Fast-Track Electric Utility Inventory- A Case Study
Larry Camp Baker GeoResearch 8120 Woodmont Ave Bethesda, MD 20814 Lonnie Fluharty Allegheny Power Company 1310 Fairmont Avenue Fairrnont, WV 26554
Introduction
As part of their AM/FM/GIS implementation project, Allegheny Power directed the work of a GeoResearch Inc. to perform electric distribution system field inventory and mapping services along 24,500 circuit miles in a 12,000 square mile service area. Over 500,000 distribution utility poles needed to be mapped in a 13-month period, representing one of the largest and fastest GPS/GIS field inventory and data collection efforts ever undertaken by an electric utility. The project required the rapid staffing and effective management of over 40 field data collectors - a challenging task. This paper will discuss the management and planning issues involved in completing such a large-scale, fast-track field data collection project on schedule. Project planning considerations, recruiting & training, management of a large field staff, field data post processing issues, and QA procedures will be addressed. The Electric Distribution mapping was part of an initiative at Allegheny Power to build a geo-spatial inventory model of their electrical facilities for their AM/FM system. This initiative also includes paper-to-vector conversion and attribution of other Division’s data. Once the database is complete, Allegheny Power will have a connectivity model of their network for load analysis, design, and outage management processes. The following statistics of this project are provided: Allegheny Power’s System:
Field Office Setup One of our first tasks was to determine the best locations to establish our local field ofilces for processing the GPS pole data, managing the local work force, and providing a GPS Correction Base Station. Allegheny Power has many “Service Centers” geographically spaced throughout their territory which provide efficient coverage to their customers. We chose 5 of those locations (Mercersburg, PA; Morgantown, Elkins, Gassaway, and Parkersburg) to serve as our local base of operations. Each of our local offices needed to provide differential corrections for the GPS field collectors in a 60-100 mile radius. The base stations operate a continuous GPS receiver which store daily files of correction factors used to remove the effects of Selective Availability from the roving field GPS receivers. We selected and rented a suitable house in each area, with a somewhat unique request to the realtors, “an unobstructed, full view of the sky”, since the inherently weak GPS signals are easily blocked and distorted by trees, walls, and buildings. Once the antennas were secured to the roofs, and the hardware & software was in place, we needed to obtain the precise coordinate of the antenna by reverse GPS surveying from a local 1st Order control monument. Each office was then outfitted with all the necessary modern equipment to perform our work including desktop GIS computers, copiers, fax machines, and software such as ArcView, Postpoint, GeoLink, MS Office, and FTP/Internet suites. Recruiting After we established the physical properties of each office, it was time to place the personnel that had been recruited during a parallel process. We were very fortunate to have several former Allegheny Power employees available to work for us during this project. Most had Design, Engineering, and Line experience and were invaluable to us as collectors, auditors, trainers, and supervisors. They also knew many of the unique characteristics and coverage areas of many of the circuits we would map. Additionally, these former Allegheny employees knew of other friends and colleagues we could draw upon for future staffing. Many also had strong professional connections all over the state as a result of long careers with the Power Co. in working with contractors and the public. Our requirements (“desires” really) of new Field Data Collectors for this task were pretty stiffi experience with personal computers, GPS, cartography/GIS, electrical systems, and perhaps most importantly, the ability and desire to work alone every day walking from pole to pole through the hills and hollows of West Virginia. Our recruiting efforts included advertisements in local and major newspapers in West Virginia, university newspapers and bulletin boards, the Geography and Engineering departments of several schools, and state employment agencies. Our initial crews included 6 former Allegheny Power staff, and several geographers, foresters, and engineering graduates. None had experience in each aspect we desired (of course, not many would), but all had at least one important aspect well covered, and with good training and teaming, all became productive and accurate in a relatively short time. Training Training was one of the most difficult of many difficult aspects of this large project for us. There were so many new concepts and specifics to learn to do this job accurately and effectively. A few of the concepts we needed to teach were:
The next training topic was the Global Positioning System and how we receive and utilize the data from it. I think we spent too much time here and got too deep into the frequencies, carrier phases, and signal processing than we really needed. It also overwhelmed many of our collectors and they had trouble concentrating on the more meaningful topics that followed. In subsequent training sessions, a lot of technical details of the Global Positioning System were eliminated. The GPS field collection software was the next training topic. It was useful to run simulations of the software demonstrating the pole collection that we’d be soon performing in the field. We were able to train each collector how to start the software (DOS commands), how to name files, and how to collect and correctly attribute each of the features we would need to map (Poles, Pad-Mounts, Secondary Poles, Transformers, Fuses etc.). Each collector received plenty of hands-on practice in-office on a wide screen monitor to gain familiarity with the software and the collection methodology. The simulated files were then translated into GIS maps that were then reviewed for connectivity and attribute accuracy. This was one of the most useful parts of our training program and prepared our new collectors for the field practice to follow. The field practice involved everyone mapping a circuit from a nearby substation. A circuit was selected which had a representative mix of features on it, and was in a somewhat rural area (to allow 20 people to safely work and walk around the poles). This was the first opportunity we had to identify the phasing of the conductors, determine the size and type of conductors, and identifi the devices on the poles. We spent nearly a week on the practice circuit, working slowly in the cold and snow of November, to give each collector experience on each aspect of the job. Again, the aspects of the job were to: 1) identify the characteristics of a pole, 2) record the information accurately, and 3) determine which pole to map next. The task began to seem much easier when viewed in these simplistic terms. One of our most beneficial training ideas was to pair an experienced former Allegheny staffer, with a computer or GPS literate employee. Often, we had a 45-60 year old former lineman or engineer who knew everything about the electrical system and nothing at all about personal computers, DOS or GIS, working alongside a 20-25 year old computer literate recent Geography or Engineering graduate. This cross-training proved invaluable, and we continued it as a model for new hires throughout the project. The data collected during our practice sessions was translated and reviewed as a group. In retrospect, we should have spent a lot more time on this aspect. It would have been useful to demonstrate (repeatedly) how such things as a single bad attribute value (like Phasing) can cause the connectivity model to fail, or show what an erroneous GPS coordinate or Backspan could do to the map. Field Data Collector's Equipment Each of our field data collectors carried with them a sizable investment in computer gear and auxiliary equipment. The following is a list of required equipment. The approximate total cost of equipment for each collector was $4300.
