ML14189A452
| ML14189A452 | |
| Person / Time | |
|---|---|
| Site: | Davis Besse  |
| Issue date: | 07/08/2014 |
| From: | Matthews T FirstEnergy Nuclear Operating Co, Morgan, Morgan, Lewis & Bockius, LLP |
| To: | William Froehlich, Kastenberg W, Nicholas Trikouros Atomic Safety and Licensing Board Panel |
| SECY RAS | |
| References | |
| 50-346-LR, ASLBP 11-907-01-LR-BD01, RAS 26164 | |
| Download: ML14189A452 (98) | |
Text
Morgan, Lewis & Bockius LLP 1111 Pennsylvania Avenue, NW Washington, DC 20004 Tel. 202.739.3000 Fax: 202.739.3001 www.morganlewis.com Timothy P. Matthews Partner 202.739.5527 tmatthews@morganlewis.com July 8, 2014 William J. Froehlich, Chair Nicholas G. Trikouros Dr. William E. Kastenberg Atomic Safety and Licensing Board U.S. Nuclear Regulatory Commission Washington, DC 20555-0001 Docket: FirstEnergy Nuclear Operating Company, Davis-Besse Nuclear Power Station, Unit 1, Docket No. 50-346-LR Re: Notification of Documents Related to the Davis-Besse Shield Building
Dear Licensing Board Members:
The purpose of this letter is to provide notification to the Licensing Board of the two attached documents. The first document (Enclosure 1) was submitted by FirstEnergy Nuclear Operating Company (FENOC), applicant in this proceeding, to the Nuclear Regulatory Commission (NRC) on July 3, 2014. The letter provides FENOCs responses to the NRC Staffs April 15, 2014 Request for Additional Information (RAI) B.2.43-4 regarding recent plant-specific operating experience regarding the Shield Building Monitoring Program. The letter identifies revisions to the Davis-Besse License Renewal Application. The second document (Enclosure 2) is FENOCs Full Apparent Cause Evaluation for Shield Building laminar crack propagation.
Respectfully submitted, Executed in Accord with 10 C.F.R. § 2.304(d)
Timothy P. Matthews Counsel for FENOC Enclosures cc: Service List
ENCLOSURE 1
ENCLOSURE 2 FirstEnergy, DAVIS-BESSE NUCLEAR POWER STATION Full Apparent Cause Evaluation Shield Building Laminar Crack Propagation Condition Report 2013-14097 Dated 9/11/2013 Apparent Cause Evaluation Report, post-CARB
Table of Contents Table of Contents ..........................................................................................................................1 List of Acronyms .........................................................................................................................3 1 Abstract ..................................................................................................................................4 2 Introduction.............................................................................................................................6 2.1 Problem Statement .......................................................................................................6 2.2 Consequences ..............................................................................................................6 3 Data Analysis .........................................................................................................................7 3.1 Methodology .................................................................................................................7 3.2 Sequence of Events ....................................................................................................8 3.3 Discussion. .................................................................................................................10 3.3.1 Plant Description and History ........................................................................10 3.3.2 Bore Monitoring and Crack Propagation Discovery..................................... 13 3.3.3 Preliminary Crack Propagation Determination ..............................................13 3.3.4 Failure Modes Analysis .................................................................................14 3.3.5 Failure Mechanism Determination: Investigation Summary.........................15 3.3.6 Ice-Wedging ..................................................................................................17 3.3.7 Water Discovery and Significance ................................................................18 3.3.8 Water Source ................................................................................................22 3.3.9 Relative Humidity of Concrete in Davis-Besse Shield Building.....................22 3.3.10 Freezing Temperatures at Crack Locations ..................................................25 3.3.11 Modeling and Analysis - Thermal Transient ................................................25 3.3.12 Implications of Ice-Wedge Cycles .................................................................28 3.3.13 Laboratory Testing and Analysis ...................................................................29 3.3.14 Structural Finite Element Ice-Wedging Model ...............................................34 3.3.15 Data Analysis Conclusions ...........................................................................36 3.3.16 Hardware Disposition ....................................................................................37 3.3.17 Comparison of Findings from RCA-1 and RCA-2 .........................................37 4 Latent Organizational Weakness Evaluation........................................................................43 5 Generic Implications .............................................................................................................45 5.1 Plant and Industry Experience ....................................................................................45 Page 1 of 80
5.1.1 Strategy .........................................................................................................45 5.1.2 Results ..........................................................................................................45 5.1.3 Conclusions...................................................................................................45 5.2 Extent of Condition .....................................................................................................46 5.2.1 Strategy .........................................................................................................46 5.2.2 Results ..........................................................................................................46 5.2.3 Conclusions...................................................................................................46 6 Apparent and Contributing Causes ......................................................................................47 7 Corrective Action Plan ........................................................................................................ 48 8 Effectiveness Review Plan ................................................................................................ 53 9 References ...........................................................................................................................54 10 Attachments .........................................................................................................................55 Attachment 1 - Event and Causal Factors Chart ..........................................................55 Attachment 2 - Barrier Analysis Worksheet ..................................................................59 Attachment 3 - Latent Organizational Weakness Form ................................................60 Attachment 4 - Failure Modes Analysis ........................................................................68 Page 2 of 80
List of Acronyms ACI - American Concrete Institute CDP - Concrete Damaged Plasticity CFD - Computational Fluid Dynamics CTE - Coefficient of Thermal Expansion EDS - Energy Dispersive X-Ray Spectroscopy FEA - Finite Element Analysis FENOC - First Energy Nuclear Operating Company NDE - Nondestructive Evaluation PII - Performance Improvement International RCA Performance Improvement International Laminar Cracking Root Cause (Davis-Besse Condition Report 2011-03346)
RCA Performance Improvement International Laminar Crack Propagation Root Cause (Davis-Besse Condition Report 2013-14097)
RH - Relative Humidity SEM - Scanning Electron Microscopy Page 3 of 80
1 Abstract The Davis-Besse Nuclear Power Station, operated by FirstEnergy Nuclear Operating Company (FirstEnergy or FENOC), discovered laminar cracking along their shield buildings outer rebar mat during the construction of an access opening for the Reactor Vessel Head Replacement during the 17 Mid-Cycle Outage. As a result, in 2011 FENOC contracted Performance Improvement International, LLC. (PII) to perform a comprehensive technical root cause assessment to pinpoint the cause(s) of the identified laminar cracking. The root cause assessment (RCA-1) concluded that a combination of rebar spacing, high moisture content, and subfreezing conditions brought on by the Blizzard of 1978 led to the observed laminar cracking.
To avoid further laminar cracking, FENOC took preventative actions by applying a water-resistant coating to the shield building in 2012 to inhibit water from penetrating the structure. In addition, FENOC established a program to actively monitor bores for changes at specific intervals.
During shield building monitoring inspections performed in August of 2013, using equipment with enhanced optics, FENOC discovered laminar cracking that had not previously been identified in the bore being observed. Subsequently, FENOC expanded the inspection to include the entire population of bores, 80 in total for signs of cracking. A total of 91 inspections were performed on the 80 bores over the next several days. Out of the 80 bores inspected, new cracking was identified in 8.
As a result of the identified cracking, FENOC again contracted PII to perform a comprehensive technical cause assessment (RCA-2) to identify the cause(s) of the unexpected crack propagation. This investigation focused on the time period from September 2011 to September 2013 which corresponds to the propagating crack conditions. It also considered the continuing validity of RCA-1 as the cause of the initial laminar cracking.
During a PII site visit in September 2013 to confirm the laminar crack propagation issue, standing water and moisture were noted in bores. Field monitoring and laboratory analysis confirmed the water source as internal to the structure and supports the observations of RCA-2.
Laboratory examination of the core samples was performed along the development of a quantitative approach to assess the crack density in the core samples. At multiple depths, evidence of Freeze-Thaw damage and evidence of water transport and excess water was confirmed.
An extracted core sample fracture surface was also examined in detail and step fractures were observed at the laminar crack location indicating progressive fracture. In re-examining the fracture surfaces of the 2011 core samples from RCA-1, step-cracking across the core was not identified; this helped in confirming the difference in crack mechanisms between RCA-1 and RCA-2.
Page 4 of 80
Laboratory testing and various analytical models were used; including thermal and structural analyses which helped determine the cause of the laminar crack propagation to be due to Ice-Wedging. Ice-Wedging requires the three following conditions to occur: 1) a pre-existing crack,
- 2) water present in the crack with localized saturation, and 3) an Ice-Wedge cycle that contained a freezing condition.
Contributing to the Ice-Wedging cause is application of the coating to the shield building. While application of the coating has effectively prevented water from entering the shield building, its application has also prevented a finite amount of moisture from leaving the structure. Until this moisture dissipates it provides the water accumulation mechanism required for Ice-Wedging.
The amount of free water will diminish over time due to absorption and disbursement.
The shield building crack propagation has been well researched as a result of the analysis performed under RCA-2. The crack growth rate has been determined, and can be predicted as a result of this analysis. Testing performed at both Purdue University and the University of Kansas shows that there are no adverse effects of the laminar crack on the capacity of the rebar and that the design basis calculation has design margin.
Therefore, appropriate corrective actions which are being implemented are focused on the on-going monitoring of the shield building. Actions include; on-going, scheduled monitoring of existing bores, determine the effects of the 2013 winter based on visual inspection, identify and record the existence of water in bores and align the existing monitoring program with ACI industry standard regarding crack growth.
Page 5 of 80
2 Introduction 2.1 Problem Statement Between October 2011 and September 2013 laminar cracks have propagated in the shield building. Eight bores exhibit evidence of new cracks discovered in the outer rebar mat within the shoulder region of the shield building during bore inspections conducted in August and September 2013. Six of the bores were last inspected in October 2011, with the remaining two bores inspected in 2012 and 2013. Six bores exhibit crack widths of less than 0.005 inches, while the remaining two bores crack width measure less than 0.007 inches and 0.009 inches respectively.
2.2 Consequences This discovery was unexpected and conflicts with the investigation performed under a root cause evaluation (CR 2011-03346) which concluded that laminar cracks were not likely to propagate.
The current operability of the shield building is supported by a design basis calculation which considers rebar splice capacity in the presence of laminar cracking based on testing performed at the University of Kansas and Purdue University. If the cracks are not passive, then over time the cracks could propagate and challenge the operability of the shield building.
Page 6 of 80
3 Data Analysis 3.1 Methodology A team of subject matter experts knowledgeable in concrete structure design, construction, examination, and modeling were assembled to review evidence and documentation associated with the discovery of laminar crack propagation in previously drilled core bores of the shield building concrete wall. This investigation was intended to determine how and why the concrete crack propagation occurred in the shield building wall.
The consulting group supporting the investigation, Performance Improvement International (PII),
is an industry recognized expert in cause analysis; and was the original root cause evaluator of the shield building laminar cracking issue discovered in 2011 (RCA-1).
PII methodology is one of the techniques endorsed by FENOC per NOBP-LP-2001, FENOC Cause Analysis to perform cause investigations; and as such PII conducted a root cause investigation of the laminar crack propagation of the Davis-Besse shield building through field observation, extensive laboratory testing and modeling analysis. The findings of PII root cause report (RCA-2) were utilized as the primary input to Full Apparent Cause Evaluation Condition Report 2013-14097.
In addition to PII methodology, the investigation performed under Condition Report 2013-14097 was conducted using the guidance described in procedure NOP-LP-2001, Corrective Action Program, business practice NOBP-LP-2011, FENOC Cause Analysis, and reference material NORM-LP-2003 Analytical Methods Guidebook. A pre-job brief was conducted with the owning Manager. Event and Causal Factor Charting and Failure Mode Analysis were methods also used during this evaluation.
PIIs RCA-2 provides in significant detail, items which may not be fully addressed within Condition Report 2013-14097 Cause Evaluation, i.e., analytical and / or testing methodologies, test results, field observation discoveries, laboratory testing and the modeling analysis phases, or other aspects related to the laminar crack propagation investigation. The entire PII report has been captured in FENOCs corrective action database, DevonWay.
The following eight Condition Reports, which are all related to the laminar crack propagation issue, have been considered as part of the scope of this cause investigation:
CR 2013-13782, 09/05/2013 - Shield Building Core Bore S5-666.0-10 Findings CR 2013-13854, 09/06/2013 - Shield Building Core Bore S7-666.0-7 Findings CR 2013-13860, 09/06/2013 - Shield Building Core Bore S7-666.0-9 Findings CR 2013-14623, 09/19/2013 - Shield Building Core Bore S13-633.0-11 Findings CR 2013-14961, 09/25/2013 - Shield Building Core Bore S4-773-16 Findings CR 2013-16204, 10/10/2013 - Shield Building Core Bore S15-646.5-8 Findings CR 2013-16210, 10/10/2013 - Shield Building Core Bore S15-674.5-3 Findings CR 2013-16211, 10/10/2013 - Shield Building Core Bore S15-777-3 Findings Page 7 of 80
3.2 Sequence of Events January 1978 The Blizzard of 1978 was the initiation event that created the laminar crack phenomenon within the shield building.
October 10, 2011 Laminar crack discovered in the shield building access opening during 17M (documented under CR 2011-03346).
February 2012 Boroscope inspections of the bores identified no indications of cracking in previously drilled core bores, no change in crack size from the previously identified condition. Water discovered in one bore in shoulder 9.
