ML20148H326

Insp Rept 99900853/87-01 on 871207-11.Nonconformances Noted. Major Areas Inspected:Validity & Basis of Closure Tests of Curtain Type Fire Damper Assemblies Performed Under Simulated in-plant Air Flow Conditions

ML20148H326
Person / Time
Issue date:	03/04/1988
From:	Baker E, Petrosino J Office of Nuclear Reactor Regulation
To:
Shared Package
ML20148H281	List: ML20148H275 ML20148H305 ML20148H326
References
REF-QA-99900853 NUDOCS 8803290395
Download: ML20148H326 (12)

Text

.

s ORGANIZATION: PULLMAN CONSTRUCTION INDUSTRIES, INCORPORATED CHICAG0,'ILLIN0IS REPORT INSPECTION INSPECTION N0.: 99900853/87-01 DATES: 12/07-11/87 nN.9fTF unMDR- 7?

CORRESPONDENCE ADDRESS: Pullman Construction Industries. Incorporated ATTN: Mr. L. H. Goldwyn President 1400 East 97th Place Chicago, Illinois 60628 ORGANIZATIONAL CONTACT: Mr. M. A. Jarigese QA Manager TELEPHONE NUMBER: (312) 374-6700 NUCLEAR INDUSTRY ACTIVITY: Pullman Construction Industries, Incorporated (PCI) is currently providing 3 hour3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> Underwriter's Laboratory (UL) rated curtain type fire dampers to NRC licensees that are enveloped within an outer flange assembly that provides for thermal expansion and seismic considerations.

, D ASSIGNED INSPECTOR: , -f / e., #2u/ Y M8 i)9depli J. Petrosino, Program Development and Date f Redctive Inspection Section (PDRIS)

OTHER INSPECTOR (S): Anthony Fresco, Brookhaven National Laboratory Joseph Ulie*, USNRC - Re ion III APPROVED BY:

Edward T. Baker, Acting Chief, PDRIS, Vendor Inspection

[

Date Branch

• Inspection participation on December 7, 1987.

INSPECTION BASFS AND SCCPE:

A. BASES: Appendix B to 10 CFR Part 50 and 10 CFR Part 21.

B.

SCOPE: The inspection was performed to evaluate the validity and basis  !

of clost.re tests conducted by PCI of their curtain type fire damper assemblies performed under simulated in-plant air flow conditions and to '

review the implementation of its QA program in certain areas.

PLANT SITE APPLICABILITY: Generic, all plants.

I I

{

w 100853 M

l

ORGANIZATION: PULLMAN CONSTRUCTION INDUSTRIES, INCORPORATED CHICAG0, ILLINOIS REPORT INSPECTION i NO.: 99900853/87-01 RESULTS: PAGE 2 of 10 A. VIOLATIONj:

None.

B. HONCONFORMANCES:

1. Contrary to Criterion I, "Organization," Criterion II, "Quality Assurance Pr6 gram" and Criterion III, "Design Control," of Appendix B to 10 CFR Part 50 and Sections II and III of PCI's quality assurance (QA) manual, PCI has failed to establish an adequate QA program to control its safety-related activities in the areas of design and QA monitoring for fire dampers. Examples of this include:

a. Objective evidence could not be provided that would indicate that PCI had performed an independent review for technical adequacy of its fire damper "suitability for application" evaluations that it has performed for several NRC licensees;

b. Objective evidence could not be provided that would indicate that PCI had perfonned an independent review for technical adequacy of its fire damper performance qualification curves;

c. PCI has not adequately delineated the authority and duties of its design engineers who perform activities affecting the l functions of safety-related fire dampers nor has it delineated associated monitoring duties for its QA engineers;

d. PCI has not established adequate engineering procedures or I instructions to provide control over its t.ctivities affecting its safety-related fire dampers;

e. It could not be determined by a review of objective evidence whether or not PCI has correctly translated its in-house fire i l

damper raw test data into PCI design parameters or instructions,  ;

and 1

f. Calculations and engineering determinations for safety-related fire dampers are typically performed by PCI engineering personnel without benefit of procedures, instructions or QA monitoring activities.

