Chris Lacey   3:04

Everyone, we'll be getting started here in just about a minute. Thank you. Hey there, everybody. Good afternoon. Welcome. Thank you for joining us today. My name is Chris Lacey, and I'm a technical marketing manager at USALCO and the moderator for the panel today. We have enabled the Q&A section in Teams for you to send in any questions, and we'd like this to be a conversation. Please send any questions that you have in through the Q&A. and we'll pivot to those as fast as we can. There will be a handful of times throughout this discussion and where we're going to ask you for input and ask you if you have anything to add or if you have any questions pertaining. And those are great, great times also to jump in. And then lastly, at the very end too, we'll have some more time. So But again, feel free to add questions and comments anytime. This is meant to be a discussion. So a couple of poll questions that we wanted to start us off on before we begin. So let me go ahead and get those going. Let's do. Here we are. So, the first one is... Do you or your team conduct physical jar testing every week to support operations? Yes or no? I'll give everybody a little bit of time to answer that one before I pop up the other one. We just have two questions. We wanted to kind of gather some information about the crowd. All right, looks like it's fairly even split so far. 55% yes, 45% no. Let's see. This is the polling's kind of new for us on Teams. So, oh, there we go. We got more responses come in. I'll give it another few seconds and we'll pop on with the other one. The. So it looks like the majority do jar test week to week. So let's go ahead and add our next poll question. And this one is what type of water do you or your team typically focus on treating?

Chris Miller   7:25

I don't see it, Chris. Yeah, it comes. Yep.

Chris Lacey   7:28

It takes, yeah, this one takes a little while.

Chris Miller   7:30

Matt.

Chris Lacey   7:34

We'll give you guys a little while to respond to that. So we have drinking water, municipal wastewater, and just any type of industrial process water. It's like we have a lot of drinking water. Folks here. So the vast majority, and then it's pretty evenly split on the other side. All right. Well, great. So that's good information there. So I'll go ahead and close these polls out so they're not hanging around. Thank you all for participating in that. So let's go ahead and jump into introducing our panel today. So Chris Miller is Vice President of Digital Innovation at USALCO. He leads initiatives that help customers leverage technology, data, and digital tools to improve operational performance. Chris has dedicated his career to developing innovative solutions for water and wastewater treatment. with technologies and methodologies adopted by utilities and industrial customers across the industry. Chris lives near Pittsburgh, PA with his family. And then we also have Brad Spangler here today. He's our vice president of sales at USALCO. He has over 25 years of industry experience, including water, wastewater, pulp and paper, and other industrial applications. He currently leads the commercial efforts, which include a group of highly technical sales personnel. This group prides themselves on solving customer problems and addressing industry challenges. Brad resides in the Baton Rouge, Louisiana area with his family. All right, so let's go ahead and dive in real quick. So let's get started with why jar testing still matters. This first question is for Brad. Brad, for anyone who hasn't run a jar test recently, remind us why it's still the gold standard and what an operator can learn from the bench that's genuinely hard to get anywhere else.

Brad Spangler   9:39

Thank you, Chris. It sounds like a lot of people on the car on the call are actively jar testing. So, but again, I think what we see is people are jar testing still today to get an idea of where to set the coagulant dose. You see many municipalities that will run a curve with their current coagulant. a variety of doses starting from low to high. And, you know, their influent water changes. And so whether they're coming in for the morning and trying to determine where to set the coagulant for the day, a good jar test and a curve will help them do that. Other municipalities have the luxury of sometimes they are able to pull water upstream of their municipality. So let's say you've got a big rain event and and you know there's some highly high turbidity water coming your way. I've seen some municipalities that are able to say test water that they know is going to hit their plant in one to two days. That's a luxury. Not everyone has that, but There's no surprises there, right? They can run a quick curve instead of upsetting the plant, they can run this jar test and get ahead of any upset conditions. So I guess in summary, what we see is people are doing that to set the coagulant dose. As you know, you have to have the jar test protocol correct to get any meaningful data out of it. You got to have your mixed energy correct. you got to have your mixing time correct. So long answer, but I think to set the coagulant dose is probably what we still see as the primary reason.

Chris Lacey   11:17

Awesome. Thank you, sir. And then for, I have a question for Chris. You've spent your career in water treatment research. What makes the physical jar test such a durable, trusted tool? And what got you interested in building on top of it with machine learning and multi-objective optimization?

Chris Miller   11:32

Yeah, thanks, Chris. Great to be here and great to see so many people today on the call. So looking forward to some of the questions and input as we get rolling. So prior to joining USALCO, I was a civil engineering faculty member at the University of Akron, interested in drinking water treatment research. I noticed a few things along the way. So kind of while the physical jar test was the gold standard for those that were looking at how do you calibrate and get some additional sort of data analytics information from that, there was a paper published in 1999 by Mark Edwards from Virginia Tech. You may have heard his name before because he's also an expert in lead and was involved with the Flint, Michigan testing. But Mark Edwards published a paper about how do you actually calibrate for different coagulants. His paper looked at alum and ferric, kind of your, you know, your core commodity products. And So looked at that. There was then a paper published about five years later that brought in, we all know that the nature of the TOC can impact treatment performance. So Edwards kind of approached looking at how do we bring a coagulant piece into this and then this paper by Castle and some others five years later, was how do you bring in, you know, the TOC and the nature of the TOC into that? And so nothing had changed then since like 2004, looking at that. And as I started to consider additional chemicals impacts of pre-oxidants, powdered active A beta carbon for TOC removal. You know, how are we going to actually not only kind of test the coagulant in traditional TOC and turbidity removal, but how can we also bring in the other things that are impacting both the turbidity performance as well as the TOC performance? So Uh, you know, how, how could we actually then maybe bring in some simulation tools and saw a gap there?

Chris Lacey   13:41

Awesome. Thank you, sir. So let's discuss the best practices now, insights and use for physical jar testing, and we'll start with Brad. Brad, across the plants you've worked with, how do operators actually end up using physical jar testing? What are the best practices, insights, and use cases?

