"Hayk's Arrow" Armenian Strategic Center
Azerbaijan’s national freight growth combines very different systems: roads, rail, shipping and a large pipeline component.
Azerbaijan’s national freight growth combines very different systems: roads, rail, shipping and a large pipeline component. Digital infrastructure is expanding as well, but route access, service quality and practical use remain separate from aggregate capacity; those distinctions determine relevance for Armenia.
The 2025 railway series separates operating network, domestic cargo, international cargo including transit, and passenger services. It provides an observed operating baseline alongside the profile’s existing corridor projects.
| Indicator | Value and reference period |
|---|---|
| Operational railway linesOperational line length reported for 2025; not proposed corridor kilometres, total track length or a capacity measure. | 2,118 km2025Reported result[16] |
| Electrified railway linesElectrified operational lines; separate from double tracking, reliability or usable freight slots. | 1,103 km2025Reported result[16] |
| International rail freight, including transitInternational category includes transit; do not add transit again to this number. | 14,270 thousand tonnes2025Reported result[16] |
| Transit rail freightSubset of international rail freight, not a separate addition to national rail tonnes. | 6,696 thousand tonnes2025Reported result[16] |
| Domestic rail freightLocal freight; with international freight sums to 16,813 thousand tonnes total. | 2,543 thousand tonnes2025Reported result[16] |
| Rail freight transport workDistance-weighted freight work, including 3,496 million tonne-km of transit; not tonnes carried. | 6,134 million tonne-km2025Reported result[16] |
| Rail passenger journeysAll reported journeys equal the local category in 2025; the international series is shown as a dash, not converted here into a measured zero. | 10,267 thousand journeys2025Reported result[16] |
| Rail passenger transport workPassenger-kilometres, distinct from journey counts; the reported average journey distance is 44.6 km. | 458 million passenger-km2025Reported result[16] |
| Average rail freight haulReported average transport distance per tonne; a network-use indicator, not train speed, delay or door-to-door cost. | 364.8 km per tonne2025Reported result[16] |
Azerbaijan’s railway baseline is strongly international: transit is already included within international freight, and both are distinct from domestic movements. For Armenia, the relevant question is which commodity flows could realistically use an Armenian route under agreed operating conditions, rather than whether an announced corridor has a large headline capacity. Armenia should request origin–destination matrices, commodity handling needs, monthly train paths and verified tariffs before assigning transit revenue to a new connection. Existing tonnes show activity; they do not establish a guaranteed future rerouting. [16]
The passenger series illustrates why one transport headline is insufficient. Azerbaijan reports many more rail journeys than its neighbours, but its published average passenger trip is short; passenger-km therefore offer a different scale comparison. Operational line length and electrification likewise describe assets without proving service reliability or spare freight slots. Armenia should compare end-to-end shipment time, border procedures and contractual access on a named route before treating neighbouring network size as either an opportunity or a competitive disadvantage. [16]
The national railway table is the comparison baseline; operator announcements may use different perimeters. International freight includes transit, so adding the two double counts cargo. Counts and transport work are not service-quality rankings. Figures use the retained official 2025 workbook checked on the review date. [16]
Armenia faces a choice between preparing a competitive transport service and committing early to facilities whose traffic depends on an unfinished international chain.
Armenia faces a choice between preparing a competitive transport service and committing early to facilities whose traffic depends on an unfinished international chain. Azerbaijan's operating western connections and separately staged Armenia-facing construction mean that route competition exists before any new crossing opens. Armenian policy should first establish the full cost and reliability of a journey it can legally use today. That evidence defines which administrative fix, maintenance item or negotiated access term could change a shipper's choice; it also reveals when a large new facility would add little value. [123][125][122]
Start with route-test because actual invoices and handover records connect geography to a decision. The exercise should use existing bookings, with no promise of experimental border access. It has lower irreversibility than a logistics park and provides a benchmark for access-schedule and document-pilot. The opportunity cost is postponing optional construction while operators identify whether delay, price, volume or legal access is the real limiting factor. [122][124][125]
Proposed horizons run from authorisation. Later stages depend on the conditions below; these are analytical choices, not adopted deadlines or funding commitments.
