Search

Naturally occurring microtremor observations measured the natural frequencies of the fluid-elastic-coupled shell plate vibration (bulging) in a large flat-bottomed cylindrical steel tank with a capacity of 125,000-m3. Five peaks appear in the observed microtremor spectral ratios of the top or midheight of the shell plate to the bottom on the tank foundation. Comparing the spectral ratios to the solutions obtained by FEM eigenvalue analysis assuming a fixed base suggests that the five peaks are the bulging modes of (m, n)=(1, 1–5), where m and n denote the vertical order and the circumferential wavenumber, respectively. The measured non-soil-coupled natural frequencies from the spectral ratio agree fairly well with those obtained from FEM analysis. The measured natural frequencies of the fundamental mode (m = n = 1) also agree well with those projected by a simplified equation developed under the assumption of a fixed base, which is adopted in the seismic codes of the Japanese Fire Service Act. This equation should provide a reliable soil-coupled natural frequency of the fundamental mode for a tank situated on firm ground in which the storage-soil-coupled effects are presumed weak. Additionally, a simple method is presented to determine the non-soil-coupled natural frequency of the fundamental mode from the observed microtremor spectral ratios without referencing the FEM eigenvalue solutions. This simple method works very well for the tank examined.

Damage and influences to oil tanks caused by severe strong ground motion due to a large earthquake (Mw6.6) that occurred in the district of Iburi-tobu, Hokkaido, Japan on September 6, 2018 are reported in this paper. In the vicinity of the seismic source region, two large-scale crude-oil storage bases are located. The neighboring two bases had in total 86 large oil storage tanks with a capacity of 115,000 m3. The oil storage bases were hit by strong ground motion with peak ground accelerations of 590 to 1,570 cm/s2 and with peak ground velocities of 50 to 80 cm/s. Shell plates of a small bunker A tank with a capacity of 306 m3 suffered diamond buckling and elephant-foot buckling. No large oil storage tanks lost their function of oil storage despite of the severe strong ground motion. However, most of them splashed oil from the gap between the floating roof and the shell plate, and many of them had damage to their pontoons, gauge poles, guide poles, rolling ladders, liquid-level meters, and shoulders of foundation. One of the 115,000-m3-in-capacity tanks was equipped with a displacement gauge system to measure uplift of the bottom of the shell plate from the shoulder of tank foundation. The system recorded a maximum uplift of 4.4 cm. This is the world’s first record of uplift of a large tank caused by a natural earthquake.

The reconnaissance just after the 2011 earthquake off the Pacific coast of Tohoku reported the pulling out of anchor bolts of an upright cylindrical tank. Regarding the tank as an unanchored tank and employing the accelerogram recorded at 2 km off the tank, uplift of the unanchored tank during the 2011 earthquake off the Pacific coast of Tohoku was forensically examined. The time history of the uplift displacement of the tank computed by the explicit finite element analysis reveals that the unanchored tank uplifts during the earthquake and its uplift displacement is more than the length of the pulling out of the anchor bolts found by the reconnaissance. This implicitly corroborates applicability of the explicit finite element analysis to analyzing the tank rocking behavior. In addition to that, the uplift displacement of the unanchored tank was estimated by some calculation methods available to date and compared with that computed by the explicit finite element analysis. Comparison reveals that the calculation methods used herein may likely overestimate the uplift displacement of the unanchored tank and need to improve their calculation accuracy.

We measured the natural frequencies of the fluid-elastic-coupled shell plate vibration excited in a large-sized, flat-bottomed, cylindrical steel tank by observing microtremors at three points of the tank. By comparing the appearance frequencies and the values of the five peaks seen in the observed microtremor spectral ratios of the top or mid-height of the shell plate to the bottom on the tank foundation with the solutions obtained from a fixed-base FEM eigenvalue analysis, we identified the five peaks as belonging to modes (m, n) = (1, 1–5), with m specifying the vertical order and n the circumferential wave number. The measured non-soil-coupled natural frequencies from the spectral ratio agreed fairly well with those obtained from theory using the FEM analysis. The measured frequencies of the fundamental mode (m = n = 1) were also in good agreements with those estimates using a simplified equation assuming a fixed base adopted in seismic codes of the Japanese Fire Service Act. This equation is expected to provide a reliable soil-coupled fundamental-mode natural frequency for a tank resting on firm ground; the storage-soil-coupled effects are presumed to be weak. Without using an FEM analysis, we present a simple approach to determine the non-soil-coupled fundamental-mode natural frequency solely from the observed microtremor spectral ratios. This simplified procedure works very well for the tank examined.

