Seismic Risk, Resilience, and Disaster Decision-Making

Introduction and Context

Earthquake engineering has traditionally focused on preventing structural collapse and reducing casualties. Contemporary practice extends this purpose to the protection of buildings, infrastructure systems, livelihoods, institutions, supply chains, and essential services. A structure may remain standing yet be unusable because of nonstructural damage, utility interruption, access constraints, or loss of organisational capability. Seismic risk management therefore requires a broader decision framework that connects hazard, structural response, damage, loss, recovery, and investment.

Seismic risk is not a single number. It is a probabilistic description of adverse consequences generated by earthquake hazard acting on exposed and vulnerable people, assets, and systems. Resilience concerns the capacity to withstand disturbance, maintain or rapidly restore function, adapt, and improve future performance. Disaster decision-making uses these concepts to select actions under uncertainty and competing objectives.

For postgraduate civil engineers, the strategic challenge is to move from “How strong is this building?” to questions such as:

• Which assets create the greatest societal or organisational risk?
• What performance objectives are appropriate for different earthquake intensities?
• Should limited funds be used for strengthening, redundancy, emergency preparedness, insurance, inventory management, or land-use intervention?
• How should short-term cost be balanced against avoided losses and recovery benefits?
• How can technical evidence be communicated to decision-makers and communities?

This chapter develops an integrated framework for answering these questions.

Learning Objectives

By the end of this chapter, the reader should be able to:

1. Distinguish hazard, exposure, vulnerability, risk, resilience, and disaster loss.

2. Explain how fragility and vulnerability functions support seismic risk analysis.

3. Formulate expected annual loss and event-loss estimates under probabilistic hazard.

4. Compare mitigation alternatives using life-cycle cost and decision-analysis principles.

5. Incorporate business continuity, recovery time, interdependencies, and portfolio effects.

6. Design resilience-oriented emergency and recovery strategies.

7. Critically evaluate uncertainty, equity, governance, and ethical issues in risk reduction.

8. Apply a structured approach to prioritising investments at building, infrastructure, organisational, and community scales.

1. The Seismic Risk System

1.1 Hazard, exposure, vulnerability, and consequences

Earthquake hazard describes the likelihood and intensity of ground motion or related phenomena, such as fault rupture, liquefaction, landslide, tsunami, and fire following earthquake, at a location. Hazard may be expressed through intensity measures (IMs), including peak ground acceleration, spectral acceleration at a specified period, duration, or a vector of several measures.

Exposure is the people, buildings, infrastructure, economic activity, cultural assets, and environmental resources located in the hazard area. Vulnerability is the propensity of exposed elements to suffer damage or functional impairment when subjected to a hazard intensity. Consequences include fatalities, injuries, repair costs, business interruption, displacement, environmental effects, and social disruption.

A useful conceptual expression is:

Risk = Hazard × Exposure × Vulnerability × Consequences.

This is not generally an arithmetic equation, because each term is probabilistic and may be spatially and temporally correlated. It is a reminder that reducing risk can involve reducing hazard where feasible, controlling exposure through planning, reducing vulnerability through engineering, or reducing consequences through preparedness and continuity planning.

1.2 Risk, disaster, and resilience

Risk exists before an event. A disaster occurs when consequences exceed the coping capacity of affected households, organisations, or governments. The same earthquake can produce a manageable emergency in one city and a disaster in another because of differences in building quality, governance, health capacity, lifeline redundancy, social vulnerability, and recovery resources.

Resilience has several related dimensions:

• Robustness: the ability to resist damage and maintain function.
• Redundancy: the availability of alternative components, routes, suppliers, or facilities.
• Resourcefulness: the capacity to mobilise finance, information, personnel, and equipment.
• Rapidity: the speed with which services and productive activity are restored.
• Adaptability: the ability to learn and change policies, designs, and institutions.

Resilience is therefore broader than strength. A highly redundant water network may be more resilient than a stronger but isolated pipeline. Similarly, a hospital requires not only structural safety but also functioning power, water, communications, medical supplies, staff access, and information systems.

