tags: [] - coffee/roasting - coffee/roasting/quality aliases: - SPC in roasting - Control charts in roasting

Statistical Process Control¶

Tags: #coffee/roasting #coffee/roasting/quality Aliases: SPC in roasting, Control charts in roasting Related: Roasting MOC | Profile Documentation | Consecutive Batch Consistency | Yield Calculation | Cropster Status: ✅ Complete

Overview¶

Statistical Process Control (SPC) is a quality management methodology that uses statistical methods to monitor and control a production process, distinguishing between normal process variation (common cause) and unusual variation that signals a process problem (special cause). In coffee roasting, SPC provides a rigorous framework for tracking key batch parameters — Drop Temperature, DTR, yield, first crack timing, total roast time — and identifying when the roasting process is drifting outside acceptable limits before it produces out-of-specification batches that fail cupping. While full SPC implementation is most common in large-volume commercial roasting operations, the underlying principles are applicable at any production scale.

Core SPC Concepts¶

Common cause variation: Natural, inherent variation in the process — the small batch-to-batch fluctuations in RoR, DTR, or yield that occur even when the process is well-managed and in control. Common cause variation is expected and should fall within defined control limits.

Special cause variation: Unusual variation caused by a specific, identifiable factor — a change in green coffee moisture, a burner component failure, an operator error, an environmental shift — that is not part of the normal process distribution. Special cause variation signals that the process is out of control and investigation is required.

Control charts: The primary SPC tool; a graph of a measured parameter (e.g., Drop Temperature) over sequential batches, with: - Centre line: The process mean (average value across historical batches) - Upper Control Limit (UCL): Mean + 3 standard deviations - Lower Control Limit (LCL): Mean − 3 standard deviations - Warning lines: Mean ± 2 standard deviations (inner limits)

A batch that plots outside the UCL or LCL, or shows a systematic trend (e.g., seven consecutive batches above the centre line), signals a special cause requiring investigation.

Applying SPC to Coffee Roasting¶

Key parameters to track on control charts for a consistently roasted production lot:

Parameter	Control chart type	Typical UCL/LCL range
Drop temperature	Individual X / moving range	±4°C from mean
DTR	Individual X / moving range	±3% from mean
Total roast time	Individual X / moving range	±45 seconds from mean
Yield	Individual X / moving range	±2% from mean
First crack start time	Individual X / moving range	±30 seconds from mean

Control limit values should be calculated from actual process data (a minimum of 20–30 consecutive batches on the same lot and profile) rather than assumed.

SPC in Practice at Different Scales¶

Small specialty roastery: Full control charts may be impractical with the small number of batches produced per lot. However, tracking key parameters in a simple run chart (sequential plot without formal control limits) still reveals trends, batch-to-batch drift, and outliers visually.

Medium roastery (50–200 kg/week): With sufficient batch volume to calculate meaningful statistics, formal control charts per lot are practical. Crop rotation (transitioning from one crop year to the next) is a key special cause event requiring chart reset.

Large commercial roastery: Full SPC implementation including automated control chart generation from roast software data is standard practice. Triggers defined out-of-specification (OOS) batches for hold and reinspection before dispatch.

Benefits of SPC in Roasting¶

Early warning: Identifies process drift before it produces failed batches; investigation and correction can occur before customer-facing quality impact
Separates noise from signal: Normal variation is expected and does not require action; SPC distinguishes signal (special cause) from noise (common cause), preventing over-correction
Objective evidence: Provides documented, statistical evidence of process performance for quality audits, certification, and wholesale customer reporting
Continuous improvement baseline: Enables tracking of the effect of intentional profile or equipment changes against a statistical baseline

Key Facts¶

SPC distinguishes common cause variation (expected, in-control fluctuation) from special cause variation (process problem requiring investigation)
Control charts plot key parameters over sequential batches with UCL/LCL at ±3 standard deviations from the mean
Track: Drop Temperature, DTR, total roast time, yield, first crack timing — all key batch consistency indicators
Formal SPC is most practical in medium-to-large operations; run charts (sequential plots without formal limits) provide partial benefit at smaller scales
Calculate control limits from at least 20–30 consecutive historical batches on the same lot and profile

References¶

Changelog¶

Date	Change
2026-04-27	Note created

This article is part of All-About-Coffee.com - The comprehensive coffee knowledgebase.