ML19256D113
| ML19256D113 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 05/09/1969 |
| From: | Dromerick A US ATOMIC ENERGY COMMISSION (AEC) |
| To: | Boyd R US ATOMIC ENERGY COMMISSION (AEC) |
| References | |
| NUDOCS 7910170568 | |
| Download: ML19256D113 (15) | |
Text
.
- "' '4 UNITED STATES S
ht ATOMIC ENERGY COMMISSION
- i
.d
- l WASHINGTON. D.C.
20S45
%,,,,,,, y#
NAY S 1569 R. S. Boyd, Assistant Director for Reactor Projects, DRL THRU:
S. Levine, Assistant Director for Reactor Technology, DRL d. 6 j
METROPOLITAN EDISON COMPANY / JERSEY CENTRAL POWER & LIGHT COMPA'IY THREE MILE ISLAND NUCLEAR STATION, UNIT NO. 2 - CONTAINMEITf &
CLASS I STRUCTURAL REVIEW We have reviewed the Preliminary Safety Analysis Report and amendments to date, and our report on this review for the above plant is enclosed.
(/,h
.{in d u A. W. Dromerick, Chief Containment & Component Technology RT-398A Branch C&CTB:DRL:EGA Division of Reactor Licensing
Enclosure:
Met. Ed./ Jersey Central PSAR review rpt.
ec w/ enclosure:
R. Tedesco R. Powell A. Gluckmann N. Davison
~
1446 162 1910170 Sh@ g
'THREE MILE ISLAND NUCLEAR STATT6N - UNIT NO. 2 1.0 Ceneral Structural Design Principal Class I and II structures will be founded on bedrock of Gettysburg shale. No liquefaction potential exists. We and our consultants find the foundation provisions acceptable.
Class I structures and components will be designed to withstand the effects of horizontal grot $nd accelerations of 0.06g (Operating Basis Earthquake) and 0.12g (Design Basis Earthquake) with sim 21taneous vertical accelerations of 2/3 of the horizontal accelerations.
These values are the same as are being used for Three Mile Island No. 1.
Tornado loading is based on the model which we have accepted on previous plants.
Combined loading conditions include the simultaneous occurrence of the maxinum earthquake and 1DCA.
A major portion of the plant, generally including the reactor, auxiliary, fuel handling and control buildings, is also hardened against direct aircraf t crashes and their side effects, such as fire from spilled airplane fuel.
Further discussion on aircraf t hardening is presented in Section 3.0.
We and our dynamic design consultants have reviewed the loading criteria presented for Class I structures and found them to be acceptable.
) I\\ h
)
. 2.0 Con t ainmen t Descrimtion The reactor centainnent building will be a prestressed concrete cylinder with a flat foundation mat and a done roof. The nat will be 10 feet thick, convenil ally reinforced, with a two foot thick, reinforced concrete slab above the bottom liner olate. The cylinder will be vertically and horizontally prestressed, with an inside diamete r o f 130 fee t, 4 foot thick wa*_'s, and a height of 157 feet from top of mat to sering line. The dome will have a 110 foot radius and be prestressed with a three way system. The liner plate material will conform to ASTM A-516, Grade 55 and will be one half inch thick in the dome, 3/8 thick at the cylindrical walls, and 1/4 inch thick
- 17. Phre b ass.
The containment is sized to provide a free volume of 2 million cubic feet.
The desien philosophy under all load combinations, except concurrent maximum temperaturs and pressure, is that the containment will be in general membrane compression and not crack, except at discontinuities.
).0, h The loading combinations have been reviewed by us and found to be acceptable.
The design of tf4 large openings (equipment hatch and personnel lock) will be based on a finite element analysis. The tendons will be draped around the openings and a thickened section will reinforce the wall at the openings. The applicant has adequately described the design parameters by which his analysis and design will be guided. Properly utilized, we feel that these can result in a satisfactory design.
Sufficient information has been presented to indicate that the criteria to be used in designing the tendon anchorage zone concrete can result in a proper design.
The liner design will incorporate sufficient conside ations for buckling through the criteria presented, and the liner anchorage system will be so designed that no chain reaction failure of anchors could take place should one anchor fail. A maximum strain limit of 0.5% has been established for the liner. We find these criteria to be satisfactory, kbh D
_ 3.0 Aircr;f t l'ardenin~
The desi:n of Three "ile #2 acainst aircraft inningement is based on the same data as was nresented in Supnlement 5 to the Unit #1 " SAD.
Cilbe rt Assoc., Inc., A/E en Unit #1,is actina as consultants for aircraft crash review and denian for Unit #2.
Burns & Doe is the design A/E for l' nit #2.
The models used in desien are all taken at 200 knots impact velocity and are listed as follows:
_ Case
'Joicht E f fe ctive Trnact Area A
6,000#
5' dian.
