ML20043A632

LER 89-060-01:on 891212,standby Gas Treatment Sys Charcoal Filter Discovered in Degraded Condition W/Possible Bypass Leakage Paths.Similar Condition Discovered on 891218.Caused by Holes in Charcoal Filter assemblies.W/900515 Ltr

ML20043A632
Person / Time
Site:	Limerick
Issue date:	05/15/1990
From:	Madsen G, Mccormick M PECO ENERGY CO., (FORMERLY PHILADELPHIA ELECTRIC
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
LER-89-060, LER-89-60, NUDOCS 9005220331
Download: ML20043A632 (10)

Text

.. 10 CFR 50073 O

PHILADELPHIA ELECTRIC COMPANY LIMERICK GENER ATING ST ATION P. O. DOX A S AN ATOG A. PENNSY LV ANI A 19464 (21!!) 3271200 say 2000 m a. u.c o ,.m ic n . ,... , .c. May 15, 1990

. .. . .. .;' 01.".'."/.*, * ". . . . ..., Docket Nos. 50-352 50-353 License Nos. NPF-39 NPF-85 U.S. Nuclear Regulatory Commission Attn Document Control Desk Washington, DC 20555

SUBJECT:

Licensee Event Report Limerick Generating Station - Units 1 and 2 This revised LER reports a condition that could have prevented the Standby Gas Treatment-System (SGTS) from fulfilling its safety function and resulted in Operations prohibited by Technical-Specifications. The event was caused by failure of charcoal filter assembly welds that resulted in possible bypass leakage paths through the filter assembly. This LER revision provides. supplemental information for the weld failures and a determination that they were due to a manufacturing deficiency.

Reference - Docket Nos. 50-352 50-353 Report Number: 3-89-060 Revision Number: 01 '

Discovery Date: December 13, 1989 Report Date: May 15, 1990 Facility: Limerick Generating Station P.O. Box A, Sanatoga, PA 19464 This revised LER is being submitted pursuant to the requirements of 10'CPR 50.73(a)(2)(1)(B), 10 CFR 50.'73 (a)(2)(v)(C) and 10 CFR 21.21(b). Changes in the revised LER are indicated by revision bar markers in the right hand margin.

Very truly yours, c

/

c 4

WAR:cah cc T. T. Martin, Administrator, Region I, USNRC T. J. Kenny, USNP.C Senior Resident Inspector, LGS 900S220331 900 DR ADOCK 050gg ,

,,em as y s. Nuctt M L 5 ; ULIT OR7 COMutEst0N Nm (Pp20VIDOMsNO 39tootM

• • ("*'; " " W

,. LICENSEE EVENT CEP:RT (LER)

DOCK t f NUM0t h m 8 AM 4 PiCILif v esAME fli Limerick Generation Stati m 1gtt 1 _

o_16 t oj o l o l 315.12 1 l0d O f 9

'" Standby Gas Treatment System in degraded condition due to eeld failures and leakanc of charcoal f rom adsorber filters. OTHEM $ AClblIlt$ INv0Lvt0 ill O VENT DAll Ill Lim NuMet h (St htPont Da t t 17e

.ac.6rt . Nawns occa t t Nyvegn g, DAY 4s IP4 NT H OAV YEAR ttAn "ONU n @y*,"$ YONTH v t A P-LCS - 2 0I Ll01010 f 31513

~ ~

1l2 1l3 8 9 8l 9 0l6 l0 0l 1 0l5 1l5 9 l0 oi 5 1 0 1 0;0, I i tuin e8,0 ? is sue.intti0 Puasva=t 50 tut alOuiniMiNt 0, so t e n i <c%. . . ..,.< ,a. ,,4.-v mi n ,im moos m n .osm so .o .i io ni.nsn .

1 u ni.im x n ni.unn.i tttmi n.o i.nuni no, 3,0i0 n .o. nin.,

.o ni.n ,

, >. n n... 1 04 = , y;,; g g mm o .oei.in n.io y u ni.ninn so ni.nsa. inn Ai n Aoei.in n..i _

en ni.nin.o _

so tr.us"*"* 10 CFR 21.21(b) to .061.In n.1 60 ni.n2niin So t3t.nsual LICE NS8 4 CONT Act P OR TMil Lt n (13.

