ML19289F551
| ML19289F551 | |
| Person / Time | |
|---|---|
| Site: | Atlantic Nuclear Power Plant  |
| Issue date: | 06/06/1979 |
| From: | NORTHERN INDIANA PUBLIC SERVICE CO. |
| To: | |
| Shared Package | |
| ML19289F548 | List: |
| References | |
| 36A59, NUDOCS 7906110173 | |
| Download: ML19289F551 (35) | |
Text
4 INSTRUCTIONS EOR ENTERING REVISICN 1 IN OPS 'IOPICAL REPORP 36A59
~
1.
Remove and insert pages in accordance with the following tabulation:
Removgage(s)
Insert Page(s)
Table of Contents (i)
Table of Contents (i)
II-1, II-2 II-1, II-2 IIT-1, III-2 III-1, III-2 III-5, III-6 III-5, III-6 III-ll, III-12 III-ll thru III-14 IV-7 thru IV-ll IV-7 thru IV-16 V-5 thru V-8 V-5 thru V-8 C-1 C-1 Appendix C - Table 3 Appendix C - Table 4 F-1 thru F-4 Table D-1 (Foldout)
Table D-1 (Foldout) 2231 009 7906339,73
,q
9 e
TABLE OF CONTENTS I.
IN1PODUCTION II.
DESIGN CONSTRAINTS AND FUNCTIONAL DESIGN REQUIREMENTS III. DESIGN DESCRIPTION IV.
DESIGN E'IAW ATIONS AND ANALYSIS V.
INFORMATION NEEDED FOR FINAL CONFIR% TION OF IADLE DESIGN VI.
ASSOCIATED SITE CRITERIA VII.
IMPACT OF EADLE DESIGN ON DOSE CONSEQUENCES VIA LIQUID PATIMAYS VIII. REFERENCES APPENDIX A CORE MELT PENETRATION APPENDIX B RADIATICN ANALYSIS APPENDIX r TECHNICAL DATA AND StirPLEMENTAL INFOR% TION APPENDIX D RELATED INDUSTRIAL EX?ERIENCE APPENDIX E COST ESTIMATE APPEh :IX F ADDITIONAL INFOR% TION REQUESTED BY NRC DURING MAY 7-8, 1979 MEETINGS 2231 010 Rev. 1
_i_
II.
DESIGN CONS'"RAINIS AND FUNCTIONAL DESIGN REQUIREMENTS FES-III recomerds issuance of a manufacturing license for manufacture of eight floating nuclear plants subject to certain conditions for protection of the environment.
As discussed in Section I, one of these corditions requires replacement of the concrete pad beneath the reactor vessel with a pad construc ted of magnesium oxide or other equivalent refractory material.
Since the replecement pad of refractory material is similar to refractory ladles employed 10 metal refining for handling high temperature molten materials, the replacement pad will be referred to as a ladle.
II.A.
General Requirements The general requirements and functional criteria set forth for the ladle in Reference 1 are:
1.
It shall provide increased resistance to core debris melt-through.
2.
It shall not react with core debris to form large volumes of gases.
3.
It r, hall be at least as thick as the current concrete pad (4 ')
and as thick as practicable.
2231 011 4.
It shall not compromise safety.
.I II-l
II.D.
Functional Requirements O
The specific functional requirements for the ladle derived frcm the general requirements and criteria listed in FES-III arrl from other FNP design constraints are as follows:
1.
We ladle design maximizes molten core debris melt-through time within the constraints of the space available beneath the reactor vessel. Specifically:
a.
Minimum delay time shall be about 2 days.
b.
Minimum ladle thickness of refractory material shall be 4 feet.
2.
We refractory materials employed in the ladle shall not react with molten core debris to form a large volme of gas which will sparge through the molten debris.
3.
The ladle shall be designed to contain the volme of molten debris resulting from melting of all the ccre materials and up to 25% of the steel contained in both the lower core support structure and bottom head of the vessel.
4.
Incorporation of the ladle within the lower reactor cavity shall not compromise the structural integrity of the containment boundary or the structural integrity of the platform.
2231 012 Rev. 1 II-2
III.
DESIGN DESCRIETION III.A.
General This section of the report addresses the materials considered and the design and performance criteria regarding the structural aspects of th a lower reactor cavity, structural boundary constraints, basic design and configuration of a refractory material ladle underneath the reactor vessel, and venting of moisture and gasses around the ladle. The discussion presented herein establishes that incorporating a ladle into the Floating Nuclear Plant to meet the requirements in Part III of the Final Environ-mental Statement (FES-III) is feasible without compromising ott.cr safety requirements.
III.B.
Design Reauirements The basic design requirements for the reactor cavity structure and the ladle of refractory material are as follows:
1.
Incorporate a ladle into the existing FNP design with minimum alter-ations.
2.
Alterations to the reactor cavity shall not compromise safety require-ments including:
2231 013 9
III-l U
a.
Structural integrity of the platform shall be maintained for all operatire aM design basis conditions prior to a postulated core melt accident.
b.
Water-tight redundancy shall be maintained between the basin and reactor cavity.
c.
Radiation shielding requirements shall be maintained.
3.
The platform structure shall withstand loading conditions for the duration of core-melt debris retention (see Section III.K).
4.
The reactor cavity structure shall not becme the weakest link of the containment pressure boundary as a result of the addition of the ladle.
5.
The ladle configuration aM material shall not comprmise other safety requirements.
6.
The ladle shall be as thick as practicable within the various design constraints but shall not be less than four feet in any direction.
7.
The ladle pool volme shall be sufficient to contain the molten core debris (see Section IV.F) during continuors basis motions (1/2 ).
