ML19344B295

Use of Hydrogen to Control RCS Oxygen During TMI-2 Recovery Period

ML19344B295
Person / Time
Site:	Crane
Issue date:	07/31/1980
From:	BABCOCK & WILCOX CO.
To:
Shared Package
ML19344B293	List: ML19344B292 ML19344B295
References
PROC-800731, NUDOCS 8008260387
Download: ML19344B295 (18)

Text

{{#Wiki_filter:- July 1980 g TASK 19 - CHEMISTRY AND RADI0 CHEMISTRY SUPPORT Use of Hydrogen to Control RCS 0xygen During TMI-2 Recovery Period BABCOCK & WILCOX COMPANY CONTPACT NO. 595-7100 1.Gl b z P.L. 30 SI Prepared By: bd b_ Reviewed Sy:- / Apr tved ByN_,%g.. 0 8008260

SIM4ARY A study was conducted with the MAKSIMA-CHEMIST con.puter code developed by Atomic Energy of Canada Limited (AECL) on the use of dissolved hydrogen in the reactor coolant system (RCS) at THI-2 to control oxygen. The study showed that dissolved hydrogen is a viable method for maintaining the oxygen below the desired level of 0.1 ppm in the RCS during the recovery period. 4 1 9 . - ~.., _

INTRODUCTION This report deals with a study to confim the use of dissolved hdyrogen (i.e., hydrogen overpressure) in the reactor coolant system (RCS) at TMI-II during the recovery period controls the oxygen fomed from the radiolytic decomposition of the water and the oxygen entering the RCS with fresh makeup. BACKGROUND Dissolved oxygen is an important factor in the chloride stress crack-ing of austenitic stainless steels. The two potential sources of dissolved oxygen in the reactor coolant system (RCS) of a nuclear power plant is the oxygen fomed from the radiolysis of the water and oxygen entering with the fresh makeup water. PWR operating experience and infomation from other sources have shown that maintaining excess dissolved hydrtgen in the RCS is an effective method for controlling the oxygen from these two sources during power operating conditions and when the plant is shutdown (i.e., the reactor is subcritical). For this reason when the RCS water chemistry speci-ficatiM were established for the TMI-II recovery ceriod it was soecified that the dissolved hydrogen should be maintained between 5-25 std cc/kg water with the upper limit being related to the concern over non-condensables in the RCS rather than to the oxygen problem. The oxygen technical specifi-l cation for nomally operating PWR plants is 0.1 max and the ocjective was to control the oxygen within this specification. As the recovery period progr9 ;ed the Nuclear Regulatory Coccission (NRC) requested that evidence M provided to verify tnis method for TMI-2 RCS oxygen control. The Md a, verification study was conducted with 2 the MAKSIMA-CHEMIST Come.4er c ue developed by the Atomic Energy of Canada Limited (AECL). CHEMISTRY CONTROL From April to December last year fresh RCS rakeup was added via the makeup ar j purification (MU) system. The makeup wiur was prepared by mixing one part of borated water saturated with air at 130 F with three parts of deioni:ed water containing abcut 25 ppb 0. The average dissolved 2 2

1 Oxygen in the resulting mixture was about 2 ppm. During December, January and February, part of the fresh makeup was supplied from the Standby Pressure / Control (SP/VC) system when the system was operated intennittently for testing Since March the SP/VC system has operated full time and has been purposes. the only source of fresh RCS makeup. The oxygen in the fresh makeup from the SP/VC system has been controlled to an average of about 0.05 ppm (50 ppb) by heating and spraying the water in the supply tank and by maintaining a hdyrazine residual in the tank. In June 1979, the average fresh makeup rate to the RCS was about 0.6 gpm and then it gradually decreased to an average of about 0.2 gpm in September. Since September the average rate has been in the range of 0.1-0.2 gpm. As shown in Figure 1 the average dissolved hydrogen in the RCS for the first year of the recovery period was 16.5 std cc/kg water based on weekly analyses of RCS samples at B&W's Lynchburg Research Center. For the three months since that period (i..e., April, May and June) the average was 14.1 std cc/kg water. It has been only ore week where the hydrogen was analyzed to be less than 5 std cc/kg and that occurred in early May 1979. WATER RADIOLYSIS The absorption of nuclear radiation by the reactor ccolant passing through the nuclear core results in the decomposition of the water through a complex series of radiolytic chemical equations. Many researcherr have outlined radiolysis models that present a varity of chemical rate equations in an attempt to predict the radiolysis phenomenon. In a recent paper by Boyd, a radiolysis model, complete with chemical rate equations, chemical 3 species, and experimentally detemined rate constants and G values was presented which was demonstrated to be acceptable for the prediction of steady-state oxygen and hydrogen concent-ations in low radiation fields and for approximately neutral solutions. This radiolysis model is pre-sented in Table 1. The initial yield of the primary radiolytic species, H, OH, H202, H2, E, H+, and OH, is usually expressed in a tem called G. The G value is the number of molecules or atoms produced for every 100 eV of radiation.

