4. Facility Location and Network Design

Introduction and Context

Facility location is the strategic decision concerning where an organization should place factories, warehouses, distribution centres, service units, offices, or other operating facilities. Network design extends this decision by determining how many facilities are required, what roles they should perform, how they should be connected, and how products, materials, information, and customers should move through the network. Together, location and network decisions shape the long-term economics and responsiveness of the production system.

A plant layout decision mainly concerns the arrangement of resources within a facility. A location decision concerns the position of the facility in relation to suppliers, labour markets, customers, transport systems, utilities, regulations, competitors, and sources of risk. The two decisions are related: a poorly selected site cannot usually be rescued by an efficient internal layout, while a well-located facility may lose its advantage through inadequate internal design.

Location decisions are difficult because they involve large, often irreversible investments and consequences extending over decades. They also involve multiple objectives. A site with low land cost may have poor labour availability; a region with attractive incentives may be exposed to political or climate risk; and a location close to customers may increase operating cost but reduce delivery time and inventory. The purpose of systematic analysis is therefore not to identify a universally “best” location, but to select the alternative that offers the most suitable balance of cost, service, risk, strategic flexibility, and stakeholder value.

Learning Objectives

After studying this chapter, students should be able to:

1. Distinguish facility location from network design and explain their strategic importance.

2. Identify global, regional, and site-level location factors.

3. Apply factor-rating, load–distance, centre-of-gravity, break-even, and optimization methods.

4. Interpret qualitative, quantitative, and risk-based location analyses.

5. Evaluate labour, infrastructure, incentives, resilience, sustainability, and geopolitical considerations.

6. Explain how digital analytics and scenario modelling support location decisions.

7. Recommend a location and network configuration while recognizing uncertainty and implementation constraints.

The Strategic Logic of Location and Network Design

A supply network consists of nodes and links. Nodes include suppliers, plants, contract manufacturers, warehouses, cross-docks, retailers, and customers. Links represent transportation routes, information flows, and commercial relationships. Network design determines the structure of this system: the number and capacity of facilities, their geographic positions, allocation of markets, sourcing arrangements, and transportation modes.

Location decisions are commonly made at three levels. At the global level, the organization evaluates countries or broad economic regions. Issues include market access, trade agreements, taxation, political stability, currency exposure, intellectual property protection, and geopolitical relations. At the regional level, alternatives such as states, provinces, industrial corridors, or metropolitan areas are compared. Labour pools, supplier ecosystems, infrastructure, utilities, and regional incentives become prominent. At the site level, specific parcels are assessed for land, access, topography, environmental restrictions, zoning, security, expansion potential, and connection to transport and utilities.

The decision should also reflect the facility’s strategic role. A plant designed for low-cost standardized production may favour scale, inexpensive labour, and efficient export logistics. A plant producing customized or time-sensitive products may favour proximity to customers and engineering talent. A regional distribution centre may be positioned to balance transport cost and delivery responsiveness, whereas a research facility may depend primarily on specialized knowledge and institutional connections.

Location Factors

The most important factors usually include the following:

• Market proximity: distance to customers, demand density, response-time requirements, and market growth.
• Supplier access: availability, reliability, quality, and geographic concentration of suppliers.
• Labour: wage rates, skills, productivity, labour relations, turnover, training requirements, and future availability.
• Infrastructure: roads, railways, ports, airports, telecommunications, power, water, waste treatment, and industrial services.
• Cost structure: land, construction, taxes, utilities, transport, inventory, compliance, and operating costs.
• Government and regulation: zoning, permits, environmental rules, labour law, taxation, customs, and incentives.
• Risk and resilience: natural hazards, political instability, cyber risk, public health threats, supply disruption, and business continuity.
• Strategic flexibility: expansion land, modular capacity, alternate transport routes, and options for future network redesign.
• Sustainability and social legitimacy: emissions, water use, community impact, environmental justice, and acceptance by local stakeholders.

These factors interact. For example, a site near a port may reduce transport distance but be vulnerable to congestion or coastal flooding. A generous tax incentive may reduce initial cost but create dependence on a government commitment that could change. Location analysis must therefore consider total cost and total risk rather than one attractive feature.

A Structured Location-Selection Process

A robust location study begins by translating strategy into decision criteria. Management should define the required service level, expected demand, facility role, planning horizon, capacity, product characteristics, and acceptable risk. The analysis then proceeds through successive levels of detail.

