Home Articles Data specification for utility GIS and corresponding cost benefits in the year...

Data specification for utility GIS and corresponding cost benefits in the year 2002

Saravanan Balakrishnan
Consultants India, Chennai, India
Phone/Fax: +91-44-4616861
[email protected]

Utility companies must consider the following to arrive at the appropriate technical specifications for their GIS data and the corresponding implementation schedule:

  • Impact of technology on the cost of data acquisition: All the technologies required for acquiring data of appropriate quality are already in place. Technology is unlikely to make any major contribution to reduce the cost of data acquisition in the foreseeable future. This will be more true for compilation of large scale maps (which we will call “Fine Mapping Product” in rest of this document).
  • Impact of maturity of GIS data vending industry: In the next few years, maturity of the data vending industry will drive down the price of GIS data and simultaneously increase the data quality
  • Impact of better and cheaper GIS software: Some of the upcoming GIS software are much more elegantly built and are so much cheaper. Hence, GIS computing will become much more commonplace and GIS database can afford to be more elaborate.
  • Cost of upgrading the landbase at a future date: Utility companies which presently start with less sophisticated land base are likely to face high upgrade and migration cost in the future.
  • When the above are taken into consideration, it shows that the optimal specification for the Geographical Information System and their implementation plan must be different from what is prevailing in the Indian industry.

The discussions of this paper have two application limitations:

  • The paper discusses the Utility GIS in the context of city based utility services only. Content of this paper will require substantial modifications to accommodate cross-country utility infrastructure like high tension power distribution
  • The paper is tailor made for developing countries like India. In most of the developed countries, GIS data of high quality are readily available at a fraction of the data acquisition cost.

Components of GIS Data
All GIS data are made of two components:

  • Geometry of geographical features like locality boundary, street boundary, plot boundary, building features, etc. These type of data are called the “geometric data”. An illustration showing a typical geometric data for an Utility GIs is enclosed in the illustration overleaf.
  • Information associated with the geometric data like locality name, street name, nature of occupation of a plot, elevations, etc. We will call this the “attribute data”

Following are the typical “Geometric Data” are generally required in an Utility GIS

Following are the typical “Geometric Data” are generally required in an Utility GIS:

  • Right of Way Edge
  • Average Road Width
  • Right of Way Center
  • Carriage Way Edge
  • Carriage Way Center
  • Median, Traffic Island
  • Property Boundaries
  • Entrances to Properties
  • Power and Telecom Poles, Pylons, Junction Boxes, Transformers, etc.
  • Overhead Power and Telecom Cables
  • Buried Power and Telecom Cables
  • Manholes
  • Underground sewerage and water supply lines
  • Locality, Sub-Locality and PIN Code Boundaries

The geometric data is usually acquired with one of the following two accuracies:

  Positional Accuracy Geometric Tolerance
Coarse Mapping
Fine Mapping
10m (using hand held DGPS or IKONOS)
2m (using Geodetic GPS and ETS)
1-2m (using Rodometer and Tape)
Sub-decimeter (using ETS)

Due to the inherent nature of the technologies deployed,

  • “Fine Mapping” is much more capital intensive than “Coarse Mapping”.
  • “Fine Mapping” is more labor intensive than “Coarse Mapping”.
  • Hence, “Fine Mapping” is usually two to three times as expensive as “Coarse Mapping”.

World over, it is the general practice to use “Coarse Mapping” specifications at the planning stage.

Whereas, India seem to stand divided in usage of “Fine Mapping” specifications. Even though labor cost is much higher in developed countries, which should make the “Fine” products much more expensive there , developed countries seem to go for the “Fine” product for utility maintenance and operations. However, in India, where utilities are just beginning to adopt to GIS, even though the “Fine” product is “only” about two times as expensive as “Coarse” product, there is a tendency to use the “Coarse” product for operations and maintenance as well.

