Hello and welcome everyone. My name is Dane Menke. I am the digital marketing manager here at Regenesis and LandScience. Before we get started, I have just a few administrative items to cover. Since we’re trying to keep this under an hour, today’s presentation will be conducted with the audience audio settings on mute. This will minimize unwanted background noise from the large number of participants joining us today.

If the webinar or audio quality degrades, please try refreshing your browser. If that does not fix the issue, please disconnect and repeat the original login steps to rejoin the webcast. If you have a question, we encourage you to ask it using the question feature located on the webinar panel. We’ll collect your questions and do our best to answer them at the end of the presentation. If we do not address your question, we’ll make an effort to follow up with you after the webinar.

We are recording this webinar, and a link to the recording will be emailed to you once it is available. In order to continue to sponsor events that are of value and worthy of your time, we will be sending out a brief survey following the webinar to get your feedback.

Today’s webinar will discuss colloidal activated carbon performance and long-term strategies for in-situ PFAS remediation. With that, I’d like to introduce our presenter for today. We are pleased to have with us Dr. Grant Carey, President of Poor Water Solutions. Dr. Carey has more than 30 years of experience specializing in the characterization, modeling, and remediation of many impacted sites across North America. He also specializes in the development of innovative modeling and visualization tools for conducting PFAS forensic assessments and for evaluating long-term remediation strategies.

He’s currently involved with five CERDEP ESTCP projects, where his main research focus is PFAS Transport and Remediation. Dr. Carey is also an Adjunct Research Professor at Carleton University in Ottawa, Ontario. All right, that concludes our introduction, so now I will hand things over to Grant Carey to get us started.

Thanks, Dane. Hello, everyone. I appreciate everybody taking the time out today to join us. We do have a lot to cover today, but I think some of this will be helpful, so let’s get started.

So, I know most of you on the phone already know the immense challenges that we face with PFAS environmentally, because PFAS have been used in many commercial and industrial products since the 1950s. They are widespread in soil, water, and air. PFAS do bioaccumulate in humans and ecological receptors, and for that reason, it’s one of the reasons why they have very, very low cleanup goals. And just to pile onto the challenges, PFAS were actually engineered to resist degradation, which means we don’t have the same remedial options with PFAS that we do with chlorinated solvents and petroleum hydrocarbons for cleaning up sites.

One remedial option that’s gotten a lot of interest over the last few years for PFAS sites is the use of colloidal activated carbon, otherwise called CAC. This colloidal activated carbon consists of very small particles of activated carbon, usually one to two microns in diameter. Because they’re so small, they’re actually fairly easy to inject into aquifers. The way this works is if there is a PFAS plume and we want to inject this colloidal activated carbon permeable reactive barrier or PRB, we use temporary wells to inject a liquid solution that contains these small particles of activated carbon.

So once that carbon’s injected, PFAS will preferentially absorb to that carbon within the PRB. Essentially, it’s a way of filtering the PFAS out of groundwater. The groundwater that flows through the PRB will be essentially clean of PFAS. The radial diagram in the top right here shows what we observed at a monitoring well. This is a site where Aptum injected colloidal activated carbon at a Navy site. They actually have monitoring wells in the barrier within the PRB. This is one of those monitoring wells where we can see before the injection that red line represents the pre-injection concentrations where PFOS was around 3000 nanograms per liter, PFHXS was similar, and PFBS was just a little under 100 nanograms per liter.

We can see after the injection, that green data series, all the concentrations have dropped by orders of magnitude—three and a half orders of magnitude for PFOS, for example. This is an example of what happens after CAC is injected. Here with the diagram on the bottom right, this is for carboxylates. PFOA and PFHXA are similar to what we saw for the sulfonates, showing orders of magnitude reduction after CAC is injected.

What’s interesting here is that we don’t see the same reduction for PFBA and PFPEA, which are the shortest chain carboxylates shown here in the radial diagram. That’s not surprising because those short-chain carboxylates absorb, I’d say, the least strongly, to the CAC. If there are other things in water, like other higher PFAS concentrations or dissolved organic carbon, that will basically compete with these shortest-chain carboxylates and limit their adsorption.

With colloidal activated carbon, I am involved with five PSTCP projects, and four of them are actually focused specifically on the use of colloidal activated carbon for PFAS. I think the U.S. Department of Defense now has seven or eight projects at least involving colloidal activated carbon, because we have seen through a number of field sites where this has been implemented over the last seven or eight years that it does work. It’s really more a question of how long does it work for, for different PFAS of concern or different regulated PFAS.

