Skip to main content

Predicting water quality risks at mine closure: a study in cold weather climates

At mine closure, waste rock cover systems in the Arctic can employ freezing processes not available in more temperate climates.

An Arctic gold mine was approved to use a thermal cover system to isolate potentially acid generating and metal leaching waste rock as part of reclamation strategies. Okane optimised the cover system design by integrating historic thermal data with robust visualization of thermal, water flow, and gas flow processes in GeoStudio. The approach led to an understanding of the long-term water quality risks and successful approval of an optimised design.



Gillian Allen
Senior Engineer – Okane Consultants


16 min

See more on demand videos


Find out more about Seequent's environmental solutions

Learn more

Video Transcript

<v ->Hi, thanks Aaron for the introduction.</v>

So as Aaron mentioned, today I’ll be presenting a case study

on modeling waste rock storage facilities in cold regions,

and I’m hoping to step a little bit away

from the technical aspects of the modeling

in this case study

and focus a bit more on big picture thinking.

And what I hope you take away from this presentation today

is that really understanding the conceptual model

is very important

and having a robust conceptual model

leads to modeling results that are better able

to inform on important management decisions

when it comes to closure of waste rock storage facilities

in cold regions.

So as I mentioned, the case study today

will focus on closure of waste rock storage facilities

in cold regions.

This is particularly relevant today

as Arctic exploration is continuing to expand

as industry looks for new deposits.

And this was going to lead to an industry-wide need

to have a good understanding of how to achieve

successful closure in cold regions.

And as of any site,

it’s important not to use a cookie cutter approach

when we’re talking about closure of the mine site,

but in particular for cold regions,

there are definitely different management challenges.

For example, the very remote nature of most of these sites

or the lack of typical reclamation materials

that you might use at more temperate sites,

but there’s also different opportunities

that present themselves in these cold regions,

such as the opportunity to use permafrost conditions

to our advantage.

So many of the management strategies

that people might be used to from working at mine sites

in more temperate regions,

just aren’t available in these cold climates

and we need to make sure that we’re thinking about these

in a site-specific way.

Specifically, today’s case study focuses

on waste rock storage facilities

from an open pit gold mine in the Arctic

that’s in the beginning stages of production.

They have some reactive waste rock at this site,

which has a potential for acid rock drainage

and metal leaching.

And so the site came to us initially

with a concept that they needed to flesh out a bit more.

So the concept for closure was that they would maintain

the reactive waste rock in a frozen state

by use of a thermal cover system.

So what this means is that the annual freeze-thaw layer,

to the layer that seasonally thaws in the summer

and then freezes back every winter,

would be maintained with a non-acid generating materials.

So the idea here is that the reactive waste rock

would remain frozen,

reducing the reaction rates, resulting in ARD,

but also limiting the liquid water

that would then transport any metal products or acid

away from the landform.

So the assumption behind this objective

that the site presented to us is that any acid generating

or metal leaching rock that’s available to be unfrozen

but available in the active layer

may lead to water quality issues at the toe

and would contribute to a loading from the lawn form.

So the first thing that we looked at was

is that really the objective is the objective

to maintain this waste rock in a frozen state?

And arguably the overarching objective here is

that we’re actually worried about water quality.

So the overarching objective

is to limit contaminant migration to the protect

so that we’re protective of surface

and groundwater quality.

So this overarching objective is actually going

to become the measurable

by which we understand if the management strategy,

which is this cover system is successful or not.

The site from this case study is located in Northern Canada

in the Arctic.

So in the Koppen Geiger climate classification system,

it’s listed as ET or a tender climate.

So this isn’t a continuous permafrost region.

So these are the different permafrost regions in the Arctic

or Northern Canada,

or some are continuous runner

in this dark blue area.

That means that the ground is frozen at depth year round.

So we were asked to flesh out

the proposed reclamation strategy.

So the 4.5 meter thermal cover system.