Brad Spangler   14:01

Well, it kind of, like I mentioned a few minutes ago, how they use it is a lot to set the dose, right? So, you know, we'll run the curve or plants will run the curve to help set their coagulant dose for the day and for the incoming raw water parameters. Some of the best practices, we always like to dose with micropipettes. As you know, it helps improve accuracy and it also feeds the product neat as opposed to stock solutions. What we found is that the coagulant will begin to work right away. So if you use stock solutions or dilute or made down coagulant solutions, you can lose some activity in the coagulant. So again, we like to feed neat. We like to use micropipette. Some of the things they're looking for, obviously, it's visual inspection of the jars. You want to, you're looking for a flock size, flock speed, how quickly does it react? How quickly does it settle? And you know, some plants have short flash mix, so they don't have a lot of time. They need a coagulant to work, right, flock quickly and settle quickly. Others have plenty of time. But those are some of the things I think best practice wise, we always recommend micropipettes for accuracy. We want to be extremely accurate in our jar test protocol, along with, you know, best practices. I believe some of the operators, in addition to final TOC removal, right, we want to We'll test the jars for how much TOC we're removing on that day, final settled turbidity numbers, but we don't just look at the final turbidity. We like to sometimes get a curve to determine how quickly it's settling as well. So those are some that come to mind.

Chris Lacey   15:47

Awesome. Yeah, I was thinking when you were saying that I've been to some large utilities that are really lucky and they have a TOC analyzer in their lab and so they're able to run a test every morning and then run TOCs on their jars and really optimize their dose for TOC removal, which is not common. Most people don't have that luxury, but that's a great use case. For the audience, I want to kick this back to you guys. Any other best practices for use cases that you would like to add or any questions about these? I'll give it a few a little while in the Q&A section. Feel free to chime in. It looks like you might.

Brad Spangler   16:10

The.

Chris Lacey   16:21

Here we go. We have a question. Do you also recommend using chemical addition disks to optimize timing between?

Brad Spangler   16:27

Hey, Chris, I think we lost you.

Chris Lacey   16:29

What was that?

Brad Spangler   16:31

Sorry, we're good now. Sorry, Chris. Yep.

Chris Lacey   16:31

Can you hear me? Okay, cool. Do you also, this is a question, do you also recommend using chemical addition disks to optimize timing between jars? So this would be for Brad. And I can also, I could also help with this as well. Can you hear me, Brad, with the disks? The question.

Brad Spangler   16:50

I'm sorry. Yeah, but you broke up for a second. Can you mind repeating the question?

Chris Miller   16:57

Yeah.

Chris Lacey   16:58

Sure, so do you also recommend using chemical addition disks to optimize timing between jars?

Brad Spangler   17:03

Finding. Absolutely. Chris, as you know, we use them all the time. We love the disk. You know, I've seen rulers set up where you can inject them all at once. But our team, I believe almost all of our team are using the disk. We input the coagulant from a micropipette right onto the disk. So yes, we like it. Anything you would add to that, Chris?

Chris Lacey   17:28

Yeah, so if your injection points are varying, right, if you have your coagulant at the very front before, like, let's say ahead of the first flash mix, and then your polyDAMAC or polyamines being injected maybe just after the first flash mix, that's a great reason to use those discs at different points in the procedure.

Chris Miller   17:28

No.

Chris Lacey   17:48

right? So you'll inject and then you'll wait, you know, your 30 seconds to a minute and a half or whatever it is, and then you'll inject the polymer. So absolutely, we recommend that. It works very well. Alright, any others? Any questions? We'll keep on trucking along here. All right, so let's discuss physical jar testing challenges. And so this one will be for Chris. What are some of the challenges with conducting physical jar tests and their adoption by water utilities?

Chris Miller   18:19

Yeah, so some of the obvious ones that come to mind is, you know, have a team that is busy and, you know, staff, you know, lots of responsibilities. So sometimes finding a dedicated person to do that. You know, we're talking about, you know, even though you can trained to this, it's not always simple and easy to, you know, be precise with, you know, the pipettes and the dose calculations, and especially if, you know, those things are changing. So, you know, really getting somebody dedicated to do that can be a challenge. Sometimes if you have a more complicated component as part of your treatment process, like ozone, or so, you know, you can't you know, you know, exactly replicate treatment. So sometimes, you know, you find ways around that or accept, you know, certain certain ways that you handle that. So, yeah, those are those are probably the few of the ones that come most to mind, Chris.

Chris Lacey   19:10

It. All right. Thank you, sir. And then, Brad, anything, anything to add to that?

Brad Spangler   19:25

No, I think I would probably piggyback off of Chris's answer there. We hear the same thing, staff challenges. So some of the municipalities that we work with, they mentioned some of the same thing. We maybe don't have someone on staff to do it every day or we don't have time to do it every day. And that's why I believe this virtual jar can be very helpful. In fact, we've even seen it with our commercial team. As you know, from a commercial standpoint, we like to run jar tests over a variety of incoming water parameters. We made jar tests for six months to a year because we want to see cold water, we want to see warm water, we want to see high incoming turbidity are low. And that's what I think is great about what Chris is doing here on the virtual jars. You know, maybe we can simulate some of these things so that and shorten the time, right? We don't have to, sometimes we'll tell the municipality, hey, call me when you experience a high turbidity event. And with the virtual jar, sometimes you don't have to wait on that. So we've seen it on our commercial side as well.

Chris Lacey   20:30

Awesome, thanks, sir. And then for the audience, any challenges that you guys are aware, you know, that you've had that we haven't already talked about, feel free to pop those in the chat and I'll bring them up as they come in. So let's go ahead and introduce the virtual JAR, because it's been on the screen now, but we haven't really talked specifically about what it is for those who don't know already. So for Chris, what is a virtual jar?

Chris Miller   20:54

Yeah, I'm going to go ahead and move back to this slide, Chris, to help kind of inform that. So the best way we like to explain it is think of it as a flight simulator. That may show my age a little bit, but I remember the first Microsoft flight simulator that you could. could do back in the day. But as Brad alluded to, is the ability to run a bunch of different scenarios that are still, you know, based in what's going on currently at the plant. But it will, what is it actually simulating? Is it simulating the different coagulant dose ranges, the different pHs that happen with that? And so we're It's handling those, the heavy lift on the calculations and doing a lot of what we know intuitively happens, but it can run those calcs and take care of that and bring into it both the, how the physical job performed, the treatability of the organics that you're experiencing. So there's this connection We can continue to elaborate on it today, but between this physical jar test, as the little graphic shows, and then there's an intermediate step that is some machine learning parameter extraction, as I was alluding to those papers that were published about 25 years ago. And then that information we can put into a simulation tool that can run many different kinds of scenarios and evaluate those for the water operator before they make a decision.

