Text

Enclosure 4-3 to NOC-AE-13002986, Risk-Informed Closure of GSI-191 Volume 3 Engineering Casa Grande Analysis, Page 161 of 260 Through End
	ML13175A241
Person / Time
Site:	South Texas
Issue date:	06/06/2013
From:	Letellier B, Sande T, Zigler G; South Texas, Enercon Services, Los Alamos National Laboratory
To:	; Office of Nuclear Reactor Regulation
References
	GSI-191, NOC-AE-13002986 STP-RIGSI191-V03, Rev 1
	Download: ML13175A241 (116)
	v • d • e

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 5.3.3 Statistical Fit of NUREG-1829 LOCA Frequencies NUREG-1829 provides a set of LOCA frequency values (corresponding to the 5th percentile, median, mean, and 9 5 th percentile) for six different break sizes (Y2", 1-5/8", 3", 7", 14", and 31") (37). The values corresponding to each break size were fit with a bounded Johnson distribution to define the full range of epistemic uncertainty associated with LOCA frequencies (8). This is illustrated in Figure 5.3.2.

100...

00 PWR 25-yr Fleet-Average Operation

-2 10.

+ Bounded Johnson S10"-4 II A

0) *

~10 10 , 951h

, mean CU 10 + 50h 10 10-t2o 0 5 10 15 20 25 30 35 Break Size (in.)

Figure 5.3.2 - Illustration of bounded Johnson fit for NUREG-1829 break frequencies The bounded Johnson cumulative distribution function and optimization model are shown in the Equation 31 and Equation 32 (8), and the fitted parameters are provided in Section 2.2.3.

F[x] = Oty + 8f[(x - Equation 31 Page 161 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 where cD[x] is the cumulative distribution function of a standard normal random variable, y and 6 are shape parameters (with y driving the distribution's skewness), k is a location parameter, X is a scale parameter, and f(z) = log[z / (1-z)] for ý < x:5 k + X.

min (F[xo.os] - 0.05)' + (F[xO.so] - 0.50)2 + (F[xo.95 ] - 0.95)2 s.t. X._- .

- UdLIUa I 32 2~+ Žx 0.95 6,, Ž 0 5.3.4 Sample Epistemic Uncertainty of LOCA Frequencies Given the fitted distribution parameters, the epistemic uncertainty of the LOCA frequency data in NUREG-1829 can be sampled. For example, ifthe 6 2 nd percentile is selected, the LOCA frequencies can be calculated based on Equation 31 and the parameters in Section 2.2.3. The calculated 6 2 nd percentile values are shown in Table 5.3.10. Figure 5.3.3 shows the LOCA frequency vs. break size for the 6 2 "d percentile assuming linear interpolation between the values in Table 5.3.10. (Note that the shape of the interpolated curves appears to be non-linear on a semi-log plot.)

Table 5.3.10 - Example calculation of LOCA frequencies vs. break size for 6 2 nd Percentile 62nd Percentile B ize LOCA Frequencies (in) (year")

0.5 1.06E-03 1.625 1.66E-04 3 6.35E-06 7 5.92E-07 14 2.74E-08 31 2.89E-09 Page 162 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1 LOCA Frequency for 62nd Percentile 1.01-02 1.0E-03 1.OE-04 4

£ 101-05 U.

1.0E-07 1.01-08 1.OE-09 0 5 10 15 20 25 30 35 Break Size (in)

Figure 5.3.3 - Illustration of LOCA frequency vs. break size for 6 2 nd percentile 5.3.5 Distribute Total LOCA Frequency to Weld Locations The overall LOCA frequencies determined from NUREG-1829 are distributed to each weld category using the relative frequencies discussed in Section 5.3.1. Based on linear interpolation of Table 5.3.10, the frequency of a 1.5-inch break is 2.65E-04. The relative contribution for a 1.5-inch break in a Category 2 weld (see Table 5.3.7) is 3.90% for four welds or 0.98% per weld. Therefore, the frequency of a 1.5-inch break in one of those welds is 2.60E-06 (2.65E-04

0.98% = 2.60E-06).

5.3.6 Sample Break Sizes at Each Weld Location CASA Grande evaluates multiple sizes of breaks at every weld in containment, and it always includes the DEGB condition for every weld. The total number of break scenarios investigated for each weld is Page 163 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision I determined based on user input for the maximum desired number of breaks in the largest pipe, NL. One of these breaks is assigned to the DEGB condition, and the remaining number, NL- 1, are assigned randomly in the large break size range. The range of break sizes for a given weld was subdivided into a number of intervals proportional to the range of the largest possible LBLOCA. The standard LOCA bins of 0.5 to 2 inches (SBLOCAs), 2 to 6 inches (MBLOCAs), and greater than 6 inches (LBLOCAs) were used; so the number of breaks in the small and medium range were determined by the following formulas:

Ns= ceil 2-0.5 6 NL) Equation 33 (max-6 NM =ceil D 6 4 NL) Equation 34 miax-6 where:

Ns = Number of breaks in SBLOCA category NM = Number of breaks in MBLOCA category NL = Number of breaks in LBLOCA category Dmax = Maximum break size (in)

The ceil(x) operator simply rounds up to the nearest integer. This guarantees that there is always at least one small break and at least one medium break at every weld that can support breaks of these sizes.

Given the desired number of breaks in each LOCA category, the conditional probability for breaks in the associated weld case was divided into an equivalent number of non-uniform bins (unequal size), and the probability weights for each bin were recorded. Random percentiles were selected from each probability bin, and the conditional probability was interpolated to find corresponding break sizes.

(Neither the probability bins, nor the corresponding size intervals are of equal size.) The set of discrete break sizes are matched with their probability weight and carried throughout the evaluation as independent break scenarios.

When this algorithm is applied to the STP weld population for NL = 10, the total number of scenarios is approximately 3,070. When NL = 5, the number of scenarios is approximately 2,250, and when NL = 3, the number of scenarios is approximately 2,100. For this evaluation, all sampling replicates were run with NL = 5.

Figure 5.3.4 illustrates the break-size selection process for Weld Case 1B, which includes the largest pipes in containment. LOCA category limits are marked with vertical solid lines. The DEGB condition, marked with a red dot, represents one of the 10 breaks imposed on the LBLOCA range. The remaining nine equal break-size intervals are separated by vertical dashed lines between 6 inches and 31.5 inches Page 164 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 (the maximum pipe diameter). Note that the size intervals only appear unequal because of the logarithmic scale. By relative proportion of their respective ranges, only two break intervals are assigned to MBLOCAs, and only one is assigned SBLOCAs. Thus, for this example, 13 breaks are simulated at each weld belonging to Weld Case lB.

Illustration of Logarithmic Sample Bins for Case 18 II A l9

~10 0

ca 10 7

C) 10 "11 101 102 100 break size (effective diam, in.)

Figure 5.3.4 - Example of non-uniform stratified sampling strategy for one weld case 5.4 Debris Generation Debris generation analysis includes calculations of the total quantity of insulation, coatings, latent, and miscellaneous debris, as well as a definition of debris characteristics (size and density). These topics are discussed in this section.

5.4.1 ZOI Model The quantity of insulation debris generated is calculated directly in CASA Grande based on the currently accepted deterministic ZOI model. As described in NEI 04-07 Volume 2, the break jet ZOI can be conservatively modeled as a sphere for a fully offset DEGB or as a hemisphere for anything less than a DEGB (i.e. a side-wall pipe break) (44). The ZOI radius depends on the destruction pressure of the Page 165 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 insulation and the size of the break. As shown in Section 2.2.15, the ZOI sizes for insulation at STP are 2D for Transco RMI, 17D for Nukon and Thermal-Wrap (assumed to be the same as Nukon), and 28.6D for Microtherm (assumed to be the same Min-K). All insulation that falls within its respective ZOI is assumed to become debris.

Figure 5.4.1 through Figure 5.4.3 show examples of the ZOls for a large 31-inch DEGB, a medium 6-inch side-wall break, and a small 2-inch side-wall break. Because of the spherical ZOI assumption, the direction of the jet is irrelevant for DEGBs (see Figure 5.4.1). The jet direction and orientation of the hemispherical ZOI for side-wall breaks is dependent on the break location radially around the pipe, but the ZOI is constrained along the axis of the pipe (see Figure 5.4.2 and Figure 5.4.3).

Page 166 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1 Figure 5.4.1 - Illustration of 17D Nukon ZOI for a 31" DEGB Page 167 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Figure 5.4.2 - Illustration of 17D Nukon ZOI for a 6" side-wall break Figure 5.4.3 - Illustration of 17D Nukon ZOI for a 2" side-wall break Page 168 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Jet formation modeling was conducted to evaluate the potential conservatism in the ZOI size and shape (10). However, the effects of realistic jets on the ZOls were not explicitly considered in this evaluation.

5.4.2 Insulation Debris Size Distribution Model To implement the fiberglass debris size distribution described in Section 2.2.16, the fiberglass ZOI was split into three sub-zones. The quantity of fiberglass insulation in each sub-zone was multiplied by the appropriate percentage of fines, small pieces, large pieces, and intact blankets as defined in Table 4.1 of the Alion debris size distribution report (45). Figure 5.4.4 shows an example of the size distribution sub-zones.

11.91D Sub-Zone Figure 5.4.4 - Illustration of sub-zones used for fiberglass debris size distribution Page 169 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 The Microtherm debris was assumed to fail as 100% fines with components of Si0 2, TiO 2, and fibers as described in Section 2.2.16.

5.4.3 Insulation Debris Using the LOCA frequency sampling strategy described in Section 5.3, three replicates of approximately 2,250 break scenarios each were sampled to illustrate the probability distribution associated with ZOI debris volume. These calculations assumed a 17D ZOI for Nukon and Thermal Wrap insulation. Figure 5.4.5 shows the complementary cumulative probability distribution function formed from the fiberglass debris quantities calculated for these scenarios with the relative initiating event frequencies included as probability weights. As shown on this figure, the maximum quantity of fiberglass debris that can be generated approaches 3,000 ft 3, but 99.9% of the scenarios generate less than 10 ft 3 of fiberglass debris.

WaNs 10 0 Dist of ZOI Debris Volume - With 102 A 10 0

E

.0 -9

> 10

-10 10 10'12 10"14 0

10"1 100 10 10 102 104 total debris volume (fis)

Figure 5.4.5 - Distribution of potential fiberglass debris quantities 5.4.4 Qualified Coatings Debris Similar to insulation debris, the quantity of qualified coatings debris is calculated based on the quantity of coatings within the ZOI. However, due to the difficulty of accurately modeling all of the coated Page 170 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 surfaces within CASA Grande, the qualified coatings debris calculations were performed outside of CASA Grande using the CAD model. As described in Section 2.2.10, bounding quantities of qualified epoxy and IOZ coatings debris were determined for break sizes of 2-inch, 6-inch, 15-inch, and 31-inch DEGB. The bounding quantities were applied for breaks less than or equal to the break size listed in Table 2.2.14 (e.g. the qualified coatings quantity for a 15 inch break was used for all break sizes greater than 6 inches and less than or equal to 15 inches).

5.4.5 Unqualified Coatings Debris The inputs for unqualified epoxy, alkyd, IOZ, and baked enamel coatings failure are provided in Section 2.2.11. For each of the unqualified coatings, the total quantity is multiplied by the failure fraction to determine the actual quantity of unqualified coatings debris generated. The quantity of unqualified coatings debris that transports to the strainers (as well as the arrival time at the strainers) is dependent on both the failure location and failure timing. Therefore, these inputs were provided in Section 2.2.11 also. The following equations illustrate the method for calculating the time-dependent and cumulative coatings failure:

Mij(t) = Mtotai,ij " Ff.il,i " F(t) Equation 35 Mij(t)

Mijcum -F(t) = M total,ij ' Ffaili Equation 36 where:

M(t) = Mass of unqualified coatings that fail during a specific time period t = Specific time period following the start of the accident Subscript i = Unqualified coating type (epoxy, IOZ, alkyd, or baked enamel)

Subscript j = Coating location (upper containment, lower containment, or reactor cavity)

Mtotai = Total mass of unqualified coatings Ffail = Total failure fraction F(t) = Fraction of coatings that fail during a specific time period Mij,cum = Cumulative mass of unqualified coatings that fail 5.4.6 Latent Debris The quantities of latent fiber and latent dirt/dust were entered as input parameters in CASA Grande based on the values specified in Section 2.2.13. The total quantity of latent debris is applicable to all LOCA scenarios.

Page 171 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 5.4.7 Miscellaneous Debris Unqualified tags, labels, plastic signs, tie wraps, etc. are assumed to fail for all LOCA scenarios. The total quantity of miscellaneous debris was entered as an input parameter in CASA Grande based on the value specified in Section 2.2.14.

5.4.8 Debris Characteristics The important debris properties were entered as input parameters in CASA Grande based on the values specified in Section 2.2.17. The parameters that are important for GSI-191 calculatio ns include the characteristic diameters of particles and fibers, the macroscopic (or bulk) density of debris, and the microscopic (or particle) density of debris.

5.5 Chemical Effects The analysis of chemical effects combines the calculated release of materials with specified solubility limits of expected chemical products to estimate whether chemical products will exist in solution. These topics are discussed in this section. Chemical effects analysis also includes an evaluation of the head loss due to the chemical precipitates accumulating on the sump strainers or in the core. The impact of chemical effects on strainer head loss is discussed in Section 5.7, and core blockage is discussed in Section 5.11.

Various metals and insulation exposed to the post-LOCA containment fluid can corrode or dissolve over time as a function of water volume, temperature, pH, and the potential development of a protective layer that can inhibit corrosion or dissolution. A generic corrosion/dissolution model was developed by the PWROG as documented in WCAP-16530-NP (73). This model identified aluminum, silicon, and calcium as the primary chemicals that may form precipitates. The WCAP model assumed a solubility limit of 0, meaning that essentially all of the chemicals in solution were assumed to form aluminum oxy-hydroxide (AIOOH), sodium aluminum silicate (SAS), or calcium phosphate precipitates1 2 . The WCAP method for creating surrogate precipitates to use in head loss testing resulted in an amorphous form of the precipitates.

Based on an initial evaluation of STP conditions compared to the integrated chemical effects test (ICET) program, the Alion integrated test program conducted at the VUEZ facility in Slovakia, and other testing, the WCAP methodology was thought to be overly conservative (15). Testing is currently being conducted at the University of New Mexico (UNM) to build off of the previous chemical effects test programs and develop more refined chemical effects models specific to STP conditions. As part of the UNM Chemical Head Loss Experiment (CHLE) test program, a number of chemical effects issues that had been previously addressed with conservatisms are being reevaluated (21).

12 Calcium phosphate precipitates are only assumed to form if TSP is used as a buffer.

Page 172 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 5.5.1 Chemical Concentration Model The chemical effects test program sponsored by the PWROG developed material release equations that have been used by the industry to inform head loss testing (73). The use of these equations have proven to produce conservative outcomes when comparing the predictive results to experimental results taken from other testing programs (15). To obtain more realistic values, the material release under STP conditions was further evaluated by the CHLE program. One objective of this research program was to obtain results specific to STP conditions that would refine the predictive equations and reduce the resulting conservatism (21). Current CHLE test results support the initial hypothesis that the WCAP-16530-NP methodology is conservative for STP conditions (18; 19). However, the current CHLE results are not sufficient to generate new predictive material release equations. Therefore, the WCAP-16530-NP methodology was used to conservatively evaluate material release for a range of scenarios (20). This is described in more detail in Section 5.7.3.

5.5.2 Solubility Limit The solubility limits for various types of precipitates are a function of temperature and pH. Solubility limits are also dependent on the form of precipitates. Crystalline forms of a given type of precipitate generally have a lower solubility limit than amorphous forms. However, since the energy required to create an amorphous precipitate is lower than a crystalline precipitate, it is more likely for an amorphous precipitate to be formed if the chemical concentration is higher than both the crystalline and amorphous solubility limits.

While it is desirable to use solubility limits for precipitates that may form under STP conditions as determined by CHLE test results, time constraints and limited experimental results required a more simplified approach. A reasonably conservative solubility limit for compounds previously identified to form in solution under LOCA conditions (73) were evaluated as function of STP temperature and pH (20).

This solubility limit was used in conjunction with the material release results to estimate product formation. This is described in more detail in Section 5.7.3.

5.5.3 Chemical Product Type, Form, and Quantity (Pool and Core)

Chemical products that form as a result of a LOCA scenario are a function of the resulting thermodynamic conditions and soluble chemical concentrations. The resulting products may be amorphous or crystalline and may precipitate in the bulk solution or develop as scale on surfaces within the containment structure. It is desirable to identify and characterize chemical products that will form under STP conditions. However, the current test results and time constraints did not allow for this.

Information from past testing and current CHLE results were used to estimate chemical product formation and associated morphology.

Page 173 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Previous testing (73) identified the formation of amorphous aluminum and calcium products. However, the CHLE test program has not observed the formation of these products in easily detectable quantities.

The most significant chemical product observed during the CHLE test program was a crystalline zinc phosphate mineral that appeared to nucleate and deposit on other test materials (19). Crystalline products are similar to particulate debris, and generally result in significantly lower head losses than is often observed with large amorphous precipitates generated using the WCAP-16530-NP surrogate recipe (73). While the zinc phosphate product may be an important parameter in the overall GSI-191 evaluation, it is not expected to result in high head losses on the strainer or in the core due to its crystalline structure.

5.6 Debris Transport Debris transport is the estimation of the fraction of debris that is transported from the location where it is generated to the sump strainers. The four major debris transport modes are:

Blowdown transport- the vertical and horizontal transport of debris to all areas of containment by the break jet.

Washdown transport- the vertical (downward) transport of debris by the containment sprays and break flow.

Poolfill transport- the horizontal transport of debris during the RWST injection phase to regions of the pool that may be active or inactive during recirculation.

Recirculationtransport- the horizontal transport of debris from the active portions of the recirculation pool to the sump strainers.

The four transport modes, potential upstream blockage, fiberglass debris erosion, and time-dependent transport are all discussed in this section.

5.6.1 Upstream Blockage Potential upstream blockage points at STP include the four 30-inch vent holes in the secondary shield wall (see Figure 5.2.7 and Figure 5.6.1) and the two 6-inch refueling canal drain lines. These potential blockage points were previously evaluated as part of the deterministic GSI-191 analysis, and it was shown that they would not be clogged with debris (64; 74).

Page 174 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1 Figure 5.6.1 - Photograph of 30-inch vent hole in secondary shield wall 5.6.2 Blowdown Transport The transport of debris during the blowdown phase is dependent on the break location. The blowdown transport fractions are provided in Section 2.2.21.

5.6.3 Washdown Transport Debris transport during the washdown phase was assumed to be negligible for cases where containment sprays are not initiated (see Assumption 6.b). For any scenarios where the sprays are initiated, the washdown transport fractions are provided in Section 2.2.22.

5.6.4 Pool Fill Transport Pool fill transport is dependent on whether the break occurs inside or outside the secondary shield wall.

The pool fill transport fractions are provided in Section 2.2.23.

Page 175 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSl191-V03 Revision 1 5.6.5 Recirculation Transport The transport of debris during the recirculation phase is dependent on the break location and size. The recirculation transport fractions are provided in Section 2.2.24.

5.6.6 Debris Erosion Pieces of fiberglass debris that are held up on grating and exposed to spray, and pieces of fiberglass debris that settle in the recirculation pool would be subject to erosion. The erosion fractions and erosion timing are provided in Section 2.2.25.

5.6.7 Strainer Transport The total transport to the ECCS strainers was determined based on the logic tree method described in NEI 04-07 (43). The transport fractions can be calculated using Equation 37 for debris generated inside the ZOI, Equation 38 for unqualified coatings debris generated outside the ZOI, and Equation 39 for latent debris.

DTFzot FBD(upper) f ((1 -) FWD(insidle) - FWD(annulus))" FErosion(spray) + FWD(inside)

[FRecirc(WDnside) + (1 - FRecirc(WDinside))" F.rosion(poot)]

+ FWD(annulus)

,[FRecirc(WDannulus) + (1 - FRecirc(WDannulus))" FErosion(pool)]l

+ (1 - FBD(upper) - FBD(lower))

- {(1 - FWD(BCinside) - FWD(BCannulus))" FErosion(spray)

+ FWD(BCinside) Equation 37

,[FRecirc(WDinside) + (I - FRecirc(WDinside))* F.rosion(pool)]

+ FWD(BCannulus)

, [FRecirc(WDannulus) + (1 - FRecirc(WDannulus))" FErosion(pool)

+ FBD(lower) f (i - 3 " FPF(sufp) - FpF(inactive))

[FRecirc(lower) + (1 - FRecirc(lower))" FErosion(pool)] + Nsumps FPF(sump)}

where:

DTFzo1 = Total debris transport fraction (for particular type/size of debris generated in the ZOI)

FBD(upper) = Blowdown fraction to upper containment FBD(lower) = Blowdown fraction to lower containment FWD(inside) = Washdown fraction inside secondary shield wall Page 176 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 FWD(annulus) = Washdown fraction in annulus FWD(BCinside) = Washdown fraction from break compartment to inside secondary shield wall FWD(BCannulus) = Washdown fraction from break compartment to annulus FPF(sump) = Pool fill fraction to each sump strainer FPF(inactive) = Pool fill fraction to inactive cavities Nsumps = Number of ECCS sumps in operation during recirculation FRecirc(Iower) = Recirculation fraction for debris initially blown to lower containment FRecirc(WDinside) = Recirculation fraction for debris washed down inside secondary shield wall FReirc(WDannulus) = Recirculation fraction for debris washed down in annulus FErosion(spray) = Erosion fraction for debris held up above the pool FErosion(pool) = Erosion fraction for non-transporting debris in the pool DTFuc = Ffail [Fupper -Fspray " FRecirc + Flower FRecirc + Freactor Equation 38 FRecirc(reactor)]

where:

DTFuc = Total debris transport fraction (for particular type/size of unqualified coatings debris)

Ffa11= Total failure fraction Fupper = Fraction located in upper containment Flower= Fraction located in lower containment Freactor = Fraction located in the reactor cavity Fspray = Fraction of coatings that would fail prior to securing containment sprays 3

FReirc = Recirculation fraction for debris washed to or initially in lower containment' FRecirc(reactor) = Recirculation fraction for debris in reactor cavity14 DTFLD = FuppeT " FWD " FRecirc + Flower

[(1-K 3"FpF(sunp) - FPF(inactive)) FRecirc + Nsumps Equation 39

FpF(sump)]

13 The recirculation transport is assumed to be the same for unqualified coatings washed down to the pool and unqualified coatings that are initially in lower containment since the locations where debris would be washed down and the locations where unqualified coatings exist in lower containment are spread out and can be reasonably treated as a uniform distribution (23).

14 The recirculation transport for unqualified coatings in the reactor cavity would be 0% for all cases except a reactor cavity break. In the case of a reactor cavity break, the reactor cavity unqualified coatings were assumed to transport the same as the unqualified coatings outside the reactor cavity (23).

Page 177 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 where:

DTFLD = Total debris transport fraction (for particular type/size of latent debris)

Fupper = Fraction located in upper containment Flower= Fraction located in lower containment FWD = Total washdown fraction FPF(sump) = Pool fill fraction to each sump strainer FPF(inactive) = Pool fill fraction to inactive cavities Nsumps = Number of ECCS sumps in operation during recirculation FRecirc = Recirculation fraction for debris washed to or initially in lower containment1 5 Figure 5.6.2 through Figure 5.6.7 show example transport logic trees for each type and size of debris generated inside the ZOI for a break in the steam generator compartments. The washdown transport fractions shown are based on the actuation of containment sprays (i.e. CS flow is greater than 0 gpm),

and the pool fill transport fractions are based on all three sumps being active (i.e. at least one pump is running on three different trains).

