Today
The plant sees the result after the variation
- Lambda is measured from periodic physical samples.
- Drivers of spread are hard to see between tests.
- Extra density often becomes the safety margin for the declared value.
Declared lambda carries variation.Density often pays for it.
Under EN 13162, declared thermal conductivity is a statistical value, not the production average. INSUS models the process conditions that widen lambda on your line so teams can review where tighter control might support the same declared performance with less material margin. Physical heat-flow-meter testing stays the basis for declaration.
01
Declared lambda is a fractile, not the mean.
02
Production spread widens the margin.
03
Physical tests stay the declaration record.
The problem
Production variation is a tax onevery declared value.
For factory-made mineral wool, the declared thermal conductivity is statistically conservative rather than the average measured lambda. Process variation becomes performance margin the manufacturer has to carry. When the line is held in specification but not managed for variance, extra density often becomes the practical insurance for the declared value.
01Mean thermal conductivity
02Variability margin
03Declared lambda
- 01Mean thermal conductivity
- 02Variability margin
- 03Declared lambda
The shift
From managing limitsto managing the variation behind them.
Quality and process teams stay responsible for declaration and production changes. What changes is earlier visibility into process drivers that widen the thermal-conductivity distribution between physical tests.
With INSUS
See variation drivers earlier
- A continuous estimated lambda signal supports monitoring between tests.
- Process conditions linked to wider spread can be identified by product family.
- Teams can review whether tighter control could support the same declared performance at lower density.
How it works
From line signalsto declared-value margin.
This fits OptiStudio as operational decision support beside the existing QC process. Soft-sensor estimates are for monitoring and advice only. Heat-flow-meter results remain the basis for the declared value.
Connect
Bring production and quality data together
Read-only links to SCADA or historian tags, production records, weight and thickness measurements, binder dosing, process parameters and LIMS or QC lambda results linked to runs.
Understand
Identify drivers of lambda variation
Learn which process variables move with measured thermal conductivity by product family on this plant's history, separating useful drivers from correlated noise.
Predict
Estimate lambda between physical tests
A plant-calibrated soft sensor estimates lambda from current line conditions and flags potential drift in mean or spread. The estimate is not a declared value.
Advise
Rank actions that may reduce variation
Rank candidate adjustments such as density or compression, curing conditions, binder distribution and line speed by modelled contribution to spread, inside product and process constraints. People approve any change.
Verify
Check estimates against heat-flow-meter results
Compare each soft-sensor estimate with the corresponding physical result. Track model performance by product family. Measured rolling standard deviation remains the operational proof that variation is falling.
Getting started
Start with one lineand one or two product families.
For factory-made mineral wool producers that already link production and heat-flow-meter history. Delivered through OptiStudio as advisory process decision support. Expert Services can help with discovery and shadow-mode validation. The aim is tighter process control of variation, not inventing a new declaration method.
- What we need. Historical heat-flow-meter lambda results linked to production runs, SCADA or historian data, density and thickness measurements, and process variables such as line speed, curing, compression and binder dosing.
- What you see first. Variation drivers by product family, soft-sensor estimates between tests and ranked process adjustments for review. No automatic declaration update.
- How it grows. After shadow-mode validation on one line, extend across product families and lines as heat-flow-meter results support it.
Related for insulation
Other insulationstarting points
Declared lambda margin sits beside melting and curing energy and retrofit compliance triage on the same insulation business.
- 01Mineral Wool Melting & Curing Energy OptimizationRecommend furnace and curing-oven settings from plant energy and process history inside melt and cure constraints.
- 02Retrofit Installation Compliance TriageRank retrofit insulation jobs by inspection risk using existing documentation and historical outcomes.
- 03Insulation industryThe wider picture for mineral wool melting energy, declared lambda and retrofit documentation.
Questions mineral wool quality teams ask us
Declared lambda under EN 13162 reflects statistical variation, not only the mean. Wider production spread can force more density as margin. This use case surfaces process drivers of that variation for quality and process review.
Heat-flow-meter lambda results linked to production runs, SCADA or historian data, density and thickness measurements, and process variables such as line speed, curing, compression and binder dosing.
No. Physical testing remains the basis for declaration. Soft-sensor estimates are for monitoring and advice only and are not treated as declarable results.
Not by default. Lines that already have production monitoring, gauging and a heat-flow-meter QC process can start from those sources. Missing continuous gauging is a separate plant decision.
No. Recommendations are advisory. Process and technical teams approve production changes and declaration decisions.
Each estimate is compared with the corresponding heat-flow-meter result. Measured rolling standard deviation of lambda remains the operational proof that variation is falling.
Through OptiStudio as operational decision support for lambda variation. Expert Services can support discovery, data linkage and shadow-mode validation against your own QC baseline.
Insulation · Declared performance
Review the variation behind your declared lambda.
See what production and heat-flow-meter history say about the spread behind your declared value, and where tighter process control could create room for better margin or lower density at the same declared performance.