Validated with three years of actual operating, maintenance, and procurement records from a fleet of ten 6,944 TEU container vessels.
01 | Research Positioning
From single-vessel experience-based purchasing to fleet-level data-driven decisions.
This research is not about whether one item was bought in excess or shortage. It addresses an integrated management problem across fleet reliability, maintenance efficiency, home-port replenishment, and procurement cost.
Thesis Topic Overview
When a shipping company relies only on requests from individual vessels or maintenance engineers, headquarters has limited visibility into future fleet-wide demand. This research proposes a calculable, traceable, and updatable method to help managers forecast maintenance spare-parts demand in advance and decide when to procure and how much to procure.
Questions This Research Answers
- How can future spare-parts demand for each vessel be forecast?
- How can demand for reconditioned parts be separated from demand for new parts?
- How can vessel-level demand be consolidated into weekly fleet demand?
- How can the optimal procurement lot size for new parts be determined?
Case Item
The study uses main-engine fuel valves as the case item because their demand is closely linked to engine running hours, scheduled maintenance, reconditioning, and scrap rates, making them suitable for validating fleet spare-parts demand forecasting and procurement planning.
Case Data Scope in the Thesis
Chapter 4 of the thesis uses a case fleet of ten 6,944 TEU container vessels, with main-engine operating records, fuel-valve maintenance records, and procurement data from 1 January 2018 to 31 December 2020. Rather than making a one-time cost comparison, the study organizes daily operation, maintenance intervals, voyage-week timing, home-port replenishment, and reconditioning outcomes into a demand model that can be recalculated repeatedly.
02 | Research Lens
Viewing the study through PDCA: spare-parts management is a continuously calibrated decision cycle.
This lens helps shipowners and ship managers see that the thesis is not just a one-off procurement calculation, but a fleet spare-parts management process that can be updated, checked, and improved repeatedly.
Analyze consumption patterns and forecast long-term spare-parts demand
Organize historical vessel operating records, fuel-valve maintenance records, route cycles, and reconditioned-part scrap rates to estimate future spare-parts demand when each vessel returns to home port.
Output: weekly fleet demand, new-part demand, and reconditioned-part demand.Execute long-term procurement planning and supplier negotiation
Use MRP and the Silver-Meal method to plan procurement lot sizes and ordering timing, while using long-term demand as the basis for supplier negotiation, delivery coordination, and spare-parts replenishment.
Output: procurement lot size, ordering cycle, and supplier agreement.Monitor spare-parts consumption, usage, and inventory status
Use MPDC and ERP inventory ledgers to track the movement of new parts, serviceable reconditioned parts, and scrapped parts, checking whether actual consumption aligns with the forecast and whether onboard safety stock and service levels are maintained.
Output: consumption deviation, inventory status, and demand fill rate.Analyze execution deviations and correct the demand-forecasting model
Feed actual consumption, scrap-rate changes, supplier lead time, and vessel maintenance conditions back into the model to adjust the next spare-parts demand forecast and procurement plan.
Output: corrected forecast, improvement rules, and the next planning cycle.03 | Practical Pain Points
The difficulty of vessel spare-parts management comes from the uncertainty of maritime operations.
Unlike land-based factories, ships cannot be replenished at any time. Spare parts usually have to be arranged around port calls, home-port returns, maintenance cycles, and reconditioning outcomes.
Shortage risk
Insufficient spare parts can affect vessel maintenance, equipment operation, and route stability; shortage risk eventually becomes operational risk.
Inventory cost
If each vessel buys conservatively and over-stocks, onboard and shore-side inventory accumulates, tying up capital and increasing holding costs.
Limited visibility
When demand information is fragmented across vessels and engineers, headquarters struggles to conduct fleet-level procurement planning, cost control, and supplier coordination.
Four management challenges identified in Chapter 1
- Replenishment of new and reconditioned parts is managed separately by individual maintenance engineers rather than integrated at fleet level, making it easy to miss home-port schedules and incur extra recovery costs.