Evaluation of the rocking motion of unanchored cylindrical tanks subjected to the ground acceleration is a main topic in the framework of the seismic vulnerability assessment of this particular kind of structures. However, despite the anchored or unanchored tanks, any calculation that attempts fully evaluate tank uplift has not been presented, while many of the analyses of the tank rocking motion and the deformation of the tank bottom plate have been respectively presented. This paper carries out a thought experiment of applicability of these analyses to evaluating uplift of the unanchored tanks. First, this paper focuses on an anchored tank at a chemical plant whose anchor bolts were pulled-out during the 2011 off the pacific coast of Tohoku Earthquake. Second, combine a few analyses of uplift of the unanchored tanks being available to date; then, calculate the uplift displacement of an anchored tank of interest as if it were an unanchored tank. Next, infer the pulling-out length of anchor bolt of the anchored tank of interest from a photo in a reconnaissance report; then, compare it with the uplift displacement calculated. Since the uplift displacement calculated is larger than that inferred. Since a discrepancy between them would be responsible for the constraint effects of anchors that have not been clearly quantified, the present analyses of the tank rocking motion and the deformation of the tank bottom plate may have a potential to give an appropriate uplift of unanchored tanks.

The Mw 9.0 2011 Tohoku-oki tsunami damaged 418 oil storage tanks and moved 157 of them. Using data on the severity of damage and the maximum inundation depth of the tsunami, a fragility curve representing the damage to oil storage tank plumbing is presented in this paper: P( η)= Φ ((ln η − 1.02)/0.574), where P is the damage rate, η is the maximum inundation depth in meters, and Φ is the standard normal cumulative distribution function. The existing method of predicting the movement of tanks exposed to a tsunami is validated by comparing the predicted damage with actual damage data from the 2011 tsunami. The accuracy (hit rate) is 76%.

This article summarizes a passive surface-wave method that uses only two sensors and its application to the estimation of deep S-wave velocity structure. Three-dimensional S-wave velocity structure to a depth of several kilometers has a large effect on long-period ground motion in tectonic basins, such as the Los Angeles (LA) Basin. Recent studies of long-period ground motion in the LA Basin (e.g., Hatayama and Kalkan, 2012) show that observed ground motion in some areas cannot be explained by the S-wave velocity models in current use. Most studies of basin velocity structure rely on geologic information, surface and borehole geophysical data, and observed earthquake records to deduce or measure seismic velocities. Geophysical data and seismic stations commonly used for velocity analysis are sparsely distributed and most well data are too shallow to characterize deep S-wave velocity structure. To establish more accurate basin velocity structure, there is a need for more densely distributed deep S-wave velocity data.

Several cracks were found on some actual floating roofs of a crude oil tank in the oil refinery located in southern Japan. We assumed that one of reasons would be due to thermal stress caused by temperature changes during the day. In order to consider whether the thermal stress could the cause damages on the floating roof, strain and temperature were measured on the actual floating roof by using optical fiber gauges. Furthermore, fracture possibility due to thermal stress was calculated to discuss whether thermal stress could cause fracture or not. As a result, the probability showed that thermal stresswasnot enough to cause fracture. Anotherexternal factor such as Typhoon could be related.

Several cracks were found on some actual floating roofs of a crude oil tank in the oil refinery located in southern Japan. We assumed that one of reasons would be due to thermal stress caused by temperature changes during the day. In order to consider whether the thermal stress could the cause damages on the floating roof, strain and temperature were measured on the actual floating roof by using optical fiber gauges. Furthermore, thermal stress analysis was carried out as effective analysis.

