Supply Chain Risk, Resilience and Sustainability

Introduction

Modern supply chains are networks of suppliers, manufacturers, logistics providers, digital platforms, financial institutions, regulators and customers. Their performance depends not only on cost and speed but also on their ability to withstand disruption and operate within environmental and social limits. A low-cost supply chain may be highly vulnerable if it depends on one supplier, one transport corridor, one energy source or one information system.

Supply chain risk is the possibility that an event or condition will prevent the chain from achieving objectives such as delivery reliability, quality, profitability, compliance or customer service. Resilience is the capability to prepare for disruption, absorb its effects, recover and adapt. Sustainability concerns the long-term environmental, social and economic consequences of supply-chain decisions.

These concepts are strategically connected. Climate change can increase physical disruption; geopolitical conflict can restrict materials; cyberattacks can interrupt production; and poor labour practices can create legal, reputational and operational consequences. Business analytics supports decisions by identifying exposure, estimating consequences, comparing scenarios and monitoring leading indicators.

Learning Objectives

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

1. Classify operational, supplier, geopolitical, cyber and climate risks.
2. Distinguish risk exposure, vulnerability, resilience and sustainability.
3. Apply risk registers, risk matrices, FMEA, stress testing and scenario planning.
4. Evaluate mitigation strategies such as redundancy, postponement, visibility and collaboration.
5. Calculate and interpret selected risk, carbon, circularity and responsible-operations metrics.
6. Assess trade-offs among cost, service, resilience, environmental performance and social responsibility.
7. Explain governance, ethical and regulatory requirements for supply-chain risk management.

1. The Nature of Supply-Chain Risk

1.1 Risk, uncertainty and disruption

Risk generally refers to uncertain events for which probabilities or probability ranges can be estimated. Uncertainty is broader: information may be insufficient to estimate reliable probabilities. A disruption is the realization of a risk that materially reduces normal supply-chain performance.

A useful distinction is:

• Hazard: a possible source of harm, such as flooding or malware.
• Exposure: the extent to which a supply-chain node or flow is in the path of the hazard.
• Vulnerability: the degree to which exposure produces harm.
• Consequence: the operational, financial, environmental or social impact.
• Resilience: the ability to resist, respond, recover and learn.

Risk is multidimensional. A supplier may have a low probability of failure but a severe consequence because it provides a unique component. Conversely, a frequent minor delay may have a high annual cost even if each individual incident is small.

1.2 Risk categories

Risks can also be independent, dependent or systemic. A local machine failure may be independent. A port closure and supplier delay may be dependent because both result from the same storm. A pandemic, regional war or financial crisis may be systemic, affecting many nodes simultaneously and invalidating assumptions based on historical data.

1.3 Network perspective

A supply chain is better represented as a network than as a linear sequence. A focal firm may have thousands of suppliers, including hidden tier-two and tier-three sources. Risk can propagate through shared suppliers, shared logistics corridors, common software providers, financial links or synchronized demand.

Network analytics can identify:

• highly connected suppliers;
• single points of failure;
• substitute paths and alternate sources;
• geographic concentration;
• common-cause dependencies; and
• nodes whose failure would disconnect important customers.

Visibility is not the same as resilience. Knowing that a supplier is exposed does not reduce the exposure unless the organization can act on the information.

2. Risk Assessment Frameworks

2.1 Risk register and risk matrix

A risk register records the risk description, causes, affected processes, owner, existing controls, indicators, response and review date. A simple risk score is often calculated as:

where P is the likelihood score and I is the impact score. If both are rated from 1 to 5, R ranges from 1 to 25. This is an ordinal prioritization device, not a true probability-weighted monetary loss. Scores should not be compared as if a score of 10 were exactly twice a score of 5.

A risk matrix is useful for screening, communication and ownership. Its limitations include subjective ratings, coarse categories, correlation between risks and the possibility that low-frequency catastrophic risks receive insufficient attention.

Where probabilities and monetary consequences are credible, expected loss may be estimated as:

where pᵢ is the probability of scenario i, Cᵢ is its consequence in monetary units, and the scenarios are mutually exclusive and collectively exhaustive for the analysis. This measure is appropriate for repeated decisions under risk, but it does not fully represent liquidity constraints, catastrophic aversion, reputation or irreversible harm.