So how we do this,

or how we like to do this before ever turning on a model

is to develop a really good conceptual model

of what we think that management strategy,

how we think that management strategy will perform.

So we do this by looking at all the different facets

that affect performance.

So things like climate, the hydrogeologic setting,

the materials available for reclamation

and in some cases,

the development of vegetation.

So understanding how all of these things interact

with the waste rock storage facility in question

lead us to have an understanding

of all the different processes.

So things like the surface water balance

or the energy and thermal balance of the landform.

So understanding these is really important

to forming a robust conceptual model,

because it means that we understand things like

that the water balance is likely to be dominated

by spring melt in this area

or that because we’re at such a high latitude,

there’s likely going to be some fairly significant slope

effects due to shading.

So the surface energy balance may look different

on different areas of the landform

or that gas transport might be a really important mechanism

in freeze back of this landform,

because it’s the dominant feature in the lab

in the landscape,

because we’re talking about a very flat tundra,

and we’re talking about an 80, 90 meter

waste rock storage facility.

So there might be significant wind effects

that either warm up or cool down the landform.

So we’ve spent our time developing our conceptual models.

So we should be ready to turn on our numerical models

right now, right?

Not quite.

We still need to do a bit more thinking in order

to make sure that we’re capturing all the processes.

So our conceptual model

is probably based on one specific snapshot in time

in our minds.

And it’s probably right at closure,

which is shown here on this graph,

which was a level of effort

that goes towards closure over time.

So it peaks sort of near closure

and then comes back down in post closure.

So we may have formed a conceptual model

for this one particular snapshot in time,

but do we understand the processes

that have led to this specific,

the state of your waste rock storage facility at closure?

So do you understand your mine plan?

Do you understand what materials are coming up when,

and so we generally do have a pretty good idea.

We do have some information that can help us understand

or constrain what we think the waste rock storage facility

will look like at closure,

but if we haven’t considered how things will change

and how things will wind up

and the processes that will lead to this point,

it makes it really difficult to even make a stab or taken,

make an estimate of how things

will behave in post closure, right?

Because we have even more uncertainty further out in time.

So things like,

do we understand how climate change

will affect our conceptual model?

Will the materials weather?

Will hydrologically, will this landform

be sort of at a steady state?

So those are all things that we need to consider

when we look at the data.

And when, before we start modeling things out into longterm.

Luckily for us at this site, we did have some data.

So we’re able to improve our conceptual model

for that one snapshot in time at closure,

by using the available information to us.

So for us, one of those things

is that we have the mine plan.

So we can look at how this landform

gets built up over time.

And what this means is that we have information

on where the most reactive waste rock

is likely to be placed.

And we might know, based on the construction schedule,

that they’re only placing material A or B in winter.

So the initial temperature is going to be lower.

So what this leads to

is better understanding of what the landform

will actually look like when we get

to this one closure snapshot.

at the end of mine life.

At this site,

we also had a nearby waste rock storage facility

that was already constructed

that had some thermal data near our surface

that we could use for calibration.

So this is the thermal data on the left here.

So we see time marching along,

along the bottom access and temperature,

year over year shown in the colors

and the depth on the y-axis.

So obviously the red is means warmer

and the blue means cooler.

So it’s great that we have this data to calibrate too,

but it’s really important to place it in time.

So if we were to say that this date,

this is what we expect the freezing and thawing

to look like at that one point in time at closure,

that may not be a correct assumption

because this reflects the evolution

of the management strategy that’s being implemented,

not how it acts specifically at one point in time.

So making sure that we understand

what the hydrology during this timeframe looks like

can be really important to making sure

that we’re getting it a good calibration by framing

this in the right timeframe.

These are just two examples of how we need to frame the data

that we do have in the right correct context of time.

So what this does is allows us to have a better constraint

on what that waste rock storage facility will look like

right at that moment in closure.

