ML20099F350
| ML20099F350 | |
| Person / Time | |
|---|---|
| Site: | Dresden  |
| Issue date: | 10/31/1984 |
| From: | Cowan R, Huff J, Kass J GENERAL ELECTRIC CO. |
| To: | |
| Shared Package | |
| ML17195A671 | List: |
| References | |
| NUDOCS 8411260383 | |
| Download: ML20099F350 (100) | |
Text
T
'NOV 13 REC 1 DRESDEN-2 HYDROGEN WATER CHEMISTRY PROGRAM
-~,,
4 Q}., AN ASSESSMENT OF PLANT OPERATION DURIllG INITI AL FUEL CYCLE THROUGH JULY 1984.
A .
ennt n . - nn ,930 NOTICE -
THE ATTACHED FILES ARE OFFICIAL RECORDS OF THE DIVISION OF DOCUMENT CONTROL. THEY HAVE BEEN CHARGED TO YOU FOR A LIMITED TIME PERIOD AND MUST BE RETURNED TO THE RECORDS FACILITY BRANCH 016. PLEASE DO NOT SEND DOCUMENTS CHARGED OUT THROUGH THE MAIL. REMOVAL OF ANY PAGE(S) FROM DOCUMENT FOR REPRODUCTION MUST BE REFERRED TO FILE PERSONNEL. g . y 7 11/9/39 DEADLINE RETURN DATE Y//4 003 //
f Iq ^
1
~
l EPORT PREPARED FOR EPRI RECORDS F ACILITY BRANCH TOBER 1984 l
0 i P
[
DRESDEN-2 HYDROGEN WATER CHEMISTRY PROGRAM An Assessment of Plant Operation During initial Fuel Cycle '~'
... EPRI PROJECTS: RP-1930-1 -
RP-1930-7 PREPARED BY THE GENERAL ELECTRIC NUCLEAR ENERGY BUSINESS OPERATION JULY 27, 1984 AUTHORS R. L. Cowan B . ' M. Gordon J. M. Huff M. E. Indig J. N. Kass W. B. Nelson J. P. Peterson J. M. Skarpelos L. L. Sundberg a -
d EEE1 PROJECT MANAGERS
'4 R. L. Jones M. D. Naughton J. T. A. Roberts s a v*
- b
r ,
TABLE OF COM2 TIS
- 0. EXEQTIVE SU199RY
- 1. INIROIUCTION
- 2. HYDROGDI WATDt GDiISTRY SPECIFICATION
.d
- 3. PLANT WATDt GDiISIRY OPERATIIG PRACTICE IMPROVEMDTiS
- 4. DRESDEN-2 EUEL CYCLE. 9 WATDI GDiISIRY PEREURMANCE
- 5. HYDROGDI WATER GDiISIRY IMPACT G1 OPERATI!G DOSE RATES, OOCUPATIONAL EXFOSURE, AND SHU'IDOWN DOSE RATES
- 6. EUELS PEREORMANCE AND SURVEILLANCE PROGRAM
- 7. APPD OIX A. TEGNICAL BASIS ECR HYDROGDI WATER GEMISIRY SPECIFICATICN
- 8. APPDOIX B. POLE OF IMPURITIES .CN IGSOC AND HYDROGDI WATER GDtISIRY- ARGCNNE NATIONAL LABCRATORIES SIUDIES 9 APPDOIX C. ION GRCMATOGRAPHY MEASURDENT PROGRAM 4
0 8
_, -7.__. , , 7 ,.~.- - - + .,,-w- - - - - - -- - - - -
O. EXEGTIVE SUNERY INIRODIK2IOS
'1he Dresden-2 IMR has been operating on Hydrogen Water Chemistry nince May, 1983 as a means to prevent Intergranular Stress Corrosion Cracking of stainless steel as well as prevent stress corrosion effects in other structual alloys. Materials studies have been conducted at the plant and in I
j the laboratory by several organizations. In addition extensive water chemistry and radiological performance data have been developed to assess the perfccmance
- - of Dresden-2 under these new operating conditions. 'Ihe key results of all of j ,,
this work are smunrized as follows.
~
4
- 1. Field stress corrosion and electrochemical potential studies provide strong evidence that stainless steel stress corrosion cracking activity at Dresden-2 has
- been arrested. A series of in situ stress corrosion tests were conducted on
- furnace sensitized stainless steel with the tests ongoing over about 2000 hours0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br />.
I No cracking was observed in either smooth or IGSCC precracked specimens tested l while the plant was operated at 20 ppb 0: oc less and 0.3 uS/cm or less.
_ Electrocnemical potential measurements on stainless steel unde at the plant for ever 3000 hours0.0347 days <br />0.833 hours <br />0.00496 weeks <br />0.00114 months <br /> revealed potentials well below -350 nfV (SHE) over 95% of the time.
l 'Ihis potential is considered to be a threshold for IGSCC for the conductivity values listed above. 'Ihese tests at the plant revealed some tolerance for oxygen adn conductivity values higher than those listed above. Coolant oxygen content
- excursion up to 200 ppb for limited periods of time up to about 15 hours1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br /> during the l constant extension rate test did not result in IGSCC. Electrochemical potentials
- during these short term excursions did not increase to the values observed under
) normal operation at 200 );pb oxygen indicating the presence of a memory effect.
! Presunably. longer term operation at 200 ppb oxygen would have. eventually resulted i in IGSCC. Perhaps the most revealing test result is the constant extension rate i
test conducted on a furnace sensitized and IGSCC precracked specimen that had seven
[ precrack phase was noted.
4 e
b 0-1
. - . = =.,L -.:_. x 2 = .: ._. _ . = , -. ==,; - - - . - - .
- 2. 'Ihe field test results are consistent with extensive laboratory test data whid have shown that no crack' initiation or growth on pre-existing cracxs '
is expected in sensitized ' stainless'stieel under Hydrogen Water 01emistry conditions. 'Ihis conclusion is based on pipe, crack growth, canstant extension rate, and other tests con &cted on several heats of material at several different laboratories, Tests also show that no cracking occurs below stainless steel electrochemical potentials of.-350 mV (HSE) if con &ctivity does not exceed 0.3 uS/m. Additional laboratory studies have. ,
shown that there is a strong interaction among oxygen, con &ctivity and coolant impurity species. Increasing canductivity requires decreasina j coolant oxygen content for prevention of cracking. In addition varicus ,,
impurity species have greater or lesser effects at given con &ctivities.
Sodits sulfate is the most severe species that has been identified out of a large rm**r of impurities studied. If only soditat sulfate is present as the impurity, no cracking is expected at:20 ppb oxygen and 0.3. uS/m con &ctivity or at 40 ppb oxygen and slightly less than 0.2 us/an con &ctivity.
- 3. Field water chemistry studies included extensive analyses of coolant oxygen and con &ctivity as well as ion chreatography to assess the impurities present in the coolant. Results of the studies show that con &ctivity has steadily improved as has the percentage of time with coolant. oxygen belcw
- 20 ppb. About 80% of hot operation time (time at above 20% power) was with
, coolant oxygen below 20 ppb and con &ctivity below 0.3 uS/an. A large amotnt of the remaining time was less than 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> where the mencry effect -
described above would prevent cracking. If this is taken into account then 93% of the time is accounted for.
t
- 4. In addition, a large amomt of the remaining time was for coolant oxygen 2
between 20 and 40 ppb but with coolant con &ctivity less than 0.2 uS/an.
. No cracking is expected for these conditions. Only a few htndred hours of l operatics were incurred where cracking would not be completely suppressed -
i and evet tnder these conditions re&ced susceptibility would be expected.
- I .
i 4
, 0-2
- 5. Ion diromatogralty studies of the coolant at Dresden-2 have shown that under continmus operation mnditions, no more than 30 Epb carbonate and no sulfate are present. 'Ihus the species making up the con &ctivity at Dresden-2 are far less aggresive than the soditan sulfate used to assess con &ctivity effects in the laboratory studies.
- 6. Additional in plant and laboratory studies have shown that stress mrrosion and/or corrosion fatigue of the other major structural alloys including Inconel 600, carbon steel, and low alloy steel are effectively eliminated by application of Hydrogen Water Gemistry. No effects were found for martensitic stainless steel. Scune increase in carbon steel general corrosion rates can be expected but resultant corrosion is well within design tolerances.
- 7. Addition of Hydrogen Water Gemistry at Dresden-2 has been found to add about 20 man rem, per. Year to the roughly 200 man rem currently incurred.
However, the advantages of red.tced exposure for IGSCE related repairs more than provides a compensating effect. For example, pipe replacement programs alone account for 1500 to 2500 man rem.
- 8. At present, no fuels performance data for a water chemistry exactly the same as at Dresden-2 has been develcped. However, based on available BWR and IWR data, no harmful effects are anticitated. A special fuels surveillance program has been instituted at Dresden-2 to confirm this with intial data becoming available after this fall's cutage.
The available information strongly supports that at Dresden-2 Nuclear Power Staticn stress corrosion activity of structural alloys, particularly sensitized stainless steel, has been effectively eliminated since the start of the current Fuel Ofcle with the Hydrogen Water Gemistry Flowsheet.
e 8
e I
I 1
l 1
0-3
- 1. IMRCWCTION Intergranular Stress Corrosion Cracking (IGSOC) of Recirculation and other pipe systems has had a significant impact on BWR plant . Iailability. Numerous remedies including alternate pipe alloys and stress improvenent treatamt that put the pipe welds in compression have been developed and implenented into
~
service. More recently, attention has been focused on Hydrogen Water Chemistry as a means to provide blanket protection against IGSOC for piping, reactor internals and, in fact, almost all structural components. We Dresden-2 BWR has been operating on IMC since Mar d of 1983. A great deal of information has been developed regarding operating plant chemistry, plant materials stress corrosion behavior, and radiation effects in the 17 months of operation at Dresden-2. 21s document will present the current mderstanding of this information and how it applies to Dresden-2 operation & ring the last Fuel Cycle.
Se operation of the Dresden-2 plant on the Hydrogen Water -
Chenistry flowsheet for a full Fuel Cycle is a pioreering step. Testing origirally done in Dresden-2 during 1982 showed that the concept of changing the reactor water swironment to provide mitigation against IGSCC was technically feasable and comnercially practical. Because of these very psitive restuls, the Connmwealth Edison Ccupany (CECO) decided to rtn the Dresden-2 plant full time on the Hydrogen Water Genistry flowsheet beginning with Fuel Cycle 9, which began in March l of 1983. In conjunction with the Electric Power Research Inetitute and
! the General Electric Company, an intensive monitoring program was
- initiated to demonstrate and quantify the benefits of Hydrogen Water i
Genistry and measure any ptential plant paraneters that could be l adversely affected, including long term Cobalt-60 in&ced radiatica l buildup, crud transport, fuel performance, plant radiation fields, and environs dose rate. his reprt quantifies the results and effects ocserved. During Fuel Cycle 9 together with a summary of related laboratory research.
1 l
l l
1-1
O 9
4 W
O O
1 9
[
l 0
2.0 HYDROGEN WATER GD1IS'IRY SPECIFICATION REQUIRDIDUS 2.1 INIROIUCTION j l
Se specification for Hydrogen Water Gemistry utilized at
'Dresden-2 during Fuel Cycle 9 has been established by extensive stress corrosion cracking testing in the laboratory and has been mnfirmed by testing in the recirculation water at Dresden-2., (1-9) The result of the
-~
General Electric labcratory and reactor testing progra are detailed in Appendix A of this document. An independant test progr a at Argonne National Laboratory looking specifically at the effect of impurities
~ '
on the requirements of the IMC specifications are detailed in Appendix B. @ e teachings of all of these studies serve as the basis for this specificaticri.
A schematic diagram of the objectives of the specifications are shown in Figure 1. 'Ite normal BWR range of oxygen concentration / electrochemical potential and reactor water impurity level lies in a region of susceptability to IGSCC. In Hydrogen Water Chemistry, careful control of water purity and red.ictim of oxygen content / corrosion p tential by feechater hydrogen addition results in creaticn of a chemistry in the IGSCC immity region.
W e
6 e
2-1
r 1 s
Figure 1 S
D.
Oxygen content or corrceson poteneel
- meresang TGSCCpoemow
% reen 'n some onwronment SWR range ~ L P
f IM poemme inctemng
- rese SWRH 2 N -
- Increamng raes IGSCC immundy repon PWRpnmery '
syneem range? "
A WEA
~-
of Conductiwr/ .
, seem e Schematic Summary of the Results of Laboratory Studies of Effect of Impurities on SCC of Stain-less Steels i 2-2
2.2 REQUIREMEN'IS EUR POWER OPERATION Table 1 describes the water quality requirements that should be met for power operation (>20% power) to meet the requirements of Hydrogen Water Olemistry. In addition, the verification measurements described below mst be made to denenstrate that conditions leading to IGSCC suppressicn have been adlieved.
Ideally. these requirements should be met at all times that the reactor is above ambient temperatures. However, feedwater hydrogen injection is only pssible when feedwater is flowing to the reactor vessel. 'Ihis constraint, ccupled with licencing requirements tied to the main steam line radiation mcnitor, allow injection only above 20%
power at Dresden-2. -Also, availability of the hydrogen injecticn system is not 100% and in addition there are also occasions when the He additica is turned off deliberately. The significance of time of oreraticn outside of the range of the specification is discussed in later chapters.
2.3 VERIFICATION MEASURDE2GS Because of pssible subtle variations in IGSC behavicr from one plant to another. it is necessary to confirm that the H NC requirements defined in Tables 1, in fact, mitigate IGSCC at a specific plant. As soon as the hydrogen injection rate to the feedwater is adjusted to satisfy the steam and recirculating water requirements given in the tables, the measurements described below should be made thereafter at or above 90% power.
- 1. Stainless Steel Corroison Potential
. . 'Ihe cerrosien ptential of a Type 304 stainless steel measuring electrode shculd be determined in recirculating water using the measurement methods described. For a fresh (lightly filmed) stainless steel electrode the measured value is required to be
<-350mV (SHE) * . If a higher value is found, the hydrogen injection rate should be increased until the corrosicn potential is .
<-350mV(SHE) .
The corrosica potentials are calculated on the Standard Hydrogen Electrode (SHE) from measurement against an Ag/AgCl reference electrode of General Electric design as described in Appendix A.
2-3
4 h e measurement is required after each refuelling outage.
(However, it is recomended that the stainless steci corrosien poten'tial be mcnitored contirmously for diagnostic purgses.)
- 2. Constant Extension Rate Test (CERT) '
A CERT should be performed on a weld-sensitized Type 304 stainless steel specimen in recirculation system water. Suitable test methods are described in references 1 to 9. A completely ductile , ~
failure is required in this test. If any IGSOC is observed, the hydrogen content of the feedwater should be increased by 10 percent increments and the CERT repeated until a completely ductile failure ,,
is obskrved.
%e CERT is required only once, following initial implenentation of Hydrogen Water Chemistry. It is recomended that the test be .
repeated periodically (e.g., once per cycle) to provide proof that IGSCC is still being suppressed during power operaticn.
4 e
4 e
2-4
r -
TABLE 1 WATER QUALITY REQUIRDENTS AND MATERIAIS EUR HYDROGEN WATER QUMISmY PARAME'IER SQUENCY OF M)MINAL MAXIMJM
, AND (UNPN MEASURDENT VAIEE (MINIMUM) VAIEE
. - REAC'ICR WRIER -
Dissolved Oxygen (ppb) Contirmous 10 >20, (<5) at Conductivity (uS/an @25'C) Contirmous 0.2 >0.3
- MAIN SIEAM -
Dissolved
, Oxygen (ppm) Contirmous 17.0 >7.0 I
6 s
2-5 i
References
- 1. B. M. Gordon, et al, " Laboratory Studies on Hydrogen Water Chemistry," paper presented at EPRI Seminar m Ccuntermeasures fcr BWR Pipe Cracking, Palo Alto,* G, November 14-18. 1983.
- 2. B. M. Gordon, et al, " Mitigation of Stress Corrosion Cracking
'Brough Suppression of Radiolytic Oxygen," Paper #50 presented at ,
the International synposita on Environmental Degradation of Materials in Nuclear Power Systems - Water Reactors," Myrtle Beach, South Carolina, August 24, 1983. ,.
- 3. E. L. Burley, et al, " Oxygen Suppression in Boiling Water Reactors-Phase 2 Final Rep)rt, DOE /ET/34203-47, NEDC-23856-7, October 1982.
- 4. M. E. Indig and J. E. Weber, " Mitigation of Stress Corrosion Cracking in an Operating BWR Via Hydrogen Injection," Paper #125, Corrosicn '83, Anaheim. G, April 18, 1983.
- 5. B. M. Gordon, et al, " Effectiveness of Hydrogen Additions for Mitigation of Stress Corrosion Cracking in BWR Environments," Paper
- 20, Corrosion '81, Toronto, Ontario, April 7,1981.
- 6. B. M. Gordon, et al, " Hydrogen Water Chemistry for BWRs - Interim Report," January,1981 - April,1983, to be published by EPRI.
- 7. W. E. Ruther, et al, "Effect of Sulfuric Acid, Oxygen, and Hydrogen in High Temperature Water en Stress Corrosion Cracking of Sensitized Type 304 Stainless Steel," Pap:2r #125, Corrosien '83, Anaheim, G, April 18, 1983.
'8. R. J. Law, et al, " Suppression of Radiolytic Oxygen Produced in BWR -
by Feedwater Hydrogen Addition," Paper 95 presented at the Third International Conference on Water Gemistry of Nuclear Reactor Systems, Bournemouth, U.K., October 17-21, 1983. ",
, 24 .
^
- = . - - - . . . . _ . - - . . .
- 3. PLANT WATER QiDtIS'IRY OPERATING PRACTICE IMPROVDIEN'IS 3.1 IN'IRCIUCTION i
Water chemistry limits normally used at haden 2 prior to the Hydrogen Water Olemistry program were based on federal regulatory requirements (Nuclear Regulatory Constission Te&nical Specifications) or fuel warranty requirements (General Electric Fuel Warranty Operating
.- Limits) whi d are adequate to meet safety concerns and avoid gross fuel damage. However, these **ahlished limits are much too high to attain the optimurn chemistry control mder Hydrogen Water Chemistry that will l *' .re&ce the probability of stress corrosion cracking of stainless steel -
components. More restrictive water chemistry limits are needed to 5 protect the Nuclear Steam Supply System piping and couponents from t
[- corrosion. General Electric strongly recossnanded the adoption of internal adninistrative water quality limits, which are more restrictive than IRC Te&nical Specifications limits, to be utilized with the Hydrogen Water N=iery program.
Important chemistry paraneters and the limits required to adieve
- adequate chemistry control are presented in Table 1 and 2. It was
, reconmended that plant management implement and strongly support
- - ~ adherence to the adninistrative limits.
i It is important that the con &ctivity of reactor water be i maintained below 0.2 us/m. Con &ctivity of Une uandensate-fee &ater should be very low. 'Ibe iron and copper leaving the Condensate Treatment System should also be very low. Come control of the corrosion pro & cts present in the feedwater and reactor water and close I
, control of conductivity in these process streams is necessary.
in BWR's during power operation and levels below 0.2 uS/m can be 4
+
Reactor water conductivities of less than 0.3 uS/m are adievable maintained by using good operational practices regardless of the types I l . or capacities of the Reactor Water G eanup and Condensate Treatment '
Systems. Figure 1 shows that over 60 percent of GE-designed BWR's j maintained reactor water con &ctivities less than 0.3 us/m for 50 percent of the time during their most recent fuel cycle. Over I* 20 percent of these plants were able to operate below 0.3 uS/cm for 90 percent of the time.
4 i
1 4 -
4 3-1 1
e --.--,n.n.n,--.-,n-.n,..,- ..,_- n .-- _ ~ n n .,- - n-,, n .-m-w_-,----~~_,-n- , - - - -
i Figure 1 ACTION LEVEL 1 ACTION LEVEL 2 1M =
0 +
! O +
0 +
0 + .
N -
g .
O +
0 + .,
so - U +
0 + .
C +
0 + -
LEGEND 70 -
g +
3 + 0 50% OF MEASUREMENTS LESS THAN INDIC ATE D U O + CONDUCTivtTY I O +
s 60 - + 90% 07 MEASUREMENTS O + LESS TH AN INDICATED I
Q g , CONOUCTIVITY 0 +
z so - a +
E O +
E O +
0 +
40 -
O +
0 +
O +
" O +
30 0 +
0 +
O +
- l 20 -
g ,
j
- O +
i l 0 + ,
10 -
O +
0 +
0+
I ' ' ' l ' ' ' ' l I I ' ' ' I 0
O 03 1.0 la WEEKLY AVER AGE RE ACTOM WATER CONOUCTivlTV (mS/cml me g na h
_. .-. - . - ~ .-- - -- . .-. . --
Dresden-2, during the previous fuel cycle (Cycle 8) , reported reactor water con &ctivity s 0.25 uS/m 50 percent of the time and f 0.42 us/m 90 percent of the time, placing it in the middle of the BWR population.
'Ihe primary indicator of reactor water purity is mn&ctivity.
Non-volatile imic impurities leaking through the condensate treatment system (or bypassing the Condensate Treatment System and entering the reactor through the Control Rod Drive System) are concentrated in the
..- reactor at some level established by the flow through the Reactor Water Ceanup System and the removal efficiency of that system. Under optime conditions, reactor water con &ctivity should be less than j 0.2 us/m, and reactor water silica concentratim should be less than 50 ppb during normal plant operation at power.
Silica entering the primary system usually is associated with
- condenser cooling water inleakage. Additional silica may be intro &ced l because of incomplete removal of silica by the plant makeup water system or by the use of diatonaceous earth in the ra&aste system.
Monitoring for silica in the reactor water and Reactor Water
- Geanup System effluent will provide valuable information on demineralizer perfcrmance. Manitoring and controlling silica levels will provide control of other anions in general. Since it is impractical to mcmitor all anions, reactive silica, which is present as
. 'a weak anion will if controlled, provide control fcr other anion impurities including fluoride and carbonate to some extent and certainly for chloride and sulfate.
Low pi conditions contribute to IGSOC and can contribute to increased general corrosion. While not exactly quantified, pH limits have been established to be consistent with the con &ctivity limits.
'Ihat is, if the con &ctivity limits in Tables 1 and 2 are met, the pH will also be under control.
! To insure that the water quality of the Dresden-2 reactor stays j ,
within the limits required for successful operation under the Hydrogen j, Water Gemistry program, the operating practices for the plant needed to te optimized.
o 2.2 DPEDW - 2 OPERATING PRACTICES To insure that the water quality of the Dresden-2 reactor stays witnin the limits required for successful operation mder the Hydrogen l Water Geistry Program, the operating practices for the plant needed i
tc te optomized. 'Ihis section describes the recomendaticns made to j cptcmize the Dresden-2 water quality related operational practices.
e 3-2
Operating practices at Dresden-2 were Assessed by examining plant operating records and cond2cting interviews with selected plant ,
personnel. 'Ihe objective of this review was to identify areas where operating practices and proced2res needed to be improved to achieve the goal of optimizing reactor water chemistry by red 2cing the input of corrosial prod 2 cts and other impurities to the reactor.
e 9
=
e e
t 3-3
Highlights of this review were presented to Dresden operations and chemistry personnel during an exit interview in the latter part of 1983. Bis was suppleented, early in 1984, by a written report.
A summary of the important findings of the review are listed below:
- 1. Demineralizer resin beds used in the ra&aste system should never be returned to the condensate treatment system. If resin beds are to be regenerated and reused, they should be dedicated for use in
,. the radwaste system mly.
- 2. A dedicated individual in operations should be assigned to monitor and supervise the operation and maintenance of the plant o
,demineralizer systems. Bis should include the demineralizers in the Condensate Treatment Systems, he Reactor Water Cennup Systems, Fuel Pool Geanup Systems, Radwaste Systems and the Makeup System.
- 3. A dedicated chemical engineer or chemist should be assigned the task of collecting and evaluating informaticn on plant water treatment systems. mis individual should prepare periodic reports reparting on the performance of the various systems and suggesting changes required to improve performance. Items that should be addressed should include:
- a. Resin Capacity (capacity before and af ter regereration if regeneration is practiced).
- b. C eaning efficiency (frequency of cleaning).
, c. Resin degradation condition (resin has a finite lifetime) .
- d. Resin loss and makeup.
- e. New resin quality.
- 4. Improve resin use.
- a. Upgrade resin specifications.
(1) Increase capacity requirments.
(2) Increase friability requirements.
(3) Increase reparting requirements.
," b. Evaluate the use of resin having an equivalent cation-to-anion resin ratio in the Condensate Treatmant System and
. the Radwaste System.
- c. Purchase premixed resin (equivalent caticn to anica resin ratio) for use in the Reactor Water Gennup System and in the Fuel Pool Cooling and G eanup System.
- 5. Enlarge and improve the wet chemistry laboratory facilities.
3-4
- 6. Improve the quality of chemistry measurements.
- a. Se;arate technician chemistry and health physics
~
functions.
- b. Minimize rotaticn of chemistry technicians.
- c. Increase techitician training.
l- d. Assign dedicated technicians to specific tasks.
- 7. Upgrade sampling and analysis equipnent and improve the maintenance of this equipment. ,
- a. Temperature control of samples.
- b. Temperature centroL of sample staticn environment.
- c. Process conductivity versus laboratory con & ctivity. .
- d. Local gi rather than laboratory pl.
- e. Sensitivity for silica.
- 8. Obtain equipnent for measuring organics in various samples.
Establish a program to monitor organics in ra&aste before it's
- returned to the condensate storage tank.
- 9. Perform a more detailed review of sampling, analytical and laboratory facilities, equipnent and practices.
e f
3-5
3.3 DRESDEN-2 OPERATIONAL GAhEES A number of changes in operaticnal practices have been made at Dresden & ring the hydrogen water chemistry program test. Probably the most important change made was the method with which Condensate Treatment System and Radwaste System domineralizers were operated.
Rackaste demineralizer beds were no longer returned to the Condensate Treatment System. All the beds in the Condensate Treatment
.- were replaced over a period of several weeks with new, regenerated resins. Se beds were prepared so that there was an equivalent mixture of caticn and anicn resin, as recxxmeded. When demineralizer beck are exhausted in the Rackaste System, they are discarded as solid waste.
We oldest bed in the Condensate Treatment System is transferred to the Radwaste System and it is replaced by a bed of new resin. Resin specifications were suggested and higher quality resin is now being used in the plant. A program for monitoring organic impurities in various plant streams is now being practiced. Other reconnendations are in various stages of study or implenentation. .
De net effect of these changes has been very positive. As is discussed in detail in a later chapter, the reactor water con &ctivity has met the desired 0.2 uS/an value for practically all of the Hydrogen Water 01cmistry program.
e e
I i
l l
3-6
--3 _ _ . . , . _ _ . _ _ _ _ - . - _ . - , , , ,
TABIE 1 REACIOR WATE QISIS'IRY LIMITS I i l I l PROGSS SIREAM AtO I I NORMAL I MACD l i AIMINIS'IRATIVE I OPGATI!G ! CPERATIIG I PLAlff OPERATIONAL CDIDITION 1 LIMITS I LIMITS I LIMITS I REACIDR WA'HR I l l l l 1 l I 1. 10WER OPERATION l l l 1 1 I I -
I CDi!UCTIVITY (uS/CM) 9 25'C l .2 1 1.0 1 5.0 1 I I I i QII4 RIDE (PPB) i 20 1 100 1 000
I I I i l SILICA (PPB) I 100 1 200 1 - . ,
I I I I i I 'IOTAL COPPm (PPB) I 10 1 20 1 -
I i l I i f i
- 2. STAR'IUP AND HCff STANDBY I !
I I I I !
I CNUCTIVITY (uS/m) 9 25'C I .2 l 1.0 I 2.0 1 1 I I I I QILCRIDE (PPB) i 20 1 100 l 100 i l i l I i 'IOTAL (DPPER (PPB) i 10 1 20 1 - i f I I f l 3. CID SHL'ITOWN
- I i i t I . . l l l l 1 QMUCTIVITY (uS/m) 9 25'C I 1. I 2.0 1 5.0 I I I I !
! CILCRIDE (PPB) I 50 l 100 l 200 !
I I I I I I pH 9 25'c -
I 5.3 'ID 8.6 I 5.3':08.6 1 4.9 'IO 9.3 I i 1 I l i l SILICA (PPB) I 100 1 200 1 - 1 I I I i !
I 'IUIAL CDPPER (PPB) I 10 1 20 1 - I l ---
-=======I l REACICR WATER CLEAEP EFFILGTf l I i l 1 i l l 4 i CDlIUCTIVITY (uS/CM) 9 25'C l .08 i .10 I .20 f.
- I I I I I l SILICA (PPB) 1 50 l 100 1 -
1 I I I i 1 1 '1 URAL CDPPER (PPB) I 1 1 2 l -
1
- 'Ihese limits also agly to fuel storage Enols containing fuel that may te reinser*ed into the reactor;.and to suppressica pool, condensate storage tank and makeup storage tank water which may be introd2ced to the reactor.
3-7 4 % g 99 % 49 4
TABLE 2 CDIDENSA7E/FEEDIATER OIDiIS'IRY LIMITS i l i i i i 150 CESS S'IREAM AND l I NOREL I MAXIMJM i l I ADlINISTRATIVE I OPERATIE l OPERATIE I l PLANT OPERATIONAL CCNDITION f LIMISS LIMITS I I LIMITS 1 l I I I i
i FEEDfATER l l i I I I I l I i i i l I
- 1 1. KMER OPERATION I I l .
I I I I i
! METALLIC IMPURITIES (PPB) 1 -
l 15 l 60 I I I I I
- I I IRON (PPB) l I l l l Insoluble i 2.0 1 10 1 40 1 i Soluble I .5 l 1 1 2 I I I I i l l 7 URAL CDPPER (PPB) I .1 I .5 1 2 I I I I I I i OXYGDI (PPB) 25 1 5 i
I 1 35 1 15 1 110 i 90 l i l i i
! CQiDUCTIVITY (uS/CN) 9 25'C 1 .060 I .065 I .10 I I I I l i I 2. STARHIP AND BOT STANDEY l i I i i l I i i
! CONDJCTIVITY (uS/CM) 9 25'c I .080 I .10 I .15 I i i i l I I SUTAL COPPER (PPB) I .2 1 1. 1 -
I I I I l i . .... . . ... ... ..
,,_ __l I CCRIDDEATE 'IREATMDR SE"IEM EFFLUENT l l l l l A:0 CQi'IROL FOD DRIVE SEl.L WATER l l I I I l i I I i i l l I I OXYGDI (PPB) i 25 5 l 35 t 15 1 110 t 90 l I I I I I I CDIDUCTIVITY (uS/04) I .060 I .065 I .1 1 I I I I I
! IPCN (PPB) i l i I I Insoluble i 2.0 1 10 1 - 1 i Soluble i .5 l 1 1 -
1 I I l l I
. I TUIAL CDPPER (PPB) l .1 I .5 1 -
I I I
. I I I
' _I C0tIDDEATE 1REATMDIT SYSTEM INFLUDIT I l i I 1 I I l
. I I i l CulDUCTIVITY (uS/Ot) 9 25'c I .070 I .090 1 -
1 I I I I 3-8
- 4. DRESDEN-2 FUEL CYCLE 9 WATER OIEMISTRY 4.1 IN20 DUCTION This section describes water chemistry parameter behavior measured during operation for one Fuel Cycle on the Hydrogen Water Chemistry specification. A statistical analysis of plant behavior with
,, regard to conductivity and dissolved 0, objectivesare discussed in detail. The identity of impurity ions are detailed and the behavior of soluble and insoluble corrosion prodicts with time are detailed.
e 0
0 o
O 4-1
i
- 4. CYCLE 9 WATER CHEMISTRY PERNRIWJG 4.2 CONIXJCPIVITY l
Reactor water conouctivity ano oxygen (alscussea later) were l monitorec continuously trom Novencer,1983 cnort.gn July 15, 1984 usiny a Keitaly Data Acquisition System ariven oy an IBM Personal Computer.
I For enis program cata signals were storea in memory at 5-nunute -
intervals, generaung 288 cata pints on a cally casis. Tne storeo inrormanon was aownloacea to prmanent alsx storage every 24 nours.