2.2 Failure Mode and Effects Analysis

FMEA examines how a process, product or service can fail. For each failure mode, the team identifies its effect, cause, existing controls and ratings for:

• Severity (S): seriousness of the effect;
• Occurrence (O): frequency or likelihood of the cause;
• Detection (D): likelihood that the failure will escape current controls, usually with a higher score meaning poorer detectability.

The traditional risk priority number is:

The ratings are normally ordinal, often 1 to 10. RPN is a prioritization aid rather than a probability model. Two failure modes can have the same RPN but very different profiles; therefore teams should also review severity separately, especially for safety, legal and environmental consequences.

Illustrative FMEA example. A chilled-food distributor identifies a refrigeration failure. Suppose S = 8, O = 4 and D = 6. Then:

A proposed temperature alarm and backup power may reduce occurrence to 2 and improve detection to 2, while severity remains 8:

The residual risk is lower, but the firm must verify that alarms are tested, backup power is maintained and employees respond correctly.

2.3 Bow-tie analysis

Bow-tie analysis connects causes, a central event and consequences. Preventive barriers act before the central event; mitigative barriers act afterward. For example, weak supplier cybersecurity and stolen credentials may lead to unauthorized system access. Multifactor authentication and least-privilege access are preventive barriers; network segmentation, backups and incident response are mitigative barriers.

The method encourages organizations to examine whether controls are merely documented or actually effective. Each barrier should have an owner, performance standard, testing schedule and escalation rule.

3. Measuring Resilience

3.1 Resilience dimensions

Resilience is not simply holding extra inventory. It includes:

1. Anticipation: sensing threats and preparing contingencies.
2. Absorption: maintaining essential service during disruption.
3. Adaptation: changing sourcing, processes or products.
4. Recovery: restoring performance within an acceptable time.
5. Learning: improving systems after incidents and near misses.

Important measures include time to detect, time to respond, time to recover, service level during disruption, backlog clearance time and financial loss. A resilience dashboard should combine leading indicators, such as overdue supplier audits and untested backups, with lagging indicators, such as disruption days and customer claims.

A simple service-loss measure over a disruption horizon is:

where Dₜ is demand in period t, Sₜ is fulfilled supply, and (x)⁺ = max(x, 0). Units are products or orders. This measure captures unmet demand but not its value, urgency or substitution effects.

3.2 Resilience strategies

Redundancy uses extra capacity, inventory, suppliers or transport options. It improves continuity but raises cost and may increase environmental impact.

Flexibility allows rapid changes in volume, product mix, routing or sourcing. Examples include adaptable manufacturing, postponement and modular product design.

Visibility and collaboration improve detection and coordinated response. Sharing forecasts, inventory and capacity information can reduce the bullwhip effect and accelerate recovery, but requires trust, data standards and cybersecurity.

Regionalization and diversification reduce geographic concentration. However, simply adding suppliers in the same country or dependent on the same sub-tier source may create an illusion of diversification.

Strategic inventory protects critical items when replenishment is slow or substitution is difficult. Inventory should be targeted using criticality, lead-time variability and service requirements rather than applied uniformly.

Design for resilience uses common components, repairability, postponement and alternative specifications. Design decisions often create more resilience than emergency purchasing can provide.

Resilience has trade-offs. Dual sourcing may reduce supplier risk but increase qualification, tooling and coordination costs. Local sourcing may shorten transport distances but may not reduce climate or financial risk. A robust decision evaluates total cost, service, risk reduction, emissions and social effects.

4. Stress Testing and Scenario Planning

4.1 Stress testing

Stress testing asks how a supply chain performs under severe but plausible conditions. Scenarios can alter demand, capacity, lead time, transportation availability, exchange rates, energy prices or recovery time.

A basic stress-test process is:

1. Define the decision, time horizon and performance measures.
2. Identify critical assumptions and dependencies.
3. Construct plausible adverse scenarios.
4. Quantify operational and financial effects.
5. Test existing controls and contingency plans.
6. Compare actions, costs and residual exposure.
7. Establish triggers and rehearse the response.

Stress tests should avoid false precision. Their purpose is not to predict the future exactly but to expose fragile assumptions and identify actions that work across several futures.

