Hang tight while we fetch the video data and transcripts. This only takes a moment.
Connecting to YouTube player…
Fetching transcript data…
We’ll display the transcript, summary, and all view options as soon as everything loads.
Next steps
Loading transcript tools…
Selenate bioreduction in a large in situ field trial | Innovinc Conferences | YouTubeToText
YouTube Transcript: Selenate bioreduction in a large in situ field trial
Skip watching entire videos - get the full transcript, search for keywords, and copy with one click.
Share:
Video Transcript
Video Summary
Summary
Core Theme
This presentation details a successful, cost-effective, and resilient in-situ biological reduction method for selenium and nitrate in saturated rockfills, demonstrating its robustness and long-term stability for mine water management.
Key Points
Mind Map
Click to expand
Click to explore the full interactive mind map • Zoom, pan, and navigate
Hi folks, Lisa Kirk here. I'm the
principal biogeeochemist with
Environment. We're a division of respect
uh located in Bosezeman, Montana about
six hours south of the Elk Valley where
this work with Tech Resources and now
the Elk Valley Resources Group um
studying in situ biological reduction of
selenium and nitrate in coalmine
saturated rockfills has been undertaken
by myself and my co-authors. I'm sorry I
wasn't able to get to Berlin to be with
you in person uh due to a family
emergency. I do want to um share with
you however what we've learned because
we're excited about this robust and
resilient approach to mine water
management. In this case we have u green
injection wells and red pumping wells.
The red pumping wells pull water away
from the injection wells in both
directions. And then the yellow and the
orange are are wells that were
constructed for monitoring purposes. Uh we
we
monitor in these locations both
chemistry using you know pumped
groundwater but also uh with bioants
that let us measure the microbial
community. Uh this facility has been in
operation for 8 years originally as a
fieldcale trial facility and now as a
commercial facility which treats 20,000
cubic um meters of water daily to reduce
um 99% of the nitrogen and 97% of the selenium
selenium
and in doing so reduce the client's cost
of water treatment by about 80% on capex
and 50% on OPEX. s uh the facility has a
30-day residence time roughly. The
treatment zone is approximately half a
kilometer long and the thickness of the
saturated zone in the subsurface is
about 40 m. Um treatment happens within
a relatively short period of time. We'll
have a look at our information about
saturated rockfills from column and
fieldcale trials. Uh here we see nitrate
and selenium affected water being
introduced via pumping with nutrient
into the saturated fill uh which has an
oxygen gradient. Um nitrate is ideally
being converted to nitrite and nitrogen
gas which is inert. Selinate is being
converted to selonite and elemental
selenium which is insoluble. We are
using the native bacteria to accomplish
this. And we've done a variety of of
studies at the column and and bench and
field scale to understand operational
performance upset responses, response to
closure, reversibility and longevity of biommineralization,
biommineralization,
speciation of these compounds etc. So um
let's have a look.
Selenium is occurs in mine water as both
selonate um or selonite. These are both
the more oxidized and more mobile forms
of selenium. Once selenium is reduced to
zerovalent selenium or to selonide, it
becomes immobile. Our goal is to reduce
selenium to a an insoluble form such as
elemental selenium. But um absorption of
selonite uh is also helpful. And uh we
want to do this with the native
communities. We want to create long-term
stable uh sequestered selenium products.
Here we have uh rock that has been
weathered in a column uh in our
laboratory over a period of 3 years. We
can see the red elemental selenium
that's formed uh on on these mineral
surfaces with the white bofilm
associated with with um the growth uh of
of the bacterial community that
supported the reduction of this selenium.
We collected samples through sonic
drilling and uh built replicate columns.
We used sight water uh and amended it
with methanol and phosphoric acid
in an upflow configuration. We saturated
these columns and we collected effluent
for analysis of multiple parameters on a
on a weekly basis. Um we obviously
looked at DOC and phosphorus but then
again nitrogen, sulfur, selenium, other
dissolved metals, some of the
geochemical parameters, redux, oxygen,
pH and conductivity. These tests are
conducted in a in a um cold room so that
we can replicate subsurface conditions
around 10° C. Uh and uh here are some of
the pumps that we use for feeding that water.