In the proposed first 90 days, select willing shippers and a repeatable cargo specification on an existing lawful route. Agree a minimal record of quotations, actual charges, waits and missed handovers. Decide which service problem will trigger a procedural remedy or commercial renegotiation; collect no personal or commercial detail beyond that purpose. [122][125]
During the proposed three-to-twelve-month phase, test one documented Armenian handover improvement and lawful document reuse. In parallel, use the measured journey to prepare an operational access annex for any authorised negotiation. Counterpart acceptance, liability and correction responsibility must be explicit before calling a digital or border procedure interoperable. [124][99][127]
Beyond twelve months, authorise optional facilities only against credible recurring traffic, an affordable operating model and useful Armenian access. Prefer a modular service using existing premises when demand remains uncertain. Recheck the baseline after openings or material tariff changes; preserve a functioning fallback rather than moving every customer to one newly available chain. [123][128][122]
The binding constraint is a complete, legally usable service chain. Armenia can improve its own processes, but cannot grant itself foreign rail slots, maritime departures or reciprocal border rights. Staff time for a small transaction study differs from capital and recurrent costs of a terminal. Keep those decisions separate. National or regional capacity figures cannot substitute for a carrier's schedule, enforceable liability and a customer's willingness to pay. [123][126][128]
Put access-schedule first if an authorised negotiation is approaching a consequential decision on jurisdiction or reciprocal use. The mandate and independent legal–operator review should then proceed while journey measurement runs alongside. An urgent defect on an operating Armenian road instead warrants qualified maintenance action before further studies. Neither exception supports optional capacity justified solely by a foreign target date or unsigned traffic forecast. [127][123][122]
Baku's existing Caspian interfaces and its Georgian and Turkish continuations give shippers usable alternatives while an Armenia-facing connection is still assembled. Armenian geography therefore creates an opportunity, not captive demand. A shorter passage can lose its advantage through handling, queues, unreliable departures or a missing return load. Conversely, better service on an existing Armenian route can create benefits before new borders open. The practical exposure is that investment and customer relationships may consolidate around operating alternatives during negotiations. Armenia's leverage is stronger when it can show a credible delivered service for a particular cargo, including documented costs and fallback arrangements, rather than asking traders to value a line on a map. [123][124][125][122]
The Azerbaijani construction record distinguishes existing BTK capacity from Horadiz–Agbend and Nakhchivan works and their future throughput targets. Completing those works alone cannot give an Armenian firm a usable transit service. Border administration, railway operations, liability, documentation and reciprocal access must also work across jurisdictions. The Armenian-published Washington declaration explicitly connects opening communications with sovereignty, territorial integrity and jurisdiction, providing a relevant negotiating reference. Armenia faces a sequencing risk if it makes irreversible commitments before complementary works and practical rights are sufficiently defined. Its opportunity is to design a connection whose benefits are shared by domestic users, not limited to moving traffic between points outside Armenia. [123][127]
Port–rail integration, coordinated train formation and customs procedures can matter as much as new infrastructure. Azerbaijan's port merger and regional action plan are evidence of attempts to address these service interfaces, although they do not themselves establish faster transit. For Armenia, interoperability creates both a bargaining issue and a domestic reform opportunity: a local exporter benefits when documents travel accurately with cargo and responsibility for delays is clear. Digital capacity is useful only if businesses can use it and data governance is acceptable. A dependency can move from a road bottleneck to a proprietary platform, scarce rail slot or unclear billing arrangement, so the commercial and administrative design deserves the same scrutiny as the civil works. [124][126][128][99]
Azerbaijan’s official transport series records 243.252 million tonnes carried in 2025, compared with 238.886 million in 2024. Freight turnover rose from 96.992 to 101.917 billion tonne-kilometres. These measures answer different questions: tonnes count cargo carried, while tonne-kilometres incorporate distance. The national totals also include pipelines and domestic transport, so neither figure is a direct count of Middle Corridor transit or container traffic. [2][3][15]
For Armenia, corridor competitiveness depends on which flows can actually use a route, their commodity mix and the service required. Bulk hydrocarbons and a container of manufactured goods place different demands on infrastructure and border handling. The corridor-specific, modal and average-distance tables therefore provide a stronger foundation for comparing routes than one headline tonnage total. Prices, schedules and access rules remain necessary complements to the physical statistics. [19][37][38]
