ML20133B272
| ML20133B272 | |
| Person / Time | |
|---|---|
| Site: | Millstone  |
| Issue date: | 10/01/1985 |
| From: | Christiansson, Hammarlund C ABB ATOM, INC. (FORMERLY ASEA ATOM, INC.) |
| To: | |
| Shared Package | |
| ML20133B270 | List: |
| References | |
| TR-UR-85-225, NUDOCS 8510030140 | |
| Download: ML20133B272 (41) | |
Text
i AG2 A- ATO M l
O O
I TR UR 85-225 i { TOPICAL REPORT g ASEA-ATOM BWR CONTROL BLADES FOR US BWRs fa l hj i Cr"ra'fE%':" .E II l
Il a a i O
, i 1
l l
1 i
i
- p r = m :..
P PDR
R:pport AO E A- ATO M R: port 7R uR 85-225 Otstreuten Fren From Datum Date fog Seda Page
. US NRC, Customers URA3/MB 1985-10-01 1 (32 )
Fjdga gygtjansson[ Wardare imm3 by C pmmarlund , ,
s* mamined by Annra ~ e - t TaetTitle Wnu.O / fWNWN
/ /
TOPICAL REPORT -
- ASEA-ATOM BWR CONTROL BLADES FOR IJS BWRs Sammantanning Abstract (3
v' This is a Topical Report on the ASEA-ATOM design for BWR control blades and their suitability as replacement control blades
(~) in US BWRs. It serves as an update of ASEA-ATOM's Report TR V BR 82-98 rev.1, and studies in particular three different control 3
j (y) blade designs as regards mechanical and nuclear design, control i blade life limitations, operational aspects, safety evaluation, and
} j manufacturing & quality assurance.
In the text, references are made to the standard type of control blade used in US BWRs with assymetric lattices, i.e. the gy BWR-2/3/4 line of reactors. The case where other control blade pg types exist in US BWRs is not taken into consideration in this u[} report.
a I f
I The three different types of ASEA-ATOM control blades analyzed in this report differ in only one aspect, which is the absorber hole j gg , depth. The report will show, that these three types are compatible to the GE type of control blade in all relevant aspects.
O o
NJ l
i
- 3
AC E A- ATO M TR UR 85-225 Pegn i
. 1985-10-01 CONTENTS 1 INTRODUCTION 1.1 General 2 ECHANICAL DESIGN 2.1 General 2.2 Description of Blades Table 2-1 Absorber Weights Table 2-2 Control Blade Weights 2.3 Stress Analysis 2.3.1 Absorber Section Q 2.4 Stiffness YI b! Table 2-3 Control Blade Stiffness k aj 2.5 Absorber Temperatures Table 2-4 Control Blade Design Data l l 4 NUCLEAR DESIGN CHARACTERISTICS OF ASEA-ATOM 4 CONTROL BLADES ll?
g 4.1 General 2 4.2 Methods and Problem Input ll 4.2.1 Table 4-1 Fuel Characteristics Fuel Characteristics 4.3 Reactivity Characteristics 4.3.1 Single-bundle Calculations 4.3.2 Four-bundle Calculations N
4.3.3 Burnup Dependence 4.3.4 Special Effects 4.4 Power Distribution Effects 4.5 Absorber Depletion 4.6 3-Dimensional Analysis 4.7 Hf-tip Irmact 5
l ti
-- A
AD E A- ATO M TR UR 85-225 Pags 11 1985-10-01 CONTENTS (Cont'd) 5 EXPERIENCE WITH ASEA-ATOM CONTROL BLADE DESIGNS 6 BLADE LIFE LD4TATIONS 6.1 Reduction of Reactivity Worth 6.2 Bulldup of Gas Pressure 6.3 Boron Carbide Swelling 6.4 Fast Neutron Fluence 7 OPERATIONAL ASPECTS Q 8 SAFETY EVALUATION I) 8.1 Control Blade Manoeuvering l{ 8.2 Mechanical 1ntegrity 8.3 Reactivity i I 8.4 Seismic Performance 4 j l
{ '
8.5 Accident Analyses U
- g* '
9 MANUFACTURING AfC QUALITY ASSURANCE l 9.1 Manufacturing of Hf-tip Control Blades 11 i 9.2 Quality Assurance and Control 10 REFERENCES 11 TABLES APO FIGURES O -
1
AOCA-ATOM TR UR 85-225 Page 2
{ 1985-10-01 .
i I
J l ,
NTRODUCTION l
1.1 General This is a topical report on the ASEA-ATOM design for BWR control blades. This report is an updating of ASEA-ATOM Report l TR BR 82-98 rev. I of June 1982, which served as the basis for l licensing of the use of ASEA-ATOM control blades in Dresden-3 ,
by the NRC. As a result, the NRC lesued an SER titled " Safety i Evaluation by the Office of Nuclear Reactor Regulation Suppor-
- ting Amendment No. 74 to Facility Operating License No. DPR-25 ,
Commonwealth Edison Company Dresden Nuclear Power Station, Unit No. 3 Docket No. 50-249".
The above referenced AA Report TR BR 82-98 rev.1 described two basic types of ASEA-ATOM control blades adapted for use in GE BWRs, herein referenced as types 1 and 2. Both types of ;
control blades were designed to have a higher worth than the '
g original GE blades by approximately nine percent reactivity (relative). One of the two types of control blades, described in TR ej BR 82-98 rev.1, contained B4 C in all of the horizontal holes in the l 1, j, w!
l a blade wing, while the second type of control blade contained l nineteen hafnium rodlets in the top six inches of the blade wings y3 ,j whereas the remaining holes were filled with boron carbide.
More recently, ASEA-ATOM has evaluated several other control
] '
l blades to meet the needs of its customers. Several of the new ,
( ]t design types involve more closely matching of the worth to the l o((} GE control blades, between 2 and 3 % (relative) on the high side,
- g by decreasing the depth of the absorber holes in the blade wings.
The number of welds in the center of the control blade, where the 3 blade wings are joined together, has been reduced in crder to 1 decrease the total weight of the control blade (cf. reference 17).
' h IE n
- We have discontinued offering the all-84C control blade design, O because we believe that the designs utilizing the hafnium tip are superior in terms of expected service life. The addition of the !
hafnium tip is a straight forward and logical extenslor, of our basic control blade design, which can reasonably be expected to
. extend control blade life.
There are no major design changes involved, other than varying ,
absorber hole depth in the four types of control blades, and the i,, reduction of center weld joints, discussed in this report. We have extended our operating experience with our control blade design ;
and gained further information based on the evaluation of that experience and related inspection results. In addition, we have performed further analyses and studies in connection with our control blade design. This report has also been updated to reflect ,
this Information.
j This report has been submitted to the NRC for review in terms of generic approval of the general use of the types of ASEA-ATOM g
l control blades described herein for use in BWRs in the United 1 y States.
s
- Is
. A'3 E A- ATO M TR UR 85-225 Pcg2 3 1985-10-01 .
2 ECHANICAL DESIGN ,
2.1 General To adapt the ASEA-ATOM control blade design for use in US BWRs the following changes were made: a slight reduction of blade span, a small increase in blade length for achieving a " gray" nose effect and mating the absorber section of the blade to a velocity limiter.
Mechanical compatibility with reactor internals is ensured by the tighter tolerances in blade geometry achieved with the ASEA-ATOM blade design. Velocity limiter and coupling dimen-Q slons are identical to those used by GE in US BWRs. Inconel X-750 is used for the coupling socket instead of 304 stainless steel in order to avoid electrolyzing of the coupling mating surface.
The ASEA-ATOM blade design has a slightly lower total weight o
Q than the GE blades in US BWRs. This ensures acceptable scram insertion speeds and allows use of the ASEA-ATOM blades on e Il existing control rod drive mechanisms.
k 0 i The choice of materlats reflects the tight requirements found in Ig [ ASEA-ATOM BWRs with regard to cobalt, nitrogen and low
. carbon contents for in-core use. In summary, the improved blades have been designed to be mechanically compatible with the GE l BWR-2/3/4 line of reactors.
lv i{}
{ 2.2 Description of Blades Four types of blades are presented in this Chapter, out of which !
@Ja, three have been delivered for use in US BWRs, i.e. types 1, 2 and b llIl t-4-
bq > Type 1: ASEA-ATOM standard design, except as modified for US BWRs, with 84C powder in horizontal 0.236-inch (6 mm) diameter holes with a pitch of 0.315 inches (8 mm). The absorber hole depth is 4.055 inches, or 103 mm. There O "'""""'"'"*'d"'"'d**'9"-
Type 2: Same as type 1, except that holes in upper 6 inches con- [
tein hafnium metal rods (with 0.236-inch diameter) in-stead of B4C powder. ,
Type 3: Same as type 2, except absorber holes are 3.583 inches (91 mm) in depth resulting in a reactivity worth closely l matched to that of the GE control blade in BWR-2/3/4 i Cores.
