ML17262A451

From kanterella
Jump to navigation Jump to search
Devonrue Calculation Re Plant Atmospheric Relief Valve (ARV) Area Ambient Temp rise-revised Methodology
ML17262A451
Person / Time
Site: Ginna Constellation icon.png
Issue date: 04/10/1991
From:
ROCHESTER GAS & ELECTRIC CORP.
To:
Shared Package
ML17262A448 List:
References
NUDOCS 9104300401
Download: ML17262A451 (48)


Text

RG&E Answers To The NRC's March 21, 1991 Station Blackout Questions Atmospheric Relief Valve Area Heatup Calculations 9104300401 910422 PDR ADOCK 05000244 P

PDF<

Devonrue Calculation Cover Sheet Project No. 8-9025.00

Subject:

ARVArea Ambient Temperature Rise During SBO Description This calculation evaluates the ambient temperature rise in the ARV area during a four hour station blackout. The calculation considers heat sources and heat sinks present in this atria and provides a methodology for considering a concrete and metal ceiling as a heat sink surface area.

The effects of opening a door to the mezzanine level of the Turbine Building is also evaluated.

NOTE:

DUPLICATE COPY RECONSTRUCTED FROM MEDIA; ORIGINAL INADVERTENTLY DESTROYED.

COVER SHEET SIGNATURES ARE BASED UPON EDIA COPY.

Prepared by Reviewed by

@WC

.E Approved by

~~~

Date Numberof Pages ~~

~/Yf Appendices Attachments

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 1 of24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology No computer calculations have been utilized in this analysis. Excel spreadsheets have been utilized in the calculation of surface temperatures of insulated piping and in the calculation of total heat generated from hot surfaces.

However, the formulas are provided in this calculation and a numerical check has been made ofthe accuracy of the spreadsheet calculations.

1.

The perimeter of the Intermediate Building is assumed to be as shown on Ginna Station Drawing 33013-2121, Rev. 0 2.

The normal max. ambient temperature in the Intermediate Building is assumed to be 104'F, consistent with the NUMARC87-00 Section 7 methodology and the maximum auxiliary building temperature listed in Ginna FSAR Table 3.11-1.

The normal maximum ambient temperature of the Containment Structure just adjacent to the South wall of the Intermediate Building is assumed to be 120'F as stated in Ginna FSAR Table 3.11-1.

The South wall of the ARV area is assumed to be poured concrete as shown on Ginna Station drawing 330103-2121, Rev. 0. The North, East, and West walls of the ARV area (the northern portion of the Intermediate Building) are solid concrete block as shown on Ginna Station drawing 330103-2121, Rev. 0.

The walls are assumed to behave in a similar manner to poured concrete walls and willbe treated similarly in this calculation.

5.

The ceiling of the ARV area is 5" thick pouted concrete, integrally bonded with 20 gauge fenestra holorib decking as shown on sketch D-523-22 and reproduced on Attachment 1.

The calculation willtn:at this construction as 5" of concrete with an inside layer of0.036" of steel and willuse the thermal properties of steel and concrete reported in the ASHRAE handbook of fundamentals.

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 2 of24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology 6.

Thermal insulation on piping is assumed to be of the type and thickness defined in RG&E Specification ME-269, Revision 0.

7.

Heating Steam piping and equipment is conservatively assumed to be normally operating at a temperattue of220'F.

8.

Under SBO conditions, immediately following reactor trip and MSIVclosure, only one Safety Relief Valve on each steam header willliftto remove decay heat as described in Ginna UFSAR Section 7.4.1.3.

Once the operators are using the ARVs to remove decay heat, it is assumed that only one ARV willbe operating.

Ginna UFSAR section 7.4.1.3 states, "One.atmospheric steam dump, which can be operated from the Control Room is sufficient for maintaining hot shutdown or to achieve cooldown of the reactor coolant system below hot shutdown conditions."

9.

The ARVs ate assumed to be actuated early enough to allow SRV tailpipe cooldown within the first hour ofthe event.

Any additional assumptions are noted in the body ofthe calculation.

1.

NUMARC87-00, "Guidelines and Technical Bases for NUMARCInitiatives Addressing Station Blackout at Light Water Reactors," including Appendix E, Appendix F and the Appendix F Topical Report, November 1987.

2.

RG&E Technical Specification, ME-269, "Pipe, Duct, and Equipment Insulation, Ginna Station," draft issue Revision 0, dated January 30, 1989.

3.

Ginna Station Plant Arrangement Drawing No. 33013-2121, Rev. 0, Cont Structure &

Intermediate Bldg Plan - Oper. Flr. El. 278'-4" & 274'-6".

4.

Ginna Station Plant Arrangement Drawing No. 33013-2129, Rev. 0, Intermediate Bldg Plans Above Elev's.293'-0", 298'-4", & 315'-4".

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 3 of 24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology 5.

Ginna Station PAID Drawing No, 33013-1231, Rev. 13, Main Steam 6.

Ginna Station P&IDDrawing No. 33013-1915, Rev.4, Heating Steam 7.

Ginna UFSAR Table 3.11-1, "Environmental Service Conditions for Equipment Designed to Mitigate Design-Basis Events."

8.

Ginna UFSAR Section 7.4.1.3, Revision 4, dated December, 1988 9.

ASHIME 1985 Handbook of Fundamentals, Chapter 39, Table 3, "Properties ofSolids" 10.

Kreith, Frank, and WilliamBlack, B i H Tr n f r, 1980.

11.

Incropera, Frank P, and David P. DeWitt, In n

H Tr n f r,1979.

The objective of this calculation is to determine the ambient temperature rise in the ARVatria during a four (4) hour SBQ-induced loss of ventilation.

