Click for next page ( 176


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 175
APPENDIX B Design Criteria In Use for Dams Relative to Earthquake Hazards CONTENTS PART 1 FEDERAL AGENCIES Ad Hoc Interagency Committee on Dam Safety Bureau of Reclamation . Federal Energy Regulatory Commission. Forest Service Interagency Committee on Dam Safety Soil Conservation Service Tennessee Valley Authority U.S. Army Corps of Engineers (for Corps Projects) U.S. Army Corps of Engineers (for National Dam Inspection Program) U.S. Nuclear Regulatory Commission PART 2 STATE AGENCIES RESPONSIBLE FOR DAM SAFETY Alaska . Arizona Arkansas . California Colorado . Georgia 175 .177 .177 .180 .181 .181 .184 .185 .187 .188 .188 .190 .191 .191 .191 .191 .192

OCR for page 175
176 Hawaii Illinois. Indiana Kansas Louisiana Maine . Michigan . `, . . . MiSSlSSlppl Missouri Nebraska . New Jersey New Mexico . New York. North Carolina North Dakota Ohio Pennsylvania South Carolina. South Dakota I' 1exas Utah Virginia Washington . West Virginia Appendix B .193 .193 .193 .193 .193 .194 .194 .196 .196 .196 .197 .197 .197 .198 .198 .198 .198 .198 .199 .199 .199 .200 PART 3OTHER GOVERNMENTAL AGENCIES City of Los Angeles, California, Department of Water and Power . East Bay Municipal Utility District, California New York Power Authority Salt River Project, Arizona Santee Cooper (South Carolina Public Service Authority) . PART 4 PRIVATE FIRMS IN UNITED STATES .200 .201 .201 .202 .203 Acres American, Inc., Buffalo, New York . . . . . . . . . . .204 Alabama Power Company, Birmingham, Alabama . . . . . . . .207 Central Maine Power Company, Augusta, Maine . . . . . . . .207 Charles T. Main, Inc., Boston, Massachusetts . . . . . . . . . .207 Duke Power Company, Charlotte, North Carolina . . . . . . . .207 PI tannin ~ R~cr`~rnh (2ornoration Denver. Colorado . . . . . . . .208 .208 .210 0~ rid 7 ~ R.W. Beck and Associates, Seattle, Washington Yankee Atomic Electric Company, Framingham, Massachusetts

OCR for page 175
Appendix B PART 1FEDERAL AGENCIES Ad Hoc Interagency Committee on Dam Safety of the Federal Coordinating Council for Science, Engineering, and Technology (From "Federal Guidelines for Dam Safety," dated June 25, 1979) 177 The 1979 report of this group, a forerunner of the present ICODS, is clirected primarily at management of organizations engaged in planning, design, construction, operation, and management of clams. The report out- lines factors to be considered and procedures to be followed in investigations and design for earthquake hazards but does not specify design criteria to be followed. The report does indicate that the design seismic event is usually the maximum credible earthquake (MCE) and defines the MCE as "the hypo- thetical earthquake from a given source that could produce the severest vibratory ground motion at the dam." Bureau of Reclamation, U.S. Department of the Interior (From letter dated June 6, 1984) The following is extracted from a description of Bureau of RecIamation's practice relating to floods and earthquakes: Prior to the early 1970's seismic loading for Bureau dams was based on the application of 0. lg ground acceleration. In 1972, the Bureau initiated the use of the MCE (Maximum Credible Earthquake) and adopted it as the required seismic loading for Bureau dams. Under this loading, Bureau dams were required to maintain adequate stability without loss of the reservoir. The MCE associated with a specific seismic source was defined as: "the maximum earthquake that appears capable of occurring in the presently known tectonic framework. It is a rational and believable event that is in accord with all known geologic and seismologic facts. In determining the MCE, little regard is given to its probability of occurrence, except that its likelihood of occurring is great enough to be of concern (emphasis added) ." The Bureau currently floes not have a formally adopted criteria regarding the questions of: 1. What probability of occurrence should be used in considering remote event earthquake loadings? and; 2. Under what circumstances, if any, would a seismic loading less than the hypothetical MCE's be considered for the seismic safety evaluation of a structure and how would this relate to criteria for evaluation of existing dams as opposed to new dam designs?

