Where California's Textile EPR Sends the Clothes Nobody Can Recycle

On 1 July 2026, the enforceable core of California's textile EPR law came down to a $1,000 annual fee and a registration form that takes about ten minutes. Any apparel producer with more than a million dollars in global sales and a single customer in California had to join Landbell USA, the producer responsibility organization CalRecycle appointed back in February, or start accruing penalties of up to $10,000 a day (per Landbell USA's SB 707 compliance guidance and CalRecycle's order). The American Apparel and Footwear Association sued in March to unwind the sole-PRO selection. Producers had to register anyway.

So the milestone landed, and it changed nothing you can see. The part almost everyone assumes the law is about, a garment getting collected, sorted, and recycled under the program, doesn't start for years. And when it does, the slice of the stream with no reuse market and no fiber recycler, the blended and coated and degraded and sometimes PFAS-laden material, becomes somebody's feedstock. Usually a waste-to-energy operator's, or a landfill's. What follows is four things people believe about textile EPR that don't survive the compliance calendar or a walk down the sorting line.

"Textile EPR means the clothes finally get recycled"

The recycling most people picture, an old shirt reborn as a new shirt, barely exists at commercial scale. EPA put the US textile recycling rate at 14.7% in 2018, out of roughly 17 million tons generated that year (per EPA's textiles material data). Nationally that breaks down to about two-thirds landfilled, a fifth combusted with energy recovery, and the remainder recycled. CalRecycle leans on a starker California figure to sell the law: 85% of the state's old clothes hit landfill.

That sounds like a floor you can build on. Actually, the recycling number flatters the picture. Most of it is export for reuse plus downcycling into wiping rags, insulation, and furniture stuffing, not fiber going back into fiber. The reuse half leans hard on export, secondhand bales shipped to West Africa and South Asia, and those markets are tightening as importing countries push back on the low-grade castoffs they end up burning or burying themselves.

Strip the reuse and the downcycling out and closed-loop fiber-to-fiber recycling runs under 1% of the stream, for reasons that are chemistry rather than effort. Around 60% of what the industry makes is synthetic or blended, and GAO's 2025 textile-waste review is blunt about the consequences: no commercial-scale textile-to-textile recycling operates in the United States today, mechanical recycling yields weaker fibers fit only for stuffing and insulation, and chemical recycling doesn't work on blends. Optical sorters read a fabric's dominant fiber well enough (Tomra's near-infrared lines are what most sorters run), but a poly-cotton-elastane knit with a PVC print doesn't resolve into anything a recycler will pay for. EPR funds collection. It doesn't conjure a recycling technology that isn't built yet, so the clothing-waste end-of-life question the law is really posing is narrower than its headline: not how to recycle this, but where the part nobody can recycle actually goes.

"We registered by July 1, so the program is live"

Registration is the first milestone, not the finish line. Walk California's textile EPR calendar, published by CalRecycle, and the distance between "registered" and "diverted" is most of a decade.

So the physical program, the bins and routes and sorting contracts and an actual destination for each grade, doesn't exist until the end of the decade. Stack the AAFA litigation on top, filed 27 March 2026 against the sole-PRO pick, and even the regulatory text isn't settled. The needs assessment due in 2027 is where the residual tonnage first gets a number, and whether that number is honest about the unrecyclable share will shape every processing contract that follows. EPR programs don't fail at registration; they fail at the plan, three years on, when someone finally has to name a home for the fraction nobody wants and finds the arithmetic never closed. For anyone running energy recovery or a landfill, that lead time is the whole story. The residual you'll be asked to take in 2031 is being designed right now, on paper, by people optimizing for producer fees and diversion headlines rather than for what actually burns or buries cleanly.

"Producers pay for it, so it's not the waste sector's problem"

The fee funds the PRO and the collection network. It does not fund an end-of-life home for the non-reusable, non-recyclable residual, and those are separate budgets. Eco-modulation, charging producers more for hard-to-recycle formats, is the mechanism everyone points to. But a garment counting as "recyclable" on a fee schedule and there being a recycler who actually wants it are not the same fact. The physical sorting still runs through charities and merchant graders whose economics already depend on selling the good third and offloading the rest; EPR money changes who pays, not the fact that a residual falls out the bottom of every sort.

