Draft
Municipal and Industrial
Water Demand Projections to the Year 2050
Central & Southern Florida Project
Comprehensive Review Study


Table of Contents


Summary
Overview
Summary Municipal and Industrial Demands for 2050 by Service Area
Demand Forecast Methodology and Procedures
About IWR-MAIN
Theoretical Basis
History
Data Collection and Model Calibration
Utility Data Collection
Calibration and Verification of IWR-MAIN Model
Summary
Residential
Nonresidential
Adjustment of County Estimates to South Florida Water Management Model (SFWMM)
        Service Area Boundaries

Projection of 2050 Use Estimates
Background
Projection A
Projection B
Population and Employment Projections
Average Per Capita Consumption
Projections Used in SFWMM Simulations
Comment

 

Draft
Municipal and Industrial
Water Demand Projections to the Year 2050
Central & Southern Florida Project
Comprehensive Review Study

Summary

Two projections of future water consumption for the year 2050 have been made for the Lower East Coast study area.  The two scenarios differ in terms of the assumed level of water use conservation.  The higher estimate, Projection A, is based on the same percentage distribution and usage of conservation flow devices, and irrigation restrictions, in effect in 2050 as in 1990.  The lower estimate, Projection B, is based on the full implementation of existing South Florida Water Management District mandatory regulations and programs.  This lower bound projection scenario represents 100% use and effectiveness of ultra-flow devices by the year 2050.

The higher Projection A estimate for the year 2050 is about 1450 MGD (million gallons per day).  The lower Projection B estimate is about 1200 MGD, approximately 18% less than Projection A.  In this study, the 2050 base condition (the "without-project" condition) assumes a more moderate application of conservation practices and effectiveness, representing a level of consumption about 12% below the 2050 Projection A estimate.

Overview

The principal analytical tool being used to evaluate the alternative plans being considered in this study is the South Florida Water Management Model (SFWMM).  This model simulates changes in South Florida hydrology which would result from these alternatives, represented by structural and operational modifications to the Central & Southern Florida Project (C&SF) Project.  It simulates a 31-year-period of hydrometeorological conditions, using historical data for the 1965-1995 period, to represent a broad range of potential conditions which could be expected to take place over the long run.  The average annual output of this simulation is used to measure the effects of alternatives being considered in this study.

A major part of the evaluation of alternatives consists of comparing SFWMM output associated with a particular alternative in place, with SFWMM output representing conditions without the alternative.  This comparison of "with-plan" vs. "without-plan" conditions is made for projected conditions for the year 2050.  These conditions include an assumed level of water use that is representative of population and economic conditions estimated to exist in 2050.

A key input to this analysis is the estimated water use requirement of the population and economy, in the area covered by the SFWMM, in the year 2050.  The SFWMM simulation process estimates as output the demands of this area for agricultural irrigation, and for the water management requirements of the C&SF project (salinity control, flood control, etc.).  But the municipal and industrial (M&I) demand is required input for running the SFWMM.  The 2050 projected M&I water use demands by the residential, commercial, industrial, and public administration sectors of the water using population and economy have been estimated using the IWR-MAIN Water Use Forecasting System.

The basis for the M&I projections used in this study is a water use demand forecast which was prepared under contract for the U.S. Army Corps of Engineers, Jacksonville District, Gulf Engineers & Consultants, Inc., Baton Rouge, Louisiana (GEC), for specific sub-areas of the SFWMM coverage area. The sub-areas for which these projections are being used in the C&SF evaluations, referred to as Service Areas 1, 2, and 3 (SA1, SA2, and SA3), and the North Palm Beach Service Area, comprise what is widely referred to as the Lower East Coast (LEC), which consists of the populated developed portions of Palm Beach, Broward, and Dade Counties along Florida’s southeast Atlantic coast.  The 2050 IWR-MAIN demand forecast for these large areas was then converted into specific well withdrawal volumes, conforming to input data format required by the SFWMM.

The North Palm Beach Service Area and SA1 coincide roughly with the eastern developed portion of Palm Beach County, but includes a small portion of northern Broward County.   SA2 consists of the rest of Broward County, plus a small portion of northern Dade County.  The demands for SA3, consisting of the remainder of Dade County, include water demands of the Florida Keys (Monroe County), since water for this area is supplied from SA3 well fields.

Summary M&I Demands for 2050 by Service Area

The Projection A average daily M&I demand for water use in the year 2050 is summarized below.  The table shows that water use is fairly evenly distributed among the LEC Counties.  The Service Areas which coincide mainly with the developed portion of Palm Beach County account for 30 percent of total forecast M&I use, with a little over 27 percent in Service Area 2, which roughly coincides with Broward County.   Service Area 3 use, representing demand in most of Dade County and the Florida Keys (Monroe County), is somewhat higher in terms of its share of the total.

