Located near Allentown, Pa., the Lehigh County Authority (LCA) is using an integrated water system model to better evaluate and optimize its distribution system and set the stage for future growth. The model automatically links to LCA's existing geographic information system (GIS) as well as the customer billing data and supervisory control and data acquisition (SCADA) systems (Figure 1). Through these flexible connections, LCA can update its model to easily replicate field conditions and process a wealth of data previously too cumbersome to handle manually. The model also can simulate a snapshot of water system conditions, ranging from steady state flows to extended period simulations for water quality evaluations.
LCA serves approximately 20,000 customers through 13 water systems and six sewer systems. The computer model was built for its largest water system, reaching approximately 90 percent of its customers and handling an average daily demand of 6.5 million gallons per day (mgd). This system includes 260 miles of distribution mains, five pressure zones, 18 wells, three booster stations, and four storage tanks. The model was strategically implemented to analyze system operations for a major LCA interconnection with a neighboring water system in the city of Allentown. The integrated approach results in a more accurate model that requires significantly less time and money to maintain.
Thanks to ingenuity and computer modeling technology advancements, LCA's model was created using original scripts and operating procedures developed by Gannett Fleming, Inc., combined with Bentley Systems, Inc., model software tools. This approach enabled LCA to closely integrate system infrastructure, operational data and setpoints, and customer billing data into a system that enhances model calibration and analytics. Development tools successfully integrated various data sets while allowing for a repeatable process that can be duplicated on future model updates. This project required more communication between the model developer and LCA, improving data coordination.
A clear understanding
Because LCA maintains up-to-date GIS mapping, including the majority of data needed to develop a water system model, it was critical to maintain the link between the model and GIS, while recognizing inherent differences between a water system GIS and model (Figure 2). GIS provides a spatially accurate inventory of water system infrastructure in an easily accessible manner. Although it requires much of the same data, a water model is intended to establish a hydraulically accurate representation of the system.
As a result, there are often pipeline connectivity issues and data deficiencies when developing a model from GIS that can outweigh integration benefits. These challenges were addressed in the LCA project through a combination of standard operating procedures. Also, customized and existing GIS and model software network topology review tools helped LCA to fully realize the benefits of model-GIS integration.
Repeatable procedures were developed to automate the assignment and future updates of customer demand in the model. Water use records are exported from LCA's customer billing system, run through a series of customized scripts, and imported into the model. This provides an accurate spatial and time-sensitive depiction of water system demands. Developed by the project team, custom scripts exported water use data with system production records to identify non-revenue water that is produced and lost before reaching the customer.
A similar automated method enabled model integration with SCADA. Data generated was then used with the SCADAConnect tool in Bentley's WaterGEMS to complete the import process. This connection enabled a larger sample of field data to be used during model calibration that allows real-time or historical field condition analysis.
Setting a higher standard
LCA's project team implemented several layers of quality control to ensure pipeline network integrity when importing GIS data into the model and to ensure proper hydraulic representation of the pipeline network. To further limit pipeline connectivity issues within GIS, scripts were developed to identify areas that did not conform to established pipeline network rules. This creative engineering approach alleviated the majority of pipeline connectivity issues prior to model import.
The effort required to review the model pipeline network became part of GIS development rather than the model import process. This approach minimized duplicate effort and allowed for more efficient data integration. Once imported into the model, the system architecture was re-evaluated using the WaterCAD/WaterGEMs Network Review utility. This procedure reduced the amount of GIS editing required following model data import and created greater symmetry between the GIS and the model.
LCA's project team established standard procedures to address connectivity guidelines, labeling schemes, and GIS tagging revisions, while focusing on segmenting pipes for large users, hydrants, and valves. Each pipe also has a GIS identification (GIS-ID) automatically populated in LCA's GIS, serving as a link between the model and the GIS and avoiding integration hassles. The WaterCAD/WaterGEMS GIS-ID allows the model to recognize and update only the pipes that have been modified or created, rather than the entire pipeline network.
