Passive mechanical safety: thermal sense rods.

[Dmytry]

So, I had this idea for passive safety. Break-apart rods (zirconium with low-melting point metal inserts for example) which go through reactor core at regular intervals, are tensioned by springs, and in the event of overheating, are torn apart, operating simple (springs, latches) mechanisms to ensure insertion of control rods (if not already inserted) and operation of pressure relief and coolant flow valves (relevant to passively cooled reactors), as well as some reliable outside indication of the overheating event.

Do any reactors use something of this sort? I know molten salt reactors can use a frozen plug as passive overheating failsafe. I know that various electrical products sometimes use connections that de-solder and disconnect in the event of overheating.

 

[rmattila]

The SECURE district heating reactor, planned between 1975-1985 as a joint Finnish-Swedish project (but never built due to the Chernobyl accident) had a venturi-type self-shutdown mechanism based on thermal locks and pressure differences. See http://www.ats-ydintekniikka.fi/lehtiarkisto/ATS_lehti_2011_1.pdf , figure on page 25, panel "D". The shutdown is based on the borated (blue) water entering the reactor due to break-down of the bottom thermal lock keeping the borated water out of the reactor as long as the primary water doesn't boil in the venturi.

 

[NUCENG]

So, I had this idea for passive safety. Break-apart rods (zirconium with low-melting point metal inserts for example) which go through reactor core at regular intervals, are tensioned by springs, and in the event of overheating, are torn apart, operating simple (springs, latches) mechanisms to ensure insertion of control rods (if not already inserted) and operation of pressure relief and coolant flow valves (relevant to passively cooled reactors), as well as some reliable outside indication of the overheating event.

Do any reactors use something of this sort? I know molten salt reactors can use a frozen plug as passive overheating failsafe. I know that various electrical products sometimes use connections that de-solder and disconnect in the event of overheating.

Would that require opening the reactor to replace the fusable links following a normal scram? If so that would be a drawback.

Just for clarity, PWR rods are raised by a CRDM. When unlatched (scram), they are pushed into the core by spring pressure and gravity. BWR rods use hydraukic pressure to withdraw rods. When unlatched and the scram headers are opened the rods are driven into the core. If a rod comes unlatched without scram pressure it will drift into the core due to the normal pressure difference.

 

[Dmytry]

The idea is that the latches can be opened either by outside SCRAM signal or by fusible link failure (thanks for the word btw, didn't know it's English name), so that in the event of SCRAM the fusible links stay intact. The idea is that fusible links would never fail in normal operation, but only in events like Fukushima when the water level ran too low, external electronics is malfunctioning, and whoever is in charge is also malfunctioning delaying any actions that would damage the reactor but keep fuel intact. The fusible links may be torsional.

Primary function would be to operate valves though. I was actually thinking of a reactor that'd be located below nearest big water source, or a big pool of water, and would have emergency coolant pipe (with backflow valve that's normally disabled) connected to that source. The fusible links would enable backflow valve and vent the reactor, then when pressure equalizes the water will open the backflow valve and flow in by gravity.

 

[rmattila]

Are you aware of the isolation condenser /core flooder design of the SWR1000 / Kerena ( http://www.areva.com/mediatheque/liblocal/docs/pdf/activites/reacteurs-services/reacteurs/pdf-plaq-kerena-va.pdf , page 5 )? There's one way to passively initiate core flooding, if the water level drops close to the core, i.e. before overheating takes place.

 

[Dmytry]

Are you aware of the isolation condenser /core flooder design of the SWR1000 / Kerena ( http://www.areva.com/mediatheque/liblocal/docs/pdf/activites/reacteurs-services/reacteurs/pdf-plaq-kerena-va.pdf , page 5 )? There's one way to passively initiate core flooding, if the water level drops close to the core, i.e. before overheating takes place.

Hmm, that is quite clever. I presume during normal operation the circulation of water through condenser is not much of a heat loss issue.

What would it do if due to some failure (pressure relief valve stuck open, pipe damage, etc) the water will keep getting lost from the reactor?
edit: nevermind, I see some flooding system with check valve. The passive pressure pulse transmitter is very clever. Just to make sure I'm getting this right: when the water flows out of the pressure pulse transmitter, the steam starts condensing there, heating up liquid in it and activating de-pressurisation valves and such? What's about presence of gasses (which would interfere with heat pipe effect), are radiolysis products a non-issue?

 

[rmattila]

Hmm, that is quite clever. I presume during normal operation the circulation of water through condenser is not much of a heat loss issue.

There's a purpose-built anti-circulation loop in the return line to prevent natural circulation as long as the core water level is high.

What would it do if due to some failure (pressure relief valve stuck open, pipe damage, etc) the water will keep getting lost from the reactor?

The core flooder systems etc. are described in that document: http://www.areva.com/mediatheque/liblocal/docs/pdf/activites/reacteurs-services/reacteurs/pdf-plaq-kerena-02-va.pdf , pages 30-39.

After the reactor depressurization, the water in the isolation condenser shell side will flow to the reactor by gravity.

 

[Dmytry]

That indeed does look very safe.

Why the first BWRs were built with so little safety, relying on pumps to remain operational to prevent an accident? The modern reactors look cheaper than the first generation ones. I know simplicity is deceiving but the first BWR don't even look easier to engineer, just much more inelegant and enormously less safe.

 

[rmattila]

That indeed does look very safe.

Why the first BWRs were built with so little safety, relying on pumps to remain operational to prevent an accident? The modern reactors look cheaper than the first generation ones. I know simplicity is deceiving but the first BWR don't even look easier to engineer, just much more inelegant and enormously less safe.

The passive components and systems tend to require more room than active ones, and set some boundary conditions for the plant layout, too. I am not aware of all the aspects that led to the design solutions based on active components, but I remember having been told by one experienced designer that the rapid scaling of the BWR:s from the 400-500 MWe class to 700-1000 MWe and beyond made it impossible to fit a suffiently large isolation condenser into the GE BWR plant without extensive redesign of the reactor building layout, and it was therefore replaced with the RCIC in later designs. I've heard similar explanations for abandoning the IC in the ASEA reactors between Oskarshamn 1 and Ringhals 1, but I've not been able to confirm if they're true.

Ever heard of the PIUS reactor? That was an attempt by ASEA in the 1980's to make a passively safe PWR. The main arguments were not, however, in the improved safety, but rather in the simplified licensing process. The version they were marketing in the early 1980's had a thermal power of 670 MW, primary pressure of slightly below 100 bar, and a rather bad thermal efficiency. The practical goal was to commersialize a 2x500 MWe twin-unit by the 1990's, but this project was terminated after the Chernobyl accident, when life extension and power uprates of the old units became the only politically acceptabe method of increasing nuclear generating capacity in most countries.

http://www.euronuclear.org/e-news/e-news-17/nps-kth.htm

EDIT: There was also a BWR version of the PIUS concept: http://www.osti.gov/bridge/servlets/purl/5148883-2dM7Re/5148883.pdf