Stock & flow model · plastic in the ocean · from 1950

How ocean plastic accumulates

A stock-and-flow model, not a forecast. Set each parameter to whatever you consider realistic — production growth, leakage rate, removal capacity, degradation — and the chart shows the accumulated stock that follows. Because plastic persists, the stock integrates: the assumptions change when a given quantity is reached rather than whether it is reached.

Parameters

Stock already in the ocean

What was already in the water by 2019. Published estimates disagree by more than a factor of two, so the model's historical leakage rate is fitted to whichever you pick — and the rate below follows it. Only the OECD figure is itself a 2019 estimate; the other two are this model carrying a 2015 estimate forward.

Of that, microplastic 15% · 34 Mt

How much of the 2019 stock above arrived as particles rather than as objects, and so was never recoverable. The share is ramped from near zero in 1950 to this value today, reflecting the growth of vehicle fleets and synthetic textile production — the ramp is an assumption, but its endpoint is the 15–31% IUCN range for ocean plastic entering as microplastic.

Production growth +2.4%

Reference: 2.4%/yr — the rate implied by the OECD baseline, 460 Mt in 2019 rising to 1,231 Mt by 2060.

Peak production year 2045

Production is held flat from this year onward. If growth is positive, that makes it the peak; if growth is negative, it is the year the decline levels off; if growth is zero, production never changes and this has no effect. Reference: 2045 — modelled peak in land-to-sea transport under the OECD's most ambitious policy scenario.

Production this year 495 Mt

Not a setting — this is what the two controls above produce for 2026, shown so it can be checked against observed global production of roughly 400–460 Mt/yr.

Leakage to ocean 1.90%

Applies from 2026. The solid marker is the rate implied by the starting estimate you chose above. The open marker at 1.5% is the rate implied by the OECD baseline, in which waste management keeps improving.

Enters as microplastic 23%

Tyre wear, textile fibres and the like, which arrive as particles rather than as objects and can never be collected. Applies from 2026. Reference: 15–31% of ocean plastic (IUCN, 2017), nearly two-thirds of it textiles and tyres. Before 2026 the share is fixed, rising from near zero in 1950 to 15% today as vehicle fleets and synthetic textile production grew; setting this slider to 15% therefore makes the series continuous.

Removal capacity 25 kt/yr

Reference: 25 kt/yr — The Ocean Cleanup's best year to date (2025), the largest such operation in the world.

Removal begins 2026

Delay costs more than it looks: macro plastic left in the water keeps fragmenting, and whatever has become microplastic by the time collection starts can no longer be recovered.

Fragmentation half-life 450 yr

How long macro plastic takes to break into microplastic. Reference: 450 yr — commonly cited persistence for common polymers in seawater. Fragmentation conserves mass: nothing leaves, it only becomes irrecoverable. A shorter half-life therefore makes the outcome worse, not better.

Project out to

The same mass, per person, per year — expressed as 1.5-litre bottles

bottles' worth of plastic used

of them end up in the ocean

recovered by every cleanup operation on earth, combined

The cost of waiting

Microplastic in excess of starting removal now — not a total. Raising removal capacity lifts this curve, because the more you could have collected, the more is forfeited by waiting.

This run, in words

Starting from the WEF/EMF estimate of 181 Mt already in the ocean by 2019, with production growing 2.4% a year and peaking in 2045, 1.9% of production leaking to the ocean, 25 kt/yr removed and a 450-year degradation half-life, the model puts 1.19 Gt of plastic in the ocean by 2100 — passing the OECD's projected 2060 aquatic stock (493 Mt) in 2048. Per person per year that is about 1,723 bottles' worth of plastic used, 33 of them leaked to the ocean, and 0.09 recovered by all cleanup operations combined.

Machine-readable model specification, including the equation, every parameter and all sources: ocean-plastic-model.json. Any set of parameters can be linked directly — the address bar updates as you change them.

How it works

Stock this year = stock last year, minus what degrades, minus what's removed, plus this year's leakage. Production is historical to 2015 and projected after that at your chosen growth rate.

The sliders act on the future only. Leakage before 2026 is held at its historical rate, so what's already in the water stays there.

The microplastic share is set separately for the past and the future, so that an assumption about what happens next cannot rewrite how much of today’s stock is already beyond recovery. For the past, the share ramps from near zero in 1950 to whatever endpoint you choose; that ramp is an assumption about the growth of tyre and textile sources, since no published historical series exists, but the endpoint sits within the range IUCN estimated for plastic entering the ocean as microplastic.