All cata nave Deen screenea cor outage perloos, oata pints oiscarceu, ~-
ano conpuea into tne statistical plot unown in Figure 1. Here, une ctanulative procaoulty or a reading Delow a given conouctivity (Y-ans) is plottea against conouctivity on a logaricamic scale (X-ans) . Tne geometric mean of tne data is inalcacea oy tne 50% 'value on ene 1er-nanu orainate; une stanoara oeviations on ene rl'gnt-nana oroinate are in increments or plus or minus 1 sigma acove or oelow cne mean. As or enis wnting, une geometric mean ror tne reprung perloa is 0.0105 l uS/cm; une reator water conouctivity nas oeen Delow 0.2 uS/cm 9n or une time. Tne station nas cleany met anu exceeusa enu goal ror une Fuel Cycle to maintain tne conouctivity oelow 0.3 uS/cm 90% or une l
time, witn a carget value or 0.2 uS/cm. Figure 2 snows enat a typical conta in 1984, rarlecting incorpranon or all Key operatiny practice recommenoacions, resultec in even cetter statistica.
3 For comparative purpues, cne average conouctivines ror Fuel Cycles 3-9 are snown Delow -
Ff f ET, CYM.R AVRDACP mMDUCTIVITTES 3 0.331 uS/cm 4 0.185 5 0.302 '
6 0.246 .
7 0.246 8 0.250 -
9
- 0.105 -
3-8 0.269
- Tnrougn July 15, 1984 l
4-2 I
t
. . .-.a. . ....c... . , . . . . ..
7_________ __ _ _
The data for Cycles 3-G represent weekly average conductiv.ities provided by Comonwealth Edison to General Electric as part of 'the Fuel Warranty documentation. While not as extensive as the data base for Fuel Cycle 9, the improvement in Cycle 9 operation is readily apparent. This improvement is not a direct result of the implementation of hydrogen water chemistry, but of a series of implemented recomendations aimed at improving overall water quality, which would be successful for a plant not operating on a hydrogen water
.. chemistry flowsheet.
4 e
m I
i n
4 6
l .
l t
e 4-3
SNO.'idIA30 Obt0N01S
=r m m - c - m m :.
n s s 1 f f* f f l l l l l l l! i -
ji j i 11 :!
l !
} ! .
, l ?
,,_ i 1 * .
6 lr !
l 1F r N ~Q [
3 = =
I .
1
~
1 l
$ v m r ' - -
W '
I1 '
Z l <
D >
ti ,
2 I i a i
- .fi I W k. I i j 2 J ~
C3 *C e
(
, t
, l
, i
! ' ~
i '
O l !l l 11 -t
~
L R-
{l l-l I !
! 6 1
2
- E
- 1 l k! ! f ll ll !' '
k
! MI i !! it
'l l
{l N !! I' ' :
! l l l N l' = !
O
- I l ,,
e -
t ,
3 N %
ll
!. =
1 O <
l l! b U ll -
' = ' .,
f l ! fx 2
> 2 **
- r ..
- I -
2 l' .
, - -)
t ,
t I, ,
2 I l4 l 3 .
ll c - 7 i
I jI! l 5 5
f $
1 I'
i l it r
~
o t/O OO oO OO OO O OOOOO O O Oo OO u & ~
m em m em o *"
te8 m
o*
.o no o o8 0 8o v oo o o o 9o
-o ec :o cm -: _ :: ::
oo do 9 8. 8 9 o o 8 o o9 o. o. 9 o. o.-c8. 9Oc o;;
co ;
c) em exn c) c) c) c) m o ooooooo -
m om oxn m e) c) c) c) c) a) o w m :r m eu .-o m l (1N3383d) A1I9I808CMd 3A: # .~V ~s I
k,,
b SNatle!A3C CbtCNels
, m m - o . c.u m.
I 1 I f i 1 ll l l l l l l l l l l ll i_ n I ll I I l !l II 5 i s l ! l ! il I i 3 i
v l l l,
I l ;
l M~
e L
e l i ~
=
ge
'i .
SI =
+
l i
( Q-o =
l C C
- E 1
rs r : -!
a <
- 5 ,
u
,y i c.
i, % >
- ) i -
l* E c O i C E
- - l % i i o i$ :2 ! % l -
,' lk
- e l
l 4 n<
c 1 ll l l ! !- "x x i !I I l I l M i l l 1 l t i I !
@[
l i li I I I i i MiI ! l ll Il
[3 ll l l !l l l l l l ll Il _! $
! l l l l l '
I Ee '
= h
, : a s i t I l i B u l
l -
!. ij.
I $ N
! l / l l 0 W i ii i I l 1 I 7 2 l
- EE= 3 iI \
c ii i i
=: -
e v
! l , ' '
A s
. i -a v I
r
!l l i 2 5 <
4: '
a w i
y x
>l , l .
h
- iI a- - -
j i i l -x
,l '
i 2 i : ! j l O
'l >
j i 3 l 2
- s
- I l'
l
!l l
i l
l; oo
! il oo o o l
o o ooo o o li :
o o oo e c. -
- e. c, . o. o. c. o. o. o. o. o. o. o. o. o. o. o .n .= =
em - m e t o oooe i.D o - tn -
7 0; Oe oo s wh CD 5 Q =* (9 f'J -
(IN300$d) A1IlI85505d 3?,Il0T Z O
.- ., n . .......w..... . . . . .
4.3 CORROSION PROIXJC BDIAVIOR Tne impimentation or nyorogen water enemistry in a BWh provices a more reoucing environmenc in ooca tne reewacer ano cne reactor water.
In ene reewater, cnis is primarily tne result or increa.wo cissolvec nyorogen mncentracions (1.0-1.5 ppt wica nyorogen acalcion,1-2 pe o witnouc nyorogen acciclon), wn11e in ene reaccor water, it is cne result or lower cissolvec oxygen concencrations. As a part or cne * -
enemistry monitoring program at Drescen-2 Ior Fuel Cycle 9, reewater ano reactor water corrosion procucts were monitoreo Irom July,1983, '
enrougn Marca,1984. In cae reewater, soluole ano insoluole corrosion '
prooucca were colleceeo continuously, wica sampes enangeo ac rougniy tnree-cay intervals. Eaca samge Iraccion was analyzeo at Vallecicos Nuclear ror Fe, Cu, N1, Co, Zn, Cr, anc Mn. . Rougnly tnree .reaccor water samples were collecceo eacn week. Tney were rirst analyzeo ac cne sace ror gamma-emiccing raaloisotopes, ano enen cransporceo co Vallecicos ror tne same enemical analyses as cne reewacer samges.
Tne cominant impurity in ene Drescen-2 reewater is insoluole iron. Figure 3 snows une concentration or insolunie Fe as a runccion or time ror cne Cycle 9 regrcing perloc. Tne 20 ppo spixes are cae result or several concensace ceminerallzer enangeouts over a snorc interval; cae 40 ppo spixe encompasses an oroerly snucoown ror ge e cracx inspeccion ano tne restartup or une reactor. No long-term aaverse consaquenccc or trycrogn accitiva are eviuent. Tne time-case gor.s ror cne ocner elenents cnac were analyzeo all snow sim11ar gccerns, wica no eatsaenc snowing an upwaro crena witn cime. Tne wncentracion or une otner insoluole metals is generally less enac 0.1 peo. Tne caole oelow snows cne average concentracion or insoAuole iron ror Cycle 9 ano eacn or une precealng 6 Fue1 Cycles.
FUEL CYrLP AVFRAnP TRON COMroNTRATION ( p_uo) ,
3 2.47 -
4 2.50 5 4.04 -
6 6.49 -
7 4.15 8 2.40 i 9 5.03 Tne Cycle 9 caca case is mucn more extensive unan ene Fuel Warrancy inrormation supp11eo ror Cycles 3-8. It snoulo ce noceo enac cae Cycle 3-6 caca coes not incluce scarcup ano snutoown sampling 4-6
, .,e ! . .. m 41- * , , - . l .. .'
aata. Ealmination or une scarcup ano snutoown samges f rom tne Cycle 9 aata case woula reouce tne average to less unan 3 ppo.
Or cne soluole species cnat were monitorea in ene reewater, Cocalc is une eltsaent or greatest concern oecause or its activanon to Co-60 in cne reactor core no accompanying potencial Ior raalaton oun aup. Figure 4 snows cae concentranon of soluole cocait in ene reewater as a runction of ume. Tne oownwara ano stacle trena or cooalt is reacuy apparent. Tnis creno pattern is also ooservea ror
.- une ocner eleaents unat were analyzeo. Because of tne operanonal pracuces enanges to une concensate creatment system unat were impienentea cunng Cycle 9, it cannot ce expilcitly concluoea unat cne crena is a result or nyorogen water caemistry. Nonerneless, unere appear to oe no awerse consequences or nyorogen water enemistry Ior soluole species in une f eewater. Cocaat monitoring in ene f eeawater is not requirea in une f uel warranty occumentauon, so enere are not comparaole cara Ior Cycles 3-8. Tne 7 ppt average ror soluole moalt is consistant witn General Electric cata trom otner aeep Dea plancs wnica are noc operating on nyarogen water enemiscry.
Monitoring or metallic impurlues in reactor water is not requirea in une General Electric Fuel Warranty Documentation. It is stui imperative to stuay tne time-oepenaent cenavlor or enese impurlues unoer concitions or nycrogen water enemistry. Figure 5 snows une concentracion or reaccor water :n insoluole cooalt as a runction or i ume. Tne aata points prior to Occooer, 1983 (rougnly aay 76) snoula
! De uisregarueu owing to samging altriculcies (lacx or aaequate ricw i races in cne sample lines) . Wnue enere is conslaeraole scatter in ene i 10-200 ppt range, une average value or 40 ppt agrees very well witn measurenents taxen at otner BWR's. .Tnere is no aerininve upwara or ,
oownwara trena in une aata ror cooalt or any of tne elenents unat were analyzea. Tne large sp1Kes occur at tne same cime perloos as une insoluole Iron cransients in cne teeawater (Figure 3) .
, Tne gamma counting or une reactor water corrosion procuct samg es
. revealea no unusual isotopes curing cne reporting perloo. Eacn sample containea une usual suite or activacion procucts, i.e. Co-60, Co-58, Fe-
- 59, Mn-54, Zn-65, anc Cr-51. Most spectra were comgetely aevolo or rission procuct isocopes. Or une activauon procucts, Co-60 was clearly tne most cominant isotope in cotn soluole ano insoluole sample rracn ons. Figure SA snows une concentration or insoluole Conalt-60 in reactor water. Again, oata prior to Octooer snoula De alsregaraea. As in ene plot ror tne insoluole cooalt precursor, enere are no ancications ror an upwara nor a oownwarc crena in ene concentranon or tne isotope. Values oetween 0.1 ano 0.2 uC1/1 are consistent witn aata t rcm otner BWR's.
4-7
.- - - . . . . . + . =:
, , . . . - ~ . . . . _, _ _ , _ , . . . . - , , , __..,--___._-,._________,___._.._m .. _ _.____. -,
Figure 6 shows the concentration of coluble cobalt in the reactor water as a function of time. The downward trend is readily apparent.
The concentration of cobalt has steadily decreased from 100 ppt at the beginning of the measurement campaign to 40 ppt.at the end of the sampling period. Downward trends were also obeserved for the renaining elements that were analyzed. Figure 7 shows-the concentration of soluble Co-60 in reactor water as a function of time. Here it appears that the Co-60 concentration is increasing with time, from typial ,
values of 0.1 uCi/l at the beginning of the sampling period to 0.2 uCi/1 at the end of the campaign. - This behavior is consistent with recent General Electric theories on cobalt transport, and does not ,
suggest that the increase is a result of hydrogen water chemistry.
Essentially. the theory predicts that with higher feedwater purity and reactor water purity, lower elemental cobalt would be expected.
Concurrently, with reduced cobalt input via the feedwater, dissolution of fuel deposit cobalt would provide a higher fraction of'the cobalt; thus, the higher Co-60.
Summarizing the corrosion product data, we have not seen any indication that hydrogen water chemistry has caused any detrimental change in the soluble and insoluble corrosion product transport in the feedwater or the reactor water. In the feedwater, insoluble spikes are common to all BWR!s, and clearly will propagate to the reactor water.
The downward trend in feedwater solubles, we feel, is the result of a gradual implementation of good operational practices centered around the management of the condensate treatment system. The concentraticns of the impurities in the feedwater and reactor water are consistent with data from other deep bed plants.
i l
l -
l 4-9 l
I
. . . . .Q y mas , . . - . . - . . . - a, e. . .
0
,4 2
M i A
R i0 0
G 2 O
R ON i
3 8
PI R ]
o t
i 0
9 1
E [
i6 ,
L 1 5
CBU 1 WO L S
i Y L
U N J 3 I
f 0 0 ] ,2 E E R [
C R 2E T 1 N
I U
G I
A S F
- W ,
S NEED p Y A
EF ,0 D 8
J S ,
E R ,0 J _
4 n i
- 0 0 0 0 . 0 0 0 0 ,
0 0 0 0 0 0 0 0 0 0 0 3 4 3 2 1 Eo.b ZO g<EZLoZOO a ZOr-c
i DRESDEb-2 HWC PROGRAu
} FEEDWATER SOLUBLE COBALT i O.06 i-
?+
- 0.05 -
m G
i b
Z O.04 -
E 9
t- f E
tc F-
. Z O.03 -
O i n ,
Z O f ,
O F- 0.02 -
l
_.s
< 0 cn
- O ,
O i t' O.01 - 1 0
0.00 i i i i i -i i i i , , i O 40 80 120 160 200 240 DAYS SINCE JULY 15, 1983 IIGURF 86
. . . e' ,
DRESD EN-2 HWC PROG 9AM
! REACTOR WATER INSOLUBLE COBALT
- O . 3 0 - -- - ---
?. \
~
0.28 -
a 0.26 -
o n 0.24 -
cn I
,, I O. 2 2 --
- t . v
) z O.20 -
- O Y F O.18 -
0.16 -
>'- z y 0.14 - ,
Z n O O.1:2 - n O
0.10 -
3 g .
< 0.08 ,7 cn O
i 1 O O.06 - U "
tb O.04 -
0.02 - U O 00 '" d "w'138 umu i i i. i i i i i , i O 40 80 120 160 200 240
. . DAVS SINCE JULY 15,1983 FIGURE S l
t.i 0
A N .
C e, m -
O 43 _hN CW ~
O$4
~
g 2 :o 10 O C O
m y -@ .
_, s- -
o CD "
~
C c eZ-
. -N O
a:
w u bb I l % M~
3 C- U Z e a >-
O _O O l L.] & to Qo4 d .
DC -
~O Q 4 _
C -
r i t i I I i I i i O
O 03 00 5 to O y eq N e O e O O O O O O O O O O O O O O O O O O O O O O, g g C, q q O, O, g g g C O O O O O O .O O O O (lW/ LOO) NOl1VMIN3ONOO 09-1lV800 z u. .. . . . . . .
. _-_ ; _ - - . u
' J"R E S J E b - 2 H WC P R O GIR A M REACTOR WATER SOLUBLE COBALT 0.G - - . - - - -
o
- 0. 5 --
E n.
Q.
O.4 -
y j Z
! 9 4
Q ti e x w -
y Z O.3 -
! O
- Z C
[
O H 0.2 -
Y ii m
O O
01 4
731gp .
Y 0.0 , , , , , , , , , , ,m ,
} O 40 80 120 160 200 240 DAYS SINCE JULY 15,1983 FIGURE 6 i
i
C
-4 N
h- _,
Ct O . .
O O w9 A 8
O to .
O
. W _.h .
I C "
/ 4 CD C
%/ 3 e 0 -f h Tw R
W O
_n y ~
h 3 -
d
=
m e l g -
- ~
Z8O 4 W< _O
=
O bwE (M -,
~
g :] .
c .
O
- G ~
74 O I i e i i i i i O (D 00 k to C 4 tra N r- O
- O O O O O O O O O O ,
O O O O O O O O O O O <
q g C, q q q O, q O, q O, O O O O O O O O O O O !
(lW/lon) NOl1V81N3ONOO 09-00 l
l i
'I
7: .
4.4 MEASURED OXYGEN CONCENTRATION At Dresden-2, primary coolant dissolved oxygen concentrations range between 70 and 280 ppb under normal operating conditions (i.e. ,
without hydrogen addition) . There is a power dependence on these values: the higher the power, the greater the concentration of dissolved oxygen (more radiolytic oxygen production) . When the hydrogen addition system is operating, the concintration of dissolved hydrogen in the feedwater largely determines the concentration of
~
dissolved oxygen in the reactor water recirculation system. For a given power level, the greater the concentration of feedwater hydrogen,
,. the . lower the concentration of dissolved oxygen in the recirculation system. The power dependence is also maintained. For a given concentration of hydrogen in the feedwater, the greater the power, the greater the dissolved oxygen concentration in the recirculation system.
The data base for the dissolved oxycjen measurements includes periods of high and low power operation, and periods when the hydrogen addition system was turned off and on. The hydrogen addition system was turned off/on many occasions for a variety of reasons; among these were maintenance activities on the hydrogen addition system itself, maintenance in other areas of the plant where there were high N-16 radiatien levels due to hydrogen addition, and fires in the offgas treatment system in which the hydrogen addition system was shutdown to extinguish the fires. In addition, during reactor startups and shutdowns, hydrogen addition is not used below 200 We (-25% power) .
From initial startup in March,1983 until the end of October,1983 the dissolved oxygen concentrations in reactor water were recorded manually. From November,1983 until the present, reactor water oxygen
~
concentratien was measured with an automatic data aquisition system at i five minute intervals. Figure 8 shows the cumulative statistics for the March through October period and Figures 9 through 17 show statistical p' lots of each individual month of data for November,1983 through July,1984. Figure 18 shows the cumulative statistics for 1983
, and 1984 and Figure 19 shows the cumulative statistics for all of 1984.
. The data presented in Figures 8 through 19 are presented in a Weibull statistics format. To find the cumulative time the variable foxygen concentration) is below a certain percentage, find the value on the "X" axis and read the corresponding percentage on the "Y" axis corresponding to the data point.
, A summary of the statistics presented graphically in Figures 8 ;
- nrough 19 is summarized in Table 1. The data is presented by month i
. 4-16 a
n,,,m,-., ,.,-e-.---.- -- -- - .-,,---
for cumulative time of reactor water oxygen less than 20 ppb, 30 ppb and 40 ppb. Also included are the monthly statistics from the plant data files showing the perceritage of time the hydrogen addition syscem was operating. When the injection system is operat'ing, it will.
significantly reduce the . oxygen concentration in the reactor warer.
This fact is shown when the statistics of reactor water oxygen 1
cencentration less than 40 ppb are compared against the plant statistics for hydrogen addition system availability. The values are ,
essentially identical for any given monthly time period and show that '
the over'all injection system availability is about 87%. During the
~
early part of the Fuel Cycle (March - December,1983), the injection .,
system, when on, did not control the oxygen concentration as well as it did during the 1984 time period. The 1984 statistics reflect hardware inprovements in the control system implemented in late 1983. For-example, in early 19.83 the injection system was "on" about 90% of the time but oxygen was controlled to less than 20 ppb for only 74% of the time. However, for the 1984 period the injection system was available for about 90% of the time and the oxygen was controlled belcw 20 ppb for 86% of the time.
It is worthwhile to investigate the nature of the time the reactor water was greater than 20 ppb oxygen because of the corrosien potential " memory effect" presented in Chapter 3. This effect shows that if the oxygen level rises to 200 ppb, the corrosien potential.
stays below the critical value for IGSCC for 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br />. The statistical data base shows that of the cumulative time, the reactor water was above 20' ppb, greater than 70% of that time was less than 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> in any given day. This analysis implies that of the total cumulative time elapsed since startup of Dresden-2 in March,1983, over 93% of the time has been in the corrosion potential region that is " immune to IGSCC".
j' Figure 20 shows the cumulative data for oxygen concentration in the main steam phase for July,1983 to July,1984 is similar to the reactor water oxygen. This data was not recorded automatically and -
represents one data point per day. ~
o I
i 4-17 .
e
c TABLE 1. CUMULATIVE STATISTICS FOR OXYGEN 00NCENTRATION IN REAC'IOR WATER DISSOLVED OXYGEN (PPB)
~-
CUMULATIVE % TIME BELON % TIME HYDROGEN
._ INDICATEQ yALUE . _ -
INJECTIQB
<20 <30 <40 AP/ MAR 83 88 UL83 87.9 AUG83 74 80 82 93.6 SEP83 96.8 OCT83 89.4 NOV83 70 76 80 80.7 DEC83 52 70 78 82.6 JAN84 91 93 96 91.4 FEB84 75 77 78 79.1 MAR 84 86 92 92.5 93.4 APRS4 91 93 94 93.3 MAY84 89 93.5 94 95.3 JUN84 85 89 89 90.7 JUL84 91 92 92.5 .
ALL 83/84 78 85 87 ALL 84 86 90 91 9
_g..,w.. - ----w-,--.w.,,.,y.. . . - - , , , , , - , , - e-
SN0!1dIA30 GbOCNUIS m m - o . m. _
4 i , . :
i i I 4 i s ;
i 6 : 6 i .
~
i ! I l I i i I l I C
obo w l ! ! l- !
5-K I l
I e.
6 '
1 a
- =
T
, i *
- lll x
, i 3
m r- ^' '
W a .1 i
, - x
- u o 8 i -
2 e 4
, , I i .
z -
w o
m
- D 1
W
=
a ', -
W o < ' k- =
1 - .
l . = w
. , . . , , , s , , ; <
1 i e i , 6 . . i i ,s . . .:. ur
+ 4 4 i i i i i i ,
i w , Ex i I !
i i % 6 -
- l l l l I1 1 8 I I L
i l l I l l 7%=> l i SE xw lo l l ,I Oc l l ,
e j >--
. au m l -c c, <
o CO r
u 01,2 a '_ %
-u H&
, i , , . . . - %<*
gr
, , , ,,,,,,,,,4
. _ g I
i l I ! t 8 r i ! t i I L
- % Eq l i e i l I i i i i
! t O
-- ?O
-z.
=d I I l '
l i i i ! i :
l ,
l I l , 5 m' I
l S e e
1 l i
o 3
- 3. o
. u.,
i I j , j i
i -
2 -
- o. o. o. o. o. o. o. o. o. e. o. o. o. o. o.
e e m e o e o o o o o o o o -
c3 mm e o e s o e a m cu -
4 (1N2003d) AillIG5203d 3AI.; 9h?:
--mm1 m p- m-%-- -p myy ---m -
$NOIl0IA30 GEOCNd1S
- m m - o - m m --
I e i i t ' t ! t 1 ii i t 4 i i I i i e i i i ii ;
i !
I II ( i I ! l i l I i 1 I t I 'g-l t I i i i i i i i i i i t ii ;
ii !I I I I I I i 1! ;
I ll Ii l l l l ll ll l l l ll Il ! d c_
i ll 8 l
1 $~
.. 3" l !
lA l l l ii r _.
=
I i I I e
! l ! l l -i .
! l 9 6 m l e r rw rI Z w .
W b b i l -
b s ,, , . i , i , a- ,
i , i 4 _, i ., .! x
= 3 ! i i ii ii i i I i i i l i !i ii i 3
$, ,, ! , ii ii ii l 6 E i , i i i i;il i <
e
- L -! ' I i! l l l l l I I l l l ! 1I I i
l l,
l l! I! ll l l l l l l ll ll ll ! 2 w
l ll ll l l l kll '
l lI l !l l $
E i !Il k i l llI .
m I t x i
li.I l e
i ! I I t t y e i ' l [ t %
e .
l f > L
' l E
i b
- e i .
i . 5
< ' *i ii ! t -i
! ! il ! ! ! ! ' ! +* i O
i i i l i ! I il !I i l 1 l i i l l I \l f I I !I i! I g W ii Ii i l l l ll l II l! I i! Ii II i
? $
c
- . t i, .
.. ., ,, =-
l i i ! l 3
. ll 4 l
3 I l l i I l li
! l $
! , I l l l ll ! I
-o c
, ; . , l , , , ,
I l l! !
l !
l i l ! !! ! ! I '
S ,=' 35
. r: :. m o o v, o o S SSSSSSS SS o o o ooooo c o o c o om
- 2
~~ :.o id oll ci 5$
c>
i c:d ddidiE5 c, co :~ w c =r m ca di -
(1N3083d1 Al W EOS^ide 3r . :~a ..
SN0IldIA30 GbtCNd1S
- r m eu - O m ., =
i e *
! ! ' ? f f f i i l I a i ii t i ! l . !
I i
,, i i . , , , . . ,
1 I I l 4 e i i i iI l l l l l l 1 i l i '
l l i i i i i P
, ,l '
i ll It
'~
l',;-
\.
i i i ,u 4 i 6 - - '
l l I I e I 1
- 1 8 f i i i I '
i.d t' l I !
5 i ! t I l 6 i i1 5 =
i 4l l i !i li e z - , s
\.,i g = l~ l\l l -
d 2
a l Il 3:
. I 3
~
o < I, ! i i =
i 71 lt -
j i =
1 ,I **
i , i 0
, i , i , s ,
t ! t i e i i i lk i & ~ 3
' ! ! ' I I ! 8 f1 f 'F 5 i e i 1 L i X I i i ! l l !
' ~
i i i ! i i i I i i 1! l i il i ! ! I i ll l l ll iI '
Te il I l l I l l i ll ,
' f z
\i i :
I,
,> w
=
I w
, v i i -
n a
- , c w
~ ~ ;7
, , , i 1 i . . !
- , i ; . e 6 e i i N * .
l ! I 'I ! ! 4 6 g ! ! $ i ! %_ N '. O
~
8 i i i G Z i 1 ! l ! ! I r ! i !
! ' i l l l I I I I i l i i l I i I ' - , ' r --
- I !I i i i I l !. I i ! * ! 5' A -
l I
h i i n l
1, ,
1 I
-i
' l f i ,
I i 1 _
Cn tno 00 OO O O O O O OO O O O O OO OO l'^ ~
O C)(D ao OO O O O O OO O G O O O OO -O^- 0 ~1 0-
- D 0)O OO AO O O O O O OO O O O O Of -
b b e3 OO bb O c>
b 00 1.f,5 a
O b se.
O b b b, :br bn .0 v -
b ~
(1N3053d) A1I118090hd 3? .!"d '
7-SN0IldIA30 CEdGNUIS
-r m ca - o - cu m I l 8 4 O
4 , i i i i i i i i f i i i i . >
5 1 e t 8 f i ) l l 6 l '
I t iI t I i l4 I I Il Il l l l 1 lI l iI ll l l l! ll !
l l!
t
- e. a.
Q C0- --
l1 1l
- l i I, l5 E t v i I
nm i ..
=._-
h , @
. at . - -
e i e 6 i Qi I
- 1 A
" l l s i i ! e 51 i L i 5 i ie i i i ai ! I i '.
- C
_ 6 l l l !! ! L l l l l li '
E 5 l ll ll l l l l l l l ll l* 5 h
5 l ll. l l N
, (/7 _
1 1 '
e sJ w l j' I l f 6 l *
t e
?
l I
i i
5
' O E I I CC
! ! ' C q
! . . i i t ' e i 1 e e e -
O 3
' ! ! ! I I 5 1
! I ' '
, i# # 1 i e i 4 i iI { t 8 i >
l, If iI t I i l ! I i l I l f l I I k
I ii iI j i l i I I1 :
I I il ii !I i
!l !!
4 4 ll ll l l l1 i l '\ **
II II I! t I 5 l
l l ! 1 I II
- l !
3! !! !! ! -
2 i I !!
5 m i !
t l h :
. t * '
I I t' '
t t E t I
C e e
. e i e e
! i ! . e j h ~ ~ - L -
, , i # i 6 s e i 4 r i e i i NIrc - 4 - E Id.
6
! l ! f I ! I f f I a 1 1 l i } { i l _E _
. _! !l ) j ! f I l I t l l l l l l l l4 ? i $b
!, i i i i, -
.! i l' 1 I I t i .!
I i i l ! i _.
i i i j i g j i< !. !l
- i. l i
l I'
i
!! ! ! ; ,I l '
- > DC CO ") O O O O O O 'J O O O O O OO : D C- ~
TO OO O O C C O OO O O C O OO C.' -C
- , m7 . ". 7<
_7 O 1, O O C O O O OO O D O O O Lfo -C CL
! k$ !!
2 m: 2:
b $>
= O mO b bbbb bb b 1.b -
w swozm to -
(1430d3d) MIM880td 3':Z Yn"~
I SNOI1 CIA 30 06cCNc1S m ca - o - n. e e i l I ! ! ! f f i i i ! I ' 4 t i t i i e
~
i j ! 6 i l l I I I I I e
~
i I I I I l l l l !
l l I I l l !' ~
- l. l l, 1 o n cr w-- r l ,~
l l 3 I I -
o i 4 i
+ , v.
i i a
3
- e 1 l i
i
_ [
, f
,k i 1 3
, g l #. l
- U I ! ! I d
{
m ! I l l l I J
' 2 x u-e - . =
x a -.
O aC ! "
i e '
i N l l.
P.-
~
1i _
, , , . m ,
iL e i e i i e -
f a f l I l 0 ! l I ! ! T' I f j f i i t i i i ! t 4 e i T
- I I i i l i l i iii i i 11 i -
i l l i
! I I l l I l l 1 l lMi a, $
z l 1 i i i i 1 I f ! ! I f I 'E O I ==
l l
w.
^
9 l .00 0
- l I I N
(! .
Z -
I E ..
- ~
j
= .
s e 8 t I ,4 e i e t *
- _ C e p i j jl ii ! i * - -' "
i I , I ! l l l 0 t ! I i E
t i I i I i i i l i ! I '
i i ,# -
ll i i l I l I i i !
i '
l 3- -
l l l l ! l !l l ! ! l' : 9 l -
. ! I i ; t .
1 I i l !
l l l ll , .
, i ! t ,
t 4
m .n o oo o o o O O O O C o o o 0 o .C, -:
-' ~
=
- a. n. a. n. o. =. a. o. o. o. o. o. o. o. o. a. .
c7) Tm mm D o o oooO ~> a o D -
og m c) c) J) O CO O .N f.D A = M CJ -
(1N3053d) AiI7I9dSO&s 2*:_r-' ~~
SNOIl0IA30 GbdGNd1S
- m cu - o - cu m
e i e e O
! ! ! $ ! $ I l 0 6 i i ! ! I . . m;3 4 e i I 4 l e I i i 4 6 6 l l 4 s 1 O
t i I I i i e i I j i i i e ie ~
6 i i i I ' 1 I i 1 i l j ! l l 1 Ii i l i ' I ! I f i i i li II !
ll l! l l l I l l l l ll l} >
li=;m lI Il l l ll
- ll ll i
' -I--
- e. l ll llTll l i !; ii : i I l l
. =
z:
i ;
.! : , i >
=- ;
gI l i '
W
.- i ;
=
i' i e t
- i 6 1 i ia e h d s i a 6 ia t de a e ii i l e i e i
m , . . i .
.8 i e i i , ,
e
~
h i I i t l l bI l l? l1 '
5 m ,
1 l !I l l 81 lll l l ll !! <
u g
a w
i ;, i f
i i, =
- , I i i u
$ f l l; ! h ' C 2, , l l s
l I
l l j
l ll -
l l
X
! I J
i ', 1 i ;
I T
i 6 ; i !' l' '2 3
. , , . s , ;
. t ! . . .
. ! , s . . . -
p 8 1 a , i { 6 i , g g Q, e e ,, i
, i 6 6 ' 6 i f i i e i i i R ! il a L
i l ! l I ! I f I l l l } h l l i l i lI t Ii i i ! l lli l l il !! li Z
ll ll i !I I!ll l i(I l ll li '
, i i i i i .
,I I 1 i I l l
! . I u
ten
!l 6 = =
b I j OC <j, '
i G -
l . . . , , ,
. t i e 4 , ! t , e t i ~ _ _
f 8
!
- I i! ' 8 8 ' ' ' ' ~ -
t t i e l t i j i i{ l l 1 i i t 2 _
I i i i f I l l l l l 1 1 I ! l6
[r l! !l l i i l l l l il ii l i !2
,: i ; i i j i
! r, 1 I '
! lI l l. I f !! ;! !
l;
' i
! I I i ! l i ! : i i lii
~
! ! ! ll ! l ! ; .
! ' ' i 1 i ! ;- ;
- ~ ~ ~ ~
i
.I -
- c aO O O c O O oec o o o oo
- - -- .".O OO C C C QC OC O .O OO c0 : /. -: ~
- c/ -
TC AC O O O C C OO O O O O OT -O LC 7 T' ! ". ! k!