Chris Lacey   22:31

Awesome. And would you like to be on this slide now, Chris, or would you want to stay back for, so let's talk about, let's talk about this. So let's get concrete about something a lot of people don't realize. How does a physical jar test actually inform the virtual jar? I understand there's a machine learning step that extracts chemistry parameters directly from the bench data. Walk us through that.

Chris Miller   22:53

Yeah, so we're actually applying the method that was published in 2004. They used Microsoft Excel's solver function, which is where you have data, you have a dose response curve, like Brad said, and you can... extract these four parameters. So that's a, you know, an error minimization, machine learning type technique. We're doing it. We've made some improvements to that. We're actually working with a couple of our customers that we may, we're going to look to probably publish some of that later this year. But two of the parameters are more or less related to the coagulant, kind of specific to the coagulant that you're using. And two of the parameters are specific to the treatability of that TOC. So if you have, you know, again, people think about SUVA, higher SUVA is easier to treat or higher TOCs, you could remove more TOC. These are actually parameters that are really specific to how much of that TOC can you actually remove via your coagulation process. So just the five steps kind of on here about how it informs. So again, we said physical jar test, get some data. You really do need either TOC or UV 254. UV 254 is a good surrogate, much easier. You can get a batch top unit. Like you said, Chris, not many people have or many utilities have a TOC analyzer on site, you know, so we've found a way to make use of if you've got a UV 254 on that step too. that we take that data, do that machine learning parameter extraction, and then that allows us to inform and build out really your digital twin or your treatment, the virtual jar that can run these simulations of the treatment plant, look at a bunch of different scenarios, and then help inform so the operator can make a decision.

Chris Lacey   24:57

Awesome. And then, so for Chris, so once those parameters are obtained, how does virtual withdraw actually forecast the turbidity in the TOC, not a dose response curve or an expected like TOC value, but a blend of coagulation mechanisms and profile of the organic carbon?

Chris Miller   25:15

Yeah, so there's a simulation component to this. So the way we set up virtual jar is that let's say that you're currently dosing the coagant in a dose of 30 milligrams per liter. You generally have a range that you're willing to move that dose in any one step, right? Whether it's two parts, 5 parts, what is that dose range? So that helps put in the guardrails of how much do you want to examine? So for my example, if you're at 30 and you're willing to go plus or minus 5 milligrams or parts in once, one move, you could look, we could look at 25 to 35 plus or minus 5 parts, and we can run all the dose, we can run it in increments and run the dosages. And then when we are calculating that, we're calculating the pH changes, which helps inform, you know, people know about enhanced coagulation, right? Suppress the pH, you get better removal. Well, the reason you get better removal in simple terms is you're generally adding more coagulant. So of course you would expect performance to go up, but it really is that that pH that moves that how a fraction of that TOC will actually absorb. And so we're doing that calculation inside of that as well, not just looking at hey, this is enhanced coagulation, you know, suppressing the pH because, you know, we know you can get good organics removal by minor pH changes, right? Using higher basicity or using blended products. It's not always just about suppressing the pH to get better removal. So chemistry calculations, absorption calculations, the TOC, and incorporating the nature of the TOC, all those things wrapped into how, you know, running those simulations and then providing them in an easy, easy way for the operator to digest so they can see those trade-offs.

Chris Lacey   27:21

Got it. Thank you, sir. Looks like we had a couple questions come in. They're both looks like both for Brad probably. So let's start with the first one. How about optimizing coagulant dosing using a combination of jar testing and streaming current monitor outputs? So, like, have you come across that, Brad? Do you?

Brad Spangler   27:46

Yeah, absolutely. I mean, I think Chris was going to touch on it maybe later in the deck where he talks about charge neutralization at SCDs. But yes, we've seen it also where you're running the jars to take at X amount, this is what dose you're going to destabilize the particles and actually

Chris Miller   27:50

Yeah.

Brad Spangler   28:09

calls it to settle out. So yes, we've seen where you can use both. Chris, anything? Chris Miller, anything you would add there on the SCD as it relates to the virtual jar?

Chris Miller   28:16

Yeah. Yeah, I think it ties into one, like you said, we were going to talk about next is that we actually, one of the things that we've recently added to the virtual jar is a net zeta potential forecast. So you take, take the, so let's say somebody does not have, so to the person who asked this question. Somebody that does not have zeta or streaming current, we are actually forecasting that net zeta off of, again, the turbidity has a negative charge. The TOC has a negative charge associated with your coagulant that's the positive contribution. we would do that calculation and forecast the net data so that, again, if they don't have it, but if you do have streaming current or zeta potential measurements, we can actually then tune the virtual jar to what you're seeing as part of that and help assess again, yeah, the charge neutralization part and have that be part of the how it's ultimately impacting turbidity removal as well as TOC.

Chris Lacey   29:28

Thank you, sir. It looks like we have another one. And this is more of a response to the challenge question from earlier, I believe. A response time can be a big challenge for people, the time delay in between the jar test results and then the process response to the chemical change out in the process, right? And so that's a big challenge that people come across a lot of times in their jar testing. So that's a that's a really good one. Thank you.

Chris Miller   29:52

Well, Chris, I think that's one of the things that's where having virtual jar comes into play. So, so again, we're talking today about if you run a physical jar and then you go ahead and, but if you know we're working with a client of ours has an active flow system, right, that's turning over every 20-30 minutes, right? You don't. So they're actually just, they're using virtual jars. They see those changes because you can run a simulation in about a minute. You can actually look and try and as that raw quality is changing and you want to actually make the change and go with it, you can do that. So, you know, or if You also know you have that delay, you have long sedimentation, may take several hours. You know, what's, you know, how do you want to kind of maybe, as you, you know, if somebody's a river source or you're getting a rain event, you know, there are ways to kind of blend using both of these tools to try and close some of that time gap. Yeah.

Chris Lacey   30:54

Awesome. Looks like we had another one just coming in for Brad. What effect, if any, do you find that pH has on performance on the bench when compared to the actual process?

Brad Spangler   31:06

Well, we try to match. I mean, I'm trying to think in the jar test, we're going to try to match what we're doing in the plant, right? So we're hoping it's the same, right? We're hoping that the incoming water is the same as what we would see in the plant. So in any jar test, whether it's mixed time, mixed energy, pH of the incoming water, we're going to try to match as close as possible to the plant so that we can get meaningful results.

