Discrete Rate Simulation — FAQ

Discrete Rate Simulation was created by Andrew Siprelle, founder of ChiAha, in 1990. He originally called the technique bulk flow simulation. It was later renamed discrete rate simulation as the paradigm was adopted into commercial simulation software in the late 2000s.

The foundational publication appeared at the 1995 Winter Simulation Conference: Modeling a Bulk Manufacturing System Using Extend (Siprelle and Parsons).

The technique was created in 1990 by Andrew Siprelle, originally under the name bulk flow simulation. The first peer-reviewed publication appeared at the 1995 Winter Simulation Conference.

The paradigm was renamed discrete rate simulation in the late 2000s as it was incorporated into commercial simulation software — the earliest peer-reviewed paper title using the new name appears to be Krahl's 2008 WSC paper Discrete Rate Simulation Using Linear Programming.

Discrete Rate Simulation (DRS) is a simulation paradigm that models system flow as a continuous rate punctuated by discrete events — rather than tracking each individual unit as its own event, as classical discrete event simulation does.

It decomposes any rate-based system into three primitives: constraints (rate-limiting points), buffers (accumulators absorbing rate mismatches), and interrupts (stochastic events that modify rates). Between interrupt events, the rate is constant and the system needs no recalculation. This makes DRS dramatically faster than discrete event simulation for high-speed production lines and rate-based systems.

Discrete Event Simulation (DES) tracks every individual unit as a separate event the simulator must schedule, execute, and clean up. On a slow line, that works perfectly. On a high-speed line running thousands of units per minute, DES becomes computationally prohibitive and dynamically inaccurate.

Discrete Rate Simulation (DRS) instead models flow as a rate; between rate-changing events — failures, changeovers, blocking, starvation — the simulator does no recalculation. The result: DRS scales independently of throughput rate, captures cascade effects like buffer-fill-and-empty oscillations exactly, and validates within 1% OEE on real production lines.

Continuous simulation recalculates system state at every time slice, whether anything has changed or not — using differential equations to integrate the system forward. Discrete Rate Simulation is event-based: it only recalculates when something actually changes (a rate event, a buffer hitting full or empty). Between events, the rate is constant and no integration is needed.

DRS combines the event-driven efficiency of discrete event simulation with the rate-aware accuracy of continuous simulation, without the recalculation overhead of either.

The three primitives — introduced by Andrew Siprelle in 1990 — are:

Constraints — points in the system that limit the rate of flow (a filler, a labeler, a triage station, a pipeline compressor).
Buffers — accumulators between constraints that absorb rate mismatches by filling when upstream is faster and draining when downstream is faster.
Interrupts — discrete stochastic events that modify rates: failures, changeovers, quality events, scheduled stops — each carrying its own time-to-failure (TTF) and time-to-repair (TTR) distribution.

Any rate-based system can be modeled using just these three building blocks.

Bulk flow simulation was Andrew Siprelle's original name for what is now called Discrete Rate Simulation. The 1995 foundational paper at the Winter Simulation Conference is titled Modeling a Bulk Manufacturing System Using Extend. The 1997 follow-up is Simulation of Bulk Flow and High Speed Operations.

For more than a decade, bulk flow was the paradigm's name in the published literature. It was renamed discrete rate simulation in the late 2000s as the technique was productized into commercial simulation software.

Computational cost in classical discrete event simulation scales with throughput — every unit moving through the system is an event the simulator must process. A line running 1,200 units per minute produces 1,200 events per minute in a DES model.

In Discrete Rate Simulation, the same line is modeled as a constant rate between interrupt events. The simulator does no work until something actually changes (a buffer fills, a failure fires, a changeover begins). Complexity becomes independent of throughput rate. For high-speed production lines, this can be orders of magnitude faster.

Yes. The 2020 Winter Simulation Conference paper High Accuracy Discrete Rate and Reliability Modeling to Drive Improvement of Plant OEE and Throughput (Lange, Fischel, and Siprelle) reported independent validation of discrete rate models within 1% OEE on real industrial production lines — accuracy that classical discrete event simulation models often cannot match at full production speed.

Tom Lange, the co-author, spent 36 years at Procter & Gamble, where he retired as Director of Modeling & Simulation in Corporate R&D.

Discrete Rate Simulation is implemented in several commercial tools:

ReliaSim — ChiAha's discrete rate simulator for manufacturing reliability and bottleneck analysis, developed by Andrew Siprelle's team. Validated within 1% OEE on real production lines.
ExtendSim — offers an "Advanced Technology" module for discrete rate, the implementation Krahl documented in the 2008 and 2009 WSC papers.
Decoupling Buffer Simulator — ChiAha's free, browser-based discrete rate simulator for exploring buffer-design tradeoffs.

Discrete Rate Simulation is used in any industry with rate-based flow systems:

• High-speed bottling, filling, and packaging lines (food and beverage, consumer products, pharmaceuticals)
• Chemical and continuous-process manufacturing
• Semiconductor fabrication
• Oil and gas pipeline operations
• Power generation
• Water and wastewater treatment
• Pulp and paper processing
• Bulk material handling (minerals, ores, powders, wood chips)

The paradigm has also been applied to non-manufacturing flow systems including hospital patient flow (ED arrivals and discharge), logistics planning, and supply chain modeling.

OEE simulation refers to using simulation to predict and analyze Overall Equipment Effectiveness — the standard manufacturing metric combining availability, performance, and quality.

Discrete Rate Simulation is particularly well-suited to OEE modeling because it captures the cascade effects of micro-stoppages and buffer dynamics that determine real-line OEE — effects that classical discrete event simulation often averages away. The 2020 Lange/Fischel/Siprelle WSC paper reported independent validation of discrete rate OEE predictions within 1% of actual production lines.