Blowdown Washdown Recirculation Erosion Fraction of Debris Debris Size Transport Transport ransport Transport at Sump

[.00 Retained on Structures 1.00 0.371 0.70 0.53 Transport Upper washed Down 0.0 Containment Inside Secondary Shield Wall Sediment 1.00 0.329 0.47 Transport Washed Down in 0.00 Annulus Sediment LDFG 0.00 (Fines) SG Compartments 1.00 0.267 0.89 Transport Active Pool (300 Sediment 0.06 0.018 0.30 Active Sump(s)

Lower Containment 0.00 inactive Surnp(s) 0.05 Inactive Cavities Sum: 0.985 Figure 5.6.2 - Example logic tree for LDFG fines (SG compartment LBLOCA) 15 The recirculation transport is assumed to be the same for latent debris washed down to the pool or initially at the pool elevation. However, the transport calculation simply assumed that all latent debris would be at the pool elevation (23).

Page 178 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Blowdown Washdown Pool Fill Transport Erosion Fraction of Debris Debris Size Iz Transport I Transport II Trnp t II atSm DbiS 0.01 0.003 0.64 Erodes to Fines Retained on 0.99 Structures Remains Intact 0.64 0.078 Transport 0.26 A A'I 0.47 Washed Down 0.003 Inside Secondary 0.36 Erodes to Fines Upper Containment Shield Wall Sediment 0.93 Remains Intact 0.58 0.027 0.10 I Transport Washed Down in 0.07 0.001 Annulus 0.42 Erodes to Fines Sediment 0.93 Remains Intact 0.01 0.002 0.74 Erodes to Fines Hetained on 0.99 Structures Remains Intact LDFG 0.33 (Small Pieces)

SG Compartments Tran6or 0.055 0.26 I Transport nn 07 n17 0.002 Washed Down Inside Secondary 0.36 Erodes to Fines Shield Wall Sediment 0.93 Remains Intact 0.64 0.128 1.00 I ransport Active Pool 0.07 0.005 0.36 Erodes to Fines Sediment 0.93 Remains Intact 0.20 N An AANA Lower 0.20 n00001 i Active Sump(s)

Containment 0.00 Inactive Sump(s) 0.00 Inactive Cavities Sum: 0.304 Figure 5.6.3 - Example logic tree for LDFG small pieces (SG compartment LBLOCA)

Page 179 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision I Blowdown Washdown Recirculation Erosion Fraction of Debris Debris Size Transport Transport Pool Fill Transport Transport I at Sump 0.01 0.001 1.00 I Erodes to Fines

- . .--. 4 Retained on 0.99 0.12 Structures Remains Intact Upper Containment 0.00 Lower Containment 0.01 0.009 LDFG 1.800 1.00Erodes to Fines (Large Pieces) Retained on 0.99 0.88 Structures Remains Intact SG Compartment 0.00 Lower Containment 0.O0 Lower Containment Sum: 0.010 Figure 5.6.4 - Example logic tree for LDFG large pieces (SG compartment LBLOCA)

Blowdown Washdown Recirculation E s Fraction of Debris Debris Size Transport Transport Pool Fill Transpor Transport I rosion at Sump 0.00 Retained on Structures 1.00 0.371 n 7fl 0.53 Transport 070 Upper Washed Down 0.00 Containment Inside Secondary Sediment Shield Wall 1.00 0.329 0.47 Transport I

Washed Down in 0.00 Annulus Sediment Microtherm 0.00 (Fines) SG Compartments 0.89Transpo 0.267 1ciePo

.00026 0.06 0.018 0.30 Active Sump(s)

Lower Containment 0.00 Inactive Sump(s) 0.05 Inactive Cavities Sum: 0.985 Figure 5.6.5 - Example logic tree for Microtherm fines (SG compartment LBLOCA)

Page 180 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Debris Size Ellowooo Transport Was--oon Transport Pool Fill Transport "Recirculation Transport Erosion Fraction iofDebris at Sump 0.00 Retained on Structures 1.00 0.371 0.70 0.53 Transport Upper Washed Down 0.00 Containment Inside Secondary Sediment Shield Wall 1.00 0.3290 0.47 5 Transport Washed Down in 00 AnnulusSediment Crud 0.00 (Fines) SG Compartments 1.00 0.267 0.89 t Transport Active Pool 0.00 Sediment 0.00.06 0.018 Lower Active Sump(s)

Containment 0.00 Inactive Sump(s) 0.05 Inactive Cavities Sum: 0.985 Figure 5.6.6 - Example logic tree for crud fines (SG compartment LBLOCA)

Page 181 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Olowdown Washdown Tecrcuatinsprrso Fraction of Debris Debris Size Transport Transport Pool Fill Transport Transport Erosion at Sump 0.00 Retained on Structures 1.00 0.371 0.70 0.53 Transport Upper Washed Down 0.00 Containment Inside Secondary Sediment Shield Wall 1.00 0.329 0.47 Transport Washed Down in 0.00 Annulus Sediment Oualified Coatings 0.00 (Fines) SG Compartments 1.00 0.267 0.89 F Transport Active Isooll 0.00 Sediment 0.06 0.018 0.30 Active Sump(s)

Lower Containment 0.00 Inactive Sump(s) 0.05 Inactive Cavities Sum: 0.985 Figure 5.6.7 - Example logic tree for qualified coatings fines (SG compartment LBLOCA)

Figure 5.6.8 through Figure 5.6.15 show example transport logic trees for each type and size of debris generated outside the ZOI for a break in the steam generator compartments. The transport fraction for the unqualified coatings is highly dependent on the initial failure fraction (which could be as high as 100%), as well as the failure timing for the coatings in upper containment. Since the majority of unqualified coatings would fail after 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months (approximately 94% as shown in Section 2.2.11), and the sprays would generally be secured within a few hours, most of the unqualified coatings in upper containment would not be washed down to the pool.

Page 182 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Debris Size Failure Fraction Initial Location Washdown Transport Pool Fill TransIport I Recirculation Tasport Fraction of Debris at Sump 1.00 0.014 0.06 I ransport Fails Prior to 0.00 0.54 Securing Sprays Sediment Upper Containment 0.94 Fails After Securing Sprays 0.43 1.00 0.198 Fails 0.46 iransport Unqualilied Alkyd Lower 0.00 Coatings Containment Sediment (Fines) 0.00 Reactor Cavity 0.57 Remains Intact Sum: 0.212 Figure 5.6.8 - Example logic tree for unqualified alkyd coatings fines (SG compartment LBLOCA)

Debris Size Initi Location Washdown T Recirculation Fraction of Debris Failure Fraction nia Transport P F Transport at Sump 1.00 0.007 0.06 Transport Fails Prior to 0oo 0.15 Securing Sprays Sediment Upper Containment 0.94 Fails After Securing Sprays 1.00 0.016 0.80 0.02 Transport Fails Lower 0.00 Containment Sediment Unqualified Epoxy Coatings 0.00 0.000 (Fines) 0.83 (Fies)0.0 Transport Reactor Cavity 1.00 Sediment 0.20 Remains Intact Sum: 0.023 Figure 5.6.9 - Example logic tree for unqualified epoxy coatings fines (SG compartment LBLOCA)

Page 183 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Debris Size Failure Fraction Initial Location Transport Fil Transport at Sump 0.41 0.003 0.06 i ranspon Fails Prior to 0.59 0.15 Securing Sprays Sediment Upper Containment 0.94 Fails After Securing Sprays 0.80 0.41 0.007 Fails 0.02 Transport Lower 0.59 Containment Sediment Unqualified Epoxy Coatings 0.00 0.000 (Fine Chips) 0.83 Transport Reactor Cavity 1.00 Sediment 0.20 Remains Intact Sum: 0.010 Figure 5.6.10 - Example logic tree for unqualified epoxy coatings fine chips (SG compartment LBLOCA)

Deris Size Failure Fraction Initial Location Wastidown Pool Fi Transport Recirculation Fraction of Debris IIII TrI Transport I Fat Sump 0.00 0.000 0.06 Transport Falls Prior to 1,00 0.15 Securing Sprays Sediment Upper Containment 0.94 Fails After Securing Sprays 0.00 0.080 0.80 Falls 0.02 Transport Lower 1.Q01 Containment Sediment Unquawied Epoxy Coatings 0.00 0.000 (Snau Chips) 0.83 Transpo Reactor Cauity 1.00 Sedknard 0.20 Remains Intact SUMi 0.000 Figure 5.6.11 - Example logic tree for unqualified epoxy coatings small chips (SG compartment LBLOCA)

Page 184 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Debris Size Failure Fraction Initial Location Washdown Pool Fill Transport Recirculation Fraction of Debris I ~~~~~ ~ ~ ~

IIITasotIITa a I.1Sump 0.00 0.O000 0.06 Transport Fails Prior to 1.00 0.15 Securing Sprays Sediment Upper Containment 0.94 Fails After Securing Sprays n.0n 0.000 0.80 0.000 Fails 0.02 Transport Lower 1.00 Containment Sediment Unqualified Epoxy Coatings .Ann O.O00 0.000 (Large Chips) 0.83 Transport Reactor Cavity 1.00 Sediment 0.20 Remains Intact Sum: 0.000 Figure 5.6.12 - Example logic tree for unqualified epoxy coatings large chips (SG compartment LBLOCA)

Debris Size Failure Fraction Washdown I Recirculation Fraction of Debris Initial Location ool Fillt Transport Transport at Sump 1.00 0.007 0.06 Transport Fails Prior to 0.00 0.15 Securing Sprays Sediment Upper Containment 0.94 Fails After Securing Sprays 1.00 0.016 0.80 Fails 0.02 Transport Lower 0.00 Containment Sediment Unqualified Epoxy Coatings 0.00 0.000 (Curled Chips) 0.83 Transport Reactor Cavity 1.00 Sediment 0.20 Remains Intact Sum: 0.023 Figure 5.6.13 - Example logic tree for unqualified epoxy coatings curled chips (SG compartment LBLOCA)

Page 185 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Debris Size Failure Fraction Initial Location Washdown Pool Fill Transport Recirculation Fraction of Debris I I I Transport I I Transport at Sump 1.00 0.046 0.06 Transport Fails Prior to 0.00 0.83 Securing Sprays Sediment Upper Containment 0.94 Fails After Securing Sprays 0.92 1.n0 0.156 Fails 0.17 0.17 Trnsport Unqualified IOZ Lower 0.00 Coatings Containment Sediment (Fines) 0.00 Reactor Cavity 0.08 Remains Intact Sum: 0.202 Figure 5.6.14 - Example logic tree for unqualified IOZ coatings fines (SG compartment LBLOCA)

Fraction of Debris Debris Size Initial Location Washdown TransportTanprPool Fill Transport Tran Recirculation Erosion Eoi o at Sump 0.00 Upper Containment 1.00 0.890 Latent Debris 0.09R Transport (Fines) Active Pool 0.00 Sediment 0.06 0.060 1.00 Active Sump(s)

Lower Containment 0.00 Inactive Sump(s) 0.05 Inactive Cavities Sum: 0.950 Figure 5.6.15 - Example logic tree for latent fines (SG compartment LBLOCA)

As discussed in Assumption 6.g, debris accumulation on the strainers is assumed to be proportional to the strainer flow split. Therefore, the debris accumulation on each individual strainer can be calculated as shown in Equation 40.

DTFS11mp(X) = DTF - Qs11n1P(x) Equation 40 QSump(A) + QSU~np(B) + QSurnp(C) where:

DTFsump(x) = Debris transport fraction to Sump(X) for a particular type and size of debris Page 186 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Sump(X) = Sump(A), Sump(B), or Sump(C)

DTF = Total debris transport fraction for a particular type and size of debris Qsump(x) = Flow rate to Sump(X)

QSump(AB,C) = Flow rate to Sump(A,B,C)

If all pumps are operating at the same flow rate in all three trains, 33.3% of the transported debris would accumulate on each strainer. However, if the pumps in two trains failed, 100% of the transported debris would accumulate on the active strainer.

5.6.8 Time-Dependent Debris Arrival Model There are several factors that must be taken into consideration to analyze time-dependent arrival of debris at the strainers or in the core. These factors were addressed in the debris transport calculation as summarized in Table 5.6.1 and illustrated in Figure 5.6.16 (23).

Table 5.6.1 - Time-dependent transport Source Time or Equation Comments t = -0 s (no curbs around inactive Assume only applies for debris blown cavity entrances) to pool and latent debris t ~~ 425 s (based t -425s on (ase onaa flow fow rate of ateof Assume only applies to debris blown Sump Strainer Fill 14,040 gpm and a pool volume of 13,325 ft3) to pool and latent debris Total Fill (Switchover) t - 20 min (LBLOCA)

Assume washdown occurs after Initial Washdown 6 s - 10 s (s); inactive and sump cavities are filled, 2 man -5 min (small pieces) but before recirculation is initiated Unqualified Coatings Assume instant washdown at time of Failure failure if sprays are on Recirculated Spray Flow t -300s Assume instant washdown Debris Washdown Recirculated Debrislashd Break Flow t < 300s Assume instant washdown Debris Washdown Spray Erosion Washdown t < 15 min Assume during pool fill Pool Erosion Recirculation 0-30 days Initial Debris in Pool at xi = blowdown + initial washdown Total debris in pool from blowdown start of recirculation (xi) debris and initial washdown Debris Recirculation Time x(t) = xie't(O/vP°°l) Q = flow rate; Vpool = Pool Volume (x(t))

Page 187 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Figure 5.6.16 - Illustration of time-dependent transport As shown in Table 5.6.1, the transient transport of debris in the pool at the start of recirculation can be calculated using Equation 41. This equation can also be applied to debris that enters the pool later in the event (i.e. failed unqualified coatings or debris circulated through the break or containment sprays).

x(t) = x - t( Q 1)

Equation 41 where:

x(t) = Time dependent arrival of debris at the strainer(s) xi = Initial quantity of debris in the pool at the start of recirculation t = time Q = Total sump flow rate Page 188 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Vpoo, = Pool volume 5.7 Strainer Head Loss Overall head loss across the strainer includes the clean strainer head loss as well as the debris bed head loss from both conventional debris (fiber, particulate, RMI, paint chips, etc.) and chemical precipitates. If the strainer head loss exceeds the NPSH margin of the pumps, the pumps would fail. Similarly, if the head loss exceeds the structural margin of the strainers, the strainers would fail potentially allowing large quantities of debris to be ingested into the ECCS.

5.7.1 Clean Strainer Head Loss CSHL is a function of the sump flow rate and the containment pool temperature (viscosity and density).

The head loss can be generically described using the following equation:

AH = apU + bpU2 Equation 42 where:

AH = Head loss a = Empirically derived viscous coefficient b = Empirically derived inertial coefficient V = Dynamic viscosity p = Density U = Approach velocity Using the test data provided in Section 2.2.27, the values in Table 5.7.1 can be plotted to derive the viscous and inertial coefficients. Since the temperature differences are minor, the changes in viscosity and density can be reasonably neglected. Figure 5.7.1 shows the measured head loss versus approach velocity with a curve fit that includes both the viscous and inertial terms. The curve fit shows that the head loss is dominated by the inertial term and the coefficient for the viscous term is a negative value, which is not physically possible. Therefore, the data was re-fit with the viscous term set to zero as shown in Figure 5.7.2.

Page 189 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Table 5.7.1 - Clean strainer head loss evaluation Test Module Approach Temperature Density Viscosity Measured 16 Flow Rate (gpm) Velocity (ft/s) (*F) (lbm/ft 3 ) (Ibf's/ft2 ) Head Loss (ft) 176.58 0.0043025 117.2 61.755 1.196E-05 0.02591 265.15 0.0064606 116.6 61.764 1.202E-05 0.05073 353.05 0.0086023 116.3 61.769 1.206E-05 0.09231 441.30 0.0107526 116.1 61.772 1.208E-05 0.14424 530.13 0.0129170 115.9 61.775 1.211E-05 0.21946 STP CSHIL 0.25 - ,

y= M1"x+M2"XA2 Value Error M1 -0.87705 0.72598 0.2 . . m2 1365 ..........

67 109 ..................................

Chisq 4.7567e-5 NA R2 0.99858 NA o* 0.15 0.1 CD 0.0 0 .05 ...........

0 1 _ L_ _ _

0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 Approach Velocity (fis)

Figure 5.7.1 - Clean strainer head loss curve fit with both viscous and inertial terms 16 The strainer approach velocity was calculated using the test module surface area of 91.44 ft 2 (17).

Page 190 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1 STP CSHL 0.25 1 1 1 r y M2*"*2 M

Value Error 0 .2 . .. hm2s 2 - 161317A ....

. 91286.1 .... .............

T ....................

Cic 6.I4922e-5 NA 2 I NA 1- .........................

o* 0 .15 . .....................

CU (D

S0 .05 . ...... ........ ...... ..... ... ....... ............. ......

. ....T".............

0 04002 0.004 0.006 0.008 0.01 0.012 0.014 Approach Velocity (rs)

Figure 5.7.2 - Clean strainer head loss curve fit with only inertial term The inertial coefficient, M2, in Figure 5.7.2 is equivalent to the inertial coefficient, b, times the fluid density, p, in Equation 42. Given an average density of 61.767 lbm/ft 3 for the test conditions and an M2 value of 1,286.1 s 2/ft, b is equal to 20.82 ft 2 "s2/Ibm. Therefore, the CSHL for STP conditions can be described using the following equation:

82 2 AHCS = 2 0. Lbmf p -U Equation 43 5.7.2 Conventional Debris Head Loss Model The NUREG/CR-6224 correlation was selected for the CASA computation of conventional debris head loss 17 across the strainer. This correlation is a semi-theoretical head loss model and is described in detail in Appendix B of NUREG/CR-6224 (75). The correlation is based on theoretical and experimental research for head loss across a variety of porous and fibrous media carried out since the 1940s. The 17 The term "conventional debris head loss" is used to distinguish between the debris bed head loss caused by typical fiber and particulate debris vs. the head loss caused by chemical effects.

Page 191 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 NUREG/CR-6224 head loss correlation was developed in support of the NRC evaluation of the strainer clogging issue in BWRs and has been extensively validated for a variety of flow conditions, water temperatures, experimental facilities, types and quantities of fibrous insulation debris, and types and quantities of particulate matter debris. The types of fibrous insulation material tested include Nukon, Temp-Mat, and mineral wool. The particulate matter debris tested includes iron oxide particles from 1 to 300 lam in characteristic size, inorganic zinc, and paint chips. In all of these cases, the NUREG/CR-6224 head loss correlation has bounded the experimental results. Due to the semi-empirical nature of the correlation STP performed confirmatory head loss tests to demonstrate the applicability of the correlation to STP conditions (24).

NUREG/CR-6224 Head Loss Correlation The NUREG/CR-6224 head loss correlation, applicable for laminar, turbulent, and mixed flow regimes through mixed debris beds (i.e., debris beds composed of fibrous and particulate matter) is given by Equation 44:

AH = A [3.5S, *am,'S(1 + 57am 3)gU + 0.65S, pUI ALm Equation 44 1i-am -

where:

AH = Head loss S= Specific surface to volume ratio of the debris t= Dynamic viscosity of water U = Fluid approach velocity p = Density of water am = Mixed debris bed solidity (one minus the porosity)

ALm = Actual mixed debris bed thickness A = Conversion factor:

A = I for SI units A = 4.1528x10s (ft-water/in)/(Ibm/ft2 -s2 ) for English units The fluid approach velocity, U, is given simply in terms of the volumetric flow rate and the effective surface area:

_ = -- Equation 45 A

where:

Page 192 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Q = Total volumetric flow rate through the screen A = Screen surface area The screen surface area (A) is the submerged (wetted) surface area of the screen. The available surface area may change with time, particularly in the case of the STP strainer design. As more debris reaches the strainer the surface area may eventually evolve to the circumscribed area as the debris starts to fill up the interstitial volume. If the debris load is sufficient to fill the entire interstitial volume, the head loss for the STP strainer is calculated using the circumscribed area with a debris load equal to the total debris load transported to the strainer less the quantity of debris required to fill in the interstitial volume of the strainer.

The mixed debris bed solidity (amn) is given by:

am = + f7 aoc Equation 46 where:

ao = Solidity of the original fiber blanket (i.e. the "as fabricated" solidity)

-= Particulate to fiber mass ratio in the debris bed (mp/mf) pf = Fiber density pp= Average particulate material density c = Head-loss-induced volumetric compression of the debris For debris deposition on a flat surface of a constant size, the compression (c) relates the actual debris bed thickness (ALm) and the theoretical fibrous debris bed thickness (ALo) via the relation:

c = co ALO Equation 47 Compression of the fibrous bed due to the pressure gradient across the bed is also taken into consideration. The relation that accounts for this effect, which must be satisfied in parallel to the previous equation for the head loss, is given by the following equation valid for (AH/ALo) > 0.5 ft-water/inch-insulation:

c = 1.3K Equation 48 (A0 Page 193 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Here, 'K' is a constant that depends on the insulation type. It is 1.0 for Nukon fiber. It should be noted that this formulation for debris bed compression may over predict compression significantly in the case of very thick debris layers (roughly 6-inches or more). Thus, in these cases, it is conservative.

For very large pressure gradients, the compression has to be limited such that a maximum solidity is not exceeded. In NUREG/CR-6224, this maximum solidity is defined to be:

65 Ibm/ft 3 am - Equation 49 Pp This is equivalent to having a debris layer with a density of 65 Ibm/ft3 . Note that 65 Ibm/ft 3 is the macroscopic, or bulk density of a granular media such as sand or gravel and clay.

Each debris constituent has a surface-to-volume ratio based on the characteristic shape of that debris type. For typical debris types, we have:

Cylindrically-shaped debris: Sv = 4/diam Spherically-shaped debris: S, = 6/diam Flakes (flat-plates): Sv = 2/thick where:

'diam' = Diameter of the fiber or spherical particle, and

'thick' = Thickness of the flake/chip.

The average surface to volume ratio for several debris constituents is calculated as follows:

XCn 2 :v)n Equation 50 where the subscript 'n' refers to the nth constituent, and vn is the volume of each constituent.

Debris Parameters Required for Head Loss Calculations The NUREG/CR-6224 head loss correlation requires the following debris parameters:

" Microscopic density, also referred to as "material" density

Macroscopic density, also referred to as "bulk" density

Characteristic size, which is the dimension to be used in computing the surface to volume ratio (i.e. diameter for fibers and particulates, and thickness for chips)

Page 194 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Based on the debris characteristics provided in Section 2.2.17, the inputs shown in Table 5.7.2 and Table 5.7.3 were used for the head loss calculations in CASA Grande.