When the oil storage tanks drift from their initial place by tsunami, they may cause severe troubles such as destruction of structures nearby and leaks of storaged oil. Model experiments revealed that the tsunami wave force and friction forces determine the slipping and drifting of the tanks, therefore we constructed a simulation model using the momentum equation for the prediction of the tank movement. The simulation results well represented the track of drifting measured by the model experiment. When the oil tanks start to drift, they tend to drift for several hundred meters unless they remain in the initial place by the force of static friction.

The 2003 Tokachi-oki earthquake (Mw 8.0) in northern Japan generated large-amplitude long-period (4–8 s) ground motions in the Yufutsu sedimentary basin, causing severe damage to seven large oil storage tanks with floating roof structures because of severe sloshing of oil. The 30,000–40,000-m3 tanks having suffered the severe damage such as fires and sinking of floating roofs experienced the sloshing with large amplitudes exceeding 3 m in which the fundamental mode was predominant. The second mode of sloshing was also excited in the 110,000-m3 tanks in which their floating roofs sank into oil, indicating that the higher modes of sloshing as well as the fundamental mode should be considered in damage prediction. The strong ground motion recordings demonstrated the earthquake dependency of predominant periods and the substantial spatial variation of the long-period shaking observed within the Yufutsu basin, meaning the necessity of source- and site-specific prediction of long-period strong ground motions. The two-dimensional numerical modeling suggested the importance of detailed structures of soft near-surface sediments as well as deep basin structure for accurate prediction of long-period strong ground motions in deep sedimentary basins.

We successfully reproduced important features of long-period (7 to 8 sec), large-amplitude ground motions observed in the Yufutsu basin, Hokkaido, during the M w 8.0 2003 Tokachi-oki, Japan, earthquake through numerical simulation of wave propagation. Recordings from the dense nationwide strong ground motion networks in Japan demonstrated significant spatial variation of long-period strong ground motions within the Yufutsu basin and revealed that the long-period shaking in the Tomakomai western port was the strongest of all the sites within the basin. In interpreting or predicting long-period strong ground motions in sedimentary basins, the importance of deep basin structures, such as the depth of bedrock with an S -wave velocity over 3 km/sec has been emphasized. However, the characteristics of the underground structure model of the Yufutsu basin that explain the observations indicate that thickness of near-surface (

The effects of the seawater of Tokyo Bay on strong ground motions are investigated in a short-period range of less than about 2.0 sec for engineering importance. A recent study based on observed accelerograms revealed that considerable short-period surface waves were included in strong ground motions in the reclaimed zone of Tokyo, and that those surface waves were mainly Love waves. However, it was not clarified whether short-period Rayleigh waves were present. The present study is aimed at assessing the excitation of short-period Rayleigh waves on land and at exploring the effects of the shallow (approximately 25 m) water layer of Tokyo Bay on Rayleigh waves in the reclaimed zone of Tokyo. We use a theoretical approach in which a 2D P-SV -wave field is calculated for a multilayered structural model. The model is composed of a sea zone and a land zone, and includes a water layer and soft-surface layers. To examine the effects of seawater, we use models without and with a water layer. As a result, it is concluded that in the reclaimed zone of Tokyo, stripping the water layer off of a typical model for shallow Tokyo Bay has little influence on the short-period ground motions around the theoretical predominant period of the ground of about 1.0 sec. It is not essential to incorporate the water layer in modeling shallow Tokyo Bay for short-period, ground-motion simulations.

The Mw9.0 2011 Tohoku, Japan earthquake tsunami damaged 418 oil storage tanks located along the Pacific coast of the Hokkaido, Tohoku, and Kanto Districts of Japan. A wide variety of damage was observed, including movement and deformation of the tank body, scouring of the tank base and ground, and movement or structural fracture of the pipe. In total, 157 of the 418 tanks were moved by the tsunami. By comparing the severity of damage with the inundation depth of the tsunami experienced by the oil storage tank, a fragility curve projecting the damage rate for plumbing is presented, and a rough but easy-to-use method of predicting tsunami damage to an oil storage tank from a given inundation depth is also presented: (i) for inundation depths of 2–5 m, tanks suffer damage to their plumbing, and small tanks (capacity < 100 m3) and empty larger tanks may be moved; (ii) for inundation depths of greater than 5 m, most tanks are moved. The validity of the previously-proposed tsunami tank-movement prediction method is first examined. A comparison of the method’s predictions with the actual damage data from the 2011 Tohoku earthquake tsunami indicates a high hit rate of 76%.Copyright © 2015 by ASME

