Status snapshot · Updated 7:00 PM PT Monday
✅ EVACUATION ZONE CONTRACTED at 6 PM PT — evacuee count dropped from 50,000 → ~16,000 tonight. About 34,000 outer-ring residents returning home this evening. OCFA released an updated incident map showing a smaller red-shaded area (new zone: Garden Grove Blvd → Orangewood, Knott → Dale) inside the prior red-outlined boundary. Residents still inside the contracted zone should NOT attempt to return — cleanup + verification ongoing. The 3-day arc: BLEVE off the table this morning → tank cooling to 93°F → evac zone reduced this evening.
How this countdown is calculated
Our countdown uses a piecewise heating-rate model: from Saturday's 90°F reading forward at 0.54°F/hr (the Friday → Saturday observed rate of (90-77)/24h) until the Sunday morning crack discovery, then 0.30°F/hr after that (estimated — the crack is venting pressure and slowing the reaction, but not stopping it). The gauge maxed at 100°F so we cannot measure the actual current temperature; the model gives a best-guess number that's consistent with everything we know.
Temperatures we've heard so far:
Fri 8:00 AM — Fri morning — first reading77°F / 25°C
Sat 8:00 AM — Sat morning — gauge re-read90°F / 32°C
Sun 5:00 PM — Sun ~5pm — OCFA via ABC7 (gauge maxed at 100°F)100°F / 38°C
Mon 8:00 AM — Mon ~8am — cooling reported, BLEVE off the table93°F / 34°C
Pre-crack rate (Fri → Sat): 0.54°F/hrPost-crack rate estimate: 0.30°F/hr
Best guess for right now118.9°F / 48.3°C
This is a guess, not a live thermometer. We only have two confirmed readings (above) and a rate firefighters mentioned out loud. The math draws a straight line from the latest reading using that rate. If the rate speeds up — which can happen as it gets hotter — these countdowns get shorter. Reaching one of these temperatures doesn't automatically mean an explosion; it means things are getting worse.
OCFA's original stated rate: OCFA Chief Covey publicly said ~1°F/hr on Saturday. That was likely a worst-case or short-term spike — the actual Fri→Sat trend was slower. Preserved here for reference.
Analysis
✅ UPDATE — Mon 6 PM PT: Evacuation zone contracted
Effective 6 PM PT Monday May 25, OCFA contracted the evacuation zone. Their updated incident map shows the new (smaller, red-shaded) zone vs the previous (red-outlined) zone — portions of the previous evacuation area have been released. OCFA: "Effective as of Monday, May 25 at 6pm, the red shaded area represents the newest evacuation zone, while the red outline represents the previous evacuation zone." Residents now OUTSIDE the smaller zone may begin returning home tonight. Residents still INSIDE the smaller zone should NOT attempt to enter — cleanup and verification continue. The 6 PM PT press conference (live-streamed by OCFA on Facebook and by @alexdatig on X) walked through the resolution: BLEVE eliminated (8 AM briefing), tank cooling to 93°F, and now perimeter contraction. The site's morning prediction of a Tue May 26 – Thu May 28 return window is being validated in real time — some residents are starting tonight.
🟡 ANALYSIS, NOT OFFICIAL. Everything below is independent reasoning from public reporting, transcribed OCFA briefings, and chemistry references. OCFA has not endorsed any probabilities. If officials say something different, trust officials.
What firefighters can actually try
- Water spray cooling (current passive baseline). Unmanned hoses + sprinkler. Was framed Saturday as buying time for the cure strategy. With Sunday's crack-venting confirmation the role shifts to keeping the tank exterior cool while the crack vents pressure — preventing thermal damage to surrounding structures rather than fighting the reaction itself.
- Controlled cure — let it finish polymerizing slowly (largely deprecated). Covey's Saturday "third option": let MMA polymerize into solid acrylic with void space absorbing pressure. After McGovern's Sunday 10:47 AM confirmation that the tank has a crack venting pressure, this strategy is largely off the table — you can't cleanly cure a leaking tank into a solid plug. Kept here for transparency about what was being considered.
