Dr Andrew Craze
Ref: GDF/T&O/2015/02
Radioactive Waste Management Limited
Head of Repository Safety and Environment
Herdus House
Westlakes Science and Technology Park
Moor Row
Cumbria
CA24 3HU
Date: 17 March 2015
AMEC Technical Report - Pre-conceptual Stage LoC Submission for the Disposability of
Contaminated Mercury
Dear Andrew,
RWM has asked if the Environment Agency’s groundwater modelling specialists could look at the types of
model RWM is currently using to assess the effects of non-radioactive contaminants on groundwater. In
particular, RWM has drawn our attention to the modelling work reported in Appendix 3 of an RWM
contractor’s report1. This letter provides some overview recommendations, set out below, related to this
modelling work and some detailed comments, presented in the attached review report. Our overview
recommendations to RWM are:
Recommendation 1: Ensure that suitable and adequate information is provided and referenced, throughout
the document, to support the modelling work.
Recommendation 2: Explain what is meant by “flow focussing”, and how this relates to the assumptions
about water balance and advection transport, and to the assumption that mercury is evenly distributed in the
backfill.
Recommendation 3: Explain in the document why advection is not considered in the modelling, because we
regard it as a main process in the GDF backfill.
1 Pre-conceptual stage LoC submission for the disposability of contaminated mercury. Amec report reference:
202017/TR/002. Issue 2, dated 10/10/14,
________________________________________________________________________________________________
Environment Agency, Nuclear Waste Assessment Team,
Ghyll Mount, Penrith 40 Business Park, Penrith, Cumbria. CA11 9BP.
T: +44 (0) 02070926416 |M: 07771940556 |e-mail: xxxxxx.x.xxxxx@xxxxxxxxxxxxxxxxxx.xxx.xx
Recommendation 4: Explain and provide evidence to substantiate the assumptions relating to scaling up and
their effects, including a consideration of alternative assumptions.
Recommendation 5. Carry out sensitivity analyses to explore the effect of varying each major assumption
and parameter.
In addition to the recommendations above, we have noted some points for clarification and areas for
improvement (in the attachment to this letter). We suggest that it may be helpful to hold an exploratory
meeting, involving our groundwater specialists, with RWM and its contractor to discuss these
recommendations and RWM’s responses to our comments in the attachment, This letter and the attached
review could serve as a framework for the meeting, to be held as part of the 2015/16 scrutiny programme.
When convenient, please contact me to arrange a suitable date for such a meeting.
Yours sincerely
Robert E Smith,
Distribution:
EA: D Ilett, S Duerden, T&O Task Leaders, T Besien, P Hart, R Evans
ONR: I Streatfield
SEPA: R McLeod
NRW: R Price
________________________________________________________________________________________________
Environment Agency, Nuclear Waste Assessment Team,
Ghyll Mount, Penrith 40 Business Park, Penrith, Cumbria. CA11 9BP.
T: +44 (0) 02070926416 |M: 07771940556 |e-mail: xxxxxx.x.xxxxx@xxxxxxxxxxxxxxxxxx.xxx.xx
Document reviewed
Author
Version
Date
Report Number
Comments
Pre-conceptual stage LoC
Amec (A
Issue 2
10/10/2014 Amec report reference:
submission for the
Tuxworth, A
202017/TR/002
disposability of contaminated
Green, M Kelly, A
mercury
Guida)
Regulators Review
Principal Reviewer(s): T Besien (E&B Senior Adviser: Water, Land and
Biodiversity), P Hart & R Evans (Permitting Technical Specialists: E&B
National Services), R E Smith (NWAT)
Approval
Date:
Date sent to RWM
D Ilett
16.03.15
17.03.15
Letter ref: GDF/T&O/2015/02
File ref
: O:\Geological Disposal\Programmes &
Projects\Scrutiny\Implementing Geological Disposal\2014-15\Approach
to non radiological hazards
1. Summary of our understanding of the risk
assessment approach
Appendix 3 provides an assessment, intended to be conservative, of peak concentrations of
mercury in pore water in the GDF backfill. The assessment assumes the mercury is derived from a
single package containing contaminated mercury which has been disposed of to the GDF. Seven
different packaging options are assessed, and the results are compared with the Environment
Agency’s minimum reporting value (MRV) for mercury. If multiple similar packages were disposed
of to the GDF, Appendix 3 argues that the results would scale proportionately.
The conceptual model is described for each packaging option, and assumes:
• mercury is evenly distributed throughout the volume of the waste that contains it
• resaturation of the GDF and failure of inner & outer containers (where they exist) occurs
immediately after the GDF is closed
• except where mercury in the wasteform is present as an amalgam, it is released immediately
into the backfill porewater (that is, the modelling doesn’t allow for diffusion within the
wasteform); where mercury is present as an amalgam, it is released slowly by diffusion
through the wasteform
GDF/T&O/2014/02_Pre-conceptual Stage LoC Submission for the Disposability of Contaminated Mercury_i1
Page 1 of 4
• once mercury is released into the backfill, there is even distribution and advective flow through
and out of the backfill; groundwater outflow from the backfill is greater than inflow (owing to
“flow focussing”, a term used but not explained in Appendix 3)
• sorption of mercury to the waste form and backfill is neglected
The mathematical representation of mercury transport is specified but not substantiated or
referenced. The assumptions made in deriving the results are stated. We consider that the
assumptions are reasonable and partially substantiated. The diffusion coefficient and solubility
limits used in the assessments are stated and the source references are given. Goldsim is used to
generate the results.
