Water Hammer in Water Systems: Analyze and Mitigate Using InfoWater Pro

YOUSSEF AL FAHHAM: Hi, everyone. Thank you for joining our AU presentation this year titled Water Hammer in Water Systems, Analyze and Mitigate Using InfoWater Pro. My name is Youssef Al Fahham, and with me today are Shawn and Nathan. And we're all going to introduce ourselves in just a few seconds.

So before we begin, just wanted to post a safe harbor statement regarding any forward looking statements that we make pertaining to any of our software. And here's a brief look at the agenda. So we're going to start by covering a little bit of background in theory pertaining to water hammer, which is also commonly referred to as transients or pressure surges. And then we're going to cover some common pitfalls that we see being around the whole concept of transient modeling or implementing transient mitigation.

We're finally going to cover the actual modeling in InfoWater Pro and how that can be executed. And then we're going to have a call to action and provide some resources to you before we wrap up with some question and answers. So as I mentioned before, my name is Yousseff Al Fahham. I'm the technical marketing manager for water products at Autodesk. And I'll pass on the mic to Shawn for him to introduce himself.

SHAWN HUANG: My name is Shawn Huang. I'm the software development manager for the InfoWater Pro. I've been with Autodesk for 12 years.

NATHAN GERDTS: And my name is Nathan Gerdts. I'm the product manager of our desktop water modeling software at Autodesk. And I've been with the Innovyze team for eight years. Back to you, Youssef.

YOUSSEF AL FAHHAM: So thank you, Nathan. And now we're ready to kick off our presentation. So starting with the background. So a little bit of background about the history of water hammer, pressure transients, hydraulic transients, pressure surges. These are all terms referring to the same thing, which is any pretty much rapid change in flow that may cause a pressure spike in your distribution system or in your piping network.

So the history of pressure transients goes all the way back to 1800s, where the classical research originated. And, actually, the fundamental water hammer equation by Joukowsky, it was researched and delivered during that time. Then later in the 1900s, there was a lot of mathematical analyses and physical experiment experimentation to validate the math behind it. And then recently in the 2000s, a lot of the computer models were commercially available for use on projects and research. And that's pretty much roughly the 200-year development of water hammer theory and in practice.

Next, one may ask, OK why are we concerned with water hammer? Why are we're needing to learn about water hammer? Why do we need to know about it? Why do we need to analyze it in our systems? The challenge is simply the availability of content in curriculums that support engineers, designers, utility planners to understand and implement corrective or protection devices for transients is somewhat scarce.

There's not many curriculum in colleges that actually talk about water hammer or hydraulic transients. There's not a handbook that strictly explains how you can, number one, understand your transient issues and how to mitigate those transient issues in your system. There's not one single source that takes you through how to analyze and look at transients in your system. There are multiple approaches that different professors, different academics, different companies, different utilities have towards the whole challenge with water hammer and transients.

And so our thought, as we develop this presentation, was trying to provide the sources that we have come across from our involvement in the industry and from our involvement as a software company that develops software for hydraulic transients. So this is a-- at the end of this slide, there's a quote that I came across reading one of the research papers, which says a little knowledge can be a dangerous thing.

And that is very true to transients because a lot of the theory and application in hydraulic transients actually challenges traditional wisdom. And it nullifies rules of thumb that are somewhat common in industry. And we're also going to be discussing all these things at a later slide.

So the consequences of either not understanding the risk of transients or not implementing the correct protection for transients can be pretty much summarized into four key points. The first one is you can experience maximum pressures, extreme maximum pressures that can rupture pipelines, that can deform pipelines, or that can cause environmental contaminants to go into the system, to go outside of the system into the environment. The other thing is you can have extreme minimum pressures, which can collapse pipelines.

It can cause air cavity formation. Or it can cause contaminant intrusion, where contaminants get sucked into or pulled into your pipeline and distributed into, say, customers at water systems. The third and fourth point is you can experience a lot of pipeline movement. If you don't protect your pipelines or your distribution network from hydraulic transients, you can also experience severe vibrations in water and wastewater systems.

