Service, demand and freight
Design the timetable before the concrete
Infrastructure is useful only when trains, feeders, fares, rolling stock and maintenance work as one operating system. The patterns below are tests for sizing the railway — not promised timetables or forecasts.
Direct trains ≠ direct tracks
A passenger should not have to change trains simply because the infrastructure corridor ID changes. Where capacity, rolling stock and operating rules allow, one train can continue across several corridors. A buildable railway still has to compete on door-to-door time and whole-life cost.
Where tracks can realistically exist
Where one train can continue
Whether the journey is competitive
Whether construction is justified
Journey-Time Gate
Each proposed passenger service needs an auditable timetable model. Separate train movement, station dwell and border processing; corridor IDs do not imply transfers. The Yerevan–Gyumri 1:35–1:50 range is a 2052 performance test subject to infrastructure modelling. Vanadzor, Dilijan and Ijevan remain service-design targets requiring engineering validation. Southern rail times are unknown until the regional route and border model are defined.
| Input | Required evidence |
|---|---|
| Distance | Surveyed or modelled kilometres |
| Gradient and curves | Ruling gradient and curve radii |
| Line speed | Realistic segment speeds |
| Running time | Pure train movement |
| Station dwell | Minutes at every stop |
| Border dwell | Processing and crossing time, separately |
| Scheduled time | Running + all dwell + recovery allowance |
| Frequency | Paths, turnback and maintenance capacity |
| Transfers | Walking, waiting and missed-connection penalty |
| Road / coach comparator | Same origin, destination, season and access time |
| Decision | Rail, coach, feeder, redesign or defer |
Operating concept
Service pattern to test
Each stage must open as a useful stand-alone service. Express and regional trains share the trunk; lower-demand extensions proceed only after reliability and ridership tests.
| Horizon | Corridor | Frequency test | Stopping pattern | Must be proven |
|---|---|---|---|---|
| 2035 | R1 Yerevan–Gyumri | Hourly all day; stronger peak if demand supports it | Fast regional with limited stops plus selected local trains | Rebuilt running time, passing capacity, four-season reliability |
| 2035 | R1 Vanadzor–Ayrum | Test two-hourly to Vanadzor and at least four daily pairs farther north | Regional stopping pattern | Georgia paths, domestic demand and maintenance windows |
| 2035 | R2 Yerevan–Sevan | Two-hourly base; hourly seasonal peaks to be tested | Regional, with express overlays in summer | Year-round demand and winter performance |
| 2035 | R4 Yerevan–Yeraskh | Two-hourly base with commuter peaks where justified | Regional stopping pattern | Infrastructure condition and southern interface choice |
| 2038 | R3 Dilijan–Ijevan | Two-hourly service test after corridor rehabilitation | Through Yerevan trains only if capacity and reliability allow | Structures, slopes, journey time and bus comparison |
Worked example, not a fleet order: a 100-minute Yerevan–Gyumri run plus 20 minutes to turn at each end creates a four-hour cycle. An hourly service therefore needs four trainsets in traffic and normally one maintenance spare. Every line must repeat this calculation with the final timetable, gradients and maintenance plan.
National demand model
Three cases, one transparent method
The model should publish assumptions and results by station, season and trip purpose. International flows are reported separately so they cannot make a weak domestic section appear viable.
| Case | What it tests | External links | Use in decisions |
|---|---|---|---|
| Low | Weaker population and trip growth, lower rail capture and slower project delivery | No unsigned border traffic | Downside affordability and minimum-service test |
| Base | Observed domestic travel plus the service and feeder improvements funded in the programme | Only contracted and operational flows | Primary appraisal case; still not a promise |
| High | Stronger tourism, land use around stations and proven mode shift | External flows shown as separate modules | Capacity and expansion test, never the sole case |
Freight business case
Do not count tonnes without a shipper
Any multi-million-tonne strategic ambition is not a forecast. Each commodity needs an origin, destination, loading method, price, contract path and alternative road cost.
| Market | Evidence to obtain | Terminal implication | Treatment now |
|---|---|---|---|
| Mining and metals | Named shipper, mine life, annual minimum, wagon type and road comparison | Separate loading loops; avoid passenger platforms | No volume in base case without a contract path |
| Construction materials | Quarry/source, receiver, seasonality and terminal handling cost | Simple bulk sidings near demand | Domestic test market |
| Food and beverages | Producer clusters, cold-chain need, batch size and delivery window | Consolidation and temperature-controlled handling | Pilot before major terminal investment |
| Containers and general cargo | Origin–destination survey, border dwell time, port route and price | Masis/Gyumri intermodal capacity | Use observed flows only |
| International transit | Binding access rules, train paths, customs, insurance and shipper commitments | Border and gauge/interface facilities | Separate upside module; excluded from base case |
One public-transport product
Ticketing, feeders and accessibility
A rail map alone does not create access. Connections must be contractually protected and affordable for daily users as well as visitors.
- One ticket and journey planner for train, guaranteed feeder and local transport
- Feeder departure five to ten minutes after the train, with a hold rule for reasonable delay
- A fare-affordability test by income group and explicit public-service compensation where needed
- Step-free routes, tactile guidance, audible and visual information, accessible toilets and assistance booking
- Bicycle and luggage space on tourism routes and safe walking access at every station
- Performance contracts covering cancellations, missed connections, cleanliness and passenger information
Traction energy
Electrification needs its own power study
No megawatt figure is credible before the timetable, train mass, gradients and electrical system are fixed. The study must compare present-network compatibility with the efficiency of 25 kV only for a separately justified new package.
| Workstream | Question to answer | Required output |
|---|---|---|
| Traction simulation | What peak and annual energy follows from the service plan? | Train-performance and load-flow model by section |
| System choice | Where should 3 kV DC remain, and can 25 kV AC lower whole-life cost of a justified new package? | Option appraisal including dual-system fleet and interface |
| Grid connections | Can candidate substations be supplied without disproportionate network reinforcement? | Utility offers, connection cost, land and programme |
| Resilience | What continues during a grid fault, tunnel incident or communications loss? | N-1 supply for critical systems, backup power and recovery plan |
| Efficiency | How much braking energy can be reused or stored? | Metered regenerative-braking case and energy KPI |
Numbers that remain intentionally blank: annual passengers, fare revenue, subsidy, freight tonnes by corridor, total fleet and traction demand. Filling them before the national model would create false precision.