An Innovative Approach to Emergency Response Planning with Innovyze

SUSAN KNEPPER: Hello, everyone. My name is Susan Knepper. I'm a water resources engineer at OHM Advisors, and I'm going to be discussing today the approach we took with the city of Livonia, Michigan, on their emergency response planning effort.

TIM MEDEARIS: And hi, everybody. My name is Tim Medearis. I'm a solutions engineer. I've been with Innovyze, now part of Autodesk, for six years. And I'll have a little bit of an addendum to Susan's presentation about where Autodesk is going in terms of evolving this type of solution, and bringing it to more folks out there.

SUSAN KNEPPER: Thank you, Tim. So we did this project for the city of Livonia, Michigan, which is actually where my home office is based. Lately, I've been working from home. But since this is where our home office is based as a company, it sits near and dear to our hearts, this project. So a little bit about the city of Livonia. The city of Livonia is located in Southeast Michigan, right outside of Detroit. This city is Michigan's ninth most populated municipality with a population of just under 100,000 residents.

We focused our emergency response planning effort on their water system. So I thought it was pertinent to provide some water system information. The city has approximately 485 miles of water main in their system, with diameters ranging from 6 inch to 36 inches. They have water main as old as the 1920s, which they're actively replacing, and a lot of corroded cast iron water main. The city purchases water from the Great Lakes Water Authority. And for my fellow water nerds, on a demand perspective, they have an average day system demand of 11 million gallons per day, with a peak hour demand almost three times that amount.

As I mentioned, the city purchases water from the Great Lakes Water Authority through eight master meter feeds. These master meter feeds basically border the city on the North and South end. And the city is surrounded by Great Lakes Water Authority pump stations. And they're on the downstream side of these pump stations, so they're seeing a lot of high incoming pressure from the Great Lakes system.

And so they have multiple pressure reducing valves throughout their system. A lot of these valves are located in the same vault as the GLWA meters. Some of these valves are located immediately downstream of those meter valves, and then they have a few internal to their system due to elevation changes.

The city has four pressure districts-- the Northwest, North, West, and East pressure districts. And these pressure districts are separated by closed valves, check valves which can be finicky sometimes, and a few of those pressure reducing valves.

The city of Livonia is no stranger to hydraulic modeling. The oldest model I could find on file in our company files was a WaterCAD model created by a previous consultant. In 2010, OHM took this model and updated a H20MAP which is now a retired software. And then in 2018, as part of a water master planning effort, we totally rebuilt their models from scratch and InfoWater. And then as part of this project, I converted that InfoWater model to the InfoWater Pro using that conversion tool, which is pretty seamless, really just to see what the InfoWater Pro was about.

So we run the city's model for multiple things. As I mentioned, in 2018, we updated the city's model as part of a water master planning effort, totally recreated it. And this was to identify capital projects to support their water main system. The city often calls us for pressure checks or fire protection checks, and unfortunately more recently we've been running their models for emergency health. So what can we do under these emergency situations they've been experiencing more frequently?

Which is really how this project started to formulate. So in 2018, the city lost one of their large master meter sources, LV-14. And this was due to a pressure surge in the Great Lakes Water Authority system. Not only did they have a break at this master meter source, they also had multiple internal city-wide breaks that were occurring. And the city, not knowing what was going on, was in this reactive phase, and put the whole city on a boil water advisory. And this break actually closed down a major highway that runs East to West along the city, eastbound 96.

So we actually have a video to show you the magnitude of this break that I'm going to play. So as you can see, it's a very large break, and they were losing a lot of water. You can't really tell in the quality, but behind this the highway is literally right behind it, where my arrow is pointing and it dipped below the ground surface you're seeing here.

We have a few more pictures to show. So as I mentioned, the city was in this reactive state. Here's some city staff trying to isolate this break. They do not know what valves to turn. They were just turning valves, hoping one of them would stop this water from gushing. Gushing over the highway, so you can see in this bottom left photo. That is eastbound 96. And if you are driving down the highway that day you probably saw something you've never seen before, which is a waterfall over that overpass. And although that was a unique sighting, it's actually a very hazardous condition.

So this is February in Michigan, which is often very cold. And you do not want to be creating Icy roads ever. You want to avoid that. So it was creating a very dangerous situation. And so isolating this brake as fast as possible was really important.

So after they are able to isolate the break, they were able to get into that vault to see what the damage was. And as you can see, the top of this PRV was totally blown off as a result that pressure surge. The PRV is obviously very old itself, which didn't help the situation. But that is really where the water was escaping out into the atmosphere.

