SWAN’s way – in search of lost water

 

The computer system then makes decisions about supply and demand and automatically sends out commands to valves, switches and pumps in order to fine-tune the network. The benefits include fewer leaks and spills, reduced energy consumption, longer infrastructure life, and lower operating costs.

 

One might think that smart networks would be eagerly adopted by the water industry, which on a global basis is worth in excess of $500 billion annually. It has also been estimated that leakage of potable water in some systems runs at 60%. Yet when it comes to adopting smart technologies, water information technology (water IT) is lagging behind.

 

“When you look at the electricity and gas utility sectors, when it comes to automation, they are 10-20 years ahead of the water sector,” says Andrew Burrows, chief technical officer for i2O Water, a UK-based supplier of smart water network (SWAN) systems. “The addressable water market is huge.”

 

Despite this, the water sector has invested billions of dollars in recent decades on software and hardware kit in order to increase the efficiency and control of water, from cleansing and distributing to collection and waste treatment. Guy Horowitz is VP of marketing at TaKaDu, an Israel-based company that supplies smart water network software. “SCADA (Supervisory Control and Data Acquisition) – in which you gather, store, manage the data and make some sense of it – has been around for a long time,” says the chairman of the recently-created SWAN forum (a meeting place for smart water network ideas and processes). “There is a lot of data being generated by water networks right now, and that data will increase by three orders of magnitude in the next decade.”

 

But most of that data, even in the most advanced water systems, is currently being sorted and analysed by humans, not computers. “There is a limit to what even the smartest human operators can do,” says Horowitz. “We can take the data and run it through very large computers using the most advanced algorithms to replicate what human operators do.”

 

 

There are several drivers to the water IT market. “Over the last several decades, we’ve seen the growth of water shortages and global warming and rainfall patterns changing,” says Burrows. “There have also been significant increases in demand due to population growth and usage patterns. Water utilities are looking at building new water treatment and desalination plants, at tremendous cost. Yet they are wasting 40-60% of the water they’re producing through leakage. Automation technology can address water shortage issues much faster, and at a fraction of the cost.”

 

The impetus for smart networks began in the UK. “In 1989, many water utilities in the UK were privatised in order to improve performance and efficiency,” says Burrows. “A regulator was set up to define performance standards and to monitor these companies against key performance indicators. In order to monitor performance and set targets, you needed to put technologies into the system to calculate and measure the percentage level of leakage, levels of service, water quality and the energy used to move water around.”

 

Technology companies devised systems to reduce leakage in order to help utilities achieve these targets. “Even though they were basic in scope, the systems required highly trained professionals, long implementation times, manual adjustments and a sophisticated degree of knowledge and skill to operate,” says Burrows.

 

About a decade ago, computers, sensors and communication devices had advanced sufficiently to allow far greater capability in smart networks, but cost barriers remained. “Ten years ago, it was not economically feasible to integrate systems,” says Manuel Parra, the director of Telvent Water Product Center. “The technology was there, but you had to do too much customising for the water sector.”

 

Recognising the opportunity, firms that were already involved in the water sector began to make smart water network components more affordable. “We at Telvent has been taking the steps towards systems integration and data analysis for several years now,” says Parra. “We have been building our in-house expertise on system integration and data management in order to add value to all information.”

 

 

One of the main benefits from a smart water network is volumetric leakage reduction through pressure control. The pressure necessary to supply potable water is governed by the demand on the system; the greater the demand, the greater the pressure required. Increase the pressure too much, however, and burst pipes and leakage rates rise to unsustainable levels. Reduce pressure, and bursts and volumetric leakage fall, but customer complaints rise. “The challenge is to optimise pressure to meet regulated supply levels, yet reduce bursts and leakage,” says Burrows. His company set out to devise a system that would automatically control water pressures. “Our target was to develop a technology that was easy to install, would automatically learn the water demand and optimise operating conditions, and would continuously react to changes in order to maintain optimised conditions.”

 

Modern potable water distribution systems are typically composed of a trunk main leaving the water treatment facility that then connects to distribution mains in district metering areas (DMAs) or pressure managed areas (PMAs). A DMA or PMA typically serves about 2,000 customers with smaller, lower pressure lines. The pressure in the trunk mains is maintained at an 8-10 bar level through gravity feeds and electronic pumps. A pressure-reducing valve (PRV) reduces pressure in the distribution mains to 5-6 bar. Frictional losses and elevation differences further reduce pressure; regulations generally call for a minimum head pressure of about 1.5-2.0 bar for customers at the most critical point, such as highest elevation or farthest distance.

 

Traditionally, water engineers use experience and judgment to optimise the pressures in order to reduce bursts and leakage, but still satisfy customer flow demand. Demand, however, can fluctuate hourly, daily and seasonally. “Summer demand in a residential community can be twice as high that in winter because people water their lawns,” observes Burrows.

