Loading...
02-05-2002 (Planning & Zoning) Agenda Packet Planning & Zoning Commission ■••■•••••100- ,1, I 1111111111111111111111111111V AM II 11111111P41111 'WM= ME 74 ir AMINE )1' /MINIM 1111111111111 MEV NAL Nt � ii a ty lie February 5 , 2002 Regular Business Meeting AGENDA PLANNING AND ZONING COMMISSION CITY COUNCIL CHAMBERS WYLIE MUNICIPAL COMPLEX 2000 Highway 78 North Wylie,Texas 75098 Tuesday, February 5, 2002 7:00 p.m. CALL TO ORDER INVOCATION & PLEDGE OF ALLEGIANCE CITIZEN PARTICIPATION CONSENT AGENDA 1. Consider and act upon approval of the Minutes from the January 15,2002 Regular Meeting. PUBLIC HEARING AGENDA 1. Hold a public hearing and consider a recommendation to the City Council regarding proposed revisions to Architectural Requirements of the Comprehensive Zoning Ordinance, including the Residential Exterior Façade Material of Section 3.4.F.1, the Repetition of Single Family Residential Unit Designs of Section 3.4.F.6 and 7, the Village Residential Front Entry Cover of Section 3.4.F.9, and the Village Residential Garage Door Offset of Section 3.4.F.11. 2. Hold a public hearing and consider a recommendation to the City Council regarding proposed revisions to the Townhouse (TH/15) District Requirements of the Comprehensive Zoning Ordinance, including the Minimum Lot Size and Dwelling Unit Size of Section 3.3.B. ADJOURNMENT Posted Friday, February 1, 2002 at 5:00 p.m. THE WYLIE MUNICIPAL COMPLEX IS WHEELCHAIR ACCESSIBLE. SIGN INTERPRETATION OR OTHER SPECIAL ASSISTANCE FOR DISABLED ATTENDEES MUST BE REQUESTED 48 HOURS IN ADVANCE BY CONTACTING THE CITY SECRETARY'S OFFICE AT 442-8100 OR TDD AT 442-8170. MINUTES PLANNING AND ZONING COMMISSION WYLIE MUNICIPAL COMPLEX 2000 Highway 78 North, Wylie, Texas 75098 Regular Business Meeting January 15, 2002 7:00 pm Notice was posted in the time and manner required by law and quorum was present. Commission Members Present: Staff Members Present: Steve Ahrens Claude Thompson, Director Mike Phillips Mary V. Bradley, Secretary Carter Porter Terry Capehart, Planner Tony Snider Gary Wiseman Commission Members Absent: William Chapman Michael George CALL TO ORDER Vice-Chairman Porter called the Regular Meeting to order at 7:05 p.m. PLEDGE AND INVOCATION Commissioner Snider offered the Invocation and Commissioner Ahrens led the Pledge of Allegiance. CITIZEN PARTICIPATION No one appeared to speak. CONSENT AGENDA ITEM 1. Consider and act upon approval of the Minutes from the January 2, 2002 Regular Meeting. Motion was made by Commissioner Phillips, seconded by Commissioner Ahrens, to approve the minutes as submitted. Motion passed 5 —O. P&Z Commission Meeting January 15, 2002 Page 2 of 2 PUBLIC HEARING AGENDA 1. Hold a public hearing and consider a recommendation to the City Council regarding a change in zoning on Tract A from Single-Family (SF-3 and SF-8.5/17) and Business- 2 (B-2) to Multi-Family (MF) and Tract B from Multi-Family (MF) to Single-Family (SF 8.5/17), Tract A, being approximately 8.175 acres and Tract B, being approximately 5.969 acres situated in the D.W. Williams Survey, Abstract No. 980 and the Allen Atterbury Survey, Abstract No. 23, City of Wylie, Collin County, Texas, being a portion of a tract of land conveyed to JCM Properties, according to the deed recorded in Volume 2223, Page 874, and part of two tracts of land conveyed to JCM Partners, L.P., according to the deed recorded in County Clerk File Number 96- 0045039, and part of a tract of land conveyed to Verna M. Poor and Michael R. Poor according to the deed recorded in Volume 4409, Page 2100 of the Deed Records of Collin County, Texas. (Zoning Case No. 01-16) Thompson stated that the applicant is requesting rezoning of the subject tracts currently zoned Multifamily, Business or Neighborhood Service and Single Family in order to relocate the Multifamily and eliminate the Business (Neighborhood Service) uses. Thompson stated that