ML20080L877

Response Opposing Util Motion for Summary Disposition of Eddleman Contention 80 Re Mixing & Dispersion Models. Statement of Facts as to Which There Are Issues to Be Heard Encl

ML20080L877
Person / Time
Site:	Harris
Issue date:	09/27/1983
From:	Eddleman W EDDLEMAN, W.
To:	Atomic Safety and Licensing Board Panel
Shared Package
ML20080L879	List: ML20080L877 ML20080L884 ML20080L888 ML20080L894 ML20080L899 ML20080L907
References
82-468-01-OL, 82-468-1-OL, ISSUANCES-OL, NUDOCS 8310030367
Download: ML20080L877 (6)

Text

Sept 27, 1983 UNITED STATES OF AMERICA DOCKETED USNRC NUCLEAR BEGULATORY COMMISSION 33 SP 30 R2:15 l

BEFORE THE ATOMIC SAFETY AND LICENSING BOAED emcg 97 gu,e Glenn O. Bright 2:" W t' Dr. James H. Carpenter -'

James L. Kelley, Chairman In the Matter of I ) Dockets 50 400 OL CAROLINA POWER AND LIGHT CO. et al. ) 50 401 OL (Shearon Harris Nuclear Power Plant, )

L Units 1 and 2) ) p,SLBPNo.82-k68-01

) OL Wells Eddleman's Resnonse to Annlicants' Motion for Summary Disposition on Eddleman 80 (Mixing and Dispersion Models)

Applicants allege many" facts"which miss the noint on this contention. It says nothing (and helps Apolicants not a bit) that l their models are " state of the art" (which the contention doesn'ta l

mention). Compliance with N90 Reg. Guides, as Annelicants know, does not establish compliance with anulicable regulations.

Given these points, most of Applicants' " facts" are wholly irrelevant,e.g.2,3,4,5,14,15,16,17,19,1922 and 22.

As to the idea that reducing radionuclide concentrations l

in air reduce dose (" facts" 8,9 and 10), of course theyi do in RG 1.109, because its air dose does not include deposition. PG 1.109 at 4-5. But this doesn't address the actual doses at ground level, which are higher, because of higher nuclide concentrations at ground level due to rainout. (McFeatters affidavit at 26 in his paper:

50 to 70 percent of a radioactive cloud will rain out in a thunder storm. Revealingly, he cites no source for this number, though he cites many sources elsewhere in his paner. I guess it 's his opinion.)

8310030367 830927 PDR ADOCK 05000400 0 PDR

'iHosiver, even RG 1.109 (see at 6, item **) sometimes requires

~

use of a ground-level dose, which rainout would increase. .

- w 'w c;:

'But note, the NRC Staff (6-24-83 response to Eddleman interrogatories at 96-97) is still analyzing this issue.

Applicants anpear to define rainout somewhat differently than I do (see " Facts" 6 and 7). I was using the term " rainout" in Eddleman 80 to account for actual radionuclide downwash and deposition (including entrainment) with precipitation of all types, whether the radioactive plume is in a precipitating cloud or below it.

ThusIdeny" fact"7astoitsuseinEddlemanCod'xntion80 e My use of the word " rainout" in this contention and responses thereon has been conssistent, I believe, and Applicants have not used my definition of " rainout" which incorporates what they term rainout and " washout".

Nonetheless this is not a very significant point. The down-washed nuclides, from " washout" or " rainout" anpear to disanpear from RG 1.109 calculations (I can find no term that adds the denosition to the ground concentration or near-ground concentration, though I find several that remove deposition from the plume). Moreover,

) as stated in BNL 3391, rainout can increase concentrations in a radioactive plume to I several orders of magnitude above those expected in a ground-level release with Gmaussian dispersion, which Applicants claim is aanmaarara conservative. BNL 3391 at 16; See BNL 3391 at 6, item 3 (a cold cloud at ground level) form comparison with the language at 16 that rainout would keen the ground concentrations from a cloud aloft the same as if the cloud was on the ground (condition 3 of page 6) or un to several orders of magnitude higher. BNL 3391 uses a continuous noint source, see at 7.

, These are the very factors (ground level release, continuous source) that Applicants say are conservatisms (" facts" 20 and 21).

Yet BNL-3391 comes to an opnosite conclusion from " fact" 10:

rainout increases contamination above that expected with a ground-level release and continuous source, even in stable conditions.

See at 16, and Figures 10,11,12,13 (rainout) connared to figs 8 and 9 ground level cloud deposition without rainout.

Also, contrary to McFeatters ' affidavit, modeling doesn't always give ground level concentrations less than observed. In test EOCR-5 in NUREG/CR 1394 (see at page 8) a concentration 13 timens that modeled was observed.

Lestyou think BNL-3391 is outdated, I cite NU9EG/CR 2239 (December 1982) which says (for accident conditions) "early fatality l

and injury worst case results ... are caused by rafmout of radio-activity from the plume onto population centers (cities) ..."

l At p. 2-102, it says " Peak early fatalities ... are generally ,

caused by rainout . . ."

! Now, what is true for a highly radioactive accident clune is still true for a less-radioactive normal clume: rainout increases the peak dose. Whatever's in the plume, it washes most of it down.

(Cf McFeatters at 26).