energy absorbed by the water. Exposure of the reactor coolant to nuclear radiation in the core region produces stoichiometric concentrations of the seven chemical species previously identified. The chemical radicals OH, H, H02 are highly reactive and will consume any free oxygen or hydrogen as illustrated by the following chemical equations: H + H2O (1) OH + H2 = H02 (4) H + O2 = H202 (14) H02 + H = OH + H2O (15) H20 + H = Chemical equation (1) shows that an OH radical, produced by radiolysis, reacts with dissolved hydrogen to produce H2O and a H radical. The H radi-cal ' reacts with free oxygen (Equation 4) to produce another radical H02, which in turn reacts with H (Equation 14) 'to produce hydrogen peroxide H202. The hdyrogen peroxide also reacts with the H radical (Equation 15) to produce H2O and forms a new OH radical which can initiate the chain again. Thus, oxygen and hydrogen a'e removed from the coolant by the radiolysis r Since there are many other rate equations in Table 1, the complete process. radiolysis model must account for all the rate equations in order to predict the steady-state equilibrium oxygen and hydrogen concentrations. MAKSIMA-CHEMIST CODE For each of the 31 chemical equations identified in Table 1, an ordinary dif-ferential equation can be derived, but integration and simultaneous solution of all these differential equations is quite difficult. Thus, the MAKSIMA-CHEMIST program developed by AECL can easily solve the simultaneous equations. The program is capable of calculating steady-state chemical concentration for any number of chemical species defined in any number of chemical rate equations. The program input consists of defining rate constants for each of the chemical rate equations. For each chemical species, provision is made to input the species charge, G values, the rate of spontaneous fomation, and the initial species concen-tration. Additionally, the irradiation dose rate for any number of times may be specified. The program output consists of chemical species concentrations at each specified time step. -4

ENERGY RELEASE RATE TO THE REACTOR COOLANT In order to reasonably predict the oxygen and hydrogen concentrations in the reactor coolant, the nuclear energy deposited into the coolant must be estimated. The energy deposition in the TMI-2 coolant from the core is primarily due to two sources: (1) gama emissions, and (2) beta emissions. The reactor core gama dose rates were calculated with the LOR 2 com-5 puter code.# LOR 2 is a modified version of the ORIGEN computer code which uses time-independent cross sections for fuel materials derived by B&W de-pletion codes. The LOR 2 code input was modeled to account for the actual TMI-2 core burnup prior to the accident, after which time all of the core gama power from both the heavy metals and the fission products are sumed at specified decay times to obtain total gama dose rates. Beta energy deposition in the coolant from penetration through the clad-ding represents only a small fraction, 2-3 percent, of the total energy de-posited into the coolant.6 The coolant beta energy deposition due to fission products escaping to the cc:.lant'(assuming 100% decosition) is three to four orders of magnitude less t;'an the core gama dose rate. Therefore, the beta deposition rate in the coolant has been neglected. Listed in Table 2 are the total core gama dose rates at various decay times after the accident. Gamma energy deposition to the coolant in the core region was estimated by Argonne National Laboratory to a first approximation to be about 12.6% of the total photons released from the core. Furthemore, 7 NRC Regulatory Guide 1.7 recomends a conservation percentage of 10% for the core gama fission product radiation energy absorbed by the coolant. Since both gamma absorption percentages agree so closely, the 10% value is used mainly for ease of calculations. CALCULATED RCS OXYGEN AND HYDROGEN CONCENTRATIONS The MAKSIMA-CHEMIST code was modeled to simulate the RCS operating conditions at various times during the recovery period in order to calculate the hydrogen and oxygen concentration. Operating conditions that are simu-lated include the RCS hydrogen concent'ation, reactor coolant makeup flow to the RCS, oxygen concentrations initiaily in the RCS and in the makeup, y m m@ =

