Venezuela: The Blackout Ledger

This is a guide to a dam with water to spare, a country rationed in the dark, and the one number nobody in charge will print.

On the night of September 30, on the highway near San Joaquín and San Mateo, National Guard officers told protesters the power collapse came from climate change, El Pitazo reported.

The same night, in La Ensenada, in the Zulia municipality of La Cañada de Urdaneta, the Guard fired tear gas into another crowd. In videos circulating on social media, El Pitazo reported, a protester shouts at the soldiers: “Hermano, hay niños y ustedes están tirando bombas lacrimógenas.” Brother, there are children, and you are throwing tear gas.

It was the sixth night in a row. El Pitazo, a Venezuelan outlet, counted at least 20 protests that night across 7 states.

The water gives the story away. In July, the reservoir behind the Guri dam stood at 269.11 meters above sea level, according to an analyst’s reconstruction from historical records and satellite altimetry. That is about 2 meters below last September’s peak of roughly 271.2 meters, and about 26.5 meters above the April 2016 low. And in mid-July, by his account, the spillway was opened, because the turbines that still worked could not take all the water coming in.

On the best evidence available, Venezuela had the water this summer. What it is short of is working machines to turn that water into power, and the people in charge do not publish the figure that would let you check.

What we know, and how well

Start with demand, because that number the government does publish.

On September 19, at 1:04 in the morning, the Ministry of Electric Energy announced a record. Demand had hit 16,030 megawatts (MW) on Friday, September 18, which it called “the highest level recorded in the last nine years.” It blamed high temperatures and asked people to save energy. The release gave no supply figure.

The record did not last a week. On September 22, according to Venezuelanalysis, the ministry reported a new high of 16,424 MW. We could not find a ministry post from that day. In remarks reported September 25, acting president Delcy Rodríguez put demand at more than 16,400 MW, the most since 2017, and tied it to heat from a “Superniño.” In a ministry post dated September 26, Juan José Ramírez, the sectoral vice president for public works and services, called the 16,424 MW reached the previous week “a historic milestone,” a level not recorded since 2017.

Supply is harder. Most supply numbers here come from company or ministry reports that El Pitazo cited and we have not seen. The newest pair: on September 30, El Pitazo reported that a Corpoelec report put demand at around 15,400 MW at the start of that week, and generation at around 12,650 MW, of which only 2,400 MW came from thermal plants. El Pitazo puts the deficit near 2,800 MW. The same day, Venezuelanalysis put “actual stable generation” at just 12,000 MW, without saying where the figure came from.

An older pair points the same way. The ministry reported a demand peak of 15,625 MW on August 10, then an eight-year high. El Pitazo, citing Energy Ministry reports, wrote that supply “does not exceed 13,000 MW,” without clearly dating that number to the same day. If both figures describe the same day, which the article does not say, the gap was at least 2,625 MW. That is our arithmetic, not an official figure. Sources linked to Corpoelec, the state power company, told El Pitazo in a September 21 report that the deficit may have topped 3,000 MW. They spoke anonymously.

That leaves the hole in the record. “Neither Corpoelec nor the Ministry of Electric Energy report the deficit,” El Pitazo wrote on September 21. Back in May, the same outlet reported no official word on why the rationing was happening or how long it would last. The closest anyone in charge has come: on September 29, Electric Energy Minister Rolando Alcalá said rising demand was widening “the gap between available generation and meeting demand.” He named the gap. He did not size it.

Sorted by how much weight each piece can bear, the record looks like this:

• Confirmed by the state: demand of 16,030 MW on September 18, then a new peak of 16,424 MW the following week (the sectoral vice president, in the ministry’s own September 26 post); the minister’s admission that a gap exists.
• Reported, sourced to company papers we have not seen: demand near 15,400 MW and generation near 12,650 MW at the end of September; supply of no more than 13,000 MW around August 10; 7 of Guri’s 20 units out in mid-September.
• Estimates: El Pitazo’s deficit near 2,800 MW (September 30); our subtraction, at least 2,625 MW around August 10, if the supply and demand figures share a date; anonymous company sources’ possible 3,000 MW or more (September 21).
• Unknown: the official shortfall, today, in megawatts. As of September 21, El Pitazo reported, neither Corpoelec nor the ministry reported it.

