ML17261A013

Research Information Letter 0032: Improvements in the Aerosol Behavior Code for Radiological Assessments of Lmfbrs

ML17261A013
Person / Time
Issue date:	08/03/1978
From:	Levine S Office of Nuclear Regulatory Research
To:	Case E Office of Nuclear Reactor Regulation
References
RIL-0032
Download: ML17261A013 (8)

Text

f'.-

~ -. *e*

e UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D. C. 20556 AUG 3 1978 MEMORANDUM FOR:

Edson G. Case, Acting Director FROM:

REFERENCE:

Introduction Office of Nuclear Reactor Regulation Saul Levine, Director Office of Nuclear Regulatory Research RESEARCH INFORMATION LETTER #32 IMPROVEMENTS IN THE AEROSOL BEHAVIOR CODE FOR RADIOLOGICAL ASSESSMENTS OF LMFPRS J. A. Gieseke, et. al., "Characteristics of Agglomerates of Sodium Oxide Aerosol Particles,"

BMI-NUREG-1977, (August 1977)

This memorandum transmits the results of completed research on the measurement of sodium oxide aerosol properties. Sodium oxide is the key aerosol constituent in postulated severe LMFBR accidents. This work was completed as part of the Aerosol Measurements and Modeling program at*

Battelle Columbus Laboratories under the direction and sponsorship of the Advanced Reactor Safety Research, Office of RES.

The work consisted primarily of the experimental measurement of the effective density of sodium-oxide aerosols as a function of agglomerate size. The effective density is an important parameter in predicting how an aerosol population will behave in an enclosed containment in terms of natural removal processes. :For *the most severe postulated LMFBR accident scenarios (HCDA and core melt), sodium-oxide aerosol represents the highest airborne mass concentrations in the containment vessel and is expected to dominate and govern the behavior of the fuel and fission product aerosol. Therefore, as a first step in improving the a~rosol behavior code, HAARM-2, separate effects work was carried out on sodium-oxide aerosol.

The results of these separate effects measurements have been incorporated into the mode~s of the aerosol behavior code.HAARM-2, and together with some additional improvements used to generate a new version called HAARM-3. The improved models in HAARM-3 provide a more realistic description of particle characteristics and thereby allow improved estimates of sodium-oxide aerosol behavior during a postulated HCDA.

The HAARM code is used by NRR for* LMFBR site radiological consequence assessment.

Enclosed is a copy of the HAARM-3 users manual for operation of the code. The code has recently been placed on the Brookhaven computer and Dr. John Long of your staff is familiar with its operation. Release of the code was expedited to be available for NRR review of the FFTF-FSAR.

The results of the research used in improving t~e code have been reviewed at Research Review Group meetings with participation by NRR staff.

Edson G. Case

.:.2-Discussion An important aspect in performing LMFBR accident analyses for siting evaluations is the postulated release from the containment of radioactivity in the fonn of aerosol particles. The predictio 1n of aerosol behavior in containment depends on the microscopic characte~istics of the individual particles. Aerosol behavior processes of most interest to LMFBR accident analysis are agglomeration and growth of partic/les and settling.

Particle shapes and densities have a pronounced effect on the settling velocities of the individual aerosols. Also, tertain mechanisms for agglomeration are known to depend on the cross/ sectional areas of the individual particles, a parameter which is di1ectly related to particle shape and density. Because of the complexity of the processes controlling aerosol behavior, it is nearly impossible to *erive information on these aerosol characteristics from integral experim~nts. Therefore, separate effects experiments were perfonned to determi*~e the physical characteristics of sodium-oxide aerosols.

A Milliken-cell apparatus was chosen for performing measurements because thennal forces as well.as agglomerate physiqal properties could be determined.

In addition to the Milliken-thermal cell, agglomerate properties were further characterized' through electron microscopy.

The primary objective of the measurements was the determination of the sodium-oxide agglomerate effective density./ The density of the agglomerate differs from that of the actual material in that as the agglomerate grows there are voids or holes between the/particles and the density is less than that of the solid material. Ther,e measurements allow determination of the real radius, which in turn affects ~he collision area, an important parameter in some of the agglomeration processes.

As a part of the determination of the effective density, t~e aerosol primary particle size distribution was determined and the 0

~alue for the mass median particle radius (0.5 µm) used in the HAA~**-3 code was verified. Also another parameter, the first order slip correction factor, was determined (a value of 1.37) and this value was inctrporated into the code.

The primary particle size distribution of the sodium oxide was determined by electron microscopic techniques.

Th~ aerosol was generated by burning sodium in air. The results indicated that the primary particle size distribution may be modeled by a lbg normal distribution with a geometric mean diameter of 0.45 µm andJa geometric standard deviation of 1.47. Knowing the primary size distri ution one may relate the equivalent mass diameter of particles.comprising an agglomerate. The equ.ivalent mass diameter is the diameter of a sph 1~re having a mass equal to that of the agglomerate. Using the equation /or a log normal distribution and the equation relating equivalent mass diameter to the primary particle size distribution parameters (Ref. 1) one can derive an expression for

~....

Edson a (the density correction factor). This equation permits a direct comparison of a measured values and the theory of Kopet al. (Ref. 1).