Table 5.7.2 - Head loss characteristics for fibrous debris Debris Type Size Geometry Size Sv Microscopic Macroscopic (m 2 /m 3 ) Density Density (lbm/ft3) (lbm/ft 3)

Fines cylinder 7 microns 571,429 175 2.4 LDFG Small Pieces18 square < 6 inches 14.4 175 2.4 Large Pieces' 8 square > 6 inches 7.2 175 2.4 Microtherm Fiber Fines cylinder 6 microns 666,667 165 15 Latent Fiber Fines cylinder 7 microns 571,429 175 2.4 Table 5.7.3 - Head loss characteristics for non-fibrous debris Debris Type Size Geometry Size Sv Microscopic Macroscopic (m 2 /m 3) Density Density (lbm/ft3) (lbm/ft 3)

Microtherm TiO2 Fines sphere 20 microns 300,000 137 27.4019 Microtherm SiO2 Fines sphere 2.5 microns 2,400,000 262 52.40'9 Qualified Epoxy Fines sphere 10 microns 600,000 94 36.6620 Qualified IOZ Fines sphere 10 microns 600,000 208 81.1220 Crud Fines sphere 15 microns 400,000 350 70.0019 Fines sphere 6 microns 666,667 124 48.3620 Fine Chips chip 15 mil thick 480 124 48.3620 Unqualified Epoxy Small Chips chip 15 mil thick 480 124 48.3620 Large Chips chip 15 mil thick 480 124 48.3620 Curled Chips chip 15 mil thick 480 124 48.3620 Unqualified Alkyd Fines sphere 10 microns 600,000 207 80.7320 Unqualified Enamel Fines sphere 10 microns 600,000 93 36.27 20 Unqualified IOZ Fines sphere 10 microns 600,000 244 95.16.2 Latent Dirt/Dust Fines sphere 17.3 microns 346,821 169 33.8019 Applicability of the NUREG/CR-6224 Head Loss Correlation to STP Conditions The NUREG/CR-6224 head loss correlation has been validated over a large range of approach velocities and debris types. However, there were specific STP conditions where the NUREG/CR-6224 head loss is LDFG smalls Sv modeled as a 0.5 inch thick cubes; LDFG large Sv modeled as a 1 inch thick cubes 19 Calculated based on a packing fraction of 0.20 for iron oxide sludge (72).

20 Calculated based on packing fraction of 0.39 for acrylic coatings debris (24).

Page 195 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 correlation had not been compared to experimental data. In particular, experimental data did not exist to evaluate the impact on the NUREG/CR-6224 head loss correlation for the following conditions:

" Low approach velocities prototypical of the STP strainers - most of the data used to develop the NUREG/CR-6224 head loss correlation was based on tests at higher approach velocities characteristic of the small conical strainers installed in the BWRs before 1992.

Buffered borated demineralized water- most of the data used to develop the NUREG/CR-6224 head loss correlation was based on tests with tap water. There were some studies done recently that suggested that water chemistry has a significant impact on head loss (76).

" Temperature - most of the data used to develop the NUREG/CR-6224 head loss correlation was based on tests at room temperature.

NEI fiber debris preparation - most of the data used to develop the NUREG/CR-6224 head loss correlation was based on tests conducted with mechanically shredded fiber debris prior to the development of the NEI debris preparation protocol.

In order to ascertain the applicability of the NUREG/CR-6224 head loss correlation to STP specific conditions, a series of vertical head loss tests were performed (24). The experiments were conducted at STP conditions including the strainer flow approach velocity of 0.0086 ft/s or less, STP-specific water chemistry, a range of temperatures prototypical of the post-LOCA conditions, and STP-specific debris loads.

The application of a head loss correlation to head loss data requires the measurements of head loss, water temperature, and flow velocity for a relatively uniform and homogeneous fibrous/particulate debris bed of known composition at relatively stable conditions. Turbidity measurements, as well as water clarity, are used to judge the completeness of the filtration process.

The correlation validation process depends on knowing the input hydraulic characteristics of each type and size category of debris introduced into the test. Debris size characterization can be used to approximate the hydraulic characteristics of simple forms of debris, such as Nukon fibers, but not for complex particulates. A typical particulate consists of roughly shaped particles of varied sizes making the analytical assessment of the specific surface area, S,, somewhat difficult and uncertain. Some insulation materials such as calcium silicate, Microtherm, Min-K, and amorphous chemical precipitates have complex forms that simply cannot be assessed analytically, and their impact on head loss has to be addressed experimentally. The solid density of a particle is based on the material properties and the particulate bulk density can be deduced by weighing a known volume of the particulate. The S, value is deduced by applying a head loss correlation to head loss test data where all parameters are known except the S, value for the material in question. As such, inaccuracies in the form of the correlation become inherent in the experimentally deduced input parameters. Therefore, the correlation and the hydraulic characteristics become somewhat interdependent.

Page 196 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 A total of eleven exploratory head loss tests were performed (24). All testing was done using fibers from a single-side baked Nukon blanket, which was processed using the NEI debris preparation process. All testing was conducted starting at 200 'F at the STP buffered and borated water conditions. The particulate types tested were green silicon carbide, iron oxide (the BWR sludge simulant used in the development of the NUREG/CR-6224 head loss correlation), tin, and ground acrylic paint. Flow and temperature sweeps were performed at the end of some of the experiments to examine the impact of different flow conditions and temperatures.

The NUREG/CR-6224 head loss correlation was used to replicate the measured head loss of the test conducted with iron oxide and a debris bed thickness similar to the test parameters used in the development of the NUREG/CR-6224 head loss correlation (24). The iron oxide S, value was adjusted until the calculated head loss matched the measured head loss. The final S, value was in reasonable agreement with the specifications of the size distribution of the sludge simulant indicating that the NUREG/CR-6224 head loss correlation was a reasonable predictor of head losses at STP water and temperature conditions. The iron oxide test, however, was limited to the lowest approach velocity of 0.02 ft/s due to equipment limitations. The NUREG/CR-6224 head loss correlation also generated reasonable estimates of the head loss experiments conducted with ground acrylic paint and extended the approach velocity down to the STP strainer approach velocity of 0.0086 ft/s.

The NUREG/CR-6224 head loss correlation, however, could not replicate the low head losses observed in the tests with tin and/or green silicon carbide. The test report provides a hypothesis for this behavior based on observations of the difference in smooth surfaces noted on SEMs of green silicon carbide and tin as compared to the rough surfaces of iron oxide and ground acrylic paint (24). Further experiments need to be conducted to confirm this hypothesis. This lack of agreement between the NUREG/CR-6224 head loss correlation and testing with green silicon carbide and tin does not impact the STP head loss calculations since there is no green silicon carbide or tin in the STP debris mixture. The green silicon carbide has been used in the past as a simulant of paint, and the tin has been used as a simulant of IOZ coatings. Most of the STP particulate debris comes from coatings, either from qualified coatings in the ZOI or from unqualified coatings elsewhere.

Another anomaly observed in the STP head loss tests was the absence of a direct correlation of the head losses observed in the temperature sweeps with the water viscosity. The test report provides a hypothesis that the temperature also impacts the compression of the fiber debris bed due to the temperature impact on the malleability of the fibers (24). An analytical model was developed to couple the compression to temperature that showed good agreement with the experimentally determined temperature sweep data. The compression algorithm implemented in the NUREG/CR-6224 head loss correlation used in CASA was not modified to incorporate the temperature dependence suggested by the tests. The experiments showed that the measured head losses at lower temperature were lower than the head losses calculated by the NUREG/CR-6224 head loss correlation, hence the CASA calculated Page 197 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 head losses are conservative. Additional experiments and analysis need to be performed to validate the temperature dependent compression algorithm prior to its implementation in CASA.

One of the tests conducted (Test 8) was designed to replicate the August 2008 ARL STP prototype test (24; 52). However, this test completely failed to replicate the head losses observed in the previous testing. Both tests used the same primary surrogates of Nukon fibers along with tin and acrylic particulates. Three differences in the tests are: 1) Test 8 had a greater thickness of fiber than was reported in the ARL test, 2) Test 8 used Alion supplied Microtherm and Marinate board particulate instead of the same materials used at ARL, and 3) the ARL fiber debris preparation protocol used a food processor whereas Test 8 used the NEI debris preparation protocol. Based on the experience of the CHLE tests (17), fiber beds with food processor prepared fiber tended to exhibit higher head losses than fiber beds prepared in accordance with the NEI debris preparation protocol. Comparisons of the beds prepared with food processor prepared debris and the NEI debris protocol revealed that the NEI protocol fibers tended to bridge the perforated plate holes and form a debris bed over the perforated plate, while the food processor fibers tended to form low porosity "dimples" at the perforated plate holes. The higher head losses observed with food processor beds was attributed to the formation of the low porosity "dimples". The food processor prepared fibers used in the ARL test could have also formed low porosity "dimples", and allowed the particulate to pack tighter in the ARL test than in Test 8 resulting in a lower porosity bed with higher head losses. The formation of "dimples" in the strainer holes instead of a fiber bed over the perforated plate could also explain the very thin bed observed in the ARL test. The lack of reproducibility of the head losses observed in the Alion vertical loop test compared with the ARL test does not impact the applicability of the NUREG/CR-6224 in calculating the CASA head losses since the differences in the results are attributable to different debris preparation methods. The NUREG/CR-6224 head loss correlation assumes the formation of a debris bed over a perforated plate as was observed with the debris beds prepared in accordance with the NEI debris preparation protocol. Therefore, the NUREG/CR-6224 head loss correlation is considered to be applicable to the debris beds formed with STP prototypical debris.

The test report also addresses the impact of the three main ACRS comments of the NUREG/CR-6224 head loss correlation (24). These ACRS comments were mainly directed at debris beds containing calcium silicate, a known problematic insulation. The test report provides suggested modifications to the NUREG/CR-6224 head correlation to address the three main ACRS concerns (24). Note that all Marinite (similar to calcium silicate) has been removed from containment at STP. Therefore, as shown in the test report, the three main ACRS comments are not significant for STP conditions (24).

Overall, these tests demonstrated that the NUREG/CR-6224 head loss correlation provided reasonable predictions of head loss for the prototypical STP debris types and loads, water chemistry, temperature, and strainer approach velocities. However, due to the generic concerns regarding the NUREG/CR-6224 correlation, the head loss calculated using the correlation was increased by a factor of five in CASA Grande to account for uncertainties in the head loss predictions.

Page 198 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 5.7.3 Chemical Debris Head Loss Model As discussed in Section 5.5, several aspects of the chemical effects evaluation model were not fully developed within this version of the analysis. Therefore, the specific conditions associated with each break scenario (pool volume, pool temperature, debris quantities, etc.) could not be explicitly linked to a corresponding chemical head loss. However, a range of conditions were evaluated using the WCAP-16530-NP calculator and estimated solubility limits for expected product formation to determine a relative comparison of the quantity of precipitants for various break scenarios (20).

For nominal temperature profiles, chemical products (aluminum and calcium precipitates) were not predicted to form for any of the small breaks evaluated. However, some of the medium and large break cases evaluated had total aluminum concentrations that were approximately equal to or slightly higher than the estimated solubility limits (20). The calcium concentration was relatively high for cases where a maximum fiberglass quantity of 2,385 ft 3 was assumed. However, for cases with 60 ft3 of fiber or less, the calcium concentration was approximately equal to the solubility limit (20). As discussed in Section 5.4.3, the quantity of fiberglass insulation debris generated is less than 10 ft 3 for 99.9% of the scenarios evaluated in CASA Grande. This indicates that even if chemical products form for the nominal scenarios, the effects on strainer head loss would be relatively benign. An evaluation of the chemical concentrations for a maximum temperature profile, however, indicated that the concentration of aluminum would be significantly higher (on the order of 20 times greater than the nominal scenarios). It is possible that these scenarios could result in significant chemical head loss. However, the maximum temperature profiles were developed based on a highly unlikely scenario where the CCW temperature is at the maximum level, four out of six fan coolers fail to operate, and all of the RHR heat exchangers fail (5). Extreme temperature profiles like this have not been fully evaluated yet, so the current limited testing does not completely preclude the possibility that chemical products may form and arrive at a debris-laden strainer in sufficient quantity to cause unacceptable head loss.

To account for the presence of extreme conditions in the scenario sample space, exponential probability distributions were defined and applied as direct multipliers to the estimated conventional head loss. The probability distributions were developed based on the current results from the CHLE testing (18; 19),

WCAP-16530-NP calculations (20), and reasonable engineering judgment. The chemical effects model that was implemented in CASA Grande is described below:

" No bump-up factor is applied if the fiber quantity on a given strainer is less than 1/16 of an inch (see Assumption 7.d).

No bump-up factor is applied prior to the temperature dropping below 140 °F (see Assumption 5.a). Note that since only two temperature profiles were implemented in CASA Grande (see Section 2.2.7), the increase in head loss would occur approximately 5 hr after the start of the Page 199 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 event for large breaks, and approximately 16 hr after the start of the event for small and medium breaks.

As shown in Table 5.7.4 and Figure 5.7.3 through Figure 5.7.5, the probability distributions for the chemical effects bump-up factors were developed with mean bump-up factors of approximately 2x for small breaks, 3x for medium breaks, and 3x for large breaks, and maximum bump-up factors of approximately 15x for small breaks, 18x for medium breaks, and 24x for large breaks.

The exponential probability density function is defined by a single parameter, the mean, and is continuous on the interval from zero to infinity. The chemical effects bump-up factor should never be less than one, and there is a practical maximum above which all events will lead to sump failure, so the following strategy was adopted. Samples of chemical factor are taken from exponential probability density functions defined using the "formal" parameters given in Table 5.7.4. Manual iteration in a side calculation is used to determine a formal maximum endpoint for each formal mean above which the cumulative tail probability is approximately 1E-5. Thus, the maximum chemical effects bump-up factor is always assigned a weight of 1E-5, and the maxima are always included in the sample design (exactly the same approach used for DEGB conditions). Sampling is performed on a logarithmic scale with an emphasis on large values. This means that a much higher proportion of samples are taken from the high end of the range, but each individual sample has a small probability contribution. Finally, all samples from the formal exponential probability density functions are shifted by one unit to guarantee that the applied factors are never less than one.

Shifting all samples by a unit of one has the somewhat unintended consequence of inflating the potential effect of chemical products more than desired. While the desired means are reported as "formal" parameters, the effective means applied in the quantification are actually closer to the "shifted" values given in the table. A more careful normalization of a truncated exponential probability density function will be used in the future to preserve both the desired mean and the desired tail probability.

Table 5.7.4 - Exponential probability distribution parameters applied to chemical effects bump-up factors for each LOCA category Parameters SBLOCA MBLOCA LBLOCA Tail Probability Min 0 0 0 ~1e-5 Formal Mean 1.25 1.5 2.0 ~le-5 Max 14.3 17.2 23 ~1e-5 Min 1 1 1 ~1e-5 Shifted Mean 2.25 2.5 3.0 ~1e-5 Max 15.3 18.2 24 ~le-5 Page 200 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1 Chemical Head-Loss Factor for SBLOCA 0.8 0-07 0ý6 I

c*

0-5 r,

CL o0.3 02 0_1 !

chemical effect factor Figure 5.7.3 - Exponential probability density function for chemical effects bump-up factors applied to SBLOCAs Page 201 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1291-V03 Revision 1 Chemical Head-Loss Factor for MBLOCA 07 0.

0.

On 0.

0 a-I O0 0-1 chemical effect factor Figure 5.7.4 - Exponential probability density function for chemical effects bump-up factors applied to MBLOCAs Page 202 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1 Chemical Head-Loss Factor for LBLOCA C

a.

10 15 25 chemical effect factor Figure 5.7.5 - Exponential probability density function for chemical effects bump-up factors applied to LBLOCAs 5.7.4 Strainer Head Loss The overall strainer head loss includes a combination of the clean strainer, debris bed, and chemical head losses as shown in the following equation:

AHs = AHcs + AHDB - BCE Equation 51 where:

AHs = Total strainer head loss AHcs = Clean strainer head loss AHDB = Conventional debris bed head loss BCE = Bump-up factor for chemical effects Figure 5.7.6 shows an example of the time-dependent head loss for random cases evaluated in CASA.

Note that the head loss spikes up at approximately 5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months when the temperature drops below 140 °F Page 203 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 and chemical precipitation is assumed to occur, and then spikes back down at approximately 6.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months when the containment sprays are secured.

Random Sample of AiP History 35 30 25 20 o1515 52 0

0 500 1000 1500 2000 2500 time (min)

Figure 5.7.6 - Typical sample of sump-strainer head loss histories generated under the assumption of exponential chemical effects factor and artificial head-loss inflation 5.7.5 Acceptance Criterion: NPSH Margin Module The pump NPSH margin is the difference between the NPSH available and the NPSH required, as shown in Figure 5.7.7 and Equation 52 through Equation 54. Note that the NPSH margin does not include the clean strainer or debris bed head losses. Therefore, the strainer head losses are compared to the NPSH margin to determine whether or not pump cavitation will occur due to loss of NPSH.

Page 204 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Figure 5.7.7 - Illustration of parameters that affect pump NPSH NPSHM = NPSHA - NPSHR Equation 52 NPSHA = + Helev Hppn ap P ngPg Equation 53 NPSHR(1_ý2 %) = NPSHR(aý=o%) (1 + O.a; ) Equation 54 where:

NPSHM = NPSH margin NPSHA = NPSH available NPSHR = NPSH required Pcont = Containment pressure p = Water density Page 205 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 g = Gravitational acceleration Helev = Head of water from the pump to the surface of the pool Hpiping = Head losses between the strainer and the pump (not including strainer losses)

Pvap = Vapor pressure ap' = Volumetric percentage of air in the fluid at the pump impeller (ap = 1O0.ap) ap = Void fraction of air in the fluid at the pump impeller As discussed in Assumption 1.c, no credit was taken for containment overpressure. The pressure was assumed to be 14.7 psia, except for cases where the containment pool temperature is greater than 212

°F, where the containment pressure was assumed to be equal to the vapor pressure. The water density and vapor pressure are determined as a function of the containment pool temperature based on standard water properties.

The head of water above the pumps is the sum of the water level above the containment floor and the elevation of the containment floor above the pumps as shown in the equation below. The water level is determined as discussed in Section 2.2.6. The elevation of the pump impellers below the containment floor is 25.83 ft for the LHSI and CS pumps, and 25.65 ft for the HHSI pumps (25).

Helev = HPOOt + Hpump Equation 55 where:

Hnoo1 = Water level above the containment floor Hpump = Elevation of pumps below the containment floor The piping flow losses include both major and minor losses, which are a function of cumulative and individual pump flow rates for each train as well as the pool temperature and piping geometry. A schematic of the ECCS suction piping geometry at STP is shown in Figure 5.7.8. The piping flow losses can be calculated using Equation 56 through Equation 58 (25).

Page 206 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 A sump C

LHSI HHSI G

Cs B

D F

Figure 5.7.8 - Schematic of STP ECCS sump suction piping 2 2

$2 2 S

+ 0. 5 8 2

HpipingQLHSI "" 2.06- fAB + 5o.005oo - . fBc (QLHs, + QHHSI + QCs)'

ft c ft ft Equation 56

+ 2.9 7 Tt5'fBC (QLHSI) 2 2 2 S S S2 +

Hpiping,HHSI = .06T--I fAB + 0.005 T-s' + 0.19 -Tt"fBD (QLHS, + QHHSI + Qcs)2 ft5 ft 5 ft Equation 57 9

+ Q.O 7-t-5"fB + o.SsT-2f) Qcs) + ft

(OQHHsI) 2 Page 207 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Hpiping'cs 2.06 $25 fA, + 0.005T-S 2 + 0.1 9 -"IB S25 "(QLsi + QHHSJ + Qcs) ft A QHI+QHS C ppft S2 S2 v (Hs1 + Qcs)2

+ 0.09 -*-"f'D + 0.19 7T- 'fDF) (QHHSI+Q Equation 58

+ 0.09-.5'AF + 0.58y-"fFG + 2.95 -y.f) * (Qcs) 2 where:

Hpiping,xx = Flow losses in piping for the LHSI, HHSI, and CS pumps respectively fx, = Friction factor for various pipe segments illustrated in Figure 5.7.8

%x = Flow rate for LHSI, HHSI, and CS pumps respectively The friction factor is dependent on the Reynolds number, and can be determined using the following equations (25).

P" (QLHsI + QHHSI + Qcs) < 2000 p " lf t then, fAB=-- 64 ReAB if, ReAB =0 (QLSI + QHHSI + QCS) > 2000 then, P. lft Equation 59 A-2---2"og 3.19 X10-s + 2.5 If t P" (QLHsI + QHHSI + Qcs)

~(

64 if, ReBc = P (QLHSI) < 2000 then, fBC-- ReBc

[t" 0.79ft -

if, ReBC = P (QLHSD > 2000 then, p" 0.79ft Equation 60 log 4.08 X 10-+ + 1.96ft 1 =--2 _ _ _.6f~

P" (QHHsI + Qs) 64 if, ReBD = _ 2000 then, fBD =-

p lft ReBD Equation 61 Page 208 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 if, ReBD = P> (QHHSj + > 2000 then,

p. lf t 1

-2" log (2.51ft 3.19 X 10-s + p(Q P'(HS 2+5Qs)

C)'

/*D B

2000 then,

<(QHS* 64 if, ReDE = fDE = -

p" 0.65ft if, ReDE -- > 2000 then,

p. 0.65ft Equation 62 JD(

= --2 log 4.86 x 10-' +

1.66f t P '(QHHSI .

It JE 64 if, ReDF -- - (QCS) <2000 then, fDF = ReDF

1. lft -

if, ReDF - p (Qcs) > 2000 then,

p. ift Equation 63

__ 2- log 3.19 +

X 10-5 2.5ift x i0 (QCS)

-+p 11 ' DI Re=p" (Qcs) 64 if, ReFG -p -[i" 0.t<

0.79f t -- 2000 then, fFG 64 ReFG P"-(Qcs) if, ReFG - p- 0.79fttQs)> 2000 then, Equation 64 JF( -2 log 4.08 x 10- +

1.96ft pag(QCS) o It 260 Page 209 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 where:

Rex. = Reynolds number for various pipe segments illustrated in Figure 5.7.8 p = Water density as a function of temperature, Ibm/ft3 p = Water viscosity as a function of temperature, Ibm/ft-sec f.. = Friction factor for various pipe segments illustrated in Figure 5.7.8

%x,= Flow rate for LHSI, HHSI, and CS pumps respectively The NPSH required is a fixed value dependent on the pump specifications. However, if gas voids are present (see Section 5.8), the NPSH required must be adjusted as discussed in Regulatory Guide 1.82 (58).

For each scenario, the time-dependent strainer head loss was compared to the time-dependent NPSH margin to determine whether any failures occur. As discussed in Assumption 12.a, the failure of one pump in any train was assumed to be equivalent to the failure of all pumps in all trains.

5.7.6 Acceptance Criterion: Structural Margin The strainer structural margin is 9.35 ft (see Section 2.2.29). If the strainer head losses exceed the structural margin, the strainer may fail allowing large quantities of debris to be ingested. As discussed in Assumption 12.b, the structural failure of one strainer was assumed to lead to complete ECCS failure.

5.8 Air Intrusion The presence of air or other gasses in the ECCS, CSS, or other systems can result in the failure of those systems to perform their intended safety functions. Gas intrusion and accumulation issues have been evaluated in response to Generic Letter 2008-01 (GL 08-01), which identifies concerns with gas upstream of pumps causing potential pump failure, gas downstream of pumps causing water hammer effects when the pump is started, and other potential issues (77). Some of these issues are directly related to GSI-191, since it is possible for air to enter the ECCS and CSS through vortexing or degasification at the strainers during recirculation.

5.8.1 Vortex Formation Vortex formation can appear to be an almost random variable since it is strongly influenced by minor variations in the local flow conditions. However, as discussed in a series of NUREGs (78; 79; 80; 81; 82),

vortex formation is somewhat related to the Froude number:

V Fr = Vry* Equation 65 Page 210 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 where:

Fr = Dimensionless Froude number v = Velocity (in the suction pipe) g = Acceleration of gravity I = Characteristic length (submergence depth of the suction pipe)

In general, vortexing is dependent on the strainer flow rate, the submergence depth, the strainer geometry, and to some extent the containment geometry (which could either induce or inhibit swirling as the flow approaches the strainer). Vortexing can be easily prevented with simple structures that disrupt swirling motion in the flow.

ECCS strainer vortexing has been evaluated at STP, and based on the strainer design, it has been determined that vortexing would not occur under even under bounding conditions (55). Therefore, there would be no air ingestion due to vortexing.

