It was not possible for a worm to take out River Bend Nuclear Generating Station. Though Entergy sold their soul to Redmond years ago, very few actual control systems were touched by M$. All important systems were under System Engineers, most of who learned to loath M$ for it's cost and poor performance. Operators like manual control and fancy controls are kind of silly in a system that's designed to run 100%. They are very adverse to having any kind of equipment that can screw up, pneumatic, electronic or otherwise, and if it could screw up, they had another one waiting to be valved in. Microsoft only made life miserable doing communications, documentation, trending and other administrative work. The system engineers in charge of a new data server got lots of late calls, but that server only served to provide redundant information storage and had no control functions.
I knew of one NT workstation at River Bend that actually touched plant equipment, the Hydrogen Water Chemistry system from GE. The actual control was done by PLCs in a "genius block". The workstation talked display system information and set a few parameters by some nasty VB interface that did not do everything it was supposed to do and had the word "testing" in the title bar. The computer used CF for disks and did well. It was not networked. The whole system could be shut down without ill effect to the rest of the plant and it had many non-computerized fail-safes and manual overrides. A DOS system was used as an improved reactor coolant flow meter It was self contained, though it had a modem the vendor called to set things! It was a redundant component, I don't think it had any control functions and it was being replaced. I can't speak for the fossil fuel world, where things are cycled daily and there are fewer restrictions on modifications. Hell, River Bend still had components with iron core memory in them. Who knows, M$ might have filtered down to the skid level in some fossil plants. Upper management was in love with M$ and spent all sorts of money on it. Never underestimate the power of one big dog. One problem is that Federal regulations require a certain offsite generating capacity for a nuke to run. In a gird failure, plants with nothing wrong with them might be forced offline. It then takes time, sometimes days, to crank the pig back up. It's not a trashcan. The whole outage seems like a grid problem not a generation problem. If the loss of a plant or two pops your whole grid, the grid is not adequate. Plants can be expected to go out for odd reasons. Lightening strikes take out critical equipment or otherwise flip shutdown circuits, animals crawl into places you don't expect them, shit happens and plants are forced offline. When River Bend goes down, you don't notice. A few more power plants in the right place might help. On 2003.08.15 16:39 Shannon Roddy wrote: > Leave it to Micro$oft to fry the grid.... > > I would not be the least bit surprised if the below turns out to be > true. In addition, I bet the power companies will shy away from M$ in > the future even if it is not true. > > Additionally, I have done some reading in the past about SCADA systems, > and it is amazing how much stuff we depend on in our daily lives is > controlled by M$ boxes. > > Shannon >
