BWR Feedwater Check Valve Integrity 1

[pwhskch]

Some BWR's seem to have problems with check, or non-return valves in their feedwater circuits. These valves either stick open or leak water when they should be closed.

What methods are used to ensure satisfactory operation of these valves?. What sort of leak rates are acceptable? Should these valves be damped so that any tendency towards water hammer is avoided?

 

[vpl]

pwhskch

I can give you some information regarding US requirements, if you want. Patricia Lougheed

 

[pwhskch]

Yes please, if you have any more information.
I recently took on a new job with a Nuclear Inspectorate (Switzerland). The plant with which I'm involved is GE BWR-6 built in 1984. There are moves to modify the (inclined) check valves so as to reduce leakage and reduce risk of stiction. Thks Pete

 

[vpl]

The requirement to have the feedwater check valves in BWRs relatively "leak free" comes from 10 CFR Part 50, Appendix A, "General Design Criteria", Criterion 57 "Closed System Isolation Valves" that says that any line that penetrates containment, but doesn't directly connect to the containment atmosphere and isn't part of the reactor coolant pressure boundary has to have at least one containment isolation valve (CIV). Criterion 57 says the CIV can't be a simple check valve, however, many US plants have an excemption to this for the feedwater system.

The requirement to test CIVs is found in 10 CFR Part 50, Appendix J "Primary Reactor Containment Leakage Testing", which specifies that the total leakage from all lines penetrating containment be less than or equal to 0.6La, where La is specified in the Technical Specifications Bases as the acceptable amount of leakage to the environment, post accident for 24 hours at peak containment pressure. (Actual number for La varies considerably by plant - from 0.5% at 56.5 psig to 2% at 45.7 psig to 0.65% at 9 psig)

Appendix J requires that valves be tested with air at the containment pressure. Feedwater checks tend to be very large (24&quot valves that don't do well under air testing. Some US plants have installed keep fill systems so that they don't have to do the testing. Others have found that very exact machining and tolerances help keep the valves in spec.

I don't have any current information about leakage rates for individual valves - you might try posting this in the valve forum and see whether there's anyone who knows. Also, I've never seen or heard about dampening systems being installed to prevent water hammer. That's not been a concern on US BWRs as far as I know.

Hope this is helpful.

Patricia Lougheed

 

[pwhskch]

Thks for your input. Maybe water hammer isn't quite the correct terminology. Hydraulic shock or pressure surge is closer. Damping the valves would tend to reduce pressure surges. From drawings it is apparent that the valves have an internal piston/cylinder arrangement.

Pete

 

[pwhskch]

By the way, what's a "keep fill system"?
Pete

 

[vpl]

A keep fill system is just that - a system to keep the piping filled with water. It generally consists of a water source, a pump, and piping. Since I only inspect these things (and not design them), you might want to ask some questions on the piping & fluid mechnanics forum if you need design type info. Patricia Lougheed

 

[vlvguy]

One of the best check valve designs for nuclear feedwater systems is the NozzleCheck. This design eliminates pressure surge, roughly 10x less pressure rise than the same size swing check. They also use very strong springs that still allow the valve to be fully open at the normal velocity, about 20-25 fps. The advantage of the strong springs is the large seat load applied during low pressure seat leakage testing, ie LLRT. I believe there are over 300 installed in nuclear feedwater systems, mostly at EdF plants. I know Fitzpatrick Nuclear Station in the US has them and they have eliminated past leakage and wear problems encountered with the originally installed swing checks.