M/M/c Standard Queue

Multi-Server Queueing with Surge Analysis & Recovery Dynamics

Mathematical Foundation

The M/M/c queue represents the canonical multi-server system with Markovian (memoryless) arrivals and Markovian service times, serving c parallel servers. This model assumes unlimited waiting capacity—patients can queue indefinitely.

The Stability Criterion

System stability requires that total service capacity exceeds arrival rate:

$$\rho = \frac{\lambda}{c\mu} < 1$$

Where ρ is the system utilization factor. When ρ ≥ 1, the queue grows without bound.

The Erlang C Formula

The probability that an arriving customer must wait (all servers busy) is given by Erlang's C formula:

$$C(c, \lambda/\mu) = \frac{\frac{(c\rho)^c}{c!(1-\rho)}}{P_0^{-1}}$$

Where P₀ is the probability all servers are idle:

$$P_0 = \left[\sum_{n=0}^{c-1} \frac{(c\rho)^n}{n!} + \frac{(c\rho)^c}{c!(1-\rho)}\right]^{-1}$$

Average Wait Time

For those who must wait, the average queue time is:

$$W_q = \frac{C(c, \lambda/\mu)}{c\mu - \lambda}$$

Real-World Applications Across Industries

Healthcare: Emergency Department Physician Staffing

Operational Context:

  • Average arrival rate: 12 patients/hour (λ = 12)
  • Average service time per physician: 15 minutes (μ = 4 patients/hour/physician)
  • Current staffing: 3 physicians (c = 3)
  • Target: Wait time < 20 minutes for 90% of patients

Analysis: Utilization ρ = 12/(3×4) = 1.0 → UNSTABLE! The system is at critical threshold. Small fluctuations create unbounded queues.

Manufacturing: Quality Inspection Stations

Operational Context:

  • Parts arriving: 180 units/hour (λ = 180)
  • Inspection time: 2 minutes/part (μ = 30 units/hour/inspector)
  • Current staffing: 7 inspectors (c = 7)
  • Target: Minimize work-in-process inventory (waiting parts)

Analysis: Utilization ρ = 180/(7×30) = 0.86. Operating near the "elbow"—adding 1 inspector dramatically reduces queue length.

Call Center: Customer Service Representatives

Operational Context:

  • Incoming calls: 240 calls/hour (λ = 240)
  • Average handle time: 5 minutes (μ = 12 calls/hour/agent)
  • Current staffing: 25 agents (c = 25)
  • Target: Average speed of answer (ASA) < 30 seconds

Analysis: Utilization ρ = 240/(25×12) = 0.80. Adequate for steady-state but vulnerable to surge events (product launches, outages).

💡 Try this: Test these scenarios in the simulator below. Enable shock simulation to test surge resilience across industries—whether it's a mass casualty event (healthcare), machine breakdown (manufacturing), or service outage (call center).

Interactive Simulation Laboratory

Run ensemble simulations (60 parallel realities) to analyze queue behavior under various conditions. Test multiple shock scenarios (2×, 3×, 5× surge arrivals) and visualize system dynamics including queue evolution and recovery time.

Capacity Optimizer Targets
🚨 System Shock Simulation

Test system resilience with surge scenarios (elevated arrivals from hour 10-12)

Interpreting Your Results

Key Metrics Explained

Utilization (ρ)

< 70%: Low stress, highly responsive

70-85%: Efficient, manageable volatility

85-95%: High volatility, vulnerable to shocks

≥ 95%: Extreme wait times, system fragility

Ensemble Distribution Width

Narrow spread: Predictable performance across scenarios

Wide spread: High operational volatility—some realities perform well, others catastrophically

The 95th percentile reveals your "worst case" planning scenario.

The "Elbow" Point

Adding servers reduces wait times with diminishing returns. The elbow is where marginal benefit drops sharply.

Before elbow: Each server dramatically improves performance

After elbow: Minimal improvement per additional server