OCR for page 175
178 Appendix B However, current thinking and practice with regard to these issues are described below: In orcler to preserve the character of the specification of the probability as an approximate estimate; to be more conservative than the value used for housing in California; and to remain in the same realm of conservatism as the U.S. Corps of Engineers criteria, a probability of occurrence of .00002 (recurrence interval of 50,000 years) is currently considered an inclusionary criteria for remote earthquake events for Bureau seismotectonic studies. This level of improbability serves as a guide to the geologist making seismo- tectonic studies rather than as a strict active fault criteria. The designation of hypothetical MCE events and their associated recur- rence interval or probability of occurrence is viewed by the Bureau as a separate function from the assignment of seismic loadings for design of the structure even though often they will be one and the same. Maximum Credi- ble Earthquakes are regarded as an actual geologic condition while (MDE) Maximum Design Earthquake (Ioading condition applied to the structure) is regarded as a parameter that may be specified according to the nature of the structure being considered and the consequences associated with potential seismically induced damages. Bureau policy since 1973 has been that the MCE events are considered in the seismic evaluation of dams and that a lesser event may be used for design of noncritical structures. In order to establish an appropriate loading level for noncritical structures on a project, as well as provide data for decision analysis studies, we are considering that seismotectonic evaluations provide a 500-year earthquake event which would be conceptually defined as the largest earthquake that is likely to occur during the life of the project and quantitatively defined as an earthquake with a recurrence interval of 500 years. The specific seismotectonic assessment made for Bureau dam sites in- cludes determination of: Hypothetical Maximum Credible Earthquakets) (MCEJ The maximum earthquake associated with relevant seismic sources is provided for each source that may produce significant earthquake shaking (greater than .OSg) at the dam site. The approximate recurrence interval of each MCE is pro- vided along with its focal depth and distance from the site. Earthquakes with a probability of occurrence up more than about .00002 are considered in establishing hypothetical MCE's. Historic seismicity For each source area from which an MCE is deter- mined, an earthquake magnitude and epicentral distance is provided that represents the earthquake from that source area with a return period of 500

OCR for page 175
Appendiix B 179 years if such an event is tectonically consistent with the source and able to be determined from geologic evidence, a projection of seismic evidence, or both. Otherwise such historic seismic information that is available is pro- vided. Surface faulting potential The potential for surface fault rupture is as- sessed at each site. Reservoir induced earthquake loading The current procedure for evalu- ation of reservoir induced seismicity in the Bureau is to consider the reservoir induced event equivalent to the local MCE event. Thus, if any active or potentially active faults are located within the reservoir regime, the reser- voir induced event as well as the local MCE event are definer! from the capacity of that fault system. The accelerogram record would, in general, be the same whether or not reservoir induced seismicity is considered. The recognition of reservoir induced seismicity only changes the probability of occurrence or frequency of occurrence of the large magnitude event but does not change the design of the structure since the structure must be capable of handling the earthquake loading regardless of when it would occur. The evaluation of the protection level is essential for formulating alterna- tives to solve the problem. This evaluation will result in one of three general cases from which to select loading conditions. Case AMaximum Loading Conditions This would be the case where the level and proximity of the downstream hazard make it clear at the outset of the problem that the consequences of dam failure in terms of potential loss of life or property damage would be unacceptable regardless of how remote the chance of failure may be. Thus, the loading conditions for the various alternatives are established at the maximum level (MCE, PMF, etc.) Case BLoading Conditions Determined byEconomicAnalysis This would be the case where the level and/or remoteness of the down- stream hazard are such that it is apparent (or becomes apparent) that incre- mental impact of dam failure would not significantly change the potential for loss of life or other nonmonetary factors, and that an economic analysis in which the costs and benefits of reducing the hazard become the primary consideration. Case C Loading Conditions as a Parameter in the Ultimate Decision Mak- ing Process This case is one where the incremental consequences of dam failure (with or without consideration of warning or other nonstructural modifications)

OCR for page 175
180 Appendix B do not clearly indicate that the dam falls under Case A or Case B. Compari- son of alternatives for this case wouicI include the economic comparison as for Case B. but would require a more comprehensive assessment of the incremental effects of dam failure on potential for loss of life (with and without warning system) as well as the incremental effects socially, environ- mentally, and politically for each alternative and load level. Response Requirementsfor Seismic Loading Under loading from the Maximum Design Earthquake, the structures of a project vital to the retention or release of the reservoir are required to func- tion (1) without permitting a sudden, uncontrolled release of the reservoir and (2) without compromising the ability to make a controllecI release of the reservoir. Under loading from the 500-year earthquake (or otherwise selected Eco- nomic Design Basis Earthquake), the project facilities not critical to the retention or release of the reservoir would be designed to sustain the earth- quake with repairable damage. The degree of damage which would be acceptable could be based on an economic analysis or on an estimate of the cost of the repair versus the initial cost to control the damage. Federal Energy Regulatory Commission (FERC), U.S. Department of Energy (From letter dated June 12, 1984) The following is extracted from material supplied by FERC: The criteria presented herein apply to both the review of designs by Com- mission staff prior to licensing and review of licensed project by independent consultants under Part 12 of the Commission's regulations. The review of project design for earthquake loading conditions utilizes two magnitudes of earthquakes; the maximum earthquake (ME) and the operating basis earthquake (OBE). Embankment structures should be capa- ble of retaining the reservoir during a ME; however, deformation is accept- able. Concrete structures should be capable of performing within the elastic range during an OBE, remain operational and not require extensive repair. During a ME a concrete structure should be capable of surviving without failure of a type that would result in loss of life or excessive property damage. The earthquake design criteria is based on the Corps of Engineers ER 1110- 2-1806.