5.8.2 Degasification Under a given set of conditions (temperature, pressure, humidity, etc.), a certain quantity of air can be dissolved in water. If these conditions change, some of the dissolved air may be released from the water. In a LOCA scenario, some air would be dissolved in the containment pool, and as the water passes through the ECCS strainer, the head loss across the strainer would cause some of the air to be released.

The following generic properties of air and water are necessary for calculating degasification:

" The composition of air is approximately 78.08% nitrogen (N2), 20.95% oxygen (02), 0.93% argon (Ar), and 0.04% carbon dioxide (C0 2) with trace amounts of other gasses (83).

" The critical temperature of water is 647.140K (84).

The molecular weight of water is 18.01528, the molecular weight of nitrogen is 28.01348, the molecular weight of oxygen is 31.9988, the molecular weight of Argon is 39.948, and the molecular weight of carbon dioxide is 44.010 (84). The overall molecular weight of air is approximately 28.97.

The quantity of air released from a given volume of water across an ECCS strainer can be determined by subtracting the concentration of air dissolved in water in the containment pool by the concentration of air dissolved in water downstream of the strainer. The concentration of air is calculated using Henry's Law:

Page 211 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 CG = KG(T)- PG Equation 66 where:

CG = Saturation concentration of air KG = Henry's constant for air at a given temperature T = Temperature PG = Partial pressure of air Henry's Constant for Air-Water Solutions Henry's constant for air (KG) can be determined based on the individual Henry's constant for each component of air (N 2 , 02, Ar, and C0 2 ). The volatility constant for each of these components can be calculated using the following semi-empirical correlation (85):

Ac Bc". (1l-T)'s ln(kc) = In(PsAT) + Lc + 0 +T*cc e(1-T*) * (T*)-0.41 Equation 67 where:

k, = Volatility constant in units of pressure PSAT = Saturation pressure at the given temperature Ac, Bc, Cc = Constants provided in Table 5.8.1 T* = T/Tc where T is the temperature and Tc is the critical temperature of water (OK)

Table 5.8.1 - Semi-empirical correlation parameters to calculate Henry's constants in aqueous solvent (85)

Maximum T Solute Ac Bc c(K)

Nitrogen -11.6184 4.9266 13.3445 636.5 Oxygen -9.4025 4.4923 11.3387 616.48 Argon -7.4316 4.2239 9.6803 568.4 Carbon Dioxide -9.4234 4.0087 10.3199 631.7 The relationship between the volatility constant and the Henry's solubility constant is shown in Equation 68.

Page 212 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSl191-V03 Revision 1 Kc= Mck PH2 0 K=Mc*kc *MH,0 Equation 68 where:

Kc = Henry's solubility constant for gas component k, = Volatility constant for gas component PH2o = Density of water Mc = Molecular weight of gas component.

MH20 = Molecular weight of water The overall solubility constant for air can be calculated using the individual solubility constants as shown in Equation 69.

KAir = KN2 FN2 + KO2 Fo2 + KAr AF+ KcoI " FCo2 Equation 69 where:

K= Henry's solubility constant for each gas component F = Mole fraction of each gas component Concentration of Air in Containment Pool The partial pressure of air in the containment atmosphere can be calculated as shown in Equation 70 using the containment pressure (P0 ) and the vapor pressure (Pv,0). Note that the subscript 0 is used to designate conditions upstream of the ECCS strainer.

PG,O = PO- PV,O Equation 70 The vapor pressure can be calculated based on the saturation pressure (PsAT) at the pool temperature, and the relative humidity in containment (ýO) as shown in Equation 71.

Pvo = 00 PSAT(TO) Equation 71 Combining Equation 71 into Equation 70 and Equation 70 into Equation 66 yields the following:

CG,o = KG(TO) [Po -0 PSATr(TO)] Equation 72 where:

Page 213 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 CG,0 = Saturation concentration of air in the containment pool KG = Henry's constant for air at the pool temperature To = Temperature of the containment pool P0 = Containment pressure

ýO = Relative humidity in containment PSAT = Saturation pressure at the pool temperature Concentration of Air Downstream of ECCS Strainer The pressure downstream of the ECCS strainer can be calculated using the containment pressure (P0 ),

the hydrostatic head of water above the strainer, and the pressure loss across the strainer (APLoss) as shown in Equation 73. Note that the subscript 1 is used to designate conditions downstream of the strainer.

P1 = PO + PL (TO) "g " HL - APLOSS Equation 73 Similar to the containment pool calculation, the partial pressure of air and the vapor pressure downstream of the ECCS strainer can be calculated using Equation 74 and Equation 75. Note that the temperature downstream of the strainer is assumed to be the same as the temperature in the containment pool.

PG,1 = P 1 - Pv,1 Equation 74 Pv, = 01"PSAT(Tl) = 01"PSAT(TO) Equation 75 Combining Equation 73 and Equation 75 into Equation 74 and Equation 74 into Equation 66 yields the following:

CGJ, = KG(To) ' [Po + pL(To) g " HL - APLOSS - 03" PSAT(To)] Equation 76 where:

CG1 = Saturation concentration of air downstream of the strainer KG = Henry's constant for air at the pool temperature To = Temperature of the containment pool P0 = Containment pressure PL = Water density at the pool temperature g = Gravity HL = Pool height above the strainer APLoss = Pressure drop across the strainer C = Relative humidity downstream of the strainer Page 214 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 PSAT= Saturation pressure at the pool temperature Quantity of Gas Released After determining the concentration of air in solution before and after the strainer, the gas released can be simply calculated as shown in Equation 77.

ACG :- CG, 0 - C0,1 Equation 77 Note that the concentration of air released is in units of mass of air per unit volume of water. Therefore, the mass rate (AmG) that air is released from the water can be calculated by multiplying the concentration of gas released by the flow rate through the strainer (0,) as shown in Equation 78.

AMG = ACG'QL Equation 78 The ideal gas law can then be used to convert the mass of gas released to a volume.

AmG R To QG- M PG,1 Equation 79 where:

OG = Volumetric flow rate of air released AmG = Mass flow rate of air released M = Molecular weight of air R = Ideal gas constant To= Temperature of the containment pool PG,1 = Partial pressure of air downstream of the strainer The void fraction (as) can be calculated as shown in Equation 80.

QG

-QG + QL Equation 80 It is important to note that this void fraction is the void fraction just downstream of the strainers.

However, the concern is the void fraction at the pump impellers (ap). Since there are three pumps per train, the void fraction would be split between the HHSI, LHSI, and CS pumps proportional to the relative flow split (see Assumption 8.i).

Page 215 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Qx atsx =QHHSI + QLHSI + Qcs Equation 81 where:

Subscript X = HHSI, LHSI, or CS pump asx = Void fraction at strainer proportional to flow split to Pump X Ox = Flow rate to Pump X Since the temperature between the strainer and pumps would be roughly constant, the volume of the gas voids at the pumps can be calculated based on the ideal gas law:

Ps tpx = P tsx Equation 82 where:

apx = Void fraction at Pump X P, = Pressure inside the strainer Ppx = Pressure at Pump X Within this version of the analysis, the void fraction at the pumps was conservatively assumed to be the same as the void fraction downstream of the sump strainers (see Assumption 8.j).

5.8.3 Gas Transport and Accumulation Depending on the strainer, plenum, sump pit, and suction piping geometry, the local flow conditions, and the size of the gas bubbles released due to the strainer head loss, it is possible that the gas bubbles would either transport through the ECCS pumps or accumulate at a high point upstream of the pumps.

Figure 5.8.1 shows an isometric view of one of the ECCS strainers, and Figure 5.8.2 shows a cross-section of the strainer and sump pit. Air bubbles that are released due to degasification would have to transport horizontally or vertically through the stacked disks into the core tube, horizontally through the core tube to the plenum, vertically through the plenum and sump pit to the ECCS suction pipe, and horizontally and vertically through the suction pipe to the pumps.

Page 216 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Figure 5.8.1 - Isometric view of ECCS strainer Figure 5.8.2 - Cross-section view of ECCS strainer and sump pit Page 217 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Bubble transport can be reasonably estimated based on the Froude number. For a horizontal pipe, partial bubble transport will occur when the Froude number is greater than 0.35, and full transport will occur when the Froude number is greater than 0.55 (see Design Input 2.2.31). The Froude number can be calculated using the following equation:

V Fr =-**. Equation 83 where:

Fr = Dimensionless Froude number v = Velocity (in the core tube, plenum, sump pit, or suction pipe) g = Acceleration of gravity I= Characteristic length (hydraulic diameter of the core tube, plenum, sump pit, or suction pipe)

The diameter of the strainer core tube is approximately 0.9 ft (see Design Input 2.2.26). Assuming a maximum sump flow rate of 7,020 gpm (see Design Input 2.2.9) split evenly between the four strainer core tubes, the maximum flow rate to each core tube would be 1,755 gpm. The Froude number within the core tubes (near the strainer plenum) is 1.14 as shown in the following calculation:

Fr = 1,755gp= 1.14 7.48 al/ft3 .60S/min . 7. ( f)O . 3 2 2. fts . O.9ft Equation 84 Since the maximum Froude number is greater than 0.55, it is possible that some air would be transported through the core tubes into the plenum. For vertical bubble transport from the plenum to the suction pipe, partial bubble transport will occur when the Froude number is greater than 0.35, and full transport will occur when the Froude number is greater than 1.0 (see Design Input 2.2.31). The diameter of the suction pipe is approximately 1.3 ft (see Design Input 2.2.26), and the maximum sump flow rate is 7,020 gpm (see Design Input 2.2.9). The maximum Froude number within the suction pipe is 1.82 as shown in the following calculation:

Fr =7,2Ogpm = 1.82 7.48 /t 60S/m

. .( ' . 3 2 2 ft/ 2 . 1.3ft Equation 85 Page 218 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1 The horizontal cross-sectional area of the sump pit is 40 ft2 (see Design Input 2.2.26). Since the hydraulic diameter of the sump pit (approximately 5.7 ft) is significantly larger than the suction pipe, the Froude number within the sump pit is only 0.03 as shown in the following calculation:

Fr = 7,-2Ogpm 0.03 7.48 t3.60S/min* 40ft 2 . 32.2 5.7ft Equation 86 Therefore, if the bubbles transported to the sump suction piping, they would easily transport to the pumps. However, at the prototypical STP flow rates, it is not likely that the bubbles would transport vertically down through the sump pit. For conservatism in the evaluation of potential pump failures due to air ingestion, and the negative effects of gas voids on the NPSH required, it was assumed that any gas voids caused by degasification would be transported to the ECCS pumps (see Assumption 8.h).

If the velocity within the strainer and sump is not high enough to transport the air bubbles, the air would accumulate at high points within the strainer or plenum. There is a small area at the top of the strainer plenum where it is possible for air to collect. It is also possible that air pockets could form at the top of the strainer disks. As shown in Figure 5.8.3, if a large enough gas void forms at the top of the plenum, air would migrate to the strainer disks closest to the plenum. If the buoyancy of the voids in the strainer disks is greater than the pressure drop across the debris bed on the strainer, the gas voids would break through the debris bed and be vented to the containment pool.

Page 219 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Figure 5.8.3 - Illustration of air bubble accumulation and venting 5.8.4 Acceptance Criterion: Pump Gas Void Limits As discussed in Design Input 2.2.32, the acceptance criterion for a steady-state gas void fraction at the pump suction inlet is 2%.

5.9 Debris Penetration Debris penetration is a function of two mechanisms. The first mechanism is direct passage of debris as it arrives on the strainer. A portion of the debris that initially arrives at the strainer will pass through, and the remainder of the debris will be captured by the strainers. The direct passage penetration is inversely proportional to the combined filtration efficiency of the strainer and the initial debris bed that forms.

The second mechanism is shedding, which is the process of debris working its way through an existing bed and passing through the strainer. By definition, the fraction of debris that passes through the strainer by direct penetration will go to zero after the strainer has been fully covered with a fiberglass debris bed. Shedding, however, is a longer term phenomenon since particulate and small fiber debris may continue to work its way through the debris bed for the duration of the event. These processes are illustrated in Figure 5.9.1.

Page 220 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Clean Strainer 0.0 Direct Passage (1-filtration)

'U 0.

W.

a, Loaded Strainer Shedding Time Figure 5.9.1 - Illustration of direct passage and shedding Debris that penetrates the strainer can cause both ex-vessel and in-vessel problems. Ex-vessel effects are addressed in Section 5.10, and in-vessel effects are addressed in Section 5.11 and Section 5.12. The most significant downstream effects concern is related to the quantity of fiberglass debris that accumulates in the core. This is a highly time-dependent process due to the following time-dependent parameters:

Initiation of recirculation with cold leg injection

Switchover to hot leg recirculation

Arrival of debris at the strainer

Accumulation of debris on the strainer

Direct passage

Debris shedding

Flow changes when pumps are secured

Decay heat boil-off The timing for initiation of recirculation, switchover to hot leg injection, and procedurally securing pumps is described in Section 2.2.1. The time-dependent arrival of debris at the strainer is described in Page 221 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1 Section 5.6.8. The decay heat boil-off curve, which defines the flow split to the core for cold leg breaks during cold leg injection, is described in Section 5.11.3. Debris accumulation on the strainer and debris penetration through the strainer (including both direct passage and shedding) are described in more detail within this section.

The various parameters associated with time-dependent debris accumulation on the strainer and core are illustrated in Figure 5.9.2, where Sn(t) is the source rate for initial introduction of debris type n, V(t) is the pool volume, m,(t) is the mass of debris n in the pool, fn(t) is the filtration efficiency for debris n at the strainer, Sn(t) is the shedding rate for debris n from the existing debris bed, Q(t) is the volumetric flow rate passing through the strainers, Y is the fraction of SI flow compared to the total flow, A is the fraction of flow passing through the core compared to the total SI flow, and gn(t) is the filtration efficiency for debris n at the core.

Figure 5.9.2 - Illustration of time-dependent parameters associated with debris accumulation on the strainer and core As illustrated by Figure 5.9.2, debris that passes through the strainer will not necessarily end up on the core. A portion of the debris could pass through the containment spray pumps, and a portion could either bypass or pass directly through the core and spill out the break. The debris that doesn't accumulate in the core may end up back in the pool where it could transport and potentially pass Page 222 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 through the strainer again. The differential rate of change for each debris type in the pool (assuming a homogenous mixing volume) can be described using the following equation (28):

dmn Q Q dt= S. - f. - M. - Y1gn(1 - M - Tnn + Sn - Ygnsn Equation 87 where all of the properties can be time-dependent and have the following definitions:

mn= Mass of debris type n suspended in the pool t = Time fn = Filtration efficiency for debris type n at the strainer Q = Volumetric flow rate passing through strainers V = Total volume of the pool Sn = Source rate for initial introduction of debris type n

s. = Shedding rate for debris type n from existing bed gn = Filtration efficiency for debris type n at the core V= Fraction of the total flow going to the SI pumps X = Fraction of Sl flow going to the core Based on Equation 87, the total quantity of debris that accumulates on the strainer or the core can be described by the following equations (28):

t MS(t) = J [I(t')mtn(t') V(t')---- sn(t')I dt' Equation 88 t

MC(t) = f y(t')A(t')gn(t') (1 -fn(t')) Q-) 7nt ) + Sn(t') dt' Equation 89 where:

Ms = Cumulative mass of debris type n on the strainer Mc = Cumulative mass of debris type n on the core t' = Dummy integration variable where t' ! t denotes all times from the start to t of interest Equation 87 through Equation 89 can be determined using the following analytical solution, where the subscript n has been dropped for simplification:

Page 223 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Equation 90 H (ti-1 Equation 91 Equation 92 N

j=1 Equation 93 NN

~ ~

S~~~~~ti-1)~~~~~~~ S (i1=f*(i1 1 -Y ti1 (

~ii 1 ... 1) ]

jj=1 Equation 94 Qj(ti-()

hj~t~l)-V(ti-l) Equation 95 Ati_ 1 = ti - ti_1 Equation 96 where:

tj = End of specific time step interval ti-1 = Beginning of specific time step interval N = Number of ECCS strainers Subscript j = Variables specific to a given ECCS strainer Sk = Source rate for initial introduction of fiber type k Each of these equations can be solved by explicit forward integration assuming that the integrands are known at the beginning of each time step and that they remain constant during each time step.

Variables such as the source rate of debris to the pool (S), the strainer flow rate (Q), the pool volume (V),

the SI and CS flow split (y), and the SI flow vs. boil-off flow split (N)are defined in other sections. The filtration efficiency for the core (g) is conservatively assumed to be 100% (i.e. all debris that transports to the core is trapped). Therefore, the primary unknowns in Equation 90 through Equation 96 are the filtration efficiency at the strainer (f) and the shedding rate (s).

The shedding rate can be defined as a function of time as described in the following equation (28):

Page 224 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 t

sn(t) = v.n,ne-'nt I f(t') tmfl(t')e'dt'dt' Equation 97 f V(e')

0 where:

v, = Fraction of debris type n that is "sheddable" (i.e. able to pass through a debris bed) tn = Time constant associated with the shedding process Similar to the analytical solution above, Equation 97 can be solved as follows where the subscript n has been dropped for simplification:

Equation 98 77 Sj (t,) = 17*.mJh (t,) Equation 99 where:

mish = Mass of sheddable debris in the bed To determine the filtration efficiency and shedding rate, a series of penetration tests were conducted at Alden Research Laboratory (ARL) (26). A combination of 100% capture filter bags and isokinetic grab samples were used to gather data regarding the change in penetration as a function of strainer loading and time. A series of sensitivity tests were also conducted at Texas A&M University (TAMU) and ARL, which showed that penetration is not strongly dependent on water chemistry (27) or debris concentration and flow rate within the range of conditions tested (26). The ARL test data was statistically evaluated to determine appropriate fitting parameters to describe the shedding and filtration terms as a function of the debris load on the strainer and time (59). The filtration equation and fitting parameters for filtration and shedding are provided in Section 2.2.33. The time constant associated with the shedding process can be calculated from the rltest/htest values using the following equation:

??test .h~i1 Equation 100 t hh(t1 Page 225 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 5.10 Ex-Vessel Downstream Effects Components and Systems downstream of the sump strainers were evaluated to assess the effects of debris-laden fluid on the ability of the ECCS and CSS components and systems to perform their post-LOCA, design basis functions (86; 87; 88).

The evaluations were developed in accordance with WCAP-16406-P (89) and the accompanying NRC SER (90). No exceptions were taken to the WCAP-16406-P methodology. Note that the WCAP methodology is a deterministic approach that evaluates ex-vessel downstream effects in a conservative and bounding manner.

5.10.1 Pump, Valve, Component Wear Concerns with wear due to ingested debris were evaluated for pumps, valves, heat exchangers, orifices, and spray nozzles within the ECC and CS systems.

ECCS and CS Pump Wear Evaluation The HHSI, LHSI, and CS pumps were evaluated in accordance with the methodology established in WCAP-16406-P. The pumps were evaluated based on three aspects of operability: hydraulic performance, mechanical performance (vibration) and mechanical shaft seal assembly performance.

The WCAP evaluation concluded that no effect on hydraulic performance is expected.

The HHSI, LHSI, and CS pumps are multi-stage and were evaluated for mechanical (vibration) performance. In accord with WCAP-16406-P, the abrasive, erosive, and Archard wear models were used to calculate the amount of wear (mils) on the suction and discharge sides of each stage of the pumps.

The evaluation showed that the combined stiffness of the suction and discharge wear rings after being asymmetrically worn by free flowing abrasive wear and Archard wear, respectively, is less than the stiffness provided by both the suction and discharge wear rings being symmetrically worn to 2 times the design clearances for the HHSI, LHSI, and CS pumps. Therefore, the HHSI, LHSI, and CS pumps pass the mechanical (vibrations) evaluation and are acceptable The mechanical shaft seal assembly performance evaluation suggested replacing the LHSI, HHSI and CS pumps' carbon/graphite packing assemblies with a more wear resistant material. However, STP has an Engineered Safety Feature (ESF) atmospheric filtration system for the building where the pumps are located. Therefore the current the carbon/graphite seal bushings are acceptable as-is.

Page 226 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Valve Wear Evaluation Valve evaluations were performed in accordance with the criteria in WCAP-16406-P. The evaluations determined that the wear impact was not critical and in no need of further evaluation.

Heat Exchanger Wear Evaluation:

Plant heat exchangers were evaluated in accordance with WCAP-16406-P. Tube failure for heat exchangers will occur when the resultant wall thickness after erosion is less than the required wall thickness to retain internal and external pressures. The evaluation concluded that the minimum wall thickness required to retain both internal and external pressures is less than the resultant wall thickness after erosion. Therefore, the heat exchangers are not expected to fail.

Orifice Wear Evaluation ECC and CS System orifices were evaluated in accordance with the methodology established in WCAP-16406-P. WCAP-16406-P states that if the orifice inside diameter is changed by less than 3% due to erosive wear, the system performance may be considered negligible. The evaluation shows that the inside diameters of all orifices change by less than 3% and is therefore acceptable.

Spray Nozzle Wear Evaluation The CS system spray nozzles were evaluated in accordance with the methodology established in WCAP-16406-P. WCAP-16406-P concluded that un-acceptable wear is when the wear of the nozzle results in an expected system flow increase greater than 10%. The STP evaluation concludes that CSS flow is changed by less than 2% and is therefore acceptable.

5.10.2 System and Component Clogging/Blockage ECC and CS system and component clogging/blockage were evaluated in accordance with the methodology established in WCAP-16406-P. The evaluation showed that the following ECCS and CSS components can accommodate penetrated particles without clogging, restricting or blocking flow:

Pipes

" Pumps

" Valves

" Instrumentation

Orifices

" Eductors

" Heat exchangers

" Nozzles Page 227 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1 5.11 In-Vessel Downstream Effects In-vessel effects include issues with impaired heat transfer on the fuel rods due to debris buildup on the cladding, as well as core blockage due to debris that penetrated through the strainer. Also, in January 2012, the NRC requested that the generic boron precipitation issue be considered as part of the overall scope of GSI-191. Boron precipitation is discussed in more detail in Section 5.12.

5.11.1 Fuel Rod Debris Deposition (LOCADM)

When debris laden water is circulated through the core, it is possible that some of that debris may be deposited on the fuel rods resulting in a layer of debris that reduces heat transfer. This buildup process is illustrated in Figure 5.11.1.

The issue of impaired heat transfer due to debris deposition on the fuel rods has been evaluated by the PWROG in WCAP-16793-NP, and a tool called LOCADM was developed to calculate the maximum thickness of debris and maximum temperature of the cladding based on plant-specific inputs (62). The acceptance criteria for this evaluation are that the peak cladding temperature must be less than 800 'F, and the total deposition thickness must be less than 50 mils. Based on an STP-specific evaluation using the conservative WCAP-16793-NP methodology, the maximum cladding temperature was calculated to be 368.90 °F, and the maximum deposition thickness was calculated to be 13.64 mils (91). These values are well within the acceptance criteria. Therefore, since the deterministic evaluation for STP shows that there are no issues with debris deposition on the fuel rods, this issue was not explicitly evaluated in CASA Grande.

Page 228 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1 Coolant with Impurities o00 Pure steam Coolant with Impurities Cladding, Cladding Oxide and Original Crud before Recirc

'*-- Pre-existing Crud

\ " Cladding Oxide Cladding goo Densification 1,00 I

- Pore Filling Deposit cracking and under-scale growth of LOCA scale 0

LOCA Scale Figure 5.11.1 - Deposit growth process assumed by LOCADM when core is boiling (62) 5.11.2 Core Blockage Scenarios The potential for core blockage to occur is largely dependent on 1) the size of the break, 2) the location of the break, and 3) the injection path. Medium and large breaks are similar in terms of core blockage (the main difference being the total SI flow rate as shown in Section 2.2.9). However, for a small break, the break size is small enough to allow the RCS to fill with water and enable natural circulation. This would be true for breaks on either the hot or cold leg side. For medium and large breaks, however, the flow path through the core is highly dependent on the location of the break (cold leg or hot leg side breaks) and the injection path (cold leg or hot leg injection).