- Going downrange to neutralize the second tank ✓ DONE. Crews went downrange Friday night with manufacturer chemists. Per OCFA: inhibitor successfully added to the 15k tank, which is now structurally sound and being monitored. The cluster cascade risk from tank #2 is materially reduced — a major operational win that hasn't been called out publicly.
- Diking, damming, and diversion. Staged for the failure scenario. Covey: divert spilled MMA to a downgrade holding area to keep it out of storm drains, the river, and the ocean.
- Drill a fresh port (hot-tapping). Textbook fix for stuck valves on reactive monomers — not publicly mentioned in any OCFA briefing. The flammable vapor around the tank makes a spark catastrophic, but specialty contractors do this for a living.
- Liquid nitrogen or refrigeration coil. Brute-force temperature drop. Heavy equipment, slow to mobilize, not in evidence.
- Foam blanket over the area. Doesn't fix the chemistry but if it does blow, foam reduces the fireball size and slows the vapor cloud.
Rough odds for each ending
Controlled spill (firefighters' preferred path)~20%
Bumped back up at 10:47 AM PT Sunday after McGovern confirmed the overnight recon found a crack venting pressure. The operational posture is now actively managing the leak: underflow dam construction, diking, and diverting captured product. Largest single bucket because containment infrastructure is staged and the team has effectively transitioned from prevention to managed-release mode.
Controlled cure or new intervention works~55%
Now the dominant scenario. As of Monday 8:03 AM PT, Covey confirmed the tank temperature has dropped to 93°F (from 100°F+), the crack has released pressure, and BLEVE risk is "off the table." The exact mechanism is consistent with controlled cure or natural pressure relief — the tank is stabilizing without major release. Cleanup and verification still required, but the worst case is closed.
Uncontrolled slow leak~15%
Valve or seam gives up on its own — toxic vapor cloud, but containable with foam and diking. Bumped up Sunday morning after McGovern confirmed a crack is already venting. The line between this and "controlled spill" depends on whether the crack stays in the Goldilocks zone the team is managing toward. If the crack grows or vents faster than containment can capture, we drift here from outcome 1.
Cooling holds, they ride it out~7%
Down from earlier. Covey effectively ruled this out — cooling alone isn't winning. Some marginal probability remains for the trend to flatten without active intervention.
BLEVE (residual — officially off the table per OCFA)~1–3%
Officially OFF THE TABLE per OCFA's Monday 8:03 AM PT briefing — Covey stated directly: "the threat of a BLEVE is now off the table. That threat has been eliminated." Tank temperature dropped from 100°F+ to 93°F overnight. Crack has released pressure. ~1–3% residual probability for completeness (OCFA could be wrong, an unforeseen failure could occur), but the explicit official position is that this pathway is closed.
BLEVE = Boiling-Liquid Expanding-Vapor Explosion — what happens when a pressurized tank ruptures and the superheated liquid inside flash-boils into a fireball in milliseconds.
Prediction history (11 earlier estimates)
Earlier snapshots, newest first. Kept for context — only the current estimate above is the one I'd stand behind right now.
After Datig reports "several cracks" Sunday eveningMay 24, 8:30 PM PT
Datig at the GKN site (8:13 PM PT) reported fire officials told her the tank now has "several cracks" — plural, not the single controlled-venting point from McGovern's morning. Additional fire hoses + engines visible for aggressive cooling. The progressive-structural-failure hypothesis (cracks-as-first-failure rather than cracks-as-relief) is now substantially more credible. BLEVE odds raised materially; cure path effectively closed.
Controlled spill (managing leak / draining)~30%
Uncontrolled slow leak~30%
Controlled cure or new intervention works~4%
Cooling holds, they ride it out~1%
BLEVE (with cluster cascade)~20–25%
After OCFA corrects "several cracks" reportMay 24, 9:45 PM PT
OCFA posted at 9:35 PM PT: there is ONE known crack on the tank, not several. The appearance of additional cracks was weather-stripping peeling on the exterior. No active leak; atmospheric monitoring confirms no chemicals leaking. This walks back the Datig-driven escalation but doesn't fully revert to the morning baseline — gauge still maxed at 100°F, McGovern's "any time" admission still stands, overnight mission still needed.
