ML12138A081

FENOC-Davis-Besse Nuclear Power Station, Unit 1 Docket No. 50-346, License No. NPF-3 Submittal of Contractor Root Cause Assessment Report-Section 6

ML12138A081
Person / Time
Site:	Davis Besse
Issue date:	05/14/2012
From:	FirstEnergy Nuclear Operating Co
To:	NRC/RGN-III
References
L-12-196
Download: ML12138A081 (177)

Text

Exhib i t 52 University of Colorado Dept. of Civil, Environmental

& Architectural Engineering Boulder College 01 Engineering and Applied Science t 303 492 8991 428 UCB Boulder, Colorado 80309-0428 1 3034927317 yunpmg.Xl@Co l o r ado edu MEMORANDUM To: Performance Improvement International 21 I I S EI Camino Real Suite 200 Oceanside, CA 92054 Attention:

Dr-Chong Chiu From: Prof. Yunping Xi

Subject:

Concrete Property Testing Results on Submitted Concrete Core Specimens

1. Introduction Concrete core samples were delivered to the University of Colorado at Boulder in Nov. 2011. The concrete cores were cut into 13 samples for testing internal relative humidity, compressive strength, splitting tensile strength, coefficient of thermal expansion, accelerated creep, and freeze-thaw resistance.

The identifications and dimensions of the samples will be described in the fo Ilowing sections together with testing resu It. 2. Internal Relative Humidity The level of internal moisture and the distribution of internal moisture in a concrete structure are important for evaluating shrinkage and freeze-thaw damage of the concrete.

Internal relative humidity (RH) distribution of the concrete was measured by using thermal and moisture sensors SHT75 from Sensirion.

The concrete core used for this test is identified as S6 11/8/11. Eight sensors were embedded in the concrete cylinder at different depths from the surface to measure continuously both internal temperature and RH. The distances of the sensors from the surface are 1.0 in, 1.5 in, 2.0 in, 2.5 in, 3.0 in, 5.0 in, 7.0 in, and 8.5 in. Distributions of RH in the concrete sample were obtained.

Test results are shown in Table 1 and in Fig. I. T ble I RH a a at I erent tunes (S.peclmen S6 1118/11)a test d t Depths from the surface (in) Time 1 1.5 2 2.5 3 5 7 8.5 11 / 18/2011 01:53PM 60.07 63.57 64.26 59.07 60.44 67.01 67.92 69.41 11119/2011 01:53PM 59.7 63.08 63.07 58.32 59.5 67.07 68.01 69.14 11/20/2011 01 :53 PM 58.63 61.94 61.42 56.82 57.91 66.55 67.55 68.42 I 1121/20 I I 0 I :53 PM 58.14 61.32 60.44 56.02 57.26 66.52 67.48 68.24 Page 1 o!' 10 E xhibit 52 100 8 0 ::I:: ----+-II II cl12 0I I 0\: 5 3 PM 60 --11 1I9 12 0 1 10 1 53 P M 1112 0120 11 0 1 :53 P M 1 112 1 120 09 01 :53 PM 4 0 0 I 2 3 6 7 8 9 Depth (inY Fig. 2 RH distributions at different times The RH values near the surface (from 1 in. to the range of 1.5 in. or 2 in.) are about 60%. The RH values at deeper locations approach to 70%, which is higher than the surface value. The values ofRH reflect the annual average RH value ofthe environment.

The gradient of RH is not large in the time period of Oct. and Nov., 2011. After the test for internal relative humidity, a ponding test was performed using the same cylinder.

The concrete cylinder was placed upright with the outer surface facing up. A water column of 13 cm was placed on top of the cylinder.

The purpose of the test was to examine the resistance of the concrete to water and moisture penetration.

The sensor at 3.0 in was damaged during the first test. So, seven sensors were used in the ponding test with the distances to the top surface 1.0 in , 1.5 in, 2.0 in, 2.5 in, 5.0 in , 7.0 in , and 8.5 in. Penetration depths 100 -Initial RH 90 -1day 80 -2 days 70 -3 days >-60 "C 50 -4 days 'E ::::l -5 days ..c 40 OJ 30 -6 days ro 20 > 7 days OJ a:: 10 8 days 0 0 2 4 6 8 10 Dept Fig. 3 RH distributions at different times during the ponding test Page 2 of 10 Exh i bit 52 Fig. 3 shows the test data. The initial RH distribution is between 30% to 45%. This is because the test started on Feb. 7, 2012, more than two months after the fLrst test. Within two days, the water penetrated to the depth of2 inches, which indicated that the resistance of the surface layer concrete to water penetration is qu ite low. The concrete at 2.5 in. and deeper portion shows much higher resistance. The high moisture region with RH > 90% reached about 2.5 inches after four days of pond ing. These results showed that the rate of moisture penetration into the concrete depends strongly on the quality of surface layer concrete, where microcracks may form due to various deterioration mechanisms such as drying shrinkage.

In order to determine moisture resistance of the concrete at different locations of the structure , more samples need to be taken from the structure and tested. 3. Com pressive Strength The compressive strength of concrete was tested according to ASTM C39. It is for the unconfined compressive strength of cylindrical concrete specimens.

Four samples were used for the test. The identifications of the samples are shown in Table 2 and Fig. 4. Dimensions of the specimens and the test data are listed in Table 2. Table 2 Compressive strength ofthe four specimens No Diameter (in.) Length (in.) fc' (psi) Specimen description 1 2.65 5.76 5444 Hallway #J 2 2.65 5.76 6342 S9680-3 3 3.39 5.88 7990 S4 11/8111 4 3.39 6.38 10508 S4 11/8111 Fig. 4 Specimen # 1 (right) and #2 (left) after the compressive strength test 4. Splitting Tensile Strength The splitting tensile strength of cyl indrical concrete specimens was tested according to ASTM C496. The maximum load recorded, P, was used to calculate the splitting tensile strength of concrete samp les based on Eq. I. Page 3 of 1 0 Exhibit 52 2P (Eq. I)j" =Trld in whichls, = splitting tensile strength; I and d are the length and diameter ofspecimen, respectively.

Test results are shown in Table 3. *1 T a ble 3 S ,p r It f mg ensl e s ren gth t est d t aa Specimen Length (in) Diameter (in) Area (in!\2) Force (kips) lsi (psi) Specimen description No. \ 4.2 3.68 10.63 67.110 957.43 S8 11/8/ 11 No.2 5.1 3.68 10.63 67.447 962.23 S8 11/3/11 No.3 5.4 3.68 10.63 58.585 835.8 S3 11/8111 5. Coefficient of Thermal Expansion (CTE) Two cylindrical specimens

(#S2 11/8/11, #S4 11/8/11) were used for the test. The diameters of the two samples are the same, 3.39 in. Thermal expansions of the two specimens were measured between two temperature ranges from noe to 40 0 e and then from to 40 0 e to 60 o e, then linear coefficients of thermal expansion were calculated based on the test data. The tests were conducted in an environmental chamber with temperature control. So, it is different from USBR 4910-92 conducted in u.S. Bureau of Reclamation , in which the specimens were submerged in water and the water temperature is varied to create the thermal expansion.

The purpose of this test is to obtain eTE of concrete used in above ground structures.

Thermal sensors were installed inside of concrete samples to double check the internal temperature in concrete samples. When the internal temperature reaches the target temperature, deformation of concrete sample was measured after two hours of holding of the target temperature.

This was to make sure that the internal temperature distribution in the cylinder reaches equilibrium (uniform distribution).

The test data of the two specimens are shown in Table 4 and Table 5. Table 4 eTE of Specimen #S2 COC) Lenfrth (in) (ill) eTE (11°C) Average eTE (11°C) 18 (from 22 to 40°C) 10.157 0.0013 0.00000711

0.0 0000875

20 (from 40 to 60°C) 10.157 0.002\ 0.00001034 38 (from 22 to 60°C) 10.157 0.0034 0.00000881 Table 5 eTE of Specimen #S4 (0C) Length (in) (in) eTE (11°C) Average eTE (\1°C) 18 (from 22 to 40°C) 10.284 0.0011 0.00000594

0.0 0000886

20 (from 40 to 60°C) 10.284 0.0024 0.00001167 38 (from 22 to 60°C) 10.284 0.0035 0.00000896 Page 4 of 10 Exhibit 52 The average value ofeTE of the two specimens

= 8.8xl0-6 /0 e. The eTE measured by USBR = 5.2x 10-6 /o F = 9.4x I 0-6 fO e. Our eTE value is slightly smaller than the eTE measured by USBR because when the concrete was heated in air , the measured thermal expansion is actually a combination of pure thermal expansion and drying shrinkage.

In add ition to the test of e T E in the temperature range above o o e, another series of tests was performed for the thermal strains in the temperature range below O°e. The purpose of this test was to examine the effect of ice formation on thermal expansion of the concrete.

The testing sample was S2 11 /8/11. The diameter of the specimen is 3.46 in. and 11.5 in. long. Two tests were performed. One is called dry test. The specimen was placed in a high temperature chamber for 14 days under 80 0 e to dry out the internal moisture, and then placed in a freezing chamber with programmable temperature control. The test started at 20 o e, the temperature was reduced to 15°e in 30 minutes, stayed at 15°e for 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> , and the strain was measured.

The process was repeated until the target temperature of -25°e. The test data are shown as the blue curve in Fig. 5. Strain-Temp relations 1 0.00E+00 -1.00E-04 C -2.00E-04 "' +' -3.00E-04 I/) -4.00E-04 -S.00E-04

-6.00E-04 30 20 10 1 0 20 -30 T/oe Fig. 5 Thermal strains under low temperatures for the effect of ice formation in the concrete (Blue curve: Dry sample Red curve: Wet sample) After the test for the dry sample, the specimen was placed in a water tank for 68 hours7.87037e-4 days <br />0.0189 hours <br />1.124339e-4 weeks <br />2.5874e-5 months <br /> for saturation. The same testing procedure was used to obtain the thermal strain under low temperatures.

The test was called wet test. T he test data are shown as the red curve in Fig. 5. Jt is important to see from Fig. 5 that the concrete contracts upon cooling from 20 0 e to o o e, starts to expand from o o e to -15°e, and then starts to contracts again. The first reversal from contraction to expansion is due to the ice formation in the concrete, because the eTE of ice is about five times higher than the eTE of concrete. The second reversal is an indication of the completion of ice formation.

Both ice and concrete contract upon a further cooling. Pa g e 5 of 10 Exhibit 52 The test results ind icated that with a high moisture content, the effect of ice formation on thermal strain of the concrete sample is significant, resulting in an expansion under the low temperature from O°C to -15°C. Because oflimited time, the internal moisture distribution in the sample may not be uniform, so the measured strains represent average values of the thermal strains. In order to determine the coupling effects among moisture content, low temperature, and ice formation, a more systematic experimental study with more samples is needed. Accelerated Creep The accelerated creep tests were performed to obtain creep strain of the concrete used in Besse Nuclear Power Station. The creep tests generally follow the procedure described in ASTM C-512 "Standard Test Method for Creep in Compression".

Three accelerated creep tests were performed under 40°C (with and without humidity control) and 80 °C (with humidity control), respectively.

Different relative humidity controls were used in the tests to find the effect of moisture level on the creep of concrete.

Some of the basic terminologies used in this section are Basic creep -The long-term strain of concrete due to load ing without drying and heating. Drying shrinkage

-The long-term strain of concrete due to drying without load ing and heating. Drying creep -The long-term strain of concrete due to loading and drying without heating. Fig. 6 The MTS machine provides a stable compressive force at 16 kips 6.1 Testing method One cylindrical specimens

(#S2 11/8111) was cut into a cylinder of 11.5 in. long. The diameter of the specimen is 3.46 in. Two contact points were installed on the top and bottom portion of the specimen.

The distance between the two contact points equals the gage length of the dial gauge to be used to measure the length change of the specimen. The specimen was capped on top and bottom surfaces. The specimen was loaded by a MTS machine mounded in an environmental chamber. The Page 6 of '0 Exhib i t 52 loading level was kept as a constant 16 kips , which resulted in a compressive stress of 1702 psi. This stress level is less than 40% of the average compressive strength of the concrete (7 , 600 psi). Fig. 6 shows the loading setup. The chamber was maintained at a constant temperature of 40°C or 80°C. For the first specimen , the temperature was kept at 40°C without humidity control. For the second specimen, the temperature was kept at 80°C, and the humidity was controlled in the range of 70% to 80%. For the third specimen, the temperature was kept at 40°C, and the humidity was controlled in the range of70% to 80%. 6.2 Test results The test results of the three accelerated creep tests are plotted in Figs. 7, 8 , and 9. Fig. 7 shows the test results under 40°C without humidity control; Fig. 8 is for the test results under 80°C with humidity control; and Fig. 9 is for the test results under 40°C with humidity control. om 0.1 1 Lo g T i me (h o u rs) 1 0 100 10 00 Fig. 7 Test results obtained under 40°C without humidity control v.vvv o v 0.01 0.10 100 1000 L og tim e (h ou r s) Fig. 8 Test results obtained under 80°C with humidity control Page 7 of 10 Exhibit 52 V .V\AJQ V.VVVV v.vVV' Ao ... Y'll' om 0.1 v I Lo g Time (hour s) 10 100 10 00 Fig. 9 Test results obtained under 40°C with humidity control 6.3 Comparisons and discussions Comparison of Fig. 7 and Fig. 9 shows the effect of drying on concrete creep. The creep reading in Fig. 7 is about 0.0005 (500 microstrain) after 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br />, which is about 2.5 times the value in Fig. 9 , which is about 0.0002 (200 microstrain).

The difference between the two tests was the relative humidity in the chamber. In Fig 9, the relative humidity was controlled at the range of 70%-80%, so the test data represent basic creep of the concrete.

In Fig. 7, there was no humidity control, so the test data represent a combination of basic creep and drying shrinkage , and that is why the measured strains are much higher. Basically, the effect of drying shrinkage is quite significant , and the creep of concrete under an arid environment (low humidity) is much higher than that in a humid environment.

I-+-80 d eg ree C v .vuvu ----4 a d e g r ee C v.vvvv c 0; .A"'!"'. !--"a-.J v /'0.0 1 0. 1 1 00 1 000 I L o g i}m e (h o ur sl O Fig. 10 Comparison of creep strains under 40°C and 80'C (log time scale) 0.0008 0.0006 " g 0.000 4"" 0.00 0 2 lL:: ..V .... ..... ...... I-+-80 d egree C 40 de gree C { a a s o 1 00 I SO 2 00 Gme (hou rs) 2 5 0 300 3 50 Fig. 11 Comparison of creep strains under 40'C and 80°C (regular time scale) Page 8 of lO Exh i bit 52 Comparison of Fig. 8 and Fig. 9 shows the effect of temperature on concrete creep. The same humidity controls were used for the two tests, and thus there is no effect of drying shrinkage.

Combin ing Fig. 8 and Fig. 9 gives Fig. lOin log time scale and Fig. II in regular time scale. Fig. II can be used to obtain creep compliance functions under two different temperatures, which can be used further to obtain the creep coefficient of concrete.

7. Freeze-thaw resistance The accelerated freeze-thaw tests were planned to obtain freeze-thaw of the concrete used in Davis-Besse Nuclear Power Station. The freeze-thaw tests generally follow the procedure described in ASTM C-666. The testing chest is shown in Fig.12. Each freeze-thaw cycle is approximately 2-3 hours. There will be 300 cycles. Two samples were used for the test: S9 3 and In Steam Room 602 #1 Fig. 12. The rapid freeze-thaw test chest After one day of testing, the temperature controller of the testing machine was broken. The test was stopped. After a new controller was installed, we did not re-start the test, because this is a long-term test and it cannot be done before the completion of the project. A summary table is shown in the next page for identifications of all sample tested at University of Colorado.

Page 9 of lO Exhibit 52 Tests at Univ. of Colorado Test Core identification Internal moisture and ponding test 56 11/8/11 Compression Hallway #1 59680-3 5411/8/11 5411/8/11 Splitting tension 5811/8/11 5811/8/11 5311/8/11 Coefficient of thermal expansion 52 54 Creep and CTE at low temperatures 5211/8/11 Freeze-thaw 59680-3 In Steam Room 602 #1 Pag e 10 of 10 Exhibit 53: C-0109 Roof Plans and Details Appendix VIII-54© 2012. Performance Improvement

.* 'Q I* :! "

i .,

Exhibit 54: C-200 © 2012. Performance Improvement International-Appendix

Exhibit 55: AC! 201.2R-08, Table 6.3 © 2012. Performance Improvement Intemational-Appendix Exhibit 55 I GUIDE TO DURABLE CONCRETE Table 6.1-Effect of commonly used chemicals on concrete Moderate Phosphoric Organic acids Alkaline solutions Miscellaneous Acetic Formic Lactic Aluminum chlori(le I Tannic Sodium hydroxide-

> 20% Ammonium nitrate Ammonium sulfate Sodium sulfate Magnesium sulfate Calcium sulfate Bromine (gas) Sulfite liquor Slow Carbonic Sodium hydroxide' 10 to 20% Sodium hypochlorite Ammonium chloride Magnesium chloride Sodium cyanide Chlorine (gas) Seawater Soft water Negligible Oxalic Tartaric Sodium hydroxide'

< 10% Sodium hypochlorite Ammonium hydroxide Calcium chloride Sodium chloride Zinc nitrate Sodium dichromate Ammonia (liquid) "The effect of potassium hydroxide is similar to that of sodium hydroxide, DePuy (1994), Taylor (1997), Skalny et al. (1998), Thomas and Skalny (2006), and Naik et al. (2006). Publications with particular emphasis on permeability and the ability of concrete to resist ingress and movement of water include Reinhardt (1997), Hearn et al. (1994), Hearn and Young (1999), Diamond (1998), and Diamond and Lee (1999). 6.2.3 Recommendations-Protection against sulfate attack is obtained by using concrete that retards the ingress and movement of water and concrete-making ingredients priate for producing concrete having the needed sulfate resistance.

The ingress and movement of water are reduced by lowering the wlcm. Care should be taken to ensure that the concrete is designed and constructed to minimize shrinkage cracking.

Air entrainment is beneficial if it is accompanied by a reduction in the wlcm (Verbeck 1968). Proper placement, compaction, finishing, and curing of concrete are essential to minimize the ingress and movement of water that is the carrier of the aggressive salts. Recommended procedures for these are found in ACI 304R, 302.1R, 308R, 305R, and 306R. The sulfate resistance of portland cement generally decreases with an increase in its calculated nate (C 3 A) content (Mather 1968). ASTM C150 permits the use of Type V sulfate-resisting cement and C 3 A with a maximum limit of 5%, and Type II moderately resisting cement and C 3 A limited to 8%. There is also some evidence that the alumina in the aluminoferrite phase of portland cement can participate in sulfate attack. Therefore, ASTM C 150 states that the C 4 AF + 2C 3 A in Type V cement should not exceed 25% unless the alternate requirement based on the use of the performance test (ASTM C452) is invoked. In the case of Type V cement, the sulfate-expansion test (ASTM C452) can be used instead of the chemical requirements (Mather 1978). The use of ASTM ClO12 is discussed by Patzias (1991). Table 6.3 provides recommendations for various degrees of potential exposure.

These recommendations are designed to protect against concrete distress from sulfate from sources external to the concrete, such as adjacent soil and groundwater.

Recommendations for the maximum wlcm and the type of cementitious material for concrete that will be exposed to sulfates in soil or groundwater are given in Table 6.3. Both Table 6.2-Factors influencing chemical attack on concrete Factors that accelerate or aggravate attack Factors that mitigate or delay attack I. High porosity due to: I. Dense concrete achieved by: i. High water absorption

i. Proper mixture proportioning' ii. Penneability ii. Reduced unit water content iii. Voids iii, Increased cementitious material content iv. Air entrainment

v. Adequate consolidation vi. Effective curing t 2. Cracks and separations due to: 2. Reduced tensile stress in concrete L Stress concentrations by:*ii. Thermal shock L Using tensile reinforcement of adequate size, correctly located ii. Inclusion of pozzolan (to reduce temperature rise) iii. Provision of adequate contraction joints 3. Leaching and liquid penetration

3. Structural design: due to: L To minimize areas of contact i. Flowing liquid§ and turbulence ii. Ponding iii, Hydraulic pressure ii. Provision of membranes and protective-barrier system(sJ II to reduce penetration

'The mixture proportions and the initial mixing and processing of fresh detennine its homogeneity and tPoor curing procedures result in flaws and *Resistance to cracking depends on strength and strain

§Movement of water-carrying deleterious substances increases reactions that on both the quantity and velocity of "Concrete that will be frequently exposed to chemicals known to produce rapid tion should be protected with a chemically resistant protective-barrier of these recommendations are important.

