[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Recent safety hazards at aging nuclear plants

Apologies to the list, but there follows a rather long response to Norms

posting which contains only a little reference to radiation safety (buried

in the eighth paragraph).  Feel free to stop reading now if you wish

although I hope some of the following might be of general interest.

I am assuming (always a dangerous thing to do) that the author of the

article below is trying to indicate is that the age of the nuclear plants in

the US is leading to increasing safety problems which in turn have meant

more unreliable plant.    In addition by referring to "the past three years"

he suggesting a worsening trend (Norm please correct me here if you think I

am wrong in my assumptions).

I have to say that I disagree both with the suggestion of increasing safety

problems AND that the trend is getting worse.

There are a number of key indicators (used both Nationally in the US and

Internationally) that can and are used to asses the performance of the US

nuclear plants, ALL of which have shown consistent long term improvement not

only for the last three years but in some cases as far back as 1980.  These

include the following.

Unplanned capability loss factor - This is the percentage of maximum energy

generation that a plant is not capable of supplying to the electrical grid

because of unplanned shutdowns or outage extensions.  A low value indicates

important plant and equipment is well maintained and reliably operated and

there are few outage extensions.  This has gone from a value of 11.6% in

1980 to a low of 3.9% in 1998, 2.0% in 1999 and 1.7% in 2000.

Unplanned Automatic Scrams - the unplanned automatic scrams per 7000 hours

critical indicator tracks the median scram (automatic shutdown) rate for

approximately 1 year (7000 hrs) of operation.  This has gone from 7.3 (per

7000 hrs critical) in 1980 to a value of 0.0 in 1998, 1999 and 2000.  In

2000 59% of operating units had zero automatic scrams.

Safety System Performance - The safety system performance indicator monitors

the availability of three important standby safety systems to mitigate

off-normal events.  For 2000 the goal was for > than 97% availability of

these three safety systems.  In 1989 only 70% of the systems achieved the

2000 goal of > than 97% availability, In 1998, 98% of the systems achieved

this target and for 1999, 95% achieved this target and in 2000, 96% achieved

this target.  Demonstrating that consistently high standards have been

achieved for the last three years.

In addition to those detailed above (which focus specifically on the safe

long term operation of the plants) Fuel reliability has increased year on

year from 1989 through to 2000.  Chemistry performance (an important part of

ensuring long term condition of the plant) has been on an upward trend since

records commenced in 1994.  Collective radiation exposures have reduced

dramatically since 1980.  Median PWR collective doses for 2000 were

approximately 20% of those in 1980, median BWR collective doses were

approximately 17.5% of those in 1980.  Volumes of Solid radioactive wastes

in 2000 were a fraction of those in 1980, 4% of 1980 values for PWRs and

7.5% for BWRs (m3 per unit - median value).  Record highs (91.1%

approximately 3 times that used generally when assessing new wind projects

which are normally based an a unit capability factor of about 30%) in 2000

for unit capability factors.  The lowest Industrial Safety Accident Rate

since 1980, 0.26 accidents per 200,000 worker-hours.   All of the above

information was released by the Institute of Nuclear Power Operators (INPO)

and published in the May 2001 issue of Nuclear News (A publication of the

American Nuclear Society).  All of the above indicates not a worsening trend

for the "ageing nuclear plants" BUT significant improvement year on year of

a number of key areas of plant operation.

Moving on I would like to review some of the incidents cited in the report

identified by Norm.

snip > January 1999: Inadequate maintenance led to a six-hour hydrogen fire

on the

roof of the control building at J.A. Fitzpatrick in Syracuse, N.Y., forcing

a plant shutdown.<snip

It is difficult to establish on the basis of the very brief summary the

seriousness of this incident.  To my knowledge there are usually only two

areas on a nuclear plant where you might find Hydrogen gas (unless the site

has an electro-chlorination plant).  One is as part of the purge system for

the reactor coolant treatment plant, used for chemistry purposes (oxygen

control in the reactor coolant minimising corrosion) and to remove

radioactive gases which are then treated in some way prior to discharge.  I

am assuming that as no mention was made in the report of radioactivity, that

the fire was not related to this system as any vent would have been

discharging radioactivity and this would have been highlighted by the person

producing the report.  the second area is in the turbine to cool the

turbo-generator.  These systems do have vents to atmosphere but I would not

have expected them to have vented above the control building.  However if

the fire did occur due to an excessive release of hydrogen from the

turbo-generator then this type of incident could have occurred at any type


large power generation plant.  It is common to find fires on the non-nuclear

side of nuclear power plants reported while similar incidents at coal or gas

fired plants are not.  While a fire on a nuclear plant should be treated

seriously, if it is confined to the non-nuclear areas it does not present a

significant danger to nuclear safety.  As a final note of caution, even

though I feel I have a reasonable grasp of layout an construction of nuclear

power plants I am still only speculating.  This illustrates how difficult it

is to assess this type of incident without much more information than was

given by the author of the original report!