and the gama dose rate absorbed by the coolant at various shutdown times. Although the RCS hydrogen during the first year of the recovery effort has averaged about 15.5 std cc/kg, the RCS hydrogen concentrations were I assumed to be the minimum guideline specification concentration of 5 std cc/kg. Pricr to operation of the Standby Pressure-Volume Control (SPVC) system, the makeup oxygen concentration was about 2 ppm. After the SPVC system was opera-tional, the oxygen concentration in the makeup was reduced to about 50 ppb (0.05 ppm). The initial RCS oxygen concentrations were conservatively as-sumed to be equal to or greater than the makeup oxygen concentrations.be-cause assuming no oxygen removal, the oxygen level in the RCS would approach the oxygen concentration in the makeup. In all cases the gama energy ab-sorption fraction was assumed to be 10%. The average pH value during the recovery period has been about 7.8. The MAKSIM code calculates the oxygen and hdyrogen RCS concentrations based on the four sets of RCS operating con-ditions listed in Table 3. The MAKSIMA code can only vary the dose rate as a function of time, hence, the code cannot simulate the changes in RCS makeup flow rates and the decrease in the makeup oxygen concentration during the entire recovery time span. Therefore, the four sets of RCS operating conditions summarized in Table 3 model the actual RCS operations at specific recovery times to detemine the RCS oxygen concentration. For example, the first two sets of RCS operating conditions represent the RCS operation prior to SPVC system operation when the makeup flow rate corresponded to 0.3 gpm and 0.2 gpm at 120 days and 250 The third set of RCS operating conditions represents RCS days, respectively. The operation one year after the accident with the SPCV system in service. last set of RCS conditions conservatively estimates the RCS operation two years into the recovery period. The results of the MAKSIMA code calculated RCS oxygen and hydrogen concentrations for the four time periods outlined in Table 3 are shown in As can be seen in each of the four figures the initial Figures 2,3,4 and 5. RCS oxygen concentration is reduced quickly to concentration levels signifi-cantly less than the 0.1 ppm oxygen control limit. Maintaining the. 1

RCS hdyrogen concentration at the minimum level of 5 std cc/kg insures the effective removal of RCS oxygen and oxygen introduced by the RCS makeup. USE OF HYDRAZINE As stated in the section on " Chemistry Control," a hydrazine residual is maintained in the fresh RCS makeup supplied from the SPVC system. Analyses of several RCS samples in March and April indicated that the RCS contained 4 ppm hydrazine. Although the normal reaction rates between oxygen and hydrogen are temperature dependent and are slow at the tempera-tures in the RCS (100-150 F), it has been shown that a radiation field actually promotes the effective reaction between oxygen and hydrazine even though the radiation field also tends to decompose the hydrazine.8In this way the hydrazine supplements the RCS oxygen control method with the hydrogen overpressure. I g@p%fN O MMWS

CONCLUSIONS Based on the MAKSIMA computer code oxygen and hydrogen concentrations results in Figures 2, 3, 4 and 5, the conclusions of this TMI-2 RCS oxygen control study are: 1. Prior to operation of the Standby Pressure Volume Control System, RCS oxygen and oxygen introduced into the RCS by the makeup was suppressed to concentrations significantly lower tnan tne oxygen control concentration limit of 0.1 ppm. 2. After placement of the SPCV system in operation, RCS oxygen con-( centration was maintained well below the RCS oxygen control concentration. l f 3. Maintaining the RCS hydrogen concentration at 5 Std cc/kg will insure oxygen suppression below the oxygen control concentration for future operating situations involving high RCS makeup rates (.1 gpm) containing ~ oxygen concentrations of 1 ppm. l f 4. The use of hydrazine to maintain a residual in the fresh RCS makeup i from the SPVC system provides an additional method to help control the RCS oxygen. RECOMMENDATIONS Continue to maintain the RCS dissolved hydrogen greater than 5 Std cc/kg and continue the present mode of controlling the oxygen in the fresh makeup supplied from the SPVC system. 8-G r m -m ,,.y _... gg* Q Q

