ML20100E507
| ML20100E507 | |
| Person / Time | |
|---|---|
| Issue date: | 03/31/1985 |
| From: | Hopenfeld J NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | |
| References | |
| NUREG-1108, NUDOCS 8504040001 | |
| Download: ML20100E507 (45) | |
Text
- __.
NUREG-11 8 i
' Radioactivity Transport Following Steam Generator Tube Rupture U.S. Nuclear Regulatory Commission
, Offica of Nuclear Regulatory Reseerch i J. Hopenfeld 1
,f* """%
v 2
i
\\,.....p/
1 J
h i
l l
8504040001 850321 PDR 1
8R
A NOTICE Availability of Reference Materials Cited in NRC Publications Most documents cited in NRC publications will be available from one of the following sources:
- 1. The NRC Public Document Room,1717 H Street, N.W.
Washington, DC M555
- 2. The NRC/GPO Sties Program, U.S. Nuclear Regulatory Commission, Washington, DC 20555
- 3. The National Tec hnical Information Service, Springfield, VA 22161 Although the listing that follows represents the majority of documents cited in NRC publications, it is not intended to be exhaustive.
Referenced document <. available for inspection and copying for a fee from the NRC Public Docu-ment Room include N RC correspondence and internal NRC memoranda; NRC Office of Inspection and Enforcement bulletins, circulars, information notices, inspection and investigation notices; Licensee Event Reports; vendor reports and correspondence; Commission papers;and applicant and licensee documents and correspondence.
The following documents in the NUREG series are available for purchase from the NRC/GPO Sales Program: formal NRC staff and contractor reports, NRC-sponsored conference proceedings, and NRC booklets and brochures. Also available are Regulatory Guides, NRC regulations in the Code of Federal Regulations, and Nuclear Regulatory Commission issuances.
Documents available from the National Technical Information Service include NUREG series reports and technical reports prepared by other federal agencies and reports prepared by the Atomic Energy Commission, forerunner agency to the Nuclear Regulatory Commission.
Documents available from public and special technical libraries include all open literature items, such as books, journal and periodical articles, and transactions. Federal Register notices, federal and state legislation, and congressional reports can usually be obtained from these libraries.
Documents such as theses, dissertations, foreign reports and translations,and non-NRC conference proceedings are available for purchase from the organization sponsoring the publication cited.
Single copies of NRC draft reports are available free, to the extent of supply, upon written request to the Division of Technical Information and Document Control, U.S. Nuclear Regulatory Com-mission, Washington, DC 20555.
Copies of industry codes and standards used in a substantive manner in the NRC regulatory process are maintained at the NRC Library, 7920 Norfolk Avente, Bethesda, Maryland, and are available there for reference use by the public. Codes and standards are usually copyrighted and may be purchased from the originating organization or, if they are American National Standards, from the American National Standards Institute,1430 Broadway, New York, NY 10018.
GPO Printed copy price:
$3.75
NUREG-1108 Radioactivity Transport Following Steam Generator Tube Rupture Manuscript Completed: December 1984 Data Published: March 1985 J. Hopenfeld Division of Accident Evaluation Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washington, D.C. 20555
,f- ~%,,
($),,
1 i
i i
4 L
m.
TABLE OF CONTENTS l
_Page ABSTRACT...............................
1 INTRODUCTION.............................
1 PREVIOUS WORK.............................
2 FORMULATION OF THE PROBLEM......................
5 DATA REQUIREMENTS............................
10 PROGRAM DESCRIPTION..........................
14
SUMMARY
17 REFERENCES..............................
18 i
APPENDIX...............................
33 I
l I
I 4
4 I
iii
k LIST OF. TABLES
.P.a_qe 4
Table 1.
Operating Experience During SGTL for Selected Plants....
21 Table 2.
Participants / Responsibilities................
22 l
Table 3.
Primary and Secondary Bundle Scaling............
23 f
4 i
J t
)
V
LIST OF FIGURES Page Figure 1.
Overall Iodine Mass Balance................
25 Figure 2.
Fraction of Iodine in Steam, F, as a Function of the Parti tion Coefficient, P.................
26 Figure 3.
Pressure Temperatures, and Mass Flow Rates Following SGTR 27 Figure 4.
Pressure, Temperatures, and Mass Flow Rates Following SGTR Wi th a Stuck Open Safety Valve..............
28 Figure 5.
MB-2 Primary and Secondary Loop Instrumentation......
29 Figure 6.
Schematic of ORNL Experiments...............
30 Figure 7.
Schematic of Carthwestern University Experiments.....
31 i
2 a
vii
i
?
4 9
ABSTRACT A review of the capabilities of the CITADEL computer code as well as plant sxperiencetoprojectradioactivityreleasesfollowingasteamgeneratortube rupture in PWR's shows that certain experimental data is needed for reliable off-site dose predictions.
This article defines five parameters which are the key for such predictions and discusses the functional dependence of these parameters on various operational variables.
The above parameters can be used in conjunction with CITADEL or they can be inserted in the appropr.iate equa-tions which then conveniently can be programmed as a subroutine in thermal-hydraulic system codes.
A joint Westinghouse, Electric Power Research Insti-tute and Nuclear Regulatory Commission Program aimed at obtaining the five parameters empirically is described.
INTRODUCTION Operating experience with PWR steam generators (Reference 1) shows that the j
tubes in these heat exchangers are sometimes prone to corrosion and mechanical-ly induced damage. When such damage results in a leaking tube, the potential exists for radioactive fission products to escape into the atmosphere whenever a secondary side steam dump occurs.
This occurrence has long been recognized and accounted for by setting operating limits on the radioactivity levels in the primary coolant so that even if a full tube rupture occurs the activity release will not exceed allowable amounts.
Indeed, although six steam genera-tor tube rupture (SGTR) incidents have occurred in the United States none has resulted in activity dotes which exceed the limits set in 10 CFR Part 100.
(300 ren to the thyroid in 2 hrs).
However, experience with the SGTR at R.E.
Ginna (Reference 2) suggests the assumption, commonly invoked in the analysis 1
of SGTR events, that only cry steam is released through the relief and safety valves may not be appropriate.
