What does the temperature of rivers have to do with operating nuclear power plants? Does it make them more likely to have problems?
Actually the river temperature problem is not specific to nuclear power plants. Fossil-fueled plants have exactly the same problem - they reject heat to water, the water is rejected to rivers, and life in the rivers can be harmed by high local or average temperatures caused by discharges from the plant.
The amount of heat they have to reject varies directly with their power output - twice the power, twice the heat to dump. With the high temperatures in Europe now, the power output demand is high, so the plants are dumping as much heat as they can.
Also, the more heat they dump, the greater the difference in the temperature (delta T) between the water coming in and the water going out. A given temperature coming in implies a particular temperature going out if the plant's power output is the same.
But now the cooling water sources are getting hotter even as the delta T's are at their peaks with the high power output. The result - the temperature of the water going back to the environment is higher than normal. And in the US there are regulations about how high that temperature can go legally. Presumably the Europeans have similar requirements.
I don't know if the temperature specs are appropriate or not. There is no question that too much heat is bad for rivers and lakes though. And if the high heat is combined with lowered river flow or lake levels, then the peak temperatures in the wild resulting from plant operation get even worse yet. So whether a specific regulation is appropriate or not, the idea is that at some point things get out of hand. (The flip side of this is that the fishing can be good all year at rivers and lakes downstream from power plant discharges).
So what can be done when discharge temperatures get too high? Assuming that the plants must operate within the regulations, then the only thing they can control is the delta T - how many degrees the cooling water is heated flowing across the plant. And they do that by reducing the power output at the plant. Which means less power for the consumers.
You might wonder why power plants must reject heat at all. Good question. Thermodynamics geniuses long ago noted that just as you can't turn a cow completely into steak, you couldn't turn heat completely into mechanical (and ultimately electrical) energy. You always wind up having to throw out some of it. Actually, most of it - well over half at contemporary fossil power plants, and more yet at nukes. It all depends on the plant's operating temperatures - the higher it is, the more efficiently the plant converts heat into power. (uh, yes, the above is somewhat simplified)
To get rid of the rejected heat, power plants use a number of big heat exchangers. The biggest by far will be the main condenser. To give you an idea of the size, the main condenser at Clinton Power Station (CPS) in Clinton, IL, is supplied with raw cooling water from Lake Clinton by two 10' pipes, and I believe the flow rate is about 180,000 gallons per minute per pipe (I'm working from a 15 year old memory, so I could be significantly off on the flow rate)(yes, I'm a PE who once worked at the plant).
This water doesn't go anywhere near the reactors. That water is in a separate loop flowing through the tubes of the condenser at a boiling water reactor like CPS, and water that seconds before was in the reactor is condensed on the other side. At a PWR like the French plants, the reactor water doesn't leave the containment building, so the cooling water is even more isolated from the reactor water. But even at the BWRs, the reactor water side is at a lower pressure (actually, a vacuum), so any leaks are from the cooling water into the plant and not vice versa. Given that such leaks increase the quantity of radwaste and water treatment volume, which are expensive to process, management deals with them rapidly by finding and plugging leaking tubes.
Whoops, what if there's a major leak - will they flood the insides of the plant? CPS had redundant alarms designed to detect flooding bright and early so the pumps sending the water into the plant could be shut off. This would be accompanied by low turbine vacuum alarms, high flow rates through the cooling water system, more current draw on the condenser cooling water pumps, and possibly high turbine vibration alarms, so operations people would know something was amiss rapidly. The plant has a very large volume, so there is plenty of time to react to problems. (as for natural flooding, that's addressed during siting).
But if you shut off this condenser cooling water, how do you keep the plant cool? No problem - the operators would trip the reactor, and entirely separate systems for keeping the reactor cool under emergency conditions would be used to bring the reactor to a cold shutdown. The systems that do this are redundant, are not subject to common failures, are served by redundant power supplies, are located inside very heavy buildings with reinforced concrete walls at least a foot thick between them and anything that might harm them, and are designed to withstand earthquakes too. And the QA is very strict.
What if the operators don't know what to do? They do. They get a tremendous amount of training, including work as teams on simulators built to mimic the plants down to tiny details. Staffing levels are maintained at no less thanand working hours are limited to maintain alertness.
What if the systems aren't ready to go when needed? They'd better be in service ready to go as required by their Technical Specifications, or else we're talking fines, shutdowns or worse. Verification includes periodic "surveillance tests", such as testing onsite diesel generators to assure that they start up, reach operating speed and voltage and pick up load in the proper sequence.
How do you know the systems perform as designed? Because they were tested to design requirements as a condition of getting an operating license. At CPS that happened around 1987. From that point onward configuration control is tight, and proposed changes are analyzed at great length before implementation.
I could go on with the questions and answers - there are several shelf feet of books on file for every operating reactor in the US devoted to stuff like this. Since 9/11 they may be less accessible, but you can probably see them for a plant near you if you are interested.
Even the most trivial events get tons of bad publicity for nuclear power plants. Nuclear power executives know that the more they goof the more they get inspected, and they know that shutdowns are incredibly expensive, so they know to keep their noses clean. And they're awfully picky about the personnel that they permit to operate multibillion dollar investments.
So the bottom line is that the lead of the story, speaking as it did of "overheating at nuclear power plants", was not about nuclear power safety as it might have appeared. The plants are as safe as they always have been.