I’ll be honest about how this article came to be. The Clean-Seas West Virginia plant in Belle had fallen off my radar. I knew the broad outline from 2023, when then Governor Justice announced a $50 million “advanced recycling” investment with the usual fanfare, and I knew the project had slipped and shrunk since. But somewhere between the delays and the incomplete applications, I stopped following it.
Eric Caruthers brought it back to my attention. Caruthers is a Belle resident who has been organizing opposition to the plant for the better part of a year, and when he reached out about the community forum he is hosting this week, the thing that stuck with me was not the outrage. It was the fact that the state had already issued the permit, and that most people in the Kanawha Valley, myself included, had not read it.
That forum is Tuesday evening, September 22, 6:00 to 8:00 p.m., at the Belle Town Hall gym, 1100 E. DuPont Ave. Organizers have sent personal invitations to 168 elected officials and community leaders, along with company representatives. Most of the conversation will, understandably, be about health, schools, and whether anyone at the county or state level can still do anything.
Before I walked into that room, I wanted to answer a narrower question: what did the state actually approve?
So I pulled Permit R13-3728, the 58-page construction permit signed by Division of Air Quality Director Laura Crowder on August 12, 2026. What follows is a plain-language walkthrough, with condition numbers cited so you can check my work. A copy of the signed permit is posted by WCHS, and the DEP’s own file is searchable through its NSR permit application database.
Why nobody in Belle, West Virginia is taking the state’s word for it.
If you want to understand why a permit fight over a 50-ton-a-day plastics plant has produced nearly 1,000 petition signatures, personal invitations to 168 elected officials, and a second packed town hall, you have to understand that the people of the Kanawha Valley have been told before that the state was watching, and have learned what that promise is worth.
This valley has hosted the densest concentration of chemical manufacturing in the country for the better part of a century. Locals call it Chemical Valley, and nobody means it as a compliment (National Geographic).
In August 2008, a pressure vessel at the Bayer CropScience plant in Institute exploded, killing two workers. Federal investigators found that Bayer management withheld information from county emergency responders while the fire burned, and that a shelter-in-place recommendation never reached 911 operators. The Chemical Safety Board’s central recommendation was that West Virginia create its own chemical plant oversight program (U.S. Chemical Safety Board). A proposal modeled on California’s program, requiring annual audits by health officials and safety experts, was put in front of the Legislature in 2009. It failed to advance (National Geographic).
In January 2010, in Belle itself, a frayed cable on a phosgene line at the DuPont plant sprayed a 58-year-old worker named Carl Fish, who died the next day. It was one of three preventable accidents at the plant inside 33 hours. Investigators cited old equipment, an ineffective mechanical integrity program, and incomplete investigations of earlier near-misses (U.S. Chemical Safety Board; BBC News). The Clean-Seas site sits on DuPont Avenue.
Then came January 9, 2014. A tank at Freedom Industries, a mile and a half upstream from the region’s water intake, had corroded through, and roughly 11,000 gallons of MCHM went into the Elk River. Some 300,000 people in nine counties were told not to drink, cook with, or bathe in their water. Freedom did not disclose that a second chemical was in the tank until 12 days later (U.S. Chemical Safety Board). The state, it turned out, had been relying on companies to self-report their risks and did not have enough certified inspectors to check (West Virginia Watch).
What happened next is the part that matters most for Tuesday. The Legislature responded with the Aboveground Storage Tank Act, passed unanimously in 2014, requiring registration and certified inspection of roughly 42,000 tanks (Mountain State Spotlight). The very next year, lawmakers passed SB 423 and began carving tanks back out (West Virginia Watch). They kept carving. By this spring, Scott Mandirola, the former DEP official who helped draft the original act, testified that the state now regulates about 4,500 of those tanks, roughly 11 percent, and that 89 percent had been deregulated (WV MetroNews). Then both chambers advanced yet another rollback, SB 641, exempting more oil and gas tanks from inspection and letting some owners in critical zones inspect themselves rather than hire an independent third party. When a senator offered an amendment to at least define what fluids the exempted tanks could hold, the Senate rejected it (Mountain State Spotlight).
