Acquiring Good Data to Improve Water Utility Services

by Laura Wainwright

Water is at once one of the most plentiful resources on earth; yet at the same time it is one of the most precious. Because only 1% of the water that covers 70% of the earth's surface is fresh, it makes sense that water utilities would take every possible measure to conserve this resource. In fact, many water utilities have implemented strategies that employ the Environmental Protection Agency's recommended methods for managing demand for water: metering, water accounting and loss control, pricing, and education.

In order to be successful, these strategies demand data, which is provided by a wide range of technologies. SCADA sensors, for example, collect data about water flow, pressures, and chlorination levels in the water-distribution system. There are numerous SCADA sensors employed by water departments to monitor water treatment facilities as well as water towers, mains, and hydrants.

Utilities also collect data from meters. Most water meters have historically been read no more than 12 times a year. However, advanced metering infrastructure (AMI), in the form of fixed networks, allows water utilities to collect data from meters daily, or even hourly. Water utilities can use this data to improve customer services.

Providing More Information for Consumers

The Boston Water and Sewer Commission's (BWSC) SmartRead program is an example of how data can be used to enhance a water department's services. The BWSC established SmartRead to improve efficiency and to supply accurate, up-to-date information to citizens about water usage.

The program employs Aclara's STAR® Network AMI system to automatically collect four meter readings daily from approximately 90,000 meters. These readings are posted online and can be viewed through a Web browser. By making more data available to customers, the BWSC is encouraging them to better control their water usage. Having access to more data also allows consumers to identify discrepancies in water usage that might indicate leaks. In fact, Boston schools saved an estimated $36,000 a year by repairing leaks found through the program.

The program also has reduced customer service calls to the BWSC by 60 percent. Now, customers who have a question about a high water bill can view online how they have been using water. Usually, this type of review will help consumers identify days on which they used an unusual amount of water — to water a garden or fill a swimming pool, for example.

Boston has also used the data to reduce the number of meter reads scheduled with customers, as well as the number of estimated meter reads. The data available has reduced the number of meter-reading related work orders by over 70% and has increased the accuracy of meter reads, resulting in recovery of about $10 million per year in revenues lost due to erroneous meter reads.

There are other ways to use data to improve water services. Some utilities use software to help customers pinpoint specific ways to control their water usage. An application from Aclara Software, for example, prompts consumers to answer questions about the types of appliances they have and how they plan to use water at their homes. The system then provides a report on costs associated with potential usage, as well as money-saving recommendations.

Data and Leak Detection

Water departments often meter water as it comes into the distribution system, and then again as it flows into homes. The difference between what comes into the system and what is used by consumers represents a large component of the system's non-revenue water loss. Well-run water systems might lose 10 to 15% a year through distribution leaks, but publications from the American Water Works Association have cited examples of losses of up to 45% in the water system.

Stemming non-revenue water loss is one of the biggest challenges facing a water utility. Leaks in the water distribution system can deteriorate for years before becoming obvious problems through disruptive water main breaks. Although a main break can spill water at a rate of 400 gallons per minute, these catastrophic leaks account for only about 1% of the water lost by distribution systems annually. More destructive are chronic leaks that may run undetected for years, losing as much water as a main break every 90 days. Finding and fixing these leaks requires collecting and analyzing data about how the system is operating.

Acoustic devices attached to water mains can detect changes in flow through the pipes. The data from these detectors can be sent back to the utility via an RF-based fixed network, where it is analyzed to determine the location of leaks.

Utilities use a variety of techniques to find leaks. These techniques require the utility to place acoustic sensors, or logging devices, in the water distribution system to identify changes in the flow of water. Each logger is programmed to collect data at specific times, usually in the early morning hours when noise in the water system caused by normal activity is lowest.

Loggers can be placed on service lines going into homes or on the water main. A disadvantage of placing the loggers on service lines, however, is that the acoustic data may be corrupted by sounds from the home. For instance, if the logger turns on at the same time that a washing machine is running in a nearby home, the data collected may indicate a false positive for a leak. Therefore, systems that collect readings from the main are more likely to provide accurate data.

A traditional method of analyzing the data collected by the loggers requires someone to physically listen to sounds picked up by acoustic sensors through headphones. Specific types of sounds may indicate a leak. The smaller the leak, the higher the frequency of the sound emitted, while large leaks are characterized by a lower frequency sound. A drawback to this method is that it is difficult to interpret the data provided by the sounds, and so actually determining if there is a leak can be challenging.

Other methods involve analyzing acoustic sounds mathematically using software. Acoustic data is collected by an RF-equipped vehicle that drives by sensor locations or is transmitted to the utility via an RF-based fixed network. The utility then analyzes the data to identify the likeliest locations of leaks.

One example of a leak-detection system that works with a fixed network is the STAR® ZoneScan product, which combines the ZoneScan leak-detection system from Gutermann International (Zurich) with Aclara's STAR Network system. Leak-detection devices placed on valves or hydrants on the main transmit the data automatically to the utility by way of the fixed network. The utility then uses special software to perform a correlation, which identifies relationships between the acoustic data sets to pinpoint the location of the leak.

To serve customers better, utilities must collect the information necessary to understand what is going on in their water distribution systems. SCADA systems are one piece of the puzzle, providing the sensor data necessary to track specific attributes, such as flow rates and chlorination levels, which affect water systems. Fixed network AMI is another piece, allowing water utilities to collect more data from the water system and its users. This data can then help utilities develop programs that improve customer services and reduce water losses.

About the Author:

Laura Wainwright is the product marketing manager for water and gas at Aclara, Cleveland, Ohio. Her background includes experience in coordinating automatic meter reading projects for a major water utility.