The railway series reports 16.813 million tonnes carried in 2025, below 18.565 million in 2024, even as overall national freight grew. The same table lists 2,118 km of operational railway and 1,103 km electrified in 2025. This divergence shows why rail performance should not be inferred from total freight growth. Sea transport has its own account and must be distinguished from cargo handled by a particular port. [16][57][17]
An Armenian route assessment should focus on the handover between rail, port and ship: wagons, containers, sailing availability and border processes can constrain otherwise adequate infrastructure. Fleet and route-length statistics describe assets, not guaranteed throughput or reliability. The practical opportunity is to compare complete transport chains through Georgia and the Caspian region, using operating evidence for each interface rather than treating a newly inaugurated segment as a complete service. [8][28][18]
The transit council lists Baku port capacity at 15 million tonnes and 100,000 TEU a year, with a further 25-million-tonne and 500,000-TEU stage still described as a development project. Its ferry terminal loads railway wagons directly onto ships. At Alat, the M2 towards Georgia intersects the M1 Russian and M3 Iranian axes. The February 2025 port–railway merger is a governance change intended to coordinate this interface; it does not itself prove faster processing. The difference between an existing terminal, an expansion target and measured service performance is commercially material for an Armenian shipper. [128]
AZCON’s 2025 operator indicators distinguish over 6.5 million tonnes of rail transit from over 16.8 million tonnes of total rail freight, and over 107,000 TEU handled at the port. TEU are standard container units, not tonnes or necessarily unique consignments. These operating quantities should be compared with their own scopes rather than added together. [129]
Roads account for 62.4% of the tonnes carried in the 2025 modal table, but that share does not mean 62.4% of international transit. The collection separates road freight, passenger transport, vehicle registrations, public roads and infrastructure quality. Some detailed engineering indicators retain a 2022 reference year, while other series reach 2025. Mixing their dates would create a misleading picture of current road conditions. [23][34][24]
For Armenia, prospective road access should be evaluated at named crossings and connecting roads, with maintenance, safety and traffic management included in the cost. A national road-length increase does not establish that a particular bilateral route is open. Passenger mobility is also a distinct service: buses, border permissions and practical demand determine usability. The statistical baseline supports these comparisons without substituting for legal or operational confirmation of any cross-border link. [39][36][22]
The World Bank approved USD 65 million for the Salyan–Bilasuvar Regional Connectivity and Development Project on 20 May 2021. Its scope combines road upgrading with roadside logistics, local-market access and road-management capacity. The estimated 250,000 beneficiaries describe the appraisal expectation. Approval and expected benefits do not establish completed works or subsequently measured income gains. [131]
The 2025 freight-turnover shares assign 48.0% to oil pipelines and 13.3% to gas pipelines. Their combined role in tonne-kilometres is much greater than their share of carried tonnes, illustrating the effect of distance. Pipeline transport is thus central to the national connectivity account, but it is a specialised system with different users and infrastructure from road or container transit. A modal total cannot represent both markets interchangeably. [21][3][2]
This matters to Armenia because energy infrastructure creates durable commercial relationships even when a general transport corridor changes. Yet participation in a road project conveys no automatic access to a pipeline, and pipeline export success does not establish a competitive freight service. Investment, revenue and service-price tables help distinguish capital committed to each mode. The foreign-policy and energy profiles are needed to interpret the associated agreements and supply relationships. [35][6][14]
The internet-network table reports 2.4119 million fixed-broadband subscriptions in 2025, up from 2.324 million in 2024. Subscriptions at 100 Mbit/s or above rose from 1.0943 to 1.4109 million. International bandwidth increased from 8,020.2 to 8,899.1 Gbit/s. These infrastructure and subscription measures indicate expansion, but subscriptions are not unique people, and available international capacity does not directly measure the connection quality experienced by a household. [85][83][77]
Armenian interests include regional data routes, service competition and opportunities for resilient interconnection, not just consumer adoption. Fixed and mobile systems should be compared with their own metrics, alongside urban-rural access and provider structure. A proposed fibre route requires evidence of commissioning and commercial use before it can be counted as redundancy. National bandwidth totals alone do not show whether traffic follows independent physical paths or shares vulnerable segments. [82][55][44]
