In soda-lime glass, limestone and dolomite are the usual sources of calcium oxide, the stabilizer that makes the glass durable. A batch house does not buy CaCO3 for its own sake: it buys a predictable CaO contribution with as little of everything else as the glass can tolerate. That is why a glass-grade limestone specification is mostly a list of limits.
Calcium basis: CaCO3, CaO or loss on ignition?
Batch calculations run on oxides, so the useful figure is the CaO the material delivers once its carbonate decomposes. A carbonate assay, an elemental-calcium figure and a CaO figure are three ways of reporting the same calcium, and they are not interchangeable; ask which one the supplier reports on each lot, and by which method. Loss on ignition belongs in the calculation too: it is the mass that leaves the batch as carbon dioxide, with any moisture and organic matter.
On AragoCor’s master sheet the figures are 53.88% CaO and 96.17% CaCO3, representative analysis by ASTM C 25, with loss on ignition of 43.50–45.00% as a published range. The glass batch chemistry brief sets out the full oxide analysis.
Iron and the trace oxides
Iron is the limit that matters most for clear glass, because it tints the melt. Flint and extra-clear glasses carry the tightest iron limits; coloured container glass tolerates more. Write the limit as Fe2O3 with the test method, rather than relying on a description such as low-iron.
Alumina, magnesia and silica enter the batch calculation as well, and some glasses limit chromium and other colourants. Strontium is worth noting for aragonite in particular: the mineral takes strontium into its crystal structure, and it belongs in the oxide balance.
| Constituent | Value | Status |
|---|---|---|
| CaO | 53.88% | Representative analysis, ASTM C 25 |
| Fe2O3 | 0.01% | Representative analysis, ASTM C 25 |
| Al2O3 | 0.10% | Representative analysis, ASTM C 25 |
| MgO | 0.12% | Representative analysis, ASTM C 25 |
| SiO2 | Below 0.01% | Representative analysis, ASTM C 25 |
| SrCO3 | 1.43% | Representative analysis, ASTM C 25 |
| Loss on ignition | 43.50–45.00% | Published range |
| Moisture as received | 5–10% | Published range |
Representative analysis is a measured result on a sample, not a guaranteed lot value. A plant qualifying a new source should agree the limits that matter in writing and check them on each lot certificate.
Moisture
Batch houses weigh and mix by mass, so water in the limestone is weight that is not batch, and a source of weighing error when it varies. Many plants set a maximum moisture as received. AragoCor publishes moisture as received at 5–10%, a published range, so a plant with a tight moisture limit should raise drying and the moisture basis when it asks for a quotation, not after the first delivery.
Particle size: why glass plants specify a grading
Glass plants generally buy limestone as a graded granular material rather than a powder. Very fine material dusts in handling and can be carried out of the furnace before it melts; oversize grains may not dissolve completely; and a raw material much finer or coarser than the sand it is mixed with tends to segregate in the batch. The usual specification is a sieve grading with a top size and a limit on fines.
That is the first question to settle with any aragonite source. GL-CAL, AragoCor’s glass grade, is sized below 0.1 mm as a published range, which is finer than many plants’ limestone gradings. A batch house set up for a coarser material should discuss custom classification to its specification; one that already handles a fine calcium carbonate can assess GL-CAL against its specification directly.
Qualifying a new limestone source
- Send the written chemical and sizing specification, with the iron limit and the calcium basis.
- Review the supplier’s analysis against it, and ask which values will appear on each lot certificate.
- Calculate the substitution on the CaO the material delivers, not one-for-one by mass.
- Trial against your current limestone as the reference, recording batch handling, melting and finished-glass results.
- Agree repeat-order limits, sampling and change notification before production supply.
Claims about melting behaviour or energy belong in a measured trial, not a sales sheet. No AragoCor batch test supports one, and none is made here.
← All articles