Constantine N. Tonias, P. E., M.ASCE (1) and Elias C. Tonias, P. E., F.ASCE (2)

BRIDGING CIVIL ENGINEERING AND GIS

INTRODUCTION


For at least 3,000 years, people have prepared and used maps for a
variety of reasons and in a variety of forms.  The early Egyptians
prepared papyri with routes and instructions to guide their great
caravans in their middle eastern travels.  Herodotus prefaced his
history of the then world, and interspersed it with descriptive
paragraphs about its geography.  Since then, mariners prepared and used
maps to sail the seas; armies used maps to conquer lands, and pirates
to hide their treasures.  Economists are now using maps to help
marketers sell their products, sociologists to locate all the woes of
the world and for politicians to add thereto; and engineers, civil that
is, to make maps so that they may build cities, highways, bridges and
other public works projects accurately.

In developing and using maps, each of these users is interested in
having a precise map with respect to extent and content.  However,
accuracy and map content are not the same among them.  While an
economist may be content with positional accuracy of a square mile or
two and market projections of a gadget to the nearest $100,000 or more,
a politician's interests might be to the nearest household in a
political district map, and an engineer always needs to know the
location of every natural and man-made feature to the nearest, many a
time, 1/100th of a foot (3 mm).  As used herein, engineer refers to
those professionals whose work relates to the planning, design,
construction and management of public works and facilities, may they be
for a public or private owner.

Up until a dozen years or so, map preparation was a tedious exercise,
and map use was limited to but a few specialized occupations.  The
introduction, though, of what is now called geographic information
systems, or GIS for short, has revolutionized the making and
proliferated the use of maps.  The ability of the GIS technology to
merge maps graphics with the management of associate non-graphic data
has introduced a multitude of new applications ranging from the
management of large areas of lands to locating slot machines in a
gambling casino.  However, a variety of reasons, have forced the
engineer to stay apart of the other GIS users and take a different
path, that of CADD; computer aided design and drafting.  Various
attempts have been made to merge the GIS and CADD technologies, some of
which have been successful to a limited degree, but on the large, the
two communities of engineers and all others, have stayed apart and
rather weary of each other.  This is unfortunate since an engineer's
product is interwoven with the work of other professions, first, by the
engineer having a need for data available to these others, and second,
by the engineer's end product providing the foundation for, and
updating that same data used by the others. Therefore, the need for
integrating general GIS and engineering applications exists,
particularly when the life cycle operation of an engineered project is
to be considered in view of deteriorating infrastructure facilities and
shrinking funds and other resources.

Towards this end, the objective of this paper is twofold, (a) to
present a methodology for structuring the GIS database of a
municipality to accommodate both engineering specific and general GIS
functionality, and (b) to remind the engineer that until the time when
such a database is available, a GIS oriented database can still be of
use.


THE CADD-GIS DICHOTOMY


There are various reasons given by many engineers for the separate
paths taken by engineering and geographic information systems users,
some of which are easy to overcome, while others do require
considerable effort.  The most common reasons given are:


It has been argued that, although the above are indeed issues that need
to be addressed, they are not by themselves reasons for separation.
Perhaps the lack of understanding of each other's needs and
capabilities is the most important reason for this dichotomy.


THE CADD-GIS INTEGRATION

For several years now, Esri and The CEDRA Corporation have been
discussing this issue and ways of bringing together GIS and engineering
applications.  Although Esri's ArcInfo and CEDRA's The CEDRA System
have been able to exchange data in their native formats, it was felt
that a more integrated approach had to be taken.  The "avenue" that
emerged to accomplish this is Esri's ArcView and its scripting language
Avenue.  Thus, born out of the need to merge GIS and engineering
technologies, ArcView based CEDRA software, such as AVland, AVsand and
AVwater, were developed to enable engineers to work in their native
environment maintaining the accuracy of their choice, as well as the
integrity of their specific linestyles and symbols, and performing
their technical endeavors in surveying and engineering design of site,
land subdivision and site design, streets and highways, geotechnical
and other project types.  In addition, engineers are provided direct
access to the various geographic and related data present in an
ArcInfo GIS database, and the ability to exercise queries to augment
design related operations.