Controlled spill (managing leak / draining)~33%
Uncontrolled slow leak~28%
Controlled cure or new intervention works~6%
Cooling holds, they ride it out~1–2%
BLEVE (with cluster cascade)~12–16%
After Datig surfaces OCFA PIO interview (receipts)May 24, 10:05 PM PT
Datig replied with the on-camera interview that was her source for "several cracks." OCFA PIO Brian Yau used "cracks" plural multiple times AND said "both BLEVE and spill risks are still on the table" plus "high risk of catastrophic failure." This contradicts OCFA's Twitter correction. Internal OCFA messaging inconsistency, not a Datig error. BLEVE odds re-revised up to ~17–22% — the Twitter correction may be damage control rather than new information.
Controlled spill (managing leak / draining)~32%
Uncontrolled slow leak~28%
Controlled cure or new intervention works~5%
Cooling holds, they ride it out~1–2%
BLEVE (with cluster cascade)~17–22%
RESOLUTION EVENT — OCFA: BLEVE "off the table"May 25, 8:30 AM PT
Monday 8:03 AM PT briefing — Incident Commander Covey (alongside Interim Chief McGovern) announced "the threat of a BLEVE is now off the table. That threat has been eliminated." Tank temperature now 93°F, DOWN from 100°F+ Sunday. Crack confirmed; pressure released; tank cooling. Evacuation still in place pending verification and cleanup, but the worst case is closed. The overnight recon mission worked. This is the resolution event we've been tracking against. Probability stack restructured around tank-stabilization (now dominant); BLEVE drops to residual.
Tank stabilizes (cure/cooling working)~55%
Controlled spill / managed drain (still possible)~20%
Uncontrolled slow leak~15%
Cooling holds, they ride it out~7%
BLEVE (residual)~1–3%
Evacuation zone contracted at 6 PM PTMay 25, 6:30 PM PT
Following the 8 AM PT BLEVE-eliminated announcement, OCFA held a 6 PM PT press conference and contracted the evacuation zone effective immediately. Updated incident map shows the new (smaller, red-shaded) zone vs the previous (red-outlined) zone. Residents now outside the smaller zone may begin returning home. Inner perimeter remains active for cleanup and verification. The morning prediction of Tue May 26 – Thu May 28 return window is being validated in real time — outer-ring returns starting tonight, inner-ring later in the window. Probability stack holds; the resolution path is playing out as expected.
Tank stabilizes (cure/cooling working)~55%
Controlled spill / managed drain (still possible)~20%
Uncontrolled slow leak~15%
Cooling holds, they ride it out~7%
BLEVE (residual)~1–3%
After Datig site report of firefighter pull-backMay 23, 11:11 PM PT
Alex Datig (verified-reliable; previously vetted on the Goodwin Co. find) reported from the GKN site at 11:07 PM PT that 20+ firefighters in full SCBA had her gear up, then told her it was "too dangerous" and they "were planning to leave." Personnel withdrawal from the perimeter is the pre-failure operational playbook. Combined with CBS LA's earlier "temperature spike" detail and 15+ hours of silence on a public tank reading, the probability mass is shifting from controlled-cure toward slow-leak and BLEVE.
Controlled spill~30%
Controlled cure or new intervention works~22%
Uncontrolled slow leak~20%
Cooling holds, they ride it out~3–5%
BLEVE (with cluster cascade)~12–18%
After 12-facility cluster + 9 PM silence-on-tank signalMay 23, 9:30 PM PT
EPA Envirofacts query found 12 TRI-listed chemical facilities within 1 mi of GKN, not just Goodwin. The 9:18 PM CRITICAL UPDATE was delivered by AQMD + EPA on air quality (clean — good news), but conspicuously skipped any tank-status update from Covey. That's the silence-on-the-number signal flagged earlier. Bumping BLEVE odds slightly up and shifting the controlled-outcomes down a touch.
