PRESENTATION OF THE OPENQUAKE- ENGINE, AN OPEN SOURCE SOFTWARE FOR SEISMIC HAZARD AND RISK ASSESSMENT

Size: px

Start display at page:

Download "PRESENTATION OF THE OPENQUAKE- ENGINE, AN OPEN SOURCE SOFTWARE FOR SEISMIC HAZARD AND RISK ASSESSMENT"

Meryl Simpson
6 years ago
Views:

1 10NCEE Tenth U.S. National Conference on Earthquake Engineering Frontiers of Earthquake Engineering July 21-25, 2014 Anchorage, Alaska PRESENTATION OF THE OPENQUAKE- ENGINE, AN OPEN SOURCE SOFTWARE FOR SEISMIC HAZARD AND RISK ASSESSMENT V. Silva 1, H. Crowley 2, C. Yepes 3 and R. Pinho 4 ABSTRACT Seismic risk assessment is a fundamental tool for the reduction of human and economic losses due to earthquakes. Performing seismic hazard and risk analysis can involve handling large databases and intensive calculations that require robust, accurate and computationally efficient tools. The Global Earthquake Model (GEM) ( developed an open-source software named OpenQuake-engine ( for estimating seismic hazard and earthquake losses. The development of this tool was initiated in 2009 within a pilot project (GEM1), in which several existing hazard and risk software applications were evaluated, with the purpose of defining the first scientific requirements. Currently the software comprises a set of calculators capable of computing fatalities, economic losses and damage distribution for a collection of assets, caused by a given scenario event, or by considering the probability of all possible events that might happen within a region and within a certain time span. Additionally, it also offers a decision-making module that allows the user to understand which seismic design is more advantageous from a financial point of view. The current development status of OpenQuake-engine and main features of each calculator are described in this paper, along with demonstrative results using a seismic risk model for the country of Peru. 1 Senior Risk Engineer, GEM Foundation, Italy, Pavia Risk Coordinator, GEM Foundation, Italy, Pavia Seismic Risk Modeler, GEM Foundation, Italy, Pavia Secretary General, GEM Foundation, Italy, Pavia Silva V, Crowley H, Yepes C, Pinho R. Presentation of the OpenQuake-engine, an open source software for seismic hazard and risk assessment. Proceedings of the 10 th National Conference in Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 2014.

2 Presentation of the OpenQuake-engine, an open source software for seismic hazard and risk assessment V. Silva 1, H. Crowley 2, C. Yepes 3 and R. Pinho 4 ABSTRACT Seismic risk assessment is a fundamental tool for the reduction of human and economic losses due to earthquakes. Performing seismic hazard and risk analysis can involve handling large databases and intensive calculations that require robust, accurate and computationally efficient tools. The Global Earthquake Model (GEM) ( developed an open-source software named OpenQuake-engine ( for estimating seismic hazard and earthquake losses. The development of this tool was initiated in 2009 within a pilot project (GEM1), in which several existing hazard and risk software applications were evaluated, with the purpose of defining the first scientific requirements. Currently the software comprises a set of calculators capable of computing fatalities, economic losses and damage distribution for a collection of assets, caused by a given scenario event, or by considering the probability of all possible events that might happen within a region and within a certain time span. Additionally, it also offers a decision-making module that allows the user to understand which seismic design is more advantageous from a financial point of view. The current development status of OpenQuakeengine and main features of each calculator are described in this paper, along with demonstrative examples using a seismic risk model for the country of Peru. Introduction The OpenQuake-engine represents one of the main products of the Global Earthquake Model [1], a global collaborative effort that brings together state-of-the-art science and national/ /international organizations and individuals with the aim of establishing uniform and open standards for calculating and communicating earthquake risk worldwide. The development of this software follows an open-source philosophy and its code is hosted on GitHub [2], a web-based public repository, under an Affero General Public License (AGPL). The main IT infrastructure and scientific requirements for the OpenQuake-engine were established based on a comprehensive evaluation of existing seismic hazard [3] and seismic risk [4] software. The OpenQuake-engine currently has the following characteristics: An open-source software license with the code available on a public repository; Detailed documentation regarding the assumptions, calculations and methodologies within each calculator; 1 Senior Risk Engineer, GEM Foundation, Italy, Pavia Risk Coordinator, GEM Foundation, Italy, Pavia Seismic Risk Modeler, GEM Foundation, Italy, Pavia Secretary General, GEM Foundation, Italy, Pavia Silva V, Crowley H, Yepes C, Pinho R. Presentation of the OpenQuake-engine, an open source software for seismic hazard and risk assessment. Proceedings of the 10 th National Conference in Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 2014.

