Say you’ve left a car sitting all summer. You finally get it running, take it out on the highway, and three days later something that has nothing to do with the engine quits on you. Meanwhile, the identical car next door, which couldn’t start while yours was running because the two of you share one fuel line, fires up and gets back on the road before you do.<\/p>\n
That’s pretty much the week Romania’s only nuclear plant just had.<\/p>\n
In October, I wrote about two Danube reactors that went dark because the river fell below their cooling pumps<\/a>. Both units at Cernavod\u0103 had been down since summer, Unit 1 since July 28 and Unit 2 since Aug. 13. They’re Canadian-designed CANDU-6 reactors rated at about 705 megawatts each, or roughly 650 megawatts once you take out what the plant needs to run itself, according to World Nuclear News.<\/p>\n Nuclearelectrica, the company that runs the plant, started bringing Unit 1 back on Nov. 4. That was day 99 of its shutdown. The reactor got back on the grid on the afternoon of Nov. 6 (day 101, if you’re counting) and hit full power at 10:07 a.m. on Nov. 8. Nuclearelectrica hasn’t given an exact time for that Nov. 6 synchronization, only that it happened in the afternoon.<\/p>\n Because they drink from the same glass, basically. Both units pull their cooling water from one intake basin at the head of the Danube\u2013Black Sea Canal, and the pumps that feed them sit at a fixed height. A running reactor pulls so much water out of that basin that the level drops, and if it drops too far, the pumps can’t get enough water and start sucking in air instead.<\/p>\n Nuclearelectrica spelled that out on Nov. 4, warning in its restart announcement<\/a> that running Unit 1 “will lead to a drop in the water level in the intake basin.” So Unit 2 had to wait for the Danube to rise and stay up.<\/p>\n How much higher did it need to go? On Nov. 7, plant director Romeo Urjan said the local reading was about 6.8 feet, which was fine for Unit 1, while Unit 2 needed another 4 to 6 inches on top of that, according to the Romanian news agency AGERPRES via Energynomics.<\/p>\n That sounds like a rounding error for a river the size of the Danube. For a pump that can’t move up or down, it apparently isn’t.<\/p>\n The river came through. Unit 2’s startup sequence began at 12:42 a.m. on Nov. 9, according to a Romanian Energy Ministry statement cited by Energynomics, and Nuclearelectrica said the water level it needed had been reached.<\/p>\n Unit 1 dropped off the grid on the evening of Nov. 9. Capital, a Romanian business paper, put it at 6 p.m. local time (11 a.m. on the East Coast), and citing data from grid operator Transelectrica, it pinned the problem on the vacuum control system for the turbine’s condenser.<\/p>\n If you’ve never had to think about one, a condenser takes the steam that just spun the turbine and turns it back into water by running it past pipes full of cold water. Doing that also keeps a vacuum on the turbine’s exhaust side, and without the vacuum the turbine can’t keep going, so the whole unit has to come off the grid.<\/p>\n Nuclearelectrica’s own wording is vaguer. Its Nov. 9 statement<\/a> says the fault was in the conventional part of the plant, meaning the non-nuclear side, and that it had no impact on nuclear safety, personnel or the environment. The company doesn’t name the component, and it says the root cause will be laid out in a later report.<\/p>\n So after three months of waiting on a river, Unit 1 lasted three days and got knocked out by equipment that isn’t even part of the reactor. That’s frankly brutal timing.<\/p>\n That same Nov. 9 statement said the plant’s technical director will be replaced on Dec. 1, and that all maintenance work will be reviewed against the plant’s procedures and international standards. On Nov. 12, Nuclearelectrica named Bogdan Deaconu, an engineer trained at Cernavod\u0103, for the job.<\/p>\n While Unit 1 was in the shop, Unit 2 synchronized with the grid on Nov. 11, and Romanian outlets reported it at full power by Nov. 12. That’s the reactor that couldn’t start a week earlier because its sibling was drawing down the water.<\/p>\n Unit 1’s comeback, meanwhile, kept slipping. Nuclearelectrica’s first estimate gave the repair two to four days. On the morning of Nov. 11, it said Unit 1 would reconnect on Nov. 12. Late that afternoon, it pushed that to Nov. 13, saying the repair work needed more time. This time the date held.<\/p>\n The reactor reconnected on Nov. 13 and reached full power on Nov. 14, in line with the schedule Nuclearelectrica announced on Nov. 11.<\/p>\n If you’re keeping score, that’s roughly four days from disconnection to reconnection, right around the top end of the original estimate. Not great, but not a disaster either.<\/p>\n Mostly, though the summer wasn’t cheap. Romanian Prime Minister Luca Niculescu put the cost of shutting both units down at \u20ac200 million, or about $221 million at the European Central Bank’s Nov. 13 reference rate, according to World Nuclear News<\/a>. He also called a project known as Bala 2, which is meant to get more water to the plant, a priority.<\/p>\n I’d argue that’s the fix that matters more in the long run. You can repair a condenser system in a few days, but you can’t do much about a river that drops below your pumps in a dry summer. Romania has already tried sinking barges in the Danube to steer more water toward the plant<\/a>, and as we covered in October, the gain it bought faded within days.<\/p>\n Nuclearelectrica put Unit 1 at full power on Nov. 14, and with both units running, the company says the two Cernavod\u0103 reactors together will be covering about 20% of what Romania needs.<\/p>\n","protected":false},"excerpt":{"rendered":" Say you’ve left a car sitting all summer. You finally get it running, take it out on the highway, and … <\/p>\nSo why couldn’t both reactors start at once?<\/h2>\n
So what knocked Unit 1 back off the grid?<\/h2>\n
Unit 2 beat it back<\/h2>\n
So is the plant back to normal?<\/h2>\n