Controlled spill~33%
Controlled cure or new intervention works~28%
Uncontrolled slow leak~17%
Cooling holds, they ride it out~3–5%
BLEVE (with cluster cascade)~8–12%
After verifying Goodwin Co. cascade riskMay 23, 8:45 PM PT
Independent reporter Alex Datig flagged Goodwin Co. (chemical contract manufacturer ~0.4 mi SW). Verified via EPA TRI: they store methanol, IPA, ammonia, and other flammables. BLEVE probability unchanged but severity worsened — a GKN explosion could cascade into a larger event at Goodwin's facility.
Controlled spill~35%
Controlled cure or new intervention works~30%
Uncontrolled slow leak~15%
Cooling holds, they ride it out~3–5%
BLEVE (now with potential cascade)~7–10%
After Covey 5:58 PM updateMay 23, 6:30 PM PT
Covey: state-level experts brought in, drones poll every 10 min, strategy shifted from "steer toward best failure mode" to "prevent any failure." Less likely to fail at all, and less catastrophic if it does.
Controlled spill~35%
Controlled cure or new intervention works~30%
Uncontrolled slow leak~15%
Cooling holds, they ride it out~3–5%
BLEVE (explosion)~7–10%
After learning about the 15k second tankMay 23, 5:15 PM PT
Transcribed Covey's morning briefing revealed a second 15,000-gal tank near the failing one, plus that manufacturer chemists were already on-site. Added the explicit "controlled cure" outcome.
Controlled spill~50%
Controlled cure — slow polymerization without rupture~20%
Uncontrolled slow leak~15%
Cooling holds, they ride it out~5–8%
BLEVE (explosion)~8–12%
Initial estimateMay 23, 3:30 PM PT
First analysis after the Saturday morning OCFA briefing confirmed the temperature was rising (90°F, +1°F/hr) and the manufacturer chemists weren't yet publicly visible.
Controlled spill (firefighters' preferred path)~50%
Uncontrolled slow leak~25%
Cooling holds, they ride it out~10–15%
BLEVE (explosion)~5–10%
What to watch next (cleanup phase)
With BLEVE off the table, the watch shifts from "will it explode?" to "when can people go home?" The next OCFA briefing's evacuation-zone language is the key signal: "partial lift" / "outer ring returning" would mean residents are days from home; "continuing" with no timeline would mean days more for verification.
Second-order signal: does the next reading show temperature still dropping? 93°F to ~85°F over the next 24h would confirm the reaction is winding down. Stable at 93°F could mean equilibrium with ambient (fine) or the crack venting has plateaued (less fine). A rise back upward would be the only way BLEVE re-enters the picture — very unlikely now but worth tracking until they declare full containment.
Is the response itself the right one?
Critique of an active emergency response is fraught. OCFA has data we don't — internal pressure readings, structural scans, manufacturer guidance, tank-supplier specs. The points below are about what's visible from public communications (including transcribed briefings), not verified failures of their plan.
What they're getting right
- Big evacuation perimeter, fast. 40k+ people moved out within hours. Textbook BLEVE response — no civilians in the worst-case zone.
- Honest public framing. Covey publicly walked back his own optimism: "Yesterday afternoon I did report… we thought we were reducing it… Unfortunately I had to say thought." Updating your own bad news live takes guts.
- Clean chain of command. Chief Covey as incident commander, Dr. Regina Chinsio-Kwong as OC Health Officer, Cal OES backstop, Governor declaration. No turf war.
- Manufacturer's chemist team on-site. Per Covey: "chemists from the emergency response team from the manufacturer" have been working with OCFA since Friday night. Right expertise, right place.
- Went offensive when defense stopped working. Friday night they put crews downrange — with manufacturer chemists — to read the internal gauge directly and start neutralizing the 15k tank. Covey: "We did put them back in harm's way last night." Active intervention, not stall.
- Spill-containment infrastructure being staged. Covey detailed prepping diking, damming, and diverting the liquid to a downgrade holding area if the tank fails — explicitly to keep MMA out of storm drains and the ocean. Environmental contingency, not just life-safety.
