Loading...
HomeMy WebLinkAbout851240.tiff state population are lower than the Colorado Division of Local Govern- ment and OBERS forecasts. This demonstrates that the differences in the Denver metropolitan area forecasts are in large part explained by dif- fering Colorado projections. The Series 1 population forecast for 2010 is 2.3 percent higher than the Colorado Division of Local Government projections. The Series 3 projection is 3.4 percent greater than the state's forecast. The Series 2030 projection is 11.3 percent greater than the OBERS high sce- nario forecast for that year. The Series 1 , 2 and 3 forecasts are adopted for two study areas and are in close agreement with the Colorado Division of Local Government's numbers. However, the local projections are higher than the OBERS fore- casts. The next section discusses, in detail, the OBERS methods and assumptions. Discussion of OEMS. Projections from the U.S. Bureau of Economic Analysis were developed from historical data from 1969-1978 and do not reflect the Census Bureau's 1980 population count. OBERS represents a national analysis that essentially disaggregates national projections to local levels. OBERS projections utilized an interregional version of the export base methodology. National projections are first prepared, then projections for every state are generated and Standard Metropolitan Statistical Area (SMSA) projections are developed from state projec- tions. For Colorado, industries were disaggregated into "basic" and "service" industries. "Basic" industry earnings were projected based on the Colorado's share of corresponding industry earnings in the nation. "Service" industry earnings were then projected based on assumptions about the basic/service earnings ratio. Employment projections are sub- sequently developed from industry earnings based on historical earnings Appendix 2 222 851240 per employee. Population projections are developed from assumptions about the labor force participation rate and the percent of the popula- tion in the labor pool. The three OBERS scenarios of future Denver population levels differ in assumptions about the Denver Metropolitan Area's share of Colorado employment. Key assumptions of the OBERS forecast include: . The unemployment rate will be 4.8 percent in 1990 and 4.4 per- cent in the year 2000. . Productivity growth to 1985 will be 2.0 percent, after 1995 it will approximate 2.4 percent. . Regional migration differences will narrow (migration satu- ration). The importance of the regional migration assumption is the prime point of differentiation between OBERS and DRCOG or State of Colorado fore- casts. Basically, OBERS assumed a very rapid decrease in regional attractiveness for prospective employers and residents after 1990. Relative attractiveness of Colorado and Denver is not differentiated from other areas of the nation after ten to 20 years. The State of Colorado and DRCOG projections postulate that some regional attractive- ness will persist throughout the projection period. The OBERS population forecasts for Denver suffer from a number of possible flaws: . The lack of 1980 Census data for callibration of the OBERS model is a major short-coming. Dozens of assumptions such as labor force par- Appendix 2 223 ticipation rates, age distributions, and employment distributions begin at an erroneous point. Extrapolation of trends therefore compounds the initial error. . The export-base methodology does not reflect local economic pat- terns in Denver, now or in the future. Assumptions about which economic sectors are basic and which are local service are not representative of the Denver area or Colorado. . A constant relationship between basic and local service sectors is unrealistic. In a dynamic economic environment, these economic rela- tionships will shift continuously, and current relationships are likely to be quite different in distant time periods. . The three scenarios of Denver area population growth are based upon a strict mathematical extrapolation of Denver and Colorado employ- ment shares. This masks the more complex dynamics of the local economy. . The rapid erosion of Colorado's and Denver's competitive advan- tage and attractiveness is questionable. These forces are not entirely based on the size of a population center. Colorado and Denver have been attractive relocation alternatives for such reasons as climate and geo- graphic location. In sum, the OBERS forecasts were deemed inferior to the DRC00 fore- casts. CONCLUSIONS ABOUT THE ECONOMIC AND DEMOGRAPHIC FORECASTS The DRCOG Policy forecasts of population, households and employment through 2000 are acceptable forecasts for the EIS. Beyond 2000, Series Appendix 2 224 1 is the most reasonable and defensible set of economic and demographic forecasts. These findings are based on the following conclusions: . DRCOG's forecasting methodologies and models address specific economic and demographic influences on the Denver region. These projec- tions. are employment-driven and yet they are adjusted for reasonableness through the migration cap. . DRCOG projections of employment, population and household growth rates generally follow past trends since 1950 and recent estimates through 1983. . The underlying employment forecasts and assumptions about the future appear sound. The top-down forecasting approach from Chase Eco- nometrics (national) to the Governor's Blue Ribbon Panel (state) to the Denver Region is logical. Based upon a methodological review and com- parison with other forecasts, these national and state projections are acceptable as a basis for Denver metropolitan area forecasts through the year 2000. . The assumptions of the DRCOG forecasting model, including mor- tality rates, labor force participation and household size are sensible, considering Denver's unusual age and demographic composition and the continued influence of in-migrants. . The alternative series of projections tend to corroborate the DRCOG projections through 2000. Alternative projections for Denver are within 11 percent of each other. The Division of Local Government and DRCOG are in very close agreement. OBERS forecasts diverge considerably in future years. However, the OBERS forecasts exhibit several possible flaws. Appendix 2 225 . While certain economic or demographic assumptions such as fer- tility rates differ from other projections, they have only a minor effect on the overall forecasts. The pace of oil shale development is an example of an assumption with little affect on the Denver metropoli- tan area. . Series 1 beyond 2000 is the most reasonable since the employment and population growth rates are most consistent with the dominant views concerning the national growth perspective. Convergence of local growth with national trends is achieved gradually over the forecasting time frame beyond 2000. Series 3 clearly fails this test while Series 2 exhibits the same drawbacks to a lesser degree. . After this exhaustive study of economic and demographic fore- casts, there remains a number of key uncertainties which cannot be resolved definitively. The quality of life in the Denver metropolitan area might exhibit an evolution in the future unlike the historical experience. A related uncertainty lies with the relative competitive economic advantage of the Denver metropolitan area vis a vis other Front Range communities and other urban regions of the U.S. The EIS assumes that quality of life perceptions about this area will not change so drastically as to cause a significant deviation in the forecasts pre- sented. The inherent uncertainties of long term economic and demographic forecasting should be well understood. The farther out in the future forecasts are postulated, the greater the chance for shifts in assump- tions and fundamental demographic relationships which form the basis of the final forecasts. Appendix 2 226 ^ CHAPTER 7 SOCIOECONOMIC VARIABLE FORECASTS N` x CHAPTER 7 SOCIOECONOMIC VARIABLE FORECASTS INTRODUCTION This chapter of the technical appendix presents the forecasts for variables necessary to develop the water demand forecasts. The deriva- tion and evaluation of the projections for each variable is also pro- vided. The economic and demographic variables discussed are: . Median household income . SF lot size . SF and MF households . Service and nonservice sector employment . Number of days of measurable precipitation Appendix 2 227 . Marginal price of water . Unmetered SF dwellings . Watering restrictions In addition, the final section of this chapter examines the internal consistency of the individual economic and demographic variable assump- tions. FORECASTS OF MEDIAN HOUSEHOLD INCOME Personal income levels are a major determinant of water demand patterns in the EIS demand area. Median household income is a variable in the EIS water demand model which, with other socioeconomic variables, must be projected to predict total water demand. Forecasts of median household income for the EIS demand area and the methodology for render- ing these forecasts applicable to individual water suppliers is described. HISTORIC TRENDS NATIONAL TRENDS Historic trends in median household incomes throughout the United States are instructive for developing EIS demand area forecasts. Real or constant dollar household income levels increased approximately 3.1 percent annually between 1950 and 1970. The 1970s had fluctuating median household incomes, as depicted in table 90. Although the average annual growth rate between 1967 and 1982 was essentially stable, annual Appendix 2 228 Table 90 U.S. Median Household Income in Constant Dollars, 1967 Through 1982 Median Household Income in Annual Constant Percentage Year (1979) Dollars Change 1967 $19,458 1970 20,457 1.5% 1971 20,274 -0.9 1972 21,081 4.0 1973 21,514 2.1 1974 20,650 -4.0 1975 19,940 -3.4 1976 20,268 1.6 1977 20,369 0.5 1978 21,000 3.1 1979 20,716 -1.4 1980 19,547 -5.6 1981 19,074 -2.4 1982 19,029 -0.2 Average annual growth rate 1967-1982 -0.1% Source: U.S. Department of Commerce, Statistical Abstract of the United States 1981, 1981. Appendix 2 229 changes in median household income fluctuated from -5.6 percent in 1980 to 4.0 percent in 1972. Much of the lack of growth can be attributed to concurrent changes in the household structure. From the long term per- spective, household incomes in the United States increased 1.9 percent annually between 1950-1982. Incomes have increased from 1982 to 1984. Real per capita income growth from second quarter 1982 to second quarter 1984 averaged 3.0 per- cent per year. Five percent growth was experienced from 1983 to 1984. Table 91 shows these trends. LOCAL TRENDS Trends in median household income for Colorado and the Denver metropolitan area are similar to the national trends. Average personal income per household in Colorado increased 2.8 percent annually from 1960 to 1970 and decreased 0.1 percent per year during the 1970s (Colorado Blue Ribbon Panel, 1981). Median household income in the Denver metropolitan area increased 0.4 percent per year during the 1970s (DRCOG, 1982). INCOME PROJECTIONS OTHER FORECASTS Income forecasts for the U.S., Colorado, and the Denver region, shown in table 92, were reviewed to develop projections of real median household income for the EIS demand area. Appendix 2 230 Table 91 Real Per Capita Personal Income For The United States, 1982-1984' Per Capita Personal Annualized Income in Growth Year Quarter 1979 Dollars!