March 2012 Issued Long-Term monitoring of the shield building laminar cracking procedure, EN-DP-01511 and schedule. Twelve bores are inspected each year to determine if there has been any change to the bore. The monitoring of the shield building is included in the Maintenance Rule Evaluation for Structures.
May 2012 FENOC completed inspections associated with the shield building laminar cracking. Boroscope inspections of the bores did not reveal any changes of discernable magnitude when compared against initial inspection records (from 17M). Water discovered in three bores in shoulder 9 and one bore in shoulder 2.
June - July 2012 Additional examination of all remaining accessible exterior surfaces of the shield building wall was performed with Impulse Response technology and confirmatory bores to confirm the findings of the root cause analysis. No evidence of crack growth.
Aug 2012 Water discovered in two bores in shoulder 9.
Aug - Oct 2012 Implemented Engineering Change Package (ECP 12-0208) applying a shield building exterior coating.
June 2013 Water discovered in two bores in shoulder 9 and one core in flute 5.
Aug 2013 A new crack was identified in one of the bores. This crack was determined to be a newly identified existing crack. The identification of this new crack was attributed to using higher quality inspection equipment. As a result of this newly identified condition, the inspection size was increased to 100%
of all assessable bores. Water discovered in two bores in shoulder 4.
Sept 5, 2013 First indication of crack propagation - CR 2013-13782 - Shield Building Core Bore S5-666.0-10 Findings. Over the next several days seven more indications are identified and documented in Condition Reports.
Page 8 of 80
Sept 11, 2013 Initiation of CR 2013-14097, Laminar Crack Propagation.
Oct 2013 Water discovered in three bores in shoulder 15, three bores in shoulder 4, one bore in shoulder 6, one bore in shoulder 8, and two bores in shoulder 3.
March 2014 Water discovered in two bores in shoulder 4, one bore in shoulder 2 and one bore in shoulder 4.
Page 9 of 80
3.3 Discussion The Davis-Besse Nuclear Power Station, operated by FirstEnergy Nuclear Operating Company (FirstEnergy or FENOC), discovered laminar cracking along their shield buildings outer rebar mat during the construction of an access opening for the Reactor Vessel Head Replacement during the 17 Mid-Cycle Outage. As a result, in 2011 FENOC contracted Performance Improvement International, LLC. (PII) to perform a comprehensive technical root cause assessment (RCA-1) to pinpoint the cause(s) of the identified laminar cracking. RCA-1 concluded that a combination of rebar spacing, high moisture content, and subfreezing conditions brought on by the Blizzard of 1978 led to the observed laminar cracking.
To avoid further laminar cracking, FENOC took preventative actions by coating the shield building with a water-resistant coating to inhibit water from penetrating the structure. In addition, FENOC established a program to actively monitor bores for changes at specific intervals.
In 2013, while monitoring several bore locations, FENOC discovered cracking in bores that had not originally exhibited cracking. On account of this discovery, FENOC examined all existing bore locations, 80 in total, and discovered 8 bores with unexpected propagating cracks. As a result of the identified cracking, FENOC again contracted PII to perform a comprehensive technical root cause assessment (RCA-2) to identify the cause(s) of the unexpected crack propagation. This investigation focuses on the time period from September 2011 to September 2013 which corresponds to the propagating crack conditions. It also considers the continuing validity of RCA-1 as the cause of the initial laminar cracking.
3.3.1 Plant Description and History The shield building is a reinforced concrete structure of right cylinder configuration with a shallow dome roof. An annular space is provided between the steel containment vessel and the interior face of the concrete shield building of approximately 4 ft 6 in width to permit construction operations and periodic visual inspection of the steel containment vessel. The volume contained within this annulus is approximately 678,700 ft3. The containment vessel and shield building are supported on a concrete foundation set on a firm rock structure. With the exception of the concrete under the containment vessel, there are no structural ties between the containment vessel and the shield building above the foundation slab.
The shield building has a height of 279 ft 6 in measured from the top of the foundation ring to the top of the dome. The dome roof is approximately 2 ft thick. The inside radius of the shield building is 69 ft 6 in and the thickness of the shield building wall is approximately 2 ft 6 in. The shield building exterior has eight vertical architectural flute reveals that are spaced 45 degrees apart. The architectural flutes consist of shoulders that extend another 1 ft 6 in outward and gradually taper back to the outer cylindrical wall of the shield building while reaching a point of tangency 17 ft 11 in from the centerline of the flute; there are a total of 16 shoulders.
Page 10 of 80
The below figure shows a plan view of the shield building. Numbers 1 - 8 refer to the flute regions of the building:
Flutes Figure 1: Shield Building Plan View The shield building is designed to provide biological shielding during normal operation and from hypothetical accident conditions. The shield building provides radiation shielding, a means for collection and filtration of fission product leakage from the containment vessel following a hypothetical accident, and provides environmental protection for the containment vessel from adverse atmospheric conditions including extreme winds, tornadoes, and tornado-borne missiles. Besides the emergency ventilation system, the shield building also interfaces with station lightning protection, and station drainage.
The Containment Purge ventilation system is responsible for supply of fresh air into the Shield Building Annulus and Mechanical Penetration Rooms during any mode of plant operation. In the event of a radiation release or an adverse pressurization of the Containment the Isolation Valves actuate in concert with the ventilation systems to isolate the Penetration Rooms and maintain a negative pressure boundary within the Penetration Rooms and annulus. Within the Page 11 of 80
annulus, heaters are in place in order to maintain an annulus temperature above 30F. [CARB Comment a]
The shield building is a structure consisting of concrete, and reinforcing steel, with other minimal miscellaneous embedded material. The shield building was designed in accordance with American Concrete Institute (ACI) 307-69, Specification for the Design and Construction of Reinforced Concrete Chimneys, and checked by the Ultimate Strength Design Method in accordance with ACI 318-63, Building Code Requirements for Reinforced Concrete.
Page 12 of 80
3.3.2 Bore Monitoring and Crack Propagation Discovery Bore inspections were performed in 2012 in accordance with EN-DP-01511, Design Guidelines for Maintenance Rule Evaluation of Structures, which consisted of the inspection of 12 bores.
These were performed per the inspection procedure as follows:
Requirement 1 - Previously un-cracked bores (a total of 6 inspections) x 3 in a shoulder regions x 1 in the steam line penetration areas x 2 at the top of the building outside the shoulders Requirement 2 - Previously cracked bores (a total of 6 inspections) x 3 in a shoulder regions x 1 in the steam line penetration areas x 2 within the top 20 feet of the building outside the shoulders No discernable changes in the cracking condition were documented in 2012. The inspections were again repeated in 2013 following Requirement 1 and 2, above.
During visual inspection of core S4-650.0-16 on August 26, 2013, as part of the 2013 inspections, using equipment with enhanced optics, FENOC discovered laminar cracking that had not previously been identified in the bore. Subsequently, FirstEnergy expanded the inspection to include the entire population of bores, 80 in total for signs of cracking. A total of 91 inspections were performed on the 80 bores. Out of the 80 bores inspected, new cracking was identified in 8.
As a result of the identified laminar crack propagation, PII was again contracted to perform a comprehensive technical root cause assessment (RCA-2) to identify the cause(s) of the newly identified cracks. Thus, this investigation focuses on the time period from September 2011 to September 2013 which corresponds to the crack conditions.
3.3.3 Preliminary Crack Propagation Determination The extent of crack propagation was preliminarily determined through a spatial analysis of the unexpected propagating cracks discovered in 2013. A review of prior core samples and bore data, along with the relationship and proximity of neighboring bores helped to establish propagation rate estimation. Three of the eight bores listed below did not have a neighboring bore to relate to for the crack rate estimation, therefore the crack propagation rate at these three bores are Undetermined.
Page 13 of 80
Table 1: Summary of Crack Propagation Rates for 2011-2013 Bore Report Maximum Crack Propagation from Plant Data S5-666.0-10 CR-2013-13782 10.25 in/year circumferentially S7-666.0-7 CR-2013-13854 Undetermined S7-666.0-9 CR-2013-13860 9.0 in/year circumferentially S13-633.0-11 CR-2013-14623 Undetermined S15-646.5-8 CR-2013-16204 0.25 in/year joining laminar cracks S15-674.5-3 CR-2013-16210 Undetermined S4-773-16 CR-2013-14961 1.25 in/year joining laminar cracks S15-777-3 CR-2013-16211 1.0 in/year joining laminar cracks Based on the spatial analysis of newly identified cracked bores, preliminarily, the crack growth appears to be bound at 6 - 10.25 in per year circumferentially.
3.3.4 Failure Modes Analysis Performance Improvement Internationals investigational efforts identified 17 failure modes as potential candidates for the newly observed crack propagation condition. Since there was previous laminar cracking adjacent to the recently identified cracks, PII developed failure modes on the basis that the observed cracks are either new cracks, pre-existing laminar cracks that are propagating (or have propagated), or a combination of both. The following table shows a list of the failure modes considered:
Table 2: Potential Failure Modes Failure Mode No. Failure Mode Name 1 Core-Drill Induced Stresses 2 Thermal Cycles 3 Settlement of a Cracked Building 4 Enhanced Thermal Gradient due to Open Core 5 Crack Misidentification Cumulative Water Damage as a result of Sealing the Structure in 6
2012 7 Moisture Gradient Induced Stresses 8 Material Property Variation 9 Stress Concentration due to Bore Holes 10 Residual Stresses 11 Freeze-thaw Damage 12 Concrete/Rebar Interaction 13 Material Degradation due to Repeat Freeze-thaw 14 Freeze-Thaw via Water Accumulation due to External Water Source 15 Freeze-Thaw via Water Accumulation due to Existing Water Source 16 Ice Wedging 17 Reverse Coefficient of Thermal Expansion Page 14 of 80
The above listed failure modes were grouped into the following categories:
Table 3: Failure Mode Groupings FM Groupings Major Contributor FMs Excess Water 6, 7, 11, 13, 14, 15, 16, and 17 Thermal Stresses 2, 10, and 12 Drilled Open Bores 1, 4, and 9 Materials 8 & 13 Building Settlement 3 Crack Misidentification 5 In general, each failure mode was technically refuted or supported by laboratory tests and examinations, state-of-the-art Finite Element Analyses, and/or other data. Some failure modes did not need to be analytically verified as deductive reasoning based on existing evidence was sufficient enough to either support or refute the respective mode of failure.
Unless there was refuting evidence against a given failure mode, it was considered as a possible contributing factor. The failure modes without refuting evidence, hereafter called unrefuted failure modes, were further quantitatively analyzed to understand their relative impact and contribution to the newly discovered laminar cracking. contains a detailed description of each failure mode along with its respective supporting/refuting evidences.
Completion of the Equipment Apparent Cause Evaluation form (NOP-ER-1001-01) was not warranted, as the higher level cause analysis conducted by PII under RCA-2 addressed this item. No additional failure modes were identified.
3.3.5 Failure Mechanism Determination: Investigation Summary The laminar crack propagation discovered in September of 2013 by Davis-Besse was confirmed by the PII root cause investigation team during its 2013 shield building inspection. In order to analyze one of the newly identified laminar cracks, core sample S5-666.0-9.5 was extracted.
This sample lies between core S5-666.0-8 and S5-666.0-10 and bisects the laminar crack.
Page 15 of 80
Figure 2: Location of S5-666.0-9.5 Relative to S5-666.0-8 & S5-666.0-10 During the PII field visit, water was discovered in S5-666.0-8 and S5-666.0-10. Further inquiries revealed that more water findings were recorded by the plant staff during previous bore inspections that took place the previous 2 years (See Sequence of Events). To confirm that water was present inside the structure, several dry bores were sealed from the outside environment for a period of time. Upon subsequent inspection, water was found inside the bores and collected. The water analysis from these samples helped confirm water transport within the structure.
A petrographic examination of the core samples was also conducted. Inspection under an SEM revealed the presence of microcracks. A quantitative approach was developed to assess the microcrack density in the core samples. At multiple depths, evidence of Freeze-Thaw damage and evidence of water transport in the form of Ettringite crystals formation and microcracks emanating from pores was found. The maximum microcrack density was near the outer most layer of the concrete (within the first 2 in). The microcracks emanating from pores at the laminar crack locations were present at a lower density than shallower locations. On account of the water detection inside the bores, the water analysis, and the presence of microcracks emanating from pores at depths up to 10 in, the presence of excess water was confirmed.
The extracted core S5-666.0-9.5 samples fracture surface was also examined in detail and step fractures were observed at the laminar crack location indicating progressive fracture. In re-examining the fracture surfaces of the 2011 core samples from RCA-1, step-cracking across the core was not identified; this helped in confirming the difference in crack mechanisms between RCA-1 and RCA-2.
Lab tests were performed to test the validity of the Ice-Wedging failure mode. The testing took Page 16 of 80
into consideration: 1) a pre-existing crack (a mechanical crack of 0.001 in width and 0.5 in long was introduced), 2) water present in the crack with localized saturation, and 3) an Ice-Wedge cycle that contained a freezing condition with a time factor (20 min in the freeze zone). The results revealed a step crack propagation after each Ice-Wedge cycle. The average crack propagation step was 0.5 in. The step fracture produced by the Ice-Wedge testing was similar to the step fracture observed in the sample extracted from the shield building which was measured as 0.6 in on average.
