ML19283B705
| ML19283B705 | |
| Person / Time | |
|---|---|
| Site: | West Valley Demonstration Project |
| Issue date: | 01/23/1979 |
| From: | Haughney C NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS) |
| To: | Rouse L NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS) |
| Shared Package | |
| ML19283B706 | List: |
| References | |
| NUDOCS 7903060630 | |
| Download: ML19283B705 (7) | |
Text
]. P D.
I t
. I 4
a UNITED STATES
["
NUCLEAR REGULATORY COMMisslON j
, j WASHINGTON, D. C. 20555
-Q JAN 2 31979 Docket No.50-201_
MEMORANDUM FOR: Leland C. Rouse, Acting Chief Fuel Reprocessing and Recycle Branch
)j FROM:
Charles J. Haughney Fuel Reprocessing and Recycle Branch LICENSEE:
Nuclear Fuel Services, Inc. (NFS)
FACILITY:
Western New York Nuclear Service Center, l
West Valley, N.Y.
J
'l
SUBJECT:
TRIP REPORT
-l i
Purpose:
To accompany NRC Region I Inspector on his unannounced inspection of.the NFS Reprocessing Plant. The purpose of the inspector's visit was to investigate the report
.]
by NFS of a defect in the 8D-2 high level waste pan.
-l
.i Place and Date: NFS Reprocessing Plant, West Valley, N.Y.
December 12 and 13,1978
Participants:
Please see attached list.
Discussion:
1
~
A.
Background
The NRC staff has an ongoing effort to provide added assurance regarding the safety of high level liquid waste storage at NFS West Valley.
In order to utilize the latest available technology and expertise available at the DOE contractor waste storage sites, the NRC requested DuPont-Savannah River Plant (SRP) to examine the safaty related information available on the NFS waste tanks. SRP examined this information and reported their findings to the NRC along with several recommerdations to obtain additional data to give increased confidence in the analysis of the safety of waste storage. On June 2, '1978, the NRC issued the SRP report entitled, " Safety Related Information Available on NFS Waste Tanks." The basic recommendations contained in the report are summarized below:
7903060(436 O
f
.j f
Leland C. Rouse 23M 1.
Inspect the wall of the tank of neutralized waste through the existing riser in the vault.
l 2.
Extend corrosion studies, including analysas of solutions in tanks and appropriate tests.
I 3.
Reevaluate assumptions on sequences of events if a waste tank leak occurs, including time avafiable and time required to empty tank.
Identify any actions necessary to make these times consistent.
4 1
4.
Study factors that influence release and movement of
~j waste into the ground surrounding the vault.
1 In order to begin to implement the third recommendation shown above, the staff wrote to NFS on June 27, 1978 and requested several i
evaluations, including an evaluation of the reliability of the waste tank pan liquid level alarm system. On October 12, 1978, j
NFS responded to our request providing several evaluations and describing some new testing that they would undertake including l)l testing of the pan liquid level detection systems.
Copies of these two letters are attached.
- 1 l
B.
Pan Level Detection System Testing Program
-i As can be seen from NFS's letter of October 12, the proposed c
j tests of the pan ifquid level alarm system included the actual addition of water to the pan accompanied by an observation of the performance of the level alarm system.
It was expected that this test would be rather simple and straightfon<ard and would prove the operability of the pan level alarm system.
j Nonetheless, as can be seen from the discussion below, this j
testing program proved to be difficult and complex. A copy 1
of an engineering drawing of the waste tank storage system is attached (CB&I Drawing VP-4413-8-D-12-ll-3).
The estimate of the amount of water necessary to produce a pan liquid level alam (1200 gallons) proved to be too small.
This initial estimate of the water necessary to cause a pan Tevel alarm failed to account for the amount of water absorbed
~
in the layer of perlite and gravel that is below the tank and above the pan. Water added to the pan can flow into the perlite and gravel layer through semi-circular holes in the annular skirt that surrounds the perlite and gravel (see Section G-G, Piece 15-1 on the CB&I drawing). Depending on the percentage of void spaces present in the perlite and pea gravel, it has been estimated that up to an additional 20,000 gallons of water would be necessary to cause a pan level alam.
-~
~
~. _
3-Leland C. Rouse Water was added to the pan via the pan sample tube (see attached drawing detail #1 from drawing number 4413-8-A-L-5). This water was added to the pan through a calibrated Badger water meter in order to keep track of total volume added. When the alarm had not been received following the expected amount of water addition, careful reexamination of the drawing showed the path l
for water to fill the void spaces in the perlite and pea gravel.
Additional amounts of water were then added.
- l
{
Even with these additional amounts of water, the pan liquid level detection instrument (LI-5) did not respond properly.
For this reason, a temporary water detection instrument was
}
installed down the 2" diameter pan sample tube that had been used to make the water additions. This temporary instrument showed a small but positive water level in the pan. Subsequent
]
installation of an additional level instrumentation probe j
through the mounting flange for the 8D-2 pan transfer pump was j
accomplished in order to better determine the amount of water in the pan.