3 Possibility to upload user-defined hazard, fragility, vulnerability, site conditions and exposure models; Capability of performing hazard and risk calculations (scenario and probabilistic) in a single software; Consideration of site amplification through the specification of Vs30 values at each site; Characterisation of epistemic uncertainties using logic trees; Compatibility with different types of assets (e.g. buildings, population); Modelling of spatial correlation of ground-motion residuals; Modelling of the correlation of uncertainty in building vulnerability; Capability to parallelize calculations, and possibility to be installed on a single processor laptop, as well as on a cluster or cloud computing infrastructure; Possibility to produce a full spectrum of hazard and risk products such as stochastic event sets, ground-motion fields, uniform hazard spectra, hazard curves and maps, disaggregation plots, damage and loss curves and maps. The main characteristics of each calculator are described herein, with special emphasis in the seismic risk components (physical damage and losses). Several demonstrative results for each calculator are provided, using the country of Peru as the region of interest. Additional information regarding the scientific features and technical details about this software can be found on the OpenQuake book [5] and manual [6]. Description of the main OpenQuake-engine Seismic Risk Calculators Currently, the OpenQuake-engine is comprised by five main calculators: a scenario risk and a scenario damage calculators that are capable of computing loss maps, loss statistics, damage distribution and collapse maps due to a single deterministic event for a collection of assets; a probabilistic event-based risk calculator that estimates the probability of exceedance of certain levels of loss for a portfolio of assets within a given time span based on stochastic event sets, a classical PSHA-based risk calculator that allows the computation of probability of losses and loss statistics for single assets, based on the probabilistic hazard; and finally, a benefit-cost ratio calculator that can support users in understanding whether employing retrofitting interventions or choosing a better seismic design might be profitable from an economic point of view. Scenario Risk Calculator The scenario risk calculator is capable of calculating losses due to a single earthquake, for a collection of assets, using a set of ground-motion fields (i.e. spatial distribution of ground motion within the region of interest), an exposure model (i.e. spatial distribution of the elements at risk) and a vulnerability model (i.e. loss ratio probabilistic distribution for a set of intensity measure levels). The use of a large number of ground-motion fields allows the consideration of the aleatory variability (both inter- and intra- event) in the ground-motion. During the generation of each ground-motion field, the spatial correlation of the intra-event variability can be considered, to ensure assets located close to each other will have similar ground-motion levels [7,8]. In Figure 1, various ground motion fields in terms of peak ground acceleration (PGA) are illustrated for the magnitude 8.0 (M w ) Pisco (Peru) earthquake of August 2007.

78 78 78 6 6 6 6 6 6 8 8 8 8 8 8 10 10 10 10 10 10 12 12 12 12 12 12 14 14 14 14 14 14 16 16 16 16 16 16 18 78 18 18 78 18 18 78 18 PGA PGA PGA 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.4 Figure 1.