For example, the engineering department of a city may access the city's
geographic database using AVsand and retrieve relevant sewer
information to build a mathematical model of its sewer system.  If any
pertinent data (inverts, material, sizes, etc.) are not available in
the database, they can be introduced by the engineering personnel in
their accustomed way to update the database without burdening the GIS
or MIS department.  Similarly, land parcel, land use information, water
consumption and/or census data may be extracted from the overall
database to develop the sewer load model.  With these models available,
hydraulic analyses may be performed to identify sewer capacities, and
assess sewer performance.  Similar functionality is availed by AVwater
for water distribution networks.  With such integrated functionality,
base information can be accessed from the database, new survey and
design work may be performed to help construct a new project, and then,
upon project completion, update the database to be ready for the
facility's continued maintenance and operation.


THE MODELS OF A MUNICIPALITY

A municipality is an ever changing and evolving living being, the
management and control of which is very complex.  Over the past many
years, a variety of mathematical models have been developed to aide in
the management of specific municipal operations.  With the possible
exception of a few, the base or foundation of the majority of these
models is a geometric model, mostly two dimensional, representing the
geographic confines within which the city is located.  Therefore, in
order to minimize costs, avoid duplicity and facilitate service to all,
this common ingredient should be one and only one, serving all models,
including engineering.

Most often, interrelated geographic models are displayed pictorially as
a series of layers overlapping each other to denote the ability to mix
and match desired features of a common geography.  Shown in Figure 1 is
a slightly different approach as might be perceived from an engineering
point of view.  Before we continue, it is felt imperative to be stated
that this approach alters no model or database structures that exist or
might be developed.  It simply views the whole issue from another point
of view, and presents an avenue by which one common base map can
eventually serve both engineering and non-engineering functions.  The
operative word in the preceding sentence is "eventually", because for
many an existing community the ultimate model structure might be a bit
expensive to build from the get-go.

As per Figure 1, the terrain model represents "mother earth" upon which
the municipality exists.  It describes the ground relief, the natural
water bodies and vegetation, as well as the various buildings, dams,
levies and bridges and other features that alter the natural shape of
the earth.  Below it, organizationally speaking, are two associate
models, the parcel model and the streets and easements model.  The
parcel model represents the division of the municipality's terrain
extent into private and public ownership parcels, and is most often
referred to as the tax map, assessor's or cadastral model.  The streets
and easements model is comprised of the centerlines (or reference
lines) and names of the streets and public easements, and is a direct
by-product of the parcel model.  These three models represent the
foundation upon which everything else is built.  The remaining models
are grouped in tiers representing the various end user function.  Each
tier may contain more than one model.  The tier position in this figure
is of no importance or consequence.  Since we address engineering, we
place its group of models at the top of the heap, and below it that of
the planning department since engineering has more of a direct
relationship with planning than with the other departments of a
municipality.  Below planning are all other model groups such as those
that might be used by public safety (police and fire), health,
political, economic development, social services, and others.  It is
noted that within the context of this paper, there is no distinction
made between engineering and public works.

The management system of any public works facility requires certain
base information in the form of maps, drawings and other records.  This
information is comprised of four major types referred to as graphic
attributes, non-graphic attributes, pictorial information referred in
computer lingo as bit maps, and associate information.


  1. Graphic  attributes are those items of information that are
represented in the form of lines, arcs, points and associated
annotation (text) as required to produce drawings, maps and the like.
A map depicts the water distribution system with lines representing the
conduits, text identifying the size and type of conduit, and special
symbols denoting valves, hydrants, pump stations and other
appurtenances.  These are graphic elements, and their linework,
linestyle, symbology, color and so on are the associated graphic
attributes.


  2. Non-graphic attributes refer to that information that augments
that of the graphic attributes, and introduces a fourth dimension,
often referred to as attribute knowledge.  To manage this distribution
system, the managing agency must keep in file such other information as
the results of a hydraulic analysis, year of installation, pipe
diameter and material, date of last repair, model and manufacturer of
valves and pumps, and similar other data.  These are the non-graphic
attributes which could be divided into the following categories:

      

    • Inventory attributes are those that identify the physical extent and
composition of the model components such as pipe material, valves,
pumps, manholes, laterals, and the like as identified above.


    • Design attributes such as flow roughness coefficients (Manning's n
or Hazen-Williams c), open-close status, pump rating curves, and the
like generally relating to engineering analysis/design function.