4.2 Scenario planning

Scenario planning develops coherent narratives rather than isolated shocks. For example, a firm might consider:

• Fragmented trade: tariffs, export controls and duplicated regional networks;
• Accelerated decarbonization: strict carbon reporting, renewable-energy investment and product redesign;
• Climate volatility: repeated extreme weather and infrastructure disruption; and
• Digital dependency: efficient but highly interconnected operations with significant cyber exposure.

A scenario should specify causal logic, not merely a headline. It should explain how external developments affect suppliers, capacity, logistics, demand, regulation and customer expectations.

4.3 Robustness and real options

A robust strategy performs acceptably across many scenarios. A flexible strategy may have a higher immediate cost but preserve future choices. Examples include maintaining a qualified secondary supplier, using modular equipment or negotiating transport contracts with flexible volumes. These actions resemble real options: the organization pays for the ability to adapt later, without committing fully today.

5. Sustainability and Responsible Operations

5.1 Sustainability dimensions

Supply-chain sustainability combines:

• Environmental performance: greenhouse-gas emissions, energy, water, waste, biodiversity and pollution;
• Social performance: health and safety, wages, working hours, human rights, inclusion and community effects; and
• Economic viability: profitability, decent employment, continuity and long-term value.

A supply chain can be operationally resilient but socially unacceptable. For example, emergency overtime may restore delivery while increasing fatigue and injury risk. Responsible operations therefore treat workers and communities as stakeholders, not merely inputs.

5.2 Carbon accounting

A basic activity-based emissions calculation is:

where Aᵢ is activity data, such as litres of fuel, kilowatt-hours or tonne-kilometres, and EFᵢ is the corresponding emission factor, such as kg CO₂e per litre or kg CO₂e per tonne-kilometre. The result E is usually expressed in kg or tonnes of CO₂e. Activity boundaries, emission factors, allocation rules and data quality must be documented.

For freight, a simplified calculation is:

where W is transported mass in tonnes, d is distance in kilometres and EF_f is the factor in kg CO₂e per tonne-kilometre. This approximation assumes the factor is appropriate for the mode, load, fuel and reporting boundary. Empty running, temperature control and shipment-specific differences may require more detailed data.

Worked example. A shipment weighs 12 tonnes and travels 800 km. If the applicable factor is 0.09 kg CO₂e per tonne-kilometre:

If consolidation reduces the mass-distance activity by 25%, the estimated emissions become:

The reduction is 216 kg CO₂e, assuming the same emission factor and no additional handling or storage emissions.

Carbon reporting commonly distinguishes direct emissions from owned or controlled sources, purchased-energy emissions and value-chain emissions. Organizations must avoid double counting when aggregating across partners and must state whether figures are measured, estimated or based on industry averages.

5.3 Circularity

A linear model extracts resources, makes products, uses them and discards them. A circular model seeks to eliminate waste, extend product life and keep materials and products in use through repair, reuse, remanufacturing and recycling.

A simple material circularity indicator may be expressed conceptually as:

This expression is only a simplified teaching metric; formal methods may weight durability, utility, losses and product lifetime differently. The numerator and denominator must be defined consistently. Recycling is not automatically preferable if it reduces quality, consumes large quantities of energy or prevents higher-value reuse.

Useful operational measures include return rate, repair rate, refurbishment yield, recycled-content share, landfill diversion and product-life extension. These indicators should be connected to quality, safety and customer acceptance.

5.4 Social and responsible-operations metrics

Possible metrics include lost-time injury frequency, worker turnover, grievance closure time, living-wage coverage, audit corrective-action closure, forced-labour screening coverage and supplier diversity. Metrics can be misleading when firms focus on audit counts rather than actual conditions. Worker interviews, anonymous reporting channels, remediation evidence and independent verification are important complements.

6. Governance, Ethics and Technology

Risk management requires clear accountability. A practical governance structure assigns:

• operational owners who manage controls;
• risk specialists who set methods and challenge assumptions;
• internal audit who provides independent assurance; and
• senior leadership or the board who approve risk appetite and major trade-offs.

Risk appetite should specify unacceptable outcomes, tolerable service loss, maximum concentration, recovery expectations and escalation thresholds. Contracts should address continuity plans, notification duties, audit rights, data protection, subcontracting and remediation.

Ethical issues include surveillance of workers, unequal distribution of disruption costs, supplier power asymmetry, privacy, misleading sustainability claims and the use of low-cost regions with weak labour protections. Analytics should not convert ethically unacceptable outcomes into acceptable averages. Safety, human rights and legal compliance may be constraints rather than variables to be traded away.

