Pure Genius

Q&A: Marcus Quigley on how the Internet of Things is automating city systems

Q&A: Marcus Quigley on how the Internet of Things is automating city systems

Posting in Sustainability | From Issue 11 February 17 & 24, 2014

The Internet of Things promises to bring us into the 21st century in a number of ways, beginning with keeping sewage out of our rivers and lakes.

The Internet of Things promises to make our lives fully automated and a whole lot easier, by doing everything from serving up our morning coffee when we wake to informing us in real time of a late train. Ultimately the Internet of Things is about devices receiving and distributing pertinent information over a wireless network. And it has greatly evolved over the last five years as cloud-based technology has become cost-effective. Much of the hyped promise has yet to hit our everyday lives, except in the case of a series of projects that have turned urban water systems into fully automated, self-monitoring objects. 

Most of the large cities in the U.S. -- like New York, Chicago, Boston, Detroit -- have sewer systems that get overwhelmed by extreme rainfall. During such storms, rain water overruns the sewers and waste water treatment plants, so much so that raw sewage from our toilets winds up bypassing the water treatment plants and is dumped directly into a nearby waterway. This leads to numerous violations of the Clean Water Act, as well as beach and lake closures.

The infrastructure engineering firm Geosyntec has transformed such century-old designed infrastructure into smart systems that know when a storm is about to hit and adjust themselves to prevent or alleviate the possible flooding of sewers.

MarcusQuigley_full.JPG
 Marcus Quigley, Geosyntec's principle water engineer, speaks about how the issue is primarily about controlling the water supply and demand, and how the Internet of Things can fix this problem that has cost the country more than $13 billion dollars in wastewater management, according to the Environmental Protection Agency. 

How is the Internet of Things helping cities to automatically control their own water systems?

Well, what is most meaningful to people is the monitoring and control of systems that have to do with helping to decrease impacts of combined sewer overflows on waterways. Combined sewers are those where storm water and wastewater combine, and waterways are any rivers, beaches and lakes that the public might use.

And most cities use combined water sewers?

In 760 cities around the United States we have a legacy combined sewer overflow issue that is a result of a historical combination of both storm water and wastewater flows in our wastewater collection systems.

Presumably this causes a pretty big problem.

It’s a problem that’s a result of timing of the flow, not necessarily the overall long-term capacity of wastewater systems. How the city drains to those combined sewer systems is something that can be greatly improved by taking advantage of these highly flexible “Internet of Things” platforms.

So how does it work?

In combined wastewater systems there are lots of places to store water to mitigate the peak [storm] flows. But they’ve historically been designed as passive tank systems.

It isn’t that the wastewater treatment plant can’t ultimately handle all the flow, it just can’t handle the peak discharges as they occur to the combined sewer.

With the Internet of Things platform, these storage tanks use cloud technology to connect with internet-based weather forecast in combination with the field data to then make decisions [to deal with the water] in advance of actual precipitation occurring.

Can you walk us through how a storage tank can self-monitor?

Let’s consider a rainwater harvesting system. So you have a tank and you collect water that’s runoff from roofs so that it can be used for something else like irrigation, toilet flushing, vehicle washing or any non-potable use. But usually the demand for rainwater is independent of the arrival of rainfall in the tank.

So you have water falling on the roof and it gets stored in the tank. But if that tank is full and the water is not drained for irrigation or toilet flushing, then during the next rainfall the system cannot function to reduce flow to the combined sewer or waterway. If the tank is able to watch both the amount of water it is storing and the weather forecast, it can choose to release stored water before the next rainfall to prevent any overflow.

But if the forecast is wrong, which we think it often is, doesn’t that create disaster in this sort of automated system?

Right. We prevent that by using the National Oceanic and Atmospheric Association (NOAA) forecast: The quantitative precipitation forecast and the probability of precipitation forecast. We use a 48-hour forecast that comes to us, the data comes in six-hour chunks that are updated every hour, and then we get probabilities for each 12-hour portion of the 48-hour forecast, and that’s also updated every hour.

