Document G161EmVD886j5k1baKObMByv

Page 2 SAoc/cP A*** Aoi*J* Examination of the agitator alarm silencing switch shoved the switch will remain in the silence^position. which it should not do,"allowing the agitator to atop without sounding the alarm. Also, the motor start-stop switch would operate with very slight pressure stopping the agitator"'la person*s clothing brushing against the switch could conceivably stop the agitator). ^J 3 a 3. The separated spool piece was constructed of 3 inch Speedline rolled flanges rated to withstand 1900-2300 psi hydrostatic pressure. The use of this type of flange was discontinued about 1965. Upon discontinuing their use all existing flanges were eitherReplaced or the insert was backwelded to the pipe as the result of a report that such assemblies had been known to separate. Poly No. 1, installed in I960, was originally equipped with this type of flange and it was backwelded. The assembly did not separate at the point where the pipe is rolled into the insert and was backwelded. Instead, the flange slipped over the lip at the end of the insert (see attached sketch). It appears moderate corrosion of the inside diameter of the flange and possibly wear on the lip of the insert allowed them to separate when subjected to the initial shock of the/ discharge. Both rupture discs in the vent line fully opened and there was no evidence of obstruction. Conclusions .... Poly No. 1 vented for an unknown reason. The possibility that the agitator stopped about the time the charge reached reacilon temperature is suspected. but not proven. The vent stack separated from the spool piece because the Initial shock of the discharge. forced the worn flange off of the serrated insert. Recommendations 0 (& Repair Poly No. 1 agitator alarm switch and agitator on/off switch. (Completed) Replace the vent nozzle spool piece with one properly constructed. (Completed) Inspect all poly nozzle and piping systems for such other rolled flangeassemblies and replace. (Completed and none found.) Inspect all poly nozzle systems for other apparent weak points and correct. (A preliminary inspection has been conducted with a few minor repairs made. A two week schedule has been initiated to thoroughly examine each poly nozzle and flange assemblies to ensure they are safe.) Evacuation Experience While the emergency was handled very'Veil by personnel on duty at the `time of the accident, some problems were experienced. In securing self-contained breathing apparatus one operator put on the training unit, which was inadequately identified, and had an insufficient air supply. his unit has since been un- mistakenly identified. A second operator stated he could not obtain sufficient air through the demand regulator', of another unit. He found it necessary to u^e NOC 15700 Page 3 the by-pas3 shortening the air supply life considerably. This unit has been checked and found to be operating properly. In^atterapt to avoid this type of problem,future inspections of the equipment will include actual operation ofthe regulator. The evacuation alarm was sounded, but there was some initial confusion about the nature of the alarm. The noise of the venting poly in addition to the area emergency whistle which had been activated made the evacuation code difficult to understand. We are considering methods of providing an evacuation alarm that will leave no doubt as to its purpose. W. A. Riley attachments (2) cc: Gourp II . F* C. Diliey H. Waltemate - with Supervisor *s I 0. F. Be.ckmeyer J. L. Nelson - R. D. Scott E. E. Mitchell- P. D. Terry E. W. Harrington - P. H. Lawrence W. E. Brodine G. Pow . W. L. Cato / a c. A p A * A oa ! P e <? dP / yf cr ^ t*s*eP A ert- j , / c* a 4 ,rc . at c- A. Ot- C. U r # p ! a n't v 6 ' f eft A e c. k tA < t-e . a' / A * d? A/yt'cPe 4 y *' *$ <2 *** p a^ Pa c. A?: u ** ."L& , 3t shb'+tcP e A /`o* c/ uJc f o tf * Pa c. Pt WAR; ss + e >- <s y *-* eh. Pitt** e j-AotolJ 4e t .0 * * a / i '0 *1 l e y ai >'s y .* i-i e ^ / 'j ` a c> A A gt / V" n S , NGC IS701 LOW ft e-ACV\ pLPSNfT Poly rjo, t v&kjt sTftcfc? sePAg.>Tc30 ~ ili-bfiz. ) ) A 6 A -e. NGC 15702 LoN A &GACW .PLA*^ ~jOl_Y NO. I V&nT -STACK SEPARATISM - i/z^/tz. SPEEDUNE iPLANSE ^` NGC 15703