One more dated fact. In 2017, with demand at similar levels and falling, Corpoelec already struggled to cover it. This time demand is setting records.

A full tank and a cracked engine

A car with a full tank and a cracked engine block will not get you anywhere. You can top off the tank all day. The car stays in the driveway.

Operations people have a name for this. Eliyahu Goldratt called it the Theory of Constraints: every system has one bottleneck, and adding more of anything upstream of it changes nothing downstream. At Guri, water sits upstream. The turbines are the bottleneck.

According to Corpoelec reports El Pitazo had access to, and engineer José Aguilar verified, 7 of Guri’s 20 generating units were unavailable in mid-September. Two had been idle since 2014. One since 2016. One stopped in 2025. One stopped on a date nobody gave. The last two went down within the previous month. The dam was producing 6,141 MW out of 10,350 MW installed. That left 4,209 MW of installed capacity unused beside a lake that, at the last reading we have, in July, was close to its recent peak. One of the seven, Unit 15, has since come back at reduced output.

The biggest failure comes down to one part you could almost hold. Unit 17 is a 770 MW machine in the dam’s second powerhouse, modernized in 2004. On August 29 it shut down in a mechanical emergency: the lower fixed wear ring of its runner, the wheel the water turns, partly came loose. The account comes from Corpoelec reports, as handled by Aguilar. The rest are his estimates, not the company’s. Rebuilding it needs a new runner and could take at least two years. He also counts a loss near 430 MW at the Caruachi and Macagua plants downstream. Add the 770 MW unit, and by El Pitazo’s count one broken ring has cost at least 1,200 MW.

El Pitazo also reports that Corpoelec’s authorities were warned about Unit 17 months earlier and that the alert was ignored. It rests on one outlet’s sourcing, so treat it as an allegation.

Then Unit 15. On Thursday, September 10, it made an emergency stop with problems in its runner blades. Even before that, it topped out at 580 MW of a rated 730 and did not run every day. Sector sources say it came back at about 550 MW by September 21. Aguilar adds that the parts to modernize Unit 15 are already sitting at Guri. The tender was won a decade ago by the company El Pitazo names as Andritz Voit, the job never went ahead because Corpoelec did not pay, and he says Corpoelec owes the company US$46 million. That is a single source. It is also a picture of the whole problem, the fix sitting at the dam beside the machine it was bought for.

The backup is worse. Venezuela’s thermal plants, the ones that burn fuel, were producing about 2,800 MW in May and again in mid-September, under 15 percent of their installed capacity. By the end of September, the Corpoelec report El Pitazo cited put thermal output near 2,400 MW. More than 80 percent of thermal units were unavailable. The national system leans about 80 percent on hydropower, and an official inspecting Guri on September 12 said the dam alone generates “more than 70%” of the country’s electricity.

On paper, the country owns far more plant than it uses. El Pitazo put installed capacity at 36.5 GW. Venezuelanalysis put it near 30,000 MW in August and about 35,000 MW in September. The figures disagree, and even the lowest is nearly double the record peak. The plant exists. It does not run.

The earthquakes are part of the story, but not the start of it. A pair of earthquakes on June 24, which killed 5,278 people by the government’s count of July 21, damaged units at the Termocarabobo plant, and Corpoelec has since recovered 3 of the 4. But the thermal fleet was already more than 80 percent unavailable on May 6, seven weeks before the ground moved.

The oven nobody can turn off

A restaurant with one working oven can never shut it down to clean it. So it runs it hot, every night, until the night it quits. Then the kitchen stops.

That is the trap Aguilar describes. El Pitazo, laying out his account, rejects the heat and El Niño as the cause of Guri’s collapse in output. Aguilar blames accumulated wear on units that run around the clock without timely maintenance. With most thermal plants out, he told El Pitazo, the system has no backup “to stop a machine at Guri for maintenance without affecting the user, who already suffers from the deficit. So they keep over-demanding them until they break.”

A system run near full capacity with nothing in reserve has no slack to absorb a failure. Every plant running flat out is a plant that cannot be serviced, and every unit that breaks pushes more load onto the ones still spinning, which speeds up the next failure. The gap feeds itself.

This mechanism rests on one engineer, quoted by one outlet, reading company reports we cannot see. It fits every number above. It is still one man’s account.

The blackout ledger

When supply falls short, somebody decides who goes without. That is triage, and every triage has a rule. The table holds what the public record shows about Venezuela’s.