Type 4: Same as type 2, except absorber holes are 3.425 inches
] (87 mm) in depth resulting in a worth still more closely matched to that of the GE control blade in BWR-2/3/4 l
l d
cores.
8 8
.I a
. ACO A- ATO M TR UR 85-225 P:ga 4 1985-10-01 .
Type 1 was supplied to Commonwealth Edison's Dresden 3 reactor in 1982, but is no longer offered on the US market. Therefore this -
study will concentrate on types 2,3 and 4.
In the following description no distinction is made between the different blade types except when explicitly stated.
The control blades consist of a cruciform absorber section contal-ning B4 C powder which is welded to a velocity limiter fitted with a coupling device for connection to the control rod drive mecha-nism of the reactor. The blade design is shown in drawing AA Hafnium-tip, entitled " Control Blade, Styrstav".
(] The cruciform absorber section is formed by four solid stainless steel sheets, which are welded together at the center. The intermittent center weld-Joint ensures straightness and required stiffness while permitting 15 cut-out sections which result in significant weight savings. Thus the total weight of the O ASEA-ATOM blades is lower than for the GE blade, in spite of the o slightly heavier absorber-containing part of the ASEA-ATOM eI h, blades.
b d l 3 The blade wings are 0.317 inches (8.05 mm) thick. The neutron
}yag absorber in each wing is contained in 454 horizontally drilled holes of 0.236 inches (6 mm) in diameter and spaced at a pitch of 0.316
! (g. inches (8 mm) resulting in a total absorber section length of 142.9 l Inches (3630 mm). The hole depth ranges from 3.425 to 4.055 la f Inches in depth as indicated above. The part of the blade that 1{i
{} Il contains the absorber coincides with the active core height,
- q except for the top one Inch of the core, when the control blade is fully inserted. The top 3.5 inches (89 mm) of the blades are free a
f holes. This section constitutes a " gray" nose which slightly y reduces local power changes, thus slightly mitigating fuel duty.
11 af a The horizontal holes are filled with natural 8 4C by vibratory pV compactors to a packing density of 70+/-3 % of the theoretical density or with hafnlum metal rodlets. The specific absorber amount for each type of control blade marketed is shown in the following Table:
p U. Table 2 Absorber Wel@ts:
Control Blade Weight of Weight of BC 4 hefnium he kg be kg Type 2 19.74 8.96 6.34 2.88 Type 3 17.40 7.90 5.59 2.54 Type 4 16.98 7.70 5.39 2.44 -
I j The resulting average 04 C content is 0.63 g/cm2 of blade wing area. Control blade design data are summarized in the enclosed g
c Table 2-4.
s
AC E A- ATO M TR UR 85-225 P:ge5 1985-10-01 The holes are closed at the outer blade edge but are connected through a narrow slit. This design allows gas pressure equalization between holes without any significant displacement of the B4C powder. The horizontal holes also render any further BA C densifi-cation after Initial filling quite insignificant. Each wing forms a separate pressure enclosure which is pressure and leak-tested after welding.
A lifting handle - cut out from the same material as the blade wings - Is welded to the top-end of the absorber section. This handle is designed to fit the grapple used for installing the control blades. At the top end of the blade are guide pads, one on each wing, which reduce the extent of any direct contact between the blade wing surfaces and the adjacent fuel channels. The guide Q pads are made of Inconel X-750, which contains less than 0.1 w/o cobalt.
The blade absorber section and velocity limiter are welded together. The velocity limiter is identical to that of GE control O blades in the United States. ,
bl k k d The blades are designed for an internal over-pressure of 2175 psi (15 MPa). The internal gas pressure is a function of average boron
[ g depletion in a blade wing. Due to pressure equalization all j [] [
I absorber holes are under the same pressure. Blades with the hafnium-tip will have a slightly lower gas pressure buildup due to
$11 the reduced axial burnup variation and no gas release from Ia I igI hafnlum. During neutron exposure a gas pressure will build up III mainly due to helium production from neutron capture in B-10, and as a result of radiolysis of any water traces present. The d
gg maximum moisture content of the B4C is specified as 350 ppm at filling in order to limit the water pressure contribution. Other
}g a reaction products in the gaseous form may be neglected. Gas y 1 35 N I pressure bulldup is not expected to be life-limiting for the blade design (cf. reference 11).
The blades with hafnium (types 2,3 and 4) are provided with a 6- -
Inch (150-mm) hafnlum-tip. The uppermost 19 holes in each wing are filled with hafnium metal rodlets instead of B 4C powder. The hafnlum is of reactor grade with 4.5 w/o zirconium. Hafnlum Q exhibits no irradiation-induced swelling and proper tolerances ensure that the hafnlum will not induce any stresses in the stainless steel sheet during irradiation. The weight increase due to the hafnium-tip is about 5.5 lbs (2.5 kg).
The total weight in air of the control blades (including velocity limiter) is shown in the following Table:
Table 2 Control Blade Weights:
Control Blade Weight j h kg i
- GE Type 225 102 Type 2 218 99
l A'OG A- ATO M TR UR 85-225 P ge8 1985-10-01 .
2.3 Strees Analysis Based on conservative assumptions regarding gas pressure bulldup (cf. reference 11), the internal design pressure was set at 2175 psi (15 MPa). Applying the same conditions and assumptions to the standard blade for equal relative B-10 depletion would result in up to 2 times higher gas pressure. This difference is due to the unique design feature of the ASEA-ATOM blades which a!!ows pressure equalization within a blade wing.
Experience with standard blades in US BWRs und with the ASEA-ATOM blades shows that premature failure (stress corro-slon cracking) is caused by B4 C swelling rather than gas pressure buildup. However, gas pressure buildup is not expected to be life-p limiting for the ASEA-ATOM blade, v
The external design pressure was set at 1375 psi (9.5 MPa), which is the maximum pressure which can occur in the reactor pressure vessel.
o n 2.3.1 -
sh' >
Absorber Section Detailed stress analyses using 2-D and 3-D finite element models were carried out for the blade design at beginning-of-life condi-tions (cf. reference 2).
IoI 32.
il !
The results indicate that a maximum external pressure of 3190 psi (22 MPa) and a maximum internal pressure of 2030 psi (14 MPa)
- 'I can be allowed without exceeding ASME-Ill criteria.
4 s' g{} I The selected internal design pressure of 2175 psi (15 MPa) has
{ been achieved by slightly tightening manufacturing tolerances in
- ] order to increase minimum wall thickness relative to the calcula-
{ tional case, 3
h ([7 s lt may also be noted, that no credit has been taken of the fact that the external pressure significantly reduces the actual pres-U sure differential by several psi or MPa for all operating conditions at high temperatures. At low temperatures, the pressure differen-tial will also be low in comparison with the design internal pressure.
-) 2.4 Stiffness Control blade stiffness was calculated for the Hf-tip design based on test results from Forsmark-3 and Oskarshamn-3 in Sweden. For the purposes of the stiffness calculation (cf. reference 4) the ASEA-ATOM (i.e. type 4) and GE designs for Millstone-1 were considered. The following results were obtained:
Table 2 Control Blade Stiffness:
Control Blade Dwg Reference Control Blade Stiffness j (N/mm) i N GE Type GE 706E855 682.99
' Type 4 AA 218 796 771.07
. 11 2
ACEA-ATOM TR UR 85-225 Pcg3 7 1985-10 01 .
The increase in stiffness,12.9 %, is considered to be moderate and is not expected to effect control blade motion significantly.
These results are volid for Types 2,3 and 4 of the ASEA-ATOM control blades (cf. referc.nce 4).
2.5 Absorber Temperatures The temperature of the boron carbide has been calculated (cf.
reference 7). The average value for a hole is 6440F (3400C) assuming a maximum power density of 44.5 kW/ litre in the adjacent fuel. The gas pressure calculations take this temperature into account.
In the case of hafnium the calculated average temperature is
(]~ 6310F (3330C) for the maximum estimated power density of 44.5 kW/ litre in the adjacent fuel assemblies (cf. reference 8).
For both 8 4C and Hf, the stainless steel of the blade wings Q maintains an average temperature below 5720F (3000C).
li 5$
Aa ou I", dI i 11, t i 4
lli l ill j IN alnlli 2 O .
O I
i
?
E I
- _ n -___-______________ ___
AS E A - ATO M TR UR 85-225 Page 8 1985-10-01 .
Table 2 Control Blade Design Data Materials Neutron absorber Blade types 2,3 and 4 8 4C powder, vibration l
compacted to 70 % of theo-retical density, in horizontal holes.
l l
Blade types 2,3 and 4, Hafnlum metal rodlets with top 6 inches 4.5 w/o Zr.
l Q Control blade wings and handle AISI 304L stainless steel with Co max. 0.05 w/o, N max.