To demonstrate the potential catastrophic consequences of hydraulic transients, just wanted to share this quick video which shows a pipe burst that happened. And you're seeing how destructive these events can be. It hasn't only impacted the water system. This has impacted an entire neighborhood or potentially an entire village or even an entire town and not only from the disruption of maybe losing your water service or compromising your water service, you're also impacting the physical world that people reside in, that communities reside in, that industries reside in. So the risk is huge. And that's why we need to be informed in how we approach analyzing hydraulic transient events and mitigating against hydraulic transient events.

So I have a few case studies that I wanted to share before we even dove deep into the theory. And these are real world case studies that are available online. You can read up on them. But I just wanted to cover some of the consequences of, again, hydraulic transients in the real world.

So in the US in a six-foot diameter pipeline in 2008, there was an uncontrolled pump shutdown, which created a sinkhole under a major highway. The sinkhole was roughly 40 feet wide, 60 feet long, and 16 feet deep, where two million gallons of water were wasted just due to that transient event that potentially could have been mitigated. Another case study was in Libya in a 13-foot diameter water main in the year of 2012, where due to a rapid valve closure, there was 53 million gallons of water that was lost and wasted.

The third case study we covered today was-- it took place in Japan in the year 1950, where an eight-foot diameter penstock, due to operating errors, spread a hydraulic transient pressure wave, which burst the penstock and led to three deaths, and up to $500,000 in damages. And then, finally, the final case study took place in Russia, where a 20-foot diameter penstock in the year of 2009 failed due to the sudden wicked gate closure, which led to 75 deaths, 40 tons of oil spilled, and 400 tons of fish kill.

So you're seeing the consequences of hydraulic transients. It's just beyond disruption in water service. It's beyond that. And the impacts reach the entire community in many different ways. So understanding transients and mitigating against transients becomes a very critical task.

And that's what we aim to achieve through this presentation is introduce the theory, introduce the risk, introduce the modeling tools that you can leverage to mitigate against transients. And hopefully that would enable you to be able to perform transient analyses and protect piping systems and pressurized systems from the risk with hydraulic transients. And with that, I'll pass it on to Shawn, who's going to be talking about the theory of water hammer.

SHAWN HUANG: For the theory, we can start with the hydraulic transient. Hydraulic transients are the transient states between the initial state and final state of a hydraulic system. As shown below, when the pump is running, the junction 11, the pressure is about 120 PSI, which is the initial state.

When the pump is off, the junction 11 pressure is about 105 PSI, which is the final state. On the way from the initial state to final state, we have many transient states with pressure going up and down a lot. So the typical duration for the transient events is several seconds or minutes. And a typical simulation time step is a small fraction of one second.

Next, we can also try to understand the hydraulic transient from the perspective of space. Hydraulic transients are the pressure waves that are generated and spread throughout the hydraulic network as shown in the right diagram. Pressure waves are mainly generated at the location of pumps and control valves. They travel at a sonic speed. And transmitting and reflecting at those junctions of pumps and control valves, as time goes, the pressure waves are dissipating, gradually. And as shown in the plot on the right, the waves are generated more and more and, finally, exists everywhere in the network.

To get a deeper understanding about hydraulic transients, we can also combine both perspective of time and space in this slide. So in the diagram on the left, we see the time axis going down, so the delta H1, delta H2, delta H3, H4, delta H5, delta H6. This is the time axis.

On the bottom, we see the x-axis. That is the distance of the pipe from the upstream end. So we see the point A and point B. And those rectangles, they are incremental pressure waves. So pressure waves, delta H1 was generated in the beginning. And it travels through our system.

So we can see the time step one, time step two, time step three until the time step six, the delta H1 in the dark blue rectangle. It's always there. It's going out of the system. And then delta H2 to the delta H6, the delta H6 is just happening. It only travels just one time step and only one rectangle there.