This is because the concentration of fission products, primarily iodine, may be higher in the wet steam.
The presence of water in the released steam can result from entrained water droplets in the steam which escaped removal by the separators due to low removal efficiency of the latter during the transient.
The inclusion of water release in the design basis analysis of SGTRs could potentially lead to stricter limits on the maximum allowable iodine concentra-tions in the primary reactor coolant.
An inadequate' understanding of radioac-tivity transport following SGTR may dictate the use of conservative calculation of the allowable iodine limits.
Should these calculated allowables turn out to be very strict they may in turn impose undue restrictions on reactor opera-tions.
In response to this background, Westinghouse, (W), the Electric Power Research Institute, (EPRI), and the NRC have recently initiated a cooperative program aimed at the generation of the required data base which would facilitate mechanistic modelling of SGTR as well as other steam generator events.
The purpose of this paper is to set forth the requirements and describe the imple-mentation of this program.
PREVIOUS WORK 1.
THEORETICAL PREDICTIONS Frevious studies (References 3, 4, 5, 6) proposed a theoretical model for the transport of radioactivity, by micron size droplets, which are carried by bubbles into the steam space.
These droplets are entrained by the rising bubbles when the primary coolant flashes into the secondary side of the steam generator.
Starting with first principles of heat and mass transfer, the authors have developed equations to describe the behavior of the droplets within the bubble and their deposition on solid surfaces upon impaction.
These equations were then programmed into a computer code (CITADEL) for use to predict radioactivity releases.
2
The CITADEL code divides the steam generator into three regions: water, steam, and moisture separators.
Primary coolant from a submerged leaking tube is j
transported to the steam space by rising bubbles' while a leak above the water level results in a flashing jet which introduces primary liqtil droplets into the steam space.
A number of features in CITADEL sust be assessed before the code can be consi-dered as a reliable tool for radioactivity release predictions.
First, the code contains numerous input parameters, such as droplet and bubble diameters, i
and' iodine partition coefficients, which themselves are subject to large uncertainties making the evaluatien of the results very difficult.
Secondly the droplet transport mechanism requires that the primary coolant flash, upon entering the secondary side, an event which sometimes may not even take place.
Flashing may occur prior to reactor scram because the primary coolant tempera-4 l
ture is higher than the secondary coolant temperature and the secondary side is in a saturated condition.
Following reactor scram, reverse heat transfer may arise and for certain accident scenarios the secondary coolant temperature becomes higher than f.he primary coolant temperature but since the secondary coolant is still saturated flashing does not occur.
Prior to reactor' scram, however, the relief / safety valves are closed so only a negligible amount of radioactivity could escape via the ejector pump and, therefore, knowledge of radioactivity transport during this period is of secondary interest.
Following reactor scram the relief valve opens periodically, j
2 but since no flashing takes place within the steam generator the CITADEL code will predict no radioactivity release due to liquid entrainment.
If mechanisms other than flashing exist for liquid carryover, the CITADEL results may underestimate the actual radioactivity release.
2.
OPERATING EXPERIENCE 1
More than 300 incidents of relatively small steam generator tube leaks have j
been recorded in this country (Reference 1).
Since plants are not required to shut down when the total primary to secondary leakage is below one gpm some c
plants have operated with leaking tubes for sufficiently long time to allow 3
s a
.-n_,
measurement of radioactivity build up and distribution on secondary side components.
9 Table 1 presents radioactivity measurements' which were obtained for various time intervals at five different steam generator units.
The Table presents J
concentration of I-131 in the primary coolant, the steam generator water and the main line steam.
The last column in the Table represents the ratio rounded off to the nearest order'of magnitude (i.e., 1, 10, 100, etc.), of the concen-4 tration of i'odine in main line steam to the concentration of iodine in the
~
steam generator water. At low iodine volatility the concentration of iodine in steam'would be predominantly due to droplet entrainment and therefore this j
ratio would oe a direct measure of liquid carryover.
On the other extreme when.
the liquid carryover in the steam is very low the last column in the Table would represent the partition coefficient which physically characterizes iodine volatility.
Table I leads to two main conclusions.
First there is a large variation from i
plant to plant in the iodine retention characteristics of the secondary system.
These large differences could be possibly reconciled by differences in leak l
location, measurement techniques and differences in system operations.
In the absence of a satisfactory explanation of the plant data one must conclude that l
present pla'nt experience is not sufficient to allow extrapolation of such data l
to the design basis SGTR.
Secondly, the large plant to plant variations in the steam to water Iodine concentration ratio does not allow one to conclude i
whether iodine reaches the steam line predominantly by entrainment or volatili-ty.
For example, since a typical W steam generator is rated at 0.2% moisture
-4 carryover, the activity ratio in Table I would not be expected to exceed 10 (dividing.2% by 20 to allow for differences in densities between water and steam), yet it does exceed this value by two order of magnitudes for both Turkey Point Units 3 and 4.
If one was.to draw a conclusion from this that iodine concentrations in the steam are governed by volatility and not be entrainment than such a conclusion would stand in contradiction with the Point Beach data which indicates tae opposite:(i.e., iodine entrainment predominates volatility).
i i
4
. - ~.
.~.
In conclusion, neither the present analytical tools nor the acchmulated steam generator experience allows reliable best estimate predictions of radioactivity releases following steam generator tube rupture.
FORMULATION OF THE PROBLEM In formulating the present program, several parameters have been identified which are deemed to be essential for assessing the CITADEL code or for laying down a foundation for a modified computer code if required.
The approach selected here was to define parameters which can be measured and at the same time are related to known physical processes.
These parameters represent everage quantities in time and space, and strictly speaking, are valid only for the range of conditions which they were measured.
This approach risks the possibility that some SGTR events could fall beyond the bounds for which these parameters were characterized; however, looking at the problem as a whole, and particularly its complexity, it was concluded that such an integrated approach would yield the most beneficial results.
The above mentioned parameters can be programmed on a subroutine in a system code such as TRAC or RELAP and they can also be used to assess CITADEL code and modify as required.