B 4,000#
3' diam.
C 300,000#
16' diam, 0
200,000#
14' dian.
The reactor building vill be evaluated for Cases A, B, C, and D, with impincement at the dome asex springline, and en the cylinder valls at or between buttresses. Attachments to the containment structure will also be evaluated for these effect of these cases.
The control buildinc vill be checked for impact on the roof and side walls.
Present design indicates that the control room floor will be isolated from the walls, in order to lessen the impact loading en equipment, ins trumen ta-tien mnd personnel. The tentrol room will also remain habitable durina and after a crash on it.
The fuel handling and auxiliary buildings will be desicned to prevent penetration or collapse and for protection acainst secondary effects due to fire, nissiles, etc.
)hh
e, The intermediate building enclosing the nain steam lines and feedwater pu:nos will also be evaluated so as to provide srotection below elevation 340'-0".
The loading and impact criteria, as well as notential secondarv ef fects have been reviewed by us, our dynamic design consultants, and bv our exoolsives and shock phenomna consultant. This review is a follow up on a similar one done for Unit #1 The criteria, as presented by the applicant, have been accepted by us and the above consultants. The desien implementation will be under continuing review during the forthcoming final design stage, at the acnlicant's option, and in any event, evaluation of this implementa-tion of the _citeria will be coenleted during the review for the operating license.
~7
. (3 I
) f ! ()
\\
. 4.0 Centainment Prest ressine System A major portion of the structural review has been directed toward the prestressing system. Several aspects of the systems nroposed differ fron systems in use or under construction for nuclear power olants in the i
L51ted States.
1.
The tendon capacity ranges fron 1280 to 2024 kips per tendon.
".h i s uoper range of latte-capacity tendons has been submitted for use on 3-? tile Unit #1, Russellville, and Rancho Seco, and Ft. St. Vrain.
Large capacity tendons have previously been reviewed and approval expressed for the large BBRV tendon system. The Freyssinet and SEEE 3
systems were accepted in principle on the Pessellville Nuclear Unit, bared on comittments by the apolicant to submit certifyine data and tests from the selected vendor. A similar apnroach is being taken on Rancho Seco with the VSL system. The Western Concrete system is essentially the s ame as the VSL. Therefore, on 3-?'ile Unit #2, the use of large capacity tendons is acceptable to us.
2 The tendons proposed are wire strands rather than individual wires.
Rancho Seco intends to use either strand or individual wire tendons and is ;he only other prestressed plant, to date, which is not comitted u a wire tendon, with the exception of H. B. Pobinson.
Strands have been reviewed on Dancho Seco and their use in the 3-Mile Unit #2 desien la acceptable to us.
) h h {)
[';
~
. 3.
All four anchorace systems proposed are friction o'r friction /bearinc
~
anchorages, as cpposed to the bearing anchorage installed to date (BB RV). The systens proposed are:
Stress Steel Western Concrete Manuf actute r Trevssinet (SEEE)
VSL Core Structures Desinnation Wire strand Uire strand Nire strand Wire strand Ultimate Capacity 1944 kips 2024 kips 1280 kips 2000 kins Design capacity 1166 kips 1214 kios 768 kips 1200 kips End Anchorage Ocen end, male Swance d,
Snlit eene Split cende indi-multiole strand threaded individual vidual strand fluted cone collar & rub strand grip-gripeer, bearine gripper bearing per, bearing into conical into female into cenical holes machined fluted conical holes nachin-into bearing hole ed in bearing plate plate The use of friction anchoranes brings up the relative ductility of bearing or wedge-gripped tendons. The significance of wedne-grip systens resulting in a ductility of Icss than 37., while straight bearing systens generally have greater than 4% elongation, must be viewed in relatien to the actual available ductility in the overall structure. Tests nade by Frick on curved 121 wire tendons with bearinc anchoraces showed an elencation of less than 3% while for straicht tendons re ults were greater than 4%.
Even a 1% elencation of the 3-Mile Unit #2157 foot vertical cylinder walls would result in a 1 1/2 foot increase in height, an order of magnitude larmer than
\\hh
_g.
anticipated. Based on the applicants' aporaisal of anticipated deflections under 115% proof test, as sumrarized on pace S6-C-5 of Supolement No. 6, we feel that the relatively lower dtctility available with wedr,e anchors should not preclude the use of wedee anchor systens.
In addition, if one accepts the applicants' proposal for aroutine of the tendons, then the comparative importance of the tendon anchorace system is reduced,thereby tending to allav intuitive misnivines about f riction anchorage systens. However, the anplicant is designing the anchorace systers to withstand at least the tendonh guaranteed ultimate strength - and conducting full scale nroof tests - thereby not takinc any anchorage credit for the arouted tendon.
1446 170
_9_
4.
The applicant has specified that the material for the tenden wires be either stress relieved wire or stress relieved and stabilized wire. Stabilized wire has not yet been officially incorporated into the ASTM standards.