My involvement with these projects is mainly modeling support, and that’s because about seven years ago, I developed a proprietary reactive transport model. I initially focused on chlorinated solvents about 20 years ago, but I’ve updated that since so it actually works with PFAS transport and simulates the effects of CAC injection. This allows us to evaluate what happens after CAC is injected and how long it takes for PFAS of concern, the regulated PFAS, to actually break through a PRB with colloidal activated carbon.

So today I’m going to talk to you about two studies I’ve done under one of the CERDA projects for the US Department of Defense. One project we’ve actually looked at specifically at source conditions at a site that’s been well characterized, South Dakota Air Force Base. I do have a paper that’s been published on that. At that site, we’re looking at how this remedy is going to work and where is the best place to actually put a CAC permeable reactive barrier at a site like that one.

The lower graphic is for a second site we’re going to talk about in this presentation, where we actually use modeling to look at the effects of tidal fluctuations and coastal geochemistry, which is a little more complicated than typical freshwater sites. We wanted to see how those factors actually affect the CAC performance when a PRB is put fairly close to the shore to prevent a PFAS plume from reaching the shore. At the end, I’ll talk a little bit about feasibility studies with a simple scenario where we can actually model not just the PRBs with colloidal activated carbon, but now with these simple modeling tools, we can actually start to look at integrated remedy components.

Does it make sense to do source containment at the same time we have a PRB, say at a downgradient battery? I’ll just show a simple example of how we can do that, and I’ll talk more specifically about what kinds of things you might want to think about if you’re considering a long-term remedy with colloidal activated carbon.

Since we’re talking about PFAS remediation with colloidal activated carbon, I think it’s important we talk about the first site, I think, in the world where in-situ remediation was actually successfully done. This is a site that Rick McGregor worked on. He’s with in-situ remediation services. It’s a central Canada site where there were hydrocarbons present, and Rick had also found that there were low levels of PFAS and PFOA that were present before remediation in the two to three microgram per liter range.

At this site, Rick injected Plume Stop, which is colloidal activated carbon, to both go after the hydrocarbons and also to absorb the PFAS, trying to mitigate the plume that was present at the site. We can see the top graph is what it looked like before remediation with the PFAS plume. After 180 days, that’s just a model result showing, basically showing what the field data show us: that after injection, PFOS and PFOA, and even all the short-chain compounds that were monitored since the injection, have all been non-detected for the last eight years.

One thing to note here is that this is an example of source treatment where the source area is about 400 square feet, and the actual CAC zone after injection was about 3,000 square feet in area. So this is a fairly small site with fairly low concentrations, and here source treatment worked quite well. We’ll look at another example in the next site where there is a much bigger source zone with higher concentrations, where source treatment would be a bigger challenge.

So we did this paper in 2019, and it’s shown with modeling what we’ve actually seen with the field data since then: that PFOS, PFOA, all the short chains are still non-detect after injection. What we’re seeing in the model is that in the source area, the model still represents an incoming PFAS flux into that source area, which could be through percolation from the vadose zone that lies above the aquifer, or it could be back diffusion from a clay layer below that source area. We’re seeing that with that flux continuing to be turned on in the model, even with CAC injected, eventually concentrations of the PFAS do start to increase as the available adsorption sites on the CAC start to get used up.

As more and more CAC is used up, we’re seeing after 10 years, 20 years, concentrations in the source area of PFAS are starting to increase. It hasn’t gone very far yet because we have a lot of CAC downgradient from the source area, which hasn’t been used up. That’s like a buffer that Rick built into the CAC zone. If he didn’t have that buffer and he only injected CAC at the source, they probably would have had breakthrough after say between 10 and 20 years. But because there was some extra room at this small site, it didn’t cost much extra to do, and they got an extra few decades before breakthrough was actually realized. So that’s just something to keep in mind; some people call them runways of CAC—basically give yourself some extra buffer so you’ve got greater longevity.

Let’s talk about the South Dakota site. There was a paper published late last year in the remediation journal that’s open access, but if you’d like a copy, I’m happy to send it to you; just send me an email and I’ll forward you the PDF. You can see my co-authors on the paper here. What we had done with this site is, because the source area was so well characterized, we wanted to use that as real-world conditions. We built a model around that, and the site actually has quite complex hydrogeology. We didn’t put all that in detail in the model; we brought in kind of the average velocity, represented the fraction of organic carbon as what they measured at the site, but the flow system is hypothetical.