Page 229 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Following a LOCA at STP, water would initially be injected from the RWST to three out of four of the cold legs (see Figure 5.11.2). Since the water from the RWST is free of debris, there would be no potential for core clogging during this phase. After the RWST has been drained (30-44 min for an LBLOCA, 44-79 min for an MBLOCA, and up to several hours for an SBLOCA (5)), the SI and CS pumps would be realigned to take suction from the ECCS sumps. At this point, some fine debris could start to pass through the strainers. Due to potential issues with boron precipitation, two of the three trains would be realigned from cold leg injection to hot leg injection 6.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months after the start of the accident 2' (see Assumption 1.i). The third train would remain aligned on cold leg injection. In the unlikely event that only one train is operating, the realignment to hot leg injection would not occur since the emergency operating procedure (EOP) requires that one train be left on cold leg injection (36).

Figure 5.11.2 - Illustration of RCS at STP 21 For some small break scenarios, hot leg switchover could occur prior to the start of recirculation.

Page 230 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Medium and Laree Hot Lee Breaks For a medium or large hot leg break, the SI flow would initially be injected in the cold legs forcing the flow to pass through the core and then spill out the break. After the start of recirculation, but before hot leg switchover (HLSO), debris that penetrates the strainer would transport with the flow and accumulate in the core. The head loss due to the debris would result in a compensating rise in the steam generator water level. Eventually, given sufficient head loss, a portion of the SI flow could pass over the steam generator tubes reducing the flow through the core. This scenario is illustrated in Figure 5.11.3.

Driving Head

\\ Hot Leg Hot Leg Figure 5.11.3 - Large or medium hot leg break during cold leg injection with partial core blockage There are two potential concerns that have been raised with the hot leg break/cold leg injection scenario: 1) if flow starts spilling over the steam generator tubes, it may cause a siphon to form that would suck all of the SI flow over the tubes and directly out the break (92), and 2) core blockage may be large enough to prevent sufficient decay heat removal from entering the bottom of the core and the remaining SI flow may preferentially spill over the steam generator on the broken loop where it could pass directly out the break.

Page 231 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 A siphon is essentially an inverted U-shaped tube where the weight of the water on the discharge side of the tube pulls down; creating a low pressure region at the top of the tube that pulls water up from the suction side of the tube. However, if the pressure at the top of the tube falls below the vapor pressure, the siphon will break. At standard conditions, the maximum possible height of a siphon (neglecting frictional losses) is approximately 33 ft. Since the temperature (and the corresponding vapor pressure) in the steam generator tubes would be significantly higher than standard conditions, the siphon would break at an elevation less than 33 ft. Elevated containment pressure could increase the maximum height of a siphon, but elevated pressure generally trends with elevated temperature, and the containment pressure drops quickly during the RWST injection phase (5). At STP, the lowest steam generator tubes are over 40 ft above the bottom of the hot legs (93). Therefore, flow over the steam generator tubes would not result in a siphon effect. This conclusion is further supported by the results of thermal-hydraulic simulations (29). Thermal-hydraulic simulations also showed that water would not preferentially flow over the steam generator tubes on the broken loop (29). This is discussed in more detail below.

After switchover to hot leg injection, two trains of SI flow would be injected in two of the hot legs, while the third train of SI flow would continue to be injected in one of the cold legs. The water injected through the cold leg would continue flowing through the same path (through the core and out the break and/or over the steam generator tubes if partial or full blockage has occurred). The water injected in the hot legs would pass through the upper plenum providing coolant at the top of the core, and exit through the broken loop. This is illustrated in Figure 5.11.4. Note that the simultaneous cold leg injection flow is not shown in this figure, but would tend to enhance mixing and the total flow through the core (unless the bottom of the core was fully blocked by debris during the cold leg injection phase).

Page 232 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Figure 5.11.4 - Large or medium hot leg break during hot leg injection Although it is possible that debris could continue to penetrate the strainer later in the event (after HLSO), it is not likely that the debris would cause any significant blockage issues during hot leg injection (see Assumption 10.a).

Medium and Large Cold Leg Breaks For a medium or large cold leg break, the SI flow would initially be injected in the cold legs, and the majority of the flow would bypass the core spilling directly out the break. However, a portion of the injected water would flow into the core to makeup the water lost due to boil-off. The debris that penetrates the strainer after the start of recirculation would transport with the flow, and a portion of the debris could accumulate in the core. Unlike the hot leg break scenario, however, the head loss due to the debris would result in a decrease in the core water level rather a rise in the steam generator water level. Eventually, given sufficient head loss, a portion of the core could be uncovered resulting in potential core melt. The cold leg break during cold leg injection scenario is illustrated in Figure 5.11.5.

Page 233 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1 Figure 5.11.5 - Large or medium cold leg break during cold leg injection with partial core blockage There are two potential concerns that have been raised with the cold leg break/cold leg injection scenario: 1) core blockage may be large enough that the water level cannot be maintained above the top of the core, and 2) debris buildup at the bottom of the core may prevent mixing with the lower plenum volume resulting in a more rapid onset of boron precipitation. The core blockage issue is discussed in more detail in the remainder of Section 5.11, and the boron precipitation issue is addressed in Section 5.12.

After switchover to hot leg injection, two trains of SI flow would be injected in two of the hot legs, while the third train of SI flow would continue to be injected in one of the cold legs. The water injected through the cold leg would flow directly out the break. The water injected in the hot legs would pass from the upper plenum, down through core and lower plenum, and exit through the broken loop. This is illustrated in Figure 5.11.6. Note that the simultaneous cold leg injection flow is not shown in this figure, but would have minimal effect since it would bypass the core and spill directly out the break.

Page 234 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Figure 5.11.6 - Large or medium cold leg break during hot leg injection After HLSO, Debris that accumulated in the core during cold leg injection would tend to get back-flushed out of the core. Some debris may continue to penetrate the strainer later in the event and could be trapped at the top of the core during hot leg injection. However, as discussed in Assumption 10.b, full blockage at the top of the core would be prevented by countercurrent flow due to thermal buoyancy.

Small Hot and Cold Leg Breaks For a small break, the capacity of the SI pumps is significantly higher than the flow that would be spilling out the break. Early in the event when the decay heat is still relatively high, the required flow to cool the core can be higher than the SI flow rate. In this scenario, however, the RCS would be mostly full of water and natural circulation would circulate sufficient flow to cool the core. This is illustrated in Figure 5.11.7.

Page 235 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Figure 5.11.7 - Small hot leg break during cold leg injection Thermal-Hydraulic Simulations A series of RELAP5 simulations were run to investigate the effects of full blockage at the bottom of the core following the start of recirculation to determine which scenarios would proceed to core damage (29). The six scenarios that were modeled included small (2-inch), medium (6-inch), and large (DEGB) breaks on both the hot leg and cold leg piping. The results of the simulations showed that if the core is fully blocked for a hot leg break (of any size), the water would spill over the steam generator tubes and pass through the upper plenum before spilling out the break. The simulations showed that this alternate flow path is sufficient to remove decay heat from the core and prevent the peak cladding temperature from exceeding 800 °F (29). Similarly, for the small (2-inch) cold leg break scenario, the break is small enough that the SI flow would fill the RCS above the top of the steam generator tubes allowing the SI flow to reach the core through the upper plenum, and prevent subsequent core damage (29). Out of the six scenarios that were run, only the medium and large cold leg breaks proceeded to core damage following the simulated blockage at the bottom of the core.

Page 236 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 5.11.3 Decay Heat Boil-Off Flow Rate For cold leg breaks during cold leg injection, the portion of the SI flow entering the lower plenum and core depends on the boil-off rate. The boil-off rate can be calculated based on the decay heat curve. As shown in Section 2.2.34, the 1979 ANS plus 2 sigma uncertainty heat generation rate was used. The actual flow into the core required to remove the decay heat is dependent on the temperature of water entering the core, the saturation temperature (i.e. the RCS pressure), and the decay heat curve. All three of these parameters can change significantly over time. The time-dependence of the decay heat is defined by the 1979 ANS curve, and the time-dependence of the inlet temperature and RCS pressure is dependent on the break size, pool temperature, etc.

The boil-off flow rate can be calculated using the following equations:

Pcore Qboila Pin . (AHtemp + AHvap) Equation 101 AHtemp = H1 ,sat - Hin Equation 102 AHvap = Hv,sat - Hi,sat Equation 103 where:

Qboi- = Boil-off flow rate Pcore = Power introduced from the core pi, = Density of SI flow as a function of the inlet temperature AHtemp = Change in enthalpy required to raise the temperature to saturation AHvap = Change in enthalpy required to change phases from liquid to vapor Hi, = Enthalpy of SI flow entering the vessel as a function of the inlet temperature Hi,sat = Saturated liquid enthalpy as a function of the RCS pressure Hv,sat= Saturated vapor enthalpy as a function of the RCS pressure The enthalpies and inlet density were determined based on standard water properties assuming that the RCS pressure is 14.7 psia, and the SI flow is saturated liquid at 212 "F (see Assumption 10.c). These pressure and temperature values were conservatively used for all break scenarios and treated as constants over the duration of the event. Table 5.11.1 and Figure 5.11.8 show the calculated boil-off rate as a function of time.

Page 237 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Table 5.11.1 - Decay heat generation rate based on 1979 ANS plus 2 sigma uncertainty Decay Inlet RCS Boll-off Time Decay Heat Generation Heat Temp Pressure (Btum) (Bum 3 Rate (hr) Rate (Btu/Btu) (Btu/hr) (F) (psia) (Btu/bm) (Btu/lbm) (lbJft ) (gpm) 0.003 0.053876 6.59E+08 212 14.7 0.00 970.13 59.83 1,414.7 0.004 0.050401 6.16E+08 212 14.7 0.00 970.13 59.83 1,323.5 0.006 0.048018 5.87E+08 212 14.7 0.00 970.13 59.83 1,260.9 0.011 0.042401 5.18E+08 212 14.7 0.00 970.13 59.83 1,113.4 0.017 0.039244 4.80E+08 212 14.7 0.00 970.13 59.83 1,030.5 0.022 0.037065 4.53E+08 212 14.7 0.00 970.13 59.83 973.3 0.028 0.035466 4.34E+08 212 14.7 0.00 970.13 59.83 931.3 0.04 0.032724 4.OOE+08 212 14.7 0.00 970.13 59.83 859.3 0.06 0.030936 3.78E+08 212 14.7 0.00 970.13 59.83 812.3 0.11 0.027078 3.31E+08 212 14.7 0.00 970.13 59.83 711.0 0.17 0.024931 3.05E+08 212 14.7 0.00 970.13 59.83 654.7 0.22 0.023389 2.86E+08 212 14.7 0.00 970.13 59.83 614.2 0.28 0.022156 2.71E+08 212 14.7 0.00 970.13 59.83 581.8 0.42 0.019921 2.44E+08 212 14.7 0.00 970.13 59.83 523.1 0.56 0.018315 2.24E+08 212 14.7 0.00 970.13 59.83 480.9 1.1 0.014781 1.81E+08 212 14.7 0.00 970.13 59.83 388.1 1.7 0.013040 1.59E+08 212 14.7 0.00 970.13 59.83 342.4 2.2 0.012000 1.47E+08 212 14.7 0.00 970.13 59.83 315.1 2.8 0.011262 1.38E+08 212 14.7 0.00 970.13 59.83 295.7 4.2 0.010097 1.23E+08 212 14.7 0.00 970.13 59.83 265.1 5.6 0.009350 1.14E+08 212 14.7 0.00 970.13 59.83 245.5 11.1 0.007778 9.51E+07 212 14.7 0.00 970.13 59.83 204.2 16.7 0.006958 8.51E+07 212 14.7 0.00 970.13 59.83 182.7 22.2 0.006424 7.85E+07 212 14.7 0.00 970.13 59.83 168.7 27.8 0.006021 7.36E+07 212 14.7 0.00 970.13 59.83 158.1 41.7 0.005323 6.51E+07 212 14.7 0.00 970.13 59.83 139.8 111 0.003770 4.61E+07 212 14.7 0.00 970.13 59.83 99.0 167 0.003201 3.91E+07 212 14.7 0.00 970.13 59.83 84.1 222 0.002834 3.46E+07 212 14.7 0.00 970.13 59.83 74.4 278 0.002580 3.15E+07 212 14.7 0.00 970.13 59.83 67.7 Page 238 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Boil-off Flow Rate 1,600 1,400 1,200 E

1,000 i

1800 S4oo

~600 400 200 J 0 1 0.001 0.01 0.1 1 10 100 1000 Time (hr)

Figure 5.11.8 - Time-dependent boil-off flow rate 5.11.4 Time-Dependent Core Debris Accumulation As discussed in Section 5.11.2, the time-dependent accumulation of debris on the core depends on the break location and injection flow path (hot leg versus cold leg side breaks and cold leg versus hot leg injection). A fraction of the debris that penetrates the strainer can be split off to the spray pumps, and a fraction of the debris transported with the SI flow can bypass the core and spill directly out the break (see Section 5.9). The debris transport to the core is defined as shown in the following equations:

Fs1(t) = Qsi(t) Equation 104 Qs 1(t) + Qcs (t) for HL Breaks during CL Injection Fcore(t) = Fs(t) Qbol (t) for CL Breaks during CL Injection Equation 105 for all Breaks during HL Injection MC(t) = Fcore(t)* MP(t) Equation 106 Page 239 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 where:

Fs1(t) = Time-dependent fraction of penetrated debris that transports to the SI pumps Fcore(t) = Time-dependent fraction of penetrated debris that transports to the core Qs,(t) = Time-dependent safety injection flow rate Qcs(t) = Time-dependent containment spray flow rate Qboi(t) = Time-dependent boil-off flow rate MC(t) = Time-dependent mass of debris on the core Mp(t) = Time-dependent mass of debris that penetrates the strainer 5.11.5 Acceptance Criteria: Debris Loads The acceptance criteria for debris loads on the core were defined based on the break location, injection flow path, and fiberglass debris loads that could potentially cause issues for debris blockage. Based on the thermal-hydraulic modeling, which showed that full blockage at the bottom of the core would not result in core damage for hot leg breaks, the acceptance criterion was set to essentially an infinite fiber quantity.

For cold leg breaks, an acceptance criterion of 15 g/FA was used based on the conservative results of testing by the PWROG (see Design Input 2.2.35).

Note that the core blockage acceptance criteria are bounded by the boron precipitation acceptance criteria (see Section 5.12.2).

5.12 Boron Precipitation Boric acid precipitation can occur in cases where there is boiling in the core. The water entering the core has a given boron concentration. As the water boils, the boron becomes more concentrated and eventually can reach the solubility limit and begin to precipitate. Figure 5.12.1 shows an example of boron precipitation during a test.

Page 240 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Figure 5.12.1 - Amorphous precipitate formation on heated surface (94)

Significant boron precipitation is most likely to occur for a medium or large cold leg break during cold leg injection. In this scenario the water in the core would be boiling and the net flow entering the core would be equivalent to the decay heat boil-off rate (see Section 5.11). To prevent boron precipitation in these scenarios, the SI flow is switched from cold leg injection to hot leg injection. The required switchover timing is dependent on the concentration of boron in the RCS/RWST/accumulators, the decay heat level, and natural mixing processes within the reactor vessel based on temperature and/or density gradients. The generic methodology used for evaluating boron precipitation has been questioned by the NRC (94), and the PWROG is currently addressing these concerns to determine whether the physical phenomena associated with temperature or density driven mixing have been appropriately modeled. The reason that boron precipitation was included in the overall GSI-191 issue, however, is that even if the physical phenomena for temperature and density driven mixing was appropriately modeled previously, the formation of a debris bed at the bottom of the core may interrupt these natural mixing processes and accelerate the onset of boron precipitation.

Based on an STP-specific evaluation using the old methodology, it was determined that boron precipitation would not occur until at least 7.0 hours0 days 0 hours 0 weeks 0 months after the initiation of the event (95). For the risk-informed GSI-191 evaluation, it was assumed that the previous methodology was appropriate, and that boron precipitation would not occur unless a significant debris bed builds up on the bottom of the core that could disrupt the natural mixing processes that were credited (see Assumption 11.a).

Page 241 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 5.12.1 Time-Dependent Core Debris Accumulation The debris accumulation on the core is addressed in Section 5.11.4.

5.12.2 Acceptance Criteria: Debris Loads As discussed in Assumption 11.d, boron precipitation is not considered to be an issue for hot leg breaks.

For medium and large cold leg breaks, the acceptance criterion for boron precipitation was assumed to be 7.5 g/FA of fiber debris on the core (see Assumption 11.b).

5.13 Parametric Evaluations One of the greater values that accrue from building a comprehensive system analysis platform like CASA Grande lies in the ability to conduct parameter studies and comparative examinations to help prioritize plant operation strategies, risk mitigation investment, and research and development activities.

Within this version of the analysis, it was not possible to run an extensive suite of parametric evaluations. However, one scenario of particular interest was a quantitative risk-benefit assessment of the strainer replacement project at STP.

Strainer replacement was performed at STP Unit 1 in October 2006 and Unit 2 in April 2007 (96). The original strainers were a flat-plate box design as shown in Figure 5.13.1. The strainer area was only 155.4 2

ft per train (97). The ECCS suction strainers were upgraded to a total of 1,818.5 ft2 per strainer using a PCI SureFlowTM design (see Section 2.2.26). The strainer upgrade increased the capacity for fibrous debris loading per unit area by almost a factor of 12, but may have concurrently increased the vulnerability to debris penetration for small debris loads that do not fully cover the surface of the new design.22 22 There may be some other variations between the strainers that could affect some aspects of the analysis, but the most important difference is the strainer area.

Page 242 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1 Figure 5.13.1 - STP ECCS strainer prior to upgrade The test-validated filtration and penetration model implemented in CASA now enables a comparison of competing effects like enhanced debris penetration from low debris areal densities on a large strainer compared to higher pressure drops caused by large debris areal densities on a small strainer. To quantify the net benefit of strainer replacement, the CASA analysis was repeated under the same set of parameter assumptions and conservative uncertainty factors for chemical effects and pressure-drop prediction, by simply changing the clean strainer surface area.

Results for LBLOCA conditional failure probabilities are presented in Table 5.13.1 to compare the previous plant condition to the present plant condition. Replacement of the ECCS strainers resulted in a minimum reduction factor in conditional failure probability for LBLOCA of 118 (see highlighted cell).

More significantly, strainer replacement eliminated vulnerabilities to small and medium break events that do trigger a large fraction of failure events when analyzed with the old strainer area. No failures are recorded for small and medium-break scenarios using the current ECCS strainer specifications.

Trends in Table 5.13.1 also confirm that the models implemented in CASA for debris filtration/penetration and circulation through the plant are behaving intuitively. For example, small strainers will load rapidly and limit debris penetration primarily to the shedding phase of debris-bed behavior. This is indicated by the absence of boron-related core-fiber load failures. All failure modes for the smaller strainers are related to higher head loss that is induced both by thicker beds and by much higher flow velocities.

Page 243 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1 The reductions in conditional failure probability shown in Table 5.13.1 cannot be equated directly with the same reduction in ACDF risk without a complementary RISKMAN analysis, but preliminary M

T indications are that significant risk reductions were achieved through the strainer replacement campaign. These risk reductions are likely to have been significant enough to move STP from Regulatory Guide 1.174 Region II down to the current quantified status in Region III (70). Thus, proactive compliance with NRC regulation through strainer upgrade achieved an important and quantifiable improvement in plant safety. Note that the core blockage failures all have a zero failure probability. This is due to the fact that the core blockage acceptance criteria are bounded by the boron precipitation acceptance criteria. It doesn't mean that it is not possible for core blockage to occur.

Table 5.13.1 - Comparison of mean LBLOCA conditional failure probabilities before and after ECCS 23 strainer replacement Failure Mode Case 1 Case 9 Case 22 Case 26 Case 43 After Strainer Replacement Core Blockage 0 0 0 0 0 Boron Precipitation 6.94E-04 1.82E-03 7.51E-05 6.15E-05 3.42E-06 Sump Failure 2.45E-04 5.39E-04 1.32E-03 9.56E-04 4.45E-03 Total 9.38E-04 2.35E-03 , 1.40E-03 1.02E-03 4.45E-03 Before Strainer Replacement Core Blockage 0 0 0 0 0 Boron Precipitation 0 0 0 0 0 Sump Failure 0.2488 0.2776 0.65 0.815 1.00 Total 0.2488 0.2776 0.65 0.815 1.00 Conditional Failure Probability Reduction Factor Core Blockage ...... --- .. --- -

Boron Precipitation N/A N/A N/A N/A N/A Sump Failure 1016 515 492 853 225 Total 265 118 1 464 799 225 23 Case 1 = Full three-train operation; Case 9 = Dual LHSI pump failures; Case 22 = Single train failure; Case 26 =

Single train failure with an additional LHSI pump failure; Case 43 = Dual train failure.

Page 244 of 260 I

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 6 Results One of the primary functions served by CASA Grande in the risk-informed resolution process is quantifying conditional failure probabilities related to GSI-191 phenomena for various states of plant operability. Failure probabilities are passed to the plant-wide PRA, which determines the incremental risk associated with GSI-191 failure modes. In this role, CASA serves very much like an elaborate fault tree that informs top-event branch fractions that are built-in to the event tree. Three new top events have been added to the PRA to accommodate composite GSI-191 failure processes: 1) failure at the sump strainer, 2) boron precipitation in the core, and 3) blockage of the core.

CASA compiles the three composite failure probabilities needed for the PRA by testing the outcome of every postulated break scenario against seven performance thresholds: 1) strainer AP _ NPSHma,g,,, ,,2) strainer A* " Pmi,,,e, 3) strainer FJ.oid >_0.02, 4) core fiber load Ž_cold leg break fiber limit for boron precipitation, 5) core fiber load > hot leg break fiber limit for boron precipitation, 6) core fiber load >

cold leg break fiber limit for flow blockage, and 7) core fiber load _>hot leg break fiber limit for flow blockage. Failure Modes 1-3 are counted as failures if any single operable strainer exceeds the thresholds at any time during the 36-hr calculation. Failure Modes 4-7 are assessed against the accumulated fiber penetration from all operable strainers, and they must exceed the threshold before switchover to hot leg injection to be counted as failures. For the present quantification, thresholds for modes 5-7 were set infinitely high so that only exceedance of the cold leg break boron precipitation loading (Failure Mode 4) was recorded as failure. This approach is reasonable because the threshold for Failure Mode 4 is substantially lower than the others, and all depend on the same internal flow distribution and fiber accumulation processes.

In the present quantification, violation of any of the seven performance thresholds is counted as an independent failure. Thus, it is possible that a single scenario can contribute both a strainer-related failure tally, and a core fiber load failure tally. After a suite of scenarios is performed, the sum of probability weights for failed scenarios within each LOCA category is divided by the sum of probability weights for all scenarios within each LOCA category to generate the conditionalfailure probabilities needed for the PRA. Table 6.1 reports the mean conditional failure probability associated with each composite failure mode for each of five pump state combinations. No failures were recorded for small or medium-break events, and later discussion will explain that only the higher range of large-break events contributed to failure. In addition to the composite PRA failure modes, total failure probability conditioned on LOCA category is also provided.

