GISdevelopment.net ---> GITA 2000 ---> Data Development and Evolution

Addressing the records standardization challenge

Stephen Newman
MidAmerican Energy Company,2811 - 5th Avenue
Rock Island, IL 61201, USA


Midamerican Energy Company
MidAmerican Energy Company is a combination gas and electric utility serving parts of Iowa, Illinois, South Dakota, and Nebraska. MEC is based in Des Moines, IA and serves approximately 653,000 electric customers and 622,000 natural gas customers. Like many of today's energy service providers, MEC was formed from the consolidation of several utility companies. MEC was formed in 1995 with the merger of Midwest Resources (MWR) and Iowa-Illinois Gas and Electric Company (IIGE). Midwest Resources, in turn, had been formed in 1986 with the merger of Iowa Public Service (IPS) and Iowa Power (IP). In addition to the mergers, MWR acquired the Sioux Falls area from Minnegasco and also purchased the Fort Madison, Algona, and Perry gas distribution systems from North Central Public Service.

Project History
The Automated Information and Mapping (AIM) project began in 1988 with a demonstration of AM/FM technology to senior officers at Iowa-Illinois Gas and Electric Company, which resulted in a subsequent request for a feasibility study by the officers. PlanGraphics was hired as a consultant for the project in 1989. The feasibility study was completed and management approval was received in 1990.

In 1991, a RFP for data conversion was issued, and Intelligraphics International was selected as the data conversion vendor. IIGE landbase conversion began in 1991 and was completed in 1993. USGS quadrangles were used as the base map. Electric facility conversion at IIGE also began in 1991. Field Data Services, Inc. (FDSI) was hired to perform a field inventory of electric facilities. The IIGE electric facility conversion project was completed in 1994, while the gas facility data conversion project began that same year.

While the conversion project was underway, IIGE merged with Midwest Resources to form MidAmerican Energy Company (MEC) in 1995. Data conversion of former MWR areas was subsequently approved in 1996. Intelligraphics International was again retained as the data conversion vendor, and FDSI was brought back to perform electric field inventory for the former MWR areas.

The merger created many new challenges for the AIM project team. Most significantly, the emphasis shifted from finding ways to save conversion dollars to an accelerated schedule. The new sponsors placed a deadline for conversion of December 31, 1999, so we now had half the amount of time to complete conversion on an area that was twice the size of IIGE.

The IIGE gas facility data conversion project continued through the merger process and was completed in 1997, with the exception of gas services that could not be placed using an automated digitization process. Subsequently, an internal gas service conversion project began in 1997 to convert the remainder of IIGE services. The landbase, gas facility, and electric facility data conversion projects for the former MWR areas, as well as the internal gas service project, were all completed in 1999.

AIM Database Design
The AIM project team knew that the key to standardization was in the design of the new system. There had to be enough attribution for each piece of equipment to leave no doubt as to what it really was. Unfortunately, costs associated to memory usage created a natural conflict with this principle. System architects from operations often found themselves warring with the computer programmers and AIM project manager over valuable disc space. In order to move forward, concessions had to be made. The system architects debated many considerations as they put the information together.

They usually found themselves asking questions like:
  • "Will the system satisfy all compliance issues from utility regulators?"
  • "Can we easily find faulty equipment in the event of mass failures?"
  • "Is there enough attribution to support system (future) interfaces?"
  • "Will the users understand the type of equipment used by the description?"
  • "Will Plant Accounting have all the information required for their records?"
  • "Is there any equipment used during construction that doesn't need to be captured?"
  • "Will we be able to cost-effectively maintain the information after conversion?"
Generally, the system architects erred on the side of too much information. A past sponsor of the AIM project likened the approach used during the system design phase as "kids in a candy store". Since that time, costs associated with memory have dropped dramatically due to hardware advances.

It is apparent that one of the best decisions made during the AIM conversion project was NOT to skimp on database attribution. The AIM system design has had a greater than anticipated positive effect on everything from system analysis and engineering studies to ease of map interpretation by facility locators.

System Design and Data Conversion Issues
MEC had over 600,000 gas service cards to convert. Original cost estimates associated with gas service conversion were very high when compared to other AIM features. The AIM project team was directed to find a way to hold down those costs.

The project team and the conversion vendor developed a plan to hold down gas service card conversion cost by using a process called automated digitization (splat). The AIM project team studied gas service card drawings and found that gas service paths conformed to 12 basic paths 89% of the time. The team decided to capture the standard service configuration (SSC) number during the gas service card rake-off process and have the splat program place the service paths. A CAD operator would digitize the service paths that did not conform to one of the twelve standard service configurations, which we captured as an SSC13.

The down side to using the SSC process was the loss of information from the service cards. Most dimensions shown on the service cards would not be captured and service fittings would not be displayed on the map. For years, persons involved with District Operations record keeping had been consistently telling the construction crews to show as much detail as possible on the service card drawings.