8.
The ladle shall be desy.m' W analyzed to remain functional for operatirg basis err /ironme.
conditions. For more severe conditions, the plant can be shutdown for inspection of the ladle.
2231 014 Rev. 1 III-2
Figure III-4 (frm reference 3) is a camparison of the relative radiation resistance of several inorganic insulators to permanent damage based upon changes in physical properties, e.g.,
tensile strength and elongation.
Magnesium oxide is included among these materials. The maximum calculamed fluence of neutrons (E>0.1 MEV) incident on the magnesium oxide floor is 16 2
3x10 n/cm over 40 years of operation. 'Ihe maximum calculated gamma dose 7
is 2 x 10 rads over 40 years of operation. These Jevels of irradiation are lower than those at the low end of the scale i" Figure III-4. One would therefore not expect radiation effects of any consequence in the magnesium oxide floor.
Table III-l PARAMETERS OF CANDIDATE LADLE MATERIALS ( *
- Melting Specific Heat of g
Pgint Densigy Heat Fusion VHAC g
3 Material
( C)
(g/cm )_
(cal /gC)
(cal /g)
(cal /cm )
Aluminum Oxide 2037 4.0 0.272 256 2929 Graphite 2760 1.9 Magnesium Oxide 2852 3.5 0.313 428 4168 Silica 1728 2.32 Thorium Oxide 2800 9.95 0.07 72 2448 Titanium Carbide 3076 4.8 0.207 283 3976 Uranium Oxide 2815 11.0 0.07 71 2675 Zirconium Oxide 2760 5.7 0.155 169 3416 (1) Volumetric Heat Absorption Capability - estimated by inverting Glueklar's " Melt Velocity per Unit Heat Flux" value; represents the relative heat absorbing capability of the material (sensible heat plus latent heat) ya 2231 015 III-5
III.E.
L 411e Configuration and Arrangement O
The configuration of the ladle has been selected to achieve an efficient utilization of the available space by makir g the ladle as thick as prac-ticable within the various design constraints, but not less 9.'n 4 feet in direction, with a pool volune sufficient to ccminodate molten core ear debris of r? proximately 920 cubic feet. The cylindrical arch has Lean selected as the appropriate configuration, within the boundary canstraints, to attain an adequate ladle thickness and capacity in addition to the following: accommodate therrul expansion, provide self support to withstand environmental loads, allow a wedging action to prevent float up of bricks in the molten material, and utilize past experience in the steel-making industry.
The ladle design in the Floating Nuclear Plant is similar to the design of crucibles and hearths used in the large furnaces of the steel industry for the past century to contain molten iron and steel for long periods of time during the refining processes.
For further discussions on steel making and canparisons of the ENP ladle with those of the steel industry, see Appendix D.
Harbison-Walker Refractories, manufacturea of high purity Mg0 brick, highly recomend the cylindrical arch configuration and give the assurance that a bricking pattern can be developed w meet the requirements of the intended application.
Their engineering function has considerable exper-tise in the detail design of bricking arrangements, sizes, and patterns required for lining furnaces in the steel and glass making industries.
2231 Ul6 Rev. 1 III-6
IiI.I.
Platform Structural Changes Some changes of the platform structure within the containment were neces-sary to obtain the required ladle thickness and capacity. Figure III-10 depicts the previous structurcl arrangement, whereas Figure III-ll shows the structures af ter the alterations. The structures that underwent significant changes are the vertical bulkhead frames, or columns 3a-1/2, 3b and 3b-1/2, which serve as major stiffening for bulkhead G and partially support the primary shielding. Aese structures now lie on the port-side of Bulkhead G to preclude contact with the molten core debris.
The reactor cavity bottom has been lowered frcrn elevation 72'-9" to 65'-0" to create as much ladle capacity and thickness as possible. This change resulted in decreasing the depth of transverse girders 3a-1/2, 3b and 3b-1/2 within the containment from 16'-9" to 9'-0" and increasing the web thickness. Longitu-dinal girder 74, which provides partial support for the primary shield, has also been changed as can be seen by comparing Figures III-10 and III-ll.
'Ihese structures have been redesigned to support the reactor vessel and withstand the other specified loads. There is, therefore, no compromise in structural adequacy and integrity due to the structural alterations necessitated by the ladle design.
III.J.
Platform Motion Considerations Plant motion will be a etnbination of motion in six degrecs of freedom i.e.
surge, sway, yaw, roll, heave and pitch. Plant m) tion could induce movement of the free surface of the molten core debris in the pool aid possibly J
2231 017 III-ll
result in spillage. The capacity of the lalle is sufficient to accomodate the postulated core-melt volune and avoid spillat 2 due to a static list of 1/2 degree or dynamic behavior of the free surface due to continuous basis motion conditions. The plant motion limits are discussed in Section 3.7 of the Plant Design Report.
III.K.
Ioading Conditions for Reactor Cavity Steel Structure and Core Ladle III.K.1 Reactor Cavity Structure a.
Normal Operation aM Design Basis Events The reactor cavity structure forming the containment pressure boundary has been designed for the loa 3ing conditions and stress criteria of Section 3.8.2 of the Plant Design Report (PDR). Additionally, primary strength members for the platform have been designed for the loading conditions and stress criteria of Section 3.12.2 of the PDR. Incorpor-ation of the core ladle within the lower rea'
- cavity shall not compromise the structural integrity of the con ainment pressure boundary or platform for operating and design basis conditions included in the above referenced sections of the PDR.
b.
