Evidence, method and full report

Show what is known, assumed and still missing

This page keeps the long-form concept available for audit. The report and thematic pages use the revised four-corridor architecture; earlier totals are archive-only comparators.

Reading discipline

Four labels prevent false certainty

A public concept becomes misleading when a planning target looks like an observed fact or a map line looks like an approved route.

Label	Meaning	Permitted use	Example
Verified fact	Supported by a dated source or observed dataset	Describe the baseline with citation and definition	An existing railway asset or published annual total
Planning target	A measurable service or policy outcome to test	Size options and define success; never report as achieved	Hourly Yerevan–Gyumri service test
Concept assumption	A provisional input used to explore feasibility	Show the range and sensitivity	Stage-level capital allowance in constant 2026 USD
Decision gate	Evidence required before a commitment	Select, redesign, defer or stop	Gate S compares multimodal southern access

Arithmetic and claim audit

What was corrected — and what remains open

The project reconciles its own totals but does not pretend that top-down screening allowances are tender estimates.

Item	Current treatment	Check	Next evidence
Architecture	Four core corridors + conditional access studies	No selected internal mountain trunk	Engineering and mode appraisal
Costs	Revised national total unknown	Earlier $18.95–29.30bn sum archived	Four-bucket bottom-up pricing
Operations / workforce	Earlier allowances retired	Not transferable to revised scope	Service, asset and package models
Journey times	Southern rail figures removed	Gyumri range is a performance test only	Journey-Time Gate
Evidence	Official statements attributed and dated	Proposal, ratification and operation are separate statuses	Published instruments and operator timetables

Version rule: figures may be revised when official datasets, surveyed quantities or agreements change. Every future release should keep a dated assumption register and a visible change log.

Priority evidence

Use primary and institutional sources first

The full register appears in Chapter 15. These sources are especially important for baseline policy, climate resilience, delivery capability and live regional dependencies.

Theme	Source	How it is used
Preparation and institutions	ADB evaluation of Armenia’s North–South Road Corridor ↗	Lessons on master planning, design standards, procurement, safeguards and client capacity
Climate and resilience	World Bank Armenia Country Climate and Development Report ↗	National hazard, resilience, energy-security and adaptation context
Transport policy	IEA Sustainable Transport Policy for Armenia ↗	Current sustainable-mobility policy context and mode integration
Baseline statistics	Statistical Committee of Armenia ↗	Definitions and national passenger/freight context; every series must retain its original scope
Railway baseline	South Caucasus Railway ↗	Existing services and operator-reported railway information
Prices and currency	Central Bank of Armenia exchange-rate archive ↗	Dated conversion only; the project’s analytical base remains constant 2026 USD
Regional interfaces	Armenian MFA: TRIPP framework ↗ Prime Minister of Armenia: railway interview ↗ Türkiye MFA: Kars–Gyumri working group ↗	Live context only; no border opening, traffic or financing is assumed in the domestic base case

Long-form record

Complete concept report

The 15-part report below uses the revised architecture, conditional delivery windows and a reset cost basis.

Executive summary

Current architecture · 9 September 2026. Independent pre-feasibility national mobility concept. Existing and restorable rail comes first: four core corridors, official regional interfaces and conditional access studies. Access needs do not preselect a railway. This is not an approved government programme.

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.

Infrastructure

Where tracks can realistically exist

Service

Where one train can continue

Time

Whether the journey is competitive

Gate

Whether construction is justified

2027–2031

Evidence first

Official GIS/DEM, asset and safety audit, demand model, preliminary profiles and engineering / hazard screening; Gate 0 and Gate S evidence

2030–2038

Fix the backbone

Justified R1–R4 renewal, stations and measured service improvements; R3 restoration only after engineering and demand validation

2030s–2040s

Open interfaces conditionally

Gyumri–Kars, Yeraskh–Nakhchivan and Nrnadzor–Agarak depend on design, agreements, border operations and funding; C1 needs revalidation

2040s–2052

Build only passed gates

Access studies may become rail, coach, feeder or road improvements, or be deferred; no fixed Kapan or Meghri railway completion year

Baseline and the access problem

2.1 What exists today

IEA describes Armenia's railway network as about 782 km and fully electrified at 3 kV DC. Armstat series use different scopes — general-purpose and total operated length — so values in the 700–800 km range should not be mixed without definitions. Secondary sources also quote 759.8 km of main lines, 693.7 km operational and 1,252.5 km including additional or abandoned track; these are not a correction unless route-km, track-km and asset scope are reconciled against the operator's current register. The main active corridor links Yerevan with Gyumri, Vanadzor, the northern border and Georgia; several other corridors are lightly used, seasonal or inactive.

