Railcar cycle time is the number of days it takes a car to go from one loading to its next loading. Most shippers who run their own or leased cars know their cycle "roughly." That is not good enough. A guess of 25 days when the real figure is 31 means you are short cars every peak and paying someone else's storage when you overcorrect. This guide shows how to measure cycle time from the car event data you already get, how to break it into legs so you know who owns each day, and how to turn the result into a fleet size you can defend to whoever signs the lease.
What Railcar Cycle Time Is
Railcar cycle time, also called turn time or round-trip time, is the elapsed days between two consecutive loadings of the same car. It covers everything in between: the time the car sits at your plant being loaded, the loaded trip, the time at the customer's siding, the empty trip back, and any wait before the car is placed for loading again.
The inverse of cycle time is turns per year. A car on a 24-day cycle makes about 365 ÷ 24 = 15.2 trips a year. A car on a 30-day cycle makes about 12. That gap is the whole story of fleet economics: the lease payment is the same per month whether the car makes 12 trips or 15, so the cost per load drops as the cycle shortens.
Cycle time is not the same as transit time. Transit is only the part where the car is moving on the railroad. In most private fleets, the days a car spends at origin and destination sidings, and waiting in yards between switches, add up to a large share of the cycle. If you only track transit, you are blind to the days you control.
Where the Data Comes From
You do not need a special system to measure cycle time. You need car location messages (CLMs), the event records railroads generate as a car moves, and a consistent way to read them. Every car on the North American network reports events under its reporting marks and number, and those events are visible to the shipper, the car owner, or both.
The common sources are:
- Railroad customer portals. Each Class I railroad gives shippers car trace and event history on their own shipments. Good for a single railroad, awkward for cars that interchange.
- Railinc RailSight. Railinc, the AAR's data subsidiary, aggregates car events across railroads. Car owners and lessees use it to see the full trip, including interchange and short line moves.
- Lessor reports. Some full-service lessors send monthly fleet reports with trip histories. Ask what event they use to start and stop a trip before you trust their averages.
- Your own plant records. Loading and unloading logs, bills of lading and switch lists. These fill gaps where railroad events are thin, such as when a car was actually loaded versus when it was placed.
The events that matter most for cycle time are placement (constructive or actual) at origin, release of the loaded car, pull from the industry track, interchange to another railroad, arrival and placement at destination, release of the empty, and the empty's arrival back at origin. If you already read these events to track shipments, you have everything you need.
How to Measure It, Step by Step
To measure railcar cycle time, pick one fixed start event, pull at least 90 days of event history for every car in the fleet, compute the days between consecutive start events for each car, and report the median and the spread, not just the average. The steps below make that repeatable.
- Pick one start event and never change it. Release of the loaded car at origin is the cleanest choice for most shippers, because it is a single, timestamped event that you trigger. Placement for loading works too. What matters is consistency: a cycle runs from one release to the next release of the same car.
- Pull a long enough window. Use at least a full quarter, and a full year if your freight is seasonal. A 30-day pull on a fleet with a 28-day cycle captures barely one trip per car and tells you nothing.
- Compute per car, per trip. For each car, list its release dates in order and take the difference between each pair. A car released on March 2 and again on March 27 has a 25-day cycle for that trip.
- Drop open trips. A car that has been released once in the window but not yet released again has an incomplete cycle. Leave it out of the average, but watch it: a car with no second release for 60 days is a lost or stranded car, not a slow one.
- Flag the outliers instead of deleting them. Trips that include a shop visit, a bad order or a long storage stretch inflate the average. Tag them so you can report cycle time both with and without them. Both numbers matter.
- Report median, average and the 85th percentile. The median shows a normal trip. The average shows what the fleet actually experienced. The 85th percentile shows how slow a bad-but-not-rare trip runs, and it is the number that drives car shortages.
- Split by lane and car type. A fleet that serves a short lane and a long lane has two cycle times, not one. Blending them hides which lane is eating cars.