So the city responding reactively, sent out a generalized communication to the entire city, even though it wasn't really necessary for everyone to be on this boil water advisory. They just really didn't know the magnitude of this break, and who was really affected, and they just knew they were getting all these calls everywhere else. And they were in a panic mode. They also sent social media updates through all their social media outlets, such as Nextdoor and Facebook, letting residents know that there was a level of service interruption.

They had critical user impacts. So Michigan Dairy, a large dairy production facility located in Livonia, was within 24 hours of not being able to provide milk to the Southeast Michigan Krogers as a result of this break. Also the city was hosting a hockey tournament around this time. And at the main, it looked like these hotels, where all the hockey players were staying, were seeing desirable pressures. But because of the elevation rise and the in the rooms up on the top story were much higher than where that water main was, they actually weren't able to flush the toilets at the hotel.

So they were getting calls from these hotels, which they never really considered a critical user until this time. And then they were contacting their consultants, which I typically run their water model, and I had happened to be out of town during this time. I don't know if that was a blessing or not. But I did get texts from coworkers asking, what can we do, what can we do to help the city out. Because this was a chaotic situation.

So then in 2018, the America's Infrastructure Act became law, which is ironically the same year they had this major break. This was an EPA requirement that required communities, based on their population size, to complete a risk and resilience assessment, and an emergency response plan.

The city saw this as an opportunity to create an enhanced emergency response plan. Because the city recognized that they had aging assets that were more likely to fail due to their age and their condition, more frequently than they have in the past. And they were experiencing this firsthand.

They realized their communication strategy wasn't as effective as they wanted it to be. And they didn't have any operational guidance in place to help inexperienced workers in the event that these situations were occurring. So they understood this didn't exist, and they wanted this to exist for the city.

So that's where they hired us to complete this enhanced emergency response plan, as an add-on to that OWEA requirement. And so we sat down with the city to identify project goals. And one of their goals was to improve their communication strategies. Another one was to understand and model the potential service disruption scenarios that could impact the city, and document the operational protocols for each one of those service disruption scenarios, so that inexperienced staff had something to reference if this were to happen again or ever.

So the city wanted to understand not only who their critical customers were, because they had already a good idea who they were-- but they were learning that they had new ones that they didn't really consider-- but what their needs were as well. And so going above and beyond the regulatory pressure requirements that states usually provide guidance on, a lot of their customers have their own operating pressure requirements to effectively operate their facilities.

The city was under a different emergency situation a little bit more recently, where they started to identify dialysis centers as critical customers. These dialysis centers needed at least 60 PSI at the water main to be able to effectively run the internal filtration system. And so really understanding all of those critical customers and their needs, and getting this in a document that can be updated routinely.

The city wanted to identify different scenarios to have operational protocols developed for. So the city wanted us to run in the hydraulic model what it looked like if they lost each one of their master meter feeds at different times, and their key city facilities which they identified as their pressure reducing valves for now. And they wanted these scenarios ran under two different demand situations. So max based system demands, a hot summer day. If you were to lose one of these, that's more of an emergency, not a planned scenario. But we want to be prepared for it. And they also wanted these ran under average day system demands.

So if they were purposely going into a meter vault to replace a meter that has aged, or purposely going into a pressure reducing vault, to replace a PRV, they're not going to do these in the middle of the summer. They're going to plan to do these outside of those high demand times, and so running those under average day system demands.

So after we sat down with the city to identify each one of these scenarios, we had more than 40 scenarios they wanted us to run. And that's a lot of scenarios. And as soon as I started getting into the modeling, I realized we needed to define some level of service goals for the project to really help guide us through the modeling, and make sure we are all on the same page with where we wanted the modeling to go.

And so we defined three level service goals. And our first goal for any scenario was we wanted to provide that minimum desired pressure at those critical customers. So identifying where all the dialysis centers are, and making sure or trying as best as we can to get them that operating pressure they need to run their facilities effectively. And if that wasn't feasible, we wanted to at least provide 35 psi and above everywhere in the city system. 35 psi is EGLE, which is Michigan's regulatory agency, minimum pressure requirement under normal operating conditions under all demand scenarios.

If this was an emergency service disruption, and meeting those first two level service goals didn't seem feasible, we wanted to at least provide 20 psi and above everywhere in the city system. And that's EGLE's minimum operating pressure under an emergency that they allow communities to operate under. And so those were our goals for guiding the modeling. And if we couldn't meet any of those goals, what capital projects could we recommend to the city to try and meet those level of service goals is where we went next.