The digital-development collection distinguishes household access, individual use and enterprise connections. It includes reasons for lacking home internet, frequency and place of use, connection speed in firms and sectoral coverage. This is a different layer from international bandwidth: a country can expand backbone capacity while some households or enterprises remain poorly connected. The tables allow access, actual use and the quality of business connections to be analysed separately. [90][92][105]
For Armenian commercial and regional research, this helps identify which sectors can participate in digital trade and which still face infrastructure barriers. It also supports more realistic comparisons than ranking countries by one adoption percentage. Enterprise access, workforce use and household use should be linked without treating them as the same population. Changes in survey wording, coverage and reference years must remain visible when interpreting a sudden improvement or decline. [101][109][97]
External context: the World Bank’s November 2023 corridor assessment identifies integrated logistics, border processes, coordinated digitalisation and Caspian/Black Sea port performance as priorities. Its January 2025 country strategy connects transport with digital development. This reinforces an Armenian focus on complete service chains: infrastructure volume alone is insufficient, and the earlier report’s future-growth scenarios must not be read as achievements already recorded. [121][122]
Azerbaijan’s eastern sea interface links Baku with Kazakhstan’s Aktau and Kuryk and Turkmenistan’s Turkmenbashi. Westward freight can continue through Georgia onto Baku–Tbilisi–Kars and Türkiye, or towards Poti and Batumi for another sea crossing. The transit council records BTK commissioning in 2017 and a capacity increase from one to five million tonnes per year after upgrading the Georgian section. That is route capacity, not the actual annual cargo carried. Two western outlets provide routing choices but retain a shared Georgian land approach; their existence does not eliminate common road, rail or border bottlenecks. [123][125][17]
Operational coordination has its own implementation record. On 1 October 2025 AZCON reported a signed Azerbaijan–Kazakhstan–Georgia action plan targeting infrastructure bottlenecks, train formations, maintenance and operating procedures, alongside a separate Azerbaijan–Georgia–Türkiye memorandum on BTK freight. This institutional work addresses the handovers highlighted by the World Bank: adding track alone cannot synchronise a ship arrival, wagon supply and customs clearance. For Armenia, the competitive benchmark is a reliable complete journey. A shorter new crossing could lose its advantage if queues, irregular departures or repeated handling absorb the distance saved. [124][122]
The transit council reports an ADY container-terminal launch at Xi’an in 2025, with 20,000 TEU annual capacity. This adds an origin-side logistics presence beyond Azerbaijan’s own rail and port assets. The published number describes terminal capacity; it does not establish actual cargo handled, a guaranteed train schedule or a measured end-to-end transit time. [130]
The proposed Armenia-facing chain contains separately governed construction segments. The transit council describes Horadiz–Agbend as a 110.4-kilometre single-track main line under construction and reports that Nakhchivan reconstruction began in Ordubad after design completion in December 2025. Its 15-million-tonne Nakhchivan capacity is a post-overhaul objective. These Azerbaijani works do not establish that the Armenian segment, border facilities, signalling interfaces or operating rights are ready. The Armenian-published Washington declaration places connectivity within sovereignty, territorial integrity and jurisdiction. That text is essential to distinguishing a proposed through-route from an extraterritorial interpretation of a corridor label. [123][127]
Road freight offers more flexible collection and delivery than fixed rail paths, but its international use still depends on permit administration and a commercially usable crossing. The existing Baku–Quba–Russian-border and Baku–Gazakh–Georgian-border axes show how road trade is organised around neighbouring gateways. Azerbaijan’s transit council assigns permit exchange and one-time foreign permits to the Land Transport Agency. For a prospective Armenian connection, reciprocal permits, customs treatment and accepted vehicle documents would therefore be economically consequential alongside pavement. Vehicle age and road maintenance affect fuel use, breakdown risk and delivery variability, making nominal distance only one part of the cost. [126][24][43]
Incomplete connecting segments and unresolved operating arrangements can postpone through-service even after substantial national construction spending. [123][127]
Repeated rail–sea handovers expose shipments to coordination failures; spare track capacity cannot compensate for missing vessel slots or delayed border processing. [124][122]
Reciprocal Armenian access could add route choice for regional traders and create local logistics demand if cross-border operating arrangements become durable. [123][127]
Higher-capacity digital networks can support regional services and business exchange where affordable firm-level connections and genuinely independent routes accompany backbone expansion. [85][99][105]
These scenarios explore how documented drivers could affect Armenia. They are conditional analytical paths, not predictions or assigned probabilities.
