9. Logistics, Inventory and Capacity Risk Decisions

Introduction and Context

Global supply chains convert time, space and capacity into customer value. Raw materials must arrive when needed, products must move through ports and distribution centres, and transport and production resources must be available at commercially acceptable cost. These conditions are not guaranteed. Weather events, labour disputes, geopolitical tensions, cyber incidents, equipment failures, carrier cancellations and sudden demand changes can interrupt the flow of goods. The resulting problem is not simply delay; it is the interaction between delay, inventory, capacity, service commitments and financial exposure.

Logistics, inventory and capacity decisions are therefore central to supply chain resilience. A firm may respond to disruption by expediting freight, changing ports, holding additional stock, reserving flexible capacity, postponing production or reallocating inventory. Each action reduces one form of exposure while potentially increasing another. Airfreight may protect customer service but destroy margin. Additional inventory may reduce stockout risk but increase obsolescence and working-capital requirements. A second port may create flexibility but add handling complexity and fixed cost.

This chapter develops an integrated approach to such decisions. It explains how to assess transport disruption and port congestion, model lead-time variability, determine inventory buffers, evaluate capacity shortages and compare alternatives using total-cost and optimization perspectives. The emphasis is managerial as well as analytical: robust decisions require accurate data, explicit assumptions, governance and the ability to execute contingency plans.

Learning Objectives

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

1. Explain the relationship between transport disruption, lead-time variability, inventory and capacity risk.

2. Distinguish cycle stock, safety stock, strategic stock and emergency inventory.

3. Calculate and interpret reorder points, safety stock and service-level measures.

4. Evaluate port, carrier and routing alternatives using total landed cost and risk-adjusted analysis.

5. Formulate basic allocation, routing and capacity decisions under disruption.

6. Design contingency-routing and capacity-reservation strategies.

7. Assess the organizational, technological, ethical and governance implications of resilience decisions.

8. Critically evaluate the limitations of quantitative models in uncertain supply-chain environments.

The Decision Problem: Flow, Time and Scarcity

A logistics risk decision arises when the normal configuration of supply, transport, storage or production may not deliver the required quantity at the required time. Four variables are particularly important.

Demand determines the quantity and timing of requirements. Demand may be stable, seasonal, promotional or highly uncertain. Forecast error becomes especially costly when replenishment lead time is long.

Lead time is the elapsed time between placing an order and receiving usable material. It includes order processing, production, consolidation, port handling, customs, transportation, inspection and internal receiving. The mean lead time indicates expected replenishment duration; its variability determines the inventory needed for protection.

Capacity is the maximum flow that a resource can process during a period. Capacity may concern vessel space, rail slots, trucks, warehouse positions, labour, production lines, containers or supplier output. A shortage occurs when required flow exceeds available capacity.

Service requirement specifies what the customer or operating process considers acceptable. It may be expressed as a cycle service level, fill rate, maximum allowable delay, on-time delivery percentage or probability of avoiding a production stoppage.

These variables are interdependent. A port closure increases lead time and its variance. The resulting inventory requirement may exceed warehouse capacity. A decision to divert freight to an alternative port may consume scarce trucking capacity. Consequently, the best decision cannot be identified by transport cost alone.

Core Concepts and Terminology

Cycle stock is inventory held to meet expected demand between replenishments. Safety stock is additional inventory held against uncertainty in demand, lead time or supply. Strategic or anticipatory stock is accumulated deliberately before a known seasonal, regulatory or geopolitical risk. Emergency inventory is positioned for rapid response to an imminent or ongoing disruption.

A reorder point is the inventory position at which replenishment should be initiated. In a basic continuous-review system:

Reorder point = expected demand during lead time + safety stock.

Inventory position includes on-hand inventory, inventory on order and, where relevant, inventory allocated to outstanding commitments, less backorders. The distinction is important: using physical stock alone can cause duplicate ordering.

A stockout is the absence of inventory when demand occurs. A backorder is demand that is accepted but fulfilled later; lost sales occur when the customer does not wait. A bottleneck is a resource whose constrained capacity limits total system output. Throughput is the rate at which the system generates completed output, while utilization is the proportion of available capacity being used.