/ Rate 1982 I $5,376 -- II 5,403 2.0% III ' 5,374 (2.1) IV 5,365 (0.7) 1983 I 5,400 2.6 II 5,460 4.5 III 5,487 2.0 IV 5,579 6.9 1984 I 5,694 8.5 II 5,731 2.6 1/ Adjusted by implicit p price deflator for personal consumption expen- ditures. Source: U.S. Bureau of Economic Analysis, Survey of Current Business, selected issues. Appendix 2 231 Table 92 Alternative Real Household Income Growth Forecasts For Selected Areas, 1980-2035 (percent) Public Blue Service United Ribbon Company of Period Bank Panel Colorado OBERS (Colorado) (Colorado) (Service Area) (Denver Metropolitan Area) 1980-1990 1/ 0.2 0.4 0.6 1.9 1990-2000 -- 1.0 1.2 1.9 2000-2035 Z/ -- -- 1.6 1.8 1/ 1978-1990 for OBERS forecasts, 1982-1990 for United Bank and Public Service Company forecasts. The later includes 29 of 63 Colorado Counties. ?/ 2000-2030 for OBERS forecasts, 2000-2008 for Public Service Company forecasts. Source: State of Colorado, the Governor's Blue Ribbon Panel, Colorado Investing in the Future 1981-2001 , Volume Two: Forecasts, July 1981. United Bank of Colorado, The Economy Over the Next 15 Years, November 1983. U.S. Department of Commerce, Bureau of Economic Analysis, BEA Regional Projections, Volume 3 Standard Metropolitan Statistical Area, July 1981. Public Service Company of Colorado Long Term Personal Income and Population Forecasts for the 29 County Service Area, March 1983 (based upon Data Resources, Inc. national forecasts). Professional judgment has been used to adjust and interpret the various data sources of income projections for applicability to the EIS. United Bank of Colorado projects 1982 to 1990 growth to be 0.2 percent per year. The adjusted Blue Ribbon Panel projections show 0.4 percent annual growth in personal income per household in the 1980's and 1 .0 Appendix 2 232 percent annual growth from 1990 to 2000. These forecasts are based upon Chase Econometrics national economic forecasts. Public Service Company of Colorado (PSCo) income forecasts for the PSCo service area exhibit the following growth rates: from 1982 to 1990, 1.2 percent from 1990 to 2000, and 1.6 percent from 2000 to 2008. The PSCo forecasts are developed from Data Resources, Inc. national projections of average earnings per employee. O.S. Bureau of Economic Analysis OBERS forecasts exhibit the most rapid increases in household incomes. Almost two percent annual income growth is forecast. These adjusted forecasts give a reasonably consistent range of alternative projections. RELATED ECONOMIC AND DEMOGRAPHIC TRENDS Employment growth, household size and aging of the population are trends which affect the incomes of EIS demand area households. A key index reflecting these changes is employment per household. Table 93 examines this ratio for 1980 to 2035. Table 93 Employment per Household for the Denver Metropolitan Area, 1980-2035 Employment/Households Year Series 1 Series 2 Series 3 1980 1.43 1.43 1.43 1990 1.44 1.44 1.44 2000 1.42 1.42 1.42 2010 1.31 .131 .131 2035 1.17 1.18 1.20 Appendix 2 233 Employment growth is strong from 1980 to 2000, declining from 2000 to 2035. Household size decreases substantially between 1980 and 2000 and is relatively stable from 2000 to 2035. As shown in table 45, the effects of these trends are largely offset from 1980 to 2000. Employ- ment per household declines from 1.43 to 1.42 during this period. Aging of the population into nonworking years is expected to reduce this ratio to 1.31 employees per household in 2010 and as low as 1.17 employees per household for Series 1 in 2035. Household income tends to increase with age of householder until 55 years of age. Median household income for the 55 to 64 age group is lower because of decreased participation in the labor force. Since most persons 65 years of age and over are not employed, incomes are much lower. As the Denver metropolitan area population ages into the 55 to 64 and 65 years and over age groups, income growth can be expected to moderate. EIS DEMAND AREA FORECAST The EIS median household income forecasts reflect the growth pro- jections reviewed and the influences of long term economic and demo- graphic changes in the Denver metropolitan area. The historical growth evident since 1950 is unlikely to be sustained. However, the slow growth experienced in the 1970s is expected to accelerate. Based upon recent O.S. per capita income growth, EIS demand area real median house- hold income is projected to increase 0.6 percent per year from 1982 to 1990. Annual growth from 1990 to 2000 is forecast to be 1.0 percent reflecting the increasing growth trend exhibited by the available fore- casts. Long-term growth is projected to stabilize at 1.0 percent per year from 2000 to 2010. Appendix 2 234 By 2035, 25.2 percent of the population would be 65 years of age or older under Series 1, compared with 22.9 percent for Series 3. The forecasts for each series reflect the effect of the age distributions on income levels. Annual household income growth for 2010 to 2035 is pro- jected to be 0.6 percent for Series 1, 0.7 percent for Series 2 and 0.8 percent for Series 3. These projections are far below the 1.8 percent annual growth rate projected in the OBERS forecasts but better reflect the local economic and demographic trends. Table 94 details the 1982 to 2035 projections for each series. Table 94 Median Household Income Projections for the EIS Demand Area Median Household Income Annual Growth Rate (in 1979 dollars) (percent) Year Series 1 Series 2 Series 3 Series 1 Series 2 Series 3 1982 20,000 20,000 20,000 -- -- _- 1990 21,000 21,000 21,000 0.6 0.6 0.6 2000 23,400 23,400 23,400 1.0 1.0 1.0 2010 25,900 25,900 25,900 1.0 1.0 1 .0 2035 30,000 30,800 31,600 0.6 0.7 0.8 Constant dollar median household incomes are expected to grow from approximately $20,000 in 1982 to $23,400 in 2000 for the EIS demand area. Incomes in 2035 are projected to be $30,000 for Series 1 , $30,800 for Series 2 and $31,800 for Series 3. Median household incomes were forecast for individual water sup- pliers the study area for application of the water demand models. The central assumption was that the current median household incomes of each water district would maintain a constant relationship relative to other Appendix 2 235 water suppliers in the EIS demand area. In other words, incomes in affluent areas will remain high while lower income areas, although experiencing gains, will remain low relative to other suppliers. SF LOT SIZE The lot size variable, or acreage per SF unit for each water supplier, was projected through 2035 for each water supplier in the EIS demand area. Sources for the lot size projection were: . Interviews with municipal and county planning officials and —. assessor's offices throughout the EIS demand area. Over 40 interviews were conducted to ascertain the viewpoints of local officials knowlege- able about residential development patterns and builder plans. . Review of land use plans and policies of various local govern- ments throughout the EIS demand area. A number of incorporated juris- dictions and counties follow land use policies which specify or limit the density of residential development. From these, the average number of SF units per acre can be derived. . Interviews with land developers for areas projected to grow rapidly, particularly in those water districts encompassing large planned communities. Appendix 2 236 . Information from water suppliers on projected SF densities for their service areas. Several recently formed districts have developed these forecasts. . An analysis of the DWD forecast of land use activities as part of their water forecasting system. These data, which extend to 2000, were evaluated or verified with local officials. . Examination of aerial photos of the EIS demand area to evaluate the potential buildout of areas and the reasonableness of the projec- tions. . Final review of new SF unit lot size estimates for reasonableness. For example, the average lot size for the incremental growth of SF dwellings was evaluated in light of past patterns and plans. Average lot sizes were not forecast to change significantly for slow growing areas. The average lot size of new SF development is forecast to be less than lot sizes associated with recent growth in the EIS demand area. Lot size projections must account for several, somewhat contradictory trends including: (1) suburbanization, which tends to increase average lot size; (2) higher densities in areas near the city center, attributable to high land costs; and (3) urban revitalization, where larger lots may be replaced with higher density units. As detailed in table 95, average lot size of new units is projected to decline from .24 acres per unit (about quarter acre lots on average) for units built between 1980 and 1990 to .20 (8,700 ft2 lots) for development between 2010 and 2035. • Appendix 2 237 Table 95 Average Lot Size of New SF Units Within the EIS Demand Area Average Lot Size for Future SF Dwellings Square Acres Time Period Feet per Unit 1980-1990 10,500 .24 1990-2000 10,000 .23 2000-2010 9,600 .22 2010-2035 8,700 .20 The current average SF lot size in the demand area is estimated to be .223 acres per unit (9,700 ft2), although new units built during the 1960s and 1970s were believed to be substantially larger, up to 12,000 square feet or more. Although new SF lot sizes are believed to be in a declining period, new lots in the 1980 to 1990 period are expected to be slightly higher than current averages. This will cause a slight increase in overall average lot size until 1990. This .223 average is projected to be relatively stable over the next 50 years, increasing slightly between 1980 and 1990 and decreasing after 1990. Differences among projection series are very slight so no differentiations are made among series. These forecasts are detailed in Table 96. Table 96 Average Lot Size of New and Existing SF Units Within the EIS Demand Area Average Lot Size Square Acres Year Feet per Unit 1980 9,700 .223 1990 9,900 .227 2000 9,900 .227 2010 9,900 .227 2035 9,700 .223 Appendix 2 238 The above data sources were utilized in preparing SF lot size projections for individual suppliers. Water suppliers which currently have small average lot sizes will experience a slight increase in the future as new growth on the periphery of these districts is more characteristic of suburban type development densities. Conversely, dis- tricts with currently large average lot sizes will decline as new developments result in higher densities consistent with maturing, urbanized areas. Average SF lot size is not necessarily indicative of average overall housing density, however, if the average SF lot size remains stable while the proportion of MY units increases over time, overall housing density will increase. SF AND HF HOUSEHOLDS The number of SF and HF households must be projected for the water demand models. The data sources for projecting the proportion of dwelling units which are SF for each water supplier in the EIS demand area are similar to those for lot size, namely interviews with planning officials, land developers, reviews of land use plans and development projections and information from water districts, including DWD land use forecasts. Appendix 2 239 EIS DEMAND AREA FORECASTS New housing construction is projected to become predominantly MF by 2035. This is likely to occur because of increasing land values, continued difficulties in qualifying for SF home mortgages, a continued high proportion of young, single adults as in-migrants and the increasing acceptance of common-wall