This led to a very difficult decision for the AIM project team and sponsors. Ultimately the expected cost savings out-weighted the data quantity. The result of the decision was not perceived well in the user community and this subject was almost always raised during the initial contacts between the project team and the users.

The project team learned that the best way to handle the situation was to be straightforward and honest with the users. We never tried to avoid discussions on the subject and we attempted work with the users while they adjusted to the change. We found that this decision made it very difficult to sell standardization.

Legacy Systems Overview
Prior to the AIM conversion project, District Operations maintained their own gas and electric mapping systems. The only consistent aspect from one service center to another was the inconsistency. Each area had its own level of detail, different equipment callouts, different symbology, etc. Timeliness of updates provided to the field ranged from monthly to annually. Persons performing mapping functions ranged from clerical positions to engineers and supervisors. They learned the mapping trade from training passed down through generations of employees who reported to their own service center. When the AIM project team interviewed the District Operations personnel in regards to their mapping system, a common reply to our questions was "I don't know, but that's the way we've always done it".

The project team found that the individual mapping systems offered clues as to the type of equipment installed, but did not tell the whole story. On the bright side, there was usually a records expert for each of the areas that could interpret the information for us. These experts were generally supervisors, engineers, construction foremen, and facility locators who were nearing retirement. In most cases, the experts had been around the area long enough to remember the construction practices used over the years.

Types of Legacy Systems
Due to the number of predecessor companies at MEC, we had an abnormally high number of mapping systems. There was a wide array of sources for AIM conversion, which included:


Some areas would use the same type of map as another area but would use different symbology and equipment call-outs. Source matrices not only had to be created for each type of map but also for each area. The same piece of equipment was often located on four of five maps. To say the least, the project team had their hands full.

Team Building
One of the single most important tasks of the project team was to find the system experts for all of the operations areas. We usually had to rely on a team for each area, since most of the team members were only experts in a specific field. We rarely found an employee or former employee that had a full understanding of all aspects of their system. Pulling together a team of experts was at times complicated. It was hard to get area experts for the amount of time required to gather all of the information needed for conversion. Most of the experts had busy schedules and found it hard to juggle helping the project team and keeping their own projects on track. However, the project team had great luck using retirees as consultants for the project. Most of them were happy to lend their expertise to the project. We capitalized on the fact that it was time to get the information out of their head and into the new system, and they were usually very proud of past accomplishments and loved to share their knowledge and related war stories.

Project Start-Up Meetings
The conversion process would usually begin with a series of "start-up meetings". The first meeting would focus on the change from past mapping practices and map products to the new system. In subsequent meetings, area operations system experts would be interviewed in a group setting to determine past construction practices and how to interpret their maps.

Believe it or not, the AIM project team was able to use the lack of mapping standards and variety of map products to their advantage. The project team had to rely heavily on District Operation's support during the conversion project, and their involvement gave them a sense of ownership in the new system. The project team also benefited from days-and sometimes weeks-of on-site communication. Operations personnel appreciated the project team's devotion to the records cleansing process. The start-up meetings helped break the ice with issues related to standardization of records, but generally area operations groups would remain skeptical until they actually had their own converted data to review. The project team realized that although this contact was very important, there was still a lot of "selling" to do.

Data Conversion Specifications
The project team created conversion tables from the start-up meeting interviews for each operations area. Some areas had different sources within their service territory and had to have multiple sets of conversion tables. These tables identified all symbology and facility call-outs used in their mapping system and related them to specific items of plant for conversion. Because each area had developed their own ways of calling out facilities, it was hard to find any consistency across the company.

Usually the symbology and call-outs were generic in nature; specific manufacturers and model numbers were rarely identified, and specific equipment information would usually be derived from one or more of the following factors: symbol type, call-out, date installed, material, size, and pressure/voltage. Most gas fittings were also derived from the connected pipe information. Understandably, the conversion tables would get quite complex, as they not only had to account for information that was present on the maps but also had to cover occasions when part of the critical information was not present. Additionally, we found that older areas of towns had the least amount of information, which spurred the project team to create synopsis tables that identified original construction practices for each town. Facilities in each town would be defaulted to the values in the synopsis tables. Typically, the only facilities called out in those areas were the modifications made to the system after original construction. The standard conversion tables would be used to convert all exceptions to the original synopsis table.

Enhancing Legacy Mapping Systems
The first and largest area that was converted in the former MWR area used a mapping system that had very little detail. We ran into significant issues affecting conversion that had to be decided in a short amount of time since senior officers had just mandated an accelerated schedule and our conversion vendor was looking for work. The electric underground maps did not show service lines, so there was no way to determine the feeding pedestal or pad-mount transformer, nor did they show phasing connections at riser poles. Both of these issues required field verification beyond the reasonable scope for the field inventory subcontractor. We had to connect the underground services to the pedestal or transformer in the closest proximity, and we had to randomly assign phasing connections at riser poles.