Summary 2050 M&I Demands by Service Area – Projection A

Area

Million Gallons Per Day (MGD)

Percent of

Total

North Palm Beach Service Area

101.25

7

Service Area 1

349.20

24

Service Area 2

422.24

29

Service Area 3

577.00

40

Total

1449.69

100

These 2050 Projection A estimates reflect a level of conservation practices that is the same as estimated to be in place in 1990.  That is, the same percentage distribution of the use of restrictive flow devices among all uses in place in 1990 is assumed to be in place for the 2050 usage, and therefore probably can be viewed safely as an upper bound forecast estimate.

Another set of forecast use estimates, representing 100% use of ultra-flow restrictive devices in place for all consumers by 2050, was also made.  The 2050 summary results of this conservation Projection B scenario, which can be viewed as a lower bound forecast estimate, are shown in the following table.  These conservation projections represent continued implementation of the mandatory SFWMD program.

Summary 2050 M&I Demands by Service Area – Conservation Projection B

Area

Million Gallons Per Day (MGD)

Percent Reduction 1/

North Palm Beach Service Area

83.66

17.37

Service Area 1

294.18

15.76

Service Area 2

345.72

18.12

Service Area 3

474.80

17.71

Total

1198.36

17.34

1/From Projection A

The IWR-MAIN forecasts have been categorized by residential, commercial, industrial, public administration, and unaccounted for uses.  The following percentage breakdown provides a profile of these uses in the study area for 2050 for Projection A.  As the tabulation shows, this profile is generally similar throughout the study area, although residential use is more heavily weighted in southern areas.

Percentage Distribution of 2050 Demand by End Use

And by Service Area

End Use

NPB

SA1

SA2

SA3

Total

Residential

47

49

56

58

54

Commercial & Industrial

36

37

28

22

29

Public & Other

17

14

16

20

17

Total

100

100

100

100

100

The demand projections made using IWR-MAIN are made by large areas because the projections are driven by economic and demographic projections, which have been made at the county-wide level.  But the SFWMM input requires that the demand input be in the form of well withdrawals, by month, in MGD, spatially identified by grid-cell location, using the SFWMM matrix of 4 square mile (2 mi. x 2 mi.) cells.  This information has been developed for existing well pumpages.  The conversion of the above projected service area water use into grid-cell based well withdrawal data has been developed using known existing well field locations, and the likelihood of future locations and operations.

The SFWMM runs that are being used for the comprehensive plan study ("The Restudy") include scenarios for the years 1995 and 2050, and represent both public and private wells.  The well withdrawals being used for the 1995 scenario are based on existing known withdrawals, amounting to a total of 803.6 MGD, and represent part (i.e., not all) of the total demand/use in 1995 for the projection area.  Excluded is golf course and commercial landscape irrigation (estimated by the SFWMM simulation as a part of the evapotranspiration simulation calculation runs), deep well withdrawals from the brackish Floridan aquifer, and some other uses which are not consumptive.  For example, water is used in rock mining operations, but it is returned immediately after use (consisting mainly of washing rock cuttings), and therefore such use is not really a consumptive use.  Instead, it is more representative of moving water from one place to another in the system.  Floridan aquifer withdrawals do not represent a withdrawal from the water system modeled by the SFWMM and are outside of the Everglades system.

Demand Forecast Methodology and Procedures

About IWR-MAIN

Theoretical Basis

Perhaps the simplest approach for predicting future water use is the use of projected gross per capita water use rates and projected population.  Gross per capita use can be observed from known water consumption and known population.  Projections can be made for different population growth scenarios combined with varying assumptions about future changes in the capita use.  This approach was not used in this study.

IWR-MAIN (Version 6.1), a PC-based software tool used to make projections for water use in this study, is based on observed relationships between water use and causal factors, or determinants, of urban demand for water.  Causal relationships have been developed separately for residential and non-residential use.  Forecasting relationships used in IWR-MAIN for the residential sector are based on the integration of approximately 60 studies of residential water demand, which contained about 200 empirically estimated water use equations.

For the residential sector, the generalized form of the equation for projected average use is (IWR-MAIN User’s Manual and System Description, April 1995, p. D-2):

Q = a Id1 MPd2 e(FC)(d3) Hd4 HDd5 Td6 Rd7

where

Q = predicted water use in gallons per day
I = median household income ($1,000’s)
MP = effective marginal price ($/1,000 gal)
e = base of the natural logarithm
FC = fixed charge ($)
H = mean household size (persons/household)
HD = housing density
T = maximum-day temperature (degrees F)
R = total seasonal rainfall (inches)
a = intercept in gallons/day
d1...d7 = elasticity values for each independent or explanatory variable

The above relationship can be approximated for any residential subsector, season, purpose (e.g., indoor, outdoor), depending on data availability.