Widen the lens beyond the valley and the pattern holds. In 1972, a Pittston Coal slurry dam failed on Buffalo Creek, killed 125 people, and left roughly 4,000 homeless across 16 coal communities; the company called it “an act of God” (WV Mine Wars Museum). In 2010, 29 miners died at Upper Big Branch, and an independent review concluded the federal mine safety agency could have prevented it (NPR). In Parkersburg, DuPont told workers and neighbors for decades that C8 had “no observed health effects” while it contaminated their water, and eventually paid $671 million to settle more than 3,500 lawsuits (PFAS Project Lab; The Conversation).
The sequence is always the same. A company gives assurances. Regulators accept them. Something breaks. The Legislature passes a law. The Legislature weakens the law. And every step of it is official, on the record, done by people West Virginians elected.
So when residents of Belle say they do not trust that the DEP, the county, or the Legislature will protect them from this plant, they are not being paranoid. They are being accurate about their own history. The burden here is not on them to prove the plant is dangerous. It is on the state to prove, this time, that it is actually watching.
Read the permit with that history in mind.
That is the lens I brought to Permit R13-3728, the 58-page construction permit signed by Division of Air Quality Director Laura M. Crowder on Aug. 12, 2026. Before I walked into Tuesday’s room, I wanted to answer a narrower question: what did the state actually approve, and how much of it depends on the company telling the truth about itself?
The short version, which I will spell out below: the permit lets the company decide for itself whether the gas it burns is a fuel or a waste, lets it check its own stacks with a hand-held meter, tests for dioxins once and then not again for five years, and requires no monitoring at all at the fence line shared with two schools. Each of those is a defensible legal position. Together they are a familiar shape.
What follows is a plain-language walkthrough, with condition numbers so you can check my work. A copy of the signed permit is posted by WCHS, and the DEP’s own file is searchable through its NSR permit application database (permit writer: Ed Andrews).
The forum is Tuesday evening, September 22, 6:00 to 8:00 p.m., at the Belle Town Hall gym, 1100 E. DuPont Ave. Organizers have sent personal invitations to 168 elected officials and community leaders, along with company representatives.
First, the basics.
Who and where. The permittee is Clean-Seas West Virginia, Inc., a wholly owned subsidiary of Clean Vision Corporation, a Los Angeles-based penny stock (OTC: CLNV) (Clean Vision press release). The permitted address is 2700 E. DuPont Ave., Suite 3B, Belle, the former 84 Lumber building, which sits a few hundred feet from Riverside High School and Midland Trail Elementary (West Virginia Watch).
What kind of permit. It is a “construction permit” under state rule 45CSR13, but don’t let the name fool you. The permit’s own description of change says it is “for the construction and operation of a chemical manufacturing process (pyrolysis units).” Once built, the company can run. The article that circulated ahead of Tuesday’s meeting got this right.
How the DEP classified it. The permit declares the plant a “nonmajor source” of air pollution. That single classification does most of the work in this document, and I’ll come back to it.
What gets built.
Section 1.0 of the permit lists every emission unit. Stripped of the ID codes:
* Two commercial pyrolysis reactor trains, each rated at 25 tons of plastic per day, each heated by a 3.5 MMBtu/hr process heater. That is 50 tons per day combined, a far cry from the 100 tons per day the company promised in 2023, or the 500 tons per day its CEO said he hoped to reach (West Virginia Watch).
* A “Training, Research, Evaluation” (TRE) reactor at 1 ton per day with its own small elevated flare. This is the unit the company has been running under a research-and-development exemption without a permit since before the application was filed (WV MetroNews).
* An enclosed combustion device (10 MMBtu/hr) that burns off gases during startup, shutdown, and malfunction.
* Eight storage tanks of 21,000 gallons each for pyrolysis oil, a truck loading rack capped at 3.96 million gallons per year, a glycol heater, a 30-ton char storage bin, and a natural-gas emergency generator.
“Burning plastic” is not quite what the permit describes.
The shorthand in most coverage, including the article promoting Tuesday’s meeting, is that the plant “would burn plastic feedstock.” That is not what the permit authorizes, and the distinction matters because the company will use it.
Condition 5.1.1.b.3 requires each reactor to run in an “oxygen-free/oxygen-limited atmosphere” with no more than 2% oxygen. Plastic is cooked, not burned. What comes out is pyrolysis oil (sold to petrochemical buyers), a non-condensable “process-derived syngas,” and solid char.
What is burned on site is the syngas. The permit lets the company fire it in the two process heaters and the glycol heater as fuel (5.1.1.c.1), route it to the enclosed combustion device during startup and shutdown (5.1.1.b.5), and flare it from the research unit (5.1.3). So critics who call this incineration are pointing at something real. It is just the combustion of gas made from plastic rather than the plastic itself. Precision here is a strength, not a concession.