A criticality classification identifies items according to the operational or customer consequence of failure. An inexpensive component can be highly critical if its absence stops a production line. Time-sensitive inventory loses value rapidly through deterioration, obsolescence or missed selling windows.

Total landed cost includes purchase price, transport, insurance, duties, brokerage, handling, storage, financing, quality costs and expected disruption-related costs. The expected cost of disruption can be approximated as the probability of an event multiplied by its financial consequence, although this simple formulation may understate severe, correlated or cascading events.

Resilience capacity is the ability to absorb disruption and continue operating, whereas redundancy refers to spare or alternative resources. Flexibility is the ability to change volume, route, supplier or mode. These are related but not identical: a firm may have two ports but lack the contractual or operational flexibility to use the second one quickly.

Lead-Time Variability and Inventory Buffers

For a system with relatively stable demand, safety stock may be estimated as:

Safety stock = z × standard deviation of demand during lead time, where z is the service-level factor. A higher target service level requires a higher z and therefore more stock. If demand and lead-time uncertainty are both material, a commonly used approximation for the variance of demand during random lead time is:

σ²DL = Lσ²D + D²σ²L, assuming demand and lead time are independent, where D is average demand per period, L is average lead time, σD is demand standard deviation and σL is lead-time standard deviation.

The formula is useful but not universal. Demand and lead time may be correlated: promotions can increase demand precisely when transport networks are congested. Non-normal demand, minimum order quantities, intermittent demand and supply constraints also reduce the validity of simple z-score calculations. In such situations, simulation, empirical percentiles or scenario analysis may be more appropriate.

A buffer should be located where it provides the greatest protection at acceptable cost. Firms may hold raw-material stock near a plant, semi-finished inventory at a postponement centre, finished goods near customers, or capacity reservations with carriers. The right location depends on product value, customization, postponement opportunities, transport time and the consequences of shortage.

Inventory pooling can reduce total safety stock because uncertainty is aggregated rather than separately protected at every location. However, pooling may increase transportation time, customs exposure or dependence on a single distribution point. The risk benefit of centralization must therefore be compared with concentration risk.

Inventory decisions should also consider ageing and obsolescence. A buffer of standard components may be inexpensive and reusable; a buffer of short-life or highly customized products may create substantial write-offs. The effective protection value of inventory is its usable quantity at the time and location of need, not its accounting quantity.

Transport Disruption and Port Congestion

Transport disruption can be classified by location and mechanism. Inland disruption may involve road closures, driver shortages, rail capacity limitations or warehouse failure. Maritime disruption may involve vessel cancellation, blank sailing, port closure, equipment imbalance, labour action or channel restrictions. Airfreight is exposed to airport capacity, fuel cost, security controls and limited payload. Customs and documentation failures can create delays even when physical transport is available.

Port congestion is a system condition rather than merely a queue of vessels. It may include berth delays, yard saturation, shortage of empty containers, slow gate operations, insufficient rail or truck appointments and customs backlogs. Congestion raises both average lead time and variance. It may also create demurrage, detention, storage and missed-connection charges.

A sound response begins with visibility. The firm should identify shipments in transit, inventory at risk, customer priorities, alternative ports, available equipment and downstream capacity. It should then classify cargo by urgency and economic consequence. Not every shipment merits diversion or premium transport.

Contingency routing is the planned movement of goods through an alternative path when the preferred route becomes unavailable or excessively risky. Routes should be evaluated on more than nominal transit time. Relevant criteria include:

• end-to-end transit-time distribution rather than its average alone;
• total landed cost, including handling and customs;
• available port, terminal, rail, truck and warehouse capacity;
• documentation and regulatory compatibility;
• frequency and severity of historical disruption;
• carbon emissions and product integrity requirements; and
• the time and cost required to activate the route.

A route that exists on paper but has no confirmed slots, inland transport or customs capability is not a genuine contingency. Firms should prequalify alternatives, establish operating procedures and periodically test them through exercises or small-volume shipments.

Capacity Risk and Allocation Decisions

Capacity risk occurs when demand for a constrained resource exceeds available supply. It can arise from a temporary incident or from structural scarcity. Common responses are to increase capacity, reduce demand, prioritize orders, substitute resources or redesign the network.

