Recovery and purity measure different outcomes in a recycling line. Recovery asks how much of the target material in the feed reaches the intended product stream. Purity asks how much of that product stream is actually the intended material. A line can increase capture while also carrying more contamination, so the correct balance depends on feed characterization, product specification, process design and verified sampling.
Neither value can be inferred from throughput alone. A 200 t/day plant describes processing scale, not recovered-material quality. Buyers should require a material balance, named sampling method and acceptance window before treating any performance figure as comparable.
Define the terms and measurement boundaries
Recovery is commonly evaluated as target material captured in the product divided by target material present in the incoming feed. Purity is commonly evaluated as target material in the product divided by total product mass. The exact calculation must define moisture basis, sampling location, contaminants, recirculation and how stock changes are handled.
These definitions reveal why a single percentage can be misleading. Sending more borderline pieces into the product stream may raise recovery but reduce purity. Rejecting ambiguous pieces may improve purity while losing target material to residue. The buyer must decide which product specification, downstream process and residue route the system is intended to support.
Feed composition sets the opportunity
A representative feed description should cover material categories, size distribution, moisture, bulk density, contamination, hazardous components, seasonal variation and the form in which materials arrive. Municipal waste, dismantled vehicles, scrap metal and waste plastics behave differently and need different preparation and separation sequences.
Average composition is not enough when variability drives the line. The design basis should include credible low, normal and high cases for moisture, fines, bulky items and target-material concentration. It should also identify upstream changes such as collection method, bagging, compaction or source separation that alter presentation to the plant.
Liberation before separation
Separation equipment works on differences in material properties, but those differences are useful only when components are sufficiently liberated. A metal fastener attached to plastic, wire inside insulation or labels bonded to containers may not report cleanly to one stream. Shredding, crushing, cutting, screening or manual dismantling can improve liberation, but they also change particle size, dust, fines and safety conditions.
The correct liberation intensity follows the target product and downstream route. Excessive size reduction can create mixed fines that are harder to separate. Insufficient liberation can leave valuable material in composite residue. Trials should therefore evaluate the full sequence rather than one separator in isolation.
Match each separation stage to a property
| Stage | Primary function | Key feed condition | Verification question |
|---|---|---|---|
| Screening | Classify by size | Controlled feed and screen condition | Where do fines and oversize report? |
| Magnetic separation | Capture responsive ferrous material | Material exposure and suitable burden depth | What remains attached or shielded? |
| Eddy-current separation | Separate eligible non-ferrous material | Suitable particle range and presentation | Are conductors liberated and stable on the belt? |
| Density or air separation | Use mass, shape or aerodynamic response | Defined moisture and size range | How do films, fines and wet material behave? |
| Optical sorting | Detect selected material characteristics | Visible, singulated, detectable pieces | Is detection linked to a product specification? |
| Manual quality control | Remove exceptions or verify product | Safe access and manageable burden | What rate and decision standard apply? |
Each stage creates at least two streams, and every stream needs a destination. Recirculation can recover misplaced material, but repeated recirculation may reduce capacity and create unstable loading. Residue must be characterized and managed; it does not disappear from the material balance.
Control feed presentation and equipment loading
Consistent metering, belt loading and particle presentation can be as important as the separator itself. Deep or uneven burden can hide target material. Wet or sticky feed can blind screens and cause adhesion. Variable speed, feeder controls and surge capacity can stabilize flow, while instrumentation can identify overloads or blocked equipment.
Settings should be linked to a defined feed window. A line tuned during a short, clean demonstration may not represent normal operation. Commissioning trials should include representative variation and record feed mass, product mass, residue, recirculation, downtime and sampling results.
Sampling and mass balance
Quality claims are only as reliable as the sampling plan. The plan should name sampling locations, increment frequency, composite method, sample preparation, test method and responsibility. Large, heterogeneous material often needs multiple increments over time; a single grab sample can miss variability.
A reconciled mass balance checks incoming feed against products, residues, retained inventory and identified losses. Moisture basis must be consistent. Off-spec material and recirculated streams should not be counted twice. Acceptance criteria can include a throughput window, product specification, recovery calculation and operating-availability test, but each criterion needs a precise method.
