Collection & Capture System — Methods, Analysis & Architecture

Draft Medium Research 7,578 words Created Mar 4, 2026

Collection and Capture System -- The Hardest Problem

> Status: Deep research -- active > Last updated: 2026-03-04 > Core question: How does the plastic actually get from the ocean onto The Claw?

This is the single hardest engineering problem in the entire project. The Great Pacific Garbage Patch is not a floating island -- it is a diffuse soup of wildly different debris types spread across 1.6 million square kilometers. There is no single collection method that handles everything. This document examines every credible approach, compares them honestly, and proposes a multi-system architecture for The Claw.

1. The Debris Field -- What We Are Actually Collecting

1.1 Composition by Size Class

The GPGP contains an estimated 1.8 trillion plastic pieces weighing approximately 80,000-100,000 metric tonnes. The size distribution tells two completely different stories depending on whether you measure by count or by mass:

Size Class	Definition	% by Count	% by Mass
Microplastics	0.5mm - 5mm	~94%	~8%
Mesoplastics	5mm - 2cm	~4%	~9%
Macroplastics	2cm - 50cm	~1.5%	~33%
Megaplastics	>50cm	~0.5%	~50%

The critical takeaway: Microplastics dominate by count (94% of all particles) but represent only 8% of the total mass. Conversely, mega-debris (ghost nets, large fragments, buoys) makes up less than 1% of particles but approximately 50% of the total mass. Any collection system must handle both ends of this spectrum -- you cannot ignore either.

1.2 Composition by Source

A landmark 2022 study found that 75-86% of plastic debris in the GPGP originates from fishing activities at sea, not land-based consumer waste. This is critical for understanding what we are actually collecting:

Ghost nets and fishing line: ~46% of total mass. Tangled masses of nylon netting, some weighing multiple tonnes. These are the single largest debris category by mass.
Fishing gear (non-net): Ropes, oyster spacers, eel traps, crates, baskets, floats -- a significant additional fraction.
Rigid consumer debris: Bottles, containers, crates, drums, buoys -- relatively rare but present.
Film plastics: Bags, packaging -- heavily degraded, often fragmenting into micro/meso.
Foam: Polystyrene fragments, dock floats -- low density, high volume.
Pellets (nurdles): Pre-production pellets from industrial spills -- 3-5mm, scattered widely.

1.3 Ghost Nets -- The Monster in the Patch

Ghost nets deserve special attention because they are the single hardest debris type to handle:

Individual ghost nets can weigh 5-10+ tonnes when waterlogged.
They form massive tangled balls that cannot be pumped, filtered, or conveyed.
They act as aggregators -- smaller debris gets caught in them, marine life gets entangled.
Between 500,000 and 1 million tonnes of fishing gear are lost or discarded globally every year.
Recovery requires crane/davit operations or heavy-duty winches -- no passive system can handle them.
The Ocean Voyages Institute has pioneered GPS tracker attachment to ghost nets, allowing them to be tracked and recovered by dedicated vessels.

1.4 Depth Distribution

Plastic does not all float neatly on the surface. The vertical distribution is driven by three factors: material density (PE and PP float, PVC and PET sink), biofouling (algae/barnacle growth increases density over time), and wind-driven mixing (storms push buoyant plastic underwater temporarily).

Depth Zone	What is Found There	Collection Feasibility
Surface (0-0.5m)	Buoyant PE/PP fragments, fresh debris, foam, some nets	Highest -- accessible to all methods
Near-surface (0.5-5m)	Partially biofouled plastics, submerged net masses, film plastic	Moderate -- boom screens (4m depth), trawls
Sub-surface (5-20m)	Heavily biofouled particles, denser plastics, sinking microplastics	Difficult -- requires trawls or pump systems
Deep (>20m)	PVC, PET, fully biofouled particles, settled microplastics	Impractical for surface vessel collection

Key research findings on depth:

Microplastic abundances generally decrease with depth, but the decrease is much steeper nearshore than offshore.
Small microplastics (<100 um) show relatively even distribution throughout the water column offshore -- they are effectively uncollectable.
Large microplastics (100 um - 5mm) show a two-orders-of-magnitude decrease with depth in offshore waters.
Biofouled particles oscillate: they sink as biofilm grows, then rise again when biofilm sloughs off in darkness at depth. This creates a dynamic "yo-yo" effect.

Practical implication for The Claw: Realistically, we can only collect from 0-5m depth. The Ocean Cleanup's System 03 screen extends 4m below the surface, which captures the vast majority of mass-significant debris. Anything deeper is largely microplastics that no current technology can economically recover at open-ocean scale.

1.5 Density Variation Across the GPGP

The GPGP is not uniform. Concentrations vary by orders of magnitude from center to edge:

Zone	Concentration (kg/km2)	Description
Core hotspots	100-1000+	Visible accumulations, debris clumps
Inner patch	10-100	Regular debris sightings, economically collectible
Outer patch	1-10	Sparse, long distances between items
Boundary	<1	Effectively background ocean levels

Average across the entire modeled GPGP area: approximately 10-100 kg/km2. But this average is misleading -- the distribution is highly uneven. The Ocean Cleanup's 2025 "hotspot hunting" initiative (using AI cameras and computational modeling) aims to map these concentrations precisely because operating in a hotspot versus a sparse zone can mean a 100x difference in collection efficiency.

At the average concentration of ~50 kg/km2: to collect 10 tonnes per day, you would need to sweep approximately 200 km2 of ocean surface. At 5 knots with a 2km-wide collection span, that takes roughly 20 hours of continuous operation -- feasible but tight. In a hotspot at 500 kg/km2, you would only need to sweep 20 km2 -- much more comfortable. In a sparse zone at 5 kg/km2, you would need 2,000 km2 -- completely impractical.

This means The Claw must operate in or near hotspots, not randomly wander the GPGP.

1.6 Seasonal Variation

Surface plastic concentrations in the GPGP vary significantly by season:

Summer (calm): Higher surface concentrations. Stratified water column keeps buoyant plastic at the surface. Best operating conditions.
Winter (stormy): Wind-driven mixing pushes buoyant plastic below the surface. Surface concentrations can drop by a factor of 6-7x. Wave heights limit operations. Worst operating conditions.
The GPGP center also shifts longitudinally with seasons, oscillating around 145 degrees W, moving east in summer and west in winter.