Thermal and Structural Finite Element Analyses (FEA) were conducted to determine the plausibility of Ice-Wedging as a failure mechanism. The thermal transient analysis determined the through-wall temperature profile for the shield building for the winter periods desired. The results from the thermal transient analysis showed that freezing temperatures at the location of the laminar cracks under consideration were present. The number of Ice-Wedge cycles calculated were multiplied by 0.6 in (measured crack propagation step width) to determine the crack propagation rate over the last two years. The Ice-Wedge crack propagation rates per location were in the same order of magnitude and less than the maximum estimated for that cores that crack propagation could be predicted.
A structural analysis was also conducted based on pre-existing laminar cracks expanding due to the 9% volume increase during the water to ice phase change. The model showed that crack propagation occurred on the same order of magnitude as that calculated in the laboratory and the field. Hence, the results from this analysis show that Ice-Wedging is a plausible failure mechanism.
The RCA-2 effort considered 17 failure modes (FMs). After ruling out many of the FMs based on testing and FEA analysis, Ice-Wedging was confirmed as the failure mechanism responsible for the observed crack growth of the shield building and therefore has been identified as a causal factor (causal factor 1).
Contributing to the Ice-Wedging cause is application of the coating to the shield building. While application of the coating has effectively prevented water from entering the shield building, its application has also prevented a finite amount of moisture from leaving the structure. Until this moisture dissipates it contributes to the water accumulation mechanism required for Ice-Wedging. Application of the coatings has been identified as a causal factor (causal factor 2).
3.3.6 Ice-Wedging Crack propagation caused by Ice-Wedging occurs when water accumulates at a pre-existing crack tip and freezes. During the water to ice phase change, there is a 9% volumetric expansion which results in a force exerted on the surrounding concrete. Due to the concretes inability to elastically absorb the resulting force, the crack propagates to relieve the induced stresses.
In order for Ice-Wedging-induced cracking to propagate in the Davis-Besse shield building, three conditions must be met:
Page 17 of 80
- 1. Pre-existing laminar cracks (Ice-Wedging Condition 1),
- 2. Water accumulation at the crack locations (Ice-Wedging Condition 2), and
- 3. Freezing temperatures at the crack locations (Ice-Wedging Condition 3)
Based on the foregoing discussion, it has been shown that the Davis-Besse shield building satisfies the first condition of pre-existing laminar cracks. Sections 3.3.7 through 3.3.11 cover the remaining two conditions that must be satisfied in order for the Ice-Wedging mechanism to occur; both water and freezing conditions at the location of laminar cracks.
3.3.7 Water Discovery and Significance Small quantities of water were discovered on several occasions in various bores during 2012, 2013, and 2014. During PII site visit in September 2013 to confirm the laminar crack propagation issue, standing water was noted in bore S5-666.0-10 (2 - 3 from the bore opening) while the adjacent bore (S5-666.0-8) was moist but with no standing water present. Consequently, water collection was requested from accessible bores along with their matching plugs. The water samples were collected for analysis to determine whether the water was a mere condensation of moist air entrapped inside the bores, rain water that had seeped in through the plug, or water in excess that had transported within the concrete and collected inside the bore. However, as will be pointed out in the plug analysis section of the report, the plugs were sealed with caulking to prevent outside water sources from seeping in.
Water Analysis Results - pH Measurements All the samples have a pH greater than 7 with varying degrees of alkalinity, typical of water that was in contact with concrete for a period of time. The presence of sodium chloride salt in the water, as detected by EDS, may act as a buffer affecting the pH measurement depending on the salt concentration present. Due to sample size limitation, quantitative analysis of water constituents including salt content could not be performed in this investigation.
Summary of Test Results The SEM/EDS results revealed the presence of sodium chloride salt, zinc oxide, spherical iron particles, and iron oxide which are not primary constituents of concrete. In addition, the results revealed calcium, magnesium, and silicon compounds that would be expected as primary constituents of concrete. Previous analysis on the concrete performed by CTLGroup and documented in FENOC document # EN-DP-01511-LTM 2012; long-term monitoring, revealed the presence of sodium chloride in the concrete with an acceptable concentration level. The presence of sodium chloride in the water means that the salt was leached out of the concrete and dissolved in the water thus making the water more corrosive in nature. The presence of iron particles in spherical form suggests that these particles were originally in the form of molten iron splatter particles that solidified and became entrapped in the concrete in the vicinity of the rebar. The EDS for these particles does not show oxygen, therefore the most likely source of these fine size spherical particles is abrasive grinding or cutting during the iron work of rebar in the initial construction of the shield building. Considering the very fine particle size of 0.09 mils, these particles can pass through any of the laminar cracks found in the structure near the vicinity Page 18 of 80
of the outer rebar mesh. The most likely source for the zinc oxide is the corrosion of the plug screw-head that is exposed to the internal environment of the bore which contains salty water vapors.
Plug Analysis The shield building contains 80 boreholes which are all plugged using Twist-Tite mechanical plugs. Figure 3 shows the plug as it is attached to the borehole. The entire circumference of each of the plugs that were sampled for water (9 total) were caulked as were the screws to ensure that no outside water or debris could enter into the bore.
Figure 3: Schematic of Plug Installed in the Bore Page 19 of 80
Figure 4: Extent of plug components damage and screw thread corrosion Figure 4 shows the plug in its disassembed configuration. Note the caulk around the plug circumference, while caulk around the screw head were removed during disassembly.
Figure 5: Plug screw head corrosion as a result of exposure to bore environment at bore S6-666-44 Page 20 of 80
Figure 6: Plug screw absence of corrosion in the plug screw threads contained in the interior space at bore S6-666-44 Examination of the plug removed from bore S6-666.0-44, Figures 5 and 6, revealed no evidence of leakage through the plug and yet the galvanized head revealed evidence of corrosion which illustrate the plug screw condition. This indicates that the outside environment did not contribute to corrosion of the screw-head. Additional plugs with this documented condition are captured in RCA-2.
Note that on all plugs the external environment did not cause corrosion as prominent as the internal corrosion of the bolt heads or threads. This is most likely due to the presence of salt in the internal bore cavity as compared to the outside environment which is salt-free.
It is important to recognize that the corrosion of the plug stem heads as well as plug stem screws in the case of gasket leakage as observed in 4 of the 7 bores was due to salt laden water vapor generated from the salt water in the bore. By comparison the water itself with salt and high pH is not conducive to generate corrosion in the rebar.
In order to confirm that the water was still transporting through the structure at the identified locations, the same bores were re-sealed on December 6, 2013 and re-inspected on March 3, 2014 for water collection. Water was found in four sealed bores which confirmed the water source as internal to the structure and supports the observations listed in RCA-2.
Water Discovery Conclusions The presence of sodium chloride salt in the water means that the salt was leached out of the concrete and dissolved in the water thus making the water more corrosive in nature. The presence of iron particles in spherical form suggests that these particles were originally in the form of molten iron splatter particles that solidified and became entrapped in the concrete in the Page 21 of 80
vicinity of the rebar. The EDS for these particles does not show oxygen, therefore the most likely source of these fine size spherical particles is abrasive grinding or cutting during the iron work of rebar in the initial construction of the shield building. Considering the very fine particle size these particles can pass through any of the laminar cracks found in the structure near the vicinity of the outer rebar mesh.
3.3.8 Water Source There are three possible water sources: wind-driven rain, rain and melted snow that has accumulated on the roof (between the parapet and dome) and seeped down through roof cracks, or condensation resulting from internal relative humidity reaching 100%.
Wind-Driven Rain This water-source mode hypothesizes that rain in combination with wind pressure, will allow water to penetrate the vertical walls of the shield building and through moisture diffusion migrate to laminar crack regions. During RCA-1 a laboratory experiment was conducted which showed rain diffusing into the concrete as a possibility.
Roof Cracks During RCA-1 cracks on the roof were observed at the parapet/roof interface. All roof cracks were repaired during coating of the shield building. The Davis-Besse maintenance rule manual states that cracks 1/16 in or less do not need to be repaired. However, a crack of this width may be sufficient enough to allow water to seep through. Correspondingly, this water source hypothesizes that water and melted snow seep through roof cracks and ultimately migrate through the structure in a top down fashion (gravity effects). However, the shield building coating was completed in October of 2012 and would therefore prevent subsequent water intrusion.
Condensation This water-source mode hypothesizes that the relative humidity (RH) in the bore holes is in equilibrium with the pore relative humidity in the surrounding concrete. Water condensation in the bore and in the pores of concrete can occur when the temperature decreases (for example, from day time temperature to night time temperature). This is because the saturation pressure of water vapor depends on temperature. When temperature decreases, saturation pressure decreases, and RH increases; when the RH increases to a certain level (oversaturation), the condensation of water vapor occurs and may accumulate at voids.
3.3.9 Relative Humidity of Concrete in the Davis-Besse Shield Building In October 2012, under RCA-1 corrective actions a water-resistant coating was applied to the exterior of the entire shield building. The rough surface of the structure prevented the use of Wet Film Thickness readings which is the industry standard nondestructive method for measuring concrete coating thickness. Based on application records which identified the quantity of coating applied and the surface area of the building a general coating thickness may be Page 22 of 80
calculated. Utilization of this method results in coating thickness values which are not greater than the 19.8 mils Dry Film Thickness required by specification A-028N for the three coat system.
Based on these application records, the discovery of four sample locations from within a 4 in.
diameter bore, observed during RCA-2 lab testing to have a coating thickness greater than the specification requirements; are not necessarily representative of the entire shield building. The coating was applied to the shield building to provide a mechanism for reduction of wind driven rain intrusion into the concrete.
The additional thickness of the coating does not impact the ability of the coating to reduce water intrusion. These values affect the permeance of the coating, i.e. the ability to allow moisture transport through the coating. Therefore, the coating has effectively blocked out external water intrusion and locked in moisture or water existing in the structure prior to the coating. This condition will have an impact on the moisture movement and distribution within the shield building wall as described:
Figure 7 shows the internal RH distribution measured on November 8, 2011 from S6. The boundary condition on the outer surface = 40% (the RH in lab). As can be seen in the figure below, the moisture variation can reach up to the depth of about 4 to 5 in (because the boundary condition is not a daily variation).
œ Figure 7: Moisture Variation in the Shield Building Wall (2011)
In 2013, a laboratory experiment was performed to determine the relative humidity (RH) versus depth of sample S5-668-17. The following figure shows the effects of the coating on the RH of the sample. As determined from a coating analysis, the coating applied had a high permeance value - prohibiting two-way water transport - resulting in the increased RH. The red line represents the RH gradient of the sample in the structure.
Page 23 of 80
Figure 8: Internal Humidity Variations at Different Depths Conclusions The discovery of water in several bores helps to explain why crack propagation is occurring now at the locations of interest, as this condition has greater relative variability. Water was discovered in the structure in February of 2012 six months prior to the application of the shield building coating. Water was also discovered in 2013 and 2014 after the coating was applied. Therefore, there are two open possibilities as to why water has accumulated at the crack tips over the last two years. The first possibility being a top-down water mechanism that, due to the rate of diffusion of moisture in concrete, has now arrived at the crack tips after a period of time.
However, the shield building coating application which including sealing all roof cracks, was completed in October of 2012 and would therefore prevent subsequent water intrusion.
The second and most likely is an increase in RH due to the application of the coating as previously demonstrated. When the saturation pressure of air decreases due to a drop in temperature, it approaches and may drop below the vapor pressure of air, resulting in RH of 100% and the formation of water. However, due to the application of the coating, external water sources have ceased and water inside of the structure is now finite and will dissipate over time.
The current relative humidity distribution will also dissipate over time until the RH has reached an equilibrium level. Hence, regardless of the source, the water volume and/or the RH distribution inside of the shield building is a decreasing function.
Page 24 of 80
3.3.10 Freezing Temperatures at Crack Locations The third condition to be met in order for Ice-Wedging to occur is freezing conditions at the location of the laminar cracking. Therefore, to determine the thermal transient temperatures throughout the shield building for the time periods of interest (2011/2012 & 2012/2013 winters),
a robust Finite Element Thermal Transient model was developed.
3.3.11 Modeling and Analysis - Thermal Transient To determine the thermal transient temperatures throughout the shield building, a robust Finite Element Thermal Transient model was developed. RCA-2 discusses the model development in extensive detail while facilitating an understanding of the shield buildings response to temperature conditions. Prior to conducting any detailed thermal analyses, a review was made of the meteorological data collected from the Davis-Besse site. The data spanned the period from September 2011 through September 2013 and included daily measurements of ambient temperature; wind speed and wind direction at 10, 75 and 100 meter elevation; dew point; relative humidity; solar flux; barometric pressure; and wet bulb temperature. All measurements were made at 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> intervals. The meteorological data revealed that the strongest winds predominantly originate out of the South and Southwest. This is especially true during the fall and winter months.
Consequently, for the purposes of determining the temperatures in the concrete shield building during a freeze cycle, a detailed FLUENT based CFD analysis model was created which incorporated not only the concrete shield Building but also the adjoining Auxiliary Building.
The geometry for this model is captured in RCA-2. Examination of the convective heat transfer contours determined that heat transfer coefficients are generally highest on the windward side.
In addition, the heat transfer coefficient peaks in the vicinity of the flutes, especially those positioned 90 degrees to either side of the 225 degree southwest wind heading. At these locations, the flow completely separates from the shield building becoming highly turbulent and improving the mixing of the air and aiding in convective heat transfer. This phenomenon occurs at every flute. The convective heat transfer coefficient is a function of wind speed.
