Water Digital Twins: Engineering and Operations Solving Problems Together

BRETT SINGLEY: Hello, and welcome to my presentation on Water Digital Twins. We're going to talk about engineering and operations solving problems together.

I'm Brett Singley. I'm a Group Product Manager over some of our hydraulic modeling and design tools inside of the water space at Autodesk. And I'm excited to share this presentation with you.

I've had years in consulting and using some of these products, as well as quite a bit of time at Innovyze and Autodesk being in charge of what new features come in next. And so I'm excited to share those with you. The agenda for the presentation today is going to start with some questions.

What is a water digital twin? What problems are we trying to solve? Then we're going to talk-- the bulk of the presentation-- about what ways some of our customers are already solving some of these problems, and interesting ways that they're using a digital twin to solve complex problems.

And lastly, we'll get a little more introspective and think about what can we do next. What's the next step for me? What steps toward a digital twin can I make to make my job and life easier?

So starting off with the definition, what is a Digital Twin? SWAN and AWWA combined forces to come up with this definition. Some things that stick out to me here are that it's a digital system, primarily. That it's dynamic. these things are moving and updating.

We've got new improvements coming into our water systems constantly, different types of storms coming in. We need a digital representation to really help us understand what's going on in places where we don't have sensors, for instance. But it also needs to enable new insights.

If all we're doing is making a prettier picture and making a better representation of what's happening, it doesn't do us any good if it doesn't give us more actionable intel. So we're trying to make sure that in any of our steps, we're really solving problems. We're not just making it prettier. We're not just making it easier for people to understand, but we're actually solving specific problems and optimizing our solutions.

So talking about the problems, what types of problems need solving? I gathered some information to make this word cloud. Hopefully, some of these things stand out to you. Some very specific one, like, how do I deal with PFAS, or more globally important, how do we manage finances. How do we get funding for the things that we really know we need to do? We've got aging infrastructure.

All these different problems are going on in our heads and have different priorities in every different location we go. But we do find that most of these problems are pretty universal, that if you're struggling with this as a water utility, the water utility next to you is probably struggling with the same things. So I put them into some categories to get us thinking about what types of things can we solve with a Digital Twin?

Obviously, wastewater treatment, stormwater management, climate change. So these are big picture things that's not really a problem. These are categories of problems. Is it a financial or budgetary problem? Is it resource management?

These are the categories. I dug in one level deeper to get some real problems. So what are your problems at your water district, or sewer district, or in your country, if you're dealing with flooding? What are some of the things that are concerning you?

Is it urban flooding? Is it cybersecurity? Is it permitting challenges, finding funding when there are shortages around? Is it the droughts and water scarcity end of the spectrum? Or is it the, I've got way too much water, I need to get rid of it? All these different water problems are all around us.

So I want you to think right now as you watch this about, what are your current biggest problems? What are some things that you really need to solve and that you're struggling to solve in your day-to-day? And I want you to think of those because I'm going to show off how a lot of your counterparts are solving these exact problems with a Dynamic Digital Twin, or a water-focused Digital Twin?

And this is going to be pretty rapid fire. And I'm going to show you things from all over the world. So we're going to have a great tour here tonight about water systems, sewer systems, flooding, early warning systems, how people are saving money, how we've opened up some recreational possibilities and some canals. Just some great wide variety of ways that people are applying a Digital Twin to their water-related infrastructure, and how it impacts many things around them, and how it really helps operations and engineers work together.

So kicking it off, we're going to go down to Australia and Southern Australia water. I was able to visit SA Water on a trip several years ago to really find out how they were using their hydraulic model in a very live way, connecting to SCADA data and doing demand forecasting. But the problem they were trying to solve originally, they were worried about securing their water supply, about lowering any kind of operational risks. If something goes down, am I going to be able to really maintain my working conditions and maintain everyone with the right pressure in their system?

But one of the biggest things-- it was one of the factors, originally, but it was one of the biggest factors in the end was reducing their electricity consumption. Their solution was to create a live model, as I mentioned, connecting their hydraulic model to SCADA and being able to do demand forecasting that can accurately project what's going to be happening in the next 24 hours and give them a real good idea about what they can expect operationally.

The impact that that has and the outcome that that has is that now they can manage shutdowns a lot more confidently and quickly. So if there's a water quality issue, or they have to take a tank offline for maintenance, or there's a break in the line and they need to shut things down, they're able to manage that on the fly and run these live simulations based on the latest and greatest updates to the model, and react very quickly.