OCR for page 175
Appendix B 181 Forest Service, U.S. Department of Agriculture (From letter dated May 23, 1984) Seismic design standards used by the Forest Service have been summa- rized as follows: The Agency requires an evaluation of earth movement potential and the establishing of appropriate design criteria on a case-by-case basis. The deter- mination of the need for detailed analyses, and subsequent design criteria, is based on factors such as the hazard presented, size of the dam and reservoir, potential ground motion at the site, site geology and the type of structure. Interagency Committee on Dam Safety (ICODS) An interagency task group established by ICODS has developed a draft paper entitled: "Proposed Federal Guidelines for Earthquake Analysis and Design of Dams." The parent body, ICODS, has not taken any action on the task force's draft. The draft discusses selection of design earthquake, ground motions from earthquakes, analyses of earthquake effects on dams, and evaluation of results of such analyses. The following material, including the flowchart shown in Figure B-l, is extracted] from the draft guidelines: The purpose of these guidelines is to develop some consistency in handling the earthquake analyses and design among the various Federal agencies involved in the planning, design, construction, operation, maintenance, and regulation of dams. They are intended to be used as general guides and are not to be considered as standards. It is recognized that the various agen- cies have differences in mission and diversified location which make agency independence desirable. It is further recognized that earthquake engineer- ing is in the developmental stage and flexibility is desirable. While the content of these guidelines generally reflects current practices, it will be necessary to make periodic revisions, additions, deletions, etc., to maintain currency with the state of the art in earthquake engineering. When the evaluation of the earthquake factor is completed, the maximum design earthquake (MDE) and the operating basis earthquake (OBE) are selected on the basis of an integrated evaluation of the earthquake factors. The MDE is the largest earthquake used in the seismic analysis of the dam and is generally equated to the controlling maximum credible earthquake (MCE) for the site. The OBE, usually smaller than the MDE, represents the maximum level of ground shaking that can be expected to occur at the site during the economic life of the dam. It may not be possible to show that all possible tectonic features have been discovered. Based on investigations,

OCR for page 175
182 DOCUMENTATION OF FOCI IIATlnN Appendix B FACTORS TO CONSIDER IN SELECTION OF DESIGN EARTHQUAKES REGIONAL TECTONIC SETTING SEISMIC HI STORY SEISMOTECTONIC STRUCTURES LOCAL OR SITE GEOLOGY SEISMIC ATTENUATION RESERVOIR INDUCED SEISMICITY _ 1 SE LECTION OF DESIG N EARTHQUAKES MAXIMUM CREDIBLE EARTHQUAKES MAXIMUM DESIGN EARTHQUAKE OPERATING BASIS EARTHQUAKE DETERMINATION OF GROUND MOTION FOR THE DESIGN EARTHQUAKES PEAK ACCELERATION VELOCITY, AND DISPLACEMENT DURATION ACCELERATION TIME-HISTORIES RESPONSE SPECTRA ~ l REQUIREMENT FOR EARTHQUAKE ANALYSI S SEISMICITY AND GROUND MOTIONS FOUNDATION CONDITIONS TYPE Of DAM CONSTRUCTION METHODS MATERIAL PROPERTIES PAST EXPERIENCE /~ YES -- RUINED --- rMETHODS OF ANALYSES | CONCRETE DAMS PSEUDOSTATIC METHOD DYNAMIC ANALYSES METHODS RESPONSE SPECTRUM TIME- HI STORY EMBANKMENT DAMS LIQUEFACTION EVALUATION PSEUDOSTATIC METHOD NEWMARK METHOD FINITE ELEMENT METHOD EVALUATION OF STRUCTURAL ADEQUACY EVALUATION OF ANALYSES RESULTS PAST EXPERIENCE Of DAMS CONSEQUENCES OF FAILURE FIGURE B-1 Flowchart depicting steps for earthquake analyses and design of dams. Source: ICODS draft of proposed guidelines.

OCR for page 175
Appendix B 183 gaps of information may exist. If so, conservatism may be desirable depen- dent upon the potential hazards associated with the dam. 1. Maximum Credible Earthquakes The first part of the investigation for selecting the MDE is to estimate the hypothetical MCE for each potential earthquake source, judged to have a significant influence on the site, from the information developed in section D. The hypothetical MCE for each seismotectonic structure or source area within the region examined is defined by magnitude and/or intensity, epi- central distance and focal depth. These MCEs are candidates for the control- ling MCE. 2. Controlling Maximum Credible Earthquake The second part of the investigation is to select the controlling MCE for the site as follows: a. Select the most conservative distance from each seismic source to the site. b. For each candidate MCE select strong motion records of earthquakes which have similar source and propagation path properties and were re- corded on a foundation similar to that of the structure or, if these site- matched records are not available, attenuate the epicentral ground motion parameters or MM intensity to the site using one or more applicable attenua- tion relationships. c. Select the controlling MCE based on the most severe ground motion parameters estimated for the site. There may be more than one controlling MCE because of the frequency characteristics on the dam and its compo- nents. 3. Maximum Design Earthquake The final selection of the MDE considers whether or not the dam must be capable of resisting the controlling MCE, which is a "worst case" situation. Usually, the MDE is equated with the controlling MCE. However, where the failure of the dam presents no hazard to life, a lesser earthquake for the MDE may be justified providing there are cost benefits and the risk of property damage is acceptable. 4. Operating Basis Earthquake The second level of design earthquake, the OBE, represents the maximum level of ground shaking that can be expected to occur at the site during the economic life of the project, usually 100 years for dams. It reflects the desired level of protection for the project from earthquake-induced structural and