2. Fragility, Vulnerability, and Loss Models

2.1 Fragility functions

A fragility function gives the probability that a component reaches or exceeds a specified damage or limit state conditional on an intensity measure. A common lognormal representation is

P(DS ≥ ds | IM = x) = Φ[ln(x/θ_ds) / β_ds],

where:

• `P(DS ≥ ds | IM = x)` is the probability of reaching or exceeding damage state `ds`;
• `Φ` is the standard normal cumulative distribution function;
• `x` is the selected intensity measure;
• `θ_ds` is the median intensity producing the damage state;
• `β_ds` is the logarithmic dispersion, representing uncertainty and variability.

The model assumes that the logarithm of the capacity associated with the damage state is normally distributed. It is applicable when the selected intensity measure adequately predicts the response and when the empirical or analytical data support a lognormal approximation. The dispersion may include record-to-record variability, modelling uncertainty, construction variability, and limitations in the damage-state definition.

A fragility curve is conditional, not a prediction of actual damage in every event. For example, a 0.30 probability of exceeding “moderate damage” at a given spectral acceleration means that approximately 30 percent of comparable components are expected to exceed that state over repeated observations or an equivalent probabilistic population.

Fragility functions may be developed from:

• empirical observations after earthquakes;
• analytical nonlinear simulations;
• laboratory testing;
• expert elicitation;
• hybrid empirical-analytical methods.

Their quality depends on representative construction, detailing, ageing, soil conditions, modelling assumptions, and appropriate intensity measures.

2.2 From fragility to damage probabilities

Suppose damage states are ordered from DS0 to DS4, where DS0 means no or slight damage and DS4 means collapse or complete loss. If exceedance probabilities are known, the probability of an individual state is obtained by differencing:

P(DS = d) = P(DS ≥ d) − P(DS ≥ d+1).

For the highest state, the probability of DS4 equals its exceedance probability. For DS0, the probability is one minus the probability of exceeding DS1. This calculation requires consistent damage-state definitions and monotonic exceedance probabilities.

2.3 Vulnerability functions

A vulnerability function relates hazard intensity to a consequence measure, often the mean damage ratio (MDR). MDR is commonly defined as expected repair or replacement cost divided by the relevant asset value:

MDR(IM) = E[Loss | IM] / V.

Here, `E[Loss | IM]` is the expected direct loss conditional on intensity, and `V` is the value basis, such as replacement cost. If damage states have representative damage ratios `r_d`, then

E[Loss | IM] = V Σ p_d(IM) r_d,

where `p_d(IM)` is the probability of damage state `d` at the specified intensity. The ratios must be defined consistently: a repair-cost ratio should not be combined with a replacement-value denominator without checking scope, depreciation, and contents treatment.

Loss models may include:

• structural repair and replacement;
• nonstructural and contents damage;
• demolition and debris removal;
• downtime and business interruption;
• casualties and emergency response;
• temporary accommodation;
• infrastructure and utility disruption;
• indirect economic and social losses.

Direct financial losses are usually easier to estimate than intangible losses, such as cultural heritage, psychological trauma, educational disruption, or ecosystem effects. Excluding them does not mean they are unimportant; it means that a particular model has a restricted decision scope.

3. Probabilistic Seismic Risk and Loss Metrics

3.1 Event-based risk analysis

In an event-based analysis, earthquake scenarios are generated from seismic sources, magnitudes, locations, recurrence models, and ground-motion models. For each event, spatially varying intensity measures are estimated, assets are exposed, damage is sampled or calculated, and consequences are aggregated.

For event `e`, a portfolio loss can be written as

L_e = Σ_i L_{i,e},

where `L_{i,e}` is the loss associated with asset or component `i`. Correlation matters: a single earthquake can damage many assets simultaneously, so summing individual expected losses without modelling common event effects can underestimate portfolio uncertainty.

3.2 Expected annual loss

Expected annual loss (EAL) is the long-term mean loss per year. If discrete earthquake scenarios have annual occurrence rates `λ_e` and conditional mean losses `E[L | e]`, then

EAL = Σ_e λ_e E[L | e].

The units are currency per year when `L` is monetary. The equation assumes that the scenario set and occurrence rates adequately represent the hazard model and that the relevant loss categories are consistently defined. If event probabilities rather than rates are used over a finite time horizon, the calculation must be reformulated accordingly.

EAL is useful for comparing portfolios and mitigation options, but it is not a forecast of next year’s loss. It also does not describe extreme tail risk. Two portfolios can have the same EAL while one has a much greater probability of catastrophic loss.