First, a long list of countries or regions is screened using non-negotiable requirements such as legal access, technical feasibility, minimum labour availability, security, and infrastructure. Second, feasible regions are compared using weighted criteria and economic models. Third, a smaller number of specific sites is investigated through visits, engineering studies, discussions with authorities, and validation of cost and risk data. Finally, alternatives are tested under different demand, cost, disruption, and policy scenarios.

The process should include representatives from operations, supply chain, finance, human resources, engineering, legal compliance, information technology, sustainability, and commercial management. Cross-functional participation reduces the risk that a financially attractive site will later prove impossible to staff, permit, connect, or operate.

Factor-Rating Method

The factor-rating method is useful when both quantitative and qualitative criteria must be considered. Each factor receives a weight representing its importance, and each alternative receives a score. The weighted score is calculated as:

Total score for site j = Σ (weight i × rating ij)

Weights should normally sum to 1. Ratings may use a scale such as 1 to 10, provided the scale is defined consistently. For example, a company comparing three regions may assign weights of 0.30 to market access, 0.25 to labour, 0.20 to infrastructure, 0.15 to operating cost, and 0.10 to risk. Scores are then developed from evidence rather than intuition alone.

The method provides transparency and supports discussion, but it does not eliminate judgement. Results can change substantially when weights or ratings change. Sensitivity analysis is therefore essential. If a site ranks first only under one narrow set of weights, management should treat the result as fragile rather than definitive.

Load–Distance Analysis

Load–distance analysis is appropriate when transportation effort is a major determinant of location. A load is the quantity moved between an origin and a destination, while distance represents the relevant travel distance. The basic measure is:

LD = Σ (load i × distance i)

The preferred location generally has the lowest total load–distance value, assuming that other conditions are comparable. Distance may be measured as road kilometres, travel time, or a weighted measure reflecting congestion and route quality. For a warehouse, loads may represent annual shipments from the warehouse to customer zones. For a plant, they may represent inbound materials and outbound finished goods.

The method is simple but limited. It normally assumes that all kilometres have similar cost, whereas actual costs depend on mode, fuel, tolls, backhauling, border delays, and shipment size. It also does not directly include land, labour, taxes, service capacity, or disruption risk.

Centre-of-Gravity Method

The centre-of-gravity method provides an initial geographic estimate for a facility serving multiple demand or supply points. If point i has coordinates (xᵢ, yᵢ) and weight wᵢ, the estimated location is:

x̄ = Σ(wᵢxᵢ) / Σwᵢ

ȳ = Σ(wᵢyᵢ) / Σwᵢ

Weights may represent shipment volume, demand, revenue, or transportation frequency. The result is a mathematical starting point, not a final site. A calculated point may lie in a residential area, protected land, a floodplain, or a location without suitable roads. The estimate must therefore be adjusted for actual geography, infrastructure, zoning, and service constraints.

Break-Even and Total-Cost Analysis

Break-even analysis compares fixed and variable costs among alternatives. For a facility with fixed cost F and variable cost v per unit, total cost is:

TC = F + vQ

where Q is output or throughput. The break-even volume between two alternatives can be found by equating their total-cost equations. A high-fixed-cost site may be attractive at large volume because of low variable cost, while a flexible site with low fixed cost may be preferable when demand is uncertain.

Location decisions should use total landed cost rather than factory conversion cost alone. Relevant components may include inbound freight, outbound freight, duties, inventory carrying cost, labour, utilities, taxes, quality losses, relocation, security, compliance, and expected disruption cost. A discounted cash-flow analysis is appropriate when alternatives differ in investment timing, useful life, residual value, or tax treatment.

Optimization Models

When many facilities, products, suppliers, and customer zones are involved, network optimization can evaluate combinations that cannot be compared manually. A basic facility-location model may use binary variable yⱼ, equal to 1 if facility j is opened, and flow variable xᵢⱼ, representing quantity shipped from facility j to customer i. A simplified objective is:

Minimize Σⱼ Fⱼyⱼ + ΣᵢΣⱼ cᵢⱼxᵢⱼ

subject to demand satisfaction, facility-capacity limits, and logical constraints linking flows to opened facilities. More advanced models include multiple products, transportation modes, service-time constraints, carbon emissions, supplier capacity, inventory, taxes, and facility closure or expansion decisions.

Optimization does not replace managerial judgement. It produces a solution under stated assumptions. If demand, costs, or capacities are inaccurate, an apparently precise answer may be misleading. Managers should examine alternative solutions, shadow prices, binding constraints, and the effect of relaxing assumptions.