OCR for page 175
184 Appendix B mechanical damage and loss of service during the project's economic life, or remaining economic life for existing dams. The OBE should be based on a probabilistic analysis which accounts for the time element involved in the definition of the OBE. A probabilistic analysis involves developing a magnitude-frequency or epicentral intensity- frequency (recurrence) relationship for each seismic source; projecting the recurrence information from regional information and past data into fore- casts concerning future occurrence; attenuating the severity parameter, usu- ally either peak ground acceleration or MM intensity, to the site; determining the controlling recurrence relationship for the site; and finally, selecting the design level of earthquake based upon an acceptable probabil- ity of exceedance and the project's exposure period selected for the design. Soil Conservation Service, U.S. Department of Agriculture (From letter dated May 21, 1984, criteria presented in Technical Release No. 60, revised august 1981) SCS has established three classes of dams as follows: Class (a) Dams located in rural or agricultural areas where failure may damage farm buildings, agricultural land, or township and country roads. Class (b) Dams located in predominantly rural or agricultural areas where failure may damage isolated homes, main highways or minor rail- roads or cause interruption of use or service of relatively important public utilities. Class (c) Dams located where failure may cause loss of life, serious damage to homes, industrial and commercial buildings, important public utilities, main highways, or railroads. Technical Release No. 60 contains the following requirement: Seismic AssessmentDams in zones 3 and 4, Alaska, Puerto Rico and the Virgin Islands and high hazard (class c) dams in zone 2 require special investigations to determine liquefaction potential of noncohesive strata, including very thick layers, and the presence at the site of any faults active in Holocene time. As part of this investigation, a map is to be prepared showing the location and intensity or magnitude of all intensity V or magnitude 4 or greater earthquakes of record, and any historically active faults, within a one-hundred kilometer radius of the site. (Obtain earthquake information for this map in print-out form from the Environmental Data Service, atten- tion D62, NOAA, Boulder, Colorado 80302. Telephone: FTS 323-6472; Commercial (303) 499-1000, Ext. 6472.) The report should also summarize other possible earthquake hazards such as ground compaction, landslides,

OCR for page 175
Appendix B 185 excessive shaking of unconsolidated soils, seiches, and in coastal areas, tsu- namis. T.R. No. 60 contains two Seismic Zone Maps, one for the contiguous United States and one for Hawaii, that are labeled as being adapted from TM 5-809-10, April 1973. On each of these maps is shown the following table of minimum seismic coefficients: Zone o 1 2 3&4 Coefficients 0.00 0.05 0.10 Base on Seismic Assessment Elsewhere in T.R. No. 60 the seismic coefficient is defined as "the fraction of a weight to be used as a horizontal force in a quasistatic analysis." Also, the minimum factors of safety for embankment slope stability with seismic forces are given as: Structure Classification (a) (b) (c) Minimum Factor of Safety 1.0 1.1 1.1 Tennessee Valley Authority (From letter dated June 7, 1984) The General Manager of TVA has pointed out that only a few of the older TVA dams were analyzed for earthquake loadings when they were designed but state-of-the-art methods have been used in the design of newer dams and in analysis of the older dams. The information supplied indicates these methods include dynamic analyses, finite element modelling and evalua- tions of liquefaction potentials. The following is extracted from a technical paper outlining TVA practices prepared by members of the TVA staff: Earthquake evaluation of dams is, and will continue to be, in the develop- mental stage as more becomes known about earthquakes and their effects on dams. Therefore, flexibility and sound engineering judgment are essential in the evaluation to reflect the specific conditions of each dam. The TVA region is located in an area of low to moderate earthquake activity. Therefore, the risk of large earthquakes occurring and affecting