So keeping the model organized was very important to keep everything organized. And so in my scenario explorer, which is when you're running InfoWater where you build your scenarios, I made sure to have two buckets, two parent buckets, one for the emergency disruption scenarios and one for the planned service disruption scenarios. And each child I created would inherit those data sets from their parent.

So for each facility I was simulating as the loss of, they would inherit the data sets from their parent scenarios, and be stored under those parent scenarios, to really keep track of everything. And then data sets on top of that are another thing that get pretty messy in modeling. And so keeping the data sets organized is really key in keeping a model clean, and so that if someone were to hop into this model later that wasn't me, they could at least be able to follow what I was doing.

So every time I had to create a new data set for control, so I was always simulating a facility offline. And so I was always closing a facility. So every time I had to create a new control data set. But any time I created a new data set, I would name that data set based on the scenario name just for seamlessly checking and making sure that I have the right data set referencing to the right scenarios. That way, I wasn't mixing up data sets because that has happened to me before, and it's not fun when you're trying to figure out where you changed something when a scenario is referencing a data set that you didn't plan for it to be.

We also worked with our GIS team to create a geoprocessing tool to improve the workflow and increase the project efficiency. And so they created a tool that basically compared baseline system information. So before any facility failure occurred or service disruption occurred, what does my baseline system look like in terms of pressure? And then you would feed that into this tool. And then we would take the output from after we've changed things in the system to counteract that failure. You'd feed that output into this tool, and then this tool would automatically compare the pre- and post-model outputs, and automatically join that to the GIS layer. So you could see that pretty seamlessly on your screen and while you're working.

And this, as I mentioned, really increase the workflow because before doing this, I would take the output from the model, export them, put them in Excel, do the calculations in Excel, rejoin it to the GIS layer to then see what does this look like. And then God forbid if I had to make a slight change and then do all that again. So this really increased the project efficiency. And it's something we use to date for other communities as well.

And GIS layers are key. Having GIS players on while you're running these scenarios was really important. Because in other models where I don't have detailed GIS information, and I get called by a client to help figure out a situation, I've often just deactivated or closed the facility, not knowing if there's really a valve downstream to really simulate what I'm doing. And that's how I started running this project. And the first scenario I hopped into I just deactivate it, and I ran, and I started making changes to counteract the loss of a facility. And I soon realized as I was putting the operational protocol together, they didn't actually have true upstream-downstream isolation valves.

And they actually had to go pretty outside of the vault that we were isolating to actually isolate and depressurize the vault. And so really having a strong understanding of where those, valves are will allow you to give better recommendations to the client.

So there's a lot of steps that I just went through. But these steps are basically showing what was going through my brain throughout this process. So here's my workflow summary. So I would generate those baseline files that I would feed into that geoprocessing tool first. And then I would simulate in InfoWater Pro, what does this look like if I lose this facility. What is the system looking like? And then I would go do the critical customer checks. Am I meeting that number one level of service goal? And then I would check and make sure I was meeting regulatory pressure requirements. Am I meeting that 35 pci goal everywhere in the city system?

And if I wasn't meeting those goals by simply taking that facility offline, what system changes could I make to meet those goals? And once I was happy with that, I would then push that through the geoprocessing tool to gather my data, and display it in a nice visual way. And then after doing that for each scenario, I then had to bundle this up into a deliverable.

And the deliverable is where things started to get overwhelming, as I started imagining what that looked like. And so when we originally scoped this project, we planned on providing the client a report with a lot of maps. And I want to emphasize a lot of maps, because we had 40 scenarios. And with the amount of detail you need, at least I feel like you would need to effectively be able to utilize this in an emergency, I'm thinking you would need four maps per scenarios to really see everything and the steps you need to take.

And so I'm thinking like 100 maps. And that's a lot of maps, especially when you realize you have to change a title, or a legend, or some type of symbology in the maps. So now you're updating a lot of maps. And that's a lot of data. And so the idea of putting this into a dashboard came to mind, and throwing this data into a dashboard.

And we sat down with the city to see if they were on board with this idea. And they graciously agreed to let us try this out, which I'm very excited about because we did create this interactive experience for them that really is something that they can build upon into the future. Any time they have new scenarios they want us to analyze, we can put it into this dashboard. And it's really providing just a user friendly experience overall.

And so I'm going to share my screen to show you what this dashboard looks like. That we've developed for the city. So here is the city's dashboard. So when I first log on to the dashboard, the first thing that pops up is this baseline system pressure bucket. And there are two different demand scenarios shown here. And so if I toggle between the different demand scenarios, you'll slowly see that the pressures, the colors of the junctions change to represent the pressure ranges under the different demand scenarios.

