ML20112D338
| ML20112D338 | |
| Person / Time | |
|---|---|
| Site: | Fermi  |
| Issue date: | 01/10/1985 |
| From: | Jens W DETROIT EDISON CO. |
| To: | Youngblood B Office of Nuclear Reactor Regulation |
| References | |
| EF2-72045, NUDOCS 8501140217 | |
| Download: ML20112D338 (10) | |
Text
T Wiyne H. Jens e
Vice Pr sdent Nuckar Operatons Ferme-2 6400 North Dme Highway
. Edison "ast";r"-
Jeneery 10, 1985 EF2-72045 Director of Nuclear Reactor Regulation Attn:
Mr. B. J. Youngblood, Chief Licensing Branch No. 1 Division of Licensing U.S.. Nuclear Regulatory Commission Washington, D.C.
20555
Dear Mr. Youngblood:
Reference:
(1) Fermi 2 NRC Docket No. 50-341 (2) Detroit Edison Letter to NRC, " Coatings Inside Containment", EF2-72778, dated August 28, 1984 (3) Detroit Edison Letter EF2-72271,
" Transmittal of Additional Information",
dated October 11, 1984
Subject:
Primary Containment Coatings Evaluation - Transmittal of Responses to Six (6) Additional NRC Staff Questions Following the initial submittal of Detroit Edison Report No.
DECO-12-2191 (Reference 2), supplemental information was provided via Revision 3 of the subject report via EF2-72271.
On December 18, 1984, Edison received, via telecopy, six additional NRC Staff questions.
In a subsequent telephone conference on December 20, 1984, Edison's response was provided to assist completion of the NRC review.
The attachment documents information provided to the NRC in this telecon.
The attachment also provides the proposed revision to Report No. DECO-12-2191 reflecting the Edison responses.
This information is provided for NRC review and approval.
Subsequent to NRC's approval of the proposed changes, Edison will reissue the subject report.
8501140217 850110 s
PDR ADOCK 05000341
() pl A
PDR i \\
i m
Mr.
B. J. Youngblood January 10, 1985 EF2-72045 Page 2 We trust that the above information will enable you to close this item prior to our scheduled fuel load date.
Any questions ragarding this matter can be directed to Mr. O.
K. Earle at (313) 586-4211.
Sincerely, f
l
[Hdk h((
cc:
All with attachments j
Mr.
P. M.
Byron Mr. M.
D.
Lynch Mr. A.
W.
Serkiz, NRC-NRR Generic Issues Branch Mr. J. C. Lane, NRC-NRR CSB Mr.
F. J. Witt, NRC-NRR CMEB Mr. J.
Norton, NRC-RIII USNRC Document Control Desk Washington, D.C.
20555 3
r.
ATTACHMENT NO. 1 TO LETTSR EF2-72045 RESPONSES TO NRC QUESTIONS
" PRIMARY CONTAINMENT COATINGS EVALUATIONS" 1.
Provide the data, or your analysis, which you used to reach your conclusion on Page 12 of your Report No. DECO-12-2191, "Enrico Fermi Atomic Power Plant Unit No. 2, Evaluation of Containment Coatings", Revision 3, dated October 1984, that "The total affected surface area due to steam and water scouring is less than 10% of the CZ-11 coated area."
RESPONSE
The statement will be deleted from the report as shown on the enclosed marked-up pages 12 and 20 of the report.
The discussion was meant to provide a quantitative margin of conservatism used in the safety evaluation.
No credit was taken in the analysis for this mitigating effect, as stated in Section 8.2 of the report.
The mitigating effect was evaluated i
on the basis of the large break design jet model, computing actual surface areas exposed to the jet.
For the Fermi 22 to containment, the largest break exposes a total of 3315 ft the steam / water jet which translates to 1.55% of the total CZ-ll and mill scale surfaces within the drywell.
2.
State the calculated velocities you assume to support your statement on Page 17 of your report that "Drywell floor and vent line/ header velocities are very low..."
Provide the calculations used to arrive at such velocities and identify the time frame following a postulated loss-of-coolant accident (LOCA) that such velocities exist (e.g.,
immediately after the postulated break, during long-term recirculation or the i
depressurization period when the operators are bringing the reactor to cold shutdown).
RESPONSE
Question Nos. 2, 3 and 4 are directed toward the discussions provided for the debris transport mechanism in the report Section 8.1.