= x ,b, bbbbbbb b Lb ~
- == = m m w nwwrm ~ -
fin 3053d) All'I M Dec 3AI. ~~^~~
SNOIld!A3G GbtCN01S m cu - C -
i c.u .~..
, , , I i 1
i i ,i , i i , i i . i a i i
t i i i e t i i ! i i j ' ' ~
l l
' ' l 1 1 I I '
l l --
i i i# i l i i i
l l ! I i
o o mg { l l l I ' -
I !
, l -
h t L
4 i i ! ' I i i . .
E =_
- i i e ! i i < r m t s -
w ! ! i i i ; ,
U b h l l i f i b r
z
= , 4 I i! ,
a
! A. I i
i ll l
E d - l Ii !
~
5 e i ! I i ,
i i l j . '
1
] l i
i i
,! e
! ; 1 t
l (m ,
l 3 - ::c
=
, ,: +; , , , , i .s . . 2 i e ! a t t ! * ' I t 'X t i .
! l ! ! ! !
I I I i t 14 2 l I I i i l I ! l ! ! l I 3 '
i , ?
i ll l 1 I ! I I I \ ! 1
- 2 l lI l !lll l l 1i' ; i ,
e I
I i. ..
T. 4 E
l %l , E
% i i :
i ,
S w
i - 3' 1 f l
(
' I r ) , _ _
i a e ! l 9 9 l t I t l l i i 8 ' 4 I ! ! l [ ,
l 'I i j l l t ! I 1 1 ! I i ! !
l t ! I I ! l ! t i i ! i ,
l l !! l l l l ! ! I - -
- ~ .
f 6
I iI i _1 I
l l llll l l I i
I ! i l l l e oo oo o o o o eoo e o o o a : -
9 9 9 '; - 5 -
- o. 99 e. o. o. 9 c. c. o o. o 9 3 $5 SS $ 8 ER3535S S " -
w e:ese u:':=cecs: r.--
J
SN0Il0IA3G CWCNOIS
r m m - o - cu m
5 i B 6 Q
. . , , , i i , _-
t .
' i . i , , =
- e. .>
i , i #
6 i i i ! ! t 4 , ,,
i:
6 , i i .. , e i . , , i i .. , -
t 6 i i i t i , j 6l t #
lI l l l I l l l l l l i ll l l l l l l ll '
-l _ . l
- m Il l l ll ll l $. :.
.- l I l !! r =
l ll l -
c t
l l
. =~. Ob- _,
a,
. . i , . . i .i i i i. 4 .
2 -
I l ii l q l i f f I I I I i i
^
w i i ! 1 i c i ii I 6 si g
- 6 l ll !l l l 11 I I I i6 i
[
E 2
! Il ll I1 l$
l1 1I I C y
}
f c fd U lll i l '?
l
- l i, I i ii =
y
!. I. I, I I
, ,l i. e i 3 l
s !t l !
v i i
. 1 i .
I I .
i c _
l 1 i , . , s , , ,
~
g
. . , . ! t i ! ! E i e i s
- I i i6 e i i i t ie i i ! t 5 t j . e - i 1
i i! ! + i I 9 l t I e i l i i Tf I I ' i i.
ii i i I i i i ! l l l l l % 8l ! ! >
k It il l I i l l l i i I l i 1 l ; i , r
! ! Il Il ll l ll ll 1 G
l ll it. ,
i i
I R 1, ! , -
I li lf , y
+
l i x
I
- I ! O i
I l ,
. r e
, I .
! 2 -
' l f ;'
~
A
. -l i
! ! j!
i i
V _= - _
. _e _ -
o e i. . . .e i ii. t -. _ _
i i i' ' ! ! ! ! ! I I I i ! _
! 'I i l l 1 l I l l I i i ! ~ s i l ! l l l l l l l l ' i 1 -
l ! l I'
~
l ' ' 0 II i ll l l l l l
. l l3 I !
- 9 1 i +
, i I 8 l1 w
~> I l 1 E
! - i . . . i ,
l l fl ll l l lI '
!l !
l l i
i ,
o o o oocco e oo oo o o o oeooo o r .c. : oc co oo : :. :_ _
- om e. co eo c ,f. -r
. N; ; -: #.' O. 90 999 . 9 C. 9 C. C.
O^ ~0 0; -
r O DO
.- r.
% 0)
. =z O J, e
a a o o cpoeeee S Q 4.D =r m c (b c
e -
(IN3Cdd ) Al[l[EU3C d 2 ' . II'* ' ~
- up
SN0!1tIA3G GECCNt19 en ca - o - c. --
9 f 1 ? 1 g i ! i l ! I f { e i ! j 9 g i i t ! i , . , . .
1 1 1 I I ( t i i , e i h l l I l t l l l l 1 e l l l l }! l l li l !l >
1 A. --
i e ,
i .
(o ,
l B e i e ; *
~
I i i '
N I w , I
- h. l l f I z g -
3 5
g 5
n l l '
l'
(
l $ l ; } W s
a ,
c < l I --
l 5 l l l ll 1
l l .
d -
l 7 l ll _
x i i i i . r N , -
i '
! i i 3 i i. ! E e e i -
L i i i t i t ! i1e *i i ! I f ! l ij i 11 i i C
! I ! l l t i i i i l I i '
5 l li i i l l l l l TI l i l l l l l l l1 I!
e=
y i l l
- E t - r.
l , :
- 1 i ;
= 7 1 i -
i s ! 1 -
i j i . . . . . - ". . a *
} ! 1 ! I t I ! ! i i i !
i 's -a 1 ! l I ! l l ! t I 8 1 ! ' - 5 .
ll 1 l l l l } i l t ! I I i ~~~b c 3 -
[r -
1 i i l l ! I i l !i l l l ll l ll l l
5 i
=
1 l -
l 1 I l ,! ! ,
l f f' l l ! I
' l ~
o 50 0 o o o o o $0 0 7o ro0 o
Ek oo
=c -
- o. c. 0 0 e. c.
- m. e. m. o. c. o. o. o.
k M M
@ OO Q Q Q @ CD 5 4 .D :r '
.) 'u -
(1N3C33d) Al[l[8DkC'dd 3A [.I'^ "
9 SN3IldIA3C CetCNOIS m cu - o -
a cu e
m e
O t
i f f f 2
. : i i . t : 5 1 ' I i i i e ! l 4 ) ! l l j , g O
~
I i i t 6 , i l i t ! t i i e i i ! t i f ! I i i ! ! l I l l l l l l l l 1 i i I .
I ! l l l I l I l l' l 1
l i i ;
YD. i i l
- . v =
- ! +% ,
~
- i E l _ =
.- i i .s .
c 3 ; - -
3 i , , i . 6 i . ,, ,
, e l 6 e i j t h 4 ' i l e i e ! is i6 4 h I l 1 I shi ! l ! I l l '
g C 5, I t ! i e I ?l I i i l ! i i i ! =
b l l ll l l l Ol g l l l l l l! '
b x l '
h
- i I l
! l- ll -
a
- 6 ,
W u
- j l
f g ! ~
! i O < l h 6
c%
- l P l i I l l i p
t i !
ij l x
' 0 l l !! l ,
l ll l . 4 i l = e s _ E i . si , , , - -
i e i i e i i it a e i i ! ..
i e i e i e i I i t i l i 6 i e i t e 3
f ! ! I i ! ! ! l ! I l, i -
I i i i i i i d
a e
li ll ! l l l lh l ll !l I
~
l ii I L.:
I'i I[
l
\
l 5
, ! l . ;
, i , s. -
i e
! r !
- 6 e g
1 i i +
i
_~-a - -
i 1 1 ! ! l l 1 i ! l } i h I i .
1 i 1 ! l I I ! l l l l l i %c ' ! - ' 'r
~
I l l l i l
l l l l 1
l l T_
l.
I I
. I l. . .
' I I
! l l T f, l 1 I l l -
! ! l l l I i ,
+
i j I ; !
.l
- 4 I
I i 1
i 1
, e n p ,
- ) Q Q Q Q Q Q Q
._y * ,
~ -
[ 3 O O O; C, O O, O, O O O 2 - '
~L
- c , s = = ;,1 = Ja= 2 =Po ru
-= = x -
l : -}, m L L % ~
(1N3)D35) di .' 3U33 D S 3*.7.9$
SN0tidIA30 OddCNd1S 3 M N ~ O ~ N M a e a f I f f - .
I ! ' 6 3 I i ' I I i .I l 6 e i I I i i i i
~
l } i l i ' i . i I l ll t l l l l ll !l
- I oc.
m ll ll 1' ' +
i l .
e *
, 'j -
~
w s e
A %
i i 66 I i I '
- 3 7 N r - - i . 6 i . 2 -, - ' '
y I i ii ! I -
$ h l I l$ li f -
z -
3 x l ll -
-l l1 I ' -
2 m i C ll l
w > i c T =<
i 4
'J) w e- I j I . !. ~
h Y k l If I C -
l l\ ' t t. ~ ~:
}
4 ii X l !> > 2
. i i _ _
1 f i e f(%
r 1 f l* I O
~
- i 1 e
i . t i % i i i e ~l l ji i! i! i ! t t i ! ' !Ti i 'i !
d1 i i e i 9 I I I I I I t % * * ' O a I ll 1 I I l i i ! I 1i i I ' E*
ll l' i l l l l I ll ll !I ' 5 l ll ll l l l l l \ ll II i!
==
U li li l
i l !t i '
ii l
l1' -
i
=C i . - .
=0 i e ,l l
2 -
il t
ex
,' l! - _
_W k
- c =
- i f i ! ! % ~
7 *
- i. . . . I 3
! . ! i h -2
!i iI i e ! ! ! I T' = A
~
I - 1+ l l '
l l l ! I 'A -. J ji -, *
~
1, f I I I i j ! i %. .
- fl I ~ ~
l l ! l ! l f! 5 ll l
l l ll l! l' l 6 l i E '-
1 I I t t O
I t s =
i 8 h I 1 i, i
l1
,! I .,
i i
! l , t i * -
Cs JC OO OO OO O O O OOOOO O O O OO CO CC :.* ~
C1 CU) WO OO 3 CU) CC) U) O OO OO O O OOO O OOOOOOOOQO v O O OO O O c' Od CC ' . -
Ow: -O -]
C-O J
CMT) ~O : :: ;
'D. CT.C) . . . C). O. @. O. O. O. O. O. O. O. Q. O. O. O. O. O D.
O) CD) '")C) 0) m CJ C) 4 O O OOOOO O O l.0 m CD CI'; @CD CT. C) C) C3 C) C1 G) XWW2mN'~ C l
(IN3333d1 AlI1[Ec90ba 3e:1F#3 l l
1
1 l
SN0i10IA3G Cd00N01S l
l
=r en cu - o -
t cu m =
4 9 e o
ii A
i . I . i 8 i
i i
i i 5
.. I o i .. . . i i ! i i ,i i i i i6 i i i ii ,i !
ll _
i ie ii i ll l ll ll l lo d ' ll ll ll l @ e;
< = +
.- l <
.. i c
l I r=-
c2x
\:
l' -- 1
- 1 I cu -
~- -
' !O N H m1 _- D <
x ~ e i . ,, i. I ut i . . 4 . ! I i. . g r m r
I I I ' I ' ' ' ' ' i' I :
u i il i! i i ! E ! . i i ! i 6 e ii >> i W
- 6 l t i i t j
[ l i j l l l l t l lt l y 5 m i !! !I i l l !!I I Il If II i 8 l
- ! Il li l I lll l ll ll l
=
m cr a
1 l . l , e
= < l l >-
t I i l x j i l i Il l l C
i i
l i i
8 l
a t
' l ; I .::-
i i l ll l <T i,
e l.
. , s , , ,
3 L-I l 1 l i - C i e i I , i I ! i i1 e i i %
i is i i I i 6 l t ! I i i .
l i t t 6 e 7
v %.L l t s i ! i i i ! I I I i 1i ! I ti ! ax
! ! ii il ll l ll l l ll ll l' !i l %
! I II II !l l l lll k ll ll ll ! e$-
If
[ , l' ! 8 I *W l.,
I t .
j l l I ll l u<
%v I
ll ! O
' ii i i l, '
r ce i
l '
-C l
l {4 11
, - i i + , ,
! , . . I e . . l !h .
i i i 1 S
, i e 6
! I f . i e I T*
, . i iii! 6 i e i & 2-
- j t + 1
! j i iiji i i%L _' _~. ?
, I I l l i i i I Ii i! '
5 0 "_
j j if ll I I 4 l
e I. I' i .
i '
- 7 u
j!
i l i ' if
! . t l ll.
l i
+
i il
! i ,
I l i i ! . ,
. m os oo o o o ooooo o o o oe or cf -
SB L7 0 oo o o o OoOC oo o o oo O i.D ~; O
". m- Y.' ? ? . ".
~
.- ce em mm. me oe tn o o ooeoo o.
cr> o cc o e in = m eu -
o e -
(1N3363d) A1I1I8080dd 3.a.Il~10a~3
- S. A E
r i SNOI10!A30 GBUCNd1S m cu - o ca f 6 i , _.*
! i i e i i i i i i i i ;
l i I i i i i i t i i -
i i i i # i
! i i
- l. I !
t '
I i r r ! O 7 i i 6-i i i i i ==
- - u l ! eI -
Y - 2 C
oo n 3 C f 1 I :
N r C w l .
- = -
a
-u g j
- i e.
r U _
a ! i i i : ~ -
g :
= - i 4 l l l 6 l - x o
I t Al i I {
i ! E7 w - I l l 1l ! l ! !
l l
l l l l hl i l
i
}[ -
l [
l I i ! l l IN_ i C i
, i . , u I l ) -
X e cr
- j 00 l , 6
- rr
=
l ! -
{
n >-
t
==
l :
i <
i 1 l l t ! - E '-
e e i . i , e i . i c
i I 6 i i ! i i i i i O e-i i i ! ! ! I I i i i !
$ i ,
_ nb i
I J l I l l l i i t i E._ ...M l l l l l l l l I i i - ..
E' .
i i t i l i i j j ,
~
Cu ii l , t i I i i l i l , ,
2_
~
i . . . .
I l . ! j I %
l !
l I i ! , -
-?. -
l, l
=
- o a
^
f I ! '
=_
i l
' I I N I ,
f!
i i ! l l l _.
- e. o. o. o. o. o. o. o. o. o. o. o. o. o. e o -
o c) en e o o o o o o o o o o -
e a c) o c) co o e o ::r m ,u -
(1N30d3d) A1I7IStS0hc 3 ail 0T#3
- 5. W C IMPACT ON OPERATING DOSE RATES AND OCCUPATIONAL EXPOSURE 5.1 ENVIRONMENTAL DOSE RATE MEASUREMEN'IS AS A FUNCTION OF HYDROGEN ADDITION One of the critical effects of the Hydrogen Water Chemistry (WC) program is the effect on environmental dose rates and personnel
,, exposure. The increase of volatile N-16 was expected with WC, resulting in a significant increase in the steam lines dose rate.
During the May 1982 W C test at Dresden-2,2 steam line radiation was monitored (Figure 1) and a single point was measured near the reactor building for environmental dose. Further studies were needed to estimate the effect of this increase on property line yearly dose rate.
During a period of 12 months, starting in May 1983, Applied Process Technology (APT) under contract to EPRI made environmental dose rate measurements at 30 locations. The net dose rates (with H -without Hz) found by APr are shown in Figure 2 2,8 as a function of distance and direction from the Dresden-2 turbine.
A series of measurements were perfcrmed by General Electric during early June 1984, to determine the levels with and without Hydrogen addition, and ten days later, measurements of the " background" with both cperaticnal plants in shutdown as an independent check.
Twenty-four locations were measured under all three conditions and varied in distance from the D-2 turbine from 400 to 2800 ft. During these tests, the Dresden-3 unit was shutdown and there were no contributions of N-16 and C-15 to the dose measured. The other
- activities from radioactive waste and long-lived activation and fission
! products, however, were still present and contributed to the dose rate.
The background rate represented the combined man made, cosmic, and terrestrial radiation.
The primary measurement instrument was a Reuter-Stokes
. Model RSS-lll area monitoring system.' The RSS-Ill consists of two
- modules:
- The high pressure ionization chamber, and 2 a readout / power supply system. It is completely bat.tery operated, and thus, completely portable. It has an upper limit of 500 uR/h and a sensitivity of 20 mV/uP/h. The readout is a strip chart recorder with auto ranging 0-50 and 0-500 uP/h and a digital light emitting diode (LED) readout. A l single point is plotted every two seconds and the chart speed is 4 i
- r../hr. , allcwing recording of individual readings. The actual l
5-1 l
l
. . . .~_ . . -- - - . - . .- . . . .
measurements of ionization chamber current are made with a sensitive electrometer (20 mV/uR/h) . The ionization chamber is.a spherical stainless steel shell filled with ultra pure argon.and provides a flar
. response from about 0.5 meV .to greater than 8 meV.
The measurements were taken over a single day for each of the conditions, both .with and without hydrogen (hydrogen flow rate of 50 scfm) at 95. to 100% full reactor power. The shutdown measurement was .,
taken 25 hrs. af ter the actual shutdown. These 24 measurements , agreed with the data shown in Figure 2 to an outstanding degree. The-few anomolies are due to movement of sources which affect the gross .-
. readings on a single day and subsequently the net values.
The effect of EC on the ALARA (man Rem) of Dresden-2 is an important consideration. .The reported total incremental dose rate (man-Rem /yr) from NC will result in an increase of 10 ma'n-Rem / year.'8 The man-Rem /yr will be a small increase when compared to average normal-BWR exposures of 1,000 man-Rem (2300 man-Rem, Dresden 1981) .
Conversely, WC could offset the need for a piping replacement effort, which is known to cost in the range of 1200-2400 man-Rem. The small potential . annual increase thus is very ALARA effective when compared to the known impact of IGSCC.on plant operations. ,
The dose rate surveys and gamma scan at Dresden-2 have been measured at five different locations, all on the 517 ft level of the drywell. The dose rates on the surface of the pipe and the gamn:a scans at distance from the pipe are measured at exactly the same locaticn
- each time. The dese rates measured are shown in Figure 3 as a function of effective full power years (EFPY) . The gamma scan measureements are i shown in Figure 4.
O 6
l l
1 5-2
- _ _ > .. ~_. _. -.. ..
i-5.2 (R$ CIRCULATION LOOP PIPING DOSE RATES y
Over the. last ten years, General Electric has evaluated the dose rate., increase with operating time for a number of BWR Units.
Basically. this. study 'was undertaken to provide an understanding of radiation buildup and control. Using this data as a base, calid.
comparisons can be made within the BWR consunity, and predictions of
--.=- radiation buildup can be made. In evaluating the effect of HWC on radiatdon buildup in the recirculation loops, these developed measurement techniques were utilized. A gamma survey and scan were performed in' March 1983, prior to the long term use of hydrogen.
- _ During short outages (24 hrs) of Dresden-2' in November 1983, and again '
. in June 1984, gamma. surveys of the recirculation loop piping in the drywell were made at that time.
- The dose rate surveys are taken with two different instruments, one a portable ionization chamber, Eberline ROSA, and a shielded directional probed Eberlire HP220 probe coupled with a digital readout,
- -. normally referred to as Directional Probe Gamma Survey Measurement
! (DPGSM) . The RO5A measurement reselts in a measurement of the -dose rate at a point unshielded and affected by surrounding sources of
^
radiation. The DPGSM'results in a dose rate at the pipeL surf ace and shielded from extraneous sources. Both of the measurements of dose are obtained at the surf ace of the insulation or pipe (i.e. , contact readings). The dose rate instrument' a tion is calibrated before and
.after each test with a' source,whose calibration is traceable'to NBS.
- The measurements are corrected by any ponlinearities and bias found in the-standardization.'
7
/
, ;s The gamma scans are performed with a highly shielded and
.- collimated IGE solid state detector, coupled with a portable battery
- . cperated multichannel analyzer. The total system has been extensively
" ' ~
calibrated to allow the conversion of count rate in a gamma peak to i i microcuries per square cantimeter on piping. The conversion allows j* co}rection for the pipe thickness, diameter, and ccunting geometry. A j st.andard source is used to check the calibration before, during, and l; af rer each series of gamma scani.
r
)
r
\
1
- 5-3 '
i.
+v.__.._.__....___.. _ _ . - . . _ . - . - - - _ - . - . - - - - - - . - . . - - - - -
F The dose rate surveys and gamma scan at Dresden-2 have been measured at five different locations, all on the 517 f t level of the drywell. The dose rates on the surface of the pipe and the gamma scans at distance from the pipe are measured at exactly the same locaticn each tbne.. The dose rates measured are shown in Figure 3 as a functicn of effective full power years (EFPY) . The gamma scan measurements are shown -in Figure 4. ..
The increases in dose rate and Co-60 activity _evels with time are not higher than would be expected in a non-hydrogen operating BWR. The ^
pre-hydrogen levels in Dresden-2 were lower than most domestic BWRs and at longer exposure than most. The slope of the increase from initial startup to the prehydrogen was less than.40 mR/h/EFPY. The increase in levels between March 1983 and June 1984 have a slope of 38 mR/h/EFPY and are certainly within the statistical variation of the measurements.
The June- 1984 measurements included a section of new pipe which had been replaced in the reactor water cleanup system heat exchanger system and a measurement of the pipe (not replaced) adjacent to it. The measured dose rates with the DPGSM showed 435 mR/h on the new pipe and 531 mR/h on the old pipe. The old pipe had seen an exposure of 8.64 EFPY and the new pipe 1.06 EFPY. This result was expected since with
- HWC conditions specific activity of Co-60 (uCi/gm) had increased by a factor of 3-4 and the corrosion film produced is thinner by a factor of 4-5, resulting in counteracting effects. It is concluded that HNC dces not increase dose rates and/or activity levels above their normal pattern.
(
l 1
l-9 l'
5-4 v
REFERENCES
- 1. " Oxygen Suppression in Boiling Water Reactors - Phase 2, Final Report," NEDC-23856-7.
- 2. JTA Roberts, R. L. Jones, M. Naughton, " Mitigation of BWR Pipe Cracking Through Chemistry Changes," Araerican Power Converence, Chicago, Ill., April 1984.
- 3. ~ E. L. Burley, "BWR Hydrogen Water Chemistry Radiological Monitoring," EPRI Hydrogen Addition Project Review, bresden Site,
,. April 25, 1984.
- 4. W. M. Lowder, D. D. Raft, G. DeP Burke, " Determination of N-16 gamma Radiation Fields at BWR Nuclear Power Status," HASL-305, May 1976.
e O l
l l
5-5 1
FIGURE 1 I
STEAM LINE RADIATION 3000 -
4- d .,4 .
s' / $ ~
L/
/A o
d d" g /e '
f'D
< + ..
(a
- _~~;.vQ a/a[-
e2000 -
24L 5 / _*~
s -
at ,I /
e 6 4
c: /
I C I 2 EXPERIMENTAL DATA:
p /
/ A STEAM LINE RADIATION z1000 -
FIELD AT 95% POWER 3
- O STEAM LINE RADIATION h / FIELD AT 83% POWER O STEAM LINE 16N ACTIVITY AT 83% POWER (scale x 0.1) i l
-- - - - a @ RECIRCULATION LOOP RADIATION FIELD AT 83% POWER
' 0 ' ' '
O 0.5 1.0 1.5 2.0 ,
H2 CONCENTRATION IN FEEDWATER (ppm) -
l t
l l .
G FIGURE 2 NET HWC CONTRIBUTION TO DRESDEN ENVIRONS DOSE RATES (pR/H)
N
,. 2 I
3 e
, .f '
\
i \
u 7 9 l
}
r I , N s ,/w + , ,
i iO o o - n- -
i 2. ! ( 3.1 \ ' 7 ." *: 1
/
. \ % < 9" ,2 / l l
~'
\
x .
=3 l /
/
l
. \
\
% 4
/
- /
\ / ,
( .' i 'N / j
/
x /nNc i
l
[
l 9 - . , - - - -- .m, s ,~.--y
FIGURE 3 AVERAGE RECIRCULATION LOOP DOSE RATE DRESDEN-2 s.
- f. 4C0 -. -*
\
a E
d
& 300 -
a 2
a o
G 200 -
M C
E i- 100 -
4 0 - * * * * . s , , ,.
O 2 + 6 8 1C EFFECTIVE FULL POWER TEARS l
l 4
I h
l e
?
FIGURE 4 AVERAGE RECIR. COBALT-60 ACTIVITY DRESDEN-2
,, 26 24 -
1
~
. et 32 -
2 c 20 -
.\.
U 18 -
@ 16 -
. I t I- 14 --
a
$ 12 -
0 U 10 -
r.
c 8-4
- ~ 6-
< 4-2-
C , , , , ,
C 2 4 6 8 10 EFFE."TIVE FULL POWER TEARS O
e i ,
e
6.0 .
N ELS PER NRMANCE AND SURVEILLANCE PRCGRAM 6.1 DRESDC-2 REL PEREPmMG Fission product release measurements are used to estimate fuel-failures in a. boiling water reactor during plant operation. ' Station chemists routinely measure six noble gas isotopes '(Xe-133. Xe-135, Xe-138, Kr-85m, Kr-87, and Kr-88) and five radiciodine isotopes (I-131, I-132, I-133, I-134, and I-135) being released from' the nuclear core.
A recent Eet of measurements made at Dresden 2 are shown in Figure 1.
The release rate has_been normalized by dividing the release rate of
- each of the six noble gas isotopes and five radioiodine isotopes by its . ,,
fission yield.and decay constant. Since the normalized release for
, these -isotopes is nearly the same, we can confidently state that there is no failed fuel in the core. This pattern was observed in -
essentially all of the samples taken during a 57 week period of Dresden Fuel Cycle 9. It is characteristic of a recoil fission source estimated to be 2.6 E 12 fissions /sec supporting the generation and release of the. measured fission products.
Recoil release is observed from uranium impurities in the fuel cladding and plateout of fissionable material on the surfaces of the core. A 2 ppm natural uranium impurity level in the cladding will yield about 30 microcuries/sec of noble gases (sum of six isotopes) .
- . This is the order of magnitude observed in new reactor cores.
! Figure 2 shows the noble gas release (sum of six isotopes) during Fuel Cycle 9 at Dresden 2. A least squares fit of the data suggest that the noble gas release was about 6500 microcuries/sec_ at 'the beginning of the fuel cycle and has been decreasing. Recent values j below 4500 microcuries/sec suggest that the fissionable material source is slowly ' decreasing by either burnup or removal by the reactor water l cleanup system. .
, we can conclude that the recoil pattern observed during the period
' April,1983 to June.1984, as well as the decreasing noble gas release, provides evidence that no failed fuel exists in the Dresden 2 core.
The fission products being produced and released are from a small amount of fissionable material plated-out on the surface of the core from failed fuel present in prior fuel cycles.
l 10-1
6.1 DRESDEN 2 FUEL PERFORMANCE Fission product release measurements are used to estimate fuel failures in a Boiling Water Reactor during plant operation. Station chemists routinely measure six noble gas isotopes (Xe-133, Xe-135, Xe-13 8, Kr-85m, Kr-87, and I-135) being released from the nuclear core.
A recent set of measurements made at Dresden 2 is shown in Figure 1.
The release rate has been normalized by dividing the release rate of each of the six noble gas isotopes and five radiciodine isotopes by its
.. fission yield and decay constant. Since the normalized release for these isotopes is nearly the sane, we can confidently state that there is no f ailed fuel in the core. This pattern was observed in ~
- ^
essentially all of the sanples taken during a 57 week period of Dresden Fuel Cycle 9. It is characteristic of a recoil fission source estimated to be 2.6 E 12 fissions /sec supporting the generation and release of the measured fission products.
Recoil release is observed from uranium impurities in the fuel cladding and plateout of fissionable material on the surf aces of the core. A 2 ppm natural uranium impurity level in the cladding will yield about 30 microcuries/sec of noble gases (sum of six isotopes) .
This is the order of magnitude observed in new reactor cores.
Figure 2 shows the noble gas release (sum of six isotopes) during Fuel Cycle 9 at Dresden 2. At least squares fit of the data suggest diat the noble gas release was about 6500 microcuries/sec at the beginning of the fuel cycle and has been decreasing. Recent values below 4500 microcuries/sec suggest that the fissionable material source is slowly decreasing by either burnup or removal by the reactor water cleanup system.
' ' We can conclude that the recoil pattern observed during the pericd
, April,1983 to June,1984, as well as the decreasing noble gas release, provides evidence that no failed fuel exists in the Dresden 2 core.
Tne fission products being produced and released are from a small amount of fissionable material plated-out on the surf ace of the core f rom failed fuel present in prior fuel cycles.
6-1 A. --% *y.- -* - ---.,,,--.m -w p-% , - - - - .r -- - - . - -
l 6.2 FUEL SURVEILLANCE PROGRAM #
Hydrogen additions to the water of Biling Water. Reactors ,(EWRs;.
creates a water chemistry condition which is'outside the experience base of either BWRs or Pressurized Water Reactors (PWRs) . Therefore, it is prudent to ascertain the effects of the altered water chemistry on fuel performance and surveillance program has been established at l Dresden-2. This General Electric /EPRI program is described below.
Four bundles (LTAs) of carefully characterized fuel components .
have been inserted at tre beginning of the first cycle of hydrogen addition. Zircalays with a known, precharacterized range of corrosicn i
. behavior representative of the range usually found with cladding ~
batches used in BWR fuel fabrication have been selected for these bundles. The preirradiation characterization report (Subtask 1.1) will qualitatively specify the expected corrosion behaviors of the rods and spacers constituting the LTAs. In addition, discharged fuel bundles
- having been exposed to various combinations of normal BWR water chemistry and hydrogen water chemistry cycles are also included in the
! test matrix.
A combination of site and hot cell examinations will provide [
required information on corrosion and hydriding characteristics of Zircaloy-2 and -4 fuel bundle components, and on crud depositien
- characteristics of the test reactor plant. Components examined will be urania fuel rods, gadolinia-urania fuel rods, spacers, and water reds.
Table 1 outlines the proposed fuel surveillance program.
As shown in Table 1, no examinations have been conducted prict to
- the first cycle of hydrogen addition; however, one bundle of three-cycle discharged fuel will be identified for future examinaticn.
- Af ter the first cycle of hydrogen water chemistry conditions,e ;
xaminaticns will be conducted on this bundle (designated as 3/0, i.e. ,
3 cycles of normal BWR water chemistry and 0 cycle of hydrogen water chemistry) , a 2/1 discharged bundle, and a 0/1 LTA bundle. Poolside
~
e:: amination may identify gross changes in the corrosion behavior due to hydrogen water chemistry, but Zircaloy hydriding infomraticn will not e e i
available until completion of the first hot cell exam in mid-1985 -
, (refueling outages are assumed to be on an 18-month cycle) . After cycle 2 another LTA bundle (0/2) and another discharged bundle (1/2)
, will be examined. Af ter cycle 3, a third LTA bundle (0/3) will be available for examination, such work is considered optional at this
- time.
6-2 I l
l I
TABLE 1. Fuel Survel11ance Cutage BWR/HWC Cycles Hot Cell S!te Examination Exam Carments 0 --
None None 1 3/0 1 Bundle 2 U, 1 Gd Discharge in Oxide Cycle 0 Visual 12 U, 4 W Crud - 4 U, 2 Gd 2/1 1 Bundle 2 U, 1 Gd Discharge in Oxide ycle 1 Visual 12 U, 4 Gd Crud -
4 U, 2 Gd 0/1 (LTA-A) Bundle 2 U, 1 Gd Replacement Oxide 2 0 RM fuel required Visual 8 U, 4 Gd 4 Spacers Crud - 4 U, 2 Gd 2 1/2 Same as 3/0 ------ -
Discharge in Cycle 2 0/2 (LTA-B) Same as 0/1 LTA - - - - - - - - - - - - - - Replacement -
fuel required l
O e
d
e FISSIONS / se:
e- e- o e 1
w w e 1 i s
d
%=
x m
c, 0 I131
& Xe133 1~
3 m-rn
.m -
2-
_ m v- e _ ;. O I133 m- :'
T
.< = -
^
- .- Xe135 5 x> ;
-EV -
, O I135
?>
5> K E5m 3
,- Kr8B g
-m e 1132 m-c s
._- Kr87 5 3 1134
, n s: , _-
W
- . ..- c r.
Xe138 13 h
. , 6 em wi tw
II.000 R = 6500 - 35.5X
- X = WEEKS FROM BEGINNING OF CYCLE 3g,gog M," R = RELEASE (pCi/sec)
-4 9,000 ..
d .