results show that it takes a while for
that biofilm to develop. And we see the
first 100 days or so that that the the
carbon's being significantly consumed by
that effort. And we see that we have
significantly noisy conditions with some
selenium reduction. We're in we're
adding one milligram per liter. But um
you know in this case we've got uh still
probably only 40 or 50% removal um maybe
60. And so uh we see that we don't
really settle down in terms of selenium
removal until we get out to about 150
days. So it takes a little while for
things to develop and settle down. We
can see that that was true for nitrogen
as well. Although uh it achieved steady
state and strong removal sooner. Um, we
did have an upset condition in the
column tests at week 320. Uh, and uh, we
can see that it took the columns a
little while to reestablish their
equilibrium, but they did. And the same
is true for for both nitrate and
selenium. Um, we can see that carbon use
increased during that upset period. So,
we we relied less on the carbon when
things were at steady state and running
nice and smoothly once the bofilm had
Once we were done with these columns, we
decommissioned them and we sampled at
the influent base, the middle and the
top of the columns. And we used 16S uh
genomic analysis of the 16S um ribosomal
RNA gene to uh compare the relative
abundance of microbial communities. And
here we are showing um things organized
by phylm which is pretty high level but
um basically we're showing the blue um
aerobic heteratroes that are consuming
carbon decreasing as we move away from
the point of carbon injection which
makes sense. Likewise, the organisms
that specifically are using methanol
like methylotinera
um and and some of these others that
fall into the um proteobacteria are
declining somewhat as we move upward in
the column and we are increasing in
abundance of some of these sulfur
reducing bacteria the purple uh and uh dulphurosperinus.
dulphurosperinus.
So um we see a shift in community as a
result of the availability of nutrient.
That's what we would expect.
If we look at the full scale SRF, we see
changes in the concentration in the
field scale trial um over time. So here
we have days and um and
during startup and here we have the
selenium concentration in micrograms per
liter and nitrate in milligrams per
liter. This is a log scale. Uh so the
black is the influent concentration and
then these are averages from those
colored monitoring arrays that we showed
you in the aerial photograph. So we see
that after about 120 days we began to
see significant selenium reduction. We
did have that upset condition um and
upset condition at around 140 days and
uh we see a temporary um slowdown in
selenium removal. but then it picks back
up. Same with the nitrate. Uh so we can
see that the system's resilient that it
can be upset but once it's upset it can
also resume operations. Uh we see as we
did in the columns that the carbon was
used to build the biomass and again to
reestablish equilibrium during the upset
condition. In both cases, we used
broomemide to um as a tracer so that we
could understand that we were getting
all of the water back. Um and as you can
see, we did did pretty well. Um if we
look at this normalized to one, which
would mean we got everything back, you
can see that most of um the water was
bromide was recovered. Um having the
tracer allows us to calculate
concentrations at C over C kn. So the
effluent chemistry over the influent
chemistry as a ratio. And what we can
see even more clearly is that selenium
reduction, the nitrate reduction um and
the influence of those upset conditions
We were able to drill a hole, a sonic
hole into uh the
treatment zone. It took us a bit to get
actually into that zone because the
treatment was so effective. It happened
so quickly that we had to really get in
close to our injection wells and
initially we didn't want to disturb
them. So we stayed farther out and we
found that we weren't intercepting the
active treatment zone. And by that I
mean that that we actually had no
nitrate and no selenium left in
solution. Makes it hard to calculate a
rate. Here we can see that the water
table's at about 57 m below ground
surface. This is a a core a drill hole
that that we uh constructed um and
completed as a monitoring well but took
the the sonic core and sampled it. Uh so
we can see uh in blue where we have um
increasing mass of DNA per gram of
sediment and in yellow or orange where
we were not able to recover any DNA.
Above this zone at 60 61 m um we have a
vados zone or a zone where we have
variable unsaturated and saturated
conditions. So we don't have much
biomass there. But immediately into the
saturated zone and where the um majority
of the methanol would actually float
would rise within the system we do see
the highest uh values of of um nanograms
of DNA per gram of sediment as much asund
asund
um as much as 15,000 grams. And so
that's quite quite a a yield. And you
can see that then it begins to diminish
um as we increase with depth and we
leave the treatment zone um entirely and
get back into a zone where there was
really no DNA yield. The highest
microbial biomass coincided with the
predictive act predicted active
treatment zone based on modeling in that
62 to 67
m below ground surface range. When we
look at this from a diversity
standpoint, we can see that we had
higher diversity in the vados zone uh
where biogeeochemical conditions are
fluctuating and below it where things
are in their native state. But in the
treatment zone itself, because we're
applying a selection pressure, because
we're adding nutrient, we're actually
selecting for certain organisms and the
diversity decreased um in that zone.
