GISdevelopment.net ---> GITA 2002 ---> Geo Soluciones

How they do overseas?
An overview of some GIT Projects in Brazil


Roberto Falco
Sisgraph LTDA
Rua Estados Unidos, 112
Sao Paulo, Brazil – 01427-000


Abstract
Besides the globalization, that has the side effect to standardize techniques and procedures around the world, each national Utilities industry still has their particular requirements, caused mainly by the local regulatory environment, highly influenced by country requirements and/or regional business drivers. This paper will analyze this environment for Brazil, and describe some key differences in comparison with the US, that greatly impact GIT application development, and in some extent also core product development. This analysis will be developed using as example some projects that were implemented, or are undergoing, in Brazil.

Introduction
As a GIT professional involved in selling and implementing solutions for Brazilian utilities based on industry-leading GIS packages, the author have been faced in the last six years with the challenge to adapt product literature, marketing material and even data model templates originally developed for the US market to the local market.

A set of key differences was identified during this period of time, in the areas of application functionality, data modeling, GIT core product characteristics, and interfaces requirements. Most of these differences can be connected to a same root, the historical drivers behind a regulatory environment evolution, which had opposite directions in both regions.

This paper is a summary of the author’s experience in this area, and will try to trace how the regulatory commandments drove the evolution of GIT for certain application areas in the US market, and compare with the opposite path that occurred in Brazil. This will be used to explain why there are in Brazil more tight requirements for applications like Outage Management and Long Term Planning, for example, when in the other hand Work Management is almost an unknown application for the IT staffs of most Brazilian utilities.

Regulatory Environment Comparisonbn

US Case

Since the early start of the Utilities industry in all countries there were a strong intervention of the government authority, in the form of a investor, owning companies, or as a regulatory entity (Krause, 1995). In the US, this intervention was developed over a strong legal apparatus put in place to support the FERC activity.

In the 80’s, when the GIT industry was getting it’s first major grow, the FERC main responsible was in the area of economic regulation, setting energy rates between the various industry players and to the final electricity consumer.

That rate regulation applied there was based on cost-of-service criteria. This criteria has the objective to keep the lowest consumer rate as possible, while guarantying the Return- Of-Investment (ROI) in an agreed level with the capitalist investor (Krause, 1995; Capeletto, 2001).

The implementation of this rate regulation criteria is highly dependent on detailed, high quality information about Utility assets value and depreciation, amount of investments, maintenance costs and operational costs, at least. That “technical” information, together with information about level of energy sales, number of customers and other costs, are the basis for the accounting practices that support the rate evaluation.

The requirement to generate that “technical” information and control the business processes that deal with that data was the key driver for GIT application development in the late 80’s and early 90’s, as will be showed forward in this paper.

The rate regulation criteria based on cost-of-service is known to favor the overinvestment and the inefficiency of the system overall, named as the Averch-Johnson effect (Capeletto, 2001; Averch & Johnson, 1963). To counter that were developed models based on price caps fixed for a period of time, and adjusted in between by a factor that includes the inflation minus a productivity gain correction, mainly adopted in Europe.

A variation of this approach is called the Performance Based Regulation, which has been in discussion by various states in the US. In this case, the capacity of the Utility to freely practice rates is conditioned to its ability to met regulator established performance goals. These goals covers various areas of the Utility activity and its customer relationship, but ones that are usually included and affects GIT applications directly are the service availability and continuity measurements, in the form of SAIDI and SAIFI numbers (or equivalents). This regulatory control over service quality aspects is relatively new, only recently being considered by GIT technology providers as an industry driver. As a matter of fact, until January 1998 the IEEE standards for these measurements (IEEE P1366/D18, 1998) were yet in a draft format.

Brazilian Case

The history of the government intervention in the Brazilian market is fairly different from the US one. Even if the Utility industry has its beginning through the creation of private owned companies (British and Canadian investments mainly), the major part of the electric utilities where owned by state governments during the 70’s and 80’s. These companies were vertical organizations covering Generation, Transmission and Distribution in the whole state geographic area. This environment, associated with the absurdly high inflation rates that were observed in the country during this period (over 100% per year), caused some distortions on the rate regulation, as follows:
  • The electricity rate, as other public service rates, were widely used as tools to control inflation, since the government as the unique controller was able to postpone or negate rises needed to face cost elevation.
  • The potential cash flow problems arisen them were covered by capital injections coming from the government treasure.
  • Also, cross subsidies between the business in the vertical organization, and across markets (commercial, industrial and residential) were fairly common, hiding the real cost structure of each step of the supply chain.