OCR for page 175
200 Appendix B West Virginia (From letter dated May 29, 1984) The "Dam Control Regulations" issued by the West Virginia Department of Natural Resources specify geotechnical investigations of dam sites, labo- ratory investigations of foundation and embankment materials and geotech- nical evaluations. Two requirements are directed at earthquake loadings: (1) a requirement that embankments have a safety factor of 1.0 under seismic loading and (2) a requirement for consideration of potential for liquefaction. PART 3 OTHER GOVERNMENTAL AGENCIES City of Los Angeles, California, Department of Water and Power (From letter dated July 3, 1984) The following is adapted from a list of procedures, criteria and standards relating to earthquake hazards furnished by the Department of Water and Power: 1. Analyze dams using maximum credible earthquakes for local and re- gional events. 2. Goal is satisfactory performance of dams when subjected to the local and regional maximum credible earthquakes. 3. Use two dimensional finite element method and Seed-Idriss approach to analyze the stability of dams. 4. Require removal of alluvial deposits and construction of dam on com- petent bedrock. 5. Require embankment fill compacted to a minimum relative compac- tion of 95 % based on DWP's Water System Standard (modification of ASTM D 1557-78, from five to three layers, resulting in 33,750 foot-pound per cubic foot). 6. Provide sufficient crest width (25-30 feet). 7. Provide adequate freeboard (a minimum of seven feet, usually ten feet). 8. Provide relatively flat slopes (3 to 1 upstream and 2-1/2 to 1 down- stream or flatter) . 9. Provide system of internal drains based on Terzaghi's filter design crite- ria and cutoffs to control seepage through the embankment and foundation to enhance the dynamic stability. 10. Provide zones in internal drainage system to serve as crack stopper to prevent rapid loss of reservoir water resulting from potential cracks in the dam from seismic shaking.

OCR for page 175
Appendix B 201 11. Require grouting of foundation and abutments to seal joint seams, fissures, and voids. 12. Encase all pipeline under the dam and construct seepage cutoffs along the pipes. 13. Provide sufficient blow off capability for smaller reservoirs (below 5,000 acre-feet), capability to blow off half of the water in seven days). 14. Provide surveillance and instrumentation to monitor dam perfor- mance. These include surveillance by the Reservoir Surveillance Group, reservoir caretakers and patrolmen. Monitoring facilities and instrumenta- tion include seepage drains, observation wells, tiltmeters, strain gages, movement and settlement points, pore pressure piezometers, seismoscope, strong motion accelerometers, and peak recording accelerometers. 15. Meet the requirements of the California Department of Water Re- source, Division of Dams. East Bay Municipal Utility District (EBMUD), California (From letter dated May 23, 1984) The following is quoted from a letter of the General Manager of EBMUD: The East Bay Municipal Utility District does not have formalized or writ- ten criteria and standards applicable to dams. We have attempted to apply current "state-of-the-art" criteria, standards, and procedures in both the design of new facilities and the analysis of existing facilities. These are ap- plied to individual dams in specific site situations and are similar, but not identical, for all dams. The District has, in most cases, utilized Professor H. Bolton Seed's meth- ods for the application of dynamic loads and the analysis of dams under seismic conditions. The determination of the seismic conditions to be ex- pected is a very vital part of such studies because new data are increasingly available. New York Power Authority (From letter dated July 24, 1984) The following is quoted from information furnished by the New York Power Authority (NYPA): The Authority is currently proceeding with a seismic reassessment of all its concrete and earth dams. In order to maintain a unified approach to the seismic stability analysis of NYPA earth embankments in the present chang- ing regulatory climate, the following procedure has been adopted:

OCR for page 175
202 Appendix B a) Determine earthquake induced ground motions from an appropriate design earthquake; b) Review and evaluate existing information and geotechnical data to determine appropriate soil/rock properties and seepage flow; c) Perform simplified analysis of typical dam cross sections for the predic- tion of dam deformation. The NYPA suggested program of studies follows a progression from a relatively simple approach using existing data to a decision point where dam performance will be determined to be either acceptable, mariginal, or unac- ceptable. More sophisticated sampling and analysis techniques could be carried out if it is required or desirable. Alternate methods currently in use attempt to determine whether or not a particular soil is susceptible to liquefaction. This includes evaluation of in- situ undrained steady state shear strength by means of laboratory cyclic biaxial tests on undisturbed specimens. A disadvantage to this approach is that it relies upon interpretation of variables involved in the recovery, han- ~ing and testing of undisturbed soil samples, and is rather expensive. The approach does not provide direct information on the likelihood that lique- faction will be triggered by a given design earthquake. Depending upon the results of the simplified evaluation program, a sec- ond phase may be required. This phase would consist of a more rigorous analytical procedure because: al certain parametric output, such as material properties of the soil, or the predicted. design earthquake ground motions, proved to be critical; and b) the simplified one and two dimensional analyses did not sufficiently define the problem for certain of these critical input values. The parametric input itself may need further refinement either by expert reevaluation of measured values or by limited field investigations. The results of these analyses, together with the associated uncertainties in both the seismic input and the predicted dynamic response of the earth embankment will stand by themselves as a deterministic evaluation (at some given safety factor) or they can be utilized to evaluate the overall seismic safety of the embankment in probabilistic terms and at various risk levels. Salt River Project, Arizona (From letter dated June 25, 1984) The General Manager of the Salt River Project (SRP) furnished material on a seismic study conducted by SRP on Roosevelt Dam, one of six Bureau of Reclamation Projects operated and maintained by SRP. This study was initi- ated to develop seismic design criteria (response spectrum) for the Maximum Credible Earthquake at Roosevelt Dam on the Salt River. The report and