Yt ttP=ONE NLV8th NAYS am8 A rQB4 G. J. Madsen. Renulatory Engineer. Limerick Cencrating Station 211 15 31217 l- 11121010 COM*Litt ONE LINI 90R B ACH COMPONENT Faitunt DischittO IN TMit AtPOmf it3i CAv$t tv87tv CovPONENT

"'t 0 Pn Cault tv5Ttv C0VPONtNT "$'g'C " fg0",','[g* '

5 I I I I .I I I I I t i l l I I I I I I I I l l 1 ! I t i SUPPLIYtNT AL mt PORT 8 x7tCito n.' v0*.'=f Dat fvtAN

$58W$5 ION YtI dit r., roepi.fe I A*f f?f O hpW35 SON 04 Ti.d No g j l AS57m&CT st.mer to 1400 spues e e appres,a .re+ Mreen s*, e so.ce t pennet.m team tt6i On December 12, 1989, the 'A' Standby Gas Treatment System (SGTS)

Charcoal Filter was discovered to be in a degraded condition with possible bypass leakage paths through the charcoal filter bed.  !

This condition provided the potential for SGTS Filter bypass l leakage to be greater than the Technical Specifications (TS) surveillance allowable limit of 0.05% and could have prevented the SGTS from fulfilling its intended safety function. Several charcoal filter assembly welds failed, creating openings (holes) at the bottom of the outer (downstream) screen of the filter assembly. On the 'A' SGTS Filter, approximately 2.5 cubic feet i of charcoal had leaked out and the level of charcoal was lower than the normal full condition. On December 18, 1989, a similar degraded condition of the 'B' SGTS Charcoal Filter was discovered j with approximately 0.1 cubic feet of charcoal having leaked out, but the level of charcoal remained at the normal full condition.

After a thorough inspection, the damaged filter screens were repaired, charcoal replaced, and the SGTS returned to operable status. As an augmented surveillance activity, each SGTS Charcoal Filter bed has been inspected monthly,.following the TS required system flow test. We have concluded that the weld failures resulted from insufficient fusion during the electric resistance welding process and is a defect reportable in accordance with 10 CFR 21.21(b).

4, . ... a.

N C

hatfeemMa4A U S t'UCLl?2 KESULLTWY COMMIT $ ION LICENSEE EVENT REPORT (LER) TEXT CONTINUATION amoveo ows =o siso-oio4

.. soises esiss  ;

eassLsN hams sti pocast huwata us **C8 tai ten huussa (Se esaa " b'.$* ' *'W*f;

• i tirrerick Cerieratinn Station. Unit 1 0 [6 l0 l0 l0 l 3l Sl2 819 -

Ol6 l 0 -

Ol 1 012 0' 0 19 l l snxiom.mnemn eme,aeso,n eucrw.w vnn i

~

Unit Conditions Prior to the Event

• l ,

Unit 1 Unit 2 .f Operating Condition: 1 (Power Operation) 4 (Cold Shutdown)

Reactor Power: 100%- 0%

These Operating Conditions existed jmmediately prior to discovery.

of the degraded condition of the Standby Gas Treatment System (SGTS)(EIIStBH) Charcoal Pijters. However both units have operated at various reactor power levels since January. 19, 1989, ,'

when the 'A' and 'B' SGTS Charcoal Filters were last inspected during routine surveillance testing. In addition, operability of ,

the SGTS was required to support Refuel Floor Secondary .

Containment operability from February'8, 1989 to' March 30, 1989 and from November 8, 1989 to December 9, 1989.

Description of the Event:  ;

During a routine maintenance activity, the SGTS Charcoal Filters were discovered to be damaged with possible bypass leakage paths through the charcoal filter bed. This condition provided the potential for SGTS Filter bypass leakage to be greater than the <

Technical Specifications (TS) allowable limit of 0.05% and could i have prevented the SGTS from fulfilling its intended Safety .

function to limit the release of radioactive material.