Subsequent to Postulated Reactor Vessel Melt-through The reactor cavity structure shall be designed to withstand, for the duration of core debris retention, the loading conditions specified I
2231 018 Rev. 1 III-12
below tich result from a core melt accident in combination with the loMires fran continuous basis environmental conditions. As defined in the PDR (Appendix 3C) continuous basis conditions are those which are not exceeded 90% of the time.
The loading coMitions resulting from reactor vessel melt-through are:
1.
Impact LoMs fran falling debris 2.
Dead loads fran weight of debris 3.
Thermal loa 3s resulting fran increased temperatures Following reactor vessel melt-through and during the debris retention period, the reactor vessel cavity shall remain intact to the extent required to cupport the core ladle.
III.K.2 Core Ladle a.
Normal Operation and Design Basis Events o
The ladle shal. be designed to remain functional (i.e. remain intact) for all nocmal and operating basis loading conditions during plant operation aM shutdo,a, inclMing transient lcais during tow. Design for the horizontal compnents of the operating basis earthq, ake will be based on static analyses of the core ladle an3 its support. Design for the vertical component will be based on a dynamic analysis of the core ladle and the supportiry girders. The analyses will represent the magnesium oxide bricks as a. homogeneous material. Horizontal forces 2231 019 Rev. 1 III-13
will be carred to the bottom of the ladle shell by friction forces be'xeen the bricks aM by the steel shell encasing the bricks. The l
}
ladle shell will be anchored to the reactor cavity floor.
The ladle shall also bi evaluated for design basis environmental loals in order to ensure *. hat there will be no gross failure which might impair the function of safety class components.
b.
Subsequent to Postulated Reactor Vessel Melt-through The ladle shall be designed to remain functional, for the duration of core debris retention, for loading conditions which result from a core melt accident in combination with the loadirgs from continuous basis environmental conditions. The loading coMitions resulting from the core melt event are: thermal aM mechanical shock fran falling debris, alditional weight of molten debris on ladle bed and thermal loads from molten debris.
2231 020 0
Rev. 1 I " -14
ve tically into the MgO bed.
These values correspond approximately to decay heat fractions of 50% and 80% being directed into the entire Mgo bed.
In cmparing Figures IV-4 and IV-5 it is clear that higher rates of heat input into the Mgo bed result in larger temperature gradients and more rapid advance of the melt front into the bed.
The basaltic concrete beneath the MgO bed begins to outgas at a temperature of approximately 90 F (evaporation of H O) and is essentially completely outgassed at 2
temperatures between 1020 F and 1470 F. 'Iherefore, the concrete will begin to outgas prior to the arrival of the melt front, the time interval between the start of outgassing and arrival of the melt front being greater for lower heat inputs.
For a heat input of 25% in the vertical direction the first inch of basaltic concrete is nearly completely outgassed (H O and all 2
non-condensable gases removed) 5 days after the start of the Mgo melt interaction while the melt front is still approximately 38" from the concrete.
For the higher heat input of 40% the first inch of concrete is outgassed after 3 days while the melt front is 29" from the Mgo-concrete interface.
When the steel box enclosing the Mgo bed melts, as much as two thirds of the concrete may be completely outgassed.
Thus, the major portion of the gas will not bubble through the melt but will be channeled along the interface between the basaltic concrete floor and the steel tank enclosing the Mgo bed and then up past tne sides of the tank in the space allowed for thermal expansion of the MgO bed.
Section III.H of this report describes the pathways provided to permit the gases to vent directly into the containment.
2231 021
. s IV-7
3 There is approximately 1900 ft of basaltic concrete surrounding the M30 bed, which if completely outgassed would add approximately 6.6 psi to the containment pressure.
Thermal radiation to the sidewalls can result in outgassire of the basaltic concrete to a depth of two feet, which is 3
equivalent to approximately 2930 feet. This would add an additional 9.8 psi to the contairment pressure for a total of 16.4 psi. This compares to a containment pressure increase of approximately 38.5 psi if it is assumed that the entire four foot thick mat of limestone concrete of the previous design were to react with the molten pool and approximately 15" of the limestone concrete sidewalls are outgassed due to thermal radiation from the pool. Significantly, in the former design, the gas released from the concrete would bubble through the molten pool, entraining fission products; whereas in the present design, most of the gas released from the basaltic concrete will escape directly into the containment without bubbling through the molten pool.
The first layer of brick in the core ladle consists of TOPEX S which can release both CO and SO when exposed to an oxidizing atmosphere. Assuming 2
3 a totally oxidizing atmosphere at standard temprature and pressure, approximately 1.73 liters arxl.34 liters of CO and S0, respectively, per 2
3 100 grams of IOPEX S could be evolved. It is highly unlikely that this a:rount of gas would be evolved since the brick will most probably be under reducing conditions. If all of the carbon and sulfur in this first layer of TOPEX S wer e released into the containment as CO and SO3 gases, the 2
containment pressure would increase by approximately 1 psi.
2231 022 g
Rev. 1 IV-8
IV.D.
Effects of Thermal Radiation on the Side Walls Since thermal radiation from the molten pool may result in a substantial heat load on the concrete walls above the pool surface, an outer layer of Mgo brick will be used to protect the concrete. An estimate of the temper-ature distribution in the side walls was made utilizing a one-dimensional model of the wall which consisted of 4" of Mgo, 36" of basaltic concrete and 1-1/8" stael bulkhead (see Figure III-3). A series of calculations was made for different pool surface temperatures because of uncertainties in the fraction of decay heat which leaves the surface of the pool as thermal radiation. We initial surface temperature of the pool was estimated by assuning that various fractions of the decay heat are directed upward through the surface of the pool.
We resulting heat fluxes wre then related to equivalent black body temperatures.