PJ g 8,000 . ,
"YD m 7.000 e x . . s -
3 e..
at 6.0m ~ s. ey . ,
of .
do ex~..e
- 3; 5. -0 -
.., p. -
4, 0 _
,% 8,#
b ast Squares Fit Line Ea 3.w0 .
2.000 -
1,000 1983 1984
.1A FE i A e AP i HY JNo it? iAll iSEiOC iNOiDElJA nFE aMAe AP nHY aJNi]LtAU ISE #0C i No DE F1GURE 2. " SUM Of~ SIX" FiSSIOtl GAS RELEASE VALUES D U R iflG THE HYDROGEft WATER CH Et1I S T R Y PROGRAM I
APPENDIX A TECHNICAL BASIS FOR HYDROGEN WATER CHEMISTRY SPECIFICATION 0.0 NCUTIVE
SUMMARY
OF PRINCIPAL FINDINGS 0.1 The initial work to quantify the HWC water quality specification
~
consisted of investigations designed to clearly demonstrate the improved IGSCC performance of Type 304 stainless steel piping in a hydrogenated BWR environment. Four heats of 10 cm (4-in.) diameter Schedule 80 Type 304 and one heat Type 316 Nuclear Grade stainless steel were evaluated in the nominal BWR operating environment [288'C (550*F) water with 200 ppb dissolved oxygen] and in simulated HWC environment (initially 50-70 and then 20 ppb 0 )2 . The metallurgical condition of the pipes was "as welded."
Two types of pipe tests were used: (1) initiation controlled pipe test (10 tests) , and (2) propagation controlled pipe tests. (4 tests) . The initiaticn pipe tests investigated piping in the as-welded condition with no intentional pre-existing flaws, while in propagation control tests were intenticnally flawed (cracked) in the nominal environment and then subsequently tested further in the hydrogenated environment.
The results of this first series of pipe tests indicated that both low dissolved oxygen and low conductivity are necessary to eliminate IGSCC. It was also determined that IGSCC crack propagation in precracked piping can be terminated if the levels of oxygen and conductivity are low enough. These
. tests suggested an initial HWC specification of 15110 ppb dissolved cxygen combined with a conductivity of <0.2 uS/cm.
Tc supplement this task, three additional pipe tests were perf ormed to determine a f actor-of-improvement (FOI) of HWC over the nominal 0.2 ppm oxygen environment. These pipes have demonstrated that HWC (low oxygen and cnductivity) prevents crack initiation. For example, in the nominal 7-4
r-environment (200 ppb oxygen) cracks will typically initiate in approximatai 300 hours0.00347 days <br />0.0833 hours <br />4.960317e-4 weeks <br />1.1415e-4 months <br />. However, in HWC no cracks have been identified af ter approximately 7,500 hours0.00579 days <br />0.139 hours <br />8.267196e-4 weeks <br />1.9025e-4 months <br />,. indicating a FOI of 25.
Two additional pipes, each precracked in 200 ppb oxygenated water for 242 hours0.0028 days <br />0.0672 hours <br />4.001323e-4 weeks <br />9.2081e-5 months <br />, were exposed under constant load in the nominal and HWC envircnments to determine the effects of cycling the stress on pipe performance. The pipe exposed to the nominal environment failed af ter approximately 1300 -
hours of exposure, while the other precracked HWC pipe did not fail af ter an additional 5310 hours0.0615 days <br />1.475 hours <br />0.00878 weeks <br />0.00202 months <br /> of exposure. .
0.2 Crack Pronacation Studies This task, which consisted of three separate subtasks focusing on icw cycle carbon steel fatigue tests, fatigue crack growth investigaticns of stainless, carbon and low alloy steels, and constant load crack growth tests of stainless, carbon and low alloy steel, was characterized by a dual 4
obj ective: (1) determine the effects of a reduction in the dissolved oxygen level in the coolant as established in the pipe test on the crack propagation rate of key BWR structural materials, and (2) determine the crack arrest and retardation behavior in this specified HNC environment.
The results of these .tudies indicate that HWC dramatically improves the f atigue crack initiation resistance of carbon stee'. and retards fatigue and constant load crack growth of stainless, carbon and low alloy steels.
0.3 In-Reactor Tests ',
Results from the initial Dresden-2 HNC demonstration and subsequent HWC ,
operation plus recent laboratory studies stongly suggest that to obtain the
" blanket" protection that HWC seems capable of providing, both the dissolved ox; sen and the conductivity of the water must be reduced. The synergism between these two critical variables defines the aggressiveness of the environment. The logic for this synergism is straight-forward: The reducticn in dissolved oxygen content reduces the driving force for the f
7-5
necessary cathodic corrosion reaction: + 4e - 40H , while the 02 + 2H 2O reduction in the conductivity of the electrolyte retards the corrosion rate.
The importance of conductivity was initially indirectly identified during the laboratory pipe tests discussed abova where, when the oxygen level and conductivity was reduced from 50-70 ppb to 15 i 10 ppb and 0.6 0.3 uS/cm to <0.2 uS/cm, respectively, mitigation of IGSCC occurred.
Similar results were identified during the first series of Dresden-2 HWC materials tests where a furnace-sensitized Type 304 stainless steel CERT
~
specimen was characterized by 35% IGSCC of the fracture surface in a reactor environment of 40 ppb 02 and 0.37 uS/cm conductivity, but had no IGSCC in an improved reactor environment of <20 ppb oxygen and 0.29 uS/cm conductivity. Although oxygen and conductivity were reduced simultaneously in both the laboratory pipe tests and the Dresden-2 CERT test, other laboratory work has isolated these two factors and has provided direct evidence for importance of conductivity on IGSCC, as described below.
CERT tests were performed on weld sensitized Type 304 stainless steel in oxygenated water at 274*C (525*F) with low concentrations of various impurities. The results of this study clearly indicated that the amount of IGSCC in sensitized Type 304 stainless steel is directly related to the specific anion and its concentration / conductivity. For example, at a constant oxygen level of approximately 22 ppb, a solution of Na2CO3 with a' conductivity of 1.0 uS/cm produced IGSCC; whereas, reducing the conductivity of the same solution to 0.3 uS/cm resulted in ductile failure. Recent results from an ongoing oxygen vs. conductivity CERT study can also be cited as evidence for the efficacy of conductivity.
During the second ECP measurement campaign at Dresden-2 (October 1953 -
April 1984), the ECPs of stainless steel was measured for 3330 hours0.0385 days <br />0.925 hours <br />0.00551 weeks <br />0.00127 months <br /> (-139 days). The potentials obtained during this extensive coverage were similar to those measured during the first demonstration campaign. The normal Oresden-2 environment was , characterized by an ECP range for Type 304 nair.less steel of -100 to -200 mVSHE, while the hydrogenated Dresden-2 7-6
l 1
l l
l
- environment reduce the Type 304 stainless steel pot'ential to -370 cc -MC j j mVSHE. It appears that during- times when HWC is terminated, a "memcry ' ,
effect occurs in that the potential does not suddenly rise into the :ange l 4
where IGSCC of sensitized Type 304 stainless steel can occur. Instead :ne
} potential rises slowly which creates a window of time prior to rising in:c j' the cracking potential region. Typically this windcw is 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br />.
1 4
i The 1983-1984 CERT test campaign at Dresden-2 was characterized by a simiin set of seven experiments. Four furnace-sensitized Type 304 stainless steel .
f specimens, one furnace-sensitized Type 304 stainless' steel specimen (precracked by IGSCC in a 0.2 ppm oxygen water laboratory autoclave and l . snipped to the site), one SA106 Grade B carbon steel and one SA508 Class II
- low alloy steel specimens were tested. The results of these CERT tests plus l
the results of the first campaign and comparable laboratory tests are
- summarized in Table 1.
I j Table 1 clearly indicates an excellent correlation between Dresden-2 and
- laboratory CERT test results. It also clearly shows that HWC provides l mitigation of IGSCC in furnace sensitized Type 304 stainless steel even in instances where hydtcgen addition'has been interrupted. For example, in f one CERT study (line No. 6) 396 hours0.00458 days <br />0.11 hours <br />6.547619e-4 weeks <br />1.50678e-4 months <br /> of test time and 45% strain were j accumulated with >10% of the test time in oxygenated water (30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br /> with no H2 and 02 >200 ppb and 15.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> with 02 at 25-41 ppb) and no IGSCC was i identified. However. in CERT No. 5, a 12 hour1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> continuous test pericd with
! the hydrogen injection terminated produced minor IGSCC damage. (This test
- result is also marred by a thermal overload problen which resulted in .
l extremely high thermal stress.) The precracked furnace-sensitized Type 304 stainless steel specimen (No. 8) (67 hours7.75463e-4 days <br />0.0186 hours <br />1.107804e-4 weeks <br />2.54935e-5 months <br /> in 200 ppb 02 ) which was ,
- carefully documented prior to in-reactor testing, failed by ductile tearing j after 301 hours0.00348 days <br />0.0836 hours <br />4.976852e-4 weeks <br />1.145305e-4 months <br /> in Dresden-2. The mechanical parameter of this specimen is L similar to non-precracked specimens. It is also important to note that this test included 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> >40 ppb oxygen. Finally, no IGSCC was identified in
- . either SA508-2 and SA533-B low alloy steel and SA106-B carbon steel in HWC at Dresden-2 despite interruptions in hydrogen injections.
7-7
Crack growth data versus time and environment on precracked furnace-sensitized Type 304 stainless steel is being obtained at Dresden-2 using the reversing DC potential drop technique as developed by GE Corporate Research and Development. The specimen was precracked in San Jose in the nominal 200 ppb 02 and then transported to Dresden-2 for testing. The K y level for the specimen bounds the crack previously identified in Dresden-2 safe-end at 27.5 MPa /m (25 ksi /in) . To supplement this program, bolt loaded WOL specimens (one each) of Alloy 600 and unclad SA508 C1. II low alloy steel have also been insertd into the autoclave.
Figure 1 presents the early DC potential drop crack growth data of the precracked furnace-sensitized Type 304 stainless steel specimen. Although the data is preliminary and is clouded by a reactor scram which occurred approximately 120 hours0.00139 days <br />0.0333 hours <br />1.984127e-4 weeks <br />4.566e-5 months <br /> into the test and some other interruptions in hydrogen injection, no significant crack growth has occurred on the specimen. This preliminary result verifies th,e result of the Dresden-2 precracked CERT test discussed above and the results of the mid-cycle in-service inspection (ISI) as presented in Table 2.
This mid-cycle ISI was perf ormed as required by the NRC t. Verify the mitigation of IGSCC growth by HWC. The inspection was performed on November 12, 1983, after approximately 5.5 months of operation with hydrogen injection. The results presented in Table 2 indicate that despite interruptions in hydrogen iniection no crack growth was detected by ultrasonic testing.
e t
t l 7-8
prerequisities for IGSCC (a sensitized microstructure (chromium depleticn n the grain boundaries) and a tensile stress above the yield stressi alsc ar +
present. A variety of IGSCC remedies has been developed and qualified wn:cn address the sensitization and tensile stress aspects of stress corrosion cracking, including Nuclear Grade Type 316 and 304 stainless steels, solution heat treatment (SHT), corrosion-resistant cladding (CRC) , heac nnt welding (HSW) and induction heating stress improvement (IHSI) . It should also be possible to suppress IGSCC by reducing the electrochemical driving -
force for IGSCC, i.e. , by modifying the BWR coolant environment.
The demonstration of an IGSCC remedy based on modifying the chemistry of the BWR coolant is the objective of this materials program. The basic concept is to reduce the aggressiveness of the.BWR. environment by adding hydrogen gas to the feedwater to reduce the dissolved oxygen concentration in the reactor water and reduce the coolant conductivity to a low value by improved operational practices. This approach appears to have the potential fcr
" blanket" proteetion of all types of BWR structural materials during power operation of the plant and may arrest the growth of incipient cracks.
To accomplish the project objective of evaluating and quantifying the beneficial effects of oxygen suppression through hydrogen addition, .
numerous laboratory and Dresden-2 in-reactor testing techniques have been used to study a broad range of BWR structural materials and corresion phenomena. The results of these laboratory materials programs and Dresden-2 operaticnal results which established the hydrogen water chemistry (HWC) specification are the subject of this chapter. Appendix A describes the ,
detailed results of the HNC materials program. -
l .
l 7-9 t
t .__ .- _ -_. - . _
Tcblo 1 Resulta of Drced;n-2 cnd I.nborotcry HWC CERT T;cto Test KI Time To Time Off Elongation Matetial Location 02(ppb) (uS/cm) Failure (h) IIWC (h) % Result 2 D-2 268 0.29 108 0 12 70% IGSCC
- 1) FS T-304
- 2) FS T-304 D-2 40 0.37 143 2 20 35gIGSCC
- 3) FS T-304 D-2 <20 0.29 >297 4 38 DF
- 4) PS T-304 D-2 5-20 0.I9 208 4 5 NM S DF
- 6) PS T-304 D-2 3-30 0.13 396 36 45 DP
- 7) PS T-304 D-2 7-19' O.09 400 25 46 DF
- 8) FS T-304PC 6 D-2 12-20 0.09 301 7 7 40 No IGSCC Extension
- 9) PS T-304 VNC O
195 (0.1 156 NA 9 17 85% IGSCC
- 10) FS T-304 VNC 15 <0.1 262 0 NA DF
- 16) 'SA 508-2 12 VNC 50 <1 44 0 29 DP
- 18) SA 106B VNC 50 <1 4G 0 29 DP
- 1) K = Conductivity 8) VNC = Vallecitos Nuclear Center
- 2) FS = Furnace Sensitized 621*C (1150*P)/24 hr 9) Not Applicable, i.e., not a HWC Test
- 5) NM = Not Measured to Date 304 Stainless Steel
- 6) PC = Precracked in 200 ppb O 11) Motor Failure, Specimen Fractured Manually
- 7) Plus 67 Ilours Precracking (36$hrtotal) 12) Creviced
Table 2.
Mid-Cycle ISI Results from Dresden-2 ISI RESULTS ,
Date Weld April 29. 1983 Movember 12, 1983 28" Safe End 1" long, 16% deep I 1" long, 13% deep.
PS2-201-1 12" Riser (Two Cracks) 0.25" long, 17% deep 0.25" long, 15% deep PDS-D20 0.25" long, 19% deep 0.25" long, 17% deep 12 Riser (Two Cracks) 0.50" long, 19% deep 0.50" Ic.ng, 181 deep PDS-D5 0.25" long, 14% deep 0.25" long, 16% deep 1
Percentage Through Wall e
0 h
. . . l Iigure I liresiten-2 IlWC Crack Growth Test on furnaco S';nsitiz:d lype 3(14 Stainless 5 teel, F. = 27.5 MPa/m
,[t1 .
k i
. 005 .
I Reactor Scram I
\
q k _. ._
.,. ,.- x
-u
= '
f""~
r, o
O. .
C
- i x
U E
I d . 005 I
c.
.4, . _ _ . . _ . . L _ . L ._ . . L.__ . .L. l . . . _ . _ -
2 IJ ') ) IE ($ (3
- .' ) TJ .J FJ iJ
.t; til . (D W
. i Iest linration (h)
TECHNICAL BASIS ECR HYDROGEN WATER CHEMIS'IRY SPECIFICATION k
1.0 INTRODUCTION
Am The recirculating coolant in BWRs is high-purity, (no additive) neutral pH i
water containing radiolytically produced dissolved oxygen (100-300 ppb) .
, This level of dissolved oxygen is sufficient to provide the electrochemical .,
driving force needed to promote IGSCC of sensitized austenitic stainlesc steel piping and similar structural components if the other two ,
e' .
prerequisities for IGSCC (a sensitized microstructure (chromium deplecicn at the grain boundaries) and a tensile stress above the yield stress] also are l present. A variety of IGSCC remedies has been developed and qualified which address the sensitization and tensile stress aspects of stress corrosion cracking, including Nuclear Grade Type- 316 and 304 stainless steels, j solution heat treatment (SHT), corrosion-resistant cladding (CRC), heat sink
, welding (HSW) and induction heating stress improvement (IHSI) . It should A' also be possible to suppress IGSCC by reducing the electrochemical driving
1 The demonstration of an IGSCC remedy based on modifying the chemistry of the BWR coolant is the objective of this materials program. The basic concept is to reduce the aggressiveness of the BWR environment by adding hydrogen gas to the feedwater to reduce the dissolved oxygen concentratien in the i reactor water, and reduce the coolant conductivity to a low value by ,
l improved operational practices. This approach appears to have the potential
[
for " blanket" protection of all types of BWR structural materials during power operation of the plant and may arrest the growth of incipient cracks. -
To accomplish the project objective of evaluating and quantifying the beneficial effects of oxygen suppression through hydrogen additien, numerous laboratory and Dresden-2 in-reactor testing techniques have been 7-10
l l
used to study a broad range of BWR structural materials and corrosion phenomena. The results of these laboratory materials programs and Dresden-2 operaticnal results are the subject of this chapter.
2.0 FULL-SIZE WELDED STAINLESS n"' EEL PIPE TESTS
2.1 INTRODUCTION
Pipe testing has been performed to quantify the effect of hydrogen water
, chemistry (HRC) on the IGSCC performance of Type 304 stainless steel piping. The general approach is similar to that used for the General Electric /EPRI Program, " Alternate Alloys for BWR Pipe Applications,"2 EPRI Project RP-968. However, instead of selecting and qualifying a particular alternative piping alloy, the thrust of the present program was to qualify -
an alternative environment using hydrogen addition to reduce the dissolved oxygen concentratich.-
1 To reliably demonstrate the improved IGSCC performance of Type 304 stainless steel in HWC in a reasonable period of time, 3 reliable, accelerated test representative of actual field piping conditions was required. The full-size welded pipe test which was developed for EPRI RP-968 fits this requirement. A statistical method was used for developing f actors of improvement (FOI) by comparing timec-to-failure for an alternate alloy with similar data for the reference alloy, Type 304 stainless steel. Similarly, the method was used to establish a FOI for HWC vis-a-vis the normal 200 ppb 02/288'C (550 *F) BWR steady-state environment by comparing times-to-failure for the reference alloy in the two environments.
2.2 RESULTS AMD DISCUSSION i
The pipes tests were conducted in simulated BWR high temperature water environments in the Pipe Test Laboratory (PTL) as described in detail elsewhere.' This f acility was modified to provide control and monitoring
.capacility for hydrogen addition and the lower oxygen and conductivity levels required f or HWC.
7-11
Initially the WC environment selected was 50-70 ppb 0 with a conductivity of 0.6 i 0.3 uS/cm. This environment was established at a time when it was not certain if lower 0 2 levels :ould economically be estacli:ned in a plant and was expected to provide conservative data. Four WC crack growth specimens, six WC crack initiation specimens and four reference
. specimens exposed to 200 ppb 02 environments were initially included in tne program. Of the four crack growth specimens, two were sectioned metallographically after precracking in 200 ppb 02 water to determine pre *
- crack depth and two were continued on test in the 50-70 ppb 0 2 NC environment. '
As described in additional detail later in this section, premature cracking occurred in the. initial 50-70 ppb 02 WC environment and it was determined that dissolved oxygen and conductivity must be lowered to 20 ppb and
<0.2 us/cm, respectively, to demonstrate improvement. Exposure of all cracked specimens was then continued in the new WC specification environment to determine if crack arrest would occur. Additionally two new crack growth specimens and .three new crack initiation specimens were added to the program for testing in the new WC environment.
The specifications for the three high temperature water environments used were as follows:
Tnitial HWC Envirnnmente Temperature 282'C i 5'C (540 *F i 10'F)
Dissolved Oxygen 50-70 ppb Dissolved Hydrogen 125 1 25 ppb ',
Conductivity at 25'C (77'F) 0.6 i 0.3 uS/cm -
i j Later HWC Envirorunanti 1 Temperature 282'C i 5'C (540 *F 10*F) l Dissolved Oxygen 15 t 10 ppb l Dissolved Hydrogen 125125 ppb
{ Conductivity at 25'C (77'F) <0.2 uS/cm o
i 7-12
Reference Environments
. Temperature 282*C 5*C (540 *F 10*F)
Dissolved Oxygen 0.2 1 0.1 ppm Conductivity at 25'C (77'F) 0.6 i 0.3'uS/cm The total test matrix is shown in Table 1. Four heats of Type 304 stainless steel were included in the program. One heat of Type 316 Nuclear Grade
, stainless steel was included in the crack initiation tests.
~
The 10-cm (4-in.) Schedule 80 pipe specimens were assembled using eleven
~
10-cm (4-in.)-long test pieces from one heat, two transition pieces, and top and bottom end caps joined by circumferential welds as illustrated in Figure 1.
The pipe specimens were cyclically loaded at one cycle per day in a pipe test stand, Figure 2. The loading waveform is shown in Figure 3. A long hold time was used to prevent fatigue failure during long exposure to develop FOI. To accelerate cracking, axial test loads were applied to attain a maximum nominal pipe section stress equal to 233.2 MPa (33.8 ksi) 253 (Sg = ASME Code allowable), as determined at the smallest cross section of the specimen, the weld counterbore.
The results were evaluated by ultrasonic testing (UT) and metallography.
Specimens were given an initial baseline UT before going on test. The pipe tests were conducted until through-wall cracks occurred or until the test cejectives were achieved and re.noved for destructive metallographic evaluation. The longer exposure specimens were periodically removed from test for UT inspection to estimate crack growth.
Cracr. growth pipe specimen AWC-1 failed by IGSCC at HAZ location designated J-1 af ter 30 cycles, 844 hours0.00977 days <br />0.234 hours <br />0.0014 weeks <br />3.21142e-4 months <br /> exposure to the 50-70 ppb 0 /0.6 2 us/cm conductivity HWC environment described above. This specimen had a precrack initiated in the 200 ppb oxygen environment estimated by UT evaluation to be 7-13
0.38 m (15 mils) deep. During the 50 to 70 ppb 0 2/0.6 uS/cm conductiv:. y W C exposure, the inter granular portion of the crack grew to a maximum depth of 4.8 mm (190 mils)- and extended to 360' around the pipe, af ter .wnian the pipe failed in ductile tension.
Because failure of specimen AWC-1 occurred earlier than expected fcr EWC exposure, testing of all pipe specimens in this environment was interrupted and UT evaluations were conducted. Table 2 lists the accumulated exposures .
of the specimens and estimated maximum crack depths from UT inspecticns.
It can be seen that all specimens tested in the 50 to 70 ppb 02 high conductivity HWC environment exhibited early cracking except specimen AWC-14, which was fabricated from Type 316 Nuclear Grade 00) stainless steel pipe. This specimen showed no crack indications during UT examinaticr at maximum sensitivity. Subsequent metallographic examination on specimen AWC-14 verified the UT results.
The remaining unfalled, but cracked specimens from this environment were transferred to the lower oxygen / conduct'ivity environment described above to determine if crack arrest would occur in this environment.
~
The results of exposure of the precracked specimens to typically 20 ppb ,
02 /0.2 uS/cm conductivity environment are also listed in Table 2. None of these specimens failed during exposure times significantly longer than the
! f ailure times for the four reference 200 ppb oxygenated water specimens listed in Table 2. All of the precracked specimens except AWC-ll and AWC-10 showed some crack extension in the lower oxygen environment. Specimens -
- AWC-9 and AWC-12 had one weld removed from each four metallographic sectioning to characterize the cracking.
The metallographic sections of specimens AWC-9 and AWC-12 showed intergranular cracks with a transition to transgranult: during the latter stage of cracking. SEM examination also showed intergranular thumbnail cracks with a transition through mixed mode to total transgranular cleavage fracture at the root of the cracks and, in most instances, showed striations r
7-14
which are indicative of fatigue associated with the cleavage fracture.
Since fatigue is not a factor in BWRs, no crack extension would be anticipated.
Two additional crack growth specimens (ANC-15 and AWC-16) were precracked ir 0.2 ppm 02 water to depths of approximately 1.27 and 1.02 mm (50 and 40 mils), respectively. Specimen AWC-15 was exposed to the low
, , oxygen / conductivity H2 WC environment for 5310 hours0.0615 days <br />1.475 hours <br />0.00878 weeks <br />0.00202 months <br /> (including 343 hours0.00397 days <br />0.0953 hours <br />5.671296e-4 weeks <br />1.305115e-4 months <br /> in 200 ppb 02) and specimen AWC-16 was exposed to the 0.2 ppm 02 water control environment at constant load to evaluate the fatigue component of crack propagation. Specimen AWC-16 failed in 1285 hours0.0149 days <br />0.357 hours <br />0.00212 weeks <br />4.889425e-4 months <br />, while the precracked specimen AWC-15 did not fail during the total 5653 hours0.0654 days <br />1.57 hours <br />0.00935 weeks <br />0.00215 months <br /> of exposure. ,
Due to premature failure of the original crack initiation specimens in the 50-70 ppb 02 water, three new crack initiation specimens (one frem each of three heats, specimens AWC-17, -18, and -19) were tested in the low oxygen / conductivity environment. Specimen AWC-17 was removed from test af ter 1693 hours0.0196 days <br />0.47 hours <br />0.0028 weeks <br />6.441865e-4 months <br /> for interim UT exanination and removal of one weld for dye penetrant testing and sectioning. No evidence of crack initiation could be seen at this time. It was repaired and returned to test. Specimens AWC-17, -18, and -19 have obtained 5122, 7580, and 7585 hours0.0878 days <br />2.107 hours <br />0.0125 weeks <br />0.00289 months <br /> of exposure respectively. with no UT indicaticns of cracking. IGSCC would be expected in the norminal environment af ter -300 hours (POI of at least 25) . Through wall failures of these pipes in the reference environment would be expected in approximately 1500 hours0.0174 days <br />0.417 hours <br />0.00248 weeks <br />5.7075e-4 months <br />.
2.3 CONCLUSION
S The welded pipe test program has revealed the following conclusions to date:
- 1. For Type 304 stainless steel welded pipe with deep IGSCC, subsequent exposure to a hydrogen water chemistry environment will arrest intergranular propagation even at stress levels of twice the ASME . code allowable. (i.e. , 2Sm)
- 7-15
- 2. The factor-of-improvement based on crack initiaticn data fer J.e HWC environment compared with the nominal environment is at leas:
- 25. Also no. failures have been observed in HWC initiatico pipe tests with exposures of -7600 hours, whereas pipes tested in :ne nominal environment fail through wall in approximately one fif th that exposure period.
i
- 3. No severely pre-cracked Type 304 stainlesc steel pipes have failer-in the HWC environment.
4.' Felded Type 316 Nuclear Grade pipe showed the expected good IGSCC performance in hydrogen water chemistry, with no cracking initiated in 4047 hours0.0468 days <br />1.124 hours <br />0.00669 weeks <br />0.00154 months <br /> at a stress of 2S,.
3.0 CRACY PROPAGATION STUDIES
3.1 INTRODUCTION
/t .
The series of tests performed under the general heading of crack propagaticn studies had two major objectives: (1) determine the effects of a reducticn in'the dissolved oxygen level in the coolant.cn the crack propagaticn rates for key BWR structural materials, and (2) determine the crack arrest and retardation behavior in this reduced oxygen environment. The types of testing techniques utilized to provide this information included low' cycle fatigue tests, constant load stress corrosion crack growth tests and fatigue
- crack growth tests.
l 3.2 RESUI.TS AND DISCUSSION
~
3.2.1 Low cvele Patione Tests on carbon Steel A series of fatigue crack initiaticn tests was performed to determine the 1 !
effect of dissolved oxygen on the fatigue lifetime of carbon steel piping in l high purity water. Tests were conducted on notched and unnotched specimens of SA 106-B carbon steel at low (11.1 cph) and high (277 cph) frequencies.
,/ 4 7-16
The specimens were cyclically loaded with a R* ratio of 0.08.
,R = minimum load EExiEuB~1oa3 The test revealed that HWC should not adversely affect the notched or unnotched fatigue crack initiation behavior of carbon steels. Indeed a HWC specification of 20 ppb 02 and 0.2 uS/cm conductivity clearly inhibits crack initiation and lifetimes approach those obtained in air tests. The
,, service lifetime of carbon steel piping in BWRs probably can be extended by controlling the dissolved oxygen and conductivity levels to low values.
~
3.2.2 Stress Corrosion Crack Growth Tests One standard IT-WOL specimen (Figure 4) was fabricated from each of four typical BWR structual materials: (1) furnace sensitized [621'C (1150'F)/12 h1 Type 304 stainless steel; (2) furnace sensitized [621*C (1150'F)/2h]
Type 316 Nuclear Grade stainless steel; (3) SA508 Class 2 low alley steel; (4) SA333 Grade 6 carbon steel. Each specimen was fatigue precracked in room temperature air to ensure that an active fatigue crack was present prior to environmental testing.
The SCC test was performed in six loading phases (three slow cyclic loading phases and three constant load phases) as shown in Figures Sa, b, c. The slow cyclic loading (SCL) phase prior to each of the three constant load (CL) phases ensured that .each specimen.had an active crack growing in. the environment prior to switching to constant load. The loading during each of the constant load phases was selected so that the stress intensity (K) for each rpecimen corresponded to the K levels used in previous 288"C (550'F)/200 ppb oxygenated water baseline tests.' Therefore, a direct comparison cf crack growth rates in the 200 ppb oxygenated water baseline environment and HWC could be made. The results of these tests are summarized in Tables 3 and 4 for the HWC and nominal environment, respectively.
7-17
The results of the SCC growth tests revealed that the HWC envil >nment was detectably less aggressive than the 200 ppb 02 environment for three of the
, four matrials tested; the Type 316 Nuclear Grade stainless steel shewed nc detectable growth in either environment. .
3.2.3 Faticue Crack Growth Tests A total of ten compact tension (CT) fracture mechanics specimens and twc .
1T-WOL fracture mechanics specimen were tested at two cyclic frequencies.
Each specimen was fatigue precracked a minimum of 1.9 mm (0.075 in.) in rocm.
temperature air to ensure that an active fatigue crack was present prior to testing in tha HWC environment. The materials tested were annealed, furnace sensitized, welded and 1ow temperature sensitized [482*C (900'F)/24 hl Type 304 stainless steel, furnace sensitized Type 316 Nuclear Grade stainless steel, SA508 Class II low alloy steel and SA333 Grade 6 carbon steel (only IT-WOL specimens) .
Six specimens were loaded at a time in a series cr. din using the skewed sawtooth waveform shown in Figure 6. The cyclical frequency was 0.74 cph (81-minute period) and 7.5 cph (8-minute period) with R = 0.6. Initial stress intensity values were selected to obtain a direct comparison of crack growth rates in HWC with growth rates generated.previously for the same materials in a 200 ppb oxygenated water environment under identical leading conditions.
l The six 0.74 cph specimens were subjected to 1349 loading cycles in HNC
( while the remaining 7.5 cph specimens were characterized by 15,000 cycles. -
Compliance-based crack length data for the twelve specimens, as well as
! eight specimens tasted in the nominal environment, were analyzed for average crack growth rate as shown in Table 5. These average cyclic crack growth rates are plotted as a function of stress intensity range ( K) in l Figures 7 through 10 for austenitic stainless steels and ferritic materials l at the two cyclic frequencies.
l l
l 7-18
Figure 7 shows that the 0.74 cyclic crack growth rates for the two sensitized Typa 304 stainless steel specimens tested in 200 ppb 02 reference environment were three times greater than the cyclic crack growth rate of sensitized Type 304 stainless steel tested in HWC. The solution annealed and weld HAZ Type 304 stai nless steel specimens which had a low degree of sensitization had HWC cyclic crack growth rates similar to the specimen which was highly sensitized. Therefore, the degree of sensitization of
,e Type 304 stainless steel seems to have little effect on cyclic crack growth rate in HWC under the loading conditions tested. The cyclic crack growth rate for Type 316 Nuclear Grade stainless steel in HNC was only slightly less than the Type 304 scainless steel data.
A more dramatic difference in cyclic crack growth rate in HWC versus the 200 ppb 02 reference environment was seen in the carbon steel and low alley steel specimens, Figure 9. The 0.74 cyclic crack growth rate in HWC was 20 times less for carbon steel and 7 times less for low alloy steel than in the 200 ppb 02 reference environment. Similar differences in cyclic crack growth in HWC versus the 200 ppb 0 2 reference enviroment was seen at higher 7.5 cph where the factor of improvement for Type 304 stainless steel was similar to that obtained at the lower frequency, but significantly higher (450 and 1000X) for the ferritic materials, Figures 9 and 10.
