CASE01 · Manufacturing · Heat treatment
Heat-treatment baselines and standard operating conditions
Production-weight baselines explain electricity use in SHAFT heat treatment and tempering. Comparable 432 kg batches reveal opportunities to reduce operating variability and excessive processing time.
01 · CASE ANALYSIS
Operating challenge
At the same 432 kg production weight, electricity consumption ranged from approximately 180–192 kWh in heat-treatment furnace #3 and 10–19 kWh in tempering furnace #1. Batch weight alone therefore did not explain every difference.
The analysis examined processing time beyond the standard duration and differences in loading, waiting and heating profiles. MES production records were aligned with equipment-level EMS electricity data; noise and influential observations were excluded from the normal-operation baseline.
02 · CASE ANALYSIS
Baseline models
Two weighted least squares (WLS) models use daily production weight Wday to predict daily electricity consumption Ebase_day. These models describe baseline energy use; they do not by themselves prove savings.
| Equipment | Daily electricity baseline | Reported model performance |
|---|---|---|
| Furnace #3 | Ebase_day = 0.4325 × Wday + 1.7336 | Adjusted R² 0.9995; RMSE 4.865 kWh; CV(RMSE) 0.93% |
| Tempering furnace #1 | Ebase_day = 0.0333 × Wday + 0.3784 | Adjusted R² 0.9898; RMSE 2.097 kWh; CV(RMSE) 4.86% |
03 · CASE ANALYSIS
Optimization measures
For furnace #3, the lower-consumption operating segment at the same 432 kg weight was selected as an initial target. For tempering furnace #1, the scenario compared the weight-based baseline with a supplementary model using a standard processing time of 150 minutes.
04 · CASE ANALYSIS
Measurement & verification
Keep the SHAFT product, 432 kg batch weight, target temperature, heating profile and loading conditions comparable. Track days exceeding the standard duration separately, and evaluate monthly savings as baseline consumption minus actual consumption.
The figures below are data-based scenario estimates. They must be confirmed through field implementation and M&V; they are not verified savings.
| Equipment / condition | Baseline | Target | Reported potential |
|---|---|---|---|
| Furnace #3 / 432 kg | 188.6 kWh | 180 kWh | 4.5% |
| Tempering furnace #1 / 432 kg, 150 min | 14.8 kWh | 13.4 kWh | 9.0% |
05 · CASE ANALYSIS
Engineering lesson
Interpret production weight, batch structure, processing duration and temperature together. A change in kWh/kg may reflect a change in production conditions rather than improved operation.
Predictive analytics or anomaly detection could be considered after sufficient operating data have accumulated. This is a possible future extension, not an AI result demonstrated by this case.
APPLY TO YOUR SITE
Connect your operating data to an actionable energy strategy.
Discuss the measurement boundary, data quality, operating constraints and verification approach for your facility.
Discuss your project All English cases