What we wanted to do was just look at whether a CAC remedy would actually work at this site because PFOA and other PFAS concentrations are much higher than the site we just saw. The key questions, as I mentioned, are: will CAC actually work? Is it a viable technology for this site given the high concentrations and the placement for putting a CAC-permeable reactive barrier? Is it at the downgradient boundary? Is it right downgradient of the source area? Maybe in between? Maybe a couple of PRBs? We just did some modeling to kind of evaluate the pros and cons of each alternative.

Through this, this remedy is being implemented so quickly because there is an urgent need to mitigate PFAS at some sites. There is a need for more information on how to characterize sites, how to do these feasibility studies, and that’s a lot of what we talked about in that paper—just suggestions and tips for what to think about if you’re doing remedial investigations or feasibility studies involving a potential CAC remedy.

As part of the modeling, the key is that with any modeling of PFAS, we need to know what the PFAS adsorption isotherms with CAC are: how strongly does PFAS adsorb? I use what’s called a Freundlich isotherm; I won’t get into the details of that. I wanted to mention I have been working, actually since 2017, with Dr. Ayn Fahm and Saifullah Hakimabadi, who’s a PhD candidate—pretty much done now—at the University of Waterloo, and they’ve been studying many different factors that affect PFAS absorption to CAC.

One thing we did, which has been really helpful for me as someone who needs input parameters for models, is we took a groundwater sample that Dr. Charles Schaeffer had sent to us from an AFFF-impacted site. It’s a groundwater sample from a plume, and they actually calculated the individual adsorption properties for the main PFAS constituents: PFOS, 6-2-FDS, PFH-XS, PFOA. That’s for the groundwater sample with the GW symbol. The Y-axis is the adsorb concentration, and the X-axis is the aqueous or the groundwater concentration.

What we can see is if you have only PFOS in a water sample and you put CAC in that sample, you get a lot more PFAS adsorbing to the CAC than if you actually have a groundwater mixture with other PFAS and with dissolved organic matter. So the PFAS adsorption for the groundwater sample has lower adsorbed concentrations than for what we call a single species PFAS only in water. So that’s key: we are actually using a groundwater sample to tell us, with all these in the water at field sites, what are the absorption properties we should be using in these models to try and predict longevity.

This paper did have a summary, actually, of 17 field sites where CAC had been implemented. It talked about initial concentrations, what were the results, and we showed with these sites that CAC was actually very effective at remediating PFAS, and the big goal was to reduce mass flux at these sites.

In terms of the South Dakota site, what it looks like. We’re showing a one-layer model in two views. There’s the plan view here showing the grid. You can see the source area here, which actually at the site was quite large. The source area was 30,000 square feet, which is about 10 times larger than that source area at the Canadian site. PFOA concentrations are about 100 times higher. So this is a much bigger site with higher concentrations. The top graph actually shows a cross-section. What’s interesting here is a lot of good work had been done at this site. One thing that had been done was Dr. Hunter Anderson, who’s with the Air Force, had worked closely with Wood, which is now WSP, doing a lysimetry study to characterize how much PFAS are actually vertically migrating downward from the VADO zone into the water table.

So we have a mass discharge measurement of 17 grams per year, PFOA, moving vertically down to the water table from the vadose zone. We know based on the data in groundwater wells around that source area, there’s approximately 270 grams per year PFOA leaving the source area. So, 270 grams per year total, only 17 coming from the vadose zone, about 6% of the total. At this site, most of the mass, not just for PFOA but the other PFAS constituents as well, most of the mass is already below the water table in terms of where the source is located.

What we believe is that a lot of that mass has probably diffused into silts and clays that are below that source area, and it’s just slowly diffusing out over time. That’s actually similar to what we’ve seen at some other sites. There are some really good papers out there, one by Dr. Dave Adamson, that characterized the distribution at an AFFF site that showed something similar with forward and back diffusion.

Okay, so we looked at three alternatives with a model, and what we’re looking at actually is the pre-remediation modeled PFOA plume. This is a very simple model. We took the 270 grams per year and just kept that constant for a 50-year period to try and just create a plume, a hypothetical plume, that we could use to actually overlay different remedial alternatives and see how that affects the plume and the transport.

So, this plume is about 1,800 feet downgradient of the source after, let’s say, in 2024. So, we’re looking at what remedial alternatives we might implement. The first alternative we looked at was actually a downgradient PRB, just a little upgrading of the property boundary. If you notice, it’s ahead of where the plume is in 2024. So, at this point, this is a site that’s lucky enough; the PFAS plume hasn’t gotten to the downgradient boundary. That’s not always going to be the case, but in this case, it gives us an option to look at what happens if we put a PRB ahead of an expanding plume and where concentrations will actually increase over time.