Where	What is reported	Source	Date
The interior, millions of people	4 to 10 hours a day without power	Corpoelec reports reviewed by El Pitazo	May 6
The capital region	”Suffers least”; sporadic cuts, mostly tied to specific faults	Same reports	May 6
Municipalities in Zulia, Táchira, Mérida, Barinas, Carabobo, Portuguesa, Guárico, Aragua	About 8 hours a day, sometimes in two blocks	El Pitazo	Sep 21
Rural Táchira and Mérida	More than 10 hours a day	El Pitazo	Sep 21
Caja Seca, south of Lake Maracaibo	Two cuts a day: 4 hours morning or midday, 4 at night	Merchant Maritza González, to El Pitazo	Sep 21
Caracas shopping malls	Asked to ration power in two 4-hour blocks, 12 to 4 pm and 6 to 10 pm, the eight most important commercial hours	Mall chamber Cavececo, on Unión Radio	Aug 19
Lara state, 53 communities	221 outages, 904 hours without power	Citizen observatory Activos por la Luz	Sep 29 to Oct 4

Maritza González sells goods in Caja Seca. The power goes twice a day. “Esto que vivimos es fuerte,” she told El Pitazo. What we are living through is hard.

El Pitazo reports that cuts used to average 5 hours. Users now report 6, 7, 8. Higher figures, 12 hours and more, come from protesters, from Zulia’s medical college, and from El Pitazo’s own October 1 report of cuts “up to 12 hours.” Venezuelanalysis says cuts can add up to 16 hours a day. Those are reported, not measured.

The scale of the cut is concrete. In May, according to company reports El Pitazo reviewed, Corpoelec was rationing up to 1,800 MW at peak hours, equal to everything Caracas, La Guaira and the Altos Mirandinos consume. In March the figure was 1,000 MW. Both numbers are months old now. Nothing newer has been published officially. El Pitazo’s September 30 read of Corpoelec reports puts the deficit near 2,800 MW.

Caracas is not spared. The government asked the capital’s malls to ration power in two 4-hour blocks, noon to 4 pm and 6 to 10 pm, which the mall chamber called the eight most important hours for commerce. Some malls report their backup generators failing after constant cuts.

The Lara log is the most useful line in the table, because ordinary people kept it. Between September 29 and October 4, 53 communities logged 904 hours without power across 221 outages. In the municipality of Iribarren alone, 38 sectors logged 155 outages. The article describes the observatory’s work only as daily monitoring, not its method. It is still more of a ledger than the state has offered.

A published shortfall would work as a public yardstick. Print a shortfall in megawatts and every citizen can measure each new plant, each repaired turbine and each week of cuts against it. Leave it unprinted and the person who sets the cuts is also the only person who knows how big the hole is. Whatever the reason behind the silence, its effect is the same: you cannot audit a number you are never shown.

What the record cannot carry is why the capital is cut least. The May reports say it is. They do not say who chose that, or on what basis, and this guide will not guess.

Two explanations, tested

Here is the government’s account, in order. On March 21 it launched a 45-day savings plan, citing the sun’s perpendicular rays and extreme heat. The plan ended May 5. The rationing did not. Reported on August 4: an “Electricity and Water Savings Plan” ahead of a “Super El Niño.” On August 11, announcing a demand record, the ministry pointed to high temperatures and to economic growth that “keeps its momentum.” On September 29, Minister Alcalá gave the fullest version: heat and a rebound in economic activity were driving demand up and widening the gap with available generation, and the June 24 earthquakes had delayed repair plans already under way. On September 30, Guard officers on a highway told protesters it was climate change.

Here is the engineer’s account: worn machines, no maintenance, no backup.

Test each one by what it predicts.

The government’s account predicts record demand. That holds. The records are official, and heat and growth raising demand is plausible, though no source we have puts a number on how much of either. The earthquake claim is partly borne out: the June 24 quakes did knock out units at Termocarabobo. But the quakes cannot explain a thermal fleet that was more than 80 percent unavailable on May 6, seven weeks earlier. And the government’s own minister now concedes a supply gap.

The engineer’s account predicts many units down and a gap that persists with water to spare. Seven of Guri’s 20 units were down in mid-September. The spillway opened in July. That fits, though it rests on reports we cannot see.