Figure 1 is a schematic representation of the various routes which fission products may take from the tube rupture site to the relief valves.
2 and 6 represent volumes containing a uniform mixture of steam and water.
3, 5 and 10 represent wall surfaces to which fission products adhere'by chemical or physi-cal sorption or by particle deposition.
9 represents the liquid in the down-comer and the liquid in the pool above the tube sheet.
A mass balance on the heated and riser sections, the separators, and the downcomer allows calculations of the amount of iodine leaving the steam gener-4 ator.
Heated section (above wa.ter level) and riser:
1.
d(Mc)H (mc)p, (mc)d - (mc)s bH
=
5
Separators:
2.
(ac), = (sc), - (ac), - S, Downconer:
3.
(ac), + (ac)f - (mc)d ~S
=
d f
In the above, M represents total mass within the specified volume, a is mass flow rate, c is the concentration of iodine per unit mass,' and S represents the rate at which iodine either deposits on or redepcsits from the walls. The subscripts are defined in the nomenclature.
It should be noted that m cj = (ac)g and the concentration c refers to iodine j
concentration in water / steam mixture or the concentration of iodine in liquid alone or steam alone when the subscripts 1 and g are employed.
1 If K is specific activity (which converts iodine mass to radioactivity), then the total offsite dose is given by:
4.
a = K f (m, c -S,)dt o
Equations 1-3 can be reduced into a more tractable form by introducing the following variables.
Fraction of primary liquid reaching the separators directly without prior mixing l
s, mass of primary liquid entering separator S.
S_
p,1
- m mass of primary liquid entering secondary side p,g Overall separator efficiency for liquid removal:
6.
r1 _ "## - ' total liquid at the entrance to separator
_ liquid entering the downcomer from the separator 1
,1,s 6
"i
. Partitiori of iodine between gas 'and liquid:
1' 7.
(a) hc p
mass of iodine per unit volume of liquid y
,g
((
mass of iodine per unit volume of vapor
~
Overallmasstransfercoeffjcientforiodinedepositiononwallsurfaces.
b 8.
h a i
g
(%
The iodine entering and exiting the separators can now be written as:
at the entrance to.the separator 9.
(mc)s = p (1-x) (mc)p + (mc)H I
p (1-x) (m c)p + (m + m )H H
=
c g
g g
at the exit from the separator into the steam dome:
(mc), = [(mc)g + (mc)f]g
' 10.
=
=[
- 2], SE at the exit from the separator into the downcomer:
11.
(mc)g = nmt,s g c
The. first terra on the right hand side of equation 9. represents the mass of 1
iodine in the primary liquid which did not participate in the mixing with the secondary side fluid.
The second term in equation 9 represents the iodine carried by the steam / water mixture (m ) in the riser section.
H r.
3 h
7
=
l The first and second terms on the right of equation 10 represent the iodine in the steam and in the liquid respectively.
The Appendix discussed in more detail the deriv.ition of equations 9 and 10.
Equations 1, 2, 3, 9, 10, and 11-represent a general mass' balance and therefore i
are not restricted to a specific accident scenario.
Some of the terms appear-ing in these equations are governed by differer,t mechanisms depending on leak
.I location and the time during recovery from the SGTR event.
For example, the total moisture content m entering the separator will be expressed different-As l
ly prior and following steam generator isolation.
In the first case'the flow l
will be characterized by a bubbly flow while in the second case by a misty flow.
As an example only we shall proceed and consider the second case when the liquid entering the separator is in a form of entrained droplets and the break site, under all conditions, is covered by the steam / liquid mixture.
This r
situation is especially of interest because it is only during this period of time tnat the steam relief value allows the escape of the contaminated steam /.
water mixture into the atmosphere.
Prior to 5.G. isolation the escape of iodine to site boundaries occurs through the steam air ejectors.
During the' time period when the relief valve is open equation 11 can be written as:
lla (mc)g = (p (1-x) m + E,)q c, p
The term p (1-x) m represents the primary liquid entering the separator p
while E represents the liquid entrained by steam rising from the s
saturated liquid pool..The term m in equation 9 can now be expressed as H
H " "g + E
- 12 m
s A substitution of equations 5-11 in equations 1-3 results in three equations for the three concentrations c
, c, and c
- c d
dI N )H =-(1-p)'(1-x) (mc)p + (mc)d - ("g )c - (hAc)H 12.
H or 8
{
k [cf (Mf. + f M )]H * "p cp [(1-p)(1-x) + f x]p +
12a.
g i
AH [E + f a ]H ~ (hAc)H cd "d -c g
13.
m pg g
((
] + 1 q[(1-x) p m + E,))c, =
p
~p (1-x) (mc),
q [E
+ (1-*) 0"p]c fH[E "g H3~
+c
+
s e
s
[(1-x) p(mc)p + cfH[E + f m ]H 3
g (hA)s pm
+m
+E p
g s
14*
d(Mc)d = q [p (1-x) m +E3C + (mc)f -(mc)d - (hAc)d p
s c
i A total of 21 parametsrs are required to solve these equations, 8 parameters (M, M, m, m, (P ), m, x) are obtainable from a mass anc. heat balance which A
g g
g p
d 9
4 can be performed by RELAP or TRAC, the three surface areas (A, A ' A ) are s
d-H known from geometrical considerations'and the two concentrations of iodine (c, c ) are given inputs.
The remaining 8 parameters p, q,_P,-P
, h, h p
f p
H s
u-E, are obtained-as discussed below.
It should be noted that p and E are also required inputs for the total mass and heat balance.
Again, equations 12 through 13 are written for a certain period of time during the transient and a certain location'of t'he break site.
For other situations-I these equations can be relatively easily modified.
3 1
I 9
.a..
DATA REQUIREMENTS 1.
Primary fluid mixing - p The degree to which the primary fluid mixes with the secondary fluid depends primarily on break location.
A break near the tubesheet will result in exten-sive mixing (p+0) in comparison to a break at the top of the U oend (p+1) where little mixing exists.
The difficulty in estimating E between these two bounds stems from the hydrodynamics of the jet at the break site.