    • Analytical attributes are the computational results of an analysis
or simulation that are associated with an inventory item such as flow
velocities, pressures, flow rates, and others.


    • Construction attributes may be considered an extension to the
inventory attributes since they further describe a model element. Such
attributes may be the bedding type of a conduit, the type of bench,
barrel, corbel and casting of a manhole, and others.

    

  3. Pictorials are graphic representations that can be accessed in
their natural form without any intelligence other than the association
to a specific non-graphic attribute most probably relating to a graphic
element.  For the maintenance of this system, it may be necessary to
access certain drawings that describe in detail the components of a
valve, pump or fire hydrant.  This is done through the display of a
drawing or photograph of the physical component for visual inspection
by the user.  Such data may be stored in the computer in bit format,
and may not be necessary for many computer models.


  4. Associate information is generally not associated directly with
specific graphic elements, but with rather model-wide global
functions.  Generally, this information constitutes individual
sub-models within the overall database.  For example, consider the
following:


    • Water supply demands as developed from area wide water consumption rates.


    • Wastewater contribution factors based on land use
codes of the land use model.


    • Population data as they may assist in water and
wastewater supply projections.


    • Rainfall gauging and groundwater monitoring data.


    • Rainfall intensity curves promulgated by the local
weather office.


    • Waste peaking factors by land use, or on a global basis.


    • Construction specifications item definition and
prevailing cost estimation data.


    • Historic data of inspections and maintenance operations.


    • Record of citizen requests (complaints) and response actions.


    • Permit application, processing and monitoring record data.


    • Various minimum design criteria and so on.

In designing the database, there is no hard and fast rule as to which
of these attributes are to be used and how they are to be organized.
The detail of the extent and organization is dependent upon the user's
operational procedures, system extent and complexity, desired degree of
sophistication, and budgetary constraints.  In any case, the design and
building of the database could be a time consuming and expensive
undertaking.  However, it can be shown that the long term benefits, if
a judicious approach is taken, will more than adequately compensate for
the initial costs.


THE BASE MAP MODELS

The top three models of Figure 1 constitute the base map of a
municipality.  Generally, the information required for these models is
available in a variety of sources, all within the same municipality.
For instance, the public surveyor and/or the planning office will have
topographic maps;  the tax assessor's office will have a set of
property maps with associated files of ownership and assessment data;
each utility department will have its own set of maps indicating the
respective system; and lastly the highway department will have its own
maps.  Quite often, these maps will carry and maintain identical
information but in duplicate form.  It is also not uncommon for the
sources for map generation to be different between the various agencies
or departments.  It is therefore necessary for all of these departments
to share one common and unified base map model upon which all
individual systems will be located.

Ownership and maintenance of the base map transcends into nontechnical
issues.  Based on years of involvement with public works projects, it
is the authors belief that the responsibility for developing and
maintaining this base map models should rest within the auspices of the
municipality's surveying office of its engineering department.

It was stated in the introduction that the content, extent and
precision of base map is debatable among the various users.  On this
issue, the authors position is that in a properly designed GIS
environment, the attainment of engineering base map requirements does
not impact adversely any other user.  The only issue of this is the
cost of attainment, which is addressed subsequently.  We will now
address the engineering requirements.

The terrain model is to contain all physical features that exist on
earth and which usually have a direct impact on an engineering project
or work.  These include:

	topography comprised of relief (contours)

	planimetric features (buildings, bridges, streets, railroads,
	culverts and other features)

	water bodies (lakes, ponds, creeks, reservoirs)

	vegetation (single trees, forests, hedges, etc.)

	geologic (soil cover/type, subsoil strata, rock, groundwater,
	etc.)

Ancillary to this model are analytical models such as slope analysis,
river flow, geotechnical and the like models that may be needed by one
or more engineering disciplines, as well as planning and other
departments. These models are considered as detailed extensions of the
other models which are described in this paper and thus omitted.

The parcel model is to be comprised of ownership property lines (public
and private), private and public easements lines, political boundaries,
and similar data that are usually available to tax maps.  Each parcel
should be identified with its centroid and a unique, permanent,
identification parcel number both acting as a reference between the
graphic and non-graphic attributes associated with this model.  The
associated attributes will generally vary from municipality to
municipality depending on record history, and governmental/legal
requirements.  Most probably these attributes will include parcel
address, owner name and address, tax number, assessment description and
assessed evaluation, and other relevant information.  Of these, the
parcel address, together with the parcel centroid and number, will be
used extensively in the maintenance of the various public works
facilities.  Although voter registration districts, police and fire
protection areas, water supply and sewerage districts, and others that
do not have a direct impact on engineering should be considered as an
extension of the parcel model since they are so directly associated
with the model.