Corridor	Current status	Planning implication
Yerevan–Gyumri–Vanadzor–Ayrum	Operating	Main passenger and international axis; frequency, journey time and asset condition require improvement.
Yerevan–Hrazdan–Sevan–Shorzha	Limited / seasonal service	A scenic route whose usefulness is constrained by infrequent service and long journeys.
Shorzha–Sotk	Infrastructure exists near a sensitive international border	No stable, regular passenger product; treat Sotk as a domestic terminus and require a security-and-demand gate.
Hrazdan–Dilijan–Ijevan	Not fully operational	Requires full survey of track, slopes, structures, power supply and stations before scope is fixed.
Yerevan–Artashat–Ararat–Yeraskh	Limited service	A strategically useful southern starting point; service and infrastructure need stabilising.
Gyumri–Akhuryan/Akhurik–Türkiye border	Existing branch; a bilateral rehabilitation working group met in April 2026	Prepare the Armenian asset case, but open only under binding border, customs, safety and operating agreements.
Yeraskh / Meghri east–west interfaces	TRIPP planning is active; the direct cross-border railway is not yet a completed operating route	Coordinate land and junctions with the national network, but keep uncontracted direct transit outside the investment case.

Baseline references: IEA, Armstat and South Caucasus Railway. Route condition must be re-audited before procurement; labels are planning categories, not engineering acceptance.

2.2 Why Gyumri is only formally accessible

Gyumri already has a railway; it does not need a duplicate alignment. The problem is the product: ordinary journeys are roughly three hours, faster services remain limited, and frequency does not support spontaneous daily travel. The project target is 1h35–1h50, hourly daytime departures and coordinated onward connections to Vanadzor, Lori and Georgia.

2.3 Why one cost per kilometre is misleading

Renewing track across the Ararat plain and constructing a new railway through Zangezur are different classes of work. In the plain, cost is driven by rail, bridges, crossings, power and signalling. In Syunik it is driven by tunnels, viaducts, seismic resistance, groundwater, landslides, portals, construction access and disposal of millions of cubic metres of spoil.

Reference project	Scope	Published reference
Southern Armenia Railway concept, 2013	About 305 km, 102.3 km of tunnels and 84 bridges	About $3.5bn in 2013 prices
Pap–Angren, Uzbekistan	123 km, including a 19.2 km tunnel and electrification	About $1.63–1.73bn in mid-2010s reporting
Sisian–Kajaran road programme	About 60 km with nine tunnels and 27 bridges in the reference design	About €1.35bn programme reference

Historical railway and road proposals provide context only. Their cost estimates are not transferred into the revised programme; road contracts do not establish railway unit costs.

Goals and measurable outcomes

3.1 Transport goals

Metric	Target
Coverage	All ten marzes included in the national access plan; each regional capital served by direct rail or a guaranteed timed feeder by 2052.
Accessibility	Planning target: 85% of the population within 45 minutes of a station or guaranteed feeder; to be tested in a national GIS model.
Frequency	Network design target: at least one train every 60 minutes on principal axes and every 120 minutes on regional routes; uncertain branches may start with a lower test service.
Reliability	At least 92% of trains arrive within five minutes; year-round mountain operations.
Passengers	No annual total is set before the national origin–destination model. Publish low, base and high cases by line, year and service pattern.
Freight	No corridor volume is admitted without a defined commodity, shipper, origin–destination pair, price and operating path. International transit remains a separate upside module.
Tourism	At least 20 major destinations accessible with one timed shuttle after the train.