The emission caps, on paper.
Condition 5.1.5 sets facility-wide annual limits, measured on a rolling 12-month basis:
Those are small numbers. They are also the reason the DEP could call this a “nonmajor” source. But read the fine print at the bottom of 5.1.5: the caps “do not include fugitive emissions sources (equipment leaks or cleaning of equipment) or the emissions released from tanker truck loading operations.” Truck loading gets its own 0.40-ton VOC allowance (6.1.2.a). Equipment cleaning gets another 319 pounds (8.1.1.a).
Hourly limits are set unit by unit. For each process heater, for example, the permit allows 0.13 lb/hr of NOx, 0.05 lb/hr of hydrogen chloride, and 0.06 lb/hr of total hazardous air pollutants (5.1.1.a). Whether a plant like this actually stays under numbers that low is exactly what county commissioners asked the DEP to scrutinize when they wrote in July asking whether the company’s emissions estimates were “based on reliable data comparable to other similar facilities” (West Virginia Watch).
What the plant is allowed to take in.
Section 4 governs feedstock, and it is stricter than I expected:
* Every shipment must be at least 85% polyethylene or polypropylene by weight, sorted and washed off-site, with a certificate of analysis. No certificate, no delivery (4.1.1).
* No bales. Shipments must arrive in sealed containers so inspectors can see what’s inside (4.1.1).
* Up to 4,500 tons of plastic may be stored on site at once, and no batch may sit for more than 12 months (4.1.3). Rejected loads must leave within five days (4.1.4).
* Any plastic the company received before the permit was issued has to be quarantined and inspected under a written plan before it can be processed (4.1.2.a). The company announced its “first commercial-scale shipment” of feedstock in August 2025, a full year before the permit (Clean Vision).
The 85% PE/PP rule is the permit’s main defense against dioxins and hydrogen chloride, which come primarily from chlorinated plastics like PVC. The stack test (5.3.1.c.7) is explicitly meant “to verify the efficacy of upstream feedstock sorting.”
Is the syngas a fuel or a solid waste?
Here is the part of the permit that I think deserves the most attention on Tuesday.
Under federal law, when you burn a “non-hazardous secondary material,” it is presumed to be solid waste unless it qualifies as a legitimate fuel under 40 CFR 241.3. If the syngas is a fuel, the heaters are just heaters. If the syngas is a waste, the heaters are incinerators, subject to the federal Other Solid Waste Incineration (OSWI) rules, with far tougher standards, and the “nonmajor” label goes away.
The permit resolves this in the company’s favor, but with a structure worth understanding:
* The company decides. Condition 5.1.1.d requires a “non-waste/waste self-determination” before startup, or a petition to EPA. The permit does not require the petition. A site-specific self-determination, documented on site, is enough.
* One report, then done. Within 180 days of startup, the company must file a “legitimacy demonstration report” (5.4.5.a). After that, Condition 5.4.5.c says “no further formal determinations under 40 CFR 241.3(d) are required unless the permittee significantly alters the accepted feedstock profile.”
* The research unit is exempt outright. Condition 5.1.3.a states that “by maintaining this operational purpose, the unit is exempt from the requirements of the Other Solid Waste Incineration (OSWI) rules.”
* The backstop. Condition 5.5.5 is the one real off-ramp. After the initial stack tests, the company must recalculate its “potential to emit” using measured data. If any single hazardous air pollutant hits 10 tons per year, or all of them together hit 25, or if “the Director determines the emission profile indicates the combustion of solid waste subject to OSWI guidelines,” the company must apply for a Title V major-source permit or implement OSWI standards.
Read that first item again. The single classification that determines whether this plant is a light-touch “nonmajor source” or a federally regulated incinerator is made by the company, about itself, in a document it keeps on its own premises. That is the same self-certification model the state was relying on when a tank nobody had inspected emptied into the Elk River.
Tom Torres of the Ohio River Valley Institute called the approval “a rollback of critical pollution safeguards put in place to keep families and children safe from the most damaging impacts of waste incineration” and said it departs “from decades of precedent” (West Virginia Watch). The fuel-versus-waste framework is what he is talking about.
What the permit does not require.
