By the time many natural graphite anode projects reach engineering, the file looks complete: a concentrate assay, a lab report showing the material “can be purified to 99.95% C,” maybe a few grams of spherical graphite. Then the engineers start asking questions that natural graphite testwork was supposed to answer, and nobody can:
- What’s the filter cake moisture after purification?
- How much spent acid and wash water does each ton of product generate?
- How long should coated material stay in the kiln, and how much volatile matter goes to the off-gas system?
- Will the fines actually flow out of a storage bin?
At that point, projects either go back to the lab and lose months, pad the design with rule-of-thumb margins and pay for it in capital cost, or build on assumptions and deal with the fallout at commissioning.
The root problem is confusing two kinds of testing: tests that prove the material can be processed and tests that give engineers something to design with. This guide covers the second kind from the perspective of an equipment manufacturer and EPC contractor: which tests you need, what design data each one produces, and when to run them.
What Testwork Is Required Before Designing a Natural Graphite Processing Plant?
Testwork for a natural graphite processing plant means lab and batch tests on representative samples that produce the data behind process design criteria, mass balances, and equipment sizing. It typically covers seven areas:
- Sample representativeness and baseline characterization
- Beneficiation testing (for ore-based projects)
- Purification testing
- Micronization and spheroidization testing
- Coating and carbonization testing
- Engineering support testing (material handling, drying, dust, wastewater, and off-gas)
- Anode material performance testing
Why “It Works in the Lab” Isn’t Enough to Design a Plant
Proving that a concentrate can reach battery grade is a prerequisite for the project, not a design basis. Engineering needs parameters it can plug into calculations, and those parameters have to feed a specific set of deliverables:
- Process design criteria (PDC): throughput, operating conditions, and design margins for each unit operation
- Process flow diagram (PFD): unit sequence, recycle loops, and by-product routing
- Mass and energy balances: feed rates and reagent, water, power, and fuel use at every step
- Equipment list and sizing: type, size, number of units, and materials of construction
- Layout, utilities, and permitting: corrosion protection, treatment system capacity, and emissions data
- CAPEX and OPEX estimates
Graphite beneficiation shows how this works. On one Australian graphite project, the team ran head and size-by-assay analysis, mineralogy, comminution, flotation, locked cycle, and tailings dewatering tests, then used the results to build the mass and water balances, process design criteria, and flow diagrams that drove the equipment list and cost estimate. Every test had a downstream user. That’s the core of design-basis testwork: decide what data the design needs, then decide what to test.
Match Testwork Depth to the Study Stage
Testwork deepens as a project moves through its study stages. In mining projects, cost estimate accuracy is typically about ±50% at scoping, ±25% at prefeasibility (PFS), and ±15% at feasibility (FS). Tighter accuracy demands broader testing, more representative samples, and more complete data.
Table 1: Testwork focus by study stage
| Study stage | Cost estimate accuracy | Typical samples | Testwork focus | Design deliverables supported |
| Scoping | About ±50% | A few representative samples | Baseline characterization and standard-condition tests to confirm the process route | Route options, preliminary scale and yield assumptions |
| Prefeasibility (PFS) | About ±25% | Composites covering the main ore or feedstock types | Option comparisons, condition tests, locked cycle tests, initial engineering support tests | Preliminary PFD, PDC, mass balance, layout |
| Feasibility (FS) | About ±15% | Composites plus variability samples | Optimization, variability testing, full engineering support tests, pilot work if needed | Detailed PFD, mass balances by feed type, equipment specs, environmental data |
For downstream anode projects, the feasibility-stage program has to cover the full range of feedstock variation, not a single “standard” sample. See [INTERNAL LINK: How Feedstock Variability Affects Natural Graphite Process Design].
📊 Image placement: Diagram showing how testwork deepens across study stages (Scoping → PFS → FS → detailed engineering), with test modules and accuracy at each stage.
Core Testwork Modules and the Design Data Each One Produces
The modules below follow ZD’s natural graphite anode route, which purifies first and shapes second. For each one, the question is the same: what do you test, and where does the data go?
Sample Representativeness and Baseline Characterization
If samples come only from a supplier’s best lot, every result downstream leans optimistic. By the feasibility stage, sampling should span multiple lots, mining areas, or suppliers.
Baseline characterization covers fixed carbon, particle size distribution, impurity chemistry and deportment, crystallinity, and morphology. Thermogravimetric analysis (TGA) is a useful quick screen, since the TGA residue flags thermally inert inorganic impurities and supports lot-to-lot consistency checks. Test methods and pass/fail criteria are covered in [INTERNAL LINK: How to Qualify Natural Graphite Concentrate for Battery Anode Production]. For design, this module produces the feedstock spec and variation range in the design basis.
Beneficiation Testing (Ore-Based Projects)
If the project starts from ore, beneficiation testwork shapes the whole front end:
- Comminution: Bond work index and abrasion index set mill selection, power, and wear estimates. Graphite adds a twist: overgrinding breaks large flakes and cuts concentrate value, so flake preservation has to be tested too (see [INTERNAL LINK: Complete Guide to Natural Graphite Multi-Stage Crushing Process Design]).
