Establishing dynamic, optimal, sewer line cleaning cycles using GIS
Through a combination of technology and business processes, West Valley Sanitation District (WVSD), Santa Clara County, Calif., sought to improve its wastewater collection system operations. The effort ranged from streamlining work practices to using advanced technology such as a complete computerized maintenance management system (CMMS) with geographical information system (GIS) connectivity and closed circuit television (CCTV) data linked to a new maintenance system to develop routine schedules, track work, and improve operations. This transformation led to improvements in the efficiency and effectiveness of operations, with increased quality as a direct result of the effective application of routine cleaning, improved work reporting and practices, application of good business tools, and integration of automated work processes.
West Valley Sanitation District, Santa Clara County, Calif.
WVSD currently maintains a 435-mile sanitary sewer collection system that serves a population of approximately 120,500. Its scope of responsibility encompasses a 29-square-mile service area that includes four cities and unincorporated areas of Santa Clara County. The district undertook a review of maintenance operations in 2002 that followed a multi-step effort: evaluate and identify opportunities, select software, and implement in 2003 both software and improvement opportunities (Figure 1).
During 2006, the district sought further improvement and developed geographic zones in the GIS to group and schedule work. Integration of the GIS with the CMMS provided valuable management information and has already shown some promising results since beginning work in the zones in 2007.
Figure 1: West Valley Sanitation District multi-step approach to reviewing maintenance operations
Previous work processes
Previously, the district performed four preventative maintenance cleaning activities: balling, high-velocity cleaning (HVC), power rodding, and hand rodding. Work was generated by service requests and monthly routine cleaning schedules for sanitary sewer lines generated out of the existing maintenance system. The service requests and routine cleaning schedules were submitted to the maintenance crews. Crews completed the work and documented work performance data on various work reporting forms. The primary system used for data input and generating line cleaning schedules was a FoxPro-based, Computerization of Sewer Maintenance Operations (COSMO) system that was initially designed in the early 1980s. This system housed 25 years of line cleaning data.
Routine sanitary sewer line cleaning was scheduled through the use of the COSMO system. The schedule documented the last date of cleaning, priority (determined by the severity rating), and next scheduled cleaning date. The preventative maintenance programs were scheduled on cycles based on the historical debris severity of each line segment. Though the COSMO system captured critical maintenance data for scheduling, it lacked flexibility to retrieve information and the capability to determine management information such as unit cost, productivity, or annual work plans.
Work performance data from the work reporting forms that were input into COSMO included labor hours, linear feet of pipe cleaned, amount and type of debris removed, and a debris severity rating. Work production information was recorded and tracked on nine forms based on the type of asset where the work was performed (main sewer, building sewer, structure) and the type of work performed (construction, maintenance, repair, miscellaneous). In addition, crews used four time tracking sheets—individual time sheet, daily time report, daily work report for miscellaneous work, and overtime work report for miscellaneous work.
Other management information related to sewer maintenance also existed within several databases. Many of these systems were independent of each other and made it difficult to collate information for decision-making purposes. Activity cost information resided in the Fox Pro Accounting system, hours and accomplishment data was tracked in COSMO, sewer line inspection data was housed within the CCTV program, cost per unit data was manually calculated and filed, and activities with associated labor hours and accomplishment was tracked in MS Excel. Further, the GIS that mapped the collection system network resided with the engineering department, yet had not been applied to operations.
Work process reengineering
District management team took on the challenge of evaluating the current operations and business practices. The goal was to determine and identify the positive aspects of the management process in place and enhance them, and to identify opportunities where work improvements in flow and or methods could be applied. Evaluation of the current operation involved interviews with key management staff, administration support staff, the operations supervisor, and field workers. Further, the major maintenance activities were observed in the field and work flow processes documented. This was supplemented with an analysis of the 25 years of maintenance data; review of GIS data in engineering; and review of operational policies, procedures, and regulations. All of this information was compiled in detail to establish a baseline of existing operations. Comparisons of baseline data versus benchmark information resulted in 26 recommendations for future success and included database consolidation, activity guidelines, increased use of HVC, development of a performance-based budget, streamlined data collection, and updated and enhanced technology.
Database consolidation—The district operated with multiple databases tracking various work information. These databases included manual records, MS Excel, Fox Pro Accounting, and Fox Pro COSMO with redundant information being collected, input, and compiled. Report generation was accomplished by transferring information from various manual files and databases and manually calculating unit cost and productivity data in spreadsheets. A goal was identified to establish one primary database where all the information would reside and where various reports could be easily produced. Also, the district was in a unique position of having 25 years of maintenance history data recorded by line segment in COSMO. This valuable information was retained in the new system and used for planning purposes. Further, the GIS mapping that existed in engineering was used and linked directly to the maintenance data.