the Single Family area and approximately 3.38 acres of Business-2 fronting Elm Road were rezoned in 1985, and approximately 8.23 acres immediately north of this B-2 were rezoned Multifamily in 1987. The remainder of this tract was zoned Single Family(SF-3). The Comprehensive Plan and Thoroughfare Plan recommend that Elm Road become a six-lane parkway with divided median as part of a new Lakeshore Drive loop system to connect Wylie's downtown core with lake-oriented developments. The higher-densities of Multifamily and the SF-8.5 Districts are more compatible with this major traffic corridor than is the rural roadway envisioned to accommodate the Country Residential classification of the Comprehensive Plan. Steve Parsons, 16810 Deer Park, Dallas, Texas, owner of the subject property, stated that the intent is to relocate the Multifamily to the frontage along Elm Road and expand from 8.23 acres to 10.44 acres. The remainder of the property will remain zoned SF 8.5/17. Parsons stated that there is no intent to develop one-acre lots as recommended by the Comprehensive Plan. The lot size will be 8,500 square feet such as are developing in the adjoining property to the north. The following citizens voiced opposition to the rezoning, and concerns for the widening of Elm Road. P&Z Commission Meeting January 15, 2002 Page 3 of 3 Kevin Sanders, 1780 Elm Drive, Wylie, Texas; Willis Dykes, 1820 Elm Drive, Wylie, Texas; Brenda Hensley, 2310 Elm Drive, Wylie, Texas; Carl Foster, 2320 Elm Drive, Wylie, Texas; Rick Foster, 2010 Elm Drive, Wylie, Texas; and Ronny Klingbeil, 821 Autumn Hill, Wylie, Texas. Thompson stated that since Elm Road was a County Road, the County is responsible for the widening and maintaining of the road. Motion was made by Commissioner Ahrens, seconded by Commissioner Phillips, to close the Public Hearing and recommend approval of the rezoning as submitted. Motion passed 4 —1 with Commissioner Snider voting no because the request does not conform to the Comprehensive Plan. Thompson stated that the rezoning would go to City Council at 6:00 p.m. on January 22, 2002, and encouraged the citizens to attend that meeting. 2. Hold a public hearing and consider a recommendation to the City Council regarding a change in zoning from Agricultural (A) to Planned Development (PD - S/F 10/19), being a certain 68.11 acre tract of land situated in the Guadalupe De Los Santos Survey, Abstract No. 1384, and being parts of Lots No. 1 and 2 of the partition of the estate of Josiah Stone among his children, and being part of a 154.31 acre tract and a 1.01 acre tract conveyed to B.B. Owen and H.A. Walker by deed dated July 30, 1960, and filed for record August 5, 1960, in the Deed Records of Dallas County, Texas. (Zoning Case No. 01-17) Thompson stated that the subject property totals 68.11 acres in size and that the applicant intends to develop approximately 155 single-family residential lots and approximately 9.4 acres of public open space. Lots will be a minimum of 10,000 square feet. Jerry Sylo, of Genlser Planning Group, 5430 LBJ Freeway, Three Lincoln Centre, Suite 400, Dallas, Texas, represented the property owner, Southwestern Medical Foundation, stated that the major site constraints limit the developable size, especially adjacency to a sewage treatment plant, and several utility corridors that will become linear parks. Sylo stated that the requirement of the SF-10 District will be followed, except the PD Conditions with all the porches closer to the street enhancing a feeling of a community. The homes would be approximately 2,300 to 2,500 square feet, and depending on the market, the development would not begin until two or three years down the road. Sylo presented to the Commissioners a handout reflecting one-acre lots, which he believes does not allow the creativity of the 155-lot presentation. The following citizens voiced opposition to the