Sunday 25 May 2014

In this penultimate part of a series of six news items examining the unique contribution NNL makes to the onigoing operation of the HALES plants at Sellafield, we look at how the different strands of work were pulled together to provide a unique output – a highly accurate assessment of the corrosion rate inside the evaporators.

Combining the output from the supporting work of coil inspection, evaporator boilingrig, thermal and structural models, NNL needed to provide the final pieces of the jigsaw by using this data to develop a quantitative relationship describing the corrosion rate as a function of operating temperature. This would enable a prediction of the corrosion rate of the evaporator base and subsequently the keystone calculation of the remaining lifetime of the evaporator. A number of key National Nuclear Laboratory peoplewere involved at all stages.

Foremost among these was Professor Guy Whillock – a world leading expert in the area of nuclear chemical plant corrosion. Guy’s work was innovative and groundbreaking – providing unique and accurate analysis from a huge range of variables.

Five key pieces of information were needed:

1. Coil inspection data enabled NNL to establish the thickness of the coils in relation to the depth of liquor they had been operated in.
2. A thermal model was used to establish the relationship between coil thickness,depth and operating temperature.
3. Over 50 years of accumulated knowledge of the corrosion rate of stainless steel innitric acid was used to provide an initial expected corrosion rate with temperature.
4. Clearly it was vital to determine the ‘as-built’ starting thickness of the heatingcoils. This was a difficult task, involving assessment of the original pipe specifications, the likely variability in coil wall thickness and the impact of bending the coils on metal thickness. This huge range of variables required the expert judgement and analysis of NNL materials specialists.
5. Known operating conditions for the heating coils. In particular, the coils had been used in three different types of highly active liquor, all of which have their own corrosion characteristics.

Through accurately combining this myriad of complex information, the National Nuclear Laboratory was able to predict a corrosion rate for the heating coils. Factors were incorporated to allow for batch to batch variability in the corrosion resistance of the stainless steel used to fabricate the evaporator and the effect of welds on corrosion resistance. The methodology development was validated by calculating thethickness of randomly selected coils and comparing the prediction with the known measurements for these coils. Very close matches were obtained. This enabled further calculations to be made to establish the thickeness of the evaporator base. To further enhance the service to our customer, statistical modelling was also used to predict uncertainties in calculations and provide upper and lower corrosionrate predictions.

The work drew heavily on NNL’s corrosion knowledge in stainless steel corrosion in irradiated nitric acid and water – a body of knowledge unparalleled internationally. Our customer was able to use this knowledge to plan for future evaporator replacement with much greater certainty.

In the final part, read an overview of the work and find out what we can do for you.