- Flotation condition tests: set the number of rougher, regrind, and cleaner stages and the reagent scheme.
- Locked cycle tests: repeated batch flotation with middlings recycled into each next round, simulating continuous circulating loads until results stabilize. One project’s locked cycle test on a roughly 200 kg composite produced a 95.9% C(t) concentrate at 91.4% recovery, with flake size in line with design assumptions (company-reported).
- Dewatering and rheology: settling, filtration, and slurry behavior for concentrate and tailings set thickener, filter, and tailings disposal design.
For equipment options, see our guide to natural graphite flotation equipment.
Purification Testing
The most common gap here is reporting only final purity. Engineering needs the process data:
- Reaction conditions (temperature, time, solid-to-liquid ratio, reagent concentration) for reactor sizing and cycle time
- Reagent consumption per ton for storage sizing and operating cost
- Washing and filtration (wash water volume, cake moisture, filtration rate) for filter selection and dryer load
- Wastewater and off-gas composition for treatment system sizing and permitting
- Corrosivity for materials of construction
Keep in mind that purifying natural graphite removes mineral impurities from material that’s already highly crystalline. It’s a different process from synthetic graphite graphitization, and parameters don’t transfer between them. See what natural graphite purification involves.
Micronization and Spheroidization Testing
Three results matter most for design:
- Yield: the share of feed that leaves the shaping and classification circuit as on-spec spherical graphite. It sets front-end throughput and is the most sensitive number in the whole-line mass balance.
- Specific energy consumption: sets electrical capacity and operating cost.
- Fines volume and properties: size the classification, dust collection, and fines handling systems and decide whether fines can be sold.
Test equipment should work on the same principle as the plant equipment, or the yield won’t scale. See [INTERNAL LINK: Where Is Material Lost in Natural Graphite Anode Production?] and our spheroidization equipment comparison.
Coating and Carbonization Testing
This stage sets the anode’s final surface properties and is the direct basis for rotary kiln design. You need data on:
- Precursor thermal behavior: TGA on pitch or other precursors gives the weight-loss curve and carbon yield, which set the coating recipe and the volatile load on the off-gas system (see pitch vs. resin coating precursors).
- Temperature profile and residence time: thermal tests in a batch or pilot rotary kiln can establish reaction temperatures, residence time, and off-gassing behavior, the inputs for kiln length, diameter, zoning, and heating power.
- Rotation speed and fill level: control how evenly particles tumble and heat, which affects coating uniformity and caking.
- Agglomeration tendency: decides whether you need deagglomeration equipment and an oversize return loop (see why spherical graphite agglomerates during coating).
- Off-gas volume and composition: size the incineration or treatment system.
The rotary kiln is ZD’s core equipment. In our experience, the more complete this data set, the more closely kiln zoning and adjustment range can match the actual material instead of rules of thumb borrowed from other products. Learn more about the coating granulation rotary kiln and rotary kiln technology.
Engineering Support Tests That Often Get Missed
Process testwork proves you can make the product. Storing, moving, drying, and safely handling it in the plant takes a different set of tests, and new projects skip them more than any other:
- Bulk and tap density: bin volumes, conveying capacity, and packaging sizes.
- Flowability: fines and spherical graphite flow very differently. Jenike shear testing under ASTM D6128 measures bulk solid flow properties, the basis for hopper angles and outlet sizes. Without it, bin bridging is one of the most common startup problems.
- Abrasiveness: drives the choice of wear-resistant linings and how much demagnetization capacity you need to handle iron pickup (see our demagnetization equipment selection guide).
- Drying behavior: drying curves and heat duty set dryer type and size.
- Combustible dust: graphite fines and precursor dust such as pitch both need evaluation. A dust hazard analysis includes characterizing the material’s moisture content, particle size, and combustibility, which then drives dust collection, explosion protection, and electrical area classification.
- Wastewater and off-gas: sample at every stage and compile plant-wide emissions data for permitting.
For downstream handling design, see the ultimate guide to final product handling for graphite.
Anode Material Performance Testing
Finally, test product samples against your target customer’s spec: purity and magnetic impurities, particle size distribution, tap density, specific surface area, and morphology. Coin half-cell testing (capacity and initial coulombic efficiency) is typically done by a third-party electrochemistry lab, and full-cell validation belongs to the cell maker’s qualification. For design, this step locks in the product spec, which works backward to constrain shaping, classification, and coating parameters.