Chris Lacey   31:35

Yeah. I'll add one of the hardest things for me when I started off in jar testing was matching A ferric sulfate with lime plant in the jar because they're feeding this lime slurry and getting a lime slurry into the jar at the right concentration takes some skill and you have to be very careful and it takes a while to get good at it.

Brad Spangler   31:36

Spring.

Chris Lacey   31:55

So that is not a pleasant thing to try to figure out, that's for sure. It's very difficult. But yeah, that's the goal. We want to match the plant in the jar. So we have two new outputs, and I kind of jumped to the screen because you were talking about Zeta potential, Chris. And so you talked about that, but we also have this new organics treatment index. You know, what hole or gap are you trying to address with these two new outputs?

Chris Miller   32:09

Yeah. Yeah, so I think we just talked about relative distributed removal and thinking about charge neutralization as the key, one of the key mechanisms that's kind of, you know, definitely treatment 101. So get this slide up here, we talked about charge neutralization. There's always Also, you know, whatever the sedimentation, particle settling, you know, again, can be totally different whether you have just, whether you have plate settlers, whether you have DAF, whether you're going to, so probably should have put their sedimentation. How do the particles actually get removed out of the system is part of

Chris Lacey   32:54

Good.

Chris Miller   33:01

what's included in the turbidity modeling. But on the TOC, Chris, you mentioned this sorbable organics and non-sorbable organics. Those are those two core fractions that have to be addressed and looked at. And so we went ahead and introduced, I talked again about net zeta, but on this organics treatment index, this is the percentage of that TOC that you can actually remove via your coagulation process. If you are also adding powdered activated carbon, we can include that in it as well. So this is showing a number of like 68 percent. What that 68 percent represents is not 68 percent. TOC removal, that 68% of the fraction that you can actually remove with your coagulation treatment process. So it's, we think it's a little more insightful about, you know, the efficiency of what you're actually doing at the moment, the treatability of that water. rather than strictly just looking at, you know, how much TOC am I removing? Because what you are trying to really assess is how efficient is your coagulation, you know, treatment process for what it can actually remove. So I see a hand up, Chris, if we want to take the. Question?

Chris Lacey   34:22

Yeah, I just unmuted Eugene. Eugene, it looks like you have a question. Go for it.

Chris Miller   34:25

K.

Chris Lacey   34:31

Play this works. Eugene, you there? You should be able to talk. And if you can't, feel free to drop your question in the Q&A too. Probably a little bit easier, probably. Oh, here we go. Let's see if that works. Go ahead and try now, Eugene.

Chris Miller   34:57

I'm still seeing a mute on his.

Jake Komarny   34:59

Done yourself.

Chris Lacey   35:00

There we go. Looks like he should be able to go. Yeah, Eugene, feel free to drop that question in the Q&A. I'll leave your mic enabled, so if you're if you're able to. Something happened to the deck. Turn it back on. All right, Chris, do you have control? I said you took control.

Chris Miller   35:49

Yeah, yeah, I got it. But you find the chat there, yeah.

Chris Lacey   36:04

Yeah, I don't see. Here we are. There we go. We draw from Lake Erie and our water changes continuously. Looks like he's still talking. Go ahead.

Chris Miller   36:13

Yeah, so we, yeah, we actually have, we have several clients that draw from Lake Erie and have been using virtual draw for several years, even also during some of those dead zone anoxic events that have happened that bring in lots of manganese. So Today, we talked primarily about turbidity and TOC removal, but we've actually added to their virtual jar manganese removal. I mean, just it only happens, you know, in rare periods during the year, but then they can adjust the oxidant as well. So yeah, we, our river sources are again, active flow, things are turned over. So I think that where the bridge from the physical jar test is that those, the nature of the TOC that is being treated, and we know again, Lake Erie, high quality water generally, but that treatability does change quite a bit as well. So Getting those historical values that you can only get from a physical jar test, those help build out some record that help involve kind of the, almost the guardrails or what's possible for what's the treatability of that water. So I know if Eugene that helped address that or not, but so we are going to talk about a workflow, you know, kind of a workflow that involves a physical jar test with a virtual, but you can run a virtual jar test as needed on demand. So if, you know, something changed over the course of several hours, We have, yeah, again, customers that will just run the virtual jar, you know, and then if they want to supplement it with the physical jar tests, there's many different ways to bring these two together.

Chris Lacey   38:09

Awesome. Thank you, Chris. It looks like we have another question that came from Andrew. So when evaluating different coagulants, how do you determine whether the new product is a higher or lower effective charge? Any standard methods to characterize charge? That's for either of you who want to take it.

Chris Miller   38:31

Take.

Brad Spangler   38:31

Chris, I mean, you talked a little bit earlier about charge neutralization.

Chris Lacey   38:31

Yeah.

Chris Miller   38:33

Yeah.

Brad Spangler   38:37

Anything you want to add on the virtual jar with different coagulant products?

Chris Miller   38:38

Yeah, yeah. Yeah, so, I mean, USACO has quite a portfolio that has, you know, different basicity, provides different charge, right, for different, you know, even for dosages, right, not all aluminum is the same, right, you know, so So we do bring that into the virtual jar component, as well as obviously, if you're testing a different product in the physical jar, again, our water treatment specialists know how, you know, how those are going to behave and whether kind of how you're balancing the different basically the species of the aluminum to address, you know, particle removal or turbidity removal versus TOC. So sometimes we want to try and match up, you know, is the customer primarily focused on, you know, let's make it simple, that mostly it's about turbidity removal or it's 50-50 or as, you know, in some cases, you know, it could be weighted towards one or the other. And that's where, you know, we'll tend to customize or guide that product selection to meet where kind of their waiting is, you know, where their focus is, if you want to call it the bigger challenge, you know, or they have to treat things equally. So.

Chris Lacey   40:04

Thank you, sir. So let's discuss an ongoing project. This is for Chris. One of our utility partners runs a physical jar test on a weekly basis, and there's been a combined physical to virtual jar workflow running there for more than three years. Walk us through what that partnership actually looks like week to week.