Operational implication: The Claw should plan for a seasonal operating window (roughly April-October) with reduced or repositioning operations in winter months. This aligns with weather windows for open-ocean operations anyway.

1.7 What Does a Collection Area Actually Look Like?

This is perhaps the most important mental model to get right. The GPGP is not a solid mass. At typical concentrations:

In a sparse zone (5 kg/km2): You might see one piece of visible debris every few hundred meters. It looks like open ocean with occasional litter.
In a moderate zone (50 kg/km2): A piece of debris every 10-50 meters. Noticeable but not dense.
In a hotspot (500+ kg/km2): Debris visible in all directions, some clumping around ghost net nuclei. This is where collection is economically viable.
Particle count reality: At 4 particles per cubic meter (measured average), a cubic meter of GPGP water contains roughly 4 tiny fragments. It is a dilute soup, not a trash heap.

2. Collection Method #1: Towed Boom/Barrier System

2.1 How Ocean Cleanup System 03 Works

System 03 is the current state of the art and the only proven large-scale open-ocean plastic collection technology. Key specifications:

Barrier length: ~2.2 km (2,200m / 1.4 miles)
Screen depth: 4 meters below surface
Towing: Two vessels hold the barrier in a U-shape
Speed: ~1.5 knots relative to water (very slow)
Sweep rate: Approximately one football field (0.005 km2) every 5 seconds
Collection mechanism: Water flows through/under the screen; plastic accumulates in the retention zone at the U's base
Extraction: Every ~4 days, the retention zone is lifted aboard a support vessel, sealed, and emptied
Capacity: ~5x the capacity of the previous System 002

2.2 Actual Performance Data

From Ocean Cleanup's 2024 operations (their best year to date):

Peak collection rate: 75 kg/hour -- an all-time high
Target rate: 100 kg/hour (not yet consistently achieved)
Total extracted in 2024: Organization reached 20 million kg cumulative (from 10M in April to 20M in November 2024 -- this includes river interceptors)
System 03 specifically: 112 extractions performed in 2024, refining efficiency, reliability, and marine life safety
Projected fleet: Modeling suggests ~10 full-size System 03 units could clean the entire GPGP
Cost estimate: $7.5 billion over 10 years, or $4 billion in 5 years with accelerated deployment

2.3 Limitations

Weather: Cannot operate in significant sea states (roughly >2.5m wave height). Winter storms force suspension.
Ghost nets: Large tangled net masses cannot be funneled -- they can damage or clog the retention zone.
Rigid mega-debris: Large buoys, drums, or container fragments may not flow into the retention zone.
Depth: Only collects from 0-4m. Anything below is missed.
Speed: 1.5 knots is very slow. Coverage rate is limited.
Escape: Some plastic escapes under the screen (especially small fragments in turbulent water) or flows over the boom in waves.
Logistics: Requires two dedicated towing vessels plus support vessel for extraction. Heavy crew and fuel requirements.
Bycatch: 99.7% plastic by mass, but some fish, turtles, and marine life do enter the retention zone (see Section 12).

2.4 Adapting for The Claw

Could The Claw deploy its own boom system instead of relying on separate vessels?

Possibility 1: Ship-deployed booms

Deploy shorter booms (500m-1km) from the vessel's own deck, towed by the ship's forward motion
Advantages: No separate towing vessels, integrated collection-to-processing pipeline
Challenges: Aframax tanker maneuverability (245m LOA, 44m beam) is poor -- cannot maintain tight U-shapes at 1.5 knots. Tankers have enormous turning circles.
Boom management (deployment, retrieval, maintenance) requires significant deck equipment and crew.

Possibility 2: Cooperative boom with support vessel

One tug-class support vessel tows one end of a boom while The Claw tows the other
More practical than solo deployment but still requires a support vessel
Boom length could be 1-1.5 km -- shorter than System 03 but meaningful

Possibility 3: Use Ocean Cleanup's system separately, receive collected plastic

The Claw acts purely as a processing platform, receiving plastic from Ocean Cleanup-style collection vessels
Simplest integration but makes The Claw dependent on external collection operations

Assessment: A self-deployed boom system is possible but suboptimal for an Aframax hull. The vessel is too large and unwieldy for the precise, slow maneuvering that boom operations require. A hybrid approach (boom deployed with one support vessel) or a feeder-vessel model is more practical.

2.5 Verdict

Aspect	Rating
Proven at scale	YES -- only method with real open-ocean track record
Throughput	Moderate (75 kg/hr proven, 100 kg/hr target)
Debris types	Surface macro/mega (not ghost nets, not microplastics)
Weather tolerance	Poor -- calm seas required
Deck space	Large (boom storage, winches, extraction gear)
Complexity	High (two vessels, specialized boom, extraction protocol)
Recommended for The Claw	As FEEDER SYSTEM only -- not self-operated

3. Collection Method #2: Conveyor/Belt Intake

3.1 Concept

Side-mounted or bow-mounted conveyor belt systems that scoop debris from the waterline as the vessel moves through debris fields. Similar in principle to beach cleanup machines, but adapted for marine deployment.

3.2 Real-World Examples

The Manta (SeaCleaners)

56m catamaran with three conveyor belts between hulls
Collection capacity: 1-3 tonnes per hour (claimed)
Annual target: 5,000-10,000 tonnes per year
Collects debris from 10mm upward, up to 1m depth
Onboard pyrolysis converts collected waste to energy
Companion Mobula vessels extend collection capacity to ~72 tonnes/day combined
Status: Still in development/construction as of 2025. Not yet operationally proven at sea.

Catamaran-type conveyor vessels

Several designs use twin-hull configurations with conveyor belts between the hulls
Debris sizes from 10mm to 1m
Maximum operational speed typically 3-5 knots

River Interceptors (Ocean Cleanup)

Barrier directs river plastic onto a conveyor belt into collection barges
Solar-powered, can remove 50+ tonnes/day in high-concentration river environments
Proven technology but rivers are orders of magnitude more concentrated than open ocean

3.3 Adapting for The Claw

An Aframax tanker is a monohull, not a catamaran, so the between-hull conveyor approach does not apply. Options:

Side-mounted conveyors: Deployed from the vessel's side, extending below the waterline. Marine conditions (rolling, pitching) make this mechanically challenging. The conveyor entrance must stay at the waterline despite vessel motion -- requires active compensation or very robust design.

