Lifelines 2022: Advancing Lifeline Engineering for Community Resilience

The Los Angeles Aqueduct provides critical supply for the Los Angeles Department of Water and Power conveyance system by transporting water from Owens Valley to the northern San Fernando Valley. The aqueduct’s crossing of the San Andreas fault (SAF) at Elizabeth Tunnel represents a major engineering challenge to risk mitigation. Future SAF ruptures have the potential to completely sever the 2.9-m-wide tunnel and disrupt its operation. Fault characterization and displacement hazard studies were performed to support development of design concepts. Following extensive field investigations, an integrated geological model was developed characterizing the SAF locally as a steeply south-dipping zone of faults that narrows and coalesces from the surface to tunnel depth. Large, calculated displacement hazard values developed in accordance with the design guidelines were compared with average and maximum surface displacements from a global data set of large, continental strike-slip earthquakes to develop final recommendations of fault displacement hazard at the tunnel crossing. Characterization of the fault zone and displacement hazard are being used to develop design alternatives to improve LADWP water supply resilience.

Field investigations can be very expensive for the liquefaction evaluation of buried pipelines, which may traverse through miles of potentially liquefiable zones in seismic regions. Liquefaction phenomenon as a natural process is inherently random and affected by the earthquake magnitude, peak ground acceleration, soil type and relative density, effective overburden pressure, thickness of liquefiable soil, and depth to groundwater and their variabilities. The state of such processes is naturally indeterminate and thus cannot be described with certainty. As such, decisions required for emergency planning of possible pipeline failure are invariably made under conditions of uncertainty. The quantification of the uncertainties and proper evaluation of their adverse effects on the performance of a pipeline can be estimated by performing a probabilistic liquefaction hazard analysis (PLHA) using the Monte Carlo simulation (MCS). MCS uses principles of probability in which the uncertainties are identified, quantified, and combined in a rational manner in order to explain the impact of uncertainties in prediction and forecasting liquefaction hazards along a pipeline. This paper evaluates probabilistically the adverse effects of liquefaction hazards on the structural integrity of a hypothetical pipeline while traversing through miles of saturated and potentially liquefiable young alluvium in the San Fernando Valley and Los Angeles Basin due to earthquakes either on the south San Andreas fault (SSAF) or Newport-Inglewood fault (NIF).

Deterministic Fault Displacement Hazard Analysis (DFDHA) commonly provides a straightforward framework to evaluate the potential extent of surface fault displacements at a lifeline crossing for specific magnitude earthquake scenarios. However, this approach is limited in quantifying the aleatory and epistemic uncertainties that come with characterizing the parameters that control the amount of slip occurring across the fault, the frequency of different size events occurring during the design life of the facility, and the distribution of the different amounts of surface displacements that occur along a rupturing fault relative to the location of the lifeline. These uncertainties cannot be captured by assessing different event scenarios in DFDHA, but a more mathematically rigorous assessment of these uncertainties requires a probabilistic approach (PA) to displacement hazard evaluations. This paper discusses a Probabilistic Fault Displacement Hazard Analysis (PFDHA) as a compliment to the DFDHA. The PFDHA accounts for uncertainties in the DFDHA by using the principles of probability and furnishes a fault displacement hazard curve at the lifeline crossing. The PFDHA results are provided in terms of annual frequency of exceedance (AFOE) or its inverse the average return periods (ARPs) versus magnitude range for fault offsets. In addition to DFDHA, the PFDHA is applied to address the recent concerns regarding the seismic vulnerability of the Metropolitan’s Colorado River Aqueduct (CRA), while crossing the South San Andreas Fault System (SSAFS) at San Gorgonio Pass (SGP) due to Mw7.8 earthquake like the 2008 ShakOut Scenario on the SSAFS. The large seismic events along the SSAFS could result a significant component of vertical displacement at SGP, drastically changing the hydraulic gradient of the CRA and potentially disrupting the flow rate at the crossing. The analysis results provide an AFOE or ARP for an uplift in excess of 2 m at SGP due to large seismic events on the SSAFS and are compared with those of DFDHA results.

The Casa Loma Siphon Barrel No. 1 is a critical water supply lifeline in Metropolitan Water District of Southern California’s (Metropolitan) water delivery system. The 148-in. diameter siphon was constructed in the 1930s and is part of the Colorado River Aqueduct system. The Siphon conveys water across the San Jacinto Valley and the San Jacinto fault zone (SJFZ) in Riverside County. Evidence of cracking and leakage due to localized non-tectonic settlement was noted in the 1960s as concentrated where the siphon crosses a strand of the SJFZ named the Casa Loma fault (CLF). The existing siphon is vulnerable to failure due to tectonic and non-tectonic displacements at the fault crossing. A geologic investigation was conducted for a retrofit project designed to improve the siphon’s seismic resiliency against potential future ground deformation. Exploratory trenching at the siphon fault crossing and a synthesis of paleoseismic and fault mapping data for the SJFZ provided input to fault displacement hazard analyses, and these results plus analysis of historical settlement data provided constraints for the retrofit design. Deterministic and probabilistic coseismic fault displacements range from 5.9–14.2 ft (1.81–4.33 m) and are consistent with other paleoseismic estimates of single-event displacements along the SJFZ. Ultimately, the pipeline was designed to accommodate 12.8 ft (3.9 m) horizontal and 2.6 ft (0.8 m) vertical of coseismic surface displacement (2,475-year return period) on the CLF and 3.3 ft (1.0 m) of differential settlement for a 50-year design life.

The ML5.7 1975 Oroville earthquake seven years after construction of Oroville Dam resulted in recognition of Quaternary activity along the Cleveland Hill fault (CHF) and Swain Ravine fault zone (SRF), located west of the Foothills fault system. A 10-km-long, north-south oriented zone of topographic lineaments including scarps, benches, and saddles identified on lidar coincide with possible northward extension of the CHF on the West Shore of Lake Oroville (WSLO) north of Oroville Dam. The lineaments are parallel with north-south oriented regional bedrock fabric. Field reconnaissance and four fault trench exposures revealed a robust correlation between differential erosion in the Jurassic meta-volcanic bedrock and the linear features. The resulting stepped topography is accentuated by side-hill-benches formed by colluvium that infills areas between resistant bedrock zones. A clay-rich saprolitic unit (175 ka) exposed across a prominent lineament in a trench in Bear Meadow was not faulted, precluding a faulting origin and seismogenic activity for the lineament zone.

In densely populated urban areas, the presence of underground tunnels may modify the input motion for the seismic design of the aboveground structures. However, neither building codes nor seismic zonation guidelines have yet considered this matter. In this paper, the effects of an underground tunnel on the ground motion are numerically investigated, in terms of the peak ground acceleration and acceleration response spectrum. Compared with existing related studies, this paper goes further, considering obliquely incident seismic waves. The results indicate that the underground tunnel influences both the magnitude and distribution of the ground accelerations. The amplification of the peak ground accelerations due to the presence of the underground tunnel can be significant, and the amplification factors are functions of the incident angle of seismic waves.

Investigations made following the 2010–2011 Canterbury Earthquake Sequence (CES) show that the underground pipes have signature failure modes that were unique to the pipe material and some types of pipes, such as those made of PVC, underwent multiple failure modes. Moreover, the damage to pipe joints appeared to show that adding axial joint load to the tests already mandated by NZ standards would greatly enhance the seismic design of pipes. For this purpose, an experimental investigation is conducted on PVC-U pipe joints to ascertain the serviceability and ultimate limit state allowable loads when a pipe joint is subjected to axial seismic-simulated actions. The results are able to replicate the observed earthquake damage by incorporating forces simulating peak ground accelerations in both the horizontal and vertical directions in the testing. From the results, a theoretical framework is developed to illustrate how seismically induced axial forces can be generated in the pipe joint.

Microbially induced desaturation (MID) is a novel, minimally disruptive biogeochemical technique for mitigating the potential for triggering earthquake-induced soil liquefaction. In MID, native microorganisms are stimulated to produce relatively insoluble nitrogen biogas, thereby desaturating the soil. Its minimally disruptive nature makes MID particularly attractive for mitigating liquefaction triggering beneath and around existing facilities, where current mitigation techniques are generally either not applicable due to their disruptive nature or too costly to be economically viable for all but the most critical facilities. Considering the large inventory of existing infrastructure built on or in potentially liquefiable soil, a cost-effective means of mitigating triggering of liquefaction in a minimally disruptively manner could significantly enhance seismic resilience worldwide. The key technical factors influencing the effectiveness of MID for mitigation of liquefaction triggering include delivery of the nutrients for generation of biogas in the liquefaction susceptible soils, the persistence of the induced desaturation, and the impact of local environmental conditions, including groundwater pH and the presence of alternative electron acceptors, e.g., sulfate. Economic analysis suggests that if these technical factors can be managed, MID can be a cost-effective means to mitigate liquefaction triggering beneath and around existing infrastructure.

