ML20028A024

Provides Final Evaluation of Sampling Program,Per .Investigations Support Conclusion That Combustible Deuterium Concentration Never Existed in Steam Generator After Draindown of RCS on 820712

ML20028A024
Person / Time
Site:	Crane
Issue date:	11/08/1982
From:	Kanga B GENERAL PUBLIC UTILITIES CORP.
To:	Barrett L Office of Nuclear Reactor Regulation
References
4410-82-L-0042, 4410-82-L-42, NUDOCS 8211160278
Download: ML20028A024 (4)

Text

I

,~

d i

GPU Nuclear Corporation fJ Nuclear

= = s 48o 8

Middletown, Pennsylvania 17057 717 944-7621 TELEX 84-2386 Writer"s Direg1 Dial Npp1ber:

(O tri L b42

$n 5:

e TMI Program Office 3

h Attn: Mr. L. H. Barrett, Deputy Program Director Eis US Nuclear Regulatory Cannission l'd c/o Three Mile Island Nuclear Station G

2 Middletown, PA 17057-0191

Dear Sir:

Three Mile Island Nuclear Station, Unit 2 (TMI-2)

Operating License No. DPR-73 Docket No. 50-320 RCS Hydrogen Evolution This letter is in response to your letter NRC/IMI-82-060 dated September 27, 1982, on the above subject. The purpose of this letter is to provide the results of GPU's sanpling program to date, along with the final evaluation of the results.

Per the comnitment provided in GPU letter 4410-82-L-0025 dated October 8, 1982, the investigations into the H2 offgassing of RCS have been completed These investigations support the conclusion that a conbustible H2 concentration never existed in the steam generator or pressurizer high points after draindown of the RCS on July 12, 1982. Since this conclusion is supported by both the results of the CRDM gas sampling program and the results of the RCS high point vent sampling, further gas sampling of the RCS steam generator and pressurizer high points will not be conducted. The CRDM gas sampling program, however, will be continued.

Beginning on Septaber 3,1982, GPU has Men continuously monitoring and sampling the gas space in the CRIE's. From these sanples, the following characteristics have been seen:

1. Both N2 and H2 gas have been released from the RCS. On a volunetric basis, about twice as nuch N2 gas as H2 gas is being released. Essentially, og

_gg no 02 gas has been liberated.

o go

2. The gas generation, both H2 and N2 have been steadily decreasing and hence m8 the source of gas production is being depleted. Over a 30 day ling h

I d ()b, t

period the total gas generation rate has decreased by a factor o (fran gJ 3

3 0.05 ft / day to 0.01 ft / day).

oo The primary sources of H2 and N2 gas from the primary system are dissolved L

gases liberated from solution and deccuposition of hydrazine. Radiolytic

' $$n.

GPU Nuclear Corporation is a subsidiary of the General Public Utilities Corporation

.DMr. L. H. Barr:tt 4410-82-L-0042 decmposition of water has been shown to be an insignificant gas generation source since although the radioactive inventory in the syst e is essentially constant, the gas production is rapidly decreasing, indicating the major contributions to gas generation are dissolution of dissolved gases and the decmposition of hydrazine.

This conclusion is further supported by the lack of 02 gas generation. Since-the radioactive and liquid inventories are less in the steam generators that in the pressure vessel, hydrogen gas generation by radiolytical decmposition of water in the stean generator and pressurizer is also insignificant.

Prior to draindown of the RCS on July 12, 1982, each steam generator contained approximately 70,000 kg of water with a dissolved gas concentration of approximately 2 cc/kg H2 and 8 cc/kg N2 This gas concentration is close to a saturated condition at atmosphere pressure. The gas space in the steam generators after RCS draindown contained mostly N2 and essentially no H2 Because of this H2 gradient between the liquid and gas the 5 ft3 of H2 gas in solution would be expected to exchange with the gas space.

