Return to Users Guide Return to Catalog

ESTIMATING COSTS FOR WASTEWATER COLLECTION AND TREATMENT

UNDER VARIOUS GROWTH SCENARIOS

James Reilly
New Jersey Office of State Planning, USA

Paul Gottlieb
Princeton University, USA

Abstract

The New Jersey Office of State Planning (OSP) has developed a spreadsheet-based
computer model that allocates municipal growth projections, estimates future incomes and
housing needs for municipalities, and then evaluates the impacts of this future growth on several
infrastructure systems. This paper describes the sewer impact portion of the model and presents
sample results.
The sewer model is based on detailed information about local conditions principally
gleaned from the Annual EPA Needs survey (A copy of the questionnaire is attached), which in
New Jersey must be completed by every operator of a licensed sewer system. OSP also collected
sewer service area maps and obtained information about water quality conditions for all streams
into which treated sewerage is discharged. If the model determined the need for a new treatment
plant, the cost for the new plant reflects both treatment capacity and the proper level of
treatment to ensure compliance with the Clean Water Act. OSP also developed several
interesting subroutines including one that calculates the likelihood for nonsewered areas to
install sewers and another one that estimates future collector costs as a function of development
density and pipe size. (Yes, it's another cost of sprawl data set!)
Results from the model have produced interesting policy discussions. Because most
municipalities are connected to regional treatment facilities, growth elsewhere in the same sewer
shed can increase cost in slower growing municipalities, or inhibit growth due to declines in
treatment capacity. In the older urbanized areas, local officials are concerned that the high costs
to rehabilitate existing facilities to meet Clean Water standards could inhibit revitalization
efforts.

Introduction

Since 1988 the New Jersey Office of State Planning (OSP) has been developing computer models
to estimate both future demographic and economic conditions and various trend and regional
planning scenarios. This paper describes that portion of the OSP computer modeling designed to
estimate the public sector capital costs associated the construction of wastewater collection and
treatment plants needed to service future populations and jobs.
The model described in this paper is a modified version of the third model constructed
for OSP's use. The initial sewer cost model was developed in the Spring of 1987 by Office of
State Planning consultants Wallace, Roberts and Todd (WRT) and Hammer, Siler, George
Associates (HSGA). The model consisted of a single spreadsheet in which two types of
residential flows (single family and multifamily) and three types of employment based flows
(commercial, office and industrial) for each municipality were calculated using per household

Paper presented at the Third Internatinal Conference on Computers in

Urban Planning and Urban Management, Atlanta, GA, July 23-25, 1993
and per employee factors unique to each of the Tiers in the Draft State Plan. (Tiers were defined
as land areas with similar characteristics, such as density, percentage sewered, etc.). The sum of
residential and employment-based flows was multiplied by a fixed operation cost ($.90 per 1000
gallons) and a fixed capital cost ($2.50 per gallon per day) to yield total municipal system costs.
There were several problems with this method. It lacked information about existing
conditions and service levels. It also assumed that all development would have new sewers.
The next sewer cost model was based on the "Comparable Cities" method developed by
Burchell and Listokin (Burchell and Listokin 1987). This method uses multipliers, developed
from a national sample of cities, to adjust municipal budgets to reflect changes in population and
shifts in the rate of growth. One of the activity costs estimated by this model was "Public
Works," which could include public sector costs for wastewater treatment and collection.
The "Comparable Cities" method was abandoned in response to criticism that it relied on
parameters derived from 1972 national information. OSP decided to develop fiscal impact
techniques using more recent and more New Jersey-specific information.
The model described in this paper utilizes extensive data sets, maintained in New Jersey
by the Department of Environmental Protection and Energy (DEPE), the agency that administers
the EPA Annual Needs Survey. In this survey each public sewer operator must annually describe
his system in detail, identifying existing and proposed service areas, and the number of
households in the service area connected (and not connected) to the system. In addition, the
operator must identify and provide cost estimates of capital improvements needed to bring the
system up to current Federal standards as well as provide cost estimates for improvements
needed to meeting the forecast year (20 year horizon) system demands. The survey requires that
engineered cost estimates be identified. Where local sewer companies have estimated
improvement costs or where future need is noted but future costs are not identified, the Needs
Survey personnel use several cost curves to verify cost projections. These costing formulas were
developed by Weston Engineers under contract to the EPA. The main engine that runs this
version of the OSP sewer model is the EPA cost estimating methodologies fueled by the detailed
systems descriptive data found in the 1988 EPA Needs Survey.
The new work performed by OSP was to program the mainframe EPA model to run on a
microcomputer. In addition, OSP modified the program to include a stochastic model capable of
estimating the degree to which a municipality might be sewered in the future, and an equation
used to modify the estimation of collector costs that would result from sub-municipal design
changes, such as the development of compact centers as recommended in the New Jersey State
Development and Redevelopment Plan.