OCR for page 175
Appendix B 203 subsequent addenda emphasize the need for consistent terminology so that the geotechnical consultant, seismologist, analyst and design engineer un- derstand the intent and limitations of design criteria. Paper supplied with the report discussed the problem of pyramiding safety failures in design. Other discussion dwelt on the probabilities of various magnitudes of earth- quakes during the life of the structure. Santee Cooper (South Carolina Public Service Authority) (From letter dated july 12, 1984) The following is quoted from letter of the President, Santee Cooper: For the past several years Santee Cooper has been conducting state-of-the- art research on our Pinopolis West Dam in an attempt to determine its reaction to a recurrence of the 1886 Charleston Earthquake. Because of this, we are very interested in the establishment of acceptable risk levels for seismicity as well as for spillway adequacy in Federal dams. As for our design procedures, criteria and standards for dam safety and inspections, Santee Cooper is licensed by the Federal Energy Regulatory Commission (FERC) and follows FERC guidelines regarding dam safety. PART 4PRIVATE FIRMS IN UNITED STATES Acres American, Inc., Buffalo, New York (From letter dated July 7, 1984) The following extracts from a paper furnished by Acres American illus- trate the organization's approach to design for earthquake hazards: Selection of Design Earthquake. The approach to the selection of design earthquake is based on the potential risks involved with the failure of the dam. For smaller dams where the associated risks are small, a design earth- quake is selected on the basis of the seismic zone in which the project is located, unless there is clear evidence of a source of seismicity near the project which will affect the project. For relatively large dams, where the risks associated with the failure could be significant (these risks could be economic, financial, loss of life and property, or environmental), design earthquake is selected on the basis of comprehensive seismic and geologic studies. In these cases, the Maximum Credible Earthquake (MCE) is deter- mined using both the probabilistic and the deterministic models. The reason for using both models is apparent lack of complete confidence in any one mode} at present and the need to assure the safety of the dam. The reason for

OCR for page 175
204 Appendix B the lack of confidence lies in the fact that it is not always possible to uncover and study all the earthquake related faults in the project area with a high degree of confidence and the historical seismic records are less than complete for a realistic probability model. Once the MCE events from both models are determined, we use the more severe of the two earthquakes in the design since the increase in cost to provide required safety measures is generally small. Selection of Design Motion. Once the maximum credible earthquake, including its source and mechanism, is defined, the key parameters such as maximum free field acceleration, predominant period and duration are determined using currently accepted correlations. The earthquake motion is attenuated to the dam site using appropriate attenuation relationships. In our experience, regional attenuation curves are only available in selected regions ant] the improvement in results using these curves is still questionable primarily due to the lack of sufficient data for moderate to large earth- quakes. For this attenuated earthquake, a response spectrum is determined and appropriate acceleration time history of the earthquake is generated. The time history is generated by either scaling an actual recorded earth- quake motion under similar geologic conditions or artificially by using mathematical formulation. In case of actual recorded time history, some- times an ensemble of time histories is used to represent a wide range of frequency and randomness characteristics. Analysis of Dam. The analysis has two distinct elements; the engineer- ing properties of the material and the mathematical calculations. The engi- neering properties are selected from the published literature and previous project during the preliminary stage of the project and from the field and laboratory tests for the final design. However, difficulties exist in character- izing large size material properties (rockfill, boulders) and Acres, like other organizations, uses published relationships supplemented with parametric sensitivity analysis to study the impact of error in material properties. The mathematical approach is selected on a project by project basis after considering factors such as dam type and height, reservoir size, type of material in the dam, regulatory requirements and any specific requirements by owner of the facility. Generally, the following approach is used for the earthfill/rockfill dams: (a) All dams are analyzed using a pseudostatic approach of analysis to study the potential of mass failure. A horizontal seismic coefficient equal to 2/3 of the maximum peak acceleration is used in the analysis. For dams with significant height, the maximum peak acceleration at various heights within