3.2 What is not a goal

A 250–300 km/h high-speed railway
A station immediately beside every monastery or natural landmark
Full recovery of passenger-network costs through fares
International links built ahead of legal agreements with neighbouring states
Replacement of every intercity bus: buses become feeders and complementary routes

Coverage rule. A marz counts as served only through a direct station or an interchange with a guaranteed timed feeder. A protected future corridor is an option, not delivered access. The interactive maps show roads and feeder links so an area without a coloured railway is not being declared unreachable or unbuildable.

Architecture of the national network

Corridor / group	Geography	Planning status
R1 · Northern	Yerevan–Armavir–Gyumri–Vanadzor–Ayrum	Core · existing / modernisation
R2 · Sevan	Yerevan–Hrazdan–Sevan–Shorzha; Sotk separately gated	Core · existing / modernisation
R3 · Tavush	Hrazdan–Fioletovo–Dilijan–Ijevan	Core · historic / restoration after survey
R4 · Ararat	Yerevan–Masis–Artashat–Yeraskh	Core · existing / modernisation
Regional interfaces	Gyumri–Akhurik / Kars; Yeraskh–Nakhchivan; Nrnadzor–Agarak / TRIPP	Official processes · design, agreements, funding and operations remain separate gates
C1	Fioletovo–Vanadzor	Previously studied connector · revalidation required
Access studies	Aragats / former R7; Ashotsk–Bavra / Akhalkalaki; north-eastern Tavush; Bazum crossing and Lori plateau	Mode undecided · rail, coach, feeder, road or defer
Earlier mountain concepts	Vardenyats / former S3; internal mountain railway to Meghri	Comparator only · no-build default; no alignment selected

South Armenia Access Gate · Gate S

What combination of regional rail reopening, road investment, direct coach and feeder transport gives Vayots Dzor and Syunik the best reliable access? No new domestic mountain railway is an explicit valid outcome.

Topography and preliminary alignment
Gradients, curve radii and realistic operating speeds
Tunnel and viaduct quantities, geology and seismic safety
Competitive end-to-end journey time
Passenger and contracted freight demand
Lifecycle finance, subsidy and affordability
Environment, climate resilience and emergency access
Comparison with road, coach, regional reopening and no new domestic mountain railway

The north–south route via Nakhchivan is the Armenian Government’s stated position of 11 March 2026, informed by mountainous terrain. It is a policy position, not this project’s independent proof that every domestic alignment is physically impossible. Cross-border service still requires agreements, infrastructure, access rights and a workable border process.

Primary source · 11 March 2026

C1 is a previously studied connector requiring a new alignment, demand, journey-time and cost case. The Bazum crossing and Lori plateau access are separate engineering and mode studies. Vardenyats and the former internal mountain railway to Meghri have no-build default.

Conditional delivery windows

The earlier six-stage automatic build-out is retired. These overlapping windows describe evidence and conditional deliverables, not construction promises. Each funded section must deliver independent value if later work is deferred.

2027–2031

Evidence first

Official GIS/DEM, asset and safety audit, demand model, preliminary profiles and engineering / hazard screening; Gate 0 and Gate S evidence

2030–2038

Fix the backbone

Justified R1–R4 renewal, stations and measured service improvements; R3 restoration only after engineering and demand validation

2030s–2040s

Open interfaces conditionally

Gyumri–Kars, Yeraskh–Nakhchivan and Nrnadzor–Agarak depend on design, agreements, border operations and funding; C1 needs revalidation

2040s–2052

Build only passed gates

Access studies may become rail, coach, feeder or road improvements, or be deferred; no fixed Kapan or Meghri railway completion year

Access to regions and destinations

All ten marzes and Yerevan remain access priorities. National access may combine direct rail, express coach and guaranteed feeders. Jermuk, Sisian, Goris, Kapan and Meghri are access objectives, not guaranteed trunk stations. Railway access through Nakhchivan is a regional scenario with independent legal and operational conditions.

A protected connection requires a coordinated timetable, one ticket, accessible interchange and a funded waiting / recovery rule. Test actual walking and waiting times; do not publish unsupported last-mile minutes. See Network & regions for all eleven access passports.

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

Technical concept

All proposed speeds, gradients, curve radii, axle loads and platform dimensions below are screening assumptions to test, not approved standards or surveyed design. No new mountain corridor is selected.