Both can hold on the same afternoon. Heat and growth push demand up while broken machines cap supply. The minister’s own word for the problem, gap, describes the supply side.

The weather risk is still coming. NOAA’s Climate Prediction Center, the US government’s forecasting office, has had an El Niño Advisory in place since at least July. Its September 10 discussion gave a greater than 90 percent chance of a very strong event this fall and winter, and a 75 percent chance that October through December 2026 brings a historic one, stronger than any since 1950. The key ocean reading was already 1.8°C above normal in August. The Morfema analyst treats El Niño as a risk ahead for the water, in 2027, not one already behind it.

Put the two clocks side by side. Rebuilding Unit 17 could take at least two years, on one engineer’s estimate. The strong El Niño is forecast for the next few months. The slow fix and the fast risk are running toward each other.

The fix runs slower than the hole

The US Department of Energy announced the biggest piece on September 2. GE Vernova “plans to bring 1 gigawatt of new reliable power online within the first 24 months, in addition to 5 gigawatts over the following four years.” Reports differ on the date the deal was signed, so this guide does not give one.

According to El Pitazo, US Energy Secretary Wright said the deal would begin to stabilize the grid in an estimated six months to a year, and that it includes bringing Guri assets back into service. Guri, he said, is producing half its design output because of worn turbines and networks. That is the same diagnosis as the engineer’s, from Washington.

Caracas has its own timeline. On September 29, Minister Alcalá promised 1,300 MW in the last quarter of 2026: 600 MW from Unit 6 at Planta Centro “in the coming days,” seven units with more than 500 MW at two Zulia plants (Termozulia and Rafael Urdaneta), and the rest from units elsewhere that he did not name. Two of the Zulia units, El Pitazo notes, had gone out of service on September 23. The minister had earlier said Planta Centro’s Unit 6 would start on October 1. Corpoelec sources told El Pitazo the work was still going. Corpoelec president Juan Fernández said “this year we have gained access to original spare parts” from GE Vernova, Siemens and Impsa. Venezuelanalysis notes that the US Treasury has issued sanctions waivers.

El Pitazo did the sum: if consumption holds, 1,300 MW covers about 46 percent of its estimated 2,800 MW deficit. Under half, and only if every unit arrives.

And some repairs outrun every promise on the table. Unit 17 could take at least two years. Alcalá says a company El Pitazo spells Eurobrush is working to reactivate Guri units 17, 18 and 20, with no dates given. The ministry’s own account of his remarks names Eurobras among its partners. Aguilar also warns, via El Pitazo, that GE Vernova may be limited at Guri, because the turbine designs there belong to other companies.

Bailing a boat with a cup works only if the holes stop. Every megawatt added gets measured against every unit that fails while the new plants are built.

What to watch

You do not need anyone’s permission to score this. Six things, each checkable:

1. Planta Centro. The minister’s October 1 date has passed. Corpoelec sources told El Pitazo on September 30 that the work was still going, and an October 3 El Pitazo report, relaying Carabobo’s governor, still described the unit’s entry as ahead. Does the 600 MW unit come online, and does anyone announce it with a date?
2. The missing number. Does any official release give a shortfall, or demand and supply for the same day?
3. Units 17 and 15. The minister has named a contractor for Unit 17. Watch for a restart date, and for Unit 15’s output.
4. The water. The next Guri level from a primary source, Corpoelec or the dam’s operator, instead of an analyst’s reconstruction.
5. The street. How many protests, in how many states. The September 30 count, per El Pitazo, was at least 20 in 7.
6. The forecast. NOAA’s October El Niño update.

Our call

Here is what we think happens, with the work shown, because a forecast you cannot check is just a mood.

We put it at 75 percent that through November 30, 2026, nobody in charge publishes or states an official national shortfall in megawatts, or national peak demand and available supply for the same date. “Nobody in charge” means the ministry, Corpoelec, the electricity minister, Corpoelec’s president, the acting president and the sectoral vice president for public works and services.

Start with the baseline. The September 19 release gave demand and no supply. El Pitazo reported on September 21 that neither the ministry nor Corpoelec reports the deficit. The call is that the pattern holds.

Then the pressure against it. Protests ran six nights straight by September 30. Officials now talk about the gap out loud: on September 29 the minister said it was widening, without sizing it. Demand records come from three voices at once, the ministry, a vice president and the acting president, and any of them can slip a supply figure into a televised address. We count that as about a one-in-four event over eight weeks. That is why the number is 75 and not 90.

