Interactive explainer
Queue simulation – interactive calculator
A team can have enough average capacity and still create long queues. This simulation explores how utilisation, variable service times and a second server interact.
Explore a shared queue with one or two servers. All 400 jobs are served in arrival order. Mean service duration is ten minutes. You can change the arrival rate, service-time variability and number of servers. The simulation runs entirely locally and always starts with an empty queue.
Queue simulation – interactive calculator
At 4.8 arrivals per hour and ten minutes of mean service, one server has 80% offered utilisation. Random arrivals can still create waiting times.
Three experiments
- Reach the capacity limit. With one server, raise arrivals from 4.8 to more than six jobs per hour. Observe mean waiting time, the 90th percentile and the time series. Read the overload explanation before interpreting the finite result as a long-run average.
- Add a second server. Switch to two servers at the same arrival rate. Arrival and service draws stay identical. You are comparing the same work under a different amount of available capacity.
- Change variability, not the mean. Set service variability to zero and then substantially higher. The theoretical mean remains ten minutes, although the realised sample mean can differ. Return to the default state afterwards.
Read the display carefully
The 90th percentile is calculated from the 400 waits. The table lists the first twelve jobs for inspection. At or above 100% offered utilisation, this model has no stable long-run mean; its finite display describes only the limited run. Breaks, priorities, abandonment and daily patterns are absent. The queueing article explains the event logic, assumptions and path towards a defensible process analysis.