NOAA Just Called a Very Strong El Niño. Here Is What That Does to a Glycol Loop, Gulf Coast to Great Lakes

On 10 September 2026, NOAA’s Climate Prediction Center issued its monthly ENSO diagnostic discussion with the alert status set to El Niño Advisory and one sentence that should be pinned above every maintenance planner’s desk in the country: “El Niño is strengthening, with a greater than 90% chance of a very strong event during the Northern Hemisphere fall and winter 2026-27.” Two days earlier, NOAA’s Geophysical Fluid Dynamics Laboratory had published its September model run, in which “all 30 ensemble members” show an event that “will compete with, if not surpass, the strongest events in the historical record.”

We ship propylene glycol and ethylene glycol out of Taylor, Texas to every state, by the quart and by the tote, and every autumn the same question arrives in the inbox in slightly different words: how much glycol do I need? This year it comes with a forecast attached, and the forecast is being widely misread in both directions. In the South one camp hears “strong El Niño” and expects a mild winter it can ignore, while the other hears “strongest on record” and expects February 2021 again. In the North the forecast of a warm winter is being read as permission to skip the walk-down. NOAA’s own pattern work supports none of those readings, and the difference matters for how you spec a loop. What follows is what the agency actually said, what an El Niño winter does region by region, and the two glycol numbers that most winterization specs get backwards.

What NOAA actually said on September 10

The ENSO diagnostic discussion is a short, dry document issued on the second Thursday of each month by the Climate Prediction Center, and the September edition is unusually direct. Equatorial sea-surface temperatures are above average across the east-central and eastern Pacific, the atmosphere over the equatorial Pacific is behaving the way it does during El Niño, and the Niño-3.4 index, the standard measuring stick for the event, reached +1.8 °C. The status line reads El Niño Advisory, which means the event is present, not merely forecast. The forecast sentence is the one quoted above: a greater than 90% chance of a very strong event, which is the top of the agency’s strength scale, persisting through the Northern Hemisphere fall and winter of 2026–27.

GFDL’s September update, published on 8 September, reads the same way with one honest hedge in each direction. The lab notes a “slightly weaker peak El Niño strength compared with last month’s forecast,” and then that every one of its thirty ensemble members still puts this event in competition with the strongest on record. It also notes that “the spread of the 30 ensemble members increases by the spring,” which is a polite way of saying that the winter is the confident part of the forecast and the exit from it is not.

What an El Niño winter does, region by region

NOAA’s climate.gov has done the historical work, and the signal is not subtle. Looking at the strongest El Niño events since 1950, the agency finds wetter-than-normal winters along the West Coast and the southern tier of the United States. The National Weather Service office in Paducah, Kentucky puts the whole map in three sentences: “The southern states from California and Arizona to Florida have an increased risk of experiencing wetter conditions compared to normal as storm systems tend to track repeatedly across these locations. The Ohio Valley and northern states from Montana into upstate New York have an increased risk of drier than normal conditions.” And on temperature: “The northern tier of the U.S. has an increased risk of extreme warmth, but the odds decrease as you head into the latter part of winter and early spring.” Climate.gov adds that over California and the Southwest the wet signal “depends significantly on the strength of the El Niño,” and that during the two strongest events of the past sixty years, 1982–83 and 1997–98, “much-above-median rainfall amounts fell across the entire state of California.” The Gulf Coast is the most dependable piece of the map, and climate.gov’s summary of it is worth quoting in full: “The most reliable of these signals (the one that has been observed most frequently) is wetter-than-average conditions along the Gulf Coast from Texas to Florida during this 6-month period. This relationship has occurred during more than 80% of the El Niño events in the past 100 years.” And for the strong events specifically: “the Gulf Coast and Southeast are consistently wetter than average.” The relationship is a cold-season one, October through March, which is exactly the winterization window.

The temperature side is where the misreading starts, in both halves of the country. In the South, the National Weather Service office in Tallahassee, which lives with this pattern every winter, describes the mechanism precisely: “The Gulf Coast generally sees cooler and wetter conditions, not because of numerous arctic outbreaks, but because of the stronger influence of the subtropical jet stream.” In an El Niño winter the flow across the Pacific and North America is more west-to-east, which, in the office’s words, “produces less Arctic outbreaks,” while the dominant subtropical jet “produces more low pressure systems, rainfall, and clouds, and it is generally cooler in the southern U.S.”