Break the Cycle Into Legs
A single cycle number tells you how many cars you need. A cycle broken into legs tells you why, and who owns each day. Use the same CLM events to split every trip into segments, then average each segment across the fleet.
Here is an illustrative breakdown for a covered hopper fleet on one lane. The numbers are made up for the example; they are not a transit estimate for any real lane.
| Leg | From event → to event | Who controls it | Illustrative days |
|---|---|---|---|
| Loading dwell | Placement at origin → loaded release | Shipper | 2.1 |
| Origin pull | Loaded release → pull from industry track | Railroad (local service) | 1.4 |
| Loaded movement | Pull → arrival at destination serving yard | Railroad(s) | 8.5 |
| Destination dwell | Constructive or actual placement → empty release | Receiver | 3.0 |
| Empty pull | Empty release → pull | Railroad (local service) | 1.2 |
| Empty movement | Pull → arrival at origin serving yard | Railroad(s) | 6.0 |
| Wait for placement | Arrival at origin → placement for loading | Shipper and railroad | 1.8 |
| Total cycle | 24.0 |
In this example, the shipper and receiver directly control 5.1 days of loading and unloading dwell, and share the 1.8-day wait at origin. That is nearly a third of the cycle. The railroad legs are harder to move, though not impossible: terminal dwell inside those legs is reported weekly by each Class I to the Surface Transportation Board under 49 CFR Part 1250, so you can see whether a slow month was your lane or the whole network.
The leg view also settles arguments. When the plant says "the railroad is slow" and the railroad says "you hold cars," the event data answers it. If destination dwell runs 3 days on most trips and 9 days on trips to one customer, that customer is the problem, and it is also probably where your demurrage and storage charges are coming from.
From Cycle Time to Fleet Size
The base formula is simple: cars needed = annual carloads × average cycle days ÷ 365. It works because, on average, each car is "occupied" for one full cycle per load. But a fleet sized only on the annual average will run short. Three adjustments turn the base number into a real plan.
1. Start with the average
Take an illustrative shipper moving 1,200 carloads a year with the 24-day cycle above. The base fleet is 1,200 × 24 ÷ 365 = 78.9, call it 79 cars. That is the number of cars that are, on average, somewhere in the cycle at any moment.
2. Size for the peak month, not the year
Freight is rarely flat. If this shipper's busiest month runs 120 loads instead of the 100-a-month average, the fleet has to cover a 1,440-load annual rate during that stretch: 1,440 × 24 ÷ 365 = 94.7, or about 95 cars. Cycles often stretch in peak months too, because receivers are also busy and yards are fuller. If the peak-month cycle runs 27 days, the requirement is 1,440 × 27 ÷ 365 = 106.5, or about 107 cars.
3. Add a spare margin you can justify
Some cars will always be out of service: in the shop, held as bad orders, or waiting on inspection and cleaning. Count how many car-days your fleet actually lost to out-of-service status over the last year and convert it to cars (car-days lost ÷ 365). That is your spare margin. Use your own number rather than a rule of thumb, because it varies widely by car age, commodity and lessor.
Choosing between the numbers
The base figure (79) is the floor. The peak figure (95 to 107) is what it takes to never short a load in the busy month. Between them sits a business decision: lease to the peak and carry idle cars the rest of the year, or lease near the base and cover the peak with short-term cars, railroad-supplied system cars, or by building inventory ahead of the peak. At an illustrative $900 per car per month, each car you carry year-round costs $10,800. Ten idle cars are a six-figure decision, which is why the cycle number deserves more rigor than a guess. Our breakdown of the true cost of leasing a railcar covers the rest of the per-car math.
Measurement Mistakes That Skew the Number
Most bad fleet decisions trace back to a cycle time measured the wrong way. These are the errors we see most:
- Mixing start events. Measuring some trips from placement and others from release adds or subtracts the loading dwell at random. Pick one event.