Chris Miller   40:26

Yeah, so there, we've been working with them now for about 3 years. It's a river source. They were in the habit of running every Monday, running a physical jar test. So they run a physical jar test on Monday. Those results would come in within a couple of hours. the operations team would meet and in a way it would help set the set the dosage for the week. Now, obviously, if there was a rain event or if there was something that happened during the course of the week, they could, you know, run another physical jar test. But that was that was kind of their workflow. So when we brought virtual jar to them three years ago, The only change to the workflow was that Monday afternoon, when they would be looking at that dose response curve, they would also run virtual jar, and it would allow them to examine some other dosages that they could look at now this organics treatment index, like where they had efficiency wise. And where that comes into play is that I know it's not the largest graph, but this graph in the left middle of the plot here or the chart is that is a TOC removal for four different dates over the course of the last over the year. They're not in chronological order. The main point of this was that even though their dosage range remains reasonably within a plus or minus 5 parts, they would see these big swings in TOC removal. Again, anybody that's looked at a lot of organics removal data sees that these things can change even though your dosage is you know, relatively consistent. So this didn't have anything to do with, you know, the coagulant or what they were doing. It just simply had to do with that. Water was harder to treat when those removals dropped from 60% TOC removal to 21% removal. So what that allowed in combination with the virtual jar was that they looked at that organics treatment index. They knew they were actually getting a reasonable amount of removal of what they were supposed to. There just wasn't much there that they could remove. So it provided that additional insight to not just keep cranking up the

Brad Spangler   42:48

Which?

Chris Miller   42:53

keep cranking up the dose, you know, to try and hit just an arbitrary TOC removal target. So, you know, they're meeting their required removals, but they like to go above and beyond because it helps them manage their disinfection residuals, helps manage their DBP levels. So they tend to shoot for even higher removal. So again, they're now doing this for about 3 years. That is the week to week flow, but they can also run virtual jar. You know, they get a rain event, they get a river event, they either will initiate another physical jar and then inform the virtual jar, they'll just run a virtual jar test. So, and again, they've been very interested in how it supports the disinfection and the DBP parts of their water quality goals as well.

Chris Lacey   43:49

All right, thanks, sir. So I have a question for Brad. A lot of your work is matching A plant's treatment needs to the right coagulant. How has the combination of a physical jar and a virtual jar test combination helped that effort?

Brad Spangler   44:04

Yeah, you're exactly right, Chris. As you know, our team, we do that quite a bit, right? We're always searching for the right coagulant. We've observed some municipalities may have a coagulant in play, but we're always asking ourselves, can we do it a little bit better? Is it the right coagulant? We're auditing ourselves, even accounts that we're feeding. And if we can improve it, we're always looking to improve it, right? So running side by side JAR, and I can think of a couple of instances here recently where virtual JAR allowed us to run the incumbent coagulant, if you will, side by side with a few different simulations. different products, right, to help predict what the outcome will be if and when you go to full-scale trial into a plant. And it was very helpful, right? It helped us simulate what was going to happen in the plant, and it was very accurate. And then it allowed that plant after the fact to it continued to aid in the decision making process. So I think the two together is very helpful as we're evaluating different products. It helps in the decision making process and basically tells you what's going to happen in the plant before you do it. And that's a so there are no surprises. That's a very helpful tool.

Chris Lacey   45:29

Awesome.

Chris Miller   45:29

And again, the only reason we can do that, right, Brad, is because we know how the different blends and how's we're trying to get that right basicity, metal content, you know, all the above, you know, matched up with their objectives. So the ability to bring that into and in a way run our own simulations to help inform that. Yeah, it wouldn't be possible unless it was, you know, one of the questions is how much of this is like just observed data, empirical data? It's not that at all. It really is because we know the chemistry and we can bring those things into that evaluation and that analysis.

Chris Lacey   46:12

All right, let's change the subject just slightly. So it's not a big change, but we're going to move away from the case study now. So physical jar test and virtual jar test adoption. So for Brad, what's the most common pushback you get from a water utility on physical jar testing or the virtual jar?

Brad Spangler   46:31

Well, I can speak on the physical jar test. The biggest pushback I think that we hear, some municipalities will say, or some customers will say that the jar, my jar test doesn't give me meaningful data, right? What I get in the jars doesn't match the plant and therefore I can't use it, right? And I talked about jar test protocol earlier. If we're using the standard out-of-the-box jar test protocol, flash mix 30 seconds plus a medium plus a slow mix and a 30 minute settle, it's probably not going to match the plant, right? So as you know, we spend a lot of time trying to make sure that our jar test protocol is correct. We spend a lot of time out in the plant measuring basins, RPMs of mixers so that we can get meaningful data out of the jar test. So I think that's probably the number one response I hear is that if there is any pushback on the jar test is that it doesn't match my plan.

Chris Lacey   47:32

Yeah. Scott Williams just put something in the Q&A. I'll read it to everybody because it's pretty good. Just a quick comment regarding how many operators conduct jar testing but are frustrated that it does not match the actual plant performance. There can be many reasons why, but working to get an accurate similitude is important. A vast majority of the time, The variations come from energy differences in the jar versus the plant. In changing the settling times in the jar test to accurately represent the plant overflow rates. I also, when making a coagulant change in the plant, grab a sample from the rapid mix and put it on a jar tester at slow mix speed. And this will shorten the time to find the right dosage in the plant. If the intermediate rapid mix sample is not producing a good flock, then most likely it will need further change. And so That's a great trick to use, like when we run trials, we'll do that because it saves a lot of time at the beginning. But thank you, Scott. That's great.

Chris Miller   48:23

Yeah.

Chris Lacey   48:25

And thank you, Brad. Chris, so if utility is skeptical but curious, what's the lowest risk way to start and where do you see this partnership between the physical and the virtual heading in the next few years?

Chris Miller   48:39

The easiest thing is to contact Brad and one of our water treatment specialists and start a conversation, right? So, that's the easiest way. But in all seriousness, one of the things that is often a limitation is a lack of organics removal information in the plan itself. So if it's like the monthly

Chris Lacey   48:51

E.

Chris Miller   49:03

you know, required TOC removal is the only data that's available. So we will, I mentioned earlier, encourage, you can get a decent bench top UV 254 unit. Sometimes people don't even realize that they've got the Hawk, depending on what model you have, you may already possess that capability. But we've got a simple SOP. want to make sure the samples are filtered, because if you have particles in the sample, it will interfere with the UV254 measurements. We have a nice SOP to help kind of get that in place. And so, because it's a large part of the coagulant demand, the coagulant performance, even when you're looking at particles, so That, that's the, you know, if we're going to really go into this path and begin balancing the coagulant for TOC and turbidity removal, that's kind of the barrier to entry to kind of get that in place and help implement that and show how that can actually become something that like 2 minutes a day and you can really assess the organics removal performance of your plant. So from a technical adoption using digital tool, there's a reason it's called Virtual Jar because, again, we're not going to show the software interface today, but it's built like a jar test. So an operator can actually put in chemical combination that they're thinking about and what they're thinking about making a change. So this has really been informed. We've been at this now for about 8 years, been informed by our customers to make it as easy and aligned with how they're thinking about making these changes, not some you know, crazy simulation tool. It actually is built to kind of respond and help inform things much like a jar, physical jar test does. So.