Bow-mounted scoop with conveyor: A forward-facing scoop channels debris onto a conveyor that lifts material to deck level. See also Section 9 (Passive Current Funneling).

3.4 Throughput Estimates

At GPGP concentrations, a conveyor system can only collect what passes through its intake area. For a 5m-wide conveyor at 3 knots:

Sweep area: 5m x 5,556 m/hr = 27,780 m2/hr = 0.028 km2/hr
At 50 kg/km2: 1.4 kg/hr -- completely inadequate
At 500 kg/km2 (hotspot): 14 kg/hr -- still very low
At 5,000 kg/km2 (extreme hotspot): 140 kg/hr -- only viable in the densest zones

The Manta's claimed 1-3 TPH rates must assume coastal/river-mouth concentrations far higher than open-ocean GPGP levels. In the GPGP, standalone conveyor systems cannot deliver meaningful tonnage.

3.5 Verdict

Aspect	Rating
Proven at scale	NO -- Manta not yet operational, river interceptors proven but different context
Throughput in GPGP	Very low without a funneling/concentration mechanism
Debris types	Surface macro, 10mm-1m range
Weather tolerance	Moderate -- depends on conveyor design and sea state compensation
Deck space	Moderate (conveyor machinery, collection bin)
Complexity	Moderate
Recommended for The Claw	Only as SECONDARY system paired with a concentration method (boom/funnel)

4. Collection Method #3: Moonpool Intake

4.1 Concept

A moonpool is a vertical opening through the hull of a vessel, providing sheltered access to the water below. Common in drilling ships, dive support vessels, and research vessels. The concept here: use a moonpool with a suction/pump system to draw water and debris into the vessel, then separate plastic from seawater onboard.

4.2 Marine Engineering Background

Moonpools are well-established technology. The REV Ocean research vessel has a large moonpool for deploying instruments and ROVs.
Typical moonpool sizes range from 5m x 5m to 10m x 15m on drilling vessels.
An Aframax hull conversion could accommodate a moonpool, but it would require significant structural modification -- cutting through the double-bottom and reinforcing the surrounding structure. This is a major shipyard operation.
Historical note: Transocean considered converting an Aframax tanker for drilling (with moonpool) but abandoned the concept due to concerns that the moonpool was too large for the hull structure and the tanker could not cope with increased facilities demands.

4.3 How It Would Work for Plastic Collection

1. Moonpool opening (perhaps 3m x 6m) in the hull bottom, forward of midships 2. Vessel moves slowly through debris field (1-3 knots) 3. Forward motion channels water and debris through the moonpool 4. Above the moonpool, a screening system (rotating drum screen, vibrating screen, or bar screen) separates debris from water 5. Water returns to the ocean; plastic moves to a conveyor leading to pre-processing 6. For microplastics, finer filtration stages could be added

4.4 Advantages

Weather-independent: The intake is internal, sheltered from waves and wind. Operations could continue in higher sea states than any external collection method.
Continuous operation: No deployment/retrieval cycles. The system operates as long as the vessel is moving.
Compact: No external booms or outriggers. All equipment is internal.
Controllable flow: Pumps can regulate intake volume. Screens can be swapped for different debris sizes.

4.5 Disadvantages

Clogging: Ghost nets, rope tangles, and large rigid debris will jam any intake system. A moonpool intake could be catastrophically blocked by a single ghost net.
Cannot handle mega-debris: Anything larger than the moonpool opening (or the screening system) cannot enter. Large ghost nets (multi-tonne masses) are completely excluded.
Energy-intensive: Pumping large volumes of seawater requires significant power. To process enough water to collect meaningful plastic at GPGP concentrations requires enormous flow rates (see Section 8).
Structural compromise: Cutting a moonpool into an Aframax hull weakens the structure and requires extensive reinforcement. Classification societies may have concerns.
Limited sweep width: The moonpool only captures what passes directly beneath the hull. Even at the vessel's full 44m beam, the effective collection width is just the moonpool opening (3-6m).

4.6 Flow Rate Analysis

At 4 particles per cubic meter and assuming an average particle mass of ~10mg (weighted average across size classes):

4 particles x 10mg = 40mg per m3 = 0.04g per m3
To collect 10 tonnes (10,000 kg) per day: need to process 250,000,000 m3 of water per day
That is 2,894 m3 per second -- roughly the flow of a medium-sized river
This is physically impossible with a moonpool system

Even at hotspot concentrations 100x higher (4g per m3):

Need to process 2,500,000 m3/day = 29 m3/s
Still extremely high. A large industrial pump moves ~1-5 m3/s.
Would need 6-30 industrial pumps running continuously.

The math does not work for microplastics. However, for macro-debris in high-density zones, a moonpool could supplement other methods by providing a continuous, weather-protected intake for moderate-sized debris.

4.7 Verdict

Aspect	Rating
Proven at scale	NO -- no open-ocean debris collection moonpool exists
Throughput in GPGP	Very low for micro/meso; moderate for macro in hotspots
Debris types	Small-to-medium macro only. Cannot handle nets or mega-debris
Weather tolerance	EXCELLENT -- best of any method
Deck space	Moderate (internal -- does not consume deck space, but consumes hull volume)
Complexity	Very high (hull modification, pumping systems, screening, clog management)
Recommended for The Claw	POSSIBLE as weather-protected supplementary system, but not primary

5. Collection Method #4: Crane/Davit Net Recovery

5.1 Concept

Traditional marine salvage approach: deploy nets or grapnels from cranes or davits, manually or semi-automatically recover large debris items. This is how fishing vessels recover lost gear and how salvage operations handle marine debris.