development, including such concerns as decreasing value of their property, overcrowding of the school, increased traffic on rural Pleasant Valley Road and potential conflict with school buses, destruction of the floodplain and County Park/Preserve, all caused by small houses on small lots. P&Z Commission Meeting January 15, 2002 Page 4 of 4 Clinton Loggins, 9512 Wells Road, Wylie, Texas; Scott A. White, 9470 Wells Road, Wylie, Texas; R. P. Miller, 101 Hunters Glen, Wylie, Texas; Travis Brock, 9427 Wells Road, Wylie, Texas; Carolyn Proctor, 4035 Avion Drive, Wylie, Texas; Becky and Carlos Sanoja, 4060 Avion Drive, Wylie, Texas; James E. Parker, 5406 Pleasant Valley, Wylie, Texas; Melinda Bento, 3525 Whitley Road, Wylie, Texas; Don G. Smith, Sr., 5102 Sachse Road, Sachse, Texas; Kim Mays, 3577 Whitley Road, Wylie, Texas; Cliff Martin, 4040 Avion, Wylie, Texas; Jacqui Makias, 3507 Pleasant Valley, Sachse, Texas; Pat Martin, 4040 Avion Drive, Wylie, Texas; Leila Makias, 3507 Pleasant Valley Road, Sachse, Texas; and Kenneth McCord, 9424 Wells Road, Wylie, Texas. The Commission took a break at 8:45 p.m. and reconvened at 8:54 p.m. Jerry Sylo stated that the land was donated to support the charitable functions of Southwest Medical and the owner is responsible to get the best return, while still being responsible to existing neighbors. If lots were one acre there would only be 65 lots, and he did not believe there was a market for one-acre lots in a subdivision layout. Commissioner Snider stated that the owner should consider larger lots and be charitable to the neighbors the land was donated. Motion was made by Commissioner Wiseman, seconded by Commissioner Ahrens, to recommend denial because the development does not comply with the Comprehensive Plan. Motion passed 5 —O. ACTION AGENDA 1. Consider a recommendation to the City Council regarding a Preliminary Plat for the LGD Properties Townhouses, being a certain 1.2388 acre tract of land situated in the S.B. Shelby Survey, Abstract No. 820, and being a part of Block 3 of the unrecorded Russell Addition, and being a part of Lot 39, Block 8, Brown and Burns Addition to the City of Wylie, Collin County, Texas, according to the Plat thereof recorded in Volume 1, Page 110-112 of the Map Records of Collin County, Texas, and being a tract of land conveyed to Marion M. McDonald by deed recorded in Volume 205, Page 653, and being a tract of land conveyed to D&M Check Cashing by deed recorded in County Clerk's File No. 95-0008356 of the Deed Records of Collin County, Texas. Thompson stated that the Preliminary Plat includes 1.2338 acres and will create eleven (11) Single-Family Townhomes constructed as zero lot line duplexes. On September 11, 2001, the property was rezoned to Townhouse District (TH/15) from Single Family (SF- 2). Larry DeBerry, 8277 PR 5397, Nevada, Texas, stated that the initial townhouse will be 1,500 square feet, and lots will be at least 4,000 square feet as required by code. P&Z Commission Meeting January 15, 2002 Page 5 of 5 Motion was made by Commissioner Ahrens, seconded by Commission Wiseman, to recommend approval of the Preliminary Plat as submitted. Motion passed 5 —O. The Commission adjourned the regular meeting and convened to a work session at 9:15 p.m. WORK SESSION 1. Consider revisions to the Townhouse District (TH/15) of the Zoning Ordinance. Thompson stated that the Zoning Ordinance was revised to combine Duplexes and all Single-Family Attached within the Townhouse District (TH-15) in July 2000. A few developers and builders have expressed concerns for the differences and unique features in each attached building-type, and required that changes be considered for this District. In attendance of the work session to discuss the matter with Commissioners was Larry DeBerry, developer of townhouses on Jackson Street and Marc Schwartz, developer of duplexes in Garland. They generally