,- On December 11, 1989, the 'A' SGTS Charcoal Filter was opened for a routine inspection and scheduled replacement of the filter bed

' charcoal. Approximately 2.5 cubic feet of a total of approximately 80 cubic feet of charcoal was discovered on the filter housing floor and the level of-charcoal in the 8 inch '

thick filter bed was lower than the normal full condition. On December 12, all of the charcoal was removed from the filter bed as planned. A thorough inspection of the filter assembly ,

revealed the source of loose charcoal to be from three openings-(holes) at the bottom of the outer (downstream) screen of the 1 filter assembly. For a diagram of the filter-assembly, refer to Figure 1 on Page 9 of this report. An engineering evaluation of

the filter screen holes determined that they were caused by ,

l i l . 73... ...

kmc ,eeen 3 eta U S CUCLln it3V*AYS7 CDuwsM14 UCENSEE EVENT REPORT (LER) TEXT CONTINUATION tmovie ove ao sm.m ,

• • t oelt8 8'3115

.A&6LsYv hamn m oocast svuesa ta' Lin huestR16: tect (si vs a "utW.  :

4'#.9 timerick Ceneratinn Station. Unit 1 ol5l0-l0jo13l512 319 -

Ol610 -

01 1 Ol3 0F 0 lo sw u . < ~ w ,-., ,m e

failure of several small spot welds connecting the screen to the .

filter mounting frame. On December 13, at 0915 hours0.0106 days <br />0.254 hours <br />0.00151 weeks <br />3.481575e-4 months <br />, station personnel _ concluded that degradation of the screen and the lower .

than normal level of charcoal in the filter bed could have created a bypass leakage path exceeding the TS surveillance allowable bypass leakage limit of 0.051. Following screen .

repairs, loading new charcoal in the filter bed and performing f required surveillance tests, the 'A' SGTS was returned to operable status at 2000 hours0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br /> on December 16, 1989. The 'A' SGTS was out of service for 134.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> total for the scheduled filter charcoal replacement and correction of the deficiency. .

After discovery of the failed welds on the 'A' SGTS filter bed, station management decided to inspect the redundant 'B' SGTS filter bed for indications of a similar degradation. On December 14, 1989 with the 'B' SGTS in operable status, a boroscope was' used to inspect the 'B' SGTS Charcoal Filter. Approximately three quarters of the charcoal bed was inspected and no evidence of similar. filter degradation was observed. The inspection was not complete because the entire filter screen surface area was not accessible with the boroscope. Since no problems were identified by the inspection, station management concluded that the 'B' SGTS was operable and would be fully inspected after the

'A' SGTS was returned to service.

After completion of repairs to the 'A' SGTS filter bed, the 'B' ,

SGTS was removed from service on December 18, and a complete visual inspection of the 'B' Charcoal Filter was conducted.

Approximately 0.1 cubic feet of loose charcoal was discovered on ,

the filter housing floor in the area not accessible during the boroscope inspection. Similar to the 'A' filter, this charcoal was due to a small opening of the filter bed outer screen where ,

it was spot welded to the filter mounting frame. Although some

! - charcoal leaked out, the charcoal filter bed was still filled '

above the screen height and there was no evidence.of. voids or potential bypass leakage paths. The 'B' SGTS was not considered inoperable as a result of this condition. However, station management conservatively concluded that the 'B' SGTS may have been in a condition that could have prevented the 'B' SGTS from performing its intended function. Following the charcoal L unloading, screen repair, loading of new charcoal in the filter I

bed and performing required surveillance tests, the 'B' SGTS was- ,

returned to operable status at 1730 hours0.02 days <br />0.481 hours <br />0.00286 weeks <br />6.58265e-4 months <br /> on December 21. The

'B' SGTS was out of service 70.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> total for the filter charcoal replacement and corrections of the deficiency.

s

=2 eosw as..

waC ,eee Je64 U S NUCLLED KEDULaTORT COMMfS510% i LICENSEE EVENT REPORT (LER) TEXT CONTINUATION s*revio on No e-m 8 41143 3-3100 sac LITy naML m pocast NVusta un ggn nyugga ag, tact tai

.... wp .

gg  !

0F Lirneritk Ceneratinn Stat f on. Unit 1 o l5 l0 lo l0 l 31512 819 -

Ol6 I O -

Ol 1 0l4 0 19 g nma - . a .- - con .- ,,m,

=

A four (4) hour notification was made to the NRC on December 13, 1989, at 0947 hours0.011 days <br />0.263 hours <br />0.00157 weeks <br />3.603335e-4 months <br /> in accordance with 10 CFR 50.72- i (b)(2)(iii)(C) because the event could have prevented the SGTS from performing.its intended function to limit the release of radioactive material. A follow-up notification was made to the .