Development of activity guidelines—The maintenance operation was found to be in need of best work management practices for the maintenance activities. These performance guidelines can assist maintenance staff, as well as management, to communicate, standardize, and document work tasks performed. Specific best management practices or performance guidelines were developed for each activity in collaboration with management and maintenance employees. The guidelines for each activity included activity definition, mix of resources required (labor, equipment, material), a method to conduct work, and expected daily productivity based on determined work unit. This information served as a baseline and resource for all of the work planning, as well as providing the additional benefit of common terminology, a device for training, and most importantly, a tool for continuous improvement.
Increased use of HVC—WVSD uses several sewer line cleaning techniques, including balling. The equipment to perform this effort consists of an assortment of sewer balls designed to fit different diameters of sewer lines, a tag line, winch, reel, and a specially modified water truck. The method uses the pressure of a water head to create high-velocity water flow around the ball. This activity was based on historical efforts and has been minimized and replaced by the HVC activity. Balling was retained for specific areas that are difficult to access where HVC is not feasible. This change to HVC results in a lower unit cost of cleaning, and increases productivity while maintaining effectiveness.
Performance-based work plan and budget—A detailed performance-based work plan was developed based on the criteria established by activity. The information developed within the activity guidelines served as the starting point. The plan included key activities performed with associated resources (labor, equipment, and material) linked to an inventory (linear feet of sewer lines), level of service (percent of inventory serviced), expected work quantity (production goals in linear feet), average daily production (expected linear feet cleaned based in a work day), anticipated crew days (work quantity divided by the average daily production), planned crew size, anticipated labor days to complete the work (crew days multiplied by the crew size), projected labor costs (including overhead), equipment and material costs, and total cost to perform the documented activity. Table 1 provides the general overview of the four major sewer-cleaning activities.
Table 1: Overview of four major sewer-cleaning activities
Costing included determination of the appropriate overhead to allocate to each activity in the work plan. Two overhead types were determined—full and avoidable. Avoidable overhead rate is that direct cost which could be eliminated if work was not done in-house. This includes employee benefits, training, uniforms, and any other cost directly related to work effort. This value was calculated at 120 percent. This cost takes into account direct productive labor, leave and non-productive labor (paid hours and overtime), benefits, training, uniforms, and partial administrative costs related to operations. The full overhead rate was calculated at 281 percent. This cost includes avoidable overhead costs in addition to full administration costs and a portion of salaries from support functions such as human resources. The avoidable overhead is applied to determine the actual cost of performing the work and for cost comparison to other agencies or contractors. The fully allocated overhead is applied to recover the full and complete costs for performing work for others.
Reduction of redundant data collection—The maintenance staff at WVSD reported work on many different forms that were developed based in conjunction with the COSMO scheduling program. Each of the forms was comparable in design and captured similar data. These work reporting forms, along with the data recording processes, have been eliminated in the new system approach. Work is currently assigned by grouping line segments based on the time to perform the specific cleaning operation and by geographic area. The new system captures crew and equipment time and associated costs, along with linear feet cleaned.
Update software capabilities and technology—A key task necessary to improve operations and business processes was selecting the software and technology that would best meet the needs of the district. The district obtained a configurable, off-the-shelf system to meet this objective. Explicit goals were outlined for the software needs of the district in the request for proposal (RFP) that matched the identified recommendations. Critical needs of the system included an asset management component, the capability to calculate a condition assessment, issue and track work orders, and the ability to manage routine and response-driven work. Other components included the ability to plan work based on defined activities, to develop quantifiable production-related goals, and to produce outputs that help the district manage the operation with efficiency and effectiveness.
The selected system implements a state-of-the-art management system that allows the district to plan, track, and manage routine and response-based work, as well as use related GIS and CCTV technology.
During 2002-2003, COSMO continued to be used as the primary system for scheduling the cleaning of sewer lines. Using this original approach resulted in 1,138,000 linear feet cleaned and included all major cleaning activities of HVC, balling, power rodding, and hand rodding.
In 2003-2004 the WVSD implemented the new approach to maintenance. This approach included new business processes and an improved line-cleaning schedule. Some improvements included selection of the most effective cleaning strategy, minimizing intensive manual efforts such as balling, determination of optimal crew size, and grouping of work. Power and hand rodding methods were used in limited and specific locations. The HVC application was increased using multiple cleaning heads for effective cleaning results and better ADP compared with other cleaning methods. Using this new approach resulted in 1,920,000 linear feet of sewer line cleaned during the first year and 2,192,000 feet cleaned in 2005-2006 (Figure 2). This includes all major cleaning activities of HVC, balling, power rodding, and hand rodding. This represented a 92-percent increase in cleaning production compared with the year prior to the process improvements.
Figure 2: Linear feet of sewer line cleaned each year by activity (1,000 feet)
Further, business process improvements included pre-populating work orders to minimize recording, production of linked location maps, elimination of redundant work reporting forms, reduction of time reporting, elimination of multiple databases with all work data input into one primary system, documented work methods and expectations for each activity, better documentation of work, and easier retrieval of data for various work reports. Further, the system directly links GIS in a two-way capability with the maintenance system. The condition ratings from the CCTV are completely integrated with the new maintenance system. All CCTV data can be directly retrieved from either the maintenance or GIS systems. The collected CCTV data now resides within the system. Capabilities also include the ability to display assigned routine work orders for select time periods (daily, weekly, monthly) and track the progress.