But the coolest thing and what was totally unique for any of my experience in modeling in the US was that Australia has energy rates that are changing constantly. So they-- in their control room, they had a running forecast of what the anticipated electricity costs were going to be, and they could compare that with their model and what they expected to be using coming up. And then they can make decisions and say, hey, electricity prices are really spiking.

I can get away with not filling up the tanks right now. I can wait till tonight, and we still have plenty of pressure. Everyone's going to be covered. And they can manage things that way and continue to provide their customers with a more reasonable price product in the end. And it was the first time I've really seen that dynamic be so tightly linked to a Digital Twin being able to affect cost and electricity so closely.

The next one is Consolidated Mutual Water. This customer's in the United States. They were struggling-- as they were combining different water systems, they were struggling because they had two different hydraulic models.

They had data coming in that was being used in these different systems from Trimble, Badger Meter, Sentryx and SCADA. They had all this data, but they wanted to put it all in one spot. They wanted one hydraulic model, and they wanted to be able to eventually connect that hydraulic model to all of that data that's coming in.

And so their outcome and what their Digital Twin is shaping up to look like is a platform on Info360 Insight that's gathering a lot of this data that otherwise was sitting dormant, and being able to apply that to their hydraulic model. This is still a work in progress so we don't have all the exciting news to say how many ways it improved it. But as they finished their implementation on a single pressure zone, they're already seeing some benefits of just having access to that data and being able to do some calculations that they weren't able to easily do when all those data sources were spread apart.

This is probably the largest scale-- actually, no, it's not quite the largest scale on our docket tonight, but it's one of the two large, large ones. London, this one's the most people affected. London runs a combined sewer and stormwater system.

They have created this Tideway Tunnel at Thames, and its purpose-- the tunnel itself's purpose is to take any of-- whenever there's a big storm, to take any of the overflows so that it can be safely conveyed away so that it won't be running into the Thames River. This is 25 kilometer distance of pipe, and it's 72 meters in diameter, as you can see that man standing at work, or standing in the middle there.

For those that don't know metric units very well, that's 15 miles long and 24 feet high. This was just a really cool project that's been going on for quite a long time. There are parts of it that are operational. There's more planned to improve upon it.

But one of the concerns as they were constructing it was, wait a minute, we're going to have people in these low lying spots in these tunnels that we want water to be in, eventually, that we expect water to be in. But they're going to be doing construction and maintenance activities. We need to make sure they're safe.

So their solution was to create a live hydraulic model that could be connected to this system. And the end result is, by running any kind of incoming storm on a regular basis, on an ongoing basis, they now can be alerted when a storm event might cause higher flows. This is an example here of an alert that got created that said, hey, within the next six hours, it's tracking pretty likely that in this specific location we're going to get depths above 1 meter.

That's super dangerous for anyone working in those conditions. We need to pull everyone out if anyone's planning on working that day. So actionable insights protecting their employees, protecting their operators and anyone doing construction in that. It's a fantastic use of a Digital Twin.

The next one is very near Thames water, just up north a bit in the Scottish Canals. I won't spend too much time here because we actually have a great BBC special that's about five minutes long. You can search it up on YouTube about building Europe's first smart canal. You can search for it on YouTube.

But I still wanted to tell the story because it's a really fantastic use of a Digital Twin. The Scottish Canals were an engineering masterpiece. They were made in 1790. It was the first sea-to-sea canal that was ever created.

And over time as people started using other modes of transportation and trade routes were changing, it started to become kind of dirty and dangerous and kind of go beyond its intended use. Again, I don't want to spoil all of it for you, but in the end, there's some amazing things that they were able to do to create real-time-control of these canals. And not only does it help with actual flood events that would happen that might be impacting places where there are already houses and businesses, but because it's a smart canal, they're able to lower the canal when a storm is incoming, and be able to create storage within that canal of 55,000 cubic meters.

And so by creating the smart canal system, they see a storm coming, they lower it. They let water run out of the canal and get it down to a reasonable level and they can capture so much of the storm just refilling the canal, not impacting anyone. The actual outcome of that process is that there's now 110 hectares, or 270 acres, of land available, which is 3,000 new homes, shops, and businesses because of that lowered risk. The areas that used to flood don't flood anymore because they can anticipate the storms.