dwelling units. As housing costs rise and interest rates remain high, prospective homeowners will experience increasing difficulties qualifying for a mortage. A chief remedy for this dilemma will likely be lower cost MF dwellings. MF and mobile home units, including townhouse development, are forecast to comprise 43 percent of new units built between 1980 and 1990, increasing gradually to 55 percent during the 2010 and 2035 period. These projections are examined in table 97. Table 97 Average Unit Mix of New Housing Units Constructed, 1980-2035, for the EIS Demand Area (percent) SF MF and Mobile Period Units Home Units 1980-1990 57% 43% 1990-2000 51 49 2000-2010 51 49 2010-2035 45 55 SF housing units currently comprise almost two-thirds of the total housing stock in the EIS demand area. As illustrated in table 98, this proportion is expected to decrease to 56 percent by 2035. Appendix 2 240 Table 98 Distribution Mix of Housing Units by Unit Type for EIS Demand Area SF MF and Mobile Home Year Series 1 Series 2 Series 3 Series 1 Series 2 Series 3 1980 65.5 65.5 65.5 34.5 34.5 34.5 1990 63.2 63.2 63.2 36.8 36.8 36.8 2000 60.6 60.6 60.6 39.4 39.4 39.4 2010 59.2 59.1 59.1 40.8 40.9 40.9 2035 57.1 56.5 56.0 42.9 43.5 44.0 Projections of EIS demand area SF and MF households are presented in table 51. Table 99 Households Residing in SF and MF Dwellings for the EIS Demand Area SF MF and Mobile Home Year Series 1 Series 2 Series 3 Series 1 Series 2 Series 3 1980 354,900 354,900 354,900 187,400 187,400 187,400 1990 470,500 470,500 470,500 274, 100 274, 100 274, 100 2000 573,200 573,200 573,200 372,800 372,800 372,800 2010 658,000 660,200 661,500 454,300 456,500 457,600 2035 740,600 774,000 803,800 555,200 595,700 631 ,400 COMPARISONS WITH NATIONAL FORECASTS SF housing units accounted for 69 percent of the United States housing stock in 1975. National projections developed for the U.S. Bureau of the Census for new construction, based on housing occupancy trends by age groups, are shown in table 100. Appendix 2 241 Table 100 Average Unit Mix of New Housing Units Constructed, 1980-2000 for the U.S. (percent) Series A-2 Series B-2 (Moderate (High Housing Consumption) Housing Consumption) New MF New MF New SF and Mobile New SF and Mobile Period Units Home Units Units Home Units 1980-1990 57 43 50 50 1990-2000 64 36 58 42 Source: Pitkin, J. and G. Masnick, Projections of Housing Consumption in the U.S. , 1980 to 2000, by a Cohort Method. Annual Housing Survey Studies NO. 9 U.S. Department of Housing and Urban Development, June 1980. SF units are expected to account for 57 percent of new construction from 1980 to 1990 under a high forecast and 50 percent under the moderate forecast, similar to the EIS demand area projection. SF units are projected to increase as a proportion of new construction for 1990 to 2000. The 1990 to 2000 Denver metropolitan area forecasts differ from national forecasts because of the relatively young age of household heads, the high proportion of single persons, a large in-migration component, and small household sizes which translate into greater demand for new MF units. Appendix 2 242 SERVICE SECTOR AND NONSERVICE SECTOR EMPLOYERS The projections of service sector and nonservice sector employees for the EIS demand area were derived largely from DRCOG forecasts through the year 2000, the last year which this regional planning agency projected employment by sector. Non-service employee forecasts were developed as the difference betwen total employment forecasts and ser- vice sector employment forecasts. For the six-county Denver metropolitan area, DRCOG projected that service sector employment would increase from 19.3 percent of total employment in 1980 to 20.5 percent of the total in the year 2000 (DRCOG, 1982). These proportions follow Governor's Blue Ribbon Panel forecasts for the State of Colorado and Chase Econometrics national forecasts. This gradual growth in proportional service sector employment was extra- polated through 2035, when 22 percent of total Denver metropolitan area employment is assumed to be in the service sector. As a proportion of total employment by place of work, service employment is currently about one percent less in the EIS demand area than in the Denver metropolitan area. Demand area employment by sector is developed from census tract information. EIS demand area forecasts of service employment mirror the trend in the proportion of service to total employment evident in the Denver metropolitan area as presented in table 101. Appendix 2 243 Table 101 Projections by Place of Work for the EIS Demand Area Percent of Service Employment Total Employment Year Series 1 Series 2 Series 3 Series 1 Series 2 Series 3 1980 144, 100 144,100 144,100 18.4 18.4 18.4 1990 206,300 206,300 206,300 18.8 18.8 18.8 2000 272,200 272,200 272,200 19.5 19.5 19.5 2010 304,100 306,000 307,900 20.0 20.0 20.0 2035 334,800 357,300 382,200 21.0 21.0 21.0 NUMBER OF DAYS OF MEASURABLE PRECIPITATION The weather conditions assumed for the 1980 to 2035 water forecasts are representative of average conditions since 1947. The number of days of measurable precipitation from May through September has averaged 44 days during the past 47 years, which is the extent of historical record for this summer rainfall measure at Stapleton International Airport. Measurable precipitation is defined as .01 inch or more in a 24 hour period. l-. Appendix 2 244 MARGINAL PRICE Because of the unconstrained nature of the water demand forecasts, the values of marginal price variable are held constant at the 1982 price for each water supplier in the EIS demand area. Water rates are assumed to increase only due to inflation. The only exceptions to this are those partially or entirely unmetered districts which currently have policies in place which will increase the proportion of units that are metered in the future. UNMETERED SINGLE FAMILY DWELLING Future metering status was determined based upon responses to the DRCOG Regional Water Study Questionnaire. Water suppliers which are expected to remain unmetered include Brook Forest, East Valley, Hazeltine Heights, Holly Mutual, Idledale, Maple Grove, Morrison and Silver Heights. Parker is currently metered but bills on a flat-rate basis. By 1990, all SF households in Parker will be billed for metered usage. No additional metering programs are assumed for Denver or Englewood for the unconstrained water demand forecasts. Historically, a small number of unmetered dwellings are metered or are lost from the SF hous- ing stock each year. The trends from 1974 through 1982 are expected to continue into the future as shown in table 102. Appendix 2 245 Table 102 Unmetered SF Dwellings in the EIS Demand Area DWD City and County Englewood Parker 1980 87,625 8,461 170 1990 85,200 8,380 0 2000 82,700 8,300 0 2010 80,200 8,220 0 2035 73,000 8,000 0 PRESENCE OF THIRD DAY, THREE HOUR OUTDOOR WATERING RESTRICTIONS Third day, three hour outdoor watering restrictions are assumed to not be implemented on a permanent basis for any water distributors in the EIS demand area. The third day, three hour restriction variable in the pooled water demand model is set at zero throughout the forecast period to reflect the unconstrained forecasting assumption. The third day, three hour restrictions only occurred in 1977. CONSISTENCY OF ECONOMIC AND DEMOGRAPHIC FORECASTS In addition to being reasonable on their own merit, economic and demographic projections, must be consistent with each other. Table 103 Appendix 2 246 C.. .. V .- e L O 1 V a Y o e °.y. H O • N 00 V c ° .4 0 0° O 139 p■ 2 o O O yq 0 ° a '0• a .... L 0 L a a °a A a a •0o. oc O O O ~7.Cre Ct.o p OO LFC0p•p ON�Lp. O `OA LLo.C La .4a y0A 2NF C ° 2 o• 0 o c o.G Ctl oo• N"2.. O L. pappp N.�i V y ° y a •D L 0 la L C O •.Oai L O a V.7. O I. O O 0 L a J c o O N ° 0. .{0 i i S C XI N y...• Y Y .0 ° a tl 0 ■ .~d O .y.a 0 0 0 0 0 .. 0 0 ° L L V .I L I C °a 0. .0 i.7 • GLOM O a 7 C ° ° y ay° 7 °ei • C °`'a L &80 ° 00 a to o.° 41.y CCo 04. L V C9 1. as L.0 m o ° y ,�. ° " °e e c c °e a o 0 .I. .°.. Lo.O C Cy .°. = goo 0.■ q0 OAL.4000 00 0 La 03ci .■. o.0 pop CV oy0ca O a°..a C O• 5 vg ca Vy S la 0O.°°.L..v Oo .°. CN 1 Ltl Otl C q $g. Cay L o °n 0c Olg N ..04 $ 00 0 L yCy V°0a Oe E I MStw°oo °° MEN. $ e' °.S Sooe 2'o �a e' 0 c453 o = ants t4 a° i S °° S : Sa 0 • a. 0 !.O p ° o ° O C a O 0O• L °0 0 8 0 a° a: w i' O O 3 .°. 0 .... L a p 0 L as y O O O at y C = A L ta OVUmI -.. L, L0 I) Na°. C ppC ° m o t oa O. D ~ O ° o ON O F Oy a a t .O y .. .� Z ■ 6 A ° Gq O a.L. p O O al ; uN La.woo L L 0 .°. O V a 88 • p m a 0 a LL V O Vp L t y L Ga LC a Va Cp • .tl. s. 0 L OOp 7 OOL CCp ►. OOA .4 1. a 0 0 .4 p0 lO.a w ■ O.p 0 CQ fl d TYUG qt a C C ; € V22O ZL C , .. y O C m a. tl {{ ~ p ..I I1 • S N ° a pM L a 0 �C m O ° 8 ° w .i ~a 0.. 0 0 4. 0°a o o C a a °� '3 a $1 0 0 • o `be �E �5�Y :ye5� Say e55� 08 $ 80. oa 00 0 O.�tla aN°m n ie n o a g � c°� e o c a L. SSi 8o S $ Se 8g al 8e t. 8e SN e 3 $ S ° 3 0 3 0 0 • 2g $ .°. .° S N a° . 3 g o E o °e 5 Oi a^a y o g a r 0 if ₹0 p. - rL0 ~ S wa .. LL 3 a i otl mp .°y N .Y°+ aO O +aCi Lo Va o Y O .i2 T a f s %a $ W OF •to 7 L. r8¢ Appendix 2 247 summarizes each key component of the EIS demand area forecasts. Many of these are forthcoming from DRCOG. DRCOG PROJECTIONS AND ASSUMPTIONS DRCOG developed projections of fertility, mortality and labor force participation rates largely based upon, but not coincident with, U.S. Census Bureau and U.S. Bureau of Labor Statistics projections. DRCOG then makes adjustments for Denver's special conditions. Fertility Rates. The apparent contradiction between increasing fertility rates and the other demographic assumptions stem from the traditional demographer practice of utilizing total fertility rates as compared with other fertility rate measures. In fact, DRCOG assumptions about increasing total fertility rates result in a future downward trend in children per household and children per employee as indicated in table 104. Table 104 Children Per Household and Children Per Employee, 1980 and 2000 1980 2000 Persons under 10 years of age 234,500 351,700 Persons under 20 505,900 679,400 Households 608,400 1,074,300 Persons Under 10 per household .39 .33 Persons under 20 per household .83 .63 Employees 869,500 1,524,700 Persons under 20 per household .58 .45 Source: Based on DRCOG data, 1983. Appendix 2 248 r The above data are based on DRCOG age distributions. Hence, these trends are consistent with declining household size and increasing employees per household. Labor Force Participation Rates. According to DRCOG, these rates will follow, but not converge upon the national trends through 2000. This disparity is attributable to high levels of employment-related in- migration and higher participation rates for older persons. Increasing employment per household and growing proportions of age-cohorts with high employment tendencies are consistent with the expanding labor force participation rates. Rates for 2010 and 2035 are based upon U.S. Bureau of Labor Statistics participation rates by age cohort. Migration. DRCOG migration forecasts are in part established external to the employment model, and this is an inconsistency in the DRCOG projections. Because of this, the employment and population models are adjusted downward to conform with the migration assumption. Migration is fully discussed in the previous chapter. Household Size. DRCOG projects household size based in part upon census forecasts of household formation rates by