snip > August 1999: A cooling- water drain line in Callaway, Mo., broke

because of severe corrosion, forcing a reactor shutdown. A subsequent

inspection revealed at least 10 areas where pipes had decayed and were in

danger of breaking.<snip

This type of statement is I believe intentionally misleading on the part of

the author.  I have to admit to becoming particularly aggrieved whenever I

see the statement "cooling-water".  I think it is used intentionally to

cause fear and alarm by those opposed to nuclear power.  WHAT COOLING WATER

SYSTEM are we talking about.  Off the top of my head I can come up with

many - component cooling water, essential service water, condenser cooling

water (often the one most talked about and having nothing to do with nuclear

safety), auxiliary service water, general service water, pond cooling water,

stator cooling water, cooling water to condition chemistry sample streams

etc. etc. etc.  Faults with several of these cooling water systems could

have required the reactor to be shut down although none directly affect

nuclear safety.  Again as no radioactivity was mentioned I am assuming that

the pond cooling water was not involved.  If I was a betting man I would

place my money on either the component cooling water or on the main cooling

water system used to condense the steam after it has passed through the

turbine.  In reality the intention of the original statement was to link in

the minds of the public the problem and the reactor.  In general if you ask

a member of the public about the "cooling water" they assume it is the water

used to cool the reactor i.e. the Reactor Coolant System.

snip> 1999-2000: Millstone in Waterford, Conn., had to repeatedly shut down

due to

> > failures of the reactor control-rod drive system, including control rods

that came loose and dropped into the reactor. The plant operator blamed

failed insulation and damaged electrical leads.<snip

All this demonstrates to me is the inherent fail-to-safety characteristics


the design.  Any loss of power to the control rod system results in the rods

entering into the core and prompting a reactor shutdown.

snip> February 2000: A steam generator tube ruptured at Indian Point 2 in

New York, contaminating 19,000 gallons of cooling water and releasing

radioactive steam into the atmosphere.<snip

This one is a difficult to regard as anything but an error on the part of

the operator and the NRC.  While the consequences have been greatly

exaggerated, the SG which ruptured was, with the others at Indian point 2,

the last of that type in service.  All SGs of that type installed at other

plants had been replaced.  The owner had replacement SGs available and had

postponed the replacement of the installed SGs several times.  In addition

there would appear to have been flaws in the SG inspection techniques and in

the approach adopted by the NRC to reviewing inspection results.  The

overall outcome was an incident which has been advanced by Norm and the UCS

as an example of how both the industry and the regulator cannot be trusted.

snip > May 2000: A failed electrical conductor at Diablo Canyon 1 in San

Luis  Obispo County triggered a fire that cut power to the coolant and

circulating water pumps that keep the nuclear core from overheating.<snip

Once again the lack of information makes it difficult to asses the real

potential for a significant incident.  Jumping in with both feet again, I

find it difficult to accept that a single fire could result in the loss of

all reactor coolant pumps and circulating water pumps.  Perhaps I am

demonstrating my ignorance but I did think that ever since the Browns Ferry

Fire important items of plant were placed on separate electrical supplies to

prevent just this type of incident.  While it may be possible for a

significant fire to remove the primary supply to these pumps and cause them

to trip (although I didn't think that should have been possible) they should

definitely have been supplied by separate, independent supplies in the event

of a loss of primary supply.  I suppose that a transitory loss of power will

however be advanced a loss of reactor cooling capability no matter how

transitory that may have been.

snip > August 2000: Peach Bottom Unit 3, in Pennsylvania, was forced into

emergency shutdown when an instrument valve failed and caused a leak of

contaminated reactor cool ant outside of primary containment. A similar

valve failure and leak of radiation had occurred May 28, 2000, but the

valves were not replaced.<snip

In order to asses this incident properly I believe the author should have

include the quantity of reactor coolant lost, the quantity of radioactive

material released and the actual rate of shutdown of the reactor.  There is

no indication of a release to the environment, not that I would have

expected one in these circumstances.  No indication of overexposure of any

workers.   There is no information on the exact nature of the valve failure.

It is usual for the utility to undertake an investigation of this type of

incident when it occurs.  Depending on the cause of the valve failure,

whether the problem lay with the maintenance or the installation or the

manufacture of the valve,  I would have expected the operator to have

reviewed the operating, maintenance and installation history of that type of

valve on the plant, if the event was significant on other plants utilising

that type of valve and with the manufacturer of the valve.  However if the

investigation did not identify the potential for a common failure mode,

either through incorrect installation, maintenance or manufacture then there

was little reason to remove other valves from service.  Again while

highlighting the original problem the author of the report does not identify

what if any failures were responsible and where any deficiencies lay.

snip> October 2000: At V.C. Summer, in South Carolina, a 29- inch diameter

coolant pipe, with walls more than 2 inches thick, suffered a crack due to

water stress corrosion, creating a leak of radioactive cooling water. Crack

indications were later found at four more reactor inlets.>snip

Little comment to make on this incident other than it is a significant event

that is receiving significant attention from the operators and  the


snip > November 2000 to April 2001: After receiving a 20-year license

extension, operators of Oconee 1, in Seneca, S.C., found 19 cracks in the

reactor where control rods pass through to the nuclear core. Radioactive

cooling water had been leaking into the containment sump. In February nine

leaks were  found in Oconee 3, which had been taken down for refuelling.