When we compare the difference between
microbial community structure in each
sample, each of these is a sample depth
and the distance between points on this
NMDS plot is showing you the magnitude
of distance between
or variation between microbial
communities um in each sample. So here
we see that the active treatment zone
samples are all clustering together in
one area and quite different from the
zone beneath the treatment zone and
above in the vados zone. So
using this dissimilarity
um Bray Curtis dissimilarity index with
these tax data we were able to really
distinguish the three uh zones with
respect to the treatment area. And when
we looked more closely at a sort of
functional level, we lumped the samples
or the identified microbes together
based on um whether they were um
methanol oxidizing or nitrate reducing
or selenium reducing. So blue and green
and red or sulfate reducing. And then we
plotted those results as a function of
depth. And two things jump out. One is
that we saw an initial increase right at
the water table of the denitrifying and
methanol oxidizing organism suggesting
that they methanol oxidation is
supporting that nitrate reduction.
Shortly thereafter with depth as water
moves down we can see that we are
getting the highest abundance of
selenium reducers right at about 65 m
below ground surface. Um, interestingly,
we commonly see that selenium and
nitrate reduction overlap, but that
selenium lags just a bit behind that
nitrate. Uh, we see that sulfate
reduction was
uh most active in the zones immediately
above and below the treatment zone. And
that's because the oxygen gradients
there are the steepest. And we know that
that selenium or sorry sulfate reducing
bacteria really can't tolerate a lot of
oxygen. So they're finding niches in
these zones where the gradients steep um
to accomplish sulfate reduction. We did
see other organisms iron cycling
hydrocarbon degradation but we don't
have time to get into that today.
Instead of thinking about it though from
a depth perspective we can also think
about it from an injection point
perspective. And so here what we've done
is plotted the um similar analysis 16s
genomics from bioons which are little
steel cages that we hung in the flow
field in those monitoring wells. We did
this we changed them every 3 months over
a period of five years. We have quite a
nice data set here at the zero point is
the point of injection. We're showing
the relative abundance of organisms and
we've we've lumped them again. So we
have uh the aerobic heterotroes in blue
which spiked initially used up that
oxygen and then became less abundant.
Thereafter we begin to see the
denitrification kick in and we see that
it continues and that makes sense too
because the mass of nitrogen is so much
greater than the selenium which kicks in
a bit later as we noted on the depth
profile um but begins to tail off within
about 30 mters. So we're accomplishing
this treatment within a a short distance
of the point of injection. We do track
iron and sulfur because those are
indicators that we may over reduce the
system. We don't need to over reduce the
system. We simply need to have a a
suboxic condition. And so um we can also
here see those the influence of of those
steep oxygen gradients uh at the point
of injection and just below the
treatment zone. So, so here we're we're
laterally moving through the treatment
system, but it looks a lot like that
depth profile.
We did use metagenomics which allow us
to car characterize more than the the
probable organisms that are present and
to infer their function. rather this
lets us extract all of the DNA from a
sample and then to look at the specific
genes that are present um and understand
using a shotgun metagenomic approach um
assembly of of the reads that we get
into contigs and we bin those and we can
actually build um some some genomes
which are hidden beneath my my little um
picture here. So, so better technology
for understanding the genomic capacity
of these samples to do the work we're
same figure again with that drill hole
depth from 55 m below ground surface to
85. What I'm showing you here is that we
chose three samples in red, the red bars
from 62, 65 and 70 m depth um for this
metagenomic analysis. So this is a lot
of data. It takes a lot of time. Um you
can see that we got lots 10 tens 5 to 10
million reads from these samples. We
assembled somewhere between 53 and 62,000
62,000
contigs and we um we were able to get an
awful lot of information
at the the gene level which genes are
most abundant and indicating what
processes are are possible and um
perhaps happening. Uh and then um it it
also allowed us to assemble some of the
genomes themselves.
that also let us map the uh pathways
that are important to the
denitrification and selenium reduction.
We were more successful with this with
the nitrate and that's really because
more is known about dnitrification.
There's more information in the
literature about the enzymes involved.
So what we've done is shown here nitrate
being reduced to nitrite nitric oxide
nitrous oxide all the way to di
nitrogenic gas and then the bubble size
indicates the relative abundance of key
genes that we know are involved in
nitrate conversion to nitrite that's
these narens nitrite conversion to
nitric oxide that's these near genes and
so forth and then you see the incomplete
reduction leading to some ammonium accumulation
accumulation
um which is um we see a little bit of of
capacity for that in the system as well.
