|
サンロケ多目的ダムプロジェクト
環境影響評価書(EIA)の分析：建設／地質／地震  by Tiziano Grifoni 5 August 1999 Major findings b. Reduce the shear strength of soils, leading to potential for soil liquefaction
and landslides. order to provide storage for floods, large volumes of water must be released from the reservoir well in advance of the forecast flood. It does not appear that there has been adequate consideration of the consequences of large water volume releases for the 1.5 million people living downstream of the dam. There does not appear to be an adequate flood forecasting system, flood warning system, spillway gate management plan or evacuation and community preparedness plan. Introduction The Updated EIA concludes that the objectives of the SRMP, i.e. generate about 345 MW of electricity, provide irrigation waters for 70,500 ha of land, improve water quality, and control floods can be achieved successfully. However, it is difficult to understand how this conclusion can be arrived at. There are too many contrasting factors described in the documents. The dam purpose of increasing irrigation water and power generation is in net contrast with the purpose of flood control. Large quantities of water and sediment will need to be frequently released to minimize sedimentation, replicate the natural flooding regime, minimize saltwater intrusion and salinization of floodplain lands, minimize the effect of groundwater supplies and maintain minimum flows for continuation of fisheries. Release of water will entail release of contaminated sediment from mining which is likely to continue without interruption. In addition, the negative impacts (socioeconomic, ecological, environmenta l) associated with the decommissioning of the dam have not been studied and have not been included in the documents. GEOLOGY/SEISMICITY Background According to the 1984 EIA, Section S.2.2.3 ≪ the probability of disturbing the stability of the geology at the site is not remote. The presence of faults will increase the risk of reservoir collapse because of the risk of reservoir induced earthquake ≫. The Luzon earthquake of July 16, 1990 has been attributed to the Philippine fault zone and its primary branch, the Digdig Fault. The SRMP is located approximately 26 km west of the Digdig Fault. The epicenter of the earthquake was reported to be northeast of Cabanatuan, and the main shock was of magnitude 7.8 (Ms). The geotechnical aspects of the 1990 earthquake were significant (Earthquake Spectra, 1991). They related to liquefaction, landslides, and ground motion amplification. Landslides and rock falls were numerous, predominantly in the northwestern mountainous area near Baguio. The Ambuklao Dam, located northeast of Baguio, reportedly suffered damage. The rains following the earthquake caused massive mudflows and further destabilized unstable slopes. Surface faulting was also observed, and most surface displacements exceeded 3 meters. The largest displacements were found on the Digdig Fault. They measured approximately 6 meters. Locally, vertical offsets on the fault exceeded 2 meters. In light of the 1990 Luzon Earthquake, the ground motion and dynamic response of the SRMP to seismic events is of great importance. The geology/seismicity review carried out as part of the SRMP risk assessment was conducted almost 4 years after the earthquake. The report states that ≪ the timing of this review is far from ideal......any geomorphic evidence indicative of very recent downslope movements related to the July 16, 1990 earthquake could have already been obliterated by now ≫. In addition, the core samples collected during the geology investigation conducted in 1977 and 1982, were not available for observation; ≪ only a few scattered cores remain and would be of no practical value for the current review ≫. Implications The ground acceleration of 0.35g recommended for the design of the SRMP appears to have been taken directly from a seismic study performed for the Sual coal-fired power plant in Pangasinan. From the review of the existing documents, there is no indication that a site specific study has been conducted to determine ground motions for the SRMP. For a sensitive project such as this, the ground motions for design should be derived from a seismic hazard analysis involving the quantitative estimation of ground motion characteristic at a particular site requiring the identification and characterization of all potential sources of seismic activity. A seismic hazard analysis can be conducted deterministically or probabilistically (Kramer, 1996). Deterministic seismic hazard analysis (DSHA) involves the assumption of some scenario, the occurrence of an earthquake of a particular size at a particular location, for which ground motion characteristics are determined. Probabilistic seismic hazard analysis (PSHA) allows uncertainties in the size, location, rate of recurrence, and effects of earthquakes to be explicitly considered and quantified in the evaluation of seismic hazards. Often, ground motion computed from DSHA are larger, or more conservative, than those computed from PSHA because the DSHA assumes that the earthquake of the largest possible magnitude occurs at the shortest possible distance to the site within each source zone. However, in highly active seismic areas, it is possible that the largest ground motion is obtained from the PSHA. For structures for which failures could have catastrophic consequences, such as large dams and power plants, both studies are carried out and evaluated. In order to conduct DSHA and PSHA, earthquake sources must be identified on the basis of site specific geology, tectonic, historical, and instrumental evidence. Geologic evidence includes the study of geologic record of past earthquake activity (paleoseismology), and the detection of offsets or relative displacements of various strata (Wallace, 1981). The geologic records may be very complex or may be hidden by thick layers of recent sediments. The identification of seismic sources from geologic evidence is a vital part of a seismic hazard analysis. Earthquake sources that may impact the dam site may include: ? Earthquakes generated from displacement of the Philippine Fault, the Digdig