This calculation uses a modification of the NUMARC87-00 section 7 methodology that accounts for the ARVama ceiling as a heat sink. The section 7 methodology is based on the assumption that pored concrete walls willact as heat sinks and that their surface temperatures remain essentially constant over the course of the 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> duration. However, the constant temperature assumption may not be valid for the ceiling because the ceiling is not as thick and has a different construction from the concrete walls. Therefore, the temperature rise in this surface must be considered ifcrediting it as a heat sink, the simplified methodology provided in NUMARC87-00, section 7, cannot be applied ditectly.

In considering the surface temperature rise in the ceiling from heat generated within the ARV room, the heat input to this ceiling from the atea above it is also accounted for.

The temperature of the ceiling after 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> is calculated by performing an energy balance:

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 4 of 24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology Q = rho cp U dT/dt, where Q = heat generation rate rho = density of the heat sink material cp is the speciflc heat at constant pressure of the heat sink material dT/dt = the temperatute change of the wall per unit time, or for this case; (Tflnal - T initial)/4 hours Once the surface temperature rise is computed for the ceiling, itwillbe considered in the simplified NUMARC87-00 equation as follows:

Tair = Tw + [Q/A] 3/4 where Tair = the resultant ambient air temperature in the ARVarea; Tw = the wall temperature after 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> computed as a weighted average of each wall or ceiling prorated on the basis ofsurface atea; Q = heat generation rate;"

A = the total surface area ofwalls and ceilings acting as heat sinks.

The following steps shall be performed to accomplish the described method for determining the ARVarea temperanue:

1.

Conservatively estimate the heat generation rate, Q, in the ARV area and in the room above. The heat generation rate willbe calculated by considering heat rejected from hot piping and equipment and using the methodology supplied in NUMARC 87-00,

- Section 7.

2.

Calculate the surface area of each wall and ceiling acting as a heat sink and determine the proportion represented by each.

3. Using flat plate heat transfer correlations, the amount ofheat transferred into the ceiling from the area above the ARV area willbe calculated.

The amount of heat transferred from the ARVarea willbe calculated based on the proportion of total heat sink surface area in the ARVroom represented by the ceiling

0',

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 5 of 24

'ubject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology

4. Determine the ceiling temperature after four hours using the equation Q = rho cp V dT/dt. The density and specific heat of the composite concrete/steel structure willbe considered in this determination.

The calculated ceiling temperature after four hours of heating will be conservatively assumed throughout the station blackout transient for the purposes ofcalculating ARVarea temperature.

5. Determine the wall temperature Tw for use in the equation Tair = Tw + I,Q/A] /.
6. Determine the ARVroom temperature after 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> using the above equation.

7.

Consider the effects of opening doors using the relationship defined in Section 7 of NUMARC87-00.

Step 1:

Determine the heat generation rate, Q, in the ARV area and in the area above.

The major heat souice in each ofthese areas is hot piping and equipment. A physical inspection of each area was performed to identify the relevant heat sources and define their characteristic dimensions such as diameter and length. In the case of insulated surfaces, the surface temperatures must be calculated.

As stated in the assumptions section, the insulation type and thickness is assumed to be as specified in ME-269, Table 1, except in the case of the HHCC Tank where the insulation thickness was field measured.

In accordance with ME-269, all insulation is treated as calcium silicate.

The surface temperature is calculated by considering the convective heat transfer between the insulated surface and the air:

q = h A dT = h (Ts - T

) 2 z rs I where h is the unit thermal surface conductance (Btu/hr. sq ft 'F), rs is the radius of the insulated

surface, 1 is the pipe length, Ts is the insulated surface temperature, and T~ is the ambient air temperature

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 6 of 24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology t

T.

the conductive heat transfer from the pipe surface through the insulation; q = (Ti - T~)

  • 1/ ( [ In (rs/ri)/2nkl ] + 1/Zxrslh}

where Ti is the pipe surface temperature and ri is the pipe radius.

Combining the 2 equations yields; Ts = ( 1/h rs * (Ti - T~) ) / ([ In (rs/ri)/k ] + 1/rsh) + T~

In this evaluation, the followingconstant values are applied:

T~ = 104 'F (the initial room temperature.

This is conservative since heat transfer from pipe surfaces willdecrease as the room temperature rises.)

h = 1.6 Btu/hr sq ft 'F (table 4.4.11 of Mark's handbook. Attachment 2 to this calculation) k = 0.045 Btu/hr ft 'F (Table 4.4.6 of Mark's handbook: Attachment 3 to this calculation)

For purposes of this evaluation, k is assumed to be constant over the range of temperatures in question.

"h" has been selected from Attachment 2 on the basis of large horizontal pipes.

This value is felt to produce reasonable results for this calculation.

The resultant surface temperatures calculated for the heat sources identified in the ARVarea and the area above are provided in Tables 1 and 2, respectively, shown on the followingpages.

h t

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 7 of 24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology Description Tl ('F)

Circ.

rl (ft)

Ins.

rs (ft) ln(ra/rl)

Ts ('F)

(rt) thk.

(ft) 30" MS header 550 6" SRV inlet 550 6" ARV inlet 550 14" MRW header 450 HHCC Tank 220 6" stm to AFWTh 550 1.25 0.25 1.5 0.182322 148.31 0.276 0.2083 0.4843 0.562304 148.47 0.276 0.2083 0.4843 0.562304 148.47 0.276 0.2083 0.4843 0.562304 148.47 0.583 0.25 0.833 0.356846 136.65 1.625 0.25 1.875 0.143101 117.72 Heating Stm pipe 220 3

0.3945 0.083 0.4775 0.190961 133.64 220 1.25 0.1159 0.083

'0.1989 0.539917 130.45 220 0.958 0.0695 0.083 0.1525 0.786069 128.48 220 1.29 0.1223 0.083 0.2053 0.517963 130.63 TABLE 1 SURFACE TEMPERATURE OF HEAT SOURCES IN THE ARV AREA>

Rcviewcr's note:

a check of the Table 1 and Table 2 T, values show that they arc high by approximately 3'F. This is partially compcnsatcd for in Tables 3 and 4 by the apparent use of Ta and Tw values that are high by 2'F. The net effect is to slightly ovcrcstimate the heat generation rate in the room, which is conservative.