As the pool grows and the decay heat decreases it is assumed that the surface temperature of the pool decreases.
Initial pool temperatures ranging from 1840 F to 3540 F were utilized which correspond to assumed available decay heat fractions of 10%
to 100%.
Figure IV-6 shows the maximun temperature of the Mgo and basaltic concrete as a function of the initial surface temperature of the molten pool. For an initial ;nol surface temperature of 3540 F the 330 brick protecting the concrete wall will not melt within the five days following accident initiation, whereas 2 to 3 inches of concrete directly behind the MgO brick may n..t under these conditions. In the same time period and for initial surface temperatures below 3290 F, no concrete will melt with a 4" pro-tective layer of_ Mgo.,This latter case corresponds to approximately 70% of 2231 023 o
Rev. 1 IV-9
the decay heat initially leaving the pool surface. The structural integrity of the steel bulkhead behind the Mgo and concrete walls is not compromised for any of the conditions assumed. After approximately five days the steel bulkhea3 in this region would not be heated above approximately 260 F.
These results indicate that a layer of B40 brick can be used to protect the concrete ::idewalls ard the steel bulkheads frcm the effect of thermal radiation from the molten pol.
IV.E.
Slag Line Attack A ptential @enomenon associated with the application of a magnesium oxide ladle for delay of core-melt debris is the potential for slag line attack, which is preferential attack of by0 by an iron oxide slag layer on top of molten iron. Preferential attack was observed in a Sandia test in which molten iron was pured into a magnesium oxide crucible, reference (5). In the test, inductive heatirg of the melt was maintained for a short time due to failure of the heating coil.
Hence, the molten iron cooled in the crucible shortly af ter pouring.
More recently, Aerospace Corporation personnel have reviewed the slag line attack pnenomenon (reference 6). Frcm a review of the phase diagram for t4go and Fe O, they observed that at the temperatures of the Sandia tests 23 (1700 C), the solubility of 540 in molten iron oxidt=., 2 ut 10%. They concluded that the molten mide layer had rapidly saturated with by0, and that further preferential attack of the F40 was unlikely. They also noted, however, that at higher tempratures the solubility of Mg0 in Fe O 23 22M 024 Rev. 1 IV-10
increases and for example is about 26 wt-% at 2000 C.
Figure IV-7 is the Fe O -MgO phase diagram.
23 Slag line attack has been observed in steelmaking furnaces. In these furnaces an oxide layer is formed on top of the molten iron. In m e furnaces like the blast furnace, the slag layer is about one-third of the total height (or approxiimately several feet) of the molten material height and increased erosion by 25% at the slag-line has been observed. Discus-sions with U.S. Steel have indicated that a 24 to 27 inch tq0 lining in a basic oxygen furnace would last about 3000 heats (at 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> each) or 3000 hours0.0347 days <br />0.833 hours <br />0.00496 weeks <br />0.00114 months <br /> of operation before slag line erosion would have to be repaired.
Furnace temperatures are at the lower end of those on the Fe O -Mgo @ase 23 diagram (1600 - 1800 C) so that solubility of It0 in Fe O w uld be 23 limited. A further consideration is the miscibility of the three principle oxides formed in the core melt-through process, i.e.,
Fe O, Mgo d ZrO
- 23 2
The phase diagram for such a three component oxide mixture interacting with MgO could well be quite different than that for the simpler Fe O -@O 23 system. In tddition, the oxide layer for a core-melt accident is probably heavier than the molten iron phase (unless most of the iron is oxidized) so that the oxide layer is likely to be beneath the molten iron after the debris enters the ladle. Additional investigation may be required to determine the exact rate of slag line attack; however, experience indicates that the core ladle can be proportioned such that the required holdup time will be achieved.
2231 025
-g Rev. 1 IV-ll
IV.F.
Estimates of Core Debris Volume I
I Following a core meltdown, a portion of the steel surrounding the core will also melt, either due to intense thermal radiation (as in the case of the lower core barrel), or as a result of coming directly in contact with the molten core (as is the case with sections of the lower head). Judging from the total amount of steel available and the relat.ively short times cal-culated for vessel melt-through, it is estimated that no more than 25% of the available steel outside the core will melt. Utilizing this conservative estimate, the total weight and volune of the debris was calculated, and the results are summarized in Table IV-1.
Similar calculations were also preformed assuning that only 10% of the ex-core steel will melt. These calculations are summarized in Table IV-2.
Both of the above mentioned tables ignore oxidation of steel and zircalloy.
l l
It is expected that most of the zircalloy will be oxidized in the early stages of the accident by reacting with water or low quality steam in the vessel. Only a very small portion of the steel is expected to be oxidized in the vessel, because most of the steel which is expected to melt will do so after the vessel has been voided of water and steam. Once the debris exits the vessel, further oxidation by oxygen in the atmosphere will be less than 1% of the steel. Overall it is considered that oxidation in the melt will not exceed 25% of the available steel and 100% of the zircalloy.
A number of more probable cases were also investigated in order to de-termine the effects of oxidation on the weight and volume of the debris. A summary of these calculations is presented in Table IV-3. As can be seen there, the volume of the worst case assumed is less than the core ladle t
i 2231 026
,,,1 IV-12
voltme of 920 cubic feet. '1he voltmes in Table IV-III are conservatively large because a high iron oxide content is assumed. Available data on the stability of iron oxides indicate that the oxides decompose to the more dense elemental iron and oxygen in the temperature range 2300 C to 2400 C.1 1Sandia Laboratories, Report 74-0332, Core-Meltdown Experimental Review, Augsut, 1975, p. 4-95.