0.08 w/o, and B max 4 w/o).
Guide pads Inconel X-750 with Co max Q 0.08 w/o and N max 0.1 w/o.
EI) f Velocity limiter Cast type stainless steel.
A2 3 Same as specifications for l Dresden-3.
Coupling socket inconel X-750 l l Dimensions and Weights i((
{
i
< Blade wing span 4.921 inches (125 mm)
Blade wing thickness 0.317 inches (8.05 mm) f} I's 1 Length of absorber section 142.9 inches (3630 mm) fI El i Overall length 174.2 inches (4425 mm)
Length of " gray" tip 3.5 inches (89 mm)
Absorber hole diameter 0.236 inches (6 mm)
Absorber hole depth type 2 4.055 inches (103 mm) type 3 3.583 inches (91 mm) type 4 3.425 inches (87 mm)
Absorber hole pitch 0.315 inches (8 mm)
No. of holes per blade 4 x 454 Hf rodlet diameter 0.236 inches (6 mm)
I i
1 L
.im ______ _ _ _ _ _ _ _ _
1985-10 01 B4C Inventory, type 2 19.8 lbs (9.0 kg)
B4C Inventory, type 3 17.4 lbs (7.9 kg)
B4C inventory, type 4 17.0 lbs (7.7 kg)
Hf Inventory, type 2 6.3 lbs (2.9 kg)
Hf Inventory, type 3 5.6 lbs (2.5 kg)
Hf inventory, type 4 5.4 lbs (2.4 kg)
Total weight in air, (approx) type 2 218 lbs (99 kg) type 3 222 lbs (101 kg) type 4 218 lbs (99 kg)
Design Pr_ essure and Temperature Design pressure, external 1380 psi (9.5 MPa)
Design pressure, internal 2175 psi (15.0 MPa)
Design temperature 5720F (3000C) 111 11
'A j
}~!i (Il 3 il il Il l O
O I
- t
! .n i .
AT A - ATO M TR UR 85-225 P ga 10 1985-10-01 ..
4
. NUCLEAR DESIGN CHARACTERISTICS OF CONTROL BLACE 4.1 General The reactivity worth of a control blade is directly dependent on the rate of neutron absorption in the blade materials - boron carbide, hafnium (when used in the types 2, 3 and 4 blades), and stainlese steel. The total absorption rate depends on several factors: ame snt and spatial distribution of the materials in the blade, and the neutron energy spectrum which in turn depends on the core cc.1ditions (temperature, steam vold content, and fuel type and its exposure history). Eventually, the depletion of the highly neutron absorbing material will also reduce the absorption
^] rate.
v One goal is to augment the control strength of the ASEA-ATOM control blades in comparison with the current standard design.
This will improve shutdown margin and scram reactivity as well as
- lengthen useful blade life but may be considered a slight penalty
- n. vi in other respects such as in reactivity increase incidents (i.e. rod ejg drop accident, rod withdrawal error).
The reactivity worth for different core states of the ASEA-ATOM ya blades relative to standard blades was calculated with the two-dimensional lattice depletion code PHOENIX. The calculations were based upon the following core geometries:
Type Reactor Country Absorber Hole Depth
{Ol
- Type 1 Dresden-3 USA BC 4 4.055 inches f j Type 2 Dresden-3 USA Hf-tip 4.055 inches Type 3 KKP-1 FRG Hf-tip 3.583 inches j f[ y} Type 4 Millstone-1 USA Hf-tip 3.425 inches b A justifiable comparison between the different types can be done using only one core geometry and one standard blade (specific calculations can be done upon request). However, between the Dresden-3 and the other calculations a new cross-section library o has been implemented. New and more reliable Hf cross-sections J are now being used. The impact of these new Hf cross-sections has led to a difference of about 4 % (relative) between old Hf calculations (type 2) and new Hf calculations (types 3 and 4).
Nevertheless, the main features from the different types of blades can be extracted from the calculations mentioned above.
An important result is that the boron sections of the ASEA-ATOM blades have a consistently higher worth for the various reactor conditions.
The goal to augment control strength of the ASEA-ATCM blades j la therefore met.
The high reactivity worth of the boron sections of the I ASEA-ATOM blades stems mainly from about 50-75 % greater g U4C loading per unit of blade wing area compared to the standard
AS E A- ATO M TR UR 85-225 Page 11 1985-10-01 .
blades. The entire contribution to the increased neutron absorp-tion rate comes from volume absorption of epithermal neutrons in the B4C. This effect is seen clearly when comparing the epither-mal absorption fractions of the different blade designs.
Other differences between the ASEA-ATOM and standard blades such as effective wing span, and the amount and distribution of stainless steel affect the relative reactivity worths in only minor ways.
The ASEA-ATOM blades have a tip (above the neutron asorber boundary) with slightly more stainless steel than in the standard
' blades. Such " gray" tips are advantageous with respect to suppres-Q sing the power gradient in the adjacent fuel above the tips.
However, generally the tip is not explicitly accounted for in core 3-D calculations. Since the differential impact between the various blade designs is quite small, no specific representation of the tip is considered.
O The blade design used in ASEA-ATOM BWRs is different in the Mg 3 3i h B4C blade section with respect to the larger blade width as compared with the ASEA-ATOM blades in this report. The assocl-ly{hf a
g ated difference is e.g. for type 1 blades about 9 % (relative) in reactivity worth. If compared to the standard blades in US BWRs ja, the worth difference is about 16 % (relative). It is pertinent to note that this higher control strength in ASEA-ATOM BWRs does Ol not pose any operational or safety concerns. On the contrary, the
{y ]
1 available shutdown reactivity in these reactors significantly feel-g litates the burnable poison design of reload fuel and provides g flexibility in reload pattern designs.
21? 4.2 IIj Methods and Problem Input 11 n h The PHOENIX lattice and depletion code is ASEA-ATOM's stan-O dard computer program to calculate reactivity quantities, bundle power distributions and homogenized neutron cross-sections for input to ASEA-ATOM's 3-D core simulator. As the PHOENIX code has been used routinely for more than a decade both direct and O indirect evidence of its performance is comprehensive. The verifi-cation is done by comparing calculated values from PHOENIX with experimental results from the so-called KRITZ experiment and experiments with p!n cell lattices, as well as more than a decade of experience (cf. reference 6).
Since PHOENIX is a two-dimensional model in x-y geometry, it will represent a cross-section of a bundle or control cell with or
' without an Inserted control blade. Reflective boundary conditions render the calculated reactivity quantitles to be valid for an infinite core with neutronic properties given by the type of fuel used. For the purpose of assessing the control strength of control j blades, especially with respect to relative worths, calculations j with PHOENIX will give reliable results. The dependence of the
- relative worths on xenon and fuel burnup may also be studied with 3 the code.
a,
.N s
. AS E A- ATO M TR UR 85-225 Page 12 1985-10-01 -
The types 2, 3 and 4 control blades have a major B4C section identical to that of type 1 and only a 6-inch Hf section at the tip.
PHOENIX calculations can only be done for one material type at a time, however, and the reactivity worth results for the Hf cases will thus represent a blade with Hf only. Of course, in the actual 3-D case the total blade worth will essentially depend on the B4 C Inventory whereas the Hf-tip section is importent mainly with respect to local power distribution around the tip.
4.2.1 Fuel Characteristics Types 1 and 2 Bundle type 9 (8x8 R) has a fulllength average enrichment of 2.65 Q w/o U-235. Average enrichment used in the calculations was 2.82 w/o for the mid-section of the bundle. Six burnable poison rods with 3 w/o Gd23 0 were used. The gadolinia content was estimated to be consistent with an annual cycle length.
O Tyne )
l ejg k A typical 8x8-1 fuel assembly, using a bundle average enrichment
- & aj of 2.62 w/o U-235. No burnable poison rods were used.
y3 Type 4 A typical 8x8-2 fuel assembly, using a bundle average enrichment
!},]a l of 2.82 w/o U-235. Six burnable poison rods with 2.0 w/o Gd23 0 4t were used.
III e
- g ] For the basic comparisons, fresh xenon-free fuel was used.
Parametric studies confirm that burnup or xenon presence will not
{ invalidate the conclusions from these basic results, 3
p jj 2[ 1] Table 4 Fuel Characteristics:
i Type Power Density Moderator Temp. Average Fuel Temp.
CZP Full Power (kW/kg U) (DF) (DF) (DF)
O 1 21.4 68 546.8 1213 2 21.4 68 546.8 1213 3 23.45 68 546.0 1254 4 19.5 68 546.8 1213 I
I
AS E A- ATO M TR UR 85-225 Page 13 1985-10-01 .