So that’s the first PRB we looked at. Then we looked at the second alternative, which included that downgradient PRB and also added in a mid-plume PRB. This was basically to say if we can stop a lot of the mass coming from the source area in the mid-plume PRB, will that increase the longevity of the CAC in the downgradient PRB because it’s going to get less mass and lower concentrations over time? So is there a benefit to actually having two PRBs at this particular site?

Then we also looked at a source area grid, which was just blanket source grid injection of CAC. Now, this is a big source area. The actual area we looked at in the model was 60,000 square feet. I actually think that’s cost prohibitive. I don’t think anybody would actually do this because it’s such a big area. But we just did it in this study to see how that affects the plume migration downgradient and how long before that plume flushes out.

So the first alternative, just the downgradient PRB, is showing the model results with the plume at seven years after CAC injection. They injected CAC here in the PRB. What we’re seeing is that plume kept moving down towards that PRB. We’ve got fairly high concentrations over 30 micrograms per liter now in the middle of the downgradient PRB, and that’s caused eventually a mass breakthrough. The breakthrough time was at about 66 years for PFOA to break through.

So with this site, one suggestion we made in the paper was if you’re looking at CAC as a long-term remedy, the suggestion would be to make sure you design it to get at least 30 to 40 years of longevity. If you get more than that, that’s great. But if you can get at least 30 or 40 years, any future injections that have to be done if you’re using that present value are actually fairly cost-effective when they’re much further ahead in the future.

This is a little complicated, but I’ll just show, because I did find the results interesting. This is looking at the concentrations across the width of that plume. So the center of the PRB is at this Y-axis, 345 meters. Going to the left, that’s actually going north along the PRB, and going to the right, that’s going south in the PRB. We can see that the PFOA concentrations actually reached the MCL of 0.004 micrograms per liter in about five years after the PRB was installed in the model. Over time, those concentrations kept going up and up because it’s an expanding plume.

So early on, there wasn’t much movement through the PRB, but as concentrations continue to increase, it started to move faster through the PRB. Over the 60 or 70 year simulation, concentrations increase by five orders of magnitude. The lesson here is if you are going to put a PRB in front of an expanding plume that’s not stable, just make sure you’re predicting what the future concentrations are going to look like. If they’re going to be big increases, that affects how much CAC dose you want to give to make sure you get the longevity you need.

Okay, and then the second alternative was actually looking at two PRBs combined simultaneously, the mid-plume PRB and the downgraded PRB, which we already saw, and the question was, will that mid-plume PRB actually increase the longevity of the downgraded PRB? The simple answer is no. We had 66 years breakthrough time with only the downgraded PRB, and we had 68 years with the mid-plume PRB installed. I’ll show you a graph in a minute that explains why.

But basically, what we’re seeing is even with the mid-plume PRB, when that was installed, there were already high concentrations of PFAS downgradient of the mid-plume PRB. Over time, we’re seeing fairly slow velocities in this simulation, a high amount of PFAS absorbed to natural organic matter, 0.25% FOC. That means we’ve got a long flushing time before those concentrations really start to go down.

So the orange line shows the PFOA concentration at the downgraded PRB when only the downgraded PRB is used. When we add the mid-plume PRB, the concentrations actually do start to come down after 50 to 60 years, but that’s too late. Most of the plume has already gone through the downgraded PRB, and that’s why we really didn’t see an effect on breakthrough time. So there was really no benefit in this case to adding a mid-plume PRB.

And just to show you the results, if we were doing source treatment or even containment, let’s say we put up a PRB just at the gradient of the source area, because the velocity is slow, we’ve got high concentrations, not only in groundwater but also adsorbed to natural organic matter throughout that plume. Because there’s high adsorbed concentrations, it takes a long time to desorb and get flushed out with a relatively slow velocity in this site example.

So that just shows if we’re using, say, PRBs, if you were thinking of pump and treat, it’s really important that we consider how fast PFAS actually desorb downgradient of where we’re implementing that remedy. Pump and treat’s going to be the same thing. If there’s a plume downgradient, we’re going to see desorption and potentially slow flushing. So that’s something to keep in mind, whether it’s from back diffusion or just equilibrium desorption or limited desorption; it may take some time to get flushing of the PFAS from downgradient of where we implement the remedy.

So we learned a lot from this work in terms of what we really need to know when characterizing sites if we want to evaluate the feasibility of a CAC remedy. I think the biggest thing that I’ve learned is the longevity of these remedies is all about the PFAS mass flux. It’s not concentrations only; it’s not velocity only; it’s the mass flux, which is a combination of those two.

