ML20137L711

Technical Evaluation Re HVAC Duct Work & Support Sys.Overall Structural Design of HVAC Sys Adequate & Provides Sufficient Margin of Safety to Failure Under Normal & DBA Conditions

ML20137L711
Person / Time
Site:	Clinton
Issue date:	01/22/1986
From:	NRC
To:
Shared Package
ML20137L692	List: IR 05000461/1985064 ML20137L688 ML20137L711
References
NUDOCS 8601280067
Download: ML20137L711 (9)

Text

. .

A// ecd mers/ 1 1

TECHNICAL EVALUATION OF THE CLINTON HVAC STRUCTURAL DESIGN OF THE CLINTON HVAC SYSTEM Staff Design Review and Audit of the Clinton HVAC System On December 8-9, 1983, the MEB Staff met with the applicant (Illinois Power Company) for the Clinton plant and its architect-engineer (Sargent &

Lundy) to discuss the design of the Clinton HVAC system. The staff reviewed the structural design adequacy of the HVAC supports, ductwork, bolting, and welds.

The mechanical design review included a review of the HVAC design specification, design criteria, design methodology, supporting calculations for standard and special HVAC designs, and HVAC drawings. This review was performed at the Sargent & Lundy offices in Chicago, Illinois. The second part of the review included a visual inspection of several areas in the Clinton plant where safety-related HVAC systems have been installed.

HVAC Review performed at Scrgent & Lundy Offices On December 8, 1983, the staff met with Illinois Power Company (IPC) and Sargent & Lundy (S&L) to review and audit the Clinton HVAC structural design.

The purpose of the review and audit was to quantify the actual design margin that existed in the HVAC supports, ductwork, bolts, and welds to determine if the strength variability of substituted materials could potentially affect the ability of the HVAC system to perform its intended function.

The HVAC design responsibility is divided into several design groups.

The HVAC Division is responsible"for the HVAC system design which includes the preparation of design specifications. The Mechanical Design and Drafting Division (MDDD) is responsible for the ductwork arrangement and the HVAC support design which uses the standard design details. The Component Qualification Division (CQD) is responsible for developing the design methology and providing the calculations for specific design problems which do not conform to the standard details. 9601290067 860122 OR ApocK 050 g 1

The Clinton HVAC design was started in the 1976-1977 period. A combined effort by HVAC, CQD, and MDDD was initiated to develop a consistent design philosophy for the HVAC structural design. This team effort resulted in developing rules for: .

1) isyout of ductwork, ductwork span criteria, and

2) ,J

3) standard design details for HVAC supports.

A generic set of calculations for supports and ductwork was established based on a worst-t:se basis to justify the standard design. The staff found that nany conservatisms are used to establish generic calculations (e.g., a maximum supported mass was used to assure that each duct span remains in the rigid rangeof33hz.) All HVAC systems in seismic Category I buildings (control, auxiliary, fuel, containment, diesel generator, and part of the screen house) were covered by the design.

The ductwork was designed using the AISI Code. The ductwork is typically ASTM A526/A527 material with a minimum yield strength of 30 ksi as required by the design specification. The ductwork meets the guidelines of SMACNA, "High Velocity Duct Construction Standard," (Third Edition). The welding is per-forced in accordance with AWS DI.1 and AWS D1.3.

The staff reviewed the HVAC Design Specification (K-2910), "HVAC Work,"

(Amendment 7). There are no exotic materials used for galvanized ductwork and HVAC supports.

Review of HVAC Design Handbook The staff reviewed the document entitled, "HVAC Duct Hanger Design Handbook." This document is a " cookbook" procedure which is used by MDDD.

This design handbook provides the hanger location criteria. For straight spans of ductwork, supports are required at a maximum of 14 feet for horizontal spans and 28 feet for vertical risers. The vertical risers use a

. .e heavier gauge sheet metal to allow the longer spans. The design handbook also provides pre-qualified support details. For HVAC supports not in confonnance with the standard design details, a S&L project instruction PI-CP-014

" Request for Special HVAC Duct Hanger Analysis" provides the guidelines to be followed for their evaluation. The seismic design was conservatively based on a 2g horizontal and 3g vertical (29 seismic + Ig deadweight) criteria.

Review of Calculations for Standard Design of HVAC Ductwork and Supports j The staff reviewed the set of generic calculations which qualified the standard HVAC ductwork and support design details. The generic calculations are documented in Calculation EMD-019620, " Design of Safety-Related HVAC Duct-Hanger System", dated September 28, 1980.

The goal of the HVAC standard design was to develop a basis for assuring that the behavior of both the ducting and supports remained rigid (fundamental fre'quency greater than 30 hertz). As a esult, the generic calculations established the weld capacity and support capacity necessary for ensuring the rigid behavior of the supports. The rigid behavior of the ductwork was established in the specified spans between supports. The calculations for the standard design identified the welds as the most limiting component in bending.