As in our response to Question No. 1, the discussions were provided to illustrate the degree of conservatism and the mitigating effects associated with the debris transport path.
In the safety evaluation, none of these mitigating effects were used.
To clarify this aspect, the assumptions used are summarized as follows:
(a)
All CZ-ll and mill scale coatings as given in Table 1 (page 8 of the report) were assumed to fail and separate from their surfaces during a LOCA.
(b)
All unqualified coatings (3 mils DFT and less, Galvanox V) are assumed to fail during the LOCA.
l
.,,_,,.-e-w.,_-.,
ATTACINENT NO.110 LEPTER EF2-72045 (Cont'd)
(c)
All hydroxide and fibrous debris generated from the failure of CZ-ll, mill scale, unqualified paints and Galvanox V are assumed to be transported from the drywell to the torus during the initial stages of the LOCA (RPV depressurization).
i (d)
All debris deposited in the torus is evenly distributed within the suppression pool water volume and held in suspension to produce the concentrations of hydroxide and fibrous debris as given in Section 8.2 of the report (0.35% and 4.1 x 10-4%, respectively).
(e)
The debris concentrations are ingested into the ECCS flow and are circulated through the ECCS flow path.
3.
Provide the data, or your analysis, which you used to arrive at the following statement on Page 18 "... average water velocities within the torus following the initial transient are below 0.25 ft/sec, but approach 0.3 ft/sec within a 4 foot hemisphere..."
RESPONSE
See response to Question 2 above.
4.
Indicate what the velocities would be in the torus pool following the initial transient if the safety / relief valves were to discharge or the automatic depressuriaation system were to be activated.
State the effect that such a valve discharga would do to paint particulates in the pool.
Indicate whether ingestion of paint particulates into the. emergency core cooling system (ECCS) recirculation trains would occur.
Provide your calculations er data, to support your conclusions on this matter.
RESPONSE
See response to Question 2 above.
5.
Indicate the effect of ingesting paint particulates into the ECCS recirculation loops on the equipment needed to bring the plant down to cold shutdown following a postulated LOCA.
RESPONSE
The cold shutdown path for the Fermi 2 reactor, following a LOCA, consists of the RHR system operating in the shutdown mode or LPCI mode.
Equipment needed to operate with I
circulating cooling water containing coating debris, consists of the RHR pumps, heat exchangers and valves. -In the initial j
stages of the LOCA and subsequent core flooding and cooling, the core spray pumps and associated valves may also be required to operate.
I t
ATTACHMENT NO.1 '10 LETTER EF2-72045 (Cont'd)
The effect of ccating particulates circulating through the cooling loops has been evaluated as follows:
(a)
RHR and Core Spray Pumps These pumps are vertical, centrifugal pumps, equipped with mechanical seals.
Experience with this type of pump and seals in fossil power applications has shown very satisfactory operation under long-term continuous duty conditions, for example, pumping general service water (such as lake and river water) which contains silt, organics and corrosion particles.
Special design features were incorporated into the Fermi 2 ECCS pumps to further l
assure long-term operability.
These features include a l
recirculating seal cooling loop which contains a cyclonic l
separator to extract any solids from the seal cooling loop before reaching the seal.
The solids are continuously recirculated to the pump suction stream to mix with the main flow.
The pump seals thus are adequately protected from the effects of circulating coating debris particulates.
Potential seal leakage and ultimate failure has previously been addressed as documented in the Fermi 2 l
SER (NUREG-0798) pages 6-21 and 15-15.
(b)
RHR Heat Exchanger The heat exchanger is a vertical vessel, 4.5 feet in l
diameter and 24 feet long.
The RHR primary fluid is l
pumped through the shell side with top entry and near bottom exit.
Flow velocities in the shell are between 6 and 10 ft/sec, far in excess of settling velocities of the particulates.
Due to the near bottom exit flow, any particulates that could accumulate in low velocity fields, would be swept out of the heat exchanger and remain in suspension in the ECCS flow.
The U-tubes are arranged vertically with the U-bend exposed to the top entry RHR flow.
Particle build-up on the tubes, therefore, is not l
considered feasible and the heat transfer capacity of the l
heat exchanger would not be affected.
(c)
Valves l
The valves used in the ECCS flow path of the RHR and core spray loops consist of large wedge-type OSY gate valves with electric motor operators.
These type of valves are standard in nuclear and fossil power plants for fluid shutoff where pressure exceeds the service rating of butterfly valves (usually about 150 psig).