As a consequence there were never tight regulatory controls over O&M costs, construction investments and network assets like there is in the US, and as a consequence not the need to develop IT support systems for these business processes (typically Work Management systems). As a matter of fact, the overall industry regulation until the 90’s was very loose, covering only a few aspects of the society/utility relationship. A department of the Energy Ministry played the regulatory role, instead of an autonomous entity like FERC. The Brazilian Electricity Regulation Agency (ANEEL) was created only in 1996, as part of the sector restructuring and privatization.

Even being a loose regulatory environment, the area of service quality received a special attention in Brazil since the 70’s, as opposed to what happened in the US. This was caused by a series of factors, including:
  • The characteristics of the network in Brazil, with a predominance of overhead networks using naked conductors and long secondary networks, in opposite with underground networks in the US. This is a much less reliable configuration, so service continuity was always an issue. Underground networks are much more reliable then overhead ones, but also more expensive. Underground networks only justify economically in high load density areas. For reference, the average customer consumption in Brazil is ten times lower than in the US. A highly urbanized area like Belo Horizonte, capital of the Minas Gerais State, shows up maximum load densities of only 23 MVA/km2.
  • The disposition of the regulator to provide customers with a “compensation” for not being able to transparently regulates rates. After the electricity price, the more sensitive aspect of the electricity service for the end user is the continuity.
There’s an established procedure to calculate service reliability indexes in Brazil since 1978. This regulation (Portaria DNAE 46/78, 1978) dictates data collection procedures, calculation methodologies, information disclosure rules, and, last and most important, service level goals to be met by the utilities, associated with financial penalties for those goals not achieved. The electric utilities where so faced with the need to automate these processes, what lead for an early development of Trouble Analysis and Outage Management systems.

Recently, these rules were revised and republished (Portaria ANEEL 024/00, 2000), creating new measurements to include not only system average indexes (DEC and FEC, very similar to SAIDI and SAIFI), but also individual service continuity indexes. The individual ones measure mean duration, frequency and maximum duration of outages (DIC, FIC and DMIC, respectively). The utility must calculate these individual indexes for each customer, and in a short run will need to publish it into the electricity bill, and provide automatic financial compensation in the following month bill to customers that had its limits violated.

Summary of Regulatory Distinctions

To summarize the differences that were appointed between the US and Brazilian regulatory environment since the 80’s, we have:
  • The economic regulation was the focus in US, creating the need for high quality and detailed information about O&M costs, construction investments and network assets, to be provided to the regulator. Today, de-regulation and Performance Based regulations are creating the need for data about service quality, mainly service continuity.
  • The service quality regulation was the focus on Brazil, what forced the utilities to develop tools to manage outage data in a detailed way. Today, this level of detail includes individual customer information. The requirement over O&M costs, construction investments and network asset data was always loose, as it’s still today.
Business Drivers And Application Requirements

From the previous analysis, we see an obvious distinction between the main business drivers of the industry in both countries during the last decade, affecting directly GIT application requirements.

As we can see from the GITA Special Achievements Awards to Illinois Power (1996) and Texas Utilities (1997), the hot issue of the GIT industry in the US at the beginning of the 90’s was the integration between AM/FM applications and Work Management tools. This integration was focused on streamlining the construction and maintenance processes, allowing for more complete, accurate and accessible data over construction investments, network assets and O&M costs.

In Brazil the GIT industry was just starting at that time, and facing other issues. The main requirements were associated with outage management, due the regulatory pressure, and engineering tools. This second driver, although not related with the regulatory environment, was common to most RFP and public bids published so, and related with another characteristic of the electric utility industry in Brazil.

The country went through a long economic crisis during the 80’s and 90’s, with inflation and high level of government debt. The investment capacity of the state-owned electric utilities was very low, when compared with the existing expansion needs. The penetration index of the electricity service was low when compared with developed countries, mostly in rural areas, and even in poor urban areas. Although, the demand growth rates in some big cities, highly urbanized, was as high as 10% to 15% a year.