GIS-ID provides a unique identifier for associating domain elements in the model to records in the GIS. This link retains association of a given element even if the model element (main) is broken within the model. A unique registry-style string of 36 characters, GIS-ID minimizes the chance for duplication or operator error.
Automation enables growth
To integrate LCA's customer billing data, project scripts were created that can be used for future endeavors. These programs import data into GIS, summarize water usage by customer type for a user-defined time period, determine closest model nodes to meter, assign model node demands, and import data into the model. The automated processing and assigning of customer billing data is a significant enhancement compared with manual methods (Figure 3).
"With the links established to our existing GIS, customer meter databases, and SCADA, we know that our model can be easily updated as our system continues to grow," said Emily Gerber, LCA GIS analyst.
A creative approach, the project team used SCADAConnect to establish initial settings within the model and the Time Series Field Data function to assign field data comparison data. The result was a more effective calibration process and timely operational data assignment, rather than traditional model data dumping from SCADA. In fact, the manual data input process for model development was virtually eliminated by integrating GIS, customer data, and SCADA with the model through use of ModelBuilder and SCADAConnect (Figure 4).
Although the concept of integrating water system models with other data sources is not new, success has been limited. The majority of integration projects have focused on GIS with mixed results, while examples of SCADA integration are few and far between. The LCA project is unique because it successfully applies integration across GIS, customer billing data, and SCADA while establishing procedures to promote process benefits throughout the model's lifecycle. This is where past integration efforts have fallen short. Eventually, water utilities experience loss of integration, system performance, and cost savings – a key consideration when developing and updating a hydraulic model and maintaining integration.
Efficiency equals savings
System computer models, GIS, customer billing data, and SCADA are often managed by separate staff within a water utility, as is the case with LCA. To compensate, project team procedures established for LCA's integrated water model enable utility staff to use separate systems effectively. This translates to effective communication between multiple data sources and allows independent management of each source. Overall, the new system model requires significantly less effort and cost to maintain when compared with traditional stand-alone versions.
Historically, the development and update of a water system model similar in scope to the LCA system has been tedious. Without the benefits of advanced integration, the standard process involved manually developing the pipeline network, locating the assignment of customer water use, and entering facility data. For a system the size of LCA, this process could take four to eight weeks. Once LCA integration procedures were established, the same process took a few days.
"Our integrated approach saves 40 hours of labor each time the pipeline network and the demands are updated in the model," said Aurel Arndt, LCA general manager. "The model also eliminates the need to duplicate efforts on future updates."
Project benefits were realized immediately. During calibration, LCA was increasingly confident that the model was appropriately developed, and that demand and facility data was accurately represented. This led to a model instantly capable of system evaluation. While the effort saved during development and calibration benefited LCA initially, true cost savings will be realized by the time saved on each future model update because of procedures put in place upfront.
From a cost control perspective, using the model to evaluate water quality within the LCA distribution system has a positive impact on the local population. Automated integration promotes frequent model updates, resulting in a more accurate and sustainable model used to determine the most effective system operation. The new model optimizes LCA facilities and identifies the most effective means to promote optimized water quality.
Project savings were presented on an international stage during the Bentley 2012 Be Inspired Innovations in Infrastructure Conference awards competition in Amsterdam, The Netherlands. A finalist in the Innovation in Water, Wastewater, and Stormwater Networks category, the integrated model is featured as an extraordinary infrastructure project in Bentley's 2012 Be Inspired publication, "The Year in Infrastructure."
View or download "The Year in Infrastructure" at www.yearininfrastructure-digital.com/yearininfrastructure/2012#pg1.
View a portion of the Be Inspired Conference presentation:
Michael Mehaffey, P.E., is a water hydraulics expert and senior project engineer with Gannett Fleming responsible for the design of water supply and wastewater projects. Michael Brown, P.E., heads Gannett Fleming's Water Modeling and Hydraulics and Water Practice Asset Management Groups. Brown and Mehaffey led the development and calibration of LCA's integrated water model. Ricardo Duarte is a marketing specialist with Gannett Fleming.