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 8 of 24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology Description Tl ('F)

Circ.

rl (ft)

Ins.

rs (ft) ln (rs/rl)

Ts ('F)

(ft) thk.

(ft) 30" MS header 550 14" MFWhdr 450 8" TB Exh. line 240 10" SRV tailppe 550 12" ARV tailppe 550 1.25 0.25 1.5 0.182322 148.'31 0.583 0.25 0.833 0.356846 136.65 0.417 0.208 0.625 0.404665 151.33 0.5 0.25 0.75 0.405465 144.5 0.333 0.125 0.458 0.318727 128.49 heating stm line 220 1.083 0.0894 0.083 0.1724 0.656886 129.48 220 1.667 0.1823 0.083 0.2653 0.375788 131.89 220 3

0.3525 0.125 0.4775 0.30354 125.36 TABLE 2 SURFACE TEMPERATURE OF HEAT SOURCES LOCATED IN THE AREA ABOVE THE ARV AREA

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page9of 24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology Note in the case of heating steam pipe where only the insulated circumference could be measured, the followinghas been applied:

Measured Circumference (ft) = 2 x rs therefore, rs = Measured circumference /2 z and ri = rs - insulation thickness Using the results from Tables 1 and 2, the heat generation rate, Q, for each area can be calculated from the followingequation found in Section 7 ofNUMARC87-00:

Q

( 01 [ 04 + 157(Ts. Ta)l/6 Dl/2

+ 1703(Ts Ta)l/3 D ] (Ts - Ta) +

1.4E-7 D (T4s - T4w) ) L Tables 3 and 4 present the resultant heat generation rates for the ARVarea and the area above the ARVarea, respectively.

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 10 of24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology Description Ts ('F)

Ts ('K)

L (ft)

L (m)

Do (ft)

Do (m)

Ts-Ta Ts4-Tw4 Q

(watts) 30" MS linc 6" SRV inlet 10" SRV tailpipe 6" stm to AFWrb 6" ARV inlet 12" ARV tailpipe 14" MFWhcada HHCC Tank Heating stm Heating stm lieating stm Heating stm 148.3 337.76 113 34.4 148.5 337.87 16 4.88 225 380.37 30 9.14 148.5 337.87 28 8.53 148.5 337.87 10 3.05 500 533.15 5

1.52 136.7 331.32 110 33.5 117.7 320.76 8.5 2.59 133.6 329.59 19 5.79 130.4 327.82 3

0.91 128.5 326.76 5

1.52 130.6 327.93 25 7.62 3

0.914 0.968 0.295 0.833 0.254 0.97 0.296 0.97 0.296 1

0.305 1.67 0.509 3.75 1.143 0.995 0.303 0.398 0.121 0.305 0.093 0.411 0.125 22.94 23.05 65.55 23.05 23.05 218.3 16.5 5.941 14.77 13 11.94 13.11 3.195E+09 3.212 E+09 1.111E+10 3.212E+09 3.212 E+09 7.098 E+10 2.23Et09 765876836 1.981E+09 1.728E+09 1.5&E+09 1.744 E+09 51079.847 2420.91 15038.544 4245.0369 1516.0846 15736.817 18543.084 895.55073 1690.7606 95.00654 111.09859 824.72502 Tr(oc)=

92.4 TF(oF)=

198 Total Q (watts)=

SA (sqm)=

597.4 112197.5 TABLE 3 HEAT GENERATION RATE ARV AREA

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 11 of 24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology Description Ts('F)

Ts('K)

L(lt)

L(m)

Do(ft)

Do(m)

Ts-Ta Ts4-Tw4 (watts) 30" MS hdr 14" MFW hdr 10 SRV (Ins) 10" SRV (unin) 12 ARV (ins) 12 ARV (unins) 8" TB cx linc heating stm heating stm heating stm 148.31 136.65 151.33 225 144.5 550 128.5 129.5 131.9 125.4 337.767 331.289 339.444 380.372 335.65 560.928 326.761 327.317 328.65 325.039 133 40.5 3

0.914 22.95 45 13.7 1.67 0.509 16.47 13 3.96 1.25 0.381 24.62 21 6.4 0.083 0.253 65.55 6.5 1.98 1.5 0.457 20.83 10.5 3.2 1

0.305 246.1 17 5.18 0.916 0.279 11.94 32 9.75 0.345 0.105 12.5 32 9.75 0.531 0.162 13.83 32 9.75 0.955 0.291 10.22 3.196E+09 2.226 E+09 3.456E+09 1.111E+10 2.872E+09 8.918 E+10 1.58E+09 1.658E+09 1.846E+09 1.342E+09 60138.9 7569.84 2737.52 10490.3 1322.69 39540.7 1072.6 844.035 1437.9 1733.95 Total Q(watts)=

126889 TABLE 4 HEAT GENERATION RATE AREA ABOVE THE ARV AREA

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 12 of 24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology Tables 3 and 4 show that the SRV tail pipes were considered to have a surface temperature of 225

'F. This value is based upon the followingjustification:

As previously stated in the Assumptions section, itis reasonable to assume that only 1 of 4 safety valves in each Main Steam header willbe automatically cycled upon MSIVclosure following reactor trip in response to the LOOP. In order to apply the NUMARC 87-00 correlations, the two affected tail pipes will be treated as 15 ft. each (30 ft. total) of uninsulated 10" diameter piping.

The set pressure of the first safety valve is 1085 psig per Ginna UFSAR section 10.3.2.4.

The steam saturation temperature at this pressure is 556 'F. The safety valve is treated as a throttling device which acts as a fiowrestrictor, leading to a pressure drop in the fluid. It is reasonable to assume that there is no change in enthalpy across the safety valve since the valve inlet and the tail pipe exit to atmospheric pressure are well separated and the exit is relatively far downsueam from the valve itself. Therefore, considering the release ofsteam through the safety valve to be isenthalpic and reducing in pressure to essentially atmospheric conditions, the resultant temperature on the constant enthalpy curve of the Mollierdiagram (Attachment 4 to this calculation) is approximately 300'F.