Page 245 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 24 Table 6.1 - Mean LBLOCA conditional failure probabilities for five states of pump availability Failure Mode Case 1 Case 9 Case 22 Case 26 Case 43 Core Blockage 0 0 0 0 0 Boron Precipitation 6.94E-04 1.82E-03 7.51E-05 6.15E-05 3.42E-06 25 Sump Failure 2.45E-04 5.39E-04 1.32E-03 9.56E-04 4.45E-03 Total 9.38E-04 2.35E-03 1.40E-03 1.02E-03 4.45E-03 Table 6.1 results can be interpreted in the following ways: Design basis accident response with three trains operable (Case 1) is estimated to incur a total failure probability of 0.09% given that an LBLOCA occurs (9 failures in every 10,000 large-break events). If only one train is operable (Case 43), this estimate increases to 0.45% (45 failures in every 10,000 large-break events). The primary reason for the increase is the additional head loss incurred at the single strainer by collecting all of the debris that is distributed in proportion to flow across three strainers under Case 1. Conversely, failures incurred by exceeding the boron fiber load are reduced (compare first and last columns), because less cumulative fiber is penetrating the single, highly loaded strainer. Blockage failure is reported as zero probability because the thresholds were set very high, partly to avoid double counting blockage failures for events that first exceed the bounding low value for fiber load thresholds related to boron precipitation.

Conditional failure probabilities reported in Table 6.1 are described as "mean" or "expected" values because five point estimates associated with independent samples of the NUREG-1829 break frequency envelope have been averaged for use in the PRA. The following discussion explains the origin and the mechanics of this averaging process.

Recall that the NUREG-1829 tables assign confidence-levels to estimates of annual occurrence frequency as a function of break size. This assignment of confidence level defines an envelope of epistemic uncertainty that was fit using bounded Johnson probability density functions at each discrete break size for which percentiles of confidence were tabulated. The purpose of these fits was to enable interpolation of the confidence bands at any intermediate break size of interest. The relationship defined by NUREG-1829 between annual occurrence frequency (events per year) and break size is presented in terms of complementary cumulative distribution functions (CCDFs). This format implies that underlying probability density functions (PDFs) have been integrated, and it is important to consider the form of the PDFs before selecting an interpolation scheme that will be applied to the 24 Case 1 = Full three-train operation; Case 9 = Dual LHSI pump failures; Case 22 = Single train failure; Case 26 =

Single train failure with an additional LHSI pump failure; Case 43 = Dual train failure.

25 The boron precipitation evaluation for Case 43 assumes that hot leg switchover will occur at 6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months .

However, if only one train is operating, it is likely that hot leg switchover would not occur and medium and large cold leg breaks are likely to have boron precipitation. This is not reflected in the value specified for Case 43, but is explicitly evaluated in the PRA model.

Page 246 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 CCDFs. Conversely, any presumption about interpolation of the CCDF has implications for the implied form of the PDF.

A PDF defined for break size must define the probabilityper unit of size that a break occurs within the interval between the discrete sizes tabulated in NUREG-1829. Without knowing the details of how fracture mechanics processes were treated during compilation of the NUREG-1829 table, it is difficult to defend any assumption other than uniform probability density between the tabulated discrete sizes.

Uniform probability density means that any break size within the interval is equally likely. Uniform (constant) break-size probability between two CCDF values is easily calculated as the positive difference between the complementary cumulative annual frequencies divided by the positive range of size across the interval divided by the total annual exceedance frequency for the smallest break size. The integral of a constant PDF needed to form a CCDF is a straight line, and this implies that linear-linear interpolation of the NUREG-1829 table is the treatment most consist with the assumption of constant underlying probability density. Alternative treatments may be justified, but insufficient evidence was available to defend them for use in this quantification.

Linear-linear interpolation of the NUREG-1829 table values leads to an interesting visual effect when plotted on log-log axes. As shown in Figure 6.1, the linear CCDF appears as a periodically looping curve on a logarithmic scale. The practical effect of linear interpolation is that break frequency (and probability) remains conservatively high across most of each size range, and only descends to match the tabulated CCDF values near the end of each interval. Figure 6.1 illustrates the extreme endpoints of the bounded Johnson fits (solid lines) and several typical random samples of the break-frequency profile that were used in this assessment (dashed lines).

Non-uniform stratified random sampling of break-frequency profiles from the Johnson PDF envelope are performed in exactly the same manner as for all other random variable. Non-uniform probability bins are predefined based on the desired number of samples and on the direction of presumed conservatism, then random percentiles are chosen from within each bin to represent, or "carry", the associated probability weights. For this study, five independent random samples were extracted from the Johnson envelope for each plant state, with an emphasis on upper percentiles of the break-frequency uncertainty envelope. Given a sample of five percentiles, the Johnson fits are then inverted to find the corresponding annual frequencies. It is important to note that all Johnson fits are perfectly correlated by using the same fixed values of the sampled percentiles. Finally, the set of annual frequencies from each Johnson fit is linearly interpolated to create the break-frequency profiles shown as the dashed lines in Figure 6.1.

Each break-frequency profile is fully analyzed in CASA using a set of three batch replicates containing approximately 2,250 break scenarios each to obtain a point estimate of failure probability for the composite modes. (Residual sampling imprecision of ~20% between the three replicates is typical of this scenario sampling size. Better precision can be obtained using larger sample sizes or more replicates as Page 247 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 time permits.) Probability weights from stratified sampling of the Johnson envelope are then used to form the weighted conditional means reported in Table 6.1 above.

Break Frequency Sampling Density

-2 10 .. . . . ,

A

-0 10-

_10 10 101 100 break size (in)

Figure 6.1 - Linear-linear interpolation of bounded Johnson extrema (solid) with non-uniform stratified random break-size profiles (dashed)

Table 6.2 below reports the five point estimates and their associated probability weights generated for the total failure probability under plant operability Case 43 (single train operable). The weighted mean is formed simply by multiplying each point estimate by its probability weight and adding the products.

Similar distributions were formed for all composite failure modes and for all plant operability states, but only the weighted means are presented in Table 6.1 above.

Page 248 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Table 6.2 - Distribution of total conditional failure for LBLOCAs under Case 43 (single train operable)

Johnson Point Estimate Cumulative of PPoil Wailt Weight Probability 0.0 0.0 0.0 3.13e-3 8.22e-1 8.22e-1 7.49e-3 4.62e-3 8.27e-1 1.03e-2 1.46e-1 9.73e-1 1.15e-2 1.00e-3 9.74e-1 1.2e-2 2.60e-2 1.0 4.45e-3 weighted mean The cumulative distribution defined for total failure probability under Case 43 (one train operable) in Table 6.2 above is plotted in Figure 6.2. This distribution reflects only the uncertainty inherent to the estimation of annual break frequency. All other random variability, including ranges on physical phenomena and decision criteria, has been integrated into each point estimate. As shown in Table 6.2 and Figure 6.2, typical variation in failure probability estimates spans a factor of 2 to 4 between the minimum and maximum values (0.012/0.003 = 3.8). This variation is caused solely by the shape of the randomly selected break-frequency profiles, which dictates the relative proportion of break frequency by size.

It is important to reemphasize that CASA Grande never makes any direct use of the annual break frequency as a time-rate quantity. All analyses proceed conditioned on the assumption that a break has already occurred. Sample profiles taken from the break-frequency envelope then describe how to partition the relative occurrence of breaks by size. CASA further redistributes the relative size probability across weld types in order to map the cumulative probability of a break as a function of size to discrete locations in the plant (see Section 5.3).

On the other side of the risk-informed analysis, the PRA samples directly from the NUREG-1829 envelope to determine the annual frequency of the initiating event. That is why it is important for CASA and the PRA to use exactly the same representation of the Johnson PDF fits. The Johnson fits are evaluated analytically in CASA to generate a table of empirical PDFs that are manually passed to RISKMAN for repeated sampling in the risk quantification. While RISKMAN may generate thousands of samples from the annual frequency envelope, CASA is not yet computationally efficient enough to match that resolution on the "outer loop." At this stage it is much more important to place available sampling resolution on the physical variables (inner loop) that drive the outcome of each break scenario, and to rely on non-uniform sampling for generating unbiased estimates of the mean failure probability.

Page 249 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1 At the beginning of this quantification, it was presumed that failure distributions like that shown in Figure 6.2 would be sampled by the PRA to generate distributions of incremental risk attributable to GSI-191 phenomena. Under this assumption, additional sampling resolution was placed in the high-percentile tails of the Johnson distributions so that the distribution of failure probability associated with the highest annual frequencies would be the most accurate. Practical experience with the PRA sampling process now emphasizes the mean failure estimate, so CASA sampling strategies may be changed accordingly in the future.

Cumulative Dist of Total Failure Probability for Case 43 0.98 0-96 3 0.94 0

C

.0-92 C

0 0_9 S0.88 U

0.86 Weighted mean445-0 082 -

2 4 6 8 10 12 14 failure probability x 10-3 Figure 6.2 - Empirical distribution of total failure probability for Case 43 (one train operable) based on five discrete samples of the NUREG-1829 break-frequency uncertainty envelope One other key piece of information that is passed from CASA back to the PRA is the conditional split fraction for cold leg breaks in each LOCA category. Recall that CASA distributes total break-size probability for a single NUREG-1829 profile across all welds in containment. This process uses the hybrid weighting scheme (described in Section 5.3) to account for the contributions of small breaks on large pipes to the small and medium LOCA categories. Each break scenario sampled from this process carries a specific size and location and a fractional weight of the total break-size probability. Before any other Page 250 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 physical parameters are considered, the distribution of probability weight can be partitioned into hot leg and cold leg events and by LOCA size.

Table 6.3 itemizes all cold leg split fractions obtained for the 15 batches associated with Case 43. These values were obtained by dividing the sum of probability weights for cold leg breaks in each LOCA category by the sum of probability weights for all breaks in the LOCA category. Hot leg split fractions are simply the complement of any single entry in the table. Recall that three replicates of 2,250 scenarios are evaluated for each of five break frequency profiles for a total of 3 x 2,250 x 5 = 33,750 scenarios per plant state. Cold leg split fractions are mildly dependent on the break frequency profile shape (note repetition in successive groups of three rows), but they are independent of plant operability state. It is interesting to note that proportion of large cold-leg breaks is substantially smaller than the 50%

proportion assumed in the preliminary quantification.

Table 6.3 - Cold leg split fractions conditioned on LOCA category for Case 43 Total Small Medium Large 4.2052034e-01 4.2962813e-01 3.8133459e-01 2.3059826e-01 4.2052034e-01 4.2962813e-01 3.8133459e-01 2.3059826e-01 4.2052034e-01 4.2962813e-01 3.8133459e-01 2.3059826e-01 4.2015626e-01 4.2933789e-01 3.8133521e-01 2.3048163e-01 4.2015626e-01 4.2933789e-01 3.8133521e-01 2.3048163e-01 4.2015626e-01 4.2933789e-01 3.8133521e-01 2.3048163e-01 4.2014556e-01 4.2932931e-01 3.8133576e-01 2.3044256e-01 4.2014556e-01 4.2932931e-01 3.8133576e-01 2.3044256e-01 4.2014556e-01 4.2932931e-01 3.8133576e-01 2.3044256e-01 4.2210420e-01 4.3087029e-01 3.8133228e-01 2.3115092e-01 4.2210420e-01 4.3087029e-01 3.8133228e-01 2.3115092e-01 4.2210420e-01 4.3087029e-01 3.8133228e-01 2.3115092e-01 4.3407111e-01 4.3931916e-01 3.8118954e-01 2.3960731e-01 4.3407111e-01 4.3931916e-01 3.8118954e-01 2.3960731e-01 4.3407111e-01 4.3931916e-01 3.8118954e-01 2.3960731e-01 CASA Grande diagnostic capabilities are still maturing for locating specific welds and specific insulation target locales that contribute dominant failure probability to the total. However, some preliminary information was compiled for this analysis. Table 6.4 lists a sample of the specific welds, break sizes, and general containment zones that are associated with one or more failure modes in Case 43. This list includes only 49 of 1,659 failed scenarios that were tallied during the analysis.

Page 251 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Table 6.4 - Partial itemization of break events that lead to failure for plant state Case 43 Weld Location System Weld Break LOCA DEGB Break Break Cat. Size (in) Size Side Compartment 12-RC-1112-BB1-2 RHR-Suction 7A 10.126 Large Yes Hot SG Compartment 12-RC-1112-BBl-1 RHR-Suction 7E 10.126 Large Yes Hot SG Compartment 12-RC-1125-BB1-9 SI-ACC-CL1 7N 10.126 Large Yes Cold SG Compartment 12-RC-1125-BB1-12 SI-ACC-CL1 7N 10.126 Large Yes Cold SG Compartment 12-RC-1212-BB1-2 RHR-Suction 7A 10.126 Large Yes Hot SG Compartment 29-RC-1201-NSS-3 RC 7E 10.126 Large Yes Hot SG Compartment 12-RC-1221-BB1-9 Sl-ACC-CL2 7N 10.126 Large Yes Cold SG Compartment 12-RC-1221-BBl-11 Sl-ACC-CL2 7N 10.126 Large Yes Cold SG Compartment 12-RC-1221-BB1-12 Sl-ACC-CL2 7N 10.126 Large Yes Cold SG Compartment 12-RC-1312-BB1-10 RH 7E 10.126 Large Yes Hot SG Compartment 12-RC-1322-BB1-1A Sl-ACC-CL3 7N 10.126 Large Yes Cold SG Compartment 12-RC-1322-BB1-2 SI-ACC-CL3 7N 10.126 Large Yes Cold SG Compartment 12-RC-1322-BB1-3 SI-ACC-CL3 7N 10.126 Large Yes Cold SG Compartment 16-RC-1412-NSS-9 Pressurizer 4C 12.814 Large Yes Hot SG Compartment Surge Line 16-RC-1412-NSS-8 Pressurizer 4 11.6767 Large No Hot SG Compartment Surge Line 16-RC-1412-NSS-8 Pressurizer 4B 12.814 Large Yes Hot SG Compartment Surge Line 16-RC-1412-NSS-7 4B 12.814 Large Yes Hot SG Compartment Surge Line 27.5-RC-1103-NSS-1 RC Cold Leg 1 3C 23.8705 Large No Cold SG Compartment 27.5-RC-1103-NSS-1 RC Cold Leg 1 3C 27.5 Large Yes Cold SG Compartment 27.5-RC-1103-NSS-6 RC Cold Leg 1 3C 27.5 Large Yes Cold RX Cavity 27.5-RC-1103-NSS-7 RC Cold Leg 1 3C 20.4323 Large No Cold RX Cavity 27.5-RC-1103-NSS-7 RC Cold Leg 1 3C 27.5 Large Yes Cold RX Cavity 27.5-RC-1103-NSS- RC Cold Leg 1 3A 24.1627 Large No Cold RX Cavity RPV1-N2ASE 27.5-RC-1103-NSS-RC-113NS RC Cold Leg 1 3A 27.5 Large Yes Cold RX Cavity RPV1RN2ASE 27.5-RC-1203-NSS-1 RC Cold Leg 2 3C 27.5 Large Yes Cold SC Compartment 27.5-RC-1203-NSS-4 RC Cold Leg 2 3C 27.5 Large Yes Cold RX Cavity 27.5-RC-1203-NSS-5 RC Cold Leg 2 3C 27.5 Large Yes Cold RX Cavity 25-C10NS- RC Cold Leg 2 3A 26.0648 Large No Cold RX Cavity RPV1-N2BSE 27.5-RC-1203-NSS- RC Cold Leg 2 3A 27.5 Large Yes Cold RX Cavity RPV1-N2BSE Page 252 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Weld Location System Weld Break LOCA DEGB Break Break Cat. Size (in) Size Side Compartment 27.5-RC-1303-NSS-1 RC Cold Leg 3 3C 21.7535 Large No Cold SG Compartment 27.5-RC-1303-NSS-1 RC Cold Leg 3 3C 27.5 Large Yes Cold SG Compartment 27.5-RC-1303-NSS-5 RC Cold Leg 3 3C 27.5 Large Yes Cold RX Cavity 27.5-RC-1303-NSS-6 RC Cold Leg 3 3C 26.9184 Large No Cold RX Cavity 27.5-RC-1303-NSS-6 RC Cold Leg 3 3C 27.5 Large Yes Cold RX Cavity 27.5-RC-1303-NSS- RC Cold Leg 3 3A 27.5 Large Yes Cold RX Cavity RPV1-N2CSE 27.5-RC-1403-NSS-1 RC Cold Leg 4 3C 23.9953 Large No Cold SG Compartment 27.5-RC-1403-NSS-1 RC Cold Leg 4 3C 27.5 Large Yes Cold SG Compartment 27.5-RC-1403-NSS-5 RC Cold Leg 4 3C 21.2034 Large No Cold RX Cavity 27.5-RC-1403-NSS-5 RC Cold Leg 4 3C 27.5 Large Yes Cold RX Cavity 27.5-RC-1403-NSS-6 RC Cold Leg 4 3C 24.1983 Large No Cold RX Cavity 27.5-RC-1403-NSS-6 RC Cold Leg 4 3C 27.5 Large Yes Cold RX Cavity 27.5-RC-1403-NSS- RC Cold Leg 4 3A 23.0525 Large No Cold RX Cavity RPV1-N2DSE 27.5-RC-1403-NSS- RC Cold Leg 4 3A 27.5 Large Yes Cold RX Cavity RPV1-N2DSE 29-RC-1101-NSS- RC-Hot Leg 1 2 22.1613 Large No Hot SG Compartment RSG-1A-IN-SE 29-RC-1101-NSS- RC-Hot Leg 1 2 29 Large Yes Hot SG Compartment RSG-1A-IN-SE 29-RC-1101-NSS-5.1 RC-Hot Leg 1 1B 10.7376 Large No Hot SG Compartment 29-RC-1101-NSS-5.1 RC-Hot Leg 1 1B 21.7244 Large No Hot SG Compartment 29-RC-1101-NSS-5.1 RC-Hot Leg 1 1B 29 Large Yes Hot SG Compartment 29-RC-1101-NSS-4 RC-Hot Leg 1 1B 26.5242 Large No Hot SG Compartment The fact that no small or medium break events have been recorded as failure for any scenario evaluated in this quantification is a strong indication that there is a minimum size break below which insufficient debris can be formed to challenge the safety systems. The same consideration explains why most failure scenarios involve the DEGB assumption of spherical ZOI - simply because more insulation volume can be involved in debris generation.

More detailed information regarding weld location and debris compositions that lead to failure can be extracted from the CASA analysis to help prioritize risk mitigation strategies such as local insulation replacement, additional ISI requirements on risk-dominant welds, plant cleanliness actions, etc.

Page 253 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 7 Conclusions The CASA Grande software provides an integrated framework for comprehensively evaluating GSI-191 phenomena based on plant-specific inputs. The STP evaluation showed that the likelihood of failure associated with GSI-191 for the current plant conditions is very low. Based on the inputs and models used, long-term core cooling was not predicted to fail for any of the small or medium break LOCAs evaluated, and was predicted to fail for only a small fraction of the large break LOCAs. It should be noted that this evaluation included significant conservatisms to address the uncertainties associated with chemical effects.

Some significant plant modifications have been made previously to address GSI-191 concerns-most notably the ECCS strainer replacement project in 2006 and 2007. The STP strainer upgrade was achieved at a cost on the order of $6.3M (96). This modification resulted in a safety improvement where the number of cases predicted to fail decrease by a factor on the order of 100 to 800 times. This reduction very possibly resulted in the desired shift from Regulatory Guide 1.174 Region IIto Region III (70). Safety improvement actions exceeding a factor of 100 are difficult to find in any mature engineered system, and the next proposed GSI-191-related major upgrade involves insulation replacement. The cost and radiological exposure estimates for replacement of insulation are significant, approaching tens of millions of dollars and hundreds of person-Rem per unit, depending on the scope of the modifications.

These comparisons suggest that STP has achieved a condition of decreasing marginal safety benefit where the costs of dramatic design changes may exceed the net risk-benefit that would be gained. With the availability of the coupled CASA Grande/RISKMAN analysis framework, dominant residual risk contributors can now be managed in a fiscally sound manner to progressively and continually improve plant safety in a quantifiable risk-management context.

Page 254 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 8 References

1. Generic Letter 2004-02. PotentialImpact of Debris Blockage on Emergency RecirculationDuring Design Basis Accidents at Pressurized-WaterReactors. September 13, 2004.

2. Memorandum from Andrew Bates (NRC) to Luis Reyes (NRC Staff). Staff Requirements - Briefing on Resolution of GSI-191, Assessment of Debris Accumulation on PWR Sump Performance, 1:30 P.M.,

2006, Commissioners' Conference Room, One White Flint North, Rockville, Maryland(Open to Public Attendance). November 16, 2006.

3. Memorandum from Annette Vietti-Cook (NRC) to R.W. Borchardt (NRC Staff). Staff Requirements -

SECY-10-0113 - Closure Optionsfor Generic Safety Issue - 191, Assessment of Debris Accumulation on Pressurized Water Reactor Sump Performance. December 23, 2010.

4. ALIaN-SUM-WEST-2916-01. CAD Model Summary: South Texas Reactor Building CAD Modelfor Use in GSI-191 Analyses. Revision 3 : November 27, 2012.

5. Texas A&M University Department of Nuclear Engineering. Sump Temperature Sensitivity Analysis.

Revision 2.0 : January 2013.

6. University of Texas at Austin. Sump Temperature as a Function of Time and Break Size. : December 20, 2012.

7. KNF Consulting Services LLC, and Scandpower Risk Management Inc. Development of LOCA Initiating Event Frequenciesfor South Texas ProjectGSI-191 FinalReport for 2011 Work Scope. : September 2011.

8. University of Texas at Austin. Modeling and Sampling LOCA Frequency and Break Size for STP GSI-191 Resolution. : September 2012.

9. Scandpower. Risk Informed GSI-191 Resolution LOCA Frequency Component Database. Revision 2:

October 21, 2011.

10. University of Texas at Austin. Calibrationand Benchmarking of Single- and Two-Phase Jet CFD Models. : May 18, 2012.

11. ALION-CAL-STP-8511-03. STP Qualified Coatings Debris Generation. Revision 0 : August 10, 2012.

12. ALION-CAL-STP-8511-06. STP UnqualifiedCoatings Debris Generation.Revision 2 : November 26, 2012.

13. ALION-CAL-STP-8511-07. STP Crud Debris Generation. Revision 0 : November 12, 2012.

14. ALION-CAL-STP-8511-01. STP Post LOCA Water Volume Analysis. Revision 1 : September 20, 2012.

15. ALION-REP-STPEGS-8221-02. Expected Impact of Chemical Effects on GSI-191 Risk-Informed Evaluationfor South Texas Project. Revision 0 : October 26, 2011.

16. CHLE-008. Debris Bed Preparationand Formation Test Results. Revision 3 : June 12, 2012.

Page 255 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis Rl-GS1191-V03 Revision 1

17. CHLE-010. CHLE Tank Test Results for Blended and NEI FiberBeds with Aluminum Addition. Revision 2 : August 19, 2012.

18. CHLE-012. T1 MBLOCA Test Report. Revision 2 : December 19, 2012.

19. CHLE-014. T2 LBLOCA Test Report. Revision 1 : January 12, 2013.

20. CHLE-016. CalculatedMaterialRelease. Revision 1 : January 10, 2013.

21. Tim Sande (Alion) and Kerry Howe (UNM). Resolution of OutstandingChemical Effects PIRT Issues.

Revision 1 : March 28, 2012.

22. CHLE-015. Summary of Chemical Effects Testing in 2012 for STP GSI-191 License Submittal. Revision 3 : January 21, 2013.

23. ALION-CAL-STP-8511-08. Risk-Informed GSI-191 Debris Transport Calculation.Revision 2 : January 21, 2013.

24. ALION-REP-STP-8511-02. South Texas Vertical Loop Head Loss Testing Report. Revision 1 : January 24, 2013.

25. ALION-CAL-STP-8511-05. STP Net Positive Suction Head Margin. Revision 0 : November 19, 2012.

26. ALION-REP-STP-8511-03. South Texas PenetrationTest Report. Revision 1 : January 24, 2013.

27. Texas A&M University. Bench Top Screen PenetrationTest (Water Type Sensitivity Analysis). Revision 2.0: December 2012.

