Priority Queue (Non-Preemptive)

Triage Systems, Variance Redistribution & Class-Specific Dynamics

Mathematical Foundation

Priority queueing implements triage logic where customers are classified into priority classes. High-priority customers are served before low-priority customers, but service is non-preemptive—once a patient enters treatment, they are not interrupted.

The Fundamental Trade-Off

Priority systems do not create capacity. They redistribute waiting time variance: reducing wait for critical patients comes at the direct expense of stable patients who absorb the "variance burden."

Two-Class Wait Time Formula

For a two-priority system, the average wait time for low-priority (Class 2) patients is:

$$W_{\text{class 2}} = \frac{E[S^2]\lambda}{2(1-\rho_1)(1-\rho_1-\rho_2)}$$

Where:

The Nonlinear Penalty

The denominator (1-ρ₁)(1-ρ₁-ρ₂) creates exponential wait time growth for low-priority patients as system load increases:

$$\text{Wait Ratio} = \frac{W_{\text{class 2}}}{W_{\text{class 1}}} \rightarrow \infty \text{ as } \rho \rightarrow 1$$

Real-World Applications Across Industries

Healthcare: Emergency Department Triage (ESI System)

Operational Context:

  • Total arrival rate: 15 patients/hour
  • ESI 1-2 (Critical): 30% of arrivals → 4.5/hour
  • ESI 3-5 (Stable): 70% of arrivals → 10.5/hour
  • Average service time: 20 minutes (μ = 3/hour)
  • Physicians on duty: 5
  • Policy: ESI 1-2 always seen before ESI 3-5

Analysis: Total ρ = 15/(5×3) = 1.0. ESI 1-2 patients see minimal wait, while ESI 3-5 patients experience extreme delays (variance redistribution).

IT Operations: Help Desk Ticket Management

Operational Context:

  • Total ticket rate: 60 tickets/hour
  • P1 (Critical): 10% of tickets → 6/hour
  • P2-P3 (Normal): 90% of tickets → 54/hour
  • Average resolution time: 15 minutes (μ = 4/hour)
  • Support staff: 18 technicians
  • SLA: P1 within 30 min, P2-P3 within 4 hours

Analysis: Total ρ = 60/(18×4) = 0.83. P1 tickets meet SLA easily, but P2-P3 tickets can wait 3-5× longer during peak periods.

Manufacturing: Rush Order vs. Regular Production

Operational Context:

  • Total order rate: 24 orders/day (1/hour)
  • Rush orders: 25% of orders → 0.25/hour
  • Regular orders: 75% of orders → 0.75/hour
  • Average production time: 6 hours/order (μ = 1/6/hour)
  • Production lines: 8 lines
  • Policy: Rush orders preempt regular production scheduling

Analysis: Total ρ = 1/(8×0.167) = 0.75. Rush orders complete in ~7 hours, regular orders may wait 15-20 hours during high utilization.

💡 Try this: Test these scenarios below to see variance redistribution across industries. Notice how priority systems shift wait time from critical to routine tasks without creating new capacity.

Interactive Simulation Laboratory

Explore triage systems where high-priority patients receive preferential service. Analyze variance redistribution dynamics—how priority schemes shift waiting time from critical to stable patients. Time-path visualizations show class-specific wait times and the wait time ratio evolution during surge events.

Capacity Optimizer Targets
🚨 System Shock Simulation

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

Interpreting Your Results

Equity Considerations

Wait Time Disparity

The simulation shows aggregate wait time. In reality:

High-priority patients: See dramatically reduced waits (often 50-80% reduction)

Low-priority patients: Experience 2-20× longer waits depending on utilization

This creates ethical questions about fairness vs. medical necessity.

System Stress Amplification

At 70% utilization: Priority has modest effect (~2× ratio)

At 85% utilization: Low-priority waits explode (~8× ratio)

At 95% utilization: Low-priority waits become extreme (~20× ratio)

Priority systems become increasingly unfair as load increases.

Clinical Implications

Appropriate use: True emergencies (STEMI, stroke, trauma) where time = life

Problematic use: VIP treatment, financial status, physician preference

Mitigation: Fast-track systems that physically separate low-acuity patients into parallel streams with dedicated providers