Capacity expansion may be physical, contractual or digital. Examples include adding a production shift, leasing warehouse space, reserving vessel slots, contracting secondary carriers, cross-training labour or using a control tower to coordinate resources. Flexible capacity often costs more per unit than committed capacity, but it can have superior value when uncertainty is high.

When capacity is insufficient, allocation rules become necessary. A simple rule based only on order arrival can produce poor outcomes. More defensible criteria include customer criticality, contractual penalties, safety implications, contribution margin, strategic importance and the feasibility of substitution. Allocation should be transparent and approved through governance rather than improvised by individual planners.

The theory of constraints emphasizes that total system performance is governed by the bottleneck. Improving a non-bottleneck resource may create local efficiency without increasing throughput. Managers should identify the constraint, exploit it through better scheduling and reduced downtime, subordinate other activities to it, elevate its capacity where justified, and then reassess the system after the constraint moves.

Capacity decisions also involve the danger of high utilization. Near-full utilization may appear efficient under normal conditions but leaves little room for variability. A modest amount of spare capacity can reduce queue growth and expedite costs. The economically relevant question is not whether spare capacity is fully used; it is whether its cost is lower than the expected cost of shortage and disruption.

Total-Cost Analysis and Optimization

The central analytical principle is to compare alternatives using total relevant cost over the decision horizon. A simplified objective can be expressed as:

Minimize total cost = purchase cost + transport cost + handling cost + inventory holding cost + shortage cost + capacity cost + expected disruption cost + environmental and compliance cost.

Not every term must be monetized precisely. However, omitting a material cost can bias the decision. For example, a low-freight route may require more safety stock, generate higher demurrage exposure and increase the probability of lost sales.

A basic scenario table can compare routes under normal, moderate-disruption and severe-disruption conditions. Each scenario should specify probability or decision relevance, expected lead time, capacity, cost and service result. Sensitivity analysis should test fuel prices, demand, port delay, exchange rates, inventory carrying cost and the probability of disruption.

Linear programming can support allocation and routing. A simplified model may define decision variables for the quantity shipped from each origin through each route to each destination. Constraints represent supply, transport capacity, warehouse capacity, demand satisfaction and flow balance. The objective minimizes cost. Mixed-integer programming is needed when decisions involve opening a facility, activating a route, hiring a dedicated carrier or selecting an all-or-nothing option.

Stochastic optimization represents uncertain demand, lead time or capacity through scenarios. A two-stage model may choose preventive actions before uncertainty is known and corrective actions after a disruption occurs. Robust optimization instead seeks solutions that remain feasible across a defined range of adverse conditions. It may produce a slightly higher normal cost in exchange for lower vulnerability.

Optimization results should not be treated as automatic decisions. Models can create false precision when probabilities are weak, constraints are missing or data are stale. Managers must review feasibility, execution time, contractual restrictions and human consequences. A mathematically optimal plan that cannot be implemented at a congested terminal is operationally inferior to a feasible alternative.

A Practical Decision Process

An integrated decision process can be organized into seven linked activities. First, define the service requirement and the decision horizon. Second, map the physical and information flows, identifying nodes, bottlenecks and dependencies. Third, estimate demand, lead-time and capacity distributions from historical and forward-looking evidence. Fourth, identify response options, including inventory, routing, mode, sourcing, scheduling and customer-priority alternatives. Fifth, evaluate options using total cost, service, risk and sustainability criteria. Sixth, authorize and activate the selected response through clear decision rights. Seventh, monitor results and update assumptions.

The process should distinguish prevention from response. Prevention includes supplier qualification, diversified routes, documentation quality, inventory positioning and capacity contracts. Response includes shipment diversion, emergency procurement, allocation, expediting and customer communication. Recovery concerns restoring normal flows and correcting the structural weakness that allowed the disruption to have excessive impact.

A useful control system combines leading and lagging indicators. Leading indicators include port dwell time, blank sailings, carrier acceptance rates, supplier schedule adherence, booking rejection, inventory coverage and forecast error. Lagging indicators include stockouts, on-time-in-full performance, expedite expenditure, demurrage, lost sales and recovery time. Thresholds should trigger predefined actions rather than merely generate reports.