- Naming the "third option" publicly. Covey explicitly broke out of the binary "spill OR explode" framing with a controlled-cure option: let the MMA polymerize slowly into solid plastic, with the void space above absorbing pressure. Whether it works is uncertain — but admitting Option 3 exists is correct.
- Bringing in state-level experts to develop new options. Per Covey's 5:58 PM update: "we brought in subject matter experts from all across the state to think completely outside the box and we had some really good productive conversations today." Escalating beyond local capacity — the right move when the standard playbook is losing.
- Drone polling every 10 minutes. Per the evening update: "We got the drones on them every 10 minutes watching for any temperature changes or spikes." Not as good as internal telemetry, but near-continuous external monitoring is materially better than I'd assumed earlier.
- Air-quality monitoring confirmed clean. Per the 9:18 PM briefing: South Coast AQMD has a mobile van measuring around the evacuation zone since Friday, and EPA has 24 stationary monitors operating 24/7. Dr. Jason Lowe (AQMD): "levels are completely normal." Harry Allen (EPA): "we have not seen any contaminants in those monitoring stations." Real measurement, multiple agencies, public release.
- ✅ The strategy worked. Three-day arc: cool the tank, neutralize tank #2 (done Friday night), let the crack vent pressure, do an overnight recon when sun exposure was lowest, monitor cooling. Outcome: BLEVE "off the table" by Monday morning, tank cooling 100°F+ → 93°F overnight, zero casualties, zero detectable chemical release, zero secondary fires. The communications during the incident were genuinely wobbly (PIO/Twitter disagreement, leadership rotations), but the operational decisions worked. Credit where due.
Open questions about the approach
✓ RESOLVED: How fast was the crack venting? Fast enough.
Monday 8:03 AM PT briefing confirmed the answer empirically: tank temperature dropped from 100°F+ Sunday to 93°F Monday. The crack vented enough pressure to cool the tank ~10°F overnight. The Goldilocks zone held — venting fast enough to prevent rupture, contained enough that no measurable chemical release occurred at AQMD/EPA monitors. The vent rate question is now answered by outcome rather than measurement.
Internal telemetry still missing
Covey's evening update says drones poll the tank exterior every 10 minutes — that's better than I credited earlier. But the gauge they had to physically enter the site to read on Saturday morning is still the only internal data point. A 34k-gallon reactive-monomer tank near a residential area should have continuous internal pressure + temp telemetry. It doesn't. Regulatory gap worth flagging after the incident.
Hot-tap option not mentioned in any briefing
Drilling a fresh port into the tank is the textbook fix for a stuck valve on reactive monomers. Covey did say they're "bringing people in from all over the country" for ideas, so it may be on the table internally. But it hasn't been named publicly. Worth a sentence either way.
✓ RESOLVED: Tank #2 successfully inhibited
Per OCFA's updated framing and federal incident reporting, inhibitor has been successfully added to the 15k tank. It is now structurally sound and being monitored. The cascade vector from tank #2 is materially reduced. This was a major operational win that didn't get a standalone briefing — buried in updates rather than announced. Leaving the question here for transparency about how the picture evolved.
✓ RESOLVED: The time-asymmetry concern didn't materialize
This was about whether the polymerization would go non-linear and outrun the cure-and-wait strategy. As of Monday 8:03 AM PT, the answer is no — the reaction stayed linear enough that the crack-venting strategy worked. Tank temperature went from 100°F+ Sunday to 93°F Monday. The autocatalytic acceleration we worried about did not occur. Preserved as historical context for the chemistry decision-making timeline.
Blast map is published but not parameterized
OCFA's Saturday briefing showed a six-zone map (3 blast damage circles + 3 toxic-impact zones — red/orange/yellow). Useful — but no actual radii in feet or miles have been published, so a resident can't tell which damage circle their house sits in. A picture isn't a substitute for distance numbers.