Additionally, the gas mapping systems in these areas also did not have very much detail. In order to improve the maps, we brought in a team of retirees to capture missing information from the construction work orders and scrub the information to the source maps for conversion. The twelve retirees who were hired all had gas system construction background, i.e. were supervisors, construction foremen, stores clerks, etc. We were pleasantly surprised with how easily they picked up the scrub process. We also deliberately overstaffed the project to allow them to work on their own schedules, so that a few of them could go golfing on nice days. They required very little supervision and in general liked being part of the team, and we benefited from their knowledge.

The scrub process finished ahead of schedule and under budget, receiving an extremely positive response from the area operations group. Our users had the chance to observe the scrub process and provide input to the process. Having the detail in AIM saves them from having to manually pull the information from work order files and allows them query capabilities that they use to make better business decisions.

Conversion Vendor Issues
Very early on, it was clear that it was impractical and inefficient to treat this project as one big job due to the significant variations between different areas of the service territory. We broke down this project into 38 smaller tracks, 13 gas, 14 land, and 11 electric tracks. The conversion vendor decided to assign three project managers to this project in charge of land, gas, and electric tracks. Due to the critical interdependencies of their efforts, the project managers had to orchestrate their efforts through an overall project coordinator. The 38 tracks were of various density and complexity. As a result, the conversion vendor had teams that were as small as 5 and as large as 100 people assigned to the various tracks. This process allowed them to have many tracks running in parallel. At one point, the combination of the conversion vendor and its four subcontractors personnel totaled over 250 technicians involved in over fifteen tracks from gas, land and electric projects. Timely communication was a major reason for the success of this project. In addition to many face-to-face meetings our project coordinator and project managers were responsible for conducting and documenting formal weekly conference calls and monthly detailed status reports. We also held a quarterly senior management project status review meeting.

Each track was implemented and completed by following the same process:
  1. MEC and the conversion vendor review all available sources (paper, digital, and tabular) for the track
  2. The conversion vendor creates a detailed source matrix at the feature and attribute level
  3. MEC and the conversion vendor develop the necessary conversion tables
  4. MEC and the conversion vendor update the conversion specifications
  5. The conversion vendor compiles all the findings and documentation and delivers to MEC
  6. MEC reviews, enhances, and approves the conversion documentation
  7. MEC and the conversion vendor select the pilot area, small enough to convert quickly, yet large enough to be representative of the data mix
  8. The conversion vendor and MEC enhance conversion specifications including the conversion tables, as issues were resolved through the pilot conversion process
  9. The conversion vendor delivers the pilot product along with all the required documentation
  10. MEC conducts a detailed and thorough pilot review process
  11. MEC and the conversion vendor hold a pilot review meeting
  12. MEC and the conversion vendor update the conversion specifications including conversion tables as a result of the pilot review
  13. The conversion vendor documents and communicates conversion specifications to its data collection and data conversion teams
  14. The conversion vendor proposes a "Detailed Delivery Schedule" for the each track
  15. The conversion vendor makes deliveries every two weeks
  16. Continuous process improvement as a result of MEC's review and scoring every delivery
User Training
We addressed the change to a standardized system in our AIM training sessions, which were usually held after the final delivery for each operation area. Training was limited to half-day sessions. Typically, at least an hour of the first training session was devoted to a discussion on the impact of changes brought about with the new system, and we spent an hour reviewing the conversion process. The rest of the time was usually devoted to learning how to use the map viewing software.

One training process that worked well involved having a trainee randomly pull all of the legacy maps and records for an area. The trainers would then compare the maps and records to the new system. We found that this process helped the users learn to interpret the new information and also helped them develop their own confidence in the quality of the new system.

Through trial and error, we found that it was impossible to cover all aspects related to the AIM system in one session. We tended to lose the interest of the trainees in early sessions by getting too technical, too fast. We also dropped a set format for training. We found that the response from training was much better when we let the questions from the trainees help drive the session.

Our goal is to spark their interest in the system with the first training session. We attempt to emphasize the positive affects that the AIM system has on their daily work processes, and we encourage them to use the new system for those processes. By using the system, they will learn at their own pace. We also let them know that we are willing to provide follow-up training on their schedule.

Conclusion
MidAmerican Energy Company has successfully implemented their AM/FM system and, in order to do so, the AIM project team had to navigate through on ocean of standardization issues.

Further, as with many other aspects of the conversion project, timely communication is a significant key to success. We involved the users from start to end of the conversion process, and they fought side by side with the project team through all of the conversion issues. In the end, the users understand the product and how it was created, and the project team is dedicated to providing continuous training and support as the users' knowledge of the systems increases.

Most importantly, the AIM project team created a product that could stand on its own. We can communicate with the users all we want, but eventually it's just the user and the system, one-on-one.
© GISdevelopment.net. All rights reserved.