The call is wrong if an official release, or an on-the-record statement by any of those six, gives a national deficit in megawatts, or states demand and supply in megawatts for the same stated date, between October 6 and November 30. It counts if it is carried by the official’s own channels or by El Pitazo, Efecto Cocuyo, El Diario, EFE, Reuters, AP or VTV. A leaked company report, like the one El Pitazo cited on September 30, does not count. If rationing ends before November 30, the call still resolves on this test. We will search the ministry’s site, Corpoelec’s official channels and those outlets, and log what we searched and when. A claim that something did not happen is only as good as the search behind it.

The lights are a separate question. This call scores whether anyone in charge tells you how dark it is.

The gap, the broken units and the cuts are sourced. The 75 percent is a call, and it goes on the scoreboard.

The bottom line

A full lake will not light a town by itself. It takes turbines that turn, backup plants to carry the load while they are fixed, and someone willing to say how much power is missing. Venezuela has the lake. It is short on turbines and backup. The third thing costs nothing but a press release.

In Caja Seca, Maritza González loses the lights twice a day, by her own account. She can count her own hours without power. The people in charge have not published the shortfall behind them.

---

Sources

• Record demand, 16,030 MW on Friday Sep 18 2026, “highest in nine years,” heat blamed, no supply figure (posted Sep 19, 1:04 am): Ministry of Electric Energy (MPPEE). Primary.
• Demand peak 15,625 MW at 2:44 pm on Aug 10 2026, highest in eight years; high temperatures and economic growth cited; no supply figure: MPPEE, Aug 11 2026. Primary.
• New peak of 16,424 MW reported by the ministry on Sep 22 (we could not find a ministry post of that date); “actual stable generation” of 12,000 MW (unattributed); installed capacity about 35,000 MW; Minister Alcalá’s Sep 29 address (1,300 MW in Q4, Planta Centro 600 MW, seven Zulia units / 500 MW, heat, economic rebound, earthquake delays to repairs); Juan Fernández on original spare parts; the outlet’s own note on US Treasury waivers; cuts of up to 16 hours: Venezuelanalysis, dateline Sep 30 2026. Secondary, quoting officials.
• Sectoral vice president Juan José Ramírez on 16,424 MW in “la semana pasada,” a “hito histórico” not recorded since 2017: MPPEE, Sep 26 2026. Primary.
• Alcalá’s Sep 29 VTV remarks as the ministry reported them, naming Eurobras, IMPSA, General Electric and Siemens as partners: MPPEE, Sep 30 2026. Primary.
• Acting president’s “more than 16,400 MW” since 2017 and “Superniño,” remarks reported Sep 25: El Periódico de la Energía. Secondary, quoting an official.
• Demand near 15,400 MW and generation near 12,650 MW (2,400 MW thermal) per a Corpoelec report; deficit near 2,800 MW and the 46 percent comparison (El Pitazo’s estimates); Alcalá’s “gap” remark, heat and economic rebound; Zulia units at Termozulia (3) and Rafael Urdaneta (4), TZ2 and TZ5 out Sep 23; Planta Centro’s earlier Oct 1 date and continuing work; Eurobrush (El Pitazo’s spelling) on Guri units 17, 18 and 20; thermal 80%+ out: El Pitazo, Sep 30 2026. Secondary; Corpoelec report unseen by us.
• Anonymous deficit estimate that may have topped 3,000 MW; “neither Corpoelec nor the Ministry report the deficit”; 2017 comparison; about 8 h/day in named municipalities, over 10 h in rural Táchira/Mérida, cuts “used to average 5 hours”; Maritza González quote; Unit 15 back near 550 MW; Termocarabobo 3 of 4 units recovered: El Pitazo, Sep 21 2026. Secondary; deficit estimate anonymous.
• Aug 10 supply “does not exceed 13,000 MW” (supply not clearly dated to Aug 10); Unit 17 failure Aug 29 and José Aguilar’s estimates (could take 2+ years, ~430 MW downstream); El Pitazo’s 1,200+ MW total; “warned months ago” allegation; Unit 15 stop Sep 10 and prior 580/730 MW; Andritz Voit and US$46 million (single source); official’s “more than 70%” on Sep 12; Wright’s six-to-twelve-month timeline and “half its design output,” via El Pitazo only; GE Vernova design limits at Guri (Aguilar): El Pitazo, Sep 14 2026. Secondary; Corpoelec reports unseen by us. The 2,625 MW gap is our subtraction.