3.3 Exceedance probability and loss curves

A loss exceedance probability curve gives the probability that loss exceeds a specified amount within a time period. The annual probability of exceedance can be converted to an approximate probability over `T` years as

P_T(L > l) = 1 − exp[−ν(l)T],

where `ν(l)` is the annual exceedance rate for loss level `l`. This Poisson-process expression assumes independent events with a constant rate over the period. It may be inappropriate where earthquake clustering, time-dependent recurrence, or changing exposure is important.

Common decision metrics include:

• Average annual loss: expected long-term loss.
• Probable maximum loss: a loss associated with a selected exceedance probability and time basis; terminology must be defined because practice varies.
• Value at risk: a loss quantile at a specified probability, which does not describe losses beyond that quantile.
• Tail value at risk or expected shortfall: the mean loss conditional on exceeding a selected quantile.
• Casualty risk: expected or high-percentile fatalities and injuries.
• Downtime risk: probability and duration of loss of function.

No single metric is sufficient. Financially modest but life-threatening risks may require priority, while low-frequency catastrophic losses may dominate resilience planning despite a modest EAL.

4. Decision Analysis and Investment Prioritisation

4.1 From risk estimates to decisions

Risk analysis becomes useful only when linked to decisions. A mitigation option may be evaluated by comparing its costs with avoided losses and improved performance. Alternatives include structural retrofit, nonstructural restraint, equipment anchorage, utility redundancy, emergency training, stockpiling, insurance, mutual aid, land-use controls, and replacement or relocation.

A basic benefit–cost ratio is

BCR = PV(Benefits) / PV(Costs),

where `PV` denotes present value over the analysis period. Benefits may include avoided repair costs, reduced casualties, reduced downtime, and avoided emergency expenditure. A BCR greater than one indicates that monetised benefits exceed costs under the selected assumptions; it does not automatically establish that the option is equitable, feasible, or ethically preferred.

For a future cash flow `X_t` occurring at year `t`, its present value under a constant real discount rate `r` is

PV(X_t) = X_t/(1+r)^t.

If benefits and costs occur over several years, their present values are summed. The discount rate should be stated clearly, and nominal cash flows should not be discounted using a real rate without adjustment for inflation.

4.2 Life-cycle cost and retrofit comparison

For an intervention with initial cost `C_0`, recurring maintenance cost `M_t`, expected earthquake loss `E[L_t]`, and residual value `S_T` at year `T`, a simplified life-cycle cost is

LCC = C_0 + Σ_{t=1}^{T} (M_t + E[L_t])/(1+r)^t − S_T/(1+r)^T.

This expression assumes that all terms are measured on a consistent basis and that expected losses can reasonably be represented by annual values. For seismic risk, event losses are often highly skewed, so a more complete analysis may use event-based simulations and scenario-dependent discounting. The residual value term is included only when it is meaningful and consistently estimated.

The incremental value of mitigation can be represented by the present value of avoided losses minus the intervention cost. Uncertainty should be tested through sensitivity analysis on hazard, fragility, cost, discount rate, occupancy, and recovery assumptions.

4.3 Multi-criteria decision analysis

Monetary analysis alone is insufficient where life safety, equity, heritage, political legitimacy, or service continuity are central. Multi-criteria decision analysis (MCDA) can compare alternatives against criteria such as:

• expected fatalities and injuries;
• reduction in EAL;
• reduction in critical-service downtime;
• implementation cost;
• technical feasibility;
• distributional fairness;
• environmental impact;
• compatibility with future adaptation;
• institutional capacity.

A weighted-sum model may be written as

S_j = Σ_k w_k z_{jk},

where `S_j` is the score of option `j`, `w_k` is the weight assigned to criterion `k`, and `z_{jk}` is the normalised performance of option `j` on that criterion. Weights must be elicited transparently, and sensitivity analysis should test whether rankings change when weights vary. MCDA is a decision-support method, not a substitute for engineering judgement or public accountability.

5. Business Continuity and Recovery Analytics

5.1 Functional performance

A resilient system is assessed by its ability to deliver required functions over time. A generic functionality curve can be represented by `Q(t)`, where `Q=1` denotes full pre-event function and `Q=0` denotes complete loss. A simple functionality loss measure over a recovery horizon `T` is

F_loss = ∫_0^T [1 − Q(t)] dt.