Verified project evidence without invented performance
China Bente’s published project portfolio provides verified scope and capacity examples: the Venezuela municipal solid-waste sorting line at 200 t/day, the Saudi Arabia scrap-metal crushing and recycling line at 150–200 t/day, the Foshan waste-aluminum recycling and ingot plant at 200 t/day, the Uganda waste-plastic line at 150 kg/hour and the USA end-of-life vehicle line at 20 t/h.
Those pages verify application, scale and stated completion status. They do not publish project-specific recovery, purity, yield, landfill-diversion, savings or revenue figures. This article does not infer any such result.
Buyer decision checklist
- Define feed sources, quantities, variability and contamination.
- State target products and buyer acceptance specifications.
- Confirm liberation needs before choosing separators.
- Map every product, by-product, residue and recirculation stream.
- Set equipment loading and feed-presentation limits.
- Agree sampling, laboratory methods and mass-balance rules.
- Test representative feed rather than a selected clean batch.
- Include utilities, dust, noise, fire, water and maintenance interfaces.
- Specify operator training, spares and adjustment procedures.
- Separate technical performance evidence from unverified commercial outcomes.
Explore Waste Recycling & Resource Recovery Solutions and the Insights library. Related decision guides include evaluating a waste-to-resource project before investment and EPC versus machinery supply.
Design the product specification with the offtake route
Purity is useful only when it connects to a buyer or downstream process. A recovered-metal stream may have limits for attached material, fines, moisture or specific contaminants. A plastic stream may require polymer, color or density separation. The recycling line should therefore be designed backward from a documented product specification and forward to a realistic residual-management route.
Product storage and handling also matter. Mixing cleaned streams, exposing material to rain or using contaminated containers can undo separation gains after the line. Sampling points should include the delivered product condition where that is the commercial acceptance boundary.
Diagnose losses by stream, not by guesswork
When performance changes, operators should compare feed composition, equipment loading, settings and the mass balance. Target material can be lost to oversize, fines, residue, dust collection, attached composites or an incorrect product stream. Sampling each major stream locates the loss; adjusting every separator at once hides the cause.
A structured test changes one justified variable at a time and records throughput, downtime, recirculation, product mass and laboratory results. The accepted settings then become operating recipes with defined feed limits. This is more reliable than treating a short vendor demonstration as universal performance.
Commissioning records and ongoing control
Handover should include sampling procedures, calibration records, equipment settings, maintenance limits, operator training and a baseline material balance. Routine checks can trend contamination and recovery indicators, but periodic representative laboratory work may still be required. When the feed source changes materially, the performance basis should be reviewed rather than assumed to remain valid.
Acceptance should also distinguish equipment availability from material performance. A line can be mechanically available while producing off-spec material, or it can meet quality during a trial but lack stable throughput. Both dimensions require explicit tests.
Frequently asked questions
What is recovery in a recycling line?
Recovery is the share of a target material in the feed that is captured into the intended product stream, based on an agreed mass-balance method.
What is product purity?
Purity is the proportion of intended material within the recovered product stream under a defined sampling and test method.
Why can recovery and purity trade off?
Capturing more borderline material may recover additional target material while also carrying more contamination into the product.
Does throughput prove recovery or purity?
No. Throughput measures processing rate; recovery and purity require representative sampling and mass-balance evidence.
Why does liberation matter?
A separator cannot cleanly divide materials that remain physically attached, embedded or enclosed in composite pieces.
Does a magnet remove all metals?
No. Conventional magnetic separation targets responsive ferrous material; other metals need suitable downstream methods.
When is optical sorting appropriate?
When the material can be detected reliably, presented as a controlled stream and separated to a justified product specification.
Why do moisture and particle size matter?
They affect flow, adhesion, screening, sensor visibility, air separation and the stability of equipment loading.
How should performance be accepted?
Agree feed boundaries, sampling frequency, laboratory methods, operating window, mass balance and treatment of off-spec or recirculated material.
Can another project’s recovery percentage be reused?
No. Feed composition, process route, equipment settings, product specification and sampling basis are project-specific.