NRC on December 18, 1989 at 2100 hours0.0243 days <br />0.583 hours <br />0.00347 weeks <br />7.9905e-4 months <br />, when it was discovered -

that the 'B' SGTS filter had a similar, although less significant degradation. Accordingly, this report is being submitted in accordance with 10 CPR 50.73 (a)(2)(v)(C). This condition may have existed since January 19, 1989 when the 'A' and 'B' filters were;1ast inspected for routine surveillance. Therefore this .

report is also being submitted in accordance with 10 CFR 50.73 (a)(2)(1)(B), "Any Operation or Condition Prohibited by the Plants Technical Specifications," because the significant bypass leakage path discovered on the 'A' SGTS Charcoal Filter-was assumed to be greater than the TS Surveillance 4.6'.5.3 allowable  !

limit of 0.05% and the required TS remedial actions were not taken. An engineering evaluation concluded that the-filter screen weld failures were due to a manufacturing deficiency that l 1s a defect reportable.in accordance with 10 CFR~21.21(b).

l l

l i

Consequences of the Event:

Tne consequences of this event were minimal. There was no release of radioactive material to the environment and the systeg degradation was discovered as a result of the established surveillance and preventive maintenance programs. The-exact duration this degraded condition existed is not known. However it did not exist on January 19, 1989 when the 'Aand 'B' filters were last inspected for routine surveillance.

There were, however, greater potential consequences of this '

condition. Normal SGTS lineup is with both the 'A' and 'B' SGTS Filter trains aligned in the automatic start mode. Had the SGTS been required to limit the release of radioactive material in response to a Loss of Coolant Accident (LOCA) or Refueling Accident, both the 'A' and 'B' SGTS trains would have initiated.

Due to the discovered condition of the 'A' and 'B' SGTS Filters, ,

the SGTS would not have performed its design function to remove 99.0% of all radioactive iodine from secondary containment effluent gases as assumed in the Design Basis Accident (DBA) safety analyces.

~ ,a u..

. i

U.$. NUCLt 3 $.StWLATORY COMM'55tDN 44C perm 384

• "'- UCENSEE EVENT REPORT (LER) TEXT CONTINUATION utovio eve No mo-o* ,

,, tocas smo

.e .

aan ava*5a m u a =U=eia

• 4 *"*

..ciu r, ... m ggan ,8

[.* b  :

<(I[0, .

Limerick Generatine Station. Unit 1 0l5l0l0l0 31'512 819 -

Ol610 -

01 1 015 0F 0l9 '

rus a n. ea . man. .. na nc , su n m> ,

l 1

I Cause of the Event:

This event was caused by holes in the SGTS charcoal filter

. assemblies created by failure of spot welds which connected:the outer (downstream) screen to the filter mounting frame.. The holes allowed charcoal to leak from the filter. bed and increased the potential for untreated filter bypass air flow. The exact. '

cause of the spot weld failures did not appear.to be the result of operations, surveillance or maintenance activities. An engineering evaluation concluded that the weld failures resulted l from insufficient fusion during the electric resistance welding process utilized in the manufacture of the filter assembly. _

-However the filter manufacturer does not agree with this assessment. They suggest that the weld defects may have been  :

caused by mishandling during shipment and installation of the filter.

1 Corrective Actions:

l 1: Following discovery of the 'A' SGTS Charcoal Filter screen l

problem, on December 13, a Nonconformance Report (NCR) was issued and Engineering Department personnel initiated an' investigation and analysis of the problem. A second NCR was issued on December 19, when a similar problem was discovered on the 'B' SGTS .

Charcoal Filter screen. Both the 'A' and 'B' filter outer t screens were repaired using a silver solder, brazing procedure.

The affected areas (openings) of the screens were re-fitted and recurely attached to the filter mounting frame. The filter manufacturer was consulted and concurred with the repair procedure used. Both filters were refilled with new charcoal and j were tested to demonstrate acceptable bypass _ leakage before they were declared operable and returned to service.