Earthquake liquefaction hazards in silty soils are a critical problem in Portland, Oregon, and other areas around the world. Recent studies suggest that liquefaction mitigation using microbially induced desaturation (MID) may provide the capability to mitigate liquefaction potential beneath existing structures in a cost-effective manner. The objective of MID is to reduce earthquake-induced excess pore water pressure generation compared to saturated soil, and thereby reduce the potential for triggering liquefaction. A field study of liquefaction mitigation using MID was performed at two sites in Portland in the summer of 2019. Low-plasticity, liquefiable silts were treated with MID by injecting a treatment solution to stimulate native bacteria for a duration of four weeks. This paper presents monitoring data that evaluate the level of desaturation before, during and after the completion of the MID trials. Monitoring included crosshole pressure-wave velocity measurements, a vertical array of embedded sensors that measure water content and bulk electrical conductivity, pre- and post-treatment seismic cone penetration tests, and direct soil sampling of the treated soils. Monitoring at both sites indicated that liquefiable silts were successfully desaturated. The persistence of desaturation was monitored for 8 months post-treatment at one site and is ongoing at the other site. The monitoring data indicate that the induced desaturation has persisted through seasonal fluctuations of the ground water table since the end of treatment. These data, which document MID longevity, are particularly important to establish MID as a viable option for liquefaction mitigation of silty soils that threatens lifelines worldwide.

The cyclic behavior of a fine-grained low plasticity silty soil (plasticity index of approximately 15) at a site in Portland, Oregon, is characterized using a field and laboratory cyclic shear test program. The field cyclic tests were performed using the NHERI@UTexas large mobile shakers T-Rex and Rattler. The laboratory tests were performed using resonant column torsional shear and cyclic direct simple shear devices. The results of this testing program are used to evaluate the soil’s potential to develop excess pore water pressure with cyclic shear strains ranging from 0.00001% to 0.5%. The results of laboratory cyclic tests are compared against the results of field cyclic tests to predict the soil’s cyclic behavior during earthquakes. The field shaking and resonant column torsional shear tests showed significantly lower pore pressure ratios at shear strains up to 0.5% when compared to cyclic direct simple shear tests. These data will contribute to the larger body of knowledge of the cyclic behavior of low plasticity silts.

This project is part of the planned reconfiguration of terminal at LaGuardia (LGA) Airport in Flushing, New York. Based on the results of subsurface exploration program, the stratigraphy of the site consists of fill underlain by organic silt/clay, underlain by sand, underlain by varved clay and silt, underlain by sand, glacial till, and bedrock. Initial liquefaction susceptibility evaluation using the New York City Building Code (NYCBC)-based liquefaction screening tool indicated that the Fill stratum may be susceptible to liquefaction. The site is assigned Seismic Site Class F, therefore in accordance with the NYCBC, it requires a site-specific response analysis to determine seismic parameters for the site. In order to develop design seismic soil profiles, Seismic Cone Penetration Tests (SCPT) were performed to collect shear wave velocity data. A site-specific ground motion analysis was conducted to develop a design acceleration response spectrum in accordance with the applicable provisions of ASCE 7 and NYCBC, using dynamic soil properties. Based on the results of the site-specific analysis, the liquefaction analysis was revised. Liquefaction risk and liquefaction-induced settlements were reduced significantly utilizing site-specific analysis results. The case study demonstrates in addition to Code compliance, performing site-specific seismic studies can obtain a better estimate of potential earthquake-induced shaking at a site, and it could potentially reduce dynamic loads, thus reduce design and construction costs significantly.

Estimating the probability that liquefaction will occur at a given site is a critical first step in calculating the seismic risk on lifelines and structures, such as lateral spreading and ground settlement. Several methodologies for estimating this probability exist. The majority of these methods are based on estimating the factor of safety against liquefaction throughout the soil profile and condensing that information into an index. The probability of liquefaction-induced damage can then be estimated qualitatively (e.g., “severe” liquefaction damage corresponds an index threshold) or quantitatively through fragility curves developed using databases of liquefaction observations. In this paper, we compare novel and traditional methods. We apply cloud analysis to a soil profile at a hypothetical site in San Fernando. This approach consists of performing multiple nonlinear site response analyses and synthesizing their results. The advantages and disadvantages of each method for the design and analysis of lifeline systems are also discussed.

The U.S. Geologic Survey (USGS) routinely updates the National Seismic Hazard Model (NSHM) for the United States and its territories. The NSHM has historically supported calculation of probabilistic seismic hazard for pseudo-spectral accelerations at three periods (0, 0.2, and 1 s) and one reference site condition (a time-averaged 30-m shear-wave velocity of V_S30 = 760 m/s) as required for the design ground motion maps of the National Earthquake Hazards Reduction Program (NEHRP) Recommended Seismic Provisions for New Buildings and Other Structures and other building-code documents. With the recent 2018 update, the NSHM expanded support in the conterminous United States to 22 spectral periods and 8 site conditions ranging from V_S30 = 1,500 m/s (Site Class A) to V_S30 = 150 m/s (Site Class E), as required by the 2020 NEHRP Provisions. Expanding support to more periods and site conditions was made possible largely by adoption of the Next Generation Attenuation project, NGA-East, ground motion models (GMMs) for the stable craton of the central and eastern United States. The USGS is now expanding support to additional intensity measure types. Here, we present the incorporation of peak ground velocity (PGV) in the NSHM. The NGA-East models include PGV, as do many GMMs used to estimate earthquake ground motions in the active crust and subduction zone tectonic environments of the western United States. For those GMMs considered in the NSHM that do not natively support PGV, we use a newly developed conditional model to estimate PGV given values of pseudo-spectral acceleration (PSA). Unlike prior models that are conditioned on a single spectral period, the period used in the new model is magnitude dependent. The model introduces a parameter T_PGV, which refers to the spectral period with the highest correlation between PSA and PGV at the given magnitude. This feature captures the magnitude scaling observed in the corner frequency of the earthquake source and also leads to an improved model of aleatory variability. The model is derived from observations in active tectonic regions, and we explore its application to other tectonic environments.

The Port of Portland has identified the deep and thick deposit of medium dense sands underlying its runways and terminals as a significant risk to post-earthquake recovery. Efforts to design mitigation of liquefaction and lateral spreading below its runways have been complicated by the inability to easily sample these soils and the lack of case history data on the performance of medium dense sands at depth. This paper describes a joint agency-industry-university collaboration to quantify the dynamic response of such soils at depth using controlled blasting tests conducted at the Port of Portland near Portland International Airport (PDX). The experimental field-testing program is described, including the geotechnical characterization of the deposit, instrumentation, and blasting program. The typical characteristics of the blast-induced ground motions are quantified to highlight the reduced significance of P-waves, owing to their frequency content, in the dynamic response of the sand. Following the presentation and discussion of typical shear strain and corresponding excess pore pressure time histories for various blasts, the relationship between shear strain and excess pore pressure is quantified in the medium dense sands at an average depth of 25 m and 2.5 atmospheres of vertical effective stress. The relationship between shear strain and excess pore pressure will be used by the Port and its consultants to calibrate constitutive models used to assess various ground improvement alternatives.

The spatial variation of ground motions affects the response of the distributed lifeline structures such as bridges and natural gas pipelines within a seismic region. At present, the number of earthquake ground motion recording stations is generally sparse. Therefore, ground motion simulation methods are deployed to estimate the input motions at locations with no available instruments. In this study, a Gaussian Process Regression model is presented to interpolate the surrounding observed ground motion time series to estimate the motion at an adjacent unobserved location. The interpolation process is done using a Gaussian Process Regression, which models the Fourier spectrum’s real and imaginary parts as random Gaussian variables. The physics-based simulated motions for the 1906 San Francisco earthquake is employed to train the model. The M7.0 Hayward fault fully deterministic physics-based simulated earthquake ground motions are used to evaluate the trained model’s performance. The preliminary evaluations illustrated that the trained Gaussian Process regression model could decently predict the ground motion time series.

We developed an efficient platform to select and consistently scale recorded earthquake ground motions. Given a set of target response spectra for horizontal and vertical ground motion components, a set of ground motions is selected and scaled such that: (1) the mean spectra of the selected and scaled horizontal and vertical motions would follow their respective target spectra; (2) the selected set of horizontal and vertical spectra would preserve the prescribed period-dependent record-to-record variability for their respective components; and (3) for each set of horizontal and vertical components, a single scaling factor is used; thus, preserving the relative amplitude and phasing of the original recorded horizontal and vertical components. Additionally, significant improvement in computational efficiency is achieved by employing a modified version of a greedy record selection algorithm.

Seismically induced slope displacements (D) are often used as a performance index in the seismic design of slope systems. Although previous studies have developed predictive relationships for estimating D, a comprehensive selection of the most efficient parameters to explain D, considering the linearity and nonlinearity in D, has not been thoroughly performed. This study uses modern machine learning-based techniques including Stepwise Selection, LASSO, and Random Forest to identify the influential features in the estimation of D in shallow crustal tectonic settings. The detected influential features are the system’s yield coefficient (k_y), initial fundamental period (T_s), earthquake moment magnitude (M_w), peak ground velocity (PGV), and the degraded spectral acceleration at 1.3T_s [S_a(1.3T_s)]. Moreover, the results indicate that there is no significant gain in accuracy beyond five features. The detected significant parameters provide insight and a basis for developing more efficient prediction models.