Parameter	Proposed basis for comparison
Gauge	1,520 mm on the national network; transfer facilities for 1,435 mm at relevant borders.
Track layout	Single track with passing-loop spacing determined by timetable and capacity modelling; selective double track on busy urban and suburban sections.
Electrification	Existing network at 3 kV DC; compare 25 kV AC with 3 kV DC for long new mountain corridors.
Rolling stock	Fleet and any traction-change interface to follow the selected electrical system, timetable and accessibility requirements; no interface site is preselected.
Speed	Test 100–140 km/h on modernised routes and 80–100 km/h in mountain sections; final operating speeds depend on geometry, signalling and stopping pattern.
Gradient	Target no more than 18‰ on through-freight sections; test up to 25‰ only with a train-performance and operating case. Steeper exceptions require a passenger-only or special operating basis.
Curves	Prefer 800 m radius or greater; 500 m minimum in constrained terrain where the operating case remains acceptable.
Axle load	Use 22.5 tonnes as the freight-interoperability baseline; test 23.5 tonnes only where the heavy-freight case and connecting infrastructure justify it.
Platforms	Standardised height and step-free access; normally 160–220 m at principal hubs and 90–120 m on low-volume branches, with protected land for extension.
Signalling	Performance-based digital control: ETCS Level 2 or compatible trunk technology where capacity and interoperability justify it; simpler fail-safe digital block on low-volume branches.
Communications	Route fibre, independent backup paths and protected tunnel radio.

7.1 Tunnels and seismic design

Avoid active faults; cross one only with a route-specific seismotectonic justification
Provide a parallel evacuation gallery or second bore for long tunnels, based on traffic and quantified risk
Place portals outside landslide, avalanche and rockfall zones wherever practicable
Design ventilation, drainage, fire zones, communications and rescue access together with the civil alignment
Use replaceable bridge bearings and structural-health sensors on critical assets
Create a digital twin for critical sections: deformation, water, temperature and seismic events

7.2 Power-system option appraisal

The existing 3 kV DC network is the compatibility baseline. A 25 kV AC option is relevant only to separately justified new infrastructure. Compare power supply, rolling stock, interfaces and whole-life costs after scope and service are fixed.

7.3 Stations as transport hubs

Place stations within walking distance of the centre or serve them by a dedicated feeder corridor
Combine platform, bus stops, taxi, parking, cycle hire and visitor information in one compact interchange
Use a modest, efficient pavilion in smaller towns — not an expensive monumental station
Safeguard land around major stations for mixed-use development, without using property income to disguise a weak transport case
Separate freight yards from passenger circulation and residential neighbourhoods

Studies, design and permissions

Evidence programme, 2027–2031. Commission official GIS/DEM and asset / safety audits, a national demand model, preliminary longitudinal profiles, hazard screening and independent alternatives appraisal before selecting construction.

Scope geological and hydrogeological investigations from ground risk; determine borehole and test-adit locations after reconnaissance. Include seasonal ecology, water, seismic hazards, heritage, land and utility interfaces, consultation and applicable permits. Study budgets and durations must be procured for the revised scope, without reusing the retired mountain programme allowances.

Gate records must show the evidence owner, source date, uncertainty, rejected alternatives, consultation response and next decision. Material scope changes reopen the relevant environmental, land, cost and service assessments.

Construction schedule and workforce

2027–2031

Evidence first

Official GIS/DEM, asset and safety audit, demand model, preliminary profiles and engineering / hazard screening; Gate 0 and Gate S evidence

2030–2038

Fix the backbone

Justified R1–R4 renewal, stations and measured service improvements; R3 restoration only after engineering and demand validation

2030s–2040s

Open interfaces conditionally

Gyumri–Kars, Yeraskh–Nakhchivan and Nrnadzor–Agarak depend on design, agreements, border operations and funding; C1 needs revalidation

2040s–2052

Build only passed gates

Access studies may become rail, coach, feeder or road improvements, or be deferred; no fixed Kapan or Meghri railway completion year

The previous workforce and person-year totals were tied to a retired construction scope and are not estimates for this revision. Build a package-level workforce curve only after quantities, productivity, procurement method and sequence are known. Distinguish peak workers, annual full-time equivalents and total person-years. Training, local supplier participation and knowledge transfer remain objectives without invented percentages.

