ML20090E627

Revised Incident Rept,Turkey Point Unit 3,Safety Valve Headers,Dec 1971

ML20090E627
Person / Time
Site:	Turkey Point
Issue date:	03/02/1972
From:	FLORIDA POWER & LIGHT CO.
To:
Shared Package
ML20090E626	List: ML20090E627
References
NUDOCS 8304040157
Download: ML20090E627 (5)

Text

. . .._ _ .._. .._. . . _ . .

I s

3 t

4 l'

4 i

1 4

i 4

i i

l 1

INCIDENT R F 'P' O R T

] Turkey Point Unit 3

)

.l Safety Valve 11endera i

)

i December 1971

],

' Revised March 8, 1972 i

3 l

4 t

4 i,

J e

/

t

'I I

4 9304040157 760225 PDR ADOCK 05000250 PDR S -

__ s 4.3 PIPE STRESS ANALYSIS The stresses in the header pipe material in the region of failure around the weldolet attachment are listed in Table 4.3-1 for the design pressure con-dition with the safety relief valves closed. The stress levels are wfthin the code allowables. i l

l l

Table 4.3-2 lists the stress levels for the design pressure condition with the safety valve full open. The stresses caused by the reaction force of 20,000 pounds are included. The actual stresses are considerably greater than the ultimate tensile strength of the material, and indicate that the header would fail due to the forces caused by the relieving of a safety valve.

The valve reaction force would also be significantly higher than the 20,000 pounds if the potential force contribution by the condensate accumulation in the mein steam line loop "A" is inc3uded. It is estimated that the re-action force could increase.to 28,000 or higher, for fully open valve re-lieving wet steam. This condition would further increase the stresses in the header.

As shown in Figure 4.3-1, the support saddle assembly is welded to the safety header at two points 24-inches apart. While the saddle itself slides on'a lubrite plate imbedded in concrete, the rigid attachment of the saddle weldment to the pipe at two points resists the thermal growth of the safety header. The stresses developed by this support arrangement have been calculated and are shown in Table 4.3-3. ,

The orientation of the slotted holes in the bearing plate is determined by -

a simultaneous growth of the steam line in a westerly direction, and of the ateam header in a north-south direction. During the hot functional test, the main steam isolation valves were closed, thus preventing a steam flow down-istream'oftheisolationvalves. This would restrict east-west movement 4

due to thermal expansion, and impose stresses on the safety' valve header until the time that the bolts fail in shear. The stresses would then be-relieved.- These stresses are also shown in Table 4.3-3.(Column'3) -assuming'.

\

~

that at the construction stage,.the. bolts were located in the slots at the most adverse position.

4.3-1L Rev.~1 - 3/8/72 v w en s -E"4<

The saddleplate welded on the pipe is 1" thick. The pipe itself has 0.56" (m) thick wall material. The presence of a thick saddle plate causes an intensification of pressure induced stresses just beyond the saddle. These stresses are also shown in Table 4.3-3 (Column 4).

All calculated stresses in the table are based on an elastic analysis. The material yielding will prevent development of this level of stresses in the pipe. The standard yield stress for SA 160-Grade B is 34,000 psi minimum.

Actual tensile tests on the pipe material indicate a yield point of 43,000 psi at room temperature. For elevated temperatures, the stress-strain curves obtained from tests by Battelle indicate a yield stress of 34,000 at 500 F for the pipe material, and 42,750 psi at 400F for the support material (see Figures 4.3-2 sheets 1 through 4).

Using the actual acress-strain curves of Figure 4.3-2, the safety factor based on ultimate resisting moment is 2.3 for the support material, and 1.2 '

for the pipe material and would probably be higher if based cn the actual

() position of the bolts in the foundation slots. The support design would cause high local stresses, but would not cause. pipe failure.

b

-\d 4.3-2 Rev. 1 - 3/8/72

C .. ..

r TABLE 4.3-2 STRESSES IN ORIGINAL SAFETY VALVE HEADER (EXISTING AT TIME OF THE INCIDENT)

SAFETY VALVE FULY,OPEN I

Pressure Hoop- 13,100 Longitudinal- 5,200 Thermal Dead Load Longitudinal 100 Seismic

• Valve Reaction Force

• Bending 188,500**

Torsion (inboard of valves, 7,800 2 valves blowing)

Ultimate Tensile Strength at 550F (Lab Test) 78,000

• Not included, since this mode did not exist at the time of the incident.

• • This is based on the minimum values from Biljaard's Method and a stress concentration factor of 1.3. If'the relies .sg force is taken as 28,000 lbs., this value will increase by 40%.

'O a

-Rev..1 - 3/8/72

• ~

. i

, L T

3 6.0 PLANT DESIGN REVIEW I 6.1 PIPING REVIEW A complete review has been made of all safety and relief valve applications ,

l on the project. Changes have been made in the header design, wall thickness, [

and/or support system to assure that the external loads caused by the valve reaction forces are adequately accounted for and incorporated in the design. .

f A thorough design audit has been performed to ascertain the completeness of r I piping stress analysic for the loads stipulated in the project design criteria. Additional stress analysis and support modifications have been made as required.

A team of desiga engineers has carried out a jobsite audit to assure that the intent of design such as the design assumptions, restraint / support locations, details, and clearances, etc., were sustained through the installation phase.

The design office audit, and the jobsite survey covered all the safeguard systems as well as other significant piping systems in the plant within Bechtel's scope of responsibility.

It is to be noted that the steam traps provided-to drain condensate from the -

bowl of the main steam isolation valves normally function automatically and i prevent accumulation of condensate in the steam line as long as the drain to the cor.denaer is not isolated. For the period of time that the condenser may not be available, a local drain has been added so that the steam trap may-continue to operate in the automatic mode.

E 6.1-1: :Rev. 1.- 3/8/72 -

' ,