Oconee 2 was later found to have four leaking control-rod nozzles.<snip

I have already circulated my view on the generic issue of cracks around the

control rod drive penetrations.  All I would add is that it is not clear

(although given that it was not stated it is unlikely) that the plant was

shut down prematurely.  If this was identified during a routine shutdown (as

were the problems on Oconee 3) then any leakage must have been within the

operating constraints.  It is not clear from the information presented here

and on the other related posting how these problems differ from long

standing issue with cracking of SG tubes.

Snip> January 2001: Failure of an 18-year-old valve at North Anna, Va.,

created a leak of radioactive coolant of more than 10 gallons per minute,

forcing a shutdown of the reactor.<snip

This incident appears to demonstrate the correct application of the

operating limits for the plant by the reactor operator.  While the age of

the valve is included to add weight to the arguments about "ageing plant"

there is no information provided by the author to indicate that the age of

the component was responsible for the failure of the valve.  Once again it

could have been a maintenance induced failure, a failure due to

incorrect/poor installation (defective welds etc.)  poor plant chemistry

leading to excessive corrosion.  If the author is trying to construct a

adequate argument to support his hypothesis on the age of the plant leading

to degraded safety he needs to provide more information.

snip > February 2001: A 20-year-old circuit breaker at San Onofre 3, near

Camp Pendleton, failed to close, creating a 4000-volt arc and fire that cut

power to coolant control systems, drowned emergency switching valves and

shut down emergency oil pumps, destroying the Unit 3 generator shaft.

Currently, 150 identical breakers remain in service at the plant.<snip

As with the previous incident there is no information provided which clearly


the age of the component as a contributory factor.  Similar points exist for

the earlier

incident where two valves of the same design failed.  There is nothing to

indicate that

there exists the potential for a common mode failure.  The investigation of

the incident

should have addressed the potential for similar failures to occur and if no

action was

required to remove the "150 identical breakers" then without additional

evidence there

is no reason why these should not remain in service.  Given that the major

item of plant

damaged by the incident was the Unit 3 generator it would indicate that the

fault occurred

on electrical systems associated with the non-nuclear side of the plant.  As

such the "coolant

control system" referred to here were probably associated with non-nuclear

coolant systems

possibly the cooling water pumps used to condense the steam in the turbine

condenser. Quite

how the loss of emergency oil pumps should result in the destruction of the

generator shaft is

not clear.  It is possible that the loss of the cooling pumps resulted in a

turbine trip.  Following a

turbine trip the bearings for the turbine shaft would have required

additional oil to ensure that the

 shaft did not come into contact with the white metal bearings.

snip > February 2001: After Arkansas 1 was re-licensed for 20 years,

extensive  cracking was

 found on the control-rod drives and thermocouple nozzles  entering the

nuclear reactor.<snip

This definitely seems to be the issue of the month for anti-nuclear

activists.  see previous comments

on this and the related posting.

snip > August 2001: Failure of a valve at Palo Verde 3, in Arizona, caused a

leak of  radioactive

cooling water from the irradiated fuel-cooling pool into  the reactor

containment building, forcing a reactor shutdown.<snip

Again there is very little detail provided by the author on the extent of

the leak neither the quantity of water

lost or amount of radioactivity released has been specified.  We can assume

that there was no environmental

release, significant exposure to personnel, significant effect on the

cooling of the spent fuel pool  or on plant within the containment building

(all of which would have been highlighted by the author).

To summarise, The author of this report can failed to produce a cohesive

argument to demonstrate that the safety of the US plant is decreasing and

more importantly in the light of the introductory statements, that the age

of the plants is at all relevant.  With the exception of the cracking around

control rod drive penetrations, the report is simply a collection of random

events at nuclear plants that have been selected in an attempt to undermine

the confidence of members of the public in the safety of nuclear power.  It

should be clear from the statistics provided by INPO that US power plants

have demonstrated consistent improvements in all key areas since records

were began and in particular that this trend has continued for the three

years referenced by the author of the original report.

Apologies for not having reviewed the original NRC reports of the incidents.

I would welcome comments from anyone who can provide more detail on any the

incidents cited above.  It was unfortunately one of those occasions where I

think if I had tried to check the detail of each incident I would have

failed to respond to the posting at all.  On this occasion I felt that

honest educated reasoning was preferable to a nil resposne.

Norm feel free to respond to any or all of the points I have raised.


            Julian Ginniver


You are currently subscribed to the Radsafe mailing list. To unsubscribe,

send an e-mail to Majordomo@list.vanderbilt.edu  Put the text "unsubscribe

radsafe" (no quote marks) in the body of the e-mail, with no subject line.