- Demand for reconditioned parts is difficult to estimate accurately because fuel-valve replacement can be affected by early or delayed maintenance, poor fuel quality, and quality differences between new and reconditioned parts.
- Different route cycles and home-port timings make it difficult to consolidate weekly demand for new and reconditioned parts; scrap decisions for reconditioned parts also require workshop confirmation, further complicating new-part demand.
- New-part procurement lacks a cost-evaluation model, so frontline crew may increase request quantities out of shortage concerns, leading to excess inventory, duplicated ordering, and higher holding costs.
04 | Research Method
A three-stage framework turns data into an executable procurement plan.
The study starts from each vessel's operating and maintenance data, consolidates it into weekly fleet demand, and then applies dynamic lot-sizing to plan when and how many new parts to procure.
Demand forecasting
Estimate each vessel's likely maintenance spare-parts demand at future home-port calls using main-engine running hours and fuel-valve maintenance records.
Fleet consolidation
Use MRP to consolidate different vessels, route cycles, and home-port return timings into overall weekly fleet demand.
Procurement planning
Apply the Silver-Meal heuristic to estimate a more reasonable procurement lot size by balancing purchase cost, ordering cost, and inventory cost.
Calculation Details from Raw Data to Procurement Decisions
Chapter 4 breaks the research process into executable procedures: first, calculate fuel-valve maintenance interval hours from the difference in main-engine running hours between two maintenance events for each cylinder on each vessel; then combine daily engine running hours with route cycles to estimate the number of maintenance events and spare-parts demand before each vessel returns to home port.
For procurement lot-sizing, the thesis uses weeks as the planning unit and sets the procurement lead time for new parts at 12 weeks. The reconditioning lead time for fuel valves can also be used to decide whether MPDC should reserve reconditioning capacity and safety stock in advance.
05 | Management Architecture
Use MPDC as the home-port hub for centralized spare-parts management.
MPDC places new parts, serviceable reconditioned parts, and scrapped parts into one logistics and inventory-ledger framework, allowing spare-parts movement and cost information to be managed together.
Onboard safety stock
Maintain essential repair capability
Each vessel still needs reasonable spare-parts stock to ensure repair safety and service level during voyages.
Centralized home-port replenishment
Reduce duplicated inventory holdings
Other spare parts are managed centrally by MPDC to improve sharing efficiency and reduce over-procurement.
ERP inventory-ledger synchronization
Track material flow and cost
Set up inventories for new parts, serviceable reconditioned parts, and scrapped parts to support auditing and continuous improvement.
Inventory-Ledger and Logistics Flow after MPDC Implementation
Chapter 3 recommends establishing clear ERP inventory ledgers so that material flows among vessels, MPDC, reconditioning workshops, and suppliers can be tracked.
The key design point is to turn “replenished onboard, sent for reconditioning, reconditioning completed, scrapped, and new parts received” into auditable transaction records.
06 | Empirical Results
The ideal model shows that past procurement was not well aligned with actual fleet demand.
The study validates the model using data from 1 January 2018 to 31 December 2020 and compares the ideal procurement model with historical procurement behavior.
Ideal Model vs. Actual Procurement Performance
Validated with three years of historical main-engine fuel-valve data from the case fleet of ten 6,944 TEU container vessels, the results are as follows:
Spare-parts procurement quantity
effectively reduced
Total cost
maximum saving
Ordering cost
maximum reduction
Inventory-cost difference
improved by 9600%
| Metric | Ideal model | Historical actual model | Relative visual | Management implication |
|---|---|---|---|---|
| Procurement quantity | 660 pcs | 943 pcs | Historical procurement quantity was clearly high | |
| Number of orders | 31 orders | 40 orders | The historical model involved more repeated orders | |
| Purchase cost | USD 198,000.00 | USD 282,900.00 | Over-procurement increased cost | |
| Ordering cost | USD 496.00 | USD 640.00 | More orders created additional cost | |
| Inventory cost | USD 145.29 | USD 14,092.84 | The historical model had high inventory waiting and holding cost | |
| Total cost | USD 198,641.29 | USD 297,632.84 | The total cost gap was significant |
Management implication: the issue is not simply whether spare parts are available. The real problem is that procurement quantity and order timing were not aligned with actual fleet demand, which amplified over-procurement and inventory cost.