Cascade risk: Goodwin Co. ~0.4 mi southwest
An independent reporter (Alex Datig) flagged Goodwin Co. at 12361 Monarch St — a chemical contract manufacturer ~2,086 ft from GKN. We verified: EPA TRI filings list methanol, isopropyl alcohol, ammonia, ethylene glycol, phosphoric acid, glycol ethers, and others on site. Goodwin's facility is 150,000+ sq ft with 42 tanks. If GKN BLEVEs, the pressure wave / radiant heat / debris reaches Goodwin. A ruptured methanol or IPA tank there could trigger a second, larger event. No public OCFA briefing has mentioned coordination with Goodwin.
✓ RESOLVED: Why OCFA kept skipping the temperature number
Covey explained directly in the Monday 8:03 AM PT briefing: "We were not doing tank temperature checks during the day while the sun was on it... We were only doing tank temperatures at night." The silence-on-the-number pattern wasn't damage control — it was risk management for firefighter safety. Daytime sun heat would have made tank-temperature checks both inaccurate and dangerous. The pattern was real; the interpretation (it's-getting-worse) was mostly wrong.
GKN Aerospace's corporate silence (partially broken Sunday)
After 3+ days of complete silence, GKN issued a four-paragraph written statement Sunday — an apology for "the ongoing disruption," acknowledgment that they are "working closely with OCFA, EPA, OSHA," and a generic "safe resolution" line. No executive on camera, no technical detail, no timeline commitment, no acknowledgment of the gauge limitation or the broken valve. Notably, GKN's statement says they are "working to mitigate the risk of a leak" — which lags reality by hours, since McGovern publicly confirmed the crack-venting leak Sunday morning. The substantive silence remains: corporate non-communication on the technical facts during an active crisis is itself a tell. OC DA Todd Spitzer's criminal probe and tip line for GKN employees (714-347-8714 / tipster@ocdistrictattorney.gov) is still active. Rep. Derek Tran has called for a "deep investigation." Multiple law firms have announced planned class-action lawsuits.
✓ MOSTLY RESOLVED: The Saturday-night "retreat" was actually planned recon
CBS LA's 10:24 PM PT broadcast Saturday described crews "pulling back after seeing the temperature reading spike" — read at the time as pre-failure withdrawal. McGovern's Sunday 10:47 AM briefing reframed this: it was an overnight specialized-team recon mission to visualize the tank, find the crack, and gather data. Datig's "too dangerous, planning to leave" observation was the recon team returning after their mission. The temperature-spike framing CBS LA reported is still unexplained — could have been a transient reading during the recon, could have been a misquote. Either way, the operation was deliberate, not panic-driven.
Tank gauge maxing out at 100°F — partly explained, still a regulatory question
Covey's Monday 8:03 AM briefing partially addressed this: OCFA was specifically limiting tank readings to overnight hours to manage firefighter sun-exposure risk during daytime. So the operational gap was risk management, not capability. BUT — the underlying question remains: why does a 34,000-gallon reactive-monomer tank have a gauge that maxes at 100°F when industrial cure-runaway thresholds start at 104°F? That's a pre-incident specification choice that limited everyone's awareness through the most dangerous part of the event. Still worth flagging for the post-incident regulatory review.
The single question worth answering publicly
"What's the current internal pressure inside the failing tank, and how much margin do we have before the void space is overwhelmed?"
The controlled-cure strategy lives or dies on the void-space-vs-pressure equation. Temperature is only a proxy. Pressure is the metric that tells us whether Covey's "third option" is winning or losing — and it hasn't been shared publicly.
What we know so far — the basics
What's in the tankA chemical used to make plastic (called MMA)
How much34,000 gallons — enough to fill a swimming pool
Second tank nearby15,000 gallons — inhibitor successfully added, tank reported structurally sound and being monitored. This means it can't be triggered if the first tank fails.
How hot it is93°F (34°C) as of Monday 8 AM — DOWN from 100°F+ on Sunday. The crack discovered Sunday is releasing pressure successfully, and the tank is now cooling. OCFA announced Monday morning that the BLEVE threat is off the table.
People told to leaveAbout 16,000
Governor's responseDeclared an emergency
Why — what's happening, in plain words
The tank holds a liquid that really wants to turn into plastic. To stop that from happening, the company mixes in a tiny amount of a "safety chemical" that keeps it calm. That safety chemical only works when the liquid stays cool and has air mixed in.