So this is where it’s really important if you can get the data to actually collect depth-specific mass flux measurements where you’re thinking of putting a PRB. You can do it in the core of the plume. If you have a long PRB, you probably want to do it in different places along that PRB. The lower the concentrations over time, the less dose you can use and the more of the fringe parts of those PRBs. Regenesis does have a flux tracer tool, so if you’re interested, I’d certainly suggest talking to Regenesis. If you’re looking at CAC, this is something that I’ve seen just from doing the modeling is really critical.

I won’t go through all the details just because we have a lot to talk about today, but obviously a lot of this is to make sure you know how the plume is moving, where it’s moving over time, what’s competing with PFAS for adsorption to CAC, and even redox conditions. Is it aerobic where you might have precursors degrading to your PFAS if concerned? That’s important to know, and you wanna make sure you’re not changing redox conditions and potentially creating more problem. This is actually a result when we’re talking about desorption downgrading of a PRB.

This is actually a field result from a different site. This is the same site I mentioned earlier where APTM injected CAC and it’s working with the Navy to evaluate performance. And I’m helping them in terms of modeling what the performance was, calibrating desorption isotherms. So just simply the groundwater is flowing left to right across this page. The PRB is here. There’s two downgrading wells. MW1 is five feet downgrading to the PRB, MW2 is 10 feet downgrading to the PRB. And you can see the symbols here represent the observed data at these wells in the shallow aquifer. And you can see kind of a slow gradual decline over time.

If I use an equilibrium desorption model, we get very rapid drops here in PFAS concentrations, completely different from the South Dakota site because velocity is a lot faster at this site. We’re starting with much lower concentrations of PFAS in the aquifer before the PRB is putting in and the FOC is a little bit less as well. So very different conditions in South Dakota site. And to actually match these data, I had to model rate-limited desorption, which means there’s some kind of a mass transfer resistance, maybe PFAS have to diffuse a little bit before they’re desorbing from the natural organic matter.

There could also be the water table at the site has been going up and down. So that might be another thing that’s causing a slow gradual decline. I’m not sure if kinetics are involved in desorption at all sites, that’s something we need more information on, but just something to be aware of. If you’re having really slow desorption, that might be one reason why is there’s a kinetic desorption effect, the rate-limited desorption occurring.

Okay, so let’s talk about the second site. This is another coastal site. I call it Coastal Site B. I’ve worked on, we can see the authors here, and the reason we wanted to work at this site is if where PRBs might be installed, let’s say next to a coast, to try and prevent a PFAS plume from getting into the shore, how challenging is that going to be and how do tidal fluctuations in geochemistry affect the CAC performance? So conceptually, without CAC, without a PRB, just because of tidal mixing, we have pretty good attenuation at the shoreline, and using a model that’s representing the effect of tidal fluctuations, we’ve got a hypothetical source built in, upgrading of the shore.

With a model, we simulated, okay, there’s about a 50% attenuation between the source and really mainly in the area pretty close to the shore, and I’ve worked on a lot of coastal sites. That’s pretty consistent with what I’ve seen with field data at other sites and with modeling other sites, usually a 50%, even a 70% drop at the shoreline. Now with a PRB being installed, let’s say 50 feet from the shore, which some sites are actually having that implemented. So you’ve got fairly strong tidal fluctuations. So at high tide, the water in the shore is high. You’ve got inward flow towards the PRB from right to left. And at low tide, you’ve got the opposite direction. You’ve got outward flow towards the shore. And after the PRB is put in, we know we’re going to have high PFAS in between the PRB and the shore early on before it’s flushed out.

So at high tide, there is going to be some PFAS because there’s an inward gradient being pushed into what I call the dam gradient end of the PRB, the right side here. In addition, there’s also PFAS coming into the left side of the PRB. So the concentration profile might look something like this, big drops of concentration of the or B, but at that right side, the downgradient end, we’ve got some PFAS accumulation. That could eventually become a secondary source. We need to be aware of that. So we wanted to see how high does that get with tidal fluctuations. And with surface water, usually we’re talking about ecological receptors much higher than MCLs.

The good news is this reverse tidal fluctuation effect didn’t get concentrations high enough to be an issue for ecological criteria, so that was good. What we did see is in terms of geochemistry in coastal sites, there’s a lot more PFAS and ZORB to natural organic matter next to the shore because there’s high ionic strength and there’s high calcium and magnesium in the water. And that just means high ZORB concentrations means longer time to flush that out after the PRBs installed. Basically what this is showing, this is data from a 2012 paper, I just plotted it this graph. And we had this case where the aquifer at this particular site, we’re using the physical hydrogeology characteristics from the site and the geochemistry, those are measured.