Read that table as a winterization planner and two different pictures emerge, with the same lesson. Across the South the average night is milder and wetter, there are more storm systems on the calendar than in a normal year, and behind each of them sits a short cold sector, usually two or three nights, landing on equipment that is wet. Across the North the average is warmer and the ground is drier, and the danger is complacency: the coldest night of a “mild” winter in Minneapolis or Buffalo is still a night that would freeze a Texas loop solid twice over. Neither is the dry La Niña shape of one enormous Arctic intrusion, and both argue for the same response: not panic, not complacency, but a concentration chosen for the coldest night your site actually sees rather than the seasonal mean.

This is not a forecast of February 2021, and that is the point

The reference every facilities manager in the South reaches for is Winter Storm Uri, the February 2021 freeze that took the grid down and, with it, most of the Gulf Coast chemical industry. It is worth being precise about that event because it is being used as the mental model for this winter, and it is the wrong one. The winter of 2020–21 was a La Niña winter by NOAA’s own classification: exactly the large-amplitude, Arctic-outbreak pattern the Tallahassee office describes as the opposite of El Niño. Uri was a single deep intrusion of Arctic air that sat on the state for a week. A very strong El Niño makes that specific shape less likely, not more.

What Uri does teach is what a hard freeze does to unprotected equipment, and the Federal Reserve Bank of Dallas wrote the clearest account. In its second-quarter 2021 Southwest Economy, Jesse Thompson reported that “as much as 80 percent of U.S. basic organic chemicals capacity was offline after the storm, and up to 60 percent was still offline in mid-March,” citing Wood Mackenzie estimates, and that the chemical sector’s 8 percent production decline in February was the largest one-month drop since 1972. The mechanism he identified is the one that should be pinned next to the NOAA sentence: “The unexpected and long-duration cold, sudden power loss and disruption of natural gas liquids supplies precluded a normal, orderly shutdown. This caused more damage that took longer to identify and repair.”

Notice what is not in that sentence. The glycol loops were not the failure. The damage that took months to find was in instrumentation, sample lines, seals and the small-bore plumbing that nobody had put glycol in because nobody thought of it as part of the cooling system. We come back to that below, because in a winter with more cold snaps and fewer deep freezes, it is the most likely place a plant gets hurt, in Beaumont or in Toledo.

Freeze protection and burst protection are two different specifications

Every glycol supplier publishes a concentration table, and almost every buyer reads the wrong column of it. The clearest published treatment is in Dow’s engineering and operating guide for its DOWFROST propylene glycol heat-transfer fluids, which opens the topic with a distinction most winterization specs never make: “There are two basic types of protection available: ‘burst protection’ and ‘freeze protection.’”

Dow defines them like this. Burst protection “is sufficient if the system will remain dormant when the temperature is below the freezing point of the solution. In HVAC applications, burst protection is considered an appropriate safeguard in systems where there is adequate space to accommodate the expansion of an ice/slush mixture and the system is inactive during the winter.” The physics is the part worth understanding: “as the temperature drops below the solution’s freezing point, ice crystals begin to form. Because water in the solution freezes first, the remaining glycol solution becomes further concentrated and remains fluid. The combination of ice crystals and fluid results in a flowable slush. Fluid volume increases as this slush forms, with the extra volume flowing into available expansion volume in the system. If the concentration of glycol is sufficient, system damage will not occur.”

Freeze protection is the stricter case: “required in systems where fluid must be pumped at the lowest anticipated temperature. Freeze protection is essential in cases where no ice crystals can be permitted to form or where there is inadequate expansion volume available to accommodate ice/slush formation.”

Here is Dow’s Table 4, the concentrations of its propylene glycol fluid required for each kind of protection. Two notes before you read it. The figures are volume percent of DOWFROST fluid, which Dow describes as 95.5 percent propylene glycol with dipotassium phosphate and water, so a bare propylene glycol figure runs a shade lower. And the table’s own footnote is the most important sentence in the guide: “Generally, for an extended margin of protection, you should select a temperature in this table that is at least 5°F (3°C) lower than the expected lowest ambient temperature.”