Capital cost and operations

The revised programme has no validated national total. Reprice the agreed scope in four buckets; do not subtract guessed savings from the earlier concept. Operating costs, revenue and subsidy also need a new service-based model.

Price first

Existing network

R1, R2 and R4; stations, signalling, power, structures and accessibility

Price by asset quantities and measured benefits

Price after survey

Restoration

R3 Hrazdan–Dilijan–Ijevan; slopes, tunnels, bridges and systems

Price after condition survey

Conditional

Conditional packages

C1, airport access and international interfaces

Price only after scope, geometry, operating model and funding responsibility are known

Study budget only

Undecided access

Aragats, Shirak plateau, north-eastern Tavush, Lori and southern access studies

Study budgets only; no assumed construction CAPEX

Earlier maximum full-build comparator — $19–29.3bn
Archived comparison only, not the revised programme budget. Its exact stage sum was $18.95–29.30bn in constant 2026 USD. The former construction scope is retired; this arithmetic does not validate quantities, risk allowances or affordability.

The cost model must identify the price base, quantities, unit rates, land, systems, vehicles, investigations and renewals. Record inflation, exchange rates, financing charges and contingencies explicitly, avoid double-counting, and publish annual real and nominal cash flows. P50/P80 labels require a documented probabilistic model.

Finance and governance

11.1 Recommended governance structure

Create a National Rail Development Agency as programme client and owner of all project data
Separate infrastructure management from passenger and freight operations, or establish transparent open access to the network
Before Stage 1, settle the legal interface with the existing concession, any TRIPP Development Company rights and the ownership and access regime for new lines
Protect non-discriminatory domestic passenger and freight access, Armenian control of safety and operational data, state emergency step-in rights and a workable termination mechanism in every infrastructure agreement
Form an independent engineering board for tunnels, seismic design, safety, environment and cost assurance
Publish base, central and stress-case cost ranges, scope changes, contracts, schedule and performance indicators every quarter; publish P50/P80 only after a documented quantitative risk model exists
Procure passenger service under long-term contracts specifying frequency, punctuality and accessibility

11.2 Funding by component

Component	Suggested model	Condition
Stages 0–2	State budget, grants and concessional IFI loans	High public value and limited direct commercial return.
Stations and development	PPP for commercial space and adjacent property	Do not transfer track, tunnel or core demand risk merely to keep debt off the public balance sheet.
Stages 3–4	Blended grants, concessional debt, budget and freight commitments	Proceed only after financial close for each independent operational section.
Freight terminals	Joint ventures with operators and shippers	Access, handling and long-term capacity charges.
Rolling stock	Lease or long-term availability contract	Payment tied to usable fleet availability and performance.

11.3 Limits of a conventional PPP for the southern corridor

A conventional demand-risk concession is unlikely to be the lowest-cost structure while geology, external connectivity and demand remain uncertain. Extensive government guarantees can transfer little real risk while increasing financing cost. Appraisal should therefore compare public or concessional ownership of the corridor, tunnels and minimum service with targeted private participation in terminals, trains, commercial space and selected operating contracts.

11.4 TRIPP and concession interface

The January 2026 Armenia–U.S. TRIPP framework is politically significant but expressly non-binding and does not itself grant rights. It envisages a development company with exclusive development rights inside designated Armenian transit routes, initially 74% U.S. and 26% Armenian ownership, while retaining Armenian sovereignty, border control and law-enforcement authority. Its relationship with the existing railway concession and a future national programme requires legal clarification before procurement, including asset, land, access, revenue, data and dispute boundaries.

Required legal outcome: a published interface agreement must identify who owns each asset, who may operate trains, who sets access charges, who controls data and dispatch, how domestic services are protected, and how continuity is maintained if an institutional or border arrangement changes.

Risks, common errors and safeguards

Public risk register. This table describes dependencies and controls without assigning severity labels, because no published probability–impact scoring matrix exists yet. The internal delivery register should add probability, impact, owner, trigger, mitigation, due date and review status for every risk.