4 PGA PGA PGA Figure 1. Ground motion fields (PGA in g) for the Pisco (Peru) earthquake of August 2007 using the ground motion prediction model by Atkinson and Boore [9]: median ground motion (left), one realization of uncorrelated ground motion distribution (center) and one realization of spatially correlated ground motion distribution (right). The set of ground-motion fields is combined with the vulnerability and exposure models, to compute the losses for each asset in the exposure model, per ground-motion field. The correlation in the uncertainty in the vulnerability functions is incorporated such that when sampling the uncertainty in the vulnerability of two assets with the same taxonomy (i.e. of a given building typology), the residuals can be sampled independently, or according to a certain correlation coefficient, defined by the user. This modeling feature aims to model the fact that buildings within a given region are likely to have been constructed with similar materials and with similar construction techniques, and thus their behavior might be correlated. The mean and standard deviation in the loss ratio distribution for each asset are calculated, and converted into absolute loss metrics by multiplying these statistics by the respective cost (e.g. structural cost, non-structural cost or contents value). It is also possible to sum all the losses per ground motion field, and estimate the mean and standard deviation of the total loss for the specified seismic event. In Figure 2, the spatial distribution of the mean economic losses due to structural damage in the residential building stock for the aforementioned seismic event is illustrated. Such results are important for activities like emergency management planning or for raising societal awareness of seismic risk.

Figure 2. Mean economic loss map for the Pisco (Peru) earthquake of August 2007.

5 Figure 2. Mean economic loss map for the Pisco (Peru) earthquake of August Scenario Damage Calculator The scenario damage calculator can be employed to estimate the distribution of damage due to a single earthquake, for a spatially distributed building portfolio. Similarly to what has been described for the scenario risk calculator, a set of ground motion fields can be employed to model the ground motion variability at each location, with or without the consideration of the spatial correlation in the ground motion residuals. In this calculator, each ground-motion field is combined with a fragility model (i.e. probability of exceeding a number of damage states for a set of intensity measure levels), in order to compute the fractions of buildings in each damage state. These fractions are calculated based on the difference in probabilities of exceedance between consecutive limit state curves at a given intensity measure level. This process is repeated for each ground-motion field, leading to a list of fractions for each asset. These results can then be multiplied by the respective value of the assets in order to obtain the absolute building damage distribution. The output of this calculator is the mean and standard deviation of the number of assets in each damage state. This damage distribution can be extracted for each asset, per building taxonomy or the total damage distribution (considering the whole collection of assets). Furthermore, the OpenQuake-engine can also use the amount of assets in the last damage state (usually collapse or complete damage) to produce collapse maps (i.e. spatial distribution of the mean number of assets that collapsed). Considering the Pisco (Peru) earthquake of August 2007, the damage distribution per building typology was calculated, as well as the associated collapse map for the whole country, as depicted in Figure 3. These results can be used in the development of post-earthquake emergency response, or to understand which building typologies are more prone to destructive damage in the occurrence of specific earthquake scenarios.

Number#of#buildings# Adobe# Wooden#Masonry# Stone# RC# Collapse# Extensive# Moderate# Slight# No#damage#

Distribution of damage per building typology (left) and mean collapse map (right) for the Pisco (Peru)

Classical PSHA-based Risk Calculator This calculator uses the classical PSHA approach [10,11] following the

The calculation of these hazard curves is done considering a logic tree structure to incorporate the

prediction models for each tectonic region [13].

6 Number#of#buildings# Adobe# Wooden#Masonry# Stone# RC# Collapse# Extensive# Moderate# Slight# No#damage# Figure 3. Distribution of damage per building typology (left) and mean collapse map (right) for the Pisco (Peru) earthquake of August Classical PSHA-based Risk Calculator This calculator uses the classical PSHA approach [10,11] following the methodology proposed by Field et al. [12] to compute a hazard curve at each site. The calculation of these hazard curves is done considering a logic tree structure to incorporate the epistemic uncertainty from a number of parameters in the source model, as well as different ground motion prediction models for each tectonic region [13]. Through the employment of linear interpolation in the hazard curves, the OpenQuake-engine can also compute hazard maps for different probabilities of exceedance in a given time span, as illustrated in Figure 4 for south America. Figure 4. Seismic hazard map using the OpenQuake-engine implementation of the probabilistic seismic hazard assessment model proposed by Petersen et al.[14].

Using the risk component of the OpenQuake-engine, these hazard curves can be combined with a vulnerability and exposure model to derive asset-specific loss exceedance curves (i.e. probabilities of exceedance for a set of loss thresholds).