Limiting only the type of cementitious material is not adequate for satisfactory resistance to sulfate attack (Ka10usek et al. 1976). The field conditions of concrete exposed to sulfate are numerous and variable.

The aggressiveness of the conditions depends on soil saturation, water movement, ambient temperature and humidity, concentration of sulfate, and type of sulfate or combination of sulfates involved.

Depending on the aforementioned variables, solutions containing calcium sulfate are generally less aggressive than solutions of sodium sulfate, which is generally less aggressive than magnesium sulfate. Table 6.3 provides criteria that should maximize the service life of concrete subjected to the more aggressive exposure conditions.

Exhibit 55 Page 1 of 2 Exhibit 55 ACI COMMIITEE REPORT Table 6.3-Requirements to protect against damage to concrete by sulfate attack from external sources of sulfate Severity of potential exposure Water-soluble sulfate in soil, % by mass

• Sulfate (S04)* in water,ppm wlcm by mass, max.tt Cementitious material requirements Class 0 exposure 0.00 to 0.10 o to 150 No special requirements for sulfate resistance No special requirements for sulfate resistance Class 1 exposure >0.10 and <0.20 > 150 and < 1500 0.50 i C 150 Type II or equivalent§ Class 2 exposure 0.20to<2.0 1500 to < 10,000 OA5 i C150 Type V or equivalent§ Class 3 exposure 2.0 or greater 10,000 or greater 0040* C150 Type V plus pozzolan or slag§ Seawater exposure -See Section 604 See Section 6.4 *Sulfate expressed as S04'S related to sulfate expressed as S03' as given In repOrts of chenucal analYSIS of portland cements as follows: S03% x 1.2 = t ACI 318, Chapter 4, includes requirements for special exposure conditions such as steel-reinforced concrete that may be exposed to chlorides.

For concrete likely to be SUbjected these exposure conditions, the maximum wlcm should be that specified in ACI 318, Chapter 4, ifi! is lower than that stated in Table 6.3 of201.2RtValues applicable to normalweight concrete.

Tbey are also applicable to structural lightweight concrete except that the maximum wlcm ratios 0.50, 0.45, and 0040 should be by specified 28-day compressive strengths of 26,29, and 33 MPa (3750. 4250, and 4750 psi),

§For Class I exposure, equivalents are described in Sections 6.2.5, 6.2.6, and 6.2.9. For Class 2 exposure, equivalents are described in Secrions 6.2.5, 6.2.7, and 6.2.9. For Class ex sure, pozzolan and sla recommendations are described in Sections 6.2.5, 6.2.8, and Portland-cement concrete can be also be attacked by acidic solutions, such as sulfuric acid. Infonnation on acid attack is provided in Section 6.5. 6.2.4 Sampling and testing to determine potential sulfate exposure-To assess the severity of the potential exposure of concrete to detrimental amounts of sulfate, representative samples should be taken of water that might reach the concrete or of soil that might be leached by water moving to the concrete.

A procedure for making a water extract of soil samples for sulfate analysis is given in Appendix A. The extract should be analyzed for sulfate by a method suitable to the concentration of sulfate in the extract solution, such as the photometer methods used in ASTM C1580. If the amount of sulfate detennined in the first analysis is outside of the optimum concentration range for the analytical procedure used, the extract solution should be either concentrated or diluted to bring the sulfate content within the range appropriate to the analytical method, and the analysis should be repeated on the modified extract solution.

6.2.5 Material

qualification of pozzolans and slag for sulfate-resistance enhancement-Tests I year in duration are necessary to establish the ability of pozzolans and slag to enhance sulfate resistance.

Once this material property has been established for specific materials, proposed mixtures using them can be evaluated for Class I and 2 exposures using the criteria in Sections 6.2.6 and 6.2.7. Fly ashes, natural pozzolans, silica fumes, and slags may be qualified for sulfate resistance by demonstrating an expansion:s;;

0.10% in 1 year when tested individually with portland cement by ASTM ClO 12 in the following mixtures.

For fly ash or natural pozzolan, the portland cement portion of the test mixture should consist of cement with lated C 3 A of not less than 7%. The fly ash or natural pozzolan proportion should be between 25 and 35% by mass, calculated as percentage by mass of the total cementitious material.

For silica fume, the portland cement portion of the test mixture should consist of a cement with Bogue-calculated C3A of not less than 7%. The silica fume proportion should be between 7 and 15% by mass, calculated as percentage by mass of the total cementitious material.

For slag, the portland cement portion of the test mixture should consist of a cement with Bogue-calculated C 3 A of not less than 7%. The C 3 A should be calculated for the sum of the portland cement plus calcium sulfate in the cement. Some processing additions, if present in sufficient proportions, can distort the calculated Bogue values. Fonnulas for calculating Bogue compounds may be found in ASTM C 150. The slag proportion should be between 40 and 70% by mass, calculated as a percentage by mass of the total cementitious material.

Material qualification tests should be based on passing results from two samples taken at times a few weeks apart. The qualifying test data should be no older than I year from the date of test completion.

The reported calcium-oxide content analyzed in accordance with ASTM Cl14 of the fly ash used in the project should be no more than 2.0 percentage points greater than that of the fly ash used in qualifying test mixtures.

The reported aluminum-oxide content analyzed in accordance with ASTM C 114 of the slag used in the project should be no more than 2.0 percentage points higher than that of the slag used in qualifying test mixtures.

6.2.6 Type II equivalent for Class 1 exposure ASTM C 150 Type III cement with the optional limit of 8% maximum C 3 A; ASTM C595 Type IS(MS), Type IP(MS), Type IS-ACMS), or Type IP-A(MS);

ASTM Cl157 Type MS; or Any blend of portland cement of any type meeting ASTM Cl50 or Cl157 with fly ash or natural pozzolan meeting ASTM C618, silica fume meeting ASTM C 1240, or slag meeting ASTM C989 that meets the following requirement when tested in accordance with ASTMC1012:

Expansion S; 0.10% at 6 months Any fly ash, natural pozzolan, silica fume, or slag used should be previously qualified in accordance with Section 6.2.5. 6.2.7 Type V equivalent for Class 2 exposure ASTM Cl50 Type III cement with the optional limit of 5% maximum C 3 A or ASTM C150 cement of any type having expansion at 14 days no greater than 0.040% when tested by ASTM C452 or ASTM el157 Type HS; or Any blend of portland cement of any type meeting Exhibit Page 2 of2 Exhibit 56: Structural and Analysis

© 2012. Performance Improvement International Appendix Exhibit 56 Redacted .. Thennal SIres,

.""P_Final Report Results & Comments Exhibit 56 -Summary Final Report -Revision dated 24 Feb 2012 1.0 Duties and Responsibilities This report summarizes the activity of to the structural thermal analysis investigation work performed at -JLlv.,,,,,, nuclear power 1.1 3DNastran_FEM Primary responsibility is the development of 3D Nastran _Finite Element Models (FEM's) for use in computing thermal transient temperature distributions due to various environmental conditions.

These 3D _ FEM's include pressure loading that result from wind . due to Tornados and other Wind conditions during the winter and summer cases. The 3D Nastran _ FEM is used in the thermal transient heat transfer analysis performed to compute solar heating/cooling for the following environmental conditions:

./ Summer Solstice (Hot w/o Wind, Hot w/34 mph Wind & Ave w/o ./ Autumn Equinox ( "" " " " " """" ./ Winter Solstice (Ave w/o ./ 1978 Blizzard ( w/105 mph ./ Vernal Equinox (Ave w/o Wind, Ave w/36 mph Note: The initial series of analysis showed Vernal Equinox conditions were not critical.

The 3D Nastran ..FEM's were also used to provide approximations for stresses & deflections throughout the Davis-Besse Shield Build due to combined effects of wind, thermal transients and 1.2 Nastran 2-D Plane-Strain

_ Idealizations During the course ofthese and evaluated . ********information.

Performance Improvement international, LLC. Page 1 of 39 2012 The total number of elernenlts and Exhibit 56 Redacted Thermal Stress Analysis:

Final Report Results & Comments ..". 1.0 Nastran Finite Element Model Definitions The analysis code used for the transient thermal and structural analysis is MDlNastran 2010 v1.3. MD Nastran is a general purpose fmite element program for performing linear, nonlinear structural analysis, vibration, dynamics and thermal analysis.

2.1 3D _ Finite Element Model Figure 2.1.1 shows an isometric view ofthe 3D Nastran _ idealization.

Key point details of the development and definition for the 3D Nastran Thermal Transient and Structural FEM are the following:

./ The 3D Nastran Nastran 360 0 3D models idealize the entire Shield Building

• idealization was Element size through the 30" concrete wall elements and node comprising the Nastran 3D The total number of degrees of freedom The reference drawings used to develop the Nastran 3D are Dwg. No. C-100, C-I04 & C-I09 The overall region idealized is from EL 567' 6" [base truncation level] to the top ofthe Dome EL 824' 3 Yz". The inside radius RIF 69'6" and outside radius ROF 72' [vertical wall thickness 2' 6"]. The Dome wall thickness

= 2' . ./ Concrete Reinforcement

./ astran _ models idealize hoop and vertical reinforcement

-The 3D Nastran _FEM ""'....VB"'"' results from other analysis *...,. _FEM is configured for use & structural model is .....,.,."Ull" a cross check validation of 3D Nastran ********information.

Performance Improvement international, LLC. Page 2 of 39 2012 Exhibit 56 Redacted 11 191== ., Th"ma' St,." Fina' Report Res"" & Commen" \ i I I \ * ,/ The NastrAn _ models Idealize entire' . \ .J 'ld'contam ment cpncrete oUl mg I \. \

• I \ \ ,/ The region is ftom EL 6" [\lase ion I ev t I] to the: t op of tDome ltL 82 4 ' 3 '/," , \ \ \ ,/ The ipsid e R,p 6":'S,d outsit radius 72' [1 ertical "'1 aU thickfes s = 2' \6"1. The Dome w thickness

= 2'l .A T , I I ./ Cono r ete I I \ \\ 3D_ Idealization:

Key Enwlope Dlme)islons I a.) Net of \lIe ve1cal walls 4 242' (EL 59 7' 6" to::L $09' 6") \ b .) Net height to the top \ofDome ,: 356' 9-1 1 2" (1\:L 567' 6" to\EL 834' 3-1/2") , ,) ,,,id, obm"" ,fV,",'" w"',,\, !l9' (1';,\69' 6") , d)y'",," ,", oil Tru'1"'2' 6" I r ,d , ' ' Ro ' 72') \ -\ Base EL 567' 6" Upper EL 824' 31/2" Truncation Level 3D Nastra _ Model Idealization page 3 of **********

information. Performance Improvement international, llC.

Exhibit 56 Redacted Thermal Stress Analysis Final Report Results & Comments "IEIJ!w ./ ./ I 3D Nastran z .., I I Figure 2.1.2 3D Nastran _ Model: Steel Reinforcement

information.

Performance Improvement international , LLC. 2012 Page 4 of 39

• Exhibit 56 Redacted Thermal Stress Analysis: Final Report Results & Comments Concr et e Figure 2.1.3 3D Nastran _Model: Typical Section Cut **********

information. Performance Improvement international , LLC. 2012 . Page 5 of 39 Exh i bit 56 Redacted Thennal Stress Analysis: Final Report Results & Comments w:t'i'1'I!

A , \1:1I(>1"lal Pr: o]>l'I"tlt'<, for Oi'('rall_Fiultt>

£1('lnPI!1

\IoII('IIF£\II hlpnllzrI I USSR VoJuc-s,nr c-mall memo r ondum 11 January 2012 RI:vi SI}(( t) 2 0 12 I Generic Steel j Rebar & Inner Steel containmLt Eo k ut &!Oroh.

& Mo.. TroM/u" I I TalJ\t' I, ('ou(,I'('I('

& $1('('1 r"Ilr""" nloliye of \peciiic reflee l l oc:.li z ed lllc!.1 >lU'elllenI S. i I Pfopcrt/J: Proper/I.. o.:;-n\tly. w I K O l ff'U SNl ty. {( S_oI,H i;J!.c,. 'E j" f hEorm3 t f Xp.ioSlon Youn,', r-to dulu'. E P o ..nOrl*s RatJ o. v {"F I (Ib/l ni I Btufhr ;n.°f) lln'fh ri I atu/lb or' ,n/l ntf (. 1 0") 1 0' ) 80 1.610 86.074 I O.Il O 0.15 680 0.30 Figure Material Properties for Davis-Besse 3D Nastran _ **********

infonnation.

Perfonnance Improvement international , LLC. 2012 P a g e 6 o f *39 Exhibit 56 Redacted Thermal Stress Analysis: Final Report Results & Comments 2.2 Nastran 2D Plane-Strain Utility Idealization Figure 2.2.1 shows an overall isometric view of the complete 2D Nastran plane-strain_

idealization and a close-up view of the rebar and concrete mesh details. Key point details of the development and definition for the 20 Nastran plane-strain FEM are the following:

./ The 20 Nastran plane-strain models idealize a section cut through the Shield Building EL 683 ' 6" . at ./ The total number of elements and node comprising the Nastran 20 plane-strain model are _elements and _nodes. ./ The total number of degrees of freedom . ./ The reference drawings used to develop the Nastran 20 plane-strain No. C-110 ion idealized is at EL 683' 6" This elevation is defined as a reference elevations 683' 6" is along the vertical walls approximately half-way between EL 567' 6" [base truncation level] to just under the ring girder EL 80 I' 6-1 /2". EL Owg. . structural model is also used *********information. Performance Improvement international , LLC. Page 7 of 39 2012 Exh i bit 56 Redacte d Final Report Results & Comments ,:Ii1@*Thennal Stress Analysis:

I. Close-U p VievJ of 2-D Plane-Strain fEM R ebar and Concrete De finiti on fu lll s ofTIetric View of 2-D Pl ane-Strain F EM Defin it i on Figure 2.2.1 20 Nastran Plane-Strain

information.

Performance Improvement international, LLC. Page 8 of 39 2012 Exhibit 56 Redacted Thermal Stress Analysis:

_ Final Report Results & Comments .'i 3.0 Step-by-Step Analysis Process *********information, Performance Improvement international, LLC, Page 9 of 39 2012

• Exhibit 56 Redacted -'ID">>-Thennal Stress Analysis:

Final Report Results & Comments Figure 3.0.1 Schematic Flow Chart Representation of"Step-hy-Step" Analysis *********information.

Performance Improvement international, LLC. 2012 Page 10 of 39

____ Transient thermal temperatures due to the Winter Solstice and 1978 Exhibit 56 Redacted. .Thermal Stress Analysis:

Final Report Results &Comments 3.1 Typical Output Results 2D Plane-Strain Utility Model The lane-strain idealization for the full 360 0 Shield Building wall was_ show examples of output results These summary plots show the distribution of maximum principal and radial stresses for the peak summer solstice condition at 7:30 pm; respectively.

In these figures the S-W facing Flutes are showing the highest magnitude of maximum principal and radial stresses.

The peak stress results occur at the outer rebar regions due to SCF effects where the overlapping rebar ends in the thick portion of the Flute. Radial stresses are plotted for each of the selected time slices as the sun traverses the sky during the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> period beginning.

One of the key aspects of this 20 plane-strain

_ is that all of the action is in the thick portion of the Flutes with peripheral, secondary action along the rebar at the OF. One of the key aspects of this study is that all of the action is in the thick portion ofthe Flutes with peripheral, secondary action along the rebar at the OF Figure 3.1.3 shows summary peak rad ial stresses during the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> period for the environmental conditions listed above. From Fi re 3.1.3 the time slices ducing the highest rad ia I stresses *********information. Performance Improvement international, LLC. Page 11 of 39 2012

.' Ex h i b i t 5 6 Reda c ted Thermal Stress Analysis: Final Report Results & Comments e of suess values sbown from mo :m stram )ccur at pt"ak SCF Th e se stress contour p l ots are concal ttme mten'3ls. M a lumum Princip a l Stress Highest In T hICk PorttOn of Flute O J e t al( View @ EL 683' 6" Close-Up View @ EL 683' 6" I , I Figure Summer Solstice Hot No Wind 7:30 pm , Constant Concrete CTE = 5.20 xlO-6 2D Plane-Strain Maximum Principal Stress N ot e: The thick regions of the West and Southwest facing architectural flutes indicate the highest magnitude of maximum principal stress values resulting from the summer solstice conditions . **********

information. Performance Improvement international, LLC. 2012 Page 12 of 39 A Exhibit 56 Redac ted Thermal Stress Analysis : Final Report Results & Comments mE.Ir_ The rn3g!l1tIldeSf SlUSS \'3Iu&s shown from the planestram __ r oce a! peak SCF effects.

stress contour plots are used to tlvely sekct cOl1c3lllme IDlerYais Maximum Radi a l HIgh e st I n Thick Portion of Flute Close-Up View @ EL 683' 6" Figure Summer Solstice Hot No Wind 7:30 pm, Constant Concrete CTE = 5.20 x I 0-6 in/in/o2D Plane-Strain Radial Stress Note: The thick regions of the West and Southwest facing architectural flutes indicate the highest magnitude of maximum principal stress values resulting from the summer solstice conditions . **********

information.

Performance Improvement international, LLC. 2012 Page 13 o f 39

---Exhibit 56 Redac t ed -:'..-.'_l

• A>Thoemal St"" Analy,,,

• Floal Report R"olts & Commeot, .: mi I UIIIIIIIi 11111111111111111111111111 Summer Solstice Hot No Wind: 7:30 P M Summer Solstice Hot M mph Wind: 6:00 AM Summer Solstice Ave No W ind: 7:30 PM ----......-,---_.-.........1_-____

1978 Blizzard Record Lov, 150 mph Wind: 5:00 AM Winter Solstice Ave No Wind: 7: 30 AM .....---,........._r___ _ --.-..-.. --_._...._-----1---.---...

1* ... _--...-_1 ... t . .. ...... __......... 111111111111111111111111 111111111111111111111111 HHlUni Autumn Equinox Hot No Wi nd: 5: 0 0 AM & 6:00 P M Au tu m n Equinox Ho t 34 m p h W i nd: 5: 00 A M Autumn Equinox A ve N o Wind: 5:00 AM Fig ure 3.1.3 S urve y R adia l S tr e s s R e sul t s: Na stran 2 D Pl a ne-St r ain r E M; H e at Tr an s fe r A nalysis; 2 4/1 H our Ti me [192 -I hour Time S li c e s] Note: **********

inform a tion. Performance Improvement international, LLC. 2012 Page 14 of 39 Exh i b i t 56 Redacted Inal Report Results & Comments Thermal Stress Ana l ysis: .".IM 3.2 Typ ica l Output R esults 3D Figur es 3.2.1 a n d 3.2.2 s how examples of output re s ults fcom th e mapped thermal tran s ient thermal stress analyses.

These s ummary plot s show the distribution of maximum principal and radial st re ss es for the peak summ er s o lsti ce condition at 7:30 pm; re s pectively . *********information. Performance Impro vem e nt internationa l , LLC. 2012 Page 15 of 39 Exh i b i t 56 Redac t ed . ),. Final Report Results &Comments -"lM1W a.\ Thermal Stress Analysis:

Overall View @ EL 683' 6" Close-Up View @ EL 683' 6" Figur e 3.2.1 Su m mer Solstice H ot N o W i n d 7:30 pm, Constant Concrete e TC = 5.20 xl 0.6 inlin/oF 3D _ F EM M aximum Principal Stress D istribution Note: N on-Symm etr ic T he rmal S tr e sses Due to Uneven/II igher H eating Gradients on So ut h Facing P a n els . **********

i nformation.