I i

ORGANIZATION: PULLMAN CONSTRUCTION INDUSTRIES INCORPORATED j

. CHICAG0, ILLINOIS l REPORT INSPECTION NO.- 99900853/87-01 RESULTS: PAGE 3 of 10 C. OPEN/ UNRESOLVED ITEMS:

The below listed issues were discussed during the inspection but were not fully and or adequately resolved. Therefore, the following items are classified as open items and will be discussed during a future NRC inspec-tion and/or subsec,uently by NRC staff:

1. Testing of fire dampers for closure under air flow at ambient tempera-tures, in lieu of elevated temperatures which would be experienced during a fire;

2. Industry curtain type fire dampers (CTFD) closure certification of dampers based on testing of one CTFD of a given size and design;

3. of the fire rating of fire dampers by the Undemriters The validity (UL), that are based on "the assumption that air Laboratory conditioning and ventilating systems are automatically shutdown at the start of a fire" (emphasis added), when automatic shutdown is not designed into the system and does not occur; and

4. The applicability to nuclear power plants of the Air Movement and Control Association, Incorporated (AMCA) standard number 500, which is the only industry standard that delineates methods to test dampers under flow and defines a fire damper as a device that closes under flow and effectively stops air flow. This differs from the UL basis for its fire duration rating.

D. STATUS OF PREVIOUS INSPECTION FINDINGS: l l

None.

l E. OTHER FINDINGS OR COMMENTS:

1. Entrance and Exit Meetings The NRC inspector informed PCI management representatives of the l scope of the inspection during the entrance meeting on Decet.oer 7, l 1987, and sumarized the inspection findings and observations during i the nxit meeting on December 11, 1987.

l

2. Background This inspection was conducted to review and evaluate CTFD closure tests that were conducted by PCI under simu!ated in-plant air flow i

' ORGANIZATION: PULLMAN CONS'.~rlVCTION INDUSTRIES, INCORPORATED CHICAG0, ILLIN0IS REPORT INSPECTION N0.: 99900853/87-01 RESULTS: PAGE 4 of 10 conditions. The testing was performed by PCI at the American Warming and Ventilating (AMV) Company test laboratory in Toledo, Ohio. The tests were conducted to qualify a PCI modification to the typical CTFD design configuration and were all witnessed and documented by AMV personne~l. The PCI modification did not change the CTFD blade or frame; instead the change consisted of adding corregated steel segments to the CTFD outside frame members, and then fabricating a dual element sleeve around the existing CTFD.

These changes add rigidity and provide room for thermal expansion for the resulting PCI fire damper assembly unit.

It was also observed that the PCI fire damper assembly would appear to better withstand rough handing and installation at the end user's facility. The PCI assembly would also appear to remove the possibility of typical warping problems during installation caused by different welding techniques used during plant installation.

However, PCI does not manufacture the actual fire damper itself, but procures safety-related CTFD's from Air Balance Incorporated (ABI),

Westfield, Massachusetts. The ABI fire damper is then encapsulated, without change to its blades or track, into the forementioned sleeve assembly which is designed by PCI to provide for both thermal expansion and seismically induced forces.

2. Purpose of Testing PCI conducted its own testing to determine the product operability parameters and effects of the following variables on damper closure under simulated in-plant air flow:

Upstream duct length and configuration Downstream duct length and congfiguration Damper and flow orientation

- Static pressure Velocity Positive and negative pressures Damper size Height-to-width ratio .

Spring tension An important aspect of the PCI testing appears to be its establisn-ment of the maximum total (or stagnation) air flow pressure at which a given damper will successfully close. It appears that the

~

~

l 1

l ORGANIZATION: PULLMAN CONSTRUCTION INDUSTRIES, INCORPORATED CHICAGO, ILLIN0IS REPORT INSPECTION NO.- 99900853/87-01 RESULTS: PAGE 5 of 10 operability of a CTFD is more dependent on the total differential pressure at the damper and less dependent on the associated air flow rates or velocities.

PCI has put its accumulated raw test data into a report which then is used to certify closure of its damper under different in-plant duct work configurations and air pressure variables.