4.3 Reactivity Characteristics A comparative study was undertaken to provide a basis for judging the reactivity characteristics of the ASEA-ATOM blades relative to the standard blades when utilized in a core such as those mentioned above. The calculations of interest are for the follo-wing operating conditions:
a) Cold critical (xenon-free), corresponding to the limiting shut- [
down condition.
b) Hot full power, zero vold, i.e. near the core flow Inlet.
1 c) Hot full power,50 % vold.
The effect of fuel burnup and xenon on the blade reactivity worth were also studied.
l
, In single-bundle calculations with the PHOENIX code, the resul-leIg I,l ting infinite multiplication constant, k w, is obtained for an l1 j infinite core lattice with all the control blades inserted when the
[ {3 ' blade is represented. This situation is similar to the reactor shut-8g3f down case. The effect of fuel burnup is evaluated by calculating
, k, as a function of burnup.
{l}]
Ia
((}y i ,2 l A core representation which is more appropriate for the reactor at high power is obtained from the PHOENIX four-bundle calcula-tions. Effectively, every fourth blade in the core is then inserted.
v
- ga {8] Power distribution changes due to differences in blade design are l}l meaningfully studied in this geometry (cf. Section 4.4).
4.3.1 h 2( ) Single-bundle Calculations The type 1 blade with B4C is worth some 7 to 9 % (relative) more than the standard blade for all conditions. In terms of absolute reactivity, this implies in the cold zero power case that the gain in shutdown reactivity is about 1 %.
The baron sections of the type 3 and type 4 blades are worth about 4 % and 2-3 % (relative), respectively, more than the standard blades for all conditions.
The Hf section of the type 2 blade has a reactivity worth per unit length which is very close to that of the standard blade, especially during cold conditions.
The Hf section of type 3 blade has about 8-10 % lower reactivity worth compared to standard blades (cf. reference 10). This lower reactivity worth as compared to the previous calculations is duo j to the new Hf cross-sections, as discussed above. However, the j total reactivity worth of the type 3 blades remains about 3-4 %
- higher compared to standard KKP blades for all operating condi-y tions.
s n . - -
AT A- ATO M TR UR 85-225 Pag) 14 1985-10-01 ..
In terms of absolute reactivity, this implies in the cold zero power case a gain in shutdown reactivity of about 0.4 %.
The hafnlum-tip section of the type 4 blade has about 10 % lower reactivity worth compared to standard Millstone-1 blades (cf.
reference 13). However, the total reactivity worth of the type 4 blades remains about 1.5-2 % (relative) higher compared to stan-dard Millstone-1 blades for all operating conditions.
In terms of absolute reactivity, this implies in the cold zero power case a gain in shutdown reactivity of about 0.2 %.
Similar calculations carried out for other plants show that the relative reactivity worth difference between ASEA-ATOM blades and standard blades is not sensitive to fuel channel wall thickness, l and not significantly influenced by the fuel assembly types (8x8 or {
9x9) adjacent to the control blades. l 4.3.2
- c. Four-bundle Calculations
- El l 55 Blade worth values will be much less in these cases than for lyn single-bundle calculations since only 1/4 of the control blades are represented. The t 32 6 to 7 % (relative)ype more1 blades control with B4 C than strength now standard have of the orderInof blades.
11 the relevant hot case this translates into 0.4 % more worth in ll 4 terms of absolute reactivity. I 4 i
((} h The worth of the hafnium section of the type 2 blade again is very
- g ] close to the standard blade worth using old Hf cross-sections. The dif ference in reactivity level is less than 0.1 % (absolute).
{
The type 3 now shows about 2.9 % (relative) more control strength h l[ l] than the standard KKP blade. Also, the type 4 blade now shows about 1.6 % (relative) more control strength than the standard Millstone-1 blade. For practical purposes this means that reacti-vity worth matching has been achieved.
Similar comparisons done for other core designs demonstrate, that
' there is no significant influence of fuel assembly type on blade worth differences in four-bundle calculations either.
4.3.3 '
Burnup Dependence Single-bundle calculations were done as a function of fuel burnup to 34 mwd /kgU. The xenon concentration was in equilibrium in the full power case. A burnup-averaged vold of 50 % was assumed in the cold cases. The comparison shows that the reactivity worth differences change only slightly with increasing burnup. The variation is somewhat greater in the cold than in the hot case. At cold zero power, the absolute difference in reactivity worth between the type 1 blade with B4C and the standard blade is j within 1.0 t 0.15 %. The corresponding difference for the type l blade section with Hf is only +0.01 to -0.16 %.
W 18 An important fin &g is that the differences in reactivity worth l between the various blade designs are nearly independent of fuel
r AD E A-ATO M TR UR 85-225 Prg315 1985-10-01 burnup and xenon concentration. For a comparison of blade worths, this consistency implies that explicit representation Is not required of the actually different burnups of fuel bundles adjacent to a control blade. Thus, calculations done to obtain results, i.e.
at zero burnup of all fuel bundles and with no xenon, are adequate for the purposes of this report.
4.3.4 Spectral Effects Due to the strong thermal neutron self-shielding in B4C, the fractional epithermal absorption is considerable, between 40 and 60 % in the studied cases.
As expected, the harder neutron spectrum in the hot cases, especially in the volded condition, results in a higher fractional epithermal absorption relative to the cold cases. The larger volume / surface ratio of the absorber for the ASEA-ATOM blades with B 4C also gives a higher relative epithermal absorption rate since epithermal volume absorption increases while the effective thermally black surface remains approximately unchanged rela- {
j
- o. - tive to standard blades. -
l
- 5I 5
l g It is seen that Hf has an absorption ratio very similar to B4C. This )
ya
]d is in spite of the very different energy dependence of the hafnium '
and boron cross-sections. It is a result of the fairly high resonance g ja integral of Hf combined with a somewhat low thermal absorption II cross-section with relatively small self-shielding.
4I 4.4
{y 3y ih Power Distribution Effects e
- g Differences in neutron absorption rate between the various con-fja trol blade designs will affect somewhat the neutron flux density distribution adjacent to the blade. It is of interest to assess the h l[ y} corresponding differences in power distribution as is shown in Figure 1 for the type 1 blade relative to the standard blade. In the four-bundle geometry all the fresh assemblies are identical and of the type described in Section 4.2.
The fuel pin power relative to average pin power in the control cell attains a maximum decrease of about 5 % at the wide-wide gap corner. In terms of linear power this corresponds to a decrease of less than 0.25 kW/ft. The largest increase is less than 2 % or 0.1 kW/ft at the remotest point from the blade. The increase is rather less in the actual core-wide case. This is i because the power distribution is renormalized over the complete core and the ASEA-ATOM control blades are withdrawn slightly more than standard blades would be.
The change in signal of the LPRM detector closest to the type 1 control blade is about 0.5 % compared to the signal obtained with l a standard rod. Thus, while the incore monitors sense only a minor I impact from the difference in absorption properties of Type 1 and i standard blades, local power variations will be accounted for by a
[, correct treatment of the blade in the core calculations.
r E
A
. AC E A-ATOM TR UR 85-225 Pcgm 16 1985-10-01 .
The fuel pin power relative to average pin power in the control cell attains a maximum increase of 6.2 % (B 4C-section) and 10.7 % (Hf-tip) respectively. In terms of linear power this corres-ponds to a change of 0.34 kW/ft (1.1 kW/m) and 0.58 kW/ft (1.9 kW/m) respectively.
l The expected change in the LPRM detector signal closest to the ASEA-ATOM hafnium-tip blade is within 0.3 % of the signal obtained with a standard KKP blade in that position. Since the incore monitors sense only a minor impact from the differences in absorption properties of ASEA-ATOM and standard KKP blades, local power distribution uncertainties will not be significantly I affected.
l The fuel pin power relative to average pin power in the control cell attains a maximum increase of 9.7 % (B 4C-section) and 10.7 % (Hf-tip) respectively, i
The expected change in the LPRM detector signal closest to the f c.
ASEA-ATOM hafnium-tip blade is within 0.5 % of the signal ,
(55dg f8 obtained with a standard Millstone-1 blade in that position. Since f
) f the incore monitors sense only a minor impact from the diffe-lllfa
}g3g rences in absorption properties of ASEA-ATOM and standard Millstone-1 blades, local power distribution uncertainties will 1
)
- g. again not be significantly affected. {
The reduction in absorber hole depth from 4.055 inches to 3.425 f!j{I 7) l
{Qp)a] inches (103 mm to 87 mm) has resulted in a slighty larger relative
{ l power level in the corner next to the inserted ASEA-ATOM g g hafnium-tip control blade.
3 This slight increase in corner blade power levels - compared to y the standard blade case -is acceptable:
II 1 a) It leads to somewhat smaller ramp rates during control blade withdrawal which is an improvement.
b) There is no thermal margin concern, since the absolute power level in the corner region is typically about 50 % of the nominal value (without control blade).