Marty Gibson, a Shrewsbury resident with asthma, told reporters after the approval that “this permit fails to require the rigorous, continuous emissions monitoring necessary to guarantee the daily safety of people like me who live right next to the site” (Earthworks). Having read it, I think that is a fair description. Specifically:
* No ambient or fenceline air monitoring. The word “school” does not appear in the permit. Nothing in it measures what reaches the property line, Riverside’s track, or the elementary school playground.
* No continuous stack monitoring for pollutants. The permit requires continuous parameter monitoring (temperatures, pressures, oxygen, feed rate) under 5.2.1, and a continuous analyzer on the syngas line for composition, hydrogen sulfide, and hydrogen chloride under 5.2.2. But the stacks themselves are checked with a hand-held portable analyzer for NOx and CO: weekly for the first six months, then biweekly, then monthly, and eventually quarterly if readings stay low (5.2.4).
* One full stack test, then wait five years. The comprehensive performance test, which is the only time dioxins and furans (EPA Method 23), particulate, hydrogen chloride, and VOCs are measured at the stack, must happen within 180 days of startup. After that, it recurs “at least once every sixty-one (61) months” (5.3.1.a).
* Visible-emission checks are by eye. Compliance with opacity limits at the combustion device, flare, glycol heater, and char bin is verified by a trained observer looking at the stack for periods ranging from one to six minutes (5.2.5 through 5.2.7, 4.2.1).
* Startup and shutdown venting is capped, not prohibited. Raw syngas may be routed to the enclosed combustion device during startups and shutdowns for up to 312 hours per rolling 12 months (5.1.2.b.5).
To be fair, some provisions are meaningful. The permit bans any “intentional, routine release” of uncombusted syngas to the atmosphere (5.1.6). It requires a leak detection and repair program with quarterly Method 21 monitoring and a 15-day repair clock (6.1.3). It requires semi-annual testing of the syngas for nine heavy metals including mercury, lead, cadmium, and arsenic, and conservatively assumes 100% of any metal found ends up in the air (5.3.7). And it requires the company to keep a log of every odor complaint (3.4.2).
But notice who does the measuring in nearly every one of those provisions. The company holds the meter, the company runs the test, the company keeps the log, and the company files the report. The DEP reads what it is sent. After 2014, the state’s own post-mortem said the system “relied on companies to self-report risks, with inadequate mechanisms for state enforcement and oversight” (West Virginia Watch). Twelve years later, that is still the architecture.
This is not the last permit.
Page 2 of the permit says the company “shall apply for a Title V (45CSR30) permit” unless the Director grants a deferral or exemption. Title V is the federal operating-permit program. It comes with its own public comment period and, unlike the construction permit, gives EPA a formal review role. Whether the DEP grants a deferral, and when the company applies, is worth tracking.
Condition 2.5.4 also says the permit “does not relieve the permittee of the responsibility to apply for and obtain all other permits, licenses, and/or approvals from other agencies, i.e., local, state, and federal.” Which brings us to the county.
The county zoning question.
This is where local government enters the picture.
Commissioner Ben Salango told West Virginia Watch in July that “we can’t zone” because of a decision made roughly 40 years ago: “They took that away from us. That was the voters” (West Virginia Watch). Organizer Eric Caruthers says nobody has been able to locate the law he is describing. Nearly 1,000 residents signed a petition asking the commission to hear them, and the answer they got was that a rule no one can produce ties the commission’s hands.
A few things worth knowing before Tuesday. County zoning in West Virginia is governed by Chapter 8A of the state code, which lets a county adopt a zoning ordinance only after it has a planning commission and a comprehensive plan, and which allows zoning to be put to voters. If Salango is right that voters rejected zoning decades ago, the record of that vote should exist in the county clerk’s election returns and the commission’s minutes. If he is wrong, and the limit is a commission policy rather than a voter mandate, the current commission can revisit it. Either way, it is a public record, and someone should produce it.
The distinction matters for a reason that goes beyond this plant. “We’d love to help, but our hands are tied” is the sentence eastern Kanawha County has heard from elected officials for two generations. Sometimes it is true. When it is offered without the document that ties the hands, residents are entitled to treat it as a choice.
The appeal timeline.
The permit says on page 2 that “any person whose interest may be affected, including... any person who participated in the public comment process,” may appeal to the West Virginia Air Quality Board. The Board’s own guidance says appeals “are to be filed within thirty days of the permit applicant’s receipt of the permit or order,” which tracks W.Va. Code 22B-1-7. For a permit issued August 12, that window closed around September 11.