And then we’re overlaying a hypothetical PFAS source and a hypothetical PRB just to help us understand under these coastal conditions what’s performance going to look like. So at a site like ours where This is the actual measured geochemistry near the shore. Calcium ionic strength is pretty high. And using this data from a paper, that means the amount of PFAS absorbed to natural organic matter might be four times higher than what we would calculate based on FOC alone.

To set up a 1D model, flow is going across the model domain from left to right. We’ve got this hypothetical source that we’ve overlaid on, a hypothetical PRB. We’re actually using the characterized hydrogeology characteristics from the site and the tidal characteristics to model the tidal fluctuations. And we’re putting in a PFOA source just using the same concentration we had at the South Dakota site, 300 microgram per litre, which means without a PRB at the shoreline, there would be about 150 microgram per litre because of that tidal mixing that occurs near the shore. We modeled the tidal fluctuations. I think that’s enough said.

This was interesting. I wanted to see how close do you have to be to the shore to have an inward velocity a couple times a day, because we have diurnal tidal fluctuations. So the horizontal line is a velocity of zero, and what we can see, the purple line is the maximum velocity at any point in time during the tidal cycle, the blue line is the minimum. When that blue line is below zero, at a distance here of 40 meters inland, that’s where we start to see inward flow. Negative velocity means inward flow, positive velocity is outward flow. So what we’re seeing is if we put a PRB 50 feet from the shore, we’re going to have pretty strong inward velocities a couple times a day, bringing that PFAS in from that stretch of land between the PRB and the shore, bringing that into that PRB boundary.

So in terms of results with this model, the symbols here above the charts show the well locations. So, the chart on the left is for an observation well in the model that’s at the downgraded boundary of that PRB. That’s where we know PFAS is going to go up a little bit because it’s coming with inward flow, it’s coming into that boundary. And the second chart on the right is for an observation well in the model basically placed at the shore. So, on the chart on the left, we can see initially initial concentrations are up here, 300 microgram per litre before CAC injection. the model actually simulates the CAC injection and the drop of PFAS that occurs. And we can see PFOS and PFOA do come down quite a bit and they don’t come down to the MCL in this case because at the coastal site, there’s a lot more mass adsorbed to natural organic matter before CAC is injected.

So there’s a lot of mass to start with once CAC is injected and that takes up some of the available sites. If this wasn’t a geocoastal site, we would have seen these concentrations come down below the actual MCL net. So what we’re seeing is both PFOA and PFOS came down, they went up a little bit, and that’s because with inward flow into that boundary carrying PFAS, that increased both PFOS and PFOA concentrations at the downgradient boundary, then they started coming down a bit as the PFAS were flushed out between the PRB and the shore. Eventually, PFOA starts to go up pretty quickly because it breaks through after 20 years. So we have, with this particular scenario that we modeled, we’d have a 20-year longevity for PFOA, much longer for PFOS because it absorbs so much more strongly to CAC than PFOA does.

Now at the downgradient well next to the shore, we actually see something pretty different. Here we see pretty quick orders of magnitude PFOA declined over that 20-year period before it breaks through. So, what we’d have to do actually is re-inject CAC before that 20-year period is hit to make sure this stays down. But we see PFAS is a much slower decline, and that’s because in this model, we’re modeling equilibrium desorption with high sorb concentrations to start with. It just takes a long time for that PFAS to be flushed out because it adsorbs so strongly to natural organic matter to begin with.

So that may be conservative. the actual PFAS decline might be a fair bit faster than what we’re modeling. A lot of these marine site studies have shown that PFAS maybe 50% of it may be irreversibly absorbed to the natural organic matter, in which case it’s not coming off. And then we will see concentrations come down more quickly.

So, okay, let’s talk about, well, how does this remedy work in the long term? I’ve talked to a lot of people about implementing CAC and I’ll talk through some of the main questions I get, and my perspective in terms of how I think about these things. So first let’s look at a simple, let’s do a really quick feasibility study check, looking not only at a CAC PRB, but also saying, well, does it help to do something to reduce the mass discharge from the source if we’re putting in a PRB next to the downgrading property boundary?

This is a different site than the South Dakota one. It’s got a faster velocity, it’s got lower mass discharge, lower concentration. and so a very different condition. So the mass discharge in this case is about 70 grams per year. Concentration at the downgradient boundary in the steady state plume before remediation, it’s about 10 microgram per liter, and velocity of about 200 feet per year in this case. So the alternatives I looked at were, okay, let’s just look at a downgradient PRB. Let’s see how that performs, and then we’ll compare these other alternatives where we try and do something to reduce the PFAS mass flux coming from the source area.