4.5 Absorber Depletion Natural baron carbide as used in the control blades contains close to 20 % of the high neutron cross-section isotope B-10 whereas the other boron isotope, B-ll, has negligible neutron affinity. A measure of absorber consumption then is to determine the frac-tional B-10 burnup which can be related to the change of reactivity worth of the blade.
The burnup of the neutron absorber in a blade will lead to decreasing neutron absorption and hence reduced control strength.
} The allowed loss of reactivity worth as a life limiting criterion is i discussed in reference 14.
s
$ The rate of reactivity decrea:e as a function of neutron exposure 8
.H depends on the initial absorber inventory (i.e. on relative effec-y - - - _ - - -
. AC E A- ATO M TR UR 85-225 Paga 17 l
l _
1985-10-01
[ tive consumption of absorber nuclides). The higher B4 C inventory l of the ASEA-ATOM blades will therefore effectively increase the-exposure necessary to achieve, say, a 10 % reactivity decrease,
, relative to standard blades. The higher reactivity worth of the ASEA-ATOM blade imp!!es a correspondingly greater B-10 con-l sumption rate but only slightly offsets the life gain due to the l Increased B4 C content.
With near 80 % more BC 4 than the standard blade, the
!- ASEA-ATOM type 2 blade will have an exposure time advantage
! of at least 60 % for a given relative reactivity decrease or l
relative depletion of B-10. It should be noted that for any chosen relative B-10 depletion, the ASEA-ATOM blade retains its higher reactivity worth compared to the standard blade. Thus, if blade end-of-!!fe were determined by an absolute reactivity worth, the exposure time of the ASEA-ATOM blade would be twice (or more) j- that of the standard blade.
i Hafnium has several isotopes with significant neutron absorption
,. cross-sections, the most important one being Hf-177. Successive ej k transmutations to other relatively large neutron cross-section
,(
f isotopes during neutron absorption serve to slow down the overall depletion of Hf. This effect gives Hf a slower depletion rate than for B4C. '
a y j2 _y 3
In order to study the relative depletion rates of hafnium and k]1 N9 l boron, calculations were performed with the HAMMER depletion j code for single cylindrical absorber rods in an appropriate neutron gl{ N]T - spectrum. Although the results are not directly applicable to the up je] i absorber geometry of a control blade, the relative absorption
- E - rates as a function of depletion give a good indication of the
&Q J
'f s
2j I
' - actual behaviour.
) g[ 1g The Hf section has a reactivity worth per unit blade length which is lower, than that of the standard B4 C blade. However, due to the slower depletion rate, the hafnium will have higi. r reactivity worth than the standard 0 4C blade at high exposures.
4.6 '
3-Dimensional Analysis The objective of the 3-dimensional core calculations was to assess the Impact of a higher blade reactivity worth on reactivity rates, ',
, power distributions, core supervision and control blade patterns.
The 3-D calculations were carried out for the cold clean state,
=I2 '
hot zero power and full power conditions (cf. reference 12). The
~
result of the 3-D core calculations confirm earlier conclusions based on. 2-D cell calculations with the PHOENIX code, i.e.
nuclear compatibility is ensured.
. 4.7i Hf-tip Impact The replacement of the 84 C with hafnium in the uppermost part j of the control blade reduces the steepness of the power gradient j in the vicinity of the control blade tip, since hafnium has about 10 % lower reactivity worth than B4 C.
g 8
S A - - - - - -
ACE A-ATO M TR UR 85-225 Page 18 1985-10-01 .
l l
The impact of the Hf-tip (top 6 inches) on the shutdown margin is negligible.
The impact of the reduced blade worth in the hafnium tip region has been assessed on a generic basis by analyzing the most limiting transient using a one-dimensional transient analysis code (cf. reference 6). The dryout limiting transient - pressure con- l troller failure in closure direction - was investigated with and I without Hf-tip control blades in a limiting end-of-life situation. ,
The influence on the reactor power history during scram is as expected very small - the equivalent insertion delay time is less than 0.01 seconds - and the impact on transient dryout margin is negligible. This conclusion is applicable to all BWR-2/3/4 type plants, where no credit is taken for fast pump runback (which would make differences even smaller).
, A similar conclusion could also be reached by simply considering the fact that 10 % reduced reactivity worth on 4 % of blade length would be equivalent to an increase by 0.01 seconds in y, scram time - disregarding the fact that the non-linearity of the
'dg l,II scram characteristics makes this increase even smaller.
5 iI
- f ioI 1 tli tiP
( A !'lI i lli
!!j 1] {
2
!Illll!
lld n l
I E
s G
E E
. E E -
Paga 19 1985-10-01 .
I s
5 EXPERIENCE WITH ASEA-ATOM CONTROL BLADE DESIGNS Control blades with the specific ASEA-ATOM design of the absorber section have been in use in operating reactors since
's 1981. Table 5-1 lists the experience as of November 1984 for reactors that were in at least their third cycle. The use of single
, control rod sequences during operating cycles has been a routine ;
operating concept since 1977. The strategy a!!ows the selection of different control cells in consecutive operating cycles which results in a comparitively even exposure distribution for control blades in the control zone of the core.
The performance experience and its implications on guide lines for determining the service life of the control blades is given in reference 14. The life limitations are also discussed in Section 6.
I Qg ill '
24!I
}it.a , i I
dis!
III !'
31]l1ljl fil I5a!.,! e
. f1
/
I s
3 E'
s -_
A E A - ATO M TR UR 85-225 Prga 20 1985-10-01 .
6 BLADE LIFE LIMITATIONS One obvious life limit has already been mentioned in Chapter 4.5, viz. that of a calculated reactivity reduction due to absorber depletion. Other potential limits relate to the mechanical inte-grity of the blade associated with buildup of internal gas pressure from helium release, baron carbide swelling, or fast neutron fluence. These various effects will be discussed in turn below.
Reference 14 gives the imp!! cations for service life considerations as based on operating experience.
6.1 Reduction of Reactivity Worth The generally used criterion which will be applied also to the ASEA-ATOM blades is that the average reactivity worth of any 3
~
ft. segment of blade must not decrease by more than 10 %
y, ,
relative. This implies an average B-10 depletion in this segment of dj{ about 42 % but with local depletion significantly above this value, especially at the very top end of the blade. Because of the higher 3l5 f B4C content of e.g. the type 2 blade it will require about 60 %
gg f longer exposure time to reach the limit B-10 depletion of 42 %
gg than with the standard blade.
ff l Due to the initially higher reactivity worth of the ASEA-ATOM j{, j blade compared to the standard one, it may be argued that a
(({s y' 2 somewhat larger reactivity decrease may be acceptable. Average B-10 depletions up to 55 or 60 % would then be reached. The dqI I extent to which such depletion extension would be utilized will 3 depend on how credit is taken for the higher worth of the blade g y and the associated shutdown margin. However, this extra-life
, hj gg g allowance, will not be credited in the current qualification of the ASEA-ATOM blades.
f Since the standard blades, using the 10 % reactivity depletion criterion, have a life insertion time equivalent to about two 18-i month cycles, it is concluded that the ASEA-ATOM blades will
( , have a correspunding life time of continuous insertion of at least three 18-month cycles. This target life time is expected to be reached with these blades even without taking credit for the higher reactivity worth of the blades.
6.2 Bulldup of Gas Pressure Neutron absorption in B-10 produces helium, most of which is retained in the crystal structure of the B 4C grains. However, a fraction of the He is released to the void volume of the compacted B4C powder. The fractional release depends on several j factors such as B4 C morphology, impurities, stoichiometry, tem-j perature, and B-10 depletion. Reference 15 gives a summary of g tnese factors. For BWR conditions and with the B AC powder s quality used in the blades, a He release of 10 % is assumed with g no depletion dependence. This may be compared to the He release
, ACEA-ATOM TR UR 85-225 Page'21 1985-10-01 .
design basis for the standard blades: He release varies linearly from 4 to 20 % as B-10 depletion increases from zero to 100 %.
For example, at 30 % blade average B-10 depletion (typical of life limit), this release model gives 8.8 % He release compared to the 10 % for the ASEA-ATOM blade. There is no gas release from hafnium.
The released He will accordingly contribute to a pressure buildup in the blade together with the other originally contained gases and any formation of radiolytic gases from traces of water (100 %
radiolyals is conservatively assumed). Since all the B4 C holes in a blade wing communicate, the pressure will equalize throughout the full void volume of the wing. For a 73 % B 4C compaction O density, the maximum calculated pressure at a blade average depletion of 30 % B-10 is less than 2030 psi (14 MPa) which is less than the design pressure of 2175 psi (15 MPa). At 30 % blade average depletion, the 42 % B-10 limit of average depletion over
_ a 3 ft. section will generally have been reached.
