M 10 . pF +E*, i .".; . _,
10 -
q- wy o ,l 4 }L*. j h, . .
-.r4d.3 W.;
1 ;
- p n-Lt g<
W: c
'n .g
- _j q, w : ; =h a y _.
..t. - . . . m.. ;
1p ',.,. C_% a :d o*
. : 2 ..
-) O-im . ;
Eph gf 4.p:
g".,;
A J
..a.~pg
.. q/ L g n__z 4; :--m ; -
m .
r! .u e ae v s.b - y m 7 r.ne : h. mm
[f99}k%
ab e
- MWW. , %wI, * .d,(~+Qi 9 It c - ,e y %,m 3@
E -
. a 4a
'9 lIIE ,.._[~. , .. ,
3 r m , + . ,.~
N~m=P
-%m - -
.fr f U: 72 0_%
g4,. v. e m.*'i . __ )
- v. .. -~' D .U -
r, ,. 1..
,M .?
> mwa .e_s J
.: g. =a . .A. o m. .M. .n. ..a p.. -
9W . F, r 7 % 'T: +- ,-
=qw g
.z. M..u.:._E. *l . -.
. n: edr--- -e4p .- .. -u .
,y. .s_h.,m y .... .
., .. . m ~ r===
.i_ .g. - ;.
- x o r a ,a =y ,3 5 n
.a . .:- - - -
.&-=_h- -m. m r Oc. - - ==d .=d- m c == .c :: . -
m?' = . = ===.
4 ---.- =pt,3c t=c x.:= -pv a , su p gmMq-q- i n1/3 !C
-5 g c ri -
l % i gmc:pApy M W 'r.e; . g nr- ,=W 10 10 7~7,,e . b=t -v v .r,n c m-2 .
5 r T , 5e i
- v. , a r }: . ,. 1 +R.~ f-yr.me.s a .
in, w. W'_.@.N n. .
syy m. t: .Q3 wg nj%m i 1 -
m ._- r.. -lpp ! . .
r.u v.J@_., .
.}(
g g ,..p 3 L.
x
..m WmM_a A 2 s
mi . y-r 1
- n y.]._.C; ; ,-r ._ -m4 nt
- ,u.tr m w
,g..~ 3 m: r -Np mm.m. mmT T, "f-r*M w'.g r
M7N 1 .
w.g -*' ' ' nm 3 T1' -
j y ' f-'D.--
' E. . 1.. ~';3 S is '; . TL '
p 7p ~ ' fT 1- T,_ ' F. t l.p ,j y 1 T gi 1 N ru .- s v.= [
Ja 0-. 't. &. ' '-!- {'e]J
. t. g Q wr rm 3 mm
,3 y_]F 7
n: uw 5 in . M4 r ps - w '
- 7 - -
OW W! :
3:W h 4:'. n_.-
% ad i '.
- q. . '.i.
JE T_ .__S_.I_.
i i U"
."m.m 2 J m "1 ::
m c d}=;-
.. -f.E_' . - -
@._H. ( c. @..
HE:. JK T_i.. ._. ;, : ,
a- --
- - - r.:
sw .- ,r ~- - ::=/ :::T x - -
= :- y .- _
- n . :.
- t= -
- - .z_
mp
-6 3 =.::s=r: -
= :-
c m w.rt;s:,1,:r: :
- w ._ m.rp s _A:m, ,,.
1 ,
+t g 1(
10 10 qrd.n mtyg p: m%
A*. .-._ _
1 9j s4 ,d.
_t.
QlJ7 m .p. r,wts.
/4- . c. g --
_- --... , . - w.- + --,p.
$, .a. / m.: ,h y-
..m. _
n q t
1.w.
-.. sti.
- . . . + .
18 N, ,_.i
' ' 16 { i u
- t : p
'i
.f %
d o ,c .u- --. L 1p iIt ,i i i.l' .i 4, _u' [ ?-:4 N r r---
4.-4; j;m._l* 7f. Jm m
- 4-r* [;-- -rd--
d}-. - 1.4'. - d' m y ._-,m ~6 , + .
- m. _7 1
[3I9 V1 Q.
7 -
14-. u1-t3
-- ,m. - .v m
AF,-- _ - -..
. M. ll3- - m ..
wyo.
cwF. n- v m ,1 i
t-
-E : ;._.+
.u-u% _-_-_- _
~
. t
- q. .a (W:r-. -
gm. x 101 s
4..-
t am.-
..J.
- l. ....-[.
]_ :
. . . . ..d. . .m !. ( .;.i. j_. .. ..
= e
..t..
,; .1h-.__ a : 3 m ...
a . _
I u. . _; . -
w
.m .-
dg ::1 1 1.- y=w
.n n c . b'-8 q.2 m:r q :mi :..F:3.::h~ ag,edc... :-M.. . ::r .
-7 2 =wl=. 1 3 : - x 3 10 10
-2
-1 10 0 "- & ' . - . 101 '<'#. '
- " 10 2
10 10 10 .s
. c. .
N3 SCALED DISTANCE Z R/W ,
mlkg I.
T: pre .- 5,
' ! b.k On B l a s t Par.3 eters for TNT k,
f