Improving the seismic performance of pipelines is fundamental to achieving resiliency of urban environments against extreme natural events. The vulnerability of pipelines is inevitably associated with their exposure to regional and local geohazards. Seismic hazard assessments of pipelines are uniquely challenged by varying local soil conditions, as evidenced by the concentration of damages in areas prone to ground motion amplification and liquefaction. This paper focuses on the hazard associated with transient seismic waves, which affects the entire pipeline system as opposed to the localized effects imposed by ground failure. Our study provides a review of seismic hazard assessments of pipeline networks and current practices to account for site effects. More specifically, we review available methods to consider the effects of local soil conditions on the seismic demands of pipelines, soil-pipe interaction, spatial variability in soils, and ground motion directionality. First, we summarize relevant observations from past seismic events and then, an evaluation of current practices to account for site effects is provided. Strengths and limitations of simplified approaches are revised and compared to methods that target the quantification of site effects and the network response on a broader regional scale. Areas of future research are identified as potential paths toward improved resiliency of pipelines to seismic hazards.

Underground structures are commonly constructed near existing or new tall buildings in dense urban areas. Tall buildings, during earthquake shaking, generate base shear that is carried by the building foundation and surrounding soils. This base shear may be transmitted to adjacent underground structures. A previous study was conducted by combining dynamic centrifuge tests and numerical simulations to evaluate the impact of highly idealized adjacent tall buildings on the seismic response of underground structures. The numerical model can reproduce the seismic behavior observed in the centrifuge including the additional loading demands imposed by adjacent building on underground structures. Furthermore, a large-scale parametric study using three-dimensional (3D) nonlinear finite element analysis with more realistic soil-structure-underground structure (SSUS) representation is performed to evaluate the impact of variability in the SSUS system on the seismic response of underground structures. The effects of different building heights, building foundations, underground structure configurations, and soil profiles are evaluated using a suite of ground motions. The results show that (1) as the adjacent building became taller and hence the base shear increased, greater dynamic earth pressures were transmitted to the underground structure, (2) the dynamic earth pressures were reduced with increasing building to underground structure distance, and (3) the racking displacements of underground structures were strongly dependent on building foundation and underground structure configurations such as basement depth, pile length, and the depth of the underground structure. For design purposes, these interactions need to be considered by modeling a realistic building and underground structure representation in SSUS system to evaluate the added demands that a given building will impose on underground structures including cut-and-cover boxes and bored tunnels.

Coherent patterns and large variations in ground shaking amplification were observed in the Los Angeles basin during the 2019 M7.1 Ridgecrest earthquake. In particular, 3 s to 6 s responses showed variations due to shallow basin geological structure that have implications for the response to large earthquakes of mid-rises, high-rises, long-span bridges, and fuel storage tanks, even if epicentral distances are several hundred kilometers. The Ridgecrest strong-motion data were recorded by seismic stations from the spatially dense Community Seismic Network, the Southern California Seismic Network, and the California Strong Motion Instrumentation Program. The mainshock observations are compared at the same locations with ground motion simulations to examine the regions that experienced the largest shaking, and to investigate the geological sources of large-amplitude shaking. The simulations were computed for the two most commonly used regional community seismic velocity models, CVM-S4.26.M01 (‘CVM-S’) and CVM-H 15.1.0 (‘CVM-H’). Both observations and simulations are used in dynamic analysis with a finite-element model of an existing high-rise with 6-s fundamental horizontal periods, located in downtown Los Angeles. The geographical variation in maximum story drift, story-level shear force, and story-level moment values suggest that the excitation of a hypothetical high-rise located in an area characterized by the largest 6-s PSA values could be significantly larger than in a downtown Los Angeles location. Ground motion simulations using the CVM-H velocity model more closely predict the long-period site amplifications in greater Los Angeles, particularly in the south-central San Fernando Valley, than simulations using CVM-S.

Many of the high voltage transmission line towers in Southern California are located in hilly terrains and in a high seismic hazard environment which requires assessment of these towers during a major seismic event. Some of these towers are located within historic landslide features making them quite vulnerable for large deformations during seismic shaking. Permanent ground deformations due to slope failures during and after a seismic event could result in losing some of these towers and their functions. Therefore, in this study, a seismic assessment in terms of slope stability was performed for several high voltage towers along a transmission power line corridor from Tejon Pass to Cajon Pass in Southern California. Incidentally, this corridor runs along the Southern San Andreas fault. Based on an initial study including review of geology maps and site visits, out of 2,269 structure locations along the corridor, 15 locations were selected for further evaluations for seismic slope stability. Field investigations consisting of borings were performed at some of these sites and soil/rock samples were obtained for laboratory testing. At some locations, field survey was performed to establish elevation contour lines in order to develop appropriate cross sections. In addition, at some locations, towers were assessed for both local slope stability and global slope stability. Local slope stability concerns were raised because some towers are located next to a cut slope which was made primarily for access roads. The analyses involve developing ground motions using UCERF3 seismic source model, developing cross sections (both local and global), assigning strength properties to each layer, and performing seismic slope stability analysis. Ground motions were developed for probability of exceedance of 5-, 10-, 20-, and 50-percent in 50 years and parametric studies were performed to establish design level earthquake. Seismic deformations were estimated using simplified method. For locations where deformations exceeded a certain threshold, different mitigation methods were proposed including but not limited to soldier pile and lagging walls and soil-nail walls. Challenges for this study include high ground motions resulting in about 1 g for PGAs, anticipated large slope deformations, and difficult terrain to implement mitigation methods. This paper will present the thought process to select the design level earthquake based on the parametric studies, results of analyses in terms of FOS and seismic deformations, and details about mitigation methods used at certain locations.

The seismic performance of the Balboa Boulevard during the 1994 M_W6.7 Northridge earthquake is examined through nonlinear deformation analyses (NDAs). This study evaluates the effect of plausible site conditions and modeling decisions not accounted for in a previous study, including: (1) stratigraphic models with different anisotropic correlation ratios; (2) constitutive model calibrations based on semi-empirical correlations for the rate of shear strain accumulation of sand-like soils, and excess pore pressure of clay-like soils; and (3) spatial variability of ground motions. Additionally, the role of the characteristic ground motion pulse of the Rinaldi Receiving Station recording on the observed displacements is studied by using non-pulse records caused by the same earthquake faulting style that have lower peak ground accelerations, but similar cumulative absolute velocities (CAVs). Findings indicate that the displacements are significantly sensitive to calibration procedures and the ground motion spatial variability, moderately sensitive to the modeled correlation ratio, and ground motion pulse, and not sensitive to the variations of the potential maximum excess porewater pressure of clay-like soils.

A series of two large-scale shake table tests were performed to investigate the efficacy of the polymer injection technique in remediating liquefaction-induced foundation settlement. In these tests, foundation-ground system response was studied first without and subsequently with polymer injected into the liquefiable stratum, resulting in a set of unique and comprehensive data. Here, a set of preliminary class C1 numerical simulations of these tests are presented using a 3D two-phase (solid-fluid) fully coupled finite element model with parameters partially calibrated from earlier studies. The physical and numerical models indicate an extensive reduction in the liquefaction-induced foundation deformations due to the injection and associated stiffening of the ground. Comparisons are made between computed and recorded engineering demand parameters (EDPs), including excess pore pressure, soil acceleration, and foundation settlement. A reasonable match is observed to the physical response, and the calibrated model is further used to run an additional scenario with sloping ground conditions. Directions for further improvement of the model are then suggested.

Several underground projects in seismic areas are currently dealing with active fault crossings, in such locations the tunnel lining performance is a key element for the project’s success. The Westside Purple Line Extension Project–Section 2 (WPLE2) in Los Angeles is crossing a zone affected by Santa Monica Fault. Advanced three-dimensional numerical models were developed to capture the tunnel lining response to several fault rupture scenarios. A special precast concrete tunnel lining (PCTL) was designed to accommodate the imposed displacements on the tunnels due to the risk of a strike slip fault rupture. This innovative design was originated as a value engineering idea put forth by the Design Builder (Tutor Perini and Frontier Kemper) for an improved tunnel lining system that was more easily constructed and reduced long-term maintenance issues. The tunnel lining incorporates numerous special seismic provisions to safely accommodate the imposed fault displacements during a Maximum Design Earthquake (MDE) event. Finally, a physical testing program was developed in conjunction with the University of California San Diego (USCD) to physically test all key aspects of the special lining, document actual behavior, and allow for refinement of the structural design.