And this is a benefit that we realized soon after that is in addition to storing all the operational protocol in this dashboard, it's also something the city can use regularly to assess are there system pressures acting normal. So if they were to get a low pressure complaint, let's say in this neighborhood, and they want to know, is this a real complaint or is this an abnormal complaint. They could go ahead and click on a junction in this neighborhood, and see that under normal operating situation, this neighborhood should expect a 60 PSI pressure.

If they're getting complaints from a resident, and they went out and tapped a hydrant and saw, oh, they're way below this normal, they might want to consider investigating a little deeper into that complaint. And so that was just an additional benefit. And that's why this is a very exciting environment to host this data, because it's something you can keep finding more things to add on to, and you can keep building upon.

So these other buckets are showing the different facilities we simulated for service disruptions. So the loss of GLWA meter feeds and the loss of key city facilities. So if I click on the loss of GLWA meter feeds, all of these asset IDs populate, and these asset IDs represent their master meter IDs.

And so you'll see here. There's these two different demand scenarios, as I explained earlier. So if I toggle between the planned disruption, this is average day. This is what they'd want to access if this was a planned service disruption. They're going in, they're replacing a master meter. Or here, it's an emergency disruption. This is something that's not planned, and happening during a high system demand time.

And since I explained the LV-14 scenario earlier, which was a real world scenario, I wanted to go and show you what this look like. So if I click on LV-14, you'll see the steps here populate that are associated with what to do if they were to lose this master meter source. So step 1A is isolate the LV-14 master meter pit. And so when I click on that step, it brings me to that location. And this little table pops up, giving me a little bit more detail. So this is great for an inexperienced operator who really doesn't know where to go if they've gotten the call that this source is offline.

The next step is contact GLWA to close the upstream valve. So anything upstream of these vaults, and really most of the things inside the vault, the city has to contact the GLWA to isolate the valves upstream to depressurize the vault. And here, we're able to provide more information, such as the contact number for GLWA that the city can use to reach out to them to ask them to close their valves.

Close the downstream valve. Close the downstream valve. And this is where it's just so great, where you can store so much information that only exists in some operators heads. And with institutional knowledge walking out the door, we're able to capture that here. So we've added a note here for this step. Close downstream valve 03857. A newer operator might think, oh, I'm closing this valve. I'm creating a dead main to the west of me, of this valve. And I'm potentially going to put all those services on that main out of water if I close this valve.

But because we've worked on the system for a while, and we've worked with a lot of the operators who are in that retirement age, we know that this valve has no services along it. And closing off this valve is not going to create an issue for any residents, because no one is tapped off this main. And so you can put that information anywhere in this dashboard, just to keep track of it.

The next step is contact GLWA to provide more pressure at LV-15, which is right downstream of one of their booster stations, so that it could supplement the loss of that facility that we're simulating, LV-14.

And this last step, which is one of the coolest steps, that was actually the client's idea on for improving their communication strategies. Is this email critical customers link, and so the city is currently populating the email addresses for the critical customers. And if this were to occur, they can click on this link to automatically populate an email that gets sent to all their customers.

So I click on this link. And now they can almost seamlessly send out an email to the customers that are going to be impacted by the loss of this facility, and letting them know that there is going to be a level of service disruption occurring.

So I also want to show you the impact of when they lose one of their key city facilities, and where we weren't able to meet the level of service goals we defined before, and so we created an additional scenario that included capital improvement projects. So you'll see here PRV-04 and PRV-04B. If I click on PRV-04B, since they haven't implemented these capital projects yet, you'll see there's a lot of pressure change and pressures here. And they're seeing pressures in the orange, which over here is between 35 and 20 psi. And not really desirable level service. And they have to go through a lot of operational steps to even get to that point.

They're redistricting pressure districts. They're changing PRV settings. Their contacting GLWA. If I scroll through all these steps, you'll see they have a lot to go through to just barely provide a decent level of service. And so this is where one of the scenarios where we looked at, if we added some new infrastructure would it help this situation? And it did. And you'll see we went from 15 steps before to just four steps with the capital projects that we recommended.

And so once the city does actually build those capital projects and have those working in their system, if this scenario would to ever occur, which it actually did happen. This is when they found out about the dialysis centers minimum requirements. They have a lot less steps to have to go through to meet a desirable level of service.