Performance Improvement international, LLC. 2012 Pa g e 16 of 39 Exhibit 56 Redacted Final Report Results & Comments IImBIe-Thermal Stress Analysis .a.'" > -..1......___ (, ..,. Radial SHess In Thick Portion of RUle '" +76 pSI Overall View @ EL 683' 6" Close-Up View @ E L 683' 6" F ig u re 3.2.S um mer S ols tice H ot N o W ind 7:30 pm, Constant Concre te C TE = 5.20 x l 0-6 irvin/o3D _ FEM Radial Str ess D Not e: N on-Sym metric T h e rmal Stresses Due to Uneve n/H igher H eat ing Gradients on So uth Facing Pane ls. **********

information. Performance Impro vement international, LLC. 2012 Pa ge 1 7 of 3 9 Exhibit 56 Redacted Thermal Stress Analysis: Final Report Resu l ts & Comments .1:,.,'&.

4.0 Summary

Re sults a n d Co mm n t s T he se results are from using the 3D FEA model for constant coefficient of thermal expansion (eT E) thermal stress analys i s. 4.1 Su m mer So l stice Cond itions Table 4.1 summarizes results from the Summe r Solstice conditions.

These r es ult s correlate with the hot daytime peak temperatures that occurred during the period from 195 9 to 20 04 in the Toledo , OH area. The " No W ind" condition r e moves heat b y convection.

The hot condition us es the high t e mp erat ur es measured for June of 10 4°F during the day 84°F a t night. The average conditions Ll se d t he average da y temperature of 83°F and 63°F at night. For a U cases gravity is also include. The contribution due to pressure loading from wind has been demonstrated to have a negligible impact on overall stress re sul t s. 2D Nastran Plane-Strain Time Slice P eak Stress 3 Nastran _ Thick Flute POliioll FEM Peak Radial Stress Architectural Notc h ID Case Description 1 Summer Solstice H ot No W i nd 7:30 PM +76 ps i -140 psi 2 Summer Solst i ce Hot 34 mph Wind 6:00 PM +46 PSi -68 psi 3 S ummer Sol s t ice Ave No Wind 7: 3 0 PM +69 p S i -126 PSi Ta ble 4.1 Summer Solstice -Summary Resu lt s for Radial Stress @ EL 683' 6" Figure 4.1.1 tlu'ough F i gu re 4. 1.3 show s ummary results listed in T abl e 4.1 for radial stress due to therma 1 transients . *********information.

P erformance Improvement international, LL C. Page 18 of 3 9 2012 Exhibit 56 R edacted EE*Iv-A ,\". Th erma l S i ress Ana l y sis: F ina l Repo rt Re sults & Co mm en ts...&a:> 'U' 0, =+ 76 psi Overall View @ EL 683' 6/1 Close-Up View @ EL 683' 6J/ Fi g ur e 4.1.Su mmer So l s tice H ot N o W ind 7:3 0 pm , Co n s t ant Co n c re te C TE = 5.2 0 xl 0.6 in/i nr3D _ F E M R a di a l S tres s Di s tribu t i oNot e: N on-Sy mm etr i c T h erm al St resse s D u e to U n e v e n/Hi g h e r II e a t ing G r ad i e nts o n So uth Fac ing P ane is . **********

i nformat i on. P erformance Im prove ment Inte rn a ti onal, LL C. 201 2 Page 19 of 39 Exhib i t 56 Redacted RliJ'ilIffiW Thermal Stress Analysi F inal Repo r t Results & Comme nts .> " +46 pSI Overall Vie w @ EL 683' 6/1 Close-U p View @ EL 683' 6" F i g u re 4.S u mm e r Solstice Ho t 34 mp h W ind 7: 30 p m , Constan t Concrete e TE = 5.2 0 x l 0-6 i nli n!3 D _ F EM R adial St ress D is t Note: No n-S ym me tri c The rma l St resses D ue to U n e v en lH ig her H e ating Gr adients o n S o ut h F ac in g Pa ne ls. ***********

inf orma ti o n. Perf ormance Impr ovemen t intern a tio na l , LLC. 20 1 2 Page 20 of 39 Exhibit 56 Redacted r.lIlilE 1eThermal Stress Analysis:

Final Report Results & Comments.', a&> C1; =+ 68psl Overa l l View @ EL 683' 6" Summer Solstice Ave No W ind 7:30 pm, Constant Concrete C T E = 5.20 xl O-6 in/in/oF 3 D _ F EM R adial S tress D is t r ibution N o te: N on-Symmetric Ther mal S t r ess e s D u e to Uneve n/H igher H ea ting Gradients on S o ut h Facing Pan els . Close-Up View @ EL 683' 6" liigure 4.1.3 **********

informati o n. Performance Impr oveme nt international , LLC. 2012 Page 21 of 39 Exhibit 56 Redacted Final Report Results & Comments .IIQM_Thermal Stress Analysis:

4.2 Win te r Solstic e & 1978 Blizzard C o n diti ons Table 4.2 s ummari zes results fi'om the Winter So l st ice and 19 78 B l i zz ard condi tions. The 1978 Bl i z zard com puted cold temperature s correlat e with the coldest daytime peak temperature of -24°F that occurred during the 1978 blizzard, 105 mph s out hwe st wind using low am bient temperatures, J 05 mph wind present, 1 20°F ste el se conda r y containment wall with grav it y includ ed. [Re ference Exhibit 65] 2D Nastran Plnne-Strain Time Slice Pe.ak Stress 3D NastraI1 1IIII F EM _ eak Ra dia l Stress Thick Flute P ortion .I\rchitectural Notch ID Ca se Description 4 Winter 197813 lizzard Record Low 5:00AM *79 p Si I +190 psi 5 Winter S o l s tic e A ve N o Win d 7: 30AM -20psi I +53 psi Table 4.Winter So l st i c e & 19 78 B li zza rd-S u m mary e s ult s for R adial Stre s s @ EL 683' Figure 4.2. 1 through Fi gure 4.2.2 show s um m ar y r esults listed in T allie 4.2 for radial stress due to thermal transients . *********information.

Performance Im provemen t internati on a l , LLC. Page 22 of 39 2012 Exhibi t 56 Redacted R.fEfiW ..A.,;., ,) Therm a l Stress Analysis: Final Report Results & Comments Radial Stress In Thick Portion of Aute Or = 70 psi O v era ll View @ EL 683' 6" F igure 4.2.1 Close-Up View @ EL 683' 6" 19 78 B li zz ard Condition 5:00 am , C onstant Concrete C IE = 5 , 20 X 10-6 i n/i ll/oF R adia l St ress D is tr i b ution ***********

information. Performance Improvement interna t iona l , LLC. 2012 Pa ge 23 of 39

. Radial Stress In ThICk Ponlon of Flute Op = -20 PS i Exhibit 56 Redacted ..Thermal Stress Ana lysis: _ Final Report Results & Comments""" JiR" Overall View @ EL 683' 6/1 Close-Up V iew @ EL 683' 6/1 Figur e 4.2W int e r S olstic e A v e Te m p e r a tures 7: 30 a m , C ons t a nt Co n cr e t e e TE = 5.20 x l 0-0 i nlinR a dia l S t ress Distrib u ti o**********

information.

Perio rmance Impro v ement intern ationa l, LLC. 20 1 2 Page 24 of 39 Exhibit 56 Redacted Fi nal Rep o rt Results & Comments .';1+11.Thermal Stress Analysis: 4.3 Autu m n Eq ui n ox Conditions Table 4.3 summarizes r e sult s from the A utumn Equinox condition s. Th e Autumn Equinox conditions li sted below correlate with th e hi g h S e pt e mb er conditions durin g the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> period w h en t e mp e ratures a re a t th e ir lowe s t and at 3:30 pm when the temperatures on the Southwest fac ing panel s are hi g he s t. The average and high September temperature s are computed with and without 34 mph wind condition present. 2D N as tIan Plane-Strai n 3D Nastrar FEM Peak Radial Stre ss ID Ca s e Descripti on Time S li c e Pe a k Str es s T hi c k Fl u te Po r ti o n Ar c hitectural Notch (; Autu m n Equinox H o t No Wind 6:00 PM +49 psi *100 psi Eguinox Hot 34 mph Win d 5:00AM *30 p si + 79 p s i 8 Autu mn E q uino x Ave No Wind 5:00 AM -20pSi + S8JlSi Ta ble Autumn Equin ox -Summary Re sults fo r R adial Str e ss @ E L 683' F igure 4.3.1 through Fig ure 4.3.3 show su mmary r es ult s list e d in 4.3 for radial stress due to thermal tran sie nt s . *********inf o rm a tion. Performanc e Im proveme n t internation a l, LLC. Page 25 of 39 2012 Exhibit 56 Redacted R{OPE* Thermal Stress Analysis: Final Report Results & Comm ents ik , , *_Ilt ,--"m, Close-Up View @ EL 683 1 6" A utum n E quinox H o t No W i nd 6: 00 pm, Consta nt C onc re t e C TE = 5.20 xl 0-6 in/in r F R adial Stress D i stribu ti o n I ..u"aa Radial Stress III Thick Porllon of FILJte 6 11 Overall View @ EL 683 1 F igure 4.3.1 ********** inf ormation.

Performance Impr ove men t international, LLC. 2012 P age 26 o f 39 Exh i bit 56 Redacted -FETr-Thermal Stress Anal y sis: Final Report Re s ults & Comment s A) . -t. .......

"' I&.H1 Radial Thick Portion of Flute =-30 ps i Overall View @ EL 683' 6/} Close-Up View @ EL 683' 6/} Figur e 4_3A utumn E quinox H ot 34 mph W ind 6:00 pm , Con st a ll l Concr e te CT E = 5.20 x 10-6 in/in r Radial Str ess

information.

Performance Impro v ement international, LLC. 2012 Page 2 7 of 39 Exhibit 56 Redacted WlOEJO-Thermal Stress Analysis:

Fi na l Report Results & Comments Radia l In Thick. Portion of Aute a. = *20 psi Overall View @ EL 683' 6/1 Close-Up View @ EL 683' 6/1 Figur e 4.3.3 Aut um n Equino x Av e N o W i n d 5:00 am, Co ns t ull f Co nc re te CTE = 5.20 xlO-6 in/i n f'R ad ia l S t ress D

information. Perform anc e Im provement i nternational , LLC. 2 012 Page 28 of 39 Exh i b i t 56 Reda ct ed Fi nal Report Results & Comments .iG'g.Thermal Stress Analysis:

4.4 Summary

R esults & Comment from 3D N ash"an Idealization The 3 D models show that the region of highest maximum principal st ress is at the outer most layers of concrete OF and inboard to the 1st rebar layer. From the OF layer ofrebar inboard, maximum principal stress levels drop off dramatically due to the high stress gradients.

T hese res ults sh ou ld in dic ate lite re g i ol1s of conc ern a t til e outer 2-3" of c oncre te 4.4.1 Su mme!' S o lsti ce C a ses >-The S/W facing panels and architectural flute s indicate the highest magnitudes maximum principal and radial > It is not believed the magnitude of radial stresses is sufficient to either delamination cracks or propagate any cracks that may be 4.....2 Win t er S ol s tice C ase s ;;.. For the normally occurring winter cold temperatures radial stresses in the thick pOition of the architectural flutes are low or compressive.

,. For the lo w te mperatures during the 1978 B liz z ard event the magnitude ofradial stresses in the " notch" cut-out of the architectural flutes is approximately 190 psi. > It is not believed the magnitude ofradial stresses is sufficient to either initiate delamination cracks or propagate any cracks that may be present. 4.4.3 Au tumn Equinox Cases )-The S/W facing panels and architectural flutes indicate the highest radial >-It is not believed the magnitude of radial stresses is sufficient to either delamination cracks or propagate any cracks that may be

• • • • • • • • • information.

Performance Improvement international, LLC. Page 29 of 39 2012 Exhibit 56 Redacted Final Report Re s u lts & Co mm e n t s Therm a l Str ess Ana lysi s: WI'EIiliW 4.5 Sum m ary Results: 3D Na stra n Idealiz atio n with Simulated 30'x30' " Cra ck" To investigate potential f or extended c rack gro wth in a pre-ex i s tin g crack re g i on, the 3 D N astran _ ide ali za tion was modifi ed to s imulate a 30' x 3 0' "C rack" The 30 ft x 3 0ft "fa iled" region It is de si r ed t o evaluate S/W facing flutes with an d w ithout the si mu lat e d "Crack" A s show n in Table 4.5 th e ma gn itude o f maximum principal stresses i nc rea sed a slig ht a moLlnt fro m C>MP= 162 p s i (No crack) t o 0M P= 18 4 p s i (w/crack). there is only a m a rginal incr ease in th e ma gn itud e of r ad ial s tress, fro m 0R= 76 psi (No crack) t o 0R= 92 psi (w/crack).

It is not b e li ev ed t hat th e in c r ease magnitudes in eith e r th e radial or m axim um pl'incip a l str esses are s uffi c ient to prop aga te c rac ks that may h ave for m e d. 2D Nastran Plane-Strain Time Slice P eak Stre ss 3D Nastrar FEM Peak Stre s s Values at ' Crack" R a dial S t r ess Ma x. Prine. Str ess ID Case Description 9 Sum mer Solstice Hot No Wind; 7: 30 PM + 76 ps i I +16 2 ps i 10 S ummer Solstice Hot No Wind; Crack 7:30 PM +92 psi J + 184 psi Table 4.Summer Solstice with S imulated 30'x30' "CSummary Re s ults for Radial Stress @ EL 785' 10Figure 4.5.1 thr o u g h Figure 4.5.2 s how v iew s o f the 3D Nastran ..FEM w ith th e simulated "Crack" region. Fi gure 4.5.3 shows summary stress re s ults li ste d in T a b l e 4.5 for m ax imum prin c ipal s tre ss due to s umm e r solstice thermal tran sien t s . *********inf orm ati o n. Perf or man ce Im prove ment int ernat ional , LLC. Page 3 0 of 39 20 1 2 Exhib i t 56 Redacted A) Thermal Stress Ana Report Results & Comments .,.. Fi gu re 4.5.1 " T hin-Crack" region intr o duced as ideali zed the "C racke d" bo undary at the O F Rebar ***********

informat ion. Performance I mprovemen t international , LLC. 2012 Page 31 of 39 Exhibit 56 Redacted .*Final Report Results & Comments A , Thermal Stress Analysis:

Simulated -Crack" Region g R= OF Rebar 859.61S"Azimuth &

Top

=

to at EL 800' 10Bottom = 190.5"to 114.S0at EL 770' F igure 4.5.2 "T hin-C rack" r e gion intr o duced as i dea li zed the "Crac k ed" boundary at t he Of Re bar ***********i nformation.

Performance Impr ovement inter na t ional, LL C. 2012 Page 32 of 39

• E x hib i t 56 Redacted Thermal St r ess Analysis: _ Final Report Results & Comments aun.. Max. Ptll'lcipal st r ess with Na crack" a w p == 162 pst Max. Pnt1clpal Stress at SHTllll.ted Crac k o,..p:; 184 pSI Fi gu re 4.5.3 Su mm er Solstic e H o t No W i nd 7:30 pm, C on s t allt Concr e te CTE = 5.20 x1 0-6 inlin f M axim u m P rincipal S tre ss Di s t r ibu***********

informalion.

Performance Improvement international , LLC. 2012 Page 33 of 39 Exhib i t 56 Redacted Final Report Resu l ts & Comments Thermal Stress Analysis .".. 4.6 Mi sc e llaneous Plane-Str a in Res ults -Overl apping Re i nf o rcement Both the plane-st rain and 3D models show regions of positive radial str e ss in the thick pOliion of the flutes. The average magnitudes of stress are about +350 psi. Sec f igure 4.6.1 The magnitude of these positive radial stresses are not believed high enough to cause cracks but the thick pOltion of the flutes is the only large region were radial stresses are positive.

It is known that there are regions where reinforcement overlaps in region s wh e r e tbe rebar either transitions to a different size or rebar of the same size is continued and the o ve d ap acts as a splice. T h e plane-strain models with overlapping rebar indicate that th e e ffe ct s of the localized stress concentration factor (SCF) around can be linked together to form a line of cracks. T he overlapping rebar also makes it difficult to fill voids due to large aggrl;gatc blocking distribution of concrete paste. Figure 4.6.1 shows results ofa parametric analysis to qualitatively view oiOtl1e eff ects when rebar is closely spaced or overlaps.

Localized cracks th a t may develop at the overlapping rebar (vertical

& hoop) could link to the adjacent SCF point since the distance to next pair of overlapping rebar isn't ve ry far. The o r lay e r is more susceptible to this crack propagation becau s e it is the OF layer where maximum pr incipal stress are highe s t. Overlapping rebar along the IF face doesn't have ma x imum principal stress available to propagate cracks. In addition, it should be noted that with the exception of the localized SCF peak tension radial stresses, regions immediately adjacent to the high tension s tresses show significantly lower stress... even negative (compression) values. It is b e lieved any localized cracks around the rebar would not propagate due to these thermal s tresses and the surrounding compressive s tres ses would arrest any localized cracks initiating due to the locali z ed SCF points . *********information.

Performance Impro ve ment international, LLG. Page 34 of 39 2012 4 , Exhibit 56 Redacted am@_ SCf effectS In Vertical

, nierface 1.290 36 Q70* PtANf FCl mut.h'

lII Pr opcrt.. 1 . Con I.*, t c n oll . Ccnc.C'tc I Plbtt Tep A ; 53 1.W1 l C' ("".dln4\1!!

S y-tC l"t l 1 t J oddJ I 3& =.96.2 1 , l c;d. U899

• 763.2745 Figu re 4.6.1 Typical Radial Stres s Contour i1-om 2D Plane-Strain Nastran W ith SC F Ef fects due to Overlap of Y ertical Re bar **********

informatio

n. Periormance Impr ovemen t international.

LLC. 2012 P a g e 35 of 39 Exhibit 56 Redacted Thermal Stress Analysis:

Final Report Results & Comments ..,.,. 4.7 Effects of Variable CTE=j{T} Two (2) of the concrete core samples from Davis-Besse Shield Building were sent to the United States Bureau of Reclamation (USBR) for mechanical and thermal properties testing. The coefficient of thermal expansion (CTE) was tested in accordance with the USBR test procedure 4910-92. The average value for CTE over the temperature range of approximately 33°p to 150 0 p given is CTE = 5.20 x 10-6 in/inJ°P... constant value. [Reference Pigure 2.1.4] The temperature range for Winter Solstice Average is indicated on Figure 4.7.1 (+27°P to +63°P) and shows that CTE= ffT} remains within the linear range ofUSBR data. Therefore during average winter conditions a variable CTE = f{T} will produce the same results as constant CTE. Figure 4.7.1 Qualitative Characteristics for nonlinear CTE= f{T} Average Winter Temperature Range Shown Figure 4.7.12 shows the assumed CTE= f{T} with the computed temperature range from 1978 Blizzard cold temperatures.

As shown on Figure 4.7.1 the temperature range from the 1978 Blizzard extends into the nonlinear range of the CTE = f{T} data suggested by Prof Xi. ********information.

Performance Improvement international, LLC, Page 36 of 39 2012 Thermal Stress Analysis:*****Final Report Results &Comments Exhibit 56 Redacted ..". Figure 4.7.2 Qualitative Characteristics for nonlinear CTE= f{T} Average Winter Temperature Range Shown For reference, recall Figure 4.2.1, "1978 Blizzard Condition 5:00 am, Constant Concrete CTE 5.20 xlO-Q iniinfF -Radial Stress Distribution", peak radial stress in the thick portion of the flutes are computed at GR= -70 psi. Figure 4.7.3 shows radial stress contour from the 3D Nastran _ FEM for the 1978 Blizzard condition assuming a variable/temperature dependent concrete CTE=f{T} similar to Figure 4.7.2. Results for radial stress in the thick portion of the flutes are GR= +470 psi for the 1978 Blizzard condition compared to GR= -70 psi when CTE is constant.

The temperature range shown during the 1978 Blizzard cold conditions

(-27°F to +32.4OP) does fall into the non-linear region when a variable CTE=f{T} is considered.

Therefore, if one views the 1978 Blizzard event as a catastrophic "once-in-a-lifetime" event then the concrete may have cracked way back then and the likelihood of another 1978 Blizzard is remote. The simulated "crack" model described in paragraph 5.0 was addressed using the variable CTE=j{T} . Analysis results did not show and significant change in the state of stress surrounding the simulated "crack" region when variable CTE=f{T} is used in place of the constant CTE. It should be noted the variable CTE=1'{T}

is based on "academic" predictions scaled to match the USBR test results and then extrapolated beyond the known test range. These results remain qualitative until conclusive data; precise material properties that is, are available.