The units are function-time, such as service-days. This measure assumes that functionality can be represented on a meaningful scale. For essential services, a weighted or threshold-based measure may be more appropriate than a simple average.

Recovery is not necessarily monotonic. Temporary repairs may produce an initial increase followed by decline, and restoration of one component may be ineffective if another dependent component remains unavailable. Recovery targets should therefore specify minimum service levels, maximum tolerable downtime, recovery time objectives, and recovery point objectives for information systems.

5.2 Business continuity planning

Business continuity planning translates risk information into pre-event and post-event arrangements. It should identify:

1. critical functions and acceptable service degradation;

2. dependencies on buildings, staff, utilities, suppliers, data, and transport;

3. single points of failure and common-cause vulnerabilities;

4. alternative facilities, suppliers, routes, and communication channels;

5. authority for emergency decisions and financial commitments;

6. procedures for rapid damage assessment and re-entry;

7. training, exercises, documentation, and periodic review.

A continuity plan that assumes utilities will be restored promptly may fail if regional damage affects the same suppliers, roads, fuel systems, and workforce. Interdependency analysis is consequently essential.

5.3 Recovery governance

Recovery has technical, financial, legal, social, and political dimensions. Early decisions about demolition, temporary accommodation, school reopening, hospital transfers, and infrastructure prioritisation can create long-term consequences. Recovery frameworks should distinguish immediate response, short-term restoration, intermediate recovery, and long-term reconstruction, while recognising that these phases overlap.

“Build back better” should not mean simply rebuilding the same exposure with marginally stronger materials. It may require relocation, revised land-use controls, improved access, social protection, livelihood restoration, and institutional reform. However, relocation can also cause cultural loss, livelihood disruption, or inequitable displacement; it must therefore be participatory and evidence-based.

6. Portfolio and Community Resilience

6.1 Portfolio effects

Risk managers often control a portfolio rather than a single building. Portfolio analysis should consider geographic concentration, common design weaknesses, shared utilities, correlated ground motion, common suppliers, and simultaneous demand for emergency resources. Diversification can reduce risk when assets are not exposed to the same event, but geographic dispersion may increase operating cost or complicate management.

A portfolio priority index should not rely only on asset value. A modestly valued bridge may be more consequential than a costly office if its failure isolates hospitals or prevents emergency access. Criticality is a function of network position, substitutability, service demand, and recovery consequences.

6.2 Infrastructure networks

Water, electricity, transport, telecommunications, fuel, health, and emergency services are interdependent networks. A power outage can disable pumps and communications; a road closure can prevent repair crews from reaching substations; water loss can limit firefighting. Network models should therefore evaluate both component fragility and system functionality.

Useful system-level indicators include connectivity, maximum flow, service coverage, unmet demand, restoration time, and the number of disconnected critical facilities. Component strengthening should be prioritised where it produces meaningful system-level benefit, not simply where component vulnerability is highest.

6.3 Community resilience and equity

Communities differ in age distribution, income, disability, housing tenure, language, transport access, trust in institutions, and ability to evacuate or recover. These characteristics influence both exposure and coping capacity. A technically optimal programme that excludes renters, informal settlements, remote communities, or people with disabilities may increase inequality.

Equity analysis should examine who receives protection, who pays, who bears residual risk, and who participates in decisions. Public communication should explain uncertainty without creating false reassurance. Risk maps and warnings must be accessible, culturally appropriate, and accompanied by feasible protective actions.

7. Illustrative Case: Prioritising a Municipal Hospital Portfolio

Background

The following is a fictional illustrative case. A municipality operates three hospitals: Hospital A, a modern reinforced-concrete facility; Hospital B, an older framed building with vulnerable nonstructural equipment; and Hospital C, a smaller facility located near a major fault and serving a low-income district. The municipality has funding for only one major intervention during the next budget cycle.

Problem

A preliminary assessment shows that Hospital A has the highest replacement value, Hospital B has the largest estimated annual financial loss because of equipment damage and downtime, and Hospital C has the greatest expected casualty consequence because of high occupancy, limited alternatives, and difficult access. A simple ranking by asset value would select Hospital A, while a ranking by EAL would select Hospital B.