3.2.4 Slow Risino Load (SRL) Fracture Mechanics Tests The H2 WC crack growth test was designed to duplicate the test conditions of a previous SRL test conducted in the nominal 0.2 ppm oxygenated water environment.' In this way, a direct comparison of results could be made
~
between the two different environments. In this task, the 1T-CT test was perf ormed in a water environment containing 230 ppb H2 and 50 ppb oxygen at 288 C (550 F) .
Ecth tests were performed using standard linear elastic fracture mechanics type crack growth specimens made from SA508 Class II low alloy steel.
However, the specimens were clad with stainless steel so that any electrochemical potential differences between the cladding and base metal 7-19
materials would be. established during testing.
In. situ compliance measurements of the IT-CT specimen tested by the sicw rising load method indicated no apparent crack growth during the entire 3:2 hours of loading in the hydrogen modified high temperature aqueous environment. In contrast, a similar specimen tested in water with 0.2 ppm 02 showed a crack extension of ~0.36 mm (0.014 in.) in a test period of 88C hours. Except for the dissolved oxygen and hydrogen levels, the conditicns -
for both . tests were the same. The crack growth data are shown in Figure II.
, Af ter test termination, the IT-cts were split apart mechanically to ,
i determine whether any pre-crack extension had occurred. The sample tested in the simulated BWR environment indicated crack growth in the base metal to approximately the extent predicted by the c.ompliance measurements, while the sample tested in the hydrogen modified environment showed no crack propagation.
3.2.5 Bolt rnaded 1/2 T-WOL Fracture Mechanics Terts Triplicate bolt-loaded 1/2 T-WOL [1.77 cm (0.5 in.)-thick wedge opening leading WOL] specimens were fabricated from carbon steel and two low allcy I
steels. The low alloy steel specimens contained Type 309 stainless steel weld metal in the precracked region and had Type 304 stainless steel plates tack welded to their surfaces to simulate a stainless steel clad component.
Duplicate bolt-loaded WOLs of the carbon and low alloy steels (6 specimens) were tested in a refreshed stainless steel autoclave system that contained ',
water at 288'C (550*F) with 140 to 258 ppb H2 and 7 to 11 ppb dissolved 02 '
The conductivity was controlled to <0.2 uS/cm. The third set of three WCLs
was tested during the hydrogen injection program at Dresden-2 from 5/19/82 L to 6/29/82.5 The stress intensity was 58.1 MPa /in(52.9 ksi /in) .
Tests from duplicate bolt-loaded specimens of three materials (SA106-B, Clad SA533-B and Clad SA508-2) showed no crack growth af ter 5521 hours0.0639 days <br />1.534 hours <br />0.00913 weeks <br />0.0021 months <br /> and the absence of crack growth was verified metallographically. Results from the 7-20
l set of three specimens which were exposed during the H2 injection program at Dresden-2 also showed no crack propagation af ter 936 hours0.0108 days <br />0.26 hours <br />0.00155 weeks <br />3.56148e-4 months <br /> of exposure, in l spite of the fact that there were occasional high dissolved 02 concentrations during the testing period. l l
3.2.6 Potential Dron Fracture Mechanics Tests
,. This investigation was designed to determine the response of SA533-B low alloy steel in HWC (50 ppb 0 2 and 125 H2 , <0.2 uS/cm) under constant load, using the potential drop technique to monitor crack growth. Compact tension specimens were machined with a 0.05 mm (2 mil) radius notch and were
~ instrumented for both in-situ crack opening displacement (for use with the compliance technique: and DC potential drop monitoring. A reversed DC potential drop technique was used which involved reversing the current flow through the sample every 0.5 to 1 sec both to minimize electrochemical effects and to reduce measurement errors associated with thermocouple ef f ects, etc. The test was initiated under constant load conditions at 55 MPa /m (50 ksi /in.) . No crack growth was observed either with the DC potential drop growth monitoring technique or with the linear variable differential transformer (LVDT) crack opening displacement gage. After approximately 50 hours5.787037e-4 days <br />0.0139 hours <br />8.267196e-5 weeks <br />1.9025e-5 months <br />, the load was increased to 47.8 kN (10,740 lb),
resulting in a stress intensity of 60.4 MPa /m (55 ksi /in.) . After one week of exposure, neither crack opening displacement nor DC potential measurements indicated any crack growth, . so the load was raised again to 52.5 kN (11,800 lb) or 66 MPa /m (60 ksi /in.). Af ter more than 1500 hours0.0174 days <br />0.417 hours <br />0.00248 weeks <br />5.7075e-4 months <br /> of exposure at dais level, there was no in-situ indicaticn of crack growth.
~
The sample was ramoved from the autoclave system, examined and photographed, snd then pulled apart af ter sawing a two centimeter notch frcm the back face -
cf the specimen and cooling the specimen in liquid nitrogen. Examination of the specimen af ter separation showed that no cracking of the metal at the sase of the notch occured.
This test shows that the low alloy steel SA533-B is extremely resistant. to
- acxing in hydrogen water chemistry at 288*C (550'F) under constant load 7-21
i conditions. This result compares favorably with these obtained on Icw alloy steel at 288'C (550*F) in water containing 0.2 ppm or 8 ppm oxygen.
At 8 ppm oxygen, crack growth was observed 'inder cyclic load and constant load at stress intensiy values above approximately 33 MPa /m (30 ksi in.1 4-In water containing 0.2 ppm oxygen, no growth was observed under constan load at stress intensity values up to 63.7 MPa /m (58 ksi /in.) , althougn under cyclic loading crack growth cccurred at stress intensity values as icw as 41.8 MPa /m (38 ksi /in.). -
3.3 EQECLUSIONS ,
The results of the various laboratory test series designed to evaluate the
. effects of HWC on crack propagation rates in typical BWR structural materials can be summarized as follows:
- 1. Reducing the oxygen content does not adversely affect the notched
, or unnotched fatigue crack initiation behavior of carbon steel and the fatigue perf ormance may improve at the low oxygen levels involved in HWC.
- 2. Under constant loading conditions, no measurable crack growth was detected in the HWC environment for furnace-sensitized Type-304 stainless steel, furnace-sensitized Type 316 Nuclear Grade stainless steel, SA508 C1.II low alloy steel and SA333 Gr. 6 carbon steel for stress intensities up to 31.2 MPa /m (28.4 ksi
/in.), 30.1 MPa /m (27.4 ksi /in.), 50.9 MPa /m (46.3 ksi /in.)
~
and 44.8 MPa /m (40.3 ksi /in.)', respectively. In the nominal ,
enviornment, measurable crack growth is observed at significantly lower stress intensities for furnace-sensitized Type 304 stainless steel (16.3 MPa /m (15.7 ksi /in.)] and is observed on SA206-B carbon steel at 44.0 MPa /m (40.0 ksi /in.) . No measurable crack growth occurs in the low alloy steel vr the Type 316 Nuclear Grade t stainless steel in the nominal environment.
7-22
- 3. The 0.74 cph cyclic crack propagation rates for furnace-sensitized Type 304 stainless steel, SA508-C1.II low alloy steel and SA333 Gr. 6 carbon steel were r6spectively 3, 7 and 20 times slower in the HWC environment than in the nominal BWR environment. At 7.5 cph, the cyclic crack propagation rate for furnace-sensitized Type 304 stainless steel was also ~3, while for the ferritic low alley and carbon steel the crack propagation rates were ~450 and 1000 times slower, respectively.
- 4. In the HWC environment no crack growth occurs in pre-cracked low alloy steel specimens clad with Type 309 stainless steel loaded to a high stress intensity, 58.1 MPa /m, (52.9 ksi /in.).
- 5. The presence of hydrogen gas dissolved in the coolant does not, increase the crack growth rates of SA106-8 carbon steel, clad SA533-B, and clad SA508-2 low alloy steel.
- 6. DC potential drop and LVDT crack opening displacement techniques used to monitor crack length in compact tension samples of low.
alloy steel SA533-B under constant load conditions at 288'C (550*F) in pure water containing 50 ppb oxygen and 130 ppb hydrogen, revealed no crack growth after 160 hours0.00185 days <br />0.0444 hours <br />2.645503e-4 weeks <br />6.088e-5 months <br /> at 60.4 MPa /m (55 ksi /in.) or af ter >1500 hours at 66 MPa /m (60 ksi /in.) .
Overall. the results indicate that materials which are susceptible to environmentally-assisted crack growth in the nominal BWR environment perform better under HWC conditions, i.e. , the hydrogen water chemistry environment is less aggressive than the nominal environment with respect to
, environmentally-assisted crack growth in BWR structural materials.
e 7-23
4.0 CORRQS101L2DTENTIAL MEASUREMENTS RQ_SIGLSTRAIU RATE SESTS__(. CERT _Ar0 SITl 4.1 EJTRODUCTIOE The objective of this part of the program was to determine whether stress corrosion or other forms of localized corrosive attack can occur in WC.
The task included three subtasks: (1) mea.surement of the corrosion potentials of Type 304 stainless steel. Alloy 600, carbon steel, Zircalcy-2, and low chromium (simulated sensitis:ed grain' boundary) Type 304 stain; i .
steel in six water environments covering the expected range of EWC dissolved oxygen and hydrogen contents; (2) Constant Extension Rate .
Technique (CERT) testing of Type 304 stainless steel, Alloy 600 and icw alloy and carbon steels, in three 0 -H2 2 water environments selected frcm subtask (1) , plus additional CERT tests on low alloy and carbon steel to s'upport 'the Dresden-2 tests and (3) limited Straining Electrode Technique (SET) testing of Type 304 stainless steel, Alloy 600 and Alloy X-750 at various test potentials simulating the nominal and HWC environments.
The objective of this part of the program was to determine whether stress corrosion or other forms of localized corrosive attack can occur in HWC.
The task included three subtasks: (1) measurement of the corrosien potentials of Type 304 stainless steel, Alloy 600, carbon steel, Zircaloy-2, and low chromium (simulated sensitized grain boundary) Type 304 srainless steel in six water environments covering the expected range of EWC dissolved oxygen and hydrogen contents; (2) Constant Extension Rate Technique (CERT) testing of Type 304 stainless steel, Alloy 600 and low alloy and carbon steels, in three 0 -H2 water environments selected from 2 ,
subtask (1), plus additional CERT tests on low alloy and carbcn steel to support the Dresden-2 tests and (3) limited Straining Electrode Technique -
(SET) testing of Type 304 stainless steel, Alloy 600 and Alloy X-750 at various test potentials simulating the nominal and HWC environments.
J 7-24
4.2 BESULTS AND DISCUSSIONS 4.2.1 Corrosion Potential Measurements The. decrease in <lssolved oxygen content in the BWR coolant as a result of cydrogen addition to the feedwater should significantly reduce the corrosion potential, which is a measure of the thermodynamic driving force
, for corrosi)n reactions. Figure 12, from Indig and McIlree* , indicates that
- there is a rapid drop in the corrosion potential of Type 304 stainless steel when the dissolvei oxygen content falls below approximately 40 ppb.
The corrosion potentials of Type 304 stainless steel, Alloy 600, carbon steel (SA333, Gr. 6) , Fe-lCNi-8.lCr, Lircaloy-2 and platinum were measured against a high temperature Ag/AgCl reference electrode in high purity water and in a 0.0lN Na2SO4 solution. The high purity water and sodium sulfate solutions contained specific concentrations of dissolved oxygen and hydrogen and were heated to 274*C (525'F). In the Na2SO4 solutions, the potential of pressure vessel steel (SA533 Gr. B) was also determined. Table 6 presents a list of the bottled gases used in this study and calculated dissolved gas concentraticns in the test environment.
i In high purity water, the electrochemical potentials for stainless steel tended to be below the potentials previously obtained when the dissolved oxygen concentration was >100 ppb. At very low dissolved oxygen concentraticns, the opposite effect occurred; the potentials were above previcus values. The other iron and nickel base alloys behaved similarly.
The net effect was to limit the response of electrochemical potential with changing dissolved oxygen concentration. The corrosion potentials in 0.01 N Na2SO4 were lower than in high purity water. In part, this decrease in pcrential is related to an increase in pH tecause the Na2SO4 forms a basic sclution in high temperature water. It appeared that hydrogen had only an
- ndirect effect on the electrochemical potential of the corrosion electrodes 4
as measured in these laboratory terts. In the reactor, the major effect of nycrogen will be in recombination with oxygen, and, as the oxygen decreases, a decrease in the chemical driving force for IGSCC will be observed.
- 7-25
4.2.2 consant Extension Rate Tests The constant e.xtension rate test (CERT) facility used for studying stress corrosion cracking behavior of alloys in simulated BWR environments was described by Clarkt, Cowan, and Danko.' To minimize oxygen gettering due :c l chemical reaction with heated surf aces, the regenerative heat exchanger was fabricated from titanium tubing and the pressure vessel and load-frame internals were made from Ti-6Al-4V alloy. .
All CERTs (apart from the Dresden-2 support tests) were conducted at 274*C .,
(525'F) in controlled aqueous environments (both water purity and various dissolved gas concentrations) at a strain rate of 2X10-5 min-I. CERTs were conducted on Type 304 stainless, Alloy _600, pressure vessel 1cw alley steel-(SA533) and carbon steel SA333, Gr. ' 6.
Table 7 presents a summary of the experimental conditions and the CERT results for the materials tested. The CERT results tabulated are the mechanical properties, failure times, and failure morphologies. Lower values of the mechanical properties and shorter CERT testing times are indications of stress corrosion cracking. Verificatien-of stress corrosion cracking is obtained by post-test examination.
The most important result of these tests was that none of the auoys exhibited any sign of stress corrosien cracking in the hydrogen water chemistry environment. Except for Alloy 600, all of the allcys exhibited some degree of stress cracking in a simulated normal BWR environment.
The laboratory CERT tests performed in support of the Dresden-2 IMC demonstration t. sed SA508 Class II, SA533 Grade B Class I low allcy steel and
~
SA106 Grade B carbon steel tested in the creviced and uncreviced condition.
The crevice was produced by wrapping and spot welding thin stairless steel shim stock around the test section. The test conditions and results are presented in Table 8. Comparison of these data with previous low alloy 7-26
.1
steel CERT data in 200 ppb 02 water, 8.0 ppm 02 water and air shows the benefit of the HWC (50 ppb 0 2, 230 ppb H ,2<1 uS/cm) .5 In contrast to the roughly 20% transgranular stress corrosion cracking (TGSCC) in 0.2 ppm 02' these HWC CERT test results reveal the 100% ductile behavior and appear to be essentially the same as 288*C (550*F) air test results.
The CERTs clearly indicated Ebat in high purity water, intergranular stress
,. corrosion cracking of weld-sensitized Type 304 stainless steel and TGSCC of pressure vessel and carbon steel can be prevented by hydrogen water chemistry.
The indications from CERTs carried out in the present and previous programs are that the low alloy and carbon steel are more resistant to stress corrosion cracking in higher dissolved oxygen concentrations than stainless steel.
Alloy 600 in the welded + LTS condition did not show any indications of IGSCC in any of the CERTs. This was not surprising, since the only instance of IGSCC of this alloy in the field was related to the presence of crevice and high concentrations of an impurity (resins) .
4.2.3 Straininc Electrode Tests The straining. electrode test (SET) is a CERT conducted under potential control rather than chemical control. Because potential control requires the passage of current through the solution in contact with the specimen, a conductive electrolyte is used to minimize IR drops. A circulating -
deserated 0.0lN solution of Na2SO4 maintained at 274*C (525*F) was the
, electrolyte. The control potentials simulated either normal BWR or HWC environments. All specimens were pulled to failure at a strain rate of 2X10-5 min-1 in an austenitic stainless steel pressure vessel. Since deaerated test solutions were used, titanium alleys in the high temperature SET system were not necessary. Figure 13 is a sketch of the SET facility and Table 9 presents the test matrix.
7-27
The results of the SETS are presented in Table 10. The critical findings ;.-
these tests were that no SCC occurred at potentials that simulated the hydrogen water chemistry environment for Type 30.4 stainless steel, A11cy 50:
or Alloy X-750. IGSCC did occur in the simulated nominal environment for Type 304 stainless steel and Alloy 600. No cracking of Alloy X-750 cccurrec in the SETS at -0.100 and -0.700VSHE. The low potential -0.700VSHE was used to determine whether hydrogen embrittlement could occur in this high strength alloy at potentials somewhat below those expected under hydrogen -
water chemistry.
4.3 CONCLUSION
S '
The results of the electrochemical measurements, CERTs and SETS indicate the following conclusions relative to the effects of HWC on BWR structural materials:
- 1. The major finding in the stress corrosion testing under chemical and electrochemical control in high purity water was that under simulated hydrogen water chemistry conditions, IGSCC of weld-sensitized and welded-plus-CTS Type 304 stainless steel and ,
Alloy 600 and TGSCC of low alloy steel were eliminated. In simulated normal BWR water, SCC of these materials occurred. SCC under electrochemical control with -a supporting electrolyte (0.0IN Na2SO 4 ) was far more severe than under chemical (oxygen) contrcl in high purity water.
- 2. HNC appears to suppress the t/ensgranular stress corrosion ',
l cracking of creviced or uncreviced specimens of SA533-B and SA508
! Class II low alloy steels and SA106-B carbon steel, which is
~
observed in the nominal environment.
- 3. No SCC was found for welded and low temperature sensitized specimens of Type 304 stainless steel and Alloy 600 at electrochemical potentials of -500 mVSHE (a simulated EWC environment), while cracking was observed for both materials in l
7-28 i
k
I l
the simulated normal BWR environment at -100 mVSHE' J
- 4. No hydrogen stress cracking was found for Alloy X-750 at potentials simulating deaerated water with a high dissolved hydrogen content (-700 mVSHE) and no SCC was observed for this alloy in the nominal BWR environment (-100 mV SHEI *
,. 5.0 DRESDEN-2 INC TEST RESULTS
5.1 INTRODUCTION
Materials tests have been conducted at Dresden-2 under HNC conditions to confirm the resui ;c of the laboratory HWC investigations. These tests
, indicate that laboratory data appear i> be directly applicable for predicting materials behavior in an operating BWR.
The first set of hydrogen injection verification studies were run at the Dresden-2 nuclear power plant of the Commonwealth Edison Co. from May 19 to July 7,1982. Hydrogen was added to the reactor feedwater via injection taps in the condensate booster pump casings. Initially hydrogen was added f or five consecutive steady-state periods during which feedwater hydrogen concentrations in parts per billion were: <5 (two days) , 200 (two days) , 400 (two days), 1000 (four hours) and 1800 (four hours). Reactor power during this sequence was held at about 83% of full power (2527 MWTh) .
The remainder of the initial demonstration testing with hydrogen was done from June 3 to June 29 to support electrochemical potential (ECP)
,' measurements and CERT tests in the reactor autoclaves. Except for power decreases for maintenance or weekend surveillance testing, reactor power was close to 100% during this time. The hydrogen addition rate was adjusted to produce the required oxygen concentrations in the test autoclaves (reactor recirculation system water) and ranged:.from 0.6 to 1.3 ppm in the reactor f eedwater. Af ter the cessation of hydrogen addition, there were 10 additional days of CERT testing under normal reactor. conditions to provide raseline comparison data.
i 7-29
In April 1983, Dresden-2 initiated full time operation on hydrogen wa er chemistry. From October 1983 through April 1984, a second series of ECP measurements and CERT tests were performed in the same reactor autoclaves..
5.2 .DRESDEN-2 MATERIATR RESULTS 5.2.1 Corrosion Potential Measurements Electrochemical potentials (ECP) of Type 304 stainless steel, Alloy 600, SA533-B low alloy steel, and platinum were measured against an AgC1 .
reference electrode in an autoclave containing flowing reactor water. In general, the ECPs decreased directly with oxygen concentrations in a manner similar to that obtained in the laboratory, Figure 14. With time, however, at any constant oxygen concentration and conductivity, the corrosion potential of film-forming materials drifted upward. The platinum potential was unique and more sensitive to hydrogen than oxygen concentration. Based on these potential measurements, and the stress corrosion cracking results (discussed below) , it appears that intergranular stress corrosion cracking (IGSCC) of furnace-sensitized Type 304 stainless steel in boiling water reactors can be prevented at corrosion potentials less than -0.325VSHE' 12 agreement with the laboratory test data presented earlier.
During the second ECP measurement campaign at Dresden-2 (October 1983 -
April 1984), the ECP of stainless steel was measured for 3330 hours0.0385 days <br />0.925 hours <br />0.00551 weeks <br />0.00127 months <br /> ( 139 days), Table 11. The potentials obtained during dais extensive coverage were similar to those measured during the first demonstraticn campaign. The normal Dresden-2 environment was characterized by an ECP range for Type 304 ',
stainless steel of -100 to -200 mVSHE, while the hydrogenated Dresden-2 environment reduce the Type 304 stainless steel potential to -370 to -460 mVSHE. Table 11 also presents the average ECP values obtained for Type 304 stainless steel and platinum during this test period while Table 12 presents some typical values for working electrcCas for a one month measurement period.
7-30
The " memory" effect noted in Table 11 is.v'ery important. It appears that during times when HWC is terminated, a " memory" effect occurs in that the potential does not suddenly rise into the range where IGSCC of sensitized Type 304 stainless steel can occur. Instead the potential rises slowly which creates a window of time prior to rising into the cracking F tential region. Typically this window is 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br />, Figure 15.
,. The ECP measurements obtained during the second campaign also revealed that at Dresden-2 there is no significant difference between the ECPs of filmed or unfilmed Type 304 stainless steel and Type 316 stainless steel. This differs from some of the data obtained by ASEA-A'IOM at Ringhals-1 which suggested that Type 316 stainless steel reaches a significantly lower potential in HWC than Type 304 stainless steel. The behavior at Ringhals-1 may not be real.
5.2.2 CERT Test Results Two low alloy steel and three furnace-sensitized Type 304 stainless steel specimens were tested in Dresden-2 environments modified by hydrogen injection, and in the normal BWR environment that contained approximately 200 ppb dissolved oxygen during the initial HWC demonstration. In the modified BWR environments, two different H2 injection rates resulted in dissolved oxygen concentrations of 40 5 and <20 ppb dissolved oxygen.
Type ~304 stainless steel tensile samplss were run. in both of the modified environments, while the SA533 Grade B low alloy steel was tested in the aquecus environment that contained <20 ppb 02 -
For Type 304 sta.inless steel, suppressing the dissolved oxygen to 40 ppb '
- resulted in a decrease in the amount of IGSCC, but elimination of the phenomenon did not occur until the dissolved oxygen concentraticn was <20 ppc. Stress versus time plots, normalized to a crosshead speed of 25 um/h
- C.001 in/h) , Figure 16, are shown for the three different reactor environments. The vertical dashed lines indicate the times the load was removed f rom the tensile specimen when the hydrogen injection system was-When the snut off. The "x"s indicate CERT data obtained in 2883C air.
7-31
curves are converted to stress-strain curves at a crosshead speed of 25 um/hr (0.001 in/h) , each 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> of the testing results in a~ 13.33
, . , specimen strain. The curves clearly show the increased ductility and .
tensile strength (decreased IGSCC) with decreasing dissolved oxygen concentration. The <20 ppb oxygen _ data. is similar to that outlined in h:.g.,
temperature air. The CERT at the lowest dissolved oxygen concentraticn was
. terminated before specimen failure.
The low alloy steel samples during the 1982 campaign, clearly shcwed the benefit of reduced oxygen concentration in stress corrosion response. Both.,
laboratory and in-reactor studies are in agreement as transgranular stress corrosion cracking was eliminated as the oxygen decreased to <20 ppb. '
Since the pressure. vessel steel was tested at 75 um/h (0.003 in/h) , each i 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> of' extension should produce about 4% strain. Transgranular cracks l were found in the sample tested in the laboratory in water containing 200 {
ppb dissolved oxygen, but no cracks were found in the samples tested in the laboratory or reactor in the low dissolved oxygen environment. Since 1cw alloy steel is more tolerant of higher oxygen concentratiens and lower water purity than sensitized stainless ~ steel, any oxygen-conductivity environment a which provides SCC protection for sensitized stainless steel will also provide protection for the low alloy steel.
The 1983-1984 CERT test campaign at Dresden-2 was characterized by a similar set of seven experiments. Four furnace-sensitized Type 304 stainless steel, specimens, one furnace-sensitized Type 304 stainless steel specimen (precracked by IGSCC in a 0.2 ppm oxygen water laboratory autoclave and shipped to the site), one SA106 Grade B carbon steel and one SA508 Class II' ;
low alley steel specimens were tested. The re::ults of these CERT tests plus - ;
j the results of the first campaign and comparable laboratory tests are *
{
summarized in Table 13. ~
Table 13 clearly indicates an excellent correlation between Druden-2 and laboratory CERT test results. It also clearly shows that HWC provides mitigation of IGSCC in furnace sensitized Type 304 stainless steel-even in instances where hydrogen addition has been interrupted. For example, in 7-32 1rv--+-7y -
- - aer-*v ev evr w r+ ry rww ~~te-w- W g-v-- er%we-v,---=vey--%~r**-e--sr+ **-tt--y, g e w
- e rm e,-t-- 4+e
one CERT study (line No. 6) 396 hours0.00458 days <br />0.11 hours <br />6.547619e-4 weeks <br />1.50678e-4 months <br /> of test time and 45% strain were accumulated with >10% of the test time in oxygenated water (36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br /> with no H2 and 02 >200 ppb and 15.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> with 02 at 25-41 ppb) and no IGSCC was identified. However, in CERT No. 5, a 12 hour1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> continuous test period with the hydrogen injection terminated produced minor IGSCC damage. (This test result is also marred by a thermal overload problem which resulted in extremely high thermal stress.) The precracked furnace-sensitized Type 304
.- stainless steel specimen (No. 8) (67 hours7.75463e-4 days <br />0.0186 hours <br />1.107804e-4 weeks <br />2.54935e-5 months <br /> in 200 ppb 0 ) which 2 was carefully documented prior to in-reactor testing, failed by ductile tearing
,- after 301 hours0.00348 days <br />0.0836 hours <br />4.976852e-4 weeks <br />1.145305e-4 months <br /> in Dresden-2. The mechanical parameter of this specimen is similar to non-precracked specimens. It is also important to note that thir test included 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> >40 ppb oxygen. Finally. no IGSCC was identified in either SA508-2 and SA533-B low alloy steel and SA106-B carbon steel in HWC at Dresden-2 despite interruptions in hydrogen injections.
5.2.3 DC Potential Dron Crack Growth Crack growth data versus time and environment on precracked furnace-sensitized Type 304 stainless steel is being obtained at Dresden-2 using the reversing DC potential drop technique as developed by GE Corporate Research and Development. The specimen was precracked in San Jose in the nominal 200 ppb 02 and then transported to Dresden-2 for testing. The K y level for the specimen bounds the crack previously identified in Dresden-2 saf e-end at 27.5 MPa /m (25 ksi /in) . To supplement this program, bolt loaded WOL specimens (one each) of Alloy 600 and unclad SA508 C1. II low alloy steel have also been inserted into the autoclave.
j Figure 17 presents the early DC potential drop crack growth data of the precracked furnace-sensitized Type 304 stainless steel specimen. Although
- he data is preliminary and is clouded by a reactor scram which occurred approximately 120 hours0.00139 days <br />0.0333 hours <br />1.984127e-4 weeks <br />4.566e-5 months <br /> into the test and some other interruptions in nydrogen inj ection, no significant crack growth has occurred on the specimen. This preliminary result verifies the result of the Dresden-2 precracked CERT test discussed above.and the results of the mid-cycle
- n-service inspection (ISI) as presented in Table 14.
7-33
This mid-cycle ISI was performed as required by the NRC to verify the mitigation of IGSCC growth by HWC. The inspection was perfcrmed en November 12, 1983, af ter approximately S.5 months of operaticn with hydroger.
inj ection. The results presented in Table 14 indicate that' despite interruptions in hydrogen injection, no crack growth was detected by ultrasonic testing.
a 9
O e
D O
- l 1
i 7-34 I
5.3 CONCTFSIONS The Dresden-2 materials tests have revealed the following conclusions relative to the effects of HWC on the structural integrity of actual BWR materials in an operating reactor:
,. decreasing oxygen content. During times when HWC is terminated, a." memory" effect occurs in that the potential of Type 304
, stainless steel does not rapidly rise into a range (> -325 mVSHE) where IGSCC can occur. This " safety" window of time appears to be
-10 hours.
- 2. HWC clearly provides mitigation of IGSCC in furnace sensitized Type 304 stainless steel and 'IGSCC of carbon steel and low alloy steel at Dresden-2. These results are in exact agreement with laboratory studies.
- 3. For pre-cracked furnace sensitized Type 304 stainless steel a Dresden-2 CERT test and preliminary DC potential drop studies indicated mitigation of crack propagation in-reactor. These results combined with laboratory pre-cracked pipe tests results and the ISI data at Dresden-2 indicate that pre-existing cracks at Dresden-2 should not propagate during HWC operation.
6.0 ADDITIONAL CORROSION TESTING ON AUSTEMITIC, MARTENSITIC AND PERPITIC BWR STRUCTURAL AT.TAYS 6.1 Im'RODUCTION Tne objective of this broad task was to assess the effects of hydrogen water chemistry on the stress corrosion, general corrosion, and crevice and galvanic corrosion of typical austenitic, martensitic, and ferritic BWR s.tructural materials. The effects of notches and. defected corrosion films are also being investigated.
7-35
)
PWR work has indicated that very low oxygen levels result in higher genera; corrosion and metal release rates for carbon and low allcy steels. Also, with hydrogen addition there are some concerns about the possiblity of-hydrogen embrittlement. Areas of the BWR reactor system of particular concern are ground (unclad) regions of nozzles, the carbon steel feedwater system and martensitic stainless steel pump and valve components.
This section of the report presents the general corrosion rates and ~
metal-to-system release rates, oxide film analysis and stress corrcsien results for the various materials and conditions af ter eight months exposuri-and compares the results in HWC and reference water environments containing 200 ppb 02 or 8 ppm 02*
6.2 RESULTS AND DISCUSSION 6.2.1 General Corrosion - Pre-Filmed Carbon and Lcw Allov Steel To investigate the effect of introducing HWC into an operating BWR with flaw'ed corrosion films on carbon and low alloy steel allcys, rectangular coupons [5.6 x 2.0 x 0.15 cm (2.2 x 0.8 x 0.06 in)) with a hole drilled in one end for hanging were used. The specimens were pre-filmed for 500 hcurs in 288"C (550 *F) water containing 0.2 ppm oxygen. To determine possible effects of a defected pre-film, one-half of the specimens were scratched by a scribe diagonally corner-to-corner on each side completely through the oxide film. Specimens were exposed in triplicate to the HWC envircnment (6 to 14 ppb 02 ,190 to 270 ppb H2 , 0.07 to 0.25 uS/cm conductivity) only. ,
Two heats of both carbon and low alloy steel were used in the "as received" .
condition. The carbon and low alloy steels showed relatively high initial corrosion rates during the first month (129 to 175 mg/dm2 ) follcwed by ,
steady-state corrosion rates of 8 to 19 mg/dm 2/mo (3744 hours0.0433 days <br />1.04 hours <br />0.00619 weeks <br />0.00142 months <br /> of exposure) .
In Figure 18, the average corrosion and release rates from zero time are plotted on oxygen dependency curves compiled by Urbanic' frcm numerous results reported in the literature. The corrosien rates fall within' the range of values compiled for 10 ppb oxygen and the metal release rates are a factor of four less than those for 10 ppb oxygen. These rates will decrease 7-36
4 with further exposure because of the decreased steady-state corrosion rate once a protective film is formed.
The corrosion and metal release rates were not significantly different for any of the carbon and low alloy steels tested. In addition, scratching of the pre-film showed no effect on corrosion or retal release rates for the materials tested.
The oxygen dependency curves of Urbanic (Figure 18) show that a substantial
. . benefit in corrosion and release rates would result if the oxygen content was increased from 10 to 20 ppb. In BWR feedwater under HNC conditions, achievement of an oxygen content 2 20 ppb might require oxygen injection.
6.2.2 General, crevice and Galvanic corrosion - Anntenitic Alloys and Un-Filmed carbon Steel
~ Coupons were machined from Type 304 and 316L stainless steel, Alloy 600 and SA333-6 carbon steel pipe wall and used as shown in Figure 19 for the general, crevice, and galvanic corrosion tests. To form the crevice specimens, two coupons were fastened face-to-face with stainless steel screws with a 25 micron (one mil) stainless steel shim placed between the coupons at each end to form a 25 micron (one mil) crevice gap. To determine crevice effects, cpecimens of like materials were fastened together.