Model Description

The sewer model consists of four major elements as illustrated in table 1.

table 1
The Major Elements of the Sewer Model

Given: Municipal Population and Employment Projections

Step 1: Estimate the Percentage of Households and Jobs Served by Sewers in the Future
Step 2: Estimate the Total Flow From Each Municipality
Step 3: Assign Municipal Flows to the Correct Treatment Facility

2
 Step 4: Compare the Existing System Capacities to the Estimated Future Demand
Estimate Improvement Costs
Step 5: Assign Collection Treatment Costs to Municipalities

The model accepts alternative municipal estimates of population and employment, and
estimates the costs for the public sewerage treatment and collection agencies to provide service
that meets the requirements of the Clean Water Act. Because the model utilizes very detailed
information about each sewer system, it produces a wealth of specific costs including capital
costs for: secondary treatment; advanced treatment; correction of infiltration and overflow
problems; replacement and rehabilitation of the existing collection system; new collectors; new
interceptors; and, finally costs to correct problems resulting from combined sewer overflows.
However, the model does not generate the total costs of public sewerage systems since it
does not evaluate operational or non-capital intensive maintenance costs. The model also does
not include an estimate of total capital costs, since it does not address costs for privately owned
treatment facilities including the cost for homes to be equipped with on-site septic systems.
Despite the model's shortcomings, we feel it is an appropriate public policy tool for
regional planning. The purpose of the model is to generate future public capital costs so that the
potential costs of providing sewer services can be considered in evaluating alternative land
development scenarios.

Step 1: Estimate the Percentage of Households and Jobs Served by Sewers

Data about wastewater treatment and/or collection facilities were analyzed to determine
which municipalities were serviced. At least one public provider was identified for each of the
State's 567 municipalities. Typically, the role of this provider was wastewater collection. Even
in municipalities where no service was provided, the DEPE records identified a public agency
ultimately responsible to provide for collection of sewage.
To determine the number of persons within the service area who are to receive collection
and treatment service, the user selects one of the three viable service conditions.
o 1988 Constant - Records provided by DEPE, based on the Needs Survey, list the
number of persons receiving collection and not receiving collection for each
wastewater provider. If the user selects this option then the 1988 percent of the
total population and employment receiving collection is used to determine the
number of persons receiving collection in the future.
o 2008 Constant - The DEPE data displays the number of persons proposed to
receive collection in 2008 and the number not receiving collection in 2008. If
this variable is chosen, the percentage receiving collection in 2008 is multiplied
by the total future population and employment.
o OSP Estimate - Several municipalities are not receiving any wastewater
collection today nor is any forecast for the future. In most cases this is due to low
growth estimates. Another problem is that growth might exceed the expectations
of the local company and might result in the need for higher levels of service
than those proposed in the 1988 or 2008 constants.
The OSP methodology uses DEPE datasets to establish a relationship between municipal
population and percent-sewered. Once such a relationship is quantified, the percentage of
residents on collectors can be projected at any level.

3
 1
First, we calculated municipal population density for the year 2008 , using the 208 plan
data set provided by DEPE. This same dataset also provided an estimate of the ratio of persons
on collectors to total population, by municipality, for the year 2008.
When graphed, the relationship between municipal population density and percent-
sewered appears to follow a pattern (see figure 1). with large concentrations of municipalities at
0% and 100%, however, the relationship depicted in figure 1 does not lend itself to modeling
using standard parametric techniques. Instead, a probability table was used to model the
relationship.
To accomplish this analysis, the 555 municipalities (out of a total of 567 municipalities)
for which there was percent-sewered data were divided into ten population density classes of 55
or 56 each. The probability distribution of municipalities in each density class was calculated
across seven classes of percent-sewered. The seven classes were: NOT SEWERED, 1 to 20%
sewered, 20 to 40% sewered, 40 to 60% sewered, 60 to 80% sewered, 80 to 99% sewered and
100% sewered. Next, the cumulative probability distribution was calculated for the
municipalities in each class. To utilize this information in the program, a lookup table was
constructed using these cumulative probabilities. The table returns percent-sewered given a
municipality's future population density and a computer generated random number. The
percentages returned by the table are restricted to 0%, 100% and the midpoints of the other five
classes (i.e., 10%, 30%, 50%, 70% and 90%).
Figure 2 shows that this simulation technique replicates the relationship shown in figure
1 reasonably well.