3. PCI Test Data Usage and Creditability Usage: Four reports were presented in a binder entitled, "Pullman Construction Industries, Incorporated - Fire Damper Data." The first is a seismic qualification of HVAC fire dampers for the Shearon Harris plant dated December 20, 1984. The second describes CTFD closure tests for ABI Model 319 horizontal and vertical fire dampers dated November 1984. The third is a report by Underwriters i Laboratories on a fire test of a 3 hour3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> CTFD by Air Balance utilizing an alternate method of installation. This report is dated December 5, 1984. Lastly, the fourth is a test report by American Waming and i Ventilating, Incorporated, on a PCI CTFD assembly, dated July 18, 1 1985.

After discussions with PCI personnel as to the specific ~ purpose of the reports, it was determined that:

Enclosure No.1 is intended to be a seismic qualification of PCI's fire damper assembly.

Enclosure No. 2 is intended to show erratic perfomance of existing dampers not supplied by PCI.

Enclosure No. 4 is intended to show actual air flow closure performance of PCI's fire dampe assembly.

Enclosure No. 3 was not separately reviewed but the report itself states that it is intended to determine whether an alternate installa-tion of a fire damper provides resistance to the paysage of fire for l 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> and to the effects of a hose str.*am following fire exposure l

as specified by the UL Standard 555 fire endurance and hose stream l tests. It was noted that each report was presented as a separate j entity with no explanation of:

The intended purpose or conclusions to be drawn from each report.

The interrelationship between each test resu'it.

I ORGANIZATION: PULLMAN CONSTRUCTION INDUSTRIES, INCORPORATED CHICAGO, ILLIN0IS REPORT INSPECTION l NO.: 99900853/87-01 RESULTS: PAGF 6 nf 10 l At this point in the inspection, the NRC team queried PCI personnel in regard to a possible misuse or misinterpretation of its test reports by either NRC staff or by licensees that do not have PCI's 1 CTFD assemblies installed.

]

PCI's management comitted to address this NRC staff concern by generating an explanation of the intention of each document and the relationship of each document to the others in a foreword signed by 1 appropriate PCI personnel. This will address the NRC concern over l possible misuse of the test reports to justify fire damper installa- l tions which were not supplied by PCI.

Credibility: After discussion with PCI personnel, it was determined that the appropriate test data for review by the team was enclosure No. 4, whir,1 PCI utilizes as a source of test data to establish the sizing and configuration of its damper assemblies.

PCI also had supported a fifth test report performed by American Warming and Ventilating, Incorporated of PCI's horizontal and vertical CTFD's. This report, dated September 26, 1985, is considered proprietary by PCI. It includes test data performed for R&D purposes where important modifications were made to the dampers and then the modified dampe s were subsequently tested for closure under air flow. I PCI personnel stated th,t its only bearing on their product is that, in some cases, the data can be used to verify that more than one sample of a given size and configuration was tested to support the data in the July 18, 1985 report.

It was revealed that only one sample of a given size and type damper is required to be tested by industry standards such as UL 555 and AMCA 500. There are no specific industry requirements concerning the minine c number of samples to be tested. For the tests reported on in tne July 18, 1985 report, it was noted that the test samples were PCI standard stock items.

In view of these factors, it was decided that only the July 18, 1985 PCI report would be considered in evaluating the acceptability of PCI's testing program for closure under air flow since its results are the most comprehensive.

Since the intended purpose of the PCI testing is to establish the maximum total (or stagnation) pressure at which a given damper will successfully close, PCI utilized the results of its testing to plot a curve of damper area versus the pressure upstream of the damper Y

--w

. ORGANIZATION: PULLMAN CONSTRUCTION INDUSTRIES, INCORPORATED CHICAG0, ILLIN0IS REPORT INSPECTION NO.- 99900853/87-01 RESULTS: PAGE 7 of 10 upon complete closure (i.e., the total pressure since the flow velocity is zero after closure with the supply fan still operating).

This curve, together with a PC-based computer program which PCI uses as an engineering tool to develop the damper specifications, is then interpreted as establishing an upper limit on the total pressure at which a given damper will close, thereby leading to certification by PCI that a damper design established by this informational input will close under air flow conditions specified by a licensee. ,

An independent plot of the curve was performed by the BNL representa-tive and the data points properly coincided with the PCI plot.