Consequently, the predicted heat transfer coefficient at any given wind speed can be calculated. This methodology was used to account for the variation in wind speed provided from the Davis-Besse meteorological data.
The convective heat transfer though the annulus was considered, however based upon the ventilation flow through the annulus it was determined that the radiant heat transfer would dominate, and dictate the temperature at the inner face of the concrete shield building. [CARB Comment a]
Radiation Heat Sources The primary intent of thermal analysis was to ascertain the variation in temperature that occurs daily as well as seasonally.
Page 25 of 80
The high Northern Latitude of the Davis-Besse facility results in the sun striking the concrete shield building primarily from the South with only minimal exposure to the Western and Eastern faces. The Northern face is perpetually shaded, receiving no direct sun light. Consequently, the highest daily and seasonal variations in temperatures would occur in the Southern half of the shield building.
In addition to solar, another source of radiant heating is attributable to the inner steel containment that houses the reactor. This steel containment operates at an elevated temperature that is well above the surrounding ambient air as well as the annular air. In late January of 2014 just prior powering down the facility, direct measurements were made of the steel containment at three different elevations and four different azimuthal positions. The data reveals that at three elevations, there is approximately a 22o F vertical gradient with little or no circumferential gradient. Using this data, an exponential curve fit was performed through the measured data to determine the vertical variation in the temperature for the steel containment.
It was further assumed that this temperature remained invariant in response to any seasonal changes in the external ambient conditions. For the purposes of incorporating the radiant heat transfer to the interior face of the concrete shield building, the steel containment was treated as an ambient source. For all of the thermal conditions examined, radiation with the external ambient was also included. This was especially important at night and in winter when the cooler night sky and surroundings would result in an appreciable heat loss even in the absence of any convective losses due to wind.
In the region where the Auxiliary Building mates to the concrete shield building, radiation to the interior of the Auxiliary Building was assumed. A nominal air temperature of 83 degrees F was estimated for the air contained in the Auxiliary building and annular passage for external ambient temperatures on the order of zero degrees F. Consequently, the radiant heat transfer from the inner steel containment would dominate and dictate the temperature at the inner face of the concrete shield building.
Ambient Temperature All of the thermal heat transfer effects associated with the steel containment are based upon the ambient temperature. This includes external forced convection and radiation to the surrounding environment.
Because water has nearly a five-fold higher specific heat than concrete, the presence of even a trace amount of moisture would greatly augment the specific heat and thermal conductivity.
Consequently, it was essential to account for the variation in relative humidity through the thickness of the concrete.
Direct measurements made of test specimens extracted from the concrete shield building revealed very high levels of relative humidity near the exterior surface. Toward the interior, the relative humidity decreased to about only 65 to 70%. Due to the profound effect that moisture has on the thermal properties of the concrete, it was necessary to use thermal properties that Page 26 of 80
accounted for this spatial variation in the relative humidity.
Transient Thermal Analysis The 6 day period extending from January 25th to January 30th 2014 was chosen for the thermal transient analysis of field data which provided a means of confirming the results obtained from the thermal transient analysis modeling. This time frame became ideally suited for examining the behavior of the concrete shield building during a typical multi-day freeze-thaw cycle, as temperatures varied greatly during this period, with temperatures as low as -7 degrees F to above freezing. This period also marked the last six days when the facility was operating at nominal power and just prior to the initiation of a refueling outage, e.g., plant shutdown. Steel containment temperatures measurements were made on February 1st and 2nd while the facility was still in a mode 3 operational state. This would have precluded any significant cool down in the steel containment, thereby permitting a more accurate determination of the containment temperatures during nominal operational conditions.
Conclusions Results from the transient thermal analysis are shown below. Here the measured and predicted temperatures at an elevation of 666 ft. and at shoulder 5 have been plotted. The results correlate with measured temperatures, especially for the interior. At a depth of approximately 8 inches, the predicted temperatures differ by less than a few degrees from the measured values. Only at the surface is there a significant difference, especially on January 28th. This was when temperatures dropped to their lowest. This may very well have been associated with the presence of wind driven snow that would have greatly enhanced the convective heat loss.
However, it can be concluded that the thermal model accurately predicts temperatures at those interior locations where cracking is most prevalent.
Page 27 of 80
Figure 9: Temperatures at Various Depths at Bore S5-666.0-8 Examination of the overall predicted temperatures reveals the dominance of external convection heat loss on the temperatures at the exterior of the concrete shield building. Note in the vicinity of all of the flutes where the predicted heat transfer coefficient is highest due to flow separation the temperatures are significantly cooler.
Results obtained from the thermal transient analysis were subsequently mapped onto more detailed 2D plane strain models that featured a highly refined mesh. Plane strain models were created at three elevations situated at 625.5, 665.5 and 741.8. This allowed a more accurate determination of the spatial variation through the thickness of the concrete near the base, just above the auxiliary building and near the top of the concrete. The plane strain model at elevation 665.5 was of particular interest since this is where laminar crack propagation was documented and core bores were instrumented for temperature measurements.
3.3.12 Implications of Ice-Wedge Cycles As previously stated, freezing at the locations of the observed cracks is one of the three requirements for the Ice-Wedging failure mode to be considered valid. The number of freeze-thaw cycles of interest are for the 2011/2012 & 2012/2013 winters, the time period for which the unexpected propagation crack propagation condition was bounded. The number of potential freeze-thaw cycles are not only of interest in the locations where cracks were identified, but in various other locations as well to help determine the spatial susceptibility of the shield building to the Ice-Wedging mechanism. Thus, PII was tasked with determining the number of freeze-thaw cycles at any location of the building, for any depth, and potentially for any winter. In order to determine the number of cycles at different locations and depths of the shield building for any time period, two thermal transient models were used in combination. The crossings of the temperature history above and below 32 degrees F can subsequently be used to estimate the number of freeze/thaw cycles that each particular location was likely to see between the Fall of 2011 and the Spring of 2013. Below are some of the findings from the modeling of freeze/thaw cycles:
S5-666-10 at 5.25 - The number of counted ice wedge cycles at the laminar crack location (5.25 in.) is 18 cycles. This translates into a total crack growth of 10.8 over the two year period.
Thus the crack growth rate would be 5.4 in. per year which falls within the estimated maximum value predicted for this location. The estimated maximum crack growth rate for this location was 10.25 in. per year.
S7-666-9 at 5.5 - The number of counted ice wedge cycles at the laminar crack location (5.25 in.) is 16 cycles. This translates into a total crack growth of 9.6 in. over the two year period. Thus the crack growth rate would be 4.8 in. per year which falls within the estimated maximum value predicted for this location. The estimated maximum crack growth rate for this location was 9 in.
per year.
Page 28 of 80
S15-777-3 at 9 - The number of counted ice wedge cycles at the laminar crack location (9 in.)
is one cycle. This translates into a total crack growth of 0.6 in. over the two year period. Thus the crack growth rate would be 0.3 in. per year which falls within the estimated maximum value predicted for this location. The estimated maximum crack growth rate for this location was 1.0 in. per year.
Conclusions From the above data, it can be seen that the extent of cracking falls within the number of freeze-thaw cycles. This reiterates the fact that in order for crack propagation to take place, freezing conditions along with accumulated water are needed.
3.3.13 Laboratory Testing and Analysis Core-Bore Fracture Analysis A detailed fracture analysis was performed on core-bore S5-666.0-9.5. This core-bore was in the pathway of the discovered propagated crack from S5-666.0-8 to S5-666.0-10 and was extracted during the RCA-2 investigation in October of 2013. The following figure illustrates its location:
Figure 10: Location of S5-666.0-9.5 Relative to S5-666.0-8 & S5-666.0-10 In 2011 and 2012 in bore S5-666.0-8 a crack was present and no crack in S5-666.0-10. In 2013 the crack had propagated to S5-666.0-10. Visual examination of the fracture surfaces from S5-66.0-9.5 revealed the presence of step fracture zones separated by ridges. These ridges were not influenced by the coarse aggregate fractures. Also, these fractures were through the coarse aggregate suggesting a driving force high enough to go through the coarse aggregate as well as the fine aggregate.
Page 29 of 80
The step fracture zones I, II, III, IV are illustrated below. Zones I and Ia represent the older fracture, F1, bound by the merging crack line that extends in the photograph from 12:00 to 6:00.
After merging, the newer crack propagated in steps through zone II, zone III and zone IV. The approximate step length is measured as 0.60 in which measures closely with the step lengths realized in the laboratory experiment.
Figure 11: Fracture Surface (S5-666.0-9.5)
SEM examination of the fracture surface revealed the presence of secondary cracks (microcracks) intersecting the fracture surface. In addition, on the fracture surfaces, Ettringite crystals (formation from sulfate reaction with calcium aluminates as a result of alternating moisture changes) were found in pores, and microcracks emanating from these pores. The significance of finding Ettringite crystals in pores in combination with microcracks emanating from these pores is evidence of first, water transport within the concrete and second, evidence of freeze-thaw cracking mechanism.
Page 30 of 80
Figure 12: Ettringite Crystals (S5-666.0-9.5)
Ice-Wedge Laboratory Experiment An Ice-Wedge laboratory experiment was conducted to test the hypothesis of Ice-Wedge cracking on the shield building concrete.
A measured crack was induced into each of the 4 specimens by a controlled mechanical method to simulate building conditions. Afterwards, the specimens were soaked in water for 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> and then exposed to a wet Ice-Wedging cycle (freezing/holding in frozen condition/thawing).
The surface cracks were then observed. Water was present only above the crack mouth in the machined notch area to ensure the presence of water inside the crack during the Ice-Wedging cycle (the wet condition). The Ice-Wedging cycles were repeated and crack extension recorded.
The samples were then fractured to expose the Ice-Wedging crack steps, the fracture surface was then examined and crack steps measured and documented.
Freeze-Thaw Cycle Two freezing rates were considered. The first one used a freezing rate relatively fast, 12 degrees C/hr., to test the hypothesis. The second one was performed at a rate within the cooling rate specified by the ASTM C666 standard (-7 degrees C/hr.) to confirm the Ice-Wedge cracking.
The mechanical induced crack varied in length per test coupon side (face). In order to examine the facture surface a saw cut was introduced in the opposite side of the crack and the sample was easily split open exposing the fracture surfaces. The fracture followed the visually observed crack during the Ice-Wedging experiment.
Page 31 of 80
Figure 13: Fractured open showing fracture following the depicted the mechanical and IW cracks.
Visual examination of the fracture surface revealed the presence of fracture steps. The crack was unevenly propagating when both faces were considered. On the obstructed faces the steps were greater. As illustrated, the mechanically induced crack did not start at the bottom of the notch but rather started at slightly higher location on the side wall of the notch due to the presence of a pore.
Page 32 of 80
Figure 14: The fracture surface of S11-671-26.5 (2) showing the mechanically induced crack and Ice-Wedge step fracture Summary of Results of High Freezing Rate 12 degrees C/Hr. Tests
- 1. Both samples tested at 12 degrees C/hr. freezing rate exhibited crack propagation (beyond the mechanically induced crack) due to Ice-Wedge mechanism.
- 2. The Ice-Wedge cracks were found to have limited crack propagation (steps) every time the samples were exposed to Ice-Wedge cycle. The crack step spacing in samples were in the 0.4 in to 0.7 in regardless of the sample mechanical loading history. In one sample the mechanical loading history affected the overall crack length as soon as the Ice-Wedge crack connected to a pre-existing crack from the splitting tension test.
- 3. For Ice-Wedge cracking to occur in shield building concrete, there are three prerequisites:
a) pre-existing crack, b) water in the crack, and c) freezing conditions.
Summary of Results of Freezing Rate -7 degrees C/Hr. Tests
- 1. The two tested samples confirmed crack propagation as a result of Ice-Wedge cycles at a lower freezing rate within ASTM standard C666.
- 2. The Ice-Wedge cracking occurred in limited crack propagation (steps) similar to that observed at the high freezing rate samples.
- 3. The crack step spacing per Ice-Wedge cycle on samples 0.3 in to 0.6 in. These values are similar with the range for the samples tested at higher freezing rate of 12 degrees C/hr.
- 4. The overall numerical average of the Ice-Wedge steps for all four samples is 0.49 in.
Page 33 of 80
3.3.14 Structural Finite Element Ice-Wedging Model An Ice-Wedging crack propagation analysis was performed via Finite Element Analysis. This cracking model uses the same Concrete Damaged Plasticity (CDP) approach as was used in RCA-1. This model is driven by inputs from several different sources. The thermal gradients are driven by a simple transient thermal model with the same mesh as the structural cracking model. They are both 2D slices, exactly the same as the structural calibration models - except the structural sub model in this report is driven with thermal gradients dimensional temperature data (X, Y, and time) at a fixed Z.
The 6 days analyzed (January 25th - 30th) gives two clear freeze-thaw cycles - one is shallow with a relatively small change in temperature. The second cycle is much deeper - the concrete temperature goes down as low as about 20 degrees F at a depth 7 in from the surface. The duration of the second cycle is also a lot longer. The approximate temperature at the depth of the laminar crack is shown as the red line in Figure 15.
Figure 15: Temperature at the Laminar Crack Plane The method of creating an initial crack is exactly the same as the procedure used in the structural submodel calibration analysis.