 

Because PRVs are buried or placed in difficult-to-access locations, pressures are set to meet the most demanding circumstances. This means that the system operates under higher than needed pressures for much of the year, leading to higher than necessary rates of bursts and leakage. There are three main categories to bursts and leakages. The first is the visible burst, in which a main ruptures and a large amount of water runs down a roadway. These bursts are relatively rare, and can be quickly repaired. The second category involves significant but invisible leaks; they require the use of location devices such as noise correlators to find. They are more common than visible bursts, but can still be economically repaired.

 

The third category is background leaks. They occur at the thousands of connections throughout the distribution network. Although individual leaks are small, the cumulative effect is large, accounting for 50-99% of all leaks in some systems. Globally, they account for 30-40% of potable water being wasted. “Because they are so small and numerous, it is uneconomical to fix them,” says Burrows. “The only way to reduce them is by managing pressures.”

 

Starting in 2005, i2O invested over £10 million working with Severn Trent Water to create a commercial smart water system that could control pressure in distribution systems. Launched in 2009, the system is a combination of hardware and software that is easy to install and runs largely by itself. Remote sensors gather pressure and flow information and send it back to a central server. There, complex algorithms sort and analyse the data in order to learn demand patterns and correlate pressure to flow. The system ‘learns’ over days and weeks what to expect, and how to react to changes. Commands can then be sent automatically to controllers installed on PRVs and pumps to adjust instantaneously to changes in demand. “It brings pressures down, but still guarantees that they will remain at or above the minimum required levels 99.5% of the time,” says Burrows.

 

The company has installed 200 DMA systems in over a dozen countries around the world. “We have seen that a 20% reduction in pressure can reduce leakage loss by 20%,” says Burrows. “Burst reduction is harder to measure, but we have seen some data in Malaysia that would indicate that a 20% reduction in pressure can lead to a 40% reduction in bursts.”

 

The City of Calgary water services department delivers 450,000m3/d of potable water through 4,500km of pipe to 1.1 million residents and several thousand businesses. The system is divided into 40 PMAs. Their water loss management programme has reduced leakage to approximately 10%, but water is scarce in the arid region, and in order to further reduce losses, they are investigating SWAN systems. Late in 2010, they began an active pressure management pilot test on two residential PMAs. They installed programmable logic controllers (PLCs), power, and a communications link to the PRVs in order to monitor demand and control pressure.

 

The project is still ongoing, but initial results are positive. “We anticipate there will be a significant benefit,” says Scott Jamieson, team leader for operations engineering at Calgary Water Services. A 20% reduction in leakage amounts to 10,000m3/d, which equates to a daily saving of Can$10,000.

 

There are over 200 PRVs in the Calgary distribution system that would have to be retrofitted. It would cost approximately Can$25,000 per chamber, for a total of Can$5 million, to cover the entire system. Software to oversee the system would be an additional expense. “It would take time and budget to implement it throughout the city,” says Jamieson. “But we’re very optimistic that the value is there, and that we can quantify and justify the expense.”

 

In addition to leakage and burst reductions, smart water systems can reduce energy consumption in distribution systems. According to the California Energy Commission, water-related energy use accounts for 19% of all electricity use in the state. “As part of pressure control, we could put smart controllers in pumps in order to reduce energy consumption. Even a small decrease can have a huge impact,” says Burrows.

 

Energy can also be saved in water and wastewater treatment plants. Schneider Electric is a major supplier of telemetry devices and control software to resources, manufacturing and utilities. “We are at the core of automation,” says Pascal Bonnefoi, a member of Schneider Electric’s water solutions team. Schneider estimates that water and wastewater treatment in Germany accounts for almost 1% of total energy consumption, and contributes 3 million tonnes per year to its carbon footprint. Better management of network information could, Bonnefoi argues, allow for more effective anticipation of treatment requirements in future, saving up to 30% in energy consumption at the plant.

 

 

Currently, the water IT segment is small in comparison to the overall water market.

 

“Right now, the SCADA market in the water sector is in the order of $2 billion, but that’s mainly for investment in sensors for data collection, some transmission and presentation,” says Horowitz. “The smart water network component is nascent, with projects in the tens of millions.”

 

Tellingly, it is IT hardware and software companies, such as Telvent, Schneider, i2O and TaKaDu making most of the headway, not companies that have water as one of their core competencies. “GE, Siemens and the other giants of the industry are interested in what we are doing, but they are not doing what we are doing,” says Burrows.

 

Several problems are impeding the implementation of water IT. A major obstacle is the publicly regulated nature of the water sector. “In the oil industry, you show them the value proposition, and they act,” says Bonnefoi. “Municipal utilities must put out a bid for an audit first. It’s a different decision process, and it slows down adoption.”

 

“Right now, water utilities are taking a step-by-step, staged approach where they integrate a few systems,” says Parra. “Integration is not foreign to them at all, but it is in its infancy. The reason is that water utilities are going through the process of internalising these concepts, and that implies the need to remove some of the historical data silos and build datasharing organisations.”