5.2 How It Works

1. Vessel identifies large debris items (ghost nets, buoy clusters, large fragments) visually or by radar/camera 2. Crane swings a grab net, grapnel hook, or lifting sling over the debris 3. Debris is lifted aboard and deposited on deck or into a collection bin 4. Crew may need to cut or section very large items during the lift

5.3 Equipment Requirements

Deck cranes: Typical offshore crane capacity of 20-50 tonnes SWL (safe working load). An Aframax conversion would likely have 1-2 cranes installed. A single marine crane (e.g., Liebherr CBB or Palfinger marine crane) costs $500K-$2M.
Davits: Smaller lifting devices for lighter operations (1-5 tonne capacity).
Nets/grabs: Specialized grab nets for loose tangled material. Standard marine lifting slings for rigid items.
Working deck: Clear deck area of at least 200-400 m2 for staging recovered debris before it moves to pre-processing.

5.4 Throughput

This method is slow and labor-intensive:

Each lift cycle: spot debris (5-15 min), position vessel (5-20 min), deploy crane/net (5-10 min), secure load (5-15 min), lift and stage (5-10 min). Total: 25-70 minutes per lift.
If average lift recovers 500 kg (a moderate ghost net section): ~500 kg per hour in a target-rich area.
In practice, finding and positioning alongside large debris items takes significant time. Realistic sustained rate: 200-500 kg/hr in good conditions.

5.5 When This Method Is Essential

Crane recovery is the only method that can handle:

Ghost nets weighing multiple tonnes
Large rigid debris (drums, crates, container fragments)
Aggregated debris masses too large for any screen or conveyor
Items that would damage or clog other collection systems

It is not the primary volume collector, but it is indispensable for the ~50% of GPGP mass that consists of large fishing gear.

5.6 Verdict

Aspect	Rating
Proven at scale	YES -- standard marine salvage technique
Throughput in GPGP	Low (200-500 kg/hr), episodic
Debris types	MEGA-DEBRIS SPECIALIST -- ghost nets, large rigid items
Weather tolerance	Moderate -- crane operations limited to ~3m significant wave height
Deck space	Large (crane footprint + staging area, 300-500 m2)
Complexity	Low-moderate (proven equipment, requires skilled crane operators)
Recommended for The Claw	YES -- ESSENTIAL secondary system for mega-debris

6. Collection Method #5: Drone-Assisted Collection

6.1 Concept

Autonomous surface vessels (ASVs) or unmanned surface vessels (USVs) patrol the area around The Claw, identifying and herding or collecting debris, extending the effective collection radius well beyond the ship's immediate vicinity.

6.2 Current State of Marine Drone Technology

WasteShark (RanMarine Technology)

Small ASV inspired by whale shark feeding
Capacity: 180 liters, 500 kg/day
Runtime: 8 hours on battery
Speed: Max 6 km/h (3.2 knots)
Deployed in 110+ locations across 34 countries -- but all in harbors, marinas, and sheltered waters
Not designed for open ocean conditions

Military/Research USVs

Saronic (US Navy contract, $392M): AI-powered ASVs capable of autonomous or swarm operations
Maritime Robotics Mariner X: Designed for long-endurance ocean surveys
These are survey/patrol platforms, not debris collectors, but the autonomy and ocean-going capability is proven

Ocean-Going ASV Limitations

Battery life: Most ASVs operate 8-24 hours before recharging. Solar extends this but limits speed.
Sea state: Small ASVs (under 5m) are limited to Beaufort 3-4 (moderate seas). Larger USVs (10m+) can handle rougher conditions.
Payload: Most existing ASVs carry 100-500 kg. Not enough for meaningful debris collection.
Communications: Reliable comms at 10-50 km range is feasible. Beyond that, satellite required.

6.3 Possible Roles for Drones Around The Claw

Role 1: Scout/Survey

Drones equipped with cameras and AI identification (similar to Ocean Cleanup's ADIS system) patrol a 20-50 km radius
Map debris density and type in real-time
Guide The Claw toward hotspots
Most practical near-term role

Role 2: Herd/Funnel

Multiple drones create a moving barrier that pushes surface debris toward the ship
Similar to how dolphins herd fish
Requires coordinated swarm behavior -- technically challenging but plausible
Speed differential only needs to be ~0.5-1 knot relative to plastic drift

Role 3: Collect and Return

Drones collect debris autonomously and return to The Claw for offloading
Limited by payload capacity (100-500 kg per trip)
At 500 kg per trip and 2-hour cycle time: 12 trips/day x 500 kg = 6 tonnes/day (with a fleet of drones)
Would need 10-20 drones for meaningful contribution

6.4 Verdict

Aspect	Rating
Proven at scale	NO -- sheltered-water systems exist, no open-ocean debris collection drones
Throughput in GPGP	Low individually; potentially meaningful as fleet
Debris types	Surface macro (scout/herd role works for all types)
Weather tolerance	Poor for small drones; moderate for larger USVs
Deck space	Moderate (drone storage, charging stations, launch/recovery)
Complexity	Very high (autonomy, ocean conditions, fleet management)
Recommended for The Claw	YES for SCOUTING. Phase 2+ for herding/collection.

7. Collection Method #6: Trawl Nets

7.1 Concept

Modified fishing trawls deployed from The Claw to sweep through debris fields. Different net types for different debris sizes: neuston nets for surface microplastics, mid-water trawls for subsurface debris, and heavy-duty trawls for ghost nets.

7.2 Net Types

Manta Trawl / Neuston Net (Research-Scale)

Standard oceanographic sampling tool
Samples top 15-25 cm (manta) or top 50 cm (neuston) of water column
Mouth opening: typically 0.5m x 0.25m (manta) to 1m x 0.5m (neuston)
Mesh size: 333-335 um (standard), though ~60% of microplastics under 1mm pass through
Tow speed: 1-3 knots
Collection rate: grams per hour -- purely a sampling tool, not a collection tool
Cannot be scaled to industrial collection without a fundamentally different design

AVANI Trawl

Can collect at speeds up to 8 knots (vs 1-3 for manta/neuston)
"Skis" across the surface -- suitable for vessels underway on transit
Still a sampling tool in terms of throughput

Industrial-Scale Modified Trawl

A wide-mouth (10-20m) trawl with appropriate mesh could theoretically collect macro-debris
Would need to be designed specifically -- no off-the-shelf product exists
Bycatch is a major concern (see Section 12)
Tow resistance increases dramatically with mouth size and speed

7.3 Scaling Challenge

The fundamental problem with trawling for plastic in the GPGP is concentration:

A 10m-wide trawl at 3 knots sweeps ~55,560 m2/hr (0.056 km2/hr):

At 50 kg/km2: 2.8 kg/hr -- pathetic
At 500 kg/km2: 28 kg/hr -- still very low
At 5,000 kg/km2: 280 kg/hr -- only viable in extreme hotspots

To match the boom system's 75 kg/hr, you would need a 270m-wide trawl at typical concentrations -- physically impossible for a single vessel.