Wildfires and earthquakes contribute to a nearly ever-present cycle of hazards that are man-aged by coastal California communities every year. Worse still, fires, and earthquakes drive slope instability, primarily in the form of debris flows, debris avalanches, and debris floods whose runout can impact environment, infrastructure, and threaten lives along the landslide path. A better understanding of future landslide runout paths, travel distance, and potential landslide depth along the path, will improve our ability to manage future hazards; however, predictive models can be difficult to implement, hard to calibrate, and/or expensive to acquire. DebrisFlow Predictor is an agent-based runout model that predicts runout, inundation, scour, and deposition along the path, of debris flows and debris avalanches. Results credible and easily verified (numerically or visually) using several built-in features. DebrisFlow Predictor is intended to better inform and constrain land management decisions where debris flow and debris avalanche hazards exist.

At the southern end of the San Andreas Fault, the complex tectonic environment near the Salton Sea generates mud pots and mud volcanoes. These features, powered by carbon dioxide gas released at depth, fluctuate in their activity level but remain stationary. In 2016, a mud pot (referred to as a mud spring) began moving southwest. Critical infrastructure, including dual railroad tracks, a petroleum pipeline, underground communication lines, and a state highway, were in the moving mud spring’s path. To maintain the infrastructure operations, we placed temporary barriers to hinder the movement, attempted to de-pressurize the mud spring, and rerouted the tracks, pipeline, and communication lines. As of February 2021, the mud spring moved over 100 m (300 ft) from its 2016 location, cutting a path through these critical infrastructure alignments. Scientific studies to understand the unprecedented movement continues as the mud spring journeys across the desert.

Earthquake damage scenarios are required to support design and analysis of spatially distributed infrastructure systems. In this paper, we develop a computationally efficient set of damage scenarios for the Los Angeles water transmission system that considers ground motion and liquefaction. Each damage scenario describes one possible realization of damage to the pipe network and includes the corresponding multihazard scenario and an associated adjusted annual occurrence probability. Each damage scenario, which specifies the damage state of each pipe in the network, is defined to be physically realistic and consistent with the associated multihazard scenario. Together, when probabilistically combined, the set of damage scenarios with their occurrence probabilities matches the probabilistic hazard and component damage distributions. The scenarios are selected to be small in number so that subsequent analysis is efficient. We combine ideas from recently developed methods to generate sets of multihazard scenarios as well as damage scenarios for analysis of spatially distributed infrastructure systems. The method applied in this paper involves simulating multiple hazards and a number of respective damage scenarios, and using optimization to select a subset of damage scenarios and assign adjusted occurrence probabilities.

As history would have it, the recent Mw7.1 Puebla-Morelos intraplate event occurred on September 19, 2017, coincidental to the 32nd anniversary of the 1985 Mw8.0 Michoacan interplate event. The former had less impact both in affected areas and in severity than the latter which had profound impacts on the nation. Factors affecting the respective impacts include: ground motion input, site setting of Mexico City with its thick lake-sediment deposits, and structural types of affected building stocks. The interplay of these factors at the two events shared general similarities with specific nuances, and the author participated in the Canadian reconnaissance teams after both events. This paper discusses the vulnerability of Mexico City to the two events that are representative of two types of source mechanisms: the recent event being moderate yet closer, while the earlier one being major but further away. In addition to the difference in earthquake magnitude and epicentral distance, they also generate ground motions of different frequency contents. Thus, site amplification differs in the two events depending on thickness of lake sediments in various seismic zones within the city, and the paper first presents an overview of the physical and seismo-tectonic setting of Mexico City. The earthquake and damage data for this pair of events are summarized in a table. Four aspects of the general earthquake impacts are then discussed: ground failure, building performance, building retrofit, and lifelines performance. Moreover, the earthquake early warning system, having its origin because of the 1985 event, has been fully operative since 1993. Up to the 2017 event, the system had recorded 6,896 earthquakes and issued 158 early warnings. Finally, a succinct summary highlights key lessons learned from this pair of seismic events.

This paper focuses on the damage to drinking water supply system in the 2018 Hokkaido Iburi-tobu earthquake. An earthquake of about 35 km in depth and magnitude M_JMA=6.7 occurred at east of Iburi region in Hokkaido, Japan, at 3:07 local time on September 6th, 2018. Maximum seismic intensity of 7 was observed at Atsuma Town. The drinking water supply system in the stricken areas was severely damaged. Water service disruption occurred at about 68,000 households in 43 waterworks bureaus just after the earthquake. Water service disruption continued for more than one month at Atsuma Town. Tomisato purification plant and service reservoir of Atsuma Town suffered severe damage by a large-scale landslide. This paper introduces an outline of the earthquake and the damage induced by the earthquake, especially, damage to the drinking water supply system. Furthermore, lessons learned from the damage by this earthquake are discussed.

The 2017 earthquake in Central Mexico (epicenter East of Ayutla, Mexico Earthquake) that terrorized Mexico City and Puebla occurred on the same month and day as the 1985 Michoacan, Mexico earthquake, which caused significant damage to lifelines in Mexico City. The official name of this earthquake is Puebla Earthquake. The differences between these two earthquakes are locations of the epicenter, the duration of strong shaking, and most importantly the magnitude. The 2017 earthquake epicenter was located about 150 km south-east of Mexico City, it happened at 13:14 local time and the magnitude was 7.1 with a duration of about 60 ss of strong motion. The 1985 earthquake epicenter was located about 330 km south of Mexico City, it happened at 07:17 local time and the magnitude was 8.0 with duration of 90 s of strong motion. Thus, all lifelines in Mexico City and areas around Puebla sustained various degrees of damage and service interruption. Electric power system experienced the most disruptions due to both substation damage and distribution system failures. Although the telecommunication system did not sustain much damage, the power system failures did cause some telecom service interruptions in areas with long duration power outages. Cellular messaging applications on smart phones reportedly were performing well when the voice call lines were saturated during the first day after the earthquake. There were scattered roads and bridges failures and also broken underground water pipelines. There was a report of liquid fuel tank failure but there was no report of cascade damage due to the tank failure. The airport was closed for a short time and there was a short section of the road around the airport terminal that cracked due to ground failure. Very minor damage was observed in the international airport terminal. There were also building failures and collapses including part of a school building. There were incidents of fire with one fire caused by gas tank explosion. Emergency services performed well. Therefore, lessons learned from the 1985 earthquake probably contributed to overall better telecommunication lifeline performance. There was no report of Central Office (CO) failures.

The 2019 Ridgecrest earthquake sequence produced extensive surface rupture affecting the Naval Air Weapons Station, China Lake, and multiple water pipelines that service the towns of Trona and Argus. This paper documents observations of surface rupture and their effects on buried water pipelines at four pipeline-fault crossings. At these crossing locations surface ruptures ranged from about 0–60 cm (V) and 2.1–330 cm (H). Some surface ruptures displayed complicated patterns. The water pipes are made of multiple materials and they are approximately 30.5 to 40.5 cm in diameter. For each crossing, the surface rupture characteristics and the observed pipe damages are described. It is anticipated that the field data presented herein will serve as a resource for subsequent research to validate and enhance existing knowledge on the behavior of faulting surface rupture and impacts on buried water pipelines.

Extensive previously and newly available data collected on two pipelines along Balboa Blvd. that failed during the 1994 M_W6.7 Northridge Earthquake are compared against numerical simulation results to evaluate the ability to predict ground deformation patterns consistent with field observations. Pipeline data consists of failure features and mode (compressional or tensional), as well as type, material, and dimension of pipelines and joint connections. A previous numerical evaluation conducted by the authors showed good agreement between the magnitude of observed and numerically estimated ground deformations at Balboa Blvd. These results are herein compared against the damage data and interpreted to provide a second order validation to the employed numerical tools and procedures by reconciling soil deformation, pipeline failure, and numerical data. Results from this study suggest advanced numerical modeling that yield reliable peak ground displacements and strain distribution can very well correlate to pipeline failure patterns.

This study presents a framework to quantify resilience of intermodal freight transportation networks disrupted by seismic events. The framework integrates available seismic fragility and restoration models for infrastructure components, and poses a port operation model along with a network model that couples ports with highways and railways to evaluate freight flow through intermodal transfer nodes, an integration that is currently lacking in existing literature. Emphasis is placed on assessing damage to intermodal systems, functionality evolution for the individual modes, and intermodal freight throughput metrics such as TEUs (twenty-foot equivalent unit). The proposed framework enables loss assessment considering physical and operational disruptions. Furthermore, this work lays a foundation for future integration of the infrastructure resilience model with community scale economic impact models by tracking the class and volume of goods disrupted across the affected region over time. A case study illustrates application of the proposed framework and opportunities for future work leveraging a hypothetical intermodal network consisting of a container seaport connected to roadway and railway networks, exposed to varying levels of scenario seismic hazard events.

