gradient background image
30 Jun, 2021

Smart water for a changing environment

Did you know that 15-30% of water that flows through our municipal pipes simply leaks into the soil? Ever wondered why when a natural disaster strikes that cuts out power, we also lose access to clean, safe drinking water?


In this episode of Radio PI, host Rebecca Ahrens sits down with Gary Wong, AVEVA’s Global Industry Principal of Infrastructure and Water, to discuss how data is solving some of the most pressing problems facing the water industry — and our planet.

Featured guest

Gary L.S. Wong, P.Eng., MBA, CPA
Gary L.S. Wong, P.Eng., MBA, CPAGary Wong is the Global Industry Principal of Infrastructure and Water at AVEVA, a leader in real-time, industrial, performance intelligence. He leads their global data centers, facilities, smart cities and water businesses and has 25 years of extensive international experience providing sustainable, strategic and cost-effective digital solutions. Mr. Wong is also the Chairman of the Smart Water Networks Forum (SWAN) Americas Alliance and holds a Bachelor’s Degree in Chemical Engineering, is registered as a Professional Engineer in Computer Engineering, holds an M.B.A. from the Queen’s School of Business and is also a Chartered Professional Accountant.

Topics covered

In this episode, you’ll hear about how big data and smart water technology are helping to keep our water systems resilient in the face of a changing climate. Learn why the infamous polar vortex that turned off the lights in Texas last winter also affected access to clean drinking water.

You’ll also discover how data can be used to reduce energy usage in water treatment plants, about water loss, and about the importance of smart water meters for leak detection. Last, Gary tells a fascinating story about the growing climate-related threat of toxic algae and how data and AI are helping to keep those toxins out of public water supplies.

It can be easy to take water for granted. That is, until clean drinking water is suddenly unavailable.


The polar vortex of February 2021

GARY WONG: In total, I think about 1,200 water systems were impacted, affecting up to about 15 million people.

That's Gary Wong, the Global Water Industry Lead and Expert for AVEVA. He's also Chairman of SWAN Americas Alliance, the smart water networks forum. Gary has been working in and studying the water industry for almost 25 years. I got in touch with Gary because I was curious about how data is changing the water industry, but also because I wanted to ask him about two really big words when it comes to water systems, resilience and sustainability.

REBECCA AHRENS: So Gary, I want to start off by asking you about a recent event that really put the resilience of the water infrastructure in America to the test, and I'm talking about the big polar vortex that hit Texas this past February. Can you just start off by reminding us what happened, how it impacted water in that state?

GARY: Sure, and that's the whole point about having to be resilient, right? It's being prepared for the next unknown and that, I think, really took a lot of people by surprise in Texas. So what happened, really, was there were ice and snow storms in Texas and that caused a widespread power outage. They were basically--

Most people know this part of the story. Basically, every potential source of power in Texas shut down at the same time.

GARY: Gas lines were blocked with ice. Wind turbines froze to a stop. So demand couldn't be met and, as a result, the grid operators had to conduct those rolling blackouts in different parts of the state.

NEWS ANCHOR: More than 620,000 waking up without power in Texas this morning.

GARY: Well, since power and water are inextricably linked, without power, the water treatment facilities weren't able to operate their pumps. They couldn't run the treatment process, so that led to boil water advisories. At the same time, pipes actually started to freeze and caused them to burst.

REBECCA: Yeah. I mean, we all remember those images of people having to boil their water just to drink, and collecting snow to flush their toilets, and everything. It's a pretty devastating situation.

The interplay between power and water infrastructure

GARY: The whole energy water nexus is critical to think about here because the typical large consumer, most likely the largest consumer of power in any town or city, is likely going to be the water, wastewater utility.

Surprising as it sounds, this is really true. Water and wastewater facilities often have things like pumps, and motors, and other equipment running 24/7.

GARY: So energy costs are typically about 1/3 of the operating budget.

That's really high. Wastewater utilities, for example, will often spend between 25% and 40% of their operating budgets on electricity. And if you think that's crazy, energy usage accounts for approximately 80% of the cost of processing and distributing drinking water.

One way utilities can reduce the economic and environmental costs of treating and distributing water is to improve the energy efficiency of their equipment. That, of course, means routinely collecting and analyzing data about how much energy each machine is using.