The SRV tailpipes ate expected to remain at this temperature for only a short period oftime (less than 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />), since the ARVs willbe used for cooldown and depressurization shortly after the onset ofthe event, precluding further SRV actuation.

In order to use a steady state temperature calculation based on the NUMARC 87-00 methodology, a constant 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> heat source term must be used.

The SRV tailpipe heat source contribution is approximated by selecting a constant 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> tailpipe temperature that willprovide a heat source contribution that conservatively approximates the contribution from the actual brief period at high temperature.

Since heat generation, Q, is directly proportional to the product of the temperature difference (between the hot surface and the ambient air) and time (dT*t),the followingis reasonable:

Since Q (300'F - 104 'F)

  • 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> ( Q' (200 'F - 104'F)
  • 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />, a surface temperature of225 'F is chosen for the tailpipes and willproduce conservative results.

0 0

Devonrue Calculation Project No. 8-9025.00 Date; April9, 1991 Page 13 of 24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology As indicated in Table 3, the surface temperature of the uninsulated 12" ARVtail pipe is assumed to be 500'F when calculating the heat generation rate in the ARVarea. 'Ihis is felt to be conservative since steam header temperature at the time when ARVoperation begins willbe somewhat less than the 550'F Main steam system design temperature.

The ARVs willnot be in operation until the operators take manual control of them.

Therefore, the tail pipe is not expected to see steam conditions for the fullfour hours. In addition, the ARVs willbe used to remove decay heat as well as cooldown the plant.

Since the coping scenario involves plant cooldown, steam header temperatures willdecrease below the hot standby operating temperattue over the course of the four hours and the temperature of the steam seen by the ARV tailpipe willalso decrease.

On this basis, the use of a 500'F tailpipe operating temperature is judged to result in a conservative heat generation rate.

Step 2

.Calculate the surface area of heat sinks The straight wall lengths on the North, East and West sides of the room are scaled from the referenced arrangement drawings (see figure).

3.05 3.04 3.35 2.74 28.04 10.06 7.92 33.33 6.40 North Portion of Intermediate Building, El. 278'-4"

Devonrue Calculation Project No. 8-9025.00 Date: April9 1991 Page 14 of24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology Using the dimensions from the figure above, the perimeter of the room is calculated by summing these lengths.

Perimeter

= 97.93 meters The Surface Area of the walls, Awalls, is computed by multiplying the perimeter by the height of the walls, Review of the referenced drawings 33013-2121 and 2129 show that the height, h, is:

h = el 298'-4" - 278'4" = 20' 6.1 meters therefon:,

Awalls = 97.93

~ 6.1 = 5974 sq. meters Since this wall will need to be treated separately later in this calculation, the area of the wall adjacent to the containment wiHnow be calculated as:

Acont = 33.33

~ 6.1 = 203.31 sq. meters The remaining wall area is then:

Aremain = Awalls - Acont. = 597.4 - 203.31

= 394.09 The area of the ceiling is calculated conservatively as 3 rectangles using the values Qom the figure above:

Aceiling = (28.04

  • 10.06)

+ (3*9) + (3*8)

Aceiling = 333.1 sq. metersno'c 2

Reviewer's note: a review of the drawing shows that the containment wall represented by the curved line in the figure actually extends upward farther than indicated. Thus, thc 28.04x10.06 rectangle takes credit for some area that is actually in the containmcnt rather than in the intermediate building. When this is accounted for, thc ARV ceiling area is approximately 244 square meters.

Usc of a hrger ceiling area should not significantly affect thc ceiling temperature determined later in this calculation. A corrected ceiling area of244 square meters is used in thc final room temperature where usc ofa larger area would bc nonconservativc.

0

Devonrue Calculation Prospect No. 8-9025.00 Date: April9, 1991 Page 15 of24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology The total surface area acting as a heat sink is therefore, Atotal = Aceiling + Awalls = 333.1 + 597.4 = 930.4 m2 Using this information, the followingproportions are determined:

Proportion of the ceiling to total = 333.1 / 930.4 = 0.36 Proportion of the containment wall to total = 203.31 / 930.4 = 0.22 Proportion of the remaining walls to total = 394.09 / 930.4 = 0.42 Step 3

Determine the amount of heat transferred into the ceiling from the rooms above and below As can be seen from Tables 3 and 4, the heat generation rate in the ARVroom is 112,197.5 watts and the heat generation rate in the room above is 126,889 watts.

As previously described, the ceiling of the ARVroom is is 5" thick poured concrete, integrally bonded with 20 gauge fenestra holorib decking as shown on sketch D-523-22 and reproduced on Attachment 1 ~ The concrete surface faces the elevation above the ARVroom and the steel decking faces the ARVroom. The calculation willtreat this construction as 5" of concrete with an inside layer of0.036" of steel as shown in the Figure below:

Elevation above ARVArea 5 Inch thick poured concrete ARVArea 0.036 Inch thick steel ARV Room Ceiling

0

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 16 of 24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-R'evised Methodology The heat transfer rate into the concrete surface from the area above willbe calculated using fiat plate convective heat transfer correlations for the upper surface of a cooled plate (i.e., the plate is cooler than the air). The followingrelationships willbe applied from section 9.6 ofIn+opera and Dewitt using the dimensionless parameters of the Nusselt number, Nu, and the Rayleigh number, Ra:

Ra = fgB ( Ts - T~) L3 ] / v a Nu '= 0.27Ra</4 Nu = hc L/k; or hc-k Nu/L Q=hcAdT where:

Ts = the surface temperature ofthe plate T~ = the temperature ofthe air in the room B = lfl'fwhere Tf=(T~+Ts)/2 L = Surface area of the plate/ Perimeter g = the acceleration due to gravity, 9.8 m/sec2 As can be seen from the above equations, the air temperature in this room must be calculated to obtain T~.

