The Bond work index (Wi) measures how much energy a tonne of ore needs to grind from a theoretical infinite size to 80% passing 100 µm. It's converted into specific grinding energy through the Bond equation, then into mill power and diameter. Because ores vary, a single composite sample can undersize equipment if it misses harder zones.
What the Bond Work Index Measures
The Bond work index is a measure of a rock’s resistance to grinding. It’s expressed in kilowatt-hours per metric tonne (kWh/t). The higher the number, the more energy you’ll spend to turn coarse feed into fine product. In Bond’s original definition, the work index is the energy needed to reduce a unit mass from a theoretically infinite particle size down to 80% passing 100 µm. That’s a useful benchmark because it strips out machine-specific effects and focuses on the ore itself.
The Society for Mining, Metallurgy & Exploration (SME) describes the Bond work index as one of the standard grindability metrics in mineral processing. It’s not a mill property; it’s an ore property. You can test the same ore in different labs and, if the procedure is followed, you’ll get comparable numbers. Soft ores like limestone sit at the low end of the scale, while hard taconite or dense sulphides demand much more energy per tonne.
How the Bond Test Works
The standard Bond ball mill grindability test starts with a representative ore sample crushed to below about 3.35 mm. You lock a measured mass of that material inside a standard Bond mill—a steel cylinder with a defined internal diameter and length, filled with a specific charge of steel balls. The mill rotates at a fixed speed, dry, and the sample grinds for a set number of revolutions. After each cycle, you screen the product to find the mass passing a chosen closing sieve, typically 150 µm or 106 µm. Then you calculate the grams of new undersize produced per revolution.
That number is the grindability index. You repeat the test in cycles, adding fresh feed to replace the undersize removed, until the net grams per revolution stabilises. It’s an equilibrium test, not a single pass. The result is a single number that feeds directly into the Bond equation.
The Bond Equation in Words and Numbers
The Bond equation is the bridge from grindability to energy. In words, the specific grinding energy (W) equals 10 times the work index (Wi) divided by the square root of the product size (P80) minus 10 times Wi divided by the square root of the feed size (F80). Both F80 and P80 are the 80% passing sizes in micrometres. In equation form: W = 10 Wi / √P80 − 10 Wi / √F80. You can see that a finer product raises the energy demand quickly, while a coarser feed reduces it.
This equation is based on Bond’s Third Theory of Comminution. The Canadian Institute of Mining, Metallurgy and Petroleum (CIM) includes Bond’s third theory in its standard comminution references. It’s not an absolute law—it’s an empirical correlation with built-in assumptions. But it’s still the most widely used starting point for ball mill sizing in feasibility studies.
Inputs That Drive the Calculation
Before you can use the Bond equation, you need four numbers: the work index (Wi), the feed size (F80), the product size (P80), and the throughput. F80 comes from the crusher circuit. If you’re feeding a ball mill directly, F80 is typically the product of the last crushing stage. P80 is your target grind, set by the downstream process—flotation might need 75 µm, while leaching could accept 150 µm. Throughput is the plant feed rate in tonnes per hour.
You can’t skip any of these. A wrong F80 from an undersized crusher inflates the grinding energy; an overly fine P80 that flotation doesn’t need wastes power. The work index ties them all together. Once you have W in kWh/t, multiply it by the feed rate to get the mill’s shaft power requirement in kilowatts.
From Energy to Mill Diameter and Installed Power
Once you have W in kWh/t, you multiply it by the required feed rate in tonnes per hour to get the grinding power draw in kilowatts. That’s the shaft power the mill needs. You don’t buy a mill by diameter alone; you buy it by the power it can transmit. A larger diameter mill has more volume and can take more power, but the specific energy per tonne stays the same regardless of mill size.
Mill diameter selection then becomes a trade-off between capital cost, footprint, and mechanical limits. For example, Xinhai reports its manufacturing range includes ball mills up to 7 m in diameter. That doesn’t mean every ore needs a 7 m mill; you select diameter based on the calculated power, the mill aspect ratio (length to diameter), and the liner design. Installed motor power is always a bit above the shaft power to account for motor efficiency and drive losses. When you’re selecting a mill, you also need to consider whether a single mill or a parallel pair will give better availability. This is where choosing a ball mill gets practical: you match power, not just diameter.
Why One Composite Sample Can Undersize the Whole Circuit
A single composite sample can lie to you. If you blend high-grade, soft ore with harder wall rock into one bucket, the composite’s Bond work index is a weighted average. When the mine feeds only the hard wall rock later, the mill will be underpowered, the grind will be too coarse, and recovery may drop. That’s why experienced plant designers insist on testing multiple geological domains, not just one blended composite.
In an EPC project, the testwork program should map hardness across the orebody. Mineral processing EPC flows usually include variability testing as part of the design basis. Xinhai’s design institute, for instance, runs approximately 200 beneficiation tests per year across more than 70 ore types according to the company’s published figures. That volume of testwork doesn’t eliminate variability, but it helps you see where the hard zones are before you commit to a mill size.
Practical Hardness Bands, Without Overpromising
You’ll often see Bond work index tables divided into soft, medium, and hard bands. There’s no single universal cutoff, but qualitative ranges help. Soft materials like limestone or gypsum need little energy; medium ores like copper porphyry sit in the middle; hard materials like taconite or some dense sulphides sit at the high end. These bands are only a guide—actual values vary even within one orebody.
No lab test, no equation, and no supplier can guarantee a mill’s performance on a new ore. The Bond work index gives you a defensible starting point, but you still need pilot-scale confirmation for very hard or highly variable deposits. Scale-up factors, wear, and classification efficiency all affect the real circuit. You also need to watch out for moisture, because a wet Bond test on a dry circuit can mislead you just as much as a single composite.
Common Mistakes When Applying the Bond Work Index
One classic mistake is using the average Wi from a deposit instead of the maximum. If half the ore tests at 10 kWh/t and half at 16 kWh/t, the average is 13 kWh/t, but the mill will see 16 kWh/t for long stretches. Another mistake is ignoring the classification efficiency in the circuit. The Bond equation assumes a certain ideal classifier performance; real hydrocyclones or screens deviate from that ideal, so you need a correction factor.
You also can’t take a Bond index from a dry test and apply it directly to a wet mill, or vice versa. The test conditions matter. And don’t forget that the work index is defined for a specific product size—using it far outside the normal range of 70–150 µm adds error. For coarse grinding or ultra-fine grinding, you’ll need a different test, not just a different Wi.
Frequently Asked Questions
What does the Bond work index tell you?
It tells you the ore's resistance to grinding in kWh per tonne. A higher index means more energy is required to reach a target fineness.
How is the Bond work index test performed?
A representative sample is crushed to about 3.35 mm and ground in a standard Bond ball mill at a fixed speed. The test cycles until the net grams of undersize per revolution stabilises, giving a grindability index.
Can one composite sample mislead mill sizing?
Yes. A composite averages soft and hard zones; if the mine feeds mostly hard ore later, the mill may be underpowered. Multiple composites by geological domain are safer.
Does the Bond work index guarantee mill performance?
No. It's a lab-based estimate. Pilot testing, scale-up factors, and classification efficiency still affect real circuit performance.