Illustrative Case: Meridian Appliances and a Congested Gateway

Background

Meridian Appliances, a fictional manufacturer, imports electronic control modules and compressors for two assembly plants. Its primary ocean gateway offers low freight cost and frequent services. The company normally carries four days of finished-goods safety stock and uses a continuous-review replenishment system. Control modules are inexpensive relative to the finished product but are highly critical: a shortage stops assembly.

Situation and Problem

A combination of vessel bunching, yard congestion and inland-truck shortages increases the primary gateway’s average clearance time from 10 to 22 days, with significant variability. Forecast demand for a new appliance model is also higher than planned. Meridian must decide whether to continue using the gateway, divert cargo to a secondary port, use airfreight for selected modules, or increase inventory.

Analysis

The supply-chain team first distinguishes critical from non-critical items. Compressors can be sourced from a qualified regional supplier within a week, whereas control modules have a long replenishment cycle and no immediate substitute. The secondary port has a higher ocean and inland cost, but its terminal and rail capacity are less congested. Airfreight is fast but limited and expensive.

The team develops three scenarios: normal congestion, prolonged congestion and gateway closure. It compares alternatives using freight, handling, inventory holding, expected shortage, line-stoppage and activation costs. It also checks whether the secondary port can accept the required container volume and whether documentation is valid for the alternative inland route.

The analysis shows that diverting all cargo would unnecessarily increase cost. Continuing all cargo through the primary gateway creates unacceptable exposure for control modules. The best segmented response is to route critical modules through the secondary port, maintain ordinary components through the primary gateway, and use a limited airfreight trigger if projected inventory falls below the production-protection threshold. A temporary buffer is positioned at the plant rather than at the congested gateway because the plant is the point at which shortage causes immediate operational loss.

Approach and Outcome

Meridian signs a short-term secondary-port handling agreement, reserves weekly rail capacity, increases visibility through milestone alerts and establishes an escalation rule based on projected days of coverage. Production planners receive a priority allocation rule approved by operations and commercial management. Customer service communicates realistic delivery commitments rather than promising universal protection.

The policy raises normal logistics cost, but it avoids broad airfreight use and reduces the probability of a line stoppage. After congestion declines, Meridian retains the secondary route as a tested contingency, reviews the appropriate buffer level and negotiates flexible rather than purely lowest-cost transport contracts.

Learning Points

The case illustrates that resilience is selective rather than synonymous with duplicating every flow. Item criticality, route feasibility, capacity confirmation and total cost must be considered together. It also shows why contingency plans require activation thresholds, governance and operational testing.

Technology, Analytics and Digital Control

Transport-management systems can compare rates, book capacity, consolidate shipments and monitor milestones. Warehouse-management systems improve inventory accuracy and support allocation by location and priority. Control towers combine carrier, port, supplier, inventory and demand data to provide exception management rather than merely historical reporting.

Predictive analytics can estimate arrival times, port dwell, stockout probability and the likely effect of alternative routes. Digital twins and discrete-event simulation can test network configurations, buffer locations and capacity scenarios before implementation. Optimization engines can recommend allocations, but recommendations should be governed by data-quality controls and human approval for exceptional decisions.

Artificial intelligence can identify patterns in disruption signals, classify exceptions and support probabilistic forecasting. Its limitations include biased or incomplete training data, unstable relationships during unprecedented events, poor explainability and the possibility of automating an incorrect objective. Human planners remain responsible for validating recommendations, protecting commercially sensitive information and ensuring that automated prioritization does not unfairly disadvantage customers or suppliers.

Challenges, Limitations and Common Errors

A frequent error is optimizing transport cost while ignoring inventory and shortage consequences. Another is using average lead time when the operational problem is variance. Managers may also count theoretical alternative capacity that is unavailable during the same regional disruption. Excessive safety stock can conceal poor forecasting, unreliable suppliers or weak process control.

Models are limited by data granularity, demand changes, dependence between risks and the difficulty of estimating rare events. Scenario probabilities can create unjustified confidence. Financial models may also undervalue safety, reputation, employee welfare and customer trust. Conversely, unstructured resilience spending can produce redundant assets with little risk reduction.

Common operational mistakes include failing to update master data after a route change, overlooking customs requirements, neglecting empty-container availability, placing buffers at a congested node, and activating premium freight without a stop rule. Post-disruption reviews should examine not only whether goods arrived but whether the decision was economically and ethically justified.