Elasticity is a measure of the relationship between water use and a given explanatory or independent variable.  For example, an income elasticity of  +0.5 would indicate that a 1 percent increase in income would result in a .5 percent increase in water use.  Some representative sample default elasticities contained in IWR-MAIN are outlined in the following table.

Representative Default Elasticities for Residential Water Use Contained in IWR-MAIN

Explanatory

Variable

Single Family

Summer

Multi-Family

Summer

Income

+0.40

+0.40

Persons per Household

+0.40

+0.40

Housing Density

-0.65

-0.30

Marginal Price

-0.25

-0.15

Fixed Charge

-0.0015

-0.001

Maximum-Daily Temperature

+1.50

+1.20

Total Rainfall

-0.25

-0.10

Forecasting relationships used in IWR-MAIN for the nonresidential sector are based on over 10 years of research on the relationship between employment and water use in over 7,000 establishments representing the eight major industry groups throughout the United States.  For the non-residential sector, the generalized form of the equation for projecting water use is:

Q = a PRd1 MPd2 CDDd3 OTHd4

where

Q = water use in gallons/employee/day
PR = labor productivity
MP = marginal price ($/1,000 gal)
CDD = cooling degree days (number of days)
OTH = other (user added)
a = model intercept (gallons/employee/day)
d1...d4 = elasticities for independent/explanatory variables

The above relationship is designed to be approximated by Standard Industrial Classification (SIC) group.  While this model is operational within the IWR-MAIN model, the elasticities for all of the explanatory and independent variables are currently set to zero, since there are currently no available elasticities for them.  As a result, this version of IWR-MAIN uses a single coefficient equation to calculate water use by industry group:

Q = GED * E

where

Q = water use in gallons per day
GED = gallons per employee per day
E = number of employees

History

The IWR-MAIN system has its origins in work during the 1960’s by researchers conducting investigations with the Residential and Commercial Water Use Research Projects at Johns Hopkins University, and for the U.S. Office of Water Resources Research.   The original system was called MAIN, which was followed by improvements in MAIN II.   These early analytical tools were based on research by Howe and Linaweaver (1967); Wolff, et. al.; and other researchers.  The initial system was developed by Hittman Associates, Inc., in the late 1960’s.

The U.S. Army Corps of Engineers' Institute for Water Resources improved upon the MAIN system, renaming it the IWR-MAIN Water Use Forecasting System, and modifying the model during the 1980’s.  Ultimately a PC version was created, and new use coefficients and computation techniques were incorporated into the new model based on newly available data and research literature.  Improvements continue to be made and are the basis for new updated versions when they are documented and become available for distribution and use.

Current IWR-MAIN development is being accomplished by Planning and Management Consultants, Ltd. (PMCL), of Carbondale, Illinois, under the sponsorship of the Institute for Water Resources, U.S. Army Corps of Engineers; Metropolitan Water District of Southern California; Phoenix Water Services Department; and Illinois Department of Transportation.

IWR-MAIN has been used for forecast studies for a number of major water utilities in the U.S.  Some of these are: Indianapolis Water Company; Phoenix Water and Wastewater Department; Metropolitan Water District of Southern California; El Paso Water Utility; Binghamton, New York; Springfield City Water, Light, and Power; Southwest Florida Water Management District; Las Vegas Valley Water District; and the City of San Diego Water Utilities Department.

Data Collection and Model Calibration

Before the water demand forecasting projections can be made, the important steps of utility data collection, and calibration and verification of the IWR-MAIN model must be made.

Utility Data Collection

Data were collected for 1990 and 1994 from utilities in Palm Beach County (representing approximately 53% of total water use in the county), Broward County (54% of total county water use), and Dade County (91% of total county water use).  Data collection was accomplished with a field survey to obtain the following water data requirements of IWR-MAIN:

Number of accounts billed monthly by user category

Quantity of water sold monthly by user category

Water production by month

Water and wastewater prices in effect

Monthly water use for the largest customers

Water conservation measures in effect

The utility companies in the data collection sample provided detailed data for a large portion of the water usage within the study area, covering a wide area, and including different types of development.  For example, data coverage includes areas with high seasonal variations, such as resort areas along the Atlantic coast, highly developed urban and suburban areas, and rural communities.  There was a high degree of cooperation on the part of the utility companies in this data collection effort.

Calibration and Verification of IWR-MAIN Model

Summary

The final result of the calibration and adjustment procedures resulted in a model-produced 1990 estimate for this water demand forecast study area, 913.05 MGD, that is within 2.4 percent of the actual use for that year, 935.52 MGD (U.S. Geological Survey).