Analysis

The engineering team develops building and equipment fragility functions and simulates earthquake scenarios. It estimates:

• structural and nonstructural damage;
• emergency department capacity;
• utility dependence;
• staff access and transport disruption;
• patient evacuation requirements;
• recovery time under alternative interventions.

Three options are considered for Hospital C: structural retrofit, comprehensive nonstructural and equipment anchorage, and a continuity package involving backup power, water storage, alternate communications, mutual-aid agreements, and pre-arranged patient transfer. The analysis shows that structural retrofit most reduces collapse probability, anchorage most reduces short-term functional loss, and the continuity package most reduces expected service downtime for moderate earthquakes. A combined programme is desirable but exceeds the available budget.

Decision approach

The municipality uses MCDA with life safety, emergency-service continuity, reduction in downtime, equity, cost, and feasibility as criteria. Stakeholders include hospital managers, emergency services, clinicians, disability representatives, community organisations, and finance officials. Weights are varied in sensitivity analysis rather than treated as fixed truth.

The selected first-stage programme combines targeted structural strengthening at identified weak elements, anchorage of critical equipment, protected emergency power and water systems, and a tested patient-transfer agreement. Hospital A receives a lower-cost inspection and maintenance programme; Hospital B receives equipment anchorage and continuity improvements funded in the following cycle.

Outcome

The chosen portfolio does not minimise one metric. Instead, it substantially reduces the probability that the municipality will lose a critical hospital serving a vulnerable population, while improving function across several earthquake scenarios. The decision is documented with assumptions, uncertainty ranges, residual risks, and a review date.

Learning points

1. Asset value is not the same as societal criticality.

2. Nonstructural damage can dominate hospital downtime.

3. Life safety and service continuity may require different interventions.

4. Portfolio decisions must account for dependencies and alternatives.

5. Transparent stakeholder participation improves legitimacy and reveals practical constraints.

6. A staged programme can outperform a single undifferentiated retrofit when resources are limited.

8. Uncertainty, Limitations, and Governance

8.1 Sources of uncertainty

Seismic risk estimates contain aleatory variability, representing natural randomness, and epistemic uncertainty, representing incomplete knowledge. Important sources include earthquake recurrence, ground-motion prediction, site amplification, structural capacity, occupancy, damage-state definitions, repair prices, recovery rates, and future development.

Uncertainty should be represented through ranges, distributions, scenario comparisons, or logic-tree methods where justified. Reporting a single precise loss number can create unwarranted confidence. Model validation should compare predictions with observed earthquake performance where data are available.

8.2 Model limitations

Fragility curves may not transfer reliably between regions with different construction practices. Loss models may omit informal economies, unpaid care, cultural heritage, or long-term health effects. EAL can hide catastrophic tail risk. Discounting can make benefits occurring decades later appear small, raising ethical concerns for intergenerational safety. MCDA results can be sensitive to subjective weights. Network models can fail if dependencies or recovery constraints are simplified excessively.

8.3 Governance and ethical responsibility

Risk decisions involve public values, not merely technical optimisation. Engineers and institutions should:

• disclose assumptions, uncertainty, and conflicts of interest;
• avoid presenting probabilistic results as guarantees;
• protect sensitive infrastructure information while enabling public accountability;
• include affected communities in defining acceptable risk;
• consider vulnerable groups and intergenerational effects;
• maintain professional competence and independent review;
• monitor whether mitigation benefits are actually realised.

Regulatory minimums are a baseline, not necessarily an acceptable societal risk target. Existing buildings, critical facilities, heritage structures, and informal settlements may require tailored strategies that combine regulation, incentives, technical assistance, financing, and enforcement.

9. Technology, Data, and Contemporary Practice

Geospatial information systems support exposure inventories, risk mapping, evacuation analysis, and portfolio prioritisation. Remote sensing, building information models, sensors, drones, and mobile reporting can accelerate damage assessment and recovery monitoring. Digital twins may connect asset condition, network dependencies, hazard scenarios, and intervention planning, but their value depends on data quality and governance.

Artificial intelligence can assist image classification, anomaly detection, demand forecasting, natural-language extraction from inspection records, and scenario generation. It should not replace engineering review. Bias in training data, poor transferability, opaque models, cyber-security threats, and automation errors are significant concerns. AI-supported decisions should maintain traceability, human oversight, version control, uncertainty reporting, and a mechanism for challenge or correction.