So what is the pressure change at the point A and point B. So for the point A, the pressure change will be delta H1, plus delta H2, delta H3, delta H4, delta H5. And, finally, delta H6 and more. And how about the pressure at the point B? It is just delta H1 because only that pressure wave arrived at the point B location.

So in other words, hydraulic transients are created rapidly, incrementally, cumulatively, and with a destructive impact. So if those positive waves build up-- and then it could lead to high pressure, and then pipe burst. If those negative waves are building up and then it leads to low pressure-- and, finally, it will be low pressure, negative pressure cavitation, and pipe collapse.

So what are the events leading to those hydraulic transients? Many system changes can lead to hydraulic transients. So some events are mild, for example, like water consumption changes or tank level dynamics. Some events are controlled.

For example, we can start a pump. We can stop the pump as we want. We can also speed up the pump or speed down the pump. And, also, we can close the valve, open the valve, or modulate the valve or even we can control the hydrant flushing. Unfortunately, there are some uncontrolled system changes, for example, pipe breaks or pump trip because of a power outage.

There are multiple ways to simulate the hydraulic transients, including this Joukowsky equation. So Joukowsky equation is the easiest way to do a preliminary analysis on the hydraulic transient. So, first, we calculate the wave speed. So wave speed c is a function of the pipe diameter, pipe thickness, the pipe material, property, the water property, and also the pipe constraint coefficient.

Next we calculate the velocity change. So we have initial state. We have final state. So we calculate their velocity. And then we are able to see the velocity change. With those wave speed c and the velocity change delta V, we are able to estimate the pressure wave for that one.

Delta H equals c over g, delta V. So c is wave speed. g is a gravity constant. Next, we can assume waves are all generated instantaneously. And there is no reflection. So with that assumption, we are able to get to maximum surge potential, maximum upsurge, and maximum down surge. So the upsurge maximum would be c over g times V0. And maximum down surge will be negative c over g, V0. V0 is the velocity in the pipe.

There are two kinds of numerical simulation methods. One is called method of characteristics, MoC. The other one is called the characteristic wave method, WCM. So with either method, we need to solve two sets of equations. The first set of equations are for mass conservation as shown on the left.

The next set of equation we need to solve is for the momentum or Newton's law. For the MoC, each pipe is divided into many computational units, as shown in the bottom left diagram. And if we see the bottom right diagram, it's an F1 circuit. So the circuit is divided into multiple sections, like straight section, like chicanes, like hairpin, or corners.

So similar way, this MoC, it does the same thing. And then the pressure and flow are calculated for each computational unit. So with that because the network could be really big, including a lot of pipes, so MoC could be quite slow. And MoC, if we look at the bottom left grid, it is created in a way that a solution can be achieved along those characteristic line.

Next is about WCM.

So for the WCM,

for example, like an F1 race, we no longer focus on the circuit. We put our eyes on the individual cars and track their position and speed. Similarly, this method is no longer focusing on pipes. So WCM targets all the pressure waves.

All waves are generated and tracked and travel along pipes. The wave transmission and reflection leads to more new waves. All waves are tracked in terms of their position and magnitude. To calculate the pressure at a desired point, we just search for all the nearby pressure waves and sum them all up.

Pretty easy, right? It was proven to be comparable with the MoC decades ago. And it is much more computational efficient. And this is what InfoWater Pro uses to do the transit modeling. Column separation in air cavity is another big topic in the transit modeling.

It plays a positive role in terms of transient modeling and also provide a negative role too. So if we do the air entrainment, those air will come into the water in the flow. And then it can absorb a lot of energy and reduce the wave speed.

Under normal conditions, the pipe usually is full. But when pressure drops, the water column separates. And water evaporates into a form of air cavity. And the air cavity will grow. And, usually, air vacuum valve will be used to omit the outside air to mitigate the negative pressure.

At a time later, water columns rejoin. And air will be forced out of the system through the air release valve. Usually, the collapse or air cavity will produce a huge pressure spike. The two-phase water and air flow is simulated through the discrete vapor cavity model.

