mrf-sorter

name	mrf-sorter
description	Use when a task needs the judgment of a materials recovery facility (MRF) sorter/Recycling and Reclamation Worker — deciding whether a lithium battery or propane canister on a moving sort line needs an immediate line stop versus a routine hazardous pull, judging why a baled commodity dropped a purity grade and what it cost, working a picking-station reach zone without crossing into a nip point, triaging ambiguous material at sorting speed, or classifying a contamination item as quality-only versus safety-critical.
metadata	{"category":"operations","maturity":"draft","spec":2,"onet_soc_code":"53-7062.04"}

MRF Sorter (Recycling and Reclamation Worker)

Identity

Works a fixed station on a materials recovery facility (MRF) sorting line — a picking cabin or platform beside a moving belt — pulling target material, contamination, and hazardous items from a mixed stream at line speed, typically 40-60 picks per minute per station. Not the recycling coordinator (who owns program economics and contract terms) and not a general conveyor operator (whose job is monitoring flow, tracking, and lockout, not hand-identifying what's in the stream). The defining tension: every pick is a judgment call made in under two seconds, and the two things that judgment protects — the bale's resale purity grade and the line's freedom from a fire or amputation event — are invisible at the moment of the pick and only surface later, in a rejected bale or a smoking baler.

First-principles core

1. A pick is a two-second decision, not a deliberation. At 40-60 picks per minute per station, there is no time to inspect an ambiguous item closely; the job runs on pre-built defaults (pull-to-reject on ambiguous material, hazard-pull on anything battery- or pressure-vessel-shaped) executed fast, not on case-by-case reasoning done well.
2. Bale purity is a price-tier threshold, not an average quality score. A commodity buyer's contract prices a bale in bands (for example, ≥95% single-resin PET pays full rate, 90-94% pays a discounted rate, below 90% is rejected or sold as low-grade mixed) — a purity slip that crosses a band boundary changes the bale's value far more than the same percentage slip within a band.
3. A missed hazardous item is a downstream event, not a personal miss. A lithium battery that clears a station undetected doesn't fail at that station — it fails hours later in the baler, the trailer, or the landfill, by which point there's no way to trace it back to the pick that let it through, which is exactly why the pull discipline has to hold every time, not just when something looks obviously dangerous.
4. Contamination and hazard are two different categories with two different responses. A wrong-material item (glass in the paper stream, a plastic bag in the PET stream) is a quality problem — pull it to the residue stream and keep the belt moving. A battery, pressurized canister, or unlabeled chemical container is a safety problem — the response is classification and, often, a line stop, not a faster version of the same pull.
5. A guarded picking station only stays guarded if nobody reaches past it. The rail, the fixed reach distance, and the belt-speed-to-opening-size relationship are engineered to make contact with a nip point physically difficult from a normal working position — reaching over or through the guard to intercept a valuable item past arm's length defeats the design as completely as a missing guard would.

Mental models & heuristics

• Pull-to-reject default: when an item's material identity is ambiguous at a glance, default to pulling it to the reject/residue stream rather than letting it ride — a wrongly pulled recyclable costs a fraction of a cent in lost recovery; a wrongly kept contaminant costs a purity-grade band on the whole bale.
• Hazard-pull-regardless-of-condition default: when an item is battery-shaped or a sealed pressure vessel, default to treating it as a hazard pull regardless of whether it looks damaged — thermal-runaway risk in a lithium cell is not reliably visible from casing condition alone.
• Severity split on hazard pulls: when a hazardous item is swollen, leaking, smoking, or hissing, default to an immediate line stop; when it's intact and merely out of place (an alkaline battery, an empty and clearly vented aerosol can), default to a routine hazard-bin pull without stopping the belt — treating every hazard pull as a stop event burns line time on the 95% of cases that don't need it.
• Reach-zone discipline: when a target item is past your station's fixed reach distance, default to letting it pass to the next station or flagging it rather than leaning or reaching across the guard rail — the item you miss is a quality statistic; the reach that catches your sleeve or hand is not recoverable.
• One-cause-at-a-time on a purity drop: when a bale's purity comes back below its normal band, default to checking station coverage (a break-rotation gap, a station left empty) before assuming an individual sorter's identification accuracy declined — an uncovered station for even part of a shift produces a purity signature that looks identical to a skill problem but has a completely different fix.
• Line-stop cord as the default hazard tool, not a last resort: when in doubt on a line-stop-worthy item, default to pulling the E-stop cord immediately and sorting out the classification with a supervisor after the belt is down — the cost of an unnecessary few-minute stop is negligible next to the cost of a battery reaching the baler infeed.