The research explores a dimension of preparedness against natural hazards by investigating disaster-induced service disruptions of critical lifelines. By using empirical survey data for regression modeling, the research establishes a methodological framework for examining protective actions and adjustments to service disruptions. Respondents were shelter-in-place households impacted by the power, water, and health facility disruptions caused by Hurricane Florence or Hurricane Michael in 2018. The study analyzes influencing factors such as sociodemographic characteristics, previous experience and expectation of service disruptions, and preparedness ability. Findings from regression models determined that number of protective actions were most influenced by expectation of power disruption, greater days of preparedness, and greater perception of preparedness, which emphasizes the importance of communicating and educating the community about preparedness measures. Additionally, socially vulnerable populations were associated with distinct adjustments such as selecting to tolerate the power disruption as opposed to leaving the residence or using a power generator.

The belief that individuals are more vulnerable to earthquakes that onset during nighttime hours is a common preconception. One reason for this belief may be a presumed limited capacity for individuals to react and to protect themselves appropriately in darkness. To better understand this issue, this study surveyed the experiences of households during the 2018 Eastern Taiwan Earthquake, a magnitude 6.4 event that struck Taiwan at 11:50 PM local time. Data were collected with regard to structural, environmental, social, and household contexts, as well as perceived shaking intensity, risk perceptions, individual preparedness, earthquake experience, and demographic characteristics, and compared against reported immediate individual protective actions taken. The results, which were then compared to data from two comparable daytime earthquakes (the 2011 Christchurch Earthquake and the 2011 Great East Japan Earthquake), indicate that approximately half of the respondents who experienced the 2018 Eastern Taiwan nighttime earthquake (49.7%) did not take any protective actions. Regression analysis revealed that respondents’ social and environmental contexts, directly associated with the earthquake occurring at night, affected the individuals’ emotional response, and, in turn, their adoption of protective actions. The results of this study highlight possible differences in people’s response to earthquake shaking under diverse scenarios, and in turn, encourage researchers and practitioners to design preparedness and response exercises that account for situational variation in earthquake shaking responses. This study also provides clear guidance regarding the response of individuals during earthquakes that can better inform policy decisions and structural regulations in earthquake-prone areas, reducing injury and promoting household and community resilience.

We present a new consequence-driven framework for earthquake scenario selection. For emergency managers, utility operators, policy makers, and other stakeholders, a scenario-based seismic risk assessment is often necessary for the purpose of emergency management and planning. In developing a scientifically defensible scenario, stakeholders can simulate a realistic event in order to pre-identify vulnerabilities in the system and support action to address these vulnerabilities. Selecting scenarios is particularly challenging for important population centers and critical infrastructure in stable tectonic environments, such as in the central and eastern United States, where uncertain long-term seismicity and unknown faults offer inadequate constraints. Notably, significant events in these so-called stable regions do occur (e.g., Nahanni, Canada, 1985, M6.9; Tennant Creek, Australia, 1998, M6.7). In regions of low seismicity, even moderate events can be consequential due to the higher vulnerability of buildings typical of such regions when compared to regions of higher seismicity. Furthermore, communicating seismic risk to stakeholders and the general public in these regions can be especially challenging due to the complexities of characterizing the hazard level. This framework has been developed to address these challenges for scenario selection in low seismic hazard regions. In this new approach, the analysis begins instead with the explicit definition of a consequence of concern to the specific stakeholder. This can range from a definition of loss (in lives, dollars, or another metric of interest), or a performance metric for critical infrastructure. The framework leverages United States Geological Survey software to run the hazard and consequence analysis. Driven by this stakeholder-defined consequence, an inversion analysis generates a complete event set of candidate scenarios that could breach this consequence. The final selection of a scenario, or family of scenarios, is then scientifically informed, but not limited by our lack of constraints in characterizing the hazard.

This paper provides a necessary analysis of community tolerance for prolonged disruptions of infrastructure services due to natural hazards. Currently, little is known about how households are affected by prolonged service disruptions. In this paper, we develop survival models for disruptions in power systems using empirical data collected in the aftermath of three major hurricanes, namely Hurricane Harvey, Hurricane Florence, and Hurricane Michael. Survival analysis was conducted to determine the household’s tolerance for disruptions of infrastructure services. Consequently, survival curves were developed, showing the proportion of households experiencing hardship when the disruptions exceed certain levels. The presented survival models could be adapted in various locations for different services to assist utility companies and involved stakeholders for developing metrics to consider the societal expectation of the infrastructure services. The introduced measures facilitate decision-making by considering service needs of the community, setting priority for restorations, and better communicating risks to service users.

This paper introduces a framework to account for physical damage to infrastructure and household disaster preparedness in estimates of potable water needs after earthquakes. A case study is presented involving the water supply system to the city of San Francisco after a M_w=7.9 earthquake. Accounting for household preparedness helps identify regions in the city where water supply is interrupted, and many people may not have personal resources to access alternative sources of water. Considering both infrastructure disruption and household characteristics may inform decisions to allocate emergency water resources across the city during emergency response.

The objectives of this study are to develop an estimation procedure for pipeline damage sites after an earthquake disaster in formulating an emergency restoration plan in the water sector, and to investigate the emergency restoration strategy for smart recovery of economical and industrial activities. In this study, a standard fragility function model on the peak ground velocity (PGV) and the damage rate is developed. In addition, the disaster resilience curves were calculated in the case study of Nishi-Mikawa region after the resumed Nankai trough earthquake disaster in Japan using a numerical simulation model on emergency restoration. We calculated an opportunity loss as an integration value of the economic activities rate of decline by the suspension of water supply with the pipeline damage in the emergency restoration period. As a result, it was pointed out that disaster resilience in water sector would require not only earthquake-resistant pipe technology, but also emergency restoration strategy for mitigation of opportunity loss.

Fire following earthquake as a cascading event can cause major losses in a community. This study quantifies induced earthquake damages to water and transportation networks and the building inventory to model fire following earthquake losses considering the response time of fire department, suppression actions, and fire spread inside and between the buildings. The fire engine assignments are completed by incorporating a decision-making algorithm to minimize losses. The model assigns fire engines dynamically over time, and incorporates the post-disaster changes in response time and water network serviceability level as a result of the earthquake. The framework is applied to a case study to investigate potential regions with vulnerability, which in-turn can be used for preparedness and mitigation as part of resilience planning. The efficiency and adequacy of fire department emergency response are shown to be dependent on the locations of ignition and building layout in the community.

The Wellington Lifelines Group (WeLG) is comprised of the lifeline utilities serving the Wellington region of New Zealand. They share a collective understanding that Wellington’s infrastructure is vulnerable to natural hazards, which was gained from previous studies based on a variety of modelling and expert opinion approaches. These initial studies provided a foundation for deeper analysis to be undertaken around the vulnerability of the region’s infrastructure to earthquakes. The focus was on the likely damage states, interdependencies, and direct and wider economic consequences of a potential major earthquake in the region. The in-depth study used computer modelling and focused stakeholder engagement to quantify the improved resilience and economic benefit of the construction of a package of infrastructure upgrades. The study informs consideration of the mitigations that could be applied to address the core issues—short-term (emergency planning works), medium term (policy change at central government level), and long-term (the potential construction of new, resilient, infrastructure). This paper outlines this WeLG project, including the objectives, process, and intended outcomes.

This study presents a novel, newly developed framework for resilience assessment of tall buildings and demonstrates its application on a 42-story PEER TBI RC core wall building. A nonlinear model of the building is created in OpenSees using a new three-dimensional model for RC walls. Building seismic response is obtained for three seismic hazard levels corresponding to SLE, DBE, and MCE-level events. Based on the results of nonlinear analyses, building damage is evaluated using FEMA P-58 tools, while the building recovery process is evaluated using a sophisticated recovery model. Building seismic performance obtained using new structural and recovery models suggest that the post-earthquake functionality of the tall building is primarily governed by the structural subsystem. For example, for a DBE-level event, the building’s functionality will be fully impaired due to significant damage to slab-column connections and shear walls; the building will regain its functionality in 8 to 13 months.

The seismic resilient infrastructure should have the capability to reduce the damage induced by earthquakes and minimize subsequent losses with an effective recovery. This study presents a probabilistic approach to evaluate the seismic resilience of the circular tunnel in soft soil. Typical circular tunnels in Shanghai are adopted as a case study, and the corresponding numerical model is established by ABAQUS. A series of ground motions are selected and scaled from 0.1 g to 1.0 g to perform non-linear dynamic analysis, aiming to obtain the tunnel responses under increasing levels of ground shaking intensity. Fragility curves are generated at the free-field ground surface, considering the main sources of uncertainties. Furthermore, based on the proposed fragility curves and the existing empirical tunnel restoration functions, the development of resilience index (Re) with the peak ground acceleration (PGA) at the free-field ground surface for circular tunnel is evaluated and quantified. The results show that this type of tunnel is good of seismic resilience.

Critical infrastructure or lifelines are an essential requirement for society, but measuring, configuring, and designing a lifeline to achieve resilience is a complex and challenging task. We examine urban water and electric power distribution resilient grid concepts using an idealized region typical of a mid-sized US city, stressed by earthquakes of increasing intensity. The damaged water network is hydraulically analyzed for demands due to leaks, breaks, and post-earthquake fires in a pressure driven analysis (PDA). Electric networks are similarly examined considering seismic retrofitting of electric substations. We find: (1) resilient water grids have a benefit-cost ratio (BCR) of about 6:1 to 8:1 for high seismicity regions, such as Los Angeles and San Francisco, with lower values for Seattle and Portland; (2) a major benefit of a resilient water grid is improved supply of firefighting water; (3) resilient power grids have a BCR exceeding 8 for the high seismicity regions; and (4) resilient grids will significantly reduce time to restore water and electric power supply.