Caruthers has said organizers are working with Fair Shake Environmental Legal Services to appeal.
The money, briefly.
The state has put real public dollars behind this. The West Virginia Economic Development Authority approved a $15 million bridge loan in July 2024, on top of a 2023 memorandum of understanding under then Governor Justice promising up to $1.75 million in performance-based loans against a pledged $50 million investment and at least 40 jobs (West Virginia Watch; WV News). Clean Vision has proposed at least 14 pyrolysis plants worldwide since 2020; as of December 2025, two were operating, in India and Morocco, and SEC filings show under $614,000 in lifetime pyrolysis revenue against nearly $58 million in accumulated deficits (West Virginia Watch).
None of that is in the air permit. The DEP’s final determination letter said residents’ concerns about the company’s finances “did not violate the state’s standards for issuing air quality permits” (West Virginia Watch). That is probably true as a matter of law. It is also a reason the permit’s monitoring gaps matter: a company this thinly capitalized has every incentive to run lean on compliance. Freedom Industries filed for bankruptcy eight days after the spill. The people of the Kanawha Valley know what it looks like when the operator is gone and the cleanup is not.
What to watch, and what to request.
If the state is going to trust the company’s paperwork, the least the rest of us can do is read it. The permit generates a paper trail. Every one of these is a public record obtainable from the DEP’s Division of Air Quality:
If you live nearby and smell something, call it in. The company is required to log it, and the DEP can ask for the log.
Questions worth asking.
For the company:
* Will you commit to fenceline monitoring near Riverside High School, which the permit does not require?
* Will you petition EPA for a formal non-waste determination rather than relying on a self-determination?
* Has the plastic you received in August 2025 been quarantined as Condition 4.1.2.a requires?
For the DEP (if anyone from the agency attends):
* Why is the comprehensive stack test, the only time dioxins are measured, required just once every 61 months?
* Will the Director grant a Title V deferral, and if so, on what basis?
* After 2014, the state acknowledged that self-reporting without enforcement capacity failed. What is different about this permit?
For the Kanawha County Commission:
* Produce the 1980s record that removed zoning authority from unincorporated Kanawha County. If it was a voter referendum, cite the election. If it was a commission decision, explain why the current commission cannot revisit it.
For the legislators who received an invitation:
* You have spent a decade deregulating the tanks you regulated after the water crisis. What would it take for you to vote to require fenceline monitoring at facilities next to schools?
One last thing. It would be easy to read all of this as an argument that nothing can be done. I don’t think that is the lesson. The lesson of Institute, Belle, and the Elk River is that the state will watch exactly as closely as its people insist. Trust in government is not something West Virginians owe. It is something the government has to earn back, one inspection report at a time, and the records table above is how you check whether it is trying. If this had fallen off your radar too, Tuesday night in Belle is a good place to put it back on.
The permit is 58 pages. I have tried to be accurate about what is in it and honest about what is not. If you find an error, or a condition I’ve misread, contact me and I’ll correct it.
Sources:
* Permit R13-3728, WVDEP Division of Air Quality, issued Aug. 12, 2026 (copy via WCHS)
* WVDEP NSR permit database
* W.Va. Code 22B-1-7
* W.Va. Code 8A-7-3
* 40 CFR 241.3
* WV Air Quality Board, How to File an Appeal
* West Virginia Watch, Aug. 13, 2026
* West Virginia Watch, July 31, 2026
* West Virginia Watch, Jan. 9, 2024
* WV MetroNews, April 28, 2026
* WV MetroNews, March 12, 2026
* Mountain State Spotlight, March 4, 2026
* Mountain State Spotlight, Sept. 18, 2025
* U.S. Chemical Safety Board, Freedom Industries final report (2017)
* U.S. Chemical Safety Board, Bayer CropScience report (2011)
* U.S. Chemical Safety Board, DuPont Belle report (2011)
* BBC News, July 2011
* National Geographic, Jan. 2014
* NPR, March 23, 2012
* WV Mine Wars Museum, Feb. 2026
* PFAS Project Lab, Parkersburg
* The Conversation, 2020
* Earthworks, Aug. 15, 2026
* Clean Vision Corp. press release, Aug. 14, 2026
* WV News
Carrie Clendening writes about West Virginia policymaking through the perspectives of elections and campaign finance money.
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit wvpolicy.substack.com