PET Bottle Sorting: Choosing the Right Technology for Recycled PET Purity

PET bottle recycling requires more than separating containers by color. Post-consumer bottle streams may contain PET bottles, HDPE and PP containers, PVC components, colored packaging, labels, metal objects, glass, and other unwanted materials. If these materials enter the wrong output stream, they can affect recycled material purity, increase downstream processing requirements, and reduce the consistency of recovered PET.

Choosing the right PET bottle sorting technology depends on the characteristics that need to be identified. Visible-light color sorting is suitable for recognizable color and appearance differences. Near-infrared (NIR) sorting is commonly used to distinguish supported polymer types. X-ray sorting can identify selected targets with distinguishable X-ray responses, while hyperspectral sorting evaluates spectral information across multiple wavelength bands for specialized material identification.

For many mixed PET bottle recycling applications, NIR polymer identification is a logical starting point, with visible-light sorting added when color separation is required. X-ray and hyperspectral equipment should be considered when a specific contaminant or material-identification challenge justifies their use. The final configuration should be validated using representative production materials.

PET Bottle Sorting Process
STEP 01
Feed
Stabilize material flow
STEP 02
Prepare
Present materials for inspection
STEP 03
Detect
Collect visual or spectral information
STEP 04
Classify
Identify target material categories
STEP 05
Separate
Direct selected materials to outputs
STEP 06
Verify
Measure purity and recovery

What Makes PET Bottle Sorting Difficult?

PET bottle recycling plants receive materials with different shapes, colors, labels, closures, surface conditions, and contamination levels. Some unwanted items are visually obvious, while others look similar to PET but are made from different polymers. An effective sorting process must distinguish between visible appearance differences and actual material-composition differences.

A clear PET bottle and a clear bottle made from another polymer may look similar under ordinary lighting. Conversely, two PET bottles may have very different colors but belong to the same polymer category. The sorting system should therefore be selected according to the property that separates the target material from the unwanted fraction.

Polymer Contamination

PVC, HDPE, PP, and other non-target materials may enter the incoming bottle stream. Polymer identification is important when the recycling specification requires a PET-rich output.

TECHNOLOGY TO EVALUATE
NIR / Hyperspectral
Color Variation

Clear, light-blue, green, and other colored PET bottles may need to be separated into different product streams to meet color-related quality requirements.

TECHNOLOGY TO EVALUATE
Visible-Light Color Sorting
Dense Foreign Materials

Metal, glass, stones, or other dense contaminants may require a different detection method from the one used for color or polymer identification.

TECHNOLOGY TO EVALUATE
X-ray, Where Suitable
Start with the Separation Objective

Different colors require visible-light inspection. Different polymers require a validated material-identification method. Selected dense contaminants may require X-ray inspection.

Choosing technology by the target property helps avoid unnecessary equipment complexity and unsupported performance expectations.

Which Sorting Technology Is Best for PET Bottle Recycling?

No single sensor technology is best for every PET recycling application. The correct choice depends on the target material, the property being measured, the physical condition of the bottles, and the required output specification.

TechnologyDetection BasisSuitable ApplicationsKey Limitation
Visible-Light Color SortingVisible color, brightness, shape, and surface appearanceSeparating distinguishable bottle colors and selected visually contrasting materialsSimilar-looking plastics may not be distinguishable by visible appearance alone.
NIR SortingNear-infrared spectral responseIdentifying supported polymer categories such as PET, HDPE, and PPPerformance depends on sensor range, surface exposure, material condition, and target polymers.
X-ray SortingX-ray attenuation associated with thickness, density, and compositionSelected dense or compositionally distinct contaminants that produce a detectable responseIt should not be assumed to distinguish common polymers reliably without validation.
Hyperspectral SortingSpectral information across multiple wavelength bandsSpecialized material identification where additional spectral features are usefulRequires suitable spectral coverage, validated classification, and representative material testing.
BEST MATCH FOR VISIBLE DIFFERENCES
Color Sorting

Use when the required separation is based mainly on recognizable color or appearance differences.

BEST MATCH FOR POLYMER IDENTIFICATION
NIR Sorting

Evaluate when PET must be separated from other supported polymer categories.

BEST MATCH FOR SPECIFIC X-RAY TARGETS
X-ray Sorting

Consider when the target contaminant has a measurable X-ray response difference.

BEST MATCH FOR SPECIALIZED SPECTRAL TASKS
Hyperspectral Sorting

Evaluate when testing confirms that additional spectral information improves material classification.

NIR PET Bottle Sorting for Polymer Separation

Near-infrared sorting is a strong candidate when a recycling plant needs to distinguish PET from other supported polymer categories in a mixed bottle stream. Rather than relying only on visible color, the system evaluates the material's near-infrared spectral response and classifies it using the available sensor and recognition configuration.