Earthquake damage to water utilities harms society far out of proportion to pipe repair costs. Water districts should be equipped to assess their resilience in terms of impacts to themselves, but also in terms of societal impacts. They should know mitigation options, costs, and benefits. We developed a tool called CUWNet to do that. We documented CUWNet’s equations and data tables in open, peer-reviewed publications. CUWNet estimates damage considering earthquake shaking and geotechnical failures. It estimates service restoration over time, considering aftershocks, repair-crew productivity, lifeline interaction, and resource limits. It reasonably hindcasted Napa’s experience after the August 2014 South Napa Earthquake. An Excel implementation was straightforward enough that Las Vegas water system engineers can use it. CUWNet enables a utility to measure and manage its resilience in terms that matter to society. Doing so could better align the interests of the utility and the community to improve the seismic resilience of both.

This study evaluates the economic impacts of a M_w7.0 Hayward fault scenario earthquake on the greater San Francisco Bay Region’s economy and the California economy as a whole using a detailed multiregional, static computable general equilibrium model. Economic impacts in terms of Gross Regional Product (GRP) losses caused by both capital stock (building and content) damages and water and electricity utilities, and telecommunications-service disruptions are estimated. The results indicate that the total losses are primarily caused by capital stock damages. In the 6 months following the earthquake, total GRP losses are estimated to be $44.2 billion (4.2 percent of California’s projected baseline GRP over the period), but this result could be reduced by about 43 percent to $25.3 billion after factoring in microeconomic resilience tactics. The GRP losses associated with lifeline service disruptions are estimated to be $1.4 billion, which can be reduced by over 85 percent when resilience tactics are implemented. The most effective tactics are the ability to make up lost production by people working overtime or extra shifts (production recapture), making greater use of processes that do not need disrupted goods or services (production isolation), and substituting for disrupted supplies and services (input substitution), though their impact varies across the various causal factors influencing GRP losses.

This paper describes how impacts to infrastructure networks within the San Francisco Bay Area may exacerbate the effects of building damage and how policies addressing these networks can improve resilience before and after the earthquake. The analysis uses existing modeling techniques that underlie the Association of Bay Area Government’s (ABAG) 2015 regional economic forecast of the San Francisco Bay region, California to estimate how a moment magnitude (MW) 7.0 earthquake scenario along the Hayward Fault, HayWired, would change the trajectory of that forecast. The ABAG forecast released in January 2015 is built on the framework of a Regional Economic Models, Inc. (REMI) model for the San Francisco Bay region and projects growth in the bay area through 2040. Using the simulation tools in the REMI model, the analysis applies the direct output losses flowing from building damages from the HayWired scenario (estimated using the FEMA Hazus model) to ABAG’s economic and demographic 2015 regional forecast. Also the analysis estimates direct, indirect, and induced effects on gross regional product (GRP), employment and population, and also highlights the effects of the physical infrastructure damage to roads, bridges, and rail to the region’s economy. Communications infrastructure, if resilient or restored, can help counteract the losses generated by building and transportation network damage. The REMI model results show that in the first year, employment would drop by almost half a million jobs, whereas GRP would decline by 8 percent. The two counties near the epicenter of the earthquake would have greater losses, of 15 percent in jobs and 13 percent in GRP. Counties with less physical damage may still have economic slowdowns due to transportation disruption. Much of the economy could recover within a few years, but a full return to the projected trajectory could take more than five years for the region and closer to a decade for the most severely affected counties. Recovery and rebuilding investments will be crucial to repairing the economic base of the region and returning it to its projected growth trajectory. State and local policies, as well as business and personal preparedness and employer flexibility in allowing remote work can reduce the length and severity of effects. Furthermore, sensitivity analyses using the model identify some critical factors that would lead to different levels of change. For example, a shortage of construction workers could result in a deeper, longer recession as rebuilding is postponed. Should major technology employers decide to relocate substantial portions of operations or expand outside of the region, the recovery period from the earthquake induced recession could stretch to six or seven years and the region’s trajectory could be permanently damped relative to the ABAG 2015 forecast for 2040.

Airborne Synthetic Aperture Radar (SAR) sensors are suitable for collecting information in an emergency response phase since they can acquire high-resolution images regardless of weather and sunlight conditions. Due to the Kumamoto earthquakes on April 14 and 16, 2016, scores of bridges were damaged and around 200 landslides occurred. In this study, we set the target area in Minami-Aso Village, Kumamoto Prefecture, Japan, which was suffered from numerous landslides. The landslide areas including several affected bridges were investigated using the pre- and post-event Pi-SAR2 airborne SAR images. As a result, the HH polarization and the surface scattering component were found to be most suitable to detect landslides by the pre- and post-event difference. The collapsed bridges were also extracted properly using the contribution ratio of each scattering component within bridge outlines.

Masonry is the most widely used material due to its availability and ease of construction in various rural parts of the world. The brittle nature of the masonry makes the buildings more vulnerable during an earthquake. Existing base-isolation techniques require specialized manufacturing processes, skilled labor, and uneconomical for masonry buildings. This study focuses on a novel isolation technique with locally available low-cost rubber as a cost-effective emerging solution for masonry buildings to overcome these challenges. A numerical study conducted on an experimentally validated one-story masonry building subjected to ground motion and responses are compared between fixed-base and spread rubber isolated-base buildings. The wall-top displacement, wall-top acceleration, and failure stress indicated a significant improvement in seismic performance of rubber isolated masonry building. The proposed spread base-isolation technique using locally available natural rubber can be a socio-economic solution for seismic protection of masonry buildings in developing countries’ rural seismic regions.

For lifeline systems, routine visual inspections play a vital role in gathering structure-level data to inform asset maintenance strategies. Current research has shown that data repeatability and objectivity may be enhanced with computer vision and robotics, whereby manual collection, cataloguing, and quantification of defects is automated. For instance, low-cost lidar sensors integrated with a robotic platform can collect 3D geometric information of a structure via simultaneous localization and mapping (SLAM), with image data fusion allowing for defect measurement. While low-cost lidars provide a fast and inexpensive way of obtaining defect measurements, maps they produce often lack the density required for accurate quantification. To remedy this, we combine depth completion enhanced point cloud data from a lidar with labelled image data to extract a more complete surface estimate in physical scale. Given a 3D map, image data with known camera poses, and camera intrinsic calibrations, our approach first transforms the 3D map to a depth map (within camera frame) through ray casting. Depth maps are then densified using a depth completion algorithm that employs image processing techniques tailored to suit sparse lidar map data. Lastly, the combination of labeled defect images and dense depth maps is exploited to extract high-resolution area measurements for defects such as spalls and delaminations. The accuracy of our method has been assessed by comparing results to those obtained from non-depth completed data and to ground truth measurements obtained using a high-resolution monocular camera with manual scale input to each image. Using a data set containing six concrete area defects, our method was shown to reduce error, on average, by 14.2%, with depth completed results having an average area error of 5.3% from the ground truth. Our results show that depth completion on sparse 3D maps can be effective in fine-scale defect quantification, providing a more complete assessment of lifeline system condition using affordable sensors.

Research on artificial intelligence (AI) and aerial imaging technologies has led to new opportunities in large-scale damage assessment for disaster management. As in other domains, the quality of AI predictions in disaster damage assessment primarily depends on the size, quality, and heterogeneity of training data. However, each disaster leaves behind a unique visual pattern of destruction that is influenced by both the intrinsic properties of the event (e.g., wind speed of a hurricane, ground acceleration of an earthquake) and the characteristics of the location it is taking place (e.g., building portfolio, terrain, vegetation cover). This makes the development and assessment of AI models for universal damage assessment extremely challenging, as the size and feature space of data collected from each disaster is distinctive. This study aims to evaluate and benchmark the generalizability of a damage assessment AI model across different hurricane events. The model is a stacked convolutional neural network (CNN) that is initially trained on aerial videos from Hurricane Dorian (2019) but is tested without retraining on unseen videos of several hurricanes in the 2017 season, including Harvey, Maria, and Irma. The testing accuracy and the square of Earth Mover’s Distance (EMD²) values are used to evaluate the model generalizability. Overall, the model achieves 36.7% precision in detecting buildings and 43.1% accuracy in damage level classification. Additionally, EMD² is found to be negatively correlated with accuracy. Results also indicate that the model generally has a better performance in predicting extreme ground truth labels (i.e., no damaged or destroyed buildings). This behavior is expected and consistent with how human annotators label buildings for damage since humans find these extreme cases more straightforward to label than cases where buildings have suffered intermediate damage levels.