The district recently grouped all sewer lines into geographic zones in an attempt to reduce travel and setup time for work crews (Figure 3). This was achieved by using GIS technology to determine total lengths for sewer lines and grouping them into relatively equal distributions of approximately 100,000 linear feet. Twenty-four "GeoZones" were developed and scheduled to be cleaned, one per month, to clean the entire network within 24 months. The 25 years of historical maintenance data that was recorded and stored in the COSMO system was used to develop this systematic routine line-cleaning plan. The data contained the following information: each segment of pipe with an associated upstream/downstream identifier, street location, segment length, hours worked, and the type (grease, roots, gravel/sand) and amount (measured in pounds) of debris removed during the cleaning operation.
Figure 3: Sewer lines grouped into GeoZones using the GIS map
After grouping the lines geographically, cleaning cycles were established for each segment in the GIS. Historical data was analyzed to determine conservative cleaning cycles of three, six, 12, 18, 24, 30, and 36 months based on the severity of debris removed and the frequency of prior maintenance. An algorithm was coded that automatically determines optimal cleaning cycles based on the defined logic. Users can update these values in the CMMS.
To overcome the challenge of having differing routines on adjacent sewer lines, conservative cycles were applied to lines that may not need to be cleaned as frequently as neighboring lines. For example, if the decision logic determined a 24-month cycle for a single line among other lines that were on an 18-month cycle, the 18-month cycle was applied to all lines to avoid a repeat visit by maintenance crews for something that could be cleaned more efficiently and inexpensively by grouping it with adjacent lines. Sewer lines that were 6 inches or less in diameter were restricted to a maximum cycle length of 24 months.
Other assumptions used in determining cleaning cycles were that a two-man crew could clean approximately 25,000 linear feet per week. This is an average production value based on more than 25 years of historical data. There are also some problem lines that are on fixed routines and were not included in the analysis.
After determining the initial cleaning cycles, a monthly work plan was developed that calculated the amount of linear feet to be cleaned each month and the location of lines to be cleaned. The district is using one crew for routine/geographic cleaning while a second crew is used for overdue or emergency cleaning. It is estimated that all overdue lines could be cleaned in about four months; the second crew could transition to cleaning lines that are due for routine maintenance but are not included in the current month’s geographic zone. The entire network is scheduled to be cleaned every 24 months.
Specific reports have also been developed that give detailed work history for individual segments and compare current cleaning cycles to cycles proposed that provide guidance for modification and optimization of cleaning. The district is now planning, after calibration, to develop system logic that in the CMMS will automatically configure and optimize this effort.
After the first few months of implementing the geographic cleaning routines, performance data was extracted from the CMMS to measure improvement. Some of the backlog work was taking longer to complete, yet noticeable differences occurred between crews. Unit costs were determined for major activities by dividing the total cost to perform the activity by the amount of work accomplished. The result is the actual cost per linear foot of main line cleaned. Crews assigned to the geographic zones showed unit costs of 15 percent less for HVC and 21 percent less for power rodding. This is likely attributed to the reduced setup costs and travel time achieved by strategically planning the work.
Productivity can also be measured by determining the amount of feet cleaned per labor hour. Crews assigned to geographic zones demonstrated 19 percent greater productivity for HVC (Figure 4) and 22 percent for power rodding.
Figure 4: Linear feet of sewer line cleaned per labor hour using high-velocity cleaning
The WVSD applied technology and multiple best business practices for operational improvements, established a performance-based budget, implemented new maintenance system software with GIS capabilities, and developed a systematic line cleaning schedule. This effort used a highly effective automated management system along with GIS and CCTV technology that made use of 25 years of maintenance history. This transformation led to an increase of work production, improved work processes, better documentation work, and the ability to provide a process for continuous improvement. Further benefits are anticipated, such as reduced travel and setup times, optimized routines for efficient cleaning frequency, and less stoppages and overflows as a result of systematic scheduling of preventative maintenance.
Overall, district customers now have more effectively cleaned lines at lower cost with fewer sewer backups or disruptions. Management has a tool to analyze and more effectively manage operations, maintenance, and capital planning. And, employees have received recognition for a job well done, including higher compensation as result of productivity increases.
Robert R. Reid, P.E., is district manager and engineer for West Valley Sanitation District. He can be contacted at email@example.com. Peter V. Sevcik, P.E., was director of engineering and operations for the West Valley Sanitation District and is currently district engineer for Nipomo Community Services District. He can be contacted at firstname.lastname@example.org. Harry Lorick, P.E., is principal of Manhattan Beach, Calif.-based LA Consulting. He can be contacted at email@example.com.