So there you could have power metering, vibration monitoring, and basically what utilities want to strive for is figuring out the optimal pump schedules. So we want to figure out when to pump water, where to pump it, at what times. Can we pump during off peak hours, for example? And overall, are all our pumps and pumping stations operating at its rated efficiencies? So there they can look at pump curves, they can measure specific energy, all these types of things.

We'll explore the so-called water energy nexus more in later episodes because this is a really important topic when it comes to sustainability. For now, we'll just point out that building resilient water infrastructure means building resilient power generation and distribution systems too. These systems are interconnected.

Like we saw in Texas, if the power goes down for an extended period of time, people can lose access to clean drinking water too. But the water utilities' dependence on huge amounts of energy for running their plants isn't the only challenge water utilities face when it comes to ensuring resilient and sustainable water systems.

NICK D’ORAZIO: Gary, remind me what the infrastructure is for how the different water systems monitor what's going on in those pipes.

That's Nick D’Orazio. You might know him as our regular host of Radio PI.

The monitoring problem in U.S. water systems

NICK: Is this the situation like the power companies, before we started seeing a lot of smart metering, people didn't really know they were having a problem with the water system until people started calling it in?

GARY: Yeah, so that's true. It really depends and varies from city to city, state to state. In the US, we have about 50,000 different water systems.

Actually, according to the CDC, there are around 156,000 public water systems in the United States. Of that 156,000, about 52,000 are what are known as community systems, meaning that they supply water to the same population year-round. These are the kind of systems Gary's referring to.

The remaining roughly 104,000 are called noncommunity systems, a category that is then further subdivided into what are known as transient and non-transient systems. Transient noncommunity systems provide water to places where people don't remain for very long periods of time. Think something like a campground. Non-transient systems supply water to the same group of people at least six months of the year, but not year-round. Some examples here might be universities, factories, office buildings, or hospitals.

GARY: So it's very, very fragmented. And a lot of the systems out there don't have smart meters or don't have any kind of water meters at all in some places.

That's not just no smart meters. That's no water meters in place at all.

GARY: However, on the water mains, for example, there definitely will be some form of metering, typically, that you see. And there, you can measure, for example, pressures and flows. But overall, I think we can do a lot better job in terms of putting more sensors out into the distribution system at least. And then as we start to put out more smart meters in people's homes, we could then easily detect more leaks.

Between 15 – 30% of community water systems are leaking water

We can start to detect water quality issues as well. And that's something that's really needed, I think. Because the amount of leaks can be as high as about 30%. And, on average, I think utilities basically report roughly about 15%. But a lot of the times, the leaks go undetected.

REBECCA: So that's-- you're saying 15% to 30% of water that's flowing through the water pipes underground in cities and towns is just leaking out into the soil. Is that right?

GARY: Yeah. Yeah, absolutely. That's crazy if you really think about it. We're taking water, we're treating it, we're pumping it, and it's leaking away easily 15%, and in some places, as high as 30%.

NICK: So Gary, can you describe-- one of the most interesting leaks I ever heard about involved a fellow that found one in his backyard. Can you describe that story?

The stream that turned out to be a leak

GARY: Sure. That was quite interesting. They actually created a nice water feature based around the water, the stream that was coming through their backyard. And it actually turned out to be a leak. So it wasn't a natural stream or spring. It was a leak that was happening for a long period of time.

And when the utility came by, they actually used technology to figure out and pinpoint that there was a leak in the area. And they were coming by and they figured out that, wait a minute, this is not a natural stream and it's really a leak. And when they turned it off, the stream disappeared.

REBECCA: Wow. That's amazing.

NICK: So can you describe what they're doing with the data to kind of narrow down where they're going to be searching for leaks?

REBECCA: Yeah. And actually, just to piggyback on that question, if you could answer that and then just explain a little bit about how-- what are some of the different methods for detecting leaks, the one that Nick just mentioned, apart from sensors, obviously? Because you said that, in a lot of places, they don't even have sensors on a lot of these pipes. So how are we detecting leaks when we do?

How DMAs are detecting leaks today

GARY: So yes. Typically, utilities, what they'll do is, within the water distribution network, they'll put in water meters in various locations. And what they do is they basically break down their entire region into smaller subregions called district metered areas or--

As an aside, District Metered Areas are often just referred to as DMAs.

GARY: And with these DMAs, they're now able to pinpoint leaks in a much smaller area, or at least get an idea that there's a leak happening in a much smaller area within the entire region. That means there are water meters that measure the amount of water coming into that DMA and also exiting that DMA. And from there, they're really doing a simple water balance within that DMA.