Prior to the RCS processing the Standby Pressure Control Systen, which provides the makeup water, contained close to 10 ppm Hydrazine. Aqueous Hydrazine even at very low concentrations is very suspectible to radiolysis and is experimentally observed to yield equal mole fractions of H2 and N2 gas (Reference 2). Using the pre-draindown steam generator water inventory of 70,000 kg and the RCS process water Hydrazine concentration, the total potential gas release from 3

radiolytic decouposition of Hydrazine is 17 ft. This is a conservative estimate of H2 production from Hydrazine hacama:

1. The CRDM sampling shows twice as nuch N2 as H2 being produced indicating reduction by Hydrazine is taking place.

2. During RCS processing the.RCS Hydrazine concentration was decreased by 1/3 by dilution.

i Cmbining the H2 source term from dissolution and Hydrazine dac-osition gives the maxirun H2 gas inventory which could have existed in the steam generators gas volume after draindown. This H2 gas inventory of 23 ft3 in each 1000 ft3 steam generator gas volune corresponds to a maximn H2 concentration of 2.3 volume %,

well below the combustibility limit of 4%. Since-the H2 gas release is not instantaneous and since the steam generator high points have been allowed to breath with the contairinent atmosphere, the expected H2 gas concentration in the stean generator would be less than the maxinom value of 2.3 V/o. A lower concentration is expected in the pressurizer high points than in the steam generator high points because, during RCS processinz,, one batch of H2 (Hydrazine) free water was processed through the pressurizer and tM inventory does not numrticate well with the rest of RCS.

These conclusions are rupported by the analysis of the high point gas sanples, which are included as Attachment 1.

These results indicate that based on the 02 concentration the high point gas volunes exchanged 60% of their volune by breathing. Using the measured 1.1% H2 concentration and the worst assurption of an instantaneous H2 release at draindown yields a maximum H2 concentration of 1.8% vs. the estimated worst case of 2.3%.

In surmary, GPU concludes that a cmbustible gas mixture did not exist in the RCS high points, nor does one exist now. It is not expected that one will exist in the future.

l

\\ _)

s Mr. L. H. Barrett 4410-82-L-0042 If you have any questions or desire further information, please contact Mr. D. R. Euchanan of my staff.

Sincerely, m

B. K. Kanga Director, IMI-2 BKK/RBS/jep Reference 1: letter from B. K. Kanga to L. H. Barrett dated October 8, 1982 (4410-82-M-0025)

Reference 2: Radiolysis of Aqueous Solutions of Hydrazine by H. A. Dewhurst and Milton Burton, Journal of American Chemical Society, Vol. LXXVII, October 12, 1955 Attachment CC: Dr. B. J. Snyder, Program Director - IMI Program Office

f C

{ *.. tM IDOP HIGH POINIS SAMPLING PROGRAM Sample 1 (%)

Sample 2 (%)

H2 02 N2 H2 02 N2 Ste m Generator A 0.1 20.9 78.7 0.7 9.4 89.3 Steam Generator B 0.3 17.6 82.7 1.1 8.9 88.9 Pressurizer 0.1 13.6 81.8 0.1 10.2 89.8 A conparison of the H2 and 02 concentration from sanple 1 and 2 and review of the video recording of the sanpling procedure indicate that for sanple 1, the 3/4" vent line to the gas spaces was partially purged. Sanple 2 results show that the 3/4" vent line had been purged for this saple as:

1.

During the purging procedure gas purging was evidenced by bubbling from the submerged underwater purge line. This bubbling was in evidence through-out the first purge.

2.

Subsequent purges are accmulative.

3.

During most of the second purging, the purge liners were withdrawn from the water. This greatly enhanced the purge rate.

4.

The pressurizer 02 concentration for sanples 1 and 2 are fairly close indicating this 3/4" line had been purged. As the same breathing occurred for each high point, the very good agreement between the H2 P us 02 l

concentration on the second sanple for all three high points (10.1,10.0, and 10.2) prove that the second sanple is acceptable.

e 5

6

_ - _. - -