In a typical application, mixed bottles are distributed through the inspection area. The NIR sensor collects spectral information from exposed material surfaces, and the classification system determines whether an item matches the target polymer category. The separation mechanism then diverts selected items into the appropriate output stream.

How NIR Polymer Recognition Works
Material Feed
Present bottles consistently for inspection.
Spectral Detection
Collect near-infrared response from exposed surfaces.
Classification
Identify supported target and non-target classes.
Separation
Direct selected items to the designated output.

When NIR Is the Right Choice

NIR should be considered when polymer identity, rather than color alone, determines whether an item belongs in the target stream. For example, a clear PET bottle and a clear bottle made from another polymer may appear similar to a visible-light camera. A suitable NIR system can use spectral differences to classify these materials when the sensor configuration supports the target polymers.

NIR performance should be evaluated for the actual bottle types, label coverage, colors, and operating conditions. Dark plastics and heavily covered surfaces deserve particular attention during trials because the sensor may not obtain a reliable material signature from every item.

NIR Application Validation Checklist
Target Polymers
Confirm which polymer categories the installed sensor can recognize.
Surface Exposure
Check the effect of labels, sleeves, dirt, and moisture.
Material Variation
Test different bottle colors, shapes, and product conditions.
Output Verification
Measure both accepted-stream purity and target recovery.

Color Sorting for Clear, Green, and Other PET Bottles

When the main objective is color separation, visible-light color sorting is a direct and practical option. It evaluates visible characteristics such as color and brightness to separate distinguishable bottle categories. This can be useful when recycling facilities produce separate clear, light-blue, green, or other specified color streams.

Color sorting can also remove selected visibly contrasting bottles or materials, but it should not be treated as a universal polymer identification method. A color sorter may distinguish a green PET bottle from a clear PET bottle while being unable to reliably distinguish two different polymers that share a similar appearance.

COLOR CATEGORY
Clear PET

Recover bottles matching the specified clear or light-colored visual category.

COLOR CATEGORY
Green PET

Separate green bottles when the color difference is clear enough for the installed system.

COLOR CATEGORY
Other Colors

Classify other specified color groups according to the required product grade.

APPEARANCE CHECK
Visible Defects

Remove selected items with distinguishable visible color or appearance differences.

Color-Based Separation Sequence
INPUT
Mixed Bottle Colors
Clear, green, blue, and other visually distinguishable categories.
INSPECTION
Visible Classification
Classify material using configured visible color criteria.
OUTPUT
Separate Color Streams
Recover or reject selected color categories according to the sorting objective.

When X-ray Sorting Is Appropriate for PET Recycling

X-ray sorting addresses a different identification problem from visible-light color sorting and NIR polymer recognition. Depending on the equipment design, X-ray systems can evaluate differences in X-ray attenuation associated with material thickness, density, and composition. This can help identify selected dense contaminants or other target materials when the response difference is sufficient.

In a PET recycling plant, X-ray sorting should be considered only when the target contaminant or material class can be reliably distinguished by the installed X-ray system. It should not be selected simply because a recycling stream contains several polymer types. PET, HDPE, and PP separation normally requires a validated polymer-identification approach rather than an assumption that X-ray response alone will distinguish them.

Potential Applications
  • Detection of selected dense foreign objects.

  • Separation tasks based on measurable X-ray response differences.

  • Specialized contaminant control where conventional optical inspection is insufficient.

Pre-Selection Checks
  • Define the exact target material or contaminant.

  • Confirm that the target creates a detectable X-ray contrast.

  • Test representative samples under intended operating conditions.

  • Verify removal performance and remaining contamination.

X-ray Suitability Decision Process
01 / DEFINE
Identify the Target

Specify the contaminant or material that must be removed.

02 / EVALUATE
Check X-ray Contrast

Determine whether the target can be distinguished from surrounding material.

03 / VALIDATE
Run a Material Trial

Measure removal, target recovery, and residual contamination.

Hyperspectral Sorting for More Detailed Material Identification

Hyperspectral sorting captures spectral information across multiple wavelength bands, providing detailed material signatures for classification. This information can support specialized identification tasks when target materials exhibit meaningful spectral differences within the sensor's operating range.

In PET recycling, hyperspectral technology may be worth evaluating when conventional visual features are insufficient and the relevant materials have spectral characteristics that can be separated reliably. Its suitability depends on the wavelength range, sensor configuration, classification method, material presentation, and diversity of the incoming waste stream.

Hyperspectral sorting is not automatically more accurate than NIR for every polymer-separation task. If a standard NIR configuration can distinguish the required PET and non-PET categories under real operating conditions, a more complex spectral setup may not be necessary. The decision should be based on comparative trials and the specific identification challenge.