Conditional scenario, not a forecast: if connecting rail segments, border systems and reciprocal operating rights are completed together, Armenia could acquire an additional international outlet and logistics demand. Actual advantage would depend on reliable end-to-end service compared with existing Georgian routes. [123][127][124]
Conditional scenario, not a forecast: if Armenian interfaces remain unresolved while BTK and Caspian coordination improve, additional regional cargo may consolidate on existing routes. Armenia’s response would depend on improving its own service costs and negotiating access, not solely on geographic distance. [123][125][122]
These are analytical recommendations based on the evidence, not adopted government decisions or commitments by the named institutions. Timelines and measures below are proposed implementation tests.
Organise voluntary trials on an already lawful Armenian export route through Georgia, using a consistent cargo specification and recording each handover.
Organise voluntary trials on an already lawful Armenian export route through Georgia, using a consistent cargo specification and recording each handover. Start with participating shippers' existing bookings; compare written quotes with actual invoices, departure times, border processing and arrival reliability.
Prepare Armenia's operational annex for any authorised connectivity negotiation: eligible domestic and transit users, customs jurisdiction, documents, tariffs, liability, settlement, emergency handling and dispute escalation.
Prepare Armenia's operational annex for any authorised connectivity negotiation: eligible domestic and transit users, customs jurisdiction, documents, tariffs, liability, settlement, emergency handling and dispute escalation. Begin with a joint legal–operator review of the Washington wording and a simulated Armenian shipment.
Pilot a standard shipment data pack across Armenian customs, a willing carrier and an existing foreign partner.
Pilot a standard shipment data pack across Armenian customs, a willing carrier and an existing foreign partner. Start with mapping duplicate fields and correction procedures, retain human-readable copies and manual fallback, and resolve consent, access and retention rules before exchanging new datasets.
Separate essential Armenian connection works from optional warehouses, logistics parks and expansions.
Separate essential Armenian connection works from optional warehouses, logistics parks and expansions. Begin with an independent demand and engineering appraisal using cargo commitments, land constraints and maintenance obligations; favour modular facilities that serve domestic users even if a new international route opens late.
The count includes distinct documents whose content was retrieved and checked for relevance to this profile. Translations, duplicate links and generic homepages are excluded. A document may support more than one topic when its content is relevant to each. Historical documents retain their original dates; they do not establish that a project or policy remains unchanged today.
Sources labelled “Excerpt checked” were reviewed only in the retained relevant passage. They can support the cited claim but are excluded from the full-document total.
Filters use the source date recorded in this library. Indicator reference periods are shown separately.
Annual freight carried by rail in the same national series.
Actual rail freight is distinct from railway capacity, all-mode national freight and cargo handled by a port.
| Observation | Value · Million tonnes | Reference period | Status | Sources |
|---|---|---|---|---|
| 2024 | 18.565 | 2024 | Reported result | |
| 2025 | 16.813 | 2025 | Reported result |
Four common-year dimensions compare actual railway activity and distinguish quantity, composition and distance-weighted work. They complement road, air and port records without adding unlike modes or treating proposed routes as operating services.