Deployment & Disbursement After all were trained and properly equipped, they were disbursed to different parts of the state and neighboring states to adequately cover Allegheny’s service territory. Their parting instructions were; “work hard, have h, map lots of poles, and bring your friends”. Since we still needed at least twice the number of initial collectors, ftiends and colleagues were one of our best sources of new labor. Handling the daily data Each field collector created about 4 megabytes of raw data per collection day. They were armed with 8MB, 10MB, or 20MB flash storage disks depending on their distance from a local office. Since most circuits took 2-3 weeks to map, it was necessary for the collectors to download their data on a regular basis (there were also many other practical reasons not to leave the data on the storage cards any longer than necessary). Many processing tasks needed to be performed on the circuit data. It was an important management decision to determine which tasks would be most effectively performed in the field office versus the home GIS lab. Our GIS processors in the local field offices worked at lower salaries than the GIS staff in the home lab, and they had a lower set of skills and GIS experience. Some of the GIS processors in the local offices were also field collectors, auditors, and supervisors, so their time had to be handled carefully. Each processing step could have been performed completely in either place, however, after a month or so, a logical separation of tasks became evident. While a circuit was still being mapped, there were many tasks that could be performed in the local offices. At a minimum, the following tasks were done on each daily data file:
After circuits were completed and the individual daily files were appended together, the data was quality checked using many QA GIS scripts created by GeoResearch specific to Allegheny’s data model. Phasing, connectivity, and attribute values were all checked for completeness, accuracy, and logic. We looked hard for any types of problems with the spatial characteristics and data attributes of each circuit. If we saw a particular error happening more than a couple times, we would attempt to create a script to “flag” that error type on all future data files. Or, in a couple cases, we modified the collection data dictionary in the field PCs (although this was a less than desired method since it halted collection while the field computers were updated). QA/QC & Auditing To confirm that our collected circuit data was meeting the high accuracy specifications specified by Allegheny Power, a QA/QC plan was needed to ensure our results. Allegheny was mostly concerned with the accuracy of attributes such as Phasing, and Conductor Size since they have such an impact on the electrical properties of the model. After running the QA scripts described above, a plot of the circuit was made with a printout of the entire data table showing the attributes for each record in the circuit file. There were usually poles or sections of circuits which the GIS processor wanted checked or clarified by a field visit, so large scale maps with questions were prepared for each location needing confirmation. As a routine QA procedure agreed upon by GeoResearch and Allegheny Power, our supervisor field audited 5°/0 of each circuit to check all attributes of the poles and devices. In addition to the 5°/0, the supervisor would visit all those poles that the GIS processors had questions about. At the beginning of the project, and for the first circuits of new hires, we typically audited 10% or more until we were comfortable with the accuracy and completeness of the data. These audits required visiting poles a second time. This was a significant investment in manpower. In a typical local office area with 6 collectors mapping data in a 60-100 mile radius, the field auditing would require approximately 10-20 hours per week (much more in Morgantown). In addition to our audit of the poles, afier delivery of the data to Allegheny, an Allegheny staff member would audit another 5°/0of the poles to help ensure that we were meeting the accuracy requirements. After the field QA was completed, the data was sent to the home GIS lab via the Internet where a few more QA processes were performed before final delivery to Allegheny by Internet. Conclusion This project was completed “on-time”, though not without considerable effort, overtime and costs. As with most large-scale projects, many mistakes occurred, particularly in the first couple months, most of those mistakes resulted in lost productivity (poles per day), and we were constantly scrambling to adjust our staff numbers to keep pace with delivery schedules, and to replace staff lost to turnover. One of our largest mistakes from the beginning was overestimating the number of poles our collectors could map each day. The rugged terrain in % of our collection area prevented reaching our estimated productivity goals on most days. However, in the relatively flat, rural area of Mercersburg, PA, our collectors averaged well over the target on most days. In addition to our overestimating of daily productivity, it was difficult to overcome the productivity losses created from numerous equipment failures throughout the length of the project (especially at the beginning). Additionally, since we were constantly adding new staff, we almost always had some number of new collectors “in training” which required personnel resources that would otherwise have been utilized as supervisors, auditors, GIS processors, and collectors. Our thoughts are that this job could have been performed much more efficiently and profitably in 2-2 % years rather than in one year. It takes a few months after training to really develop a smooth and efficient working environment, and it takes time to absorb the significant start-up costs. The technology for this type of project is here, and it works --- very well. The key to success is good management starting from the estimating and proposal stage through recruiting, training, planning and production. | ||
|
|