Using space-borne images and their time series for disaster-induced damage detection and mapping has been a classic topic since the advent of remote sensing (RS) technology. The traditional methods usually start with a change detection step that extracts structural changes followed by a pattern classification model. By exploiting recent advances in deep learning, the authors developed a novel metric learning-based damage classification model, termed Enhanced Triplet Network for Damage Detection (EnTriNet-DD). The model enables a dual-phase training and learning of inter- and intra-class variances of damage levels in bi-temporal image pair for application in remote sensing-based disaster damage mapping. Compared with other existing damage-learning methods in the literature using remote sensing images, this model first model inter- and intra-class variances explicitly, significantly improving learning efficiency, damage-classification accuracy, and robustness to uncertainties. With this preliminary research, we argue that in the advent of remote sensing big data, deep-learning and RS-based damage mapping can contribute to the lifecycle management of disasters both for lifeline systems and communities.

The increasing use of tall buildings worldwide has led to rising demand for strategies to control lateral displacements in these buildings. Viscoelastic Damper (VED) devices are passive systems that depend on energy dissipation through a constrained viscoelastic material. This study aims to study the effect of angle orientation of the infill pattern on the energy dissipation of the 3D-printed Thermoplastic Polyurethane (TPU) VED device. The mechanical and viscoelastic properties of 3D-printed TPU with a rectilinear infill pattern printed at three angles were investigated using cyclic uniaxial tension tests. The VED with 3D-printed TPU was subjected to sinusoidal signals with varying frequencies and strain amplitudes to examine its energy dissipation capabilities. The tests showed that the 3D-printed VED is capable of dissipating energy with good efficiency.

During nuclear power plant (NPP) operations, operators selectively focus their attention on alarms, indicators, and control switches. Failures in allocating attention at the right time on the right parts of the workspace have been observed to cause 35% of situation assessment errors. Existing methods are inefficient in preventing such failures due to a lack of comprehensive attention allocation processes in various NPP operations. Such characterization is critical for the real-time identification of gaps in human attention for monitoring operational hazards. This paper presents formal knowledge representations for characterizing visual attention allocation processes of NPP operators. The proposed knowledge representations systematically mapped out attention drivers such as task structures, objects, and action types in operating procedures. Such a formalized representation of NPP operation knowledge can support automatic reasoning methods to identify critical objects and safety-related properties during reactor startup processes. Capturing this operational knowledge and understanding operators’ visual attention allocation principles would help establish a human-centric computing framework for ensuring the safety and efficiency of NPP operations.

In this study, damage to old type steel conduits is predicted using machine learning. We used a supervised machine learning algorithm with a gradient boosting method. To construct the estimation model, we use inspection data from five past large earthquakes: the 1995 Southern Hyogo Prefecture Earthquake, Mid Niigata Prefecture Earthquake in 2004, the Niigataken Chuetsu-oki Earthquake in 2007, the 2011 off the Pacific coast of Tohoku Earthquake, and the 2016 Kumamoto Earthquake. The damage was checked by pipe inspection camera. The model was constructed using 75% of the 23,833 records of the inspection results, and a test was carried out using the remaining 25% of the records. Each record has a label (damaged or no damage), and has 14 variables. Using this method, we constructed a model which has high prediction accuracy, an area under the receiver operating characteristic curve of 0.87 with the test data. The converted displacement in this model contributed the most by confirming the contribution of the respective variables. Old steel conduits with screw joint cannot absorb displacement, and the estimation model may be interpretable.

To ensure the resilience of our society, the risks associated with complex natural hazards, such as a strong earthquake followed by aftershocks must be evaluated for essential structures. To handle the computational challenges caused by uncertainties in aftershock hazards, a probabilistic framework using a Bayesian network (BN) is recently proposed. In an effort to apply the BN-based framework to complex structural systems, this paper investigates a reinforced concrete (RC) flat slab building, which is a typical structural type in Mid-America. Multiple measures of structural responses including maximum inter-story drift ratios and seismic intensities of ground motions can be modeled by a BN to perform probabilistic inference based on the corresponding probabilistic relationships. To overcome the challenges that arise when handling complex systems in the BN, the overall modeling and inference processes are efficiently executed by the Matrix-based Bayesian network. The proposed framework is expected to contribute to the maintenance of a resilient community in the face of hazardous events by evaluating the risk of sequential earthquakes and making risk-informed decisions.

Shear wave velocity of soil is an important parameter for earthquake analysis and civil engineering designs. Correlations between shear wave velocity and SPT-N values based on the simple power-law regression model were provided by previous studies. However, due to the inherent uncertainties of the SPT, estimation of shear wave velocity may be improved by incorporating other site investigation data in addition to SPT-N values. In this study, artificial neural networks were used for the correlation analysis between the shear wave velocity with SPT and CPT data, based on recent field test data obtained from the site investigation program of a major transportation project at Macau. The results of this study indicated that incorporating the additional soil parameters in the correlation model would improve the prediction performance. When combined with SPT-N and CPT-q_c to form a neural network model, a better prediction would be obtained than that using SPT-N or CPT-q_c alone.

Bridges are indispensable elements in resilient communities as essential parts of the lifeline transportation systems. Knowledge about the functionality of bridge structures is crucial, especially after a major earthquake event. In this study, we propose signal processing approaches for automated AI-equipped damage detection of bridges. Mel-scaled filter banks and cepstral coefficients are utilized for training a deep learning architecture equipped with Gated Recurrent Unit (GRU) layers that consider the temporal variations in a signal. The proposed framework has been validated on an RC bridge structure in California. The bridge is subjected to 180 bi-directional ground motion records with sampled scale factors and six different intercept angles. Compared with the benchmark cumulative intensity features, the Mel filter banks resulted in 15.5% accuracy in predicting critical drift ratios. The developed strategy for spatio-temporal analysis of signals enhances the robustness of damage diagnosis frameworks that utilize deep learning for monitoring lifeline structures.

Peak floor acceleration (PFA) can be useful immediately after an earthquake to potentially optimize recovery resources, especially in cases of invisible damage that can be detrimental for essential lifelines. While code equations estimate PFA based on the design spectrum, simplified methods, available in the literature, have proven to be cumbersome for real-life structures. To mitigate this, a machine learning (ML) approach using Artificial Neural Networks (ANN) was investigated to quickly predict PFA using ground motion parameters and structural properties without the need for time-history analysis (THA) or providing ground motion time series. While previous studies fed ground motion time-series data to the ML model, this study investigates the use of ground motion parameters instead. Data from linear elastic THA were used to train the model. Results illustrated that a shallow ANN predicted PFA with coefficient of determination (R²) reaching 0.8; however, additional studies are underway to enhance the model performance.

Identification of modal parameters of a structure is useful for verification, validation, and improvement of Finite Element Analysis (FEA) models, identifying structural degradation and damping appropriate to anticipated structural response levels. Accurate identification of actual modal properties consequently leads to more accurate structural analysis and response prediction. Modal properties are defined as resonant frequencies, mode shapes, and modal damping. The effective values of these modal properties may vary depending on the excitation levels. The effective modal property variations with amplitude are gradual but shed light on the nature of these non-linearities and their relation to structural damage. These experimental studies focus on various experimental techniques to identify modal properties of bridges, viaducts, and other civil structures. These techniques include ambient vibration surveys, impulse excitors, eccentric mass vibrators producing sinusoidal forcing, and servo hydraulic vibrators producing impulsive, random, or sinusoidal forcing. Example results and applications are presented with emphasis on railway bridges and viaducts, but also dams and nuclear power plant structures. Many of the test and structural model identification techniques were developed at the University of California Los Angeles following the 1971 San Fernando Earthquake.

Rapid progress in the field of sensor technology has led to acquisition of massive amounts of measured data from structures being monitored. The data, however, contain inevitable measurement errors which often cause quantitative damage assessment to be ill conditioned. Attempts to incorporate a probabilistic method into a model have provided promising solutions to this problem by treating the uncertainties as random variables usually modeled with Gaussian distribution. However, the success enjoyed by the probabilistic method is limited by the lack of adequate information to obtain an unbiased probabilistic distribution of uncertainties. In addition, the probabilistic surrogate models involve complex and expensive computations, especially when generating output data. In this study, a non-probabilistic surrogate model based on wavelet weighted least squares support vector machine (WWLS-SVM) is proposed to address the problem of uncertainty in vibration based damage detection. The input data for WWLS-SVM consists of selected wavelet packet decomposition (WPD) features of the structural response signals, and the output is the Young’s modulus of structural elements. This method calculates the lower and upper boundaries of the changes in the Young’s modulus based on an interval analysis method. Considering the uncertainties in the input parameters, the surrogate model is used to predict the output of this interval bound. The proposed approach is applied to detect simulated damage in the four-story benchmark structure of IASC-ASCE SHM group. The results show that the proposed method can perform well in uncertainty-based damage detection of structures with less computational efforts compared to direct finite element model.