The streets centerline model is in essence an extension or direct
by-product of the parcel model.  Its purpose is to relate the terrain,
parcel and the various public works models (see later on) to each other
via the street address system of the municipality.  A key constituent
of this model is the geometric location of the street centerline.  A
baseline or reference line could be substituted in easements of odd
shape.  Alleys, quite prevalent in older cities, may be treated as
streets or easements, or alternatively as a distinct feature.  For ease
of reference, they are all referred to as streets and centerlines.  The
graphic representation of these centerlines may be computed fairly much
automatically once the parcel map model has been established.  Their
graphic representation and non-graphic attributes should include:

	A code as to whether the line is a street, easement or alley.

	The street name and a unique identification street number.

	Ranges of street numbers from intersection to intersection.  A
	coordinated point at the start and end points, and at each
	curvature and change in linear direction.

	Engineering horizontal curvature data for each curvature along
	the street.  This could be expanded to also include a design
	vertical alignment, as well as reference(s) to typical street
	section templates.  Alternatively, this information could
	constitute a separate mathematical model.

	A coordinated point at each street intersection.  Optionally,
	the cross street identification number could also be included.
	Alternatively, street intersections could be computed on the
	fly by the end use software when needed.  This would save disk
	storage space; but, at the expense of compute time.

	A stationing system which may coincide with the federal or
	state mile post system.

	Designated traffic directions and classification as to type of
	street, arterial, major, minor, local, etc.

Additionally, this model may be expanded to incorporate all or some of
the attributes of the pavement model addressed subsequently.  In
working with the models of the various utilities of the municipality
and associated software, it will be found advantageous if a street
definition convention is established under which all street centerline
figures and stationing advance.  For example, following the federal
highway system, all streets would advance from south to north, and from
west to east.  Although not mandatory, a similar system could be
employed in the assignment of the street numbering system.


BASE MAP EXTENT AND ACCURACY

It was stated in the introduction that the base map extent and
positional accuracy are two issues of concern to engineers in the use
of a GIS database.  The extent was addressed in the preceding section.
In that, the terrain model appears not to be out of the ordinary as
used by most GIS applications.  Possible exceptions such as the ground
relief or geologic attributes, for which many a GIS application may not
have a use, can always be added at any time and used by select users.
The same also applies to the parcel model, but in the street centerline
model we begin to detect some deviations which can easily be overcome
provided the accuracy issue is adequately resolved.  In addressing this
issue, it is noted that base map accuracy is dependent upon two
elements, geographic position and parcel integrity.

Geographic position refers to the location of the various city features
as described by the base map models above, and the other models to be
discussed later on, upon the earth's surface under one unified spatial
coordinate system.  For this, the need exists for a sound geodetic
horizontal and vertical control program.  As a country matures and its
transportation and communication networks expand and advance, its
geodetic control also expands.  New major technologic advancements in
surveying instrumentation expedite such control with regards to time
and money involved.  In spite of these advancements, however, cost
still remains a major factor which many municipality's find difficult
to meet.

Parcel integrity refers to the discrepancy problems inherent to the
property information available in the form of property maps and deed
descriptions.  The identification of these discrepancies, many of which
may already be known, is relatively easy, although it could be time
consuming.  Their resolution, however, is difficult, potentially
requiring litigation, and therefore also time consuming and expensive.
These inherent property discrepancies could be attributed to a variety
of sources including, by not limited to, instrumentation, human error,
undetected property law violation, and others.  Regardless, though, of
the cause, it is safe to assume that the streets' rights-of-way will
not significantly change as a result of any property resolution.
Therefore, it becomes necessary to locate the streets as accurately as
feasible.   Concurrently, parcels can be located geometrically or by
digitization, depending on the best available information.