Technology can strengthen risk management through control-tower platforms, event monitoring, digital twins, supplier portals, satellite data, predictive maintenance and machine-learning forecasts. However, AI outputs depend on data quality and model assumptions. Important controls include data lineage, access management, human review, bias testing, explainability appropriate to the decision, model monitoring and fallback procedures. A model that predicts supplier distress but cannot explain its data sources may be unsuitable for automatic termination decisions.

Cyber resilience requires asset inventories, identity controls, multifactor authentication, network segmentation, secure backups, patch management, supplier-access governance, incident exercises and tested recovery objectives. Cybersecurity is a supply-chain responsibility because attackers may enter through logistics, software or supplier connections.

7. Illustrative Case: Northstar Medical Devices

Background

Northstar Medical Devices produces infusion-pump assemblies. Its final plant is located in one country, but a specialized microcontroller comes from a single approved supplier in another region. The supplier itself depends on a second-tier semiconductor foundry. Northstar has low inventory because the component is expensive and has a long shelf life. The company also faces new customer requirements for carbon disclosure and responsible sourcing.

Problem

A geopolitical dispute creates a credible risk of export restrictions. At the same time, the supplier reports a cyber incident and a nearby region experiences drought that threatens electricity reliability. Northstar must decide whether to qualify a second source, increase inventory, redesign the controller or accept the existing exposure.

Analysis

The risk team maps the network and identifies three common dependencies: the single approved design, the regional semiconductor foundry and a shared logistics corridor. A risk matrix ranks export restriction as high impact and medium likelihood. FMEA identifies controller unavailability as a high-severity failure mode. A stress test assumes a 12-week supply interruption and a 30% demand increase due to a hospital procurement program.

The team estimates that normal weekly demand is 1,000 units. Existing inventory covers four weeks, while an alternate supplier would require eight weeks for qualification. Under the disruption scenario, the first four weeks of demand consume 4,000 units. The increased demand is 1,300 units per week, so the remaining eight weeks require 10,400 units. Total demand during the 12-week horizon is therefore 14,400 units. Existing inventory covers only 4,000 units, leaving 10,400 units exposed unless supply is restored or demand is rationed. This is a service-risk calculation, not a forecast that the disruption will occur.

Approach and solution

Northstar adopts a portfolio of actions:

1. qualify a second supplier using a technically compatible design;
2. hold targeted strategic inventory for hospitals with critical service needs;
3. redesign the controller interface to permit future substitution;
4. require the primary supplier to disclose relevant sub-tier dependencies;
5. implement cyber-access controls and a tested offline recovery plan;
6. use a supplier scorecard covering quality, continuity, emissions and labour practices;
7. establish trigger points for inventory replenishment, regulatory alerts and logistics rerouting; and
8. include worker safety and responsible sourcing clauses in supplier contracts.

The company does not simply move all production to a new country. It evaluates the alternate source for technical quality, labour conditions, carbon intensity, water exposure, financial stability and geopolitical risk.

Outcome

In the illustrative scenario, qualification is completed before the disruption occurs. Northstar accepts a modest unit-cost increase and additional supplier-management work but reduces single-source exposure. During a later transport interruption, the firm uses the qualified supplier and strategic inventory to protect critical hospital orders. It also discovers that its carbon data had been incomplete because supplier electricity and expedited freight were excluded; the data-governance process is therefore improved.

Learning points

The case demonstrates that resilience is a system design problem. Inventory alone would not solve a shared sub-tier dependency. Supplier diversification without compatible specifications would not be actionable. Sustainability, cyber controls and continuity planning must be integrated because a decision that reduces one risk may increase another.

8. Implementation Road Map and Best Practices

A practical implementation sequence is:

1. Define objectives: service, safety, financial, environmental and social outcomes.
2. Map the network: include critical tier-two and tier-three dependencies where feasible.
3. Segment risk: classify suppliers and flows by criticality, substitutability, exposure and consequence.
4. Establish data governance: define owners, sources, update frequency, quality ratings and access rights.
5. Assess risk: combine risk registers, FMEA, scenario analysis and quantitative evidence.
6. Select controls: prioritize prevention, detection, response and recovery barriers.
7. Test plans: conduct simulations, cyber exercises, supplier drills and recovery tests.
8. Monitor indicators: use dashboards with thresholds, escalation rules and trend analysis.
9. Review after events: capture near misses, control failures and learning.
10. Report responsibly: disclose methods, boundaries, uncertainty and corrective action.