Recommendation 1: RWM should ensure that suitable and adequate information is provided and
referenced, throughout the document, to support the modelling work.
Results are given in Table 15 and compared to the Minimum Reporting Value (MRV) of 0.01μg/l.
The results are at least two orders of magnitude below the MRV. Some of the assumptions made
in deriving the results are clearly conservative, such as: instant metallic mercury release into
backfill porewater; immediate failure of inner and outer containers on GDF closure; instant
saturation of backfill and waste on GDF closure; and exclusion of adsorption processes. However,
less conservative assumptions are also made, such as uniform distribution of mercury in the
backfill.
2. Detailed observations
We have a number of queries and observations which we would like RWM to clarify through our
ongoing discussions and some suggestions to help RWM improve its approach, as follows:
2.1. A2 Assessment of Consequences
The 3rd paragraph states that the 7 mercury packaging options share a number of similar features
but does not say what these are. Inclusion of a list of the common features would help to clarify the
conceptual model, since the modelling approach may also be the same, or similar, for these
features.
In the same paragraph “backfil porewater” is referred to but it is not clear from the document
where the backfill is located. RWM should clarify this (probably by amending Fig 10).
2.2. A2.1 Approach to modelling the near field
In the paragraph beginning “The simple conceptualisation consists of a number of components”,
the second bullet says, “An outer container that includes all of the components listed above”. RWM
could clarify this by referring instead to Table 13, which sets out the conceptual model for each of
the seven options.
GDF/T&O/2014/02_Pre-conceptual Stage LoC Submission for the Disposability of Contaminated Mercury_i1
Page 2 of 4
The document states that: “Once in the backfil the mercury is transported out of the GDF by
advection, on account of water from the host rock that flows through the GDF. Note that, in
general, the flow rate out of the GDF backfill (QC) is greater than the flow rate in (QR), on account
of the effects of flow focussing.”
Recommendation 2: RWM should explain what is meant by “flow focussing”, and how this relates
to the assumptions about water balance and advection transport, and to the assumption that
mercury is evenly distributed in the backfill.
2.3. Fig 11 conceptual flow diagram
RWM should amend this figure to clarify whether “wasteform” in this context includes the inner
container, in those options where such a container is present.
2.4. A2.2 Mathematical model
From the information given, the basis for the mathematical modelling and the equations is unclear.
The document should include references to supporting information (see Recommendation 1). The
equations quoted are dissimilar, for example, to diffusion equations set out in our report2.
Recommendation 3: RWM should explain in the document why advection is not considered in the
modelling, because we regard it as a main process in the GDF backfill.
We would welcome further discussions with RWM to increase our understanding of the approach
and equations.
2.5. Results
Appendix 3 states that scaling up from one package to multiple packages is in direct proportion to
the number of packages. While we recognise that this is likely to be conservative, we recommend
RWM considers alternatives. In particular, we suggest that if just one volume of water passes
through the first waste package to the next, and so on, there would be an incremental increase in
discharge concentration for each package. However, if every waste package is instantaneously
saturated the concentration would not increase after the first waste package, but would stay the
same.
Recommendation 4: RWM should explain and provide evidence to substantiate the assumptions
relating to scaling up and their effects, including a consideration of alternative assumptions.
2 Environment Agency. Contaminant fluxes from hydraulic containment landfills: a review. September
2004. ISBN 1844323153, LIT 1879
. https://www.gov.uk/government/publications/contaminant-fluxes-
from-hydraulic-containment-landfills-a-review
GDF/T&O/2014/02_Pre-conceptual Stage LoC Submission for the Disposability of Contaminated Mercury_i1
Page 3 of 4
2.6. Sensitivity analysis
We recognise that certain assumptions in the assessment (e.g. containers fail immediately on GDF
closure) correspond to a conservative (“worst case”) approach, whilst other assumptions (for
example that mercury is evenly distributed in the backfill) may not.
Recommendation 5. RWM should carry out sensitivity analyses to explore the effect of varying
each major assumption and parameter. These could include, for example:
• Assessing the effect of organic materials on mercury mobility. In this work, organic materials
are assumed to be absent in the wasteform, but present in the backfill. The assessment
assumes enhanced mercury solubility in the backfill, to account for the presence of organic
materials and, conservatively, no adsorption to the wasteform or backfill. Elemental mercury
and mercury sulphide have a very low solubility limit in the absence of organic materials, but
the solubility is significantly increased in their presence. In addition, adsorption to the backfill is
reduced. It might be helpful for RWM to explore more realistic and less conservative scenarios
by considering different levels of mercury solubility in, and adsorption to, the wasteform and
backfill.
• Assessing the effect of variations in pH / Eh values on discharge concentrations.
• Assessing the effect of varying groundwater volumes and flow rates on dilution and discharge
concentrations. The dilution provided by assuming immediate saturation of the GDF backfill by
groundwater (in whatever volume may be necessary to achieve this) to the mercury metal
immediately released from the package or the mercury more slowly released by diffusion from
amalgam (depending on which of the seven options is being considered) appears to influence
the results.
• Although not directly relevant to the disposability of contaminated mercury, a concern we have
is the possible presence of contaminants such as mercury or chromium VI, arising from grout
additives such as ground granulated blast furnace slag (GGBS) or pulverised fuel ash (PFA),
especially near the boundary of the GDF with the host rock. We recommend that RWM should
consider modelling contaminant diffusion from the grout in its future studies.
GDF/T&O/2014/02_Pre-conceptual Stage LoC Submission for the Disposability of Contaminated Mercury_i1
Page 4 of 4