MINING BLOCK MODEL SOFTWARE
STREAM takes the block model your geologist already produced and puts it where the mining decisions are made: against the surveyed ground and against the pit you designed. It gives you the ore/waste split, tonnage and average grade bench by bench, and a classification breakdown you can defend — without asking you to re-estimate anything.
A block model divides a deposit into a regular 3D array of blocks, each carrying an estimated grade, a density and usually a resource classification. It is produced from drillhole data by an estimation method such as kriging or inverse distance weighting, in a geological modelling package.
The model on its own is a table of numbers. What turns it into a mine plan is intersecting it with two surfaces: the ground as surveyed today, and the pit design. That intersection is what decides which blocks are actually recoverable, how much waste has to move to reach them, and what the grade of the ore you send to the plant will be.
A four-step wizard reads the file: delimiter, decimal separator and encoding are detected for you to confirm, columns are mapped with a proposal you approve, and the unit is never guessed — you cannot leave that step without stating it, because a wrong unit is a ten-thousandfold tonnage error that nothing else catches.
The model is intersected with your surveyed DEM and with the pit design, so what you read is what is actually inside the designed excavation — not what is in the deposit.
Tonnage and average grade per bench band, in a table you can export as CSV or Excel, with a SHA-256 traceability block that ties the numbers back to the source file.
Raise the cut-off and tonnage falls while average grade rises; blocks below it are not deleted, they move to waste. The same pit gives you two different reserves, and you can see both.
You map the model's resource class column onto Measured + Indicated and Inferred yourself, and the breakdown reports them apart, so an inferred tonnage never quietly gets counted as a reserve.
Draw the orebody as a 3D body or shade it onto the terrain in plan, with the pit's crest and daylight lines on top — ore left outside the design only becomes visible when you see both together.
Gaps and overlaps in the density table, repeated block identifiers, geometric overlap, blocks outside the grid definition and truncated spreadsheet exports are each reported with the value and what it means for the tonnage.
Point the wizard at the block model table. Confirm the delimiter, decimal separator and encoding it detected, then the column mapping it proposed.
Choose the grade unit explicitly — it is never guessed — and tell STREAM the coordinate system, because a block model file does not carry one.
Warnings do not block the import; they are written into the traceability record so you sign off knowing exactly how clean the data is.
Pick the pit from your Grading and Benches designs. STREAM does not draw a second pit — it reads the one you already designed.
Enter the cut-off grade, or pick it off the grade-tonnage curve. Tonnage and average grade update together so the trade-off stays visible.
Assign the model's class categories to Measured + Indicated and Inferred, and check row by row which category is being counted.
Send the bench plan to CSV or Excel, the ore polygons per bench to DXF, the block dump to another package, and the report to PDF.
No. STREAM imports a block model that was estimated elsewhere and evaluates it against your terrain and pit design. Building the model from drillholes is the job of a geological modelling package.
No. Estimation is deliberately out of scope. STREAM reads the estimated grades that are already in the model and reports what the pit recovers from them.
Not in this module. The pit design belongs to Grading and Benches; the block model module reads whichever design is active so the ore and the excavation are evaluated against each other.
Delimited text — CSV and TXT — with any delimiter and either decimal separator. The wizard detects the dialect and the encoding, and you map the columns. Sub-blocked models are recognised and reported.
Because getting it wrong is invisible. A model in ppm read as a percentage is off by a factor of ten thousand, and every downstream number — tonnage, grade, revenue — stays plausible while being completely wrong. Nothing else in the pipeline catches it, so the wizard will not let you past that step.
The file does not carry one and it cannot be inferred from the coordinates, so you state it during import. That is what lets the model line up with your DEM and your design.
Only if you say so. You map each class category yourself, and Measured + Indicated is reported separately from Inferred, so the two never merge by accident.
Yes. The block dump exports every block that passed your filters row by row, and its tonnage total matches the bench plan, so the two can be reconciled.
Free beta for Windows x64. Works fully offline with GeoTIFF, LAS/LAZ and DXF.
Download STREAM free