And this model is with a lot of assumptions. Because of those assumptions, the simulation results, usually, they are not so reliable after cavitation occurs. Therefore, when we see the cavitation occurs, we need to know how cavitation happens and how to prevent it.

To mitigate the search impacts effectively, we first need to fully understand how a specific water system behaves. So, first, we need to identify three lines as shown in the diagram for this hydraulic system. So there's a great line, which is the elevation change along the pipeline.

And there's another one called the hydraulic grade line on top. And we have another line we can call it like maximum surge potential, which here is called negative c over g times V0. So if we apply-- if we combine those hydraulic grade line in then the maximum down surge, here we are able to get the dashed line.

And the dashed line will intersect with the grade line. So from this intersection, for those points above the dashed line-- now the grade line is above the dashed line. They will be low pressure, negative pressure cavitation, and pipe collapse. So, luckily, the assumption of that Joukowsky equation is not totally true because we have another important term, time.

So there are three important times that we can play with. They are reflection time, pump run down time, and valve closure time. So what is the reflection time? The reflection time is defined as the waves travel forward and back to the original location, how long it takes.

So usually it is 2 times the distance over 2L over c. The pump rundown time is the time for the pump to run down to zero speed. Valve closure time is the time for the valve stem change from fully open to fully closed. So why do we want pump rundown time or valve closure time to be larger than the reflection time?

The reason is we want the hydraulic system to produce incremental pressure waves slowly and then also wait for the reflected negative wave to come back and cancel out the newly generated positive waves. The large pump inertia means a large rundown time. And pump bypass line, it's useful to supply water from suction side to the discharge side when the hydraulic condition allows.

So, next, we can install some surge protection device to fight with the surge or transient. So, first, we can install a closed surge tank next to the pump station in order to protect the pump station. And it also covers high pressure, low pressure, and also very effective to cover the whole system.

Secondly, we can install a pressure relief valve. When the pressure is really high, just let the water out. Third, we can install the air valve in those high points. Those high points, they have low static pressure, really easy to get to the negative pressure and even cavitation. So those air valve will emit or release air in order to protect the local area.

Number four, we can also install the open surge tank in those high points. But it requires maintenance, a lot. Lastly, we can also install some check valves because those check valves are able to be served to serve as a boundary and then stop the wave to propagate into the neighboring zones. OK, then I will pass it to Youssef.

YOUSSEF AL FAHHAM: All right, so now that we've covered the background of water hammer and hydraulic transients and also the theory and the mathematics behind water hammer and hydraulic transients and some of the computation methods, numerical methods that are available to us, it's a good point for us now to talk about some of the common pitfalls in industry when it comes to the implementation of transient analysis on projects. So to begin, there are some traditional wisdom that you might come across or you might encounter in your journey of analyzing hydraulic transients.

The first one is that demand driven approach is applicable or is suitable or is fit for purpose in transient analysis. Now, it's good for us to take a step back and maybe understand what demand driven approach is. So when we talk about demand driven approach, we're talking about the assumption that demands can be met under all operating conditions.

So independent of the pressure-- so if you have negative pressures or if you have zero pressures, you're assuming that the demand will always be met. Now, the reason why this should not be, I guess, modeled in a transient analysis and the reason why demand driven approach is not suitable for transient analysis is because research has showed us that the demand driven approach, which is also sometimes referred to as the pressure insensitive demand--

It tends to overdesign transient protection devices, which could result in unnecessary additional cost without really offering that much of value when it comes to protection. And so it's just good to be aware of that. As you get into transient analysis, you always want to be going in with pressure dependent demand to make sure you're correlating the pressure with the demand.

Now, the second thing that you might encounter is pertaining to skeletonization. And you might hear people referring to this in different ways. But skeletonization is the process of selecting certain parts of your hydraulic model for inclusion in the analysis.