figure 1
Percent of Population on Collectors as a Function of Log(Population Density)
percent served

Log of Population Density in People/Sq. Mile

4
 figure 2
OSP Simulation of Population on Collectors as a Function of Log(Population Density)
percent served

Log of Population Density in People/Sq. Mile

The final result of Step 1 is the identification of people and employees receiving
collection for each municipality-based collection provider and the total number of persons and
employers served by each treatment plant.

Step 2: Estimate the Total Flow from Each Municipality

Total wastewater flows, expressed in millions of gallons per day (MGD), are used during
the next step as major input to cost estimating formulas. The general formulas used in the model
are:

Residential MGD = TPS x (GPCD x (1-CW) )/1,000,000 (1)

Employment MGD = TES x (GPED x (1-CW) )/1,000,000

TPS = Total Population Served

TES = Total Employees Served
GPCD = Gallons per Capita per day
GPED = Gallons per Employee per day
CW = % water saved by conservation methods

To a larger degree, the formulae for flows are derived from the EPA cost estimation
methods. OSP added the water conservation element, since some type of water conservation
methods will likely to implemented in the future. Three of the formula elements are user-selected
variables.
1. Gallons Per Capita a Day - three alternatives are presented:

5
 o 1988 Constant - The DEPE data records 1988 flows as gallons per capita
per day. This number is derived by taking total flow, subtracting industrial flow
and dividing the remainder by the number of persons receiving collection and/or
treatment. This facility-specific 1988 GPCD is used as a constant if this
alternative is selected.
o 2008 Constant - The constant in all equations is the facility-specific
GPCD reported in the Needs Survey for 2008.
o EPA numbers - EPA recommends 95 GPCD to serve populations of less
than 5,000; 105 GPCD to serve populations greater than 5,000 and less than
10,000; and 115 GPCD to serve populations greater than 10,000 individuals.
2. Wastewater per employee per day - This factor represents the total industrial flow sent
to public collection and/or treatment facilities divided by the number of employers served by the
facility. It does not represent total industrial wastewater discharged in a service area. Much
industrial wastewater is treated and discharged by business itself. Options available for selection
by the user range from 30 gallons per day per employee (GPED) to 90 GPED, in increments of
ten.
3. Water saved through conservation - Many parts of New Jersey already use more water
than that provided by locally available sources. Northeastern New Jersey is dependent on water
from the Delaware River. Parts of Southern New Jersey are exhausting groundwater supplies to
such an extent that the water table (piezometric level) has dropped as much as 50 feet. Therefore,
the implementation of water conservation techniques is likely to be commonplace in the near
future. Such reductions in water consumption also result in reductions to wastewater flows.
User-selected variables to simulate this phenomenon are expressed as percentage reductions to
the wastewater flow ranging from 0% reduction to 20% reduction.

Step 3: Assign Municipal Flows to the Correct Treatment Facility

Two alternative structures were plausible. First, wastewater collection agencies could
operate their own treatment plant. Alternatively, they could be grouped into service areas and
provided with a single regional treatment plant designed to treat the collective wastewater of its
constituent collection agencies.
We assigned municipal flows to the treatment plant intended to serve these areas in the
future, in accordance with the current 208 and 201 plans submitted by local agencies to DEPE..
Where a municipality was served by more than one facility, the proportion of the municipality
served by each agency in 1988 was assumed to continue. For example, if 20% of Anyplace
Township is served by Someplace Sewer Company and the rest of Anytown is served by Noplace
Sewer Company, then 20% of the future population and employment is assigned to the
Someplace Sewer Company and the rest of Anyplace's population and employment is assigned to
the Noplace Sewer Company.
To perform this assignment process OSP used named formulas to link municipal flows to
treatment facilities. This results in a program that works, but which is very difficult to check or to
make corrections or change. It would be better to use a database program to store these
relationships and then to link the database to the sewer program using DDE methods.