There is a great variation among municipalities as to the means for
attaining a base map sound enough for engineering.  There are those
that over the years have meticulously developed sound base map
information, and there are those that must start from scratch.  Each
municipality must stand on the merits of its own available information,
and take appropriate measures to develop the base map as needed.  Thus,
municipalities may be grouped into:


  1. Those that have a sound and accurate base map;     

  2. Those that have a sound database, but of questionable positional 
accuracy and/or geometric parcel integrity; and

  3. Those that need to begin from ground zero.

The municipalities of the first group are in a good position to
commence an engineering-GIS integration, if they have not done so
already, while those of the third group are in a position to properly
design and implement a phased integration plan that meets local finance
constraints.  Those of the second group may feel a bit of dilemma in
wanting to go all the way but may find it difficult to justify
additional expenditures.  For these municipalities, a gradual approach
may be considered.  Under this approach, the overall engineering
database such as suggested by this paper is put into effect, bearing in
mind that the accuracy is not as good as it should.  Then, as ground
control becomes available on an engineered project by project basis,
and property problems are resolved, the basemap is gradually updated.
Progress may also be enhanced by additional work as personnel resources
and/or funds become available.

It is important for the engineering department to realize that,
although an accurate base map database is most definitely desirable and
its pursuit should not be forgotten, it must not wait to take advantage
of the GIS potential until a mapping database is available.  Although
direct engineering street and site design cannot be had without
accurate mapping, there is a multitude of engineering functions,
particularly in the operation and maintenance of the constructed public
works facilities that could make excellent use of GIS applications.
This in itself would justify the building of the engineering database.
When properly designed, a database sharing the resources of the various
departments of a municipality could produce operational savings that
could be used to eventually create the base mapping to the desired
engineering accuracy.

Returning for a moment to the database extent, depending on the size
and complexity of the municipality, there could be several utility
models required by the engineering department, only a few of which have
been identified in Figure 1.  The potential models could include
pavements, structures (bridges and culverts), wastewater sewers,
stormwater sewers, open drainage facilities, street signage, street
lighting, water supply and distribution, other utilities, survey
monumentation control, and others.  Some of these are addressed below.


THE ENGINEERING MODELS


When referring to engineering models, there is a need to differentiate
between two types of models, inventory models and analytical models,
both of which share the same database.  Inventory models help the user
maintain an inventory of the facilities by accounting for each
component of a system.  Analytical models are those that help the user
simulate the operation or intended function of the utility.  For
example, a water distribution system is comprised of conduits and
associated appurtenances such as valves, pumps and the like.  Its
geometric configuration, material composition and geographic location
constitute the inventory model.  The simulation of the water flow
within the water system accounting for flow rates, velocities,
pressures, and similar functions are dependent on the mathematical
model that accounts for engineering principles.

The models needed by a municipality depend upon the responsibilities
delegated to it, and their physical composition.  A city located at the
confluence of two rivers, each with various tributaries, will need
river related models more so that a city in high ground with nary a
need for flood protection.  Similarly, the extent and complexity of the
models will be different between two cities, one with a population of
less than 100,000 and the other of more than a few million.  In the
latter example, although the database extent greatly differs, model
requirements may not vary greatly.  Presented below with a capsule
description are some of the models that could be needed.  It is noted
that certain of these models are an extension of the terrain model,
some of which may not fall under the administrative jurisdiction of the
engineering department, but which may have an impact upon engineering
operations from time to time.  In developing these models, provision
must be made for the incorporation of periodic inspections, and the
processing of the inspection observations to update the base model
database.


  1. The pavement model should be comprised of right-of-way widths,
pavement edges, composition of pavement, shoulder and sidewalk, type of
curb and/or gutter, traffic signalization appurtenances, and names.
traffic designation (arterials, primary routes, alleys, etc.) and
capacity, ownership (city, county, state, federal), and associated
maintenance data.  In addressing this model, it is noted that this
model could be:

	Integrated with the street centerlines model previously
	discussed,

	Expanded to either include or relate to an inventory of street
	signs, street lights, street landscaping (trees and
	beautification features), and traffic signalization
	appurtenances, and

	Associated to a traffic simulation model for the assessment of
	traffic capacity and signalization.

	Incorporated into an overall pavement management program (it is
	noted that the definition of pavement management ranges from a
	mere facilities inventory to a complex knowledge based
	management and maintenance system).