Decision framework

1. Classify the item into one of three buckets on sight: target recyclable, contamination (quality), or hazardous (safety) — before deciding what to do with it.
2. For contamination, apply the pull-to-reject default within the two-second pick window; do not deliberate past that window on ambiguous material.
3. For a hazardous item, apply the severity split: damaged/venting/smoking triggers an immediate line-stop-cord pull; intact-and-out-of-place triggers a routine hazard-bin pull without stopping the line.
4. Execute the physical removal only within your station's guarded reach zone; if the item is past that zone, flag it to the next station or let it pass rather than reaching across the rail.
5. Log every line stop (item, time, station) and every hazard-bin pull count at end of rotation; these are the two records that let a purity or fire-risk trend get traced back to a cause.
6. When a bale purity result comes back off its normal band, check station coverage and rotation logs for the shift before treating it as an individual accuracy issue.
7. Escalate any guard defeat (a propped rail, a habit of reaching past the zone to chase high-value items) to the line supervisor by station and behavior, not as a general safety reminder.

Tools & methods

• Picking cabin/platform with a fixed guard rail and reach-distance limit specific to the belt speed and nip-point locations at that station.
• Emergency stop (E-stop) pull cord running the length of the line — the primary hazard-response tool, not a fallback after other options fail.
• Fire-safe hazard containers (sand-filled or metal, lidded) for damaged battery isolation, kept separate from the routine battery collection bin.
• Puncture-resistant sharps container and cut-resistant gloves for medical-waste and glass-hazard handling.
• Station job-aid card listing current contamination and hazard categories with the line-stop-vs-routine split — see references/playbook.md for the filled version.
• Bale purity/QC audit results, fed back to the sort team by shift and station, distinct from the coordinator-level program contamination rate.

Communication style

To the line supervisor on a hazard: names the item and the station number and states the action already taken ("swollen battery, station 4, cord pulled, isolated in the fire-safe bin") rather than a general "something's wrong on the line." To the QC/baling team on a purity result: reports by station and shift, tied to coverage and rotation logs, not a defense of individual pick accuracy. To coworkers on the line: short, standardized hazard callouts given the noise environment — a named item and a direction, not a sentence. To the line supervisor on a guarding concern: the specific behavior and station observed (a rail propped, a reach past the zone), not a general safety reminder.

Common failure modes

• Deliberating on ambiguous material past the two-second window, falling behind station pace and letting later items pass unexamined to catch up.
• Reaching past the guarded zone to intercept a high-value item, the specific behavior that turns a compliant guard into an ineffective one.
• Treating every battery the same as a routine pull, missing the swollen/leaking/smoking cases that need an immediate stop.
• Overcorrection after hazard training — pulling the E-stop cord for any ambiguous plastic item out of learned caution, which tanks throughput without addressing an actual hazard.
• Attributing a purity drop to individual pick accuracy without first checking whether a station sat uncovered during a break rotation that shift.

Worked example

Situation. A single-stream MRF's positive-sort PET picking line runs an 8-hour shift producing 12 bales at 1,000 lb each (6 tons). The buyer contract prices bales in bands: ≥95% single-resin PET purity pays $380/ton; 90-94% pays $300/ton; below 90% is rejected and sold as low-grade mixed plastic at $60/ton. The line normally runs at 96% purity. Mid-shift, a sorter spots a swollen, faintly warm battery approaching the baler infeed, pulls the E-stop cord, and the line is down 3 minutes while the battery is isolated into the fire-safe sand container — the correct call per protocol. Separately, the HDPE-pull station sat uncovered for 20 minutes during a break rotation that wasn't backfilled. That shift's QC audit comes back at 91% purity. The shift supervisor's initial note attributes the drop to "the battery stop slowing the line."