Prior t.o reactor scram when the secondary side is in the circulating mode a leak at the top of the tube bundle, will cause a flashing jet of primary fluid to mix with a flowing two phase mixture of secondary fluid.
A leak at the bottom of the tube bundle, near the tube sheet, will cause a flashing jet of primary fluid to mix with a flowing subcooled secondary liquid.
Following reactor scram and steam line isolation, a leak located anywhere in the tube bundle, will result in either a flashing or nonflashing jet of primary fluid, to enter stagnant pool of saturated liquid.
The direction of heat flow between the primary and the secondary coolants will determine whether flashing will or will not occur.
At the beginning of the transient the primary fluid is hotter than the secondary fluid, however, as the primary fluid is cooled by steam release through the intact nonleaking steam generators, the secondary side of the faulted steam generator becomes hotter than the primary fluid and therefore reverse heat transfer takes place.
The enthalpy of the secondary fluid is now higher than the enthalpy of the primary fluid and therefore no flashing would occur.
Because of the complex flow geometry it is difficult to estimate how much liquid will actually reach the separator and, furthermore, it is even more difficult to determine what fraction of this liquid was originated during flashing and what fraction originated from the secondary side.
Data reported in the literature on aerosol scrubbing by suppression pools, jet mixing in simple geometries, droplet distribution from flashing jets, were all obtained under condition which are significantly different from those existing in a steam generator during shutdown following a tube rupture.
10
2.
Separator Efficiency q
PWR steam separators consist of two devices commonly known as the primary and the secondary separators. The primary separator employs a swirl vane which utilizes centrifugal force to separate water droplets from steam.
Such separa-tion depends on flow velocities, droplet size, droplet size distribution, and droplet concentration.
The secondary separator, also commonly referred to as the steam dryer, is constructed from corrugated plates and separation is achieved by impaction of droplets on these plates.
As steam velocity increases, a point is reached where the steam reentrains water droplets from the wall liquid film
. and consequently the dryer loses its effectiveness.
Separators are very effec-tive in drying steam near or at their design conditions.
The SGTR events, however, may give rise to flow velocities and particle loadings which are outside the design envelope and, therefore, tests are required to establish separator efficiencies for these off-design conditions.
Of particular concern here are (a) loadings with droplets which are less than 5 microns in diameter, (b) large steam velocities immediately following steam line break, (c) very high liquid loadings, and (d) flow reversal in downcomer.
Because of the uncertainties in scaling laws, separator efficiencies under these off-design conditions ca.1 be obtained only from prototype testing.
3.
Pool Entrainment - E When the condenser is not available for steam durp following SGTR, after a brief period of time the secondary side pressure reaches a certain level and is maintained at that level by periodic steam releases through the relief valve.
Each time the valves open the water pool level swells due to rapid void expan-sion on depressurization.
The mechanism for liquid entrainment under these conditions may be different than the mechanisni for liquid entrainment from a boiling pool at constant pressure where entrainment results from the breakup of liquid ligaments and protrusions following bubble bursting. at the surface.
If the entrainment mechanism during depressurization indeed is different than the entrainment mechanism during boiling, the available data on entrainment rates (Reference 5) maybe not be applicable.
Tests are required to check this e
point.
11
L 4.
Mass Transfer coefficient - h Molecular; iodine, I, and possible 10$ can be either physically or chemically -
i' adsorbed on surfaces which are immersed in water or. steam. Metal composition, coolant chemistry, temperatures and Reynold numbers are the main parameters which could control =the deposition of these species on steam generator surfaces.
i 5.
' Feed Water; Concentration - cf 4
The concentration of iodine in feed water depends on the. history of the SGTR event under' consideration.
Prior to reactor scram, iodine contaminated steam i
from the faulted unit is the source of iodine in feed water.
A mass balance on the entire secondary system is required to determine the feed water concentra-tion c.
Therefore, correlations for h should also include the appropriate f
conditions existing in the secondary system outside the faulted unit.
ii 6.
Partition Coefficient - P i
When the steam _ at the outlet from the steam generator is very dry (X=1), steam will be the sole carrier of iodine into the' environment, and its amount is l
governed solely by the partition coefficient.
When the steam is wet, the magnitude of the partition coefficient will determine whether-for a given steam quality water or steam will be the predominant carrier.of iodine.
Figure 2 i
compares the relative contributiuns of iodine in steam and iodine in water to the total radioactivity release.
Unless the moisture in the steam exceeds 1%
most of the iodine is carried by steam for P<1000. 'If_the steam, on the other
_ hand, contains more than 15% moisture (x<.85) entrained liquid will control I
s iodine release.
For P>10 even minute amounts'(x,. 99) of entrained liquid contribute significantly to iodine release.
Presently P is not sufficiently I
known to determine which mode of iodine predominates.
i i
Following'a steam generator tube rupture' accident, the: aqueous / gaseous iodine
{
partitioning will be determined by iodine hydrolysis and the effect of radia-l tion.
Both hydrolysis and radiolysis'are, in turn, influenced by system pH, temperature,. concentration, and impurities.
The overall reaction for iodine
~
hydrolysis may be written as 12-t.
n,
.-----+,.a.-
l l
2 + 3H O = SI + 10
+ 6H+.
~
3I 2
3 Rapid iodine hydrolysis is favored by high pH.
Styrikovich et al., Reference 7,
.have published the results of studies of the partitioning of iodine between steam and water.
The measurements were made on saturated steam and pressures of 0.19, 0.4, and 1 MP.
Their results show that I2 volatility decreases with 3
increase in pH.
The partition coefficients measured in these studies with an initial form of 12 varied from 17.0 to 25000 depending on the pH (6-10) temper-ature (118 C, 143 C, 179 C), and time.
Reference 8 indicates that several hours may be required before the partition coefficient reaches a steady state value.
Since the SGTR shut down event takes much less time, the use of equili-
~
brium partition coefficient as published in the literature may not be justified.
In addition, radiation effects have the potential to override hydrolysis in 4
controlling iodine volatility (Reference 9).