Step 4: Estimating Improvement Costs

In developing this portion of the model, several Rules had to be incorporated into the
program to decide when and how to use the cost estimates produced by the local companies as
opposed to using costs projected by OSP or EPA methodologies. First, if future municipal flows

6
were equal to or less than the 1988 wastewater flows, then no costs were added to meet future
needs. Second, if future needs exceeded the 1988 flows, but were equal to or less than the 2008
estimated flows, then the local agency estimates could be used. One of the data sets provided by
DEPE indicates a basis upon which these locally produced costs were estimated. If these costs
were the products of engineering studies, the costs were directly used in the model. If the costs
were based on the EPA cost-estimating equations, then new costs were generated using the EPA
equations and the exact new flow requirements. Third, if future flows exceeded the 2008 Needs
Survey flow estimates and the 2008 costs were locally engineered, these engineered costs were
used up to their design capacities then the EPA cost equations were used to cost out the residual.
Where the 2008 costs were not engineered, all growth costs were estimated using the EPA cost
equations.

Backlog Costs. All Needs survey identified costs intended to alleviate existing "backlog"
problems were included in all cost estimates reported by the model.

Collector Costs. Collector costs are estimated using the following equation provided by the
EPA.
2
(Future Collected - Existing Collected) x 16 ft. x $59 x CDA x (RCA ) (2)
where:
CDA = constant dollar adjustment from the EPA base year of 1985
RCA = Regional Cost Adjustment

The model assumes 16 feet of pipe per person and that the pipe costs $59 per foot. It
became evident to us that the constant of 16 feet of collector pipe per person had to incorporated
a constant assumption about residential density. To make the collector cost model more robust
(and more sensitive to the clustering policies included in the Plan), we undertook a statistical
analysis in order to identify the relationship between density/design on the one hand, and per unit
road requirements on the other. It was assumed that most collectors are built beneath or parallel
to local roads, as are some other infrastructure systems, especially water distribution lines.
The general approach was to generate a large and diverse sample of residential
development patterns using 1986 aerial photoquads at 1:24,000 scale. To measure the
development we used clear plastic sheet sheets with 20-acre grids, each then subdivided into 16
squares of 1.25 acres each. the twenty-acre grids were placed on the photoquads. Center-line road
lengths (in linear feet) within each grid were calculated using an electric digital curvimeter
(naturally, driveways were excluded). The number of housing units in each subsquare was
counted and recorded, along with the photoquad number, sample number, and road length. (The
number of units divided by the 20 acres in each grid resulted in gross density.
Samples were chosen based on clarity and recognition as single-family housing
developments. The intent of this exercise was to count housing units, not industrial, commercial,
or office uses. since no field work was conducted, it was necessary to restrict the sample to
single-family residential developments, which are easy to identify in aerial photos. One must
therefore exercise caution before extending the results of this analysis to developments
containing mixed-use or multi-family structures (even though intuitively, multi-family
developments could be expected to increase infrastructure cost-efficiencies).

7
 Approximately three 20 acre observations were taken from each photoquad. Exceptions
to this occurred in cases where the photoquad only showed a small amount of New Jersey land.
Eventually, 368 observations were collected.
In addition to density and road length, the variables COVER was estiamted. COVER, is
the percentage of subsquares in each sample that contain any development. Holding density
3
constant, the COVER variable clearly serves as a measure of cluster .
Linear regression was used to analyze per-unit length of road (dependent variable) and
the various measures of density and cluster (independent variables). the coefficient on DENSITY
is hypothesized to be negative; and COVER, positive.
4
At the outset, it was recognized that density and cluster are likely to be highly collinear .
In some sense, whether a land use pattern is dense or highly clustered is simply a matter of scale.
this problem has not been satisfactorily resolved in the analysis.
Plots of the data suggested using a log-log form for the regression equation. If one had to
choose among the collinear independent variables, it appears that DENSITY provides the
strongest confirmation of the hypothesis. some of the regression results are displayed in table 4.

table 4a
Regression of Density to Road Length
(Log-Log form)
Dependent Variable: Log of Per-Dwelling Unit Road Length
Independent Variables:

Parameter Estimate T-Statistic PROB > T S.E.

Intercept 4.66 259.48 0.0 .02
Log Density -0.64 -39.40 0.0 .02
N = 367
F-VALUE: 1552.39
PR>F: 0.0
R-SQUARE: .809

table4b
Regression of Density and Cover to Road Length

Dependent Variable: Log of Per-Dwelling Unit Road Length

Independent Variables:

Parameter Estimate T-Statistic PROB > T S.E.