28. LA-UR-13-20079. ParametricModel of Debris PenetrationThrough Sump Strainers with Concurrent Filtrationand Shedding. :January 2013.

29. Texas A&M University Department of Nuclear Engineering. Core Blockage Thermal-Hydraulic Analysis. Revision 1.0 : November 2012.

30. Technical Specifications Section 3/4.5.1. Accumulators. : Unit 1 Ammendment No. 188; Unit 2 Ammendment No. 175.

31. Technical Specifications Section 3/4.3.2. EngineeredSafety FeaturesActuation System Instrumentation. : Unit 1 Ammendment No. 116; Unit 2 Ammendment No. 104.

32. OPOP05-EO-EO10. Loss of Reactor or Secondary Coolant. Revision 20: April 28, 2011.

33. OPOPOS-EO-ES13. Transfer to Cold Leg Recirculation. Revision 10 : July 1, 2008.

34. OPOPO5-EO-ES11. SI Termination. Revision 14: May 13, 2010.

35. Email from Tim Sande (Alion) to Kerry Howe (UNM) and Ernie Kee (STP). Best-Estimate Time for Spray Operation. : February 23, 2012.

36. OPOPO5-EO-ES14. Transfer to Hot Leg Recirculation. Revision 7 : July 1, 2008.

Page 256 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1

37. NUREG-1829. Estimating Loss-of-Coolant Accident (LOCA) FrequenciesThrough the Elicitation Process. : April 2008.

38. STP-2699325-O-03.

Subject:

On the Frequency of Success States Involving Different Numbers of Pumps Operating. : December 18, 2012.

39. STPEGS UFSAR. Chapter 6.3: Emergency Core Cooling System. Revision 15.

40. MC-6220. SI & CS Pump NPSH. Revision 4 : February 5, 2002.

41. 5N109MB01024. Design Basis Document ContainmentSpray. Revision 3 : November 17, 2004.

42. ALION-CAL-STPEGS-2916-002. GS1 191 Containment RecirculationSump Evaluation:Debris Generation.Revision 3 : October 20, 2008.

43. NEI 04-07 Volume 1. Pressurized Water Reactor Sump PerformanceEvaluation Methodology.

Revision 0 : December 2004.

44. NEI 04-07 Volume 2. Safety Evaluation by the Office of Nuclear Reactor Regulation Related to NRC Generic Letter 2004-02, Nuclear Energy Institute Guidance Report "PressurizedWater Reactor Sump Performance EvaluationMethodology". Revision 0 : December 2004.

45. ALION-REP-ALION-2806-01. Insulation Debris Size Distributionfor Use in GSI-191 Resolution. Revision 4 : May 20, 2009.

46. SFS-STP-PA-7101. South Texas ProjectUnits 1 & 2 Sure-Flow Strainer Module Details. Revision 5:

September 5, 2006.

47. TDI-6005-01. SFS Surface Area, Flow and Volume Calculations.Revision 1 : August 31, 2006.

48. SFS-STP-GA-00. South Texas ProjectUnits 1 & 2 Sure-Flow StrainerGeneral Arrangement. Revision A : September 7, 2006.

49. SFS-STP-PA-7103. South Texas Project Units 1 & 2 Sure-Flow StrainerSections and Details. Revision 2 : August 4, 2006.

50. 2F369PS10572 Sheets 3, 4 & 6. Safety Injection 'SI'.

51. 5L019PS0004. Specificationfor Criteriafor PipingDesign and Installation. Revision 23 : s.n.

52. 66-9088089-000. South Texas Project Test Reportfor ECCS StrainerTesting. Revision 0 : August 29, 2008.

53. EC-PCI-STP-6005-1001. AES Document No. PCI-5473-S01 Rev 2 "StructuralEvaluation of Strainersfor ContainmentEmergency Sumps". Revision 2 : January 7, 2010.

54. EC-PCI-STP-6005-1004. AES Document No. PCI-5473-S03 Rev 0 "StructuralEvaluation of Strainersfor Containment Emergency Sumps for Long Term Post LOCA Case". Revision 0 : January 7, 2010.

55. TDI-6005-07 (STP Document 0415-0100057WN). Vortex, Air Ingestion & Void FractionSouth Texas Project Units 1 & 2. Revision 3 : November 17, 2008.

Page 257 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1

56. WCAP-16631-NP, Volume 1. Testing and Evaluationof Gas Transport to the Suction of ECCS Pumps.

Revision 0 : October 2006.

57. DRN 0250-0019-14 (STP Document VTD-G927-0001). Units 1 and 2 Acceptable Gas Void Volumes in ECCS and RHR Suction Piping. Revision 1 : October 25, 2011.

58. Regulatory Guide 1.82. Water Sources for Long-Term RecirculationCooling Following a Loss-of-Coolant Accident. Revision 4 : March 2012.

59. University of Texas at Austin. Filtrationas a Function of Debris Mass on the Strainer:Fitting a ParametricPhysics-BasedModel. s.l. : June 5, 2013.

60. SN079NB01000 (WCAP-12381). STPNOC Design Basis Document Accident Analysis. Revision 15 : July 29, 2009.

61. Technical Specifications Section 1.27. Rated Thermal Power. Unit 1 Amendment No. 154; Unit 2 Amendment No. 142 : s.n.

62. WCAP-16793-NP. Evaluationof Long-Term Cooling Considering Particulate,Fibrous and Chemical Debris in the RecirculatingFluid. Revision 2 : October 2011.

63. Technical Specifications Section 5.3. Reactor Core. Unit 1 Amendment No. 104, Unit 2 Amendment No. 91.

64. ALION-CAL-STPEGS-2916-005. GSI-191 Containment RecirculationSump Evaluation:CFD Transport Analysis. Revision 3 : October 21, 2008.

65. NUREG/CR-6808. Knowledge Base for the Effect of Debris on PressurizedWater ReactorEmergency Core Cooling Sump Performance. : February 2003.

66. Email from Larry Jones (STP) to Tim Sande (Alion) and Ernie Kee (STP). RE: Switchover to Hot Leg Injection. : December 6, 2012.

67. LA-UR-99-3371. Pressurized-Water-ReactorDebris Transport in Dry Ambient Containments--

Phenomena Identificationand Ranking Tables (PIRTs). Revision 2 : December 1999.

68. McCain, William D. The Propertiesof Petroleum Fluids. 2nd Edition : PennWell Publishing Company, 1990.

69. Website http://en.wikipedia.org/wiki/Compressibility_factor. Compressibility Factor.s.. :

Retrieved August 7, 2012.

70. Regulatory Guide 1.174. An Approachfor Using ProbabilisiticRisk Assessment in Risk-Informed Decisions on Plant-Specific Changes to the Licensing Basis. Revision 2 : May 2011.

71. University of Texas at Austin. A Frameworkfor UncertaintyQuantification:Methods, Strategies, and an Illustrative Example. : January 21, 2013.

72. ALION-REP-STP-8511-04. Verification of CASA Grande Calculations.Revision 0 : January 25, 2013.

Page 258 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1

73. WCAP-16530-NP. Evaluation of Post-Accident Chemical Effects in Containment Sump Fluids to Support GSI-191. Revision 0 : February 2006.

74. 0415-0100022WN (WES010-CALC-001). South Texas ProjectPost-LOCA Containment Water Level Calculation.Revision A: November 24, 2008.

75. NUREG/CR-6224. ParametricStudy of the PotentialforBWR ECCS StrainerBlockage Due to LOCA GeneratedDebris. : October 1995.

76. Duke Energy. GSI-191: The Effect of Water Type on Debris Bed Head Loss Across ECCS Sump Strainer Modules. : July 2011.

77. Generic Letter 2008-01. Managing Gas Accumulation in Emergency Core Cooling, Decay Heat Removal, and ContainmentSpray Systems. : January 11, 2008.

78. NUREG/CR-2758. A ParametricStudy of Containment Emergency Sump Performance. : July 1982.

79. NUREG/CR-2759. A ParametricStudy of Containment Emergency Sump Performance: Results of Vertical Outlet Sump Tests. : October 1982.

80. NUREG/CR-2760. Assessment of Scale Effects on Vortexing, Swirl, and Inlet Losses in Large Scale Sump Models. : June 1982.

81. NUREG/CR-2761. Results of Vortex Suppressor Tests, Single Outlet Sump Tests and Miscellaneous Sensitivity Tests. : September 1982.

82. NUREG/CR-2772. Hydraulic Performanceof Pump Suction Inlets for Emergency Core Cooling Systems in Boiling Water Reactors. : June 1982.

83. Website http://en.wikipedia.org/wiki/Earth'satmosphere. Earth'sAtmosphere. : Retrieved July 20, 2012.

84. Lide, David R. Handbook of Chemistry and Physics. 75th Edition : CRC Press, 1994.

85. Harvey, Allan H. Semiempirical Correlationfor Henry's Constants over Large Temperature Ranges.

Pages 1491-1494 : AIChE Journal, Vol. 42, No. 5, May 1996.

86. 0415-0100007WN (WES006-PR01). Evaluationof Containment RecirculationSump Downstream Effects for STPEGS. Revision A: March 28, 2006.

87. 0415-01000011WN (CN-SEE-05-76). STP Sump Debris Downstream Effects Evaluationfor ECCS Equipment. Revision C: December 11, 2008.

88. 0415-0100012WN (CN-SEE-05-73). STP Sump Debris Downstream Effects Evaluationfor ECCS Valves.

Revision A: March 28, 2006.

89. WCAP-16406-P. Evaluation of Downstream Sump Debris Effects in Support of GSI-191. Revision 1:

August 2007.

Page 259 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1

90. Letter from James Gresham (NRC) to Gordon Bischoff (Westinghouse). FinalSafety Evaluationfor Pressurized Water Reactor Owners Group (PWROG) Topical Report (TR) WCAP-16406-P, "Evaluation of Downstream Sump Debris Effects in Support of GSI-191 ", Revision 1 (TAC No. MD2189).

December 20, 2007.

91. 0415-0100086WN (CN-SEE-1-08-66). South Texas ProjectLOCADM. Revision A : December 9, 2008.

92. Nuclear Regulatory Commission Official Transcript of Proceedings. Advisory Committee on Reactor Safeguards Thermal Hydraulic Phenomena Subcomittee. : September 23, 2008.

93. Drawing 6488E72 Sheet 1 of 4. South Texas Nuclear Power Plant Replacement Steam Generator GeneralArrangement.Revision A : August 31, 1999.

94. WCAP-17047-NP. Phenomena Identification and Ranking Tables (PIRT) for Un-Buffered/Buffered Boric Acid Mixing/Transportand PrecipitationModes in a Reactor Vessel During Post-LOCA Conditions. Revision 0: May 2009.

95. SEC-LIS-3908-C4. Hot Leg Switchover Analysis to Support RSG Program. Revision 0 : July 16, 1997.

96. Email from Wes Schulz (STP) to Tim Sande (ENERCON). STP Sump costs. : January 21, 2013.

97. MC-5758. Containment Emergency Sump Protective Screen - Total and Open Area. Revision 0:

October 28, 1984.

98. SFS-STP-GA-O0. South Texas Project Units 1 and 2 Sure Flow StrainerGeneralArrangement. Revision A : September 7, 2006.

99. Email from Dave Morton (UT Austin) to Tim Sande (Enercon). RE: penetrationanalysis. : January 15, 2013.

Page 260 of 260

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 Appendix 1 The following pages document a typical input deck for CASA Grande that was used to generate the quantification discussed in this report. A detailed explanation is not provided here, but variable names are largely mnemonic, and the information is grouped in associated blocks, so even a casual reader will recognize much of the information and the intent of the data. A reader with some experience with MATLAB programming constructs will immediately recognize the syntax used to define lists (vectors) and tables (arrays) of input values.

Many of the figures presented in the main body of this report can be reproduced using values specified in this input deck. Of particular interest are the random values that that can be sampled during the course of each analysis. These are highlighted in yellow, but not all of these options were exercised. It is relatively trivial to add new random variables to the analysis, so eventually most (if not all) of the input parameters will be handled generically as random variables. When point values are desired, the standard deviations of the associated probability distributions are set equal to 0 so that the probability distribution collapses to a point and random sampling simply returns the mean for every sample.

The example input deck illustrates that a CASA analysis is data intensive, but all of the input information derives from engineering calculations and from test data that are readily available to plant engineers and analysts. There is very little information here that would not be found in a detailed hand calculation of similar complexity.

Page 1-1 of 1-16

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 T~ l-1 t

c casefolder = ' C: \sers\Us(: r\My D( uments\MaLab\ ,th Tehxas Project \Casa Grande\South Texas Projec t analysisfolder 'Analyt ic Relsu ts';

runsubfolder 'PRA FIailure Proh - Jan 2013';

runsubsubfolder = 'Parametei Variations';

runsubsubsubfolder = 'ThemFac Figs';

cadfolder = 'CAD Files';

concretesubfolder = 'Concrete cData';

equipsubfolder = 'Equip Data';

gratingsubfolder = 'Grat ing Data';

pipesubfolder = 'Pipe Data';

freqfolder = 'Freq Data';

breakfreqfile = 'LOCA Frequency and Weld Inpu ts - ]2-"-12 Rl.xisx';

breakfreqtab ='LOCA Data';

weldcasefile 'LOCA Frequency and Weld Inpu t S - 12-7-12 RI.xlsx';

weldcasetab = 'Weld Table';

unitconcrete = 1.0; unitequip = 1.0; unitgrate = 1.0; unitpipe = 1.0; i* [t :, -s~s ye t with LOCAbin = [0.5 2.0 6.0];

icad = [0; 0];

intro = 0; iplt = [0; 0;

0];

ZOIzoom = 50; ZOIintvl 25; 1/2 T~--~ 7,-i Page 1-2 of 1-16

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 1fies , I r ' UI N, .-, 4 ,4 41 delL = 6; .r r Ii' ir Nangbin = 12; -~ V 2. 4]

Por opt = 1; pF I'-I F1d/

glht 4n

rsCr 1r 1 II - ýs Ii t Q: DI t t 1 F-4 4 AN Ci N 1- C , NA 4 1 4-.i .si N 1no no - J w -,N a valvelabel - {'Valve' 'VALVE' 'MOV' 'XRH' 'FCV'};

hangarlabel = {'Hangar' 'Hanger' 'HL' 'AF 'GU' 'SS' 'SHI' 'RR' 'RHI' ...

'Work Point' 'work point' 'Work point' 'work Point' 'WORK POINT'};

nonstdweldlabel = {'FW' 'Weld' 'WELD' 'FS'};

r m IFT fi. I i haN all points in o ecomponentw 4 hv sae d toý S.yst(eTm' as fi st t xt strIn: in fieldl SGlabel {'SG' 'Steamienerator'};

RCPlabel = {'RCP' 'ReactorCoolantPumnp'};

PZRlabel = {'PZR' 'PRZR' 'Pressurizer'};

RHRlabel = {'RHR' 'ResidualHeatRemoval'};

~'F lilAN 4 setrand = 2; setrandfile = NN~ Al ,,I-1; -f 4-Nmaxbrk = 5; fIt I L~- 4 I ii L.LOCAl",

t t44 Flrr<vl r 4

.4 N I 4~. .:~A I-,,r k., s 1-ir ..2-~4 4 Mrep = 3; 44 III 4 0 - 14 1 Frep = 5; 4 i -4 t- I 41' 44 Fbase = 2; FuprPtile=0. 999; 4 h 44intepl Nfreqpts = 1000; BSbase = 10; 14- e4 j~ 1' "1 4oa it h ,,

N - 4 .4 F 244 ' 44 4 4 I +~ F..4-a4<w4 44a,.44'N 44 44 ..4 44 14 N'- - 4 i4, 4 .44 4-Page 1-3 of 1-16

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSl191-V03 Revision 1

{j II ] ry - i e,r ai s daaiq o R/ D tuppe

)

Rldfg = [7 11.9 17];

Pldfg = [.2 .13 .08;

.8 .54 .07;

.0 .16 .41; I )r !

.0 .17 .44];

ZOT dsYirir f roperT i ets trgtlabel = {'NUKON'; 'NUKON '; 'MICROTHERM';

'RMI'; 'LEAD'; 'THERMAL WRAP'; ...

'IOZ'; 'ALFYD'}; - 9-_7 I <,l- r I iLDFG =[1 1 0 0 0 1 0 0];

iMicroTh = [0 0 1 0 0 0 0 0];

Rdmg = 17 0 0 99999 1 0 0; ~oTIo'Th 17 0 0 99999 1 0 0; S I 28.6 0 0 99999 1 0 0; 1 0 0 99999 1 0 0; 1 0 0 99999 1 0 0; 17 0 0 99999 1 0 0; 1 0 0 99999 1 0 0; 1 0 0 99999 1 0 0];

1y 9 n .i-i .ii f it be Y t[ in riL 11< t Cv r t heI Jia[1i 1ý 1"1<'ch ý t91 5- >79I[199

' I at 'D/Iaým' 'Rt, native cal(: ll111*il AI u , I'bll ,f ' ' Ii l/ f t1 i ' ' kg/imP P, DebrisProp = { {'LDFG - fines' 2 7 175 2

{'LDFG - small' 2 2.78e5 175 2.41;

{'LDFG - large' 2 5.56e5 175 2.41;

{'uTherm - filaments' 6 165 15 };

{'uTherm - SiO2' 1 20 137 27.4};

{ 'uTherm - TiO2' 1 2.5 262 52.41;

{'QualCoat - epoxy' 1 10 94 6.661;

{'QualCoat - IOZ' 1 10 208 1.121; f Crud' 1 15 350 70.0};

Page 1-4 of 1-16

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1

{' UQCoat - epoxyfine' 1 6 124 8.36};

{'UQCoat - epoxyFchp' 1 1.25e4 124 8.36};

{'UQCoat - epoxySchp' 1 1.25e4 124 48 .36};

{'UQCoat - epoxyLchp' 1 1.25e4 124 48. 36};

{'UQCoat - epoxyCrls' 1 1.25e4 124 48 .36};

{'UnQualCoat - alkyd' 1 10 207 80. 73};

{'UnQualCoat - enamel' 1 10 93 36. 27};

{'UnQualCoat - IOZ' 1 10 244 95. 16};

{'Latent - particulate' 1 17.3 169 33. 80}.

{'Latent - fiber' 2 7 175 2.4} };

MI f I' C FI I rm col Is titueI I f w i t uThRhornfc = 15; Jd s uThFfiber = 0.03; mass ffac' I bil/f t3 uThFSiO2 = 0.58; 1f uThFTiO2 = 0.39; ma [sf de , r i s ss a r t a nd s oi p t rimeS mi '

i1f del s 1 bro iers P. 1 I- 1 T1 Tstartsrc [ 0 0 0 0 0 0 0 0 0 10 10 10 10 10 10 10 10 0 0];

Tendsrc = [10 10 10 10 10 10 10 10 10 2.3e4 2.3e4 2.3e4 2.3e4 2.3e4 2.3e4 2.3e4 2.3e4 10 10];

-' 5 thi f If <II i- i hr1 .

.1 ni Ks ]<, Iin debrsIL qua ti ie u fc i*IS aYe pre sent1y uncf QualEpoxy = [105 0 0 99999 1 0 0];

QualIOZ = [39 0 0 99999 1 0 0];

CrudFines = [24 0 0 99999 1 0 0]; KI hisill; UnQualEpxyFine = [234 0 117 234 2 0 0 117 234 0.5 0.5];

it UnQualEpxyFchp = [709 0 355 709 2 0 0 355 709 0.5 0.5];

UnQualEpxySchp = [180 0 90 180 2 0 0 ... mam hV 90 180 0.5 0.5]; 1un for Ifua

UnQualEpxyLchp = [391 0 196 391 2 0 0 196 391 0.5 0.5];

UnQualEpxyCrls = [391 0 196 391 2 0 0 luIf t~

196 391 0.5 0.5];

UnQualAlkyd = [271 0 0 99999 1 0 0]; 4nu 1Kepo UnQualEnml = [267 0 0 99999 1 0 0];

UnQualIOZ = [369 0 0 99999 1 0 0];

1 ff LatentParts = [170 0 0 99999 1 0 0];

LatentFiber = [12.5 0 0 99999 1 0 0];

Page 1-5 of 1-16

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 PIIi YIL ýný ý , " rI ,L i T

t;Iti .11qIi u iI ckpreýbb aF I I1,s Ie I f 11 w I t t for Tmax =36; Trecirc = [1.5 2 4 6 8 12 27.5; ime 337 79 56 44 38 31 30];

Tsproff {[ 0 0 0 99999 1 0 0]; ... (lLSpeI r ii 20 5 0 99999 1 0 0]; ... I~ ~ pr I IifI[pI 1

171Ub il n 20 5 0 99999 1 0 0]};

Tallspr {[390 5 390 420 1 0 01; ... IT Si

[390 10 390 420 1 0 0]; ...

[390 15 390 450 1 0 0]};

Ttrnoff {[99999 0 0 99999 1 0 0]; ... 1. ,, - I reý raiin MiII)

IS',v, I')

[99999 0 0 99999 1 0 0]; ... pI.

4 , n evU> Vpp-'en keep a,-

[99999 0 0 99999 1 0 0]}; i FýeýsJ a Oun!2S Tchem = {[0 0 0 99999 1 0 0]; ...

[0 0 0 99999 1 0 0];

[0 0 0 99999 1 0 0]);

Thlinject = {[360 0 345 360 2 0 0 b ýiT 345 360 0.5 0.5];

[360 0 345 360 2 0 0 ...

345 360 0.5 0.51;

[360 0 345 360 2 0 0 ...

345 360 0.5 0.5]};

I,-* d r I es ChemTemp = [140 5 0 99999 1 0 0];

<<_* or W;e t Tl hel IL T<

I I i r I ChemBump = {[1.25 1/1.25^2 1 15.3 3 1 101; .

[1.50 1/1.50^2 1 18.2 3 1 10 ...

I I -Is tl11-1 2 0.5 0.5]; ...

[2.00 1/2.00-2 1 24.0 3 1 10 ...

1 10 1 0.5]};

S i 'IY I $--QUvI FbrpFAblockHL [99999 0 0 0 1 0 0 ... F I I~k5 17 60 85 1 31 ;

FbrpFAblockCL [99999 0 0 0 1 0 0 ...

140 165 1 3];

Page 1-6 of 1-16

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSl191-V03 Revision 1 I ' , I11 FA.l fI b I , s rn d FbrpFAboronHL [99999 0 0 0 1 0 0 .

0 0 0 0];

I. 7t FbrpFAboronCL = [7.5 0 0 0 1 0 0 ...

7.5 20 1 31; Pbuckle = [9.35 0 0 0 1 0 0 ... I i 1 1ý uIick 9 9.5 0.5 0.5];

GasLimit = [0.02 0 0 0 1 0 0 ...

0.02 0.03 3 1];

~rI -S- J 1'~

/ ) i ~t 1(~ L +1 1 I I '1 Ill I I~1 I~T1~

1r' s r) Mp t ýI PrJr 1; r 1 Irlt

<z~

[: ki

]FS 1,, I tl oi PlantState I:S J sIF ay

[0 1 1; 1 1 1; 0 0 0];

II; trai IfI 1t 1 a HSI ] la j s ?6

[1r, NRCP = 4; I ~ ~ 2 F-NPZR = 1; ii <F Page 1-7 of 1-16

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 NRHR = 3; NSG = 4; delT = 5; r : I I Ai Amisc = [100 0 0 99999 1 0 0];

ontaiJ et-Foverlap = .25; f on clebi ielIap ThinBed : 1/16; ClipZOI = 1; A 7A)J A P hL Aid II 1/0 /n) gcore = 1.0; Hstrain = 39/12; n 1ri r um I-~

phiO = 1.00; phil = 1.00; f IF, 1 ,

Nfuel = 193; CrankdelP = [5 1 1 10 1 0 0];

mEiAA f I w n(eeded to cool core Qboiloff= [0.0028 0.0042 0.0056 0.0111 0.0167 0.0222 0.0278..