The results are best explained graphically. Figure 16 shows the results of the structural submodel with Thermal Gradients driven by the first 6 days of data.
Page 34 of 80
Figure 16: Results In chronological order:
- 1. The dark blue line is the lab data from the Ice-Wedging testing done at University of Colorado.
- 2. The red line is the final calibration selected from the Ice-Wedging lab sample calibration analyses.
- 3. The dark reddish brown line is the calibration from the structural submodel without thermal gradients.
- 4. The light blue is the first 6 days of the data driving the structural submodel, using exactly the same calibration parameters as the structural submodel and Ice-Wedging model calibrations.
- 5. The purple line is an estimated lower bound for the crack propagation (see below).
Discussion Interestingly, the response changed with the thermal gradient evolution in the model. The total amount of cracking after 2 cycles, however, is roughly the same.
A second analysis was undertaken to predict a lower bound for the crack growth if the concrete is able to retain more of its fracture toughness and stiffness during the first step. Essentially, the calibration parameters from the second step were used for both steps. This produces only about a 20% change in the total crack growth over two cycles. This change is small considering that changing the maximum allowable damage from 0.99 to 0.90 is equivalent to changing the residual strength and stiffness from 1% to 10%. Thus, the difference in the residual fracture toughness and stiffness is potentially a factor of ten different. Yet, the results are only reduced by about 20% despite the large change in the most important input parameter. This indicates Page 35 of 80
that the model is relatively stable and insensitive to potentially large changes in some assumptions.
Conclusion The average crack propagation over 2 cycles is calculated to be 0.48 in. A bounding set of analyses shows a lower bound on the crack length of 0.36 in per cycle. One cycle resulted in a maximum crack propagation length of 0.81 in. This average value of 0.48 in agrees favorably with laboratory tests. The range from 0.36 to 0.81 in also agrees favorably with the laboratory examinations. The data and the analyses are consistent, both confirming a typical crack propagation of roughly 0.5 in per cycle.
3.3.15 Data Analysis Conclusions The laminar crack propagation discovered in 2013 by Davis-Besse was confirmed by the PII root cause investigation team during its 2013 shield building inspection. In addition, water was discovered in S5-666.0-8 and S5-666.0-10 during the same visit. Further inquiries revealed that more water findings were recorded by the plant staff during previous bore inspections that took place in the previous 2 years. Water was confirmed to be present in the structure when several bores were sealed from the outside environment and monitored for water accumulation.
A petrographic examination of the core samples was also conducted. At multiple depths, evidence of Freeze-Thaw damage and evidence of water transport in the form of Ettringite crystals formation and microcracks emanating from pores was found. The extracted core samples fracture surface was also examined in detail and step fractures were observed at the laminar crack location indicating progressive fracture.
Lab tests were performed to test the validity of the Ice-Wedging FM. The testing took into consideration: 1) a pre-existing crack, 2) water in the crack, and 3) Ice-Wedge cycles. The results revealed a step crack propagation after each Ice-Wedge cycle. The average crack propagation step was 0.5 in. The step fracture produced by the Ice-Wedge testing was similar to the step fracture observed in the sample extracted from the shield building which was measured as 0.6 in on average.
Thermal and Structural Finite Element Analyses were conducted to further determine the plausibility of Ice-Wedging as a failure mechanism. The thermal transient analysis determined the through-wall temperature profile for the shield building for the winter periods desired. The results from the thermal transient analysis showed that freezing temperatures at the location of the laminar cracks under consideration were present for the time period of interest. The number of Ice-Wedge cycles calculated were multiplied by 0.6 in (measured crack propagation step width) to determine the crack propagation rate over the last two years. The Ice-Wedge crack propagation rates per location were in the same order of magnitude and less than the maximum estimated. In addition, a structural analysis was also conducted based on pre-existing laminar cracks expanding due to the 9% volume increase during the water to ice phase change. The model showed that crack propagation occurred on the same order of magnitude as that calculated in the laboratory and the field.
Page 36 of 80
Causal Factors Ice-Wedging (failure mode 16) has been determined to be responsible for the observed crack propagation. Ice-Wedging is identified as Causal Factor 1. Based on pre-existing laminar cracks, water confirmed to be present in the structure, freezing temperatures during the 2011-2013 winters at identified crack locations, fracture analyses of core samples, applicable laboratory experiments, and FEA results, the three conditions required in order for Ice-Wedging to occur have been confirmed.
Contributing to the Ice-Wedging cause is application of the coating to the shield building. While application of the coating has effectively prevented water from entering the shield building, its application has also prevented a finite amount of moisture from leaving the structure (failure mode 6). Until this moisture dissipates it provides the water accumulation mechanism required for Ice-Wedging, and therefore is identified as Causal Factor 2.
3.3.16 Hardware Disposition The initial condition assessment of the laminar crack propagation determined that the shield building was functional and within design conformance. The initial condition assessment concluded that no compensatory actions or operating restrictions were required due to the shield building concrete laminar crack propagation. A review of engineering analysis documentation developed following the initial laminar crack condition, demonstrated that the shield building remained structurally adequate for the controlling load case(s) and is in compliance to the current design and licensing bases.
3.3.17 Comparison of Findings from RCA-1 and RCA-2 Based on the RCA-2 findings showing Ice-Wedging as the failure mechanism for those regions examined, PII revisited RCA-1 to determine the extent to which the Ice-Wedging mechanism may have played a role in the observed cracking of RCA-1 samples.
The laminar cracks discovered in 2011 during the 17TH re-fueling outage led Davis-Besse to bore 80 cores as previously discussed. Of the total cores, PII examined approximately 30 during RCA-1 and re-examined 16 during RCA-2. The fracture surface of the cores listed in the table below were all re-examined during RCA-2 for step-cracking. Two of the 16, marked by an asterisk, represent the cores that were examined for microcracks as well:
Page 37 of 80
Table 4: RCA-1 Samples Examined by PII in 2011 and those Re-Examined during RCA-2 2011 Sample ID Final Sample ID F3-1 F3-1-666.0-3 S5-1* S5-666.0-8*
S5-2* S5-666.0-10*
S7-1 S7-666.0-7 S7-2 S7-666.0-9 S7-3 S7-667.0-25 S7-656.6-6 S7-656.6-6 S9-1 S9-666.0-11 S9-2 S9-666.0-12 S9-653-11 S9-653-11 S9-785-22.5 S9-785-22.5 S11-1 S11-663.75-30 S11-2 S11-663.75-32 S12-1 S12-666.0-2 S12-2 S12-666.0-4 S16-3 S16-613.0-46 At the time of the RCA-1 investigation, the fracture surfaces underwent a fracture and petrographic analysis for which no indications of step-cracking were identified across the core.
The following figures show fracture surfaces of select RCA-1 samples examined and re-examined by PII:
Page 38 of 80
Figure 17: S5-666.0-8 Fracture Surface with no Evidence of Step Cracking Figure18: S11-663.75-30 Page 39 of 80
However, fracture surfaces from Ice-Wedging laboratory tests match the fracture surface of RCA-2 sample S5-666.0-9.5 in both topography and crack magnitude, which makes this dissimilar to what was observed in RCA-1 samples. Since examination of RCA-1 fracture surfaces showed cracks passing through the aggregate and with no signs of step-cracking across the core, the indication is a sudden laminar crack occurring during one event. Therefore in retrospect, for those samples examined by PII, Ice-Wedging was not the mechanism that caused the observed cracking in RCA-1. Consequently the conclusion from RCA-1 still stands, that is, for the examined regions of the building, the 1978 Blizzard was the single event that led to the observed cracking.
Furthermore, during RCA-2, some of the RCA-1 samples were re-examined through Scanning Electron Microscopy for which microcracking was discovered. Contrary to failure mode FM 3.16 of the RCA-1 report, optical microscopes were used during that investigation and microcracking was accurately reported as not discovered due to the resolution of optical microscopes.
However, re-examination using SEM images shows that RCA-1 samples did in fact possess micro-cracks emanating from the pores, which confirms freeze-thaw damage. The following figures show the difference between an optical image and an SEM image of the same RCA-1 sample (note how microcracking is seemingly invisible in the optical image):
Figure 19: Sample S9-653-11 under Optical Microscope (Microcracks Invisible)
Page 40 of 80
Figure 20: Sample S9-653-11 under Scanning Electron Microscope (Microcracks Visible)
It is critical to note that Ice-Wedging damage will also show freeze-thaw damage (due to pore-saturation surrounding water-filled crack tips) but freeze-thaw damage does not necessarily indicate Ice-Wedging (as no pre-existing cracks are needed for this phenomenon). As proof, core S5-666.0-10 which was extracted in 2011 showed significant freeze-thaw damage (core re-inspected during RCA-2) but did not possess a laminar crack until the crack from S5-666.0-8 propagated through it; the propagation was post-2011 and had to have occurred before 2013 as further described in section 14.04d of RCA-2. However, examination of 2011 core sample S5-666.0-8 shows significantly less freeze-thaw damage than S5-666.0-10 although it possessed a laminar crack in 2011.
Recall, in order for Ice-Wedge cracking to occur 3 parameters must exist: 1) pre-existing laminar cracks, 2) water at the crack tips, and 3) freezing temperatures at the location of the crack. Water discovered during RCA-2 led the root cause team to expand the failure modes based on a phenomenon that occurs in nature and is familiar to the geological sciences, Ice-Wedging.
Conclusions In summary, the conclusion of RCA-1 that all the examined cracks were a result of a one-time event, i.e., the 1978 blizzard which produced extreme high stress to cause general delamination, is still valid. Furthermore, the conclusion in RCA-1 that the general delamination is not likely to propagate under thermal fatigue is also valid. Due to the fact that there was no step-wise fracture surface or microcracks at the aggregate-cement interfaces in all the bores examined, PII concluded in RCA-1 that the general delamination was not likely to propagate. Note that in RCA-Page 41 of 80
1, Ice-Wedging was not considered because it had not been known to be involved in concrete crack initiation or propagation.
In RCA-2, through SEM examination of the fracture surface of one additional core at the lower portion of the shield building (S5-666.0-9.5) which was involved in crack propagation between RCA-1 and RCA-2, and the observation of water in core bores, PII is able to hypothesize that Ice-Wedging could be a credible crack propagation mechanism. The Ice-Wedging failure mode is localized at the portion of the shield building where the water most likely accumulates in pre-existing cracks. This hypothesis of Ice-Wedging was later supported by results from both laboratory experiments and detailed computer analysis.
Page 42 of 80
4 Latent Organizational Weakness Evaluation A Latent Organizational Weakness (LOW) evaluation was performed in accordance with NORM-LP-2003, Analytical Methods Guidebook, Attachment 14.
In Davis-Besses laminar cracking root cause analysis under RCA-1, the failure mechanism was determined to be the laminar stress induced by outer surface ice volume expansion as large volumes of water were driven into the shield building just before the 1978 Blizzard. As a preventative corrective action, a protective coating was applied to the shield building between August and October 2012.
RCA-2 determined that the failure mechanism for crack propagation to be caused by Ice-Wedging which occurs when water accumulates at a pre-existing crack tip and freezes. During the water to ice phase change, there is a 9% volumetric expansion which results in a force exerted on the surrounding concrete. Due to the concretes inability to elastically absorb the resulting force, the crack propagates to relieve the induced stresses.
In order for Ice-Wedging-induced cracking to propagate in the shield building, three conditions must be met:
- 1. Pre-existing cracks
- 2. Water accumulation at the crack locations
- 3. Freezing temperatures at the crack locations The shield building has met all three of these conditions and that there was enough moisture sealed within the shield building as a result of the coatings application to contribute to Ice-Wedging. While application of the coatings has been determined to be a contributing cause to laminar crack propagation, the decision making process that went into coating the shield building does not represent a Latent Organizational Weakness.
The presence of moisture is inherent in any concrete structure, and as in the case of the shield building, it was not believed to pose any challenges to the coating effort. Water discovered in plugged bores prior to coating application was believed to have entered from the outside environment, however the possibility of existing water within the shield building was posed as feasible. However, the belief was that had the water come from inside of the shield building, the amounts discovered were small enough to present no adverse effect to the shield building.
The decision making process which was followed as part of the corrective action development, in order to prevent wind driven rain from entering the shield building concrete was in accordance with industry standards and specifications for the application that was required; that being the ability to prevent rain driven water from entering the concrete.
Page 43 of 80
In addition, based on the extensive analysis performed under RCA-2, Ice-Wedging in concrete was not previously known as a potential failure mode or part of a failure mechanism.
As part of the coatings selection process specification A-028N was developed and identified coating system characteristics such as resistance to 98 mph wind driven rain per ASTM D6904 to bound historical wind speeds in the area of the plant, freeze-thaw resistant per ASTM D2243 to prevent failures during the winter, water-vapor permeance not greater than 18 perm per ASTM D1653 and sufficient flexibility with no cracking per ASTM D522.
In October 2012, under RCA-1 corrective actions a water-resistant coating was applied to the exterior of the entire shield building. The rough surface of the structure prevented the use of Wet Film Thickness readings which is the industry standard nondestructive method for measuring concrete coating thickness. Based on application records which identified the quantity of coating applied and the surface area of the building a general coating thickness may be calculated. Utilization of this method results in coating thickness values which are not greater than the 19.8 mils Dry Film Thickness required by specification A-028N for the three coat system.