So let's say if you have a 100-pipe network, you might say that, OK, I have 25 dead ends. Then I'm just going to remove because they're not really going to contribute much to my analysis. And I will say while this might have some reasoning behind it or some logic behind it in hydraulic simulation, so in steady state or EPS hydraulic modeling, this does not apply in the world of hydraulic transients because of exactly what Shawn talked about earlier pertaining to wave reflection and the complications associated with that.

And so skeletonization in hydraulic transients, we've come across several research papers that talk about how skeletonization and demand reallocation can lead to poorly designed and inadequate transient mitigation or in some cases even overdesign. And the reason behind that, again, is just the complex interactions in the pipes due to the hydraulic chains in event. So the solution or the path forward to overcome any challenges with skeletonization is to perform system wide transient analysis with limited or even no skeletonization.

One of the core values of InfoWater Pro is that you have the ability to do that because we leverage the wave characteristic method which Shawn mentioned earlier as well. Another traditional wisdom that you might encounter in industry or you might come across at your discussions about water hammer is the whole thing about looped network configuration.

So when we talk about looped network configuration in the context of hydraulic transients, it can be perceived that the more gridded or the more looped your system is, I guess, the more protected you are. And because you're going to have multiple loops or grids in your piping system-- and so velocities are going to be lower.

And, hence, the pressure change is going to be a less catastrophic. Or it's going to be a smaller rise in pressure due to the lower velocities. However, again, with hydraulic transients, it's more than just the initial pressure condition that you're in or the initial pressure state. It's the complex cycles of wave propagating and waves reflecting constructively, which could cause far greater pressures than one might be able to just assume.

And so that's why when we talk-- when we talk about loop network configuration, it's a challenge that cannot be undermined in performing transient analysis. And the solution to that or overcoming that requires the performance of system wide transient analysis without overlooking these looped or gridded systems that might be available.

And then, finally, a common traditional wisdom is that-- this is somewhat typical in engineering is that you always want to induce or put a factor of safety. You always want to-- if you need, let's say, 10 inches of concrete, why not make it 14, right? When you over design, it's just-- mentally, we associate it with safer, with better. That's going to be what's best for us and our clients and our end customers.

However, with hydraulic transients and water hammer, it's actually counter-- it's counter to that, which is counterintuitive. But over design and hydraulic transients can lead to a greater pressurized or a greater minimum pressure, which can cause greater or more destructive pressures present in your system. And that is because of, again, the constructive wave transmissions and reflections and these complex interactions that happen between the pressure waves.

So these are key traditional wisdoms that we've encountered from our experience in industry. And we just wanted to summarize them and present them to you so that any time you encounter a project pertaining to hydraulic transients, you're always want to keep in mind the whole concept about demand driven approach, which is what we would refer to as pressure insensitive. You want it to be-- you want your approach to be pressure sensitive demands, the whole concept of skeletonization and model size reduction, and then looped network configuration and overdesign.

So in addition to the traditional wisdom, there are some rules of thumb that you're going to come across. And my good friend Shawn had mentioned the use of Joukowsky equation for just a-- we're going to call it a desktop calculation of just getting a sense of some of the numbers and how those might play out in your system there. Joukowsky comes with some limitations.

And let's see 1, 2, 3, 4, 5, 6, 7, 8, 9, 10. I've put down there roughly 11 limitations. And there's more. So when you're using Joukowsky, don't use it in the context of this is absolutely my answer and I have no problem with my system. Use it as just getting a sense of what you might be encountering in your system. And then use that information to then build a transient model and perform a system wide transient analysis that would be the best route for you.

Some of the limitations pertaining to Joukowsky is that theoretically-- and I'm going to say these are the theoretical limitations that research papers have told us about. The first one is that it has to be a single pipeline with constant material, constant diameter, constant thickness and with uniform pipe friction. The system cannot have any wave reflection or cavitation or gas release or trapped gas or line packing.

And the list goes on. I just didn't want to bore everyone with all these theoretical limitations. I just wanted to expose not in a negative way but just inform all of you attending some of the risk associated with just strictly relying on Joukowsky so that as you're approaching your transient analysis projects, you just keep those in mind and everyone would be doing the industry big favors that way. So with that being said, I'll pass it on to my good friend Shawn, who's going to talk about hydraulic modeling and performing transient analysis in InfoWater Pro.