For the HVAC ductwork, the staff reviewed the calculations qualifying the 14 ft, horizontal span and 28 ft. vertical span. The ductwork allowable stress was based on 0.6 x material yield strength. The material yield strength of the A526/527 sheet metal was assumed to be 30 ksi. For rectangular ducts the maximum bending stress was found to be 10,609 psi (with an allowable stress of 18,000 psi).

For the vertical risers where a span of 28 ft. was allowed, the critical failure mode was found to be buckling. As a result, any ducting with one side less than 12 inches in width was, required to be provided with a lateral support eve ry 14 ,f t. For large ducts (120x40 inches), it was found necessary to provide a longitudinally stiffened stud to prevent the sheet metal from buckling.

e

4

. .c Review of Special HVAC Duct-Hanger Analysis The S&L project instruction, PI-CP-014, provides procedures for preparation, review, and disposition of requests for special HVAC duct-hanger analyses. The staff selected a sample calculation for'several HVAC supports which did not conform to the standard HVAC support design details. The calculation reviewed by the staff was identified'as Calc. No. CQD-008566, revision 0, dated July 15,1983.

The calculation followed the design rules of the AISC Code.

for example:

allowable shear stress = 0.4 Sy allowable bending stress = 0.6 Sy weld allowable = (0.3)(60)(.707)(3/16) = 2386 lbs/in.

The calculation allowed a 33 percent. increase for seismic conditions (both OBE andSSE). A tabulation of the calculated stresses for the support steel, welds, and' connection bolts are shown in Attachment A to this report.

Review of Expansion Anchor Bolt Design The staff reviewed the desigii of expansion anchor bolts used in the HVAC support embedment plates. The expansion anchor bolt design criteria was developed by the Structural Engineering Division (SED). Design charts were provided to the mechanical (CQD) group by SED. The charts were based on.the lesser of the manufacturer's ultimate capacity or the test results established by S&L for the response to IE Bulletin 79-02. A safety factor of four was used for both wedge and sleeve anchors. However, mostly wedge anchors (Hilti and Phillips) are used at Clinton. The mechanical group standardized the use of expansion anchor bolts based on the standard support size, the building in which the bolt is located, and the anchor plate size. For those expansion anchors where the standarized chart is not applicable, the loadings are sent to SED to determine the appropriate expansion anchor design.

h

. .c 1 .

4 The staff reviewed many support details and noted that expansion anchor bolts are not extensively used. Instead, the staff found that it was more

coninon to attach the HVAC supports to structural building steel with A490 j bolts. The staff finds the use of A490 bolts to be more effective than ,

j expansion anchor bolts from an installation standpoint.

1 1

Review of Design Criteria HVAC Ducts and Supports

! The staff review of the calculations for both the standard HVAC design details and the special (non-standard) cases found the design stress allowables

acceptable and in accordance with the AISC Code. However, the staff review of j the design criteria, DC-ME-1C-CP, " Design Criteria for Design

Verification of Structures and Components in Category I Buildings", found that inSection2.11(HVACDuctsandSupports),theductworkallowablestress(for bending + membrane) in the faulted condition was shown to be 1.5 Sy. The staff

) could not find an adequate basis for allowing the 1.5 Sy limit. However, l because the design calculations did not use the 1.5 Sy allowable but rather used a conservative allowable of 18 ksi (or 0.6 Sy) for ductwork, the staff finds the design to be acceptable. S&L stated that the design criteria would be changed to reflect the design' criteria actually used.

In a letter from D. I. Herborn to A. Schwencer dated March 12, 1984, the applicant provided the staff with its revised design criteria for HVAC design.

j The applicant comitted to meet the intent of the staff's position on the HVAC j ductwork stress limits for the faulted condition by revising the design j criteria to limit the HVAC ductwork stresses to 0.95 of the material yield

! stress (Sy). The revised stress limits for ductwork and duct supports were

) included in the new FSAR paragraph 3.9.3.5 and Table 3.9-14. Furtherinore, the Sargent and Lundy design criteria, DC-ME-16-CP, noted above, was also revised

]

j to reflect the revised allowable stress limit.

Based on.the revisions to the design criteria and FSAR and because the design l calculations did not actually use the 1.5 Sy stress limit as indicated above, I

i

the staff concludes that the design criteria used for the HVAC ductwork and supports are acceptable.

HVAC Review Performed at the Clinton Plant Site On December 9,1983, the staff met with representatives from Illinois Power Company, Sargent & Lundy, and the constructor (Baldwin Associates) at the Clinton plant site. The purpose was to visually inspect several areas of the plant where safety-related HVAC systems are installed in order to gain a better understanding of tha HVAC structural design and to identify any critical areas where the design assumptions could be potentially deficient.