On smaller screens, scroll the table sideways to compare all three countries.
| Indicator | 🇦🇲 Armenia | 🇦🇿 Azerbaijan | 🇬🇪 Georgia |
|---|---|---|---|
| Rail freight carriedThousand tonnesRead with stated qualifications | 1,852.5 thousand tonnes2025Reported resultSource | 16,813.0 thousand tonnes2025Reported resultSource | 13,257.6 thousand tonnes2025Reported resultSource |
| Domestic share of rail tonnes% of all rail freight tonnesRead with stated qualifications | 32.4%2025Reported resultSource | 15.1%2025Reported resultSource | 7.9%2025Reported resultSource |
| Rail passenger journeysThousand journeysRead with stated qualifications | 511.4 thousand journeys2025Reported resultSource | 10,267.0 thousand journeys2025Reported resultSource | 1,717.9 thousand journeys2025Reported resultSource |
| Rail passenger transport workMillion passenger-kmRead with stated qualifications | 59.3 million passenger-km2025Reported resultSource | 458.0 million passenger-km2025Reported resultSource | 393.3 million passenger-km2025Reported resultSource |
Annual national reported rail cargo tonnes, including domestic, import/export and transit where reported; this is not tonne-km or corridor capacity. Common calendar year and units. National railway reporting and network geography differ; totals compare the scale and composition of reported activity, not equal-length routes, distinct people, service quality or spare capacity.
Armenia’s smaller reported cargo base makes commodity-specific continuity and commercially realistic traffic forecasts more useful than matching neighbouring headline corridor capacities.
Annual passengers carried as journey events across the national railway series; repeated trips are included and these are not distinct people. Common calendar year and units. National railway reporting and network geography differ; totals compare the scale and composition of reported activity, not equal-length routes, distinct people, service quality or spare capacity.
Use the journey count with passenger-km and route coverage before proposing Armenian service expansion; a large commuter market and a long-distance market have different needs.
Annual passenger journeys weighted by distance within each national railway reporting system; not train-km, passenger count, punctuality or capacity. Common calendar year and units. National railway reporting and network geography differ; totals compare the scale and composition of reported activity, not equal-length routes, distinct people, service quality or spare capacity.
Georgia’s passenger work is much closer to Azerbaijan’s than the journey counts suggest. Armenia should compare route distance and service purpose before drawing an investment lesson from either neighbour.
Common calendar year and units. National railway reporting and network geography differ; totals compare the scale and composition of reported activity, not equal-length routes, distinct people, service quality or spare capacity.
Geostat lists transit separately from international non-transit freight; Azerbaijan includes transit within international freight. The domestic-share calculation uses total rail tonnes in all three countries, avoiding that category mismatch.
These observations are for 2025. They do not certify today’s cargo permissions, route availability, prices, border queues or performance; a 2023 logistics index would not do so either. Network kilometres are omitted from the numeric comparison because the national line/track definitions are not harmonized here.
Several countries or indicators can share one underlying data release. Multiple source links do not mean independent confirmation.
These dated records preserve the earlier research edition. Later source releases and statistical revisions are reflected in the detailed analysis above.
Reference period: Calendar year 2024; ITA guide published 12 January 2026
The US International Trade Administration reports Baku port cargo turnover of 7.6 million tonnes in 2024 and a 73% rise in container traffic. These are recorded flows, distinct from port design capacity or announced expansion.
Reference period: 2023 study; conditional scenarios to 2030
A 2023 World Bank study identifies logistics coordination, border procedures, digitalisation and Caspian port performance as priorities for Azerbaijan, Georgia and Kazakhstan. Its growth and travel-time scenarios are conditional projections, not achieved results.
This is a curated starting collection, not an exhaustive country database or a live feed. Each record identifies its reference period and links to original evidence. “Announced” does not mean funded, ratified or operational; those distinctions are stated in the text. Armenia implications are editorial analysis, not source quotations. Census counts, population estimates and forecasts must not be treated as interchangeable. Source-check dates do not change the underlying data period.