Galvanic . effects were investigated by fastening specimens of different materials together. The general corrosion specimens were exposed in triplicate and the others in duplicate.
O Types 304 and 316L stainless steel specimens showed low corrosion rates and metal-to-system loss in both the simulated normal BWR and HWC environments.
~
The creviced / galvanic condition only resulted in slightly lower initial weight loss for these alloys in both environments.
Alley 600 showed significantly higher corrosion rates in 200 ppb 02 water nan.in the HNC environment, but showed low initial weight loss in both environments. As anticipated, the corrosion rates were higher than for 7-37
Types 304 and 316L stainless steel. No effect cf crevicing or galvan;.c coupling was observed for Alloy 600.
Corrosion rates were very low for the unfilmed carbon steel in the reference environment with relatively high initial weight loss, indicating that the corrosion film that forms becomes very protective in the higner oxygen environment. However, corrosion rates were up to about 40 times
~~
greater in the HWC environment with correspondingly larger metal-to-system loss. Cervices and galvanic coupling appear to have no significant effect.
6.2.3 Oxide Film Studies This task involved the analysis of the corrosion film formed on highly polished Type 304 stainless steel, low alloy steel, and carbon steel specimens exposed to nominal (200 ppb 02 ) and low 02 (50 ppb 02 and 125 ppb H 2) environments. The oxide films were analyzed for thickness and compositional differences after 700 hours0.0081 days <br />0.194 hours <br />0.00116 weeks <br />2.6635e-4 months <br /> of exposure in each environment.
Small amounts of cobalt (1 ppb) were added to the test solutiens to determine its relative incorporation into the nominal and low 0 2 films.
The results of this study indicated that although some differences in the corrosion films on Type 304 stainless steel, SA533-B 1cw alley steel and SA333 Grade 6 carbon steel can be detected as a result cf changes in water chemistry, these differences are not marked af ter 700-hour exposures. The most significant difference seen was the fctmation of a thinner film en stainless steel under hydrogen water conditions. The cobalt concentratien ,
in the surf ace films seems to be independent of water chemistry for all -
these materials, implying no significant change in initial radiaticn buildup, in the low 02 environment. Somewhat icwer concentraticns of cobalt are ,
found in the corrosion films that form on alloy and carbon steels than on Type 304 stainless steel under both water chemistry conditions.
7-38
~
l 6.2.4 Constant Load Tests _ _.fJirbon ancL1qw_ Alloy Steel )
l Smooth and notched tensile specimens were used in the constant load tests.
The lower strength SA106-B and SA333-6 specimens were loaded during test to a rat rection stress of 190% of the 288*C (550*F) yield stress. The higher 1
strength SA533-B and SA508-II specimens were loaded to a net section stress of 110% of the 288*C (550*F) yield stress to avoid exceeding the ultimate l
i I
e f
J i
h e
7-39
tensile strength. One-half of the specimens were notched, the notch being -
designed to result in a stress concentration f actor K , of 2.0.
t These constant load specimens were exposed to the same HWC envircnment as that used in the general corrosion studies. No evidence of hydrogen stress cracking or other adverse effects were observed in either of these two j ferritic alloys af ter 5359 hours0.062 days <br />1.489 hours <br />0.00886 weeks <br />0.00204 months <br /> of exposure. It appears that these alleys should continue to exhibit the same excellent SCC resistance in HNC as is
- currently being observed in BWR service.
{ 6.2.5 rnnnennt rnmA Tests - Anatonitic A11ovs l . The constant. load tests on Type 304 and 316L stainless steel and Alloy 600 were performed on the materials in two conditions: (1) furnace-sensitized at 650*C (1202*F) for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> and (2) welded and then low temperature
- sensitized at 500*C (932*F) for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. For the welded specimens, the pipes were welded with a standard V-groove butt weld. The constant lead tensile specimens were then machined transverse to the welds so that the pipe 1.d. surface, including the weld and both heat affected zones, was
! located in the gage section.
1 i
The welded tensile specimen configuration is shown in Figure 20. The furnace-sensitized specimens were of similar configuration except for the t absence of the butt weld. A load of 4.7 kN (1060 lb) was applied to the specimens af ter environmental specifications as described in Section 6.2.1
{- were achieved, the width of the cross-section being varied for each material, i
to result in an applied stress equal to 136% of the 288'C (550*F) yield ,
stress. The specimens were exposed in quadruplicate. ,
l The only austenitic alloy which experienced IGSCC was a furnace-sensitized
- Type 304 specimen which cracked af ter 1530 hours0.0177 days <br />0.425 hours <br />0.00253 weeks <br />5.82165e-4 months <br /> of exposure to the 200 ppb oxyger environment. No materials cracked in the HWC environment. Although this result is not significant statistically, it does support other observations indicating that HWC mitigates the IGSCC phenomenon.
7-40
I .
6.2.6 Martensitic stainless steel Bent Beam Tests The objective of this brief task was to examine the effect of HWC and high oxygen (8 ppm) high purity water environments on the stress corrosion cracking (SCC) and hydrogen stress cracking (HSC) propensities of various wrought and cast high strength martensitic stainless steel alloys typically used in the BWR for pump shaf ts and impellers, valve trim, etc. The test
.- materials included two wrought alloys (t/pe 410 stainless steel, 3 heat treatments and Type 422 stainless steel, 2 heat treatments) and two cast
,. alloys (CA15 stainless steel, 2 heat treatments and CA6NM, 7 heat tr eatments) .
The configuration of the bent beam specimens used in the tests is shown in Figure 21. The specimens were bent over a stainless steel fixture having a radius of curvature to give 1% strain in the outer fibers and were held at each end by carbon steel bolts. The specimens were removed from the autoclave at approximate total exposure times of 1, 2, 4, 8, and 16 months and examined for cracks (without removal from the fixture) using a stereo microscope at magnificaticns up to 40X.
The results of this scoping test program revealed that wrought martensitic stainless steels Types 410 and 422 and cast martensitic stainless steels CAIS and CA6NM, in several welded and non-welded, annealed, and tempered 593'C (1100*F) to 677'C (1250*F) conditions showed no evidence'of cracking in 8 ppm oxygen 288*C (550*F) bigh purity water environment. These same alleys and heat treatment conditions also showed no evidence of cracking in
,' a simulated HWC environment (high purity 288'C (550*F) water containing nominally 10 ppb oxygen and 230 ppb hydrogen) . Therefore, it appears that the presence of higher amounts of dissolved hydrogen in the coolant (- 10X nominal) does not result in hydrogen stress cracking / hydrogen embrittlement in these materials over this test period.
7-41
6.3 CONCLUSION
S These test results indicate the following conclusions concerning the effects of HWC on the general, crevice, galvanic and stress corrcsicn of BWR structural materials:
- 1. The corrosion rates and metal-to-system release rates are not significantly different for the two carbon steels and the twc lcw alloy steels tested in HWC. However, the corrosien rates were significantly higher in HWC than in the nominal environment. -
Scratch defects in the oxide film of carbon and low alloy steels showed no significant effect on corrosion or metal-to-system release rates on exposure to HWC..
- 2. Types 304 and 316L stainless steel corrosien and metal release rates are very low in both HNC and reference 200 ppb 02 water.
- 3. As anticipated, Alloy 600 corrosion rates are much faster than those for Types 304 and 316L stainless steel, but initial weight gains are lower. Corrosion rate and metal-to-system loss are lower in the BWC environment than in the nominal EWR environment for Alloy 600.
- 4. The initial carbon steel corrosien films are_ less protective in the HWC environment, resulting in up to 40 times higher corresion rates than in the nominal environment and correspondingly higher ,
metal-to-cystem loss. However, this acceleration in corrosion .
rate can be controlled in practice by injecting oxygen into the f eedwater system. ,
- 5. Crevices appear to have no significant effect on corrosion rates or metal-to-system loss except for slightly reduced rates for Types 304 and 316L stainless steel.
7-42
- 6. Oxide film studies revealed that thinner films are formed on stainless steel in low 02 environment while the films on carbon and low alloy steel are unaffected. Information derived from the injection of cobalt into the water indicated that low 0 has no effect on the cobalt concentrations in the initial films formed on the steels.
- 7. SA106-B and SA333-6 carbon steels loaded to 190% of the 288'c
.. (550*F) yield strength and SA333-6 and SA508-2 low alloy steels loaded to 110% of the 288'C (550*F) yield stress show no cracking af ter exposure for 5,359 hours0.00416 days <br />0.0997 hours <br />5.935847e-4 weeks <br />1.365995e-4 months <br /> in the HWC environment or up to e,
11,371 hours0.00429 days <br />0.103 hours <br />6.134259e-4 weeks <br />1.411655e-4 months <br /> in reference 0.2 ppm 02 water environment. .
- 8. Types 304 and 316L stainless steel and Alloy 600 showed no evidence of cracking when stressed at constant load to 136% of the 288'C (550'F) yield strength for 5,359 hours0.00416 days <br />0.0997 hours <br />5.935847e-4 weeks <br />1.365995e-4 months <br /> in HWC environment.
However. exposure to the nominal environment did result in cracking of one furnace-sensitized Type 304 stainless steel specimen af ter 1,530 hours0.00613 days <br />0.147 hours <br />8.763227e-4 weeks <br />2.01665e-4 months <br />.
- 9. Bent beam tests on martensitic stainless steel specimens indicated that excess hydrogen in the BWR environment does not induce any loss in ductility in high strength alloys. No cracking was observed in either the HWC or the control environment; the latter is in agreement with the excellent fielii perf ormance of this f amily of alloys.
A O
9 7-43
7.0 SUMMAPY OF CONCLUSIONS AND REMMMENDATIONS Overall. the results of these laboratory and Dresden-2 WC materials programs are highly encouraging. Hydrogen water chemistry clearly improves the stress corrosion cracking performance of several BWR structural materials as follows:
,. 1. WC prevents IGSCC of sensitized Type 304 stainless steel and Alloy 600 and 'MSCC of low alloy and carbon steel.
~
- 2. WC clearly provides mitigation of IGSCC in furnace sensitized Type 304 str.inless steel and 'MSCC of carbon steel and low alley steel at Dresden-2. These results are in exact agreement with laboratory studies.
i
- 3. For pre-cracked furnace sensitized Type 304 stainless steel a Dresden-2 CERT test and preliminary DC potential drop studies indicated mitigation of crack propagation in-reactor. These results combined with laboratory pre-cracked pipe testis results 3 and the;ISI data at Dresden-2 indicate that pre-existing cracks at Dresden-9 should not propagate during HWC operation.
- 4. The ECP of all the tested BWR structual materials decrease with decreasing oxygen content. During times when WC is terminated, a " memory" effect occurs in that the potential of Type 304 stainless steel docs not rapidly rise into a range (> -325 mVSHE) where IGSCC can occur. This " safety" window of time appears to be 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br />.
s( -
y
- 4. WC prevents initiation and propagation of IGSCC in welded Type 304 stainless steel piping at high stress levels (twice ASME Code allovable). A EDI of >25 has been achieved on crack initiation for full size Type 304 stainless steel piping in the
. WC environment.
I 7-44
_ __ _ _ _ _ - __ _ . . _ _ . ~ . . _ - . _ . _ _ - _ . _ _ _ . . . . . _ . - _ . _ ~
4 l
l
- 6. For Type 304 and Type 316 Nuclear Grade -stainless steels, icw !
alloy steel and carbon steel, no measurable crack growth
observed in HNC even at high stress intensities. Measuracle
, crack growth is observed for Type 304 stainless steel and carcen j steel in the normal BWR environment.
- 7. Cyclic crack propagation rates are significantly reduced in HWC for stainless, low alley and carbon steel even at high stresc '
intensities.
~.
- 8. No detrimental effects such as hydrogen embrittlement have been found for high strength materials such as Alloy 600, Alloy X-750 and martensitic stainless steels in the HWC environment.
- 9. The general corrosion rates of Alloy 600 and stainless steel are reduced in the HWC environment, but there is an acceleraticn in general corrosion rates of carbon and low alley steel. (This potential problem can readily be addressed, if deemed necessary, by oxygen injection into the feedwater system.)
The results obtained from in-reactor studies and from the laboratory tests conducted under this program correlate well, and it appears that the laboratory results are directly applicable for predicting materials behavicr in Dresden-2.
All evidence available to date suggests that HWC has been effective in ,
mitigating IG5CC during reactor power operation at the Dresden-2 staticn. -
6 7-45
REFERENCES
- 1. J. Alexander, et al., Alternate Alloys for BWR Pine Acolications. EPRI, October 1982 (EPRI NP-2671-LD).
- 2. J.C. Danko, et al. , "A Pipe Test Medbod for Evaluating the Stress Corrosion Cracking Behavior of Welded Type 304 Stainless Steel Pipes,"
, Properties of Steel Weldments for Elevated Temperature Pressure
. Containment Applications, MPC-9, ASME Winter Meeting, San Francisco, California, December, 1978.
- 3. D.A. Hale, A.E. Pickett, " Materials Perf ormance in a Startup Environment First Semiannual Progress Report May 1981 - January 1982."
NEDC-23492-1, EPRI Contract RP-1332-2, April 1982.
- 4. J.D. Heald and R.M. Horn, " Slow Rising Load Test Under In-Service
' Environment: Inconel 600 and SA508 Class II Low Alloy Steel," Final Test Report 81-509-67, September, 1981.
- 5. E.L. Burley, et al. , "Orycen Suncression in Boiling Water Reactors -
Phase 2 Final Report," DOE /ET/34203-47 (NEDC-23856-7) , October 1982.
t
- 6. M.E. Indig and A.R. McIlree, Corrosion, 25 28, 1979.
- 7. W.L. Clarke, R.L. Cowan, and J.C. Danko, " Dynamic Straining Stress Corrosion Test for Predicting Boiling Water Reactor Materials Performance," Stress Corrosion Cracking - The Slow Strain Rate Technique. G.H. Ugiansky and J.H. Payer, Eds, ASTM 1979 (ASTM STP
. 665).
E. J.E. LeSurf, "The Water Chemisty of CANDU PHW Reactors," Paper Uc. 21, L presented at International Conference on Water Chemistry of Nuclear Reactor Systems, October 24-27, 1977.
7-46
Table 1 TEST MATRIX FOR HYDR 0 GEN WATER CHEMISTRY PIPE TESTS Test Condition / Specimen Number Heat Normal 8WR Water Hydrocen Water Chemistry Material Number 200 ppb 02 50 pob Op- 2G ::: ::
Reference Soecimens Type-304 SS 04836 AWC-5 Type-304 SS 04836 AWC-6 Type-304 SS 51416 AWC-7 Type-304 SS M2152 AWC-8 Crack Growth Specimens Type-304 SS 04836 AWC-1 (Precrack) ---* AWC-1 Type-304 SS 04836 AWC-2 (Precrack)*
Type-304 SS 04836 AWC-3 (Precrack) -
- AWC-3 = AWC-3 Type-304 SS -04836 AWC-4 (Precrack)*
Type-304 SS 04836 AWC-9(Precrack) - AWC-9 Type-304 SS 04836 AWC-10(Precrack) --> AWC-10 Type-304 SS 04836 AWC-11(Precrack) --* AWC-11 Type-304 SS 51416 AWC-12(Precrack) --* AWC-12 Type-304 SS M2152 AWC-13(Precrack) --*- AWC-13
- Type-304 SS 04836 AWC-15 (Precrack)
- AWC-15
- Type-304 SS 04836 AWC-16 (Precrack)
AWC-16 =
' Crack Initiation Specimens .
~-r .
- Type-304 SS 04836 AWC-17 -
- Type-304 SS M1989 AWC-18
- Type-304 SS M2152 AWC-19
- For metallographic crack depth determination
- New tests
- u. .. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . . _ _ . _ _ _ . _ _ _ . _ . _
.7 s' * ,- ,
Table 2 ESPOSURl!S AND RESUI.TS OF llWC PIPE TESTS ( )
200 PPB 02 Water 50-70 PPB 02 Water 20 PP8 02 Water (IMC)
Specimen llea t llours" Haxi' mum Hours Maximum Hours Maximunit~
2 Nmnber Number Exposure 2
Depth nun __{ mil s1. Exposure Depth, etn (mils)2 Exf_o_sure Depth, nun (milsj _
Ee_f e_rence_ Spec im_ ens AWC-5 04836 2044 Failed AWC-6 04836 1645 Failed AWC-7 51416 1059 Failed AWC-8 M2152 1339 Failed.
Crack Growth Specimenss AWC-1 04836 308 0.38(15)3 844 Failed AWC-2 04836 308 1.40(55)3 877 1.52(60) 3955 2.54(100)
AWC-3 04836 1075 0.25(10)3 .-
AWC-4 04836 654 1.12(44)
AWC-9 04836 ,
847 1.27(50)f 3974 2.67(105)4 AWC-10 04836 847 1 . 9 51 77 4ila 1.95 f77 AWC-Il 04836 848 1.42(56)3 ) 3082 1.42 L56)1 AWC-12 51416 847 2.54(100)3 3911 >5.08(>200)
AWC-13 M2152 848 >5.08 (>200)3 3082 >5.W W AWC-15 04836 343 1.27(50)3 5310 (5)
AWC-16 04836 343 1.02(40)3 1285 Failed Crack initiation Specimens _
AWC-14 03165 8*/ No Cracks 4047 No Cracks (UT) '
AWC-17 04836 5122 No Cracks (UT)
AWC-18 M1989 7580 No Cracks (UT)
AWC-19 M2152 7585 No Cracks _(UT)
(1)~'llGscept AWC-14 are Type-304 stainless steel. AWC-14 is Type-316 NG stainless steel.
(2) As estimated by UT inspection unless.noted otherwise.
(3). Piecrack depth.
(4) Measured on metallographic section.
(5) UT complete, metallography to measure crack growth.
i; u'L;[: ,
Table.3 SCC CRACK GROWTH TEST RESULTS IN HWC (20 2 15 ppb 02 , 125 2 25 ppb H,, <0.2 pS/cm)
HNC -
Material Growth Rate Stress Intensity
'FS* Type 304 No Growth ** K 1 31.2 MPa 5 (28.4 ksi /in.)
FS Type 316 NG No Growth K 1.30.1 MPa 4 (27.4 ' ksi 'Tii. )
SA508 No Growth K$ 50.9 MPa 5 (46.3 ksi .'iii. )-
SA333-6 No Growth K$ 44.8 MPa 5 (40.3 ksi ,'Tii.)
- FS = Furnace sensiti ed 621 C (1150 F/12h)
- Below detectable limit, i.e., growth rate <8 x 10~ mm/s (3 x 10 in./s).
Table 4 SCC CRACK GR0hTH TEST RESULTS IN 200 PPB OXYGEN WATER Material Growth Rate Stress Intensity FS Type 304 8.9 x 10 -8 m (3.5 x 10~9 in.)/s K = 17.3 MPa '5 (15.7 ksi ,TR.)
FS Type 316 NG No Growth K 5 26.7 MPa 5 (24.3 ksi .TR.)
SA 508 Incipient Growth K = 49.1 MPa 5 (44.7 ksi .TR.) ,
SA 106-8 1.4 x 10-7 m (5.2 x 10-9 in.)/s K = 44.0 MPa 5 (40.0 ksi .'Tii. ) ,
. Table 5. PATIGUE CRACK GROWTH TEST RESULk'So
- Crack Growth Rate Cyclic Stress Inter sity mm/cygle- (in/cycge) Factor of MaLet ial ELCQuency-s _cph En.vitonment HEa 6 (KsIt in) x10 ( x10 ) ImprovemenL i
'S Ty pe 304 0.74 HWC 30.9 (28.1) 69 (2.7) 3 200 ppb 02 30.3 (27.6) 254 (10) --
7.5 HWC 25.8 (23.5) 16 (0.63) 3 4
- 200 ppb 02 36.5 (33.2) 50 (2.0) --
1 SA 508-2 0.74 HWC 34.1 (31.0) 7.6 (0.3) 7
- 200 ppb 02 37.6 (34.2) 50.8 (2.0) --
7.5 HWC 33.8 (30.8) 1.7 (0.07) 450 200 ppb 02 33.7 (30.7) 760 (30) --
i 4
SA 333-6 0.74 HWC 30.4 .(27.7) 5.6 (0.22) 20 200 ppb 02 30.6 (27.8) 109 (4.3) --
.t 7.5 HWC 24.2 (22.0) 0.33 (0.013)- 10eq 200 ppb 02 23.7 (21.6) 330 (13) --
- Since there is no contr ol data for Type 316tJG stainless s* eel, no cor"'.r t son c 'uld be rade.
Table 6 CONCENTRATION OF GASES Gas Phase (Pressurjzed , Tanks) Acueous Phase 0 H 2 2 . 0 2 N5 ppb (caiculated) 2.5 'O 1000 0 0.50 0 200 0
- 0.50 6.5 200 102 0.125 0 50 0 0.125 7.7 50 120 0.09 36 0.08 7.7 32 120 S
+
Table 7 CERI RESULTS FOR MATERI ALS TESTED IN tilGli PURITY WATER AT 274"C (525'T)
Dissolved Gases (ppbj actum(g)
Inlet Effluent Material and -- "5 NS(2) I3I n t1
- 0 11 T R.A.(4I Elonga tion Condition 2 2 2 2 MPa ksi MPa ksi I(h ) (%) Fracture Morphology _
_(%)
T-304SS W 200 0 180 0 610 88.6 429 62.2 246 29.8 29.2 20% IGSCC; 80% Ductile .
T-304SS, W+LTS 200 0 180 0 600 87.1 409 59.3 244 31.5 26.8 30% IGSCC; 70% Ductile T-304SS, W+LTS 200 100 100 10G 694 100.7 403 58.5 236 41.9 27.1 25% IGSCC,**75% Ductile T-304SS, W+LT5 45 112 19 112 1570 228 409 59.3 294 74 36.3 Ductile SA533B 210 0 205 0 895 129.9 601 87.3 150 32.8 17.2 16% TGSCC; 84% Ductile Shallow pits, deep
, auxiliary cracks SA533B 200 100 65 100 1700 246.7 637 92.5 203 62.8 21.9 100% Ductile SA5338 45 112 13 112 1699 246.6 623 90.4 188 63.3 24.4 100% Ductile; minor pits A 600, W+LTS 187 0 180 0 1072 155.6 559 81.1 362 47.9 t.2. 8 100% Ductile-A 600, W+LTS 30 125 '10 125 1103 160.1 525 76.2 342 52.4 40.7 100% Ductile SA333 Grade 6 40 125 5 125 1076 156.2 471 68.4 262 56.2 29.6 100% Ductile (1) Maximtsn load / failure cross section (2) Maximum load / original cross section (3) Failure time in nours (4) Reduction in area ,
- 11 c ncentration calculated for gas input 2
51 TGSCC also notel W = Welded
! LTS = Low Temperature Sensitized [500 C (932'f)/24 h]
Table 8 -
LABORATORY CERT TEST RESULTS IN LOW OXYGEN WATER Spec I.D. Strain Rate Time to Failure Max. Stress Material (min-1) RA Elongation (h ) MPa (ksi) M (2) Fracture Mode A106-1 SA1068III Carbon Steel 1 x 10'* 46 496 (72.0) 41.6 28.6 1001 Ductile A106-2 SA106B I}
Carbon Steel 1 x 10~4 39.5 494 (71.7) 35.6 26.3 1001 Ductile A533-1 SA5338 Cl IIII Low Alloy Steel l'x 10'4 47 571 (82.8) 70.4 32.0 1001 Ductile A508-1 5A508 C1 IIIII Low Alloy Steel 1 x 10'4 43.5 581 (84.3) 75.0 28.6 1001 Ductile Notes:
- 50 ppb 02 , 230 ppb H2 , < 1 pS/cm c nductivity at 288*C (550*F)
(1)
Specimen was creviced by wrapping and spot welding thin stainless steel shim stock around the 3.1 asu (1/8 in.) diameter test section.
(2) This specime. was tested witimut a crevice.
I i
e . O %
~ -
-: =; -
i
- Table 9 SET TEST-MATRIX IN 0.0lN Na 2 50 4 AT27d*Q
.' UNDER POTENTIAL CONTROL, STRAIN RATE 2 x 10~* min-1 Controlled Potentials V
SHE
- Type'304 SS, welded + LTS* (500'C/24 hours)-
-0.100, -0.500 Alloy'600, welded + LTS (500"C/24 hours)
-0.100, -0.500
= Alloy ~x-750, three-step heat treated (1038'C/1 hour, -0.100. -J.700 900'C/24 hours, 704'C/20 hours)J
- Low Temperature sensitized 4
'I
=
9 e
x -m+c y.--,,' -
Table 10 l
l SET RESULTS IN 0.0lN Na 2 40 AT 274*C (526 F)
I (STRAIN RATE 2 x 10-5/ MIN)
I l
Material and Electro- III 83I Metallurgical Chemical Fractare Stress III UT5 R. A.I4I Condition Potential 'Til Ma ksi MPa T'( h ) (1) Elonlation t) Fracture morphotoa, Type-30455. W*LT5 -0.100 35.1 242 33.7 232 36 3.9 4.8 1005 ICSCC Type-30455. WELT 5 -0.500 118.3 1228 68.3 471 277 61.1 36.4 >955 Ductile, some TG5CC initiation I Alloy 600 W*LT5 -0.100 111.3 767 77.3 533 233.5 30.6 31.1 M IG5CC. 305 Ductile Alloy 600 Weli5 -0.500 180 1240 83.2 573 320 53.8 43.3 1005 Ductile Alloy I 750 -0.100 275 1895 168.5 1861 306 39.1 31.2 1005 Ductile 3-step h st treated Alloy X-750 -0.700 258.4 1780 165.4 1140 283 36.0 31.5 1001 Ductile 3-step heat treated (1) Man sawn load /f ailure cross section (2) Maa tsua load / ort <3:nal tro u secteun (1) latlure time in hours (4) Reduction in area
' e' .
i TABLE 11 ECP MEASUREMENTS AT DRESDEN NPS 10/83 THROUGH 4/84 o TOTAL TIME MEASURING.
ELECTROCHEMICAL P0TENTIALS = 3330 HOURS (139 DAYS)
.. o TIME MEASURING ECPS WITH WC OPERATING = 3160 HOURS
~
o WC NOT OPERATING DURING ECP MEASUREMENTS = 17 0 HOURS (95% EFFICIENCY)
AVERAGE ECP MEASUREMENTS Vggg 304 ss el WC WITH HYDROGEN ON -0.430 -0.690 WC WITH HYDROGEN OFF -0.300(l) -0.010 NORML BWR -0.200Cl)
Cl}ECP OF 304 SS DURING NORML HIGH OXYGEN CONDITIONS
> -0.200 V (SHE). " MEMORY'EFFECT OF RECENT WC OPERATION DECREASES OBSERVED VALUE
i 4 l.
V Table 12 4
j.
Range of Riectrochemical Fotential and Oxygen j
im Hydrogea Water Chemistry, Nov. 19 to Dec. 18, 1983 at 273I2*C i
l Fotentials Dissolved Volte (SHE)
Oxygan Coad setivity -
304 SS
{ (PPb ) ( pS/cm) 304 SS (filmed) Cold Platinus Autoclave i
30 to 6 0.16 to 0.11* -0.40 to -0.46 -0.44 to -0.53 -0.52 to -0.62 -0.69 to -0.73 -0.39 to -0.46 1
4 f
- 0.29 pS/cm transient on 12/2 i
i I
+
e e
Table 13 Itesults of Dresden-2 and Laboratory HWC CERT Tests Test K I
Time To Time Off ' Elongation Material. Location 02 (ppb). (uS/cm) Failur e (h) HWC (h) % Result 2 T-304 268 108
- 1) 12 FS D-2 0.29 0 70% IGSCC
- 2) PS T-304 D-2 40 0.37 143 2 20 35gIGSCC
- 3) PS T-304 D-2 <20 0.29 >297 4 38 DF i 4) PS T-304 D-2 5-20 0.19 208 4 5 NM 5 DP t
' 5) FS T-304 D-2 5-23 0.17 181 15 NM Minor IGSCC along gauge i 6) PS T-304 D-2 3-30 0.13 396 36 45 DP
- 7) PS T-304 D-2 7-19 0.09 400 25 46 DF
- 8) FS T-304PC 6 D-2 12-20 0.09 301 7 7 , 40 No IGSCC Extension i 8 9
- 9) PS T-304 VNC 195 <0.1 156 NA 17 85% IGSCC
- 10) FS T-304 VNC 15 <0.1 262 0 NA DF
) 12) SA 533B D-2 5-20 0.29 63 0 24 DF
- 13) SA 533B VNC 200 <0.1 43 NA 11 40% TGSCC I 14) SA 533B VNC 12 <0.1 60 0 22 DF 3 15) SA 508-2 D-2 12-18 0.08 52 11 0 NM DF
{ 18) SA 106B VNC 50 <1 40 0 29 DP i
- 1) K = Conductivity . 8) VNC = Vallecitos Nuclear Center
- 2) PS = Furnace Sensitized 621'C (1150*P)/24 hr 9) Not Applicable, i.e., not a HWC Test j
- 4) Thermal Overload Ended Test SA 508-2, SA 106B and 1 mil /h for Type 4 5) NM = Not Measured to Date 304 Stainless Steel i 6) PC = Precracked in 200 ppb O 11) Motor Failure, Specimen Fractured Manually
- 7) Plus 67 Heurc Prceracking (36 hr total) 12) Creviced 1
1 i
i
TJ ~1 l
~.
l Table 14 Mid-Cycle ISI Results from Dresden-2 ISI RESULTS
Date '
Weld , April _29. 1983 November 12, 1983' 28" Safe End l" long, 16%' deep 1" long, 13% deep PS2-201-1 12" Riser (Two Cracks) 0.25" long,-17% deep 0.25" long, 15% deep PD5-D20 0.25" long, 19% deep 0.25" long, 17% deep 12". Riser (Two Cracks) 0.50" long, 19% deep 0.50" long, 18% deep PDS-D5 0.25" long, 14% deep 0.25" long, 16% deep 1
Percentage Through Wall I
D e
e V
o
O
. Tdj:1 .
o I .
O'T H i.-]
o a ( :s i
4 g 3% '
! Ed I!
k je en f5$
,I \ / p, ,-
s;
~
he o#g w il og O 2 C
!!!- IIc g
~ *C w
,O M
!!u EEE l,
- 0 ik 5 I
= i'
) tr
.c
]
o
~
U w' 9 t $ SM n 3 u
E5
$IN $
5!
!.ss e w-
-~
Ei
.f l'eE r MT:
.o ; g 5 y
i'
$3
- ! t
, (
EEE> {E20 h h
$5 55! 3 33
- U Oy!y E u $U C b
- - 4
" M *2 C3
/ \( 5;; )
4L -
6
. 2 m=
g . .
- k 2
$$I T E$ I l ]f --v v . = g: z ,
. n
- L e t $ 5E $ "h-* * *
- T 2 -50 t l7,; $
- 91 5 . I EE 5 - it ,
EI E-n E3 n
h/4j* .e .
3, 3
= ,
W
l' ENVCONMENTAL WATER STRAIN G AGE &
f OUTLET THET.MOCCUPLE OPENINCs 1 ._
ENVIRONMENTAL WATER
- INLET g I e
I s <j kt
,, vy M f k'/ ,
I '/
\ !
- .i ,
/ 10 cm (4 m.1 PIPE TEST N
- a / SPECIMEN TEST WELOS h
y I
i c )
,ll sN / SAFETY HOUSING s !
n'f s [
Q~:+ '
c.+
l
, =
l J TH ERM A L INSULATION k,,):'! -
5 k '
1 hf.s.','
e .
- I/
/
e s4 l a /
';Is* = . / 8 lt ! ,
PRESSURE 1 1
N ? -
P l
SWITCH 'lld y (4 , ,
i/
' 4 i f ',,
CHECK [' ' '
- O VALVE j y < ,N f
- s. c
- d. / ,
',r, t '
/
/
STEAM VENT lL >g ,
VENT ,
TO ORAIN V 7 HYDR AULic FLugo gNLgy Qv I I
/ .
MYOR AULIC FLUID OUTLET s
10 cm (4.n i VERTICAL PtPE TESTS .NO Figure 2. Pipe Test Specimen Loading Stand 1
_ _ - _ _ _ _ _ _ _ . _ _ _ . _ _ _ _ - _ _ - _ _ _ _ _ _ _ _ _ _ - - _ _ - _ - - _ - _ _ _ _ _ - _ - = - _ - . . - .
1 R
2su r .
/
- -5" = = 20 H .+su; : .s%
, i Figure 3. Loading Waveform used for Hydrogen Water Chemistry Pipe Tests.