7.4 Verdict

Aspect	Rating
Proven at scale	For FISHING, yes. For PLASTIC COLLECTION, no
Throughput in GPGP	Very low -- fundamentally limited by concentration
Debris types	Surface/near-surface macro and meso
Weather tolerance	Moderate (standard trawling limits apply)
Deck space	Moderate (net handling, winches, stern ramp or gallows)
Complexity	Moderate (proven fishing technology)
Recommended for The Claw	NO as primary. POSSIBLE as supplementary for targeted meso-debris recovery

8. Collection Method #7: Pump and Filter

8.1 Concept

High-volume seawater intake through hull-mounted sea chests or suction inlets, passed through multi-stage filtration to remove plastic particles, then clean water returned to the ocean. Essentially industrial-scale water treatment at sea.

8.2 The Math Problem

This method fails on basic arithmetic at GPGP concentrations.

GPGP plastic concentration: ~4 particles per cubic meter, averaging ~0.04g per m3 of water (roughly 40 micrograms per liter).

To collect 10 tonnes (10,000,000 g) per day:

Water volume needed: 10,000,000 / 0.04 = 250,000,000 m3/day = 2,894 m3/s

For context:

A large ship's ballast pump moves ~5,000 m3/hr = 1.4 m3/s
You would need ~2,000 large ballast pumps running simultaneously
The power requirement at ~3 kWh/m3 (minimum, for coarse filtration only) would be 750 MW per day -- roughly the output of a nuclear power plant

Even in a hotspot with 100x concentration:

Still need 2,500,000 m3/day = 29 m3/s
~20 large pumps, ~7.5 MW continuous power
Technically possible but enormously energy-intensive for modest collection rates

8.3 When It Could Work

The only scenario where pump-and-filter makes sense is post-concentration: if another method (boom, funnel, herding) has already concentrated plastic into a small area of water, pump-and-filter can extract the concentrated debris. This is essentially what happens in the Ocean Cleanup retention zone -- the boom concentrates, then the concentrated mass is removed.

8.4 Verdict

Aspect	Rating
Proven at scale	YES (for water treatment) but NO for open-ocean plastic at GPGP concentrations
Throughput in GPGP	Physically impossible at ambient concentrations
Debris types	Microplastics (with fine filtration) -- but volume problem prevents it
Weather tolerance	Good (internal system)
Deck space	Enormous (pump rooms, filtration banks, water discharge)
Complexity	Very high
Recommended for The Claw	NO as standalone. YES as post-concentration processing step

9. Collection Method #8: Passive Current Funneling

9.1 Concept

Using the vessel's own forward motion to channel water and debris into an intake. A bow-mounted scoop, funnel, or widening channel narrows toward an intake point, concentrating debris as the vessel moves forward. Analogous to a whale shark's feeding strategy or a baleen whale's lunge-feeding.

9.2 Design Approaches

Bow scoop: A rigid or semi-rigid scoop extending forward and to the sides of the bow, tapering to an intake. Width: 10-30m at the forward edge, narrowing to 3-5m at the intake.

Hull-mounted deflectors: Angled plates welded to the hull sides that channel surface debris toward a centerline intake (potentially a moonpool -- see Section 4).

Towed funnel: A floating funnel structure towed ahead of the vessel, connected by conveyor or pipe to the deck.

9.3 Throughput Estimates

A 20m-wide funnel at 3 knots:

Sweep: 20m x 5,556 m/hr = 111,120 m2/hr = 0.11 km2/hr
At 50 kg/km2: 5.6 kg/hr
At 500 kg/km2: 56 kg/hr
At 5,000 kg/km2: 556 kg/hr

Better than a conveyor alone (because of the concentration effect) but still modest except in hotspots. A wider funnel (40-50m) approaching the vessel's beam would roughly double these numbers.

9.4 Advantages

Simple and reliable: Few moving parts. Passive concentration via geometry.
Low energy: Only requires vessel propulsion, no pumps or motors for collection itself.
Continuous: Operates whenever the vessel is underway.
Scalable: Width can be adjusted. Could deploy wider funnel arms in calm weather, retract in storms.

9.5 Disadvantages

Speed-limited: Must move slowly (1-3 knots) to prevent debris from washing over or around the funnel.
Cannot handle mega-debris: Ghost nets would jam the funnel throat.
Surface only: Funnel depth limited to ~1-2m before drag becomes excessive.
Low throughput at typical concentrations: Only viable in moderate-to-high density zones.

9.6 Verdict

Aspect	Rating
Proven at scale	NO -- concept stage for ocean plastic; proven for river debris (Interceptor)
Throughput in GPGP	Low-moderate (depends heavily on concentration zone)
Debris types	Surface macro/meso, 10mm-50cm
Weather tolerance	Moderate (funnel can be retracted in storms)
Deck space	Moderate (bow area, funnel storage, intake screening)
Complexity	Low-moderate (relatively simple engineering)
Recommended for The Claw	YES -- strong candidate for PRIMARY continuous collection

10. The Multi-System Approach

10.1 Why No Single Method Works

The debris field analysis (Section 1) makes clear that the GPGP contains:

Micro/meso particles (94% by count, 8% by mass) -- too small and diffuse for any practical open-ocean collection
Macro-debris (2cm-50cm, ~33% by mass) -- collectible by booms, funnels, and conveyors
Mega-debris including ghost nets (~50% by mass) -- requires crane/davit recovery

No single method handles all three. The economically recoverable fraction (macro and mega, ~83% by mass) requires at least two different approaches: an area-sweep method for distributed macro-debris and a targeted lifting method for large items.