T~ willbe calculated using the NUMARC87-00 equation, T~ = Ti + [Q/A]3/4 where T

= the resultant ambient air temperatutc in the room above the ARVarea; Ti= the initial temperature of the walls considered as heat sinks Q = heat generation rate; A = the total surface area ofwalls and ceilings acting as heat sinks.

From arrangement drawing 33013-2129, it can be seen that the perimeter of the room is virtually identical to that of the ARVroom, i.e., 97.93 meters. and that the height of these walls is 17 feet, or 5.2 meters.

The area of the ceiling in this room willbe neglected to ensure conservative results.

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 17of24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology It should be noted that the construction of the ceiling in this room is identical to the ceiling of the ARVarea.

'Iherefore, surface area of the walls = perimeter *height or.

A = 97.93

  • 5.2 = 509.24 sq. meters Assuming that the initial air temperature is 104'F or 40'C, and that the surface is in equilibrium with the air, T~ is:

(126,889 watts./509.24 sq meters) 3/4+ 40 'C T

= 62.7 + 40 = 1Q2.7'C or 217'F or 375.85'K KnowingT, Tfand B can be calculated based upon B = 1fffwhere Tf= g~+ Ts ) /2 Assuming that the initial surface temperature of the plate is in equilibrium with the air, 40'C or 313 15 K:

Tf (375 85.K + 313 15 K) / 2 344 5 K B Q QQ29OK-1 Calculating L= Surface Area /perimeter, where the surface area was calculated in step 2, L= 333 sq meters / 97.93 meters = 3.4 meters The constants v and a are taken from Table A.4 from Incropera and Dewitt (provided as to this calculation). The values for air at 350'F from the table willbe used since the air temperature is 344.5 'K as calculated earlier in this step.

Therefore, a = 2,99E-5 and v = 2.092E-5 Calculating the Rayleigh number, Ra, is the next step.

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 18 of 24

Subject:

Ginna Atmospheric Relief Ualve (ARV) Area Ambient Temperature Rise-Revised Methodology Knowing Ra= [gB (Ts-T~) L3]/vtx, Rais:

(9.8) (.0029) (62.7) (39.3) (1.6E9)

Ra = 1.12E11 Knowing that the Nusselt number, Nu, is 0.27 Ral/4, Nu = 156.2 Using Nu and k, taken from Attachment 5, as 0.03, hc can be calculated:

hc = (003) (1562) / (3.4) = 1.38 W/m2 'K The heat transfer rate into the plate from the room above is then calculated using the equation Q =

hc A dT as follows; Q (138W/m2 oK) (333 m2) (62,7 oK)

Q= 28,813.2 W into the concrete surface from the area above Having determined the heat transfer rate into the ceiling from the room above, the heat transfer rate into the ceiling from the ARU area itselfmust be calculated. A slightly different approach willbe used to make this determination, since the surface area of the ceiling is to be treated as a heat sink.

As previously calculated in step 2, the surface area of the ceiling represents 36% of the total heat sink surface area.

Therefore, 36% of the heat generated in the room willbe transferred to this surface or:

Q = 0.36 ( 112, 197W) = 40,390.92 W This the total heat fluxinto this surface is shown on the followingfigure:

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 19 of 24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology 28,813.2 W Elevation above ARVArea ARVArea 40,390.92 W The total heat fluxinto the ceiling is:

40, 390.92 W + 2S,S13.2 W = 69, 203.4 W Step 4

Determine the ceiling surface temperature after 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> The surface temperature ofthe ceiling after 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> willbe calculated using the relationship:

Q = rho cp V dT/dt, where Q = heat generation rate into the ceiling rho = density of the heat sink material cp = specific heat at constant pressure of the heat sink material dT/dt = the temperatute change of the wall per unit time, or for this case; (T final - T initial)/4 hours.

As previously calculated in step 3, the total heat transfer rate, Q, into the ceiling volume is 69,203.4 W. The total volume of the plate which is the ceiling is computed as follows: (note the following calculations are made in English as opposed to SI units for the convenience of the preparer)

Devonrue Calculation Prospect No. 8-9025.00 Date: April9, 1991 Page 20 of 24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology Volume = Surface Area of plate>> Thickness of plate V= (333 m2

  • 1 ft / 0.0929m2 )>> (5.036 in'2 in/ft.)

Vtotal = 1504.3 ft3 Because the ceiling is constructed of two materials, a composite density and specific heat willbe calculated based on the volumetric proportion of each material in the ceiling. As a first step, we must calculate the proportion of steel and proportion concrete to the total volume as follows:

Vconcrete/ Vtotal = (thickness concrete

  • surface area) /(total thickness
  • surface area) or.

Vconctete/ Vtotal = thickness concrete / total thickness, and similarly Vsteel / Vtotal = thickness steel / total thickness resulting in:

Vconcrete / Vtotal = 5 / 5.036 = 0.993 Vsteel

/ Vtotal = 0.036 / 5.036 = 0.007 Using these values the composite density and specific heat are calculated from the values for each individual material taken from the ASHRAE Handbook ofFundamentals:

rhoconcrctc 145 lb/ ft3 rhostcei = 487 lbs / ft3 cp(concrete) = 0.156 Btu/ lb 'F cp(steel) = 0.113 Btu/ lb 'F Therefore, the composite density and specific heat of the ceiling is:

rhoceiling = o993 (145) + 0.007(487)

= 147.4 Ibs / ft3 cp(ceiling) = 0.993(0.156)

+ 0.007(0.113)

= 0.1557 Btu/ lb 'Fn<<e 3

The next step is to convert the heat transfer rate into units consistent with the english units used in this portion of the calculation as follows:

Reviewer's note:

actually, the cp of the ceiling should bc mass weighted from the components rather than

~

~

volume weighted.

This results in a cp of 0.1550 rather than 0.1557, which affects the dT of the ceiling by approximately 0.1'F. The effect on thc final calculated room temperature willbc negligible.