Best practice is to combine centralized standards with local knowledge. It also requires cross-functional participation from procurement, operations, logistics, finance, information security, sustainability, legal, quality and human resources. Risk management becomes ineffective when it is treated as a yearly document rather than an operating capability.

9. Limitations and Contemporary Developments

Historical data may underrepresent unprecedented events. Supplier ratings can be biased toward large firms with better reporting systems. Carbon factors vary by geography and technology. Audit data may be incomplete or manipulated. Scenario models can create confidence without validity. These limitations should be recorded rather than hidden.

Contemporary developments include climate-adjusted network design, mandatory sustainability disclosures, digital product passports, traceability technologies, renewable-energy procurement, regional manufacturing, low-carbon fuels and greater attention to just transitions. Regulatory requirements differ across jurisdictions, so organizations must monitor applicable reporting, product, labour, data-protection, customs and due-diligence rules.

The strategic question is not whether a supply chain can eliminate all risk. It is whether the organization understands its material exposures, has proportionate controls, can act quickly under stress and creates value without transferring unacceptable harm to workers, communities or future generations.

Chapter Summary

Supply-chain risk arises from uncertain events that can affect service, cost, quality, compliance, people or the environment. Important categories include operational, supplier, logistics, geopolitical, cyber, climate, market and social risks. Effective assessment distinguishes hazard, exposure, vulnerability and consequence and recognizes that risks may propagate through networks.

Risk matrices provide a screening tool, while expected-loss calculations are appropriate when credible probabilities and consequences are available. FMEA prioritizes failure modes through severity, occurrence and detection ratings, but its ordinal RPN should not replace judgment. Stress testing and scenario planning reveal fragile assumptions and support robust, flexible decisions.

Resilience involves anticipation, absorption, adaptation, recovery and learning. Redundancy, flexibility, visibility, collaboration, diversification, strategic inventory and resilient design can improve continuity, but each creates cost and sustainability trade-offs. Sustainability measurement includes activity-based carbon accounting, circularity indicators and social metrics. Data boundaries, quality and uncertainty must be transparent.

Governance assigns accountability, sets risk appetite and tests controls. Cyber resilience, ethical sourcing, human rights and responsible use of AI are integral to operations. The most effective organizations manage risk as a continuous, cross-functional capability rather than as a one-time compliance exercise.

Key Terms

Supply-chain risk: The possibility that uncertainty will prevent supply-chain objectives from being achieved.

Hazard: A source or condition capable of causing harm.

Exposure: The extent to which a supply-chain asset or flow is subject to a hazard.

Vulnerability: The susceptibility of an exposed system to harm.

Disruption: An event that materially reduces normal supply-chain performance.

Resilience: The capability to prepare for, absorb, recover from and adapt to disruption.

Risk register: A structured record of risks, controls, owners and responses.

Risk matrix: A visual or tabular tool that combines likelihood and impact ratings.

Expected loss: The probability-weighted consequence of possible scenarios.

FMEA: Failure Mode and Effects Analysis, a method for examining potential process or product failures.

RPN: Risk Priority Number, traditionally calculated as severity × occurrence × detection.

Stress testing: Evaluation of performance under severe but plausible conditions.

Scenario planning: Development of coherent alternative futures to test strategies.

Redundancy: Additional capacity, inventory, suppliers or routes that provide alternatives.

Carbon footprint: Greenhouse-gas emissions associated with an activity, product, organization or supply chain.

Emission factor: A coefficient that converts activity data into estimated emissions.

Circularity: Keeping products, components and materials in use while minimizing waste.

Responsible operations: Operations that address legal, ethical, social, safety and environmental obligations.

Risk appetite: The amount and type of risk an organization is willing to accept.

Control tower: An integrated capability for monitoring supply-chain events, exceptions and responses.

Review Questions

Short Answer Questions

1. Distinguish hazard, exposure and vulnerability using a supply-chain example.
2. What are the three ratings used in traditional FMEA?
3. Why is supplier diversification not always sufficient to create resilience?
4. State the activity-based emissions formula and define its variables.
5. Identify two leading and two lagging resilience indicators.