Observation of these temporary instruments and their response following additions of water to the pan indicated that water levels would drop slowly following the water addition. This drop was felt to have been caused by capillary action in the a
perlite and gravel void spaces. Up to this time in the testing I
program, no indication had yet been observed on the vault liquid level instrument (LI-15). The vault level instrument, which measures the liquid level in the annular space between the outside of the pan and the inside of the vault, is completely separate from the pan level instrument.
On a routine night shift reading following a day of water addi-tions to the pan, an operator noticed a level reading on the vault's level instrument. This was the first positive reading observed on this instrument. A sample of water taken from the l
pan indicated less than 2x10 6 pCi/ml gross beta activity.
Comparison of the level reading in the vault and that on the j
temporary pan level instruments showed an approximately one foot difference in water heights between the pan and the vault. The vault instrument probe is mounted on a common flange with the steam supply line to the vault eductor. The mounting flange bolts to an above-ground riser in the tank fann.
Because of the earlier problems with the installed pan level instrument and a suspicion that the eductor could provide a syphon path, a decision was made to remove the vault level instru-mest and vault eductor for inspection and measarement. The subsequent removal and examination of the vault instrument
,e
. e
.,w*e e,
e S
,i l
Leland C. Rouse 2 0 I379 probe and vault eductor showed their lines to have been con-structed one foot shorter than required by the construction drawings. The vault instrument probe and vault eductor were reinstalled.
Testing was conducted to determine if water could j
transfer between the vault and the pan. The results of this testing clearly indicated a connection path between the vault and the pan. This testing also indicated that the location of f
the hole or leak or defect in the pan is probably less than 2".
pan height.
Testing of the spare waste tank (80-1) pan and pan and vault i
}
1evel instrumentation was then conducted. This testing indi-cated that the pan for 80-1 has no such leak as is present in 8D-2, but this testing also indicated that the vault instrument probe and vault eductor steam supply line for 8D-1 were built one foot shorter than required by plan.
i C.
Cause, Location and Size of the Defect in the Pan
]
At the time of this trip, the cause, location and exact size O
of this defect are not known; but active investigm.', to f
determine this information is underway.
If the defect is located under the bottom of the tank, the cause, location and size may never be known.
b While the cause of this defect is presently unknown, several possibilities are thought to exist and are under investigation.
These are:
]
1.
Possible construction error.
Please refer to the previously discussed CB&I drawing of the waste tank system and to the note in Section G-G of this drawing. The requirement of this note that the i
semi-circular holes in Piece 15-1 be field fabricated j
following tank stress relief is puzzling.
If these holes i
were built in accordance with this note, that would mean that the tank, pan and ski-t construction, including all welding, must have been completed prior to the fabrica-tion of these semi-circular holes. This construction sequence would have been necessary because stress relief of the tanks had to have been accomplished after all welds were finished. The amount of space available between the inner edge of the pan and the outside of the skirt (Piece 15-1) is very limited. This space limitation would be cumbersome for any construction worker who would have had the task of cutting some 225 semi-circular holes in the skirt.
5-Leland C. Rouse Examination of the construction records has failed to disclose whether a standing water test was ever conducted l
e for the pan. Thus, the possibility exists that in field j
cutting those semi-circular holes, a hole was also inadvertently cut in the bottom of the pan.
2.
Accelerated local corrosion.
j If a corrosive substance had been inadven'tently added to the pan at some time since its construction, then a defect could possibly have been created through stress corrosion cracking or some other type of local corrosion. At this time, no evidence to suggest that such an inadvertent addition of corrosive materials has ever taken place into the 80-2 pan.
3.
The floating incident.
I The floating incident in April 1965 could perhaps have j
induced stress in the pan that either caused or contri-
]
buted to this defect.
4.
Other possible causes may exist and will be pursued as the investigation of this defect proceeds.
D.
Intended Further Actions by NFS The actions NFS intends to take to further investigate this problem are listed below. These actions may be modified as the investigation proceeds:
1.
Make preparations and enter the 8D-1 tank-vault annulus to examine condition and as-built construction details.
2.
Continue examination of construction records to review the documented post-construction testing on the pan.
At this point, the only known documented testing of the pan is that of radiography of field welds.
3.
Investigate availability of remote inspection equip-ment that can be used to inspect the 8D-2 annulus.
e
0 7979-Leland C. Rouse 4.
Extend the piping for the 8D-1 and 2 vault eductor and levelometer probe to conform with the construc-tion drawing.
5.
Consider furthc? upgrading of the leak detection systems.
s s
Charles J. Haughney
(
Fuel Reprocessing and Recycle Branch Division of Fuel Cycle and Material Safety i
Enclosures:
'l
- 1. List of Participants
- 2. NRC ltr of 6/27/78 i
- 3. NFS ltr of 10/12/78
- 4. Engineering drawing -
Waste tank storage system a
0 I
=
LIST OF PARTICIPANTS j
A 1
- l NRC t'. Kinney, IE - R-I C. Haughney, FCRR NFS W. Oldham J. Duckworth A. Pierce E. Erlandson (12/13 only)
I c
I e
h t
1
@