O It is concluded that He pressure buildup will not become limiting h I.l prior to reaching the reactivity end-of-life.
5 h 6.3 yaf Baron Carbide Swelling
- Swelling of the boron carbide as a result of neutron irradiation has J ) been identified as a potential mechanism for causing blade failure. The volume expansion of the B4 C will exert a strong q{}
{ internal pressure which may induce cracking of the containment 4 o
- p1.l] f wall. The swelling is due to the helium and lithium retained in the microcrystals.
'ta 09l The swelling rate for solid B4C (hot pressed or sintered) is given
~ h (({} in reference 16. A similar swelling rate would be expected for
.n particles in B 4C powder, and a value of the order of 0.2 % volume V increase per % of B-10 depletion has been measured. In the case of contained powder, the rate of volume swelling is unknown since some of the 30 % porous volume will be occupied by the swelling grains. However, due to the hardness of B4C and an observed caking of the powder into a fairly solid mass, swelling of the B4C N. may eventually cause high stresses in the stainless steel wall.
At present there is no adequate model to describe the mechanical interaction between the 84C and the steel wall. Only operating l experience with a given design will accordingly provide reliable
~
data on the range of B-10 depletions above which onset of failure may occur. Currently, it is concluded that the potentially critical l B-10 depletion is of the order of 45-50 % " local" depletion defined ,
as the average depletion along a single hole. This depletion level l Is based on calculations while measurements of B-10 depletion i give higher values (cf. reference 14). Depletions in excess of the j threshold level have been reached with the ASEA-ATOM design in g BWRs. in Sweden. A pattern of cracking occurrence has been g identified.
8 8
. AD E A- ATO M TR UR 85-225 Pcg3 22 1985-10-01 .
It should be noted that after crack initiation, B-10 depletion is expected to increase somewhat before significant loss of B4 C occurs. The size of this increase is not known for the depletion level and geometry of the ASEA-ATOM blades but based on experience with the standard blades, a reasonable value would be in the range up to 15 % increase of B-10 depletion.
Factors that promote a high critical B-10 depletion for the ASEA-ATOM blades are the use of:
low-carbon steel which has low susceptibility to stress-assisted intergranular corrosion. Such corrosion has been the main failure mechanism of control blades; o
U minimum steel surface in contact with the reactor coolant.
Combined with the absence of crevices this further mini-mizes corrosion risk; p
0 gas pressure equalization throughout the complete blade y, wing. The contribution to local stresses is minimized; .
- El
]O -
short horizontal holes for the B 4C. This design eliminates any ag concern resulting from continued densification, if any, of the jggh B4 C powder which may aggravate the effect of carbide sweih,ng.
]
gh Experience shows that the occurrence of cracking is strongly
]l related to the local depletion. When the uppermost B4C hole g{} 5 exceeds 45-50 % average B-10 depletion (calculated), there is a e
- lE risk of crack initiation. The probability increases with depletion and at about 53 % depletion it is in the order of 0.5. The cracks propagate along the holes across the blade wing. Eventually, very 9
3 small black B4C grains will penetrate the crack and some will h gyg adhere to the thin oxide and crud layer of the blade wing surface.
n This makes crack occurrence easily visible when inspecting the U blade surfaces.
Such inspections have shown that cracking invariably starts in the very top segment of the absorber length. This is a result of the n very strong neutron fluence peaking in this segment relative to U the remaining part of the blade. Figure 6-1 depicts two examples of neutron fluence distribution along a centrol blade. In the case of an all-B4C control blade, the fluence threshold level for crack initiation is generally exceeded before the nuclear (or depletion) life limit is reached.
This experience demonstrates the need for hafnium as a neutron absorber in the . blade top end. Hafnium has no known deleterious effects at high depletions and has the advantage of the slower rate of reactivity reduction. Thus, for each 3-ft segment below the Hf blade segment, attainment of the nuclear life limit is j possible without breach of mechanical integrity.
I C
S E
a R
. AS E A- ATO M TR UR 85-225 Page 23 1985-10-01 .
Based on the ASEA-ATOM control blade experience and a hot-cell examination carried out on four.high-exposure blade segments, the results allow the followlag conclusions (cf. reference 14):
Visual inspection of ASEA-ATOM control blades in the reac-tor pool is sufficient to check blade integrity.
ASEA-ATOM control blades develop cracks at about 60 %
local B-10 depletion based on actual measurements.
B4C washout occurs in absorber holes with cracks. Washout which can occur during one cycle is limited such that the impact on shutdown margin is negligible.
O 6 Fast Neutron Fluence The irradiation by high-energy neutrons (>l MeV) will accordingly l embrittle the stainless steel of the blade. However, the mechani-cal strength requirements are satisfied for saturated conditions at j
Q high fluences. These requirements concern stiffness as regards seismic strength and handling. Hence high neutron fluence will not eIg [(,; limit blade life.
a4 y yI a li f i
,i 14 l Il
{} "h!
nilli O
O s
i 1
S I
a s -
ACEA-ATOM TR UR 85-225 Pcg2 24 1985-10-01 ..
7 OPERATIONAL ASPECTS I The incorporation of ASEA-ATOM control blades with matched worth (types 3 and 4) into a reactor core which otherwise has the GE blade has very small real impact on reactor operation. This is to be expected for two reasons, namely the mechanical compati-bility of the ASEA-ATOM design with core components and the -
nearly matched (types 3 and 4 blades should be worth 2-3 %
(relative) more than GE blades) reactivity worth between the blades. In addition, the impact on power distribution is small and, hence, tolerable.
g Nevertheless, it is pertinent to discuss items here that are related to core management. Most of them have been treated elsewhere
~
in this report and will be included for the sake of completeness.
It has been shown that when the ASEA-ATOM control blades are part of a Single Rod Sequence, they will be inserted into the core for a large part of each operating cycle. The higher control '
,;Q aj l strength of the higher worth ASEA-ATOM blades will very
] j slightly change the Insertion pattern relative to a core with 3 -standard blades only. The difference will amount to one notch of
)y withdrawal for some blades which does not significantly affect sequence planning.
J j The ASEA-ATOM blades can be correctly treated in the calcula-g ] tions for core management done by the reload fuel vendor. Thus the specific power distribution surrounding the ASEA-ATOM gg) I og f blades will be accounted for as well as global effects on the core ejJ a power distribution.
22 f j'llj With regard to the calculation of reactivity shutdown margin j gQg there is no need to ~ explicitly include the properties of the ASEA-ATOM blades since calculations based only on standard 0- biades wiii be conservative.
Current utility practice for the accounting of control blade exposure will be satisfactory also for tracking of the exposure of n the ASEA-ATOM blade. The exposure limit is suggested to be set
'L) at 42 % B-10 average depletion for any 3 ft. section of blade. The corresponding depletion level according to the core surveillance
- P3 program (valid for GE control blades) is given in reference 14.
~
8 SAFETY EVALUATION It is the objective to provide ASEA-ATOM control blades which are mechanically compatible with the existing blade ' drives and core geometry. Whereas the coupling device and velocity limiter section of the ASEA-ATOM blade are identical to those of the currently used standard blades, the blade section has, however, a j different design (although the blade envelopes are very similar).
[ The reactivity characteristics are purposely different for the two 5
g designs. It is of interest to evaluate the implications of these y design differences on the safe operation of the ASEA-ATOM E control blades.
.t ,
, ACCA-ATOM TR UR 85-225 Psge 25 1985-10-01 8.1 Control Blade Manoeuvering Mechanical compatibility of the ASEA-ATOM blades with core and rod drive components must be assured in all respects in order that normal manoeuvering and fast insertion (scram) shall be satisfactory. The following requirements must be met:
Control blade shape must essentially conform to current standard blades to ascertain that no significant increase in friction of movement or in flow resistance is obtained.
Coupling shall be made to existing rod drives.
- Handling of control blades shall be done with existing tools and grapples.
Control blade weight should not be significantly higher than Q for standard blades in order to ensure proper scram times.
N These requirements are met for the ASEA-ATOM blades. Specifi-kl ylg j h cally, the coupling and velocity limiter section of the blades are Identical to the existing ones. Blade span, blade thickness, and
)y g handle features are also equal to those of the standard blades. The ja dimensional tolerance envelope of the blade is well within that 11 currently required. The weight is about 7-12 lbs less than for I standard blades.
4 3 t Differences between the ASEA-ATOM and standard blades which
' g((} may affect function are:
a a
- Overall blade length is greater by 0.4 inches. This extension is well within limits of overtravel requirements for blade j j[ 1] movements in connection with reactor scram.