Although most seismic damage occurs within 10-20 km of an event, long-period waves can cause destruction at much greater distances. The long-period component of seismic ground motion generated by earthquakes attenuates more slowly with distance than higher frequencies and so persists to much greater ranges. Large-scale structures such as tall buildings and big tanks can resonate with the long-period ground motion because their own natural periods are in the same frequency band. The severity of shaking can be such that panic is produced for people in high-rise buildings from earthquakes several hundreds of kilometers away (Pan, 1995; Pan and Sun, 1996). Such long-period ground motions can even be damaging in some circumstances; the worst example occurred with more than 20,000 fatalities in Mexico City at a distance of 400 km from the 1985 Michoacan earthquake (Beck and Hall, 1986). A further example was provided by Tomakomai, a coastal city on the island of Hokkaido in northern Japan, which was hit by long-period ground motion from the Tokachi-oki earthquake 250 km away on the morning of 26 September 2003 (local time = UT + 9 hours). This Mw 8.3 earthquake was the largest event in the world in 2003 (National Earthquake Information Center, 2004), but damage was not extensive because the location is offshore. No one was killed, although two people fishing during the tsunami are missing. We note that no intensities larger than 6- were observed on the JMA scale, where the highest ranks are 7 and 6+ (Japan Meteorological Agency, 2003). The intensity at Tomakomai was only 5-, which corresponds to no structural damage to houses. Many oil tanks were damaged by the long-period ground motions, however, and two of them caught fire (Coffin and Hirata, 2003), as shown in Figure 1. This article will …

We compare the distribution of damage to housing caused by the 1995 Hyogo-ken Nanbu (Kobe) earthquake with those estimated for several source models proposed for this earthquake. This comparison aims at identifying source models that can provide loss estimates that are most appropriate for planning emergency response activities just after earthquakes and/or for preparing effective countermeasures for mitigation of future earthquake disasters. The results suggest the necessity of finite-fault slip models that can reproduce or predict accurately strong ground motion within a frequency range closely related to damage. The loss estimation just after earthquakes based on source models can be counted on in areas without dense strong-ground-motion observation networks. Even with the dense networks, source models will also be useful for accurate loss estimation in the immediate vicinity of earthquake source faults.

Summary This paper aims to present a new method to calculate surface waves in 3-D sedimentary basin models, based on the direct boundary element method (BEM) with vertical boundaries and normal modes, and to evaluate the excitation of secondary surface waves observed remarkably in basins. Many authors have so far developed numerical techniques to calculate the total 3-D wavefield. However, the calculation of the total wavefield does not match our purpose, because the secondary surface waves excited on the basin boundaries will be contaminated by other undesirable waves. In this paper, we prove that, in principle, it is possible to extract surface waves excited on part of the basin boundaries from the total 3-D wavefield with a formulation that uses the reflection and transmission operators defined in the space domain. In realizing this extraction in the BEM algorithm, we encounter the problem arising from the lateral and vertical truncations of boundary surfaces extending infinitely in the half-space. To compensate the truncations, we first introduce an approximate algorithm using 2.5-D and 1-D wavefields for reference media, where a 2.5-D wavefield means a 3-D wavefield with a 2-D subsurface structure, and we then demonstrate the extraction. Finally, we calculate the secondary surface waves excited on the arc shape (horizontal section) of a vertical basin boundary subject to incident SH and SV plane waves propagating perpendicularly to the chord of the arc. As a result, we find that in the SH-incident case the Love waves are predominantly excited, rather than the Rayleigh waves and that in the SV-wave incident case the Love waves as well as the Rayleigh waves are excited. This suggests that the Love waves are more detectable than the Rayleigh waves in the horizontal components of observed recordings.