The seating surfaces are perpendicular to the flow stream, thus l
avoiding a build-up of particulates.
During the closing
ATTACHMENT NO. 1 TO LETTER EF2-72045 (Cont'd) of the valve, the fluid velocity across the seat increases nearly proportional to the decrease of available flow area across the seat, thus promoting a self-cleaning action.
Particulates entrained within the flow stream would not affect the operability of the valve, in either the closing or opening phase.
The seating surfaces (body seat rings and wedge surfaces) are hard-surfaced with stellite No. 6 to prevent seat erosion.
Such gate valves are indeed well suited for the intended service, including slurries with particulate concentrations far above those postulated to result from coating debris.
(d)
Reactor Vessel The behavior and affects of particulate concentrations within the reactor vessel have been previously evaluated by General Electric Co., as discussed in Response No. 3 of the Addendum to Report No. DECO-12-2191, Rev. 3.
6.
Provide the breakdown of the total surface area of mill scale and varnish in the drywell and in the torus.
Provide the range of particle sizes and the distribution of the 25.22 cubic feet of mill scale indicated in Table 1 on Page 8.
RESPONSE
There is no mill scale and varnish within the torus.
Some minor items previously identified in Section 2.4 of the report will remain uncoated until the first refueling outage.
.Though no mill scale is present, the surfaces are oxidized (rusted).
These surfaces represent about 150 ft2, Within the drywell, the mill scale and varnish is generally evenly distributed as part of the= structural steel and a few piping sections.
Structural steel consists mainly of hanger attachments, brackets, reinforcing gussets and similar small items which are difficult to clean and coat.
These uncoated surfaces are also present on the inside of rectangular structural tubing, not generally accessible to spray guns, brushes or wire wheels.
From a series of tests conducted at the Detroit Edison Research Department, it was determined that the measurable mill scale particle sizes ranged from approxi-mately 4 to 60 microns.
As stated in the-response to Question 2, all the mill scale and varnish are assumed to separate from their surfaces and be transported to the torus.
Response to two additional questions was requested during the teleconference on December 20, 1984.
7.
The amounts of unqualified coatings shown on Table 1 (page 8 of the report) are not consistent with the responses provided in the Addendum.
ATTACHMENT NO. 1 TO LETTER EF2-72045 (Cont'd)
RESPONSE
The unqualified coatings shown on Table 1 have been reduced in quantity due to the removal of unqualified coatings thicker than 3 mils DFT and due to recategorizing unqualified inorganic zinc primers.
The remaining unqualified coatings consist of:
Valve yoke, bonnets, bodies - 460 f t2 at 0.7 mills =0.027 ft3 MSIV top works
- 100 ft2 at 2.1 mills =0.018 ft3 Conduit labelling
- 350 ft2 at 2.5 mills =0.073 ft3 62 ft2 at 2.0 mills =0.010 ft3 Terminal box labelling Total
- 972 ft2 0.128 ft3 Table 1 will be revised accordingly as marked on the attached copy.
8.
Provide documentation regarding failure modes of unqualified coatings with DFT of 3 mils or less as stated on page 11 of the report.
RESPONSE
The mode of failure of unqualified thin film j
coatings under accident conditions involving steam exposure at 340*F can be postulated based on an examination of test panels following exposures.
Coating manufacturers and utilities have conducted many tests at Oakridge National Labs, which included coatings unsuitable for nuclear service.
The failure of cured thermoset type films is normally in the form of blistering.
Preradiation at lx109 rad causes some disintegration of non-radiation resistant films.
Nominally blisters average less than 1/4 inch in size.
Blisters form due to the reaction of the steel substrate with steam to form hydrogen.
The film is soft at 340*F and blisters which form, cause the film to be reduced to a fraction of the original thickness.
If any cracking or removal occurs, there is insufficient film tensile strength in the resulting 1 to 2 mil film to maintain any film integrity, and the resulting debris is present as a fine sludge.
However, in most cases the blister formation relieves the hydrogen pressure and this results in no further force acting to cause film removal, except in areas where steam may directly impinge the surface.
Discussions with coating experts associate] with nuclear coating manufacturers indicate that tensile strength measurements would be of little value unless measurements could be made on film heated in the 20C' to 340' range.
There is no conventional equipment or test method known to perform such a test in a reliable manner.
At the temperatures predicted, unqualified coatings at 1 to 2 mils thickness such as alkyds are too fragile to be mounted for test.