Therefore, the discovery of four sample locations from within a single 4 in. diameter bore measured during 2014 laboratory testing to have a coating thickness greater than the specification requirements are not necessarily representative of the entire shield building. The coating was applied to the shield building to provide a mechanism for reduction of wind driven rain intrusion into the concrete. The additional thickness of the coating does not impact the ability of the coating to reduce water intrusion.
A review of this documentation concluded that the coating system met the design specification requirements and the coating was determined to have been applied in accordance with manufacturer requirements. No human performance issues were identified as a result of the coatings application.
The processes used to determine the corrective actions to be taken under RCA-1 were not deemed insufficient because the failure mode of Ice-Wedging in concrete was unknown outside of geological circles, not known in concrete and the corrective action taken to coat the shield building was deemed an industry standard. While the coating has been determined as a contributor to the laminar crack propagation the decisions that resulted in its use do not represent a Latent Organizational Weakness.
Page 44 of 80
5 Generic Implications 5.1 Plant and Industry Experience 5.1.1 Strategy During the investigation to RCA-1, fleet, regulatory and industry resources were researched for similar symptoms of laminar cracking. A second search was conducted for plant and industry experience with concrete cracking to gain knowledge regarding similar potential failure modes, causes, corrective actions, and generic problems. These reviews determined no plants have reported experiencing similar laminar cracks and the laminar cracking of the shield building is unique with respect to reinforced concrete.
As a result of the identified laminar crack propagation in 2013, FENOC contracted PII to perform a comprehensive technical cause assessment (RCA-2) to identify the cause(s) of the unexpected crack propagation. As part of PIIs assessment, industry resources were researched for similar symptoms of laminar crack propagation.
5.1.2 Results Research conducted during RCA-2 identified the failure mode of Ice-Wedging for laminar crack propagation. This research concluded that Ice-Wedging is a phenomena known to the geological sciences through observations of rock cracking, however, it has not been known or studied as related to concrete cracking. In order for Ice-Wedging to occur, three conditions are required: 1) Pre-existing laminar cracks, 2) Water at the location of the laminar crack, and 3)
Freezing temperatures at laminar crack depths. Based on the fact that Ice-Wedging in concrete is not possible without pre-existing laminar cracks, of which RCA-1 determined laminar cracking of the shield building to be unique to Davis-Besse; combined with the extensive research already performed by PII during RCA-2, additional resource review is not applicable.
5.1.3 Conclusions There have been no similar issues with design specifications, coatings, or design interfaces to indicate a generic problem. There have been no similar previously identified events with concrete laminar crack propagation at Davis-Besse or within FENOC; however the effectiveness of prior corrective actions, specifically application of the shield building coating is discussed under the Latent Organizations Weakness section of this evaluation.
Page 45 of 80
5.2 Extent of Condition 5.2.1 Strategy The shield building concrete laminar crack propagation was reviewed for the extent of condition relative to other applicable programs/processes, equipment/systems, organizations, environments, and individuals. In order for laminar crack propagation to occur, three conditions are required for the Ice-Wedging mechanism to occur: 1) a pre-existing crack, 2) water present in the crack with localized saturation, and 3) an Ice-Wedge cycle that contained a freezing condition.
5.2.2 Results As determined in RCA-1, acute freezing and expansion of moisture in the shield building concrete was the only scenario capable of generating radial stresses large enough to initiate laminar cracking. The blizzard of 1978 was determined to be the only event during the life of the shield building that integrated the moisture content, wind speed, temperature, and duration necessary for development of radial stresses large enough to enable the concrete laminar crack initiation. RCA-1 determined the extent of the laminar cracks to only be associated with the shield building; therefore no other concrete site buildings or structures are affected, as the required Ice-Wedging mechanism of a pre-existing crack can be ruled out. The evaluation, causes and corrective actions implemented under RCA-1 have not been invalidated by the results of RCA-2.
5.2.3 Conclusions Without the existence of laminar cracks it can therefore be concluded that no other concrete structures are undergoing or have undergone Ice-Wedging. Therefore, the Ice-Wedging condition has been determined to be limited to the shield building.
Page 46 of 80
6 Apparent and Contributing Causes The Apparent Cause for the shield building concrete laminar cracking propagation is due to Ice-Wedging. The three conditions required for Ice-Wedging to occur; pre-existing laminar cracks, water accumulation at the crack locations and freezing temperatures at the crack locations have been demonstrated to occur based on field, laboratory and modeling analysis and data. [Cause code T22, failure attributable to the environment]
While application of the coating has effectively prevented water from entering the shield building, its application has also prevented a finite amount of moisture from leaving the structure. Until this moisture dissipates it provides the water accumulation mechanism required for Ice-Wedging, and therefore is identified as the Contributing Cause to the laminar cracking propagation. [Cause code OG6A, methods failed to uncover causal factors of consequences]
Page 47 of 80
7 Corrective Action Plan Ice-Wedging requires three necessary conditions (i.e., water, freeze-thaw cycles, and preexisting cracks). To prevent it, would require one or more of the three constituents to be eliminated:
Removal of existing cracks, prohibit freezing temperatures and/or removal of water are not being recommended for implementation as they are not reasonable options. Additionally, if implemented could have unintended consequences. Therefore, the most reasonable and appropriate corrective actions being implemented focus on monitoring the crack propagation condition.
No cause codes have been selected for the two corrective actions being implemented as the actions are not designed to mitigate the causes. Instead of mitigation, the corrective actions are designed to monitor the causes. Monitoring has been determined to be the appropriate means to address the causes. In addition the design basis calculation; C-CSS-099.20-063, Shield Building Design calculation contains sufficient margin to support this conclusion. Therefore all three corrective actions have been assigned cause codes of N/A.
Page 48 of 80
Corrective Action Cause Cause Corrective Action Description Corrective Group Due Date or Notification Code Action Responsible Number Type 2013-14097-001 Apparent N/A In order to determine the effects of LTCA DB-EN- 10/01/2014 Cause the 2013 winter: DESN In accordance with procedure EN-DP-01511, Design Guidelines for Maintenance Rule Evaluation of Structures, inspect a total of 4 previously uncracked shield building bores in shoulders 4, 8, 9 and 12 for evidence of laminar cracking propagation following the winter of 2013. These 4 bores shall be adjacent to areas of known cracking.
2013-14097-002 Contributing N/A FENOC has identified a population LTCA DB-EN- 11/01/2014 Cause of 20 bores for long term monitoring DESN of the Shield Building Laminar Cracking condition. In order to provide adequate monitoring to ensure conformance with the plant design and licensing basis:
Revise the Shield Building Laminar Cracking Long Term Monitoring Program, under procedure EN-DP-01511 to:
1). Require inspections of, at a minimum, the 20 bores identified for long term monitoring prior to September 1 in the following years:
2015, 2016, 2017, 2018, 2020, Page 49 of 80
Corrective Action Cause Cause Corrective Action Description Corrective Group Due Date or Notification Code Action Responsible Number Type 2022, 2024, 2026, 2030, 2034, and 2037.
2). Include the inspection of 3 laminar crack leading edges as identified by changes in 2013 as part of the population for each inspection cycle at the same periodicity as the bore inspections.
Install new bores as required during each inspection cycle to bound crack limits at these locations.
3). Include the identification and recording of water as found in the bores when inspected.
4). Establish the analyzed conditions identified in Calculation C-CSS-099.20-063 Shield Building Design Calculation as acceptance criteria. (i.e. planer limits of cracking, and crack width) 2013-14097-003 N/A N/A Review C-CSS-099.20-063, Shield LTCA DB-EN- 10/01/2014 Building Design Calculation for the DESN identified shield building crack propagation.
Revise the calculation as required to:
- 1) reflect the new condition and Page 50 of 80
Corrective Action Cause Cause Corrective Action Description Corrective Group Due Date or Notification Code Action Responsible Number Type
- 2) modify the analysis (if required).
In concert with the calculation revision and associated 50.59 documents; the Updated Safety Analysis Report (USAR) shall be reviewed and revised as necessary to reflect the crack propagation.
Note: The results of this CA may affect the acceptance criteria information to be added to procedure EN-DP-01511, under CA 2013-14097-002.
DB-L-10-221- Apparent Revise the Structures Monitoring, DB-EN- 01/01/2016 LRAA.1-20 Cause and Shield Building Laminar DESN Cracking Long Term Monitoring Program, under procedure EN-DP-01511 to reflect the ACI 349 recommendations for review of cracks.
This action is a commitment FENOC is implementing as part of the License Renewal process. This commitment is tracked as Commitment DB-L-10-221-LRAA.1-20 and is therefore being tracked outside of the Corrective Action Program.
Page 51 of 80
8 Effectiveness Review Plan At the discretion of the Owning Manager, and in accordance with NOBP-LP-2011, FENOC Cause Analysis, an Effectiveness Review is not being performed, based on the fact that there are no Corrective Actions being implemented to mitigate adverse conditions. The actions being implemented under this Condition Report are all related to shield building monitoring activities, which will be tracked through the Corrective Action program.
Page 52 of 80
9 References Condition Reports CR 2011-03346, Crack Discovery Above Temporary Access Opening of Shield Building CR 2013-13782, Shield Building Core Bore S5-666.0-10 Findings CR 2013-13854, Shield Building Core Bore S7-666.0-7 Findings CR 2013-13860, Shield Building Core Bore S7-666.0-9 Findings CR 2013-14623, Shield Building Core Bore S13-633.0-11 Findings CR 2013-14961, Shield Building Core Bore S4-773-16 Findings CR 2013-16204, Shield Building Core Bore S15-646.5-8 Findings CR 2013-16210, Shield Building Core Bore S15-674.5-3 Findings CR 2013-16211, Shield Building Core Bore S15-777-3 Findings Calculations Calculation C-CSS-099.20-063, Shield Building Design Calculation Drawings C-100, Shield Building Foundation Plan & Details C-109, Shield Building Roof Plan and Details C-110, Shield Building Roof Plan, Wall Section & Details C-113, Shield Building Details, Sheet 2 C-114, Shield Building Dome Framing Plan and Details Design Changes DBNPS Engineering Change Package 10-0458, Install Shield Building Construction Opening DBNPS Engineering Change Package 12-0208, Exterior Coating Systems for the Shield Building Walls and Dome Procedures DB-PF-03009, Containment Vessel and Shield Building Visual Inspection EN-DP-01511, Design Guidelines for Maintenance Rule Evaluation of Structures EN-DP-01512, Shield Building Concrete Examination EN-DP-01511, Design Guidelines for Maintenance Rule Evaluation of Structures Specifications Specification A-028N, Design Specification for Shield Building Coating System Descriptions SD-0022D, Containment Vacuum Relief System Work Orders FENOC Order 200008657, Restore the Containment Shield Building FENOC Order 200502209, Exterior Coating Application for Shield Building Walls and Dome Page 53 of 80
Industry Documents American Concrete Institute (ACI) 307-69, Specification for the Design and Construction of Reinforced Concrete Chimneys ACI 318-63, Building Code Requirements for Reinforced Concrete ACI 349-06, Code Requirements for Nuclear Safety-Related Concrete Structures &
Commentary ASTM D522, Standard Test Methods for Mandrel Bend Test of Attached Organic Coatings ASTM D1653, Standard Test Methods for Water Vapor Transmission of Organic Coating Films ASTM D2243, Standard Test Method for Freeze-Thaw Resistance of Water-Borne Coatings ASTM D6904, Standard Practice for Resistance to Wind-Driven Rain for Exterior Coatings Applied on Masonry Other References Performance Improvement International, Root Cause Analysis - Laminar Crack Condition of Davis-Besse Shield Building, 6/18/2014 Performance Improvement International, Root Cause Assessment - Davis-Besse Shield Building Laminar Cracking, 4/18/2012 Performance Improvement International, Root Cause Assessment - Davis-Besse Shield Building Laminar Cracking, 2/27/2012 Page 54 of 80
10 Attachments Attachment 1, Event and Causal Factors Chart January 28 - 31, 1978 January 28 - 31. 1978 Winter 1978 - Winter 2013 October 10, 2011 Blizzard of 1978 Laminar cracking occurs Ice-Wedging results in Discovery of laminar crack propagation cracking CF Wind driven rains at over 100 mph Pre-existing laminar cracks Saturated concrete Moisture at crack locations Sudden drop in temperature below 0 degrees F Freezing temperatures at crack locations Blizzard conditions Page 55 of 80
Winter 2011 February 2012 May 2012 June - July 2012 Water discovered in bore FENOC boroscope Impulse Response testing Ice-Wedging & confirmatory bore located in shoulder inspections phenomenon results in inspections crack propagation CF 12 bores inspected Extent of condition Pre-existing laminar performed - shield cracks building & other structures Crack growth not Moisture at crack detected locations Water discovered in bores Freezing temperatures at crack locations Page 56 of 80
August 2012 August - October 2012 Winter 2012 Bore inspections Coatingsappliedtoshield Ice-Wedging building phenomenon results in crack propagation Provides a barrier Pre-existing laminar No crack growth to restrict moisture cracks detected entry Moisture at crack Coating restricts locations Water discovered in internal moisture from bores located in evaporating shoulder CF Freezing temperatures at crack locations Page 57 of 80
September 11, 2013 June & August 2013 September 2013 Bore inspections Bore inspection Discovery of laminar crack propagation Newly identified crack which was a pre-existing condition New boroscope with improved optics / greatly improved resolution Finding results in 100% verification of bores Page 58 of 80
BARRIER ANALYSIS WORKSHEET Attachment 2 CONSEQUENCE/ADVERSE EFFECT BARRIER THAT SHOULD HAVE PRECLUDED BARRIER CONSEQUENCE ASSESSMENT (WHY THE BARRIER FAILED)
Ice-Wedging leading to laminar crack Ice-Wedging was unknown in the concrete community, and There was no barrier propagation the corrective action to coat the shield building was based that existed for this on the industry practices, with input from industry experts condition.
during RCA-1. There was no barrier that existed for this condition. The presence of moisture is inherent in any concrete structure, and as in the case of the Shield Building, it was not believed to pose any challenges to the coating effort. Water discovered in plugged bores prior to coating application was believed to have entered from the outside environment, however the possibility of existing water within the shield building was posed as feasible.