And so this chart here is showing the pressure differences or the pressure ranges in the scenarios between the baseline, so before there's any changes and after you make those changes. So you can understand how many customers are seeing a large pressure change. And so greater than 10 psi we have as red here, because generally at least the city of Livonia, starts to get complaints from residents or customers when their pressure change is in the 10 psi or greater. And so that is where that tool also came in handy to help be able to build this chart and show those results on the map.

So we're in the part of the project where we're actually implementing this enhanced emergency response plan, and reviewing these scenarios and recommendations with city staff and their newer workers. And we always recommend a trial run of these scenarios. Because during those trial runs is when you actually find that, oh, I had a GIS error. I don't actually have a valve in this location that GIS says. I need to go and install a valve to isolate a vault.

You also determine under the trial runs if you have a faulty valve. In that scenario where I was describing there was so many steps that they had actually went through, we had provided recommendations to them originally. And they went out to redistrict a pressure district, and learned that some of the valves that we asked them to turn and open were not actually turning, and that they needed to be replaced. So doing these trial runs is really important to understand your system limitations.

And that is what we helped the city of Livonia basically develop. And it's something that is just very exciting, because you can keep growing upon it as time goes on. And it's just a handy place to store their data. And it's super user friendly.

TIM MEDEARIS: Thanks, Susan. I love hearing your story here at Innovyze, a part of Autodesk, we hear kind of both sides of your experience quite a bit, where utilities are trying to become more proactive with emergency response planning, and where consultants are trying to deliver more dynamic solutions to the utilities that they're working with. I loved your example about trying to provide something a little bit more dynamic and dashboard-like compared to 120 maps that you would have to create for all those different scenarios.

So this falls very neatly along the lines of what we've seen through utilities and consultants around the country. And so just a little bit here on what next, and what we're working on here at Autodesk. So I'll be talking a little bit about a specific software product at Autodesk. So here's our safe harbor statement. Everything here is currently available, but just in case there.

So what we'll be talking about is a solution very similar to Susan's workflow that's built into we're calling Info360 Insights. So this is a tool that's all about taking live, data often SCADA data information that utilities may have in making sometimes better use of that within a wet infrastructure kind of hydraulic modeling perspective. If you can take some of this information and use it in innovative ways there are a lot of methods to be able to reduce some of these problems you see here on the left.

And one of those tools is very similar to Susan's presentation here today. So looking at that emergency response for those particular main break events that can be very dramatic, as shown in that kind of highway example there, and cause a lot of issues. The idea is you can take that InfoWater Pro model, publish it up to the cloud, put it in a space where it's very accessible and everybody can make these simple kind of what-if modeling approaches here.

So once you're able to publish your InfoWater Pro model one time up into the cloud with Info360 Insight, you have that model there. And you can simply simulate these pipe break assessments and what valves you might need to close in the case of a particular break. So really hydraulic model is required, customer location points, and those isolation valve locations that were so critical to Susan's workflow, and making sure those were accurately represented there in GIS.

Once you have that information publishing up one time, instead of maybe 40 different scenarios, to Info360 Insight, you can then start to create these what-if main break assessments very easily to see, again, what customers might be affected, how long you might be out of service, how long you might be below that minimum required pressure that you want to hit there even during those main break events.

So again, simple inputs here. This isn't a full hydraulic model workflow, where you've got tons of different options and dials that the utility consultant might be doing for master planning. Here, we're really looking for simple things like pipe ID, how long do you expect that main break to be. And then because info360 Insight can be tied to your SCADA information, we can apply those real time conditions that might be occurring at that point that you specify that main break is occurring. It's not a stagnant model. It's looking at of live data feeds for how your pump stations, your tank levels, or your different valves might be operating at that particular time.

And again, so the process from there is then it's going to report to different customers that are affected, how long they stayed below that minimum service level, minimum pressure you see there. And then it's going to locate automatically what were those valves that you need to close. And you can plan different scenarios for what if I get to this valve three hours earlier. How much is the simulation improved? Is this break impact assessment improved, if I can get to these valves earlier? What if a particular valve isn't in operation? And it can give you these results quickly in the cloud, in this very water specific dashboarding tool here.

All right. So again, there's lots of different applications for info 360 insight. I could go on and on about how it looks at non-revenue water, pump performance, water quality analysis. But Susan's presentation I think really highlights an example, a problem and solution, that we're looking to continue to improve and iterate on here at Autodesk Innovyze to make that process a little bit easier for all sides tackling these important hydraulic problems today.

All right. So with that, we appreciate any questions you might have, appreciate hearing from you. And that is our presentation.