10.2 Proposed System Combination

Debris Type	Primary Method	Secondary Method	% of Collected Mass
Ghost nets (>1 tonne)	Crane/davit recovery	Manual cutting + winch	~30%
Large rigid items (50cm-5m)	Crane/davit recovery	Passive funnel (if fits)	~15%
Macro-debris (2cm-50cm)	Passive bow funnel + conveyor	Boom barrier (with support vessel)	~40%
Meso-debris (5mm-2cm)	Passive bow funnel (fine screen)	Post-concentration filtration	~10%
Micro-debris (<5mm)	Not practically collectible in open ocean	Incidental capture only	~5%

10.3 System Integration on the Vessel

Bow zone (forward 50m):

Passive funnel structure (deployable arms, 20-40m width)
Funnel throat leads to screening/dewatering station
Conveyor from screening station to pre-processing area

Midships zone (amidships 100m):

Pre-processing area (port side): shredding, sorting, de-salting, drying
Buffer storage (center): 3-5 day feedstock reserve
Plasma processing (starboard side or below deck): reactor, syngas handling, power generation

Aft zone (aft 80m):

Deck crane(s): 1-2 marine cranes, 25-50 tonne SWL, positioned for port and starboard reach
Ghost net staging area: open deck for large debris before cutting/sectioning
Drone operations: ASV storage, charging, launch/recovery

Moonpool (optional, forward of midships):

Weather-protected intake for continuous moderate-debris collection
Screening drum above moonpool
Bypass/isolation gate for clog clearing

10.4 Deck Space Allocation

Aframax tanker main deck: approximately 245m x 40m = ~9,800 m2 gross deck area. After superstructure, piping, walkways, etc., usable deck might be ~6,000 m2.

System	Deck Space Required	Notes
Passive funnel + conveyor	~400 m2	Bow area, mostly overhanging
Pre-processing (shred/sort/dry)	~800 m2	Port side midships
Buffer storage	~600 m2	Below deck in converted cargo tanks
Plasma reactor + auxiliaries	~1,200 m2	See vessel-internal-layout.md
Crane(s) + staging	~500 m2	Aft deck
Drone operations	~200 m2	Aft deck
Crew/accommodation	~800 m2	Existing superstructure
Boom storage (if used)	~400 m2	Optional, aft deck
Total	~4,900 m2	Fits within available deck area

10.5 Crew Requirements

System	Crew Needed	Role
Vessel operation	8-12	Bridge, engine room, navigation
Passive funnel + conveyor	2-4	Monitoring, clearing jams, maintenance
Crane operations	3-4	Crane operator, riggers, signalman
Pre-processing	4-6	Sorting, shredder operation, monitoring
Plasma reactor	3-4	Reactor monitoring, maintenance
Drone operations	1-2	Mission planning, maintenance
Total	21-32	Typical for offshore operations vessel

11. Pre-Processing Pipeline

11.1 Why Pre-Processing Is Critical

The plasma gasification reactor cannot accept raw ocean debris. Marine plastic is:

Wet: Saturated with seawater (salt content ~35g/L)
Mixed: Plastic, organic matter, marine life, barnacles, algae, metal, rope, glass
Variable size: From 5mm fragments to 10-tonne ghost nets
Contaminated: Biofouled surfaces, potentially hazardous chemicals (antifouling paint, fuel residues)

The reactor needs relatively uniform, dry, sized feedstock. The pre-processing pipeline bridges the gap.

11.2 Stage 1: Gross Sorting (Deck Level)

Purpose: Remove items that cannot enter the processing chain.

Marine life (returned to sea immediately -- see Section 12)
Metal items (drums, cans, fittings) -- separated for recycling or storage
Hazardous materials (fuel containers, chemical drums) -- isolated for proper disposal
Oversized items requiring cutting/sectioning
Non-plastic debris (glass, wood, natural fiber rope)

Method: Manual sorting on a conveyor belt or sorting table. 2-3 crew members.

Throughput: Matched to incoming collection rate. At 400-500 kg/hr incoming, this is manageable.

11.3 Stage 2: Size Reduction

Purpose: Create uniform particle size for downstream processing.

Ghost nets and large items: Industrial shredder with pre-cutting. Ghost nets are extremely tough (nylon, HDPE) and resist shredding. Require:

Pre-cutting with hydraulic shears or circular saws to reduce to ~1m lengths
Heavy-duty shredder (e.g., twin-shaft industrial shredder, 50-100 kW) for primary size reduction to ~50mm
Secondary granulator for further reduction to ~10-20mm if needed by reactor

Rigid plastics: Standard single-shaft shredder handles bottles, crates, fragments. Output: 20-50mm chips.

Film plastics: Tend to wrap around shredder shafts. May need a dedicated film shredder or agglomerator.

Power requirement: 50-150 kW total for shredding operations.

11.4 Stage 3: Washing and De-Salting

Purpose: Remove salt, sand, organic matter, and surface contaminants.

Why salt matters: Chlorine from salt (NaCl) in the plasma reactor produces hydrochloric acid (HCl) in the syngas. This is corrosive to downstream equipment and must be scrubbed. Reducing salt input reduces HCl scrubbing load. Additionally, inorganic salts increase slag production and decrease syngas quality.

Method:

Freshwater rinse (can use desalinated seawater from onboard RO system)
Friction washer for surface cleaning
Density separation (float/sink tank) to separate different plastic types and remove sand/grit
Estimated freshwater consumption: 2-5 m3 per tonne of plastic processed

Note: Perfect de-salting is not required. Plasma gasification can handle some salt -- the scrubber system handles residual HCl. But reducing salt input is worthwhile.

11.5 Stage 4: Dewatering and Drying

Purpose: Remove water to improve reactor efficiency. Wet feedstock reduces syngas production and increases energy consumption.

Methods in sequence: 1. Gravity drain: Simple but slow. Conveyor with perforated bed. Removes bulk water. 2. Squeeze/press: Mechanical dewatering via screw press or roller press. Reduces moisture to ~20-30%. 3. Thermal drying: Waste heat from plasma reactor used to drive a rotary dryer or belt dryer. Reduces moisture to ~5-10%.