mr

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 21 of 24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology Q = 69, 203.4 W ~ (1 Btu/hr. / 0.2931 W) = 235,939.5 Btu/hr Inserting all values into the equation Q = rho cp V dT/dt, yields:

235,939.5 Btu/hr = (147.4 lbs / ft3) (0.1557 Btu/ lb 'F) (1504.3 ft~) (dT/4hrs) which results in a dT = 27.3'F. Ifwe assume, as stated earlier in this calculation, that the initial temperature of the ceiling is in equilibrium with the air at 104'F, then after four hours, the temperature willbe:

Tceiling 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />

= 104'F + 27.3'F = 131.3'F = 55.17'C This temperature willconservatively be used as the ceiling temperature throughout the station blackout transient when calculating the ARVarea temperature rise.

Step 5

Determine the representative wall temperature, Tw In step 4, the surface temperature of the metal/concrete ceiling was calculated since it cannot be assumed to remain constant over the four hour heat-up scenario.

In addition to considering the temperature of the ceiling surface, the temperature of the South wall which is adjacent to the containment structure must be considered.

The South wall surface temperature inside the Intermediate Building will be calculated to be the average between the normal Containment maximum temperature and the Intermediate Building normal maximum temperature:

T(South wall) = (120'F + 104'F) / 2 = 112'F The other walls in the Intermediate Building willbe assumed to experience no temperature rise during the four hour loss ofventilation and are therefore assumed to be at 104'F.

The wall temperature to be used in the equation, Tair Tw + [Q/A]3/4 willbe computed as a weighted average of each wall or ceiling prorated on the basis of surface area as follows:

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 22 of 24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology From step 2, the followingproportions have been calculated:

Proportion of the ceiling to total = 333.1 /930.4 = 0.36 Proportion of the containment wall to total = 203.31 /930.4 = 0.22 Proportion of the remaining walls to total = 394.09/930.4 = 0.42 Therefore, Tw is calculated as follows:

Tw = 0.36(131.3'F)

+ 0.22(112'F)

+ 0.42(104'F)

Tw = 115.6'F = 46.4'C Step 6

Determine the ARV Room Temperature As previously stated, the ARV room temperature is calculated based on the simplified equation

'supplied in Section 7 ofNUMARC87-00:

Tair Tw + [Q/A] 3/4 where Tair = the resultant ambient air temperature in the ARVarea; Tw = the wall temperature after 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> computed as a weighted average wall or ceiling prorated on the basis ofsurface area; Q = heat generation rate; A = the total surface area ofwalls and ceilings acting as heat sinks.

of each Substituting the values determined throughout this calculation, the resultant air temperattue is:4 Tajr = 46.4 C + (112ql97 / 841.3)3/4 Reviewer's note:

thc corrected total wall areas arc used for thc remaining calculations, as discussed in footnote 2.

a I

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 23 of 24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology Tair = 39.2 + 46.4 = 85.6'C = 186'F These results show that after four hours the temperanue in the ARVaa:a willrise to 186'F.

Step 7

Calculate the effects of openings doors NUMARC87-00, section 7 provides a methodology forcalculating the effects ofopening doors to allow the removal of heat from an area through natural circulation. However, Appendix E to this document clarifies that the methodology presented in Section 7 has been tested only within the followingparameter ranges:

24,000W < Q < 100,000W 0 C<dT<50'C Examining these ranges we find that Q just slightly exceeds the limitat 112,197W, but that dT is within the range at (85.6 - 40)= 45.6'C.

Based on engineering judgement, the correlations presented in Section 7 willbe applied in this case even though Q exceeds the parameter range.

Since Q only exceeds the tested parameter range by 12%, the effect on the accuracy ofthe results is judged to be minimal.

The Section 7 methodology consists of the development of F, the door factor using the following equation:

F = H3/2 W, where H is the door height (m) and W, the door width (m)

Once the door factor is calculated it is applied as follows:

Tf= 4 + Tw + [(}

/ A3/

+ 16.18 (F)0.8653]

The F factor for the ARVroom is based upon opening the door to the Turbine Building on the East side of the North wall. As shown on the arrangement drawing 33013-2121, this door opens to the

Devonrue Calculation Project No. 8-9025.00 Date: April9, 1991 Page 24 of 24

Subject:

Ginna Atmospheric Relief Valve (ARV) Area Ambient Temperature Rise-Revised Methodology Turbine Building mezzanine, a large open area with roof ventilators without steam piping, which is not expected to experience any significant heatup during the loss ofventilation event. Therefore, it can be assumed that this area willhave an ambient temperature ofno higher than 104'F or 40 'C at the time the door is opened.

In addition, the door to the south portion of the Intermediate Building could be opened which leads to a stairwell. Due to the relatively small volume of the south portion of the Intermediate Building as compared to the ARv area (north portion), the effects ofopening this door willnot be calculated.

However, it is expected that opening this door to the stairwell could have beneficial effects.

From the referenced arrangement drawings, the dimensions of this door are found to be 3 ft wide by 7 ft. high. Converting these dimensions to meters, the F factor is found to be:

F = (2.13)3/2 (0.914)

= 2.85 Applying this value in the above cited equation yields:

Tf = 4 + 46.4 + (112,197)3/4 / [(841.3)3/

+ 16.18 (2.85)0 8653]

Tf = 4 + 46.4 + 31.2 Tf= 81.6oC

= 178.9'F

C gf(Y~

gdfrcl~> T g.)8-<

46, TIP ~"~

>~ ~JoLoa.iS QEc.P

~,~ h~

(

p,~'

,k

~t

~

Bcrcnson finds better agrccm>>nt with Ih>> data if Ka 0.09.

For very small wires, th>> heal flux will exceed that predicted by this flat plate formula. A reliable prcdiction of Ihc crili>>al temperature is not availablc.

For nucleate boiling accompanied by for>>cd convection, Ibc heat flux may bc approximated by Ihe sum of thc heal flux for pool boiling alone and ihe heat flux for forced convection alone.