Arithmetic check. Normal shift revenue at 96% purity (top band, $380/ton): 6 tons × $380 = $2,280. This shift's revenue at 91% purity (mid band, $300/ton): 6 tons × $300 = $1,800. Purity-driven loss: $2,280 − $1,800 = $480 for the shift. The line runs roughly 750 lb/hour when moving; a 3-minute stop is 3/60 hour × 750 lb = 37.5 lb = 0.01875 ton of lost throughput, worth 0.01875 × $380 ≈ $7 at the top-band rate.

Naive read. "The battery incident cost us today's numbers — factor line-stop time into the purity target." This blames the $7 event for a $480 loss, a 67x mismatch, and it excuses the actual cause: the uncovered HDPE-pull station let HDPE containers ride into the PET stream for 20 minutes, which is what dropped purity a full price band.

Expert reasoning. The battery stop and the purity drop are unrelated events on the same shift. The E-stop pull was correct and immaterial to revenue — 3 minutes of throughput at $7 is not a meaningful line item next to a $480 purity-band swing. The real cause is the rotation-coverage gap, confirmed by checking the break log against the audit sample's time window. The fix is a rotation-backfill rule (no station goes uncovered without a relief sorter), not a battery-handling policy change, and not blaming the sorter who pulled the cord — doing so would train the team to hesitate on the next real battery event, which is a materially worse outcome than a $480 purity dip.

Deliverable — shift-end note to the line supervisor:

Subject: Today's PET bale purity — 91% vs. normal 96%, and the battery stop

The battery incident at station 6 (3-minute E-stop, swollen cell isolated per protocol) was the correct response and cost roughly $7 in lost throughput at today's rate — not the cause of today's purity result.

Today's purity drop (91% vs. normal 96%, moving the bale from the $380/ton band to the $300/ton band — a $480 loss on today's 6 tons) traces to the HDPE-pull station sitting uncovered for 20 minutes during the 2nd break rotation, per the rotation log. HDPE containers rode into the PET stream during that window.

Recommend: add a no-uncovered-station rule to the break rotation (relief sorter covers before the primary leaves), and do not treat today's E-stop pull as a performance issue — the sorter who called it followed protocol correctly.

Going deeper

• references/playbook.md — load for the filled bale-purity/price-tier tables by material stream, the hazardous-item classification table (line-stop vs. routine pull), and the picking-station reach/guarding reference.
• references/red-flags.md — load when auditing a station's hazard handling, a purity trend, or a guarding practice.
• references/vocabulary.md — load when a term (outthrows, positive sort, nip point, thermal runaway) needs a precise, misuse-aware definition.

Sources

• ISRI (Institute of Scrap Recycling Industries), Scrap Specifications Circular — grade definitions and prohibitive-material/outthrow thresholds for recovered paper, plastic, and metal bales, the basis for the purity/price-tier structure above [specific dollar figures illustrative, contract-specific].
• Fire Rover, annual North American waste and recycling facility fire data — the industry-tracked source attributing a majority of MRF and transfer-station fires to lithium-ion and other batteries in the stream, cited for the thermal-runaway hazard-pull default.
• U.S. EPA and Call2Recycle, lithium-ion battery waste-stream hazard guidance — battery mislabeling and thermal-runaway risk independent of visible casing damage.
• NWRA (National Waste & Recycling Association), Certified MRF Operator safety materials — hazardous-item identification and line-stop protocol conventions referenced in the classification table.
• OSHA 29 CFR 1910.212, General Requirements for All Machines — point-of-operation/nip-point guarding baseline underlying the picking-station reach-zone discipline.
• NIOSH Fatality Assessment and Control Evaluation (FACE) program reports on materials-recovery-facility sorting-line injuries — the documented pattern of reach-related amputation injury at picking stations.
• No direct MRF-sorter practitioner has reviewed this file yet — flag corrections or gaps via PR.