These qualitative results do suggest strong evidence to support the hypothesis that the 1978 Blizzard event could be one of the primary contributors to the cracks. Exhibit 61: explores variations on the variable CTE concept . ********information.

Performance Improvement international, LLC. Page 37 of 39 2012 Exhib i t 56 Redact e d Thermal Stress Analysi Final Report Results & Comments A' " ;> Radial Stress if! ThlcJc of Flute o a =t 470 pS4 Overall View @ EL 683' 6" Close-Up View @ EL 683' 6" Figu re 4R adia l St re s s Contour from 3D Nast ran 1_ F E1 9 78 B lizz ard Co nd ition , V ar ia ble C oncr et e CT E = f{T} @ 93% S

inf ormation. Performance Improveme n t internationa l , LLC. 2012 Page 38 of 39 Thermal Stress Analysi Exhibit 56 Redacted .,.,..Final Report Results & Comments ./ The 3D Nastran models indicate the regions of interest for highest radial stresses are in the thick portions of the Flutes. The magnitudes of radial stresses from any of the thermal transient stress analysis are not sufficient to initiate or propagate cracks that may have formed \vithout another mechanism for crack initiation

& crack growth present. [Reference pages 19,20 , 21,35 & 38] ./ The plane-strain and other sub-models show localized peaks in the radial stresses resulting from stress concentration factors (SCF) around discontinuities.

These SCF effects can result :B:om (a.) overlap of adjacent reinforcing bar, (b.) abrupt change in stiffness l2teel-Concrete]

and (c.) thermal gradients with abrupt change in coefficient ofthermal expansion

(\":TE). [Reference page 35] ./ It is not unusual to have peak stresses at points where SC F's are known to e x ist and some localized dis-bonding of the concrete to the rebar may result. With the exception of these localized peak tension radial stresses , regions immediately adjacent to the high tension stresses show significantly lower stress ... even negative (compression).

L ocalized cracks that may develop around the rebar due to these SCF would not propagate due to thermal stresses alone and the surrounding compressive stresses would arrest any localized cracks initiating due to the localized SCF points . ./ The 3D Nastran _ models indicate stress gradients exist due to thermal transient conditions.

T he maximum principal stresses are largest at the outer most 2" -3" of concrete at the outer rebar layer. Thermal stress gradients lead to significantly lower stress as one move inboard the radial direction fi"om the OF toward the IF ofthe Shield Building wall. At approximately 6" -8" inboard of the outer most layers, radial stresses drop off to levels that would clearly not initiate cracks. [Reference Figures 4.1.1,4.1.2

& 4.1.3] ./ S ome qualitative results suggest strong evidence to support the hypothesis that the 1978 B lizzard event could be one ofthe primary contributors to the cracks. These qualitative resu Its indicate the low temperatures during the 1978 Blizzard may be a catastrophic "in-a-lifetime" that may have cracked concrete.

At date of release of this report these results remain qualitative and academic until conclusive data in the form of precise material properties are available, i.e. CTE=j{T}, allowing for a quantitative re-assessment of the 1978 8 1izzard condition.

[Reference Figure 4.7.3 ] *********information . Performance Improvement international , LLC. Page 39 of 39 2012 Exhibit 57: Temperature Dependent eTE © 2012. Performance Improvement International-Appendix Exhibit 57 Temperature dependent coefficient of thermal expansion (CTE) Under low temperatures, concrete may expand (instead of contract) during a cooling period. This possible expansion is due to ice formation in the concrete.

During a severe cooling process, the temperature of concrete in outer layer of the cylindrical wall is lower than that of inner layer. So, ice may form in the outer layer of the wall resulting in an expansion and ice may not form in the inner layer of the wall leading to continuous contraction.

This special outer-expansion-and-inner-contraction deformation pattern can result in a tensile stress in the radial direction of the wall. Delamination cracking may occur in the case of excessively high radial tensile strength.

The key issue here is the coefficient of thermal expansion (GTE) of concrete under low temperatures.

Did the Davis Besse concrete in outer layer of the wall expand during the blizzard?

If yes, how much did it expand? Was the expansion high enough to cause the cracking?

These questions will be discussed in the following sections.

The effect of temperature on CTE of concrete GTE of concrete depends on temperature.

The reason is that the state of moisture inside of concrete depends on temperature.

Under elevated temperatures (e.g. fire), liquid water turns into vapor which generate high vapor pressure.

This case is not within the scope of this project, and thus will not be discussed further. Under low temperatures, liquid water or vapor turns into ice which is associated with a 9% volume expansion.

After all moisture freezes, the effective GTE of concrete is a mixture of GTE of concrete and GTE of ice. The GTE of ice is about 5 times of the GTE of concrete without ice. Therefore, at a very low temperature, the GTE of concrete is not the same as the GTE of concrete at room temperature.

Accurately speaking, the effective GTE of concrete depends both on temperature and on ice content, and thus on moisture content. The effect of temperature will be discussed first, and the effect of moisture will be discussed later. _ 40 ., .c: 'E g BO g <ii t;; 160 TIME {h) Fig. 1 A typical strain and temperature chart (ASTM G671) Literature review showed that the ice formation starts at QOG and completes at about -15°G or lower depending on the microstructure of concrete.

This freezing process is due to the fact that the freezing point of water depends on pore size in concrete.

The freezing point of water in large air voids is close to QOG, and thus the water in large air voids freezes first; and the freezing point in small pores could be well © 2012. Performance Improvement International Page 1 Page 1 of 7 Exhibit 57 below O°C, and the water in small pores freezes after the water in large voids. For a continuous cooling process, the strain of a concrete specimen is shown in Fig. 1, contracting first, expanding due to ice formation, and then contracting again after the completion of ice formation.

In Fig. 1, the slope of the strain vs. tem perature curve is CTE. It is very clear that the slope of the curve, the CTE is not a constant.

Depending on concrete mix design and cooling condition (temperature range and freeze-thaw cycles), the curve could be significantly different, as shown in Fig. 2. The shape of the curve, Le. the temperature dependency of CTE is closely related to the internal structure of concrete.

4 LENGTH CHANGE MEASUREMENTS AFTER MULTIPLE FREEZE-THAW CYCLES (MORTAR BEAMS, w/c; 0.32. AGE TWO WEEKS)

TEMPERATURE (0C) Fig. 2 Test data of concrete length changes under low et aL 1988; see Refs. 18 and 19 for test results from T.C. Powers and R.A. Helmuth) Experimental studies showed that properly air-entrained mortarf; contract upon freezing, while entrained mortars or improperly air-entrained mortars expand. The expansion of the latter is attributed primarily to hydraulic pressure, owing to the rapid growth of ice,'which nucleates at low temperatures in laboratory samples (Sun and Scherer 2010). Moisture in concrete Because of very small pores in concrete, water in the pores exists as a mixture of liquid and vapor in above ground concrete structures.

Internal relative humidity (or pore relative humidity), RH of concrete are often used to represent internal moisture state of concrete.

The internal RH can be correlated to the moisture content in concrete (the weight of moisture in concrete) by adsorption isotherms.

Adsorption isotherm is a relationship among weight of internal mOisture, temperature, and RH. Fig. 3 shows experimental results of adsorption isotherms in the literature in comparison with the predictions of a theoretical model developed by Xi et al. (1994). In the figure, the horizontal axis represents RH (where p is water vapor pressure and Ps is the saturation pressure at a given temperature);

and the vertical axis stands for moisture content in gram of moisture in gram of concrete.

So, with a given concrete mix Page 2 © 2012. Performance Improvement International Page 2 of 7 Exhibit 57 design, temperature, and RH, the moisture content in the concrete can be obtained from the adsorption isotherms.

It is important to note that RH = 100% does not mean all pores are filled up by water, it means the vapor pressure in the pore reaches the saturation pressure of vapor at the given temperature.

At this stage, the concrete reaches its adsorption capacity, which is different from absorption capacity.

There may be only limited layers of water molecular covering the surface of pore walls at RH ;:: 100%. When all pores are filled up by water, the concrete reaches its absorption capacity_

0."" ....----------, 0 . .10 C).o.:a o .5 s: 0.10 "-"" 0.0 '1110" 0582 Typ/!

• c:ernet\1 T .. 2$4'K t-26d1ya o:z I'OWIII'I lit aI.

0 .* 1.0 0.20 '1110-0.45 T -2911" K t-2555 day. C.! w/c";'O.ll Type I CIlment 1-7day3 MUc,,*1 et aI. (tIil15) 0." .,-------------, 0 ..

---

, Fig. 3 Adsorption test data and comparisons with predictions pf a theoretical model (Xi et al. 1994) For above ground structures such as Davis Besse containment structure, it is more suitable to use RH and adsorption isotherms for estimating the internal moisture content. For under water structures, the absorption capacity is a better indicator.

The effect of moisture on CTE The general trend is that the higher the initial moisture content before ice formation, the larger amount of ice formed in the concrete, and possibly the larger dilation of concrete during the ice formation process. The following figures show available test data in the literature.

© 2012, Performance Improvement International-Page 3 Page 3 of 7 Exhibit 57 4 ; .r ,,,....0'4';',,.

l (lflf!. 'Of , , * . o Z #tIC .. 0' SO -0\ PlAIN Mj'X($ 15--20 Fig. 4 Temperature vs. dilation of concretes at various levels of saturations (Grieve et al. 1987) .!. ... .! Q -100 300 200 100 0 ...'" :; I ...; " "" Ie.perature

("I::) 100 95 .!. :::: IS :;; "Iii j Q , .. "" -100 l "" : 0:: .::. 1: 2 :0.65 -200_ 20 10 15 20 Te.,>er.tur.

("c) 100 95 *r gO ....!. ... IS :;; ... 85 j.. Q ...; 80 To.perature

("c) Fig. 5 Dilatation of concrete and internal relative humidity (Zhou and Mihashi 2008) © 2012. Performance Improvement International Page 4 Page 4 of 7 Exhibit 57 In Fig. 4, there is almost no expansion when the saturation level in the concretes is below 90%. However, in Fig. 5, there are significant expansions at low temperatures even the initial RHs in the concrete samples are below 90% (see the first and the second figures, the RHs are about 88%). Fig. 6 shows the dilations of concrete under dry and wet conditions.

One can see from Fig. 6(a) that the concrete at wet-freezing condition expands significantly (near 8 degree C), while the concrete at freezing condition does not expand at all. Therefore, the moisture content has a major effect on the expansion of concrete during ice formation process, and the extent of moisture effect depends on internal structure of the concrete.

The internal structure of concrete depends on mix design, curing conditions, and age of the concrete.

In Fig. 6(b), even if at wet-freezing condition, the concrete with proper entrainment only expands slightly near 8 degree C). o .(j ... *2(Wl .g

-. *600 Iro, ;:..'i: .." " C *700 --w-ct -Ib:'l,.'zing (a) wlc = 0.5 non-air-entrained concrete (b) wlc =0.35 air-entrained concrete.

Fig. 6 Influence of saturation on dilation of concrete (Zuber and Marchand 2004) In summary, at a sufficiently high RH level and under a continuous cooling process, there are three possible types of temperature-dependent thermal strains for concrete:

Type 1 -Contraction, significant expansion, and contraction, as shown in Fig. 1 and Fig. 6(a). Type 2 -Contraction, slight expansion, and contraction, as shown by solid squares in Fig. 2 and Fig. 6(b). Type 3 -Contraction, as shown by hollow circles in Fig. 2. From Exhibit 52 Univ. of Colorado lab test report, the relation of thermal strain and temperature of Davis Besse concrete follows Type 3 when the sample is dry, and Type 1 when the sample is wet. So, Davis Besse concrete does expand under low temperatures.

/' ,i -:tIJ\) .:§ © 2012. Performance Improvement International-PageS Page 5 of 7 Exhibit 57 The resulting temperature dependent CTE at the temperature ranges are: T> 23°F (-5°C), CTE = 5.2 x 1Q-6 rF (The same as USBR test 8.6°F (-13°C) < T < 23°F (_5°C), CTE = -4.94 x 10-61.4°F (-1rC) < T < 8.6°F(-13°C), CTE =-43.1 x 1Q-6Below 1.4°F (-1rC), CTE = 5.2 x 10-BrF Crhe same as CTE under room Conclusions CTE of concrete depends on temperature, internal moisture, arid internal structure of concrete.

Some concretes may expand during the ice formation process if their internal moisture content is sufficiently high. Davis Besse concrete showed expansive strains under low temperatures (Exhibit 52). Therefore, a tem re CTE was developed

© 2012. Performance Improvement International Page 6 Page 6 of 7 Exhibit 57 References Bazant, Z.P., Chern, J.C., Rosenberg, A.M., and Gaidis, J.M. (1988) "Mathematical Model for Thaw Durability of Concrete", Journal of American Ceramic Society, 71(9), 776-783. Grieve, R., Slater, W.M., and Rothenburg, L. (1987) "Deterioration and Repair of Above Ground Concrete Water Tanks in Ontario, Canada", Research Report to Ontario Ministry of the Environment, Golder Associates and W.M. Slater & Associates, Inc. Sun, Z., and Scherer, GW. (2010) "Effect of Air Voids on Salt Scaling and Internal Freezing", Cement and Concrete Research, 40, 260-270. Xi, Y., Bazant, Z.P., and Jennings, H.M. (1994) "Moisture Diffusion in Cementitious Materials:

Adsorption Isotherm", Journal of Advanced Cement-Based Materials, 1,248-257.

Zhou, Z.Y., and Mihashi, H. (2008) "Micromechanics Model to Describe Strain Behavior of Concrete in Freezing Process", Journal of Materials in Civil Engineering, ASCE, 20(1), 46-53. Zuber, B., and Marchand, J. (2004) "Predicting the Volume Instability of Hydrated Cement Systems upon Freezing using Poromechanics and Local Phase Equilibria", Materials and Stmctures, 37, 257-270. © 2012. Performance Improvement International-Page 7 Page 7 of 7 Exhibit 58: Carbonation Lab Testing © 2012. Performance Improvement International Appendix Exhibit 58 Performance Improvement Providing a competitive advantage through research and applications To: From: Date: 02/27/2012

Subject:

Laminar Cracking of Davis-Besse Shield Building -Concrete Sample Testing for Carbonation

-Based on my observation and examination of concrete-core samples received from the Besse Shield Building, my findings for Carbonation are detailed in what follows. Page 1 of 9 Exhibit 58 Carbon atio n in Concret e C a r bonati on in Concr e te Laboratory tests and examinations were conducted on several concrete core samples to determine the extent of carbonation within the samples The cracked concrete samples, which are vulnerable to carbonation , were isolated a nd fractured in a plane perpendicular to the original cracked surface. Figures A1 and A2 show examples of the carbonation depth as m easured from the exterior surface. The exterior surface is the portion of the shield buildin g that is exposed to the elements; it is the outer diameter surface. P ag e 2 Page 2 of 9 Exhibit 58 figure A 1: Fracture Sample with Carbonation Laver (Core F2-790.0-4.S)

Figure A 2: Fracture Sample with Carbonation layer (Co r e F3-1) Page Page 3 of Exhibit 58 The following table shows the nominal carbonation depth as measured from the exterior surface. The table lists the 16 samples used in determining the average nominal carbonation depth which , as previously stated, is 8.57 mm. Table A 1; Nomina l Carbonation

-layer Depth from Exterior Surface (Carbonation Rate Determination)

Core Sample F3-1 5 11-1 511-2 S1 2-1 5 12-2 516-3 S5-1 55-2 57-1 I Nominal Ca r bon ati on Depth From Exterior Surfac e (re f erence), m m '11.7 9.33 10.00 8.59 8. 33 7. 73 7.90 7.87 7.7 5 57-2 57-3 S9-1 5 9-2 S-7-656.5-6.5 5-9-6 53-1 5 785-22.5 AVERAGE 9.07 7.56 10.05 7.65 8.84 8.65 6.06 8.57 Longitudinal Fr acture Carbonation analysis was conducted for both longitudinal and transverse cracks. Figure A3 shows a longitudinal crack for reference

.. As can be seen , the longitudinal cross-section is defined as the plane that is parallel to the longer dimension of the core sample. Several longitudinal cracks with no evidence of carbonation were evidenced.

For example, Figure A4, Core F2-790.0-4

.5 , shows that a distance of 7 inches from the surface , no carbonation is detected. ' Page 4 P age 4 of 9 Exhibit 58 Davis Bes s e Nuclear Plant F2*790.0-4.5 "* .. " . . II ' . \I ** ,. , .. t l Iimiml Figure A 3: longitudinal Crack (Reference)

Figure A 4: Longitudina l Crack with no Evidence of Carbonation Pag e Page 5 of Exhibit 58 The Following table shows several samples with longitudinal cracks at various distances from the exterior surface. For the samples in the table listed below , ihere is no carbonation layers formed at any of the various distances within each re sp ective sample. Table A 2.: Carbonation Results from several Samples with longitudinal Cracks (No Carbonation)

Core Samp le Crack Dista n ce From E x1e rior Su rface Maxim u m Carbonation Depth, mm F2-790.0-4.S 17" Lo ng C rack Longitudinal Cracl< #1 T' 0 F2-79 0.0-4.S 17" Long C rack Longitudinal Crack #1 1005" 0 F2-7 9 0.0-4.S 17" L o n g C rack Longitudinal Crack #1 13" 0 F2-7 9 0.0-4.S 1 7" Long C r ack Longitudinal Crack #1 16" 0 F2-7 90.0-4.S 17" Long Crack Longitudinal Crack #1 20" 0 F2-79 0.0-4.S 5.5" Long Cr ack 45 Degree Crack #2 21"t022" 0 F2-790.0-4.S 5.5" Long Crack 45 Degree Crack #2 23" to 26" 0 F4-794.0-3.S 19" Long Crack Longitudinal Crack #1 9" 0 F4-79 4.0-3.S 1 9" Long Crack Longitudinal Crack #1 13" 0 F 4-794.0-3. S 19" Lon g C rack Longitudinal Crack #1 17.5" 0 F4-794.0-3.S 1 9" Long Crack Longitudinal Crack I #1 20" 0 F 4-7 9 4.0-3.S 1 9" L ong Crack Longitudinal Crack #1 23" 0 F4-7 9 4. 0-3.S 19" Long C ra ck Longi tudi nal Crack #1 26': 0 F S-791.0-4 9" Long C r ac k Longitudinal Crack #1 7.5" 0 FS-791.0-4 9" Lo ng C rack Longitudinal Crack #1 9'" 0 FS-79 1.0-4 9" Long C rack Longitudinal Crack #1 11 " 0 F S-7 9 1.0-4 9" Lon g C rack Longitudinal Crack #1 14.5" 0 8 2-798.S-4.S 5" L ong C rack Longitudinal Crack #1 1" 0 5 2-798.5-4.S 5" Long Crack Longitudinal Crack #1 2.5" 0 5 2-798.5-4.5 5" Long Crack Longitudinal Crack #1 4" 0 P age 6 Page 6 o f 9 Exh i bit 58 The following table shows the results from carbonation analys i s on several core samples with longitudinal cracks. Table A 3: Carbonation Analysis on Longitudinal Cracks Core 5ample -. Crack Longitudinal Crack Length Maximum Carbonation Depth, mm F3-1 #1 7" 5.4 511-1 N/A N/A N/A 5 11-2 #1 0 5 11-2 #2 2.09 512-1 #1 5 0 5 1 2-1 #2 1 Yz" 2.70 512-2 #1° 1 % 2.99 5 1 6-3 N/A N/A N/A 55-1 N/A N/A N/A 55-2 N/A N/A N/A S 7-1 #1 2 0 5 7-2 N/A N/A N/A 57-3 N/A N/A N/A 5 9-1 N/A N/A N/A 59-2 N/A N/A N/A ' N/A = N o Lo ngitu d in al C r'Cr a c k Fou nd U po n Sectio n iTra n sv e r se Fractu re Figure A6 shows an example of a transverse fracture sample , Core S5-2 , which was determined to have a measured carbonation layer of 0.455 mm. For comparison with Figure A6, Figure A7 shows an example of a transverse fracture sample, Core S7 -656.5-6.5, which was determined to have no carbonation layer. Pa ge 7 Page 7 of 9 Exhibit 58 Figure A 5: Carbonation Detected on a Transverse Fracture Sample (Core 55-2), 1.8 yrs. 8F i gure A 6: Transverse Fracture Sample with No Carbonation (Core 57-656.5*6.5) Page Page 8 of Exh i bit 58 The following table is a list of transv e rse fracture core samples along wi t h their associated ca r bon a tion depth. Core samples with an asterisk in their identification number are t hose wi t h transverse c r acks as identified by the plant IR Inspection.