SHAWN HUANG: Thank you, Youssef. This section will cover the single pipeline modeling in a full system modeling. So for the single pipeline modeling, we will introduce a few basic steps of the transient modeling. So, first, we need to identify the boundary conditions.

So, for example, on bottom of the slide, there is a single pipeline from a reservoir pumping water to another reservoir. So this pipeline capture the low and high points. And also it has a reservoir as their boundaries. We also need to determine the initial state to start from.

And, usually, for those analyses, we like to focus on the worst scenario. So under those scenario, the system will be under huge stress. For example, there is a high demand and the pump is pumping at full speed. And next we need to specify a few parameters, for example, because the pipes they have-- with different diameter and different thickness and different sizes materials, so we need to calculate the wave speed for those individual pipes.

And then here in the screenshot on the right, it's just a global wave speed if a pipe doesn't have a wave speed. So this number will be assigned to those pipes. We also need to specify the length tolerance. That's just the computational unit, which is-- here is 10 feet.

And then the info order pool will use those two parameters to calculate the simulation time step. Besides that, we also need to specify cavitation head because the pressure cannot drop unlimited-- there is a ceiling. There is a floor for the pressure to touch. We also need to specify the simulation duration here in the run managers search tab. Usually, it's seconds or minutes.

So once the simulation settings are done, we can go ahead to set up some transient events. So, first, I will here cover the pump operation change. So to set up a pump transient event, we need to go to the pump search data window on the right. So there we can check a few options, if this pump has a bypass, if this pump has a check valve, if this pump has a pump file option.

And, next, we can go to the pump operation change window. From there, bottom right, we have the speed option and also the trip option. So those two options are quite different. So for the speed option, we are able to give it a speed change curve to represent the pump start up or a speed change curve to represent the pump shut down. And it's all controllable.

There is another option, which is a trip. So if we select the trip option as operation change, then we don't need to specify the speed change. We just need to specify the trip start time. And then the InfoWater Pro will calculate the pump speed on the fly for us.

Next, we are going to talk about how to model a control valve operation change. So it's a little bit more complicated than the pump. So InfoWater Pro use a transient modeling term of the active valve. Or we call it AV. It's not a new device or new type. It's just a term for the transient modeling.

It means the valve is changing the stem position during the simulation. So, first, we need to change the valve type to TCV and then assign a minor loss coefficient to the valve. And that minor loss coefficient curve described how the valve restricts the flow with a given percent open position.

In the next pump, we need to check if this control valve has any bypass line option or check valve. And after that, we can activate this valve as an AV, active valve, and then assign the stem change curve. And the same change curve is just a curve to describe how the valve changes its stem position. So it could be from fully open 100% to 20% or from 50% to 100%, depends on the situation that we want to model.

So after the transient model is set up-- and then we can run it. And then we can view the model results. So the first thing we need to check is initial state verification. So in order to do this in the run manager, there is an option called hydraulic run only. Check it on, run it, and we are able to see the initial condition for the search run in the model explorer.

So in the bottom right, there are two Model Explorer screenshots. One is for the hydraulic simulation results. One is for the surge hydraulic run only results. So we can pick the same pump. And we can compare their flow and hack in to see if they are consistent or not. If they are different, we need to figure out why.

And we can also use this tool to view the pressure profile along the pipeline to see how the profile changes as time goes. So it will go up and down. And we are able to see the top maximum pressure and then the bottom low pressure.

As long as the initial condition is good-- so we can go ahead to review the hydraulic the trending modeling simulation results. There are a lot of tools that we can use to view the results. It could be a single element time series graph. It could be a multiple element tabular report.

And then we can put those output results into the map for a symbology chain, color coding. And this slide shows two widely used review tools. One is called the search node range report. So it is just a tabular report. And then it can show the maximum value for that element for that junction. Or it could show the maximum pressure value for all the junctions.