• Guri fleet: 7 of 20 units unavailable, idle dates, 6,141 of 10,350 MW, 4,209 MW not produced; thermal ~2,800 MW; Aguilar’s maintenance claim and quote: El Pitazo, Sep 14-15 2026. Secondary; reports unseen by us.
• Reservoir level 269.11 m in July 2026, reconstructed from historical series and satellite altimetry, ~2 m below the ~271.2 m peak seen in September 2025; spillway relief from mid-July; El Niño framed as a 2027 risk: Morfema Press, analyst Nelson Hernández, Jul 12 2026. Secondary; no primary level located. The analyst writes that the level is “almost 13” meters above 2016; his own April 2016 low of 242.56 m gives about 26.5 m, and the article also gives a second 2016 low of 241.35 m. We use 242.56 m.
• Load shed up to 1,800 MW (Caracas, La Guaira, Altos Mirandinos equivalent); 1,000 MW in March; capital “suffers least”; interior 4 to 10 h/day; thermal ~2,800 MW (14%), 80%+ units unavailable; ~80% hydro; 36.5 GW installed; no official word on causes or duration; Mar 21 savings plan: El Pitazo, May 6 2026. Secondary, five months old.
• Installed capacity ~30,000 MW; Termocarabobo 600 MW lost, 300 MW restored; “Electricity and Water Savings Plan” ahead of a “Super El Niño,” Aug 4: Venezuelanalysis. Secondary. Installed-capacity figures differ between sources and between this outlet’s own articles.
• Earthquake toll, 5,278 dead by the official count of Jul 21: EcoAvant. Secondary, reporting the government count. Later, higher counts exist; we use the dated official figure.
• Caracas malls asked to ration power in two 4-hour blocks, 12 to 4 pm and 6 to 10 pm, “the eight most important hours” for commerce; generator failures (Cavececo’s Claudia Itriago, Unión Radio, Aug 19): El Diario. Secondary.
• Lara outage log, Sep 29 to Oct 4 (53 communities, 221 outages, 904 hours; Iribarren 38 sectors, 155 outages), Activos por la Luz: El Informador, Oct 5 2026. Secondary; observatory method unseen. Three municipalities reported outages without durations, so we do not compute an average. The article’s fluctuation counts are internally inconsistent, so we leave them out.
• Planta Centro’s Unit 6 entry still described as ahead, relaying Governor Rafael Lacava (“mejorará con la entrada de la unidad 6 de Planta Centro”): El Pitazo, Oct 3 2026. Secondary.
• Zulia medical college on 12-hour daily cuts: El Pitazo, Oct 5 2026. Secondary.
• Sep 30 protests (sixth consecutive day, at least 20 in 7 states, Guard officers’ climate-change remark on the highway near San Joaquín and San Mateo, tear gas in La Ensenada, La Cañada de Urdaneta, the protester’s words from videos on social media, cuts “up to 12 hours”): El Pitazo, Oct 1 2026. Secondary.
• El Niño discussion of Sep 10 2026, under an El Niño Advisory (90%+ very strong; 75% historic in Oct-Dec; Niño-3.4 +1.8°C in August; next update Oct 8): NOAA Climate Prediction Center. Primary.
• GE Vernova pledge, 1 GW in 24 months plus 5 GW over four years: US Department of Energy fact sheet, Sep 2 2026. Primary. The DOE says Wright traveled to oversee the signing on Sep 2; El Pitazo gives Sep 4. Neither date is stated in the text.
• Theory of Constraints: Eliyahu M. Goldratt, The Goal (1984).
• What we could not verify, and left out: any internet-outage (IODA) measure of the blackouts; a primary Guri reservoir level, including a reported Aug 31 reading; Venezuelanalysis’s claim that Guri produces over 10,000 MW, which conflicts with the 6,141 MW above; US Energy Department supply and demand estimates relayed only by a secondary outlet; corruption and “rationing for the refineries” claims; a ministry post dated Sep 22 for the 16,424 MW peak; the raw Corpoelec load-shed reports and the Activos por la Luz dataset.