Thus, various loss exceedance curves are also calculated for each asset, and a user has the possibility of extracting results from a specific logic tree branch, the mean loss exceedance curves, or

7 Using the risk component of the OpenQuake-engine, these hazard curves can be combined with a vulnerability and exposure model to derive asset-specific loss exceedance curves (i.e. probabilities of exceedance for a set of loss thresholds). Due to the employment of logic trees, it is possible to have multiple hazard curves at each location (one hazard curve per logic tree branch). Thus, various loss exceedance curves are also calculated for each asset, and a user has the possibility of extracting results from a specific logic tree branch, the mean loss exceedance curves, or loss curves corresponding to a specific percentile. In Figure 5, loss exceedance curves for low-code and high-code reinforced concrete buildings in the province of Lima (Peru) are presented. Probability#of#exceedance##in#50#years# 1.00# 0.90# 0.80# 0.70# 0.60# 0.50# 0.40# 0.30# 0.20# 0.10# RC#?#low#code# R RC#?#high#code# R 0.00# low Economic losses Figure 5. Loss exceedance curves for low-code and high-code reinforced concrete buildings in the province of Lima (Peru) for a 50 years time span. The collection of loss exceedance curves can then be used to compute loss maps (i.e. expected loss for a given return period). Such analyses are useful for comparative risk assessment between assets at different locations, or to understand the areas where mitigation actions should be concentrated. high Figure 4. Economic losses for a probability of exceedance of 10% (left) and 2% (right) in 50 years.

8 Probabilistic Event-based Risk Calculator The probabilistic event-based risk calculator uses stochastic event sets and the associated ground motion fields to compute loss exceedance curves for each asset contained in an exposure model. For each ground-motion field, the intensity measure level at a given site is combined with a vulnerability function, from which a loss ratio is randomly sampled for each asset. Spatial correlation of the ground motion residuals and vulnerability correlation in the loss ratios within assets of the same taxonomy can be accounted, as described in the Scenario Risk Calculator. The distribution of loss per asset across all ground motion fields is used to estimate the rate of exceeding a number of loss thresholds, which can then be converted into a loss exceedance curve assuming a Poissionian distribution. Once again, a logic tree structure can be employed in the calculation of the ground motion fields, which leads to a number of loss exceedance curves per asset equal to the number of branches considered in the hazard calculations. Besides the possibility of deriving loss exceedance curves and maps, the consideration of the spatial correlation in the computation of the ground motion allows the aggregation of the losses from all the assets per ground motion field. This list of total losses (one per event) can then be used to estimate a total loss exceedance curve, representative of the whole exposure model. Retrofitting Benefit-cost Ratio Calculator The Retrofitting Benefit-cost ratio calculator allows users to understand if from an economical perspective, a collection of buildings should be retrofitted. This calculator uses loss exceedance curves that can be calculated using either the Probabilistic Event-based Risk calculator or the Classical PSHA-based Risk calculator. These loss exceedance curves need to be calculated considering two vulnerability models: one with the original asset vulnerability, and a second one using the retrofitted vulnerability configuration. The average annual losses considering both vulnerability configurations are calculated, and employed to estimate the economic saving during the life expectancy (or design life) of the assets. This benefit is modified by an inflation rate in order to model the variation in the asset value throughout time. Then, the resulting benefit is divided by the retrofitting cost, thus obtaining the benefit-cost ratio. A ratio above one indicates that a retrofitting intervention would be advantageous from an economic point of view. Currently the OpenQuake-engine performs this assessment considering only losses due to structural damage. However, including losses due to damage in the non-structural components, contents and loss due to business interruption is in its future developments plan. Conclusions The characteristics of the main seismic risk calculators within the OpenQuake-engine have been presented. At present, the OpenQuake-engine is comprised of five main calculators: two capable of computing loss and damage distribution due to single events, two with the purpose of estimating probabilistic seismic risk considering a probabilistic description of the events and associated ground motions that might occur in a given region within a certain time