2231 027 Rev. 1 IV-13
TABLE IV-1 FNP CORE MELT CONSTITUENTS MATERIALS SOURCE WEIGHT (LB)
EcNSITY(LB/FT )(#} VOLUME (FT )
00 Fuel (100%)
222,739 550 405 2
Zircalloy-4 Cladding (100%,
50,913 375 136 Guide and Instrtrnen-tation Thimbles (100%)
Steel Lowcr Reactor Vessel Head (25%)I 22,500 400 56 Lower Internals (25%) (c) 60,000 400 150 Core (100%)(d) 11,500 400 29 I
Silver *I Control Rods 5,000 540 9
TOTAL 372,652 785 (a) Valtes represent best estimate for these materials in the 2500 C to 3000 C range (4532 F-5432 F)
(b) Mostly Carbon Steel (with thin layer of S.S. cladding)
(c) S.E. 304 (d) Mostly S.S. 304 with small quantities of Inconel-718 (e) Inditzn ard Carinitrn will also be present but in relatively negligible quantities
- Lower Reactor Vessel Head is assumed here to inc1tde 1/2 of the transition ring 2231 028 Rev. 1 IV-14
T_ABLE IV-2 FNP CORE MELT CONSTI'IUENTS MATERIALS SOURCE WEIGHT (LB)
DENSITY (LB/FT ) (a)
VOLUME (FT )
U0 Fuel (100%)
222,739 550 405 2
Zircalloy-4 Cladding (100%,
50,913 375 136 Guide and Instrtmen-tation Thimbles (100%)
Steel Lower aaactor Vessel I '*
Head (10%)
9,000 400 23 Iower Internals (10%)I#I 24,000 400 60 Core (100%)(d) 11,500 400 29 I
Silver *I Control Rods 5,000 540 9
'1 DIAL 323,152 662 (a) Values represent best estimate for these materials in the 2500 C to 3000 C range (4532 F-5432 F)
(b) Mostly Carbon Steel (with thin layer of S.S. cladding)
(c) S.S. 304 (d) Mostly S.S. 304 sith small quantities of Inconel-718 (e) Indium ami Cadmium will also be present but in relatiiely negligible quantities
- Lower Reactor Vessel Head is assumed here to include 1/2 of the transition ring 2231 029 se, 1 IV-15
TABLE IV-3 CORE DEBRIS VOLUMES FOR VARIOUS PERCENTAGES OF STEEL AND ZIRCALIDY OXIDATION
% Steel
% Zircalloy Oxidation Oxidation Case A(a)
Case B(b) 0 0
785 662 0
75 848 725 0
100 869 746 10 75 865 732 10 100 886 753 25 75 890 744 25 100 911 765 3
Note: ThedgnsitiesofZrO2 xY and Fe O were taken to be 313 lb/ft and 300 lb/ft, respectively (a) Case A corre3 ponds to Table II-1, i.e., it is assumed that 25% of the steel in the lower vessel head and the lower intervals is included in the debris.
(b) Case B corresponds to Table II-2, i.e., it is assumed that 10% of the steel in the lower vessel head and the lower intervals is included in the debris.
2231 030 Ret. 1 IV-16
c.
Slag Line Attack i.
See Section IV.E for description and state of knowledge on this subject.
ii.
Importance to Ladle If a substantial layer of iron oxide (slag) were to form on top of the molten debris in the ladle (which seems unlikely) and if extensive attack at the slag line were to occur, preferential lateral dissolution of the ladle at the slag line could occur thereby reducing melt-through delay times.
iii.
Information Required Available information indicates that slag line attack rates are low in steel making furnaces. Additional confirmatory testing with the mixed oxides present in Corium may be required.
2.
Refractory Bed Behavior a.
Floatup i.
Description
'Ihe refractory ladle described in Section III is constructed of fixed Mgo bricks with a density less than half of that of 2231 031 Rev. 1 V-5
molten debris. If the nolten debris penetrates around the bricks arri if the bricks are not restrained, they will float on the denser molten debris.
ii.
Importance to Ladle Extensive floatup could lead to relatively rapid penetration of the refractory ladle thereby shortening the melt delay time.
iii. State of Knowledge For the refractory beds employed in furnaces and ladles in the metals refining industries, floatup has rarely been experienced. Like Corium, steel melts are also more dense than the MgO refractories used to contain them.
In addition, floatup has not been observed in furnaces used for refining nickel melts whose densities are as high as those of Corium.
In the metals refining industries, the furnace and ladle beds are usually constructed in an inverted arch shape so that the arch wedging action reduces floatup tendency.
iv.
Additional Information Required The FNP refractory ladle will be constructed in an inverted arch configuration like that employed for refractory furnace and ladle beds in the metals refining indastry. In addition, a tongue and groove brick configuration is planned to further restrict the tendency for floatup.
Based on experience in the metal refining industries and the planned tongue and 2231 032 V-6
goove construction, this area is considered to be adequately treated and no testing is required.
b.
'Ihermal Expansion and Crack Penetration i.
Description As it30 is heated it expands.
A crack may be left between bricks to accomodate brick expansion during heatup (see Section IV).
A small gap may continue to exist between some bricks as a result of less than perfect construction or as a result of non-uniform temperatures across the bed.
Thus penetration of cracks between bricks by i~1 ten debris nuy occur.
ii.
Importance to Ladle Extensive crack penetration could lead to brick floatup and enhaiced bed penetration.
iii. State of Knowledge Freezing of the debris melt as it flows into the narrow cracks betwecn bricks is likely to limit crack penetration.