Contemporary resilience practice increasingly emphasises compound and cascading hazards, climate-related interactions, supply-chain disruption, cyber-physical systems, and recovery equity. Climate change may alter rainfall, landslide, flood, wildfire, and infrastructure deterioration conditions that interact with earthquake consequences, even though tectonic hazard itself is not generally controlled by short-term climate variation. Planning should therefore use multi-hazard and systems approaches without conflating distinct hazard mechanisms.

10. Best-Practice Workflow

A defensible seismic risk and resilience programme can follow these steps:

1. Define objectives, decision horizon, stakeholders, and acceptable service levels.

2. Establish the hazard model, including relevant secondary hazards and uncertainty.

3. Build and quality-check an exposure inventory.

4. Select appropriate fragility and vulnerability models.

5. Estimate event losses, annualised losses, casualties, downtime, and tail risk.

6. Model dependencies, network effects, and recovery constraints.

7. Identify intervention alternatives across engineering, planning, financial, and organisational measures.

8. Compare alternatives using life-cycle cost, MCDA, and distributional analysis.

9. Test sensitivity to key assumptions and identify value-of-information priorities.

10. Select staged actions with responsibilities, funding, performance indicators, and review dates.

11. Exercise emergency and continuity plans.

12. Update models and plans using new data, inspections, exercises, and earthquake experience.

Performance indicators should include more than expenditure or number of buildings strengthened. They may include reduction in collapse probability, expected downtime, restoration time, critical-service coverage, participation of vulnerable groups, exercise performance, and progress in closing identified risk gaps.

Chapter Summary

Seismic risk management integrates hazard, exposure, vulnerability, and consequences. Fragility functions estimate the probability of reaching damage states at specified intensity levels, while vulnerability functions translate damage into economic, functional, or social loss. Event-based analysis supports portfolio loss estimation, expected annual loss, exceedance curves, and tail-risk assessment.

Resilience extends beyond structural strength to include robustness, redundancy, resourcefulness, rapid recovery, and adaptation. Business continuity requires analysis of critical functions, dependencies, alternative resources, recovery objectives, and governance. Infrastructure and community decisions must account for interdependencies, equity, and the different meanings of criticality.

Life-cycle cost, benefit–cost analysis, and multi-criteria decision analysis help prioritise limited investments, but results depend on assumptions, discount rates, data quality, and value judgements. Ethical practice requires transparency, stakeholder participation, protection of vulnerable groups, and clear communication of uncertainty. Effective programmes combine engineering retrofit with nonstructural protection, land-use planning, preparedness, continuity arrangements, financial mechanisms, and learning-oriented recovery.

Key Terms

Hazard: The probability and intensity of an earthquake or related phenomenon at a location.

Exposure: People, assets, services, and activities located in areas subject to hazard.

Vulnerability: The propensity of an exposed element to suffer damage or loss.

Seismic risk: Probabilistic potential for adverse consequences from earthquake hazard.

Resilience: Capacity to withstand, recover from, adapt to, and learn from disruption.

Fragility function: Probability of reaching or exceeding a damage state conditional on intensity.

Vulnerability function: Relationship between hazard intensity and expected consequence or loss.

Damage state: Defined level of physical or functional damage to an asset.

Mean damage ratio: Expected loss divided by a specified asset value.

Expected annual loss: Long-term average loss per year.

Loss exceedance curve: Relationship between loss magnitude and probability or rate of exceedance.

Tail risk: Potential for rare but exceptionally severe losses.

Criticality: Importance of an asset or function to people, systems, or recovery.

Business continuity: Arrangements that maintain or restore essential organisational functions.

Recovery time objective: Target time for restoring a function to a specified level.

Life-cycle cost: Present value of acquisition, operation, maintenance, loss, and disposal costs.

Multi-criteria decision analysis: Structured comparison of alternatives using multiple criteria.

Interdependency: Relationship in which the operation of one system affects another.

Residual risk: Risk remaining after mitigation measures are implemented.

Aleatory variability: Natural randomness in hazard or response.

Epistemic uncertainty: Uncertainty caused by incomplete knowledge.