The main areas on which the staff focused included:

1) the HVAC system above the control room, and

2) the areas where there were long, vertical risers of HVAC ductwork.

The staff review of the control room HVAC supports found no use of expansion anchor bolts. All the HVAC supports were found attached to the structural building steel with 1-inch diameter A490 bolts. The 14 ft.

horizontal span appeared to have been met and there were no long vertical risers in the area.

The Control Building at elevation 825 ft. however, had several long, vertical risers of HVAC ductwork. The staff noted that local buckling of the sheet metal occurred at the top of the riser where the support was located. An inspection of the mirror-image ducting on the other side of the building revealed the same type of failure. The applicant stated that the failure was probably caused by installation. However, the applicant stated that a review of the ductwork design would be performed to determine whether the failure could have occurred in other long, vertical risers of HVAC ducting.

Subsequently, in a letter from D. I. Herborn to A. Schwencer dated March 12, 1984, the applicant provided the staff with the findings from its review.

The applicant found that an inadequate duct gauge thickness combined with the HVAC installation sequence caused the buckled condition. The two long vertical

4

. l ducts were constructed of 22 gauge material instead of the 18 gauge material specified in design.

j l Sargent and Lundy evaluated the stress in the two ducts and found that the f analysis predicted a buckling of the 22 gauge material. However, the analysis j also showed that buckling of the specified 18 gauge inaterial under similar

} construction methods would not have occurred. In its review of the design i calculations for the two HVAC risers, Sargent and Lundy verified that the l ductwork can withstand all nonnal and accident loadings, including the loading due to the construction sequence used for long vertical risers. A program has

}i j been initiated to replace the buckled 22 gauge material with the proper 18

]

guage material. Furthermore, all vertical risers will be reviewed for proper i material gauge.

1 l Based on the actions taken by the applicant as discussed above, the staff

finds that our concerns have been adequately addressed.

l *

! Staff Conclusions l

In nur review of the typical design margins available in the structural design of the HVAC ductwork and supports at the Clinton plant, the staff has i found that many conservative assumptions are used in the generic calculations which qualified the standard HVAC ductwork and support design details. For

the special (non-standard) designs, the staff also found the use of conserva-j tive assumptions in the design calculations. Based en these conservatisms, the l

l staff concludes that there is an inherently large margin between the stress or j

load level that was calculated for nonnal and design basis accident conditions '

j and the stress or load level that would result in actual structural failure t

) and loss of functionality of the HVAC ductwork and support systems. Therefore, the staff finds that the overall structural design of the Clinton HVAC systems j is adequate and provides a sufficient margin of safety to failure under normal and design basis accident conditions.

1 s

l .

! Attachment A Tabulation of Calculated vs. Allowable Stresses Calc No.: CQD-008566 (Rev.0)

Ratio of Calculated Stress to Hanger No. Description '

N Allowable Stress

P-4007 L li x li x i 0.52(bending)

L li x li x i 0.53(shear) weld 0.40 weld 0.39 A-0011 L li x li x i 0.94 (bending)

L li x Il x i 0.93(shear) weld 0.63 weld 0.61 i R-7036 L2x2xi 0.61 (bending) l L2x2xA 0.46

C 4 x 5.4 0.30(bending)

! C 4 x 5.4 0.23 (shear) i weld 0.27 weld 0.49 A490 bolt 0.35(shear) l A490 bolt 0.31(tension) 1 i

o I

I l

l .

i i

I

.. . _ = _ _ - - . - - _ _ _ _ _ - . . - . _ =

Attachment B i

Document Reviewed

1) Design Specification K-2910. "HVAC Work," Amendment 7.

2) DesignGuide,"HVACDuct-HangerDesignHandbook,"(Revision 10).

3) Design Calculation, " Design of Safety-Related HVAC Duct-Hanger System,"

EMD-019620, dated 9/28/80.

4) Project Instruction, PI-CP-014. " Request for Special HVAC Duct-Hanger Analysis," Revision 2, dated 11/20/81. .

5) Project Instruction, PI-CP-031 "HVAC Seismic Duct-Hanger Design,"

Revision 0, dated 7/8/82.

6) Design Criteria, DC-ME-16-CP, " Design Criteria for Design Verification of Structures and Components in Category I Buildings," Revision 0, dated 10/24/80.

7) Drawing No. M 14 D-2115. Revision C, dated 11/15/83(SupportNo.D-2115) 4

8) Design Procedure " Generic Procedure for Seismic Qualification of HVAC Ducts," Calc. No. CQD-009727.

9) Special Analyses, Calculation No. CQD-008566 Revision 0, dated 7/15/83.

I

\

'