(ft = minutes, H = hours, Sg = ASi1E Code Allowable Stress) e
V
/
Y k COO ascu x -
s/
Figure 4. Typical WOL or Compact Tension Specimen for Crack Growth Rate Study O
e ei imm mi eiiie.M
LOAD SCL SCL PHASE 2 PHASE 3 l CL CL PHASE 3 SCL(1) A 2 > ,
PHASE 1 ,
,. CL(2)
PHASE 1
M L l3 <
=
TIME (1) SCL - SLOW CYCLE LOADING (2) CL - CONSTANT LOAD Figure Sa. Stress Corrosion Test Loading History SLOW CYCLE LOADING CONSTANT LOAD
. WAVEFORM j
{ O k 6.2 KN (1400 LB)
- - 3 UNLOAD / LOAD CYCLES PER DAY TO MEASURE 80 MINUTES SPECIMEN COMPLIANCE 1 MINUTE TIME Figure 5b. Slow Cyclic Loading Figure Sc. Constant Load Detail Wavefom Detail
I f
MAXIMW A
- MINIMUM LOAD
- 0.8 LOAD MAXIMUM LOAD i
i
\
t s's' J- .- .- -
l 5% M ' 90% 2 M 5% MINIMUM
! LOAD i
Figure 6. Cyclic Loading Wavefom i
9 I
{ '
1 I
l I
l
STRESS INTENSITY RANGE, AK [MPa AJ
-a 10 20 30 40 50 10 1 I I I l -
~
10 '
~ ~
10 * -
o.
g D t -
2.
-3 1
' 10 e
& I W -
e : -W g -
E
-s
=
E 1O 2
f O open Symbols (200 ppb 02 )
~
h e
4 Solid Symbols (Hydrogen Water Chemistry) m O SS-21 Type-304 SS [11500 F(6210C)/16 hr], O x -
EPR = 21 Heat 03580 2
-4 M
m Q SS-7 Type-304 SS [1150 0 F(621 0 C)/2 hr], ~~"
1O
" . EPR = 13, Heat 03580 -
y
$SS-50 Type-304 SS [19500 F(10660 C )/1 hr], .~
S - EPR = 1.1, Heat 04904 . N j E SS-51 Type-304 SS [11500F(621 0 C)/16 hr],
d
>- EPR = 35.0, Heat 04904 >-
U g -S G SS-03 Type-304 SS Weld Heat Affected Zone ~
+ [9000F(4820C)/24 hr), EPR = 3.4, Heat 808228
- $ SS-86 Type-316 NG SS [11500 F(6210C)/16 hr),
~
EPR = 0.36, Heat 59076
~
10 "
~
I I I
- 2 '7 I O 10 20 30 40 50 STRESS INTENSITY RANGE, AK CKs t En]
Figure 7. Comparison of Cyclic Crack Growth Data (0.74 cph, R=0.6) in HWC versus !!orminal Environment for Furnace Sensitized Stainless Steel w- g*- - - - - - ---------------------ee- e-e +-----wW wc ?---T+r--e-u--w- --'--*w -w -* - - -c- - - - w -- - - ' = - - - - -------m-F -
STRESS INTENSITY RANGE, AK [MPa82
_3 10 20 30 40 50 10
- 1 I I I I -
10'
-4 - -
10 -
e -
e u
o x
- .- V _ ,0 f
c -
e E
~
W -
A -
W E
I -5 - -
2 10 -
o -
o r
x -
x e -
e x -
8 - 4 x
~ 10 W A $
u . Y ~
u J
~
J N
U O open Symbols (200 ppb 03) N
-6 9 Solid Symbols (Hydrogen %'ater "
U 1O I Chemistry) .
6 CS-16 SA508-2 Low Alloy Steel ~
A LAS-37 SA508-2 Low Alloy Steel 7 CS-4 SA333-6 Carbon Steel = .
T CS-84 SA333-6 Carbon Steel ---
.O~ .
~_
gg -7
~
l l I l O 10 20 30 40 50 STRESS INTENSITY RANGE, AK CKst E 3 Figure 8. Comparison of Cyclic Crack Growth Data (0.74 cph, R=0.6) in IIUC versus Nominal Environment for Carbon and Low Alloy Steel
1 l
STRESS INTENSITY RANGE, AK [MPaA]
-3 10 20 30 40 50 10 _
g ;
i i i ,
10 '
10* -
o o -
0
~
10 '
S _ : I m . m d -
d E
-5 ~
i 10 --
E
- g
- t5 @
e d $
10
-4 c -
- h 2
u u ,
O Open Symbols (200 ppb 02 ) ~
u
] 9 Solid Symbols (Hydrogen Water Chemistry) _J u O SS-14 Type-304 SS [11500F(6210C)/16 hr),
~
U _g EPR = 27, Heat 03580 -
M u
10 -
O SS-u, Type-304 SS [1150 F(6210C)/2 hr),
EPR = 10.5, Heat 03580 -
E SS-52, Type-304 SS [11500 F(621 C)/16 hr),
_ EPR = 38.4, Heat 04904
~
_ $ SS-53, Type-304 SS [11500F(6210C)/2 hr], -
1O '
EPR = 5.4, Heat 04904 ~
- A SS-87, Type 316 NC SS [11500F(6210C)/16 hr], -
~
EPR = 1.4, Heat 59076 ~
~
V SS-88, Type 316 NC SS [11500F(621 0 C)/2 hr],
EPR = 0.11, Heat 59076 ~
-# I I I '
10 O 10 20 30 40 50 STRESS INTENSITY RANGE, AK [KsIEE3 Figure 9. Comparison of Cyclic Crack Crowth Data (7.5 cph, R=0.6) in HWC versus Nominal Environment for Furnace Sensitized Stainless Steel
STRESS INTENSITY RANGE, AK EMPaE]
-3 10 20 30 40 50 10 .
~
10 "
-4 -
10 --
7 : -
7 u -
u x
~
10 C - -
3 w .
u r
k -
E
!10 -5 ~
h [
8 : -
e o .
w x -
-4 x E -
O open Symbols (200 ppb o ) 7 IO 6 -
9 Solid Symbols (Hydrogen Water ~ h U
y Chemistry) -
y H -
A CS-15 SA508-2 Low Alloy Steel -
~
d A LAS-38 SA508-2 Low Alloy Steel -
d y ~
7 CS-3 SA333-6 Carbon Steel .
U 10 " -
T CS-81 SA333-6 Carbon Steel
~
g .
-5 10
~
10 '
~
I I I I O 10 20 30 40 50 STRESS INTENSITY RANGE, AK [KsiEn]
Figure 10. Comparison of Cyclic Crack Growth Data (7.5 cph, R=0.6) in HWC versus Nominal Environment for Carbon and Low Alloy Steel
Ia'"'") 3DNVH3 H1DN31 )t3VW3 I E R 8 S S R R g a y y R I i i i i i i i , 'I i
E e' '
g m
O O a 5
0 0 8 O g _ g 5
~
O E d O *= a 8
~.-
m O E5 5 O
0 0 g$g _ g g O i2 0 o I" l *3m C =
_ g g .C O y O 2 O o e SO o 'e o _ g 3 6D o af O o O
- E =
s cn o f. Oi<-le * '-
' a
- 2 " ';; %C 5
d be $M u: O O O 2 C -
3 8
' ! ! l l l l l l
,4 -
e : e ! - e ,
O O
- O T
- O I** 'WI 3DNVH3 M1DN31 )saygg i
, . - . - - ~ . - _ . ._. - ,,. -- . , . . . . - - . . - , _ - _ - - - . , . , - . - - . _ _ . , - _ , _ _ . . , . - . . . - . , , , _ . , , , , -
0 -
h BWR 3 -200 - h SCC y ~ SCC w
4 - h NO SCC
-800 HIGH PL/RITY WATER 2740C l
@ ' ' ' f f fffaet! , , ,,,,,,
0 10 100 t000 02(Doba i
Figure 12. Relationship Between Dissolved Oxygen and Potential to IGSCC of Welded Type 304 Stainless Steel e
e l
f I
1
8r DRIVE R E DUCIN'G 'DI A L GEARS WORu MOTOR INDICATOR GEAR .
h l_ I
\ylT x
O l N"M L'WN\\M I f"3 LOAD CELL " u,,
4 #
' 8 # AND PRES $URE
/ FITTING UNSTRESSED SPECIMEN % % (
,r r
ELECTRCCE % l
~
.r
[ _ p ELECTROLYTE PULL RCD g
- i N , d =d k ==+ E F F LUE NT l WM s
\bk , s;.
$3 h
.\W#
[s l w .
N .
- L,1
]
SPECIMEN g \ '
I'L ' "
i LE AD WIR E j g. ,,5
\ CHLOAl E
' ', REFER ~NCE i
\
6 i E LECTRODE TINSILE SPECl JEN.
ELECTRICALLY /
\ @1-
_% m
@ N ,
INSULATED '. N SUPPORT FROM GRIPS -
\ COLUMNS w
+, J AUTOCLAVE INFLUENT t
f
~
N. N 7' "V N. N yN.y N.
s _,N. y xN.x N. w y_ N. .v5 T. x. x N. N. ' %. 5. N."N J'7 .y ~
Figure 13. SET Facility i .
2o0 - '
O -
~
NOfteAAL G TV svust svATER 525*F
-tn
-2o0
- --- ms iGsec $
1982 y j, CERT g
E -300 h 36%IG8CC e
g E -325 h DUCTILE O g
High Purity Water (2740C) .
, Dresden Hydrogen Addition:
1982 Data: 8 266 0C m 80% Power O 269 0C 95% + Power 1983 Data: O 273 0C a
-400 1 I I O to 100 I D00 10 0LO O 28#3 1
Figure Ill. The Effect of Dissolved Oxygen on the Corrosion Potent 101
' of -Type 304 Stainless Steel in Ill9h Purt tv. Wate: at
'27fl0 C (5250F) * *
- \l lli) 1 1
{ ,l1 l) 0 0 0 6 1 2 " 0 0 0 0 2 2 2 ' 6 4 2 0 0 0 0
' 1 1 0 8 6 0 1 1 4 2 sS '[,c .
O nJ 4 1 0
a 0 6
s s
) i b
r a e s t d e l e 0
2 o G
r a 0
0 T 4 AQO 4
0 a
0 2
86
- 2 C
W l
2 i
g . 0 n 0" -
0 i 4 r ' 2 u 2 D f l
t c
t e
i L
y r
O -
4 n
2 0
0 0 2 o
m le e f - p y
iL L
0 t
a 0
u 2 0
e d
I e l O m p G i l
u T
v.
L 0
0 C C .
8 C
i C S 1
\ S G I" G nVI t
e + ,
I v O i
N
. 0
. 0 6
1 4
8 5
0 2
. 0 1
1 f
f
" O fl3' 0 0
. 2 1
- gv $
I( -
o 0 0 0 0 5 0 1 0 2 0 0 0 2 3 3 0 0
- 4 5 0 0
- - - 6 7 lI1 i { !
i.
- Figure 16. ICSCC BEHAVIOR OF SENSITIZED .
STAINIESS STEEL. IN DRESDEN-2 TESTS ,
1 i 100 <
i
- == = - INDICATES f tMES yn4FN LOAO ntMOVID l , ,
l X Indicates 288*C Air Data I
i i
x
- 1 i -
- x l g < M ws x l ,
!" $ l i N l ti 4015 PPD l en -
g 3I I
ll l
'I
- I I l ll a _ /
I l
I lI l 1
al 11 I i
I 11 f3! :
- , ,l I il l i
n a ll
,l x . .....
N I
o a i ,
o .. ,
Tit 1E (h)
O % g 4 6 9 % i
l itysre 17. Dresden-? 'llWC Crat.k Growth Test on Furnace Srnsitized
~
~
Type '304 Stainless Steel, K =.27.5 MPa/m
,(y } .
0 0 b .- '
i
- Reactor Scram .
e
- I
' % M^
/. n. e ~% ~ e n -~ -
]
m O ,
C
. i
.J .
x .
O L. t l
<] .
. 005 --
t
.i l
.b . L. . . .L. ..__L........ '!
{1 i; o ca c) rJ N o i; rJ fJ
.J O LN il ir O.1 OJ :- CD (D .l i i
'.lest. Duration (h)
I too f
l 1 yfffgfj CORR 00lON IN NEUTRAL.
OXYGENATED H2O
.. RELEASE IN NEUTRAL.
j OXYGENATED H2O '
.. CORROSION IN 10 = '
MATED H O2 1 l 9 COR e 0=ia n.C. !
, Q THitPAPER .
. l i
j RELEASE IN H2WC, 3 TH18 PAPER 1
N4 in '
5$ In gpm \
u s e\
1 8 A\ l
\
l
\
o.1 -
\
\ /////#nwa
\ __---
l i RANGE OF SWR
Figure 18. Corrosion Rates and Release Rates for Carbon Steel at Different Oxygen Concentrations for 1,000 h Exposure Time.s at Temperatures in the Range 200-300*C (392-572'F). (Compiled by Urbanic, V. F., fe m orginal corrosion .
l data and numerous results published in the literature. .
l t
.- l l
l NO.9 HOLE 1.0 cm 0ABem e 10A0 k.)
me>-
(3/16 en.)
q
//
%J
-b- '
i1 c 3 'N- _
-> e
~
0 .152 mm toAGO in.)
UNCREVICED i 1- I I
1 1 CREvlCEO ' ,, ..
Rf 0.03 mm 40D01 M.3 SHsM AT SOTH ENOS l .
Figure 19. General. Crevice and Galvanic Corrosion Specimens
9 #
P i
3
% "% tbl COMPLETED PIPE WE LD (el PtPE WELD PREP WITH DEEP COUNTER 80RE l
0.32 cm
- 10.12 .nJ s 1',
s g j
l I I, JL 1
L i (cf PIPE WELD WITH "
CROWN S DE MACHINED 1
l 0.32 cm (0.12 in.)
y "
j t
1/320 THREADS.
PIPE WALL
" f
- SOTH ENDS i 0.32 cm 6 (NOMINA U !' j 10.12 m.) *
, a-a r h, T-~- $
,2g --
+ -
0.51 crn j ; s.n e -
(2.2s .n.:
a.: piNISHED SPECIMEN WITH INTACT ROOT-3:DE SURPACE i
- Figure 20.
Fabrication Sequence for Welded Coastant Load Tensile Specimen L. - . . - . . - -
1% 5TR AIN
._, SPECIMEN
\ e l
e F
- - 2& _
~
l 2M* MIN HOLDER CONTACT
- LENGTH
> w ~
-d a
e 2.72 A AM
- 7 " OLE:
127 1f 7
%+) %+)
c n s.71 =
c 10.48 2 0.76 r'i i i
- ! !e !
k m I
' DIMENSIONS IN CENTIMETERS Figure 21. Bent Beam Specimen and Test Fixture
, 4 APPENDIX B. '
ROLE OF IMPURITIES CE IGSCC AND HNC -
ARGONNE NATIONAL LABORA'IORIES STUDIES Extensiva studies to assess the effectiveness of Hydrogen Water Chemistry in mitigating stress corrosion cracking of stainless steel have been conducted at Argonne National Laboratories.1
.- Three basic types of evaluations have been conducted. First, constant extension rate tests were conducted under controlled potential
,. and conductivity to assess the critical potential and conductivity where IGSCC is eliminated as a function of conductivity. The studies were conducted on solution annealed Type 304 stainless steel, lightly sensitized Type 304 stainless steel, and heavily sensitized Type 304 stainless steel. The conductivity was adjusted by application of controlled amounts of NaH SO.. The constant extension rate tests were were conducted in water at 288'C. The results of the studies are shown in Figure 1, a plot of corrosion potential vs. conductivity at 25'C.
The corrosion potential was measured as part of each test. In general, ?
th'e data show that as conductivity increases, the corrosion potential required to prevent IGSCC becomes more negative. Measurements of stainless steel corrosion potential at Dresden-2 have shown that under Hydrogen Water Chemistry conditions, values considerably less than -350 mV (SHE) can be expected for long term operaticn. These corrosion potential conditien values result in IGSCC mitigation if the -
conductivity is 0.3 uS/cm or less with Na:SO. as the impurity. These tests are cuite relevant to reactor operations since this type impurity can enter the reactor recirculation system through leakage of res'in beds. The data curve shown is for heavily sensitized Type 304 l stainless steel but very similar results were found for -lightly sensitized stainless steel.
In addition to the controlled potential CERTS described above, additional CERTs were conducted wherein coolant oxygen and conductivity l
were systematically varied. As in the potential controlled tests, conductivity was varied by Na*SO' additions. Results were similar to .
1
_ hose in the controlled potential tests in that when coolant oxygen was reduced to about 20 ppb, corresponding to a corrosion potential of -350 l
1 mV (SHE) , no IGSCC was obtained if conductivity did not exceed 0.3 uS/c=.
( 8-47 l
. . - . , ~ . - . _ _ - _ . ....- . _ . _ . . - _ _ _ _ _ _ _ , . _ . _ _ - . . _ , . . m-~ ,. _ _ _ . _ _ . . . _ _ , . . _ , . . . - - , . _
The second bacic type of ctudy involvcd an evaluation of a ;er; broad rango cf coolant impurity species. At fir st . nine differen types of impurities were evaluated as shown in Figure 2.
Significantly, none of the impurities evaluated were more severe in promoting IGSCC than the Na:SO. used to conduct the studies described above. These results were followed up by a study o' f sul' fate with several different types of cations. Results, shown in Figure 3, indicate that although some cations are slightly more severe than Na:SO., the main result still holds that no IGSCC is expected if the corrosion potential is below -350 mV (SHE) if conductivity does not .
exceed about 0.3 uS/cm.
The third basic area of study was a crack growth study on both .
sensitized and solution annealed Type 304 stainless steel. Both -
heavily and lightly sensitized material were evaluated. As shown in Figure 4, no crack growth was observed in the sensitized stainless steel when oxygen in the coolant was reduced from 200 ppb to 20 ppb oxygen. Crack growth rates were re-established at their original rates when the coolant oxygen was increased frcm 20 ppb to 200 ppb.
In summary, the results of these studies show for a broad range of coolant impurities, no IGSCC is expected if the corrosion potential is below -350 dV (SHE) and the conductivity does not exceed 0.3 uS/cm.
These corrosion potentials are achieved in Drcsden-2 when the coolant oxygen is less than 20 ppb. Moreover, the crack growth data shcw that pre-existing cracks can be stopped if coolant oxygen is reduced to 20 ppb.
T. Kassner, paper presented at General Electric, Argonne t!ational Labs. , NRC, EPRI Seminar on IGSCC Studies. San Jose. California, March, 1984.
8-48
l Ibul(t 1 ,
CERT EXPERIMENTS ON SENSITIZED TYPE 304SS iN SIMULATED BWR - QUALITY WATER AT 289 C i (DATA FROM ARGONNE NATIONAL LABORATORY) 300 I 200 -
l g 100 i I l
E 0 -
E ~ ~
I IGSCC SUSCEPTABILITY
! $ - 200 -
! F 1 2
- m H - 300 -
. O .
E y I
-400 mm Emm j g ,,,,,,,,,,,i op.,. tina -1 FOR MOST DELETERIOUS
- y
" *8' " 8 IMPURITY (SULFATE)
! 8 - 500 HWC REGION _
l - 600 _
NOidSCC 1
- 700 1 10 100 O.1 CONDUCTIVITY AT 25 C (uS/cm) , , , ,
ARGONNE NATICNAL LARS STimy 160 - , S EN51T!Z ATIC N~
'[, EPR = 2 C/c:n I
./'.j P' 'f] 7 140 - 1 2 .
[' { y,- ,d 02= 0.2 ppm I -
e = 1x10-e s -t w 120 - g/d.d;gg,d gp .;
s
~
.'?.!
d
~
/jli(hb Pej:,'4dy/'1hh'by//
> /. ,
~ '
/') '// h.'.h ('.'/
</
O 80~ l/4' Y9$k'.i'/{
IV f/
'S s', 'VA' (f)'2)
~
p i - .
w 00~ I//)'t E /)
2
~
f '$
- V/ , --
. f$')fb/$/$l',7fls)$Y.,.
Q 40~ I$ ;',
c i I'.'f.)
fY /
y f- 3
- c. 3 '
~
74* rt C iy 4 '"/A /* jT 1' /
, '. . C 4) 0 li '*A 1 '-
i 19 (f/
4//A 20- ; =gg p g ;[ a ,,ggg 3; gg g p g"j ,;g 0
Fig. 2 EFFECT OF VARIOus ANION SPECIES AT A CONCENTRATION OF 0 1 PPM WITH SODIun IN 289'C WATER CONTAINING 0 2 PPM Diss0LvEn OXYGEN ON THE I!ME TO FAttuRE OF l.IGHTLY SENSITIZED (EPR = 2 C/cM2) TYPE 304 SS SPECIMENS IN i '
l CERT EXPERIMENTS AT A STRAIN RATE OF 1X 10-6 S -1 .
1 o i .
l
[
16 0 -- [,--
/, SEN9?T Z AT'O'
// E:o = 2 C, c.
/
140 - i c -
//; 0, = 0.2 pp r.
- 7
,// .
~
120 - ./' c = !>..a.-5 3 -
Y */
/
e- i/ -
5 <n I< -
.. .su '
/ ,,
_ 's
'i
,//'
/
- 0-
~
A Q -
//
/ y/ '
> 50-f*' '
//,// ,
$-e i
$Vf.f/)7\
41 , <TJ/[/ '7h3 F ' ;.A 43 % ., ,/
0'4 V[/ JA .l f/ 7.i o}:/ A,, ' A<'Td 0'3 qu '
3=u /. / /_k i
- $. = y fA- * ' .
- n . ./^ A
- c. o f
- 20 -
F t d';i . C 'J- ', 7:g g=
=n r
=
og f o[!]f ,m' o. 3"3 3,'< $ ,.
A 4C 5 90' Sh.,3g!j!E1~.' ' L[ $'
$I.';j1'a .d ' '53 4: ' =J '
. t.gj r ac . a .t 03 lS O tl .
0 Fig. 3 Iffec: of 0.1 pas Sulfa:e wi:5 cifferen: Ca Tailure of I.igh:17 Sensiti:ed (IPR = 2 C/cs;) ions Type on 30t.:he SS ri= :o Specinens' in CER"" Experisen:s a: a Strain Ra:e of 1 x 10- s -1 in 239'C ".ta:er with 0.2 pps Dissolved Oxygen.
e e
e
4
- i .
FIGURE 4 1
RdbNNE NATIONAL' LABORATORY DATA:
t l
g r m ..iisi: e w
_ ; a -
m o E ,
,/ trersn gj YPM_ C/e- '
g; l
, " ^ o :
- - D 2 5 a: o to x ;
o
( '
o e ,;
pu .- -
on(Ptol: - '
c.: ; not -:- o.:
. e , , , , ,
S 500 1000 1800 2000 2500 3000 TIME (h) ss
~
HWC ARRESTS CRACK GROWTH IN SENSITIZED T-304 SS UNDER SLOW CYCLIC LOADING .
e O
6
APPENDIX 3.
ION CHROMOTAGRAPHY MEASUREMENT PROGRAM IntIqduct1QD Ion chromatography has proven to be a valuable research tool in many industrial applications. It has the advantages of both qualitative and quantitative analyses for a variety of ionic species.
,. Moreover, the technique affords very good sensitivities for most chemical species: frequently, concentrations of less than 0.1 ppb can be accurately measured with a high degree of confidence. Samples can also be completely characterized in a reasonable time frame. with determination rates on the order of several samples per hour. Because of the many advantages afforded by this tec.m_que. ion chromatography is rapidly finding diverse applications in many BWR chemistry laboratories. Many BWR chemistry groups now routinely use the ion chromatograph for daily reactor water chloride measurements, and have abandoned the core cumbersome techniques of colorimetry. turbidimetry.
and electrochemistry. Others have used ion chromatography for assessing the performance of the condensate treatment and reactor water cleanup systems for a variety of chemical species. Condenser leak rates may also be assessed by low-level sodium and chloride measurements.
l i
9-50 L_
l l
l
. ION CHROMATOGRAPHY MEASUREMENT PROGRAM l During a- one-month demonstration test conducted in 1982. and for
{ the entirety of Fuel Cycle 9 (April,1983 to present), the Dresden-2
- reactor has operated on hydrogen water chemistry. In a very repearable fashion, both the reactor -water conductivity and the reactor water pH increase when the hydrogen addition system is turned on
- conversely.
when the hydrogen addition system is shut down both parameters " '
decrease. Calculations have shown that the changes'in conductivity can be attributed to the changes in pH. regardless of the status of the
- hydrogen addition system. i.e. on or off. ' These changes are thought
- to be the result of a change in the concentration of the equilibrium radiolysis products when the core changes from an oxidizing to a more L
reducing environment when the hydrogen addition system is turned on, or, the reverse when the hydrogen addition system is shut down.
The dramatic change in the core oxidizing potiential as a result I
of hydrogen addition may produce some chemical changes that would not be detected by ordinary chemical monitoring instrumentation. Possible I reactions that might occur with the implementation of hydrogen water chemistry would be the conversion of ferric- to ferrous iron, ' the reduction of nitrate ion to nitrite and/or ammonium ions, or to gaseous nitrogen oxides, or the reduction of hexavalent chromium to insoluble l
chromium trioxide. The documentation of these reactions (or their I absence) would further assist in the complete understanding of the implications of hydrogen water chemistry. Accordingly. the Electric Power Research Institute contracted with the General Electric Company to perform ion chromatography measurements at the Dresden-2 reactor. . ,
The first measurement- campaign was perfermed in February and March of .
this year. The results summarizing this campaign are the basis of this 1 report. A second series of measurements is planned for the first quarter of 1985. These data will be compared to the 1984 campaign 'to see if any changes will have occurred as the result of an additional year of operation with hydrogen water chemistry.
Equignent_and Ins:tallatiQD I The mainframe instrument used for this series of measurements was a Dionex Model 2020i Liquid Ion Chromatograph "'hree operating 9-51
channels were used in this configuration nonovalent cations, anions, and transition metal ions. Detectors for the anion and cation channels were Dionex Conductivity Detectors, while the transition metal ion detector was a Dionex Opti-Ion UV/ Visible single-beam spectrophotometer . For all three channels, eluents were pumped through the columns and detectors using Dionex high-pressure Analytical Pumps.
Anion and cation ~regenerant chemicals were introduced to the system by gravity feed. The PAR reagent used in the detection of transition metals was housed in a one-liter glass bottle, and introduced to the system using pressurized argon. Columns used in the cation channel were HPIC-CSI separator, HPIC-OGl guard and concentrator and CFS suppressor, all from Dicnex. Columns for the anion channel included HPIC-AS4 separator, HPIC-AG4 guard and concentrator, and AFS suppressor. The fiber suppressor systems are extremely convenient, as they are configured for continuous regeneration. For the transition
- metal channel, a HPIC-CS2 separator and HPIC-OG2 guard and concentrator columns were used in conjunction with a postcolumn membrane reactor.
Customized automatic sample changing modules were built for this test. For the cation and anion channels, 'an Eldex Model #AA100S dual-piston high pressure pump delivered sample water to the concentrator columns: for the transition metal channel, the concentrator column was loaded by an Eldex Model #B100S single piston high-pressure pump.
Automatic sample changing was controlled by air-actuated 10-port rotary valves. Minarek programmers were used to sequence the operation of the sample changer and the ion chromatograph. The basic steps of the program consisted of a diree-minute.line flush, a 10-to-15 minute period for loading the sample onto the concentrator columns, a 15 l minute period for column elution and generation of the chromatogram,
. and sample changing. The program was written to accomodate 10 sample
(~ changes, and then shut the sytem down automatically, thus allcwing
( -
unattended operation. Anion and cation chromatograph outputs were 1
. monitored on a Linear Model 585 strip chart recorder, while the transition metal chromatographs were maintained on a Hewlett-Packard j Model 3885 integrating strip chart recorder.
- 9- 52 l
l l
Reitgents_and_ Standards All chemicals used in this study were reagent grade quality or better. Eluent and regenerant solutions were prepared in 8-liter quantities, and were stored in polypropylene Nalgene Lowbouys. The eluent solutions were 0.008M hcl for the cations, 0.003M Na:CO:/0.003M -
NaHCO for the anions, while the transition metal eluent contained 1.26 g/l oxalic acid,157 g/l citric acid, and 1.22 g/l LiCH. The ,
regenerant solutions were 0.025M H SO. for the anions, and 0.02M tetramethylammonium hydroxide for the cations. The photochromic solution for tne transition metal ion de.tector contained 207 ml 30%
ammonium hydroxide, 57 ml glacial acetic acid, and'0.1 g 4-(2-Pyridylazo)resorcinol monosodium salt, monohydrate per liter of solution, hereafter referred to as PAR.
Mixed cation standards were prepared from Nacl. NH.C1. and K CO .
The stock solution contained 1000 ppm of each cation, and was prepared and stored in a plastic volumetric flask. Mixed anion standards were i
prepared from NaC1. NANO: . NANO . K SO. , K:HPO. , and Na:C 0. . again i-l 1000 ppm concentrations for each anion, and stored in a plastic flask.
Individual transition metal standards for Fe, Cu, Ni, and Zn in 1000 l
ppm concentrations were obtained frcm Spex Industries. Instrumental working standard concentrations were 5 and 10 ppb of each species, and were prepared by appropriate dilutions of the stocks in plastic volumetric flasks. These working solutions were prepared daily. The water used in the dilution of the' primary standard also served as the ',
daily blank. Scme problems were encountered finding good-quality water for the preparation of standards and blanks. The concentrations of sodium and chloride impurities in the laboratory demineralized water I
supply were very dynamic, and often contained salt concentraticns in l
l excess of 10 ppb. Generally. the most suitable water in the 71 ant for the preparation of standards and blanks was the effluent of the condensate treatment system Samg1e_ Joints Seven sample points were monitored on a daily basis. The list of acronyms associated with these points is as follcws:
9-53 J
[ ~(1) CDI--Condensate Demineralizer Inlet (2) CDE-Condensate Demineralizer Effluent (3) FN--Final Feedwater (4) RCI--Reactor Water Cleanup System Inlet (5) RCEA "A" Reactor Water Cleanup Demineralizer Effluent
.- (6) RCEB "B" Reactor Water Cleanup Demineralizer Effluent (7) MST-Main Steam
,. CDI, CDE, and FN samples were obtained from grab sample taps in the turbine building sample sink. RCI. RCEA. and RCEB samples were '
obtained from the reactor building sample sink. MST samples were obtained from a grab sample tap off the General Electric process instrument line located immediately outside the X-Area wall. Several precautionary measures were taken to obtain representative.
uncontaminated samples. Sample containers were 250 ml plastic culture bottles. These bottles were filled with water from the respective sample points, and were soaked in for a three day period prior to use.
Sample bottles and their caps were labelled and dedicated to their respective streams. The CDI and CDE sample taps were manifolded to a common effluent at the turbine building sample sink, as were the RCI.
.RCEA, and RCEB sample taps in the reactor building sample sink. Prior to obtaining samples, each line was flushed for 10 minutes at a flow rate of roughly 1 liter per minute.. Each bottle was filled, ' capped, shaken, and sample discarded four times before the actual sample was collected. Duplicate samples were obtained at each location-one for transition metals, the other for the cation / anion system.
Icalyses Roughly two hours were required for the acquisition of duplicate samples. Once obtained, they were transported to the ion chromatograph which was set up near the Unit-2 air compressors. Samples were segregated between transition metal and cation / anion botttles, and were positioned at the appropria'.e rotary. valve ports. All lines between
- ne grab sample bottles and the rotary valves were 0.125 in o.d. X 0.028 in wall teflon tubing. The following instrumental parameters were used for the three channels in the ion chromatography determinations:
9-54
Cation Anion Trans. Metals
- Eluent Flow (ml/ min) 2.4 2.4 1.2 Sample Pressure (psig) 410 920 510 Lo Pres. Ala'rm (psig) 300 500 300 Hi Pres. Alarm (psig) 90 0 1200 1300
~
Samp]e Flow (ml/ min) 2.90 3.19 3.16 Sample Volume (ml) 29.0 31.9 47.4 '.
Detector Sensivity 3 uS/cm 3 us/cm 0.2 A Baseline Response 3.5 uS/cm 18.0 uS/cm --
Detector-Temp. Comp. 1.7 1.7 ---
Blanks and standards were analyzed in an identical fashion to the i
samples. The net signal for each ion, standard minus blank, was used to calculate the concentrations. Blank readings were not subtracted from the sample readings. Owing to the dynamic impurity concentratiens
~
in the water used to prepare standards and blanks, it was not possible to determine the contribution of the analysis system to the samples.