Review Questions

Short Answer Questions

1. Distinguish hazard, exposure, vulnerability, and consequence in seismic risk analysis.

2. What information does a fragility function provide?

3. Why can two portfolios with the same expected annual loss have different risk profiles?

4. Identify four dimensions of resilience.

5. Why is nonstructural damage important in hospital continuity planning?

Long Answer Questions

1. Explain how fragility functions can be converted into expected repair loss using damage-state probabilities and damage ratios.

2. Discuss the strengths and limitations of expected annual loss, probable maximum loss, and loss exceedance curves.

3. Explain how life-cycle cost analysis can support comparison of seismic retrofit alternatives.

4. Discuss the role of infrastructure interdependencies in community disaster recovery.

5. Critically evaluate the ethical and governance responsibilities associated with probabilistic seismic risk models.

Analytical / Application Questions

1. A municipality must choose between strengthening a high-value office building and improving the continuity of a lower-value emergency facility. Develop a decision framework that considers life safety, downtime, criticality, equity, uncertainty, and life-cycle cost.

2. Design a simplified risk study for a hospital portfolio. Specify the hazard, exposure, fragility, loss, recovery, and decision models required, and identify at least five sensitivity analyses.

3. An organisation has limited funds for structural retrofit, equipment anchorage, backup utilities, and emergency training. Explain how MCDA and scenario-based analysis could be used to select a staged investment programme.

Multiple Choice Questions with Answers

Question 1

What does a seismic fragility function most directly estimate?

A. The annual repair budget of a building

B. The probability of reaching or exceeding a damage state given an intensity measure

C. The number of occupants in a building

D. The duration of post-earthquake reconstruction

Correct Answer: B. The probability of reaching or exceeding a damage state given an intensity measure

Question 2

In the lognormal fragility expression, what does θ_ds represent?

A. The median intensity associated with damage state ds

B. The annual earthquake occurrence rate

C. The replacement value of the asset

D. The standard normal density

Correct Answer: A. The median intensity associated with damage state ds

Question 3

Expected annual loss is best described as:

A. The loss guaranteed to occur next year

B. The maximum possible earthquake loss

C. The long-term mean loss per year

D. The loss from the largest historical earthquake

Correct Answer: C. The long-term mean loss per year

Question 4

Which measure is most directly concerned with restoration of organisational function over time?

A. Mean damage ratio

B. Functionality curve

C. Ground-motion intensity

D. Asset replacement value

Correct Answer: B. Functionality curve

Question 5

A benefit–cost ratio greater than one indicates that, under the stated assumptions:

A. The project has no residual risk

B. Monetised present-value benefits exceed monetised present-value costs

C. The project is automatically equitable

D. The intervention will prevent every earthquake loss

Correct Answer: B. Monetised present-value benefits exceed monetised present-value costs

Question 6

Which is an example of redundancy in a resilient system?

A. Using a single larger transformer

B. Removing emergency supplies

C. Providing an alternative water-supply route

D. Reducing inspection frequency

Correct Answer: C. Providing an alternative water-supply route

Question 7

Why can asset value alone be a poor basis for seismic investment prioritisation?

A. Asset value is unrelated to construction cost

B. Low-value assets can have high criticality or network consequences

C. High-value assets never suffer damage

D. Asset value cannot be expressed in currency

Correct Answer: B. Low-value assets can have high criticality or network consequences

Question 8

The expression P_T(L > l) = 1 − exp[−ν(l)T] assumes, among other things, that events follow:

A. A constant-rate Poisson process

B. A deterministic annual cycle

C. A uniform damage distribution

D. A zero-loss model

Correct Answer: A. A constant-rate Poisson process

Question 9

What is the main purpose of sensitivity analysis in seismic decision-making?

A. To eliminate all uncertainty

B. To identify whether conclusions change under plausible assumptions

C. To replace stakeholder participation

D. To guarantee a benefit–cost ratio above one

Correct Answer: B. To identify whether conclusions change under plausible assumptions

Question 10

Which statement about artificial intelligence in seismic risk practice is most appropriate?

A. AI output never requires engineering review

B. AI removes the need for uncertainty reporting

C. AI can assist analysis but requires validation, oversight, and governance

D. AI is useful only for structural design calculations

Correct Answer: C. AI can assist analysis but requires validation, oversight, and governance