Extrrience from the metals refining industry indicates crack penetration is not a problem although the range of temper-atures associated with such application is not as great as may be experienced for a molten core debris ladle. Calcula-tions indicate that the maximum penetration of the core ladle would be about one layer of brick.
2231 033 Rev. 1 V-7
iv.
Information Required The extent of crack penetration for a range of crack sizes and temperature needs to be determined so that acceptable brick spacing can be identified, c.
Thermal Shock i.
Description When molten materials at high temperature are poured on refractories at room temperature, large taiperatures gradi-ents and thermal stresses are inJuced.
The thermal stresses can lead to cracking and/or spalling.
ii.
Importance to Ladle Fracturing of the brick as a result of thermal shock can reduce the penetration time of the refractory ladle.
iii. State of Knowledge In the metals refining industry, melts at high temperature are poured into cold ladles.
Extensive fracturing of the refractory lining of the ladles does not occur. However, materials less susceptible to thermal shock than Harklase brick are usually employed for such applications.
iv.
Information Required An upper layer of chemically bonded MgO brick which is less susceptible to thermal shock than burned Harklase brick will 2231 034
~
V-8
APPENDIX C TECHNICAL DATA AND SUPPLEMENTAL INFORMATION TABLE C-1 Data on TOPEX S, Chemically-Bounded Magnesite Brick TABLE C-2 Data on HARKLASEr Burned Magnesite Brick TABLE C-3 Data on HARMIX FE Brick TABLE C-4 Data on OXIBOND Mortar FIGURE C-1 Linear Expansion vs. Temperature for HARIGASE Brick FIGURE C-2 Thermal Conductivity vs. Temperature for HARKLASE Brick FIGURE C-3 Expansion Under Load and 50 Hour Creep for HARKLASE Under 25 psi ATTACHMENT Memo From Harbison-Walker Refractories 2231 035 Rev. 1 C-1
q Yg - }'X:1 HARBISON-WALKER REFRACT 0 RIES j y ':j'- lQ Y-K.2 Dresser Industries. Inc.
f::.
s Y-2 Gateway Center. Pittsburgh Pennsylvania 15222 APPENDIX C - TABLE 3 HARMIX FE Techni.~ l Data:
Physical Properties:
(Typical)
English Units SI Units Maximum Service Temperature 4,000 F 2,204 C 3
3 lb/ft kg/m Weight Required For Ramming 164 2,620 Bulk Density After Drying at 230 F (110 C) 164 2,620 2
Modulus of Rupture lb/in kPa 0
After Heating at 3,000 F (1649 C) 800 to 1,000 5,500 to 6,900 Cold Crushing Strength 0
After Heating at 3,000 F (1649 C) 4,000 to 6,000 27,600 00 41,400 Permanent Linear Change, %
After Heating at 3,000 F (1649 C)
-0.4 to -0.6 Chemical Analysis:
(Approximate)
(Calcined Basis)
Silica (SiO2) 0.8%
Alumina (A1 03) 0.2 2
Iron Oxide (Fe 02 3) 0.2 Lime (Ca0) 0.5 Magnesia (Mg0) 98.3 All data subject to reasonable deviations and therefore should not be used for specification purposes.
ASTM Test Methods, where applicable, used for determination of data.
2231 036 Revision 1 technical data GI-2 (Continued)
l]
13 r
~
. i l ' IM:;,aD k j j jg
.J J
~-
_ _ m _-. m _.)
O HARMIX FE (Cont'd)
==
Description:==
A magnesite ramming mixture of exceptional purity and stability.
Features:
Closely controlled grain sizing, optimum for good ramming, and dense, stable structure.
Highly resistant to chemical attack by ferrous metals and oxides.
Uses:
Designed primarily as a lining material for coreless-type induction furnaces melting ferrous alloys.
Shipping Data:
Shipped in multi-wall moisture-proof sacks of 100 pounds (45.36 kg.) net weight.
O 2231 037 O
3
?p:pq HARBISON-WALKER REFRACTORIES I
11:/w &!f Dresser Industries. Inc.
- .~~,.-,. lf"
. ~ :.-
2 Gateway Center. Pittsburgh, Pennsylvania 15222 AP"cSIDIX C - TABLE 4 OXIBOND
==
Description:==
Extremely refractory, dry, heat-setting magnesite mortar.
Uses:
Used dry for leveling and grouting Basic 0xygen Furnace working linings.
Mixed with water, OXIBOND is ideally suited for laying H-W MAGNESITE, H-W PERIKLASE, and other types of basic brick in applications where a heat-setting basic mortar is required.
Technical Data:
Physical Properties (Typical)
Approximate pounds required per 1000 9" Equivalent. Dry for B0F linings -
80 to 85 If used vet, approximate amount of water required for trovelling consistency (per 100# mortar) 3-3/h to h-1/4 U.S. Gal.
Approximate Pounds Wet Mortar required per 1000 9" Equivalent:
Brick Laid Dry then Grouted 300 to hoo Brick Laid using Thinly Trevelled Joints 500 to 600 Refractoriness Test: Mortar does not melt or flow out of joints when heated for 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> at 2910CF.
Chemical Analysis:
Silica (SiO )
h.2%
2 (Typical)
(Calcined Basis)
Alumina (A1 0 )
0.3 23 Iron Oxide (Fe2 3) 0.3 0
Lime (Ca0) 0.9 Maenesia (Me0) 94.3 NOTE: All data subject to reasonable deviation.
Shitting Data:
Shipped dry in 100# moisture-proof sacks.
ASTM Test Methods, where applicable, used for determination of data.