The tank system watches minute-by-minute and if there is some change that NOAA makes to the forecast, it adjusts accordingly.

The algorithm we create can be complex, adjusting depending on how long a specific tank takes to drain, how much of the forecast we want to use, and how much we want to rely on that forecast.

And it’s been successful?

Very successful. In terms of metrics, we’ve been able to reduce the amount of flow by over 70 percent.

Can you give us a picture of what happens when the systems are flooded. How bad can it get?

It can result in beach closures, and water-quality violations in swimmable and fishable waters.

The Clean Water Act in particular requires that there is control of combined sewer overflows. The bill for compliance with the Clean Water Act in the United States, as it impacts just combined sewers for the next two decades, exceeds $100 billion. This is money out of the public’s pocket that shows up on utility fees, water fees and water bills.

The information we can gather from the cloud and [have it matched] with physical control solutions allows us to dramatically improve performance. And we are doing this with existing infrastructure. There is no need to build anything new.

So you really don’t have to do much to the barrels, actually just add a sensor and wiring that creates a valve or a pump?

Yes. The marginal cost for the integration of the additional components to make systems smart is negligible relative to the capital expenditure for the infrastructure itself. On any storm water system it’s the construction of tanks or green roofs that are the major capital expenditure.

Your firm started working on these new types of systems two years ago?

Right. The idea to monitor systems remotely has been around for decades but the necessary cloud-based resources would have made our current solutions completely impractical from a cost perspective. But the technology has progressed so rapidly and now we can use the integration of forecast to get control of civil infrastructure.

How far reaching are you with the technology, how many systems are we talking about?

We have over 50 different systems deployed nationwide.

Where are some of the these systems?

In the City of Omaha we have retrofitted two porous pavement parking lots. These are parking lots that function much in the way that I’ve described tanks.

They have storage beneath the porous pavement surface. We’re able to control the discharge of water from the storage base into the combined sewers.

The Austin Public Library had an existing rainwater harvesting system that we retrofitted to include the forecast integration.

There is also a pond in Pflugerville, Texas that provides peak water flow control. We have turned that outlet into a controlled outlet that is linked up with these highly distributed real-time solutions for monitoring and control.

So we can decide dynamically how long to keep water in the pond to improve water quality without compromising the flood storage capacity. This holds great promise because there are many similar ponds all over the United States and around the world, and they've been built for flood control but they serve little purpose for improving the runoff quality.

Could a cloud-connected system have prevented something like the disaster hurricane Katrina caused in New Orleans?

Absolutely. I think within two decades these types of systems will be ubiquitous; we’ll absolutely use them in situations like levy and dike protection.

Warning systems about potential failure of critical systems is one of the really interesting uses, particularly with [complex] levies, where we might have literally tens of thousands of miles of systems that we’re trying to watch at once during major events.

Apart from water and runoff management, what other systems are next?

It's going to change things like parking and traffic, and energy management.

We can’t always rebuild our way out of the problems we got ourselves into at city-wide scales. The Internet of Things for infrastructure really holds that promise. It's actually achievable right now. It’s not really something we have to wait for. 

Top photo: New York City's Hudson River (HBarrison/Flickr); Bottom: courtesy of Marcus Quigley

Share this

Christie Nicholson

Contributing Writer

Christie Nicholson produces and hosts Scientific American's podcasts 60-Second Mind and 60-Second Science and is an on-air contributor for Slate, Babelgum, Scientific American, Discovery Channel and Science Channel. She has spoken at MIT/Stanford VLAB, SXSW Interactive, the National Science Foundation, the National Research Council, the Space Studies Board and Brookhaven National Laboratory. She holds degrees from the Columbia University Graduate School of Journalism and Dalhousie University in Canada. She is based in New York. Follow her on Twitter. Disclosure