So this is really great because it lets
us map the system and it lets us
understand which genes are present and
maybe being put to work. If we were to
have an upset where we had a real
problem with with function, we'd be able
to look back at these genes and their
expression and understand what we've
lost and what we need to fix. So this is
a really useful tool.
We're less able to do this with selenate
reduction because there's so little
known about selenium reduction. Um the
selanate reductase enzymes that we do
know about come from an organism called
tower selinatus. Um there's three um
different genes s and c. They're very
similar to nitrate reductase um enzymes
known as nar GH and I. And so we, you
know, we we don't have as many of the
known genes to use in our mapping, but
we were able to use the patterns that
underly these um sequences, the sequence
patterns from these genes and build a
database that lets us understand a
little bit more about what may be
happening in our system. And so um what
we've done with these same three samples
that we've talked about um is looked for
a sequence um that is similar to the s A
gene, the s B gene and the s C. And then
these are the different elevations 62 m
65 and 70. So you can see we have quite
a few um organisms that appear to have
similar capacity uh
genes in organisms that are present that
seem to have capacity to accomplish this reduction.
I don't really have time to tell you
about our closure study today, but I do
want to tell you that we've put quite a
lot of effort into that because of
course once we stabilize the selenium,
we want it to remain stable so that it's
not an environmental problem going
forward. We've conducted a column study
of the SRF biommineralization that
accreted in the columns when we ran them
over a period of months and years. Um,
we simulated closure by turning off the
water and turning off the food, uh, the
carbon supply and the phosphoric acid.
And then we watched to see if we would
observe any release of nitrate
or nitrite or selenium. Now, I'll just
show you how we did this.
Every six months we would push one
syringe, you know, roughly 180 mls of
water into the column. And because it
was such a small volume, we weren't
mixing through the column. We were just
displacing the next 200 mil volume out
the top in essence. And so it took us a
few years to push most of that water
through. You can see in this shaded box
that's kind of the limit of the of of
that period of time when we were
displacing that initial pore volume. And
what we can see is there was some upset
at the time, not terribly significant,
but some um at the time that we stopped
the flow of nutrient and then it settled
back down. The other thing that's really
interesting here is that the sight water has
has
nitrate and selenium in it. And so we um
we can see that even though we're not
feeding these columns, there's
sufficient carbon and biomass being
recycled so that the the system is able
to continue reducing nitrate and also selenium.
selenium.
So we don't see reversal, we don't see
remobilization, and we do see ongoing
capacity to handle small amounts. So
let's say there was a storm and it
rinsed some selenium into that closed
SRF which is no longer fed. These data
suggest that this will continue to serve
as a robust and resilient sink for those elements.
elements.
This is the graphical abstract from our
paper in we published in Stoton this
year about the closure work. Um I guess
I'll just point to the minology work
that we've done. We have shown that the
mineral precipitation and sorption that
occurs. Um we see precipitation of of
selenosulfide complexes and we seeption
of selonite to the iron oxyhydroxide
minerals. Um we know that selonate won't
sorb but selonite will in this pH range.
And so um we showed not only did these
removal mechanisms occur but they also
were not reversible. We tried to reverse
them under changing oxygen and nitrate
exposure pH and we were we were
unsuccessful. So we're quite confident
that once this is re reduced and made
stable that these materials will remain
stable postclosure.
In conclusion, our successful um
selonate and nitrate reduction in column
and fieldcale uh saturated rock trials
has shown us that this is possible that
it's cost-effective that it's robust and
resilient and sustainable. We have shown
that selenium is accreted in two forms
absorbed and precipitated. Both forms
are stable under postclosure conditions.
And this technology has now become
commercial with a technology readiness
level of seven under review for level
eight in British Columbia.
These are some publications which we
would suggest you turn to with further
questions but I am happy to um speak
with any of you offline um you know by
phone or or online however would be
convenient. I'm sorry I wasn't able to
Click on any text or timestamp to jump to that moment in the video
Share:
Most transcripts ready in under 5 seconds
One-Click Copy125+ LanguagesSearch ContentJump to Timestamps
Paste YouTube URL
Enter any YouTube video link to get the full transcript
Transcript Extraction Form
Most transcripts ready in under 5 seconds
Get Our Chrome Extension
Get transcripts instantly without leaving YouTube. Install our Chrome extension for one-click access to any video's transcript directly on the watch page.