age group. Household formation rates account for increasing fertility rates and decreasing mortality rates. Household size is also dependent upon migration levels since in-migrants tend to be concentrated in high household forming age groups (young, single adults). Each of the U.S., Colorado and Denver forecasting series identified for comparative evaluation indicate declining household size, increasing total fertility rates and decreas- ing mortality rates. Appendix 2 249 OTHER PROJECTIONS Median household income, SF lot size, and housing mix forecasts are consistent with DRCOG projections. Median household income is forecast to exhibit slow growth in the 1980's, increasing through 2010, and declining as the population ages into retirement years. The early increases are consistent with smaller household size, expanding employment opportunities on a per household basis and higher labor force participation. Fertility rates also increase with rising economic expectations. SF lot size is forecast to decrease for new units based upon the increasing absorption of land in proximity to the city center and available land use plans. As a finite resource, urban area land is likely to rise in real price faster than the moderate levels of real personal income growth projected. Hence, homeowners will be encouraged to accept smaller lots. Smaller household size will also be consistent with smaller lots and more MF unit living. The average lot size for all SF dwellings initially increases because of the location of growth. The proportion of dwelling units which are SF is projected to decrease. This is attributable to the same basic justification for the trends in lot size. Again, this is consistent with the forecast age distribution, income growth, and household size. Appendix 2 250 CHAPTER 8 EATER DEMAND FORECASTS FOR THE EIS DEMAND AREA j i i � CHAPTER WATER DEMAND FORECASTS FOR THE EIS DEMAND ►RBA CHAPTER 8 WATER DEMAND FORECASTS FOR THE EIS DEMAND AREA INTRODUCTION This chapter presents water demand forecasts for the EIS demand area. These forecasts are based on two water demand models detailed in Chapters 4 and 5 combined with forecasts of the demographic variables in Chapter 6 and the socioeconomic measures in Chapter 7. Figure 13 illus- trates the respective water demand forecasts for the EIS demand area. These forecasts have not been adjusted for natural retrofit, which is discussed in this chapter. Appendix 2 251 O LY W J O J W O W O O W U O 6 f O LLL.r Ln O N O - O N O _ N O .c Y Ln L LO N O NN Y LT en in r1 } v • O L L W O1LI.Q LL y C C C O O O O L In _ O 4J O s N O _ 01 01 O 1 I 1 I I I I 001 O O O O O O O O O O r 01 O O O O O O 00 O n ID !11 tt II N ;aa}—any jo spuosnogj Appendix 2 252 Because of their importance, certain qualifications about the water demand forecasts are reiterated. A careful understanding of precisely what the EIS water demand forecasts represent is essential to interpre- tation or application of the results. First of all, water demand fore- casts are uniformly expressed as the amount of water delivered to the end user. Distribution system losses or other losses from the point of diversion are not accounted for. Unconstrained demand is defined as extrapolation of conditions which currently exist. Prices for water are assumed to remain constant in real terms. Weather conditions are normalized so that forecasts represent average weather conditions. Con- servation programs and practices currently in existence, such as the ET Program, are assumed to have the same effect on future consumers as they have on present consumers. These qualifications are fully discussed in Chapter 1 of this Appendix. POOLED MODEL WATER DEMAND FORECASTS FORECASTING PROCEDURES The pooled model forecasts utilized the following information: . Pooled, aggregate water demand model derived in Chapter 4. . Projection of independent variables specified in the model for each supplier. Appendix 2 253 r . The 1981 and 1982 residuals, weather adjusted, for each supplier to adjust for existing conservation activities and any other unique individual water supplier characteristics (a residual is the difference between actual demand and demand estimated by the model). . Water savings for natural metering of existing unmetered dwell- ings (not related to additional conservation programs). The pooled model forecasts were developed for each year on a supplier-by-supplier basis. The first step in the forecasting procedure was to project g.h.d. for each water supplier for 1990, 2000, 2010 and 2035. G.h.d. was calculated by multiplying each coefficient for each independent variable by the projected values of each independent variable: household income, average SF lot size, percent SF household size, service and nonservice employees per household, weather (normalized), presence of watering restrictions and marginal price. The variable coefficients as well as the constant term were determined through the historical water demand modeling exercise discussed in Chapter 4. Derived g.h.d. is adjusted by the 1981-1982 g.h.d. residuals for each water supplier and for natural metering. Once g.h.d. has been projected for each water supplier, demand is forecast by multiplying g.h.d. by the number of days in the year and by the number of households projected for each individual supplier. These annual water supplier water demands are then summed to calculate total EIS area water demand. r Appendix 2 254 SUMMARY OF FORECASTS FROM THE POOLED WATER DEMAND MODEL TOTAL WATER DEMAND Water demand forecasts have been prepared for population projection series 1, 2 and 3 applying the pooled water demand model. Under all series, water demand supplied by public water systems is forecast to grow about 235,000 acre-feet by the year 2000 reaching 548,000 acre-feet in that year. Publicly supplied demand is expected to increase to 789,400 acre-feet by the year 2035 under Series 1 . As shown in table 105 and subsequently explained, the annual rate of water demand growth is projected to decline from 1980 to 2035. Table 105 Publicly Supplied Water Demand in the EIS Demand Area, 1980 to 2035, Based upon the Pooled Model Year (percent) 19801/ 1990 2000 2010 2035 Series 1 Acre-feet (000's) 308.1 420.3 547.5 654.1 789.4 Average Annual Growth Rate -- 3.2 2.7 1.8 0.8 g.h.d. 507 504 516 525 543 g.c.d. 191 210 221 226 234 Series 2 Acre-feet (000's) 308.1 420.3 547.5 656.9 833.6 Average Annual Growth Rate -- 3.2 2.7 1.8 1.0 g.h.d. 507 504 516 525 543 g.c.d. 191 210 221 226 234 Series Acre-feet (000,$) 308.1 420.3 547.5 658.9 875.2 Average Annual Growth Rate -- 3.2 2.7 1.9 1.1 g.h.d. 507 504 516 525 545 g.c.d. 191 210 221 226 235 1/ This column is not directly comparable to actual 1980 consumption, which was 316,000 acre-feet. The above figures assume average weather conditions, conservation programs existing in 1981 and 1982, and 1982 marginal price for water. Appendix 2 255 Series 1, 2 and 3 forecasts diverge from 2000 to 2035 because of the differing underlying economic and demographic projections. Water demand forecasts for 2010 vary by less than one percent among the series. By 2035, however, Series 3 forecasts are 86,000 acre-feet, or 11 percent higher than the Series 1 forecasts. G.h.d. remains fairly steady between 1980 and 2000. G.h.d. increases slightly after 2000. A number of different trends in the Denver metropolitan area bring about changes in g.h.d.: . Increasing income levels in the Denver metropolitan area tend to increase future water demand. Incomes are projected to increase more rapidly from 1990 to 2010; income growth has the greatest effect on water demand during this period. . Growth in the employment base and business activity will tend to increase area water demand expressed in g.h.d., particularly for those water suppliers projected to be centers of future economic growth. . The projected shift of new residential construction from SF units to MF units will tend to reduce water demand in the EIS demand area. . Although lot size for new SF units is expected to decrease, overall projected average lot size of SF units is representative of the existing development. This is attributable to the large base of existing SF lots which are assumed to remain at their current size under the unconstrained forecasting assumption. Trends vary, however, among individual suppliers as discussed in Chapter 6. r Appendix 2 256 . Average household size declines from 1980 to 2000, stabilizing after 201O. This trend reduces consumption per household through 2000. . Several factors do not change over time for the unconstrained forecast. Constant average weather conditions are forecast for each year and no imposition of three day, three hour maximum outdoor watering restrictions is assumed. Water conservation programs existing in 1981 and 1982 are assumed to continue into the future. Marginal price for water is also assumed to remain constant, after adjusting for inflation. When considered simultaneously, these influences tend to slightly reduce demand per household from 1980 to 1990 and moderately increase demand per household from 1990 to 2035. FORECASTS BY SUPPLIER The primary purpose of the demand forecasting task is to project aggregate water demand for the EIS demand area as opposed to detailed forecasts by water supplier. Even so, the approach adopted for the EIS forecasts requires individual water supplier projections, recognizing the characteristics of these local areas, and an accumulation of these to achieve demand area totals. Exhibit 2-A appended to this report pro- vides water demand forecasts for individual suppliers by population series and demand model. Analysis of future water demand for groups of suppliers is useful in demonstrating the changing geographic pattern of water use likely over the next 50 years. As table 106 indicates, water demand (unadjusted for natural retro- fit) is projected to increase slowly in the DSA from 1980 to 2035. Demand increases to 283,000 acre-feet by 2000 and slows thereafter; demand in 2035 is forecast to be 369,400 acre-feet. The DSA's share of total publicly supplied water decreases from 61 percent to 42 percent Appendix 2 257 a rim a CO aa1 t11 CO taCO Ft()N. 0 .i a1 V I O O CO ( P. O ON O. Cr* ? �a N C O L . N N N L7 C C L O a O O.ra. 7 O .i O CI 20 r sal "'I L C 9 C C1 4.3 Q L t. 42 O .01 0 Cl 0 ty O 0 0 0 0 0 0 0 0 0 0 a co 7. Cl Oy1 0 0 0 0 000 000 C1 a O k. N O. # M N O1 in - y C y < 0 A '4 n .C SO. N IAA a N C P.O N m 6 •- N N U 7 CO as a1 .i C! 7 C m TI .. O C C 01 a C) O 59-I 41 el ei O 4.1 "'1 a > .i 0 20 CU m .0 0 .O Oa Oa a% C-Co 0a g 'd a .C L 7\ reit O E,OE_, 0 0 0 000 .- •- 4.4 a f. t 0•-1 L en en M MMM MM M s. 0 0 0 a1 CL. L y a. O aa) C) L 2 0 v-I..Oj aL1 DI CO a 2 7 0 +1 0 m y01 C. Oa CO L L a L aads1 CI O 4i a1 0 C1 f' 0 0) O O O 000 000 C O X F C7 a1 10 d1 0 0 0 0 000 000 a.1 .-I a al a M W N S CO CDS . O 0 O CO 0 V �,.O N S '0 Z e ? .O 2- N M S e- .O at a m as 0 7 C O L Oa N .0 000 In.0t- as r O O L .3 r.-1 O NNN NNN O 0 a0 k W 40. 0 CJ O. 4 f" .-i CO .Oi 4 o O 7 a s O ^+ mC 0 Iv C. N 4 > O L W • C. -^ .0 fa T O U d C fi C) .O O ' N 44-. L .5 1 .i \ d 0) �C e d a1 C7 8 d NI a+ in W vCi in O d.C1 m 7 C a.1 M in M N .- .O b e- A a1 al a1 C 20 0 0 • Y5 O .i C1 PLC L .i 0 0 F .O (A IA n 4? in • .1 Oo7 5. a 551 e1 O B a) O L y 4r a min m 0 N 43 4.3 SF 4.2 O 0 y0 a .1 0aO i > (ICp da CO al > 6 NaOa L. a .-1 as O > N F el 7 -1 0 O 0 O N Cl a1 +7-10 C a 7 � • .C > 2 7 C y O O 0 0 0 0 000 000 Oa'1 .a 4 �E OCNO e 0 0 0 0 O O O 0 0 0 C O O O N '0 O Oa CO a 4. N to en in T M .O N . .i C .