Two things in that table decide most Gulf Coast specs. First, the gap between the columns is not small. At 0 °F, which after the 5 °F margin covers a design low of 5 °F, a loop that must keep pumping needs half again as much glycol as a loop that only has to survive. Second, look at the burst column stop climbing. From −40 °F down, it sits at 36.6 percent no matter how cold the table goes, which is why Dow’s text can say flatly that “for burst protection, a 35 percent (volume) solution of propylene glycol” “is usually adequate.” A little over a third glycol and the slush will not break the pipe at any temperature in the table, which reaches well past anything the lower 48 produce.

More glycol is not more protection

The single most useful counterintuitive fact in this business, which we wrote up at length in our complete guide to ethylene glycol, is that the freezing-point curve of a glycol and water blend is U-shaped. Pure ethylene glycol freezes at about 9 °F. A 50/50 blend with water protects to roughly −34 °F. Past about 60 percent glycol the freeze point starts climbing back up. “Just run it straight” is the one mix that fails in both directions: it freezes warmer than the blend, and it pumps like syrup.

Propylene glycol behaves the same way in the range that matters. Look again at Dow’s freeze column: each step colder costs more concentration than the last, from 18.8 percent for 20 °F to 62.8 percent for −60 °F, and the guide stops the table there. Dow rates its fluid’s water solutions for “freeze protection to below −60°F (−51°C) and burst protection to below −100°F (−73°C),” which is more winter than any customer of ours will see. The practical lesson is the same for both glycols: there is a concentration that buys the protection you need, and going past it buys viscosity, cost and a worse chiller, not safety.

The line that freezes first is the one with no glycol in it

This is the section to hand to whoever does the winter walk-down, because it is where the Dallas Fed’s “damage that took longer to identify and repair” actually lives. A glycol loop, by definition, has glycol in it. The plumbing around it frequently does not:

None of these are exotic, and every one of them fails on a two-night cold snap that the main loop, correctly filled, sails through. In the El Niño winter NOAA is describing, that two-night snap is the likely shape of the risk. The glycol order is the easy half of winterization. The walk-down is the half that decides whether the plant is running on the third morning.

A worked example

Take a 500-gallon closed cooling loop on an outdoor skid east of Houston. The owner looks up the coldest night the site has actually seen in recent memory and calls it 10 °F, then applies Dow’s margin and designs to 5 °F, which puts the spec on the table’s 0 °F row.

• If the loop must run through the snap (a process chiller that cannot stop), the freeze-protection column applies: 37.7 percent of the fluid volume, so roughly 189 gallons of glycol fluid in a 500-gallon system, the balance water.
• If the loop can sit idle (a comfort-cooling loop shut down for the winter, with expansion volume in the system), the burst-protection column applies: 25.1 percent, roughly 126 gallons.

Now move the same 500-gallon skid to the Upper Midwest, where the owner’s coldest recent night is −25 °F and the margin puts the design on the −30 °F row. Freeze protection asks 52.4 percent, roughly 262 gallons of glycol fluid; burst protection asks 34.6 percent, roughly 173 gallons. NOAA’s forecast of a warmer-than-usual northern winter changes none of those figures, because the table is entered with the coldest night the site can see, not with the season’s average.

Sixty-three gallons is the difference between the two readings of the same table in Texas, ninety in Minnesota, and it compounds. The 25 percent loop pumps easier and moves heat better for the whole of the next summer; the 38 percent loop is the only one of the two that can be running at 5 °F. Neither is right in the abstract. The question that decides it is whether the pump is on when the front comes through, and only the owner knows that.

Propylene glycol or ethylene glycol

Both do the job in the table above, and the choice between them is usually made for you by what the loop is near rather than by the freeze point. Ethylene glycol is the more efficient heat-transfer fluid and the cheaper one per gallon of protection; it is the base of most engine antifreeze and most industrial chiller charges, and our engineer’s guide to ethylene glycol heat-transfer systems covers the design side in depth. Propylene glycol is specified where a leak could reach food, beverage or drinking-water contact, in glycol chillers for breweries and dairies, and in HVAC loops inside occupied buildings; it is also what a purple-dyed heat-transfer fluid like our Arctic Assist is built on. The regulatory and handling differences between the two glycols are laid out in the ethylene glycol guide linked above, and that is the right place to make the call. What both need for a Gulf Coast winter is the same: an inhibited product, a concentration read from the correct column, and a margin below the coldest night you believe in.