Summary of IWR-MAIN Estimated 1990 Water Use (MGD)1/

For Service Areas Comprising Palm Beach, Broward, and Dade2/ Counties

 

Service Area

Sector

North Palm Beach

Western

Palm

Beach3/

Service Area 1

Service Area 2

Service Area 3

Total

Residential

23.86

3.20

87.66

150.20

187.79

452.71

Commercial & Industrial

23.64

4.41

86.60

92.45

105.04

312.14

Public & Other

8.43

1.76

30.95

32.82

74.23

148.19

Total

55.93

9.37

205.21

275.47

367.07

913.05

1/ Totals may not add due to rounding.
2/ Dade County demand includes Florida Keys (Monroe County); the Keys are supplied from Dade County well fields.
3/ Western Palm Beach County M&I demand, representing 1% of the total, is not included in the SFWMM calculations.

USGS 1990 Water Use Estimates1/ for Palm Beach, Broward, and Dade2/ Counties (MGD)

Use

County

Three County

Category

Palm Beach

Broward

Dade

Total

Public Supply Domestic

112.80

132.25

212.36

457.41

Public Supply Commercial

22.52

28.14

44.33

94.99

Public Supply Industrial

9.59

15.45

30.00

55.04

Public Other

3.24

2.25

6.51

12.00

Public Use

16.47

14.44

32.42

63.33

Public Subtotal

164.62

192.53

325.62

682.77

Self Supply Domestic

21.34

9.60

10.75

41.69

Self Supply Comm/Industrial

32.17

1.63

40.34

74.14

Self-Supply Subtotal

53.51

11.23

51.09

115.83

Golf Courses

44.39

21.31

10.39

76.09

Thermoelectric

0.00

0.05

2.26

2.31

Other Irrigation (Comm)

15.61

25.49

17.39

58.49

Total

278.16

250.61

406.75

935.52

1/ Data is from Water Withdrawals, Use, Consumption, and Trends in Florida, 1990, USGS Water Resources Investigations Report 92-4140, and unpublished USGS data.
2/ Dade County demand includes Florida Keys (Monroe County); the Keys are supplied from Dade County well fields.

IWR-MAIN estimates demand (not supply source), and therefore, as shown in the table, the sector breakdown for the IWR-MAIN estimates is categorized by use.  The USGS estimates are in terms of slightly different use categories, and partly also by supply (i.e., self-supplied vs. utility-supplied).  Although the individual breakdown categories may not be completely comparable, the overall totals are.  The USGS and IWR-MAIN totals are comparable because they both represent estimates of total use, even though the category breakdown formats are different.

A good choice for a base year for use in creating an IWR-MAIN forecast is a census year because of abundant socio-economic data availability for model calibration.  Hence, 1990, a census year, was chosen to be the base year.  The fact that 1990 was a dry year with some mandated water use cutbacks, and the fact that the verification backcast year, 1994, was a wet year without such cutbacks (but with lower irrigation demands), combined to test the predictive capability of the model.  As discussed below, the 1990-calibrated model turned out to be a relatively good predictor of 1994 demand.   This was due in part to the fact that 1990 and 1994 weather variables, which would be expected to account for some of the variation between the two dissimilar years’ consumption patterns, were used to estimate use in 1990 and 1994 in the calibration and verification procedures, respectively.  Long term projections through to the 2050 planning horizon used long term average weather variables.

Residential

Data collected from the utilities surveyed were entered into the IWR-MAIN system and 1990 use estimates were estimated by the system for the sampled utility service areas.   Initial test runs proved that the model estimates are closer to the utility service area estimates when summer elasticity values are used year round.  The elasticity values were adjusted appropriately.  The calibration procedure involves adjusting the coefficient (intercept) so that the model will estimate the actual water use for 1990.

The next step was to use this calibrated model to estimate 1994 use for the same utility service areas, as a test of model effectiveness.  For the sampled utilities for which this initial model was constructed, the overall average residential MGD use predicted by the model was 6% below actual use for the 1994 verification.  This differential varied by residential subsector and by season.  For such small areas, it is highly unlikely that the IWR-MAIN equations will predict values that are identical to actual values in a verification procedure such as this one for 1994 with a model reflecting 1990 conditions.  As a rule of thumb, according to IWR-MAIN procedural guidance, differences in the 3 to 5 percent range indicate good performance; differences exceeding 10 percent usually mean further calibration is needed.  The 6% difference in the verification procedure was deemed to be reasonable, and certainly well below the 10% difference threshold suggesting further calibration.

Additional calibrations were made to reflect residential water use patterns in areas within the SFWMM study area, but outside of the utility service areas whose sampled data was used in the initial calibration.  Such calibrations were intended to reflect differences in unaccounted for losses, pumpage vs. treatment quantity differences, and per housing unit water consumption patterns.  County-wide 1990 USGS data were used for these additional calibration adjustments. These adjustments helped to make the overall calibration much closer, and enabled extrapolation of water use estimates to the rest of the county areas.