Testwork-to-Design Reference Table
Table 2: How each test feeds the design
| Test | Key results | Design deliverable | Typical consequence if missing |
| Baseline characterization | Fixed carbon, impurity types and deportment, PSD, crystallinity | Feedstock spec in the design basis | Wrong process route or misjudged margins |
| Comminution | Work index, abrasion index, flake preservation | Mill selection, power, wear | Undersized mills or flake damage |
| Flotation and locked cycle | Grade, recovery, circulating load | Flotation PFD, mass and water balances | Grade or recovery short in continuous operation |
| Dewatering and rheology | Settling rate, cake moisture, slurry viscosity | Thickeners, filters, tailings | Filtration bottlenecks |
| Purification | Reaction conditions, reagent use, wash water, emissions | Reactors, corrosion protection, treatment capacity, permits | Undersized environmental systems, OPEX overruns |
| Spheroidization | Yield, specific energy, fines volume | Front-end capacity, classification, power | Bottlenecks or idle equipment |
| Coating and carbonization | Carbon yield, temperature profile, residence time, off-gas | Rotary kiln sizing and zoning, off-gas system | Uneven coating, caking, mismatched kiln capacity |
| Material handling | Bulk density, flowability, abrasiveness | Bins, conveying, wear materials | Bin bridging, iron contamination |
| Drying | Drying curve, heat duty | Dryer type and size | Product moisture out of spec |
| Combustible dust | Combustibility and related properties | Dust collection, explosion protection, electrical classification | Safety and compliance risk |
| Product performance | D50, surface area, tap density, magnetic impurities | Product spec and process constraints | Failed customer qualification |
When Do You Need a Pilot Plant?
Not always. A pilot usually makes sense when the process or step combination is new, when feedstock comes from several sources or varies widely, when customers want ton-scale qualification samples, or when batch tests can’t reproduce continuous circulating loads or equipment wear. For planning and scale-up, see From Lab to Pilot to Plant.
Five Common Gaps in Testwork Programs
- Unrepresentative samples. Testing only the best lot designs the plant for the best case.
- Lab sequence that doesn’t match the plant. Shaping before purifying in the lab, then purifying first in the plant, makes yield and fines data unusable.
- Results without process data. “99.95% purity” with no reagent, water, or off-gas figures leaves engineers with nothing to size.
- No engineering support tests. Missing flowability, abrasiveness, and dust data surfaces at commissioning.
- Reports that don’t line up. Different labs and vendors using different units, bases, and sample IDs means no one can build a single mass balance.
How to Organize a Testwork Program: An EPC Perspective
Work backward. List the design deliverables, then the data each one needs, and only then set the test list and assign labs. Testwork is usually split among mineral labs (characterization and beneficiation), process labs (purification, shaping, and coating), equipment suppliers (machine-specific tests), and electrochemistry labs. What matters is that one party owns the whole-line view and aligns scope, data formats, and the mass balance basis, which is where an EPC contractor adds the most value early on.
For how testwork fits into the full project, see how to build a natural graphite processing plant and our natural graphite processing plant equipment list.
FAQ
What is design-basis testwork for a graphite plant? Design-basis testwork is testing aimed at what plant engineering needs. Beyond proving a feedstock can make on-spec product, it supplies process data such as reagent consumption, yield, residence time, flowability, and emissions for process design criteria, mass balances, and equipment sizing.
What is a locked cycle test in graphite flotation? A locked cycle test runs repeated batch flotation rounds in the lab, recycling middlings into each next round until results stabilize. It gives grade and recovery figures closer to continuous plant performance than open-circuit tests do.
Is a pilot plant required before designing a graphite anode plant? Not always. With a proven process and one consistent feedstock, thorough lab and batch testing is usually enough. Pilot work is generally recommended for new processes, multiple feedstock sources, or ton-scale customer samples.
What’s the difference between concentrate qualification and design testwork? Concentrate qualification asks whether a feedstock is usable and focuses on pass/fail criteria. Design testwork asks how to build the line and focuses on parameters engineers can calculate with. See [INTERNAL LINK: How to Qualify Natural Graphite Concentrate for Battery Anode Production].
What tests are needed to design a coating and carbonization rotary kiln? Precursor TGA (weight-loss curve and carbon yield), batch or pilot thermal tests (temperature profile, residence time, rotation speed, and fill level), agglomeration tendency, and off-gas volume and composition. Together they set kiln dimensions, zoning, heating power, and off-gas treatment.
Conclusion: Decide What the Design Needs, Then Decide What to Test
A natural graphite plant design is only as good as its testwork data. Too little testing leaves the design running on margins and assumptions, and the wrong testing produces data that never reaches the design. For new entrants, building the test program backward from design deliverables, before equipment and capital are locked in, is the cheapest way to manage project risk.
As an equipment manufacturer and EPC contractor for natural graphite anode lines, ZDZN can help at several points:
- Planning your testwork? Send your test plan or existing reports to ZDZN’s engineering team. We’ll check them against the design deliverables and flag missing data and tests.
- Have samples? Send them for anode material performance testing (purity and magnetic impurities, PSD, tap density, specific surface area, and morphology). We’ll benchmark results against your customer’s spec and help turn your coating and carbonization trial data into rotary kiln design parameters.
- Ready to launch? Request a full-line proposal review. We’ll build a mass balance from your testwork data, configure the purification, shaping, coating granulation rotary kiln, and carbonization sections, and lay out a phased EPC plan.