Governance, Ethics and Professional Responsibility

Resilience decisions affect suppliers, employees, customers and communities. Allocating scarce capacity solely to the highest-margin customer may conflict with contractual commitments or create unfair treatment. Emergency sourcing may expose the firm to labour, environmental or product-safety risks. Expedited transport can increase emissions, while deliberate inventory disposal creates environmental waste.

Governance should specify who may authorize premium freight, alter customer priorities, release strategic stock or activate an alternative supplier. Decisions should be documented, auditable and consistent with contracts, trade controls, customs law, sanctions requirements, competition law and product regulations. Data sharing with carriers and technology providers should observe confidentiality and cybersecurity requirements.

Professional judgement requires balancing continuity with responsible conduct. A resilient supply chain is not one that transfers risk invisibly to weaker suppliers or communities; it is one that identifies exposure, shares information appropriately and manages trade-offs transparently.

Emerging Developments and Best Practices

Current practice is moving from static safety stock toward differentiated, risk-based buffers and dynamic capacity portfolios. Firms increasingly combine long-term committed capacity with options for surge volume. Regionalization and nearshoring may reduce some transit risks but can increase unit cost or dependence on concentrated local suppliers. Multi-modal networks, inland ports and rail alternatives are receiving greater attention as maritime reliability becomes more variable.

Best practice includes:

• segmenting products by criticality, value, substitutability and time sensitivity;
• measuring lead-time distributions and service outcomes rather than relying only on averages;
• evaluating routes through total landed cost and scenario analysis;
• maintaining genuinely executable alternatives with contracts, documentation and capacity;
• locating buffers at the point where they protect the most consequential failure;
• combining committed and flexible capacity;
• establishing transparent allocation and escalation rules;
• using digital visibility with data-quality and cybersecurity controls; and
• conducting post-event reviews, simulations and periodic contingency tests.

Chapter Summary

Logistics, inventory and capacity risk decisions require an integrated view of time, flow, scarcity and service. Transport disruption and port congestion increase lead-time variability, which raises inventory requirements and may create production or customer shortages. Safety stock, reorder points and strategic buffers provide protection, but their location and size must reflect criticality, cost, obsolescence and network concentration.

Contingency routing should be evaluated through feasible capacity, end-to-end reliability, total landed cost, regulatory requirements and activation effort. Capacity shortages require explicit allocation rules and attention to the system bottleneck. Total-cost analysis should include inventory, shortage, capacity, environmental and expected disruption costs. Linear, mixed-integer, stochastic and robust optimization can improve decisions, but model outputs require managerial validation.

Resilience is strengthened by tested alternatives, flexible capacity, digital visibility, clear governance and disciplined learning. The objective is not to eliminate all uncertainty or maximize spare resources, but to achieve an economically and ethically defensible level of continuity under plausible disruption conditions.

Key Terms

Capacity risk: The possibility that available resource capacity will be insufficient to meet required flow.

Contingency routing: A preplanned alternative path used when the normal route is unavailable or unreliable.

Cycle stock: Inventory held to satisfy expected demand between replenishments.

Expected disruption cost: Probability-weighted financial consequence of a disruption.

Fill rate: The proportion of demand fulfilled immediately from available inventory.

Inventory pooling: Consolidating inventory to reduce aggregate uncertainty and safety-stock requirements.

Inventory position: On-hand and on-order inventory adjusted for allocations and backorders.

Lead-time variability: Fluctuation in the time required for replenishment.

Port congestion: Excess demand or operational impairment causing delays in port and connected inland activities.

Reorder point: Inventory position that triggers replenishment.

Robust optimization: Optimization seeking solutions that remain acceptable across adverse conditions.

Safety stock: Additional inventory held against demand or supply uncertainty.

Service level: A measure of the probability or extent to which demand is fulfilled as promised.

Strategic stock: Inventory accumulated deliberately to protect against a known future risk or event.

Total landed cost: Complete cost of acquiring and delivering goods to the required destination.

Throughput: Rate at which a system produces completed output.

Utilization: Proportion of available capacity being used.

Bottleneck: Resource whose constrained capacity limits overall system performance.