Results__nd a DiscuSE1QD Reactor power has always shown a pronounced effect on vessel-chemistry at the Dresden-2 reactor. In general. the lower the power, the lower the vessel conductivity. Lower power results in lower impurity input per unit time, by allowing more efficient removal of
~
impurities by the condensate treatment system (lower gpm/sq. f t) , .
assuming the reactor water cleanup system flew rate remains constant.
Figure 1 shows the reactor power a'id the reacter water conductivity (as read frcm a General Electric process instrument) during the ion chromatography measurement period. The symbols on the curves correspond to the actual data points when samples were collected and analyzed. The correlation between the two parameters is apparent.
i Full power at the Dresden-2 reacter is roughly 2527 MWt. and the plant operated at neraly full power for the last week of the testing pericd.
The maximum vesrel conductivity that was observed was 0.14 uS/cm.
Contributing-to the achievement of low vessel conductivity is the 9-55
- , . - - . , , , , , - - , ,,-,,,y
~ design basis of the reactor water cleanup system which was operating at 6.6% of rated feedwater flow (about 1200 gpm) . From a conductivity standpoint, the Dresden-2 reactor currently has one of the best vessel chemistries of all domestic BWR's.
The dissolved gas concentrations in the reactor water, main steam,
, and final feedwater are shown in Figures 2 through 5. respectively.
The data point symbols correspond to the time periods when sar.ples were collected. The hydrogen addition system was shut off on four separate days during the testing interval, as indicated by concentrations of less than 20 ppb dissolved oxygen in the primary coolant (Figure 2),
and feedwater dissolved hydrogen concentrations less than 40 ppb i
(Figure 5).- Subtle differences in the concentration of ionic impurities as a result of hydrogen addition status may not be readily apparent during such a short testing interval. The dissolved oxygen concentration in the feedwater ranged between 15 and 59 ppb during the measurement campaign, generally within the Puel Warranty Operating Limit range of 20-200 ppb.
f.Qadensate-Feedwater Results Tables 1 through 3 show the ion chromatography measurenents for l the CDI, CDE. and FEW samples, respectively.
Several ions that constituted the makeup of the standard solutions are conspicuously absent from the table. The transition metal analyzer can simultaneously detect.both ferrous and ferric iron along with copper. nickel and zinc. The sensitivity for both iron species is .
~
nearly tenfold less than the sensitivities for the remaining elements, and the sample sizes used for these analyses were probably not
. sufficient to allow good characterization using a full scale absorbance value of 0.2A. Although the 5 ppb ferric iron standard always generated a strong analytical signal- there were some problems in the resoluticn of this peak with the copper peak (partially resolved do"blets). Real samples never showed doublets in this region of the enrematcgram, and the sample retention times of the singlet peaks were 9-56
n t p m -
. \ /
\ /
1 a., -2
< \ -
EB 9
2 Q50 \
3 h ' -
- O o- --
b -2 32 l
a
. . ~
T* ".
t .-
14>- -
- "c
! b N1 O1H
- o: >-
j L NE $-
1e &
E. . .
O t-
= y gN
- e Ax q s.
- to i e
- - 1.d N N -
=
l H N N l
Z$ % x
'S -
. Lt.1 e s I Ck \ '
-T bh Vif we N
t i
e, -
k
=
2 \,. \ v -
\ . ..
O C -N G
\s ,o t a p D
l ,-
i **
i i i i i l I i I i i i i I
! O C W T N O @ to 4 N O 03 (C 4 N C TO N N N N N w w - - w C C C C C l d d dd d d d d d d d d d 6 6 d i
(IMW) 83 mod '(wo/Sn) GNCO 10
DTESDEN 2 F W6 P ROG'R A W RECIRC WATER DISSOLVED GAS CONC.
300 _ _ . . _ . _ _
280 -
260 -
n 2 4 0 --
m
$ 220 -
a.
200 -
(n y
180 -
Q 160 -
C f 140 -
./
Z La 120 -
}
0 100 -
& 80 -
O 60 -
40 -
20 -
w C a C a .
O , , , ., , , i i , , , i i O 2 4 6 8 10 12 14 DAYS SINCE FEBRU ARY 21, 1984
- o RECIRC 02 (PPB) + RECIRC H2 (PPB) 4
1:1G11111! 3 DRESDEN 2 HWC PROGRAM MAIN STEAM DISSOLVED GAS CONC.
20 19 -
18 -
, 17 -
16 -
r^ m I 15 -
- n. ,
g, 14 -
g 13 -
y 12 -
F 11 -
E 10 -
- F-w z g_
O 8-Z O- 7-O 6- - .,
n N 5-0 4-3-
2- .
, ; p__ ,_ : : :u O i , i i , , i , , , , , ,
O 2 4 6 8 10 12 14 DAYS SINCE FEBRUARY 21, 1984 o MST O2 (PPM) + MST l12 (PPM)
ljI i I \1l\'l\!l! 1l' l l 4
. 1 i
2 M i 1
A R i GCN OOC i 0 1
4 8
9 C' .
1 d
PE G i 1
Y 2 X
CO i8 Y R
A WE D V i U
R
- L B O E S F 2S D I i6 E C
N R I E S NTA i S
EW Y A
DDEE ,4 D SF E i
'R D -
i2
/B i O
0 0 0 0 0 0 0 6 5 4 3 2 1 L
mC1v zO &4cWZwOzo0 c 'zN>XO C.
,l 'l' l!' l !1ll ll! l ll l L
n .
, ,, j' ~
, 7 v.,mn s JP _ HWC PROGRAM
- J :- ~' .., SOLVED HYOROGEN CONC.
1.50 g ,
1.40 -
1.30 - y 2 1.20 - .- am ,
a -
m m
g .
v 1.10 - _ .
Z-
>o . 1.00 -
P >
0.90 -
it ' -
F- ~
g 0.80 -
w ,
- E O.70 -
o -
O O.60 - ,
.- ]
2 w
~
0.50 -
O r. , ,e
@ 0.40 - u .;
O p _, O . 3 0 -?
x- .,.
., 0. 2 0 - -
0.10 - :
,s o' -
s
~
" '/
0.00 , , , ,
i .. i , v7 , , ,
O 2 4 . 6 8 10 12 14
.!: , , a DAYS SINCE FEBRU ARY ,21, 1984f
~'
- ./ ..
/ ',w i p+ s , 8_ e ,
p .. a ;,
- d. _
,li \\\\ il 4
1 49 i
i 12 M
A ,
R i
,. GEC \ i 10 4 W ONA ,
8F 9
F
,.R M 1 R
- POF i 1
o
- R 2 E
CP [N i8 Y R
A WMU I i UE R
D D FD O ^
N E C S F N
2 i i6 E+
C t
i-M-
s m H l
S
.E- i
-,..- .,N D
.< ,ED .
f M S Y
A1 D NO i4 DD C
SC E /_- i 0
.R . .
s D .
i2 f
. 8 .
i
_ - - - ~ - - - -
O n 7 . g- 5 4 3 2 1 0
, ton
, o.." r 6P4%$wOzO 2 6 a; +
. r '-
~ j .
r[;\\\j\i\ \\; \ > l llIl1lll llL
I1ll1 l1 4
1
~
~
l
?
M l 1
A R E 3 gl w
GCN 0 W
/
OAM 4 s 71 4
8 9
'l I
RRO \ 1 PF R l 1
c E 2 CPE l 8 Y R
A WD I R / UE RD HOL - l D
E C H 'l 7
C 2
i t O l 6 E C+
t l
l i
G I
N I N
- ' I 1
M S U ,, l bE D S E U Y
AI DNO s l
4 DD C
SC E l o
R ',
D -
l 2
I
- - - - - - ~ ~
O 8 7 6 5 4 3 2 Taa' i .
2Op4(HZWOZOU WOE 9b k
f1 hllll;l!\lll l !j t
, COND. DEMlH. SULFATE PERFORMANCE 4.0 - --
3.5 -
,3,
[ 3.0 - y'-
. /'
' f 2.5 -
/ +
- p 4 D tr C U 2.0 -
0 N ), s 4
z \
O o
'O 1.5 - q N/
) % / ,./ N
'N s i
g D
> 1.0 -- /g/ \ ,
16 \N i O v1 Nf l
0.5 - .;
O.0 - , , , i i i i i , , , , i O 2 4 6 8 10 12 14 D AYS SINCE FEDRUARY 21. 1984
4
. 1
~
O n.- a l
M' 1
?
I A
R ~
s i GEC /,
'o
' W ONA
- i 0 1
4 P, iI RMR 9
- 1 .
POF -
i 1
o R - 2 E
CP .
A e
8 Y R
R A WE P -
Y i U:
RID
- P O
- C 9
O C
/ t l
P.
IR 2 N s
6 E C+
G m'-
I I
F M H l
E S ND * +1 S
E D Y
AI DNO i 4 DD C
SC E i o
R t D , 2 ,
B ,
O 0 0 0 0 0 0 0 0 0 0 0 0 9 8 7 6 5 4 3 .
2 1
0 1 0 0 O 0 0 O O O O 0 s
, m 0- .Q v Z O- i < & 6 - '
- g o Z O O x W o. .QO O 8
I1lll ! llI !' l l
Tzble 1. CDI Ionic Impurity Concentrations (ppb) .
DATE TIME CU NI ZN NA CL SO4 22284 1100 0.21 0.17 1.88 2.42 3.03 2.34 22384 1100 0.23 0.32 1 80 3.31 4.03 0.93 22484 1500 0.98 1.91 1.09 22684 1300 0.74 0.28 1.58 1.12 0.98 1.49 22784 1115 0.09 0.J5 1.43 0.80 0.81 0.98 22884 1200 0.36 1.09 1.75 0.30 0.54 1.79 22984 1130 0.39 0.56 1.53 5.24 5.76 2.69 30284 1045 0.33 0.65 2.05 1.42 1.59 1.93 30384 1045 0.18 0.62 1.40 0.64 0.72 2.35 30584 1045 0.10 2.38 4.46 1.01 1.26 0.89 Missing entries: No samples were collected.
Table 2. CDE Ionic Impurity Concentrations (ppb).
4 DATE TIME CU . NI ZN NA CL SO4 -
22284 1100 0.07 0.23 1.21 1.33 1.25 1.74 22384 1100 0.08 0.79 1.72 7.12 7.07 2.75 22484 1500 1.72 2.09 1.98 22684 1300 0.03 0.42 1.80 2.64 1.84 1.63 22784 1115 0.00 0.13 1.72 2.76~ 2.50 3.67 22884 1200 0.30 1.06 1.31 1.34 0.65 1.22 22984 1130 0.12 0.36 1.41 3.40 2.88 1.58 30284 1045 0.09 0.69 1.78 1.75 2.00 3.25 30384 1045 0.16 0.96 1.62 1.44 0.87 0.90 30584 1045 0.13 1.49 2.67 1.24 0.77 2.64 Missing entries: No samples were collected.
Table 3. F&l Ionic Impurity Concentrations (ppb) .
DATE TIME CU NI ZN NA CL SO4 22284- 1100 0.31 0.24 2.32 1.24 1.13 1.51 22384 1100 . 0.12 0.45 1.98 2.25 2.26 0.93 i- 22484 1500 1.21 1.41 2.59 22684 1300 0.11 0.89 2.04 1.03 0.51 3.08 22784 1115 0.05 0.24 1.71 2.82 3.21 1.84
- 22884 1200 0.18 1.32 1.03 1.04 0.64 1.61
- 22984 1130 0.23 0.53 1.41 3.65 1.16 1.38 30284 1045 0.21 1.20 2.08 0.94 0.42 0.75 30364 1045 0.09 1.03 1.83 0.54 0.10 0.85 30584 1045 0.17 1.05 1.23 0.86 0.77 1.91 Missing entries: No samples were collected.
9-72
-. r - ,w e-,r-, --,---,-e-r,- -
, ---,,.--..,e,- . --,--. ..
-,-------,,e - -
.y,--. . - - - , -#- ---
-=-
within 0.02 min of-the copper standard. Farrous iron peaks, which elute relatively lata compared to the other species were not observed in any samples. Potassium and ammonium ions were also not observed in
- any samples, nor were nitrate, nitrite, or oxalate anions. Perhaps these could have been detected using larger samples sizes, or a more
- sensitive detector attenuation. With the configuration used in this test, the~ concentration'of each species is estimated to be less than 0.1 ppb.-- In a previous study at Dresden-2, the oxalate ion, or some species that elutes with the same retention time as oxalate, was observed in many stages of extraction steam and most primary system sunples. The potassium ion is not normally encountered in BWR chemistry- .Somewhat surprising was the absence of the common -
nitrogenous ions in CDI water, or reactor water samples (discussed later).
o Figures 6 through 9 show the performance of the condensate treatment system (CDI vs. CDE) in the removal of sodium, chloride.
sulfate, and copper. The feedwater concentration is also plotted to trend the release or plateout of metallic impurities along the f eedwater train (CDE' vs. FfW) . In many cases, the eff1'uent concentration (CDE) of each ion is greater than the concentration of
, the demineralizer inlet (CDI) . If these data are real it indicates the condensate demineralizers are sloughing impurities. If there were no net deposition or release of impurities down the feedwater train, the concentration of each species in the condensate demineralizer j effluent and the final feedwater sample streams should be the same.
The figures indicate that the CDE and FfW trend lines cross one another on several occasions, favoring neither the release er deposition mechanism.-
To the best of our knowledge, the concentrations of sodium and sulfate have never been measured in the Dresden-2 reactor. The logical
[ source of these ions in the condensat'-faedwater e train is the condensate demineralizer ion exchange resins, as sulfuric acid and
- sodium hydroxide are the regenerant chemicals. Hotwell sources (CDI) include inputs from the condensate storage tank and radwaste. Chloride
~
concentrations in these streams are rarely monitored: usually, lower .
limits of detection (<20 ppb) are reported. The Dresden reactors should have 1cw concentrations of chloride throughout the system, owing '
to the non-saline cooling water source. .
Copper and nickel are routinely monitored in the feedwater as a requirement of the Fuel Warranty Operating Limits. In addition,
{ General Electric has ben sampling feedwater corrosion products on a 9-73 I
nn.n__, --v~,w ~ . - - - - - - - ~ - - - - - - * * - - " ' ' ^ ^ ^ ^ ^ ^ ' -
continuous basis for the entire cycle of operation on hydrog:n water chemistry. Soluble matallic impurity samples are collected on icn exchange membranes, using sample volumes on the order of 3001. These measurements show an average copper concentration of 0.1 ppb, with occasional spikes approaching 0.3 ppb. Thus, the copper concentrations reported by the two methods are in reasonable agreement. Feedwater soluble nickel concentrations reported by the ion exchange membrane technique are on the order of 0.05 to 0.10 ppb, an order of magnitude below the concentrations determined by the ion chromatograph. A part of the problem for the nickel determination may be the stainless steel heads in the Eldex sample pumps, or some other stainless steel surface
, between the pump and the ion chromatograph concentrator column. The concentration of soluble zinc in feedwater using ion exchange membrane filters ranges between 0.01 and 0.02 ppb. This is nearly two orders of
/ magnitude below the concentrations reported oy the ion chromatograph.
All three sample streams, CDI, CDE. and FEW, show zinc concentrations between 1 and 2 ppb on a routine basis. The concentration of zinc in the reactor water cleanup inlet and effluents, discussed in the next section, is also typically 1-2 ppb. This strongly suggests that a fixed level of zinc contamination was being introduced to each sample.
The source of this zinc contamination is under investigation.
Figure 10 shows the conductivity readings fcr these three process streams during the time periods when samples were acquired for the ion chromatograph analyses. Here it appears that the condensate demineralizers are doing an effective job in the removal of impurities, with the average removal efficiency on the order of 70%. pH measurements were not made on these sample streams, and the actual removal efficiency of impurities may be somewhat greater, if the inlet or effluent pH is slightly different than 7.00. In most cases, the final feedwater conductivity is greater than the condensate demineralizer effluent conductivity. Perhaps indicating a solublization of impurities down the feedwater train.
Table 4 compares the conductivities from the three process instruments with the calculated conductivities, which are based on the sum of the equivalent conductances of the separate ions, Na, C1, SO. ,
Cu, Ni. and Zn, at infinite dilution. The calculated sums also include
. the contribution of theoretically pure water, 0.0550 uS/cm at 25*C, having a pH of 7.00.
'9-78
FIQlRI! 10 DRESDEb 2 H WC P ROG R A vi CONDENSATE-FEEDWATER CONDUCTIVITIES )
0.0660 0.0650 -
0 .0 6 4 0 -- ,
m O.0630 -
E o 0.0620 -
N M O.0610 -
0.0600 -
g 0.0590 -
g 0.0580 - .
^
[
i
[ O.0570 -
o , ,
l h 0.0560 -
S 0.0550 -
O Z O.0540 -
O O O.0530 -
1 0.0520 -
0.0510 -
1 l 0.0500 i i i i i i i i i i i i i l 0 2 4 6 8 10 12 14 l DAYS SINCE FEBRU ARY. 21, 1984 l
O CD1 + CDE o ITW
( . .
s
Table 4. Cond:nsate-Feedwattr Conductivitics (uS/cm) .
CDI CDI CDE CDE FFW FFW l DATE TIME IONCHROM METER IONCEROM METER IONCHKOM METER 22284 1100 0.0744 0.0583 0.0659 0.0565 0.0673 0.0565 22384 1100 0.0763 0.0591 0.0946 0.0566 0.0705 0.0566 22484 1500 0.0630 0.0586 0.0665 0.0563 0.0650 0.0577 22584 1100 0.0586 0.0563 0.0580 22684 1300 0.0663 0.0592 0.0711 0.0564 0.0685 0.0591 22784 1115 0.0627 0.0596 0.0755 0.0567 0.0744 0.0578 22884 1200 0.0651 0.0618 0.0658 0.0572 0.0655 0.0579 22984 1130 0.0874 0.0584 0.0743 0.0559 0.0713 0.0573 30184 1200 0.0606 0.0567 0.0598 v 30284 1045 0.0697 0.0597 0.0727 0.0570 0.0649 0.0578 30384 1045 0.0655 0.0602 0.0660 0.0574 0.0627 0.0586 30484 1200 0.0614 0.0566 0.0563 30584 1045 0.0728 0.0611 0.0708 0.0565 0.0658 0.0565 In every set of paired measurements, the conductivity as read from the process instrument is less than the calculated conductivities from the ion chromatograph measurements. This again suggests that the samples might be picking up some contamination between acquisition and analysis. The contamination might originate from the atmosphere, handling operations, or perhaps from some internal components of the instr ument.
None of the condensate-feedwater ionic impurity measurements appear to correlate with the on or off status of the hydrogen addition system. The dissolved hydrogen concentration in the CDI and CDE sample streams is less than 2 ppb with or without hydrogen addition. The hydrogen is injected into the condensate booster pumps, located Hydrogen carried o'ver
~
downstream of the condensate demineralizers.
with the steam is recombined in the offgas treatment system. The concentration of dissolved oxygen in the condensate-feedwater train is reduced as a result of hydrogen addition because less oxygen is carried over with the steam, resulting in a lower equilibrium
. concentration in the condenser. No ionic effects were observed in the final feedwater (ca. 1 ppm H with hydrogen addition, 2 ppb without) during the measurement period. Any changes that are occurring might be evident only in a long-term testing program.
9-80 -
l
, - , , ~ _ _ _ - -
f t
Reactor.Jater..Besults Tables 5 through 7 show the conctntrations of ionic impurities as measured by the ion chromatograph in the RCI, RCEA. and RCEB sample streams, respectively.
It was somewhat surprizing that nitrate, nitrite, and ammonium
~
ions were not detected in the reactor water samples. If present, their concentratlons are estiinated to be less than 0.1 ppb. Perhaps this implies that the dominant forms of nitrogen in a reactor are gaseous molecules, either N*, NO, or NOz. The concentrations of sodium and sulfate ions have never been measured in reactor water prior to this campaign. Previous chloride measurements at the Dresden-2 reactor again suffer from the lack of measurement sensitivity. Concentrations s of less than 20 ppb are typically reported. Station personnel do not measure the concentration of transition metals in reacter water. The General Electric sampling program measures the transition metals in '
reactor water, as taken from the reactor water recirculation sample point, but no monitoring is perf ormed on either effluent from the cleanup system. Ion exchange filter samples are collected three times per . week, typically with 10-liter volumes. By this technique, the average copper concentration in the reactor water is 1 ppb, with occasional spikes up to 4 ppb. For copper, the ion chromatography data are in reasonable agreement with the filter samples. We have previously reported soluble nickel concentrations in reactor water that range between 0.1 and 0.2 ppb. These values are roughly a factor of five lower than the ion chromatography measurements, and again we suspect sample contamination by the system. The reactor water zinc measurements by the ion chromatograph are an order of magnitude higher than results using the filter technique (1.5 vs. 0.15 ppb) .
Figures 11 through 14 show the concentrations of sodium, chlcride.
sulfate, and copper as a function of time for the reactor water cleanup system inlet and effluents. There is a strong correlation between the concentration of sodium in the reactor water (Figure 11), and the reactor water conductivity and power (Figure 1) . The data indicate that at full power, a 90% removal efficiency for sodium is achieved. ,
For all four species, the concentrations in the "A" and "B" ef fluents .
are comparable, indicating similar removal efficiencies for each cleanup demineralizer. No correlation was found between the
- concentration of any species measured by the ion chromatograph and the .
on or off status of the hydrogen addition system. Sodium and chloride ions, each having only one commor valence state, should not be affected 9-81
Table 5. RCI Ionic Impurity Conc;ntrctions (ppb) .
DATE TIME CU NI ZN NA CL SO4 22284 1100 0.37 0.22 1.83 2.25 3.40 2.91 22384 1100 0.90 0.77 3.02 6.34 3.89 1.66 22484 1500 4.05 2.35 2.51 22684 1300 0.75 0.62 2.21 6.31 3.04 4.21 22784 1115 0.67 0.21 1.21 12.80 4.50 3.51 22884 1200 1.38 1.31 1.42 11.40 1.68 3.05 22984 1130 0.86 0.70 1.40 14.40 2.65 3.27 30284 1045 0.62 0.81 1.57 10.10 1.23 1.93 30384 1045 0.49 0.74 2.06 10.20 1.92 3.12 30484 1200 1.67 0.68 5.11 10.40 17 .3 2.96 30584 1045 0.90 1.32 1.58 10.80 7.85 2.30 Missing entries: No samples were collected.
,a Table 6. RCEA Ionic Impurity Concentrations (ppb) .
DATE TIME CU NI ZN NA CL SO4 22284 1100 0.08 0.03 1.48 0.29 0.00 1.40 22384 1100 0.11 0.56 2.33 1.01 1.49 0.85 22484 1500 1.29 2.42 1.01 22684 1300 0.12 0.40 1.85 0.76 0.59 1.45 22784 1115 0.05 0.21 1.14 1.71 2.14 1.18 22884 1200 0.15 1.07 0.71 0.92 0.96 1.79 22984 1130 0.09 0.69 1.60 1.39 1.18 1.50 30284 1045 0.02 0.76 2.11 0.76 1.03 0.75 30384 1045 0.09 1.18 1.33 0.49 0.27 1.20 30584 1045 0.15 0.77 1.36 0.43 0.77 1.06 Missing entries: No smnples were collected.
Table 7. RCEB Ionic Impurity Concentrations (ppb) .
TIME CU NI ZN NA CL SO4 DATE 22284 1100 0.07 0.11 1.33 0.17 0.00 0.91 22384 1100 0.11 0.72 2.59 1.41 2.11 0.81 22484 1500 2.24 3.80 1.21 22684 1300 0.13 0.20 1.86 1.22 0.78 1.22
- 22784 1115 0.09 0.00 1.53 0.57 0.94 0.94
. 22884 1200 0.21 0.79 1.21 0.30 0.60 2.62 22984 1130 0.14 0.49 1.13 0.80 0.96 1.04
- 30284 1045 0.17 0.76 8.60 2.46 3.23 0.80 20384 1045 0.01 1.03 5.87 1.08 1.20 0.98 30584 1045 0.20 1.13 1.59 1.35 2.27 1.02 Missing entries: No samples were collected.
9-82 .
FICIRl! 11 DRESDEN 2 HWC PROG RAk 15 CLEANUP SYSTEM SODIUM PERFORMANCE 14 -
13 -
s
,, 12 - \ \
m ct a 11 - 'if \ s v
/g Z 10 - g O
P 9-4
$ 8-z W
.a e O 7-Z O 6- /
o .-
,i 3- \
\
/
/
5 4- h' o
m 3_
/
2-
/ - N's A
/\
N 1- .
p / (-
' ~- ~ g* N* '+- 1
~~ .y %
0_
T
- m4s'y M +
I I I T- 1 I I I I I I I '-
0 2 4 6 8 10 12 14 D AYS SillCE l'Ef1RtlARY 21, 1984 a RCI -
i RCEA o RCI'll T
k I
JRESDEN 2 l- W C P R O G R A h/
CLEANUP SYSTEM Cill.ORIDE PERFORMANCE 18 --
r) 1/ -
16 -
15 -
4
@ 14 -
I b 13 -
Z 12 -
9
- y 11 -
'. %, 10 - \
U O 9-
- o Z 8- U O
'O 7-o 6-l h 5-b 3- '\f #\ O .-
2- ,,- h / E3 1 I- 9/ A fM -t-I I I I I I I I T I I I I O 2 4 6 8 10 12 14 L DAYS SINCE FEDRUARY 21, 1984
- O RCI + RCEA o RCEO
l'IGiltli I3 DRESDEN 2 FWC PROGRAM 5.00 CLEANUP SYSTEM SULFATE PERFORM ANCE g 4.00 -
n_
b z
~
y 3.00 -
/
tr b- e.
d d
o \sm-a 0 2.00 -
f !'
I g \ 'N
- /\s a y ul 1.00 -
s'x/ o y
0.00 --
i i r~ i i 1 i i i l~
I O 2 4 6 8 10 12 14 D AYS SINCE l'EllRIJ ARY 21, 1984 0 RCI + RCI:A o RCl ^ t )
,s .' .
,< e, s DRESJEb 2 FWC PROGRAM CLEAllUP SYS~lEM COPPER PERFORMANCE 1.7 - - - -- - g 1.6 - ,
1.5 -
1.4 -
[ 1.3 -
b ~1.2 -
z 'll-O Q 1.0 -
oc I t- 0.9 -
O.8 -
o 0.7 - N O
O.6 -
[ O.5 -
n.
O O.4 -
0 0.3 -
f% o O.2 - +
0.1 - g#
9 k' A vN Ny i
0.0 i i i i i i i i i i i i 6 8 10 12 14 O 2 4 D AYS SlHCE FEBRUARY 21, 1984 o RCI + RCEA o RCEB
by tha oxidation-reduction characteristics of the core. The sulfate '
ion could po:sibly ba reduced to some icwer valent species, but no spurious peaks in the chromatograms were observed, nor did the retention time of the eluting sulfate ion shif t on a day-to-day basis.
There is no strong evidence that the concentrations of copper, nickel, or zinc are being affected by hydrogen addition.
In spite of the possibility of sample contamination, we again have attempted to reconcile the reading from the reactor water process conductivity monitor with the sum of the conductivities obtained frcm the ion chromatograph. using the equivalent conductances of the separate ions at infinite dilution.
These data are shown in Table 8.
s Table 8. Reactor Water Conductivity Balance.
e INSTRUMENT CHROMATOGRAPH CONDUCTIVITY CONDUCTIVITY DATE TIME RCI PH (uS/cm) (uS/cm)
~
22284 1100 6.67 0.076 0.105 22384 1100 7.07 0.076 0.086 22484 1500 7.17 0.099 0.071 22684 1300 7.00 0.092 0.088 l 22784 1115 7.30 0.129 0.104 22884 1200 7.25 0.123 0.095 22984 1130 7.10 0.115 0.100 30284 1045 7.16 0.121 0.086 i 30384 1045 7.44 0.126 0.104 30484 1200 6.50 0.147 0.194 30584 1045 7.25 0.107 0.106 i
The chromatography conductivity column in the above table includes
! the contributions from hydroxide and hydrogen ions which were obtained from the pH measurements. In the reactor water, we see several instances where the process instrument conductivity is higher than the s sum of the individual ionic contributions. In most cases, the process -
instrument value is higher when the indicated conductivity is greater taan 0.1 uS/cm. This is reasonable in that the contributions of the ,
various contaminants to the senple are less when the actual impurity '
concentration is increased. With calibrated process instrument meters, cells, and temperature indicators, and appropriate temperature compensation of the conductivity readingr, the use of the equivalent
' conductances of the separate ions provide a good cross check on the adequacy of the ion chromatograph measurements.
9-87 l
l l
l
( -
A second crocs chack of tha ion chromatography measuremrnts is a curration of the cation and cnion equivclents in tha sample: this requires the additional neasurement of pH, since the ion chromatograph i does not detect hydrogen or hydroxyl ions. If the ion chromatograph is accounting for all the ionic impurities in solution, the number of cation equivalents should equal the number of anion equivalents. The results of this excercise are shown in Table 9.
Table 9. Cation / Anion Equivalents Balance.
SUM METER CATION ANION CAT-AN HCO2- COND. COND.
DATE TIME EQUIV EQUIV EQUIV PPB uS/cm uS/cm 22284 1100 3.87E-07 2.03E-07 1.83E-07 11.2 0.113 0.076 22384 1100 5.08E-07 2.62E-07 2.46 E- 07 15.0 0.097 0.076 o' 22484 1500 2.44E-07 2.66E-07 -
0.070 0.099 22684 1300 4.87E-07 2.73E-07 2.13E-07 13.0 0.098 0.092 22784 1115 6.72E-07 4.00E-07 2.72E-07 16.6 0.116 0.129 22884 1200 6.83E-07 2.89E-07 3.95E-07 24.1 0.113 0.123 22984 1130 7.99E-07 2.69E-07 5.30E-07 32.4 0.124 0.115 30284 1045 6.03E-07 2.19E-07 3.84E-07 23.4 0.103 0.121 30384 1045 5.83E-07 3.95E-07 1.89E-07 11.5 0.113 0.126 30484 1200 1.00E-06 5.81E-07 4.19E-07 25.6 0.213 0.147 -
30584 1045 6.47E-07 4.47E-07 2.00E-07 12.2 0.115 0.107 The cation equivalents represent the sum of sodium, hydrogen, '
nickel, copper, and zinc ions, while the anion equivalents represent the sum of chloride, sulfate, and hydroxide ions. With one exception, all the measurements show a significant anion deficiency. Because the anion eluent used for these measurements is a mixture of carbonate and bicarbonate ions, the concentrations of these species could not be measured. The most logical sources of these ions in the BWR are condenser air inleakage, where carbon dioxide hydrolyzes to carbonic acid, which then breaks down to bicarbonate and carbonate ~ ions, and' the decomposition of organic matter to carbon dioxide. followed by the same hydrolysis and equilibrium dissociation reactions. We have ascribed the equivalents difference totally to the bicarbonate ion, owing to the neutral pH of the water. The carbonate ion would make a greater
. contribution at higher pH values. This somewhat crude method shows the concentration of reactor water bicarbonate ion ranges between 10 and 30 ppc. The addition of the bicarbonate conductivity contribution to the sum octained using the equivalent conductances of the separate ions makes the acreement with the process conductivity meter somewhat better f or rhose cases where dae meter reading was originally higher.
9-88
Rassmendations for Future _Studit:c In spite of all the precautions that were taken for this study, there is strong evidence tnat many samples suffered from contaminaticn.
either in the acquisition, analysis, or both facets of the determination. Contamination has been, and always will be the Achilles heel of grab sampling characterizations at the part per billion level.
Using the automatic sample changers that were designed for this test. '
sample lines should be hard piped directly to the rotary valve ports, with additional valving provided to accomodate high velocity bypass flow immediately upstream of the sample changer. Sample contact with fresh metallic surfaces should be minimized. Sample pump heads should $-
be fabricated from Kel-F or some other inert, non-leaching material:
incoming sample lines should be Teflon as much as possible. At Dresden-2, hard piping of sample lines will require that two separate i, instrument packages be provided, since condensate /feedwater sample lines are located in the turbine building, and reactor water cleanup sample lines are located in the reactor building. This configuration would permit more sample analyses per day on any given stream, and provide a better opportunity to detect transients in a timely fashion, should they occur.
O 9
9 9
9-89
.