So from there, as shown in the bottom left, we are able to see there are a few junctions they have a cavitation. The negative 14.39 is the cavitation head. And then there is a maximum pressure of 310. Yeah, so those are the things that we can locate. We see their location. And we can just right click to bring those element IDs into the map and see where are they.

Another tool is search node group graph. That is the graph that shows a bunch of junctions. And from this graph, we are able to see how the pressure wave travel and propagate. We are also able to see when the high pressure will happen and also when the low pressure will happen and which one is the troublemaker to have the cavitation first.

And, next, we will talk about how to mitigate cavitation. So in the slide, we see a dashed red line. That is the cavitation head, is the floor for the pressure. Pressure cannot go below. And we can change the cavitation head as needed, depending on the situation. But usually it doesn't change a lot.

And InfoWater Pro also provide this tool called a surge animation. So it imitates the pump station, the reservoir, and then the relief in the pipeline along the landscape. So from there we are able to see the pressure profile going up and down. When it hit the elevation, pressure go really low. There will be a cavitation.

So with that animation, we are able to see there is a cavitation. So where to locate the cavitation? We can go to the search node ranging report to see which junction is having cavitation. And when to have cavitation, we can use a surge node group graph to figure out which element has the cavitation first.

In order to mitigate the cavitation or to prevent it, to avoid the cavitation, we have to ensure the surge protection devices. So, usually, next to the pump station, we like to install the closed surge tank to prevent the whole system. So when there is a high pressure, it will absorb the water. When there is a low pressure, it can provide water to the system.

So, usually, the system will be protected pretty well. But there are still a number of high points probably not able to be covered. In those situations, we can install an air valves in those locations, high points. So the air valve will emit air or release air in those locations, hopefully, to protect the neighboring areas. With that, I will pass it over to Nathan.

NATHAN GERDTS: So that was an example of an individual pipe profile. And now we'll take a look at a running surge analysis on a full system. And to hopefully bring things all together, to bring the different ideas, I'll do this as sort of a software demo to walk you through from start to finish the process.

And so I'm going to walk through five different phases. First, I'll take an existing model run for typical planning purposes with normal standard hydraulics and we'll adapt that model. And then we'll configure our surge event. We will assess the impact using some of the tools we just looked at. We'll configure different scenarios to try to resolve the surge analysis to alleviate cavitation issues. And then I'll show some tools how we can publish and share those results out with our team.

So starting first, let's take a look at adapting a model. This is a smaller system that I'm going to run the surge analysis on. And to get started, of course, we'll need to make sure that we have a calibrated model, number one. So you'll want to make sure that, of course, the flows and the pressures are matching up on typical average operating conditions that you're running your surge analysis for. Obviously, if your model isn't calibrated, then your surge analysis isn't really going to be reliable. So that's a common step one.

And once we've done that, you'll also need to configure the valves. Shawn showed a few ways how valves can be configured. This does require a different interface from the normal valve. So you can set up PRVs that will reach a certain target set point.

And then you can also set up interactive valves using a throttle control valve. These can include bypass and check valve. And with these you can give them a time series of that stem change time series. So you could close it, open it. You can vary the valve throughout the simulation. So that's a useful valve type. And that's typically what you'll turn to if you need to apply changes during the simulation.

All right, and then one of the next things you'll need to make sure with your model is ensure that you don't have any negative pressures in the model. So one way you can do that is take a look at summary maps. Right now all of my pressures are OK. You can also, of course, check out range reports on the junctions. Make sure your pressures don't go below zero.

You can end up with negative pressures. Sometimes the typical planning models, when you have groundwater pumping and you have those points upstream of pumps-- and, normally, that's not really an issue in typical modeling. But when we're actually modeling surge and we're looking at cavitation, obviously, those negative pressures can start to cause problems when we're actually trying to simulate that cavitation side.