  2. The water distribution model could encompass the entire water
supply and distribution system for inventory purposes, but it should be
divided into smaller interrelated models comprising the primary supply
trunks, primary distribution mains and secondary service areas for flow
simulation purposes.  Due to its nature, the flow simulation of a large
system is very unyieldy and cumbersome to handle even with modern day
computing horsepower.  Each of these models should be comprised of
pipes, valves, hydrants, pump stations, reservoirs and other
appurtenances, each tagged with its material, size, roughness (c
factor), and manufacturer and model where applicable.  Separate
associated models could include the physical boundaries of various
demand areas much in a similar manner to those of the contribution
areas of the sewers models (see later on).


  3. The sewers model could be comprised of up to three distinct
systems depending on the sewerage character of the municipality.  Older
cities would most probably be comprised exclusively of combined sewers
with a moderate amount of separate wastewater and stormwater systems,
while newer municipalities will have only separate facilities.
Incorporated into this model(s), as far as The CEDRA System is
concerned, could also be all open drainage systems since provision for
their flow modeling is built in AVsand.  The use and composition of
this model is addressed separately in the subsequent section.


  4. The water courses model could include rivers and tributary
creeks, lakes, reservoirs, coastal lines, wetlands and similar
features, as well as the delineation of watersheds from an inventory
point of view.  The modeling and specific engineering analysis and
modeling of such elements is quite diverse and considered beyond the
scope of this paper.


  5. Other utility models such as gas mains, electric lines,
telephone lines, and cablevision lines are generally the responsibility
of private utility companies.  In spite of this, it is of interest to
the entire community, and all entities involved with public and private
utilities, if a common base map exists which serves all utilities
regardless of individual responsibility.


  6. Other base map models could be created to serve various needs
not only of the engineering department, but also those of other
departments and agencies.  Some of these models are:

	The geologic models relating to drainage, soil protection,
	water resources pollution control, geotechnical, agricultural
	and other application needs of the municipality.  Major
	constituents of these models are the top soil stratum and its
	cover, the subsurface soils and rock, groundwater, and
	vegetation.  Extensions or relations thereto could include
	environmental sensitive areas such as landfills, wetlands,
	historic sites, natural preserves and others.

	The land use model, generally maintained by the planning
	department, in addition to being of use in the planning and
	socio-political aspects of a municipality, may be used in the
	definition of load or demand models for sanitary sewer and
	water distribution systems since they generally pertain to the
	population or land use they serve.  Therefore it is possible to
	create a separate layer containing land use information and
	relate it to the various nodes of the respective system upon
	which they are to be applied.

	The water consumption model containing data by household or
	block areas could be used in the conduct of infiltration/inflow
	studies of sanitary and combined sewer systems, as well as load
	determination to verify land use based water demand and waste
	factors.

	The population census model can be used in conjunction with the
	previous models in the estimation of wastewater contribution
	and water consumption projections.



THE SEWER MODEL

Sewage collection and treatment systems represent one of the most
critical components of a municipality's infrastructure because of their
direct impact upon the health of the citizenry and the environment.
The major issues of concern for their proper performance are age, funds
and system management.  Like most public works constructed facilities,
the life expectancy of sewer systems can be quite lengthy.  Sewers that
have been properly designed, constructed and maintained, have been
found to function well for fifty and even many more years.  However,
due to various reasons, and primarily due to changes in loading
conditions and neglect, age has had a very deleterious effect upon most
old sewer systems.  The neglect of these systems is not the  result of
a "do not care" attitude, but rather of a continuous competition for
the limited supply of available funds with other governmental services
and responsibilities.

Sewer system management pertains to the ongoing paper work associated
with the daily operation of the sewerage system, and to its maintenance
operations.  Rather broadly, sewer system management may be divided
into two categories, flow management and capital system management.
The former pertains to the task of flow storage - conveyance -
treatment, while the latter addresses the task of system-user
interface, maintenance, administration, capital improvements, and the
like.  Fundamental to both is the proper recording and maintenance of
the overall system inventory, because it is upon this, that most all
decisions must be based.  For the purpose of this paper, we will
address the general aspects of the relevant database.  For more
specific reading on a system addressing the overall sewer management
subject, the reader is referred to Sewer Information Management and
Simulation System, 8th European Esri User's Conference 1993.