Actions Taken-to Prevent Recurrence:

With the charcoal removed, each of the filter screens were examined ~ completely for evidence of degradation (attachment weld failures). The only problems identified were on the outer (downstream) screen, where it connected to the lower mounting t

,g. .a -...

A h

U.3 kgCLt33 3.E1941T%7 COMMt3SION JemC poem anta

• smovso oue =o sm-om LICENSEE EVENT REPORT (LER) TEXT CONTINUATION soints nin  :

eacauty hawa ti, m attkuussa m ggm wpuega 16p 'ACS

• u w.':'  %*;:

~

vs a 0 15101010 l31 512 819 -

Ol6 l 0 -

01 1 016 0F 0 19 Limerick Generatine Station. Unit 1 .

TLIT t# anoep spece a reewed, asse admasaW 4#C #eam 3RM s/ (173 frame. The affected areas were limited to three welds on the 'A' filter and one weld on the 'B' filter. This failure was reported on the Nuclear Plant Reliability Data System.

. As an augmented surveillance activity, a complete inspection of each SGTS Charcoal Filter bed is currently performed monthly, following the TS required system flow test. These special inspections performed since January 1990, have revealed no .

further evidence of charcoal leakage or screen weld failures.

Based on the results of these inspections, the frequency of inspection will be decreased accordingly. A return _to the normal 18 month frequency will be implemented when it is evident that the filter silver brazing repairs and spot welds do not degrade because of filter service time or the alternate repair technique.

This is appropriate based on the Engineering Evaluation that the spot weld failures resulted from a manufacturing. deficiency and 1 not from inservice use.

All of the plant system charcoal filters at the Limerick Generating Station, were manufactured with a similar screen spot weld design. Maintenance History records were reviewed and no previous similar weld failures have occurred. Therefore as concluded in the engineering evaluation, this' event was limited to individual defective welds and is not a potential generic-l problem.

l Previcus Similar Occurrences:

l 1

None Cause Codes: B12 Manufacturing Error 1'

Additional 10CFR21.21 Information:

r

, The 'A' and 'B' SGTS Charcoal Filters ~were supplied to the  ;

' Limerick Generating Station by the American Air Filter. Company. '

Failure of three welds on the 'A' filter and one weld on the 'B' filter were due to a manufacturing deficiency that is a defect reportable in accordance with 10 CFR 21.21(b).

gig.... nu  ;

+

4

'=ac e .kA vi e; cts:.2 zeovatoav comueo=

• • u :ovio ove =o. mso-oio.

LICENSEE EVENT REPORT ILER) TEXT CONTINUATION

.. socas e ucs p_ciufvnauttu GOCKE1 NUweaa al 43 a muuggn asi PAGE (38 vsaa "#'.;;P  :

• 'a*.f I Limerick Generatinn Station. Unit 1 0 l5 j0 l0 l0 l 31512 819 -

Ol6IO -

01 1 Ol7 0' 0 19 ,

sus tu ,., mm. w.e u..a w unc w w a me l

ENGINEERING EVALUATION FOR FAILURE OF THE WELDS IN THE SCREENS OF  ?

THE SGTS CHARCOAL FILTERS CONSTRUCTION OF THE SGTS CHARCOAL FILTER L The SGTS charcoal filter consists of a cylindrical tank enclosing two concentric perforated screens extending to the full height of the tank. An adsorbent charcoal fills the space between the screens. The cylindrical screens are located eccentrically relative to the centerline of the tank, with a separation that varies 3" to 3' from the tank wall. The air enters the tank ,

through the inner core of the screens; then the air is filtered as it flows radially through the charcoal-filled screens; and exits the tank through the space between the outer screen and'the-tank wall.

The inner and outer screens are supported circumferential1y by four (4) stainless steel hoops spaced between the floor and the roof of the tank. The screens are spot welded to the hoops at intervals of about 1". The hoops are welded in-turn-to 3" x 3" x 1/4" angles vertically arrayed around the hoops and welded.to the tank at the floor and roof. The charcoal and air bypass is sealed off by welding the hoops at the floor and roof to the tank. The screens are 26 gauge stainless steel and the hoops are 1/4" thick stainless steel that vary in width from 1/2" to 5 l 3/4".

l WELDING OF THE SCREENS Spot welding is one of the resistance welding processes which creates a fusion of the parent metals through the heat generated

, by the resistance to the high amperage electrical current at the l joint. The electrical current is delivered by the electrodes l- clamping the parent metals. The soundness of the spot weld depends on the accurate control of current to produce melting of parent metals at the joint, timing of current delivery to provide '

the right conditions for fusion, and the proper application of ,

pressure at the clamps to provide the required resistance. In view of the variety of factors that have to be blended precisely to obtain a sound weld, the spot we3 ding process is usually performed with automatic resistance welding machines.