Addressing this anomaly adequately will create/save economic value running into tens of crores of Rupees in the country within the next few of years.
Criteria for an Optimal Data Model
For arriving at an Optimal Data Model, we will be required to consider the following:

  • Cost of Geometric Data, as varying over time
  • Cost of GIS software and hardware of different capabilities
  • Benefits derived from different Systems
  • Cost of future upgrade, if any

Cost of Geometric Data
The prevailing market prices for the “Fine Geometric Data” is almost twice as expensive as the “Coarse Geometric Product”

  • Within the next few years, no new technology is expected to reduce the present cost of acquiring “Fine mapping” data products
  • In the third/fourth years from now, certain evolving technologies are expected to reduce the “Coarse Mapping” data cost by some 20%.
  • Presently, many of the data products available in the market are of suspicious quality. If the buyers become more quality aware, the market prices are likely to harden. While the quality awareness of the buyer is not likely to change drastically in the immediate future, it might happen in about two/three years time, after the buyer has acquired a substantial experience in using the data products.
  • As the market matures, data vendors will begin to distribute the cost of compiling their base data to more than one customers. This will have a large contribution in decreasing the market price of the data products. This is likely to reduce the cost of the map product by some 40% after two/three years

Cost of GIS Software and Hardware
Desk Top GIS is just coming of age the world over. However, almost all the utility GIS installations in the country are already desk top GIS. Desk top GIS today has the following four disadvantages:

  • Desk Top machines are not yet adequately equipped for handling complex real-world GIS. Users often notice crippling performance deficiency with Desk Top GIS. However, this will change in the coming years as the Desk Top hardware and software become more capable. As per Moore’s Law, our Desk Top machines will be at least 10 times more capable within the next 5 years. That kind of capability will be good enough for handling all perceivable problems with Utility GIS

  • GIS application products are presently priced at ridiculously high levels. Some of the developments that are happening in the industry will reduce the cost of GIS software by a factor of 10 with the next two/three years.
  • Utility GIS applications involve development of a wide variety of custom solutions, which are presently being offered by specialized international vendors. However, some of the Indian companies, with their low cost base, have already started entering this space. These companies will be in a position to offer good quality custom solutions at one third or one fourth price in about the next three/four years time. This will allow the utilities to dramatically extend their GIS capabilities.
  • In about three year’s time, real-time decimeter GPS will be available for about one fifth the today’s cost. This, when combined with all the factors indicated above, will modernize the cable/infrastructure maintenance way beyond today’s capabilities.

Benefits from GIS Systems built with Coarse and Fine Mapping Products
The advantage of the “Coarse” geometric data product is, evidently, that this product is cheaper than the “Fine” data products. Other than this, it usually takes less time for get the “Coarse” mapping done, though this doesn’t seem to matter in Indian scenario.

Whereas all that can be done with “Coarse” product can also be done with “Fine” product, the “Fine Product” has a few additional capabilities:

  • Last Mile Planning: With the kind of accuracy available with “Fine” product, the last mile planning, including customer connectivity, is very accurate. The planned quantities and the theoretically most optimal quantities can hardly be more different than 1%. Traditionally, this figure is seldom lower than 3%.
  • Pilferage Control: As the GIS data model is very close to the actual scenario on the ground, pilferages and misappropriations of network infrastructure material are totally eliminated. This could save between 1 and 2% of the network cost.
  • Repairs and Maintenance: With the help of high precision map product, coupled with decimeter GPS, repair and maintenance works can be automated to a large extent and the cost drastically reduced. This can save up to 70% of the cost of the repairs and maintenance of the network.
  • Resale of Map Data: Often, Indian utilities are found to be footing the full cost of data acquisition. With the high precision “Fine” data product, utilities can recoup more than their investment by selling the map data even across industry. This is extremely difficult with “Coarse” mapping products as many GIS users will eventually find it inadequate.

Cost of a Future Upgrade
Unlike upgrading the GIS Hardware or Software, upgrading the underlying data will be an expensive proposition because the utilities will also be spending resources towards (1) purchasing a high resolution base map, (2) mapping mostly buried utilities to higher precision and (3) matching the old GIS data with the new geometric data.

While this additional investment will certainly increase the quality of the GIS, the inertia, coupled with not having a pressing need, will lead to delays in such upgrade projects, resulting in opportunity loss.

The Decision Making
The chart on the right hand side shows the cost of two different Geographical Information Systems for an area of 100sqkm, one implemented using “coarse” geometric data and the other implemented using “fine” geometric data.

At the indicated prices, the GIS installed with “Fine” map data has provisions for investing in decimeter GPS, comprehensive attribute data collection, extended custom GIS solutions, mobile GIS (hardware and software) for maintenance crew, etc.