Evaluating Uncertainty, Risk, and Resilience

Traditional location studies often use average cost and expected demand. Contemporary network design must also ask how the network behaves when conditions are abnormal. Resilience is the ability to anticipate, absorb, adapt to, and recover from disruption. It may be improved through multiple suppliers, geographic diversification, strategic inventory, alternate transport modes, modular capacity, emergency contracts, and visibility systems.

Risk analysis begins by identifying hazards and estimating their likelihood and consequence. A qualitative risk matrix can classify threats such as flooding, labour unrest, border closure, power interruption, cyberattack, or regulatory change. Quantitative approaches may use expected loss:

Expected loss = probability of event × financial consequence

This estimate should be supplemented by scenario analysis because low-probability events can have catastrophic effects. Stress tests may examine a major supplier failure, a 30 percent demand increase, a prolonged port closure, currency depreciation, or a sudden energy-price increase. Robust solutions may not minimize cost in the most likely scenario, but they may perform acceptably across a wide range of plausible futures.

Real-options thinking is useful when uncertainty is high. A firm may deliberately lease rather than purchase land, begin with a smaller facility, reserve expansion rights, or use a contract manufacturer while demand is tested. These choices have option value because they preserve future flexibility. The cheapest immediate alternative is not always the most economically rational over the full planning horizon.

Global and Regional Considerations

International location decisions require analysis beyond wage comparison. Total cost may be affected by customs procedures, rules of origin, tariffs, transfer pricing, tax obligations, foreign-exchange movements, local-content requirements, and the reliability of cross-border logistics. Political relationships can alter access to technology, components, capital, or markets. A country may be economically attractive but strategically unsuitable if critical inputs depend on a politically vulnerable corridor.

Labour analysis should examine not only hourly pay but also productivity, absenteeism, turnover, training time, supervisory requirements, and industrial-relations conditions. Infrastructure should be evaluated through actual reliability rather than nominal availability. A site with a grid connection may still require backup generation if voltage fluctuations or outages are frequent. Similarly, an apparent motorway connection may provide poor performance during seasonal congestion.

Incentives can influence a decision but should not dominate it. Grants, tax holidays, subsidized land, infrastructure support, and employment credits should be assessed for duration, conditions, clawback provisions, administrative reliability, and political durability. Ethical governance requires that the organization report incentives transparently and avoid decisions based on unrealistic promises to public authorities.

Digital Tools and Data-Supported Decisions

Geographic information systems can combine demand locations, road networks, labour catchments, property data, hazard maps, utilities, and emissions information. Advanced analytics can estimate travel time rather than straight-line distance, model congestion, and compare alternative service territories. Digital twins can represent a proposed network and test demand, capacity, inventory, and disruption scenarios before capital is committed.

Artificial intelligence and machine-learning tools may improve demand forecasting, site-risk classification, and pattern detection in logistics data. However, algorithmic output is only as reliable as the data and assumptions behind it. Historical data may underrepresent rare disruptions, encode existing geographic bias, or overlook informal constraints. Governance should require traceable data sources, human review, sensitivity testing, cybersecurity, and clear accountability for the final decision.

An Illustrative Case: Selecting a Regional Distribution Centre

Background

Orion Appliances, a fictional manufacturer, serves five metropolitan markets from a plant located in the western part of a country. Demand has grown in the eastern markets, and delivery times have become inconsistent. Management is considering three possible distribution-centre regions: Northport, Central Junction, and Eastgate.

Situation and Analysis

Northport has excellent rail and port access and the lowest land cost, but it is far from the largest customer cluster and has a limited skilled-labour pool. Central Junction is close to the geographic centre of demand, has strong highway access, and offers moderate labour and land costs. Eastgate is closest to the largest market and has an available workforce, but land is expensive and the region depends heavily on one congested highway.

The team first applies a factor-rating model using market access, labour, infrastructure, cost, expansion potential, and disruption risk. Central Junction receives the highest weighted score. A load–distance analysis also favours Central Junction when annual customer shipments are used as loads. Break-even analysis shows that Eastgate would become financially attractive only if eastern demand exceeded the current forecast by a substantial margin. A risk scenario involving highway closure reduces Eastgate’s service performance, while Central Junction retains two feasible alternate routes.

Approach and Recommendation

The team recommends Central Junction as the primary site, with a smaller contingency-stock arrangement near Eastgate rather than a full second distribution centre. The recommendation is subject to confirmation of land drainage, utility reliability, labour availability, and contractual protection for expansion. A two-stage investment is selected: lease and fit out an initial building, then expand when demand reaches a defined threshold.

Outcome and Learning Points

The illustrative case demonstrates that no single method is sufficient. Factor rating incorporates strategic and qualitative criteria; load–distance tests transportation effort; break-even analysis examines volume sensitivity; and risk analysis reveals the value of alternate routes and staged investment. The final recommendation is a network decision, not merely a site ranking.