-1 2 d 0 L L>D f0a 0 ...3O1 V O C) N in Men.* r 1I1 as. U L co M CO NNN .0 .O .Oa as > O] a C O a e- N N MMM MMM C .00. .a 1 C 01 .1 m 01 0 L .a Pal _ W .ai .ICi > 0 CO1 m CO p 7 U O 0 r• a L .a1 Oa U e a O O .i L a .i al O 0 .i v) 0 COO � G a1 a1 m el ...iii 5.-1 .4 CV M •- N M O C ..Ci .0 > • '-0 a4.a ref CO/ L es: C 6 ^'I pC1p 1 IC . I V .ei O O. L C co O1 C CO G 0) .Oa m C. .�� gp a in g a s a Caa) S d l L O Y N 7'I PI e m 0 .Ci C m z° m L 0 0 0 0 IA 01 a a 0 0 0 eI NI mi SI Appendix 2 258 over this period. These DSA forecasts include the City and County of Denver, total service, read and bill, and those master metered suppliers relying solely upon the DWD for treated water supplies, as previously defined. The DSA forecasts do not include additional DWD water sales to other area water suppliers which are accounted for as individual enti- ties. Water demand for the other major water suppliers within the demand area is forecast to increase to almost 168,000 acre-feet by 2000, a 78 percent gain over 1980. Year 2035 water demand increases to 280,000 acre-feet. These suppliers' share of total water demand is relatively stable from 1980 to 2035. Water demand in currently sparsely populated water suppliers in the demand area is projected to grow considerably in future years. Water demand will increase by 72,000 acre-feet by the year 2000 to 97,500 acre-feet. While these suppliers represent only eight percent of current water demand, their share will expand to one-quarter of total publicly supplied use by 2035. 0.4E FACTOR WATER DEMAND W)DEI. FORECASTS FORECASTING PROCEDURES The use factor model forecasts applied the following information: . The use factor model derived in Chapter 5 Appendix 2 259 . Projections of demographic parameters for each supplier . Projections of independent socioeconomic variables related to SFM for each supplier The use factor model forecasts were developed on a supplier-by- supplier basis. SFM water demand by supplier was calculated by multiplying each socioeconomic variable coefficient by each supplier- specific projections of that variable. The result served to adjust the overall 478 g.h.d. SFM use factor. Other sectoral use factors were multiplied by supplier-specific forecasts of those demographic parameters. Finally, water demand from each sector was accumulated to derive total demand for the supplier for that year. SUMMARY OF FORECASTS FROM THE USE FACTOR MODEL The use factor model forecasts water demand for water suppliers in the EIS demand area using a model disaggregated by consuming sector. Results of applying the model to the projected independent variable data base are summarized in figure 13. As shown in the application of the model to historical data, predictions for the overall EIS demand area are expected to be more accurate than predictions for individual supplier's demand. Table 107 presents a summary of the use faactor model forecasts and the per unit trends for g.h.d. and g.c.d. Appendix 2 260 Table 107 Publicly Supplied Water Demand in the EIS Demand Area, 1980 to 2035, Based upon the Use Factor Model Year (percent) 1980 1.9M 2000 2010 2035 Series 1 Acre-feet (000's) 313.7 413.6 514.6 597.6 703.0 Average Annual Growth Rate -- 2.8 2.2 1.5 0.7 g.h.d. 516 496 486 480 484 g.c.d. 195 207 208 207 209 Series 2 Acre-feet (000's) 313.7 413.6 514.6 599.9 739.8 Average Annual Growth Rate -- 2.8 2.2 1.5 0.8 g.h.d. 516 496 486 480 482 g.c.d. 195 207 208 207 208 Series Acre-feet (000's) 313.7 413.6 514.6 601.5 773.6 Average Annual Growth Rate -- 2.8 2.2 1.6 1.0 g.h.d. 516 496 486 480 481 g.c.d. 195 207 208 207 208 These forecasts exclude the effects of natural retrofit. Little variation in demand is noted between the three series for the projected years. The average annual growth rate does show a vari- ation. The average annual growth rate for all three series for the cur- rent 20-year period (1980-2000) is projected at 2.2 percent. A slower growth rate is projected for the subsequent 35 years (2000-2035) . Series 1 projects an average annual growth rate of .89 percent; series 2-1.04 percent; and series 3-1.17 percent. Appendix 2 261 The distribution of future water demand by user class is shown in table 108. Proportionate demand by SFM households and MF households will increase. Demand by SFF households will decrease since there is no projected growth in the number of unmetered households. Commercial/industrial and public demand will remain essentially unchanged. The percent of total water demand by SFM and SFF households will remain constant at 62-65 percent for the entire period. Table 108 Distribution of Future Water Demand by User Class (Percent)1/ Projected Consuming Sector Current 1990 2000 2010 2035 SF (metered) 42 48 49 51 55 SF (flat) 23 16 13 11 8 MF 16 16 17 18 19 Commercial and industrial (employment) 12 13 14 13 11 Public (population) 7 7 7 7 7 Total 100 100 100 100 100 1/ Series 1 data. Series 2 and Series 3 yield similar results. FORECASTS BY SUPPLIER Individual water supplier forecasts are presented in exhibits 2-A. Water demand for the City and County Denver, the largest district, is projected to remain essentially unchanged for the 35 year period, 2000- 2035. In contrast, Aurora's water demand is projected to grow at an average annual rate of .73 percent during the same time period. Appendix 2 262 CRITICAL ASPECTS OF OSE FACTOR FORECASTS Three sensitivity analyses were performed using the series one pro- jection data. They pertained to the following: . Marginal price. . Income . Lot size model coefficient. . Weather. The marginal price analysis examined the effects of changes in mar- ginal price on water demand in the EIS area. Three alternatives were tested with the results shown in table 109. In the first alternative, marginal price is reduced for every supplier by 20 percent. This resulted in an overall increase in demand of 3.5 percent. Conversely, a 20 percent increase in marginal price resulted in a decrease in demand of 3.4 percent. A 40 percent increase in marginal price resulted in a 6.9 percent decrease in demand. Total water demand is also moderately sensitive to changes in the income forecast. A ten percent increase or decrease in household income would result in a 1 .9 percent change in demand. Appendix 2 263 Table 109 Sensitivity Analyses of Series 1 Data Percent Difference EIS Demand From Variable Sensitivity Area in 2035 Base Case (thousand acre-feet) Base case 703.0 -- Marginal price Reduce marginal price by 20% 727.8 3.5 Increase marginal price by 20% 678.8 -3.4 Increase marginal price by 40% 654.9 -6.9 Income Increase income by 10% 716.3 1.9 Decrease income by 10% 689.7 -1.9 Lot Size Double lot size coefficient to 854 gallons/acre/day 701.9 -0.2 Weather High irrigation requirements 773 +10.0 Low irrigation requirements 633 -10.0 The lot size coefficient sensitivity analysis was prompted by con- cern that the implied elasticity of water demand for lot size was lower than intuition might suggest. However, table 109 shows that even doubl- ing the coefficient results in little change in overall demand, in this case a drop of 0.2 percent. Finally, the sensitivity of the forecasts to certain weather variations was analyzed. The model was established to forecast demand for a typical weather year or expected average demand over a series of years which together represent average weather conditions. However, should conditions be significantly different from average for several consecutive years, it is important to estimate the resulting effect on demand. Appendix 2 264 Water demand estimates were made for a year with high irrigation requirements (drier than normal) and a year with low irrigation require- ments (wetter than normal). The definition of the high and low year are such that 95 percent of historical weather data would fall in between the two cases. Using the results of the weather analysis discussed in Chapter 3, it is expected that these weather conditions would increase water demand up to 10 percent in dry years and decrease it as much as 10 percent in wet years. Since this assumes that MF, E and P demand are affected to the same extent as SF households, actual variation would probably be less. SELF-SUPPLIED WATER DEMAND PROJECTIONS OF SE<.F-SUPPLIED WATER DEMAND Self-supplied water demand from industrial companies is forecast to increase from 32,610 m.g. or 100,100 acre-feet in 1980 to 38,590 m.g. or 118,400 acre-feet in the year 2000. As presented in table 110, demand is expected to remain essentially unchanged from the year 2000 through the year 2035. The major future self supplied water users are Public Service Company of Colorado, Adolph Coors Company, Gates Rubber Company, sand and gravel companies, and agricultural users. This use is in addition to water these companies might consume which is supplied by public water districts. Appendix 2 265 Table 110 Forecast of Self Supplied Water Demand of Major Industrial Consumers (m.g.) Public Service Adolph Gates Sand and Total in Company of Coors Rubber Gravel Acre-feet Year Colorado Company Company Companies Total (000's) 1980 18,000 14,200 300 120 32,620 100.1 1990 15,000 20,000 150 130 35,280 108.3 2000 12,300 26,000 150 140 38,590 118.4 2010 12,300 26,000 150 150 38,600 118.4 2035 12,300 26,000 150 170 38,620 118.5 Demand for the presently self-supplied households in the demand area, numbering approximately 10,100, was assumed to remain at 1982 demand levels throughout the forecasting period. These household are assumed to continue relying upon groundwater supplies. NATURALLY OCCURRING RETROFIT Both publicly supplied water demand forecasts presented in this chapter have not been adjusted downward for naturally occurring retrofit. Based upon the 1983 Household Users Survey, an estimated 434,000 housing units currently lack water-conserving plumbing fixtures. However, such fixtures are standard for replacement of old toilets and shower heads. Since the expected life of existing high water using shower heads is about 15 years (U.S. Department of Housing and Urban Development, 1982); all inefficient fixtures are assumed to be replaced by low use fixtures by 1995. r Appendix 2 266 The life of a toilet is about 40 years (U.S. Department of Housing and Urban Development, 1982) . By 2020 all existing high water-using toilets will be replaced by low-use facilities. Water savings for shower head retrofitting is 13.5 g.c.d., and toilet retrofitting saves 9.0 g.c.d., according to the preliminary conservation analysis of this EIS. Total savings are calculated by multiplying daily per capita savings by average household size and by the number of naturally retrofit households. These data are provided in table 111. Table 111 Assumptions for Calculating Water Savings from Natural Retrofit Additionally Per Capita Average Naturally Daily Water Persons Retrofit Savings per Household Households Shower Heads E. 1980 13.5 2.65 0 1990 13.5 2.39 289,000 2000 13.5 2.34 434,000 2010 13.5 2.32 434,000 2035 13.5 2.32 434,000 Toilets ?