TABLE 1 E
Om R
SUMMARY
OF PRIMARY CONTAINMENT COATINGS Oh d dc 54 Approx.
mM Total Dry Film Total Total m0 Type of Qual.
Average DFT*
Surface Density Volume Mass O
- g Coating Coatings (mils)
(ft2)
(t/ft3)
(ft3)
(1bs) m g
Yn C
Ccrbo-Zinc 11 No 7
125,000 217 73 15,841 y@
Q
-e Plasite 7155 Yes 12 67,000 150 66.9 10,035 ud O
A2eron 66 and Yes 1/16" plus 7,380 125 44.6 5,575 m
t Surfacer 10 mills 2
Galvanox V No 5
775 202 0.36 73 m
l Mill Scale No 3.4 89,000 350 25.22 8,827 n
cnd Varnish a
o o,
m 3
h 3
0.7 to[
h782**
90 to 150 0.48 72 l
Unqualified No gS$
Pcints 2.C 97z o, g gg jg 2
g 3
0 O
e 2
DFT (Dry Film Thickness) measurements were taken with a Positector 2000 w
gauge, Serial #30531, calibrated with NBS shims #22272.
Q*FordetailsseeItem5ofAddendu mO I f
os N
e I
l MULTIPLE DYNAMIC 3 CORPORATION l
SUSJECT:
DATE ISSUED DOC. NO.
EVALUATION OF CONTAINMENT Oct. 1984 DECO-12-2191 l
COATINGS FOR FERMI 2 REVISION:
3 PAGE 12 all the unqualified coatings was conservatively a s s umed.
The DBA test report No. 56878 issued by Carboline (Reference 1) demonstrated that Carbo-Zinc 11 is not lost in flakes, but rather in particles of a size less than 20 microns.
The report further states that the particles do not dissolve in water and do not clog screens.
The density of CZ-11 dry film coatings is between 3 to 4 times that of water, and particles are expected to settle to the bottom of the drywell and the suppression pool, concentrating as sludge in low l
velocity areas.
The particle separation mode is a result of continuous scouring action of steam and water spray as simulated in DBA testing programs.
In a typical BWR containment, direct scouring occurs only in the immediate vicinity of the postulated pipe break and within a few feet of the containment spray headers (if used).] The total affected surface area due to steam and water scouring is less than 10% of the CZ-11 coated
~
~
,9 '
areas.
Temperature resistance of CZ-11 'up to 750*r is fde censidered excellent by the manuf acturer, as described g/
in Reference 2.
The failure mode for Galvanox v is considered similar to that of CZ-11, based on the physical properties provided by the manufacturer (Reference 3).
- However, no detailed test data exist to cuantify the rate of deterioration.
The tensile strength of Galvanox V dry film coating is such that if flaking occurs, the flakes
MULTIPLE CYNAMIC3 CORPORATION
SUBJECT:
DATE ISSUED DOC. NO.
EVALUATION OF CONTAINMENT Oct. 1984 DECO-12-2191 COATINGS FOR FERMI 2 REVISION:
3 PAGC 20 8.2 ECCS Performance
- 7 conservatively evaluate the postulated performance of plant systems and equipment, it was assumed that the hydroxide type debris, consisting of CZ-11 and mill scale particles, are completely suspended in the pool water during the early turbulent phase of the LOCA.
Under this assumption, the total hydroxide particle concentration in the pool is calculated to be less than 0.35%.,f~A more realistic assumption, using a 10%
( fraction of the debris as discussed in Section 6.0 j
would produce a hydroxide debris concentration of,]
0.035% in the pool water / In accordance with the 3
conclusions and guidelines given in NUREG-0897, Section FAN 3.2.2.4, a solid hydroxide concentration of less than 1% of mass does not affect pump performance.
The solid particles are less than 60 microns in size, and will therefore freely pass through the 1/8" holes in the suction strainers of the RHR, Core Spray and HPCI lines.
The fibr ous debris, generated by failure of unqualified 3
thin coat paints, is also transported to and distributed in the pool water volume.
The total volume of 0.5 cu f t (72 pounds), as shown on Table 1, results in a volumetric concentration of 4.1 x 10-4 %, well below the acceptable limits of 4% given in NUREG-0897 for fibrous debris.
The particles are expected to be 3
small enough to freely pass through the strainers.
In addition, the ECCS pumps and system piping has been
(~~~ -
'