Challenges, Limitations, and Common Errors

Several errors recur in facility-location studies. The first is treating initial capital cost as total cost. The second is using straight-line distance where travel time and route reliability matter. The third is assuming that incentives are permanent. The fourth is relying on average demand and ignoring peaks, growth, and product mix. The fifth is underestimating ramp-up time, recruitment difficulty, and the cost of transferring knowledge to a new site.

Other problems arise from false precision. A model may produce coordinates or an optimized network to several decimal places even though demand forecasts and cost estimates are uncertain. Decision makers should present ranges, confidence levels, assumptions, and sensitivity results. They should also distinguish hard constraints, such as legal restrictions, from preferences, such as a lower land price.

Organizational bias is another danger. A proposed site may be favoured because of executive preference, political pressure, familiarity, or a desire to justify a decision already made. Independent review, documented criteria, and an audit trail improve governance. Site visits and direct verification are indispensable because databases rarely capture every operational reality.

Ethical, Regulatory, and Sustainability Considerations

Facility location affects employees, communities, suppliers, public infrastructure, and the natural environment. A responsible analysis considers emissions, water consumption, waste, noise, traffic, land conversion, worker safety, and the distribution of benefits and burdens among communities. Environmental permits and impact assessments should be treated as substantive decision inputs rather than administrative obstacles.

Organizations should avoid exploiting weak labour regulation or shifting environmental harm to jurisdictions with limited enforcement. Procurement and location decisions should include responsible labour standards, anti-corruption controls, and respect for community consultation. Sustainability can also improve resilience: renewable energy, water reuse, energy-efficient buildings, and local sourcing may reduce both environmental impact and exposure to volatile resources.

Emerging Directions and Best Practices

Current network design is moving toward regionalization, multi-echelon flexibility, and greater visibility of tier-two and tier-three suppliers. Firms are reassessing extreme concentration in distant low-cost locations and are considering nearshoring, friend-shoring, postponement, and dual sourcing. The appropriate response is not universal reshoring; it is a careful comparison of cost, service, risk, technology, and strategic control.

Best practice is to:

• align the location decision with competitive strategy and facility role;
• separate mandatory screening criteria from weighted preferences;
• use several complementary analytical methods;
• include total landed cost, working capital, and disruption exposure;
• test demand, cost, policy, and hazard scenarios;
• evaluate labour and infrastructure through verified evidence;
• preserve flexibility through phased investment and expansion options;
• document assumptions, uncertainty, and governance responsibilities; and
• review the network periodically rather than treating location as permanently settled.

Chapter Summary

Facility location and network design determine where operating capacity is placed and how facilities, suppliers, markets, and transport links are connected. The decision involves global, regional, and site-level analysis and must reflect cost, service, labour, infrastructure, regulation, sustainability, resilience, and geopolitics. Factor rating supports multi-criteria comparison; load–distance and centre-of-gravity methods examine geographic relationships; break-even analysis compares cost structures; and optimization models evaluate complex networks. Risk and scenario analysis are necessary because average-cost solutions may fail under disruption. Digital tools improve visibility and modelling but require reliable data and human governance. Effective decisions combine analytical discipline with managerial judgement, staged investment, and continuous network review.

Key Terms

Facility location: The selection of a geographic position for an operating facility.

Network design: The configuration of facilities, flows, capacities, and market assignments in a supply network.

Load–distance method: A technique that compares locations using the sum of shipment loads multiplied by relevant distances.

Centre of gravity: A weighted geographic point used as an initial estimate for facility placement.

Factor rating: A weighted scoring method for comparing alternatives across multiple criteria.

Break-even volume: The output level at which two alternatives have equal total cost.

Total landed cost: The complete cost of supplying a product, including production, transport, duties, inventory, and related expenses.

Fixed cost: A cost that does not vary directly with short-term output or throughput.

Variable cost: A cost that changes with volume or activity.

Facility capacity: The maximum practical output or flow that a facility can handle.

Resilience: The ability of a network to withstand, adapt to, and recover from disruption.

Scenario analysis: Evaluation of performance under alternative future conditions.

Expected loss: The probability-weighted consequence of a risk event.

Real option: The value of preserving flexibility to make a future investment or operating choice.

Geographic information system: A digital system for analysing spatial and geographic data.

Regionalization: Designing supply activities around regional markets rather than relying exclusively on distant global sources.

Incentive: A financial, tax, land, or infrastructure benefit offered to attract investment.