Whereas, it is technically inappropriate to add these additional facilities to “Coarse” map data. Hence, the “Coarse Product” prices indicated in this chart does not provide for any of these facilities.

Cost Benefit Analysis
Let us consider two utility companies, installing a new GIS infrastructure for an extent measuring some 100 Sqkm:

  • The first utility, “Utility C”, let us say, is basing their GIS on “Coarse” product.
  • The second utility, “Utility F”, let us say, is basing their GIS on “Fine” data product

Let us assume that each of these utilities suffer a cost of finance of 14%.

The “Utility C” will be investing Rs 75 Lk on the new GIS and “Utility F” will be investing Rs 160 Lk.

By restricting the investment level to Rs 75 Lk on a GIS database with “Coarse” product, the “Utility C” will have the advantages of

  • Conserving the scarce financial resources
  • Opportunity to upgrade to “Fine” product at a later date, may be in about four years from now, after the product becomes much cheaper than today.
  • Opportunity to gradually include additional features like GPS, mobile GIS, etc., after these technologies mature and become cheaper than they are today

Whereas by Investing Rs 160 Lk, “Utility F” will get:

  • A high resolution base map with resale value (with a potential to earn the entire investment)
    Initial Investment

    Cost of GIS Facility 160
    Saving in Network -60
    Net Cost 100

    Financials after 2 Year

    Cost carried over 34
    Add 14% cost 5
    Saving in Maint. -80
    Net Cost -41

    Financials after 4 Year

    Cost carried over -127
    Add 14% cost -18
    Saving in Maint. -80
    Net Cost -225
    Financials after 1 Year

    Cost carried over 100
    Add 14% cost 14
    Saving in Maint. -80
    Net Cost 34

    Financials after 3 Year

    Cost carried over -41
    Add 14% cost -6
    Saving in Maint. -80
    Net Cost -127

    Financials after 5 Year

    Cost carried over -225
    Add 14% cost -31
    Saving in Maint. -80
    Net Cost -336
  • A set of decimeter GPS and mobile GIS hardware and software so the fault and maintenance works can be fully automated from the day one.
  • Better last mile planning and better cost control of the project execution.

However, on the first year of investment, “Utility F” will achieve a saving of 2% of the cost of last mile infrastructure as against “Utility C”, computed at of 2% of Rs 30 Cr , being Rs 0.6 Cr. Also, every year from the date of implementation of the GIS, the “Utility F” will also save 40% on the cost of maintenance of the network as against a non-GIS Utility, computed at 40% of Rs 2 Cr per year, being Rs 0.8 Cr.

Initial Investment

Cost of GIS Facility 75

Financials after 2 Year

Net Investment 46
Add 14% cost 6
Saving in Maint. -40
Net Cost 12

Financials after 4 Year

Net Investment -26
Add 14% cost -4
Saving in Maint. -40
Net Cost -108
Financials after 1 Year

Investment 75
Add 14% cost 14
Saving in Maint. -40
Net Cost 46

Financials after 3 Year

Net Investment 12
Add 14% cost 2
Saving in Maint. -40
Net Cost -126

Financials after 5 Year

Cost carried over 38
Add 14% cost 5
Saving in Maint. -30
Net Cost -37

With the cost of funds pegged at 14%, following will be the cost/benefit achieved by “Utility F” over the period of five years:

Whereas, “Utility C” will be spending much less during the first year, yet get a 20% saving on maintenance cost (i.e. Rs 40 Lk).

However, growing needs and technological developments will demand that this company ports its GIS to “Fine” product by the fourth year. This upgrade, as per the discussions we have had earlier, is likely to be in the order of Rs 108 Lk.

The graph below shows the investments and paybacks of “Fine” and “Course” products.

The following conclusions can be drawn from the analysis above:

  • While the “Fine Mapping” is twice as expensive as the “Coarse Mapping”, it pays for the entire investment in a matter of two years and brings in a 2X profit (after fully amortizing the investment) before 5 years of operation.
  • This may be reason why utilities operating in countries with abundant resources go for “Fine” product, even though the initial investment is much higher.
  • Other than this, if we also consider the commercial benefits that “Utility F” will enjoy due to having a superior GIS (and hence better customer relationship), the balance is further tilted towards building a GIS with the “Fine” product