Nonresidential

The nonresidential sector includes commercial, industrial, and public administration use.  Model estimation is simply a matter of employment combined with per employee average water use coefficients.  These use coefficients exist in the model for the eight major industry groups, and by further more detailed breakdown into employment industry subgroups at the 2- and 3-digit SIC code level.  There are 432 such categories of nonresidential water use per employee coefficients available within the version of IWR-MAIN used in this analysis (the most recent available at the time).

A test using Dade County employment data and nonresidential water use for 1990 was conducted to determine the best SIC level for predicting nonresidential water use.   The 3-digit SIC code employment and water use coefficients came closest.   While there are no reliable employment projections at the 3-digit SIC code level, the 3-digit water use coefficients were calibrated to reflect the 3-digit code employment mix within each major group in each county, using 1990 data.  This calibration produced adequate estimates for 1990 except for Palm Beach County, which required additional adjustments to account for high water use for golf courses, parks, and sports facilities.

Adjustment of County Estimates to SFWMM Service Area Boundaries

The economic demand models within IWR-MAIN are driven by independent demographic and economic variables that must be specified for the base year of the forecast (1990) to reflect conditions within the study area.  Criteria for selection of a base year are that data must be available for both actual water use, and for the economic and demographic independent variables to be used in the forecast.  Most of the required demographic and economic input variables are available from the census, making the latest census year a good choice for a base year.  Census employment data is by residence and IWR-MAIN requires employment data by place of work; consequently, other data sources were used for employment.

The SFWMM service area boundaries do not coincide with county boundaries, which required manipulation of the data.  This was accomplished by overlaying SFWMM service area maps onto census maps to determine which census block groups and which portions of census block groups fall into each service area. To estimate the portion of each census area included within each service area, census TIGER files were used to develop maps, broken down by block group boundaries, of the counties within the study area.  The SFWMM service area boundaries were overlaid onto these maps, allowing land area portions within each block group located within each service area to be identified.  For the urbanized developed areas, population and housing data for block groups that straddle service area boundaries were estimated to be in proportion to the area breakdown between different service areas. For the less populated areas, data for block groups bisected by service area boundaries were apportioned in accordance with locations of communities known to be population centers.  This was necessary because in such areas it is significantly unlikely that population would be proportional to area.  For example, 10 percent of the area of a rural census block group could contain nearly 100% of the population.

For employment, 1990 county estimates were taken from The Florida Long-Term Economic Forecast, Volume 2, State and Local Counties (University of Florida, Bureau of Economic and Business Research, 1995).  The county level data were adjusted to represent SFWMM service area boundaries using the percent of population estimated for each service area at the block group level as a proxy for the employment percentage distribution.  Housing types for population in each area were estimated based on known categories from water use data collected from utilities.  Household income and persons per household data were based on census county-level data.

Projection of Use Estimates to the Year 2050

Background

The next step was to use the model to estimate water use to the year 2050, using economic and demographic projections.  The 2050 use estimate is necessary input for the planning evaluations being used to compare alternatives, which are based to a great extent on SFWMM simulation calculations representing estimated conditions, including M&I water use, in 2050.  M&I water use affects plan formulation and recommendations, because of the relationship between M&I water use and the other water needs of South Florida.  These effects are incorporated into the SFWMM simulation procedures and calculations, and are reflected in SFWMM output.

As explained previously, the 2050 use estimate is the only forecast estimate actually used as input in the SFWMM simulation runs.  The simulations which use 1995 M&I demand as input are based on USGS estimates of actual use.  The 2050 forecast estimate is part of an IWR-MAIN set of projections that include use estimates from the present through 2050.  Two sets of projections were made: a high projection (Projection A) based on levels of conservation in 2050 that are the same as now (i.e., no change in intensity of conservation); and, a conservation projection scenario (Projection B) based on more aggressive conservation measures.

Projection A

The Projection A demands for selected years are outlined in the following table.

Projection A – Average Forecast Estimated Total Use (MGD) For Indicated Year

Service Area

1995

2000

2010

2030

2050

N.P. Beach

62.02

72.44

88.56

96.39

101.25

SA1

216.02

251.66

306.45

332.82

349.20

SA2

294.99

328.86

380.19

407.13

422.24

SA3

367.36

424.28

511.43

562.82

577.00

Total

940.49

1077.24

1286.63

1399.16

1449.69


As discussed previously, the IWR-MAIN model verification procedure revealed that for the sampled utilities, the 1994 average residential use as predicted by the model was about 6% below actual use, considered to be reasonable.  USGS estimates available for 1995 total actual use, when compared with 1995 model projections for the entire area, differ similarly.  The model estimated projection for the 4 service areas, 940.49 MGD, is about 6% below USGS actual use data for 1995, just over 1,002 MGD, for the three county area.  Considering that about 1% of the three county area (Western Palm Beach County) use is not included within the projection area boundaries, the model estimate is closer to 5% below actual use.  USGS 1995 actual use estimate data for the three county area is shown in the following table.