Dependency	Current uncertainty	Required project response
Geology	Groundwater, weak rock, landslides and faults can change alignment and cost	Boreholes, test adits, quantified risk analysis and independent review.
Seismic safety	Tunnels, bridges and slopes require route-specific seismic design	Microzonation, ductile details, monitoring and route redundancy.
Demand	Passenger and freight volumes may be below planning targets	Phase independently useful sections; use shipper contracts and coordinated bus feeders.
External connectivity	International access, border facilities or traffic rights may be delayed or changed	Deliver domestic value first and exclude unsigned flows from the base case.
Institutional interface	The relationship among the existing concession, any TRIPP rights and new national assets is not yet defined	Settle ownership, land, dispatch, data, access and continuity rules before Stage 1 procurement.
Southern international interface	Regional security, sanctions and compliance requirements may affect finance, insurance, suppliers and traffic	Keep Agarak conditional; screen compliance and retain a stand-alone domestic operating case.
Eastern terminal operations	Access or service at the border-adjacent Sotk terminus may be constrained	Design a domestic terminus with controlled access, an emergency plan and no assumed transit.
Environment	Potential effects on Sevan, Dilijan, Zangezur, water and biodiversity	Avoid sensitive areas, complete four-season surveys and publish monitoring.
Cost and schedule	Inflation, scope change, geology and currency exposure remain unresolved	Use reference-class checks, quantitative risk analysis, change control and phased tenders.
Skills and supply chain	Tunnel, railway-systems and specialist supplier capacity may be limited	Training, knowledge transfer, packaging strategy and supplier-market testing.
Operations	Infrastructure could open without a useful service level or renewal budget	Approve a funded service contract, operating plan and renewals reserve before construction.
Policy and stakeholder continuity	Priorities, community support, land needs or governance may change over the programme	Use a statutory corridor process, published alternatives, consultation, local-benefit agreements and oversight.
Cyber and data governance	Dispatch, customs or passenger systems may be disrupted or depend on external control	Keep critical operational data under Armenian jurisdiction, segment systems, retain offline modes and audit independently.
Heritage	Archaeological finds and protected buffers may alter construction	Survey early, apply a chance-find procedure and carry time and cost allowance.

12.1 International dependencies — status reviewed 8 September 2026

Interface	Relevant development	Required before investment or operation
Armenia–Azerbaijan / TRIPP	The January 2026 framework is a dated, non-binding anchor. On 24 August the Prime Minister described a later framework agreement as being in ratification; entry into force and operating rights require the published instruments.	A completed railway and border facilities, relevant demarcation work, infrastructure access, reciprocal operating rules and contracted traffic.
Türkiye	A joint working group for rehabilitation and operation of Kars–Gyumri met in April 2026.	Border, customs, safety, gauge/interface, capacity, commercial and operating arrangements.
Existing railway concession	The current network, operating organisation and staff provide a possible implementation base; public concession terms were not independently verified in this review.	A legal and economic interface for investment, asset ownership, access, operational responsibility and continuity. Do not assume a concession change or extension before a signed instrument is published.
Iran	Regional Iran connectivity via Nakhchivan is a conditional scenario; a direct Agarak–Iran railway is not selected.	Acceptable regional-security, sanctions, financial-compliance, insurance, supplier and cross-border-volume conditions; no such traffic is assumed in the base case.
Georgia	Georgia remains Armenia's working international rail gateway.	Renewed assets, dependable train paths and a resilience plan for concentration on one operating gateway.

12.2 The twelve most likely project mistakes

Draw an attractive line on a map without a vertical profile or geology
Start with the most expensive southern tunnel before improving existing trains
Price the entire programme with one average cost per kilometre
Book transit revenue for which no interstate agreements or freight contracts exist
Add too many stations: local service is necessary, but express trains must skip some stops
Leave a single-track railway without enough passing loops and discover later that frequency cannot grow
Put a station outside town without a protected feeder route
Underfund drainage, portals, slope protection and monitoring
Buy incompatible rolling stock before the electrification choice is settled
Fail to create a major-renewals fund and recreate a slow network within ten years
Split contracts so that nobody is accountable for integrating track, power, signalling and trains
Move the alignment after ESIA and land purchase without reassessing the consequences