- Using the lessor's number without asking how it was built. Some reports measure only loaded-plus-empty movement and leave out siding time at both ends. That number will always look better than what your plant experiences.
- Averaging across lanes. A short lane and a long lane averaged together produce a cycle time that matches neither. When the long lane grows, the fleet comes up short and nobody can explain why.
- Ignoring the tail. A fleet with a 24-day average and an 85th percentile of 38 days behaves very differently from one with a 24-day average and an 85th percentile of 27. The wider the spread, the more spare cars you need to avoid shorting loads.
- Deleting shop trips. Out-of-service time is real fleet time. Report it separately, but do not pretend it did not happen.
- Measuring once. Cycle time drifts with season, network conditions and customer behavior. Recompute it monthly and trend it; the trend is more useful than any single figure.
Where Shippers Can Actually Cut Days
The fastest cycle-time gains are almost always at the two ends of the trip, where the shipper and receiver control the clock. Railroad legs improve with the network and with good routing, but siding days improve the week you decide to work on them.
- Load and release on the local's schedule. A car released an hour after the local switch has passed waits a full day, or longer on a lane with service a few days a week. Match loading shifts to the switch schedule.
- Hold receivers to an unloading standard. Put a target for destination dwell in the sales or supply agreement and share the event data with the customer monthly. Receivers rarely know how long they hold cars until someone shows them.
- Send empty instructions before release. Empties released without clear routing can be delayed or sent the wrong way. Our guide to empty railcar return covers the tariff defaults and what changes cost.
- Watch for cars that stop moving. A car with no new event for 48 to 72 hours should trigger a call, not a note in next month's report. Catching a stranded car in two days instead of two weeks is the cheapest cycle time you will ever buy. This is the core of railcar exception management.
- Fix bad orders fast. Know which repair shops serve your lanes and authorize routine repairs up front so a car is not held waiting on approval.
- Review the route. An extra interchange adds a yard handling, and every handling adds dwell. Sometimes a different routing on the same lane shortens the cycle even if the miles are similar.
If your team does not have the hours to pull CLMs, compute cycles by leg, and chase the cars that stall, that is exactly the work our managed rail service takes on: tracking, exception follow-up and fleet reviews that tell you whether you have the right number of cars. See our full rail logistics services, or start with the free rail freight courses if you are building this skill in-house.
Frequently Asked Questions
What is railcar cycle time?
Railcar cycle time is the number of days between two consecutive loadings of the same car. It includes loading dwell, the loaded trip, unloading dwell at the receiver, the empty trip back, and any wait before the car is placed for loading again.
How do you calculate railcar cycle time?
Pick one fixed event, such as the loaded release at origin, and measure the days between consecutive occurrences of that event for each car. Use at least 90 days of car location data, drop incomplete trips, and report the median, average and 85th percentile by lane.
How many railcars do I need?
Start with annual carloads times average cycle days divided by 365. Then size for your peak month's loading rate and cycle time, and add a spare margin based on how many car-days your fleet actually lost to shop and bad-order time last year.
What is the difference between cycle time and transit time?
Transit time is only the time a car is moving on the railroad. Cycle time is the full round trip, including days at the origin and destination sidings and waits in yards. Siding days often make up a large share of the cycle and are the part a shipper controls.
How can I reduce railcar cycle time?
Load and release cars to match the local switch schedule, hold receivers to an unloading standard, send empty instructions before release, follow up on any car with no new event for 48 to 72 hours, and pre-authorize routine repairs so bad-ordered cars are not held.
Sources: Surface Transportation Board, Railroad Performance Data Reporting, 49 CFR Part 1250 (weekly Class I reporting of terminal dwell and related service metrics); Railinc, RailSight car location data service. All fleet, cycle, lease and leg figures in this post are illustrative and are not estimates for any specific lane or railroad.