Energy: Thermal drying can use reactor waste heat (available from syngas cooling and slag quenching). This is an efficient heat recovery loop.

Target moisture content: <10% for optimal reactor feed.

11.6 Stage 5: Buffer Storage

Purpose: Decouple the variable collection rate from the steady processing rate.

Collection is inherently variable:

Good day in a hotspot: might collect 15-20 tonnes
Bad day (weather, sparse area, transit): might collect 1-2 tonnes or zero
Crane recovery of a large ghost net: sudden spike of several tonnes

The reactor needs steady feed: 5-10 tonnes per day, delivered at a constant rate.

Buffer sizing: 3-5 days of feedstock = 30-50 tonnes of processed (dry, shredded) plastic. At a bulk density of ~200 kg/m3 (shredded mixed plastic), this requires 150-250 m3 of storage. An Aframax cargo tank section can easily accommodate this.

11.7 Complete Pre-Processing Flow

Ocean debris (wet, mixed, variable)
    |
    v
[Gross sorting] --> Marine life (return to sea)
    |                Metal/hazmat (store)
    v
[Size reduction] --> Pre-cutting (large items)
    |                Shredding (50mm)
    |                Granulation (10-20mm, if needed)
    v
[Washing] --> Freshwater rinse
    |          Friction wash
    |          Density separation
    v
[Dewatering] --> Gravity drain
    |             Screw press
    |             Thermal drying (reactor waste heat)
    v
[Buffer storage] --> 30-50 tonne capacity
    |                 3-5 days reserve
    v
[Reactor feed] --> Metered conveyor to plasma reactor

12. Bycatch and Environmental Impact

12.1 The Problem

Any system that collects floating debris will also encounter marine life. The GPGP ecosystem includes:

Fish (especially juvenile stages that shelter under floating debris)
Sea turtles (attracted to floating objects for shelter and food)
Seabirds (resting on floating debris)
Marine mammals (dolphins, whales transiting the area)
Invertebrates (barnacles, crabs, worms colonizing floating plastic -- the "plastisphere")
Pelagic organisms (jellyfish, siphonophores, salps, neuston community)

12.2 Ocean Cleanup's Bycatch Data

From System 002's first 12 operational trips (the best available real-world data):

Total plastic collected: 193,832 kg
Total bycatch: 667 kg
Ratio: 99.7% plastic, 0.3% bycatch
Bycatch composition: Mostly fish, sharks, mollusks, and sea turtles
Conclusion: Very high selectivity, but not zero bycatch

12.3 Mitigation Measures (Proven and Planned)

Currently deployed by Ocean Cleanup:

Tow speed of 0.5-1.5 knots (gives mobile animals time to avoid)
Escape opening at the bottom of the retention zone (fish swim down and out)
Drone monitoring flights to watch for large marine life
Underwater cameras for real-time observation
Manual rescue operations when animals are spotted in retention zone
Visual cues (bright colors) and audio deterrents to help animals detect the system
Multiple breathing ports in retention zone for air-breathing animals

Under development:

AI-assisted detection of turtles entering the retention zone
Active turtle exclusion mechanisms
Improved exit routes
Shark decoys to deter turtles

Standard fishing industry techniques applicable to The Claw:

Turtle Excluder Devices (TEDs): Metal grids in nets that allow small debris through but deflect turtles through an escape hatch. Well-proven technology from shrimp trawling.
Acoustic deterrents (pingers): Emit sounds that repel marine mammals.
Night/day scheduling: Some species are more active at certain times. Data-driven scheduling could reduce encounters.
Depth-selective collection: Staying in the top 0-2m (rather than 0-5m) reduces interaction with some species.

12.4 Environmental Monitoring Requirements

The Claw should maintain:

Continuous bycatch logging (species, count, condition, outcome -- returned alive vs. mortality)
Marine mammal observers (MMOs) on watch during collection operations
Environmental impact baseline monitoring (water quality, plankton sampling)
Reporting to relevant authorities (NOAA, IMO, flag state)
An Environmental Management Plan approved by classification society and flag state

12.5 The Plastisphere Dilemma

A growing body of research shows that GPGP plastic hosts a unique ecosystem -- the "plastisphere" -- of microorganisms, algae, and invertebrates. Removing plastic also removes this ecosystem. Some scientists argue this is a net negative for biodiversity in the short term. However, the consensus view is that ocean plastic is overwhelmingly harmful (entanglement, ingestion, toxin transport) and removal is net positive.

12.6 Practical Protocol for The Claw

1. All collection intakes fitted with escape mechanisms appropriate to the method 2. Gross sorting (Section 11.2) includes immediate release of any marine life 3. Trained marine life handlers on crew (basic veterinary triage for turtles, seabirds) 4. Collection operations paused if whale or large marine mammal sighted within 500m 5. Bycatch data transmitted to shore daily for transparency 6. Annual third-party environmental audit

13. Collection Rate vs Processing Rate

13.1 The Fundamental Bottleneck

The Claw must balance two rates:

Collection rate: How fast debris comes aboard (highly variable)
Processing rate: How fast the plasma reactor consumes feedstock (relatively constant at 5-10 TPD)

If collection outpaces processing, buffer storage fills up and collection must stop. If processing outpaces collection, the reactor sits idle (wasting fuel/energy). The system must be tuned so that average collection meets or slightly exceeds average processing demand, with buffer storage smoothing the variability.

13.2 Collection Rate Estimates by Method

Based on analysis in Sections 2-9, estimated sustainable collection rates in a moderate-density zone (~50-200 kg/km2):

Method	Estimated Rate (kg/hr)	Hours/Day	Daily Yield (kg)
Passive bow funnel (20m wide, 2 kn)	15-60	20	300-1,200
Boom barrier (1km, with support vessel)	30-75	16	480-1,200
Crane/davit (ghost net recovery)	200-500	8	1,600-4,000
Moonpool intake (supplementary)	5-30	20	100-600
Drone-assisted herding (fleet of 5)	Additive (improves funnel/boom rates by 20-50%)	--	--
Combined total (realistic)	--	--	2,500-7,000

13.3 Can Collection Keep Up?

At 5 TPD processing: Combined collection of 2,500-7,000 kg/day means the system can keep up most days, especially in moderate-to-high density zones. Buffer storage smooths gaps.