USGS 1995 Water Use Estimates1/ for Palm Beach, Broward, and Dade2/ Counties (MGD)

Use

County

Three County

Category

Palm Beach

Broward

Dade

Total

Public Supply Domestic

124.85

144.57

245.76

515.18

Public Supply Commercial

28.88

38.56

60.68

128.12

Public Supply Industrial

5.12

6.54

11.93

23.59

Public Other

2.99

2.84

4.24

10.07

Public Use

25.04

29.79

49.92

104.75

Public Subtotal

186.88

222.30

372.53

781.71

Self Supply Domestic

17.19

2.16

12.71

32.06

Self Supply Comm/Industrial

22.95

0.34

43.38

66.67

Self-Supply Subtotal

40.14

2.50

56.09

98.73

Golf Courses

43.10

23.44

13.75

80.29

Thermoelectric

0.00

0.49

1.35

1.84

Other Irrigation (Comm)

8.28

28.64

3.04

39.96

Total

278.40

277.37

446.76

1,002.53

1/ Data is from Water Withdrawals, Use, Consumption, and Trends in Florida, 1995, USGS Water Resources Investigations Report 98-4140, and unpublished USGS data.
2/ Dade County demand includes Florida Keys (Monroe County); the Keys are supplied from Dade County well fields.

Projection B

In the Projection A estimates discussed and outlined above, the only increases in use efficiency are due to the effect of the increasing block rate structure used by most utilities (higher price per additional units of water used).  Projected future water consumption patterns also change because of changes in the socio-economic profile (employment mix, housing, income, etc.).

The conservation Projection B estimates reflect the additional reduction of water consumption by end use water fixtures.  IWR-MAIN estimates of water demand per employee and per housing unit includes the identification of twenty end uses, such as toilets, showerheads, bathroom faucets, dishwashers, and lawn irrigation.  Three categories of use are defined and used in IWR-MAIN.  Default values within the model for these three categories of use for toilets, for example, are: 5.5 gallons per flush for nonconserving toilets, 3.5 gallons per flush for conserving toilets, and 1.6 gallons per flush for ultraconserving toilets.

The Projection B conservation scenario use estimates are based on all new construction and remodeling using ultra-low flow water fixtures.  Those fixtures included in this conservation scenario are toilets, sink faucets, and showerheads.  These projections also assume the gradual changing over to ultra-low flow devices throughout the period of analysis so that by 2050, all fixtures in all units will be ultra-low flow devices.   In addition, Projection B includes conservation practices for lawn irrigation designed to achieve a 10% savings between conserving and ultraconserving housing units.   These practices are comprised of lawn irrigation allowed only during the period from 5 p.m. to 9 a.m., and the requirement that automatic sprinkler systems be equipped with rain sensors.

Area-specific usage rates for nonconserving, conserving, and ultraconserving fixtures, available from the larger utilities, were used in the projections for the 1990 base year.   Otherwise, the IWR-MAIN default values were used.  For the four service areas, the assumed levels of end use conservation measures in place in 1990 are outlined in the following table.

Assumed Percent Distribution Level of Conservation in 1990 by Housing Type

 

Housing Unit type

SA1/

Single Family

Multi-Family

Mobile Home

 

Conserving

Nonconserving

Conserving

Nonconserving

Conserving

Nonconserving

NPB2/

36%

64%

40%

60%

NA

NA

SA1

36%

64%

40%

60%

NA

NA

SA2

19%

81%

25%

75%

36%

64%

SA3

16%

84%

8%

92%

NA

NA

1/ Service Area
2/ North Palm Beach


The ultraconserving end use rates assumed for each service area are outlined in the following table.

Ultraconserving End Use Rates By Service Area

Service Area

Toilets

(gallons per flush)

Faucets

(gallons per minute)

Showerheads (gallons per minute)

North Palm Beach

1.6

1.6

2.5

Service Area 1

1.6

1.6

2.5

Service Area 2

1.6

2.0

2.5

Service Area 3

1.6

2.0

2.5


The Projection B conservation demands for selected years are outlined in the following table.

Projection B - Average Forecast Estimated Total Use (MGD)

For Indicated Year

Service Area1/

1995

2000

2010

2030

2050

NPB

61.17

68.85

80.09

82.59

83.66

SA1

213.24

240.26

279.90

289.57

294.18

SA2

284.98

308.54

341.71

346.39

345.72

SA3

359.55

403.00

464.06

482.49

474.80

Total

918.94

1020.65

1165.76

1201.04

1198.36

1/ NPB = North Palm Beach, and SA = Service Area.


The percentage reductions in total average use within each service area resulting from the Projection B conservation scenario estimates, compared with the Projection A estimates, are shown in the following table.