Radiation can change the chemis-try of the system by producing a host of products such as OH, H 0, H, and 22
~
e,q that can react with iodine species.
At low pH (<N7), the hydroxyl radical, OH, or hydrogen peroxide, H 0, can react with I-to form 1.
Conversely, at 22 2
high pH tne hydrogen peroxide can convert I2 to I_
It should M noted that-iodine concentration in the primary coolant can vary by two orders of magnitude following power-changes (Iodine spike) and, therefore, one can not be certain that equilibrium iodine concentrations have been reached during the SGTR event.
Coolant impurities,.pH, and radioactivity levels are the main variables to be investigated.
i 7.
THERMAL-HYDRAULIC The thermal-hydraulic history of.the SGTR event provides the boundary condi-tions and flowrates req,uired to solve Equations 1-3.
Figure 3 shows typical mass' flows, temperatures and pressures during an SGTR event calculated by TRAC.
These boundary conditions strongly depend on the accident scenario as can be seen from Figures 3 and 4 which make comparison of two different accidents.
In the base case the temperature of the secondary coolant is higher than the 13
. ~.-
temperature of the primary coolant and consequently no flashing occurs when the relief valve is opened.
In the second case (safety valve stuck open),
the primary temperature exceeds the secondary liquid temperature and-flashing does' occur. These two different accident scenarios could result in different
-radiological consequences if-flashing is the main source of radioactivity ~
release as currently calculated by the CITADEL code.
On the other hand, if flashing is not the predominant mechanism for iodine release, less pronounced dependence on accident scenarios is expected.
- PROGRAM DESCRIPTION' The variables governing h, p, rl, E, P together with the expected occurrence of a given SGTR accident provide the required. input for the implementation of the overall program objectives:
(1) obtain the required dat.s base to assess the ability'of CITADEL to estimate offsite radioactivity relaases following various,
postulated SGTR events; and (2) optimize the mix between full size prototypic tests and small scale laboratory tests to minimize cost.
The scope of the present program is limited to the W type steam generator.
The partition coefficient P is equally applicable to once through steam generators however, the mixing parameter p and the surface deposition coefficient h most likely are different for the two' designs.
A separate study will be required to determine p and h for the once-through design.
Table 2' outlines how the experimental work is divided between the three program participants. Westinghouse's main responsibility.is to conduct tests using their prototypic steam generator facility MB-2.
The data will be used to check correlations'which were obtained in laboratory tests and to provide data which cannot be obtained in small scale tests.
Oak Ridge National Laboratory (ORNL)
_ responsibility is to elucidate the chemistry of radioactive iodine under simulated SGTR environments.
Northwestern University (NW) responsibility is to generate correlations for liquid carryover for a wide range of conditions.
Los Alamos-National Laboratory (LANL) provides analytical support to the above organizations,'with current emphasis on droplet formations-and breakup.
Battelle Columbus Laboratory (BCL)'is responsible for updating the CITADEL code 14 o
l l
j as the data becomes available.
A brief description of the three facilities is presented below.
Westinghouse - MB-2 Facility The O Model Boiler No. 2 (MB-2) facility is approximately 1% power scaled model of the Model F steam generator. The MB-2 is geometrically and thermal-hydraulically similar to the Model F; however, since its total volume is scaled to correspond to its power a full size tube rupture cannot be simulated by one to one scaling because this will overfill the unit in a nonprototypic manner.
Table 3 presents a comparison between MB-2 and SG-F with respect to the number of tubes, tube specifications, volumes and unit heights and separator specifi-cation.
Figure 5 presents a schematic of secondary and primary loop instrument-ation.
The MB-2 presently is equipped to simulate SGTRs at two elevations only; one is near the tube sheet elevation, and the other is at the U-bend elevation.
The primary liquid injection is controlled by a removable tube which penetrates the shell. A calibrated ventury tube, differential pressure transmitter, system pressure transmitter and fluid thermocouples are used to measure injec-tion rates.
106 thermocouples are positioned at various axial and radial
~
locations in the bundle to measure the primary and secondary liquid tempera-l tures and the outer wall tube temperature.
The secondary side includes 36 pressure taps of which 9 are located in the tube bundle region, 12 in the downcomer pipes, and 15 in the upper shell region.
Steam pressure measurements are made at three different locations in the outlet piping.
Primary, feedwater and steam flow rates are all measured with a calibrated orifice and narrow, middle and wicle range differential pressure transducers.
A collection tank is used to weigh and analyze any carryover liquid.
Different nonvolatile traces for the primary and secondary fluids are used to characterize primary / secondary fluid mixing as discussed above.
ORNL - Facility The apparatus shown schematically in Figure 6 is located in a hot cell in the ORNL radiochemical complex.
This apparatus consists of (a) 3-%" diameter vessel,-5' long; (b) tube, bundle and separators; (c) relief valve; and (d) a condenser.
Radioactive iodine will be injected into the vessel which I
15
contains saturated water at the 1000 to 1200 psi range.
The residence time of iodine in the liquid will be varied by means of the relief values, thus allow-ing the measurements of the partition coefficient as a function of time.
I-131 and Ba-140 are used to measure iodine deposition and liqaid carryover.
Procedure for measurements and valve activation will be developed.
Northwestern University Fig ~ure 7 is a schematic of the NWU facility which in main is very similar to the ORNL facility but unlike ORNL it employs nonradioactive iodine (sodium-Iodine) and operates at relatively low pressures.
This facilitates the attain-ment of a large number of data points required for entrainmen: (E) and mixing (p) correlations.
Briefly, the NWU facility consists of a 10 in., 6-ft long vessel containing approximately 40 3/4-in. tubes.
The primary liquid pressures, up to 20 bars, can be injected into this vessel after the desired steam flow conditions have been established.
Traces in the primary and secondary liquids are used to determine the respective amounts of liquid in the carryover.
The vessel is also equipped with viewports to allow photographs c.f the flashing jet.
Resistivity probes at various elevations allow measurenents of local i
void fractions.
The validity of the correlations will be tested against the prototypic MB-2 data.