Hyperspectral Identification Workflow
STAGE 01
Acquire Spectral Data

Capture information across the available wavelength bands.

STAGE 02
Build Material Profiles

Establish reference data for target and non-target samples.

STAGE 03
Classify Materials

Evaluate spectral patterns using a validated recognition method.

STAGE 04
Verify Separation

Check whether the output meets the required quality specification.

When to Consider Hyperspectral Instead of Standard NIR

A hyperspectral evaluation may be appropriate when a recycling application contains difficult-to-classify material categories, when additional spectral information may improve differentiation, or when the target specification requires investigation beyond a standard NIR setup. The first step is to test representative samples and determine whether the available spectral data consistently separates the materials of interest.

The evaluation should include normal variations in bottle color, labels, surface contamination, moisture, and material condition. Classification should be assessed on representative samples beyond those used to develop the recognition model, so the trial better reflects performance on unfamiliar production material.

How to Combine Sorting Technologies in One Recycling Line

Some PET recycling lines benefit from more than one sorting stage because different contaminants require different identification methods. Combining technologies should be driven by the actual process requirements rather than by the assumption that installing more sensors will automatically improve quality.

For a typical mixed-bottle stream, NIR can be used to identify supported polymer categories, while visible-light color sorting can separate selected color grades. Additional equipment may be installed upstream or downstream if a defined contaminant cannot be controlled adequately by the primary sorter. X-ray and hyperspectral systems should be included only when their specific detection capabilities match a verified process need.

Example Technology-Matching Process
INPUT MATERIAL
Mixed Post-Consumer Bottles
Define target polymers, colors, and contaminants
↓
POLYMER SEPARATION
NIR Sorting

Recover target PET from supported non-PET polymer categories.

COLOR SEPARATION
Visible-Light Color Sorting

Separate selected color categories when the output specification requires it.

↓
OPTIONAL ADDITIONAL STAGE
X-ray or Hyperspectral Sorting

Evaluate only if testing identifies a specific contaminant or material class that warrants additional detection.

↓
FINAL QUALITY CONTROL
Verify Purity, Recovery, and Residual Contamination
Sorting StagePrimary ObjectivePotential TechnologyReason for Use
Color RecoverySeparate distinguishable bottle colors.Visible-light color sortingThe target distinction is visible in the inspected material.
Polymer RecoverySeparate target PET from supported non-PET materials.NIR sortingPolymer identity is the main separation requirement.
Dense Contaminant ControlRemove specified foreign objects with distinguishable X-ray responses.X-ray sortingThe target has a suitable measurable X-ray contrast.
Specialized ClassificationResolve a verified identification challenge using additional spectral information.Hyperspectral sortingTesting shows that spectral data can distinguish the target classes.

How to Measure PET Sorting Performance with Real Data

The correct way to compare sorting technologies is to test them against the same material and the same output requirements. A high removal rate alone does not establish that the accepted PET stream has the required purity, and high throughput does not prove that the machine is separating the target materials correctly.

During a material trial, the recycling operator should record the incoming mass, target-material mass, contaminant mass, accepted output, rejected output, and operating time. The accepted and rejected fractions should then be sampled and analyzed using a consistent method.

Accepted-Stream Purity
Target Mass in Accepted Output ÷ Total Accepted Mass × 100%

Shows the proportion of the accepted stream that meets the target-material specification.

Target Recovery
Target Mass Recovered ÷ Target Mass in Feed × 100%

Shows how much target material is recovered in the intended output.

Contaminant Removal
Contaminant Mass Removed ÷ Contaminant Mass in Feed × 100%

Shows how much of a defined contaminant is removed from the target stream.

Throughput
Processed Mass ÷ Operating Time

Shows the material-processing rate during the defined trial period.

Illustrative PET Sorting Data
Hypothetical example for explaining performance calculations

The following dataset is hypothetical and is included only to demonstrate how a recycling plant can compare output purity and target recovery. It is not a measured result for any specific machine and should not be used as an equipment performance guarantee.

MetricTrial ATrial B
Incoming feed mass1,000 kg1,000 kg
Target PET in incoming feed800 kg800 kg
Target PET recovered in accepted output760 kg740 kg
Total accepted output800 kg760 kg
Accepted-stream purity95.0%97.4%
Target PET recovery95.0%92.5%
Accepted-Stream Purity
Trial A                        95.0%
95.0%
Trial B                        97.4%
97.4%
Target PET Recovery
Trial A                        95.0%
95.0%
Trial B                        92.5%
92.5%
How to Interpret the Results

In this illustrative comparison, Trial B produces a purer accepted stream but recovers a smaller proportion of the incoming target PET. The preferred setting depends on the required product specification and the economic balance between purity, recovery, and material loss.