/ 1980 9.0 2.65 0 1990 9.0 2.39 108,500 2000 9.0 2.34 217,000 2010 9.0 2.32 325,500 2035 9.0 2.32 434,000 1/ Assumes conversion from 6.0 gpm to 3.0 gpm shower head, 6 minute showers and 0.75 showers per person per day. ?V Assumes conversion from 5.5 gallons per flush to 3.5 gallons per flush, 4.5 flushes per day. Source: EIS Conservation Analysis. Appendix 2 267 As shown in table 112, adjustments for natural retrofit increase from 13,100 acre-feet in 1990 to 25,400 acre-feet by 2035. Table 112 Acre-Feet Adjustments for Natural Retrofit Annual Year Water Savings 1980 0 1990 13,100 2000 20,500 2010 22,800 2035 25,400 Water demand forecasts can be adjusted for natural retrofit for consistency with the unconstrained water demand forecast definitions. COMPARISON OF POOLED & USE FACTOR MODELS The pooled model predicts g.h.d. for each water supplier based upon the following independent variables: . lot size . family income . household size . marginal price . service employees/household . nonservice employees/household Appendix 2 268 . percent SF households . number of days measurable precipitation . presence of three-hour/three-day restrictions (0 or 1) The pooled model is based on ordinary least square regression using water demand data at the indiviudal supplier level and DECOG, Census, and other local data for the independent variables. The use factor model predicts total gallons for each water supplier based upon the following data for each supplier: . number of metered SF homes . number of unmetered SF homes . number of MF homes . number of employees . population Metered SF demand is modified to reflect the following socio- economic data: . marginal price . lot size . household size . family income The use factor model is based on analysis of nonresidential use factors, household water demand, and lawn irrigation requirements. The use factor model is also based on ordinary least squares regression using household survey data to incorporate the socioeconomic data into the projections. Appendix 2 269 BAC[CASTING CAPABILITY The performance of the two models on historic data is presented in table 113. Table 113 Performance Comparison of Use Factor and Pooled Model on Historical Data ?/ Use Factor Parameter Model Pooled Model Percent variation of overall district demands explained 99.8 99.5Z/ EIS demand area accuracy (percent) -1.4 2.0 DWD demand (excluding master meters) accuracy (percent) 6.9 3.9 Average absolute error all providers (percent) +23.5 +8.9 ?/ Both models were applied to the historical data base from which they were derived. The use factor model was analyzed on 48 observations while the pooled model was analyzed on 254 observations. ?1 The percent variation previously reported in Chapter 4 was for the equation unit demand (g.h.d.). This table shows comparisons for: . percent of explained variation--This is calculated in a manner similar to the coefficient of correlation. . aggregated demand accuracy--This is the total actual demand for all observations minus the total calculated demand, expressed as a per- cent of actual demand. Appendix 2 270 . DWD demand accuracy (except master meters)--This is calculated similarly to aggregated demand. . average absolute error for predicting individual suppliers (from the largest to the smallest)--This shows the absolute value of the residuals for all observations. The two models perform similarly in most respects on the historical data. The pooled model performs somewhat better for individual dis- tricts, due largely to its residual adjustment feature. A comparison of the elasticities for the two models is shown in table 114. Table 114 Comparison of Elasticities for Use Factor Model and Pooled Model Use Factor Pooled Model 1/ Model 1/ Marginal price -0.34 -0.15 Income 0.26 0.41 Lot size 0.22 0.30 Household size 0.22 0.21 Percent SF 2/ 0.06 Service sector employment per household 2/ 0.08 Nonservice sector employment per household 2/ 0.09 Days of measurable precipitation 2/ 0.32 Presence of 3 day/3 hour restrictions 2/ 0.01 1/ Elasticities for the use factor model apply to SF metered demand while those in the pooled model apply to overall g.h.d. 2/ Not in use factor model. Appendix 2 271 It is difficult to compare the elasticities directly because the use factor model elasticities apply to SF use while the pooled model elas- ticities apply to all use. Both models exhibit elasticities for marginal price that are consistent with those found in the literature. Given this difference, the use factor model elasticities for income, lot size, and household size are less than the pooled model elasticities if they were both on a total water demand basis. A primary difference in the two models, and a major reason for the lower demand forecasts from the use factor model, lies in its much lower income elasticity as compared wtih the pooled model. The other five variables shown in the table are not part of the use factor model so no comparison was possible. Except for days of measurable precipitation, the elasticities for these variables were small. r-. SELECTION OF USE FACTOR MODEL The use factor and pooled models represent two alternative tech- nical approaches to forecasting water demand. As shown earlier in this chapter, the two models also result in different forecasts for the EIS demand area through the year 2035. Use factor model and its forecasts are selected as the preferred approach for this EIS for two major reasons. The first reason for this is that the use factor model is dis- aggregated by consumer sector: SFM, SFF, MF, E and P. This is essen- tial for performing the wide range of conservation analyses conducted in Task 4. Many of these analyses can be performed only with the use factor model. For example, the demand reductions associated with meter- ing unmetered homes must be completed using the use factor model because it makes a distinction between the metered and unmetered demand sectors. The affect of price on SFM households can only be examined by the use factor model. Appendix 2 272 The second major reason for the selection of the use factor model is that it is based on a broader range of data sources. (See discussion of use factors in Chapter 3.) This helps to identify and de-emphasize data which is at odds with other data sources. In contrast, the pooled model relies on a single data set (one also used for the use factor model) using a single technique (linear regression) which requires little understanding of underlying relationships in the data. The importance of using diverse data sources, when they exist, is that a higher degree of confidence can be placed in estimates of future conditions (forecasts) when a model is not tied to a single account of present conditions. Also, the process of justifying diverse sources contribute to a more thorough understanding. Both demand forecasting models contribute to a better understanding of demand in the EIS area, now and in the future. However, in light of the eventual uses of the model, to serve as a basis for conservation analyses, the use factor model is chosen as the most appropriate for this EIS. As discussed in Chapter 6, the series 1 data were selected as the most likely for the EIS area. Therefore, using the preferred use factor model and the preferred demographic and socioeconomic data (series 1) , the following projected water demands are considered as the best esti- mates available and will be used in the subsequent conservation analy- sis. The final demand forecasts are presented in table 115. Appendix 2 273 Table 115 Unconstrained Water Demand Forecasts • EIS Forecast Year Demand!/ (acre-feet) 1980 314,000 1990 414,000 2000 515,000 2010 598,000 2035 703,000 1/ These forecasts have not been adjusted for natural retrofit. r Appendix 2 274 • -a EXHIBIT 2-A WATER DEMAND FORECASTS BY WATER SUPPLIER Table 1 Eater Demand Forecasts by Water Supplier Pooled Model, Series 1 1/ (acre-feet per year) Water Supplier 1 80I/ 120 2000 2010 2215 Arapahoe 83 1,752 3,734 4,265 4,741 Arvada 15,195 19,205 23,851 28,377 34,692 Aurora 29,534 42,590 61,077 77,106 98,760 Beverly Hills 12 12 12 12 15 Brighton 2,759 4,446 7,223 9,426 12,851 Brook Forest 31 37 46 49 52 Broomfield 3,090 5,670 8,628 10,577 13,470 Charlou Park 68 83 107 110 114 Castle Pines 0 660 1,528 2,197 2,341 Castle Rook 988 4,523 9,313 13,154 19,423 Chaparral 0 199 479 503 546 Con. Mutual 12,955 14,842 17,349 18,165 19,613 Cottonwood 49 1,531 3,323 3,685 4,078 Crestview 2,243 3,127 4,198 1,440 4,796 DWD Service Area?/ 188,779 237,370 283,332 323,529 361,617 _ Denver S.R. Sub 512 1,635 3,237 4,382 5,124 Dolly-O-Denver 0 9 25 34 49 Eastlake 25 25 25 28 28 E Cherry Creek Val. 423 4,673 10,353 12,731 15,130 East Valley 31 64 104 107 114 Englewood 8,147 9,270 10,540 11,734 12,396 Erie 132 221 325 405 528 Evergreen 752 871 1,003 1,114 1,123 Florence Gardens 9 37 71 95 114 Forest Hills 18 18 18 18 18 Genesee 184 230 276 316 451 Glendale 782 1,721 2,676 3,173 3,562 Golden 2,716 3,882 5,146 6,122 6,919 Hazeltine Heights 31 34 37 37 40 Hi-Land Acres 52 58 68 71 77 Holly Mutual 31 49 74 77 80 Ideldale 18 25 34 40 43 Lafayette 773 1,356 2,099 2,657 3,713 --. Louisville 1,102 2,228 3,605 4,673 6,017 Louviers 83 120 166 175 187 Maple Grove 12 12 12 12 12 Mission Viejo (Highlands Ranch) 15 4,026 9,178 13,167 15,968 Morrison 175 264 353 414 451 Mt. Carbon 0 8o 166 215 261 North Table Mt 1,283 1,724 2,191 2,581 3,102 Northglenn 5,324 5,888 6,658 6,843 7,312 Orchard Hills 52 68 83 86 92 Parker 117 875 1,850 2,476 2,752 Ro:borough Park 28 393 890 1,270 1,596 Sedalia 28 49 77 101 123 Silver Heights 52 52 55 55 61 South Adams Co. 3,882 5,170 6,658 7,901 8,745 Stonegate 0 801 1,789 2,338 2,633 Thornton 7,923 12,163 17,091 22,538 30,804 Thunderbird 92 92 92 95 101 View Ridge 9 9 12 12 15 Weisner Estate 21 28 31 31 34 Westminster 8,251 11,568 15,848 19,527 25,357 Willows 2,123 3,025 3,964 4,317 4,750 Balance of Area 7.094 11.905 17.429 28.082 55.545 Total (acre-feet) 308,088 420,765 548,509 652,645 792,536 1/ End use level, does not include system losses. Adjustments for water savings attributable to natural metering and retrofit have not been made to the above projections because of difficulties in allocating such sav- ings to individual districts. 2/ Does not include Arvada, North Table Mountain, Consolidated Mutual, Broomfield, or Crest View; does not include Sable and Cherry Creek Valley districts after 1980. _ p Not directly comparable to actual 1980 consumption. Assumes average weather conditions, conservation programs existing in 1981 and 1982, and 1982 marginal price for water. • Appendix 2 275 Table 2 Water Demand Forecasts by Water Supplier Pooled Model, Series 2 1/ 0 ' (acre-feet per year) Water Suoolier 121122/ 1222 2000 2010 1.2.11 Arapahoe 83 1,752 3,734 4,299 5,121 Arvada 15,195 19,205 23,851 28,512 36,637 Aurora 29,534 42,590 61,077 77,628 106,217 Beverly Hills 12 12 12 12 15 Brighton 2,759 4,446 7,223 9,435 14,130 Brook Forest 31 37 46 49 52 Broomfield 3,090 5,670 8,628 10,601 14,547 Charlou Park 68 83 107 110 114 Castle Pines 0 660 1,528 2,197 2,341 Castle Rock 988 4,523 9,313 13,164 21,740 Chaparral 0 199 479 503 546 Con. Mutual 12,955 14,842 17,349 18,177 19,755 Cottonwood 49 1,531 3,323 3,704 4,302 Crestview 2,243 3,127 4,198 4,409 4,817 - DWD Service Area?