The measurements indicated a reasonable agree~ent to the theory for an average primary particle density of 2.27 g/cm

• Evaluation and Application The characteristics of sodium oxide agglomerates, available from the measurements, have been incorporated into the computer code.

Also, prov~sions.were made to the code to account for the effect of nonspherical shapes of aggl_omerates on the mobility and on the collision cross section.

Calculation of the values of a allow the determination of a new radius instead of using a radius based upon the theoretical density.

The

• increased radius has the effect of enhancing the collision area for gravitational agglomeration and increasing the settling velocity.

Comparison calculations were made showin*g the differences between HAARM-2 and HAARM-3 predictions of airborne mass concentrations and leaked mass for various assumed initial aerosol releases in a typical LMFBR containment. The results of these comparisons are attached in Table 1 and Figures l and 2.

Using the HAARM-3 code with the improved models there is a reduction in the amount of leaked mass as a function of time (Figure 2). Also included, Figure 3 is a comparison of predictions of suspended aerosol mass concentration vs. time for each of the two codes with an integral aerosol test. This test, the first of a series being perfonned at the Containment Systems Test Facility (CSTF) at HEDL for DOE, represents the largest known aerosol test performed to date.

The CSTF vessel is about a 1/2 scale model of reactor containment, with respect to vessel height, the key geometry parameter in the agglomeration and settling of aerosols.

As can be seen in Figure 3, the HAARM-3 code provides much better agreement with the experimental results.

Future Work Currently work in this area is being directed towards the characterization of fuel (UO?) aerosols which includes the detennination of the aerodynamic properties of the particles. With the completion of this work, program emphasis will be placed on the interaction of mixed aerosols (sodium-oxide and fuel) primarily in large integral tests. The goal of this research is the verification of the HAARM-3 code.

A code verification plan is currently being developed to implement the accomplishment of that goal.

I *, :... *.

.. *.. *-~* -----~>---* ***"-.. ""_..,,...; _____.. _,._

J Edson G. Case

,_4_

Recommendations The HAARM-3 code will allow NRR to perform improved estimates of the depletion of sodium and fuel aerosol in the reactor building and other containment spaces for use 'in radiological assessments of postulated LMFBR accidents. It is recommended that NRR use the HAARM-3 code as opposed to earlier versions in performing aerosol calculations in view of the information presented in this RIL.

For further information on the application and use of HAARM-3 contact Dr. John Larkins in RES.

~

.. ~

Office of Nuclear Regulatory Research

Enclosure:

HAARM-3 Users Manual

.. -.. -**---*-**--*-----~

~ *-'--*.--

~... ~... -

LEAKED. MASS PREDICTED IY BAARM-2 AND HAARM-3 CODES Initial Concentration, iag/em3 2

10 100 Time 2 hrs 1 day 30 days 2 hrs 1 day 30 days 2 hrs 1 day 30 days Pertinent Input Data:,

Kass Leaked, g

{Accumulated)

HAARM-2 HAARM-3 15 15 68 58 76 65 72 65 255 110 269

• 113 480

. 158 865 160 870 161 Leak Rate - 0.1 percent per day Initial Rso

• o.s µm Initial OR

• 2.0 Cas Temperature - 310 X.

~..

... *--*- - ------- -. ___... _____ ;._*-*--------~~

e

~*~.,--TITTITI1---irT1TTmi---r-,-,-rrmr~,...-,...,-,rmn----r--.-'l""l""9._..*

10!1::-------------------..:.-..-----

c*

0 *-

4-e 4-c

~ 10°1=--

c 8

'"C

~

• B_ 10*.__ lnltiol R50= 0.5 p. m

~

lnitio I a-v = 8.0 Gas Temperature: 310 K

!). = 104 cm 16 1-- No source

. 101 101 Time, sec FIGURE 1 SUSPENDED AEROSOL MASS CONCENTRATION WITH TIME FOR THE CRBR CONTAINMENT GEOMETRY AT THREE INITIAL CONCENTRATION LEVELS 10 1

I.. l!lt._.

• ,.. l,**.'",

~100 c

E.

-c Cl>

~*

c.,.

..J e*

2hours

---

.HAARM 2

---*HAARM3 Leak Rote :0.1 % per day Initial R50= 0.5p.m Initial tTv = 8.0 Gas Temperature : 310 K No source 30 days l

l0~2;-----'--'u..u.J~!--L-J...J....LJ..uJJ.~J......J...W...U..U~....L.J....J....LJ..J.l.l.L-l~'LLl.J.llU 10 104 105 10~

107 Time, s.ec FIGURE 2

LEAKED MASS WITH TIME FOR THE CR!R CONTAINMENT GEOMETRY' AT THREE INITIAL CONCENTRATION LEVELS I i

,_.,, ____ *'~*- *-'-*-- -* --.

,,~ "'"l". ~

c

.2 e -

c

~

810°..-

---

HAARM-3

---

HAARM-2 "C

0

~ &

HAARM-3 (with reduced source s ~

• rate)

Q, 9, 0

Experimental Do ta -*

0 10 2

0

~--'--'~----"'-----4-----..1----__;L------'----~

IOO 200 300 400 500 600 700 Time After *Sodium Spi II, minutes FIGURE 3 COMPARISON OF PREDICTIONS AND EXPERIMENTAL RESULTS; AIRBORNE CONCENTRATIONS, CSTF TEST AB-1