Service continuity in Swedish grocery distribution under disruptions - Evaluating microbuffers across disruption scenarios

Abstract

Sweden’s grocery distribution is routed through a limited number of upstream hubs and relies on corridors and ferry links. When disruptions affect a hub or a critical link, reallocations propagate fast and travel times increase, with some areas losing coverage temporarily. Decentralising the whole system is costly, so targeted resilience options must be evaluated.

The purpose of this thesis is to assess whether strategically placed microbuffers can help maintain acceptable access to staple groceries in Sweden under disruption scenarios affecting supply nodes and critical transport corridors.

To address this, the thesis builds a travel-time network representation that combines road and ferry links and uses locality-level demand from official settlement statistics. Distribution supply is represented through a simplified set of regional supply nodes obtained by aggregating demand locations, enabling nationwide scenario evaluation without relying on confidential facility data. Disruptions cover facility outages, critical-corridor slowdowns, and partial capacity degradation. Candidate microbuffer hosts are generated around the most affected areas, and a fixed installed-capacity budget is allocated under a recourse policy. Site portfolios are selected using two heuristic decision rules: a continuity-first lexicographic rule and a compensatory weighted-sum rule used to explore trade-offs and test methodological robustness. Performance is assessed across scenarios using threshold-based service continuity and transport-effort indicators, supported by tabular exports and Geographic Information System (GIS) visualisation.

Across the evaluated disruption scenarios, strategically placed microbuffers improve service continuity under the threshold definition and reduce the exposure of the most affected localities. The weighted-sum exploration confirms that similar continuity improvements can be achieved under a compensatory decision rule, suggesting that the main conclusions do not depend on a single selection logic. The lexicographic solution is kept as the reference because it implements a non-compensatory continuity-first policy without requiring weight calibration. Sensitivity tests indicate that smaller capacity units improve spatial coverage under a fixed installed-capacity budget, whereas larger units increase tail risk.

Popular Abstract

Supermarket shelves look stable until they don't. A cancelled ferry, a road slowdown, or an outage at a major distribution hub can force extreme detours and leave some areas with severely degraded access to supply. Sweden's grocery network is efficient precisely because it is concentrated: high volumes flow through a small number of key hubs and corridors. That concentration is its strength. In a crisis, it can become its weakness.

One way to limit these failures is to add microbuffers, small facilities placed in advance to store emergency inventory and support replenishment when normal flows break. They are not consumer pickup points; they are a logistics backup layer, quietly waiting until needed.

In disruptions, resilience is not a feeling, it needs a rule that can be tested. Here, it is measured as service continuity: if the travel time to reach workable supply exceeds 240 minutes under disruption, the situation is treated as unacceptable. That threshold is a severe stress test, not a service promise. It is chosen so that exceedances point to genuine remote failures, rather than labelling large parts of Sweden as failed even without disruptions.

The study models travel times on a national road-and-ferry network covering 2,017 Swedish localities, from dense cities to remote northern regions where detours can become extreme. The supply side is simplified. It reflects a centralised structure without claiming to reproduce a specific retailer network. Disruption scenarios target critical hubs and corridors, and microbuffers are assessed as a limited backup layer, not a redesign of the whole system.

Selecting buffer locations nationwide is a large-scale siting problem, so the study uses a heuristic search that builds a solution step by step. Two decision logics are compared: one that prioritises continuity, and one that balances continuity against transport effort. Transport effort here is a proxy for truck time and detours. This reveals a real choice, between protecting continuity and keeping transport effort manageable.

Across the tested disruption cases, microbuffers reduce how often localities cross the threshold. The biggest gains consistently appear in the same places: bottlenecks, sparse northern regions, and ferry-dependent areas like Gotland. These locations keep reappearing as valuable across very different disruption types, which makes them strong candidates for contingency planning.

The practical output is a shortlist of geographically valuable sites across disruption types. It is a planning input, not a guarantee, and a reminder that in a concentrated network, a little targeted preparation goes a long way.

• author: Ollé Puig, Pau
• supervisor: Sandeep Jagtap
• organization: Engineering Logistics; Production Management
• course: MTTM02 20252
• year: 2026
• type: H2 - Master's Degree (Two Years)
• subject: Technology and Engineering
• keywords: Swedish grocery distribution, service continuity, disruption scenarios, microbuffers, supply chain resilience, facility location, heuristic optimization, contingency planning
• other publication id: 6054
• language: English
• id: 9223999
• date added to LUP: 2026-03-11 17:38:29
• date last changed: 2026-03-11 17:38:29