Chris Lacey   51:01

Awesome. We had a question come in for you, Chris. Will the virtual jar, this is probably from like one of our coagulant distributors who resell our products. Will the virtual jar test be available to be publicly available for your vendors or will it only be available internally to use Salco?

Chris Miller   51:04

Yeah. Yeah. Yeah, so we, so Bonus Blue was the company that merged with USACO. We've now been on this journey for about 13 months. So we, I think we have a pretty good process down for our water treatment specialists and team. And I think one of the things we're excited about is talking about how do we introduce it to our distributors and that. So I think that's a, I would love to have you follow up with us. And if you have some ideas and thoughts on that, keep in mind that each virtual jar is a custom build for a treatment plant. So because they may have potassium permanganate or sodium permanganate or free chlorine is an oxidant. What's your coagulant? So part of this is figuring out how, you know, how do we actually bring that to support specific clients, et cetera. And I think, you know, Brad may have some thoughts on that. Well, like I said, we're really You know, trying to think about that in the next few months and how we can make that make that available and and and help support our distributors on that front, so.

Chris Lacey   52:28

Awesome, we've had two awesome.

Chris Miller   52:29

Stay tuned. Yeah, stay tuned and follow up with us. And again, let us know your thoughts on that.

Chris Lacey   52:37

And we had two more come in. Jose is asking, we've seen a few new portable water plants using DAF technology and also did a project in Asheville to get emergency treatment in place to use DAF. Jar testing on DAF is hit or miss unless you have specialized equipment. Is there a capability in the virtual jar to simulate the DAF process yet?

Chris Miller   52:58

Yes, so again, we didn't go knee deep in this or get really into the weeds, but one of the advantages, if you will, or one of the gaps the virtual jar can close is we will use the actual, the measured performance to help tune or inform So while we use physical JAR to inform what's possible, you know, under the conditions you tested, virtual JAR will actually use in-plant performance data. So it's an advantage when you have ozone or in the case like here, like DAF, and you're looking at what is the performance of the DAF units, will actually, one of the inputs will be yeah, what's actually happening now. And so it can tune to, you know, some of those empirical external factors that are outside of the chemistry realm and that are impacting, you know, again, we're not modeling bubble size and all this kind of, you know, we're actually we'll just be like, hey, here's how.

Chris Lacey   53:57

Good.

Chris Miller   53:57

Here's how those units are performing. We can bring that in along with that chemistry part and get good simulation or good forecast to what if the operator made a change, what should happen. That's it allows us to make that forecast with confidence using the real plant performance as well as the chemistry. of whatever quag that you're using.

Chris Lacey   54:22

Sounds good. So we have three more questions so far. All tests on raw water will help with what dosage to start on jar tests, such as turbidity, color, and alkalinity. So I guess, generally speaking, when you're beginning with brand new raw water, what dosage would you start with? And to target turbidity, color, and upline removal, or not upline removal, but turbidity and color removal, probably. Brad, you or Chris, yeah, go ahead.

Chris Miller   54:49

I think like what I was just going to say, I think like what Brad said is that, you know, oftentimes we hear, hey, come in the summertime, come in the winter, part of that's to capture the temperature, but it also is to capture different water quality conditions, right? Whether it's a TOC level or even you'll see alkalinity shift, right? reservoir turnover, these, you know, different, different things. So I think that, again, what Brad alluded to is that regardless of when we go in and test, once we get an idea of, I think it's more important to see, like, you know, are we matching the right, coagulating it with probably what some of those goals are. Once we get one of those physical jar tests, you could actually give us, hey, this is, you could give us some historical data and we could actually run different historical dates. So I'm not showing the Pittsburgh, oh, I meant sorry who it was here. But this large utilities results here, This, they wanted us to actually examine these were four different scenarios that had different SUVAs, had different, you know, water quality conditions, pH. And so you can actually go back and run some of that and see, like, depending on what your targets were for to, you know, maybe meet Required TOC removal, et cetera, and and see how much that that would that would change things, so... Hopefully that connected that bridge that, but we still want to know how the coagulant behaves on at least, you know, one physical jar test helps us anchor and then we can bring in some of the other what-ifs simulation on historical dates. So.

Chris Lacey   56:40

Got it. And just for the audience here, we're at the point now for Q&A. So feel free to fire away on that. We're going to keep going. So we have another question from Rosalyn. She says, wastewater utility here. Can this system be used with a chemically enhanced high rate primary treatment system? We use jar tests to adjust our dosing. to treat wet weather flow to a TSS standard.

Chris Miller   57:06

Yes, that'd be the short answer. So, and again, where I go with yes is that that the what I was mentioned a few minutes ago about the using the actual plant performance data. So, you've got and again, whether whether you're using a you know a surrogate for TSS or however you're getting, you know, whether it's turbidity, you know, however you're capturing.

Chris Lacey   57:08

It.

Chris Miller   57:29

the actual performance removal live. So you can make again, I think this gets to that heart of how do you make decisions quick when things are moving quickly, like during a wet weather event. So hopefully answered answered that, you know, so again, where we're going to lean on is lean on the actual performance and then I was saying about being able to make a change with confidence, it will tune to that real data in the real plan in the moment. And then do you need to, you know, again, I like to say, you know, which direction you would probably adjust the coagulant, like if you're not getting the solids removal during that high weather event, you're going to turn it up. But the question is how much? I think that where the virtual jar comes in place, it help you quickly assess, you know, is it 10 parts, 20 parts? What is it that we need to go up to get to the performance in real time? Yeah. But very familiar with that. Taught at the University of Akron. Akron has, they haven't had to use it much yet, but they have the high, they have that high rate treatment in place as well during their real large storm events.

Chris Lacey   58:42

Awesome. Our next one, what is the difference between AI plant simulation and the virtual jar?

Chris Miller   58:51

What was that again, Chris? An AI? Was that it?