How to Compare Different Technologies Fairly

A meaningful comparison requires the same incoming material, equivalent sampling methods, clearly defined target classes, and comparable operating conditions. If one test uses clean, well-separated bottles and another uses heavily labeled mixed packaging, their results cannot be attributed to sensor technology alone.

For a fair trial, record the target polymer, color category, contamination profile, throughput, accepted-stream purity, target recovery, and reject composition. Repeat the test with representative material variation. If the application involves X-ray or hyperspectral sorting, define the exact target and verify that the relevant physical or spectral signal supports the intended classification.

Common PET Sorting Problems and Practical Solutions

Even a suitable sorting technology can produce inconsistent results when material presentation, inspection conditions, or classification settings are poorly matched to the incoming feed. Diagnosing the problem helps determine whether the plant needs process adjustments, different sensor technology, or additional upstream preparation.

Observed ProblemPossible CauseRecommended Check
Different polymers remain in the PET streamVisible appearance is similar, or polymer identification is insufficient.Evaluate NIR or a suitable hyperspectral configuration with representative polymer samples.
Incorrect bottle color separationOverlapping colors, inconsistent lighting, labels, or unsuitable color thresholds.Review material presentation, lighting conditions, and color-classification settings.
Selected dense contaminants remainThe target is not sufficiently distinguishable using the installed detection method.Identify the contaminant and evaluate suitable X-ray or dedicated separation equipment.
Results change between batchesChanges in polymer mix, labels, bottle color, moisture, or feed rate.Compare batch composition and repeat validation using representative samples.
Good PET is rejected excessivelyClassification settings may be too aggressive, or material presentation may be unstable.Inspect the reject stream and adjust settings while monitoring purity and recovery.
Performance falls with heavily labeled bottlesLabels or sleeves obscure the surface used for material identification.Evaluate label removal, bottle preparation, and sensor capability on the exposed material.
Sorting Troubleshooting Workflow
STEP 01
Identify the Error

Determine whether the issue is polymer contamination, color mixing, foreign material, or product loss.

STEP 02
Check Detection Suitability

Confirm that the installed sensor measures a characteristic that distinguishes the target.

STEP 03
Validate the Correction

Repeat the material test and compare purity, recovery, and reject composition.

How to Select the Right PET Bottle Sorting Solution

Before selecting equipment, recycling operators should define the required output, target polymers, bottle colors, foreign-material risks, expected throughput, and acceptable product loss. These details help establish whether color sorting, NIR, X-ray, hyperspectral technology, or a combination of suitable processes should be evaluated.

Project RequirementTechnology to Evaluate FirstWhat Must Be Verified
Separate clear and green PET bottlesVisible-light color sortingColor-category separation, recovery, and accepted-stream appearance.
Separate PET from supported non-PET polymersNIR sortingPolymer classification under actual bottle, label, and contamination conditions.
Remove selected dense contaminantsX-ray sorting, where the target has suitable contrastDetection capability, contaminant removal, and the effect on the main product stream.
Resolve a specialized spectral identification challengeHyperspectral sorting evaluationWhether available spectral data reliably distinguishes the target material classes.
Improve several different quality characteristicsA staged process using matched technologiesWhether each additional stage provides a measurable benefit that justifies its cost and complexity.
Equipment Selection Checklist
Material Information
  • Whole bottles or prepared flakes

  • Target polymer categories

  • Color distribution and label coverage

  • Foreign-material types and quantities

Performance Requirements
  • Required accepted-stream purity

  • Minimum acceptable target recovery

  • Normal and peak throughput

  • Sampling and quality-verification method

Conclusion: Match PET Bottle Sorting Technology to the Actual Material

Effective PET bottle recycling starts by identifying the property that separates the target material from unwanted items. Visible-light color sorting is suitable for distinguishable color and appearance differences, while NIR is a practical technology to evaluate when polymer identity is the main concern. X-ray sorting can address selected contaminants with detectable X-ray response differences, and hyperspectral sorting may help with specialized identification challenges when testing confirms a useful spectral distinction.

For many mixed PET bottle applications, NIR polymer identification combined with color sorting where required provides a logical starting point. Additional X-ray or hyperspectral equipment should be selected only when the target contaminant or material category justifies it. No technology should be assumed to detect every contaminant or guarantee a particular purity without validation.

A representative material trial remains the most reliable way to choose the right configuration. By measuring accepted-stream purity, target recovery, contaminant removal, throughput, and reject composition, recycling plants can compare equipment objectively and develop a sorting process that meets their actual production and recycled-material quality requirements.

Contact Us