Despite the accuracy of which the automatic machine welding is capable, the machine is presumed-to be unable to distinguish and '

compensate for the dissimilarities in thickness of the parent metals (1/4" hoops and 26 gauge screen). In the case.of the 4 80mu 3.f m

=ac e .'assa v.s ::vetsu c s196r.'onv couwism  !

LICENSEE EVENT REPORT (LERl TEXT CONTINUATION emova9 ove ao *m-**

.. ex css o w n

.acitetv maus m putati sevmasa tal 'aCs (36 -

tan huweta (Si saa "t!.@ -

"%* D 1 01610 lo 10 l 31512 819 -

016l 0 -

01 1 018 0F 0 19 Limericka ceneta. tine sw w a,. < , w uncstation.

on.m .nmUnit defective welds, the greater mass of the hoop may dissipate heat faster than the screen, thus, the melting in the hoop would lag behind the screen and would be unable to fuse successfully with the opposite melt in the screen. This could result in a defective cold joint spot weld.- This phenomena is evident from the pattern of' spot weld failures which shows most of the defects occurring near the joints between the bottom hoop and the vertical angles, where the large volume of metal would provide an ideal path for dissipating heat. ,

i CAOSES OF TACK WELD PAILURE When a charcoal filter is in operation, the pressure. imparted by the flowing air on-the charcoal is transmitted by the charcoal to +

the outer screen. The pressure tends to bulge out the screen and i pull it away from the hoops. At the bottom hoop, where the ..

screen ends, a broken weld would tend to open a gap between the screen and the hoop. The mechanism that created the gaps in the filter apparently started with a typical cold joint described above. Deprived of support, the load on the cold joint would be transferred to adjacent weld,-further stressing the adjacent welds and causing more failures of welds with inadequate fusion.

REPAIR OF THE SCREEN '

The repair by continuous silver brazing of.the gap between the screen and the bottom hoop produced a stronger. bond by providing I.

a continuous seamless joint. In rejoining the screen to the ,

l hoop, the additional stress imposed on the adjacent welds has been relieved, thus blocking further propagation of the separation.

VENDOR'S POSITION When the American Air Filter Company was contacted for'their views on the possible causes of the weld defects, they contended that they do not fully agree with the assessment that the defects were due to a manufacturing deficiency. They suggest that the defects may have occurred through mishandling during shipment and installation of the filters. However, in our judgm Jt defects-l, caused by a failure mechanism due to mishandling would have been manifested differently than the defects that we observed. ,

e

~[

g e.... ..

.Rc . . vs metin novatar co==inio= )

• unovio ove ao sino-o**

.- UCENSEE EVENT REPORT (LER) TEXT CONTINUATION si,iais sm

[

8 ACILily hAMt its 90Catt hvuasmtas 46R NUutta16 tatt (Si

. . . a gt,m*6 a';=.g:

Lit 2 rick ~ Generating Station, Unit 1 0F 0 19 o l610 lo lo 13 l Sl 2 81 9 -

0 1610 -

0 l1 0 l9 tort in - . . .-,. .u- ==c e- w., un ,

i F::GURE a .

1 1

LOCATION OF _  !

MELD FAILU9,ES IMPECTION CDVER  !

,n

//# \ 'l 'g l all I \ ' ' .

I g l- * )I '

e l

e.

i\ // ( ,I / !-

1

, /

( \ \'d_s _ ,

f

~ -

[

V:: N W.AN AIR FLOkt 1 m RIR FLObj

! INLET T -

OuT LET i

j h ,. .

i i e i

l l

li l i ,

I ( ,'

LOCRTION OF 8'8  !!

I I

'Wg Ii i! h FILTER SCREENS MELD FRILORES l l. l; [

, y!  !! L, 2AONT ELENRTION g ,.. . ..