Long Answer Questions

1. Explain how operational, geopolitical, cyber and climate risks can interact in a supply-chain network.
2. Compare risk matrices, FMEA, stress testing and scenario planning, including the limitations of each.
3. Discuss the trade-offs between redundancy, efficiency, carbon performance and customer service.
4. Explain how governance, contracts and assurance support responsible supplier risk management.
5. Evaluate the role and limitations of AI, control towers and digital traceability in resilient supply chains.

Analytical / Application Questions

1. A firm transports 20 tonnes over 600 km using an emission factor of 0.08 kg CO₂e per tonne-kilometre. Calculate emissions and explain two assumptions that should be checked.
2. Construct an FMEA for a food, pharmaceutical or electronics supply chain. Select three failure modes, assign ratings, calculate RPNs and recommend controls while discussing severity separately.
3. Design a 12-week stress test for a company dependent on one overseas supplier. Specify scenarios, performance measures, trigger points and a resilience portfolio that also addresses carbon and labour risks.

Multiple Choice Questions with Answers

Question 1

Which statement best distinguishes resilience from redundancy?

A. Resilience is only the amount of inventory held in a warehouse.

B. Redundancy is one possible resilience strategy involving additional alternatives or capacity.

C. Redundancy eliminates all forms of supply-chain uncertainty.

D. Resilience applies only to environmental disruptions.

Correct Answer: B. Redundancy is one possible resilience strategy involving additional alternatives or capacity.

Question 2

A risk has a likelihood score of 4 and an impact score of 5 in a simple risk matrix. What is its screening score?

A. 1

B. 9

C. 20

D. 45

Correct Answer: C. 20

Question 3

In traditional FMEA, a higher detection score generally means that:

A. The failure is easier to detect before reaching the customer.

B. The failure is less severe.

C. The failure is less likely to occur.

D. The failure is more likely to escape existing controls.

Correct Answer: D. The failure is more likely to escape existing controls.

Question 4

A shipment weighs 10 tonnes, travels 500 km, and has an emission factor of 0.1 kg CO₂e per tonne-kilometre. What are its estimated emissions?

A. 50 kg CO₂e

B. 500 kg CO₂e

C. 5,000 kg CO₂e

D. 50,000 kg CO₂e

Correct Answer: B. 500 kg CO₂e

Question 5

Which is a leading indicator of supply-chain resilience?

A. Number of customer claims after a disruption

B. Total revenue lost last year

C. Percentage of critical suppliers with tested continuity plans

D. Number of orders that were late last quarter

Correct Answer: C. Percentage of critical suppliers with tested continuity plans

Question 6

Why should an organization identify tier-two and tier-three suppliers?

A. Hidden sub-tier dependencies can create common points of failure.

B. Only direct suppliers affect supply-chain performance.

C. Sub-tier suppliers are always legally responsible for the focal firm.

D. Mapping sub-tier suppliers automatically reduces procurement cost.

Correct Answer: A. Hidden sub-tier dependencies can create common points of failure.

Question 7

Which control is primarily preventive in a cyber supply-chain bow-tie analysis?

A. Offline recovery from a tested backup

B. Post-incident customer communication

C. Multifactor authentication

D. Emergency product substitution

Correct Answer: C. Multifactor authentication

Question 8

Which statement about an FMEA RPN is most accurate?

A. It is a precise monetary estimate of expected loss.

B. It is an ordinal prioritization aid based on severity, occurrence and detection ratings.

C. It proves that risks with equal RPN have equal consequences.

D. It removes the need for expert judgment.

Correct Answer: B. It is an ordinal prioritization aid based on severity, occurrence and detection ratings.

Question 9

Which practice most directly supports circularity?

A. Increasing single-use packaging without recovery arrangements

B. Designing products for repair, reuse and remanufacture

C. Reducing supplier visibility to protect commercial information

D. Choosing the lowest-priced material regardless of lifetime.

Correct Answer: B. Designing products for repair, reuse and remanufacture

Question 10

What is a central purpose of scenario planning?

A. To predict one exact future with certainty

B. To replace all operational data with narratives

C. To test whether strategies remain effective across coherent alternative futures

D. To guarantee that disruptions will not occur

Correct Answer: C. To test whether strategies remain effective across coherent alternative futures