Intercept 4.61 113.77 0.0 .04
Log Density -0.58 -12.92 0.0001 .04
Log Cover -.010 -1.45 0.1496 .07
N = 367
F-VALUE: 1557.11
PR>F: 0.0
R-SQUARE: .810

8
 Further analysis to control for colinearity between density and cover were made by
dividing the sample into density classes. Regression analysis of COVER within each class to
road length was made, holding density more or less constant. It was found that the effect of
COVER on Road length was largely insignificant. Therefore, the relationship between road
length and density was the basis for the OSP algorithm to calculate collector lengths.
Trend cost estimates are produced using the density to road length equation (table 4b)
and the existing density of the municipality in lieu of using the EPA constant of 16 feet per
person. If the user selects to test a Plan scenario, an additional dialog box prompts the user to
estimate the extent to which development under Plan would be more (or less) dense than would
the same development under a Trend growth scenario. This density factor than is multiplied with
the mean density of each municipality to simulate the mean municipal density for new
development under Plan.

Interceptor Costs. - To calculate interceptor costs a multi-step process is employed. First, total
future flow is subtracted from total existing flow to yield total new flow due to growth. The
model always assumes that this increased flow will be handled by constructing a new interceptor.
This flow is then compared to an EPA generated table to determine the diameter of pipe needed
to carry the flow. A second EPA table reveals the cost per foot of the selected diameter pipe.
Finally, a third EPA table estimates the length of interceptor required based on the change in
population served by the system.
Based on this information the following equation is used.

Interceptor Cost = Length x Cost/sq ft x (CDA) x (RCA) (3)

Again, where engineered interceptor costs are provided, they are preferable to numbers
generated by the EPA cost methods.

Treatment Plant Costs. The following briefly describes the EPA cost equations.
o Treatment Costs - These equations estimate capital costs of sewer
treatment plants. The method uses separate cost equations for ten different types
of plants, ranging from 'plain vanilla' secondary treatment facility to highly
sophisticated advanced treatment plants. Data provided by DEPE identified
which of the ten plant types would be needed for the plant to meet the
requirements of the Clean Water Act, given the water quality of the discharge
stream or river. In applying the EPA cost equations, OSP assumed that the
"backlog" costs, if any, would bring the existing plants up to requirements.
Therefore, the cost equation only estimated those plant improvements necessary
to serve the added flow created by growth and the costs of improving treatment
methods, if required. The EPA model also provides equations to "recover" costs
to reflect efficiencies from re-using existing plant facilities. This cost-recovery
method was not used in the OSP model, nor is it used by DEPE, due to the use
of "backlog" costs to improve existing plant estimates.
o Plant Capacity Variable - Federal law stipulates that new treatment
facilities should commence the design/build process when flows reach 80% of
plant capacity. Many plants operate in violation of this requirement. This
variable allows the user to select the operating level at which the cost of a new

9
 facility is estimated. The range is 80% to 120%, with a default setting of 100
percent.

General Cost Adjustment Variable. For all cost estimates, resulting costs are adjusted by a user-
selected variable to reflect local labor and materials price conditions. The cost adjustment
variables differ from the Regional Cost Adjustments, in that the regional cost adjustments reflect
cost differences that DEPE reported as a function of density, congestion, and the degree to which
new improvements have to be installed into utility right-of-ways serving other systems, such as
water or electric lines. The price adjustment variables are:
o Constant = 1
o New York Region = 1.52
o Trenton Region = 1.09
o Philadelphia Region = 1.21

Step 4: Reassignment of Costs to Municipalities

During this final step, all backlog and improvement costs are reassigned to the
municipalities served by these systems. Costs are assigned proportionate to each municipality's
percentage of total flow to the system. Multiple costs can accrue to any municipality. This cost
aggregation allows for the cost of regional facilities, possibly primarily resulting from growth in
another municipality, to be assigned to each municipality in the sewer service region.