Battelle's Columbus Laboratories (BCL) i The CITADEL code was developed by BCL for the NRC to analyze the transport and j
deposition behavior of iodine in the secondary loop of a PWF. during SGTR acci-dents.(5,6) The previous version of the code did not include the iodine partitioning with steam from the secondary water, and had some numerical problems.
BCL is currently modifying the code to include the iodine partition-t ing and to correct the numerical problems.
Following the code modifications BCL will use the code to asse.d the radio-activity releases to the environment from the three vendors steam generators; Westir.ghouse, Combustion Engineering and Babcock and Wilco)., for four accident sequences, namely, (1) SGTR with loss of offsite power, (2) SGTR with a PORV stuck open for 20 minutes after the PORV opens initially, (3) SGTR with an ARV l
16
stuck open for 20 minutes after the ARV opens initially, and (4) SGTR with ARV stuck open for accident duration.
BCL will also keep abreast of the experi-mental data as it comes in to compare the correlations in the code with the data and to see to what extent the code should be modified.
The resulting code could then be used by itself or as a subroutine for dose calculations in a thermal-hydraulic systems code.
SUMMARY
The relatively high probability of SGTR accidents dictates that a better understanding of the mechanism of radioactivity release is required.
Determin-ation of Technical Specificatiori limits for radioactivity levels in the primary and secondary requires reliable calculations on radioactivity transport for steam generator tube rupture events.
A program aimed at the generation of the required data to test and modify if required the CITADEL code has been defined and is now being implemented.
The experimental results will also be used to program special iodine transport subroutines for thermal-hydraulic system codes such as TRAC or RELAP.
t i
17
REFERENCES 1.
C.Y. Cheng " Steam Generator Tube Experience," NUREG-0886, February 1982.
2.
Safety Evaluation Report Related tc the Restart of R.E. Ginna Nuclear Power Plant, NUREG-0916. May 1982.
3.
A.K. Postma and P.S. Tam, " Iodine Behavior in a PWR Cooling System Follow-g ogated Steam Generator Tube Rupture Accident," NUREG-0409, 4.
A.K. Postma, G.M. Hesson and D.W. Faletti, " Iodine Transport Predicted i
for a Postulated Steam Line Break With Concurrent Ruptures of Steam Generator Tubes," NUREG/CR-2659, December 1982.
5.
S. Raghuram et al., " Iodine Behavior in Steam Generator Tube Rupture Accidents," NUREG/CR-2683, April 1982.
6.
S. Raghuram et al., "CITATEL:
A Computer Code for the Analysis of Iodine Behavior in Steam Generator Tube Rupture Accidents," NUREG/CR-2682, April 1982.
7.
M.A. Styrikovich, 0.I. Martynova, K. Ya. Katkovskaya, I. Ya. Dubrovskii, and I.N. Smirnova, " Transfer of Iodine frcm Aqueous Solutions to Saturated Vapor," Sov. At. Energy 17(1), 7.35 (1964).
8.
J.T. Bell et al., " Predicted Rates of Formation of Iodine Hydrolysis j
Species at pH Levels, Concentrations and Temperat Anticipated in LWR l
Accidents, NUREG-2900, November 1982.
9.
" Iodine Volatility" E.C. Beahm.and W.E. Shackly " Proceeding of the Fission Product Behavior and Source Tern Research." To be published.
10.
C.A. Pelletier et al., " Sources of Radioiodine at Pressurized Water Reactors," EPRI, NP-939, Nov. 1978.
11.
J.W. Handler et al. "In-Plant Source Term Measurements at Turkey Point Station - Units 3 and 4," NUREG/CR-1629, August 1980.
18
Nomenclature A - surface area c - iodine concentration per unit mass E - entrainment h - mass transfer coefficient defined by equation 8 m - mass flow rate M - total mass within a given volume P
partition coefficient, EQ 7 S - iodine removed from fluid due to deposition t - time 1-x - liquid mass fraction upon flashing Subscripts t
d - downcomer f - feed water g
gas phase H - volume bounded by separator and liquid pool l
o - exit from steam generator 1 - liquid phase 19
.t p
primary i
s - entering separator Greek p - mixing parameter, EQ 5 c - entering downconer
{
q - separator efficiency, EQ 6 p - density i
f l
h 4
-l J
i 4
i 20 4
4
._._..,,_2:
.. ~.
Table 1 OPERATING EXPERIENCE DURING SGTL FOR i
SELECTED PLANTS Steam Iodine 131 Concentrations, Leak Flow pCi/cc M.S./
Wate GMS/
Reactor 5.G.
Main S.G.
Plant Date GPM CC(5) Coolant Liquid Steam Liq.
Source Paint Bch.
2-20-76 0.76 3.75
.007 1.39 x 10 4 3.78 x 10 8 10 s Reference 10 Turkey Point 11-09-77
.05-4.00
.06 1.8 x 10 4 2.9 x 10 8 10 2 Unit 3C 11-17-77.13 4.00
.014 4.3 x 10.s 3.6 x 10 7 10 2 Reference 11 1
Unit 4A 1-18-78 4.00
.007 2.7 x 10 8 5.6 x 10 s 10 2 1-20-78
.01-4.00
.007 3.9 x 10.s 1.0 x 10 7 10 2 l
1-24-78.33 4.00
.007 1.3 x 10 5 2.1 x 10 7 10 2 1-26-78 4.00
.047 7.4 x 10 4 1.1 x 10 5 10 2 7.0 x 10 7 Pvt.
Ft. Calhoun 2-18-84
.001 Communic tion
.014 2.5 x 10 4 6.3 x 10 8 10 4 Pvt.
San Onofre 6-14-84
.1 Communication N.A.
3.5 x 10.s N.A.
6-19-84
.2 4
I 21
8 Table 2 PARTICIPANTS / RESPONSIBILITIES Main feature of Parameters to Variables Begin/
ORGANIZATION work be characterized Complete Prototypic (very q, E Leak Loca-Aug 84 W
glose), Geometry of tion May 85 W steam generator Irradiated Iodine, P,h,E pH; (4 to 10) Sept 84 Prototypic, pres-1-131; (10 5 -
ORNL sures, temperatures, 10 10 molar)
Sept 86 coolant chemistry Ba-140 and pH.