Two pools are tracked. Macro plastic can in principle be collected; microplastic cannot. Fragmentation moves mass from the first to the second and never back — it destroys nothing, so the microplastic pool only ever grows. Removal acts on the macro pool alone. Some leakage arrives as microplastic already and never passes through the macro pool at all.

Removal is a flat annual tonnage from the year you choose. After the peak year, production is held flat rather than declining.

Reference masses

Ocean Cleanup, cumulative to 2025 — ~53 kt
Great Pacific Garbage Patch, floating — ~100 kt
Estimated ocean plastic stock, 2015 — over 150 Mt (WEF / Ellen MacArthur Foundation, The New Plastics Economy, 2016). Widely requoted as 165 Mt, which is the same figure in US short tons
Marine plastic pool, 2015 — 263 Mt (Kaandorp et al., Science Advances, 2025)
Every adult alive, combined body mass — 287 Mt in 2005 (Walpole et al., BMC Public Health, 2012)
OECD projected stock in rivers and oceans, 2060 — 493 Mt
All plastic ever made, all compartments — ~8 Gt (Elhacham et al., Nature, 2020)

Growth and peak-year references: OECD, Global Plastics Outlook: Policy Scenarios to 2060 (2022).

Leakage-pathway figures: Kole et al., Int. J. Environ. Res. Public Health (2017) for tyre wear; European Environment Agency, Microplastics from textiles (2022) for synthetic fibres.

What this is not

Not a forecast. It's a single well-mixed pool with no sediment burial, no fragmentation tracking and no regional detail. Mass only — particle counts rise as material fragments, so exposure grows faster than this curve shows.

Bottle equivalents assume 35 g per 1.5 L PET bottle and a world population of 8.2 billion. Nothing here claims all plastic is bottles, or that the leaked share is packaging — it is a unit of mass you can picture.

Production series 1950–2015 after Geyer et al. (2017).

Two of the three starting figures are not measurements. They are what this model produces by 2019 when its historical leakage rate is fitted to a published 2015 estimate, and each inherits every weakness of the estimate behind it. The estimates themselves disagree by more than a factor of two, which is the honest state of the evidence.

Which plastic this counts

All of it. The model is not specific to packaging: the leakage rate is an effective, all-source rate fitted to estimates of total plastic in the ocean, so it already includes fishing gear, agricultural film, textile microfibres and tyre wear alongside packaging.

Two of those are worth naming because they behave differently. Tyre wear releases an estimated 0.81 kg per person per year globally, of which Kole et al. put 5–10% of all ocean plastic; synthetic textiles shed an estimated 25–500 g per person per year in Europe. Both arrive as microplastic already, without ever having been a bottle or a bag, so neither is reachable by deposit schemes, beach cleans or surface collection. The bottle figures above are a unit of mass, not a claim about what the plastic was.

One inconsistency to note: tyre rubber is an elastomer and sits outside the plastics production series this model grows, so tyre-derived ocean plastic is expressed as a share of a production figure that excludes it. At 5–10% of ocean input the effect is small against the factor-of-two spread in the stock estimates, but it means the leakage rate slightly overstates the loss rate of conventional plastics.

What this model does not argue

It makes no comparison between materials and no case for substituting one for another. Leakage is treated as a property of the collection and waste-management system, not of the polymer — a well-managed system leaks little, a poorly managed one leaks a lot, whatever the material. The model's conclusion is that closing the leakage pathway is the only intervention that changes the endpoint rather than the schedule, which is an argument for investment in collection, recovery and closed-loop design.

Reference values are drawn from published sources whose authors have neither reviewed nor endorsed this tool.

Why this runs ahead of the OECD line

Two assumptions, both yours to change. The starting stock: published estimates of what is already in the water differ by more than a factor of two, and the toggle at the top lets you pick. And the leakage rate: this model holds it constant, where the OECD baseline assumes waste management keeps improving — the mismanaged share falling from 22% to 15% by 2060 without any new policy. Drag leakage to the open marker to adopt that view.

Set both to the OECD position and the curve still passes half a gigatonne and keeps climbing. Their figure also covers rivers as well as oceans, which works the other way.

Illustrative model, not a forecast. This tool simplifies a complex and contested system. Its outputs depend entirely on the assumptions selected by the user and on published estimates that disagree with one another, and it is provided for general information and discussion only. It is not professional, technical, regulatory, investment or policy advice, and should not be relied upon for any decision. Verify the cited sources independently before using any figure. To the fullest extent permitted by law, Hoskins Circular accepts no liability for any loss or damage arising from use of, or reliance on, this tool or its outputs.

Reproduction is welcome with attribution to Hoskins Circular and a link to the source page. The underlying data belongs to the organisations cited above.