O -
The centrat structure of the blade is fairly open in the ASEA-ATOM blade design. This will very slightly increase flow resistance during scram. The blade surface of the ASEA-ATOM blade is smooth whereas the standard blades
_m have a perforated sheath. This will give slightly lower flow V resistance for the ASEA-ATOM blade. Since flow resistance accounts for about 1 % of scram insertion force, these i differences will have negligible impact on scram time.
- The use of guide pads at the blade top instead of pins and 1 rollers. Because of small lateral forces, any increase in )
friction between guide pads and fuel assembly channels over l that obtained using pins and rollers has negligible effect on l control blade motion for normal operating conditions.
Thus, the ASEA-ATOM control blades are fully compatible with j the core and fulfl! the performance requirements with respect to j
g manoeuvering and scram for normal operating conditions. Manoe-uvering and scram tests will be performed with the / ~EA-' ATOM g blades prior to cycle startup for demonstrating accer Sie func-8 tion.
- h
AC E A- ATO M TR UR 85-225 Pcge 26 1985-10-01 .
8.2 MechanicalIntegrity Conceivable failure mechanisms not directly associated with structural strength of the control blades are dimensional changes and corrosion.
Significant blade warping as a result of stress relief due to temperature cycling and neutron irradiation would increase blade to fuel channel friction. As a result normal manoeuvering and scram could be slowed down and ultimately the blade might get stuck between fuel channels. However, the ASEA-ATOM blades have been stress relieved through heat treatment during fabrica-tion and no significant warping is therefore expected. This is confirmed by experience in Swedish BWRs where no warping has so far been found.
The cracking of the blade surfaces by stress-assisted corrosion due to B 4C swelling has already been discussed. Here, the implications of such cracking will be discussed.
C)
U 7 When cracks develop along a hole, some BA C will escape through e$g k, the crack. There are two mechanisms for this process but both 5 h depend on repeated water exchange through the crack as a result a
of external pressure variations. One mechanism is simply washout hy2 of very small B4C grains and is evident from B 4C traces found outside the holes. The other mechanism is a gradual leaching process.
] Examination of cracked control blade wings show that loss of B4 C 4{}
{ $ is a gradual process. Only the immediate few adjacent (uncracked) til holes show significant loss. The resulting effect on the shutdown
- I reactivity margin is negligible (cf. reference 14).
f i Water intrusion is of little consequence with respect to pressure h j[ y} conditions. Water that enters during cooldown will upon the n ensuing heating be partly expelled from the blade. Even if
! J expulsion were very slow, Internal pressure would not build up to more than about 150 to 300 psi (1 or 2 MPa) above the external, pressure. The internal pressure would then correspond to satura-tion pressure of the water in the blade which is at most 18 to o 360F (10 to 200C) above the reactor coolant temperature. The
_,1 differential pressure should be compared with the design pressure of 2175 psi (15 MPa).
8.3 Reactivity The decrease of reactivity as a function of B-10 depletion is discussed in Chapters 4 and 6. The maximum calculated decrease is within 10 % relative for any 3 ft. segment of blade.
Loss of B4C due to teaching or washout may significantly reduce reactivity worth of a blade. This could have an impact on ccre shutdown reactivity. Although large-scale B4 C loss is highly j unlikely to occur, it is instructive to estimate the reduction in j shutdown reactivity upon complete or partial loss of the absorber.
g A study was carried out for Swedish BWRs and is indicative also E for US BWRs. Core reactivity was computed for a case with only 8 one (stuck) blade withdrawn and with different assumptions of E
R
AOOA- ATO M TR UR 85-225 Pcg3 27 1985-10-01 .
84C loss in a blade . adjacent to the vacant position. With total loss of B 4C in the complete blade, shutdown reactivity decreased by less than 2 % (absolute). If only one wing of the blade experienced B4C loss, the decrease was less than 0.5 %. These values are approximately halved for the cases where only the top quarter segment of the blade is empty of B4 C.
It is highly unlikely that loss of B C 4 would decrease the shutdown margin by more than 1 % which is the minimum calculated requirement. Considering shutdown reactivity verification at be-ginning of cycle and the fairly slow process of B AC escape, there is consequently no safety concem related to the hypothetical loss of B4C.
O The higher worth of the ASEA-ATOM blades relative to standard blades (cf. Chapter 4) will improve on shutdown and scram reactivities.
Notch reactivity worth of the ASEA-ATOM blades will also be
,O slightly higher than for standard blades. This is of interest in a k connection with local criticalities to the extent that such are 4 j j done. The increase is less than 10 % relative and would not significantly aggravate approach to criticality procedures.
I 8.4 Seismic Performance The current US standard blade design has been quallfled by l seismic testing performed with unirradiated control blades and 4 using static simulation. Control blades were scrammed into gaps ggy b] formed by fuel channels, which were held rigid at the middle with y i varying degree of deflection and bow. The control blades did fully
- E insert up to a critical value of bow.
Seismic tests with vibrating fuel assemblies and control blade j g{ gj scram insertion have been performed by another BWR supplier.
The results Indicated that scram insertion was accomplished in a O rachet-type of motion up to a certain critical amplitude of oscillatory motion of the fuel assemblies.
Differences between the ASEA-ATOM blade design and standard blades which might impact seismic performance are.
Blade stiffness Friction forces between blades and fuel channels Blade strength ]
1 During scram insertion under seismic conditions hydraulle friction i forces can be neglected compared to blade / channel friction. The '
ASEA-ATOM blade design has guide pads instead of pins and rollers on the standard blade tip. The friction force caused by guide pads against the fuel channels will be equal to rollers sliding (due to contact with both channel walls) against the fuel channels j since it depends only on the friction coefficient and contact
>g force. However, if the contact force increases to the point where deformation occurs on the channel surface, the friction force will
{s increase. The large contact surface of the guide pads as compared to rollers means smaller contact pressure, less surface deforma-
.aa _ . _ _ _ _ _ __ _ _ _ _ __ - . _ _ _ __- -- - - - - - - - - - - - - - - - - .--
, A'2 3A-ATOM ,TR UR 85-225 Pcg3 28 1985-10-01 tion risk and hence equal or smaller friction forces during scram insertion under seismic conditions.
A comparison for the case of rotating rollers (assuming contact agsinst one channel wall only) and guide pads shows that pads result in slightly higher friction forces. This increased friction is small (about 2 %) both compared to fuel assembly weight and available blade insertion force.
The ASEA-ATOM blades have a smooth surface as compared to the standard blades which have a rather rough surface caused by about one thousand 1/2-Inch holes punched out of the sheath which forms the blade surface. Therefore, the ASEA-ATOM blade Q
U is expected to cause less resistance due to blade surface / channel contact durlag scram insertion under seismic conditions.
Comparison of ASEA-ATOM and standard blades with regard to blade strength during seismic loading shows that both blade y
(}
designs will have about the same maximum strain occurring in the outermost blade fiber. A conservative estimate gives about 0.6 %
ej 1 which is well below maximum strain for the ASEA-ATOM blade,22
] h maximum allowable strain of 1 % at EOL (10 n/cm2, E >_1 MeV) for the ASEA-ATOM blade material, t
.p In summary, it is concluded that the seismic performance of the 11 ASEA-ATOM blade is comparable to the standard blade.
F 8.5 4{}
g I! Accident Analyses The types 3 and 4 control blades have a marginally higher up reactivity worth (1-3 % relative) than previous standard blades.
85 The implication of this close matching is that the reactivity increase accidents are not significantly affected.
jj g{ f The effect of the lower worth of the top end segment with hafnium has been assessed in reference 6 with respect to scram O reactivity requirements. (Cf. Chapter 4.7.)
O e
I s
E 11
AOGA- ATOM TR UR 85-225 Paga 29 1985-10-01 ..
9 MANUFACTURING APO QUALITY ASSURANCE 9.1 Manufacturing of Hf-tip Control Bladea ASEA-ATOM operates a Fuel and Component manufacturing fact-lity with a capacity corresponding to two complete first cores for large reactors including 225 control blades a year. The facility went into operation at the current location in 1972. Technical personnel competence supporting fuel and component manufac-ture dates back to the late 1950m The first control blade for Swedish BWRs was manufactured by U^ ASEA-ATOM 17 years ago in 1968. Status of manufacturing experience as of September 1985 can be seen in Table 9-1. (The type of control blade denominated "CR-82" in Table 9-1 is equivalent to the Hf-tip blade discussed in this report.)
y The control blades are manufactured in a production line which is a k equipped with machines that meet the extensive and unusual 2 al quality requirements.
yi The main steps in the fabrication process are:
f'I l - Receiving inspection of material I '
3 -
Gun drilling of the blade wings NI -
Milling of the blade wings
'22E Cleaning of the blade wings f
OE l g[hl - Boron filling of the blade wings O - Seai roiiina of the biade wings Seal welding of the blade wings p - Helium leak testing of the blade wings O
- Welding leak test hole ,
- N.D.E. test of sealed leak test hole
- Welding of cruciform wing assembly in fixture
- P.T. test all welds Stress relief heat treatment of cruciform assembly j - Dimension inspection in reference gauge I Welding velocity limiter to cruciform assembly g -
t
, AOOA-ATOM TR UR 85-225 Pcg2 30 1985-10-01 .
