Economy · Part 5 · About 11 minutes
Travel demand, traffic and capacity
Demand describes the journeys people and goods want to make. Traffic is the part assigned to your network. Capacity determines how much each route can carry before service deteriorates or demand spills elsewhere.
The one-minute explanation
Settlements generate passenger and freight demand according to their population, economy, resources, development and distance from one another. The network then tries to connect each origin and destination through available routes.
Travellers and freight do not divide evenly across every option. Roads, operating rail lines, air services and water routes compete in one assignment. Faster and cheaper choices usually attract more traffic, while eligibility and capacity limit what each mode can carry. If one option fills, the remaining movement can move to another usable choice; movement with no connection or capacity remains unserved.
Demand is not traffic
Potential demand is the desire to move between two settlements. It can exist before any connection has been built. Time and price effects leave retained demand: journeys that take far longer than they should are partly given up, and so are journeys whose every useful option is charged too heavily.
The “too long” benchmark is explicit. A passenger trip is expected to take about half an hour plus the distance at 60 km/h; a freight shipment about two hours plus the distance at 45 km/h. Only time beyond that benchmark suppresses demand, and passenger demand reacts to the excess roughly twice as strongly as freight does. The network then divides retained movement among eligible routes and services; the result is carried demand and the traffic placed on physical links.
Passenger trips and cargo movement wanted between an origin and destination.
Potential movement left after pair-level excess-travel-time and price suppression.
Movement that finds a usable connection and fits within available capacity.
Where demand comes from
The game calculates demand separately for passengers and cargo. Both respond to the strength of the settlements at each end, but they represent different reasons for movement.
Passenger demand
Larger settlements tend to generate more trips, but demand does not grow one-for-one with population. This prevents the largest cities from overwhelming every other relationship. Economic strength increases the amount of travel, while distance reduces it.
Demand is directional. A large regional centre can generate a different number of outbound trips to a smaller city than the smaller city generates in return. The model therefore keeps both directions even though some displays summarize the pair.
Freight demand
Freight is shaped by the relationship between production and consumption. Resource-rich settlements create stronger outbound cargo potential, while developed settlements attract inputs and distribute finished goods. Minerals, energy and agricultural resources can strengthen particular flows.
Freight declines with distance more slowly than passenger demand. A distant resource centre may therefore remain an important cargo partner even when passenger demand between the same places is modest.
Distance, economy and policies
Distance weakens demand rather than imposing a hard maximum. GDP raises the scale of movement, but its effect is bounded. Policies assigned to a settlement can also increase or decrease the passenger or cargo demand originating there.
To keep a large world manageable, each settlement retains local partners, separate passenger and freight leaders, and mandatory administrative and freight backbones. Remaining pairs are not discarded: they are aggregated into a long-tail total by origin and destination region. The tail remains visible in the national demand total, but it is not routed pair by pair and is not counted as network-unserved demand.
The detailed matrix is recalculated quarterly. Network, service and price changes can alter assignment between those updates, but they do not rewrite the current quarter’s potential demand.
How demand finds a route
For roads, the network first looks for ways to connect the origin and destination. A journey may use a direct link, a continuous named corridor or an indirect path through intermediate settlements. Rail participates through active operating lines; air and water participate through active scheduled routes with suitable endpoints. This is why traffic between A and C can load the A–B and B–C road links even when B is not the journey’s destination.
When several options can serve the same pair, they compete through a generalized cost:
A quicker route is more attractive. A toll or fare makes a route less attractive, but the effect depends on travellers’ value of time: wealthier endpoints are generally less sensitive to paying for time savings. Mode and service preferences are applied inside the same choice calculation; they do not create a second demand source.
Route choice is therefore relative. Improving one corridor can draw traffic away from another even when total demand between the settlements has barely changed.
Diversion is different from suppression
A charge can produce two different responses:
- Diversion: movement shifts to a cheaper available route.
- Suppression: some movement is no longer made because every useful option is too slow or charged too heavily for the time it saves.
If some usable option is both free and fast enough for the distance, the pair loses nothing to suppression—demand is redirected instead. Otherwise, each mode contributes its own best option to the pair’s retention: a strong second mode alongside a tolled road keeps more of the demand alive, even when it is not the cheapest choice. Suppression is applied once to the pair, so adding several parallel tolled links of the same kind does not repeatedly erase the same demand—within a mode, only its best option counts.
This creates the basic toll trade-off: a higher rate earns more per vehicle but can reduce the number of vehicles willing to use the charged network.
What capacity means
Road capacity is stored as a one-direction design-hour capacity, and the game reads it on two different scales:
- For congestion, assigned flow is compared with a peak-equivalent monthly allowance. This test asks whether the busy hours are saturated, and it is what drives travel-time deterioration.
- For how much the link can carry and bill in a month, the ceiling is the design rate over the full model month. A link can therefore move far more monthly traffic than its peak allowance suggests—at the price of congestion—before assignment refuses to place more on it.
Rail is different again: a railway’s binding limit is scheduled train-path slots—how many round trips per day the track and its grade can host—not a vehicle-equivalent flow. The infrastructure view shows this as round trips per day and slot occupancy.
Passenger and cargo movement also consume capacity differently. Cargo is converted into passenger-car-equivalent units, with each cargo unit occupying more capacity than a passenger movement.
Capacity is directional, so saturation uses the direction with the higher combined passenger-and-cargo equivalent flow.
Passengers and cargo travelling either way can pay tolls or fares.