This environment lead to a need to optimize investments as much as possible, as opposite to the overbuild tradition of most US electric utilities. To the GIT providers, this meant integration with engineering tools to support design and development of long term planning tools and functionality.

In a visit from a Brazilian electric utility to Texas Utilities in 1998, the differences between the business drivers’ priorities in both countries become clear. They decided to visit TU because its award-winning system is based on the same GIS platform that this company used to develop a Long-Term Planning tool. But their main interest was over engineering functions to support the design process, that in fact were absent of TU implementation, instead of the estimating and work order life-cycle management support that was the main feature of the system.

Another curiosity about this visit was that, after going through the scheduled program, the Brazilian manager asked about the possibility to visit a Distribution Operations Center and know what type of system was used there. At that time, the visited DOC (at Forth Worth area) had only printed paper maps of the feeders it was responsible for, and the trouble analysis was done manually, over paper trouble tickets printed at the DOC room. The Brazilian company already had at that time an automated trouble analysis system implemented since 1995, integrated with the CIS system, an IVR unit and the facilities database, this last one running at a mainframe based system.

The evolution of the Brazilian electric utility industry on the last few years, with privatizations and restructuring, did not change the scenario in terms of GIT application needs. The new sector model is focusing competition, through the separation between Generation, Transmission, Distribution and Commercialization business, creation of an independent Regulatory Agency (ANNEL), establishment of a spot energy market, and creation of the status of “Free Customer”, initially for high demand/high voltage customers, but to be expanded to low voltage costumers until 2005.

The end-user customer rates are based in a price-cap model, subjected to a five-year revision where a factor of productivity will be taken into account to reduce rates in favor of the customer. The service quality regulation become more rigid and detailed, and the service continuity indexes limits have been reduced from year to year, forcing the utilities to improve quality continually.

The recent GIT projects undergone in Brazil show how these drivers affect application requirements. The typical list includes:
  • Built-in engineering aids to support network design, focusing system optimization
  • Integration with engineering tools to support long-term planning, demand forecast, feeder topology optimization, capacitor placement, substation planning, etc.
  • Automation of outage management related processes like trouble call, trouble analysis, emergency field crew dispatch and management, outage data collection, and regulatory report preparation.
Project Descriptions

The project descriptions provided hereafter are non-official, and based on the author’s involvement with each one, and personal discretion. For confidentially purposes, the companies will be referred by fictitious names, and details that could lead to its identification, like region served for instance, will be omitted.

Company A is one of the five biggest utilities in Brazil, serving a few million customers, and also a GIT pioneer in the Country. Their initiatives goes back to the early 80’s, when they tied their paper maps with their mainframe based facilities database through the addition of UTM X and Y coordinates as attributes to the feeders conductor segments description.

In the late 80’s they started researching the application of graphic technology to support distribution management processes like network design and planning. During the early 90’s, they accelerated these studies, and attended AM/FM International Conferences before doing the technology selection. They purchased an industry-leading GIS package in the end of 1994, and started an internal development project, supported by some external consulting service, with the goals of:
  • Build a long term planning application able to do statistical demand forecast, feeder load flow calculations, simulate primary network (medium voltage) changes and improvements, and consolidate the feeder and substation construction plan for a 5 years forecast.
  • Build a geospatial facilities database, integrated with a network design tool capable of doing primary and secondary network electrical calculation and optimization. This tool and database would be the heart of future GIT initiatives, and replace the mainframe based-UTM referenced facilities database used so.
The first phase, the long term planning tool, was accomplished between 1995 and 1996, and is in production since then. The second phase was never accomplished completely, at least as originally planned.

Company A started in parallel another project to develop its own core geospatial functionality, in the form of a very customized and proprietary data model and a library to be used with low-level programming languages. After a lot discussion, the product of this effort was elected as the main platform for its GIT initiatives in 1998.

The new platform was used to replace the facilities database with a geospatial database during 1999 and 2000, including the improvement of data, support to landbase maintenance and some electric analysis capabilities (load curve analysis, voltage drop calculations, etc.). The support to the design process was postponed, mostly due difficulties to add long term transaction support to the core library.