However, this should not be construed to be a verificatior, of the technical adequacy of the PCI test data or design program.

Conclusion:

PCI's fire damper testing methodology appears to have simulated some worst-case in-plant air flow conditions, using a sound engineering rationale, and its test results and conclusions appear to be crediable based on the inspectors review. The PCI documented testing observations also appear to have identified one of the two major reasons why CTFDs industry wide are experiencing ,

difficulties or failures during air flow closure attempts. One reason appears to be the total differential pressure as discussed above. The other reason appears to be that the CTFD's were not originally designed to close under air flow conditions, because licensees did not specify air-flow closure requirements on purchase orders in the time frame of 1970-1980.

4. Industry Standards Currently, the only industry standard that has been typbily imposed by NRC licensees in regard to CTFD is the Underwriters Laboratory (UL) Standard number 555 (UL-555), "Fire Dampers and Ceiling Dampers." The UL-555 intent states, in part: "It is the intent that tests made in confonnity with the test methods described herein will develop data to enable regulatory authorities to determine the acceptability of fire damper assemblies for use in locations where fire resistance of a specified duration is required."

The UL-555 scope also states, in part, that: "Closing reliability of fire dampers is evaluated on the basis that air-conditioning and ventilating systems are automatically shut down when a fire occurs as described in the various provisions of the Standard for the Installation cf Air-Conditioning and Ventilating Systems, NFPA Number 90A, Therfore, the _ ratings are applicable to fire dampers and ceiling dampers installed in systems where air movement is effectively stopped at the start of a fire" [ emphasis added.]

. i ORGANIZATION: PULLMAN CONSTRUCTION INDUSTRIES, INCORPORATED l

CHICAG0, ILLIN0IS 1 REPORT INSPECTION NO.: 99900853/87-01 RESULTS: par # A nf in The Air Movement and Control Association, Incorporated (AMCA)

Standard Number 500 (AMCA-500) was adopted by the AMCA membership on October 18, 1973 and its stated purpose is to establish uniform test i methods for relating pressure drop and/or water penetration to sir  ;

flow rate for louvers, dampers and shutters. The AMCA-500 definition of a fire damper is: " ...a device arranged to interrupt air flow automatically through part of an air system so as to restrict the passage of flame, and is installed...to close automatically in the event of a fire in order to maintain the integrity of the fire-rated separation."

The AMCA-500 standard has several different test setup configurations for fire damper closure tests under air flow, with guidelines for l associated instrumentation types / configurations for the monitoring  ;

and recording of flow rates, static pressure drops, seating torque, '

and other related measurements. l l

PCI conducted its own testing program primarily because UL-555 is the only industry standard that has been historically imposed and applicable to fire dampers and ceiling dampers installed in systems 1 within nuclear plants. Since the PCI product is intended to provide advantages in installations required to meet seismic requirements, seismic tests were conducted in addition to closure under flow and structural integrity tests. Furthermore, the tests provided R&D ,

information on the usefulness of various modifications intended to  ;

improve performance leading to justification of the final product design. Ultimately, the test results establish a data base which PCI utilizes to specify product size and configuration (i.e., whether a single or multiple damper is required for a given size and flow condition). )

5. Design Control The NRC inspectors reviewed documents and conducted discussions to determine whether PCI's design control activities in regard to its fire daroper assembly test program are adequately controlled. It l was determined that PCI has not adequately controlled its design  !

control program as evidenced by the lack of objective evidence to assure that the quality activities were satisfactorily accomplished, j The basis and the results of the fire damper test program were reviewed to ascertain several aspects which included:

(1) Whether the test result raw data [ design basis] was correctly translated into design requirements;

O

. ORGANIZATION: PULLMAN CONSTRUCTION INDUSTRIES, INCORPORATED

. CHICAG0, ILLIN0IS REPORT INSPECTION NO.: 99900853/87-01 RESULTS: pant o nr in i

(2) Whether the design requirements were independently reviewed and  !

approved for technical adequacy; and i (3) To verify whether an adequate review for "suitability of applica-  ;

tion" is being performed. i 1

The review of the above three areas revealed that PCI had failed to adequately control its quality program activities. As an example, during a review of PCI's "suitability of application" design activities ,

it was revealed that the same PCI engineer typically performs all i the design activities and approves his own work. The design j activities include a review of a licensee's ventilation system design parameters and characterics for determining the particular size, type, and/or configuration of the PCI CTFD that will adequately perform their function within the licensee's ventilation system.