Table A 4: Transverse Fracture Carbonation Analysis Core Sample Fracture (Crack) Distance From Surface, inch e s Maximum Ca rbonation Depth, mm F 3-1 #T1 7 0.58 2 S 1 1-1* #T1 8 0 S 1 1-1* #T2 5 1'2 0 5 1 1-2 #T 1 8 0 S 1 2-1 #T1 21 0 S1 2-2* #T1 51'2 0.264 S1 2-2* #T2 16 1'2 0.6 8 6 S1 6-3* #T1 14 1'2 0.343 S 1 6-3* #T2 15 1'2 0 S 16-3* #T3 2 1 1'2 0 S 5-1 * #T1 9 0.500 S5-1* #T2 91'2 0.897 S5-1* #T3 14 1'2 0.60 4 S 5-1* #T4 16 0.893 . S 5-2 #T1 12 "/8 0.445 S7-1* #T1 0 S7-1* #T2 10 0 S7-1 * #T3 16 % 0 S7-2 #T1 15 % 1.4 2 57*3* #T1 6 1'2 0.710 S 9*1 * #T1 4 1'2 0.329--*._-0 S 9*1* #T2 12 S 9-2 #T1 10 0.388 " C o re sa m p les with transverse cracks as identifie d b y the pl a nt I R in spection P a g e 9 Page 9 of 9 Exhibit 59: Test Report from the United Bureau of Reclamation Appendix VIII-60© 2012. Performance Improvement United States Department of the Interior BUREAU OF RECLAMATION P.O. Box 25007 Denver, Colorado 80225-0007 IN REPLY REFER TO: JAN 192012 86-68180 RES-3.40 MEMORANDUM Performance Improvement 2111 S EI Camino Suite Oceanside, CA Attention:

Dr. Chong Katie Bartojay, P.E., Civil Engineer, Materials Engineering and Research Laboratory Group (MERL)

Thermal Properties Testing Results Materials Engineering and Laboratory Report No. MERL-2012-02 INTRODUCTION Six concrete core samples were delivered to the Bureau of Reclamation's Materials Engineering and Research Laboratory (MERL) on December 19, 2011. The specimens were identified as F4-791-2.5

1. 1 through #4 and S7-782.0-8.5

1. 5 and #6. AU six specimens were approximately 2.5-inches in diameter and 4-inches long. The submitted samples were tested for thermal diffusivity, specific heat, and thermal coefficient of linear expansion testing on concrete cores. Conductivity was calculated using the specific heat and diffusivity results. CONCLUSIONS AND DISCUSSION Thermal Diffusivity Thermal diffusivity measures the rate at which temperature changes take place in concrete and is defined as an index of the facility with which a material will undergo temperature change [i]. Thermal diffusivity was tested in accordance with Reclamation's test procedure USBR 4909-92, "Thermal Diffusivity of Concrete" (with modifications to account for upgraded equipment).

Two concrete core specimens marked S7-782.0-8.5

1. 5 and S7-782.0-8.5

1. 6 were tested over three temperature ranges: 35°P to 75°P; 75°P to 115°P; and 115°P to 155°P. A small diameter hole was drilled from one end to accept a Exhibit Page 1 of 5 Exhibit 59 thermocouple to be located at the approximate center of the specimen.

The hole was filled with epoxy before testing. Specific Heat Specific heat is the amount of heat required to raise the temperature of a unit mass of material one degree [i]. Specific he a t was tested in accordance with Reclamation's t s t procedure US B R 4907-92, "Specific Heat of Aggr egates, Concrete, and Other Materials" (with modifications to account for upgr a ded equipment).

Two c n crete core specim e n s marked F4-791-2.5

1. 1 and F4-791-2.5

1. 2 were tested over a t e.mp erature range of approximately 35°F to 150°F. Thermal Conductivity Conductivity is the rate at which heat is transmitted through a unit thickness of material.

The coefficient of thermal conductivity (K) represe nts the uniform flow of heat though a thickness of material when subjected to a unit temperature difference bet w ee n t wo faces [i]. Thermal conductivity was calculated from the specific he a t (c), diffusi vity (0), and concrete density (P). The harden e d den s ity determined from this study was used in this calculation.

K =cpo [ii] Thermal diffusivity, speci fic heat and conductivity tests results are summarized in Tabl e 1 and reported graphically in the Attachment.

Table 1 -Summary of tbermal properties of select cores " Temperature

.... . (oF) . SQecific Heat (c} Btu/(lbm*oF)

F4-791"2.5

1. 1&#2 Diffusivitv (d} felhr '. S7*782.0*8.5 Conduc.ivi!ll l!!ll Btu/(tf*hr*oF/ft) i--' Calculated

__ 50 100 150 0.478 0.428 0.378 0.054 0.049 0.044 3.79 3.08 2.44 Typical ranges of these thermal properties for nonnal concrete[ii]

are approximately:

• 0.02 to 0.06 ft2/hr for Diffusivity

• 0.20 to 0.28 Btu l lb per OF for Specific Heat

• 0.8 to 2.1 BtuJft 2 lhr °F/ft for Conductivity The spec ific heat values measur ed for the submitted specimens were not in the typical range for normal concrete.

The calculated conductivity was also outside the ran ge for normal concrete.

Coefficient of Linear Thermal Expansion Thermal coefficient of linear expansion is the change in a unit length per degree of tempera ture change of the concrete [iii]. Thelmal Coefficient of E xpansion was tested in accordance with Reclamation

's test procedure USBR 4910-92, "Coeffic ient a/Linear Thermal Expansion" (with modifications to account for upgraded equ ipment). Two Exhibit 59 Page 2 of 5 E xhibit 59 concrete core specimens marked F4-791-2.5

1. 3 and F4-791-2.5

1. 4 were tested over a temperature range of approximately 33°F to 150°F. Coefficient of linear thermal expansion tests results are summarized in Table 2 and reported in the Attachment.

Table 2 -Summary of coefficient of linear thermal expansion Specimen 10 .Average Coefficient Of* .. LinearThermal .

• Expansion. (InchllnchrF)

F4-791-2.5

1. 3 5.2 x 10-6 F4-791-2.5

1. 4 5.1 x 10.6 Average 5.2 x 10.6 The coef fic ient of linear thermal expansion of concrete varies greatly with aggregate mineralogy and can be as low as 4-x 10.6 per degree F to as high as 13 x 10.6 per degree F[ii]. The values determined by this testing are in the range for normal concrete.

The test results derived from this work shall not be used to imply endorsement b y the Bureau of Reclamation or the U.S. Government and cannot be used for advertising or commercial purposes.

Attachments cc: Dr. Yungpin Xi, University of Colorado, yungpin.xi@colorado.edu (electronic copy) CJ Concrete, Mindess and Young, Prentice-Hall , lnc., 1981 [U] " Properties of Concrete, Fourth Edition" A.M. Neville, Pearson Education Limited, 2009. (;;;] "Concrete Manual, Part I, Eighth Edition", A Water Resources Technical Publication , U.S. Department of Interior , Bureau of Reclamation, Denver, CO, 1988 Reprint. Exh i bit 59 Page 3 of 5

---

Exh i b i t 59 PII Co r e Specific H e at T est 0--E ..0 ::J Ci:J .... '" Q) J: '0 (\J r./) 0.055 0.050 0.045 .... .::; 'Vi OJ 0.040 :t: is 0.035 0.65 0.60 0.55 0.50 0.45 0.40 0.35 0.30 0.25 0.20 I I y =-O.OOlx + 0.5277 1 -I . ..... R2 =0.2978 1 I .... I ./ *j u " I ... .......-* --.".", --r.' *

• I , 40 60 80 100 120 1 4 0 160 Te m g erat ure n

• Core 1

• Core 2 -Linear (Average) 0.060 P II C or e Diffu s i v i ty --,-__-,____ -+------1 1------\-----+----

1-----1 ......ii+/-::=_---I-----+_---+---__l I-------------+-t-----+.-]f--.-I------l 0.030 40 60 80 100 120 140 160

• Core 5

• Core 6 Average -Linear (Average)

P II Core Co n du c tivit 4.50 ,-----,----,-4.00 3.50 3.00 2.50 -f------*/-----f-----r----+----

---O-----I 40 60 80 100 120 140 160 Calculated Exhibit Page 4 of 5 Exhibit 59 "J CI lJ L....r wITtl 0.0007 I 0.0006 -\,---Specimen)

=

Specimen 4 =

Dl ' *r e 1 l! If: n LlIlIH J ,.JL d U. -1 II, I=' 8 0.0004 .,.... 'j./ M 0.0003 l h '/ --Specimen 4 Specimen 3 P -Average ..J 0.0002 0.0001 -j --l 0.0000 \ " '/ o 20 40 60 80 100 120 140 160 180 cJLL r IT ILJI GJ ::T r HL T ,-1. ' Project: PI I Cores Test Date: 12/21/2011 Test Age: not known Bureau of Reclamation Materials Engineering and Research Laboratory Exhibit 59 Pag e 5 o f 5 Coefficient of Thermal Expansion of High Moisture Concrete The coefficient of thermal expansion (CTE) of high moisture concrete is a highly nonlinear function of temperature.

This is associated with the 9% volume expansion of the freezing of entrapped water. The freezing of water in small concrete pores does takes place at a lower temperature than 32°F due to surface tension which prevents the rearrangement to form ice. The end effect is that water in concrete freezes at varying temperatures depending on the pore size. This nonlinear dependence of the CTE with temperature is shown in Exhibit 57 and used as an input to the finite element analysis presented here. Tests of moisture penetration were also performed at the University of Colorado at Boulder, which showed that a 1-0 depth of water penetration up to 3 or 4 inches is possible when there are winds in excess of 90mph (such as during the 1978 blizzard).

Page 4 Figure 1 below shows the location of ********************

I Vicinity of Flute Studied in Figure 1-Shield Building with Flute Numbers and Azimuth Locations

_ Page 5 Exhibit 60: Test Report from the University of Colorado Appendix VIII-61© 2012. Performance Improvement International Exhibit 60 University of Colorado Dept. of Civil, Environmental.

& Architectural Engineering College 01 Engineering and Apphed SCience t 303 492 428UCB f Boulder, Colorado 80309-0428 Exhibit Page 1 of Exhibit61:

Stress State during the 1977 and 1978 Blizzards

'! Appendix VIII-62© 2012. Performance Improvement Stress State during the 1978 and 1977 Blizzards Table of Contents Summary of Results .................................................................................................................................Modeling Summary .................................................................................................................................Overall Approach .................................................................................................................................Finite Element Software ......................................................................................................................Temperature Conditions

......................................................................................................................................................

................................................................................... Modeled Geometry .............................................................................................................................Material Properties

..............................................................................................................................Coefficient of Thermal Expansion of High Moisture Concrete ....................................................................................................................................................................................... Circumferential Temperature Distribution at O.F. Horizontal Rebar ..................................................... ....................................................................................................................................................................................................................................1978 Blizzard Condition

..; ..........................................................

....................................................... 1977 Blizzard Condition

.....................................................................................................................-. Page Summary of Results The results of the analysis presented in this report can be summarized as follows: The blizzard of 1978 produced stresses above the tensile strength in the hoop direction, likely resulting in damage. The area exceeding the tensile strength is confined to a circumferential plane at the depth of the outer face main cylindrical wall under the raised shoulders. The 1977 blizzard shows significantly lower stress compared to the blizzard of 1978. The hoop stress approached the tensile strength of the concrete and it is limited to a small area. For these reasons only minor damage, if any, is predicted.

Modeling Summary Overall Approach * -Page 2

---

.. Finite Element Software was used exclusively in the finite element analysis presented here. Temperature Conditions The following two temperature conditions are presented in this report. The details of the temperature conditions and the selection of the time of day are summarized separately in the Root Cause Analysis Report: I} Low temperature during the 1978 blizzard (105 mph wind, winter solstice, 5:00 2) Low temperature during the 1977 blizzard (76 mph wind, wintersolstice, 5:00 Expansion of concrete due to freezing of entrapped moisture was studied in the _ _ . This model is utilized to determine the stress state in a subsection of the structure spanning from the middle of one panel to the middle of the adjacent panel. The raised shoulders and the flute geometry are included in the model. Nominal steel reinforcement is included using a technique called The detailed stress concentration at the steel and concrete interface is not included in the model. Modeled Geometry The drawings used as geometry input for this model are:

• Drawing No: C-100 Rev. 5 "Shield Building Foundation Plan & Details SH. I"

• Drawing No: C-110 Rev. 6 "Shield Building Roof Plan Wall Section & Details" All vertical reinforcing bars in the containment shell section are modeled as rebar #10 (diameter 1.270") at 12" center to center spacing. The inner face horizontal rebars are #8 (diameter 1.000") at 12" spacing. The outer face horizontal rebars are # 11 (diameter l.4!0") at 12" spacing. The vertical and horizontal rebars in the shoulder sections are #8 at 12" spacing. Material Properties The material properties used as input to the finite element analysis in this report are summarized in the following documents attached to the Root Cause Analysis Report:

• Exhibit 56, Figure 2.1.4: Material Properties for Davis-Besse 3D 11I11IIII11III111 I Model

• Exhibit 56, Section 4.7: Effects of Variable CTE

• Exhibit 57: Temperature dependent coefficient of thermal expansion (CTE) Page 3 ture Distribution at O.F. Horizontal Rebar (See Exhibit 65) The temperature profiles around the Shield Building at the outer face horizontal rebars are shown in Figure 2. The figure shows 8 sets of double peaks for each temperature profile. The double p eaks represent the warmer temperature under the shoulders.

The temperature is warmer under the shoulders becau se there is a thicker layer of concrete at those locations which reduces the heat loss to the exterior during the blizzards.

Temperature (OF), Mid-Height, Outer Face Horizont al Re bar Depth 1 6 . -1917 Blizzard Temperature Calculation (Worst C Clse) -197 8 Blillard Temperatur e Ca l cula li on (W orsl Case +2 0' F) 22.S 67 5 " , " 2 1 7 5 ". ,.. Figure 2 -Circumferential Temperature Distribution at the O.F. Horizontal Rebar Depth Page 6 Figures 3 through 5 below depict the geometry and finite element mesh of the-Figure Geometry and Rebars Page 7 Figure Detail of Flute Region Figure 5-. Detail of Flute Re g ion with Mesh Pa ge This section summarizes the results _ is used to make predictions about the delamination propensity due to the two blizzard conditions.

This model does not attempt to make predictions of stress concentration effects around the included reinforcing bars due to lack of detail at the concrete/steel interface.

The tensile strength of the Davis-Besse concrete i s in the range of 836 to 962 psi. The contours in the stress figures in this section are assigned an upper limit of 900 psi. A tensile stress exceeding 900 psi is indicated by li ght grey contours in the stress figures. The interpretation of any light grey area in the contour plots below is that damage may occur in that area. The damage that results from any tensile. stress above the strength of the concrete depends on 3D stress state as well as the strain ene r gy available to open the crack. Low strain energy results in microcracks and high strain energy results in more microcracking and eventually a structural crack. The stress contour results shown in this sect ion can be summarized as follows: Higher tensile stress and larger stressed areas is predicted in the 1978 blizzard compared to the 1977 blizzard Blizzard of 1978: Tensile stresses high enough to damage the concrete is predicted The high stresses are distributed over l arge areas in the observed crack locations under the thick sections of the shoulders and not in the thinner sections in the flute and panels Blizzard of 1977: Tensile stresses are lower or equal to the strength of the concrete The highest tensile strength are confined to small areas under the thick sections of the shoulders Page 9 1978 Blizzard Condition The result **************

due to the 1978 bli z zard condition is shown in this section. The temperature contours can be s een in Figure 6 and the stress results is shown in Figures 7 through 12. rHll + S 210e+O-+4.7 4 1e. +4 276e-+3 , 80g e +.. + 3.J42-+2.87 5 e -+2.4 0 7e +01 +1.0 06 e +0-+S. + 7. t4 !le-Ol -3.95 1e+O O O DS: m 1 22 J , od b Ab a qusjS t;., n d a,d 6.10-3 Feb 21 18: 00: 59 P;, cifi c Ti me 20:1 2 S tep: Step-l Incr e ment 1: Step Time::: Primary V lIr: N TD efo rmed V a,: U D e f o rmac o n Sc a le Fact or: +S.Figure 6 -Temperature (OF) during the Blizzard of 1978; Deformation Scale Factor 500X Pag e 10 S , (*lu:*:. Prlncl pa t (Avg : 7S%) +4.72.2e+03

+9.00 Q'H 02 +7. 5D Oe+02 +6.+4. . +3.0 0 06++1.S0 (}@t+0 .00 0..--1 , S O O e -l 02 -3.000e-4.S 00 e +-6.0 00'H --9.0006+-1. 87 G e+008: m1 2 23.odb Aba qu:;: jS tandard 6.10-) Tue Feb 2118'00;S9 Pacific Standard 1lm e 2 012 y S tl.\p: Incre m ent 1: Step Ti me = L x P rl ma ry Var; S , '"'lax. D efo r med Var: U Deform a ti on Scale FClctor: +S.

Figure 7 -Max PrinCipal Stress (psi) during the 1978 Blizzard; Deformation Scale Factor 500X S , (Avg: +4.722 e -+1.2 0-

-+8. 0-

-2.00 Oc +-

-1 .200e+0-1. B76e+0008: rn1223.od b A baqu sj Standard 6. 10-3 Tue Fe b 2118;00;S9 PacifIC Standard lime 2012 y Step: Step-l Incr emen t 1: S tep To me = 1.000 L x Pr im ary V a r: S , P rln c lp ,,Defor rr.ed Var: U Sca le factor: +5.Figure 8 -Max Principal Stress (psi) during the 1978 Blizzard; Deformation Scale Factor SOOX; Contour Range (+/-1200 Page 11 5,511 (Cy1) (A vg: 75%) +3. 8S1 e+0+9.+ 7.S

+4.S0 0e+3.00De+0+1. +O.OOO t H -1.50 0e +D-3.0 0OC + 02 -4.S COc +0 2 -7. 500e+-9.000e -3.8 3 Qe+03 008: m1223.odb Abaqus/Standard 6.10-3 Tue F eb2 118:00:59 Pacific S tand a r d Time 2012 y Step: Step-! Inc re m ent 1: Step TIme = 1.000 L x Primary Var: 5, Sl1 Deformed Var: U Oe for mi:l t io n Sca le F act o r: +S.O O D e Figure 9 -Radial Stress (psi) during the Blizzard of 1978; Deformation Scale Factor 50 OX 5, 522 (Cy1) (Av g: 7 5 0/0+1.9 32e+9.0+7.S+6.000 e +02 +4.

+1.5 00e+ D2 +O .o-1. -3.000e+02

-6.000-7. S 00e+0-9.. -

---

. 008: m1223.o db Ab a qu s/Sta nda rd 6. 10-) Tue Feb 2 113:00:59 Pacific Sttlndard Tl m e 2012 Step: S tep-l Increment l' TIme = 1.000 L x V,:,r: 5, 522 (CyD efo r med Var: U Deformation Scale Fa c t or: + S.000 e+Figure 10 -Hoop Stress (psi) during the Blizzard of 1978; Deforma t ion Scale Factor SOOX Page 12 5, 522 (Cyl) (Avg: 75 (:.'0) +1.9 3 2e+03 +1 .2 0 0e+ 03 +8.000e+0 2 + S.OOOe +02 +4.00C e *r*02 +2.0 00e+02 +O .OO O\!+O O -2.0 0Qe + 02 -4.0 00c +02 -S.O OOo::!of 02. *8. 000..+ 02 -1. 000Q+03 -1.20Qe+OJ

-3.47Ce+03

---

ODS: mI 2 U .odb Ab Dq u sjS tandtl r d 6.10-3 Tue Feb 2118:00:59 Pacif,c Standard llme 20 12 y S te p: S lO p-l Ir.crc ment 1: S te p nme = 1.000 P n ma ry Var: S, 52 2 (C yl) D e fo rmed Var: U Defor m ati on SCille Factor: + 5.0 00e+02 Figure 11-Hoop Stress (psi) during the Blizzard of 1978; Deformation Scale Factor 500X; Wider Contour Range (+/-1200 psi) 5 , 5 3 3 (C yl) (Avg: 75%)

+9.

