Redundancy Concept and ECCS in NPP

[kollier]

Hi Folks
I'm studying about redundancy concept in nuclear power plants.
I want to find a list of all nuclear power plants in which the number of safety(eccs) trains are mentioned.

 

[theCandyman]

For my senior design project, I was in charge of overseeing the safety of our reactor design. Are you only interested in ECCS?

The reactor we designed was a sodium cooled LMFBR. A system system could be considered implemented as part of the design (passive - such as having a large pool of coolant to prevent boiling off all the coolant and exposing the core) or an extra system that is implemented (active - like having a backup coolant pump in case of losing flow).

 

[kollier]

Although now I'm gathering case information about redundancy concept with focus on PWRs and I need information about ECCS systems in PWR plants but any other information about redundancy in other plants will be useful.
I found http://mit.edu/canes/publications/book.html" very useful and I recommend it.By the way it's information about the number of ECCS trains in different PWR plants is not enough.
I tried to find a website in which these information brought but I fail!

 

[gmax137]

For the commercial nuclear plants in the US the answer is "they all have two trains of ECCS." There may be a very few exceptions to this (very old plants - but most of them had redundant ECCS systems installed, or they have been decommissioned). The US NRC General Design Criteria (in 10CFR50, Appendix A) require a redundant ECCS that meets the single failure criterion. A two train system meets this requirement. If you are interested in the actuation systems the answer is more complicated, since a design goal was to provide a single-failure-proof assurance that spurious actuation signals would not be generated. I cannot answer your question for plants outside the US or for plants not yet licensed in the US.

 

[kollier]

Thanx gmax
In the book I have mentioned it is written that for two loop plants in the US there are 2 or 3 trains of eccs systems.
Although they're old where can I find their information?

 

[gmax137]

I'm not sure you & I mean the same thing by "train." Please tell me what you mean by this. I can't think of any plant in the US that has three trains. Also, when you say "two loop" do you mean two steam generators? While it is true that the Westinghouse "two loop" units are the old ones, the CE designs always used two loops - they just built bigger & bigger SGs (including the 3800 MW "system 80" plants, Palo Verde in Arizona). I just looked at the link to the book you mentioned - I may be confused because it looks like that book considers plant designs that were never built, whereas I'm limiting my comments to actual units (those built & operating).

I think some of the really early units had less than two trains, as originally designed. Units such as Yankee-Rowe, San Onofre Unit 1, etc. If you are in school you may be able to find the FSARs for some old units. They may be available in the NRC PublicDocument Room but they will be hard to find.

I'm not sure why you're looking for this info, but there are a few books that discuss the history of the regulations and the plant designs. "Containing the Atom" by Sam Walker is an excellent read, so is "Nuclear Reactor Safety" by David Okrent (though that one is hard to find).

 

[Astronuc]

There is some discussion about ECCS here in this article - http://www.npjonline.com/NPJMain.nsf/504ca249c786e20f85256284006da7ab/c370a4ae38717df686256af5007abdff?OpenDocument

I believe most plants have to independent (redundant) ECCS systems, although perhaps some of the 3-loopers have three.

There is another text on NP safety, Nuclear Power Reactor Safety (1977) by E. E. Lewis. I don't know about the availability though. Okrent's book, mentioned by gmax137, is the other one that comes to mind.

The individual NPP ECCS information is available in the plants/units UFSAR. Some of the detail used to be available on-line at the NRC, but most or all was removed after Sept 2001 over concerns that saboteurs could use the information.

 

[kollier]

Yes I mean two steam generators by 2loop.
and by eccs train I mean high head safety injection system(HHSI) and low head safety injection sysytem(LHSI).I want to know how many hhsi and lhsi systemes they use.
I found that number of accumulators are equal to the number of loops except german design that use more redundant systems(for example 8 accumulators for KONVOI design).
And I know that in french designs they use 2 ECCS trains.
I know also that in recent generation designs (EPR, APWR and...) they use four trains of safety system(i.e. N+3: one systems in not working due to maintenace one fail to start and one fails after starting therefor always one sysytem is available)
And I know that 2 train of ECCS fulfills the CFR criteria.
In that book(page 96, 1st edition)) it is written that 2loop (and 3loop) Westinghouse disgns can have 2 or 3 ECCS trains and 2 or 3 redundant electrical trains.
I have found some FSARs of old plants but they are 4loop plants.

 

[kollier]

I know also German designers use four 50% eccs trains( except accumulators that are 8) (i.e. N+2 ,N=2)

 

[chayced]

South Texas Project has three trains. (with four loops) It was designed around having the ability to easily take one out for maintenance at power. Also adds a nice extra margin of redundancy.

Don't know offhand of any others like that.

Navy Carrier PWRs have a very interesting kind of ECCS, but I'm not sure if I can talk about them. Think I was supposed to forget everything when I left the service. :(

 

[gmax137]

There really isn't much connection between the number of SGs (ie, "loops") and the number of ECCS trains. In the evaluation of a two-train, four-loop PWR, you would postulate rupture of one loop, followed by a failure of one train. The flow delivered to the ruptured loop is "spilled" to the floor, and you are left with 3/4 of the flow from one trian delivered to the RCS & core. The flow split between the four loops is essentially equal due to the presence of high pressure drop valves in each of the four ECCS lines.

EDIT: Wait a minute - in the Combustion Engineering design, each of the two high head pumps is piped to each of the four cold legs; whereas each of the two low pressure pumps is piped to two cold legs. So for a cold leg break (and assuming one "train" of pumps fail), you lose (spill) 1/4 of the HPSI flow and 1/2 of the LPSI flow. Is this what the OP was asking about? The piping arrangement? I had been thinking only of the number of redundant pumps...

The mention of the three train ECCS at south texas (STP) is intriguing: Does the STP license allow them to run with one of the three trains out of service indefinitely? If so, it would "really" be a two train system as far as the ECCS evaluation is concerned.

 

[chayced]

Yeah, for ECCS evaluation it is a two train system with an extra train.