Detailed Findings from the Empirical Data
Chapter 5 shows that fuel-valve maintenance intervals for the ten 6,944 TEU container vessels were generally concentrated around 800 to 900 hours, and daily main-engine running hours were around 16 hours. However, historical procurement quantities differed significantly, indicating that procurement behavior did not necessarily match actual equipment operating demand.
Sensitivity analysis also shows that when fuel-valve maintenance interval hours (RHPM) fall below 80% of the historical level, total cost may change noticeably. RHPM is therefore not only a model parameter but also a monitoring indicator for fleet maintenance quality and cost risk.
07 | Practical Value
This model makes spare-parts management more predictable and controllable.
For shipowners, ship managers, fleet technical managers, and procurement managers, the value of the study is that it turns experience-based replenishment into an auditable management cycle.
Integrated fleet management
Allows headquarters to understand future fleet-wide demand and reduce duplicated vessel-level procurement and information fragmentation.
Data-driven demand forecasting
Uses engine running hours, maintenance records, route cycles, and scrap rates to estimate demand instead of relying only on subjective requests.
Synchronize new and reconditioned parts
Manages parts sent for reconditioning, successfully reconditioned parts, scrap decisions, and new-part replenishment at the same time.
Procurement management cycle
Can connect purchase requests, ordering reminders, delivery checks, quality acceptance, and PDCA improvement.
Balance cost and service level
Supports a more robust decision balance among shortage risk, maintenance demand fill rate, and total cost.
Integrated fleet management
Data-driven demand forecasting
Centralized home-port replenishment
MRP
Silver-Meal Heuristic
MPDC
Procurement and inventory cost control
Operational risk management
Management Decisions Supported after Implementation
The thesis concludes that the model's value is not only in calculating a lower total cost. It also enables managers to understand procurement plans and equipment operating conditions in the short term, and to estimate overall cost and supply risk in the long term. The forecast data can also support procurement negotiations with suppliers, capacity reservation, and delivery scheduling.
08 | Extensions
Potential Extensions
The study uses fuel valves as the validation case, but the methodology can be extended to broader vessel equipment and operating scenarios.
Integrating other main-engine related spares
Equipment such as exhaust valves and pistons has similar maintenance patterns, but longer maintenance intervals, higher spare-parts cost, and in some cases no reconditioning option. Integrating demand forecasting for related main-engine spares can greatly improve cost-control precision.
Generator spare-parts management
Container vessels are equipped with three to five generators that share internal-combustion characteristics with the main engine. If the complexity of multi-unit parallel operation can be addressed, this would fill a gap in managing the second-largest equipment cost area onboard.
Route and schedule transition analysis
In real operations, route assignments change frequently and schedules are significantly affected by port congestion. Future work should study how route changes affect main-engine running hours and home-port replenishment planning, improving the model's dynamic adaptability.
Automated monitoring and dynamic adjustment
Continuously and automatically feeding maintenance records and main-engine operating data into the calculation model would enable real-time monitoring and dynamic adjustment, reducing manual intervention and process cost.
Conclusion
Vessel spare-parts management is not just a procurement issue; it is an integrated management issue involving fleet reliability, maintenance efficiency, and operating cost control.
By using data and modeling, this research makes fleet spare-parts demand more predictable, home-port replenishment and procurement planning more controllable, and helps shipowners and ship management companies maintain operational safety at lower cost.
In one sentence: use vessel operating and maintenance data to forecast maintenance spare-parts demand across the fleet, and reduce shortage risk, over-procurement, and inventory cost through centralized home-port management and procurement lot-sizing.