The tank is heating up. As it heats, the air gets pushed out and the safety chemical stops working. Once that happens, the liquid starts turning into plastic — and that process gives off heat. More heat means it turns into plastic faster. Faster means even more heat. That's the loop firefighters are trying to break.
Why is 90°F dangerous? That just sounds like a warm day.
Reasonable question — 90°F (32°C) on its own isn't alarming. That's a typical summer afternoon. The problem isn't the number, it's the situation around it.
The chemical industry's storage guideline for MMA says: keep it between 18–40°C (64–104°F), preferably under 30°C (89°F). At 90°F the tank is already at the warmer end of "safe." The standard industrial-evacuation threshold is 60°C (140°F). The tank is heating at about 1°F per hour, with no working cooling. At that rate it crosses the 140°F evacuation threshold in about two days.
Worse: the "safety chemical" mixed into the MMA needs air mixed into the liquid to work. As the tank heats, the air bubbles out. Once the safety chemical stops working — which may have already happened — the MMA can start turning into plastic on its own. That reaction gives off heat, which speeds up the next round, which gives off more heat. So the temperature is supposed to rise linearly at 1°F/hr — but once the feedback loop really kicks in, the rate can double, then double again.
So 90°F is dangerous in the same way "falling 6 inches" is dangerous if you're at the edge of a cliff. The number itself is fine. The trajectory and the lack of an off-ramp is what makes it a crisis.
Wait, there are two tanks?
Yes. Per OCFA Incident Commander Craig Covey's Saturday-morning briefing (we transcribed it from his X video update — it's not in most press coverage):
Tank #1 (the failing one): ~7,000 gallons of MMA still in the original 34,000-gal storage tank. Heating up, can't be accessed normally because of the broken valve. This is the one the countdown is about.
Tank #2 (nearby): ~15,000 gallons. Friday night, crews went "offensive" with the chemical manufacturer's emergency response team to start neutralizing this second tank — meaning making it inert so it can't be triggered if Tank #1 explodes. This is good and important work, because the worst-case scenario (Tank #1 BLEVE setting off Tank #2) would be much bigger.
What does the chemical smell like? How would I know if it reached me?
MMA has a strong, sweet, fruity smell — OCFA's own description. If you're outside the evacuation zone but suddenly notice a sweet/fruity odor in the air, that may be vapor drifting from the tank. Don't stand around to investigate: move indoors, close windows, turn off any AC that pulls in outside air, and call the OC public info line (714-628-7085).
The fruity smell is detectable at very low concentrations — you'll smell it well before it reaches a level that's actually dangerous to breathe. So smelling it isn't the same as being in immediate harm — but it is a signal to take precautions.
What do the colors on OCFA's map mean?
OCFA's briefing map has six zones — three circles for "if it explodes" and three outer zones for "if the gas leaks":
Blast damage circles (closest to the tank):
• Innermost — severe structural damage
• Middle — moderate damage
• Outer — light damage still possible
Gas/vapor zones (broader area):
• Red — flammable; could catch fire or flash if vapor reaches this area
• Orange — immediate health danger; breathing this would cause injury
• Yellow — odor threshold; you'd smell it but it's not harmful at this level
OCFA hasn't published the actual distances yet — there's no way to tell from outside which circle your address falls in.
Where is the broken valve? Can't they just fix it?
Officials haven't shown a picture or said exactly which valve. All we know from press conferences is that OCFA Chief Craig Covey described it as "broken and gummed up," and that the damage stops crews from safely getting to the chemical inside.
Big chemical tanks like this one have several valves — usually on top (for adding chemicals or venting pressure), on the side (sampling ports), and on the bottom (draining). For this kind of tank in an aerospace plant, the broken valve is most likely the one they would normally use to drain the tank or add a stabilizer. Without it working, they can't offload the liquid, can't pump in fresh "safety chemical," and can't relieve the pressure that's building up inside.