OCR for page 175
Appendix B 205 the dam is calculated from response and analysis (either simplified such as Newmark method or more detailed using one-dimensional or two-dimen- sional mode! and design time history). A minimum factor of safety of 1.0 or larger is considered acceptable. The soil properties used are those represent- ing the undrained behavior under consolidated stress conditions. (b) If the dam is constructed] of materials which are considered not to experience more than 10 to 15 percent loss of strength due to shaking, a more detailed dynamic analysis is not usually performed. The results may be supplemented by calculating permanent deformations using simple meth- ods such as Newmark analysis and a post-earthquake stability analysis using effective stress parameters and excess pore pressures generated during earth- quake. The Newm ark type analysis is performed to assist in the development of adequate freeboard. (c) For materials which may experience significant strength loss due to shaking and for dams, say, higher than 300-400 ft. a more comprehensive dynamic analysis is performed using state-of-the-art techniques such as de- veloped by Seed et al. Sometimes these analyses may be performed even for smaller structures when required by the owner or the regulatory agency. The techniques are well documented in the engineering literature and we main- tain a library of software and an analysis group to perform these studies. These analyses may include determination of overall factor of safety based on stress levels, anticipated deformation along potential failure planes or identification of zones that may have potential for tensile stresses. As stated earlier, perhaps the single most important element of difficulty is the deter- mination of the in situ behavior of the coarser soils and rockfill materials, and for this reason alone, we consider these analyses to be good guides and not the specific tools to prepare a safe design. For the concrete dams, the approach is much more defined. Again, the judgment for the type of analysis is made on a project by project basis. In many cases, a pseudostatic analysis is performed to determine the overall stability and to calculate the stresses. For arch dams and gravity-arch dams, this analysis is supplemented with dynamic response analysis. As a first step, a response spectrum analysis is used to assess the stresses in the structure. If the stresses are within the allowable limits, no further analysis may be per- formed since the response spectrum analysis yields conservative results. However, if the resulting stresses exceed the allowable stresses, then a time history response analysis is performed using finite-element modelling devel- oped by the USER which analyzes linear deformation. This analysis is fur- ther refined by performing a single cantilever time history analysis incorporating nonlinear springs in beam elements to more accurately esti- mate tensile stresses in the cantilever.

OCR for page 175
206 Appendix B The results of these analyses are critically reviewed for appropriateness of input data and reasonableness of results. Design of Structure. The safety of the dam against damage and/or fail- ure during an earthquake is assured through prudent design. For the earthfill/rockfill dams, the safety of the dam is assured by using appropriate construction materials and procedures and by performing nec- essary foundation preparations. The use of materials that will dilate during seismic events is encouraged and the upstream zoning is provided with high permeability materials. Any liquefiable soil in the foundation is excavated to found dam on competent material. Liberal use of cohesionless filters and appropriate drainage provisions are made within the body of the dam. The use of liquefiable or otherwise degrading materials during earthquake is avoided in critical zones. Liberal freeboard is provided to accommodate slumping during earthquake and to provide protection against seiches caused by seismic event. The internal zoning and shaping of abutment is done carefully to minimize tensile stresses and a careful quality control is exercised during construction. In Acres, we believe that it is essential to make provisions for corrective measures should the dam experience damage due to earthquake. Thus, we give a serious consideration to facilities that could be used to evacuate the reservoir (such as low level outlet) and need for galleries to improve, monitor and rehabilitate drainage and grouting. Similarly, the safety of concrete dams is enhanced by including specific construction details, utilizing higher strength concrete and providing ade- quate foundation preparation to improve compatibility between the foun- dation and the superstructure. Further, concrete structures are not normally preferred on sites where a known or potentially active fault is discovered in the foundation. Conclusions. The rapid advances in the field of seismicity, computa- tional procedures and material testing and study of failure case histories have greatly increased the understanding of behavior of dams during strong earthquakes. However, many of the parameters are still derived from se- miempirical relations and the historical records are less than complete. Therefore, the design of a safe dam for earthquake as well as other loads still requires a great deal of engineering judgment and experience, knowledge in material behavior and construction techniques and a sound engineering approach. Any step in formulating a universally acceptable approach would help bring consistency in the profession and make it easy to relate the experi- ence on other similar projects.

OCR for page 175
Appendix B 207 Alabama Power Company, Birmingham, Alabama (From letter dated July 18, 1984) The following is quoted from letter of Alabama Power Company, which reports it has eleven hydroelectric projects on two river systems within Ala- bama: In the area of seismic analysis, our approach is simplified. All of our projects are located in a region of very low seismicity, and we have tradition- ally used a pseudo static load factor of O.lg for purposes of analysis. This method of analysis has been supported by our independent consultants and approved by the FERC. Central Maine Power Company, Augusta, Maine (From letter July 31, 1984) Structural analysis sheets prepared by Charles T. Main, Inc. for five hy- dropower projects were furnisher! by the Central Maine Power Company. The analysis sheets indicate that the analyses used the pseudostatic method for accounting for earthquake forces with horizontal forces ranging from 0.05g to 0. lOg and, in one instance, a vertical force of 0.03g. Charles T. Main, Inc., Boston, Massachusetts (From telephoned response by Thomas Neff, June 15, 1984) The Charles T. Main organization has used the seismic design criteria of the U.S. Army Corps of Engineers in its work in other countries and has encouraged the use of those criteria in countries that had no or less conserva- . . ~ . . ~ tine seismic Resign criteria. Duke Power Company, Charlotte, North Carolina (From letter dated duly 19, 1984) Information supplied by Duke Power Company indicates its standards for dams comprise regulations of the Federal Energy Regulatory Commission supplemented by standards and criteria issued by a number of federal and state agencies.