/ 188,779 237,370 283,332 323,882 365,213 Denver S.E. Subs 512 1,635 3,237 4,422 5,127 _ Dolly-O-Denver 0 9 25 34 49 Eastlake 25 25 25 28 28 E Cherry Creek Val. 423 4,673 10,353 12,731 15,118 East Valley 31 64 104 107 114 Englewood 8,147 9,270 10,540 11,767 12,470 Erie 132 221 325 408 568 Evergreen 752 871 1,003 1,114 1,135 Florence Gardens 9 37 71 98 120 Forest Hills 18 18 18 18 18 Genesee 184 230 276 319 494 Glendale 782 1,721 2,676 3,203 3,903 Golden 2,716 3,882 5,146 6,161 7,156 Hazeltine Heights 31 34 37 37 40 Hi-Land Acres 52 58 68 71 77 Holly Mutual 31 49 74 77 80 Ideldale 18 25 34 40 43 Lafayette 773 1,356 2,099 2,657 3,995 Louisville 1,102 2,228 3,605 4,710 6,545 Louviers 83 120 166 175 190 Maple Grove 12 12 12 12 12 Mission Viejo (Highlands Ranch) 15 4,026 9,178 13,188 16,226 Morrison 175 264 353 417 463 . Mt. Carbon 0 80 166 215 261 North Table Mt 1,283 1,724 2,191 2,593 3,289 Northglenn 5,324 5,868 6,658 6,843 7,340 Orchard Hills 52 68 83 86 92 Parker 117 875 1,850 2,479 2,801 Roxborough Park 28 393 890 1,283 1,596 Sedalia 28 49 77 101 126 Silver Heights 52 52 55 55 61 South Adams Co. 3,882 5,170 6,658 7,944 8,785 Stonegate 0 801 1,789 2,350 2,789 Thornton 7,923 12,163 17,091 22,544 33,446 Thunderbird 92 92 92 95 101 View Ridge 9 9 12 12 15 Weisner Estate 21 28 31 31 34 Westminster 8,251 11,568 15,848 19,534 26,913 Willows 2,123 3,025 3,964 4,333 4,922 Balance of Area 7.094 11.905 17.429 29,442 74,624, Total (acre-feet) 308,088 420,765 548,509 658,416 836,713 1/ Bud use level, does not include system losses. Adjustments for water savings attributable to natural metering and retrofit have not been made to the above projections because of difficulties in allocating such sav- ings to individual districts. 2/ Does not include Arvada, Worth Table Mountain, Consolidated Mutual, Broomfield, or Crest View; does not include Sable and Cherry Creek Valley districts after 1980. j/ Not directly comparable to actual 1980 consumption. Assumes average weather conditions, conservation programs existing in 1981 and 1982, and 1982 marginal price for water. Appendix 2 276 Table 3 Meter Demand Forecasts by Water Supplier Pooled Model, Series 3 1/ (acre-feet per year) Water Supplier 1 8 / ..-..190 2000 2010 2012 Arapahoe. 83 1,752 3,734 4,330 5,545 Arvada 15,195 19,205 23,851 28,647 38,613 Aurora 29,534 42,590 61,077 76,153 113,802 Beverly Hills 12 12 12 12 15 Brighton 2,759 4,446 7,223 9,442 15,422 Brook Forest 31 37 46 49 52 Broomfield 3,090 5,670 8,628 10,626 14,876 Charlou Park 68 83 107 110 114 Castle Pines 0 660 1,528 2,197 2,344 Castle Rock 988 4,523 9,313 13,176 23,587 Chaparral 0 199 479 503 546 Con. Mutual 12,955 14,842 17,349 18,190 19,923 Cottonwood 49 1,531 3,323 3,722 4,547 Crestview 2,243 3,127 4,198 4,412 4,842 DWD Service Area 188,779 237,370 283,332 324,259 369,353 Denver S.R. Subs 512 1,635 3,237 4,458 5,140 Dolly-C-Denver 0 9 25 37 46 Eastlake 25 25 25 28 28 I Cherry Creek Val. 423 4,673 10,353 12,731 15,112 East Valley 31 64 104 107 114 Englewood 8,147 9,270 10,540 11,804 12,556 Erie 132 221 325 408 608 Evergreen 752 871 1,003 1,129 1,148 Florence Gardens 9 37 71 98 126 Forest Hills 18 18 18 18 18 Genesee 184 230 276 322 525 Glendale 782 1,721 2,676 3,234 4,277 Golden 2,716 3,882 5,146 6,198 7,419 Hazeltine Heights 31 34 37 37 40 Hi-Land Acres 52 58 68 71 77 Holly Mutual 31 49 74 77 80 Ideldale 18 25 34 40 43 Lafayette 773 1,356 2,099 2,660 4,284 Louisville 1,102 2,228 3,605 4,750 7,097 Louviers 83 120 166 175 193 Maple Grove 12 12 12 12 12 Mission Viejo (Highlands Ranch) 15 4,026 9,178 13,213 16,520 Morrison 175 264 353 420 476 Mt. Carbon 0 80 166 218 313 North Table Mt 1,283 1,724 2,191 2,608 3,480 Northglenn 5,324 5,888 6,658 6,846 7,373 Orchard Hills 32 68 83 86 89 Parker 117 875 1,850 2,485 2,857 Roxborough Park 28 393 890 1,295 1,596 Sedalia 28 49 77 104 129 Silver Heights 52 52 55 55 61 South Adams Co. 3,882 5,170 6,658 7,987 8,837 Stonegate 0 801 1,789 2,363 2,961 Thornton 7,923 12,163 17,091 22,553 36,109 Thunderbird 92 92 92 95 101 View Ridge 15 Weisner Estate 21 28 12 31 31 34 Westminster 8,251 11,568 15,848 19,543 27,861 Willows 2,123 3,025 3,964 4,348 5,112 Balance of Area 7.094 11.905 17.429 29.871 91.838 Total (acre-feet) 308,088 420,765 548,509 660,355 878,286 1/ End use level, does not include system losses. Adjustments for water savings attributable to natural metering and retrofit have not been made to the above projections because of difficulties in allocating such sav- ings to individual districts. 2/ Does not include Arvada, North Table Mountain, Consolidated Mutual, Broomfield, or Crest View; does not include Sable and Cherry Creek Valley districts after 1980. 2/ Not directly ocmparable to actual 1980 consumption. Assumes average weather conditions, conservation programs existing in 1981 and 1982, and 1982 marginal price for water. Appendix 2 277 • • Table 4 6KYtA Water Demand Forecasts: Series 1 (acre-feet) Water District 1980 1990 2000 2013 2035 ALAMEDA 17792 309 — 2133944 _— sN37733 —322— 394 ARAPAH01 374? 4115 ARVADA AURORA DE It 26 12]31 57657 70220 33579 DEM OF AR VEIDONVULTTHILLS 5720 ,743 9164 E132 3115 12771 61R6 2670 +9,43 7716 2011-MAR 22 6/ 2 8318 11711 BRIGHTON FOREST p 2839{ 4170 150 150 0 MF 177NF1f 77'36 139 075 10176 12405 *5437 A97Lp64211 19 12599 17691 E ROCK 10818 MERRY OU►ARKL 137 24gfp0 190 197 718 CCCCso .p.pERRYMOOR !T a90 227 1660s2 578 774 61 181 CRE974IEYDl 141,4 3069 1+1+5s61p0 130006152 17611 18448 16513 DENVER S.E. S 608 lb30 2785 7911 3993 BD,ug c-O-DENVE 0 14314? 154517 16554699773 1{{584736 BR TOTAL}li SER 129798 31621 41191 19961 21031 16995 20569 E CHERRY LL CRK. s EIAASTTlVALLEY 423 49 76 76 76 EDOEVATER 771 85i 929 911 908 fN01 EYOOD 9190 10009 20938 11661 11764 ERIE 19d 320 472 593 754 EVERGREEN 1196 1664 2150 2537 2858 FLORENCE BARD 11 48 83 111 125 FOREST HILLS 4 24 24 24 24 Mud 6g !79 294 374 468 gOFuK7CE 063 1113 '164 2500 2687 6 !"EN JJ4? 1�3'� 6090 6494 GREEN MTN 4576 3177 7939 13966 HAZELTINE HEI 78 42 45 45 45 HI-LAND ACRES 75 84 94 96 .n3 995 961 1091 1147 !!1 OH VIEW 789 HIGHLANDS RAN 19 4323 9731 5 Mil, MUTUAL 25694 15- 23 31 31 3 IOIEOALE 7B 47 {{6{ZZ0 6 6 279 KEKEN-CARYLON IRNT 1087 12140 296, 3763 . 16287 LAFAYETTE 1455 LAKEHURST If 191 2871 3269 70 3496 0 7687 V`LOUISVILLE 11111 3168e7 5136 6690 0 y0 8716 MMEEADOVBROOK 290 717 44 161 776 MORRISON NT. CARSON 140 1M91 76 263 309 NORTH TABLE 9 1772 1787 117 2542 3023 NORTH WASHNOT 2109 4528 28883 2T7 3121 NORTHOLENN - 3713 4005 4763 4403 1269 NORTHEIDE 444 630 863 13 ORCHARD HILLS 54 9 PANORAMA PARK 19 p 7 PARKER 139 1066 d 2117 2797 3053 9E^6116U8N PA 26 134 17`a' 229 265 68 68 SILVER HEIGHT C 31020 6`744y7 8416 497 10583 3NORNTDM 7866 11716 13917 20210 2.4479 THUNDERBIRD 91 91 3! ppV�A�EELLEY EE 936 1171 1321 1722 1795 YEISNEPDEgg4TAT t4 �p77p999�2 2 210 78 0 S 19 18 YVIEEATTRMSTER ROOR 17461 12775 15710 17]17 1685 •+6LOV 2286 2876 7419 7605 3171 TOTALS 313,749 413,586 5144, 18 597,614 703043 'figures are for 1980 with typical weather and should not be compared directly with actual 1980 demands. Appendix 2: 278 Table 5 MU Water Demand Forecasts: Series 2 (acre-feet) ' _ _ i4atxr Xistrlct_ 1990 2000 _ 2010 2035 ee 30 -31 �1p1 ��P3 37777- _. 394 AR63 43541 VADA 219'6 �2 i1 254006 19061 35570 AURORA 299 6 7733 57 70697 92559 ANOFT-CLOV 5720 73 81622 !63557 26003 67162 NVUEY HILLS NNAAggy lit jj 336 eNTFNREDT "tp' 4111 6;�2 ins 1790'0 DROOMFIELD 88 7150 7055 II 6 150 150 T I 8 ! 16659 CCCNETE2Rµ00RyL 1028! I7iNARLOU PARK 131t�g222 922 12666608 *9773 2e9 CHERRYNO R ST 12}0 lz7 1787 166 Isi 90 651 X RRY CRK 8s CON. MUTUAL 14124 15660 13000 eCeggOEET9TOOTgNqWEOEOD 8 17218 13727 DEB9FRTE. $ { +s, 3291 `37 77 DOLLY-0-8EMVE 3049 601 lilt 2721 2464 1923 8YD CITY+COUN 8 145147 154517 36 DOD READ 4 DI 11873 154744 162141 DOD TOTAL SER 14374 31621 MR 21031 52910 0 20697 E CHERRY CRK. 173 5144 11079 13455 16206 EgEAAAyST VAEELI EY 23 49 76 76 76 EEEDB`E` 2T£R32 ''22 3 ERIE 9190 9 90 10096 01 109]0 32 33117 109 76 13 ERIgp EEEE {771 JJ 11693 11813 fLORENCENDARD 1186 1644 2150 25552 2876 FOREST HILLS 24 1 112 eSEEENEMp9^LE 0i 131718 2164 2 9 24 378 s2'+ BOLDEN IR63 1794 25 6114 6811 GREEN MTM 1576 5177 5730 7938 13967 NHAAZELTINE HEI 9 NI-LAND ACRES 75 e14 H NIGH VIEW e♦♦ 9d !0. HIGHLANDS RAM 71q 4577 97951091 MOLLY MUTUAL 13597 17891 IlLEDALE 30 91 t6Z0 368 1 111 KEELTTON MEIBXT 127 135 143 5373 lnseDi REM-[ARYL 1507 1955 5861 LAFAYETTE E { p CS LAREYO0DT 4 3459 4626 LOUISVILLE 1552 327 23 '700 19 1687 LOUVIERS�Ep Igi1k 31447 519732 202 9218 "'RIS0N��pp 29149 E 0 22207 203 364 304 NORTH Ma N 177? 187 222!16 2532 3333 205 NN�pOORRRTHOOyLLIENNENRT - 1713 40025 4365 4405 888 2976 4785 ORCM�RODMIL`f 444 610 801 1107 1217 PANORAMA PARK 6t 25 PPAARRKKEERRQ 159 1066 '211 2800 3093 "�D�ALI^ M} 426 f$$ 51175 1250 1266 MINESOUTH ADAMS C 8102 6`711, 8416 11313168 120620 7866 11791 15913 20215 78704 TNUNDERDIRB 1 84 �VOiELLELiy1LL91ERYR yygg���� 933 1127„1 - 13*78 1` 1713 1822 _____IN5TERA7 1 aj 177776 177!! 2101820 217+7422 Imo` 2406 2876 04 2780 17118 5213 +027 3405 I LLSYlR TOTALS )13,749 413,586 514,618 539,933 739,822 TMfigures are for 1960 with typical weather and should not be compared directly with actual 1980 demand. Appendix 2 • 279 Table 6 ^ 6KY&A Water Demand Forecasts: Series 3 (acre-feet) _ Water District 19801/ 1990 2000 2010 2035 ALAPANOE 3II09 344 373 377 394 AULLROR29 AA EE 79932 1500 OI R .25406 2922s0e* 3401 3784 7474 AMCROFT-CLOY 5720 4 71174 6803 '71700 1164 12771 7 77919 � RULEY MILLS 116 7616 63145 t}737 6 16 17 BRIGHTON 2934 41,0 6;23 _^t�3109 423 179 OON ELD B9 9 7950 7079 3 1097608 'S0 31 14130 16717 OOK FOREST 7¢i :ITII Ms 0 1949 N R`OCUA►AIIKL 1080 '0 III; 7105 12617445 595 216391 + HENRY CRK s. 1 ' 378 197 il1 NE MO T 4278 570 7 162 MUTUAL 13069 124460 1308 a0 te1 TTOM 17625 1 15 E9TVTEY 1333 6152 5 X1,9 645P DENVER S.E. 9 { DOLLY-0-DENVE 409 10 21 4004 . ICITY*COUN 129799 143147 154517 16 76 aBwYnn READ + DI 18736 31621 41181 15299 111750 DODMTOTAL CSER 14574 19309 21038 52992 9 827 A TTT VALLEY 475 2 5144 AKE 328 110776 13456 16220 EINDI 32 WATER 9770 10099 10979 * 7 9 76 11 ERIE BREEN 196 328 472 11724 4 I1 873 FLORENCE BARD 1186 1648 2150 2566 7131 FOREST HILLS11 1s24 2!9 24 s24 '^ 111111IE 3R63 •797 21644 2091 3111 GREEN MTN NAZELTINE NEI 4576 5172 5730 6147 13960 NI-LAND II `` yyACRES VIEW .5 41 �4 96 45 45 HOHILLYAMUTUALN 715 472, 7711 1361e9 995 961 4 161099 IDLEDALE 39 • 60 1 31 31 KELTON HEIGHT s3� 135 II!! 263 279 LA[MERRE 105 1755 786+ J95J 10746 tAK9Y0GDTT y6 LOUISVILLE OI`tE[( 9RL 1552 2676 3269 8 J 02 ]1 7 IP100YORL101E( 1 f 0e 77 1441/7 0003 690y0y 9585 MORRISON 290 387 4 6.0 771 NT. CANON 149 227 p 366 yy NORTH ►ECOS 311 7`)} 914 360 434 MONTH TABLE 9 1372 17 7 2196 2568 NORTH WASHNOTzees 3396 RORTHOLENN ' 2109 7 4005 4365 4407 4 11 NORTMSIDE 444 630 801 1167 1271 ORCHARD HILLS 54 69 IS PANORAMA PARK 119 22 21 PARKER PA 526 1066 117 . 