This proecdurc will nol be satisfactory al high qualities, and no satisfaclory corrcbtion exists for the maximum heal flux.

For a given liquid and boiling prcssure, thc nature of thc surface may substantially influcnce the flux al a given (AI)~

Tabl>> 4.4.l0. These data may bc used as rough approximations for a bank of submerged tubes. Film coeliicients for scale deposits arc given in Table 4.4.g.

For forced~!ion eva poraion, vapor binding is also encoun-ter>>d. Thus with liquid benzene entering a 4 pass steam jack.

cled pipe at 0.9 fps, up to thc point where 60 percent by <<eight was vaporized, the maximum flux of 60,000 Btu/h/fl'as For comparison, in a natural convection evaporator, a >>au mum flux of 73.000 Btu/h/ft <<as obtained at (4I)~ of I(O Ii Combined Convection and Radiation Coe!II>>ion!a cas>>s of heal I~. such as that from bare and insulated p t6,

<<here loss is by convection Io the air and radiation to the vab of th>> enclosing, space it is convenient to usc a combined coo'..

veclion and radiation cocilicient (h, + h,). The rate of b>>at loss thus bccom>>s

~0 q m (h, + h,)A(SI)r (4.42$ )

where (AI), is lhc Icmpcralure ditfcrcncc. deg F, bctureea tba surfa>>c of the hot body and thc walls of the space. In cvaluat'. ';

ing, (h, + h,), h, should bc calculatcdby thc appropriate coo vection formula fsec Eqs. (4.4.llc) Io(4.4.IIS)f and h, froln tb>> y equation h, m 0.00685!(T/)00)l Tabl ~ 4.4.10 Maximum Rux and Coneapondinq Overall Temperature Oiflerence for Uqulda Boiled at 1 ~tm wilts a Submerged Horizontat Steam Heated Tube j'ni J

/an >~

4 90 TRAHSMISSIOH OF HEAT BY COHDVCTIOH AHD CONYECTIO JQ /

appears to bc Ihe best available. Zubcr also pcrformcd an anal-o Iained at an overdII temperature dill'erence of 60 F.

ysis for subcoolcd liquids and pfoposed a modification which is this point, the co>>%>>icnt and flux decreased rapidly, 4 p also in excellent agrcemcnt with expcrimcnt:

ing the values obtained in sup>>!heating vapor. s>>c Bq (4

)

ala all (q/r(),

K,p,fk + C gfau ri))[

I

]

(

j faguAu P.))

/

Pi Pi Pr f I +

5 33{P C ihr) (I ri)f ga(P! - a)n! )

Zuber's hydrodynamic analysis of thc Leidcnfrost point yields Iri (q/A),

K~[

0.144 < K, (O.I77 rrgagr(Pr Pr)

'. (4.4.144) 3g

'<O.S

'ir0.4 I+i~

J~

0 I

e

.4'

~ ~

~

Uquid Ethyl aec!aae........

Beaaeae............

Ethyl alcohol........

hlrihyl alcohol......

Die!it!rrt>>atec......

Aluminum Cos>ier Chtunuunr

~

!baaed iopaee I

9/4 (4!I IOOO il)l 55 UlI~,

!alla I

Iar)r 9'A I

9/A IOOO I

IOou j

l000 70

~

4I, 40 5l 10 40 45!00,'$

1)0 4$

11

$ 5 7) too I2i 45 IIO IIO 350 75 t00 I IO l$0 l5$

ilo I

"sha-k r ~FJ.~r t

Table 4.4.11 Yaluea of (h, + 4,)

For ho!iron!al bare or insula!cd saandard s!cel pipe of various sires and for har p4!es in a room aa SO'F Hominal!:Iie diam, la tart>>!cniperarurc diltrrencr. dre f. Ironr surface ao room 50 I I00.1$ 0 100')0 300 i00 '500 400

~ 100 400 900 I IOOO

~ IIOO I!00

2. Il2.IS 2.14 ). IO ).il ).7$ i.il5.)0 S.llil.1$ l. io 9.7)ill.lo Il.ll,li.4$

2.0) 2.)d 245 2.94 3.29 ).42 i.)) 5. I4 4.071. I I 4.15 9.57>ll.oi 124$ lb.id I.93 2.21 1 ~ 52 2.45 3. Ii ).ili. Ili. 99

$.49 d. 92 4.01 9.)4 l0.45 Il.idIi.ll I.bi l. Id!.st 2.71 3.01 3 )) i ~ Ol 4.43 5.12 d.ls 7 ~ 49 9.ll IO ~ 44 llllli.09 I.ld 2 04 129 2.40 249 3.20 ).Sb i.bd 5.51 d.do 7.1) 905 lb. 50 ll.Io I)93

l. 7 I 2.01 1. 2i 2. 5i 2. Sl ) e I) ). 43 i.Sl 5 ~ 50 4 a 52 l. 45 4 ~ 94 10 i2 l2.0) I).bi I.di !.9) 2.!52.452.12).0) 3.10 i.il$.)14.391.524.4) lb.1b I!.90 I),10 I

I I

I i

f I

I I.bl l.l) l.io'1.10 2.99 3.304.00 i.!9 5.10'd.12'l.dd 9. Ib lb.di ll.l5'Ii.od 1.00 1.351.4S1.91 ).24 ).59 i.)l 5. Il4.0i 1.01 4.1! 9.)i I I.OI I!.I)Ii.iS

$6 I

lllli Fane Paaaaa Yer!leal,........,....