The idea here is that utilities break their distribution systems down into these small areas. That way, they can look at the water coming into a specific area and the water coming out and compare it to both nearby areas and to historic averages.

In other words, what that water flow in and out has looked like in the past. And if something looks a little fishy, like there's more water flowing into a certain area than being metered flowing out, then the water utility has a pretty good idea that there's probably a leak somewhere in there.

GARY: That's right. So in terms of how they detect the leak, once they're at the DMA, then they usually will have to send teams out to pinpoint and identify the exact location of that leak. So in situations where there's going to be more metering in place, for example, where you're going to have some household meters, we're able to much more quickly determine the rough location within maybe several hundred feet of the leak. And if they don't have it-- or actually, it really doesn't matter if they do have it or not. There's still going to have to send people with acoustic leak detection. So--

Acoustic leak detection equipment is basically a device with a microphone that you place on the ground that allows you to listen to the sound water is making as it flows through a pipe. Leaks can create sound waves due to pressure changes in the water as it approaches and flows past a leak. And based on what water sounds like, you can pinpoint the location of a leak.

GARY: Yeah. And then from there, they're going to have to dig up the ground and then find that leak. They'll see the leak physically, and then they'll have to repair it from there. So we talk about the runtime of that leak, meaning how long is that leak running before we can find it and then fix that leak.

How White House Utility District used sensors to improve leak detection

That's why it's really critical to have a lot more sensors in place so we can identify which areas are leaking and then send people out there to find that leak and then fix it right away. A great example is White House Utility District in Tennessee.

White House Utility District is a water utility-based northeast of Nashville. They're the largest water utility by geography in Tennessee and oversee about 1,000 miles of pipe. Several years ago, over 30% of the water White House Utility District was pumping and processing was being lost as what's called nonrevenue water.

Nonrevenue water is the difference between the amount of water that's produced by a water utility for consumption or use and the amount of water that is actually billed to the customer. According to one report published by the World Bank Group, the total cost to water utilities caused by nonrevenue water on a worldwide basis is estimated to be a staggering $141 billion per year.

Part of the problem for folks at White House Utility District was that they didn't have a proactive approach to leak detection, but they knew that needed to change. So they launched a program to install more water meters and sensors and set up 33 district metered areas from which they could capture flow data. And--

Why White House Utility District did not know about its $300,000-per-year-leak

GARY: And within 3 and 1/2 days, they found a $300,000 a year leak that they never knew about.

REBECCA: So how do they just not know about it? I mean, that just seems mind blowing.

GARY: Yeah. So not all leaks surface. Meaning leaks occur in the ground and they leak further deeper into the ground. So we never see, for example, wet patches of soil or things like that.

NICK: Right. I've heard similar things. If it's next to the seashore, people just think, OK, because it's not pulling up anywhere. It just runs off into the ocean, so there's-- nobody complains about it.

GARY: That's right. So you don't see it. And if you don't have, for example, meters in place, or you don't have the ability to detect it or see it, then you just don't know.

In the case of that White House Utility District leak, it was leaking at a rate of 80 gallons per minute and could easily have been going on for a year or more because it was in a remote forested area leaking downward, deeper into the ground, so there were no visible signs of it on the surface.

Since then, White House Utility District has gone on to use some of the same sensors and technologies they adopted for this initial DMA leak detection program to do a lot of really cool things like enabling remote work during the pandemic, improving the general efficiency of their operations, and learning things about their water system that they never knew before.

Now, to be clear, this method of using district metered areas to proactively search out leaks and narrow down where they might be occurring is not something that the majority of water utilities are currently doing, especially not smaller water utilities like White House Utility District.

Why water utilities must be proactive about leak detection

According to Gary, most utilities are still generally reactive when it comes to leaks. That is, they only go out with their leak detection equipment when someone calls in to say they had a sudden drop in water pressure or that their backyard is suddenly turning into a swimming pool.

It doesn't take a lot to realize that faster leak detection means less water wasted. But what many people might not realize is that for every drop of drinking water lost in a leak, we're not just wasting water. We're also wasting all the energy and cost that went into purifying and transporting that drop through miles of pipe, only to have it flow away underground undetected.

So we've talked about energy usage, and leak detection, and minimizing the amount of resources we're wasting, but what about those other two big words when it comes to water resilience? I'm talking, of course, about climate change. Changing weather patterns and changing environmental conditions. What sorts of challenges are these trends posing to the water industry?