Figure 2 depicts a diagrammatic of the life cycle of a sewer system
from inception.  For most municipalities, the initial phase has long
past.  For them, the stage of initiating the inventory database of the
now existing sewer system varies dramatically from municipality to
municipality depending on the availability of records and degree of
prior computerization.  However, for new municipal sewer systems, the
establishment of a formulated inventory database at the onset of
development is considered to be a very advisable requirement.

Described below is a basic sewer model database, which assumes the
existence of the terrain, parcel and street centerlines models as
described previously.  This sewer model could be that of a wastewater,
stormwater and/or combined sewer system.  The stormwater model could
include closed conduits as well as open drainage ditches.  The basic
primary difference between these three types of models is in the
determination of their load contributions.  Although the discussion
below pertains to sewers, it can be extrapolated and made applicable to
a water supply and distribution system, as well as to other similar
utility systems.

The objective of the sewer model is to provide the engineering (DOE)
and wastewater (DOWW) maintenance staffs of a municipality the
necessary tools to:

The primary tool to attain this objective is AVsand.  The development
of the sewer model is to comprise three elements or sub-models, the
geometric model, the services model and the load model.  Of these, the
services model is optional, and its development may be delayed.
Depending on the size of the municipality, and the physical composition
of the sewer system(s), it is possible to stage the model development
in one of various methods.  For example, in separate waste and storm
sewer system, one could be developed first, then than the other.
Additionally, small sections of a sewer system could be developed to
completion individually, instead of completing an entire model all at
once.  It is also necessary to take into consideration the ultimate
intent of the model.  If the purpose of the model is strictly its
hydraulic modeling for sewer assessment, infiltration/inflow analyses,
and the like, it may be appropriate to create a model of only the major
sewer lines of, say, 18 inches (450 mm) and higher.  For the ultimate
purpose, though, of overall system maintenance and management, the
model may have to include all sewers and laterals.  With respect to
very large metropolitan municipalities, it may prove advantageous to
have two separate models, one for system management and one for
hydraulic analyses.

The geometric model of a sewer is comprised of its nodes and connecting
conduits.  Conduits are generally circular pipes; however, a multitude
of other configurations are available, as well as the ability for
custom design of closed and open shapes.  Depending on the type of
model, nodes can be manholes, inlets, pump stations, siphon intakes and
outlets, points at which a change in the conduit shape or alignment
occurs, or at points at which analytical or modeling results are
desirable.  Points at which laterals or services are connected to the
sewer could be treated as nodes; this practice, however, is discouraged
(see later on).  In this model, the conduit connecting two nodes is
referred to a reach.  A series or chain of reaches from a downstream to
an upstream end is referred to as a strip.  Strips may be defined in
any user desired order.

The development of the geometric model is perhaps the most tedious task
of the overall sewer model, and as said before, it can vary from
municipality to municipality depending on the source of the required
data to build the model.  Generally, most municipalities have a graphic
representation of the sewer system in some form of maps.  The
development time will be greatly simplified, if the entire sewer system
appears on a uniform, well matched set of map sheets.  With such maps
at hand, the geometric model could be developed by either of three
methods:

Of these approaches, the one to be opted should be based on an
evaluation of cost and logistics.  The logistics relate to the handling
of the documents for scanning purposes.  Since large scanners are not
portable, either groups of originals will have to be removed from the
DOE offices, scanned and returned.  This may impose a risk as well as
an imposition to the municipality.  High quality prints can be deemed
adequate for scanning purposes, but their production imposes an added
time and expense.

Once the horizontal layout of the sewer has been defined, its vertical
orientation, in the form of invert and top of grate elevations, as well
as, the conduit sizes, hydraulic properties and all other elements of
its inventory must be introduced in the model.  If all of this
information, or portion thereof is available in some sort of digital
file format, the model building with AVsand will be greatly expedited.
Otherwise, there is no alternative to the manual entry, either on a
batch process of a pre-design file format, or directly via AVsand.

The services model connects each household and business to a sewer
reach thereby relating each service to a parcel of the parcel model.
It is possible for a parcel to have more than one service connection
(lateral).  In GIS speak, this connection is made via the parcel's
unique  identification number and street address.  Although either will
suffice, both are used because generally maintenance personnel work
with addresses.  Parcels connected to sewers located in a back alley
are referenced to both the fronting street and to the back alley.