Fault, and all other known faults located a few tens of kilometers from the
dam site. Based on the above discussion, it is recommended that the most severe level of earthquake, the maximum credible earthquake (MCE), be considered as the design earthquake for the SRMP. A site specific seismic hazard analysis could well yield ground motions ranging from 0.2g to 0.8g. From the review it is not clear whether the value of 0.35g represents the MCE. Consequences of MCE The SRMP should be designed to withstand the MCE with only minor damage and no possibility of dam break. The following parameters should have been considered in the design: ? Seiche hazard. Earthquake-induced waves in enclosed bodies of water can be developed by long-period seismic waves that match the natural period of oscillation of the reservoir. A 7.8 earthquake similar to that occurred in 1990 could last up to 90 seconds. Seiche waves could reach several meters in height, and could increase erosion of banks and upstream slope of the dam. ? Weakening instability. Through a process of pore pressure generation, earthquake-induced stresses and strains can reduce the shear strength of soils. Weakening instability can occur when the strength drops below the induced stresses in slopes. Weakening instability is usually associated with liquefaction. Liquefaction can lead to landslides or lateral spreading. ? Landslides. Strong earthquakes can cause large landslides, and large landslide, if occurring into the reservoir, could produce waves which must be considered in the design of the dam freeboard. Consequences of the 1990 Earthquake Reportedly, the site geology is characterized by the presence of intensely fractured diorite, sandstone, and shales. Faults and structural discontinuities are also present. The diorite is the most common rock in the project area and underlies most of the dam foundation. The strong ground shaking developed by the 1990 earthquake could have dilated the rock fractures and weakened the rock formation at the dam site. Weakened rock conditions could lead in the future to massive landslides. Large landslides occurring upstream of the dam into the reservoir could generate waves capable of overtopping the dam. Apparently, the present freeboard is 7 meters. This value should be checked against the maximum waive height as a result of landslides occurring into the reservoir. FLOOD CONTROL One purpose of the SRMP is flood control. The dam will have a live storage of 530 mcm. The documents state that large floods can be stored in the new reservoir. However, because the reservoir will always be filled at maximum capacity for power and irrigation purposes, it is likely that not enough storage will be available to store large quantities of water resulting from major storm and flood events. In order to provide storage for floods, large volumes of water must be released from the reservoir well in advance of the forecast flood. Are the consequences of large water volume release understood? Will there be enough time to release water from the reservoir prior to major storm/floods? Has a flood forecasting system, flood warning system, and spillway gate management been adequately prepared? What about the evacuation and community preparedness plan? How reliable will the flood forecasting be? The answers to these questions are of paramount importance for the safety of the 1.5 m illion people living downstream from the dam. These questions have not been adequately covered in the existing documents. The 1984 and 1994 EIAs stress the importance of proper reservoir management with respect to water releases. Reservoir management, including provisions for gate operations, is an important factor to avoid dam overtopping during extreme storm events. Dam overtopping can lead to dam failure with catastrophic consequences. This type of failure can be avoided by conservative spillway design, attention to the possibility of large landslides into the reservoir, and generous freeboard. Due to the importance of the project and the consequences of failure, the spillway design should be based on the Probable Maximum Flood (PMF). The PMF can be estimated with a rainfall-runoff model calibrated with site specific data. Presently the gated spillway has been designed to release a PMF of about 12,800 m3/s, which has been linked to a 1:10,000 probability event. The 1984 spillway was designed to release a PMF of 15,300 m3/s, and a study performed by ELC in 1979 estimated a PMF of 11,200 m3/s (ERA 1997). These estimates are at variance. The ERA 1997 pointed out that the hydrologic data recorded at San Roque were of poor quality and therefore not suitable for design. It is not clear which data and which model was used to design the present spillway. Once the spillway has been designed appropriately, it will be possible to develop a proper management regime, which includes gate operating rules. Recommendations 1. Search for geologic evidence of earthquake sources by identifying faults
through review of published literature, interpretation of air photos and remote
sensing, field reconnaissance including logging of trenches, test pits and borings,
and geophysical techniques. Stephen L. Kramer, 1996, “Geotechnical Earthquake Engineering”, Prentice Hall International Series in Civil Engineering and Engineering Mechanics, pp. 106-139. Wallace, R.E., 1981, “Active Faults, paleoseismology, and earthquake hazards in the Western United States”, in D.W. Simpson and T.G. Richards, eds. Earthquake Prediction: An International Review, Maurice Ewing Series 4, American Geophisical Union, Washington D.C., pp. 209-216. Earthquake Spectra, October 1991, The Professional Journal of the Earthquake Engineering Research Institute, Philippines Earthquake Reconnaissance Report.
|