These quantities should not be expected to match. A highly directional commuter corridor may be close to capacity during its busy direction while its billable total includes journeys in both directions. The same directional-capacity, bidirectional-billing split applies to water routes as well.
Overflow and unserved demand
When parallel direct routes or continuous corridors compete for the same demand, the assignment first follows route attractiveness. Capacity is then resolved for every origin–destination pair at once rather than pair by pair, so no market gets first claim on a shared link; if a route exceeds its available capacity, excess passenger and cargo demand is redistributed toward alternatives that still have room.
Redistribution cannot create capacity. If every practical option is full—or if no connection exists—some demand remains unserved. Adding capacity at the wrong place may not solve the problem: every segment of a corridor has to support the movement passing through it, and the narrowest or slowest part can remain the constraint.
Rail does not use road-style congestion in the same way. Its carried load is capped at capacity rather than being shown as indefinitely accumulating on the railway. Excess movement is served only where the route assignment can place it on another available choice.
How congestion develops
Road congestion appears when directional peak-equivalent traffic becomes large relative to capacity. As saturation rises, travel time increases. The slower time then raises the route’s generalized cost, making alternatives more competitive in later assignments.
Rail has a price feedback instead of a time one: as an operating line approaches saturation, its automatic fare will not fall below what covering the line’s operating cost requires, so a crowded railway defends its finances rather than slowing down. Lines run under a public-service stance are exempt and may keep losing money by design.
The published route load is adjusted progressively toward the current assignment rather than snapping completely between competing paths every month. This helps traffic settle when two similar routes would otherwise alternate between “empty” and “overloaded.”
Congestion matters beyond the traffic display. Poor road service can reduce transport efficiency, weaken the service associated with affected settlements and create development or migration pressure. Capacity is therefore an economic constraint as well as an engineering number.
Capacity is not the same as service
A route can have high nominal capacity and still provide weak service. Capacity answers how much can pass; service also depends on whether the link is in the right place, connects the right settlements, offers competitive travel time and avoids bottlenecks along the full journey.
- Coverage
- Can demand reach the network at both ends?
- Continuity
- Does the entire path connect, including every intermediate segment?
- Attractiveness
- Is the route competitive in time and price?
- Capacity
- Can its busiest direction carry the assigned passenger and freight load?
- Reliability
- Will congestion or a narrow segment undermine the promised journey?
This is why widening an uncongested feeder can achieve little while a short saturated segment continues to constrain the corridor.
Air and water services
Air and water routes draw on the same underlying demand matrix, but their service and revenue are constrained by each scheduled route’s own capacity—held separately for each direction, so a route’s effective two-way throughput is twice its one-direction figure—and by the processing capacity of the airports and ports themselves, which is shared among every route using the facility. Passenger and cargo suitability also differs by mode and service type. Water route panels report load factors per direction, and the suggested frequency follows the busier direction rather than the two-way sum.
Unlike a road or rail corridor, an air or water connection requires an active route rather than only the presence of endpoint infrastructure. An airport or port can therefore exist without carrying the potential demand between its settlement and another place.
When comparing modes, do not read a large origin–destination demand number as a promise that every mode can capture it. Availability, capacity, preference, charges and mode-specific eligibility determine the part that becomes actual traffic.
Reading the demand panel
The detail view follows the same funnel used by the assignment kernel:
Demand satisfaction is carried divided by potential demand. Capacity service rate is carried divided by retained demand, so it isolates the network after time and price effects. A market can therefore have a perfect capacity service rate but lower demand satisfaction when suppression is the only loss. Likewise, “network unmet” means only unconnected plus capacity-constrained movement; it never includes time or price suppression or the unexpanded long tail.
The mode chart uses actual carried movement, not nominal capacity or route preference. Expand a mode to see its stable operating-service or path contributions. Direction tabs keep A→B and B→A separate; the combined view adds their absolute quantities and then recalculates ratios from the combined totals.
How to diagnose a corridor
Use this sequence when a route is empty, congested or less profitable than expected:
- Confirm that demand exists. Check passenger and cargo movement for the origin–destination pair.
- Confirm the connection. Look for a complete path, not merely infrastructure near each endpoint.
- Inspect through traffic. A link may be busy because longer journeys pass through it.
- Compare alternatives. Faster, cheaper or preferred road, rail, air or water options may be taking most of the flow.
- Check time and charges. A pair whose best option is far slower than the distance warrants loses demand to time suppression, and tolls can divert traffic or suppress it when no free, fast option exists.
- Find the constrained segment. Capacity added elsewhere will not remove the active bottleneck.
- Separate directions and purposes. Peak saturation, billable traffic, passengers and cargo answer different questions.
- Check the dates. Potential demand updates quarterly; route loads move progressively toward equilibrium after network, service or price changes.
Common misconceptions
“Building a road creates its own traffic immediately”
The road connects and distributes existing demand. Longer-term economic and population changes can later alter demand, but the initial traffic still depends on the places and journeys already present.
“A link only carries demand between its two endpoints”
Indirect journeys can pass through intermediate links. A seemingly modest local segment may be carrying an important national corridor.
“Capacity is the sum of both directions for the month”
The displayed saturation test uses the busier direction against directional peak-equivalent capacity. Billing can still count movement in both directions.
“An empty route proves there is no demand”
Demand may exist but choose a faster or cheaper alternative, fail to find a continuous path, be suppressed by excessive travel time or charges, or exceed the service’s usable capacity.
Continue the collection
The next part looks at the charges inside route choice: when pricing earns, diverts or suppresses. To follow network performance back into settlements instead—service scores, accessibility, growth and metro clusters—see Settlements, transport service and metros.