( REFERENCES 1. J.H. Hicks, TMI-2 Recovery Project, B&W Water Chemistry Manual, NPG0 TRG-79-11, October,1979. 2. M.B. Carver, D.V. Hanley and N.R. Chaplin, Atomic Energy of Canada, Ltd. Report, AECL-6413,1979. A.W. Boyd, M.B. Carver and R.S. Dixon, " Computed and Experimental 3. Product Concentrations in the Radiolysis of Water", Radiat. Phy;. Chem. 1980, 15, 177. 4. A.Z. Livolsi and H.C. Gabler, LOR 2-Isotope Generation and Depletion Code, Babcock & Wilcox Report, NPGD-TM-497, March 1979. 5. ft.J." Bell, Oak Ridge Mational Laboratory Report', ORML-4623,1973. 6. J.R. Honekamp, S. Gordon, K.H. Schmidt and D.J. Mallory, "An Analysis of the Hydrogen Bubble Concerns in the Three-Mile Island Unit 2 Reactor Vessel", Argonne National Laboratory, Argonne, IL 60439. Regulatory Guide 1.7, " Control of Combustible Gas Concentrations in 7. Containment Following a Loss-of-Coolant Accident", Rev. 01, Sept.1976. P. Cohen, Water Coolant Technology of Power Reactors, P.119, Gordon and 8. Breach Science Publishers, New York, NY (1969). _g. Oh e

TABLE 1: WATER RADIOLYSIS REACTIONS AND YIELDS USED TO CALCULATE PRODUCT CONCENTRATIONS Rate Constant, Liter-mol-I -I -sec Chemical Reaction H + H2O 3.6E + 07 1. OH + H2 = 2. OH + H202 = H02 + H2O 3.3E + 07 02 + OH-9.0E + 09 3. OH + 02- = H02 1.8E + 10 4. H + O2 = H02-2.0E + 10 5. H + 02- = 02-1.9E + 10 6. E- + 02 = 7. E- + H202 = OH + OH-1.2E + 10 1.3E + 10 0( = H02 + OH-8. E- + H 2.2E + 10 9. E- + H = H + OH-2.0E + 01 10. E- + H2O = 11. E' + H02 = 0- + OH-3.5E + 09 H2O + 02 1.2E + 10 12. OH .+ H02 = H202 5.5E + 09 13. OH + OH = H202 2.0E + 10 14. H + H02 = 15. H + H202 = H2O + OH 9.0E + 07 E- + H2O 2.1E + 07 16. H + OH- = 02 + H02-8.9E + 07 17. H02 + O2- = H202+ 02 2.0E + 06 18. H02 + H02 = H02 4.5E + 10 19. H+ + 02- = H+ + 02-8.0E + 05 20. H02 = 21. H+ + H02 = H202 2.0E + 10 H+ + H02-3.56E - 02 22. H202 = H2O + 0-1.2E + 10 23. OH + OH = OH + OH-1.7E + 06 + H2O 24. 0+ = 1.43E + 11 25. H + OH-H{0 = H + OH-2.6E - 05 26. H2O = H2O 2.0E + 10

27. H

+ OH = H2 1.0E + 10 28. H + H = H2 + OH-2.5E + 10 29. E- + H = H2 + OH + OH-5.0E + 09 30. E- + E- = OH-3.0E + 10 31. E- + OH = Primary Yield E-H+ H H2 OH OH H202 H02 P02 02 02 G(Molecules /100eV) 2.7 2.8 0.55 0.45 0.1 2.75 0.7 6. >, _., -mh -__u-*m. 'j** W _7- -

TABLE 2: TMI-2 CORE GAMMA DOSE RATES AFTER THE ACCIDENT Decay Time Total Gamma Power Hours eV sec-1 liter-1 0 2.9 E + 22 4.80 E + 01 1.7 E + 21 2.40 E + 02 8.2 i + 20 1.44 E + 03 2.0 E + 20 2.88 E + 03 1.0 E + 20 6.00 E + 03 2.9 E + 19 8.76 E + 03 1.1 E + 19 '1.50 E + 04 2.7 E + 18 1.75 E + 04 2.1 E + 18 l l