Service territory: The set of customers assigned to a particular facility.

Network node: A facility or entity at which materials, products, or information are received, transformed, stored, or dispatched.

Review Questions

Short Answer Questions

1. How does facility location differ from internal plant layout?

2. What is the purpose of a factor-rating method?

3. State the formula used in load–distance analysis.

4. Why is the centre-of-gravity result only an initial estimate?

5. Define resilience in the context of network design.

Long Answer Questions

1. Explain the global, regional, and site levels of facility-location analysis.

2. Compare factor rating, break-even analysis, and optimization as location methods.

3. Discuss how labour, infrastructure, incentives, and regulation influence location decisions.

4. Explain why total landed cost is more useful than factory conversion cost alone.

5. Discuss the role of scenario analysis and real-options thinking in uncertain networks.

Analytical / Application Questions

1. A company has three candidate sites with different fixed and variable costs. Explain how you would determine the preferred site at low, medium, and high demand volumes.

2. A proposed warehouse has the lowest load–distance score but lies in a flood-prone area with one major access road. Evaluate whether it should be selected and identify additional analyses required.

3. Design a decision framework for a manufacturer choosing between a low-cost offshore plant, a regional plant, and a flexible contract-manufacturing arrangement.

22. Multiple Choice Questions with Answers

Question 1

Which statement best distinguishes facility location from plant layout?

A. Location concerns internal equipment arrangement, while layout concerns country selection.

B. Location concerns geographic placement, while layout concerns arrangement within the facility.

C. Location concerns only labour cost, while layout concerns only transport cost.

D. Location is short term, while layout is always permanent.

Correct Answer: B. Location concerns geographic placement, while layout concerns arrangement within the facility.

Question 2

A location has a factor-rating score that is highest only when labour receives a very high weight. What should managers do first?

A. Accept the result without further review.

B. Discard all qualitative factors.

C. Perform sensitivity analysis on the weights and ratings.

D. Replace the method with centre-of-gravity analysis only.

Correct Answer: C. Perform sensitivity analysis on the weights and ratings.

Question 3

In load–distance analysis, a shipment of 500 units moved 20 kilometres contributes what value to the basic measure?

A. 25

B. 480

C. 520

D. 10,000

Correct Answer: D. 10,000

Question 4

What is the principal purpose of the centre-of-gravity method?

A. To calculate employee productivity.

B. To provide an initial geographic estimate based on weighted demand or supply points.

C. To determine environmental permit requirements.

D. To calculate the exact break-even volume.

Correct Answer: B. To provide an initial geographic estimate based on weighted demand or supply points.

Question 5

A facility has high fixed cost but low variable cost compared with another site. It is most likely to be preferred when:

A. Expected volume is high.

B. Expected volume is zero.

C. Demand is guaranteed to remain very low.

D. Variable cost is irrelevant.

Correct Answer: A. Expected volume is high.

Question 6

Which item is most appropriate for a network optimization model?

A. Only the manager’s personal preference.

B. Facility-opening decisions, shipment flows, capacity, and demand constraints.

C. Only the colour of facility buildings.

D. A single unweighted qualitative opinion.

Correct Answer: B. Facility-opening decisions, shipment flows, capacity, and demand constraints.

Question 7

Why might a slightly more expensive network be strategically superior?

A. It has no transportation requirements.

B. It may provide better resilience and alternate routes under disruption.

C. It eliminates the need for demand forecasting.

D. It guarantees that all risks disappear.

Correct Answer: B. It may provide better resilience and alternate routes under disruption.

Question 8

Which statement about government incentives is most accurate?

A. They should always dominate the location decision.

B. They are irrelevant to total cost.

C. Their duration, conditions, reliability, and clawback provisions should be evaluated.

D. They permanently eliminate operating costs.

Correct Answer: C. Their duration, conditions, reliability, and clawback provisions should be evaluated.

Question 9

A calculated centre-of-gravity point falls in a protected wetland. The most appropriate response is to:

A. Build there because the formula is mathematically optimal.

B. Ignore geographic constraints and use straight-line distance.

C. Use the result as a starting point and search for feasible nearby sites.

D. Cancel all network analysis.

Correct Answer: C. Use the result as a starting point and search for feasible nearby sites.

Question 10

What is a major governance risk when using AI for location analysis?

A. It always produces too few alternatives.

B. Its output may reflect biased data or hidden assumptions and therefore requires human review.

C. It cannot process geographic information.

D. It makes scenario analysis impossible.

Correct Answer: B. Its output may reflect biased data or hidden assumptions and therefore requires human review.