The preparation of this model could be accomplished by either of the
three methods above described.  Similarly as for the main sewer, if the
system of laterals is available in a digital file format, provisions
can be made for expediting the data  entry.  In many a municipality, a
lateral, in addition to a street address, is identified by its distance
from the first downstream street intersection.  A feature in AVsand
enables the mass entry of such information to automatically establish
the services model.

The load model helps the user define the wastewater contributions into
the sewer system.  AVsand provides for the definition of wastewater
contributions as the product of area, density per unit area, and flow
contribution per unit density.  By using unity (1) for either of these
product components, various load scenarios may be constructed.  For
this, the load model can be established in either, or all of the
methods below.


Of the above, the selected option should be based on the available
data, local engineering practices, and overall goal of the
municipality.


THE USER INTERFACE

One of the objectives in establishing a sewer model is the development
of the necessary tools to enhance the maintenance of the physical
system itself.  An engineering-GIS integration utilizing ArcView
capabilities enables DOE and DOWW personnel to easily access a database
of engineering and conventional GIS attributes, review its contents,
perform pre-programmed or on-the-fly queries as needed, and update the
database so as to always have current record data.  For example, a user
may specify a desired address and direct AVsand to:

Having responded to a citizen call, and taken the appropriate action,
the system could be accessed again to report the results of such action
and update the database.  Figure 3 depicts this as perceived by the
Division of Sewers of the City of Dearborn, Michigan.

Under future expansion of the engineering-GIS interface system, it is
possible to equip maintenance vehicles with AVsand on portable
computers, remotely connected to the database, to expedite service
calls, facility inspections, and maintenance operations.  Additionally,
a knowledge based system could be built into the system to recommend
remedial actions to problems encountered, plan and monitor maintenance
operations and help plan long term budgets for utility maintenance and
expansion as may be needed.

Shown in Figure 4 is a potential user interface displaying a portion of
a city with its sewer system, its sewer capacity database table, and a
query dialog box.  This query box may be used to display sewers of
inadequate capacity, excessive or non-cleansing velocities, dates of
last inspection or maintenance, and so on.  In Figure 5 a sewer
skeleton model, as may be used for sewer flow modeling, and its
geometric attribute table are displayed, while in Figure 6 a sewer
geometric model superimposed upon the services model and on the parcel
model, together with attribute tables containing sewer information as
may be desired by DOWW, are displayed.


THE HYDRAULIC ENGINEERING DATABASE

Another of the objectives of a sewer model is the flow assessment of
the sewer facility under various flow contribution loads with respect
to sewer capacity and backwater profiles, design sewer improvements,
and prepare construction drawings.  The Hydraulic Engineering Database
is comprised of three relations associated with the conduct of
hydraulic analyses and results thereof.  The relations and their
attributes are:


SUMMARY CLOSURE

The recent advancements in personal computer (PC) processing and high
end GIS, such as ArcView 2, is providing surveyors, engineers and
facilities managers, at affordable cost, desk-top information handling
capabilities that were previously available to large computing
systems.  The bridging of GIS and Engineering-CADD technologies, such
as CEDRA's AVland, AVsand and AVwater modules, enables both private and
municipal owners to better inventory, operate, maintain, plan, design
and budget for sewer, drainage, water, street and other facilities.

The merging  of GIS and Civil Engineering technologies, is a result of
developers looking down the road in an effort to provide the end-user a
truly unified database serving multiple disciplines.  The growing
popularity and use of GIS will inevitably result in the integration of
GIS and other technologies in the coming years.

The implementation of a computer aided comprehensive sewer management
system as that provided by AVsand can prove a very useful tool in the
understanding of the flow behavior of a sewer system by its technical
management staff, and in the maintenance of accurate and up-to-date
records for the maintenance of sewer systems.  Additions and
improvements to a sewer system can be analyzed in the planning stages,
and checked in the design stage to assure adequate service.

Additionally, once the overall system has been implemented and track
records of costs become available, the system can be expanded to
incorporate decision making techniques to assist the community in its
budgeting, planning and optimized use of its funds.

The database as necessary for the implementation of the sewer
management system is of use to the various departments of the overall
municipality, as well as to various public service and private
organizations.  Under a judicious shared resource cost sharing plan,
the initial implementation and the maintenance of the database can be
recovered quickly, and in the long run provide overall community wide
economic savings.