Rolling stock: winning a tender and then living with it for decades

Tenders reward documented compliance more than cleverness

Public procurement for rolling stock is scored against a specification, and a technically superior offer failing a stated requirement loses to a compliant one. Bid teams therefore spend more effort demonstrating conformity, evidencing references and pricing risk allowances than on engineering novelty. The specification itself is often written around what the operator already runs: depot dimensions, platform interfaces, signalling equipment, driver training arrangements. Suppliers who engage while requirements are still being drafted shape what compliance means. Those who arrive at publication compete on price inside somebody else's design envelope. Most of the work of winning happens well before the tender documents are published.

Each network authorises the vehicle separately

Authorisation to place a vehicle in service is granted for defined infrastructure and operating conditions, covering structural strength, braking, fire behaviour, electromagnetic compatibility with signalling, and interaction with track and power supply. Even where common technical rules exist, infrastructure differences force testing on the actual network, which needs track access at times the infrastructure manager can offer. Track access, not engineering, frequently becomes the schedule constraint. Fleets destined for several networks need staged authorisation, and a late design change can invalidate completed testing, which is why configuration freeze matters so much here.

Local assembly is usually a condition of the award

Awards frequently require final assembly, a proportion of local sourcing, or a maintenance depot in the purchasing country. That obligation shapes the industrial plan: a carbody may be built in one country, fitted out in another and commissioned by a locally recruited team. Establishing such an operation for a single fleet is expensive, and the facility's future then depends on winning follow-on work in the same market. Manufacturers handle this by treating local plants as regional platforms rather than contract-specific sites, but the first contract carries the establishment cost and rarely recovers it in full.

Maintenance commitments outlive the build by decades

Many contracts bundle vehicle supply with maintenance, spares provisioning or availability undertakings measured across the fleet's life. The manufacturer is therefore pricing decades of unknown failure behaviour at bid stage. Design decisions taken to win on capital cost can become expensive under an availability regime, because a component cheap to fit but awkward to reach is paid for repeatedly. Firms with mature service businesses feed field data back into design; those without generally discover their assumptions were optimistic once the fleet enters heavy use, when the contractual remedy sits with the operator.

The cash valley on a bespoke fleet build

These contracts are hungry for cash at the front. Design, prototypes, authorisation testing and long-lead components consume money before deliveries begin, and milestone payments rarely match the spend profile. Where advance payments are made, the operator requires bonds, which consume banking facilities. A delay in authorisation therefore strikes twice, deferring revenue while fixed costs continue and possibly triggering penalties. Bidders who model the contract as a series of vehicle sales miss this entirely; the useful model is a project with a long negative cash phase financed at a cost belonging in the bid price.

Frequently asked questions

Why does authorising a new train take so long?

Because the vehicle must be shown safe and compatible with specific infrastructure, and much of that evidence can only be produced by running on the network itself. Test paths compete with commercial traffic and are usually offered at night or during engineering possessions. Add structural, fire, braking and signalling compatibility testing plus assessment by an independent body, and the sequence is long even when nothing goes wrong. Any design change during the process risks invalidating evidence already gathered.

Can a manufacturer reuse a vehicle platform across different countries?

Partly. Structural design, bogies, traction packages and interiors can be shared, and platform strategies exist precisely to spread development cost. What rarely transfers unchanged is the interface with infrastructure: gauge, platform height, power supply, signalling equipment and national operating rules. Adaptation is genuine engineering work rather than configuration, and it needs its own authorisation evidence. Manufacturers who oversell platform commonality during a bid usually pay for the difference during delivery.

Who carries the risk when a fleet underperforms in service?

It depends on contract structure. Under a plain supply contract, the operator carries availability risk once warranty expires. Under availability-based arrangements, the manufacturer or maintenance provider is paid according to how many vehicles are serviceable, so the exposure sits with them. Those regimes normally carve out infrastructure faults and vandalism. Operators should also check how spares obsolescence across a long fleet life is handled, because that argument arrives late and costs real money.