Reduction in Total Average Use Resulting from Projection B Conservation Scenario1/

Service Area2/

1995

2000

2010

2030

2050

NPB

1.37%

4.96%

9.56%

14.32%

17.37%

SA1

1.33%

4.53%

8.66%

13.00%

15.76%

SA2

3.39%

6.18%

10.12%

14.92%

18.12%

SA3

2.13%

5.01%

9.26%

14.27%

17.71%

Total

2.29%

5.25%

9.39%

14.16%

17.34%

1/ Extent to which conservation Projection B is lower than Projection A.
2/ NPB = North Palm Beach, and SA = Service Area.

Population and Employment Projections

Water demand forecasts using IWR-MAIN require, as input, projections of housing, employment, and income.  Housing projections have been derived from population projections.  Population, employment, and income projections are available from the Bureau of Economic Analysis (BEA), U.S. Department of Commerce, and from the Bureau of Economic and Business Research (BEBR), University of Florida, State of Florida.  The BEA and BEBR projections are based on different methodologies and assumptions, and as a result, differ in projected growth.  In general, the BEBR projections tend to reflect slightly higher growth than the BEA projections for South Florida.  The BEBR estimates tend to be closer to actual growth in Florida, and consequently have been used for planning purposes in Corps of Engineers water resource planning studies in Florida.   This represents a slight departure from the agency practice of using BEA projections for most feasibility studies and investigations.  This practice insures that to the extent projections affect study conclusions, all such studies will have been done using a set of nationally consistent projections.

The small area projections made available by BEA are governed by national and state control totals.  The use of BEBR projections instead of BEA projections only marginally affects study outcomes, at the same time making projections used in planning studies closer to the mark.  The BEBR county level projections follow the same control total principle at the state level, as that which is used in the county level BEA projections, relative to national and state totals.

At the time these water use forecasts were prepared, BEA projections were available at the county level to the year 2040.  Available BEBR projections extended to 2005.   The BEBR employment and population projections were extended to 2050 using BEA growth rates for these "out" years.  This resulted in a BEBR projection higher, but somewhat "parallel" to the BEA projected growth track in the later years of the planning period.  The 2040-2050 rate of growth was based on an extension of the BEA rate from 2020 to 2040, the last two projection years for the BEA county level projections.

The BEBR population projections used in this analysis, from The Florida Long Term Economic Forecast, Volume 2, State and Counties (University of Florida, 1995), represent BEBR’s medium range projections (low and high estimates are also developed in BEBR’s population studies).  In the case of Dade County, additional adjustments have been made reflecting growth closer to, but slightly below, the BEBR high range estimate.  These higher growth trends reflect the anticipated impact of the 1994 United States agreement with Cuba permitting increased immigration from the island.   These growth estimates have been made by the Metro-Dade County Planning Division, which has been carefully tracking ongoing growth and immigration trends for the county.

Estimated population and employment projections for the four service areas used to forecast water M&I demands for the year 2050 are summarized in the following table. The basis for the estimates is as described above.

Projected Population and Employment in 2050, Florida Lower East Coast Service Areas1/

 

Service Area

Category

North Palm Beach

Service

Area 1

Service

Area 2

Service

Area 3

Total

Population

342,758

1,245,829

2,203,306

2,952,565

6,744,458

Total Employment

131,147

481,845

947,333

1,141,435

2,701,760

Construction

6,655

24,169

40,840

42,012

113,677

Manufacturing

6,473

23,709

46,239

65,993

142,414

Transportation

4,205

15,702

40,386

85,717

146,009

Wholesale

5,783

21,913

60,133

100,112

187,941

Retail

27,348

99,456

172,438

191,864

491,106

Finance

7,503

27,505

54,029

76,041

165,078

Services

56,754

205,010

330,396

400,327

992,547

Government

13,516

49,886

106,453

160,023

329,878

Other

2,910

14,495

96,418

19,286

133,110

1/ Some totals may not add due to rounding.


Summary population and total employment projections for the service areas for selected years from 1990 through 2050 are shown in the following two tables.

Estimated Population by Service Area for Selected Years, 1990-2050

Service

Year

Area

1990

2000

2010

2030

2050

North Palm Beach

187,642

234,737

280,183

317,881

342,758

Service Area 1

691,746

860,090

1,021,462

1,156,557

1,245,829

Service Area 2

1,406,411

1,643,637

1,866,904

2,074,926

2,203,306

Service Area 3

1,719,946

2,037,509

2,466,588

2,814,351

2,952,565

Total Population

4,005,745

4,775,974

5,635,137

6,363,715

6,744,458

Estimated Employment by Service Area for Selected Years, 1990-2050

Service

Year

Area

1990

2000

2010

2030

2050

North Palm Beach

86,877

110,845

128,931

130,975

131,147

Service Area 1

322,456

409,742

475,138

481,719

481,845

Service Area 2

698,318

847,559

961,531

960,137

947,333

Service Area 3

836,935

972,367

1,150,017

1,174,510

1,141,435

Total Employment

1,944,586

2,340,513

2,715,617

2,747,341

2,701,760


The above projection estimates, along with associated companion income and housing projections, were used as input in the IWR-MAIN model, which resulted in the projected water demands by service area.