I AAAA 0.0417 0.0556 0.1111 0.1667 0.2222 0.277.8 0.41 67 ...

0.5556 1.1111 1.6667 2.2222 2.7778 4.166 7 5.55.56 ...

11.1111 16.6667 22.2222 27.7778 41.6667 111.1 ill ...

166.6667 222.2222 277.7778; ...

1414.7000 1323.5000 1260.9000 1113.4000 1030.5000 ...

973.3000 931.3000 859.3000 812.3000 711.0000 ...

654.7000 614.2000 581.8000 523.1000 480.9000 ...

388.1000 342.4000 315.1000 295.7000 265.1000 ...

245.5000 204.2000 182.7000 168.7000 158.1000 ...

139.8000 99.0000 84.1000 74.4000 67.7000];

specialweld '31-RC-1102-NSS-i. 1';

U!,-

VPool = {[0 0 0 0 2 0 0 ...

AlA- I I1 43464 61993 .5 .5]; . 'A

[0 0 0 0 2 0 0 ...

39533 69444 .5 .5]; ...

[0 0 0 0 2 0 0 ...

45201 69263 .5 .5]}; LariIL If> f Apool = 12301; Aclean = 1.8185e+003; C. C r, 'c' 0ii1 Iw z<"u i I dPclean = 0.22;

mq r u, (,ut j if,ý r l s fqo,,ýIm Qhpsimax 1620; : - r Qlpsimax 2800; I u Qsprymax {[0 0 1932 2600 2 0 0 ... -o' . A 1932 2600 0.5 0.5]; ...

Page 1-8 of 1-16

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1

[0 0 1932 2600 2 0 0 ...

1932 2600 0.5 0.5]; ...

[0 0 1932 2600 2 0 0 ...

1932 2600 0.5 0.5] };

r al. 1i S iKI~flf YI iS r E.LaI' V K'-t

-i 71K7 17K K t 41-:1K itable = 1; it>'< ht aK if itable Gtable I 0 0 1.8185e+003 8.1790e+001 5.00OOe-001 4.1900e+002

8. 1800e+001 5.01OOe-001 4.1931e+002

2. 8016e+002 8.1421e+000 4.4718e+002

4. 7853e+002 1.5783e+001 5.9256e+002

6. 7689e+002 2.3424e+001 7.4768e+002 8 .7526e+002 3. 1065e+001 9.1253e+002

1. 0736e+003 3. 8706e+001 1.0871e+003

1. 2720e+003 4. 6348e+001 1.2714e+003

1. 4703e+003 5.3989e+001 1.4655e+003

1. 6687e+003 6. 1630e+001 1.6692e+003

1. 8671e+003 6. 9271e+001 1.8827e+003

2. 0654e+003 7. 6912e+001 2.1060e+003 2 .2638e+003 8. 4553e+001 2. 3389e+003

2. 4622e+003 9 .2194e+001 2. 5816e+003

2. 6605e+003 9. 9835e+001 2.8341e+003

2. 8589e+003 1. 0748e+002 3. 0962e+003 3.0573e+003 1. 1512e+002 3. 3681e+003 3 .2556e+003 1 .2276e+002 3.6497e+003

3. 4540e+003 1. 3040e+002 3. 9411e+003

3. 6524e+003 1. 3804e+002 4 .2422e+003 3 .8507e+003 1. 4568e+002 4. 5530e+003

4. 0491e+003 1. 5332e+002 4. 8735e+003 Page 1-9 of 1-16

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 4.2474e+003 1.6096e+002 5.2038e+003 4.4458e+003 1.6860e+002 5.5438e+003 4.6442e+003 1.7625e+002 5.8935e+003 4.8425e+003 1.8389e+002 6.2530e+003 5.0409e+003 1.9153e+002 6.6222e+003];

else

. 1 5e p ate p!a ap ro ,o o].d s! Ya",

Vgeom = linspace(0,2500,50)';

Ageom = Aclean*ones(size(Vgeom));

xgeom = Vgeom./Ageom*12; Gtable = [Vgeom xgeom Ageom];

end 17 ri ea

1. wJ it I ,]* :

171 7 117 1< 1~

~1 arr ula 1717 by s11ze, ea :1h ('crVi'. - i -, s brkfrqtitle 'Present-Day Exceedance Frequency';

tablebreak = [1/2 1+5/8 2 3 6 7 14 31] ';

tablefreq = [6.8e-5 5.0e-6 3.69e-6 2.1e-7 6.30e-8 1.4e-8 4.1e-10 3.5e-11; 6.3e-4 8.9e-5 6.57e-5 3.4e-6 1.08e-6 3.1e-7 1.2e-08 1.2e-09; 1.9e-3 4.2e-4 3.10e-4 1.6e-5 5.20e-6 1.6e-6 2.0e-07 2.9e-08; 7.1e-3 1.6e-3 1.18e-3 6.1e-5 1.98e-5 6.1e-6 5.8e-07 8.le-08]';

tablepval = [0.05 0.5 NaN 0.95];

JohnsonParam = [1.650950E+00 5.256964E-01 4.117000E-05 1.420000E-02; 1.646304E+00 4.593913E-01 2.530000E-06 3.200000E-03; 1.646308E+00 4.593851E-01 1.870000E-06 2.360550E-03; 1.646605E+00 4.589467E-01 1.200000E-07 1.220000E-04; 1.646403E+00 4.566256E-01 3.000000E-08 3.965000E-05; 1.645739E+00 4.487957E-01 6.023625E-09 1.220000E-05; 1.645211E+00 3.587840E-01 2.892430E-10 1.160000E-06; 1.645072E+00 3.343493E-01 2.636770E-11 1.600000E-07];

47<74KV ~<7 Atst = 91.44; Page 1-10 of 1-16

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1 nu = [0 0 0 0 2 0 0 ...

0.00956 0.0272 0.5 0.5];

eta = [0 0 0 0 2 0 0 ...

0.008236 0.0546 0.5 0.5];

mfilt = [0 0 0 0 2 0 0 ...

0.000339 0.003723 0.5 0.5 Mcut : [0 0 0 0 2 0 0 ...

790 880 0.5 0.5]; njjlI 1f r 1m IT 1 bfilt = [0 0 0 0 2 0 0 ...

0.656 0.706 0.5 0.51; Itp afilt = [1 0 0 0 1 0 0]; -j Im-al <ii I'-

X2 <66Th dfilt = [0 0 0 0 2 1 0 ... file 0.0011254 0.0013078 0.031 0.10000 0.45000 0.1000];

e< Fa6:tctor s e 1 >S&eI at i \'v VIf ii r e dereFai svue her

{LDL L Iies FG,s I . , 1'tF(G , 1vi1 hefl I'll I i EýS , 6JItjI 2 "!, I f1 lic1>5, Tw AlI WAYSLe1

-VU If ist1 l\Ll1JI S - e ordes i t 1t") I 11-DTFzoi = .

[0.70 0.60 0.22 0.70 0.70 0.70; 0.30 0.25 0.00 0.30 0.30 0.30; 0.53 0.27 0.00 0.53 0.53 0.53; 0.47 0.19 0.00 0.47 0.47 0.47; 0.00 0.27 0.00 0.00 0.00 0.00; E7 1

~

0.00 0.00 0.00 0.00 0.00 0.00; 0.02 0.00 0.00 0.02 0.02 0.02; F II 0.05 0.00 0.00 0.05 0.05 0.05; 1.00 0.64 0.00 1.00 1.00 1.00; 1.00 0.64 0.00 1.00 1.00 1.00; F W' 1.00 0.58 0.00 1.00 1.00 1.00; 0.00 0.01 0.01 0.00 0.00 0.00; 0.00 0.07 0.07 0.00 0.00 0.00];

1 Ii>>

6 I, 1 1 <I6 11.1 t Al

66. I i Page 1-11 of 1-16

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 DTFuc = ...

[0.80 0.80 0.80 0.80 0.80 0.43 0.43 0.92; 0.15 0.15 0.15 0.15 0.15 0.54 0.00 0.83; 0.02 0.02 0.02 0.02 0.02 0.46 1.00 0.17; 0.83 0.83 0.83 0.83 0.83 0.00 0.00 0.00; 0.06 0.06 0.06 0.06 0.06 0.06 0.00 0.06; 1.00 0.21 0.00 0.00 1.00 1.00 1.00 1.00; 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00];

I ALV<~ r ~

pF(~- LI ~'L1tIJ(& 1/4 I DTF1d 0.00; V 1.00 1.00; b F-1.00 1.00; 0.02 0. 02; I- - - F 0.05 0.05; Iz~

1.00 1.00] ; F -~

I I iF K hi I I LIII I I I it v-. § I~I I Ith ,7 Ttemp -[0.0000 r 0 6 0ti.088 f I 0.089 TtempS =[0.0000 0.0847 0.0864 0.0881 0.0897 0.0914 0.0931 0.0947 ...

0.0964 0.0981 0.0997 0.1014 0.1031 0.1047 0.1064 0.1081 0.1097 0.1139 0.1306 0.1472 0.1639 0.1806 0.1972 0.2139 0.2306 0.2472 0.2639 0.2806 0.2972 0.3139 0.3306 0.3472 0.3639 0.3806 0.3972 0.4139 0.4306 0.4472 0.4639 0.4806 0.4972 0.5139 0.5306 0.5472 0.5639 0.5806 0.5972 0.6139 0.6306 0.6472 0.6639 0.6806 0.6972 0.7139 0.7306 0.7472 0.7639 0.7806 0.7972 0.8139 0.8306 0.8472 0.8639 0.8806 0.8972 0.9139 0.9306 0.9472 0.9639 0.9806 0.9972 1.0139 1.0306 1.0472 1.0639 1.0806 1.3611 1.6944 2.0278 2.3611 2.6944 3.0278 3.3611 3.6944 4.0278 4.3611 4.6944 5.0278 5.3611 5.6944 6.0278 6.3611 6.6944 7.0278 7.3611 7.6944 8.0278 8.3611 8.6944 9.0278 9.3611 9.6944 10.0278 20.0833 32.0833 44. 0833 56 0833 68.0833 80.0833 92. 0833 104.0833 ...

116.0833 128.0833 140.0833 152.0833 164.0833 ..

176.0833 188.0833 200.0833 212.0833 224.0833 ..

236.0833 248.0833 260.0833 272.0833 283.3333 ..

297.2222 308.3333 319.4444 333.3333 344.4444 ..

355.5556 369.4444 380.5556 391.6667 402.7778 ..

416.6667 427.7778 438.8889 452.7778 463.8889 ..

Page 1-12 of 1-16

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GS1191-V03 Revision 1 475.0000 488.8889 500.0000 511.1111 525.0000 ...

536.1111 547.2222 561.1111 572.2222 583.3333 ...

597.2222 608.3333 619.4444 633.3333 644.4444 ...

655.5556 669.4444 680.5556 691.6667 702.7778 ...

716.6667];

T-m ipS r I [1 t TempS = [11I9.6000 131.2987 140.1689 150.3314 156.1240 ...

159.2343 162.1567 164.5680 166.6937 168.5685 170.2457 171.7175 172.9577 174.0415 174.9570 175.7084 176.3081 177.5299 164.4935 132.7076 124.0848 123.6914 123.5988 123.5641 123.5529 124.4938 127.6399 129.7484 131.0391 149.8002 158.2393 162.7694 165.4960 167.3851 168.6688 169.7687 170.9814 171.9993 172.8771 173.7150 174.4595 175.0903 175.6074 176.0061 176.2923 176.4625 176.4855 176.3916 176.2055 175.9468 175.6184 175.2411 174.8243 174.3902 173.9374 173.4284 172.8459 172.2319 171.6143 171.0143 170.4548 169.9507 169.5034 169.1086 168.7661 168.4824 168.2551 168.0847 167.9707 167.9020 167.8705 167.8665 167.8947 167.9451 168.0131 168.0978 170.0607 170.9606 171.4105 170.8721 169.8110 168.7942 168.1132 165.3090 164.1228 163.0112 161.4436 159.9385 158.1298 158.4517 156.5706 151.6937 163.7090 160.9624 158.1118 156.1579 154.6151 153.2333 151.9641 150.8191 149.7667 148.7924 147.8649 136.2080 129.0230 124.9790 122.1450 120.1310 118.4710 117.3160 116.4980 115.6160 114.7100 113.8960 113.1730 112.5210 111.9240 111.3580 110.8590 110.3930 109.9930 109.5770 109.2090 108.9100 108.5930 108.2810 107.9680 107.7100 107.4730 107.1620 106.9430 106.7150 106.4770 106.2500 106.1240 105.8930 105.6660 105.5410 105.3160 105.1930 105.0690 104.8440 104.7250 104.6070 104.3770 104.3660 104.1400 104.0230 103.9050 103.7910 103.6730 103.5660 103.4520 103.3350 103.1450 103.1000 102.9130 102.8680 102.6810 102.6450 102.5250 102.5160];

LAi w--

ý Jrrrur i ti oI iii K ~stq '~sr -.~ rr ft I

'IK

~

P~ ( '.IF >ýSL~ 4p TtempM =[0. 0000 0.0847 0.0864 0.0881 0.0897 0.0914 0.0931 0.0947 ...

0.0964 0.098 1 0.0997 0.1014 0.1031 0.1047 0.1064 0.1081

0. 1097 0.11339 0.1306 0.1472 0.1639 0.1806 0.1972 0.2139 0.2306 0.247 2 0.2639 0.2806 0.2972 0.3139 0.3306 0.3472 0.3639 0.380)6 0.3972 0.4139 0.4306 0.4472 0.4639 0.4806

0. 4972 0.51339 0.5306 0.5472 0.5639 0.5806 0.5972 0.6139 0.6306 0.647 2 0.6639 0.6806 0.6972 0.7139 0.7306 0.7472 0.7639 0.780)6 0.7972 0.8139 0.8306 0.8472 0.8639 0.8806 0.8972 0.91339 0.9306 0.9472 0.9639 0.9806 0.9972 1.0139 Page 1-13 of 1-16

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 1.0306 1.0472 1.0639 1.0806 1.3611 1.6944 2.0278 2.3611 2.6944 3.0278 3.3611 3.6944 4.0278 4.3611 4.6944 5.0278 5.3611 5.6944 6.0278 6.3611 6.6944 7.0278 7.3611 7.6944 8.0278 8.3611 8.6944 9.0278 9.3611 9.6944 10.0278 20.0833 32.0833 44.0833 56 0833 68.0833 80.0833 92 .0833 104.0833 ...

116.0833 128.0833 140.0833 152.0833 164.0833 176.0833 188.0833 200.0833 212.0833 224.0833 236.0833 248.0833 260.0833 272.0833 283.3333 297.2222 308.3333 319.4444 333.3333 344.4444 355.5556 369.4444 380.5556 391.6667 402.7778 416.6667 427.7778 438.8889 452.7778 463.8889 .

475.0000 488.8889 500.0000 511.1111 525.0000 .

536.1111 547.2222 561.1111 572.2222 583.3333 .

597.2222 608.3333 619.4444 633.3333 644.4444 .

655.5556 669.4444 680.5556 691.6667 702.7778 .

716.6667];

% Temperature(F) profile for medium breaks TempM = [119.6000 131.2987 140.1689 150.3314 156.1240 ...

159.2343 162.1567 164.5680 166.6937 168.5685 .

170.2457 171.7175 172.9577 174.0415 174.9570 .

175.7084 176.3081 177.5299 164.4935 132.7076 .

124.0848 123.6914 123.5988 123.5641 123.5529 .

124.4938 127.6399 129.7484 131.0391 149.8002 .

158.2393 162.7694 165.4960 167.3851 168.6688 .

169.7687 170.9814 171.9993 172.8771 173.7150 .

174.4595 175.0903 175.6074 176.0061 176.2923 .

176.4625 176.4855 176.3916 176.2055 175.9468 .

175.6184 175.2411 174.8243 174.3902 173.9374 .

173.4284 172.8459 172.2319 171.6143 171.0143 .

170.4548 169.9507 169.5034 169.1086 168.7661 .

168.4824 168.2551 168.0847 167.9707 167.9020 .

167.8705 167.8665 167.8947 167.9451 168.0131 .

168.0978 170.0607 170.9606 171.4105 170.8721 .

169.8110 168.7942 168.1132 165.3090 164.1228 .

163.0112 161.4436 159.9385 158.1298 158.4517 .

156.5706 151.6937 163.7090 160.9624 158.1118 .

156.1579 154.6151 153.2333 151.9641 150.8191 .

149.7667 148.7924 147.8649 136.2080 129.0230 .

124.9790 122.1450 120.1310 118.4710 117.3160 .

116.4980 115.6160 114.7100 113.8960 113.1730 .

112.5210 111.9240 111.3580 110.8590 110.3930 .

109.9930 109.5770 109.2090 108.9100 108.5930 .

108.2810 107.9680 107.7100 107.4730 107.1620 .

106.9430 106.7150 106.4770 106.2500 106.1240 ..

105.8930 105.6660 105.5410 105.3160 105.1930 ..

105.0690 104.8440 104.7250 104.6070 104.3770 ..

104.3660 104.1400 104.0230 103.9050 103.7910 ..

103.6730 103.5660 103.4520 103.3350 103.1450 ..

103.1000 102.9130 102.8680 102.6810 102.6450 ..

102.5250 102.5160];

Page 1-14 of 1-16

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 4i e- vr ti 1aYgI brea ~k TtempL = [0.0000 0.0847 0 .0864 0.0881 0.0897 0.0914 0.0931 0.0947 .

0.0964 0.0981 0.0997 0.1014 0.1031 0.1047 0.1064 0.1081 0.1097 0.1139 0.1306 0.1472 0.1639 0.1806 0.1972 0.2139 0.2306 0.2472 0.2639 0.2806 0.2972 0.3139 0.3306 0.3472 0.3639 0.3806 0.3972 0.4139 0.4306 0.4472 0.4639 0.4806 0.4972 0.5139 0.5306 0.5472 0.5639 0.5806 0.5972 0.6139 0.6306 0.6472 0.6639 0.6806 0.6972 0.7139 0.7306 0.7472 0.7639 0.7806 0.7972 0.8139 0.8306 0.8472 0.8639 0.8806 0.8972 0.9139 0.9306 0.9472 0.9639 0.9806 0.9972 1.0139 1.0306 1.0472 1.0639 1.0806 1.3611 1.6944 2.0278 2.3611 2.6944 3.0278 3.3611 3.6944 4.0278 4.3611 4.6944 5.0278 5.3611 5.6944 6.0278 6.3611 6.6944 7.0278 7.3611 7.6944 8.0278 8.3611 8.6944 9.0278 9.3611 9.6944 10.0278 20.0833 32.0831 44.0833 56.0833 68.0833 80.0833 92.0833 104.08- 33 116.0833 128.08K 140.0833 152.0833 164.0833 176.08: 33 188.0833 200.08K 212.0833 224.0833 236.0833 248.08-33 260.0833 272.08K 283.3333 297.2222 308.3333 319.4444 333.3333 344.4444 355.5556 369.4444 380.5556 391.66667 402.7778 416.6667 427.7778 438.8889 452.7778 463.88E39 475.0000 488.8889 500.0000 511.1111 525.0000 536.11 i1 547.2222 561.1111 572.2222 583.3333 597.2222 608.33333 619.4444 633.3333 644.4444 .

655.5556 669.4444 680.55556 691.6667 702.7778 716.6667];

i emperaTru e ( I- rgCk b PI,, K TempL [119.811 L3 213.9295 2'42.3104 255.0268 2 55.7907 253.1617 ..

252.9372 252.5390 251.9023 250.9733 249.7169 245.8894 235.985E 224.0051 212.9495 203.5499 195.7225 179.5894 199.8048 174.8143 174.8276 177.3518 180.7405 183.2333 185.1644 186.4925 187.2579 187.8270 188.1924 188.4266 188.560E* 188.5934 188.5042 188.3375 189.3187 189.7570 189.0922 188.5202 188.0148 187.5621 187.4103 187.0671 186.733C 186.4249 186.1559 186.7640 186.5012 186.2557 186.055E 185.9119 185.8265 185.8062 185.8495 185.9526 186.1092 -* 187.8900 187.9673 187.9196 187.9119 187.9385 187.9954 188.0710 188.1647 188.2538 188.3385 188.4003 189.099E 188.9199 188.7439 188.5614 188.3622 188.1314 187.8597 187.5387 187.1667 186.7559 178.4091 171.8762 166.5421

162.2238 158.1410 154.9818 151.7673 148.9234 146.0834 143.7967 141.6054 139.5251 137.9892 136.4819 134.886E 136.9000 136.6489 135.3569 134.3103 133.2941 132.4452 131.9467 132.0536 132.1915 131.3055 130.7946 130.276E 123.0489 118.1991 114.9095 112.4170 110.4096 108 .729C 107.2834 106.0152 104.8855 103.8671 102.9399 102.089C 101.3027 100.5720 99.8894 99. 2491 98.64( 51 98.0763 97 .5362 97.0229 96.5339 96.0669 95.6474 95.1520 94.7720 94.4055 ...

93.9649 93.6254 93.2967 92.9000 92.5932 92.2953 92.0057 91.6547 ...

91 .3822 91.1168 90.7942 90.5432 90.2982 89.9998 89.7671 89.5396 ...

89.2620 89.0452 88.8328 88.5733 88.3703 88.1712 87.9276 87.7368 ...

87.5494 87.3198 87.1398 86.9628 86.7457 86.5753 86.4076 86.2427 ...

86. 0401]

I - KI t IT I<K eu ~ d in 111> Ti- i I Page 1-15 of 1-16

South Texas Project Risk-Informed GSI-191 Evaluation Volume 3: CASA Grande Analysis RI-GSI191-V03 Revision 1 t e, ', tcM I I I k Nforr rihpeinonitec S, H, Tpool = { [TtempS; TempS]; ... I:,M

[TtempM; TempM]; ...

[TtempL; TempL]};

fJPSF ci r :

(sp-,c'fK ,

ajor 111 v eari eS epsruf = 0.00015; pi fil(f HeaderDiam = [1.27 .99 1.27 .84 1.27 50 24. 911 (jl7(tl p1 r 1't HeaderL = [66.96 25.41 12.00 25.46 11.

depth of Common headeil (It PumpDepth = [25.83 25.65 25.83] I, Si ý,SP[ Y o f l bo -ws,, ees, en - nces1, i f:

[ (if (it 90 degree) (# of 45 (I ý ", (11 if of q-a11( C(-sI) (41 cte7t~r1es mHLtab le = [4 2 1 1 0 0; 3 0 0 0 0 1; 0 0 0 0 1 0; K lii 3 0 0 0 0 1; .7 ii --- I 0 0 0 0 1 0; 3 0 0 0 0 1]

Page 1-16 of 1-16

'A

NOC-AE-13002986, Enclosure 4-3 to NOC-AE-13002986, Risk-Informed Closure of GSI-191 Volume 3 Engineering Casa Grande Analysis, Page 161 of 260 Through End

Text

Subject:

Navigation menu

NOC-AE-13002986, Enclosure 4-3 to NOC-AE-13002986, Risk-Informed Closure of GSI-191 Volume 3 Engineering Casa Grande Analysis, Page 161 of 260 Through End

Text

Subject:

Navigation menu

Search