9.2 Guelity Assurance and Control The control blade fabrication at ASEA-ATOM including design and purchase of material is performed in accordance with a quality assurance program which is in compliance with the requi-rements of the USNRC 10 CFR 50 Appendix B that is documented in a QA manual which has been submitted to and accepted by Commonwealth Edison Co, Westinghouse Electric Corp., Tennes-see Valley Authorities, Philadelphia Electric Co., and Northeast !
Utilities.
O O '
nltt t
I 1
[
jj}f!h lei W [g
!!ll!il O
<O I
i il I
- a __ __ __
I L
, AOCA-ATOM '
TR UR 85-225 Page 31 1985-10-01
- i 10 REFERENCES 1. TR BR 82-98, rev. I
" Performance Verification of an Improved BWR Control Blade Design" Orjan Bernander, Johann Lindner l 2. KUA 82-115 "EPRI Control Rod Stress Analysis" Enrique Pallards
- 3. KUA 82-160 "EPRI Control Rod End-of-life Properties and Seismic Consi-derations" Q Bo Huggstr5m
- 4. KUA 85-145
" Millstone Control Rod Stiffness" Per R5nning Hansen
,OI 5. KUA 85-155 -
e} "KKP Control Blade Stif fness" A ( Per R5nning Hansen
'3 6. KPA 85-156 ja , " Impact of ASEA-ATOM Hafnlum-tip Control Rods on Tran-II l sient Reactor Performance in BWR-2/3/4 Plants" l Stig Andersson 4
( 7. BR 83-265 o [8 "B4C Temperatures in ASEA-ATOM Control Blades"
- El Leif Holm t2
{ lll 8. UR 84-115 l ((} "Hafnlum Temperatures in ASEA-ATOM Control Blades" Leif Holm
- 9. UR 85-032
" Reverification of PHOENIX" Torg5t Berling, Sture Helmeroson
'O
- 10. uR 85-054
" Nuclear Design Characteristics of ASEA-ATOM Control ,
Blade with Hafnium-tip for KKP-1" Jan-Erik Christiansson, Johann Lindner .
- 11. UR 85 055
" Gas Pressure Bulldup in ASEA-ATOM Control Blades" Leif Holm
- 12. UR 85-063 "KKP Impact of ASEA-ATOM Control Rods on 3-Dimen-I sional Core Calculations" Sven-Birger Johannesson l
k
- k
ACOA-ATOM TR UR 85-225 P:g2 32 1985-10-01 .
- 13. UR 85-092
" Nuclear Design Characteristics of ASEA-ATOM Control Blade with Hafnlum-tip for Millstone Unit 1" Jan-Erik Christiansson
- 14. UR 85-093, rev. I
" ASEA-ATOM Control Blades for BWR-2/3/4 - Service Life Limit Recommendations" Johann Lindner
- 15. EPRI NP-1974 (Nov 1981)
" Control Rod Meterials and Burnable Poisons" Prepared by S M Stoller Corp., N.Y.
O 16. Nuclear Technology 60 (March 1983),362
" Operational Experience with and Postirradiation Examina-tions on Bolling Water Reactor Control Rods" N. Eickelpasch, et al e
O 17. KUA 84-406 -
- $ " Control Rod Welds Between Absorber Blades and Filling 1l a Pieces"
- 3 Per R6nning Hansen, Jeffrey Smith
't a j l hI i Igl 1 lil a
lij 88 1IIf I nb O
O I
i ii 8
_11-----_------___------------
I N C.U W.e.#n m=~N
. -mm U p .
O O ~ =.::=:== == pU U
% .u_~ e.. = = - - -
)
e M
Table 5 OPERATNG EXPERENCE WITH ASEA-ATOM CONTROL BLADES AS OF NOVEheER 1984> 4 i 0 E
Plant Not Power Cycle No. of Gross Electr. Full Power Average Core No. Blados Production Hours Burnig W e) TWh mwd /kg U Oskarshamn 1 440 12 112 33.6 73 800 55.3 Oskcrshamn 2 570 9 109 33.6 65 200 58.4 l
l Ringhals 1 750 7 157 35.7 47 500 39.2 Bctseb5ck1 570 8 109 ~ -4 33.1 58 400 52.3 g W Barseb5ck 2 570 6 109 Y g 29.6 52 200 46.8 Es l TVOI 660 6 121 is O 26.0 38 900 36.5 TVO II 660 4 "hw 121 16.8 25 800 24.2 j Forsmark 1 900 4 161 24.3 26 100 24.6 Forsmark 2 900 3 161 18.8 19 600 18.4 '
.]
o
. k e
t
?
ACC A- ATO M TR UR 85-225 Figure 6-1 1985-10-01 .
AXIAL DISTRIBUTION OF B-10 DEPLETION A
UPPERMOST B4C HOLE g .%
e
.I lj 11 A 3 ~
CONTROL BLADE WITH HIGHEST.
15-BLAoe rip oEPLETioN
}
4 h0 (BLADE AVG = 9%)
"~
ji , AVERAGE BLADE DEPLETION
<- BLADE AVG = 21%)
n IE- '~
O .
O e
5-ed I ~
% B-10 DEPLETION (AVG IN B4C HOLE) o 10 2o 3O 4O 5O 60
, AOCIA-ATOM TR UR 85-225 Tabla 9-1 1985-10-01 .
ASEA-ATOM CONTROL BL ADES - Manufacturina Experience September 1985 Reactor Country Not Power Number of Year of Type of Output Control Blades Delivery Control Blade Oskarshamn 1 Sweden 440 112 (+ 4 spare) 1970 l
Ringhals 1 Sweden 750 157 (+ 2 spare) 1972 Oskarshamn 2 Sweden 570 109 (+ 2 spare) 1974 Barsebeck 1 Sweden 570 109 (+ 2 spare) 1974 Barsebeck 2 Sweden 570 109 1976 TVOI Finland Q Forsmark 1 Sweden 660 900 121 (+ 4 spare) 161 (+ 4 spare) 1978 1978 TVO II , Finland 660 121 1979
, Forsmark 2 Sweden 900 161 1979 a ,
Oskarshamn 1 Sweden 440 5 (replacement) 1979 -
Barsebeck 1 Sweden 570 4 (replacement) 1979 Forsmark 3 Sweden 1060 169 (+ 4 spare) 1983 g
R Dresden 3 USA 800 8 (replacement) 1983 4 CR-82 Oskarshamn 2 Sweden 570 10 (replacement) 1983 4g y Oskarshamn 3 Sweden 1060 169 (+ 4 spare) 1984 lglJ E K R B 11 C FRG 1250 2 (replacement) 1984
$l KKI FRG 870 2 (replacement) 1984 g KKP-1 FRG 860 4 (replacement) 1984 g l KKB FRG 770 4 (replacement) 1984 p Oskarshamn 1 Sweden 440 10 (replacement) 1984 CR-82 Ringhals 1 Sweden 750 4 (replacement) 1984 CR-82 Barsebeck 1 Sweden 570 10 (replacement) 1984 CR-82 Barsebeck 2 Sweden 570 6 (replacement) 1984 CR-82 O KKi fro 870 31 (,epiacement) 1985 22 CR-82 KKP-1 FRG 860 5 (replacement) 1985 CR-82 ,
Oskarshamn 2 Sweden 570 26 (replacement) 1985 24 CR-82,2 CR-85 _
Barsebeck 1 Sweden 570 15 (replacement) 1985 13 CR-82,2 CR-85 Quad Cities 1 USA 809 29 (replacement) 1985 CR-82 Millstone 1 USA 660 13 (replacement) 1985 CR-82 Subtotal: 1712 g
On Order La Crosse USA 50 20 (replacement) 1986 CR-82 l
j' KKK FRG 1260 20 (replacement) 1986 CR-85
,g TVO I Finland 710 36 (replacement) 1986 CR-82
-n- JXmWrnA . IUWA . v4 fit __-- -_____flArMmrriimid XM_____R_
(
l OVERSIZE l
! DOCUMENT l
! PAGE PULLED 1
l 1
i SEE APERTURE CARDS l i NUMBER OF PAGES: /
3 ACCESSION NUMBER (5):
1 26looso / po-og_
1 i
I l
l l
l
! APERTURE CARD /HARD COPY AVAILABLE FROM RECORD SERVICES BRANCH,TIDC FT5 492-8989 i
l l
l