What this means if you buy glycol from us

We stock both glycols in Taylor, Texas, in the forms a winterization actually calls for: concentrate you blend to your own number, and pre-mixed blends when the number is already known.

Need the Certificate of Analysis for a lot? Just ask; we send it over at no charge. And if you are sizing a loop and want a second pair of eyes on the arithmetic, send us the lowest ambient you design to, the system volume, and whether the pump runs through the winter. We are glad to talk through the concentration and which of the two columns you are actually in.

Common questions

What did NOAA say about El Niño in September 2026?

In its ENSO diagnostic discussion of 10 September 2026, NOAA’s Climate Prediction Center set the alert status to El Niño Advisory, reported the Niño-3.4 index at +1.8 °C, and stated that El Niño is strengthening with a greater than 90% chance of a very strong event during the Northern Hemisphere fall and winter of 2026–27. NOAA’s GFDL laboratory reported on 8 September that all 30 of its ensemble members show an event that will compete with, if not surpass, the strongest on record.

Does a strong El Niño mean a cold winter, and where?

Not in the sense of more Arctic outbreaks anywhere. NOAA’s historical work shows the southern tier from California to Florida and up the East Coast runs wetter than average, and the Gulf Coast mildly cooler, because the subtropical jet stream brings more storm systems, clouds and rain; the National Weather Service describes that cooling as coming not from Arctic outbreaks but from the stronger subtropical jet. The northern tier from Montana to upstate New York and the Great Lakes typically runs warmer than usual and drier than normal. What the southern pattern brings is more frontal passages, each with a short cold sector behind it; what the northern pattern brings is a milder average that still contains the coldest nights of the year. In both cases the coldest individual night is what the glycol has to be specified for.

What is the difference between freeze protection and burst protection?

Freeze protection means the glycol solution stays liquid and pumpable at the lowest temperature the system will see, with no ice crystals forming. Burst protection means ice crystals are allowed to form but the mixture stays a flowable slush that expands into available space instead of rupturing the pipe; it is appropriate only for a system that sits idle when it is below the solution’s freezing point and has expansion volume. Burst protection needs much less glycol: on Dow’s DOWFROST table, 25.1 percent of the fluid by volume at 0 °F against 37.7 percent for freeze protection.

How much propylene glycol do I need for freeze protection at 0 °F?

Dow’s Table 4 calls for 37.7 percent by volume of its DOWFROST fluid, which is 95.5 percent propylene glycol, for freeze protection at 0 °F, and 25.1 percent for burst protection. Dow also advises selecting a protection temperature at least 5 °F below the lowest ambient you expect, so a site that has seen 5 °F should read the 0 °F row. In a 500-gallon loop that is roughly 189 gallons of glycol fluid for freeze protection or 126 gallons for burst protection.

Is more glycol always better?

No. The freezing point of a glycol and water blend follows a U-shaped curve: pure ethylene glycol freezes at about 9 °F while a 50/50 blend protects to roughly −34 °F, and past about 60 percent glycol the freeze point rises again. Over-concentrating also raises viscosity, lowers heat transfer and increases pumping cost for the whole year. Choose the concentration that gives the protection you need at a temperature 5 °F below your design low, and make sure the inhibitor level is adequate, which Dow notes needs separate attention below 30 percent glycol.

Should I use propylene glycol or ethylene glycol for winterization?

Both provide freeze and burst protection at the concentrations in the tables. Ethylene glycol transfers heat more efficiently and is the usual base for engine antifreeze and industrial chillers. Propylene glycol is specified where a leak could reach food, beverage or drinking-water contact, in glycol chillers for breweries and dairies, and in loops inside occupied buildings. In either case use an inhibited product, read the correct column of the concentration table for whether the loop runs or sits, and winterize the plain-water lines around the loop, which are what fail first.