TABLE 3: MAKSIMA-CHEMIST RCS OPERATING DATA USED TO CALCULATE OXYGEN AND HYDROGEN RCS CONCENTRATIONS TIME (DAYS) 120(3) 250(3) 365(4) 730 PARAMETER 5(I) 5(2)- 5(2) 5(2) RCS H Concentration, cc/kgm 2 RCS 0 Concentration, ppm 2 2 1 1 2 0 Concentration makeup, ppm (l) 2 2 0.05 1 2 Makeup Rate to RCS, gpm(l) 0.3 0.2 0.2 1 7.8 7.8 7.8 7.8 pH ~I.1~I 1.0E19 2.9E18 1.1 E18 2.1E17 Gamma Absorption Dose Rate, eV-s Gamma Dose Absorption Fraction 0.1 0.1 0.1 0.1 (1) Constant Concentration or Rate (2) Initial Concentration (3) Prior to operation of SPCV System l (4) After operation of SPCV System l l O l l - \\

1. 6 ene e L

, w1

a e e e f f .f t 9 9 f _k t t f .e 9 9 m r.... 4 C tr 8 C sa r a 5ll u C2 s -- m ta L ee 3 a t s C 8 u Z u g C o a = y e 0 h m r-a "g = v Lt c0 e C u C w e a U s N u k n C s o C U Z e m = r-ec = o r w v v c b C a r-m c u a x La m N z z t-1 a l r-8

• L3 p

3 D g = a 0 e r-cs m ._D s 3 = ~3 C = p 6 m N a = i m; u m r-m C s I a t C a N 8 * - re m rt. C 22 ca tv

=

m + + W + L w SM/oc *t:aSo.:pAH Babcock & Wilcox

---

a,,,. -..

Concentrations Figure 2. RCS H2 and 02 120 Days Into Recovery l 0 -'1_ i i i l I i t i i 1 F I a a _g...-. 10-4 a = -, -n., - ~ 10-'- 0 CONTROL LEVEL (0.1 ppm) g e a % J L, 's 1 g o -* E 02 H2 N 10 -'- 2 m __J O ] bs.i.-_. i r 10-*-. ua Z n-1:: c J L) b LT. z 10-" - Z LL1]D.-12 -- l o .n_. O u 10-".. e 0--e l 10-" -.- I i i i l i l t i f f _) 30' Id# id Id ' T)MF tSFr0NDS1 1

Figure 3 RCS H2 and 02 Concentrations 250 Days Into Recovery I I j I j i i i j i 10-3 .. m.g,... l N 10 -' 3r 10 3 0 CONTROL LEVEL (0.1 ppm) { 2 l w 10-5 Of ~ 1 d 10 '- u2 O "o n l- _r O ~ r 10 ~*- }Q'9-- t u,z l S 10-'*- I-- C 5 10'" - z wu 1 0-'z z-Q U 0- e a i l ) 0- -- I l i f I I I I I l 1 1 30' Id * !d' id' . TIME ! SECONDS 1 n CD M

r Concentrations Figure 4. RCS H2 and 02 1 Year into Recovery 10"!- = - mnww -_- 10 4 t 4

:.t e_ea n.e 10.s; 0 CONTROL LEVEL (0.1 ppm) j l

2 _1. tu l o -5 c: H 7 -_310 s 1 wa 10, 02 . H2 O 0 r 10 "-- cn 10..go -- z o-1 O'" - W-cc C 1 0 * -- -z Wu 1013 z au 10-"__ _9__ l 10.is 2 30' 10' 10' 10 ' 2 t 2 TIME (SECONDS) 16-

Concentrations Figure 5. Projected RCS H2 and 02 2 Years Into Recovery l i t I I I I I I 10 -3 l . :..=. = : =-. 10"- u

: :- - e::::=-

2 (ONTROL LEVEL C 0 0.1 ppm) 10 ~* -- y K W, n -6 _ .-. 1 LJ ) _,,j 1 0 1 LJ 02 H2 9 ~ " ' " a e m -S _ _ Oz 10 -9 a $ 10-' - x ~ p. -rC O -- u Z i ou 10-'. I 1 i-t t l f_ l i I I l 3 -l 1 0-"-{~j 8 2 5 0 ~ 6-2 5 3 10' 10, 13 10 2 5 TIME toECONDS) . i t s --n + r- . - - - - - + ,my- ,m_ -w.- -,-,}}