Prototypic, steam E,6 Leak Loca-Sept 84 and break tion and size, Sept 86 NWU velocities steam veloc-ities, depres-surization rates.
LANL Analytical charac-Sept 83 terization of Jets Sept 84 and sprays BCL CITADEL code April 84 modification, Oct 84 licensing calculations
~
22
Table 3 PRIMARY AND SECONDARY BUNDLE SCALING SG-F Ms-2 MB'12/SG-F Number of U-Tubes 5646 52 0.00921 4
Tube Specifications 0.D. cm (in.)
1.75 (0.688) 1.75 (0.688) 1.0 I.D. cm (in.)
1.54 (0.608) 1.54 (0.608) 1.0 2.49 (0.98) 2.49 (0.98)
- 1. 0 Pitchcm(in.})
Volume m3 (ft 18.41 (650.)
0.139 (4.90) 0.00755 Secondary Bundle Volume m3 (fts) 44.88 (1585) 0.283 (10.0) 0.0063 23
SAFETY VALVES RELIEF VALVE FEED TO H.P.
TURBINE g
~ ~ " ' ~ ~ "
' ' ' ~ "
- l STEAM 7
DOME TO STEAM JET DOWNCOMER b
AIR EJECTORS y
t g __, q
< SEPARATORS g
o (S)
HEATED AND RISER SECTIONS BREAK (2-PHASE REGION)
/
m
/
Q-SUBCOOLED LIQUID POOL 4
4 (1-PHASE REGION)
I i
PRIMARY I
COOLANT
@ lodine entering secondary side (mc)P from broken tube (2)todine eness in volume bounded (Mc)H by separators and liquid pool
@ heated and riser sections lodine deposited on walls in (S)H
@ lodine entering separators (mc),
@ lodine deposited on separator (S),
walls
@ lodine entering steam dome (mc),
( ) lodine from separators entering (mc)g downcomer
@ lodine from feedwater entering (mc)g downcomer
@in liquid pool lodine mass in downcomer and (Mc)d h lodine deposited on downcomer (S)d and liquid pool surfaces
@ lodine from downcomer entering (mc)d 2-phase region Figure 1 Overalliodine mass balance.
25 i
.l
1.0 l
0.9 3
CONSTANT LIQUID 0.8 CARRYOVER, 0.7 u.
0.6 E
m E
O.5 10 1.0 0.1 0.01 u.
O Z
O 0.4 E<
E I
0.3 0.2 SATURATED STEAM 0.1 AT 285'C AND 1000 psia LODINE CONCENTRATION:
10-10 MOLAR I
I I
0 2
3 8
s 7
1.0 10 10 10 10" 10 10 10 PARTITION COEFFICIENT, P i
Figure 2 Fraction of lodine in steam, F, as a function of the partition coefficient, P.
26
ob5 39u.
E 002 l
~
yro ta ro
)
N ia ba s
L Ep R
l N
a 5
U1 no S1 R
S5 it a
E1 T
N R1 G
N 1
P(
0 S
s
\\
N i
05 g
I o
E n
1 i
m R
w l
YU a
A RT o
M_S l
)
l
\\
s C
o o
DR f
E L
NE s
A S
CP e
d, CM
(
t W
M a
\\
a t
EE r
O A
w s
ST lo L
R F
o B
C l
\\
f S
S s
W A
R sa O
A_
J M
T m
C d
e A
n c
E a
N i
0 s
r 0 R u
^
J 0 R e
o I
S r
1 E
u T
t E
F a
A r
R e
U E
p M_
I e
g T
M m
Y A T
t R R A
E e
TP MEM ru I L s
R NE s
PIT erP 3
I W
ig n
e i
r u
F 00 i
5 A
~
_0 03 a
0 E
25
_ ! 5 K 3 < = 's s sO3 or~
Il.
,',i lj llI I,
i I
I I
l j
4 Source: J. W. Bolstad. Los Alamos National Laboratory l
550 PRIMARY INLET TEMPERATURE E
N C
N SECONDARY TEMPERATURE g
500 N
g%
N e
E N
O w
w t-N m
450 o
200 g - 1500 M
5 e
1 9
N B
l#
PRESSURE N
3-E u.
a I
m 100 m -
500 400 4
cc E
n.
MASS FLOW 0
l l
I i
350 300 500 1000 1500 TIME AFTER REACTOR SCRAM (SEC) i Figure 4 Pressure, temperaturcs and mass flow rates following SGTR with a stuck open safety valve.
I t
STEAM FLOW CONTROL VALVE FW HEATER
^
--Q T-2201.DP.-220 P-255- ---9.--k,
g
- + - - -
(- P-211 I
g g
STEAM FLOW fop.
I I I
(NAR., WIDE) i I l FW FLOW (DP-201A)
I P-91 g.
T VENTURI (NAR., MID, WIDE) 3 AMBIENT
[
TEMPERATURES DP-41C l g
S AM f-----------------"
FLOW T-603 e MB-2
^l pDP-299 CONDENSER
-&j T-299 AUX FW P-2101
- --+-I o P-299 SURGE TANK V
+
T-2101 o T-290 T
T-602 e p
g FW PUMP l
ll T-601 e t_.
. __j,_
+
' D P-150 DEMINERAllZER SUBCOOLER SGTR INJECTION SYSTEM h
P-12 RTD T-1150 (NARROW)
COLD HOT T-1151 (WIDE)
RTD T-1250 NARROW
^*
BYPASS V
LINE A
PRESSURIZER LOW PRIM.
FLOW @
PRIMARY D P-111 PUMP PRIMARY FLOW (DP-109)
HEATERS T - TEMPERATU RE ANNUBAR (NAR., MID, WIDE)
P - PRESSURE DP - DIFFERENTIAL PRESSURE
~
Figure 5 MB-2 primary and secondary loop instrumentation.