9 span, and a last one that uses loss exceedance curves to carry out retrofitting benefit-cost analysis. The results from the various calculators are fundamental for the development of seismic risk reduction measures, such as post-earthquake emergency management planning, identification of the regions with higher seismic risk within a certain country, or evaluation of the building typologies more prone to damage on which retrofitting interventions should be prioritized. Despite the large spectrum of capabilities already implemented in the OpenQuakeengine, there are several other functionalities planned for future development or currently being implemented such as: the extension of the logic tree structure to the vulnerability/fragility modeling (i.e. more than a function can be associated to each asset); inclusion of aleatory uncertainty in the economic cost of the assets; possibility to calculate other types of loss exceedance curves (e.g. loss curves considering either only the maximum loss or the aggregated loss per stochastic event set); employment of Nonlinear Static Procedure-based methodologies for the estimation of building response, which can be related to damage distributions (e.g. N2 method [15] or Capacity Spectrum Method [16]); and consideration of other elements such as networks or infrastructures. The OpenQuake-engine is currently being intensively utilized by several institutions and research projects throughout the world for the calculation of seismic hazard and risk (e.g. SHARE: calculation of seismic hazard for the Europe; EMME: calculation of seismic hazard and risk for the Middle East), which allows the development team to better understand the regional requirements and the weak areas of this software where improvements should be prioritized. References 1. Pinho R. GEM: a participatory framework for open, state-of-the-art models and tools for earthquake risk assessment worldwide, Proceedings of the 15th World Conference on Earthquake Engineering, Lisbon, Portugal, OpenQuake-engine public repository (last accessed on 10/17/2013) Danciu L, Monelli D, Pagani M, Wiemer S. GEM1 Hazard: Review of PSHA software, GEM Technical Report , GEM Foundation, Pavia, Italy, Crowley H, Colombi M, Crempien J, Erduran E, Lopez M, Liu H, Mayfield M, Milanesi M. GEM1 Seismic Risk Report: Part 1, GEM Technical Report GEM Foundation, Pavia, Italy, GEM OpenQuake Book (last accessed on 10/17/2013) GEM OpenQuake Manual (last accessed on 10/17/2013) 7. Crowley H, Bommer JJ, Stafford PJ. Recent developments in the treatment of ground-motion variability in earthquake loss models. Journal of Earthquake Engineering 2008; 12(1): Jayaram N. Baker JW. Correlation model for spatially distributed ground-motion intensities. Earthquake Engineering and Structural Dynamics 2009; 38(15), Atkinson GA, Boore DM. Empirical ground-motion relations for subduction-zone earthquakes and their application to cascadia and other regions. Bulletin of the Seismological Society of America 2003; 93(4): Cornell CA. Engineering seismic risk analysis. Bulletin of the Seismological Society of America 1968; 58(5) McGuire RR. Seismic hazard and risk analysis. Earthquake Engineering Research Institute Publication No. MNO-10, Second monograph series, 2004.

10 12. Field EH, Jordan, TH, Cornell CA. OpenSHA: A Developing Community-Modeling Environment for Seismic Hazard Analysis, Seismological Research Letters 2003; 74(4), Monelli, D, Pagani M, Weatherill G, Silva V, Crowley H. The hazard component of OpenQuake: The calculation engine of the Global Earthquake Model. Proceedings of 15th World Conference on Earthquake Engineering, Lisbon, Petersen M, Harmsen S, Haller K, Mueller C, Luco N, Hayes G, Dewey J, Rukstales K. Preliminary Seismic Hazard Model for South America, Proceedings of Conferencia Internacional. Homenaje a Alberto Giesecke Matto Fajfar P. Capacity spectrum method based on inelastic demand spectra. Earthquake Engineering & Structural Dynamics 1999; 28(9), Freeman S. Review of the development of the capacity spectrum method. ISET Journal of Earthquake Technology 2004; 41(1), 1-13.

Development and Application of OpenQuake, an Open Source Software for Seismic Risk Assessment

Development and Application of OpenQuake, an Open Source Software for Seismic Risk Assessment V. Silva University of Aveiro, Portugal H. Crowley, M. Pagani, R. Pinho GEM Foundation, Pavia, Italy D. Monelli