Where does that tonne go once the fee is paid and the diversion credit is booked? Once collection scales past thrift-store cherry-picking, the sorted-out residual is a large fraction of the intake, and it needs somewhere to land. That is a landfill under RCRA Subtitle D, at California tipping fees north of $70 a ton (industry estimate), or it is a combustor. The producer's fee doesn't follow that tonne to the grate. So the diversion mandate lands, in practice, on operators running zero-waste-to-landfill solutions and energy recovery, who inherit a feedstock they never specified. A brand books its diversion credit and files a clean ESG-compliant number; the tonnes still show up at a scale house downstream. That gap, between the booked claim and the physical tonne, is exactly what an ESG auditor should test and usually doesn't.

"Textile-to-energy is the obvious home for the residual"

Why wouldn't the residual just go to a combustor and disappear? It's a home. Not an obvious one. Textiles fight the reactor: low bulk density, moisture that swings with the load, and a chemistry problem I keep circling back to on the compliance side. The energy content is real once the material is dry, but chlorine from certain dyes and PVC prints raises the same acid-gas and corrosion questions any chlorinated feed does. I've laid out before why old clothes resist the reactor, so here I'll take the policy cut, which is shorter and meaner. California's AB 1817 has banned intentionally added PFAS in new apparel since 1 January 2025, yet the garments already in the waste stream, plus everything imported before that, still carry it. Feed a PFAS-uncertain textile fraction into a combustor and you have walked into monitoring territory regulators are still drafting.

What that looks like in practice: on a PFAS air-permit review I ran in 2024, a monitoring condition the state dropped in mid-cycle stalled the file for months, because the plant's installed CEMS simply couldn't produce the number it demanded. The problem was never a dirty stack. It was that you cannot certify compliance against a value the instrument cannot read, and no amount of clean operating data fixes an unreadable metric. Now picture that condition propagating across every unit expected to absorb textile residual. It's the same monitoring exposure I flag under the federal 40 CFR Part 60 Subpart Eb limits: the rule you meet at issuance is rarely the rule you meet at renewal.

And then there is where these units sit. Environmental-justice screening will reshape US siting faster than any emissions limit will, and in California that screen is CalEnviroScreen. Most of the mass-burn capacity people quietly assume will absorb this residual sits in exactly the census tracts it now flags, which is why a Martin-style grate three miles upwind of a flagged community is not a permit you win on thermodynamics. Downcycling the cleaner cotton fraction on an Andritz tearing line soaks up a little of the pressure, but tearing lines want clean single-fiber input, which is the opposite of what the residual is. Even the cleanest modern waste-to-energy technology, pyrolysis included, has to clear that siting gate before a single bale of residual moves.

So the residual doesn't vanish because a producer paid $1,000 and filed a form in July. It arrives at a grate or a scale house around 2031, blend-heavy and PFAS-uncertain and bigger than anyone's diversion slide admits, and the permit that has to accept it is being written right now under monitoring rules I don't think half the current fleet clears at its next renewal. California didn't solve textile waste this summer. It scheduled the invoice, and mailed it to the waste sector.

Sources & Notes

• The registration mechanics, a $1,000 annual fee, the July 1 deadline, penalties reaching $10,000 a day, and Landbell's February selection, come from Landbell USA's SB 707 compliance page, checked against CalRecycle's order.
• For the timeline, needs assessment in 2027, regulations by 2028, an approved collection and processing plan by 2030, and implementation in 2031, see CalRecycle's textile EPR program page.
• The disposal shares (roughly 66% landfilled, 19% combusted with energy recovery, 15% recycled) and the 14.7% recycling rate are EPA's, from its textiles material-specific data.
• On the fiber-recycling wall, the under-1% closed-loop share, the absence of any US commercial-scale textile-to-textile plant, and the blend problem, I leaned on GAO's 2025 report on federal textile-waste efforts.
• California's PFAS-in-apparel ban runs under AB 1817; the mid-cycle air-monitoring episode is from my own 2024 permit work and isn't in the public record.