Company A also had a trouble analysis system in place, that used the connected model of the mainframe facilities database to analyze incoming trouble calls and generate emergency crew services to probable outage devices. Using the proprietary core libraries, this system was replaced by a GIT based outage management and dispatch system, integrated with the Call Center, the IVR unit and a satellite AVL (Automated Vehicle Location) and Mobile Data Communications system, for crew tracking and data communication. This system is implemented through the companies Distribution Operation Centers, and generates all information needed for regulatory outage reporting.

Company B Company B is a medium size electric utility that was formed by the split of the former statewide company. As a young company, it started building its IT infrastructure almost from scratch, and decided to invest in GIT to support distribution management processes. They selected an industry-leading platform and an implementation partner to built a system that covers the following functionality:
  • Construction of a geospatial facilities database and connected model of the primary network, from the substation to the distribution transformer, integrated with database maintenance tools, Intranet visualization tools, and electrical analysis packages.
  • Implementation of an outage management and workforce management system to automate the Distribution Operations Center (DOC), including the calculation and reporting of the regulatory service continuity indexes. Integration with the CIS and a satellite AVL and Mobile Data Communication system.
The first phase, the construction of the geospatial facilities database and data maintenance tools was accomplished in a record timeframe of six months, with the aid of a template data and application model, that was adapted after a gap analysis. This was crucial to the success of the project, since was done as part of the Y2K bug actions because the legacy facilities database was not compatible. The integration with electrical analysis packages was done along the year 2000, together with the Intranet viewing and query application deployment.

The second phase was done in 2001, being the more critical in terms of gaps between the local requirements and the resources offered by the choose COTS application, what clearly reflects the distinction between the US and Brazilian regulatory requirements in the service quality area. The major gaps were found in the following functional areas:
  • Field capture of detailed outage data to support reporting;
  • Data consistency over event closure;
  • Trouble analysis algorithm
Company C

Company C is one of the five big electric utilities in the country, serving a few million customers also. Even being recognized by its leadership in various technologic areas, the company does not have any GIT application implemented till now. The company started to evaluate the use of geospatial technology with consultants from one of the “Big Fives” consulting firms in 1999. During 2000, they contracted a recognized GIT consulting firm from the US to help them build their RFP. Although the selection process was not yet finished when this paper was wrote, some very interesting aspects of it must be highlighted:

  • The recognition of the validity of the investment in GIT tools, based on a very favorable ROI analysis done taking into account the Brazilian reality, lead to a “stateof- the-art” system design, done together with the US consultant.
  • The system design includes all the pieces of an integrated Distribution Management System as offered by leading vendors. It was composed by a geospatial facilities database, a design tool, an outage management piece, and a dispatch and workforce management piece capable to integrate dispatch for various service types (emergency, construction, maintenance, etc.) and also able to support mobile GIT tools.
  • Interfaces where designed to the CIS and also to modules on a SAP R/3 system.
  • The system design also emphasize some typical Brazilian requirements, as a distribution planning tool capable to support long term feeder and substation planning.
  • That was one of the first times in Brazil that emphasis was put on typical work management processes support. The RFP specified an integrated job estimation workflow to be supported by the geospatial design toll and the SAP R/3 system. The question is how much the US consultant influenced this, at least in the format it was designed to work.
Conclusion

The experience shows us that even if at a first sight the procedures and processes of the electric utilities are the same around the world, there are important details, great part of them influenced by the regulatory environment, that causes significant distinctions in GIT application requirements between countries.

Specifically in the Brazilian case, the main distinctions are in the area of outage management and emergency service dispatch, since the regulatory requirements in this area are much more developed in Brazil than in the US, were the industry-leading products are developed and have their major marketplace.

References
  • Averch, H. & Johnson, L. L., 1963, “Behavior of the Firm under Regulatory Constrain”: The American Economic Review.
  • Capeletto, G. J., 2001, “Aspectos Relevantes na Qualidade dos Serviços de Energia Elétrica: Casos Europeu e Brasileiro”: Proceedings of the Brazilian Regulation Congress, paper 062.
  • Krause, G. G., 1995, “Mercado Spot e Energia Elétrica: Uma Commodity Como Outra Qualquer ?”: Revista Brasileira de Energia, Volume 4, Number 1
© GISdevelopment.net. All rights reserved.