The county projections which are the basis for the service area projections are outlined in the following tables.

Estimated Population by County for Selected Years, 2000-2050

 

Year

County

2000

2010

2030

2050

Palm Beach

1,080,244

1,289,384

1,462,868

1,577,347

Broward

1,463,687

1,642,700

1,816,256

1,933,303

Dade

2,294,751

2,778,002

3,169,671

3,325,335

Total

4,838,682

5,710,086

6,448,795

6,835,985

Estimated Employment by County for Selected Years, 2000-2050

 

Year

County

2000

2010

2030

2050

Palm Beach

510,049

593,166

602,503

603,295

Broward

764,949

861,713

857,062

847,897

Dade

1,095,131

1,295,210

1,322,795

1,285,545

Total

2,370,130

2,750,090

2,782,360

2,736,737


As mentioned earlier, Monroe County water use is supplied by Dade County wellfields.   Monroe County population is currently just over 80,000, and is projected to reach 125,000 by 2050.

Average Per Capita Consumption

Per capita rates of use were not used as an input variable in this study, as explained earlier.  But the estimated future water use totals can be divided by the population projections for the respective service areas, revealing the resulting average per capita use.  The following table shows the average gallons per capita daily consumption for total average use resulting from the two projection scenarios.  There is variation among the different using sectors, and from area to area, but these overall averages provide an indication of the underlying trends resulting from the forecast procedures.

Average Per Capita Water Use Resulting From Projections A and B

(Gallons Per Capita Per Day)

Year

Projection A

Projection B

2000

226

214

2010

228

207

2030

220

189

2050

215

178


Projections Used in SFWMM Simulations

As discussed previously, the higher Projection A estimates include only nominal use efficiencies in the future, representing the effects of higher prices for incrementally higher levels of water use, and the level of conservation practices already in effect (i.e., with no changes in this pattern/profile of use).

A more realistic future water use scenario would likely include the effects of the SFWMD’s mandatory water conservation program.  The conservation projections (Projection B) discussed above reflect 100% participation in this program by 2050, including full use and effectiveness of ultra-flow devices, and result in forecast water use estimates that are about 18% below the Projection A estimate for 2050.  A more moderate application and effectiveness of the practices and techniques that make up the Projection B conservation scenario, would result in about two-thirds of this conservation, or estimated water use about 12% less than the 2050 estimate for Projection A.  This level of conservation in use was determined by relaxing some of the assumptions within the IWR-MAIN model, so that less than 100% of fixtures are operating with fully effective ultra-low flow criteria.

In the initial formulation procedure, the Projection A figures were used as SFWMM input.  The 12% reduction discussed above, a more credible and more reasonable scenario, was used in the final formulation to represent the 2050 without-project condition.  This level of water consumption and conservation represents the most likely future scenario without implementation of a comprehensive plan resulting from the Restudy.

Comment

These projections are based on fairly reliable relationships between water use and it’s known causative factors.  The confidence of long term projections of such key determinants of water demand as population and employment is diminished the farther into the future the projections are estimated for, and the smaller the area for which the projections are being estimated.  These projections have been made for both a relatively small area, and for a fairly distant future. Although they are "best estimates," it is important to recognize their limitations, in view of these observations .

The higher Projection A scenario presumes a continuation of existing use patterns.   If conservation practices were to be assumed in the projection scenario, consisting of ultra-low flow devices being in use in 100% of the area by the year 2050, this could result in 2050 forecast demand ranging from 15% to 18% less than Projection A. Both projection scenarios reflect a degree of use efficiency, as discussed above, and as demonstrated by the decline in per capita use during projection period.  Such figures and comparisons, while helpful in gaining an understanding of trends, should be viewed with caution since they include water use throughout the economy, and are broad average relationships whose true meaning is limited and perhaps impossible to really fully understand.  As mentioned earlier, such simplistic relationships as per capita average use rates are not used in the projection methodology of IWR-MAIN.

Finally, the projection scenario is only one of many possible outcomes for the population and economy of South Florida 50 years from now.  The scenario being used is believed to be a reasonable mid-range best estimate of the set of conditions expected to influence water use in the year 2050.  Consequently, the resulting range of estimated future municipal and industrial water use requirements is very likely a reasonable projection to use in this study.


Point of Contact:  [Bill Hunt]  904-232-1020
Page Created:  06/98
Page Last Updated:  07/02/98 

PageMaster

csflogos.gif (5223 bytes)