The Lei affidavit, assertir.g ': hat breathing dose dominates all doses, is directly contradicted oy 113 of NFC translation 15 520, which Ak shows external doses of $1 mrem 1. ~~ound radiation, 1

31 mrem internal and external dose from noble gases, anu deeds or thousands of mrem doses from food chains.

l l

Applicants anpear to assert that buildings disnerse wakes

(" facts 31-13) more than Gaussian diffusion would.4 This is not necessarily so.

-n-Richard H. Thuillier of Pacific Gas & Electric, in en article in the Journal of the f.1r Pollution Control Association,1982, at p 526, notes that the Gaussian dispersion parameter adjustment for building .

wakes doesn't week well. Indeed, he finds wind speed to ik be the l

best predictor of ground level concentrations, and finds "'In general, popular models of dispersion in the less of structures nerform poorly with a distinct trend toward increasing under-prediction with increasing wind speed ... Site specific studies annear to be indispensible for fairly accurate or even conservative modeling p 526.

of dispersion in the lee of complex structures." Not only does this contradict Applicants' claims in " facts" 11-13 re Gaussian affirms plume modeling as it relates to structures, it that site-specific studies are needed for " fairly accurate or ...

conservative modeling".

Applicants concede (resconse to Interrogatory 8G-8(a) by me, their resnonse of 7-29-83 at 22, that no such studies

have been done at SENPP. This fact alone, added to Thuillier's scientific paper concluding that site-specific studies are needed for conservative modeling of dispersion in the lee of connlex structures, blows CP&L's motion out of the water. The issue of model accuracy for Harris is alive and well. Note that Thuillier based his article on studies of two nuclear nower olants, Millstone and Duane Arnold (ibid at $26, cony attached.

Also, as explained above, contrary to Annlicants ' areuments, motion at 8-10, Lei affidavit, both BNL-3391 and NUDEG-09 2239 conclude that rainout increases ground-level doses from radioactive plumes. BNL-3391 says it can increase them orders of magnitude over those from a ground-level release (which they, like CP&L, model as continuous with wind in one direction). (see at 16). NRC Translation $20, at 113, cony attached, directly contradicts Lei also.

For the above reasons, Applicants motion must be denied.

_5 In cddition, I h vo acquirsd the English version of the naner i published in Germany, Strontium 90 released in TMI venting, i

l ,

by J. Harvey, Picconi and Pisello, which states (p.2) that an NRC licensee underestinated ground-level concentrations of Kr-85 released by a factor of 4 million and released concentrations i ..

of that nuclide which exceeded the regulation limit (MPC) [often".

This is a powerful fact showing that NFC-annroved dispersion models

[

l can give quite inaccurate results and MPCs can be exceeded when such models say they won't be. This is further reason to deny Applicants' motion. '[7 Wells Eddleman Note: Auplicants refer to unsuonlemented "I don 't know"

( answers re RG 1.109 re Eddleman 80, but have intheir nossession extensive resnonses re that Reg. Guide's deficiencies on Eddleman 29, 37B, Joint II and so on. Consider these, niease, as undates to that past response. As noted above, mere comnliance wi th these Reg Guides doesn't heln Anplicants' motion anyway, as it j does nothing to establish connliance with annlicable rules.

l l

l l

l l

I l -__ - - _ .

STATEMENT OF FACTS AS TO WHICH THE9E ARE ISSUES TOBE HEARD RE EDDLEMAN CONTENTION 80 1, Rainout increases radionuclide concentrations of gases and

,i particulates on and near the ground, where they may be inhaled, give external dose to persons exposed, and be denosited not only on the ground, croos, gardens and forage, but on nersons also.

2. Dispersion in the wake of complex structures such as nuclear plants requires site-specific studies to give conservative modeling or fairly accurate modeling of radionuclide disnersion.

3. CP&L knows no models which model snow, ice storms, or hail Pemoving radionuclides (Response 7-29-83 to Interrogatory 80-4 at 19, items (b) (iii, iv, v).

, 4. CP&L has done no site specific studies of diseersion at Harris, same interrogatory resnonses at 232, to Interrogatory 80-8.

5 " Popular models" (including NRCs) of atmosnheric disnersion are inaccurate, and have a tendency to increase undernredict!on with increasing wind speeds.

6. CP&L models do not include the effects of rainout.

7. CP&L model (NRC model) includes a sector average concent for normal doses, averaging dose over a 22.5 degree sector.

Actually, radioactive clumes give much higher doses than this " average". See Pisello et al, Mobile Monitoring of Airborne Radioactive Effluent from the Oyster Creek Nuclear Generating Station.

8. NRC models can underestimate earticulate doses at ground level by a factor of 4 million, exceeding the MFC for such nuclides.

See Pisello, Picconi and (lead author) Joan Harvey, Strontium-90 released in TMI venting, at 2.

9. NRC Staff is still reviewing Eddleman 80 (6-2h-83 resnonse at 96-97)

10. NRC models are inaccurate.

11 Treatment of dry and wet denosition in NFC models does not establish that the deposition doses (1) are accurate or conservative -

(2) are properly included in doses to pecole.

12. Observed ground level concentrations of radionuclides which exceed the MPC were found by Harvey at al, (see 8 above) even though the NRC approved nodels of dispersion of such radionuclides indicated the MFC would not even be anoroached.

, [3, VOSf,lh N f{ btsfwStw NShf CtleY

.b q (t CCoh (n fag {y (Adw,> d f)ff,-dgui[z

\

Sv .S u,m tvtam dA2 f 05 d%v 9///23 ED &

l/N O IV'L Of(U y fv% C f Y&lwp p y V gtn)O k .

_ -