Site amplifications during aftershocks of the 1995 Hyogo-ken Nanbu earthquake were examined in and around the disaster area of Higashinada ward, Kobe City, Japan. Five temporary stations together with the KOB station (CEORKA) were installed just after the main shock to arrange an array for strong ground motion observations across a severe damage band about 1.5 km wide. Several aftershock records were obtained from these stations. The peak values of the observed ground accelerations and velocities at the soil sites were 3-20 times larger than those at the rock site. Spectral ratios between the soil sites and the rock site have peak frequencies characteristic to each station pair. We also estimated a thickness of about 1 km for the sedimentary basin just under the soil sites using SP-converted waves recorded at all soil sites. Waveform simulations using a dislocation point source in a horizontally stratified medium were performed to explain the observed records at both the rock and soil stations. We found that large amplifications at the soil sites were generated by a thick sedimentary layer including low-velocity surface layers. The heavy damage caused by the main shock is therefore strongly connected with the sedimentary structure.

Well-developed basin-induced waves with a predominant frequency of about 0.3 Hz were detected as later arrivals of a direct S-wave in seismograms observed in the Osaka basin located in western Honshu, Japan from three deep earthquakes occurring near Japan. The wave type and the secondary source area of the basin-induced waves were examined using a semblance technique and polarization analyses for small aperture array data in the center of the basin. The results well agree with the characteristics of the Love wave such as particle motion perpendicular to a propagation direction and dispersed apparent phase velocities. The arrival direction was evaluated to be almost due east. Seismograms from the east-west linear array crossing the basin show that the observed Love wave propagated from the east to the west after being excited at the eastern edge of the basin by incidence of the direct S-wave. Numerical simulations of two-dimensional SH problems were performed for the east-west cross section models of the Osaka basin subject to incidence of the direct S-wave. The synthetics agree well with the observed Love wave in a predominant frequency and a group velocity.

Several cracks were found on some actual floating roof of the crude oil tank in southern Japan. We assumed that one of causes is due to thermal stress during the day. In order to figure out whether the thermal stress could cause damages on the floating roof, strain and temperature were measured on the actual floating roof by using optical fiber gauges. Furthermore, thermal stress analysis and fracture possibility estimation were also carried out.

It will be difficult to prevent the damage of oil storage tanks caused by a large earthquake even if extreme care has been exercised. Therefore, rational emergency responses will be required for preventing expansion into secondary disaster. In this paper, we propose a system that can estimate ground motion distribution in a whole of petroleum stockpiling base using a seismic record and evaluate damage of oil storage tanks just after an earthquake in order to support rational emergency responses. Spectral ratios of horizontal and vertical component of microtremors were used for evaluation of relative surface soil amplification. The system can assess the hazard for circumferential shell stress, axial shell stress, seismic capacity and liquid sloshing wave height immediately using the ground motion at each tank site estimated from both the relative soil amplification factors and a seismic record at a petroleum stockpiling base just after a large earthquake.Copyright © 2004 by ASME

In this study, we aimed to explore the S-wave velocity structure of shallow soils using microtremors in order to estimate site responses in Tekirdag and surrounding areas (NW Turkey). We collected microtremor array data at 44 sites in Tekirdag, Marmara Ereglisi, Corlu, and Muratli. The phase velocities of Rayleigh waves were estimated from the microtremor data using a Spatial Autocorrelation method. Then, we applied a hybrid genetic simulated annealing algorithm to obtain a 1D S-wave velocity structure at each site. Comparison between the horizontal-to-vertical ratio of microtremors and computed ellipticities of the fundamental mode Rayleigh waves showed good agreement with validation models. The depth of the engineering bedrock changed from 20 to 50 m in the Tekirdag city center and along the coastline with a velocity range of 700–930 m/s, and it ranged between 10 and 65 m in Marmara Ereglisi. The average S-wave velocity of the engineering bedrock was 780 m/s in the region. We obtained average S-wave velocities in the upper 30 m to compare site amplifications. Empirical relationships between the AVs30, the site amplifications, and also average topographic slopes were established for use in future site effects microzonation studies in the region.