FG-5
- TOC l1I1lCQ! d QIQ
APPENDIX F ADDITIONAL INFORMATION REQUESTED BY NBC DURING MAY 7-8, 1979 MEETINGS 1.
Provide a3ditional information concerning. the long-term effects on 330 of moisture absorption caused by high atmospheric huidity.
Resoonse:
Based on experience with installation of basic brick (HARKLASE and
'IOPEX S), hydration effects have not been observed even in hunid climates for onsite storage periods up to a few months prior to installation. However, the bricks are normally stored under cover to prevent their direct exposure to precipitation.
It is recognized that magnesia containing brick will undergo hydration if exposed to humid atmospheres for an extended period, but specific data regarding the exposure period which leads to significant hydration are not available. ExEnrience however indicates the time period for significant hydration during continued exposure to humid at:nospheres at ambient temperatures, is of the order of a few years or longer.
2231 039 4k Rev. 1 F-1
2.
Estimate the maximun amount of gas evolution (SO ' CO ) from IDPEX S 2
2 bricks.
{
)
Response
Refer to Section IV.C.
3.
Provide a bibliography of technical references on MgO refractories.
Response
Most of the information on use of Mg0 Lcick has been published in journal references. A few good sumary sources are listed below.
a.
J. H. Chester, Refractories For Iron and Steel Making, The Metals Society, London.
b.
H. E. McGannon, ed, The Making, Shaping and Treating of Steel, U.S.
Steel Canpany (1971), (Chapter 2, Refractories).
c.
Harbison-Walker Refractories Company, Modern Refractory Practice; 3rd edition, (1961).
d.
American Ceramic Society, Phase Diagrams for Ceramists, with Sug lements.
2231 040 0
Rev. 1 F-2
We above summary references adequately treat the use of Mgo as a refractory in the metals refining industries.
4.
Provide a contact in steel-making industry who is familiar with operations involving nickel melts in contact with Mgo.
P.esponse:
te name of a contact has been provided to the NRC.
5.
Provide quantitative information on the extent of slag-line attack of Mg0 in steel-making operations.
Response
Refer to Section IV.E. Harbison-Walker was unable to provide 2dditional quantitative information on slag line attack in steel making operations beyond that provided at 'he May 7-8, 1979, meeting held in Jacksonville t
and sumnarized in the NBC Report of that meeting (NRC Report by A. R.
Marchese issued May 17, 1979). Harbison-Walker has stated since the meeting that, in their judgement, RNE brick will perfon atis-factorily at the slag-line in the core ladle.
6.
Provide a list of quantities of core melt constituents, incltding a range of possible gaantities of iron oxide that can be formed.
2231 041 Rev. 1 F-3
Response
9 Refer to Section IV.F.
7.
Identify loading coMitions and parameters for the lower reactor cavity structures aM the core ladle.
_Resoonse: Refer to Section III.K.
8.
Offshore Power Systems will prwide the following additional informa-tion during detailed design of the core ladle:
a)
Evaluation of the need for instrumentation (thermocouples, moisture monitors) in the core ladle region; sufficient justification will be provided if the instrunentation is fouM not to be necessary.
b)
As part of more detailed heat transfer calculations, evaluate the need to: (a) protect the upper reactor cavity concrete with Mgo, and (b) increase the ladle sidewall thickness of Mgo such that the melt-through delay time is extended.
c)
Provide design details of the Mgo ladle perimeter (sidewalls and floor) reg ion, including the interface between the Mg0 and the steel shell.
d)
Provide contacts in the steel-making industry regarding visiting aM observire related steel-making operations.
2231 042 g
i Rev. 1 F-4
I
,=
1 I
COMPARISON OF STEEL
- MAKING FUR IPOTO OR APPPatl'%TE LEr4GTH OF F
11PE FtnR' ACE DESCkiPil0N FUtiCTION D!tJMM
@ *./JE TOR CAPArtif TIPT M AT T
~
REf ERE!.CE LADtt) SIZE (ErdH MLLT) m wm==.w.-=
Continaously 1,000 to 3,000 5 tack type f ur-tons per day nace filled with operating Blast Furnace liwstone & tron Entracts iron Fig. D-4 sizes 20 Tt. to~
(averaq**to Con tinuotes Iajers of coke, furnace.
Varies. Furnace
^
large stre).
operation 3e r *e. Pot air &
from tron ore.
40 ft. dia. n Over 10,000 f or years,
100 f t. or mre tons per day ony9en blast (largest site atove hearth at high.
furnace) bottom.
Crucible type
% aries from 5 ton furnace partially filled with rol.
Pcfires molten to 25 ton capac-Fessen r Converter ten iron. Air irr.n into steel Fig. D-5 ity. A 25 ton 25 tons un.
si hr.
'Prieratic Furpace) blown thru rniten and/or refines furnace is 7 ft.
irc.n far dia. n 14 ft.
iron fror jets at bottoti of C0 erc ial use.
hig s appros.(insid-furnace.
+ re u)vable bottom Crucible tyre Varies from 80 furnace Iartiall/
ton to 300 ton Basic Onygen filled with r,st.
Fefines rolten furnace ten tron. 0.ymn i ron t r.to s teel Fig. D-6 capacity A E0 14 to 300 ters 45 rinutes su n s w (OneuNtic Furnace) f orced teru tron g
g,,
alt m t cog letely f ror. jet JWe 9
rs laced Lesuc er liquid retal, lined with igO.
Varies from 100 ton to 600 ton Large rectar ga-lar shallow capacity. For 1% t, 309 tons 8 hrs usirq 3
Own Hearth Furnace hearth filled Refines rollen Fig. D-2 &
300 ton furnace,
( weray) 00 osy-fuel, 10 hrs.
with rolten iron fron into steel D-3 actual 5 earth size tans en witt wt esygon 4
& scrap st(cl.