Climate change and the rise of harmful algae blooms in drinking water

GARY: There are so many other scenarios as well. In terms of resilience, we're starting to see, for example, harmful algae blooms. With harmful algae blooms, basically, what's happening is the temperature of water is rising in some of our lakes. And because of that increase in temperature, algae starts to form. And sometimes these algal blooms can be harmful to us. They can create cyanotoxins.

Just so you know, cyanotoxins are chemicals produced by little creatures called cyanobacteria. If humans are exposed to these cyanotoxins, they can have a large number of nasty side effects ranging from headaches and skin rashes to fever and vomiting. And, in some cases, even death.

Gary told me about this one water utility in the City of Salem, Oregon. In the summer of 2018, they had a cyanotoxin scare when harmful algae blooms suddenly invaded a Detroit lake. The local utility wanted to find a way to predict when these sorts of blooms were going to happen so that they could take proactive measures to protect the residents of Salem. So what did they do? They put sensors on a pontoon, of course.

GARY: Multiple sensors on the pontoon that are on the lake measuring things from lake turbidity, temperature of the water.

They also had depth measurements, weather data, satellite imagery.

GARY: Different-- I think they had eight different-- yeah, they have a different data streams coming in. And they're taking all that data and they're sending it off to a third party up in the cloud.

From there, sharing it with ecologists and mathematicians.

GARY: And using artificial neural networks and AI to predict if there's going to be harmful algae blooms, what are the odds of having a harmful algae bloom and the chances of having cyanotoxins. That way they can inform the public.

Thanks to all that data, the City of Salem can now see ahead of time what the algae and cyanotoxin levels are likely to be in their water source and take appropriate action, like changing pumps, adding filters, or redirecting water to make sure they've always got clean, safe drinking water for the City of Salem.

GARY: They also post all that information online. So it's very transparent and the public is able to see exactly what's happening at the lake. And that's really important. That that's all about-- that's part of being resilient, and being transparent, and sustainable.

REBECCA: What are some of the other concerns that people in the water industry have that are particularly related to changing environmental conditions?

GARY: So there are a lot of environmental challenges that the industry faces. For example, droughts, storms, including hurricanes that we see every season, wildfires, harmful algae blooms. What we're really seeing is more extreme weather events, including higher or lower temperatures and rainfall. That makes it very difficult for utilities to be resilient and sustainable.

The Eastern seaboard sees a lot of hurricanes come in during hurricane season. And the whole challenge there is to be able to predict and try to forecast where we need to send crews. Where does the utility need to send their crews in order to basically maintain their level of service? So it becomes a challenge because they need to take in information such as wet well levels, the amount of water being consumed, the weather patterns, radar forecasts that come in, data from NOAA.

That's a lot of data coming from a lot of different places.

Using predictive analytics to mitigate the impact of extreme weather

GARY: And combined, they then have to take a look at and predict where that storm is going to travel, what's the path of the storm, and how that's going to impact their facilities and the services that they provide. So they're starting to get into more predictive analytics. Of course, they also take in rainfall data, which is really important, because we see, a lot of times, rainfall can be very localized. And from there, they're able to then figure out and determine where are the high priority areas. JEA is a great example.

JEA is a not-for-profit, community-owned utility located in Jacksonville, Florida. They oversee about 4,400 miles of water lines.

Mapping real-time water system data is critical for disaster preparedness

GARY: They have the ability on a map to see in real time what are the wet well levels, how much fuel is there. Because just in case we do lose power, we can send crews with backup generators to run those pumping stations, for example, to keep wet well levels, so we prevent sewer spills.

REBECCA: Because part of the problem in a storm is that you have a lot of extra runoff that's going into the sewer system, essentially, and could cause it to overflow?

GARY: That's right. That's definitely a part of it. So during a hurricane, you can have a lot of rainfall. And that runoff will then go into the system. A lot of times we call those combined sewers. And you can create situations of a combined sewer overflow where that runoff goes into the same wastewater collection system and that goes to the plant.

The plant's not able to treat all that runoff. And therefore, they have to discharge without treating that wastewater. So in situations where a storm comes in, there's going to be a lot of rainfall runoff, they might lose power, which means a lot of that wastewater is going to start to get backed up and could create an overflow.

REBECCA: So this is all about preparation, really, for a disaster scenario? Like, looking at all those data sources you mentioned, weather data, wet well level data, and basically making an educated guess about where help is going to be needed most.