At 10 TPD processing: Combined collection falls short on many days. Would need:

Consistent operation in high-density hotspots (>200 kg/km2)
Full crane utilization (multiple ghost net recoveries per day)
Boom system (with support vessel) in addition to passive funnel
Drone-assisted scouting to minimize time in sparse zones

Conclusion: 5 TPD processing is well-matched to achievable collection rates. 10 TPD is ambitious and may result in reactor underutilization unless operations consistently target hotspots.

13.4 Buffer Storage Sizing

Based on collection variability:

Good day: 7,000 kg collected
Average day: 4,000 kg collected
Bad day (weather/transit): 500 kg collected
Processing consumption: 5,000 kg/day (at 5 TPD)

A 5-day buffer at 25 tonnes (50 m3 at 500 kg/m3 compressed, or 125 m3 shredded) provides adequate smoothing. This allows 2-3 consecutive bad days without reactor shutdown.

14. Comparison Table

Method	Throughput (kg/hr)	Debris Types	Weather Limit (Hs)	CAPEX Estimate	OPEX (crew)	Deck Space (m2)	TRL	Priority
Towed boom (self)	30-75	Surface macro	2.5m	$2-5M	4-6	400	8 (proven)	Secondary
Conveyor belt	5-20	Surface macro	2.0m	$500K-1M	2-3	200	5 (unproven at sea)	Tertiary
Moonpool intake	5-30	Small macro	4.0m+	$3-8M	2-3	Internal	3 (concept)	Phase 2
Crane/davit	200-500	Mega, ghost nets	3.0m	$1-3M	3-4	500	9 (proven)	Primary (mega)
Drone ASVs	Additive	Scout/herd	2.0m	$2-5M (fleet)	1-2	200	4 (unproven)	Phase 2
Trawl nets	3-30	Macro/meso	3.0m	$200-500K	2-3	200	7 (proven for fish)	Not recommended
Pump and filter	N/A at GPGP	Micro (theory)	4.0m+	$5-10M	2-4	800	6 (for water treatment)	Not viable
Passive funnel	15-60	Surface macro/meso	2.5m	$500K-2M	2-4	400	4 (concept)	Primary (macro)

TRL = Technology Readiness Level (1-9 scale, 9 = fully operational)

15. Recommended Architecture

15.1 Design Philosophy

Accept three realities: 1. Microplastics cannot be economically collected in the open ocean. The concentrations are too low and the volumes too vast. Focus on macro and mega debris (92% of mass). 2. Ghost nets are half the mass but require completely different handling than distributed macro-debris. Two systems minimum. 3. Hotspot targeting is as important as collection technology. A 10x improvement in collection efficiency comes from operating in the right zone, not from better equipment.

15.2 Phase 1: Minimum Viable Collection (Year 1)

Primary System A -- Passive Bow Funnel

Deployable funnel arms, 20-30m width at forward edge, tapering to 4-5m intake
Screen depth: 1-2m
Funnel throat feeds onto dewatering conveyor leading to gross sorting station
Operates continuously while vessel is underway at 1.5-3 knots
Handles distributed macro and meso debris (2mm-50cm)
Estimated yield: 300-1,200 kg/day depending on concentration

Primary System B -- Deck Crane Recovery

One 30-tonne SWL marine deck crane (port side, aft)
Grab nets, grapnel hooks, lifting slings
Dedicated to ghost net and mega-debris recovery
Targets large items identified by bridge watch or drone scouts
Estimated yield: 1,600-4,000 kg/day

Support -- Scout Drones

2-3 ocean-capable ASVs with cameras + AI debris identification
Patrol 10-20 km radius, mapping debris density in real-time
Guide vessel navigation toward hotspots
No collection capability in Phase 1 -- pure intelligence gathering

Phase 1 Combined Yield: 2,000-5,000 kg/day Matches: 5 TPD processing target on good days Buffer storage: 25 tonnes (5 days reserve)

15.3 Phase 2: Optimized Collection (Year 2-3)

Additions:

Boom System (with support vessel)

1-1.5 km barrier deployed cooperatively with a tug/support vessel
U-shape tow at 1.5 knots, periodic extraction into The Claw
Adds 500-1,200 kg/day in suitable weather
Requires charter or purchase of one support vessel (tug class, 40-60m)

Second Crane

Starboard side, enabling simultaneous recovery operations
Doubles mega-debris throughput potential

Drone Fleet Expansion

5-8 ASVs with herding capability
Create moving barrier that funnels surface debris toward the vessel's funnel intake
Increases effective sweep width from 30m to 100-200m
Potentially doubles passive funnel yield

Moonpool (if hull survey permits)

3m x 5m opening forward of midships
Rotating drum screen for continuous moderate-debris extraction
Weather-protected -- operates when other systems cannot
Adds 100-600 kg/day

Phase 2 Combined Yield: 4,000-10,000 kg/day Matches: Full 10 TPD processing capacity in hotspots

15.4 Phase 3: Fleet Operations (Year 4+)

If The Claw proves the concept, the model scales to a fleet:

Multiple processing vessels (The Claw class) stationed at different GPGP locations
Dedicated collection vessels (smaller, more maneuverable) feeding debris to processing vessels
Drone swarm networks covering hundreds of km2
Shared hotspot intelligence across the fleet
This aligns with Ocean Cleanup's modeling that ~10 large systems could clean the GPGP

15.5 Critical Success Factors

1. Hotspot intelligence: Invest heavily in scouting and prediction. Operating in a zone with 500 kg/km2 vs 5 kg/km2 is the difference between success and failure. Partner with or license from Ocean Cleanup's mapping data.

2. Ghost net capability: Do not underestimate this. Ghost nets are ~50% of mass. Without robust crane recovery, The Claw misses half its potential feedstock.

3. Buffer storage: Oversized is better than undersized. Variable collection is a certainty -- the buffer is what keeps the reactor running during bad days and transit.

4. Seasonal planning: Operate April-October for maximum collection efficiency. Use winter months for maintenance, transit, resupply, and hotspot survey.

5. Adaptability: The optimal collection method may change based on operational experience. The vessel design should allow adding/modifying collection systems without drydocking.