Enter the arrival rate and service rate into the calculator to determine the average number of customers in the waiting line (average queue length, Lq). The calculator uses the M/M/1 (single-server, FCFS) steady‑state model: Poisson arrivals, exponentially distributed service times, and λ < μ.
Related Calculators
- Little’s Law Calculator
- Expected Frequency Calculator
- Frequency Index Calculator
- Failure Rate Calculator (Simple, MTBF, & Fit)
- All Statistics Calculators
Queue Length Formula
The queue length calculator estimates the average number of customers waiting in line for a single-server queue operating under the classic M/M/1 model. This result measures the waiting line only and does not include the customer currently being served.
L_q = \frac{\lambda^2}{\mu(\mu-\lambda)}In this formula, queue length is driven by the balance between how fast customers arrive and how fast the server can complete service. As the arrival rate gets close to the service rate, the average line grows rapidly.
Variable Definitions
- Lq = average number of customers waiting in the queue
- Arrival rate (lambda) = average number of arrivals per unit of time
- Service rate (mu) = average number of customers served per unit of time
Use the same time unit for both rates. If arrivals are entered per hour, service must also be entered per hour. Mixing minutes and hours will produce the wrong result.
Equivalent Form Using Utilization
It is often helpful to express queue length in terms of server utilization. Utilization shows what fraction of the server’s capacity is being used on average.
\rho = \frac{\lambda}{\mu}L_q = \frac{\rho^2}{1-\rho}When utilization is low, the line stays short. When utilization approaches 1, even small changes in demand or service speed can create large increases in waiting line length.
When This Calculator Applies
- There is one server.
- Arrivals are assumed to be random over time.
- Service times are assumed to vary randomly around an average rate.
- Customers are served in arrival order.
- The system is analyzed in steady state, meaning average inflow stays below average service capacity.
If your process has multiple servers, batch service, scheduled arrivals, or priority classes, a different queuing model may be more appropriate.
Stability Requirement
The formula only works when the server can keep up with demand on average.
\lambda < \mu
If arrivals are equal to or greater than service capacity, the line does not settle to a stable average. In practical terms, the backlog tends to grow over time.
How to Calculate Queue Length
- Determine the average arrival rate.
- Determine the average service rate using the same time unit.
- Confirm the service rate is greater than the arrival rate.
- Substitute the values into the queue length formula.
- Interpret the result as an average, not as a guaranteed exact line size at every moment.
Example
Suppose a service desk receives 10 customers per hour and can serve 15 customers per hour.
L_q = \frac{10^2}{15(15-10)} = \frac{100}{75} = 1.3333This means the system has an average of about 1.33 customers waiting in line. Sometimes the line will be empty, and sometimes it will be longer, but over time the average waiting line is approximately 1.33 customers.
Why Queue Length Can Increase So Fast
Queueing systems are highly sensitive when utilization becomes high. A small gap between arrival rate and service rate can produce a disproportionately large waiting line.
| Situation | Interpretation |
|---|---|
| Arrival rate is much lower than service rate | Short lines and low waiting time are more likely |
| Arrival rate is close to service rate | Lines become volatile and average queue length rises quickly |
| Arrival rate equals or exceeds service rate | No stable average queue length exists for this model |
Related M/M/1 Measures
If you are analyzing the full service process, queue length is only one of several useful metrics. These related measures are commonly used alongside this calculator.
| Metric | Meaning | Formula |
|---|---|---|
| Utilization | Average fraction of time the server is busy | \rho = \frac{\lambda}{\mu} |
| Average number in system | Waiting customers plus the one in service | L = \frac{\lambda}{\mu-\lambda} |
| Average waiting time in queue | Time spent waiting before service begins | W_q = \frac{\lambda}{\mu(\mu-\lambda)} |
| Average time in system | Total time waiting plus service time | W = \frac{1}{\mu-\lambda} |
| Little’s Law for the queue | Links average queue length to average waiting time | L_q = \lambda W_q |
Interpreting Fractional Results
A result such as 1.33 customers is normal. Queue length in queuing theory is an expected average. It does not mean one-third of a person is in line; it means the line fluctuates over time and averages to that value.
Common Input Mistakes
- Using different time units for arrival rate and service rate
- Entering total customers instead of an average rate
- Applying a single-server formula to a multi-server operation
- Ignoring that the result excludes the customer currently in service
- Using the formula when demand is at or above capacity
Ways to Reduce Queue Length
- Increase service capacity so the service rate rises
- Smooth arrivals through appointments, scheduling, or load balancing
- Reduce service variability with standardized workflows
- Add self-service or pre-processing steps
- Shift demand away from peak periods
FAQ
Does queue length include the person being served?
No. Queue length refers to the waiting line only. If you want the average number in the entire system, use the system-size measure instead.
What happens if the arrival rate is greater than the service rate?
The system is overloaded. The single-server steady-state queue length formula is no longer valid because the line does not stabilize around a finite long-run average.
Why is my queue length much larger than expected?
This usually happens when the service rate is only slightly higher than the arrival rate. Near capacity, queue length grows nonlinearly and waiting lines can become large very quickly.