--+4.S0Qe+02 +3.00Ce+02

+1 SOOe+0 2 +O .OOO e+OO -1. 500e+02 -3.000e+02 -4.S00e+02 -G.OO Oe+02 -7.500e+02

-9.000e+02

-3.30ge+03 006: m122 3.od o Ab aql lsjStandard S. 10-3 Tue F eb 2. 1 18:00:59 PacIfic Standar d li m 2.0n y inc r em en t 1: Stl!P Tim e = Pr i ma ry Var": S, 5 33 V.::l r; U Defo r m at ion Scale Factor; Figure 12 -Vertical Stress (psi) during the Blizzard of 1978; Deformation Scale Factor 500X P ag e 13 The result 1977 Blizzard Condition due to the 1977 blizzard condi t ion is shown in this section. Figure 13 depicts the temperature distribution in the model. Figures 14 through 17 show the stres s state in the max principal, radial , hoop , and vertical directions, respectively.

NT+4.6 76e+4.225e+O+3.+3.323 e +2.B 72+2.4 2.1e+O+1.970 e + +I.S1 Se+ O+1.0 6 9c-t-O+6.177 e+O+1.667 e +O* -* * --2 .84 20-1*0 0 -7.351.; +0 0 006: m12 2 2.o db A ba q u$/Stand ard 6.10-3 Tug F e b 2 117:5 0: 3 9 Pac i fic SUin.:l al"d Tim e 201 2 y Lx S te p-1: St e p Ti m e = 1.0 0Va .. : NT!l D eform e d V a r: U Deform a ti o n SC,;ll e Fa c tor: +$.OOO e+02 Figure 13 -Temperature (OF) during the Blizzard of 1977; Deformation Scale Factor 500X Page 14 S, f'la x. Pnncipal (A v g: 75 0/0) -+ 1.632 (H03 -+9 , OO oe+02 . +7.S0 0e+02 . + 6. 000e+02 -+ 1.S00e+D2 -+3.000 e +0 2 -+ 1. 5 QC.e+ 02 -+O.o OOe +OO -LSO oe..*02 --4.5 00e+ 02 -6. Q OOt! +0 2 --7. S0 0e+0 2 -9.0 0 0e+ 02 -1. ?14e+D 3 008: m1222.o d b Aba qus/Stdn dard 6.10-3 Tue Feb 2117:50:39 PiJciflC S tllndD rd Time 2012. Ste p : Step-In c rerr e nt 1: Step 'TIme :; V ar: 5, 1'>111x. Princi paDeformed Var: U Oeforma Uo n Scale Fact O f-; +S.Figure 14 -Max Principal Stress (psi) during the Blizzard of 1977; Deformation Scale Factor 500X 5, 511 (e y(Avg: 7 5+9.

+

+4 .S0Ce+3

+1 .S0 0-+O

--1. -3.

--7

-3.037008: m1222.odb Abaqus/Standard 6.10-3 Tue Feb .2 117: 50:39 Padffc Stt: ndard T,m e 2012 y S tep: Increment 1: Step Time = L x Pri ma ry V ar: 5 , S l1 De for me d Var: U 5cale Fact or: +5.

Figure 15 -Radial Stress (psi) during the Blizzard o f 1977; Deformation Scale Factor 500X Page 15 S , S22 (Cyl) (A v g : 75%) +1.067e+03

.. +9,O C OO+02 +7.SQ Oe+02

+ 4.50 01.: +3. Q O Oe+0 2 +1. 5 0 0e-i +O.O O Oc+-1.S0Q (H 0-3.QOOe+02 --1, 500e+0 2 :Max 930 -9.0 00Q+0--3.60 1 e+0ODS: m 1 222.odb Abaqus/Sta nda r d 6.10-3 Tue feb 2 11 7:50: 3 9 P a cific T ime 20 12 y S t ep: SteIn crem e nt 1: S t e p TIme = p ri lT\.i'llJ'"l I V a r: 5, 5 2 2 (ey lDe for m ed V a r: U De for m ati on Scal e Facto r: +S.DOOe+0Figure 16 -Hoop Stress (psi) during the Blizzard of 1977; Deformation Scale Factor 5 00 X 5 , 533 (cyl) (A v g: 75%) +1.00 2 e +D+9.0 a O e +0+ 7.S0 0 e+0+6.QOOe ++ 4,5 00e+02 +3.+1.50 00+ D. QOQe*..O O --1.500e+0 2. --3.000*+0--4,SOOe+0--6.0QOe+0-7.500e+02 Max 720 -*9.QOOe+02 --S.S13e+03 ODS: m12 22.odb A baqus/Stundard 6.10-3 T u e F eb 2117: 50: 39 Pacifi c 5 n d ar d llme 201 2 y S t ep: SInc r emo n t 1: Step TIm == Pnm lJ ry V a r: 5, 533 (Cy lDefo r me d V ar: U C e f ormatlo n Scale Fac t or: +S.OFigure 17 -Vertical Stress (psi) during the Blizzard of 1977; Deformati o n Scale Facto r 500X Page 16 Exhibit 62: Stress Analysis due to 105 MPH Wind load Appendix VllJ-63© 2012. Performance Improvem e nt Internation al Exhib it 62 Stress Analysis due to 105 mph Wind Load Summary of Results The results of the analy s is pre se nted in this report can be summarized as follows:

• The wind pre ssu re does not produce stre sses capable of delaminating the structure.

• The 105 wind pressure load re su lts in a max principal stress of about 55 psi

• The 105 wind pressure load re su lts in a r adia l s tress of les s than 1 p si Modeling Summa ry was used exclusively in the finite e l ement analysis pre sen ted here . . II d I tli Il t 1" The drawing s u se d as geometry input for thi s model ar e:

• Drawing No: (-100 Rev. 5 "S hie l d Building Foundation Plan & Details SH. 1"

• Drawing No: (-110 Rev. 6 " Shi e ld Building Roof Plan Wall Section & Details" All vertical reinforcing bars in the containment s hell section are modeled as rebar Ino (diamet e r 1.270") a t 12" spacing. The inner f ace h orizonta l rebars are tl8 (diamet er 1.000") at 12" s pacing. The outer fac e horizontal reb a rs are 1/ 11 (d ia meter 1.410") at 12" spacing. The vertical and horizontal rebars in the sho ulder sec tion s are #8 at 12" spacing. Page 1 Page 1 of 6 Exhibit 62 1ateriCll Properties The material properties us e d for this analysis are summarized in Exhibit 56, Figure 2.1.4: Material Properties Figure 1 show s the radial di s placement due to gravity and the 105 mph wind l oad. Th e maximum radial deflection is about 0.07 inch inward on the side of the st ructure facing the wind and about 0.07 inch outward on th e sides of the structure that are parallel to the wind direction.

U , I,J 1 (C }' ji -. ..

• 7 S t;* 5.EJ o..-O l ..3.3 7 Sc-'12 S"l e-O l + 1. 12641 -02 4 .0i Ooo -0]6 -1. U6e-O'l :S.37 89oo-4 , S04 -(12 -S. &3 C -0;: -6.756e-0:2 ';:Ilp. l'l t ',' -(! n lJ "'I'i nd -l iOS ITlF L 5 t-= p = PrI "", ,.., \I" r : U , ti l (Cy l) L.J 0 SL.d c Direction Figure 1 -Radial Di s placement (inches) due to Gravity and 105 mph Wind Load; Deformation Factor =: Page 2 Page 2 of 6 E x hi bi t 62 Figure 2 depicts the max principal stress due to gravity and the 105 mph wind pres s ure load. The maximum s tress due to the wind load i s 55 psi although some lar ge r stresses can be seen in the ring girder area. The st resses in the ring girder are a re su lt of the dome weight and not due to the wind load. $, {:t el '!:. n ( A" \iI: 7 V" ,) ...

1-0:;:

+1.J.He+'J 2

+-1.D S3<< f ..;"}.C:S2.p)1

+S .ll<'"!+O l +J.jli A<l +O i 2 .... 1 l.e *' 01 +" 1.05.72 + :0 1 2.!>&l e !XI 55 psi y Z ii r L >!

1 SC=[r 1'1 m p. '" 1. 1)(\0 . Prlnl ll r,t S, PrJ n Cl **,1 ' t:. cJ \'d r'. U 0 U F Eh:...l(-: " I _ Figure 2 -Max Principal Stress (psi) due to Gravity and 105 mph Wind Load; Deformation S ca le Factor = 2000X P age 3 Page 3 of 6 Exhibit 62 Figure 3 -Location Page Page 4 of 6 Exhib i t 62 Figures 4 and 5 below depict the max principal and radial stress _ The max principal stress shown in Figure 4 correlates well with shown in Figure 2. Both the location and magnitude of the max principal stress are in agreement.

Figure 5 indicates that the radial stress is very low due to the combined gravity and wind load, The only location that experiences any s ignificant radial stress is the corner of the flute. However , the corner of the flute location is a singularity due to the sharp angle between two elements in the finite element mesh. In the region of i nterest the rad ia l stress is below 1 psi. Page 5 Page 5 of 6 Exhib i t 62 :;

.... T t... Ie "iJ I * .....

.. . -l . :tl=..;e TlJ l .J 1 lSJ a ....{I.:. -*2 6 7.: ' *. 000.-+ ,-S.:rn. j.(>> . """ p: SbJj:.-J * ... ,.,*fll" ,...1. Sr:&IP T :110 J. "... , "'-)0' Jl r :to r:1 1/ '.ip:::lJ St.llNi F""1 .'*,I , +! C0l.1("' + Figure 4 -Max Principal Stress (psi) due to Gravity and 105 mph Wind S , S! l '9: ;" ... .! "J.. : .651.:: f -.:-; ,'lC.)f: -i.Ble , '!'" i).;')GhIl O[! -3. !:11C-'J 7 &.ci6 7"'-tC i':> *1.))" +\', .:.goo., t-,

• 3.1:i;:;. "T'l:' "1 . -1" ne-.o1. tt.;p li ne L 00 Prirn:.f)' .... =, ::: 11

(");" f.', V.I IJ I t/nn., I +OJ I").oi+fC.I Figur e 5 -Radia l Stress (psi) due to Gravity and 105 mph Wind Page 6 Page 6 of 6 Exhibit 63: CFD Analysis of Shield Building © 2012. Performance Improvement International-Appendix CF D ANA LYSIS OF DAVI S-BESSE CO N TAIN ME NT TOANA LY S IS P ERFORMED BY: JAN UARY 1, ****inforrration.

Perfo r rnanc:e Improvement international, LLC. ****** 1 Davis-Besse C ontainment Tower Requirements The CFD analysis perfo rm ed f o r thi s re port in cl ud e s: Pages

• No surrounding buildings

• 34mph from the No r t h w est (summe r) 5-9

• 34m ph from the So ut hwest (winter) 10-14

• 72mph fro m the Southwest (winter) 15-22

• With surrounding buildi ngs

• 34mph from the Northwest (summer) 24-28

• 72mph from the So u th west (winter) 29-34

• 105mph from th e Southwest (winter) 35-40

• Tornado 41-44

• Catego r y F 2

• Tr a v eled fr om t h e Northwest to Southeast Bo undary C on d i tions for the problem consisted of:

• Winter

• Ambient temperature o f -13°F.

• Te m perature of the containment tower remained at a con stant 7°E

• S ummer

• A mbi ent tempera tu re of 104°F.

• Te mperature o f the containment tower rem a i n ed at a cons t ant 13 0°F. Re s ults extrac ted f r om the CFD:

• Pressure distributions on t he surface.

• Heat transfer coefficients.

• Vorticity shedding calculated o n t he 72mph case. ,. ***********

inform2t i o;L Peric:rmJl1Ce Imp rovement intel*nation;;;i

.. LLC ******

Model Creation *The CFD m esh c o n sis t e d of 3.6 million cells to cr eate t he a ir v o lu m e.

• Total size o f the ai r v olu me w as a 2 ,5 0 0 ft. diameter and a h eig ht of 6 7 0 ft.

• Using a large air v o l u me eliminates any wall effects. _ sing a sma ll mesh size allows t h e vo r tici t y shedding to be captur ed mo re a c c urateCONTAINMENT TOWER CFD MESH Solution Method *The CFD program used for this analysis was Fluent version 13, industry standard and proven analytical -Incompressible ideal gas law was used, because the wind are below Mach -The containment tower analysis without the buildings was using a steady state

-The containment tower analysis with the building was done using transient analysis information.

Performance I LLC._**_

a:: w ::l V) ..........

l-V) S w a:: W ::J: a:: LL. .J: M LJ) 0 ..J --' .,.., c Q.) E Q.) >0 I.... 0. E (!) u c ro E I.... 0..,.... Q.) 0.. c 0 '.0 ro E 0 C Dav i s-B esse Containment Tower C FD Results 34mph Northwest Summer C onditions

-1 .33e-0 2 -1.86 e-02 -2 .4 0e-02 -2 .94e-02 -3 .47e-02 4.01e-02 4.54e-02 -5 .08e-02 -5.62e-02

-6.15e-02

-6 .6ge-02 -723e-02 DIRECTION

-1

-1.86e-02

-2.40e-02

-2 .9 4 e-02 -3.47 e -02 4.01e-02 4 .54e-02 -5.08e-02

-5 .62e-02 -6 .15e-02 -6.6ge-02

-723e-02 -7.76e-02

-7.76e-02

-8.30e-02 -8 .30e-02 FR O NT B ACK PR ES S URE CON TOU R S (psi) ***********

in'i0l"1TI3t ion ,

LLC. ****** 6 HEAT TRANSFER C OEF FICIENT HAND C AL CUL ATIONS FOR ANA LYTICAL COMPA RI S ON TOWER = 130°F AIR TEMP = 10t;°F (40°TEMP AVERAGE = llrF (42.22°v =0.1693 cmA2/s k =0.027 w/m*k Pr =0.71 U =15.20 m/s (34mph) D=44.73m Re =U* D I v Re = 40 , 159,24 4 Nu =h* D/k l\Ju = 0.3 + (0.62*Re A O.5*Pr A O.33)/([1+(OA/Pr)AO.67)AO.25) " [1+(Re/282,OOO)AO.625)AO

.8 Nu =38 , 092 h =(38,092*0.027 w/m* k) I 44.73m h = 22.99 w/mAYk

• 0.1761 BTU I hr"ftA2* o f h =4.05 BTU I hr*ftA2* of ( This number compares to the front surface of the tower (slide 8) . Region of comparison is the light blue and cyan) This indicates the CFD model has predicted the correct s urface heat transfer coefficients.

information. Perforna!Kb I rnprovement interndtiufl

,I, LLC ******

I D a v is-B esse Containment Tower CFD Results 34mph Northwest Summer C o nditions 1.6 0 e+01 1.4ge+01 1.37e+01 1.26e+01 1.15e+01 1.03e+01 9.16e+00 8.02e+00 6.87e+00 5.73e+00 1.15e+00 O.OOe+OO DIREcnON F R ONT 1.60..'*:01 1.498+01 1.378+01 1.26 e+01 1.15e+0 1 1.03e+01 9.16e+00 8.02e+OO 6.87e+00 5.73e+00 4 .58e+00 3 .44e+00 2.2ge+00 1 .15e+00 O.OOe+OO H x AREA O F COMPARISON Heat Tran s fer W a ll Co e fficients (B tu/ h r-ftJ\2-0F) ***********

inforT natinn. Perform21lce impl"OVerne'1L intern at ic n;:;i LL C. ******

D av is-B e s se Co ntainment Tower C F D Re s u3 4mph Northwest Summer Con d itio ns 2.;"-02 1.8ge-02 1.35e-02 8.18.-03 2,82.-03 .7J11.-03 *'33e-02 *1.88H)2 *2.400-02

• 294.-02 -3.7.-02 -6D8e-02 -6B2.-o2 -4.15e>llZ 4 ...-02 *7.i6.-0 2 *8.3 0e-0 2 CROSS SECTIO N PR ES S URE CONT OUR S (ps i) CRO SS S E CTION VELOCITY CONTOURS (ft/s)

• The cross section picture of the pr essure c on t our sho ws a steady gradient p r essu re bu il d u p in fro n t o f the to wer.

• At slow w ind speeds the flow mainly st a y s attached except along the top front and aft ed g e.

• The flow tries to stay attached , but fl ow s e paration hap pens at the botto m l1al f d ue to til e l o w pre ssure region.

• The top dome has a profound effect on th e flow s eparat io n.

• Another contributor of fl ow separ ation i s the archite ct ural flut e s l oc ated o n the side of the bu ildi n g. ***********

lnformat i on. Performance Improver: ,ent interllatondi , L LC. ****** 9

-ex:: UJ -Z -t-V) UJ ::I: :J 0 til UJ ::I: 0 ex:: LL .t:. c.. E -.:::t M L i -..l -J ro c 0 o-' ru '-' "'-' c .:-' C (l) E (l) > 0'c..,... .5 0) to ) \",j t OJ , c: 0 ro E 0:::

a vi s-Bes s e Con tainment Tower CFD R esult34mph Southwest Winter Conditions DIRECTION

-327 e-02 -4.12e-02 4 .98e-02 -5.83e-02 -6 .68e-02 -7.54e-02

-8.3ge-02

-925e-02 '! , , -..-1.01e-01 .. -1 .10e-01 -1 .18e-01 -127e-01 L -1.35e-01 FRO NT 1 'I r I I . I \ .' , \ PO, " :M: 1 I, ,'t , 'I ' , I .I -2.42e-02

--327e-02 4.1

-8 .-9 .-1 .-1.1' X -1 .35e-01 BACK P RESS U RE C ONTOURS (p***********

info r m ation. P er f o l T nance I m p r ov em ent in te l'llJ t lOnal , LL C. ****** 11 I HEAT TRANSFER COEFFIC I ENT HAND CAL C ULATIONS FOR ANALYTICAL COMPARISON TOWER =-13°F AIR TEMP =TF (-13. TEMP AVERAGE =-3°F (-19.4°v = 0.1168 cm A 2/s k = 0.02248 w/m*k Pr=O.72 U =15.20 m/s (34mph) D=44.7 3m Re=U *D/v Re =58,210,273 Nu =h*D/k Nu = 0.3 + (O.62*Re A O.5*Pr A O.33)/([1-:-(O.4/Pr)1I0.67]1I0.25) * [1+(Re/28 2 ,000)1I0.625]1I 0.8 N u =55,111 h = (55,111

• 0.02248 w/m*k) /44.73m h =27.7 w/mIl2*k '" 0.1761 BTU / hr*ft A 2* OF h =4.87 BTU I hr*ft" 2* OF ( This number compares to the front su rfa ce of the tower (slide 13). Region of comparison is the light blue and cyan) Thi s indicates the CFD model has predicted the correct surface heat transfer coefficients.

l infofrnatio\1 , Perfo r man c e Improvement inte r n2tion;;I, LLC. ****** 12 I Davi s-B esse C o nt ainment Tower CF D Results 34mph Southwest Winter C o n d it ions 1.96e+01 1.81e+0 1 1.66e+01 1.51e+01 1.36e+01 121e+01 1.06e+01 9 .07e+00 1.51 e+OO O.OOe+OO Hx AREA 1.96e+01 1.818+01 1.66 e+01 1.51e+01 1.36e+01 1.21 e+01 1.06e+01 9 .07e+00 7.56e+00 6.05e+00 4 .53e+00 3.02e+00 1.51 e+OO O.OOe+OO BACI< Heat Tran s f er Wa ll Co e fficients (Btu/h r-f tA 2-0 F) ***********

infnITnation. Per f ormance Improvemen t internai.i o nal, l.L C. ****** 13 I Davi s-Besse Containment Tower CFD Results 34mph Southwest Winter Conditions

),56...02 2 70

• 1.85e-02 8'-.0' lA3e-O) '7.11.'()3 ., 56e'()2 *2A2...o2 -327...02

.. ,21t,\12 ".9a.0,2 -6.83...02

-6.668.02

• 754* .oZ -6.3e.-OZ

-9,.26.-OZ

.101...01 *1'1 0.-01 *1 .18.-01 *1 27e.Q1 *1.3 5.-01 , l CROSS SECTION PRESSURE CON TO URS (p s i) CRO SS SECTION VELOCITY C ON TOU RS (ft/s)

• The cro ss s ec t io n pic t ure o f the pressure contour s hows a s maller p res sur e bu il dup in front o f the b ui l din g .