Review Questions

Short Answer Questions

1. Why is lead-time variability often more important than average lead time when setting safety stock?

2. Distinguish cycle stock, safety stock and strategic stock.

3. What costs should be included in a total landed-cost analysis?

4. Why may a theoretically available alternative route fail as a contingency route?

5. Explain the relationship between a bottleneck and system throughput.

Long Answer Questions

1. Explain how transport disruption, inventory buffers and capacity shortages interact in a global supply chain.

2. Discuss the assumptions and limitations of safety-stock calculations based on demand and lead-time variance.

3. Describe how a firm should evaluate and implement a contingency-routing strategy.

4. Compare linear, stochastic and robust optimization for supply-chain risk decisions.

5. Examine the governance and ethical issues involved in allocating scarce logistics capacity.

Analytical / Application Questions

1. A component has average daily demand of 500 units, average replenishment lead time of 8 days and safety stock of 1,200 units. Calculate its reorder point and explain how the result should be reconsidered if lead-time variability increases substantially.

2. A primary port has low freight cost but high congestion risk, while a secondary port has higher cost and reliable inland capacity. Design a scenario-based decision process for determining when to divert critical components.

3. A plant faces a production bottleneck and cannot fulfil all customer orders during a disruption. Develop an allocation rule using customer criticality, contractual commitments and shortage consequences, and explain how the rule should be governed.

22. Multiple Choice Questions with Answers

Question 1

Which factor most directly increases the safety stock required for a given target service level?

A. Lower demand variability

B. Higher lead-time variability

C. Lower product criticality

D. Shorter order-processing time

Correct Answer: B. Higher lead-time variability

Question 2

Why should a contingency route be tested before a major disruption occurs?

A. To guarantee that its freight rate is always lowest

B. To eliminate the need for inventory

C. To confirm capacity, documentation and operational feasibility

D. To ensure that all shipments use the alternative route

Correct Answer: C. To confirm capacity, documentation and operational feasibility

Question 3

Which expression best represents a basic reorder point?

A. Average demand minus safety stock

B. Average demand during lead time plus safety stock

C. Annual demand divided by inventory value

D. Transport cost plus purchase price

Correct Answer: B. Average demand during lead time plus safety stock

Question 4

A firm diverts only critical control modules to a secondary port while keeping ordinary components at the primary port. What principle does this illustrate?

A. Uniform allocation regardless of risk

B. Selective resilience based on item criticality

C. Elimination of all route concentration

D. Maximization of transport utilization

Correct Answer: B. Selective resilience based on item criticality

Question 5

Which cost is most likely to be omitted in a narrow freight-cost comparison but included in total-cost analysis?

A. Expected shortage cost

B. Product description

C. Supplier name

D. Order number

Correct Answer: A. Expected shortage cost

Question 6

What is the principal concern with operating a resource at nearly full utilization during uncertain conditions?

A. It guarantees lower lead-time variance

B. It leaves little capacity to absorb variability

C. It eliminates bottlenecks

D. It reduces the need for scheduling

Correct Answer: B. It leaves little capacity to absorb variability

Question 7

In a capacity shortage, which allocation approach is most defensible?

A. Give all capacity to the earliest orders without review

B. Allocate randomly to avoid managerial responsibility

C. Use transparent criteria such as criticality and contractual obligations

D. Prioritize only the largest customer by revenue

Correct Answer: C. Use transparent criteria such as criticality and contractual obligations

Question 8

What is a key limitation of using average lead time alone in inventory planning?

A. It cannot measure purchase price

B. It ignores the dispersion and tail risk of delays

C. It always overstates demand

D. It prevents inventory pooling

Correct Answer: B. It ignores the dispersion and tail risk of delays

Question 9

Which model is particularly appropriate when preventive decisions occur before uncertain disruption scenarios and corrective decisions occur afterward?

A. Two-stage stochastic optimization

B. Simple descriptive accounting

C. Fixed reorder-point calculation only

D. Historical average costing

Correct Answer: A. Two-stage stochastic optimization

Question 10

Which governance practice is most appropriate when activating premium freight?

A. Allow any planner to use it without a stop rule

B. Authorize it through defined thresholds and decision rights

C. Use it for every high-value product permanently

D. Exclude customer-service and finance functions

Correct Answer: B. Authorize it through defined thresholds and decision rights