Model Results and Findings

The sewer model has been used by OSP to evaluate various public policies, with interesting
results. In addition, the model was adopted for use by the Rutgers University Center for Urban
Policy Research, when they conducted the official evaluation of the New Jersey Development
and Redevelopment Plan, as mandated by Act of the New Jersey Legislature.
Rutgers reported that the possible sewer capital cost savings that would accrue from
implementing the State Plan would be $379 Million over a 20 years period. While this is a great
deal of money, it is not as large as many might have anticipated. However, it is important to note
that this result is primarily due to the relatively conservative population growth forecast which
Rutgers used in their analysis.
The Rutgers finding also foreshadows some policy myths exposed by the model. It had
been taken as an article of faith, by proponents of the State Plan that there were areas of the State
with substantial excess capacity and that the concentration of new growth into these excess
capacity areas and into areas already provided with existing service would result in substantial
public cost savings. The spatial expression of these concepts was that new growth should be
focused into the older urbanized portions of the State, which experienced population and
employment declines (and therefore should have excess treatment capacities and the existing
collectors and interceptors to support growth), into vacant land in suburbanizing areas served or
designated to be served by sewer, and into existing towns and villages, presumed to have excess
capacity so that only collectors and interceptors would have to be built.
Unfortunately, the model and its datasets demonstrated that for all practical purposes
there was no excess capacity in New Jersey. The older urbanized areas typically had collector
systems in need of substantial reconstruction and undersized plants providing very low levels of
treatment. The costs to bring these areas into existing day conformance were so high that some of

10
the counties feared that this information might act as a detriment to future development. It also
became evident that adding capacity to these older systems might be more costly that building
new systems in more rural areas of the State. Because of this result the utility underpinning of the
Plan's urban policy was subtly changed to argue that new development should be allocated to
these urbanized areas since major investment would be required in any case to service the
existing residents and business.
Another political policy problem resulted from the concept of locating suburban growth
into areas designated to be sewered. Earlier versions of the State Plan utilized sewer service
boundaries to identify areas where future growth was to be encouraged. The sewer model dataset
revealed that many municipalities had received funding from the Federal government to provide
collectors into areas designated in their 208 Plans, but instead had used the money to provide
service to areas where the market had directed growth. (In New Jersey there is no requirement
that Master Plans and Zoning conform.) Because of the Federal grant, many municipalities now
were obligated to sewer the 208 areas with their own money as a pre-condition to receiving any
additional Federal sewer construction grants. Therefore, if growths were directed into many areas
designated to be sewered, substantial local costs would be incurred. The policy resolution of this
problem was the downgrading of the role of sewer service areas played to define areas where
growth is desired. In addition, increased emphasis was place on the Plan's policies to require or
encourage private developers to install sewer service.

11
 Bibliography

Burchell, Robert and Listokin, David. 1987. The Fiscal Impact Handbook. ( New
Brunswick, NJ: Center for Urban Policy Research).
Burchell, Robert. 1992. Impact Assessment of the New Jersey Interim State Development
and Redevelopment Plan. ( New Brunswick, NJ: Center for Urban Policy
Research).
Gottlieb, Paul. 1990. Density, Design and Infrastructure Costs, Part II: Physical Survey
of New Jersey Development. (Trenton, NJ: New Jersey Office of State Planning).
New Jersey Office of State Planning. 1988. Communities of Place, A Legacy for the Next
Generation: the Preliminary State Development and Redevelopment Plan for the
State of New Jersey. (Trenton, NJ: New Jersey Office of State Planning).
New Jersey Office of State Planning. 1992. The New Jersey State Development and
Redevelopment Plan (Trenton, NJ: New Jersey Office of State Planning).
Real Estate Research Corporation. 1974. The Cost of Sprawl. (Washington, DC:
USGPO).
U.S. Environmental Protection Agency. 1988. 1988 Needs Survey, Appendix A, User
Manual. (Washington, DC: EPA).
Wallace, Roberts and Todd. 1987. Draft Preliminary State Development and
Redevelopment Plan. (Trenton, NJ: New Jersey Office of State Planning).

1
Year 2008 data was used for this analysis because of its ready availability from DEPE. An alternative
would be to use 1988 data. Since we are simulating for the years 1995 to 2010 in the model, one could
imagine using either 1988 or 2008 as a standard, depending on whether the target year is early or late in
the forecast.
2
Three regional cost adjustments are provided and user selected: urban = 1.13; Suburban = .95; Rural =
.80)
3
A second varible representing the standard diviation of the number of units in each subsquare also was
developed and explored. Results from this variable are not discussed in this paper.
4
Pearson correlation coefficient of .759 between DENSITY and COVER in this sample.

12