 

Finally, many utilities simply don’t know about the advantages of SWAN. “There is a lot of difficulty to get agreement to publish success stories regarding savings of costs and energy,” says Bonnefoi.

 

 

To that end, proponents of SWAN have begun to organise themselves in order to get the word out. “We started pitching ideas to other people: the pump optimisation guys, the leak detection folks, the pressure guys and the SCADA people,” says Horowitz. The result is the SWAN forum, which now has 20 members internationally. “The SWAN forum tries to promote change. Our goal is two-fold: firstly to use the data we have now and optimise it, and secondly to invest further in better equipment.”

 

Their first annual conference was held in May 2011, in Paris. “It was a very good event, with lots of good presentations,” says Horowitz. “The vendors are pushing the concept, and the utilities are pulling.”

 

Telvent presented a case study from South America. “We did a large integration project in Brazil,” says Parra. “A publicly owned water utility serving three million customers wanted to reduce maintenance and service outage downtimes, as well as water losses. Telvent helped it to implement its SCADA, hydraulic model and geographical information system, together with a demand management application, and integrate them together with its existing maintenance system. They realised their goals and also boosted the image of the company.”

 

“The SWAN forum is important because to meet the demands of the future, it is going to take a concerted effort,” Burrows maintains. “The ultimate goal is to have an integrated system so that water companies can monitor and control all parts of the water and wastewater network, right from source to treated effluent discharge. It is impossible for any one company, even one as large as GE, to achieve that goal by themselves. You need to collaborate.”

 

Consolidation also forms part of the picture – Schneider Electric made a $2 billion bid to acquire Telvent at the end of May, which has every chance of succeeding, especially given that the move has the support of Telvent’s 40% shareholder Abengoa.

 

The future looks rosy for smart water networks. There are over 50,000 water and wastewater utilities in the US alone, and some 150,000 worldwide, and all can benefit in some fashion. “Analysts have looked at what they call the ‘water IT’ potential, both hardware and software for municipal and industrial use, and they see huge potential, something like $10-20 billion per year in terms of ultimate size,” says Horowitz.

 

And as great as water IT’s potential is, utilities stand to benefit even more. Worldwide annual operating expenditure for water and wastewater utilities stood at about $222 billion in 2010. “If we can take 30% off that, then it will be a significant achievement,” says Bonnefoi.

 

 

A smart water network (SWAN), also known as water IT, uses data-driven software and components to automatically optimise the performance of reservoirs, pumps, water and wastewater treatment plants and distribution pipes.

 

* SWAN systems could reduce water leakage by 20% and pipe bursts by 40%.

 

* Worldwide annual capital expenditure in hardware (SCADA, telemetry and sensors) for the water sector stands at approximately $2 billion. The proportion accounted for by SWAN investment is currently around 10% of that.

 

* The annual hardware and software potential for SWAN networks has been estimated in the $10-20 billion range within the decade.

 

* Worldwide annual operating expenditure for water and wastewater utilities stood at about $222 billion in 2010. SWAN technologies could shave 30% off this figure.

 

 

The Smart Water Networks Forum has six founder members. What do they bring to the table?

* TaKaDu’s software monitors water distribution networks, giving utilities real-time information on leaks, pipe bursts, DMA breaches and other network inefficiencies. Guy Horowitz is VP of marketing for TaKaDu, and chairman of the Smart Water Networks Forum. Tel: +972 52 542 4848 E: guy.horowitz@takadu.com

 

* i2O Water’s technologies optimise the pressure in water distribution networks to reduce energy consumption, leakage and pipe bursts. i2O completed a £10m fundraising round in February 2011. Andrew Burrows is CTO of i2O Water. Tel: +44 238 011 1423 E: andrew.burrows@i2owater.com

 

* Serck Controls (part of Schneider Electric) provides monitoring and control systems for water, desalination and wastewater facilities. The company integrates hardware and SCADA software to enable real-time asset monitoring and reduced operating costs. Pascal Bonnefoi is the water segment director for Schneider Electric. Tel: +334 7657 9202 E: pascal.bonnefoi@schneider-electric.com

 

* Telvent’s proprietary Water Management Suite offers smart network control, allowing energy optimisation through power consumption monitoring, as well as demand forecasting, leak detection and water quality management. Manuel Parra is the director of the Telvent Water Product Center. Tel: +34 678 644 647 E: manuel.parra@telvent.com

 

* Derceto’s Aquadapt software offers real-time energy optimisation by operating pumps so that they use less energy to move more water. Aquadapt schedules pumping to occur at the cheapest times and finds the lowest-cost delivery path. Simon Bunn is CTO of Derceto. Tel: +64 9 373 7100 E: sbunn@derceto.com

 

* Echologics’ diagnostic technologies enable municipalities to locate leaks, and to assess the structural thickness of pipes using non-invasive acoustic wave analysis. The company was acquired by Mueller Water in December 2010. Marc Bracken is VP and general manager of Echologics. Tel: +1 416 249 6124 E: marc@echologics.com