Soil contamination due to pipeline leakage has an adverse effect on environment. Geoelectrical methods can be used in assessing the hydrocarbon contamination within soil medium. This study experimentally investigates the efficiency of Electrical Resistivity (ER) and Induced Polarization (IP) in detecting and monitoring hydrocarbon soil contamination as a result of buried pipeline leakage into clayey backfill material. The work presents the results of a laboratory-scaled free field model contaminated by a petroleum hydrocarbon light non-aqueous phase liquid (LNAPL). Localized areas within the soil layer were injected with a 10% used motor oil by weight. The Electrical Resistivity Tomography (ERT) and IP surveys were conducted in 2D, 3D, and 4D formats to verify the location of the contaminated area as well as its dimension and its progress with time. To better understand the geoelectrical characterization of contaminated model soil, a series of proctor compaction tests are carried out using a modified plastic mold. The results of these specimen tests are used to interpret the outcome of ERT and IP surveys. Based on these results, geoelectrical tools may be used in subsurface investigations for delineating the hydrocarbon plume location along pipeline projects. It is also discussed how LNAPL detection by these methods can help in designing sampling strategies to save cost and time.

In various civil engineering applications (CEAs), most vibration responses vary with time and space and are characterized by nonlinearities and uncertainties that are not accounted for during data acquisition. Furthermore, these responses may be contaminated by various sources, which may affect the damage identification process. The main challenge is how to denoise these data in order to acquire a sensitive feature for damage identification that is insensitive to noise and environmental effects. Wavelet Transform (WT) has been proven to be useful for denoising in the field of structural health monitoring (SHM). However, its efficiency is affected by the selection of wavelet parameters. The questions related to the best approach for utilizing the most suitable parameters have not been adequately answered. This study attempts to address this issue by proposing a new denoising algorithm based on the Discrete Wavelet Transform (DWT) technique. The proposed technique provides a strategy to choose the right decomposition levels for denoising and selecting proper mother wavelets. The proposed algorithm uses separate noise thresholds for negative and positive coefficients at each level and applies denoising to detail and approximation components. Datasets from actual civil structures have been analyzed. According to the presented experimental results, the proposed technique exhibits promising results for signal denoising using “db3” compared with traditional techniques. In addition, “db3” and “sym3” are shown to be the best choices for the mother wavelet.

This paper puts forward a real-life assessment of a novel inverse filtering methodology to extract bridge features from acceleration signals recorded on smartphones in the passing vehicles. The vibration of a moving vehicle is affected by various features, such as suspension and speed. This study focuses on filtering out these effects from the signals to extract bridge frequencies as the vehicle crosses the bridge. Hence, the spectrum of the vibration data recorded on the vehicle when moving off the bridge is employed to form an inverse filter which removes the vehicle-related frequency content. Since the speed of the vehicle is found to be one of the most effective factors in the filter design in our previous studies, a database of the off-bridge vibrations is built for different speeds. Later, when the same vehicle is moving on the bridge, the corresponding inverse filter is applied to the recorded on-bridge data to suppress the vehicle frequencies and amplify the bridge frequencies. All the required data are recorded using the built-in accelerometer and GPS sensor of the smartphone, eliminating the need for any extra instruments. In addition, this approach considers each data source separately and designs a unique filter for each data collection device within each vehicle, which makes it robust against device and vehicle features. As a result of the proposed methodology, it would be possible to monitor a large number of bridges using crowdsourced data collected from the smartphones in the vehicles. Such methodologies are expected to improve the sustainability and resiliency of our future infrastructure systems and future cities.

ShakeCast is a US Geological Survey (USGS) software application that automatically retrieves ShakeMap shaking estimates and performs analyses using fragility functions for buildings and lifelines. The ShakeCast system aims to identify which facilities or lifeline segments are most likely impacted by an earthquake—and thus which ones should be prioritized for inspection and response—and sends notifications to responders in the minutes after an event. By focusing inspection efforts on the most damage-susceptible facilities in the severely shaken areas, ShakeCast can improve critical lifeline inspection prioritization and reduce response time in the aftermath of a significant earthquake. Overviews and technical specifications of the ShakeCast system and software have been presented at earlier conferences; here we discuss ShakeCast users and their response protocols to provide further insight into the use of the ShakeCast system. We focus on case histories of ShakeCast users who are responsible for monitoring and response for critical infrastructure. We emphasize the inventory, fragility, and notification issues pertinent to these users, their efforts in developing protocols for post-earthquake inspections, and response.

Quarrying operations between the late 19th and early 20th century transformed the once moderately rolling, stable eastern slope of San Francisco’s Telegraph Hill to a slope with near-vertical rock faces prone to rockslides. One such failure occurred following an intense rain event in December 2019 on a portion of slope directly beneath the Filbert Steps, a popular tourist attraction leading to the iconic Coit Tower. As part of the emergency response, geological mapping using three independent methods was performed. These methods were manual measurements using a geological compass, unmanned aerial system (UAS) data capture, and terrestrial Light Detection and Ranging (LiDAR) data capture. The results of this study showed structure from motion (SfM) models derived from UAS photogrammetry to be an effective method for characterizing a rock mass, capable of obtaining similar results to more expensive pieces of equipment (i.e., LiDAR) and eliminating the safety concerns associated with mapping unstable slopes using a geological compass. The large dataset of geological structure orientations obtained from the UAS imagery-based SfM model facilitated relatively robust analyses at an early stage in the project, and provided a basis for slope stabilization recommendations both for emergency response and permanent restoration. This paper presents a comparison of the three methods used for geological mapping and provides commentary on the effectiveness of UAS mapping for provision of efficient slope stabilization recommendations.

With an increased effort to improve public safety during an earthquake event, several cities and regions have developed and/or are developing unreinforced masonry (URM) retrofit policies. One of those cities is Portland, Oregon. In October 2019, the City Council of Portland continued its process, with origins in 1995, towards developing a voluntary seismic retrofit policy. This paper reviews Portland’s process and develops a socio-economic equity analysis framework that can complement the process. The framework synthesizes socioeconomic data and Portland’s URM Building Database to assess the relationship between socio-economic status and the spatial distribution of existing URM buildings. The framework is also used to identify aspects of Portland’s URM policy recommendations that can potentially have negative impacts on low-income residents. Finally, results from the analysis are used to put forth suggested policy considerations. The findings of this work could inform Portland in developing a robust and equitable URM policy that mitigates the severity of earthquake damage, without imposing an undue burden on low-income areas in Portland.

Our existing utility network is a system of systems, whereby a service disruption to any portion of one utility system has the probability of a cascading effect, not only on other utility systems, but also other lifelines such as transportation. Accurate records of these systems are poor to non-existent. Active utility systems may have records, but the quality of those records is uncertain in regards to accurate location and attributes information. Abandoned utilities usually have no recoverable records. This lack of accurate records leads to recovery issues after destructive events. The American Society of Civil Engineers put together two standards to assist in solving this problem. ASCE 38, Standard Guideline for the Collection and Depiction of Existing Subsurface Utility Data, prescribes a series of actions that result in affixing a value judgment of utility presence and location uncertainty to buried utility segments. ASCE 75, Standard Guideline for the Recording and Exchanging of Utility Infrastructure Data, prescribes a series of actions that result in the documentation of the accuracy and measurable attributes of above and below ground observable utility segments. Placing these data into a central secure database after project development for access by utility owner and emergency personnel is challenging in the US, due to proprietary and security concerns. Public officials need to weigh the risks versus rewards for keeping an up-to-date comprehensive central repository of our underground lifelines.

In response to a Congressional mandate to recommend “options for improving the built environment and critical infrastructure to reflect performance goals stated in terms of post-earthquake re-occupancy and functional recovery time,” the National Institute of Standards and Technology and Federal Emergency Management Agency have initiated efforts to address the functional recovery performance of buildings and lifeline infrastructure systems. This paper focuses on recommendations for lifeline systems, emphasizing the need to: (i) establish a consistent national performance framework across all lifelines to enable consistent functional recovery objectives; (ii) provide national guidance on regulatory authorities to enable consistent implementation; (iii) evaluate existing lifeline infrastructure to assess expected performance and identify and prioritize upgrades; (iv) develop national standards for orderly recovery at component, system, and system-of-systems levels; (v) create regional councils to identify lifeline interdependencies and manage agency coordination; and (vi) enable recovery through planning, education, and financial resources.

Engineers must above all else protect and advance the health, safety, and welfare of the public. Professional societies create codes and standards that guide development of the built environment to serve the public needs. Elected officials are, ultimately, responsible for the allocation of public funding to ensure the resilience and overall performance of new and existing infrastructure. This intersection of public policy and engineering is a place where engineers should take a more active role. This study outlines a process for public accounting using value analysis of the regional infrastructure, including benefit-cost analysis for new structures as well as existing infrastructure. CRISP, Community Resiliency Infrastructure Protocol, is founded on probabilistic estimates, stochastic analysis, and accepted discounting assumptions. The ASCE Infrastructure Resilience Division has initiated a project to further develop this concept. The goal is to create methodologies that can be utilized for decision-making, improving the resilience and sustainability for infrastructure systems and communities.