OCR for page 175
208 Appendix B Planning Research Corporation (PRC), Denver, Colorado (From letter dated June 19, 1984) The following is extracted from a description of the design criteria for seismic standards used by PRC: Whether working in the United States or overseas, we normally define the seismic design parameters by analyzing each specific site. For U.S. projects, we use published design criteria such as the Aigermissen Map or building code recommendations as a check, but not as the only method of determining the seismic design parameters. Our designs are normally based on a site specific Design Basis Earthquake (DBE) which is defined as the largest earthquake which would be expected to occur once during the expected life of the project. We design a project to withstand the DBE with minimal or no possibility of damage occurring and then check to ensure that catastrophic failure will not take place when/if a Maximum Credible Earthquake occurs. With regard to reservoir induced seismicity (RIS), we normally assess the potential for this phenomenon to occur based on site characteristics, reser- voir size, and reservoir depth. A comparison of the particular site under study with projects that have experienced RIS in the past is made and if it is judged that RIS could occur at the site in question, appropriate instruments are incorporated in the design and installed prior to project construction. This monitors the pre and post reservoir filling conditions to determine if, in fact, RIS is experienced and the magnitude of the reservoir induced earth- quakes. If the potential for RIS is judged to be low or nonexistent, we usually recommend that a much simpler instrumentation network be installed as a normal part of the overall project instrumentation package in order to moni- tor the performance of the structure under seismic loading. R.W. Beck and Associates, Seattle, Washington (From letter dated June 12, 1984) The following are extracts from statements of "Design Criteria" and "Methods of Analysis" for an arch dam at Swan Lake Hydroelectric Project, Alaska, supplied by R.W. Beck and Associates to illustrate the firm's prac- tice: Design Earthquake DE) The DE is defined as the largest earthquake that would be expected to occur during the economic life of the dam (recurrence interval of once in 100 years) . ("Largest earthquake" implies the earthquake producing the greatest

OCR for page 175
Appendix B 209 loading on the structure.) The magnitude of this event is determined from magnitude versus frequency of occurrence relationships. The dam is re- quired to safely withstand the loads due to the DE although some repairable damage is acceptable for this loading, if it occurs. Those systems and compo- nents important to safety must remain operable. Maximum Credible Earthquake (MCEJ The MCE is defined as the earthquake that would cause the most severe vibratory ground motion capable of being produced at the site under the currently known tectonic framework. It should be a rational and believable event which can be supported by all known geologic and seismologic data. It is a judgment based on the maximum earthquake that a tectonic region could produce, considering the geologic evidence of past movement and the re- corded seismic history of the area. The dam was analyzed to ensure that it can withstand the loads from the MCE without any sudden or uncontrolled release of the reservoir even though damage may occur. Dynamic Analysis a. General Because the project site is in a seismically active region, a detailed dy- namic analysis of the dam was essential. Two different approaches for calcu- lating stresses due to earthquake loading were used, the response spectrum method and the time-history !,.ethod. The former gives more approximate and generally conservative answers but is convenient and economical. The time-history method is more complex but has the advantage of showing stresses varying with time. The maximum stress values resulting from each method, however, correspond reasonably well. A stress analysis of an arch dam acting elastically under earthquake load- ing is performed as follows: (1) Define the mass and stiffness characteristics of the dam taking into account the effective hydrodynamic mass considered to move with the struc- ture as it vibrates; (2) Determine natural frequencies and mode shapes; (3) Calculate for each mode considered the inertial forces and stresses due to a unit spectral acceleration at each point of the structure; (4) Determine stresses contributed by each mode based on an actual input ground motion; and (5) Determine the total stresses. b. Response Spectrum Method A response spectrum is a plot of the relationship between the maximum response of a series of single degree-of-freedom systems, each having a differ-

OCR for page 175
210 Appendix B ent period of vibration. The response can be plotted as acceleration, velocity, or displacement. Total stresses are determined by taking the root-mean- square value of the sum of the stresses contributed by each significant mode of vibration. c. Time-H'story Method The time-history method differs from the response spectrum in methodol- ogy and in that stresses are computed for discrete times during the earth- quake. The stress contributed by each significant mode at a given instant of time is determined and total stress at each mode at a given instant of time is deter- mined by mode superposition. Thus, it is possible to obtain stress distribu- tions throughout the dam at any given instant of time. Yankee Atomic Electric Company, Framingham, Massachusetts A company representative has made available a report, dated December 1982, titled "Supplemental Seismic Probability Study Yankee Atomic Electric Company, Rome, Massachusetts." The report documents extensive studies of the probability of earthquakes affecting the site of a nuclear power station, specifically to develop estimates of earthquake shock effects having annual probabilities of 10-3 and 1O-4. The report does not discuss design criteria for earthquake hazards as such, but shows the problems in attempt- ing to analyze earthquake potentials in the eastern United States.