(overal O 21 ft.
lined with M 0.
wide a 70 ft.
9 long appron.
Circular enclosee varies from small 3
furnace. Faises Refines noiten terp. of rolten iron irto 'recial; of to 40.3 tons y
Electric Arc iron charge &
grade steels, tig. D 7 cap. city. A 290 Larcest elec.
2 to ) brs.
t ton fernace is arc furnc es r
furnace relts scrap steel alloy steels &
22 ft. dia. x 14 to 420 tons witn elec. arc stainless steels 20 ft. high using 3 approx.
elec trodes
'for 25 tons (liquid volu"e) 5 f t. dia. m Sane capacitf fbiten Fetal in Circular steel Conveys rolten 5'-3" deep.
as furr. aces.
Lajle furnace shell lired with retal to & from flg. D-3 For 500 tons Fror 25 tons to 15 mir, to I hr.
Ladles (large) silica & alumina furnaces.
15 ft. ob
=
600 tons re fractory. 2-ger 15 f t. deep furnace.
approa.
Large rectan-j galar snell with I
(c n
)
. Delay of Fig. !!!-9 26 * -6" x 36 * - 3" Up to 200 tons Appros. 2 days I "
f flip Ladle Core-nelt debris rectangle filled with fMO brick 3 covered I
with a steel liner plate.
( I ) The type of refractory used depends largely on the chaistry & terperature of the noiten metal Generally only the tginer occasionally needs to be repaired or replaced.
(2)
The " permanent I tning neaTto the steel shell, ',eldom (l' ever) needs to be repaired.
2231 043 T
i y
1 ABLE D-1 CE CONDITIONS WITH FNP LADLE CONDITIONS 1
ArrR01. HEIGHT OF APPROI. FE IGHT OF TYPE & APIROX.
FREE FALL 0F COLD
%LTE4 ?TTAL ENFECTED LIFE OF
'T PEFRACTORY THICCJ55 0F STEEL SCLAP ONTO POUR FPON LADLE TEEP *'AL 5 HOCK Elf 6ACTWY tIhl%G TE R AT URE SURFACE SEFFACICRY W.0 L IMING DURING C'.:3 F J. ACE FLnGR 14 COVACT w!TH TLHPERATt'PI tlNING (1)
L6A0 LNG 0FFURNACE OR FLP'JCL TO LADLE pol 1EN ME T At.(2)
Warious silica Fbiten iron p:u ed
%sh" ( flene area) 15 r
& alumina re-teto ladles 13 ft. to repaired every 3 to 5 s500 F at Top fractories 20 ft. free fall, ye a rs. The hearth 0
3F throughout.
f.A after ficwin; teru Cold ladies receive lining is re;, aired 0
3500 F at Thickness varies sloping t'rics tr Nghs 2000GF tron from 10 to 20 yrs.
0 air jets Some furnace 20 f t. tr. 83 f t.
hearths are long carbon.
Silica lining.
Fron iron ladle to jPermatele bottom lining, itA s 3f000F 16" to 24" at s10' furnace slo f.
12C30F t ron.
,25 heats.
Lininq frun fu nace to steel
' teel poured out of
- at st& s. up to.0u0 bottom.
r ladle up to 23 ft.
furrace at s30%CF heats.
trio cold ledics.
kj0 lining throughout.
25 f t. to 30 f t. free fall. Lengtm of slicei Cre-irco ladle to Fu e -e receives AA s3000 F 3'n' eaa.
or turtile f rom tea of { f erace 125 f t.'
1 2EUfF iron.
0 Mt om nests from f arrue to steel l Stet 1 need out of 3t botton &
raised charg1'i<1 i
l')wer sides hopper to turnace
' ladle 25 ft. to
- furnace at s300FF
! 73 ft.
into cold ladles.
floor approx. 50 ft.
P 0 lining.
9 s3000 F in From tron ladla into Furore receives 0
- F to contact with
.3k'at stepped deflector Ct0JT 1ron. Entir-mol t en re ta l.
center of
$ ft. to 8 ft.
15 ft. to 18 ft.,
furcate tust withstand 200 to 300 heats x)* F Apprgaching hearth from furnace to steel fla e te ;s. up to
.(Gecord
- 700 to 830 450f F at flam ladle 15 Ft. to 45FF for 22 hrs.
i heats) blast 20 ft.
per day. Steel poured 0
out at s 3000 F.
P$ lining.
Jerp.
For chrore f + ar:
alloys s3500 F s3Y at from tron ladte to Furnace rec.cives
- p. can U tottom &
20 ft. to 25 ft.
furnace 120 ft..
ircn at Q6000F.
150 to 300 heats isad ruch lower sides from furnace tu steel Steel poured cut of fu nace at s30000F i
Iddle 1 20 ft.
r per into cold ladies.
Various silica Lining of steel ladies is
& alurina Cold ladles receive
' repaired every 10 to 25 re f ra c torie s.
iren at 260FF. Cold' heats to remove slag f;A Up to 3500 F 4" to 12" NA (See 8D0'e) 18dles deposits & to tre<.:nt at 3gGO[eceive steelF & alfo s at tinclusions of slag &
s 3500 F to 3600 F liner particles t'eing mined with the refined steel.
Mg0 Cold ladle receives Or.e heat for actor vessel faf ts at s 20000f.0 4,
M
'll, Corium at s3400 F appron. 2 days 0
crerature of sielt up to s4000 F following a core-melt thick accident.
" " ~ " '
2231 044 I
(