12.3 Place-specific challenges

Area	Main challenge	Design response
Shirak and Lori	Winter, snow, seismic exposure and ageing bridges	Snow protection, strengthening and redundant communications.
Ashotsk–Bavra high plateau	Severe cold, drifting snow, icing and uncertain demand	Road-and-feeder reference case; winter operations trial and demand threshold before rail feasibility.
Stepanavan–Tashir	Competing M3 corridor, rolling terrain and uncertain ridership	Compare express feeder, bus priority and rail on equal terms.
Dilijan–Ijevan	Landslides, forest, national park and disused infrastructure	Remain in the existing corridor wherever practicable; retaining structures and robust drainage.
Lake Sevan	Protected ecosystem and shoreline erosion	Set back from water, control runoff and avoid new shoreline embankments.
Shorzha–Sotk	Border-adjacent operation, controlled access and uncertain all-year passenger demand	Stage the service from Sevan eastward and apply a separate security-and-demand gate before Sotk.
Vayots Dzor	Narrow valleys, heritage sites and vineyards	Compare road, direct coach and feeder access before selecting any structures.
Vardenyats Pass	Snow closures, elevation, avalanches and probable long tunnelling	Keep M10 as the base access route; investigate rail only as a separate high-cost option.
Syunik	Deep gorges, faults, protected terrain and border proximity	Mode-neutral access appraisal; independent terrain and operating review before any rail selection.
Yerevan and Ararat plain	Urban development, level crossings and utilities	Corridor safeguarding, grade separation and selective double tracking.

Decision gates and stop criteria

Six programme gates remain: 0 Strategy, 1 Renewal, E East, S South Armenia Access, F Investment and O Opening. Journey-time validation is an evidence requirement across relevant gates, not an extra automatic construction stage. Every gate can select, redesign, defer or stop.

South Armenia Access Gate · Gate S

Topography and preliminary alignment
Gradients, curve radii and realistic operating speeds
Tunnel and viaduct quantities, geology and seismic safety
Competitive end-to-end journey time
Passenger and contracted freight demand
Lifecycle finance, subsidy and affordability
Environment, climate resilience and emergency access
Comparison with road, coach, regional reopening and no new domestic mountain railway

Primary source · 11 March 2026

Journey-Time Gate

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

Map system and confidence levels

Direct trains ≠ direct tracks

Infrastructure

Where tracks can realistically exist

Service

Where one train can continue

Time

Whether the journey is competitive

Gate

Whether construction is justified

Corridor / group	Geography	Planning status
R1 · Northern	Yerevan–Armavir–Gyumri–Vanadzor–Ayrum	Core · existing / modernisation
R2 · Sevan	Yerevan–Hrazdan–Sevan–Shorzha; Sotk separately gated	Core · existing / modernisation
R3 · Tavush	Hrazdan–Fioletovo–Dilijan–Ijevan	Core · historic / restoration after survey
R4 · Ararat	Yerevan–Masis–Artashat–Yeraskh	Core · existing / modernisation
Regional interfaces	Gyumri–Akhurik / Kars; Yeraskh–Nakhchivan; Nrnadzor–Agarak / TRIPP	Official processes · design, agreements, funding and operations remain separate gates
C1	Fioletovo–Vanadzor	Previously studied connector · revalidation required
Access studies	Aragats / former R7; Ashotsk–Bavra / Akhalkalaki; north-eastern Tavush; Bazum crossing and Lori plateau	Mode undecided · rail, coach, feeder, road or defer
Earlier mountain concepts	Vardenyats / former S3; internal mountain railway to Meghri	Comparator only · no-build default; no alignment selected

Map 01 shows infrastructure and engineering status on schematic relief. Its six statuses distinguish existing / modernisation, historic / restoration, official project / design, previously studied, mode-undecided studies and no-build comparators. Map 02 shows one-seat passenger patterns; dashed services depend on infrastructure and operating approvals. External regional links are explicitly distinct from Armenian infrastructure.

Terrain zones indicate investigation priorities, not surveyed hazards, tunnel portals or railway geometry. Map coordinates are unsuitable for engineering, cadastral or jurisdictional decisions.

Sources, assumptions and audit trail

Current revision, 9 September 2026: the seven-line build-out, fixed southern completion years and precise southern rail times have been retired. The revised national total, employment, operations and subsidy remain unpriced. The established sources below retain their own dates and scopes; they do not validate new alignments or make historical estimates current.