Hence, PCI cannot show that someone other than the person who performed the design activities independently reviewed it for technical adequacy (Nonconformance B.1 was identified in this area).

6. Procedures Instructions The NRC inspection team reviewed the PCI quality assurance manual design section and attempted to review PCI's design procedures and instructions. It was revealed that PCI has failed to adequately establish its design control program and failed to establish proce-dures that would assure that important engineering and quality aswnnce activities have been satisfactorily accomplished in regard to its i1re damper assembly program (Nonconformance B.1 was identified in this area).

The NRC review revealed that PCI has failed to adequately implement its QA program requirements in regard to assuring that quality activities are adequately controlled by procedures and instructions that delineate the applicable criteria. As an example, the NRC inspector requested to review the applicable procedure that delineates how the PCI engineer performs his review for suitability of component application at a licensee facility. The inspector was shown how each review is performed; however, PCI did not have a p a cedure that defined the job functions and responsibilities of the engineer for this aspect. It was also revealed that each of the other engineering activity areas in regard to PCI's test program were not delineated in regard to its QA program control and monitoring.

. i l

- j ORGANIZATION: PULLMAN CONSTRUCTION INDUSTRIES, INCORPORATED

~

CHICAGO, ILLINOIS l

REPORT INSPECTION '

NO.: 99900853/87-01 RESU!.TS: par,F 10 nf in

7. PCI Personnel Contacted l

• L. H. Goldwyn, President

• J. P. Goldwyn, Sr. Vice President

• R. F. Cekanor, Executive Vice President

• G. Zielinski, Vice Prasident

• M. E. Stucier, Project Manager

• T. I. Stewart, Project Manager i

• M. A. Jarigese, QA Maniger l

• K. Schaeflein, Shop Manager I

• S. Szykowny, Welding foreman i

• R. Baker, Welder 1

• Attended exit meeting.

- - ,- r,,-

Pullman Construction Industries, Incorporated t i

DISTRIBUTION:

iRIOS; Code;IE:09)

VIB Reading DRIS Reading ,

-JPartlow

• BGrimes ,

JStone EBaker JPetrosino (2)

LParker RMoist Julie, RIII EMerschoff, RII RGardner, RIII WXane, RI LReyes, RII CNorelieus, RII LCallan, RIV DKirsch, RV LZerr JGuillen JTaylor, BNL WShire, BNL AFresco, BNL 0FC :A

VIB:QflS

___...:..mg-/..__.....:..CHED

. V IS :AC/VIB:DRIS  :  :

. ........__: ..y. ____:..______. .................__..___......

NAME  :  : aker :JCStone

...__.:d.._.h___rosino:mgc  :

DATE :S /'//88  : / /88 ___:3/M/88

^) /1S/88  :  :

n

i \

~

l

, March 22, 1988 i

Pullman Construction Industries, Incorporated  !

l l

DISTRIBUTION:

iRIDS Code-IEiO9)

VIB Reading I DRIS Reading JPartlow I BGrimes JStone EBaker i JPetrosino (2) '

, LParker RMoist JVlie, RIII EMerschoff, RII l RGardner, RIII i WXane, RI l LReyes, RII '

CNorelieus, RII LCallan, RIV l DKirsch, RV l LZerr '

JGuillen JTaylor, BNL WShire, BNL AFresco, BNL l

l i

r 0FC :VIB:QRIS . CH ED :A V IS :AC/VIB:DRIS  :  :

......:...yh L /____...__:... ..........: ._y.

NAME :qJ}ggrosino:mgc  :  : aker . . . :JCStone

. : . . . . . . . 7..............................

. .(:  :

DATE :3 / //88  : / /88 :3$/i/88/  :$ /1 1/88  :  :

. . . _ . . -. _ _._ __