However, the belief was that had the water come from inside of the shield building, the amounts discovered were small enough to present no adverse effect to the Shield Building. The specification developed which was the basis for coatings used did however provide a positive barrier in that the coatings did perform as required by keeping water from entering the structure. In addition, the bore monitoring program functioned as intended by FENOC discovering the crack changes; the unintended consequences of the coatings, instead of the issue presenting itself or by discovery outside of FENOC.
Page 59 of 80
LATENT ORGANIZATIONAL WEAKNESS FORM Attachment 3 (Attachment 14 to NORM-LP-2003)
CR 2013-14097, Laminar Crack Propagation Problem Statement:
Between October 2011 and September 2013 laminar cracks have propagated in the Shield Building. Eight core bores exhibit evidence of new cracks discovered in the outer rebar mat within the shoulder region of the Shield Building during core bore inspections conducted in August and September 2013. Six of the bores were last inspected in October 2011, with the remaining two bores inspected in 2012 and 2013.
Six bores exhibit crack widths of less than 0.005 inches, while the remaining two bores crack width measure less than 0.007 inches and 0.009 inches respectfully.
Step 1: Event and Causal Factor Chart: YES Step 2: Listing of Inappropriate / Unintended Actions:
IA # Date Who? What? Requirement or Where?
Expectation?
1 Aug to Oct FENOC No awareness of Requirement. NOP-LP-2003.
2012 - date the effects of coating (unintended con- CAs shall be application sequences) of written coating the shield SMARTER building R = Reviewed for unintended consequences.
No knowledge or experience within There were no the concrete known con-industry sequences regarding Ice-(Ice-Wedging)
Wedging as a failure mode The coating was applied based on industry best practices.
Step 3: Description of the initiating action(s):
Coating of the shield building results in increased moisture.
Page 60 of 80
Step 4: Determine the context of the decisions made by the individual(s) committing the action:
Initiating Action Decision to coat the Shield Building Goal Eliminate water intrusion from exterior of Shield Building Focus Determine best method / material to prevent intrusion Awareness Awareness based on eliminating water intrusion. Industry standards followed in determining best materials, application and vendor.
Step 5: Provide GEMS information for each inappropriate action identified in step 2 above.
IA # SB RB KB Discussion Justification 1 X Coating was selected per Systemic approach went industry standard / spec. Ice- into decision making wedging failure mode unknown process to eliminate outside of geological circles (not possibility of wind driven known in concrete) rain entering Shield Building. Product requirements, selection of product, vendor.
Step 6: Consider conducting a Human Performance Gap Analysis as described in 7. This human performance gap analysis should be conducted by the cause evaluator or team rather than by an individual who reports to the site Training Department.
A performance gap analysis is not necessary based on no individual inappropriate actions. No individual error precursors were identified related to task demands, individual capabilities, work environment, or human nature. Coating was selected per industry standard / spec. Ice-wedging failure mode unknown outside of geological circles (not known in concrete).
Step 7: Describe error precursors (undesirable, prior conditions that reduced the opportunity for success at the job site) that contributed to the event:
The presence of moisture is inherent in any concrete structure, and as in the case of the shield building, it was not believed to pose any challenges to the coatings effort. Water discovered in plugged bores was believed to have entered from the outside environment, not from inside of the structure.
The decision making process which was followed as part of the Corrective Action develop-ment was in accordance with industry standards and specifications for the application that was required of the shield building, that being the ability to prevent water driven rain from entering the concrete.
Page 61 of 80
Error Precursors that contributed to Basis / How did it breakdown and contribute to event?
event Unknown risk Eliminating future water intrusion from wind driven rain was of perception concern with the Corrective Action. The amount of water contained in the Shield Building is finite and not considered as a potential failure mechanism or part of a failure mechanism. The presence of moisture / water contributed to the Ice-Wedging mechanism.
Assumptions Water discovered in plugged bores prior to coating was believed to be from the outside environment, or if from shield building it was believed to be minor amount which did not present an adverse condition.
For each error precursor, identify whether it was ever evaluated for risk and, if so, the results of that evaluation.
Error Precursors that contributed to Was it ever evaluated for risk? Results?
event Unknown risk It was unknown at the time, therefore no it was not evaluated for perception risk. Industry best practices were followed. Unknown failure mode within concrete community.
Assumptions Yes. The presence of moisture is inherent in any concrete structure, and as in the case of the shield building, it was not believed to pose any challenges to the coating effort.
Water discovered in plugged bores prior to coating application was believed to have entered from the outside environment, however the possibility of existing water within the shield building was posed as feasible. However, the belief was that had the water come from inside of the shield building, the amounts discovered were small enough to present no adverse effect to the shield building.
Page 62 of 80
Step 8: Select any of the following error precursors that apply and address the questions that follow:
Task Demands Individual Capabilities Time Pressure Unfamiliar with task/first time High workload Lack of knowledge Simultaneous, Multiple tasks New technique not used before Repetitive actions/monotony Imprecise communication habits Irrecoverable Actions Lack of proficiency/inexperience Interpretation Requirements Indistinct problem solving skills Unclear goals, roles, or responsibilities Can do attitude for crucial task Lack of/unclear standards Illness or fatigue Other ________________ Other ________________
Work Environment Human Nature Distractions/interruptions Stress Changes/departure from routine Habit patterns Confusing displays/controls Assumptions Work around/OOS instrumentation Complacent/overconfidence Hidden system response Mind set (intentions)
Unexpected equipment conditions Unknown risk perception Lack of alternate indication Mental shortcuts (biases)
Personality conflicts Limited short term memory Other ________________ Other ________________
Did the work evolution require that a pre-job brief be performed? If not, why not?
Yes Were the error precursors identified and discussed during the pre-job brief? If not, why not?
While the development of SMARTER CAs would be part of a pre-job brief, the brief would be limited to that extent. The brief would not discuss actual proposed CAs, as that would come out of the CR evaluation process itself.
If error precursors were identified during the pre-job brief, what defenses were discussed to address the error-likely situation?
N/A Page 63 of 80
Step 9: Identify all hazards, targets and barriers and evaluate each group:
BARRIER ANALYSIS WORKSHEET CONSEQUENCE/ADVERSE BARRIER THAT SHOULD HAVE PRECLUDED CONSEQUENCE BARRIER EFFECT ASSESSMENT (WHY THE BARRIER FAILED)
Ice-Wedging leading to laminar Ice-Wedging was unknown in the concrete community, and the There was no barrier crack propagation corrective action to coat the shield building was based on the that existed for this industry practices, with input from industry experts during RCA-1. condition.
There was no barrier that existed for this condition. The presence of moisture is inherent in any concrete structure, and as in the case of the Shield Building, it was not believed to pose any challenges to the coating effort. Water discovered in plugged bores prior to coating application was believed to have entered from the outside environment, however the possibility of existing water within the shield building was posed as feasible. However, the belief was that had the water come from inside of the shield building, the amounts discovered were small enough to present no adverse effect to the Shield Building. The specification developed which was the basis for coatings used did however provide a positive barrier in that the coatings did perform as required by keeping water from entering the structure. In addition, the bore monitoring program functioned as intended by FENOC discovering the crack changes; the unintended consequences of the coatings, instead of the issue presenting itself or by discovery outside of FENOC.
Page 64 of 80
Flawed Defects/Defenses - defects in measures designed to protect plant equipment or people against hazards or to prevent the occurrence of active errors; functions include awareness, detection, warning, protection, recovery, containment, and escape.
Select any of the following that apply:
Pre-job Brief Inadequacies Place keeping deficiencies Procedure Use & Adherence weaknesses Document Preparation Ineffective communications Supervision/Management Turnover Brief deficiencies Workmanship Planning weaknesses Post-Job Critique Inadequate Verification Practices Tool/Equipment Use Insufficient Self check/STAR Safety Insufficient Peer check Radworker Practice Insufficient First Check Outside of Procedures Inadequate Questioning attitude Flagging Flawed Defenses Questions:
What defenses/barriers broke down that allowed the event to occur? Why were they allowed to fail? List and justify. (Relate the answer to the above choices. Only select and justify the most pertinent defenses.)
Defenses/Barriers Basis/How did it breakdown and contribute to event?
that broke down N/A N/A Defenses/Barriers Why were they allowed to fail?
that broke down N/A N/A Describe effective barriers that could have mitigated or prevented this event?
Had Ice-Wedging been known in concrete at the time, effective barriers could have affected the decision making process.
Step 10: Advanced analysis (e.g., TapRooT) completed: Yes Summarize the results:
PII advanced cause analysis was completed as part of this CR evaluation.
Step 11: Safety Culture Evaluation completed: Not Required Summarize the results:
A Safety Culture Evaluation is not required for a Full Apparent Cause evaluation.
Page 65 of 80
Step 12: Evaluate each error precursor to determine the organizational source(s)
(Latent Organizational Weaknesses - LOW) that originally established the error precursor at that job site. Identify the results of the risk evaluation and whether appropriate defenses or barriers are in place to reduce or eliminate the precursor.
Eliminating future water intrusion from wind driven rain was of concern with the Corrective Action. The failure mechanism of Ice-Wedging was unknown in the concrete community at the time. There were no latent organizational weaknesses that established error precursors at the job site. While appropriate defenses are, and were in place; as part of decision making; corrective action development (SMARTER), this unknown failure mechanism was unknown.
Step 13: Evaluate each barrier to determine the organizational source(s) (Latent Organizational Weaknesses - LOW) that allowed the barriers to fail or that did not include the missing barriers.
N/A, as described in Step 9 of this Attachment.
Step 14: Identify any organizational, programmatic or management issues or drivers that were factors in the event.
Latent Organizational Weaknesses - undetected deficiencies in the management control processes (e.g., strategy, policies, work control, training, and resource allocation), or associated values (shared beliefs, attitudes, norms, and assumptions), that create work-place conditions that provoke error (precursors) or degrade the integrity of defenses (flawed defenses).
Select any of the following that apply:
Basic Work Practice weaknesses Inadequate Program Control Inadequate Control of Personnel Resources Human Factor Deficiencies Inadequate Documentation Radiation Work Practice Deficiencies Equipment Challenges Security Work Practice deficiencies Inadequate Housekeeping Tools & Equipment Use Challenges Safety Work Practice Deficiencies Training & Qualification deficiencies Management Ineffectiveness Inadequate Work Planning & Execution Material Unavailability Inadequate Work Scheduling Oversight Weaknesses Other (provide details)
Latent Organizational Weaknesses Questions: (Relate the answer to the above choices.
Do not choose too many. Only select and justify the most pertinent latent organizational weaknesses.)
Were there latent organizational weaknesses that were identified as contributing to this event? List and justify (e.g., workarounds became accepted practices; habits allowed to develop over time; pre-job briefs were felt to be unnecessary).
Page 66 of 80
Latent Organizational Bases / How did it contribute to the event?
Weakness that failed N/A No latent organizational weaknesses exist.
Latent Organizational Why had it not been previously addressed?
Weakness that failed N/A No latent organizational weaknesses exist.
Step 15 Consider to the following questions:
These questions do not need to be answered explicitly; however, the evaluators should consider the questions.
- 1. Are there organizational factors that lead to negative conditions prior to the inappropriate actions or problem occurrence (prevention)?
- 2. Are there organizational factors that created negative conditions or actions in the course of the inappropriate actions or problem occurrence (prevention and detection)?
- 3. Are there organizational factors that created conditions or actions following the inappropriate actions or problem occurrence (detection) that didnt allow the organization to detect the problem?
- 4. Are there organizational factors that degraded barriers which could have prevented or minimized the problem occurrence?
- 5. Is the event a result of repeat minor failures? For example, did multiple valve or similar valve failures occurred prior to the initiating event or did the results of multiple procedure adherence analysis documents fail to correct a procedure adherence problem?
- 6. Did the organization not adequately address precursors to this event? Consider whether this event could be the symptom of a larger organizational issue.
- 7. Were error precursors to the event adequately evaluated for risk and were appropriate defenses in place when the event occurred.
- 8. Are there organizational/cultural influences that resulted in poor decision-making?
Or less than adequate ownership?
Page 67 of 80
Attachment 4 - Failure Modes Major Contributor: Excess Water Failure Modes: 6, 7, 11, 13, 14, 15, 16 and 17 FM 6: Cumulative Water Damage from Coating the Structure in 2012 Verified Supporting Evidence:
- 1. Water present in bores
- 2. Moisture level in concrete increased from 65% as measured in 2011 to 90-100% as measured in 2013
- 3. Presence of water damage in the outer most layer of the concrete wall
- 4. Evidence of moisture transport from cross-section analysis
- 5. Water analyses support water transport within the concrete Verified Refuting Evidence:
- 1. Core bore S5-666-10 had exhibited water damage, yet laminar crack did not develop until later suggesting that excess water can cause Freeze-Thaw damage and microcracks but not laminar cracks.