HFV.................,

IIFV,I>>rrunnrrt,lrctnd up>>ard; IIFD, aonannrrt nrrnl do>>n>>an>>

...,..... '. 5 4 I. 45 2. 09 1. 34 1. 4 3 2. 9) 3. 4 I i. ) 4 5. 17 4. 11 7. i0 d. 7 I I0. Id I I. 14 I). 51 h

I f~~

Peer n

v, >>

lnn Ã'n

~

~

~

~

I

/

I I

~ ~

I

~

~

~

~

~

I ~ I I I

~

ll I I

I I

J

~

II

~

I I I

~

. I Il ~

~

I

\\

'I ~

~ ~

~

III

~

I I ~ I

~

~

I II ~

~

~

~ ~

>> ~

~ ~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~ ~

~

I I ~

>>C I

q~t i 0+vVI~II~g 4w P>> <<>>

% /VS

<'I~IJxw~~g I'I~l~hP YW+im~lk

,LJ aP v

'>> f'.

~

. +>>'fag'~9 ) Q' P

e

~,I lp ~*

SUPPLEMENTARY TABLES AND FIGURES (USCS UIIITS) 881 1550 1500

~ '

IIS! ~ t

~

~@+

st nt trmecralwc Itrc. f..I 1 000 950 79 30'6 1g 1.2140 1.1813 I.I529 1.1280 1450 '

~O

~ ~

O 1400

.r g

'r r

BSO'OO tso 700 tsla)

L5

).8

'I 0 I,S sia) l.0547 I.0749 l.0904 I. I092 1300

1250, I ~

g 1350

~

I

. r+

~

0

~

uI)S

~

~

eP L

650'00'50 4 I'9 500

'50 utu~.crc f 350 stsnt trmnr r-1

,4 I

1.0108

<.2 t

I.O39R OBl rt 66-104, 1966.

1200 1150 1100 1050 1000 0"0

~

0 I

0 0 r

0

'L

!o P tn' IIr

~ %'0 lo I

~

'b.

300' I

~

200.

Iso'0 100

, 2.0

, 2.1 950 900 1.3 1.4 1.5 1.6 1.7 1,8 Ip] 9 r

Entropy, Btu,!(lb. R)

Figurc A-25 Mollier diagram for steam. Source: J. H. Keenan and J. Ke>es, "Tbcrmodynamic Properties of Stcam," Wiley, New York, 1936.

Is"!s'Is-5'D CCJ~A p4hQE p lq I

',"Il '

>g.'-'iI'IS!rp < tt gtn ~ tI 'c~t'

/ g

Appendix A Therrnophysiext Properdes of Mxrrer 681 Table A.4 Thermophysical properties of gases at atmospheric prcssure 8

>3 59 SO 40 2040 1945 1340 2890 775 810 830 1105 745 3

.6

.'7 2010

)$5 97 I

7

.5 I

2720 2385 2805 IERMAL SPECIFIC JCTIVITY,k HEhT, e

'm K)

(J/kg K)

(K) hir 100 150 200 250 400 4$0 500 550 600 650 700 7$0 800 850 900 950 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500 P

(kg/ms) 3.5562 2.3364 1.7458 1.39 1.1614 0.9950 0.8711 0.7740 0.6964 0.6329 0.5804 0.5356 0.497$

0.4643 0.4354 0.4097 0.3868 0.3666 0.34S2 0.316'.2902 0.2679 0.2488 0 %31'r 0.2177

0. OI9 0.193$

0.1833 0.1741 0.1658 0.1582 0.1513 0.1448 0.1389 0.113$

ev (kJ/kg K) 1.032 1.012 1.007 1.006 1.007 1.009 1.014 1.021 1.030 1.040 1.051 1.063 1.075 1.087 1.099 1.110 1.121 1.131 1.141 1.159 1.175 1.189 1.207 1.230 1248 1.267 1.2&6 1.307 1.337 1972 1.41'7 1.478 1.55&

1.665 2.726 71.1 103.4 1325 1$9 184.6 208.2 230.1 2$0.7 270.1 288.4 305.&

322.5 338.8 354.6 369.&

384.3 398.1 411.3 424.4

. 449.0 473.0 496.0 530 557 584 611 637 663 689 71$

740 766 792 818 955 2.00 9 34 4.426 13.8 7.590 18 I 223 15.89'6.3 2

~

30.0 26.41 33.8 32.39.

37.3 38.79 40.7 4537 43.9 52.69 46.9 60.21 49.7 68.10 52.4 76.37 54.9 84.93

$7.3 93.80 59.6 102.9 62.0 1122 64.3 121.9 66.7 141.8 71.5 162.9 76.3 185.1 82 213 91 240 100 268 106 298 113 329 120 362 128 396 137 431 147 468 160 506

!75 547 196 589 222 841 486 2.54 5.84 10.3 15.9 22.5.

29.9'3 47.2 56.7 66.7 76.9 87.3 98.0 109 120 131 143 155 168 195 224 238 303 350 390 435 482 534

$89 646 714 783 869 960 1570 0.786 0.758 0.737 0.720 0.707 0.700 0.690 0.686 0.684 0.683 0.685 0.690 0.695 0.702 0.709 0.716 0.720 0.723 0.726 0.728 0.728 0.719 0.703 0.685 0.688 0.685 0.683 0.677 0.672 0.667 0.655 0.647 0.630 0.613 0.536 fr 10r v ~ 10<

k. 10r
a. IOe (N.s/m').~m/s)

(N/m K)

~m/s)

Pr 9

,4 2385 2720 3M 320 340 360 380 400 420 0.6S94 0.6448 0.6059 037I6 0&10 0.5136 0.4888 0.4664 Ammonia (NHs) 2.158 2.170 2.192 2221 2287 2.322 2.357 101.5 109 116.5 124 131 138 145 ISL$

14.7 24.7 16.6 0.887 16.9 27.2 19.4 0.870 19.2 29.3 22.1 0,872 21.7 31.6

~

24.9 0.872 24.2 34.0 27.9 0.869 26.9 37.0 313 0.853 29.7 40.4 35.6 0.833 32.7 43.5 39.6 0.826