In other respects, the wire shall conform to ASTM-A-416, Specifications for Uncoated Seven Wiere Stress Relieved Strand for Prestressed Concrete, with minimum ultimate strength of 250,000 psi and the seven wire strand shall also conform to ASTM A 416, with minimum ultimate strength of 270,000 psi.
A review Of Appencix SK to the PSAR - a discussion on relaxation losses by Shupack & Associates, and previous contact and discussion with CF&I on their LOK-Stress (stabilized) strand has led us to the conclusion that either stress relieved or stress relieved and stabilized strand can be used for the Three Mile Unit No. 2 containment structure prestres? system. Predicted losses in the 85 F range are 14% for stress relieved and 4% for stabilized. The degree of predictability and conservatism in the stabilized strand relaxation estimate is at least equivalent to that for stress relieved strand.
It is our conclusion, therefore, that, if the same design conservatism will be used for stabilized strand as for stress relieved, that the use of either strand is acceptable. Moreover, stabilized strand is being used at Ft. St. Vrain and abroad.
144b j"/t 1
. 5.
Grouted tendons are proposed. For a summary of our review of this feature, please refer to Section 5.0.
1446 172 5.0 surveillance and Testing Grouting of the stranded tendons with a cement grout has been proposed by the applicant, instead of unbonded, greased tendons.
Bonded tendons with cement grout have only been used to date on a U. S.
nuclear containment at H. B. Robinson, in conjunction with vertical prestressed steel rods. As a design feature, grouted tendons are recognized as providing certain advantages in controlling structural cracking. However, it is felt that certain other aspects demand a reliable surveillance program for the operational life Of the system.
These aspects include:
1.
Quality of field groucing of all tendons.
2.
Reliability of corrosion protection for tendons.
The applicant, recognizing a lack of existing data on grouting of large capacity tendons, has conducted a three-phase program of testing.
Phase I investigated whether normal grouting procedures could be effective, Phase II whether effective special grouting procedures were required and could be developed, and Phase III checked reproducibility of results. Vertical, horizontal and draped (around simulated opening) tendons were test grouted. One of the reasons for conducting these tests was the predominant opinion among structural engineers that successful grouting of stranded tendons would prove to be much more difficult and less reliable than for individual parallel wire tendons.
4
12 -
Test samples all showed good grouting results, in spite of difficulties encountered with the system during an observed portion of the Phase III program. Of great concern to us, however, is the fact that hundreds of tendons, certainly more than 100,000 linear feet, will have to be very effectively grouted in the field, under scheduling and adverse environmental and location conditions which can significantly alter the reliability of the test program. It is also almost impossible to actually verify that a tendon duct in the containment is properly grouted. For this reason alone, some future structural verification would be desired, since the results could affect the bonded characteristics of the tendon and the potential for corrosion, as well as reduce the level of prestress.
There have been studies made of prestress tendon failures which show that failures have occurred due to errors in placement, inadvertent inclusion of corrosive elements, and s lack of knowledge, due to the state of the art, that a failure mechanism was present. Grout mixtures available to date release either hydrogen or nitrogen gases. Embrittlement due to hydrogen is now well documented, but the long term effects of nitrogen on high tensile wires is not as well known. We are unable to derive the same assurance as the applicant that the cement grout will provide a corrosion protection environment sufficiently superior to a greased environment to warrant the loss of inspectability presently inherent in the greased system.
\\ h 6 6
- \\ ' [;
7 However, recognizing that the grouted system does have some practical engineering advantages, as well as stimulating a technical diversity, we could accept a properly designed grouted tendon system if a sophisticated surveillance program were incorporated to verify the tendon conditions for the useful life Of the structure. On this point, we have been unable to come to an agreement with the applicant.
He does not consider that periodic in-service surveillance of the reactor containment building is necessary, but has presented the following program. A test pressure of 69 psig (115% of design pressure) will be applied at 2,10 and 20 years af ter start of comercial operation of the plant. Visual inspection of cracks, and displacement measurements with dial gauges and theodolites will be used to evaluate the pressure effects. It is our belief, which is not shared by the,apolicant, that some permanently embedded instrumentation can and should be utilized to be able to detect more than a gross structural deterioration. This is based fundamentally on our lack of faith that the corrosion protection is absolute.
Instrumentation would yield more information than the very small gross deformations anticipated under test conditions.
For the initial proof test, we concur in the 115% overpressurization with deformation and crack observations and strain gauge readings.
P00R OR'GINAL 1446 175
. 6.0 fonclusion With the exception of the proposed surveillance program for grouted tendons, we feel that the criteria presented by the applicant adequately enable the proper design, construction, and use of the containment and other Class I structures so as to cresent no undue risk to the public health and safety.
AAh I
b e
J 4
g g