GARY: It's really combining all those different pieces of data to making an informed decision. And they have to do it in real time, and a lot of the times they have to make a best guess effort in advance of that situation. And that also holds true for any type of natural disaster that might occur, including wildfires or droughts.

REBECCA: And so, I mean, you're kind of saying that these sort of extreme events, whether they be hurricanes, or large storms, or droughts, or fires, they're going to become more frequent. And so it's just becoming even more critical that local water utilities have the capability, have the technology and the capability to deal with it in this way. Because I assume not all water utilities probably are set up for this right now.

GARY: That's right. So-- and this holds true, for example, even in the pandemic. So we've seen a rise in remote operations, remote monitoring, all because of the pandemic has changed the way they are able to and have to work.

Mapping real-time water system data is critical for disaster preparedness

NICK: And Gary, you're saying-- you listed a couple of things. You said you have to have the data, it has to be in real-time. And I keep imagining these are on maps. These are always on maps. How important is it to actually be able to map these things to a map, like an RTIS display or something, something in real time?

GARY: Well, having the geospatial component is critical because these utilities manage very large areas. There's a lot of people that do like to use geospatial data because it's easy for them to understand on a map. They know what they're looking at and where it is. Because a lot of times, this data comes in forms where they might have some archaic tag name that doesn't really make sense to them, so they really don't know what data they're looking at.

When you do put it onto a map, and you provide context around it, and you know that it's a pumping station or you know that it's a water quality-- for example, you know that it's chlorine residual in a certain location, that makes sense to people. So that makes it very easy to understand what they're looking at and provides another level of context.

So just to recap here, we've got water systems that are highly dependent on large amounts of electricity. We've got changing water temperatures in lakes giving rise to toxic algae blooms. We've got an increasing number of extreme weather events. We've got old, leaky pipes, some of which are telemetered and some of which are not.

And what's the one thing that can help with all of these issues? Real-time sensor-based data. And it doesn't hurt to throw in a little fancy AI and analytics too. So where is all this technology and data leading us?

Predictive analytics, neural networks, AI, digital twins on the rise in water utility management

GARY: Smart water is all about an ecosystem of having sensors, controls, data analytics, and visualization that are all needed to make informed decisions, mostly in real-time. The challenge is that there are many disparate systems, many data silos as well.

You can imagine that there's different pieces of hardware and software from many different manufacturers within a water utility. They might all speak different languages or protocols and they don't communicate with each other. So what's really important is to have a data abstraction layer, a data core, a data hub, and from there, that's-- we talked about the one source of the truth in terms of data management.

And a lot of this operational data is needed to make these informed decisions. So not only do you have real time data, but you have a long-term historical data set as well. What are some of the uses of that data? We're starting to see a lot more use of digital twin, predictive analytics, using neural networks and artificial intelligence on these big data sets.

For example, water utilities can use data coming from their pumps to predict when the sewer system might overflow during a storm. They can look at pump data to do predictive maintenance before pumps fail. And with enough smart water meters in place, they can use AI to pinpoint leaks and reduce the amount of time it takes to detect and locate them. Just imagine what a difference it would make reducing the amount of water that leaks out of our pipes every day.

I'm thinking back to the beginning of the show when we talked about how little of the planet's water is actually in a form that we can use. It's such a small percentage. And to think that we're losing, in some cases, up to 30% of it due to leaks, it just-- it boggles the mind.

According to the EPA, the US loses around a trillion gallons of water a year due to household leaks. 1 trillion gallons. That doesn't even take into account all the water lost due to leaks in water mains or pipes that have to move water across mountains and deserts to bring it to communities.

And this is only in the United States. The World Bank suggests that the annual global value of nonrevenue water, that is, water produced and lost by utilities, is close to around $14 billion. And six years ago, the World Economic Forum ranked the water crisis as the number one global risk based on impact to society as a measure of devastation.

Now, I know that's a lot to think about. The good news is that we do have tools to do a lot more than we're doing right now to help mitigate this crisis. As always, it's just a matter of getting the right data and information to the right people at the right time. And that actually might be easier than you think.

That's our show for today. Special thanks to Gary Wong and to our co-host, soon to become host emeritus, Nick D’Orazio, who's retiring this month. We will miss you and your decades of wisdom and experience.

graphics design elements

How can we help you?

We’ve been delivering reliable data for over 40 years. In fact, 24 of our first 25 customers still use our platform. Let us know how we can help you.

©2021 OSIsoft, LLC