• A co ld dense air h a s a tend e ncy to s he d fro m struc t u re s m ore ea s ily du e to a h igh er R e y n ol ds number.

• D uring w int e r c ond i ti ons, the flow separat es co m pletel y fr o m th e tow er a t 34 m ph. A r esu l t is vo r ticity sheddin g.

• An effect of the flow separation a t low er sp e e ds w ill ca u se a c yclic p r e ss u r e loa d s on th e co n t ai nm ent tower .

• The top dome has increase d th e eff ec t of flow s e parati on.

• Another contributor of fl ow separa tio n is the arch i tect u ral flut e s l ocat ed on the si d e o f the bu il d i n g. 1*-1 i nfor m at i on. Performancf' Irnprovemeni

internationa l , L LC ******

---

0:: UJ -Z sl-V) $ UJ ::I: ::J 0 V) UJ ::z:: 0 0:: u. ..c Q. N Li rl u -l ....J (0 C () ..;::; I\l C OJ ...... c OJ cOJ 0 >-Co r-C OJ u C .'0 E 0 (U 0... c 0 0';:; <] 0 ,c L =:==:==:.11 D avis-Besse Containment Tower CFD Results 72mph Southwest Winter C o ndi tio ns DIRECTION FRONT PRESSURE CONTOU RS (psi) .. 6 *****i nfcrrnation.

PerFOl'lnanl21mprovemeni i rternaiion31, LLC. ****** -1.06e-02 -1.06e-02 -3.4Se-0 2 -S.83e-02 -3.4Se-02 -S.83e-02 BACK H EAT TRA N S F E R COE F FICI E NT HAND CALCULATIONS FO R A NA LYTICAL COM PARISON TOWER = -13°F (-25°C) AIR TEMP = TF (-13.9°C)

TEMP AVERAGE = -3°F (-19.Ll°C) v = 0.1168 cm A 2/s k =0.02248 w/m*k Pr =0.72 U = 32.63 m/s (72mph) D = 44.73m Re=U*D/v Re = 124,960,607 Nu = h*D/k Nu = 0.3 + (0.62*Re A O.S*Pr A O.33)/([1+(0.4/Pr)AO.67]AO.2S) * [1+(Re/282,OOO)AO.62S]AO.8 Nu = 97,032 h = (97,032

• 0.02248 w/m*k) / 44.13m h = 48.76 w/mA2*k

• 0.1761 BTU / hr*ft A 2* of h =8.587 BTU / hr*ft I\2* of (This number compares to the front surface of thetower (slide 18). Region of comparison is the light blue and cyan) This indicates the CFD model has predicted the correct surface heat transfer coefficients.

inhJl'm3cion. Perfol"fY1anc 2 1mprovernent i nLernaUonal, LLC. ****** 17 D a vis-B esse Containment Tower CFD Results 72mph Southwest Winter C o nd itio n s L FR O NT 327e+01 305e+01 2.83e+01 2.61e+01 2.40e+01 BA C K 3.27e+01 3.05e+01 2.83e+01 Hx AREA CO MP ARIS OHeat Tr a nsfer W all C o effi cie nt s (B tu/hr-ftI\2-0 F) **********

l lnformaUi)n. PE:rformance Irnproven'lent internation<Jl , Ll C. ****** 18 Da v i s-B esse C ontainment Tower CF D Res ult s 72mph Southwest Winter Co nd i t ions , 32e-02 -1 O6e-02 -345e-02 -S.83e-02

-8.22e-02 -1 , Q6e-01 Large of CROSS SECTION PRES SURE C ONTOU R S (psi) CROSS SE CTION VELO CITY CON TOURS (ft/s)

• The pressure conto u r has stay e d the same fr o m the 34mph , but the pres su re loa d a n d suction h as i n c rea se d.

• A cold dense air has a tende ncy t o sh ed fro m structures m ore eas ily due t o a higher R e yn olds numbe r.

• During winter conditions, the flow separates completely from the tower at 72m ph.

• An effect of the flow separation at h igher speeds will cause more cy clic pres s ure l o ad s on the co ntainment tow er,

• The top dome has increased the effect of fl ow separation.

• Another contributor of flow separation is the architectural flutes located o n the side o f the building.

19 ***********

informa ti on. rJerfol'lnance I mprollement mtemationClI , LLC. ******

I Davis-B esse Co ntainment Tower CFD Re sul ts 72mph Southwest Win te r Con dition s Velocity path l.ines a t 1/3 do wn f ro m th e to p of the con tainm e n t to wer (ft/s) ***********

I informati Xl. Per f ormance I mprovemant inte.:-n2lio.:"JrlJi , .****** 20

, Davis-Besse C o n tai nm en t T owe r CFD Re s ults 72mph Southwe st W int e r Co n dit i ons Velo ci ty path l i n es ha l f way f ro m the top of the con tainm ent towe r (f t/s) ***********

infoI"lT'latiol1. Performan c e Improvement intelTlationa i, LLC ****** 21 Davis-Besse Containment Tower CFD 72mph Southwest Winter Conditions Vortici ty S heddin g at 48ft from c y l inde r T h e flow tries t o stay attached , but the f lute causes separation. V elocity con tours 1/3 down from the t op at 72mph Vorticity s h eddin g fr equ en cy = (108 ft/s) / (4 8ft) = 2.2 5 hz 9 .91e+01 9 .01 e+01 811e+01 721e+01 6 .3*le+01 5.41 e+O'I 4 5 1 e+01 360e+01 .70e+01 1 80e+01 ***********

1 information. Performan c 21mprovemem International, ******

Model Creation of Containment Tower with Buildings

-The CFD mesh consisted of 3.26 million cells to create the air volume . -Total size of the air volume was a 2,500 ft. diameter and a height of 670 ft. wall effects. Using a small CONTAINMENT TOWER WITH BUILDINGS AIR VOLUME WITH

-r j :::.J c '" (} m E , 2 QJ a.. ,c .*,." E o .

D a vi s-B esse C ontainment Tower CFD Results 34mph Northwest Summe r C onditions

.1 , 78.,.02 -Z94e-O' .1948002 409e-Ol -4

-5 :!4e-01 .5.24e*01

.6 3ge-0? -839e-01 -1S4e-O . 754 e*02 -8 70e-02 *870e-D:!

-9

-98Se-02 -I IOe-O I

• I I Oe*OI .1 22e-O I -1.2 2e*O I -133... 0 1 -1.33 e-O I I -1 4S e-01 -I -I S 6e-Ol

• 1 S Se.Ol -1 68e-0 1 -1 7g e-O I -1.7g e-0 I FRONT BACK P R ESSU R E C ON T OUR S (p s i) With the addition of the s urr o undin g b u i l din gs , the pr es s ure h a s increased b y O.0 2 7psi. ***********

i nformation. Performance Improvemenl internati o nal, L lC. ****** 25 I Da vi s-Besse Containment Tower CFD Results 34mph Northwest Summer Conditions I 78e+Ol 164e+01 151e+Ol , I 137e+OI I 1 23e+OI \ II 1 10e+Ol B 21e+0 0 II 6801e+0 0 48e+00 4 11e+00 274e+00 1 37e+00 o OOe+OO 1 92e+OI I I 78e+OI 1 64e+OI 151e+Q l 1,37e+0 1 I 23e+OI 1,10e+O I 9's8e+OO 821e+00 664e+OO I H I (' .. I' ' I ':11 i [" ____ J 548e+00 I 4,lle+00', 1 , 1 2 740+00 ,1 370+00 0 0 00+00 FRONT BACK Heat Transfer Wall Coe f ficients (Btu/hr-ft"2-0F) 2G ***********

i nformati o n. Performan ce :mprovernen i: international LLC ******

I Dav is-B esse Co ntainment Tower CFD R esults 34mph Northwest Summer Conditions CROSS SECTION PRE SSURE CON TOURS (p si) CR OSS SECTION VELOC ITY CO NTOUR S (f t/s)

• The pressure co nt o urs h a ve d r amatically ch ange d wit h the additi o n of sur ro und i ng b ui l d i ngs.

• There is a large low pressure r egion lo cated above the building on the aft sid e of the contai nm ent to w er.

• The velocity vectors are disr up ted fro m t h e buildings causing the flow to separate at l o w er wind speeds. 27 ***********

in'formation. Perf o rmance Improvement international , LLC ******

J D av i s-Besse Containment Tower CFD Results 34mph Northwest Summe r Co nditions Large area of separated f low ll tl 4e+Ol 636e+O I VEL OC I TY VECTORS (ft/s) ***********

l inforrn3jun

.Pel-kxlYlancE:

Improvemem ir, t emt {io!" ,!I, lLC ****** -8

-0::: UJ -Z -V) l!) zCl ....J :J CQ :::I: V) UJ :J: :J o V) UJ J: I-a 0::: u. ..c: c.. r:1 C () ro c '-' c 'C (lJ E iJ) > 2 a.. E (!) fJ C ;0 i= c o

• 1 19e-01 -, SOe-OI -19Se.01

• 23Be*OI -'27ee-OI

-3. lee-01 -3.S9e-01 -3 Sg e. 01 -4 3ge-01 -4 .7ge-01 -51ge-OI -55ge-01 -5 SSe-Ol Davis-Besse Containment Tower CFD Results 72mph Southwest Winter Conditions -I -198e*-238e-O

-3.I-3 Sge--39ge*O-4.3g e-01 -4 .7g e*O l -5 1ge-O 1 -5.5ge-0 1 -5 .9ge-O 1 -640e-O l -6.40e-Ol FRONT PRESSU R E CONTO U RS (p s i)

• W ith the addition of the su rrou n di n g bu ildin gs , the pres su re ha s inc reas e d by O.0 5 4psi. B***********

irrformatior"l.

Perfor, n anC2lrnprovemen1.

inte n-,ati o i1.:;\, U C. ****** 30

--D a vis-Be s se Co ntainment Towe r CFD R es ul ts 72mph Southwest Winter C ond i ti o ns 1*82e*OI 'I r'1' ' , . t'I i

• I 1 ' " I ' 168...01 \1 1 1f: ,; I' I 54e<Ol I ' t ' 14De ' OI }",1: 1 26<: -01 \ .' I 12e+OI 9 8le-OO (. 1'1 j ' '", -I Ol e-O n *1 -.... "" S 61 p'DO ', p' Ol) I &2e+Ol 168e+Ol I 54e+Ol 140e+Ol I 26e+Ol 112e.OI 982e* OO 842e+OO 7.0 1e+OO 56 1 e*OO II " ',j J ele-C I O / *281e+00 1*10 ,,' 0 0 V""'-1 40e+00 o OO e. O O O.OOe+OO BACK F RONT H e at Tran sfer Wall Coefficie nt s (Btu/h r-ftJ\2-0 F) 3J. ***********

inform at ion. Per f ormance Improvemc=nt i n te rnati on ClI , LLC ******

I D av is-B esse C on tainment Tower CFD Results 72mph Southwest Win te r Co nditi o ns CROSS SECTION PRE S S URE CONT OU RS (p si) C ROSS SECTION VELOCITY CONTOURS (ft/s)

• The stagna t ion pre ss ur e re g ion has sh i ft e d u p t o w ards t he top of the c on ta i nment t owe r. Th is is a r e s ul t o f th e b uilding s being in front o f contai nment t o wer.

• The flow o n t h e aft si de of t h e to w er is t u rbule nt c om pared to t he case w i t h no b u i l d ings. ***********

l lnfonn3tion. Perlorrnan(2 Improvement intemationai , LLC. ****** 32 Davis-Besse Co ntainmen t Tower CFD Result s 72mph Southwest Wint er Cond i tions 154e+02 1.42e+0 .. 1.30 e+0 2 1 1ge+02 1.o7e+02 94ge+01 8 30e+01 711e+01 5.93e+01 4.74e+01 3'5 6e+0'1..' 237e;+ 01 *1 1ge+01 :\.

VELOCITY VECTORS (ft/s) ***********

l inforlTJation. P erform a n c e Imprc.vement intern ational, Lt C ****** 33 Da v is-Besse Containment Tower CFD Results 72mph Southwest Winter Conditions 1 7ge+0 2 1 8 6 e'*" 02 1 54e+O:! 142e+02 1..3 tJ e+02 1 18e+02 107e+(t s! 948e+01 8.30e+01 7 11e+Ol 5.9 3e+0 1 4.74e+0*\ V ELOC I T Y VECTORS (f t/s) ***********

I informauon. Performan ce Improvement mternation(:ll , LLC. ****** 34

--Lii (Y) -0:: w -Z -V') c:J -Z e ...... :;:) CO :I: V') W :I: :;:) 0 V') w :I: 0 0:: LL c. Ln 0 \"""t i -' . ...J -Ir' e 0 OJ c. QJ E QJ >CJ0. -ill u !: m E '-0 IJ) e 0 ".;::i IV .. ...0-. c c I Da v i s-B esse Containment Tower CFD 10Smph Southwest Winter Conditio n s 5.23e*03 -844e-1r.1 I ., 74e*OI *264e-OI *353e-Ol -443e-01 -5.33e-Ol -622e-Ol *712e-Ol -802e*Ol -8.91 e-O I -9.81e-Ol -I07e+OO -I 16e+OO -1 2 Se+O O -1 3 4e+OO -1 .4 3e +OO \ I i'! . i ? ..... :..-.'.1. I 1 ::-; . ;.I I) " '---.,/ -Il I -.

....i:ij I. ___ 'I /'I ,

1/; ----.. '1/--.. FRONT *B.

.

-8

---

I0 7 e+O-I 16e+O-' 2 -1 .3 4 e-, 43BACK PR E SSU REC ON T OURS (psi..., .. Jt' ***********

i nfol'mation. Pel"formance

.nnrov*ment mtern.:rtic)nal , r..LC. ******

D av i s-Besse Containment Tower CFD Results lOSmph Southwest Winter C on diti o n s 270e*01

!32e+Ol ::1 12e<-01 , 93e+Ol 1 74e+OI 155e+Ol 1 35e+Ol 2.51e*Ol e-Ol ::1 121.'-01 1.93e* Ol 1 155p-Ol I )5e.Ol 1 96&e.OO 713e+OD 7ge* OO 3 abe' GO I 93e. O O 0 00.,* 00 . ...., t

• 4, '.* J: I . -' .....

• I I 1 r-x " ./. ,-; " :::"';::" " 'I __ ",.... " 'iL r" i I FRO NT H e at T r a ns f e r Wall C o efficients (Btu/hr-ft

" 2-0F1 16e+01 9 SBe+OO 773e+0 0 5.7g e+00 3 8 6e+00 I 93e+00 D.O O e+OO B ACK 3 , ***********

l i nrorrnaLion. Performallce Improvement international, L LC. ******

Davis-B esse Containment Tower CFD Results 10Smph Southwest Winter Con d itio ns CR OSS S EC TION PR E SSURE CONTOURS (psi) CR O SS S E CT ION V ElOC I TY CON TO U R S (ft/s) w Th e st a g n a tion p ressu r e region has shifted up to wards the to p o f the co nta i n m ent tower. This is a r e s u lt of th e bu ild i n gs being in fro nt of c onta inme nt to wer.

• The flo w on the aft side o f the to wer is u ns tead y an d turb u le nt.

• The additi o n o f the bu i l d ings h as c a use d the fl ow to r ise do t o the pressure i nc re a se j ust b ef o r e re ac h i ng t he b uilding.

• This results in a higher pressure re gio n a t the midpoint causing a larger ove rturni n g mo m e nt. 38 ***********

1 info r m at ion. P 2rfo r rn i in C e Improvement intemat , o ll :: iI . LlC. ******

I Dav is-B esse C ontainment Tower CFD 105m ph Southwest Winte r Con ditions '2 2 17 e+02 200e+0 2 1.84e+02 I.B7e+02 1.S0e+02 1.34e+02 1.17e+02 I .OOe+02 8.35e+OI 334'6: +01 '. '1 : G ,i e+Q.l \ ' VELOCITY VECTORS (ft/s) ..,9 ***********

i n f o rmat io n. Pe r forman c 21m p r oIJe ment i n t ernati o nal, LLC. ******

Dav i s-Besse Containment Tower CFD Results lOSmph Southwest Winter C onditions 250e+02 _ 34e+02 '2 t 7e+0 2 200e+02 184e+0 2 1.67e+02 1.50e+02 1.34 e+02 1 , 17e+02 100e+02 B , 3 5e+O 1 V ELOCITY VECTORS (ft/s) ***********

1 inforr(l3ti JI1 , PerfO!T (lanCe I mprovemen t

• • • • • • 40

-(,/) <.( z <.( C I.L. U o c <C z 0::: -" o ;;;; ,0 C UI -I-' c:

========1 Tornado Cond it i on s

• F2 Tornado with winds between 113 and 157m ph.

• Tornado touched down just west of the Davis-Bessie power pla nt b e tween 84S pm and 900pm on June 24, 1998 . .. Tornado was 100 yards wide an d t raveled southeast for 3 miles.

• Considerabie damage was noted along this path with some b a rn s totally de s troyed.

• Slide 43 and 44 shows pressure contours on the buildings a n d co ntainment tower as the tornado passes. 4.62e-+02 4.33e+02 4 , 04E'-+02 375e-t02 3.46e+02 3 17e-t02 288e-+02 2.60 e-t02 2 31 e-t0 2 2.02 e-+D 2 1.73e+0 2 1.44e+0 2 1.158+02 8.65 e+01 .77 e-t{J 1 2.8%+01 6.988-03 VELOC I TY VECTO RS S IM ULAT I NG A TOR NADO (ft/s) ************

inf i xmation. Pe rrOr'TI a nee Irn provl?l1"1ent in ternal io n::ll.:..LC ****** II J D a v i s-B e s se C ontainment Tower CF D Results CFD Analysis of Tornado Passing by Containment Tower .1 Hff-O l I ' )< ' 00 .. r -I.... ;"1 1 8.36sec 'I 9.36sec w:. QI .nl..9 f 1. [ I I 0\10

• 00 j

.' -".b* 12.36sec 1l.36sec , I ...