Chris Lacey   58:54

Yeah, what is the difference between an AI plant simulation and the virtual jar?

Chris Miller   58:59

So, like an artificial intelligence thread, the AI is out of this on there, yeah.

Chris Lacey   59:03

Yes, yes, sir.

Chris Miller   59:05

So, again, my colleagues have heard this story many times, but 2015 actually published a paper using neural networks where you take just plant performance data, fit a neural network to it, you know, so whether it's like AI or you know, take historical data and do that. and, you know, published the paper. It was the first like kind of V1 that we were doing with virtual jar, and it started to hallucinate. So while we're familiar with how large language models, ChatGPT can hallucinate or make things up, our virtual jar would say that something was mathematically possible. So we went back to the drawing board and made sure that the foundational parts are chemistry. So my answer is that, and again, our competitors, or there are others out there that may just take that machine learning, artificial intelligence approach, and it's going to hallucinate. Much like, you know, while we can trust it, you know, 80% of the time or something, we need something that you can trust 100% of the time is going to be, you know, guide you the right direction. And so foundationally, it has to have chemistry at its core, otherwise you will have...

Chris Lacey   1:00:33

And it could take.

Chris Miller   1:00:33

That's the primary difference, yes. Ours is based on fundamentals and chemistry. And then we do a little tuning, like I was talking about, but, you know, we really bring that core chemistry expertise into this. And that's really the differentiator between what's out there.

Chris Lacey   1:00:54

And I'll let everyone know we're going to keep going until we don't have questions. So we have a couple more to go, but we're not going to stop in 3 minutes. So if you'd like to keep going along for the ride, feel free. So we'll go on to the next one. Machine learning works best on data that is similar to the training data. When the virtual jars experience anomalies in water quality, how is their accuracy affected? Do they provide a metric such as confidence intervals to indicate its accuracy?

Chris Miller   1:01:23

Yeah, so I think there's two parts to that question, I think, that I want to make sure to try and isolate. So the first part is, yeah, we talked about how on a given day or in a given time window, the physical jar test can inform, you know, what that water quality, that treatability, So that's the only historical data that's going into virtual jar and is becoming part of its catalog or guidance. But in the moment, so this is the second part. So the treatment plants, whatever, it could be a calm day. And it's like, how is virtual jar working on that calm day? It could be a calm day that It has a different alkalinity combination. It's not looking at the past to help inform what it's going to do that day. It's saying, hey. We are going to measure what the current plant performance is for turbidity removal and for organics removal. And we're going to compare it against what for the given raw conditions, pH, alkalinity, temperature, the coagulant, the oxidants, all those chemicals that are in there and what were forecast

Brad Spangler   1:02:38

Everyone.

Chris Miller   1:02:39

forecasting and whatever that difference is, we will tune or we'll make adjustments. But how we make those adjustments then allows us with confidence to deal with that day and that specific condition. So that we don't, again, if you increase the coagulant, you consume more alkalinity. The question is how much and how far does the pH go? which is going to, you know, how much will it increase the efficiency of removal? Those, that's what I mean by that. Those, all those calculations are being done by virtual jars so that we know will things will move the right direction and not, you know, push things into a zone where you wouldn't want. Wanted to go, so.

Chris Lacey   1:03:27

Thank you, sir. So we've got a version of this question before. So does this mean fewer jobs for operators? Like how do you answer that? This would be for Brad or Chris, you know, Brad, what do you think?

Brad Spangler   1:03:41

Well, I know the answer, but I'm going to let Chris answer this one. This is a good one.

Chris Miller   1:03:45

Yeah, that's it. Yeah, it laughs. Yeah, that of course not, right? So this just allows, again, this is a core function, but it's only one of, you know, 50 functions or whatever an operator has to do, or, you know, 30 functions, you know, they've got lots of things that they have to do. This just really, it does the heavy lifting of these like chemistry calcs, pH changes, all that. It also does calculate the cost of treatment. So it brings that, you know, you add a little more polymer, you do whatever, you know, if you're modifying all the chemicals. So just takes away some of that heavy lifting. I think the best analogy I've used is we've all probably forgotten what it was like to take out your Atlas and figure out how you were going to get from, you know, Chicago to St. Louis, right? That required some energy. You still had to drive. That was the job, drive from Chicago to St. Louis. This just makes it easier and faster to know, hey, here's the pass, here's the thing, you know, and you can make these changes with confidence quickly, easily, and then go about those other 45 functions that you have to do. So we intentionally do not look for a virtual jar to control the chemical dosing intentionally because the operator ultimately legally responsible for making those decisions. It just helps inform that so they can make those changes with confidence. So our intention is not to build out an autonomous chemical dosing system and take that Yeah, again, things we've talked about today, physical observation of the flock, other things that matter and where, you know, the operator can bring that into the mix. This is just meant to be, you know, a superpower, a super tool to put in their tool belt to make it so they can do it even better.

Chris Lacey   1:05:43

We have another one here. How many years of data do you feel is needed for the virtual jar test to predict accurately? Also for plants that get a lot of low turbidity water, like below 1/2 a NTU, it can be difficult to determine optimal dosages from a physical jar test because subtle turbidity is so close to the raw water turbidity. Do virtual jar tests perform well for predicting dosages for low turbidity raw water? So I'm thinking like Colorado right now. It's like really low. Yeah.

Chris Miller   1:06:07

Yeah, yeah, exactly. Yeah. Well, or even Brad knows one of the ones that we talked about there, NTU is about one or less, generally even.9 NTU, right? So I think this gets back to one of those earlier questions about, do we need a lot of historical data? And the answer to that is no, because it's not calibrating to a bunch of historical data. So I think, Chris, that was kind of just kind of hinted at. But so then to the, you know, how does it perform with, you know, low turbidity waters? And that part is indeed tough. That's, I think, where the charge neutralization part comes into play. And where I'm going with that is that, you know, what is the TOC? That's why you've got to get that UV 254 or something in there, because that's going to create that additional negative demand. So, you know, are you hitting near neutral? Are you getting where you want to be? Sometimes that's masked by, you know, hey, here's half a part of NT, you know, half NTU, not half a part, but.5 NTU, but what's that organics part so you can make sure you're at least getting charged neutralization. So, I mean, the reality is you generally don't need much coagulant, but you want to make sure you're getting the TOC removal tends to kind of be almost the focus, even if it's like you said, Chris, high quality water, which Colorado has, that it is that organics part that's moving that's causing those swings. And so

Chris Lacey   1:07:30

Yup.