25 27 X77 1172 SEDALIA 48IGHT l0174 $ 69 ILVER ADAMS C 5102 6747 9476 9470 769 1pls 2�t269 T61 HUNDERBIRD 766 11;1 15903 202'7 7112, YY 99 VtEEAtLLEY EEB8gg[ 936 11}7711 1335 17+g8 WESTMINSTERAT 1 ±4 13776 177 fb 1481 VILLSR1960K 77401 27831 33� 2704 2e+e6 2286 287` 7419• , 362 8125 TOTALS 313,749 113,586 314,618 601,478 773,623 *Figures are for 1930 with typical weather and should not be compared directly _with actual 1980 demand. Appendix 2 280 �. _ EXHIBIT 2-B - POOLED !DDEL RESIDUALS AND INCOME ADJUSTED'? FACTORS Table 1 Pooled Model Residuals, 1981 and 1982 Average Pooled Model Water Supplier Residual (g.h.d.) Arapahoe 0 Arvada 13.627 Aurora 42.40067 Beverly Hills 0 Brighton 96.8518 Brook Forest 0 Broomfield -62.4754 Charlou Park 0 Castle Pines 0 Castle Rock 88.28565 Chaparral 0 Con. Mutual -24.3071 Cottonwood 0 Denver S.E. Suburbs - Dolly-0-Denver 0 Eastlake 0 E. Cherry Creek 69.62205 East Valley 0 Englewood 71.3164 Erie -12.2973 Evergreen 0 Florence Garden -79.8727 Genesee 0 Glendale -85.8494 Golden -61.5024 Hazeltine Heights 47.5058 Hi-Land Acres -86.3415 Holly Mutual -287.267 Idledale 0 Lafayette 0 Louisville -87.7686 Louviers 0 Maple Grover 0 Morrison 0 Mt. Carbon 0 Northglenn 190.525 Orchard Hills 0 Parker -62.0352 Roxborough Park 0 Sedalia -272.82 Silver Heights -91.2837 South Adams County -34.7283 ^ Stonegate 0 Thornton 76.75775 Appendix 2 281 Table 1 (Continued) Pooled Model Water Supplier Residual (g.h.d. Thunderbird 0 View Ridge 0 Westminster -32.8498 Willows 33.3694 Balance of Area 0 DWD City + County 72.01421 DWD Total Serve 65.4017 DWD Read + Bill 79.91875 Green Mountain -43.8136 Wheatridge -31.2777 Edgewater 3.79372 Bow-Mar 12.71081 Panorama Park -252.442 Alameda 0 Kelton Heights -54.5231 Bonvue 0 High View 0 Valley 2.33475 Lakehurst -18.9166 North Washington 132.2136 Cherry Creek Village 128.004 Meadowbrook -75.9929 Willowbrook -55.0607 Northside 22.15884 North Pecos 0 Cherrymoore Sth 0 Key Caryl 151.014 Bancroft-Clover -26.9038 Lakewood -119.975 North Table Mountain 68.5763 Crestview -63.2132 Notes: A description of the residuals and their application to the pooled model is provided in Chapter 4 and Chapter 8. Suppliers with zero either (1) did not submit data for 1981 and 1982; (2) were outliers; or (3) were not yet in existence. Water demand forecasts for several of the outlier suppliers were adjusted to reflect reported 1981 and 1982 data. Appendix 2 282 Table 2 Pooled Model Income Adjustment Factors Series 1 Series 2 Series 1980 1.000 1.000 1.000 1990 1.006 1.006 1.006 2000 1.041 1.041 1.041 2010 1.032 1.031 1.031 2035 0.937 0.934 0.931 Note: These constants should be multiplied by each district's income forecasts to create a system-wide weighted average household income projection at growth rates specified in Chapter 7. This adjustment is necessary because the pooled model is based on a per unit, GHD calculation without regard to the number of households in district itself. Appendix 2 283 EIRIBIT 2-C ALTERNATIVE POOLED IADEL PROJECTIONS ALTERNATIVE POOLED MODEL PROJECTIONS Water demand forecasts applying the alternative interaction term pooled regression model are presented in table 1 . This model was iden- tified and described in Chapter 4. 'The underlying economic and demo- graphic forecasts are the same as those utilized for the preferred pooled water model forecasts reported earlier. Table 1 EIS Demand Area Water Demand Forecasts Applying the Alternative Pooled Model, Thousands of Acre-Feet Year Series 1 Series 2 Series 1980 307.5 307.5 307.5 1990 415.0 415.0 415.0 2000 534.5 534.5 534.5 2010 632.1 634.9 636.8 2035 747.1 790.0 830.3 Appendix 2 284 REFERENCES CITED American Water Works Association. No Date. The Story of Water Supply. AWWA, Denver, Colorado. American Water Works Association. 1975. Water Conservation at Home. AWWA, Denver, Colorado. Agthe, E. and R.B. Billings. 1980. Dynamic models of residential water demand. Water Resources Research (June): 476-480. Beattie, R. and H.S. Foster. 1980. Can prices tame the inflationary tiger? Journal of the American Water Works Association (August): 441-445. Berry, D.W. and G.W. Bonem. 1974. Predicting the municipal demand for water. Water Resources Research (December): 1239-1242. i-- Appendix 2 285 Billings, R. 1982. Specification of block rate price variables in demand models. Land Economics (August): 386-393. Black and Veach/Consulting Engineers. 1981. Treated Water Planning Study. Board of Water Commissioners, Denver, Colorado. Boland, J., et al. 1981. An Assessment of Municipal and Industrial Water Use Forecasting Approaches. U.S. Army Engineer Institute for Water Resources, Fort Belvoir, Virginia. Camp, R.C. 1978. The inelastic demand for residential water: new findings. Journal of the American Water Works Association (August): 453-458. Carver, P.H. and J.J. Boland. 1980. Short and long-run effects of price on municipal water use. Water Resources Research (August): 609-616. City of Aurora Utility Department. 1983. Unpublished Water Use Data. Aurora, Colorado. City of Englewood. 1983. Unpublished Billing Records Data. Englewood, Colorado. Colorado Demographic Section. 1981. Adjusted Population Projections of Colorado Counties: 1981-2010, Colorado Population Reports. Colo- rado Division of Local Government, Denver, Colorado. Series CP-25, No. 81 (A)-1A, computer printout. Corps of Engineers. 1978. Water and Related Land Resources Management Study: Supporting Technical Report Appendices, Volume V, Appendix J, Volume 2-6. Omaha, Nebraska. Danielson, R. E. 1979. An analysis of residential demand for water using micro time-series data. Water Resources Research (August): 763-767. Denver Regional Council of Governments. 1982. Employment and Popu- lation Policy Forecasts for the Denver Region - 1980-2000. Denver, Colorado. Denver Regional Council of Governments. 1982. Employment and Popula- tion Projections for the Denver Region - 1980-2000. Denver, Colo- rado. Denver Regional Council of Governments. 1982. Household and Household Size Forecasts for the Denver Region, 1980-2000. Denver, Colorado. Appendix 2 286 Denver Regional Council of Governments. 1978. Regional Water Study. Denver, Colorado. Denver Regional Council of Governments. 1983. Task 2.1 Report, Popu- lation and Employment Projections, Regional Water Study. Denver, Colorado. Denver Regional Council of Governments. 1983. 1983 Regional Water Study Questionnaire. Denver, Colorado. Denver Water Department. 1981. Annual Report. Denver, Colorado. Denver Water Department. 1975. Metropolitan Water Requirements and Resources, 1975-2010, Three Volumes. Denver, Colorado. Under the direction of the Denver Regional Council of Governments. Prepared for the Colorado State Legislature, Metropolitan Denver Water Study Committee. Denver Water Department. 1978. Population and Land Use Report. Denver, Colroado. Denver Water Department. Selected years. Tabulation of Water Rates. Denver, Colorado. Department of Commerce. Selected years. Local Climatological Data. Denver, Colorado. Denver Water Department. Selected years. Three Inch Metering Study. Denver, Colorado. Dworkin, D. and D.D. Baumann. 1974. An Evaluation of Water Reuse for Municipal Supply. U.S. Army Engineer Institute for Water Resources. Fort Belvoir, Virginia. Flack, J.E. 1982. Urban Water Conservation. American Society of Civil Engineers. New York, N.Y. Flack, J.E., W.P. Weakly, and D.W. Hill. 1977. Achieving Urban Water Conservation: A Handbook. Colorado Water Resources Research Institute, Colorado State University, Fort Collins, Colorado. Com- pletion Report, No. 80. Foster, S. Jr., and B.R. Beattie. 1979. Urban residential demand for water in the United States. Land Economics (February): 43-58. Appendix 2 287 Gallagher, D.R. et al. 1981. Methods for Forecasting Urban Water Demands. Department of National Development and Energy, Australian Water Resources Council. Australian Government Publishing Service, Canberra, Australia. Gibbs, R.C. 1978. Price variable in residential water demand models. Water Resources Research (February): 15-18. Gottlieb, M. 1963. Urban domestic demand for water: a kansas case study. Land Economics (May): 204-210. Governor's Blue Ribbon Panel. 1981. Colorado: Industry in the Future, 1981-2001, Final Report, Volume 2. Denver, Colorado. Grima, A.P. 1972. Residential Water Demand: Alternative Choices for Management. University of Toronto Press, Toronto, Canada. Hanke, S.H. and L. de Mare. 1982. Residential water demand: a pooled, time series, cross section study of Malmo, Sweden. Water Resources Bulletin (August): 621-625. Heim. Personal communication. Rocky Mountain Arsenal, Denver, Colo- rado. Hibbert, J. 1983. Adolph Coors Company, Golden, Colorado. Hogarty, T.F. and R.J. Mackay. 1975. The impact of large temporary rate changes on residential water use. Water Resources Research (December): 791-794. Howe, C.W. and F.P. Linaweaver. 1967. The impact of price on resi- dential water demand and its relation to system design and price structure. Water Resources Research (1st Quarter): 13-55. Hughes, T.C. and R. Gross. 1979. Domestic Water Demand in Utah. Utah Water Research Laboratory, College of Engineering, Utah State University, Logan, Utah. UWRL/P-79/04. larger, E. Personal communication. Gates Rubber Company, Denver, Colorado. Maidment, R. and E. Parzen. 1981. A Cascade Model of Monthly Municipal Water Use. The Texas A&M University System, Texas Engineering Experiment Station, College Station, Texas. Metcalf and Eddy, Inc. 1979. Wastewater Engineering: Treatment Disposal, Reuse. McGraw-Hill, New York, N.Y. Appendix 2 288 Metropolitan Water Roundtable, Water Development Task Group. 1982. Report. Denver, Colorado. Milliken, J.G., et al. 1975. Systems Analysis for Wastewater Reuse, A Methodology for Municipal Water Supply Planning in Water-Short Metropolitan Areas. Volumes 1 and 2. University of Denver, Denver Research Institute, Industrial Economic Division, Denver, Colorado. Morgan, W.D. 1974. A time series demand for water using micro data and binary values. Water Resources Bulletin (August): 697-702. Morris, J.R. and C.V. Jones. 1980. Water for Denver: An Analysis of the Alternatives. Environmental Defense Fund, Denver, Colorado. Morris, J.R., C.V. Jones and D.F. Luecke. 1981. Response to Denver Water Department Staff Review of Water for Denver: An Analysis of the Alternatives. Denver, Colorado. Mugler, L. 1983. Personal communication. Denver Regional Council of Governments, Denver, Colorado. Oregon State Water Resources Board. 1969. Oregon's Long-Range Require- ments for Water. Sharpe, W.E. 1981. Water Conservation Devices for New or Existing Dwellings, Journal AWWA, pp. 144-149. Turnovsky, S. J. 1969. The Demand for water: some empirical evidence on consumers response to a commodity uncertain in supply. Water Resources Research (1969): 350-361. Q.S. Bureau of Economic Analysis. Selected years. Survey of Current Business. U.S. Department of Commerce, Washington, D.C. U.S. Bureau of Economic Analysis. 1981. 1980 OBERS BEA Regional Projection, Volume 10, Region 7, Rocky Mountain. U.S. Department of Commerce, Washington, D.C. U.S. Bureau of Reclamation. 1972. Final Environmental Statement. Central Arizona Project. Q.S. Department of the Interior, Washing- ton, D.C. Q.S. Bureau of Reclamation. 1979. Final Environment Statement. Central Utah Project, Bonneville Unit. U.S. Department of the Interior, Washington, D.C. Appendix 2 289 U.S. Bureau of Reclamation. 1977. Final Environmental Statement, Second Stage, Southern Nevada Water Project. U.S. Department of the Interior, Washington, D.C. U.S. Bureau of Reclamation. 1976. Final Environment Statement, San Felipe Division. U.S. Department of the Interior, Sacramento, California. U.S. Bureau of the Census, 1982. 1980 Census of Population, General Population Characteristics, Colorado. U.S. Department of Commerce, Washington, D.C. PC80-1-B7. U.S. Bureau of the Census. Selected years. Number of Inhabitants, Colorado. U.S. Department of Commerce, Washington, D.C. U.S. Bureau of the Census. 1981. Statistical Abstract of the United States. U.S. Department of Commerce, Washington, D.C. U.S. Water and Power Resources Service. 1980. Final Environmental Statement, Animas - La Plata Project, Colorado-New Mexico. U.S. Department of the Interior, Washington, D.C. U.S. Water and Power Resources Service. 1981. Final Environmental Statement, Colorado - Big Thompson, Windy Gap Projects. U.S. Department of the Interior, Washington, D.C. U.S. Corps of Engineers. Wiley, R. 1983. personal communication. Denver Water District, Denver, Colorado. Wong, S.T. 1982. A model of municipal water demand: a case study of Northeastern Illinois. Land Economic: 34-44. Young R. 1973. Price elasticity of demand for municipal water: A case study of Tucson, Arizona. Water Resources Research (August): 1068-1072. Appendix 2 290 Hello