One obvious one I didn’t model, because we kind of looked at that before with South Dakota, was we could have looked at a second PRB, but I wanted to try a couple of different things here. So I tried a partial wall scenario where groundwater has to flow around, but not through the source area. The thinking being that less groundwater flow through the source, you would think that would reduce the mass discharge. That’s actually, after modeling, that wasn’t the case. I’ll show you why that is. And then I looked at, well, what if you cover the source and prevent the infiltration? That’s gonna reduce, let’s assume in this scenario, 50% of the mass, PFAS mass is coming from the VADO cell. So a cover would reduce the mass discharge by 50%.

How does that affect things? And how does that increase the longevity of that downgrade in PRB? And then finally, we’ll look at a full wall scenario, but you can’t just put a full wall around a source area. You also have to put a cover on. That way the water inside the wall doesn’t kind of build up and eventually come out through the ground surface. So it’s a cover plus full wall type scenario. So this shows the plumes at 30 years after start of remediation. So the top one is the PRB only, and you can see the higher concentrations, one to 10 microgram per liter, kind of getting through close to the downgrading edge of the PRB that actually breaks through four years from the simulation at 34 years. And we can see with the partial wall scenario, There’s a lot of red here, which now that I look at it, it’s intuitively obvious, because we have much slower velocity inside the source area. But we still have infiltration bringing mass to the water table. We still have back diffusion. We just have less water flow. It’s more stagnant water. And that results in much higher concentrations within the source area inside this partial wall.

So the effect of trying to reduce the groundwater flow actually resulted in much higher concentrations and actually ended up, it’s about the same mass discharge at the end of the day versus nothing being done in the source. So that partial wall scenario, at least this configuration didn’t work well. I didn’t try and optimize and try and look for other options, but that’s the kind of thing that we can use simple models for. And then finally, the PRB with a cover only over the source and PRB down here, That did cut the mass discharge by 50%. We still have a plume getting to the PRB and it increased the longevity by about 20 years. And then finally the full wall and cover at the source, basically the mass discharge. There’s a little bit in the bottle that’s seeping through the wall.

In reality, these walls have a finite lifetime. The model might be having some, might be some dispersion artifacts. I have to check that. But I do expect even with the wall around the site, we are gonna have fairly high concentrations left inside and there are gonna be places where these walls are not gonna get a perfect seal, where you’re gonna get some mass coming through. So I think this is actually representative. We might even have more mass coming through with a wall scenario. So what we’re seeing is all four alternatives, they all look the same, the gradient of the PRB. So these extra source treatment options to reduce mass discharge didn’t affect things beyond the property boundaries. So yes, it does extend the longevity of the PRB, but these alternatives two through four are upfront costs.

The other alternative is you go with a shorter longevity, still pretty good at 30 years, and then just re-inject new CAC before that breakthrough actually occurs, another PRB. And that way it’s in the future, net present value is pretty low cost if we look at it with net present value. In this case, that’s probably the best benefit. Assuming there’s no receptors on the site, there’s no extra benefit here by trying to do something with the source. The key is that downgradient PRB to prevent PFAS from moving offsite. And just to show you what concentrations look like over time, the purple triangles are just an observation well put in the model.

I should have mentioned as well. So when I modeled mass discharge from a source over say 50 years or 100 years, I actually represent a decline half-life of typically 30 years. I had done that 2019 paper I talked about shows that with back diffusion, we still get a decline. It’s a slow decline, but if that’s the main driver for source mass discharge, that declines at a half-life of about 30 years. That’s why the concentrations do go down over time for the PRB-only scenario, and that’s a 30-year half-life. And then they drop more quickly by 50% when we put a cover on, but they still stay elevated well above MCLs by a thousand plus times the MCL. And the purple one does drop pretty quickly with the full wall, but we still have mass getting down gradient just because some walls may leak and that really depends on what kind of wall you build and how well you sustain it.

So in terms of that partial wall, this just shows, so initially concentrations at that well were in the yellow one to 10 microgram per litre range. Over time, we got higher and higher concentrations in the source area because less dilution was occurring and that created a more intense plume moving towards the PRB. So the key is whatever we do we want to make sure we’re not making concentrations higher in the source area.