So that we have a fuller picture of what our starting point

might look like.

Knowing that there still is going

to be some uncertainty there,

but making sure that we have constrained that

as much as possible

before we start modeling

these waste rock storage facilities

way out into the future,

because we’re likely going to have a lot less confidence

in what are our predictions are,

what our models show us way out in the future,

because there is guaranteed to be more uncertainty.

For example, in the Arctic,

we’re dealing with some fairly significant

climate change conditions that are possible

in the future and understanding what the climate

looks like in 150 years is a much more difficult thing

to predict than what the climate will look like in 10 years.

So it’s really important for us

to frame all the existing data

or the data that we have in the short term

needs to be really framed in context for us

to be able to model long-term things more effectively.

So linking it all together,

we finally did turn on some numerical models.

So we use the Geostudio suite of models

using air/w, temp/w, CW NC tran,

because there are lots of different processes

as you saw when I went through the conceptual model

that affect the thermal conditions near surface.

So we have here is sort of the average annual temperature

near surface for this waste rock storage facility

in the long term.

So we have depth along the y-axis,

depths from surface.

Along the x-axis, we have an average annual year

and the black dash line

throughout the middle is the interface

between the cover system and the reactive waste rock.

So you do see that the active layer does extend

into the reactive waste rock.

So, if we think back to what our original objective was

and what our original conceptual model was,

we, if we were going to be asked,

does this meet the objective

of maintaining the active layer within the cover system?

The answer would have to be no.

However, we are missing a few important bits,

as I mentioned at the beginning.

So if we think only of the objective

as to maintain the freeze-thaw layer

within the cover system,

we’re missing the larger overarching objective

of achieving good water quality

and the assumption underlying that original conceptual model

or that original objective

that any speck of acid generating rock

that can thaw could lead to load is what’s driving

that flaw that the modeling is pointing out.

However, we’re not considering the time aspect, right?

So the assumption that any,

any speck of waste rock that could contribute to load

will contribute to load might be correct at sometime

in the future.

But what we see from the modeling is that

that’s not a correct assumption in the timeframes

that we’re looking at.

So understanding that if our conceptual model is static

and we don’t understand the processes

that are leading to the actual impact,

that we won’t have a good understanding

of whether the management strategy that we’re looking at

is successful or not.

So basically when we don’t consider the timeframe

in our modeling,

we lose all the color of the solution.

So really considering our conceptual model

in the context of time,

lets us move from that 2D drawing

to a more colorful picture.

So when we go back and look at the same results

in the context of time and the overarching objectives.

We can see that the anticipated performance of this

can’t be measured only by understanding

if the cover system remains frozen,

you need to have a greater look at what the impact of it

not being totally frozen all the time has on water quality.

So it could be that our management strategy

of the cover system works perfectly, stays frozen

and there’s no reactive waste rock that that ever thaws,

but we still could have water quality issues

if we think about this, how it was framed originally,

but that wouldn’t mean

that the management strategy is successful

because you still have a water quality issues.

So when we thought about the modeling

from a different perspective

and really thought about the conceptual models

and the different processes that are occurring

and linked it back to water quality,

and didn’t look at performance on just a frozen,

not frozen basis,

we’re able to have a fuller picture

of what successful closure might look like at the site.

So really what it boils down to

is that these successful closure designs

that might happen far out in the future

really need to have a strong conceptual model of performance

to be able to make successful management decisions.

So to do this,

you need to understand the effect of time on performance,

not just understand what might drive performance,

you need to context contextualize the field data

that you might have to make sure

that you’re making the right assumptions

with the data you have.

And during all this will allow you to project future

conditions with a bit greater confidence

than you would just thinking about your conceptual models

in a static space.

So thank you all for listening to this quick talk.

I hope you found something of interest

that you can take away.

And if you have any more questions

about this particular case study

or any other problems that you might be facing.

Feel free to reach out to us at [email protected].