This is the other really large scale one. This is countrywide. We've got a great example of how having a Digital Twin can actually save lives and really impact a lot of people all at once.

In Malaysia, they are working on, in various phases, a forecasting system so that people can know when a flood is coming. So in about 10% of the total land area of Malaysia, people are at risk to flood annually. That's about 5.6 million people. And sometimes they're not even aware that the storm's coming in or how it might or might not impact them this time.

By creating a hydraulic model that is connected to live storms coming in, they are able to do a seven day forecast and keep track of that seven days. So now, they see a storm that's seven days out and they say, OK, that might be a big deal. It gets a little bit closer, oh, it's a bigger deal than we thought. Oh, it's a bigger deal. Let's get people moving. There's flash floods potential. Let's make sure the public is aware.

They post all of this directly on their website. You can see there's a little ticker at the bottom, if you go to that website, that is constantly moving across and telling you where you've got high levels already, whether or not it's a current storm, or just maybe there's other reasons why there's high levels at certain areas. But all of that's just at your fingertips. And it allows them to get the word out significantly faster.

The background here is one of those warning systems that just says, hey, when it goes above 6.5 meters in this specific location, it's dangerous and we need to warn people to stay out of it and stay out of the way of that area. And so these events can trigger people being able to get to safety and take care of their belongings, and their houses and, potentially, sandbag, and things like that, to prepare for these storms as they come.

Next one is City of Oakland back here in the States. This one's quite a bit different than the other ones. This one, they use hydraulic modeling, but this one is more about their asset renewals. They have about 1,000 miles of sewer pipeline that's all kind of getting old at the same time because that's what happens in big cities. They grow rapidly in one decade and they often have their assets start failing all at the same time.

So we need to start figuring out better ways to prioritize it better than just age. So their problem was, we need to get a good plan. We need a defensible plan. And their solution is to create a risk and rehab analysis tool to prioritize and plan for their replacements. And they're going to use CCTV inspections and other existing information.

Now, the City of Oakland has been doing this for a while with us. They've used some of our desktop tools. They are now moving on to some of our cloud-based tools and they're-- at this same conference, there's a whole presentation for this project. So I'm not going to get into a lot more detail. But it's exciting to see how these web-based, cloud-based tools are able to help operations and engineers really speak the same language, and really see how their system is reacting in different situations.

And one of the best ways-- or one of the best ways to improve and to have a plan is to make sure that it's not just based on some engineer's judgment only, that there's a defensible way that an engineer helps set it up, and the operations team helps set it up, but that it's repeatable and that it doesn't matter who's working that day, you're going to get the same results either way.

The last one of these I have is Colorado Springs Utilities. A group of our team just recently got to go and visit them and kind of understand how they're using their Digital Twin. They've been customers of ours for a long time.

Their original reason for jumping into live hydraulic modeling was a new water treatment plant back in 2014. So they knew that having a new plant, figuring out where it was going to go, and how that was going to impact where water was coming into their drinking water system, it was important to them that they were forward thinking and thought about a model that they could trust moving forward, a model that they could update on a regular basis really easily. And by creating that live operational model, they've got a model that runs every night at midnight.

And so when the engineer comes in, he can see how it's been tracking for the last eight hours, and can track it throughout the day and doesn't feel as inclined to do things manually when things are following a pattern. So what the result of this is that we don't have as many water quality problems. Chlorine residuals improved once they had this more regular controls and they could tweak those controls based on water quality now instead of just telling people to keep the tank full, or fill it up.

I've heard in an operations room before, oh, I can tell so-and-so was working today because he always likes to fill up the tanks for his next shift, for the next person coming. That is not the most efficient way to run a water district, believe it or not. So they've really found some savings in chlorine residual improvements.

They can validate their model much easier because SCADA's connected to it. They do fewer control adjustments throughout the day. And ultimately, they have just-- if you talk to them, they're just excited about what more things they can keep doing. They know they've made some great strides, but there's more things that they want to keep doing.

So that's our whirlwind tour of a lot of different locations and what they're using Digital Twins for. Now, we're going to get that introspective part of the presentation. So whether you are in charge of a water supply system, stormwater, wastewater, whether you're in asset management or using the digital data and in charge of the IT side of things with SCADA and other IoT devices, I want you to think about your current situation and your role.

