Water hammer events can result in service issues for utilities, as the sudden increase in pressure can contribute to pipe fatigue, eventually causing pipe failure—a costly situation for municipalities.
Negative pressure events (transients) can have significant impacts on water quality as well: The low pressure created inside a pipe can pull stagnant water into a pipe through leakage points, cross-connections and faulty seals or joints. When this stagnant water enters a pipe, it can bring waterborne disease molecules and harmful biologics into the main. This can contaminate a municipal water supply with nonpotable water, potentially making thousands of residents sick.
Finding the sources of these events and understanding their severity can allow water utilities to mitigate potentially hazardous situations. Knowing these sources can help municipalities guard against costly, often destructive water main breaks.
Locating the source
Pumps that turn on and off and valves opening and closing too rapidly are usually the source of transient events, which are often produced by large industrial customers.
Generally, water distribution managers know what could be causing pressure events. The challenge comes in gathering the data they need to determine exactly where and when these events take place so they can recommend an action plan for changes in these areas.
The solution to acquiring this data is simpler than many utilities realize. Hydrant pressure recorders and line pressure recorders monitor water pressure throughout distribution systems, capturing the waveforms of water hammer and negative pressure events. Pressure recorders developed by Telog Instruments, for example, can store up to 450 events of variable duration that may occur over many months of onsite monitoring.
These recorders can be placed upstream or downstream of the predicted problem site, collecting data during transient events at rates of up to 20 measurements per second. Field personnel can then transfer the data using a portable PC or a dedicated transfer unit and download the data to the utility’s information management system.
The resulting comparisons of trend and impulse data provide dramatic graphical illustrations of the impulse event with the accompanying numerical data that documents the pressure change and the duration of the event. This data provides proof of the event and its impact on the system components through which it passed. It can also be used to calibrate a distribution system model to help predict the impact of future events.
How impulse recorders process data
To illustrate this process, we begin with a recorder configured for trend recording and impulse detection with the following parameters:
• Sample rate: 50 ms
• Trend recording interval: 5 minutes
• Pressure change threshold: 5 psi
• Pressure change within: 10 seconds (scan window)
• Store preimpulse data for: 15 seconds
• Store postimpulse data for: 20 seconds
Based on these settings, the recorder will sample the pressure 20 times per second and compute the difference between the maximum and minimum values within a 10-second rolling scan window.
If the difference is greater than the pressure change threshold, the recorder immediately triggers an impulse condition. As determined by the preimpulse data setting, the recorder stores 50 ms data for the 15 seconds prior to the impulse detection, as well as for the current sample.
The recorder will continue to store 50 ms samples as long as the difference within the scan window is greater than the pressure change threshold. Once the difference drops below the pressure change threshold, the impulse event will be considered over and the recorder will continue to store data at the 50 ms sample rate for 20 seconds, as determined by the setting in the postimpulse data window.
If the recorder detects another impulse at any time, the device goes back into impulse mode and repeats the process.
When the data is downloaded into the accompanying information management software—in this case, Telogers for Windows—users can view impulse data along with the trend data. For durations in which impulse data was detected, the software displays the impulse data on the chart in a different color. Since impulses are asynchronous, the software will not connect the points between the trend and impulse data transition on the chart.
The software also displays an extra bar at the bottom of the chart to indicate an impulse event. In the tabular data, the impulse data is displayed under the average column, and the rows that contain the impulse data are highlighted to distinguish them from the trend data.
The next step: modeling
Some water distribution modeling companies have developed software that may allow water utilities to predict water transient events. These packages work with standard analysis software such as Telogers Enterprise to simulate transient phenomena and predict the impact of events with a wide range of causes.
The critical element to modeling of this kind is calibration of the model, to be sure that the model reflects the likelihood of actual events. Data recorded by Telog’s impulse detection devices can be instrumental in calibrating the modeling program, improving a utility’s ability to forecast water hammer and negative pressure events and plan practical solutions that may prevent these phenomena from damaging its water distribution system.
Using real data collected during pressure events, utilities can verify the events’ sources and make the case for mitigation of their impact on system components. With this empirical data, users also can predict the impact of future events, calibrating a model to reduce or eliminate the impact of such events before they happen.