So in terms of the biggest question I get, the most common question I get is what happens with these PRBs 10, 20, 30 years down the road? These are effective for adsorbing PFAS but that means PFAS are adsorbing to the CAC and it’s building up in that barrier. So what happens down the road? Well, option one is we could actually inject another PRB just slightly downgradient, get another 30-40 years of longevity. For me personally, I’m quite confident that in the next 10 to 20 years, we’ll probably have destructive technologies available for in situ PFAS treatment. I’m an optimist, I guess. But even now, there are thermal technologies that are being tested, which might actually work pretty well for treating PFAS in a CAC zone, when that CAC is close to being spent. So we can inject down a gradient and treat in the CAC zone that’s been replaced.

I think that’s where we’ll probably be, but at the end of the day, I see this as a way to slow down what’s leaving the site and give us time to catch up and come up with potentially more destructive technologies. And the key that I think is important for feasibility studies is looking at net present value. That really opens up a lot of options about do you spend more upfront or do you spend enough now to get good longevity for 20, 30 years and then spend more later when the net present value is actually less? That’s something to consider for feasibility studies.

Okay, and then we do have as of last month, we now have final EPA MCLs for PFAS. Sure, everybody’s aware of it, I’ll just go over. So there is one compound HFPODA, that’s from the GenX class. Most AFFF impacted sites don’t have that. So I typically, I’m just gonna talk about the four main ones we see at AFFF impacted sites. Obviously these MCLs are quite low. Background conditions at sites might actually be higher than these MCLs. So that’s one thing is all these sites, we’re gonna need to characterize background pretty well. And the key is these four with MCLs are all long chain PFAS. They all adsorb pretty well to CAC. The one that doesn’t, that initially in the proposed MCLs was a bigger deal was PFBS. That’s a short chain. It still absorbs fairly well. There’s one site I’m working on where at least that in barrier monitoring we’re doing, PFBS is still held up in the barrier, hasn’t broken through.

I think PFBS probably will be the first to break through compared to these others because it’s a shorter chain. It just absorbs less strongly than the others, but this also tends to be lower in concentration and it has a much higher health-based water concentration. The key thing is for hazard index, for PFBS to even be relevant when we’re looking at remedial goals for groundwater, this hazard index is only if you have a mixture of two or more PFAS of these four PFAS present. If we have PFNA and PFHXS being held up in a CAC barrier and they’re not breaking through, even if PFBS is breaking through, as long as we don’t have PFNA and PFHXS at the degrading end of the boundary, then that’s okay. It’s when you have a mixture present where PFBS becomes part of the mix in the hazard index calculation. And it’s still early days. To be honest, I’m not exactly sure how this will be implemented, but I think it’s an important change that EPA made compared to the proposed MCLs last year where it wasn’t based on the mixture, it was based on any one of these compounds being present. So that’s just important to know, I think. I’m going to go past this.

If PFBS is an issue, there are ways. I think the long-chain compounds, competitive absorption modeling is not as critical because they absorb pretty well already. PFBS is short-chain. It actually does have enhanced adsorption downgradient of where the long-chain compounds get held up in a barrier. So you can model that and you can model the more realistic longevity. PFPS is important, but hopefully because of the Hazard Index Calculation, that’s not going to be an issue at a lot of sites. So in terms of modeling CAC performance, the key for me, I actually learned a lot just from running the models. Every site I work on, because this is such a new way of remediating, I’m still learning new things. And they’re great teaching tools. These visual aids we can put together, if someone’s not as sure how this works, maybe they don’t understand it as well. This creates the graphics we need to help support that. And we can use models to support feasibility studies, remedial design, performance assessments, even to evaluate remedial action objectives. If we’re trying to quantify what the attenuation is going to be, say, downgradient of a PRB at a site boundary and a downgradient receptor that might be thousands of feet downgradient, modeling tools might be useful for that kind of thing in some cases.

Okay, so Pore Water Solutions, if you are interested in some of the graphics I showed with Radial Diagrams, we do have a tool that does that, this poorwater.com PFAS HTML page. We are starting to put together e-learning courses, including a number on PFAS, PFAS Transport and Remediation. We have a questionnaire available. If you’re interested at all, I’d really appreciate if you could fill out that questionnaire, so we learn what the industry needs as we put these courses together. And on that note, I will say Poor Water Solutions is a proud sponsor of the Battelle Conference coming up in Denver. We’re at booth number 244. So if you’re there, please come by. I’d love to have a chat with you. And if you have any questions, if you need any information, if you want a copy of the slide deck, feel free to email me. I’m happy to send that out. And my phone number is there as well, if anybody’s interested. And on that note, Dan, I’ll end there and see if anybody has any questions. All right. Thank you very much, Grant.