And if you aren't in the industry, think about-- you could think about things in your industry, or you could think about things that are happening around the world that we just talked about. We can think about things that are happening in our country, in our city, on our streets. What are some water-related things you've seen in your neighborhood, and are these things that you could bring up in a town hall meeting?

The opportunities for improving our water resources are pretty endless. So I want you to think about your situation and how-- and what you have control over and what you can impact. And our first step is, which of my problems need solving first?

I think about all the different things I did as a consultant, and the things that I interact with customers, and find out that they struggle with. I want you to think about, is this a financial thing? Is it a asset and data problem? Is it public health, energy cost reductions?

Think about some of those things. What is it? Is it, oh, I'm frustrated because my demands for my model are never up to date? Or, gosh, our city's growing so much, are we even going to be able to keep up with this? Or our flooding in this one particular part of my city is really causing us to be in the newspaper every week, and that's been really hard for us. So I want you to think about what problem can you think of that you can solve first.

Next up, we're going to assess our current situation. So when was the last time I updated my hydraulic model? Was it five years ago that I calibrated it? Do I use it often?

Maybe your role doesn't really require it, maybe that's not a big deal. What process or task takes longer to do than it should, or is harder to do than it should be? We've all got those types of tasks.

Think about some of the data that you deal with and the information, is some of that useful to other people in your organization? Maybe it's not. Maybe you have some data that you know is being collected that no one is using and you don't know what it's for. It's worth the conversation and the thought. Is there an improvement that you know of that you've been putting off?

These are some questions that we need to think about our existing role and what our part is, and what we can improve. And I want you to think about your tools that you're using, whether you're on the engineering side, or the operations side. As we move toward this Digital Twin, we've got different problems and we've got different tools to solve those problems.

I haven't gone in-- I've purposefully tried not to mention every time InfoWorks ICM, or InfoWater Pro, or InfoWorks WS Pro, and Insight and Asset and Plant. I'm trying not to inundate you with when each of these are being used because I don't want you to worry about I don't have that tool right now, or I need to-- I want you to think about the problems. So with the problem that you're thinking about, what do you know about your tools? Do you need an additional tool? Do you know that your tool doesn't cover it?

We've got people at AU. We've got people you can reach out to anytime to talk more about those tools. But these are some of the tools that we have at our fingertips for creating a dynamic Digital Twin for water systems.

The other question there is, do you have the right tool but you need to use it more, or you need to find out how to use a certain aspect of that tool that you weren't aware of? On that note, I've got-- I wanted to put in a little plug here that sometimes there are opportunities to change how we work with our existing tools. Just this year we have released Cloud Modeling Service that's connected to each of these hydraulic modeling and drainage design software pieces. And it is changing the way people work.

And I've got this quote here from Denver Water. He says he's up and running, and the latest InfoWater Pro update, and his fire flow sims went from 1.5 hours to five minutes. That's running it on his local machine, took an hour and a half. And because of the beauties of parallelization in the cloud, he was able to get the same results in five minutes on the same model.

On the storm sewer flood side, we've got a customer that was using cloud modeling inside of InfoWorks ICM that relayed a little bit longer of a story about having a project where he was working on six models, that each were taking two to three hours to run. And on his own machine, those had to be done back-to-back. They couldn't be parallelized. And it was taking 12 plus hours to get through all of that simulation time.

And once he moved his project to the cloud, each simulation only took about two minutes and they were all running simultaneously. Now, I am making no promises that you will see those same results. But depending on the type of analysis you're doing, there could be huge ways that you could be improving the way you're using the tool that you're already using.

So I want you to reach out. I want you to find out what we're doing. My team has been working on so many really fun things that really will change the way that we think of Digital Twins. I hope today has kind of inspired you in thinking about what you might be able to do, what next step you can take.

And the last thing that you can do now is to reach out and ask questions. Feel free to reach out to me, anyone on my team, your local rep. We would love to talk to you about Digital Twins. We'd love to talk to you about new things that we've put in the software that you already have that will help you on your way to Digital Twins.

We'd love to talk to you about ways that your neighbors might be using it, or what other tools we have in our quiver to supply you with ways to solve these problems. I'm really excited about just the changes that I've seen in my almost 16 years in the industry and just how powerful of problems that we can solve as engineers and operators. We're doing really important work, and I hope everyone that's involved in this work knows what an impact we can have.

And again, thank you so much for spending this time with me.