AIR OPERATED LlOUID RELIEF VALVE, ADDITION 1200 - 1000 psi ARGON VENT ORIFICE 17
-T Y
L J
L J
ONE LITER CONDENSER, 3.0 in. ID, 9 in. L.
9.5 LITER
[ " AUTOCLAVE, s
3.5 in. ID, 60 in. L.
SAMPLE THREE-ZONE CONTROLLED FURNACE SAMPLE t
Figure 6 Schematk of ORNL experiments.
t 30 t
v
t EXHAUST l
i i
) (VENTURI SOLUTION n s M AK E-UP e
em q'
CONDENSER
[)l N%
2
/
~
SIGHT VIEW LEVEL GLASS PORT GAUGE r%
r- - - - - - 4 1!
l v
I COLLECTION I
l TANK g
I bl ic T
RESISTIVITY j'
" PROBES
- 4
-_ ql I
L _ _ _ _ _ -ai SPECTRO i
STEA PHOTOMETER N
HEATER C g
K DRAIN i
^
SAMPLE J
Figure 7 Schematic of Northwestern University experiments.
i,
I l
APPENDIX i
l l
33
- rm -
(
b "L,o* I ~*1; s "g,o**g,s**g m =nm c <.,, <<,,
o 1,,i <,
g g,
SEPARATORS i
s y,,i b
Separator /
d Secondary
- i,s Boundary B
d o
m
=m m1,H c,H 9,s(1-x) pmp Ty (1-x)m
((1-x) (1-p]m
^
p p
m M =M 4M g
g g j
t, j
xm Secondary Volume H p
p y Primary / Secondary Boundary a
- d Schematic of Mass Balance on Control Volume H and Sepaiator l
l m
= (1-x)m primary liquid f rom the break entering secondary side
=
p p
which contains a mixture M of steam M and water M.
g g
t f
(xm)p mass of primary liquid which flushes and mixes with the
=
steam / water mixture.
(1-x) Em mass of primary liquid which reaches the' separator by short
=
p circuiting mixing.
[(1 x) (1-p]m mass of primary liquid which undergoes mixing.
p a
35 1
p=
total steam generation in control volume due to evapora-u tion and flushing from primary fluid.
).
m
=
total water flow in the secondary side as dictated by mass and 1,H heat balance considerations.
i j
2.
Comments on Equations 9 and 10 Equation 9 4
J l
If the iodine concentrations in the primary steam and liquid immediately upon flushing p denoted'as c',p andcj,p then
]
1.
(ac),=(xac'),+[(1-x)mcj]p j
2.
c=xcf+(1-x)cj=[xp c'+(1-x)]cj p
p p
i I
i
= [x
, (1-x )]cj p
p l
In reference 3 Postma pointed out that since.the time required to flash vapor from superheated liquid is very short liquid phase diffusion may be neglected.
Under these conditions the interface concentration during flashing Pt l
builds up to a value which corresponds to 1 i
EPg i
thus from equation 2 above' p
f,p =cj,p c
=c The use of c in equation 9 implies that the concentration of iodine in p
the liquid was not alt.ered due to flasning.
l i
9 i
{
36 t
- y
Equation 10 (ac)o = [(ac)g + (ac)g]o 1.
utilizing the definition of P=cg pg c p g g we get:
2.
(ac)o = [m
+ m ]o g,o e
g g
The liquid exiting the separator and carried by the steam is assumed to have the same iodine concentration as the liquid which flows into the downconer.
- Thus, 3.
cg,o = cg,g = cc 4.
(mc)o = [m
+ m ]oc g
g g
i I
i s
5
(
f I
37 Y
y
/
.. C., o....
.5m.....>...s..-,,cs
...c.s...........
C-.
-~
ses BIBLIOGRAPHIC DATA SHEET NUREG-1108 j
2 t
,,,,..s3,...,<,
. m.. s15 v ~...
Radioactivity Tr sport following Steam Generator Tube Rupture c
De mber 1984
..,_e.,.
f..,0.,,,m.o
, o.,
7 March l....
Joram Hopenfeld 1985
...a.o o,~C,0 c. w.,,0,~. w.so.....sc x.u. um i, C,
Division of Accident Evalu. f on Office of Nuclear Regulator Research
'o''''a U.S. Nuclear Regulatory Com sion Washington, DC 20555 11 5 0%50ma%4 CaGa%'la f 0% Navf AND Y A'Li%s ADOat SS sea.
la C es 6
12e 7**t 05 REPOMT Same as 8, above.
echnical
,,,,,i,OoCo..
ou~,,,..
Oct 1,1983-Oct 1,1984 13 5LPPL IYt %T AM V NQf f 5
.J s,
C, am n.,
g A review of the capabilities of the CITADEL omputer code as well as plant experience to project radioactivity releases follo/ng
- eam generator tube rupture in PWR's shows that certain experimental data is need@ for'; liable offsite dose predictions. This article defines five' parameter %hich aps the ke for such predictions and discusses the functional dependence of these paramgteis on vagious operational variables. A joint Westinghouse, Electric Power Researqh Institute,hnd the Nuclear Regulatory Commission program aimed at obtaining the fiv parameters emkirically is described. Present status of the CITADEL code is also reviewed.
(
l ji
,Se at v WOAD 5 A%o DOCuv.4T ANAL v5s1 l:.
15e DE SCm iP T O A 5 Steam Generators il Thermal Hydraulics f
CITADEL Computer Codel 4
16 AvestasiLITD 5'AftYtNT
'? SE CUR'Iv CL A55
- sca f *0N tu Nuveta OF eaGES Unlimited cT5ssi fied 19 SE CURif y CLat318 eC ATION JQ PRt(g "0n'"5Tassi fied s
UNITED STATES counts etass unit MUCLE!AR REGULATORY COMMISSION WASHINGTON, D.C. 2X55
'8'^y,*j"".^*
wagsoc OFFICIAL BU!ilNESS PENALTY FOR PFllVATE USE,6300 5
E E
I l' 05 % Jidd 7 7 1 1A?;1dl 5
PU)
T aP-POR NUPT G cail.uam x
20sss
?
2 s
5
_ _ _ _ _ _ _.