Why not just send someone in to swap the valve? Three reasons: (1) the area right next to the tank has poisonous fumes leaking out, so anyone close enough to touch the valve is in serious danger; (2) the gas around the tank is flammable, so a single spark from a wrench could start a fire or set off the explosion everyone is trying to prevent; (3) the heat may have warped the tank itself, which can seize the valves into the metal so they can't be removed even with the right tools.
Why is it heating up in the first place?
A valve on the tank broke. That let a tiny bit of warming start — maybe from sun on the tank, maybe from a small reaction inside. Once the liquid got a little warmer, the air mixed into it started bubbling out. The "safety chemical" only works when there's air mixed in, so it stopped doing its job.
Now the liquid is slowly turning into plastic, which gives off heat. That heat pushes more air out, which makes more plastic, which makes more heat. It's a feedback loop — the warmer it gets, the faster it gets warmer. Firefighters can't open the tank to pour in fresh safety chemical because the broken valve is in the way and the gas inside could catch fire.
Why are they spraying water on it?
To pull heat out faster than the chemical is making it. If gas starts leaking, the water also catches the gas in the air so it doesn't drift over neighborhoods.
What could actually go wrong?
The tank leaks: a seam pops and 6,000–7,000 gallons spill into the parking lot. The fumes are poisonous and can catch fire, but it's a smaller problem.
The tank explodes: if the heat loop wins, pressure builds up inside until the tank bursts. That's a much bigger blast, with a much bigger poison cloud.
If it actually explodes, what would happen?
If the tank ruptures, the pressurized MMA inside flash-boils from liquid to vapor instantly, expanding ~1,500× in volume in milliseconds. That vapor ignites almost immediately because the hot tank exterior is above MMA's autoignition temperature and there are sparks and hot metal everywhere. The result is a fireball roughly 260–330 feet across that lasts 8–15 seconds. Raw chemical energy: ~240 GJ (≈57 tons of TNT chemically), but only ~5–6 tons TNT's worth converts to actual blast pressure.
By distance from the tank: within 300 ft, total destruction and lethal heat. From 300–1,000 ft, severe structural damage and 3rd-degree burns. From 1,000–2,000 ft, 2nd-degree burns and heavy roof/window damage. From 2,000–5,000 ft (≈1 mile), window breakage and 1st-degree burns. Beyond a mile, mostly vibration and noise. Tank fragments can also fly 1,000–3,000 feet — in the 1984 San Juanico disaster in Mexico, tank pieces flew 10,000 feet (3 km).
The bigger concern is what's nearby: 12 EPA-listed chemical facilities are within 1 mile of GKN. The two that materially change the worst case are Goodwin Co. (1,400 ft southwest, stores up to 1.9M lb of ethylene glycol plus methanol, isopropyl alcohol, ammonia, and other flammables) and Holcim (stores toluene diisocyanate). A GKN explosion delivers ~10–15 kW/m² of radiant heat to Goodwin — enough to ignite roofing and damage tanks via heat alone. If Goodwin's methanol or IPA tanks rupture in that heat, you get a secondary explosion 30–60 minutes later. If Holcim's TDI is involved in a fire, it can release hydrogen cyanide (HCN) — a toxic gas lethal at very low concentrations that travels with wind.
Combined worst-case (GKN plus cluster cascade): roughly 7–8 tons TNT equivalent over a square mile, with a toxic plume that could extend 5–15 miles downwind if HCN is involved. OCFA's evacuation perimeter is roughly 1 mile from GKN — explicitly sized to put residents outside the lethal-blast and lethal-thermal radius for both the primary fireball AND the cluster cascade scenario. The hospital and 11 nursing homes that were relocated to DHS Metropolitan in Norwalk extended that footprint further. The reason the evacuation zone looks 'too big' for a single tank is that it isn't sized for a single tank.
Important framing: this is the scenario if the tank explodes — not a prediction that it will. With current odds at ~20–25% BLEVE risk and ~60% managed-leak / controlled-drain, the most likely outcome remains a semi-controlled resolution. But everything authorities are doing right now — the evacuation footprint, the FEMA 14-day mass-care planning, the hospital relocations, the aggressive cooling expansion — reflects that they are actively planning for this worst case. If you're inside the zone, you should be too.