Detailed control room schematics?

swbrown

I'm toying around with the idea of writing a simulator, ideally for a less automated and relatively operator workload-heavy reactor like the pre-Chernobyl RBMK design (as watching a machine run itself is no fun), and I'm trying to find information on the function of an actual control room's various buttons, switches, and displays as I want to focus on a tactile simulation of control room operations from an operator's perspective rather than modelling reactions in the core as existing simulators do.

I've had very little luck finding that information, even finding photos with a high enough resolution to read the markings on dials is difficult. I've run into the IAEA's documentation on simulator design which is very useful for modelling the core, but nothing about what is essentially the user interface for the operators.

Does anyone know of a source for that kind of detail for the old RBMK design or even any other type of reactor if that's not available? One option would be to tour Kursk's NPP as the training room they apparently let people poke around in is an analogue of the old-style RBMK control rooms but I'd still be lacking the manuals to make sense of it all.

gmax137

Try searching for info on 'full scope simulator.' that's what we call the simulator that looks just like the control room (all of the panels, switches, indicators, etc. are there just as they are in the real control room). The simulator software models all of the plant (not just the core) and it drives the panels, switches, indicators, etc.

Every plant in the US has such a simulator that mimics that plant's control room. Developing the software for these simulators was a huge job. There were several vendors, including simulator companies (eg, Singer), as well as the reactor vendors who had expertise in modeling the power plant systems and designing the actual control rooms.

Doing this on your own is a daunting task, and finding information on the control room panels is the least of it. You're going to need detailed descriptions of the systems in order to develop software to drive your instruments and so on. Maybe you should concentrate on a limited area first. For example you could begin by building an RCS model and see if you get the flowrates and temperatures right. Then add a pressurizer model and see if you can make that work (trust me, it is not simple). Good luck.

jim hardy

a real control room simulator is thousands of I/O points and scores of man-years.

Years ago i knew some enthusiasts who wrote a pretty nice little simulation of a CANDU power plant. It ran on a TI-99 and used every single byte of memory.
They were engineers at a simulator factory so had a big head start.

i'd say start at nucleartourist.com

and my Mom used to take a magazine called "Soviet Life"
the issue a month or two before Chernobyl had a feature article on the RMBK plant, fwiw

swbrown

It would be a much easier task in my case as I'm only interested in simulating the feel - the model only needs to be as accurate as is necessary to support the workflow, and where things are too complex or time consuming to model they can be fudged. It'd certainly not be useful as a training simulator, but as the intended audience would be people interested in simulators as games, it would be fine. The problem is, there is a lot of available information on accurately simulating a reactor, at least at a high level, but (as far as I've found) a complete lack of information on user interface and workflow of a specific plant design. That situation is the opposite of what I'd need.

Ideally, I'd be able to find documentation on the layout and function of the controls of an interesting to simulate existing or historical plant along with documentation on operator workflow for various tasks (startup, startup with new fuel, refueling, optimizing power distribution, etc.) and then spend a couple years implementing enough of it to give someone the impression that they're operating an analogue of the real thing.

gmax137

You lost me there. What do you mean by 'feel?'

Borek

I can only guess - but think of early car race game (like Test Drive III) - it had a steering wheel, it looked like driving a car, but it was not a real simulator, just an arcade game.

swbrown

Putting the user in a position where they can easily imagine that they're a human physically controlling the real thing. Think of what I'm trying to build as like a flight simulator game for a NPP. What matters most is the authenticity of that human-machine interface, not how the machine actually functions under the hood.

a.ua.

http://forum.pripyat.com/showthread.php?t=416&page=15
http://tryapichka.narod.ru/viur.jpg
http://tryapichka.narod.ru/oblozhka.jpg
http://tryapichka.narod.ru/new_sim.jpg
http://tryapichka.narod.ru/perv_zagr_react.jpg
http://tryapichka.narod.ru/viur.jpg
http://tryapichka.narod.ru/viub_viut.jpg
http://tryapichka.narod.ru/inform_schit.jpg
http://tryapichka.narod.ru/peregruz_react.jpg