ML20206P747

Offers Observations & Comments Re Util Results of Exams of Poison Insert Assemblies Removed from Spent Fuel Storage Racks. Util Should Submit Surveillance Plan W/Test Procedures to Determine Condition of Damaged Areas

ML20206P747
Person / Time
Site:	Point Beach
Issue date:	04/14/1987
From:	Tony Brown AFFILIATION NOT ASSIGNED
To:	Lear G NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM), Office of Nuclear Reactor Regulation
References
NUDOCS 8704210260
Download: ML20206P747 (3)

Text

,,

9 April 14, 1987 Document Control Desk U.S. Nuclear Pegulatory Cmmission Washington, D.C. 20555 Attention: Mr. George Icar, Director PWR Project Directorate #1

Reference:

Cmments on Dockets 50-266 and 50-301 Results of Examination of Poison Insert Assmblies Rmoved frm the Spent Fuel Storage Racks-Point Beach Nuclear Plant, Units 1 & 2. Sutmitted February 11, 1987 by Wisconsin Electric Power Cmpany

Dear Sir:

I have studied the Wisconsin Electric (WE) document referenced above in sme detail and would like to offer sxe observations and cmments for your consideration. These cmments are prmpted by my concern that a potentially serious probim has been discovered at the Point Beach Nuclear Plant (PBNP) and the investigation into the problem was superficial. I am also concerned that this specific case indicates a generic problem may exist which effects other fuel racks incorporating this particular neutron absorbing material.

My initial concern is with the limited scope of the follow-up investigation after the destruction of the poison surveillance sam;les. Only one full size poison panel which had been exposed to radiation was examined.

I believe this sample size is too small to provide data on the condition of the general population. I would suggest a larger number of samples be examined. These should be taken frm both high level and " average" radiation exposure regions.

WE came to the conclusion that the surveillance samples overstate the radiation damage suffered by the fuel rack panels after examining only one irradiated panel. Examination of one panel is not a sufficient basis for this conclusion since the failure mechanism is not well understood.

Secondly, Wisconsin Electric has assumed that the extent of the radiation damage is a function of the total radiation exposure. This is certainly an important factor, however it ignores the results of all the scientific investigations into environmental qualification test methods. The NRC sponsored research by Drs. Clough and Gillen at Sandia Labs indicates that dose rate, length of exposure, tmperature and availability of oxygen are also important factors. This research shows that the degradation in mechanical properties cannot be predicted solely by the total integrated radiation dose.

Since these synergistic factors were not taken into account, WE does not have a sound basis for extrapolating 20 years life for the average panel.

WE has recmmended the next sample be taken in five years. I believe that time period is too long as the rate of change of properties, or growth rate of the danaged area, has not been established. Based on the difference in exposures between the surveillance samples and the poison panel, and the difference in amount of damage between the two; the sampling period should be 8704210260 PDR 870414 ADOCK 05000266 PDR I \

t

.~ o Mr. George Icar 2 April 14, 1987 shorter. The data seems to indicate that once a threshold for radiation damage initiation is reached, the material deteriorates quickly. I feel it is unwise to assume the damage is progressing at a constant rate.

WE has concluded that the amount of damage observed on the full length poison panel is acceptable; however they give no technical basis for this.

They examined the panel visually and observed sare areas that appeared similar to sane of the surveillance coupons. I believe they should have taken measuranents at the damaged areas and canpared them to measurements taken at sound areas. They could have Ineasured neutron attenuation, thickness, specific gravity, hardness. Sane data is presented, however the significance of the data is not clear.

The report states that sane surveillance coupons decreased in thickness 30% and weight 50%, yet the neutron attenuation only changed frcm 99% to 96%.

Due to the nature of this material, a uniform mixture of boron carbide particles dispersed in a Inethylated polysiloxane binder, the loss of thickness indicates loss of boron carbide. The small change in attenuation for a relatively large change in thickness demonstrates the exponential shape of the attenuation vs. boron curve. Even though the change in neutron attenuation is an imprecise test, It will give an indication of gross boron loss.

In addition to the measurement of the physical properties of the damaged areas, mechanical properties should have been examined also. Since these areas were " powdering", mechanical tests might show whether these areas would spall or fracture during normal fuel handling. Fuel handling accident conditions should also be examined. Clearly, this visual examination and the few measuranents taken cannot be used as the basis for concluding the panel is currently serviceable and will ranain serviceable for 5 years.

In sumnary of my cannents on the investigation into possible fuel rack damage, Wisconsin Electric should:

Increase the sample size of poison panels examined to support their conclusions on the general condition of the fuel rack poison panels.

Decrease the sample interval to one year to attenpt to determine the rate of change of degradation. This i may lend credibility to their forecast of expected  ;

service life.  ;

i Subnit an adequate surveillance plan, with test procedures to determine the condition of the damaged areas.

Determine by analysis, the extent of damage (boron loss) the poison panels can withstand before the fuel rack fails to rneet technical specifications.

Mr. George Lear 3 April 14, 1987 The Wisconsin Electric report discusses the differences in design between surveillance coupon holder and the design of the poison inserts. The coupon ,

holder was well vented while the poison panels were sealed all around with

. vents only at the top. WE speculates that these differences in design may have contributed to the. differential damage. This. factor may have relevance in . evaluating the surveillance programs for other spent fuel rack installations.

The difference in the method of encapsulation affects the degree to which spent fuel pool water can infiltrate into the enclosure. The design of the sample holder allowed SFP water to circulate around the edges of the coupons.

'Ihe design of the poison inserts allowed SFP water to enter only at the top, and thus the water within the insert was stagnant. I suggest that, after radiation, this is one of the principle factors affecting the differential damage. Refreshed SFP water could make more oxygen available to the poison matrix and also may cause convective heatircJ. 'Ihe damage initiated at the edges of the poison and these were the areas with the greatest access to refreshed SFP water.

All fuel racks, designed to use Boraflex, allow for venting gases evolving frm the material. Sane of these designs are quite open and may allow SFP water to circulate around the faces as well as the edges of the i poison. These designs may have a greater potential for suffering radiation damage than the BPNP fuel racks, or other racks which limit infiltration of SFP water.

I believe the NRC should address the issue of whether a generic problen exists. BPNP sems to have received early warning that the fuel rack poison has less than the anticipated life. Other nuclear plants may have surveillance programs which understate actual damage to the fuel rack, rather than overstate it. These programs should be re-evaluated to determine their adequacy in light of the BPNP experience.

Sincerely, Thcmas B. Brown 1312 Kingwood Ypsilanti, MI 48195 (313) 484-3878

_ .