Chris Miller   1:07:41

It's good to get that UV254 to get in there and so you can assess that.

Chris Lacey   1:07:48

We're just going to go for a few more minutes here. So we have another one. Would the virtual jar be able to simulate changing to changing to different treatment chemicals, i.e. aluminum sulfate to ferric or various powder activated carbon products?

Chris Miller   1:08:02

Yeah, so that's right up our alley on both parts of that. So that's obviously our core business is helping make sure our customers have the right, the right coagulant, right? And we've got the whole gamut. So it's not like we're just trying to, you know, sell one. one version of coagulant. And then on the powdered activated carbon part, city of Akron, they leverage their powdered activated carbon as a key part of the TOC removal. And so it's part of the virtual jar. There's A coagulation part and we actually are now One of their questions, actually we just met last Friday, is that how can Virtual Jar easily show them like how much lift can be done by the powder within the allowable dose range they want versus how much the lift can be done with their album. So, and as the nature of that TOC changes, that answer changes. So the powder can do a pretty good job on some of the lower molecular weight parts of the TOC and the coagulant does a fantastic job on the higher molecular weight parts.

Chris Lacey   1:09:16

Awesome. I don't think we've covered this one yet. When you're, well, we have, but not this particular question. Specifically, when you're running into plants that are saying the jars don't match the plant, is the frustration usually that subtle turbidities don't match closely what the plant produces, or that if they try the dose associated with the point of diminishing returns for turbidity, UV254 or TOC, they don't get good performance.

Chris Miller   1:09:42

You want to take first shot, Brad, and then I can...

Brad Spangler   1:09:44

Yeah, so I think both. They're saying basically, does the turbidity not match or does the TOC not match? And we've seen both, right? We've seen, I'll go back to the jar test protocol. If someone settles for 30 minutes in the jar, reality may only be 12 minutes in the plant. So the turbidity is not going to match. We take a lot of time to make sure that the jar test protocol matches the plant. Settling time is a big part of that. So if you don't know how much settling time you have in real life, then the jar test, there'll be a disconnect there. And same on the TOC side. I've seen TOC not match the jars. in the plant, sometimes the mix energy is off. We're not getting enough mixing or contact time in the jars and then you get into the plant, it gets much better mixing and TOC removal goes up. So we've seen it both ways, TOC and turbidity.

Chris Miller   1:10:42

Yeah, and I think the only thing I'd piggyback on there, Brad, is the... We, one of the, I mean, the main part of what we're extracting from the physical jar test for virtual is that TOC behavior. And the TOC behavior is a little less sensitive to that. And the reason why is because the absorption is fast. You know, so like turbidity really, really, I mean, the energy stuff is critical, critical, critical. And where you sample on the

Chris Lacey   1:11:05

Good.

Chris Miller   1:11:13

and where do you like, you know, to get, there's a lot of that. Fortunately, the TOC, you know, as long as you're dosing the right dosage, you should be close. Now, if you do see a difference, then it just may be that two hours, you know, or whatever the time difference, when you sample the raw, the raw could just be changed. you know, during that window, I mean, believe it or not, you know, could cause those. So, you know, people always say, hey, our TFC is not changing much, and then you'll still see it move 10 or 15 percent. So even if it's 3 milligrams per liter, it's only moving between, you know, 3.4 or 2.7. It seems like that's not much. But, you know, that can significantly change, obviously, your observed removal, so...

Chris Lacey   1:12:02

Oh, you got just another one just came in that's really interesting. Can the virtual jar test predict filterability indices between different coagulants or just the characteristics of what the settled water quality turbidity would be? That is, if you were trying to look for the new coagulant, can the virtual jar test predict which coagulant will perform best or yield the best filter? Water turbidity.

Chris Miller   1:12:24

Yeah, so we're, that's on the roadmap and where I'm, that's part of this net zeta piece and where I'm going with that is that, you know, filters. how you backwash them, how you treat them, you know, I mean, all those, how old's the media, right? Like, I mean, there's all the, all these things that I'd say filters are like children, right? You don't know which, you know, I mean, some, and where I'm going with that is that, that if we understand the first principles of making sure are we not, overcharge to where those particles aren't going to behave as well in the filter process or not. So we're trying to figure out how to navigate that, you know, but ultimately, if you are removing the right TOC, not overdosing your coagulant, depending on what the targets are, those are things that should foundationally then translate to you know, better filterability performance. So last thing on that, we just added, we have one client, we're still working this out, is that is forecasting unit filter run volume. So obviously, the other challenge with filters is their performance is an aggregate of hours of what's going on then, right? And it, even if the turbidity moves, the nature of what you're putting on it, turbidity, I mean, the flock wise. So it's an aggregate thing that you're looking at versus, you know, trying to match the treatment of what the water coming in and meeting your target. So We're making progress on that, and but that's they challenged us to basically do that is that, hey, we've got kind of target unit filter run volumes and want to make sure we're matching those. It's in virtual jar, and they're the first ones they've been operating for about four months. So We think we'll have that in there. That's how we're handling it right now. Yep, is going on a unit filter run volume basis.

Chris Lacey   1:14:32

Really good comment. I'm going to read it real quick and then we're going to close. One said, sometime the non-match is not due to the jar test protocol and it is really related to the plant conditions is how high the solids level is in the settling sedimentation basin or some of the filters need to be backwashed or they're not in good condition.

Chris Miller   1:14:34

Yeah.

Chris Lacey   1:14:51

And you know what Juan is actually with the Salco, and that was such a good comment, and it's something that oftentimes we forget is, is you know, if you know if the solids in the sedimentation basins are too high, that can impact performance quite a bit, and a lot of people tend to forget that when they're when they're looking into optimizing the coagulant.

Chris Miller   1:15:05

Yeah.

Chris Lacey   1:15:11

or program. So I want to say we appreciate you all joining us this afternoon. Some of you are longtime customers, some are brand new to USALCO and are just getting to know us. If you would like a specific follow up or anything, you could drop, you know, drop that into the QA or QA, or you can e-mail webinar at usalco.com. And we also have Chris and Brad's emails listed on that prior slide as well. So feel free to reach out to them. Reach out to us at webinar at usalco. We really appreciate you all joining us this afternoon. Thank you so much. We'll see you all.

Brad Spangler   1:15:46

Thank you.

‍