10.36sec 13.36sec ***********

i nfo r rnatioll. Perfo r man c e Impr ov em e nt interna t ion al , LL C. ******

I Davis-Besse Containment Tower CFD CFD Analysis of Tornado Passing by C ontain m ent Tower t :1.. ="'--i:."., I '. . to .' ...... .--..... ., .' 1_ 1 I J * * !)Il I:...... * *1 [,(),. 0Il 1 64.* 00 107... . .JI ...f t r l' l! J U.'l' o 14.36sec : ",," , 15.36sec 16.36sec I J.k I I)) l li> , 1(, 1 lil o OO I ' lY.l [ gi'!'" * " ._ f t;l..7 Jl. , ,," 17.36sec

l inf o rm Jti o n. Pel"fon n a l 1Ce Improvement LLC. ******

Exhibit 64: Thermal Stress Ana l ysis Gravity and Wind © 2012. Performance Improvement International Appendix Exhibit 64 Thermal Stress Analysis with Gravity and Wind Load Table of Contents Summary of Results .................................................................................................................................Modeling Summary .................................................................................................................................Overall Approach .................................................................................................................................Finite Element Software ......................................................................................................................Modeled Geometry .............................................................................................................................Finite Element Models .............................................................................................................................Descriptions

................................................................................................................................................................................................................................................................................................................................................ Thermal Stress Screening

.........................................................................................................................Thermal Stress Screening Results .........................................................................................................Combination Load Cases......................................................................................................................Analysis Based on Measured Properties

................................................................................................ Circumferential Temperature Distribution at O.F. Horizontal Rebar ................................................... ..........................................................................................................................Stress State during Hot Summer Condition

............................................................................................ Stress Analysis Results Summary .......................................................................................................Shoulder 10

.............................................................................. Azimuth 225" Location ............................................................................... Pagel Page 1 of 19

---

Exhibit 64 Summary of Results The results of the analysis presented in this report can be summarized as follows: The temperature and wind conditions found to maximize the radial stress are not sufficient to delaminate the structure alone Thermally induced radial stresses is maximized at the hot summer temperatures At the location of the outer face horizontal rebar, the maximum radial stress due to temperature gradients, gravity, and wind is about 300 psi Modeling Summary Overall Approach Page 2 Page 2 of 19 Exhibit 64 Finite Element Software analysis presented here. was used exclusively in the finite element Modeled Geometf'y The drawings used as geometry input for this model are:

• Drawing No: C-100 Rev. 5 "Shield Building Foundation Plan & Details SH. 1"

• Drawing No: C-110 Rev. 6 "Shield Building Roof Plan Wall Section & Details" All vertical reinforcing bars in the containment shell section are modeled as rebar #10 (diameter 1.2701J) at 12" center to center spacing. The inner face horizontal rebars are #8 (diameter 1.000") at 12" spacing. The outer face horizontal rebars are # 11 (diameter l.4lOIJ) at 12" spacing. The vertical and horizontal rebars in the shoulder sections are #8 at 12" spacing. Finite Element Models focusing on two distinct different geometric Page 3 Page 3 of 19 Exhibit 64 Figures 1 through 3. Figure -Shoulder/Flute; All Mesh Shown Page Page 4 of Exhibit 64 Figure -Shoulder/Flute; Mesh with Rebars Exposed Figure -Shoulder/Flute; Rebar Mesh -------

PageS Page 5 of 19 Exhibit 64 Figures 4 through 6. Figure Shell Section; All Mesh z x Figure Shell Section; Mesh with Rebars Exposed Page Page 6 of Exhibit 64 Figure Shell Section; Rebar Mesh Thermal Stress Screening In order to understand the effect of the various thermal conditions that the containment structure may be subjected to, a screening analysis was performed.

The screening analysis was performed using preliminary material properties before the official material properties were obtained.

The screening analysis considered a total of 32 thermal conditions.

They included the summer and winter solstice, the spring and autumn equinox, windy and calm condition, as well as average and hot/cold ambient temperatures.

The six thermal conditions resulting in the highest radial stress in the screening analysis is analyzed with gravity and wind pressure loads in the next section. Page 7 Page 7 of 19 Exhibit 64 Page Page 8 of Exhibit 64 Combination Load Cases The result of the screening analysis identified the thermal conditions most likely resulting in the highest These combination load cases were again solved with the preliminary material properties since the official values had not yet been obtained.

Page Page 9 of Exhibit 64 Analysis Based on Measured Properties The six cases predicted to result in the maximum radial stress is analyzed using measured material properties from samples taken from the Davis-Besse containment structure.

The material properties used for the analysis are summarized in a separate section in the Root Cause Analysis Report (Exhibit 56, Figure 2.1.4: Material Properties The conditions analyzed using the measured material properties are the same six conditions presented in Table 2. They are listed below along with the time of day determined to produce the highest radial Page 10 Page 10 of 19 E xh i bit 64 Circumfcr ntiLll TCmpCI.IlLi e Dislributiun.1 O.F. HOI-izontal Rebar The temperature profiles for the six conditions resulting in the highest radial stress based on the screening analy s is are shown in Figure 7. The temperatu r e profile s are plotted i n the circumferential direction a round the shield building at the outer face ho r izont a l r ebar depth. Temperature (OF), Mid-Height, Outer Face Horizontal Rebar Depth LL:' Q III :l-L-60 a;"' III -No Wind, Summer So l s tice, Hot Temperature, 7:30 PM :I 40 -No Wind, Summer Solstice, Average Temperature, 7:30 PM No Wi n d, A u tumn Eq u i nox, Hot Temperature, 6:00 PM 80 -No Wind , Autumn Equinox, Average Tem p erature , 6:00 PM 34 mph Wind, Summer Sol s tice, Hot Temperature, 6:00 PM I 34 mph Wind, Autumn E quino x, Hot Temperature , 5:00 PM '---o 22.5 45 67.5 90 112.5 135 157.5 1 8 0 202.5 225 247.5 270 292.5 315 337.5 360 Azimuth n Figure 7 -Cir c umferential Temperature Distribution at O.F. Horizontal Reb a r For each of the s ix temperatur e profiles shown in Figure 7 eight set of double valleys can be seen. The valleys represent the lower temperature under the thick se ctions of the shoulders.

These are as are covered by thicker layer of concrete so it takes longer fo r them to heat up due to the hot exterior conditions. Figure 7 also s hows th a t the azimuth 225° location corresponds to the hottest location around the structure.

The condition resulting in the hottest t e mperature at th e outer face horizontal rebar depth is labeled " No Wind, Summer Solstice, Hot Temperature, 7:30 PM." This is the t e mperature condition studied i n the following sections.

Page 11 P age 11 o f 19

---

Exh i bit 64 The south to south-west side has the highest thermal gradient do to the solar heating during the day.* Figure 8 shows the location of the flutes, shoulders, and the azimuth convention for the Davis-Besse containment structure.

Figure 8 -Shield Building Flute Numbers and Azimuth Locations Pctge 12 Page 12 of 19 location Exh i bit 64 Stress State during Hot Summer Condition The results shown in this section describes the detai l ed stress state in the hottest location around the structure for the hot summe r condition (No Wind, Summer Solstice, Hot Temperature, 7:30 PM) Figures 9 through 13 show the r'esults from the shoulder 10 location Figures 14 through 18 depict the same result s from the azi muth 225 0-, For each of the two l ocations the result is presented in f i ve figures, The first figure shows the temperature distribution and the following four figures depict the stres s state: 1, Temperature 2, Max Principal 3, Radial 4, Circumferential (Hoop) 5 , Vertical The stress state is presented at the mid-height section , The contour range is set to +/-300 psi for a l l the stress figures so that they can be compared more easily, Stress 11.(1)

Smllmary The maximum stress is confined to the top and bottom of the outer face horizontal rebars, The maximum tensile stress is about 300 psi and not enough to crack the concrete, Pa ge 1 3 Page 13 of 19 Exhibit 64 Shoulder 10 Location The temperature distribution, max principal stress, radial stress, hoop stress, and vertical stress in shoulder 10 ar-e depicted in Figures 9 through 13, I-espectively.

Figure 9 shows that the shoulder surface is hotter than the flute sur face. This is the result of more solar exposure on the shoulder su rface compare to the flute valley. Also, there is more surface area at the corner of the shoulder re sult ing in h igher temper atu re during the h ot a mbient condition.

rlT1 +1. 1 7 5e+0. -. -+1.16 QQ +O Z +1.14Se+02

+1.1 2ge+Ol +1.1 1 4e+OZ +1.09ge+0 2. +l.06qe+O l +1.06ge+02. +1.0 3 9@ ++1. 0 2% +9.936e+Ol C0 8: m 5tRp: St e;l -l T n(.t#.m o ll t'-1 : St ep =-1. " T1l Defo rmed V n r, U C-efo rm()oon s.:a l e F':'Kl:or:

+5. Figure 9 -Temp erat ure Di s tribution n) Figures 10 through 13 depict the stress state using the max pr i ncipal stress and the three st re ss components in a cylindrical coordinate system located at the containment st ructure center. The max principal a nd r ad i a l stresses are highest at the outer face horizontal rebar. The figure s also indic a t e an in the Shoulder 10 Location area of high stress on the left edge of the model. This ha s been identified to be a si ngularity Comparing the stress i n the three radial, hoop, and vertical directions (Figures 11 through 13 respectively) indicates th at the radia l component has the highest tensile s tres s. As shown in Figure 11, the radial tensile stress i s below 300 psi which is less than the tensile s trength of the concrete.

It is concluded that the hot summer temperature condition is not capable of delaminating the structure in the flute/shoulder location.

Page 14 P ag e 14 o f 19 Exhib it 64 S, I"lax. Princip'" (Alfg: 7S Q ,+4.8

+2. +2.-+1 .+1.

+O .oo Oe -5.00 00-1.000oe+-1. 5000...0-2.00 pe .,.-2,5 0 0@+O-3.00 0e+0ODB: 5 09:45:12 PlKltilt Standard Tim e 2D12: stJ>p. S 'p-1 1: StQ P nme :;; Pn m,.'II), V03 r: 5, r*u, :.:. Principal Oef o rmed Vi:lr: U De form"uo n Sc-ztle Factor: + Figure 10 -Max Principal Stress (psi) in the Flute/Shoulder 5 , 51: (e \,l) (Al,"g: 75%)

+2.a O()a +1.+1.+S , O OOe+O.*S.

-

-2.-2.500e+0-3.0006+02

-7.228e+02 OD6: ml.,W"r;*" 09:45:1:

Stan dard lim IJ :>01.:. SIl..\f> : I ncre ment 1: Step Time Pli m<lt"" Va r; S , Sl1 D l!!!o r m ed Vu: U [)Qrol rn ll tJon SC.,lc Fz,ctor-: +5Figure 11-Radial Stress (psi) in the Flute/Shoulder Page 15 Page 1 5 of 19 1 : $b: p TIme = DeformM f on Scale FlIctor: +5.0 00e-, 0 2 Exhib i t 64 S, 52:;' ( C I'I) lAvg: 75 /1', ) +3.000e+0 +2 .SOOC! +0 +2.0 0/)o;;!+ +l.SQOe +O+5,0 00.+ 0+o.aoOe+-S.OO-1.O O-1.

-2.00 0e +-1. 5 0 0e.+0. -1. 1 6 50 +03 ODS: step: In cr e m ePI 1f'l"IlJr l 5 , 5 1 :2 (eDefo rmed VM : U Figure 12 -Hoop Str ess (p s i) in the Flute/Shoulder 5, 53 3 r C\'I) (Av Q: -+ 3.2E e+02 -+3 , 000.+0+25 00e+0+2.+1 .S00e+02 +1.+5.0 0 De+ + O.(I COe+O-. S.OOoe.,.O-l.OO Oe+ -1. -Z .ooOe+OZ -2.500e+0-3.

ODS: 09: 4 5: 12 Pa c ific TIme 0 1:2 S t c.p-I n c.r em en t 1; S tep Time Pl imar'y Va..: S, 5.33 (CylC efO fm e d VM: U ec r o rm atlon So.:al c factor: +5 Figure 13 -Verti ca l Stress (psi) i n the Flute/Should er PClge 16 Page 16 of 19 Exhibit 64 Azimuth 225 ' Location The temperature distribution, max principal stress, radial stress, hoop stress, and vertical stress in the shell area at azimuth 225° are shown in Figures 14 through lS, respectively.

Figure 14 shows that the exterior surface is hotter than the interior.

This is the result of the hot ambient daytime condition and the colder nighttime condition. + 1.20 St!+O l + 1.i76e+0 2

-+1.147e+. +1 , 1 03e+1 , O SCe+0+1.0 7

+1.0Sge+-+1.0+4e++1.030e+ODB: mlS34-r eb31'S-b.odb Abaqus/Sta ndard 6.1 0-3 V)ed Feb 1S 12:16:58 Time

< y StEp: Sre p-}-f'W', .... IND_HO I _SS_ T Ef'l P _RTP _730P1'1 Incre men t 1:

lime:: 1.00 0 PI1mll'y rnl1 x De f ormed Var: U Scale F"cto r: +5.000e+02 Figure 14 -Temperature Distribution (OF) at the Azimuth 225 0 Location Figures 15 through 18 depict the stress state using the max pr i ncipal stress and the three stress components in a cylindrical coordinate system located at the containment structure center. The max principal and radial stresses are highest at th e outer face horizontal rebar depth (see Figures 15 and 16). Comparing the stress in the radial, hoop, and vertical directions (Figures 16 through lS, respectively) i ndicates that the radial component has the highest tensile stress. As shown in F ig ure 16, the radial stress is below 300 psi which is less than the strength of the concrete.

It is con c luded that the hot summer temperature condition is not capab l e of delaminating the structure in t he shell section l ocation (middle of a pane l). Furthermore, Figures 17 and 18 show that the hotter exterior surface temperature results in compression stresses in both the hoop and vertical directions due to expansion of the outer layer. Pctge 17 Page 17 of 19 E xh i b i t 64 :., (A'/g : 75 n+7.

+2. .

+S.OOOe+0.0 0

< y X Figure 15 -Max Principal Stress (psi) at the Azimuth 225° Location S,S11(C),I) (A'.*9,75%)

-+3.193e+02.

-

+2.+1 .5+1.

-+O.ooOe+OO --S.OOO*

.1. *

--3 .000e+02 --8.640e+02

< y x ODB:

A.t:laqtJ!Oj Standard 6.10-3 Wed is 12:16:58 Pacific StandarC1 lime .2.012 Step; rEt-I P _RTP _7)O P I'I t n c l£: me n[ 1: Step nmp = 1.0 00 V..,,-: S, r*lzI:.;. P..C e f o rmed U Defr.-rme'tt on 5ci:lle +S.000e+02 008: ml S34-rebars-b.odb Abaqus/St.:lndard G.10-3 w03 d Feb 15 12;16;50 PaCific Standard li me 20 1:2 Step: Stgp-l-I-IOWlfJD H O T SS i E I'l RTP 730Pf>1 Incremen t 1:

Tl rn e == 1. 000 Plimar)' Vi!Jr: S, 511 (Cyl) Defo.-nleD V lIr : U CercI n1!.tlo n Scale FlI ctor: + 5 OOOe+02 Figure 16 -Radial Stress (psi) at the Azimuth 225° Location Page 18 Page 18 of 19 Exhibit 64 S, 522 (Cyl) (A"g: 75%) +1.726e+02.

+3. 000e.+02 +2. . S0 Qe+DZ +<.OOOc +02 +l.SOOe+OZ +1 .OOOC-f-02 +S .OOoe +Ol + O.OOOe +OO -S.OOOe+O l -1.00Ce+D2

-1.500e+0 2 --2.000e+02

--2. S00e+02 --3.000e+0 2 . -1. 54;le+03 00 8: m t53 4-rQbar s-u.odb Abaqus/S tanda,-d 6.10-3 WQd Fat> 1S 1 2: 16:58 Pa c l l , c $t.a nd old Tlnu i 2 012 Srep. Step-l -NOWWO_'IOT_SS_T E HP _RTP _73 0PN Increment 1: $tep lime;:: 1.0 0 0 PI1mary Var: 5, S2.1 (Cyl) Def Q rmed Vl.Ir: U Sc<ll e

+5.000e+02 at th e Azimuth 225° Lo c ation Figure 17 -Hoop Stress (psi) S , 533 (e\, l l (A': g : 7 5%) +6.+3. +2 .S+2.0+1 .5. +1.0o-+5.0 0-+D.--1.OOOe+0-1.500 e+-2 ,O-L.5aOe+02

-3 .00 0e+02 -1.066e+03 < y x COG : 6.10-3 Wed 1 5 1Z:16:56 PaClf.r.:. S tandard TIrn u '2 0 12 Ste p-1-HOV..'WO HO T S5 T I: HP RTP 7)OPH I n c re me nt 1: Ste p Time;: 1.0 0 0 Pll nl J ry V.)r: S , 533 (C yJ ) Qe (o rm Q d V l1r: U DQ fc I SC:tl-l l

+ s. oo oe+02 th e A zimut h 225° Location Fig ure 18 -Vertical Stress (psi) Page 19 Page 19 o f 19 Exhibit 65: Thermal Analysis Append ix VIII-66© 2012. Performance Im provement Intern ationa E x hi b it 65 DAVIS-BESSE THERMAL ANALYSIS 1.0 ANALYSIS MODEL For the purposes of assessing the seasonal and daily variations in the temperature of the outer concrete shield building of the Davis-Besse reactor, a detailed 3D transient thermal analysis finite element based model was generated.

This model was derived from the same 3D NASTRAI'J structural" model that was used for this effort and utilized the same node and element numbers. Additional surface flux thermal elements were added to the interior faces of the concrete in order to improve the capture of radiation heat transfer from the interior steel containment as well as convective heat transfer by the passage of the annular air. By preserving the same node numbers, this permitted directly mapping temperatures onto the NASTRAN structural model without having to interpolate temperatures between dissimilar meshes. This ensured that temperatures were accurately specified for all structural analyses performed with the NASTRAN model. Similarly, preserving the element ID numbers ensured proper specification of thermal properties for all of the materials present. The thermal analysis model that was used for this effort and is shown in Figure 1. The majority of the concrete was modeled using linear 8 node brick elements.

The steel rebar reinforcement was explicitly modeled using 1D bar elements that share the same nodes with the 3D solid elements used to represent the concrete.

Only in the dome region OF and IF rebar were membrane elements used to approximate the smeared thermal properties of the rebar and concrete based upon a volumetric averaging of their properties. All pre-and post processing of the thermal analysis model was performed using MSC MD.PATRAN version 10.2. MSC MD.PATRAN is an open ended pre-and post -processor that facilitates the creation and post-processing of results for a number of different CAE solvers. This includes MD.NASTRAI\J and ABAQUS, the two structural finite element analysis (FEA) solvers used for this effort. This enables models and results derived from one the analysis code to be converted into its equivalent in another code. In this way, the thermal models and results files could be converted into an equivalent ABAQUS version. Page 1 Page l of 23 Exhib i t 65 By doing so, thi s obviated the need to generate a separate ABAQUS based thermal analysis model. FIGURE 1. NASTRAN TRANSIENT THERMAL ANALYSIS Page Page 2 of Exhibit 65 2.0 THERMAL BOUNDARY CONDITIONS

2.1 RADIANT

HEAT TRANSFER SOURCES The primary intent of thermal analysis was to ascertain the variation in temperature that occurs daily as well as seasonally.

To accomplish this, it is essential that the variation in the position of the sun as it transits across the sky be properly modeled. This entails specifying the zenith angle, Z, or the angle of the sun relative to a normal pointing directly overhead.

The zenith angle is a function of both the latitude as well as the time of year. It is derived from the following relationship (see http://edmall.gsfc.nasa.gov/inv99Project.Site/Pages/sciencebriefs/ed-stickler/ed-irradiadiance.html

): Z::: Zenith Angle::: The angle from the zenith (a point directly overhead) to the Sun's position in the sky. The zenith angle is dependent upon the latitude, solar declination angle and time of day. Z::: cos-1 (sin tP sin 0+ cos tP cosH) q, =Latitude H Hour Angle = 15° x (Time -12) (Angle of radiation due to time of day where time is given as the hour of the day from midnight) = Solar Declination Angle Solar Declination Angles for the Northern Hemisphere Vernal Equinox March 21/226 0° Summer Solstice June 21/226 =+23.5° Autumn Equinox September 21/22 6 =0° Winter Solstice December 21/22 6 -23.5° The solar radiation that strikes the earth, also known as Insolation, is then simply given by 1= S cos (Z) I::: Insolation or solar flux S::: solar flux -1000 Watts/ m 2 -2.2 Btu/Hr -in 2 (Clear day insolation perpendicular to the incident solar radiation)

Z::: Zenith angle Page 3 Page 3 of 23 Exhibit 65 The solar insolation that strikes the earth is strongly affected by the angle of incidence with the surface being radiated by the sun. The more oblique the angle, the lower the flux. Consequently, in the latitudes farther from the equator, the solar insolation will be lower. In addition, seasonal variations will cause the solar declination to change by 47 degrees between the winter and summer solstices.

Thus, the solar flux will be least during the winter and greatest during the summer in the northern hemisphere.

The UV spectrum of sunlight is principally responsible for solar heating. It is strongly affected by the angle it passes through the atmosphere.

This is shown in Figure 2 1.4 June solstice 1.2 1.0 i .5 0.8 <<> Sept """" Yeal a: Mar. equinox 0.6 0.4 "'"" Dec. solstice 0 10 20 30 40 50 latitude (0) FIGURE 2. SEASONAL AND ANNUAL VARIATIONS IN RELATIVE SOLAR UV-A RADIATION (340 nm) FOR DIFFERENT LATTITUDES (BASED ON JOHNSON ET AL 1976) Page 4 Page 4 of 23