- 2. Water discovered in 2012 prior to coating the structure.
Discussion: In order for a laminar crack to propagate taking into account the presence of excess water and Freeze-Thaw conditions a pre-existing laminar crack had to be present in order for Ice-Wedging to take place. Therefore the application of the coating on the vertical surfaces, thus in turn entrapping the water in the outer most surface layer, alone may have not caused laminar crack propagation unless a pre-existing laminar crack is present.
==
Conclusion:==
Only feasible when combined with existing laminar cracks and freeze-thaw conditions. This is categorized as a contributing factor.
Page 68 of 80
FM 7: Moisture Gradient Induced Stress Verified Supporting Evidence:
- 1. The measured RH revealed a gradient of moisture from around 90-100%
Verified Refuting Evidence:
- 1. No abrupt moisture gradient to develop appreciable pore pressures
- 2. Cracking is in the circumferential direction Discussion: In general, for a hardened concrete of 40 year old under ambient environment, drying shrinkage does not cause cracking.
==
Conclusion:==
Refuted FM 11: Freeze-Thaw Damage Verified Supporting Evidence:
- 1. Microcracks propagating through fine and coarse aggregate.
- 2. Evidence of moisture and moisture transport through multiple inputs
- 3. Thermal analysis via FEA, revealed freezing temperatures at multiple depths inside the concrete Verified Refuting Evidence:
- 1. Maximum microcracking was found at the outermost 2 of the structure, yet no laminar cracks were observed at that depth.
Discussion: The presence of high relative humidity, 90-100% as measured in the first 8 inches of the outer most layer, in combination with Freeze-Thaw temperature exposure will result in internal microcracking.
==
Conclusion:==
Only feasible when combined with existing laminar cracks and freeze-thaw conditions. This is categorized as potentially a contributing factor.
Page 69 of 80
FM 13: Material Degradation due to repeat Freeze/Thaw Verified Supporting Evidence:
- 1. Freeze-thaw damage identified in core-bore S5-666.0-9.5 and also recently identified (during RCA-2) in other core samples removed in 2011.
- 2. The FEA performed to determine occurrence of F/T cycles at crack depths in addition to review of historical F/T cycles confirmed the occurrence of F/T cycles at laminar crack depths.
- 3. Water analysis confirmed water transport within the structure.
Verified Refuting Evidence:
- 1. Measured compressive strength from samples collected during RCA-2 efforts revealed compressive strengths above the Shield Buildings design compressive strength of 4000psi.
Discussion: As a result of repeat exposure to Freeze-Thaw cycles, microcracks were developed. However, recent material property testing revealed no reduction in the mechanical properties as a consequence of F/T cycles.
==
Conclusion:==
Refuted FM 14: Freeze-Thaw via Water Accumulation due to External Water Source x This FM has been consolidated with FM 11.
Verified Supporting Evidence:
- 1. Water analysis revealed that water collected had evidence of water transport within the structure in the vicinity of the rebar.
Verified Refuting Evidence:
- 1. The water introduced in the 1977/1978 blizzards would have reached ambient equilibrium over the 36 years of operation.
Discussion: The roof section of the structure (flat surface in between the parapet in the dome) is considered a good snow-collecting surface. Upon melting of the snow and with cracks present (as observed during 2012 inspection, prior to adding sealant) near the parapet/surface interface, water can seep into the structural wall. However, shield building coating was completed in October of 2012 and would therefore prevent subsequent water intrusion.
==
Conclusion:==
See FM 11 for Freeze-Thaw Conclusions Page 70 of 80
FM 15: Freeze-Thaw via Water Accumulation due to Existing Water in the Structure x This FM has been consolidated with FM 11.
Verified Supporting Evidence:
- 1. Thermal gradient will drive the moisture normally found in the concrete to the outer most layer and concentrate it due to lack of breathing surface due to presence of sealing coating Verified Refuting Evidence:
- 1. Water was discovered in bores prior to coating Discussion: Typical cured concrete will have a moisture content in the range of 45-65%. The presence of thermal gradient across the concrete (ID hotter than OD), will tend to drive the moisture to the outer most layer and saturating it in that area. The presence of sealant coating will prevent the driven moisture from leaving the structure and saturate the moisture in the laminar crack zone (within the outer most layer).
==
Conclusion:==
See FM 11 for Freeze-Thaw Conclusions FM 16: Ice-Wedging Verified Supporting Evidence:
- 1. Tested core sample revealed very high moisture content, between 90-100% in the first 8 inches
- 2. Water analysis revealed water transport within the concrete.
- 3. Step fracture found in core sample S5-666.0-9.5 Verified Refuting Evidence:
None Discussion: Ice-Wedging occurs when water accumulates in a cracked section of concrete and expands by a volume of 9% upon freezing. The force exerted by the Ice-Wedge on the adjacent concrete faces causes existing cracks to propagate.
The Ice-Wedge lab tests revealed the criteria for Ice-Wedge crack propagation; three elements has to be present : 1) a pre-existing crack, 2) water present inside the crack and localized water saturation in the crack tip area, and 3) Freeze-Thaw condition.
==
Conclusion:==
Causal Factor Page 71 of 80
FM 17: Reverse Coefficient of Thermal Expansion Verified Supporting Evidence:
- 1. Newer material test data shows that the concrete has a different CTE than previously measured. The difference in CTE was attributed to change in moisture content. This difference will minimally contribute to this failure mode.
Verified Refuting Evidence:
- 1. There have been no events with the required conditions similar to the 1978 Great Blizzard Event over the last 2 years that would cause a CTE reversal.
Discussion: Review of historical temperature data in the last two years at laminar crack depth did not reveal any temperature drop to 10o F at which reverse CTE damage will occur.
==
Conclusion:==
Refuted Page 72 of 80
Major Contributor: Thermal Stresses Failure Modes: 2, 10 and 12 FM 2: Thermal Stress Cycles Verified Supporting Evidence:
- 1. Stress analysis confirms that in special circumstances (such as the inside of the concrete being colder than the outside) the radial stresses at a crack tip can exceed the fracture strength of the concrete)
Verified Refuting Evidence:
- 1. Thermal analysis of the Shield Building revealed the exposure to multiple Freeze-Thaw cycles
- 2. Microcracks were found emanating from internal pores in the concrete indicative of Freeze-Thaw damage. Microcracks would not result from Thermal Cycle cracking.
Discussion: The occurrence of special circumstances (such as the inside of the concrete being colder than the outside) at which the internal temperature (~105 degrees F) is colder than the outside temperature is considered a rare occasion.
==
Conclusion:==
Refuted FM 10: Residual Stresses Verified Supporting Evidence:
None Verified Refuting Evidence:
- 1. There is no repetitive mechanical loading on the standing structure
- 2. Typical residual stress in concrete is found in structures such as pavements and landing strips
- 3. No industry recognized accurate measurement technique known to quantify residual stress in concrete Discussion: Data search on this subject revealed recent approaches to measure residual stresses, these efforts were primarily focused on residual stresses in pavement concrete that experiences repetitive mechanical loading that can cause residual stresses.
These methodologies are still in the research stage. Cracks arrest when the motivating forces can no longer overcome the fracture energy (Gf). After crack propagation, the residual energy will be some fraction of Gf. Therefore, the residual energy is smaller than Gf by definition, and cannot by itself propagate a crack.
==
Conclusion:==
Refuted Page 73 of 80
FM 12: Concrete/Rebar Interaction Verified Supporting Evidence:
- 1. Laminar cracks were found in the vicinity of rebar.
Verified Refuting Evidence:
- 1. Finite Element Analysis results of the Concrete/Rebar interaction shows a 180 psi stress which is incapable of producing or propagating a laminar crack alone.
Discussion: The 180 psi radial tensile stress at the outer face horizontal rebar is unable to crack the concrete or propagate an existing crack alone. Typical concrete tensile strength is a factor of 3 greater.
==
Conclusion:==
Refuted Page 74 of 80
Major Contributor: Drilled Open Bores Failure Modes: 1, 4 and 9 FM 1: Core Drill Induced Stresses Verified Supporting Evidence:
None Verified Refuting Evidence:
- 1. Core locations such as S5-666.0-10 did not develop any laminar cracks when drilled in late 2011 until September 2013 (two years after core-drilling)
- 2. No surface fragmentation was observed on the bore internal surface.
Discussion: Poor core drilling would result in immediate cracking of the concrete as induced stresses from the drilling process would be immediate. Latent effects such as stress concentration effects are considered in other FMs.
==
Conclusion:==
Refuted FM 4: Enhanced Thermal Gradient due to Open Bores Verified Supporting Evidence:
- 1. Presence of an open bore will allow the temperature of concrete at depth to interact with the outside temperature.
Verified Refuting Evidence:
- 1. The bores are aligned in the radial direction which is the predominant direction for heat transfer, thus, it will have a minimal impact on the thermal gradient inside the bore.
Discussion: The presence of bores will not significantly enhance the surrounding thermal gradients that induce radial stresses. The bores are aligned in the radial direction and will have minimal impact on the thermal gradient surrounding the bore in the radial direction.
==
Conclusion:==
Refuted Page 75 of 80
FM 9: Stress Concentration due to Bore Holes Verified Supporting Evidence:
- 1. FE analysis provided a stress multiplier (Kt) of 2.9 as a result of geometrical change from bore drilling.
- 2. FE analysis shows that under some circumstances hoop stresses exceed the concrete failure stress at the top and bottom of the bore holes.
Verified Refuting Evidence:
- 1. Bore holes do not produce the necessary direction of stresses to cause laminar cracks or their propagation.
- 2. The effect of stress concentrations produced by bore holes has a very limited area of occurrence.
Discussion: The stress concentrations, due to the bores, magnify hoop stresses which are perpendicular to the laminar crack. Thus, cracking ensuing from the resulting stresses would be radial in direction. However, a laminar crack requires a cracking stress in the radial direction
==
Conclusion:==
Refuted Page 76 of 80
Major Contributor: Materials Failure Modes: 8 and 13 FM 8: Material Property Variation Verified Supporting Evidence:
- 1. The IR crack pattern reveals an area around the circumference at elevation 680 free of laminar cracks. This area was found to have initial compressive strengths ~12% greater than the rest of the structure.
- 2. The IR map revealed excessive cracking where the initial compressive strength was the lowest.
Verified Refuting Evidence:
- 1. Uncertainty analysis Discussion: Although the poured sample compressive strength varies, none of the measured values were found below specifications. Freeze-Thaw cracking will occur in the concrete regardless of strength.
==
Conclusion:==
Refuted FM 13: Material Degradation due to repeat Freeze/Thaw Verified Supporting Evidence:
- 1. Freeze-thaw damage identified in core-bore S5-666.0-9.5 and also recently identified (during RCA-2) in other core samples removed in 2011.
- 2. The FEA performed to determine occurrence of F/T cycles at crack depths in addition to review of historical F/T cycles confirmed the occurrence of F/T cycles at laminar crack depths.
- 3. Water analysis confirmed water transport within the structure.
Verified Refuting Evidence:
- 1. Measured compressive strength from samples collected during RCA-2 efforts revealed compressive strengths above the Shield Buildings design compressive strength of 4000psi.
Discussion: As a result of repeat exposure to Freeze-Thaw cycles, microcracks were developed. However, recent material property testing revealed no reduction in the mechanical properties as a consequence of F/T cycles.
==
Conclusion:==
Refuted Page 77 of 80
Major Contributor: Building Settlement Failure Mode: 3 FM 3: Settlement of a Cracked Building Verified Supporting Evidence:
None Verified Refuting Evidence:
- 1. Shield Building is on bedrock Discussion: Based on the bedrock foundation, it is not likely that building settlement has occurred.
==
Conclusion:==
Refuted Page 78 of 80
Major Contributor: Crack Misidentification Failure Mode: 5 FM 5: Crack Misidentification Verified Supporting Evidence:
None Verified Refuting Evidence:
- 1. Existing cracks identified as authentic cracks Discussion: This FM addresses the potential of misidentifying the newly developed cracks. Based on review of plant inspection (review of obtained photos) in addition to PII team visual inspection of some of the newly developed cracks, the new cracks were identified as genuine laminar cracks.
==
Conclusion:==
Refuted Page 79 of 80
UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE ATOMIC SAFETY AND LICENSING BOARD
)
In the Matter of )
) Docket No. 50-346-LR FIRSTENERGY NUCLEAR OPERATING COMPANY )
)
(Davis-Besse Nuclear Power Station, Unit 1) ) July 8, 2014
)
CERTIFICATE OF SERVICE I hereby certify that, on this date, a copy of the Notification of Documents Related to the Davis-Besse Shield Building was filed through the Nuclear Regulatory Commissions E-Filing system.
Signed (electronically) by Stephen J. Burdick Stephen J. Burdick Morgan, Lewis & Bockius LLP 1111 Pennsylvania Avenue, N.W.
Washington, DC 20004 Phone: 202-739-5059 E-mail: sburdick@morganlewis.com Counsel for FENOC DB1/ 79921045.1