And then, fourth, do we skeletonize the system? "Skeletonize" is that term where we simplify, trim off dead ends. And as Youssef was explaining, if we do that, we actually can change and artificially alter the way that pressure waves will reflect and bounce through the system.

And because we're using the wave characteristic method that Shawn explained, we can actually do it pretty efficiently throughout the whole system without skeletonization. So in this case, nah, I'll skip that. And we do have case studies of customers running surge analysis directly on the full system, no skeletonization needed.

All right, so now we're ready to configure the event. So what are we going to simulate in this model? I'm going to model a simultaneous power outage on four different groundwater pump wells on this western side of the system. So this is a bigger event having a simultaneous power outage at multiple different pumps that are all getting tripped at once. But it's something that could foreseeably happen.

So to do that, you can just select the pump. You come over to the pump controls within surge. And I'm going to check on the pump disturbance type. So here I can choose speed. And then I can define a curve. In this case, it's a simple time series. It goes from 1 to 0. I'm giving it one second for that pump change.

Obviously, the slope does matter. So you can play with that. And in the interest of time, I've done that already for the other pumps as well. So if I just select the pumps with data, the tool can show you where those pumps are in the model. And once we have that event set up, let's go ahead and run the analysis.

So make sure that you're on the surge tab of your run manager. I'm going to turn on pressure sensitive demand, which was explained earlier. That better represents demands during these low pressures. And then you're ready to click Run. And so now the simulation will run.

And once you're done, you can go ahead and load the results. And so now we've just configured the surge event. So let's take a look at assessing the impacts. So just like most software, you can select any point in your system, pull up a point graph. Here I can see the pressure at that junction. Obviously, it's going negative. This is indicating cavitation.

We can also pull together a profile, a series of objects. So how do you do that in a system? Well, I like this tool. You can just select from one node to another, and it will automatically grab the shortest path of pipes in between. It can be useful when you're grabbing profiles within a bigger system.

And then, of course, you can animate that. I'll let that run really quick since you already saw that tool in a previous session. And then one of the more interesting things when you have a whole system is to actually plot a map, for instance, the minimum pressure since in this case with a pump shutdown, we have low pressure waves going throughout the system.

And so here, you can see at a bird's eye view I'm coloring all of those junctions red and large if they go negative. So this helps you quickly understand spatially where the issues are. And so, obviously, our problems are fortunately isolated to that one area where the power outage was.

Sometimes you can be prone to assume that the problems are worse worst right at the pump station. But that's not always the case if you have higher elevations elsewhere. Sometimes the worst problems are not right at the source of the event.

So now that we have assessed the impact, let's start taking a look at how do we mitigate this, how do we look at some scenarios. So let's bring back up that graph that illustrates the problem. And within InfoWater Pro, we can create additional scenarios on the fly pretty easily. We can give each scenario its own surge data set. So once you activate that and apply it, I can make any changes I want, and it will then track those changes for a new independent scenario.

So here, I select right downstream of the pump. I can activate a surge protection device. That's an example of a vacuum valve. But if I switch to a closed surge tank-- this is a useful way that you can add pressure or allow water to flow into the system to mitigate that down pressure surge.

And that's how you set up surge protection devices within scenarios. And then you can simulate. And now let's go ahead and compare the results. So if I click Refresh, you can see now that time series just changed. It still goes negative but not as bad as it was.

Let's go ahead and compare it with several other scenarios that I've already run. So here I have the worst case in yellow and then various different surge protection designs all laid on top of that of that one simulation. So that's a quick look at configuring some of those.

And just a brief review, I know we've already covered some of these. These are just some of the different surge protection devices. Obviously, open surge tanks, these are typically good when they're up on a hill or somewhere at a higher elevation because you can't have it too low where there's higher static pressure.

Closed surge tanks are a great solution downstream of pumps. They can add pressure as well as alleviate the high pressure spikes. So they're a great all-purpose use case. Feed tank is a case where you allow one way flow to just handle those lower pressure surges. Similarly, a bladder surge tank, you can set it at a certain pressure so it'll only kick in once that pressure is achieved.