Currently a megaproject called ITER (International Thermonuclear Experimental Reactor) is being built in France. The device would use magnetic containment of plasma reacting at temperatures of over 100 million Kelvin. What would happen if the magnetic field containing this reaction were to fail? Realistically?
Realistically, what you have is a very small amount of superheated hydrogen gas escaping in the environment.
Other than ruining a very expensive piece of machinery, and a loud noise, just about nothing will happen.
Fusion reactions cannot take place under natural conditions on Earth, so the reaction will stop as soon as containment is breached. The contents of the reaction vessel is a few grams of superheated heavy hydrogen, which will do what hydrogen does; a part of it will burn producing a tiny amount of heavy water, the rest will raise up in the atmosphere and eventually be lost in outer space.
(For comparison, a cubic meter of ordinary sea water contains about 300 grams of heavy water.)
Then they will have to find the funds to repair their expensive machine.
ITER themselves have the following to say about this topic:
Although 100 million degrees Celsius is an extremely high temperature, the density of the plasma (atoms per cubic metre) is very low—about one million times less than air—and the total energy in the plasma is not very great. The very rapid release of the energy could cause superficial damage to some plasma-facing components (i.e., surface melting) but would not be sufficient to produce structural damage.
Expanding my comment out into an answer: Temperature is not heat.
Imagine you have a boiling pot of water on the stove. You can reach through the steam to turn off the element. Steam has a high temperature (100C), but it doesn't have much heat, because there's not much of it. You could also stick your hand in the boiling water, but I wouldn't recommend it; boiling water is also 100C, but there's a lot more mass, so the amount of heat is much greater. (Steam gets dangerous under pressure; the amount of steam in a given place increases and so the heat increases as well.)
Here we have 10g of matter at 100M degrees. A calorie is the amount of energy needed to raise 1g of water 1 degree, so there are 1 billion calories of energy in this sample. (EDIT: as noted, the actual number is higher; Hydrogen at room temperature carries at least 3.5 times more heat than water of the same mass, and things get screwy at 100M degrees.)
A bathtub has about a 300L capacity. Conveniently, that's also 300kg of water. That 1 billion calories of energy is enough to bring 33 bathtubs from freezing to boiling. Which is impressive, but 33 bathtubs isn't that many.
An olympic-sized swimming pool has 2,500,000L, or 2,500,000kg of water. That 1 billion calories of energy is enough to raise the temperature by 0.4 degrees.
So, how much danger is there? I wouldn't want to stand next to it if it broke, but it's unlikely to hurt anyone outside the actual building. A heated swimming pool is going to use more energy.
There would be no ecological impact; nothing would get out of the building.
Even in terms of thermal pollution your car has more effect.
The ecological impact of having excavated such a huge site and hauling in the 30,000 tons of concrete that went into it in the first place, were, however, massive.