Graphite Weight Loss at AGR Reactors

There have been several reports in the news1 about EDF wanting to raise the limits of the amount of graphite which can be lost from the moderator of the AGR nuclear reactors. As usual the reports seem to me to not give some very important information. Here are just a few points I think it is important to know.

  • The graphite blocks act as a moderator to slows down the neutrons which makes it more likely that they will cause nuclear fission.
  • In the AGR reactor they not only act as a moderator but are also vital structural components. The graphite blocks have channels in them for the fuel rods, control rods, emergency shut down systems as well as allowing the CO2 cooling gas to pass effectively over the fuel.
  • The figure of 8% is just one of the limits which EDF wants lifted. This figure is for the core as a whole. Some areas can have over 40% weight loss.
  • That the graphite blocks are also cracking as well as loosing weight.
    Short of decommissioning the reactors it is very difficult to accurately determine the weight loss and cracking in the bricks.
  • This level of weight loss was not expected when the reactor was originally designed and the weight loss and cracking is still not adequately understood.

None of these points should be new to EDF or the ONR. However, what has been expressed in the media has been incomplete and sometimes more than just ‘economic with the truth’.

For example EDF has claimed2:

However, this is very different from the reports of meeting between EDF and the ONR3 (emphasis added):

However the interventions earlier in 2013 gave ONR cause for concern about the methodology being used to calculate weight loss itself. This concern applied both to the processes being used and the apparently small margins that existed between weight loss and the limits.

The problem of loss of graphite and cracking has been studied for a very long time since it was also a problem with the older Magnox reactors. Recently there has been several research projects to provide data to convince ONR that the new limits are safe.

First of all let us have a look at some of the possible safety problems since it gives an idea of why this area is problematic.

Block Cracking and Expansion

The brick as noted above are structural components. If they are weakened by too much weight loss they may be unable to support the bricks above. When they crack they also change dimensions which may impeded control rods or emergency shutdown procedures.

An accident may cause a sudden shock which may suddenly change the block alignments. Such problems have been discussed in an excellent report by Large Associates4 in 2006.

Under Moderation

Since the graphite’s main purpose is to slow down (called moderate) the neutrons then if there is less graphite then the reactor is under-moderated as compared with the original design. This can be compensated for in several ways such as increasing the enrichment of the fuel or moving the control rods out further.

One of the possible accident scenarios considered in the original design was water ingress into the reactor. Since water is a good moderator then this could cause a very sudden increase the power output (explosion). This effect with be much larger if the reactor is under-moderated.

Change in Neutron Density and Energy Distribution

Since the reactor is not as good at moderating neutrons then there will be changes to the energy distribution of the neutrons as well as their density since the loss of graphite is not homogeneous. The changes in the neutron distribution can change the way in which various components of the reactor age – including the moderator.

This brings us to some of the problems of measuring and determining the affects of the graphite weight loss.

It is possible to stake samples of the graphite blocks and determine the extent of cracking and weight loss from this. However, to determine the amounts of weight loss complicated statistical analysis had to be used5:

“We had all of this data, we actually had to bring in some real experts in statistics to help us rationalize the data to make it useful”,

Stephen Grant, Fraser-Nash power and energy director

I will not go into some of the problems of using statistical analysis when the underlying mechanisms are not fully understood. However, the fact that the amount of weight loss and cracking cannot be directly observed is worrying.

One of the problems is that these affects are not independent. For example if a crack or weight loss occurs this can affect the local temperature and coolant flow which may increase or decrease the amount of weight loss and cracking5.

“Graphite is non-linear. You can’t say, ‘this is what happened in the last six months, this is what will happen in the next six months.’ “

Stephen Grant, Fraser-Nash power and energy director

1 Dungeness B nuclear plant operator wants safety limit raised, BBC, 4 June 2014 (

2 No concern over possible change to Dungeness nuclear safety rules, says regulator, Utility Week, 4 June 2014 (

3 Level 4 Meeting on NP/SC 7623 43% Graphite Weight Loss Safety Case (Hinkley Point B, Hunsterston B, Heysham 1 and Hartlepool Power Stations)Office for Nuclear Regulation (ONR), September 2013 (

4 Brief Review Of The Documents Relating To The Graphite Moderator Cores At Hinkley Point B And Other Advanced Gas -Cooled Reactor (Report Ref No R3154 – Graphite), Large Associates, 28n June 2006 (

5 For the longest time, Nuclear Engineering International, 14 June 2013 (


Leave a Reply

Your email address will not be published. Required fields are marked *

Captcha: * Time limit is exhausted. Please reload CAPTCHA.