explained that ownership of the various attached housing types differ and this dictates differences in design. Until land and purchase costs increase with demand, the Wylie Market cannot support the larger attached dwellings. Thompson led the group in a discussion of the requirements of the lot and dwelling size for a townhouse, reviewing the limits of the regulation and various alternative modifications. An open and comprehensive discussion followed. Although various alternatives were discussed, general consensus was to call a Public Hearing to consider the following revisions: • Reduce the minimum dwelling size to 1,000 square feet • Require enclosed garage with 20 foot setback ■ Keep 4,000 square feet minimum lot size, but allow 3,000 square feet minimum size for interior lots when three or more lots are aligned with attached, zero side yard buildings. Commissioners directed staff to bring these revisions to a Public Hearing at the earliest possible date. P&Z Commission Meeting January 15, 2002 Page 6 of 6 ADJOURNMENT Motion made by Commissioner Ahrens, seconded by Commissioner Wiseman, to adjourn at 10:10 p.m. Motion passed 5 —O. aZ.,f Carter Porter, Vice-Chairman Mary V. Bradley, Secretary iiiir f 4 My of Wylie Public Hearing Item No. 1 Revisions to Residential Districts Of the Zoning Ordinance Planning and Zoning Commission Meeting: February 5, 2002 Applicant: City of Wylie Summary: Council adopted new Residential Districts in July, 2000 with the desire to encourage flexibility of design as well as raise the standards of construction. From the time of adoption no developers have fully implemented the new flexible point system. Recent developments have been zoned as Planned Development Districts which establish unique development requirements or have not yet initiated construction under the new requirements. At the Work Sessions held on December 18, 2001 and January 2, 2002, including discussions with numerous developers and builders the Planning and Zoning Commission directed that a Public Hearing be called to consider several issues concerning the architectural requirements for Single Family Districts. The following proposed revisions were suggested for possible recommendation to the City Council for adoption. 1. Revise the current requirements of Section 3.4.F.1.2 for Exterior Facade Materials from 100% brick or stone to 100% masonry on the front façade and 80% masonry on the remaining walls (masonry to include brick, stone and composite cementitious products as approved by the Building Official). 2. Revise the current requirements of Section 3.4.F.6 pertaining to Repetition of Same Floor Plan and Street Elevation to require skipping 4 lots on the same side of the street and 2 lots skipped on the opposite side of the street before repeating the same elevation and no floor plan repeated side by side. 3. Revise the current requirements of Section 3.4.F.9, Village Residential Front Entry Cover, from required 40 sq. ft. minimum front porch with 4 ft. minimum depth to 20 sq. ft. minimum front porch with a 5 ft. minimum width. Revisions to Residential District Requirements of the Zoning Ordinance Public Hearing Item 1, February 5,2002 Page 2 of 3 4. Revise the Village Residential Garage Door requirements of Section 3.4.F.11 to define the required offset as 12 inches. Issues: 1. Exterior Façade Materials The requirement for 100% brick or stone laid units currently allows non-masonry products to be used on portions of the structure where there is no structural support for masonry as well as to establish an historic or architectural style. Composite masonry products now offer the same durability as brick and stone as well as added flexibility in design. The required 100% brick or stone tends to produce subdivisions with very little visual diversity, and appears arbitrary the goal of the new flexible point system. A broader masonry definition is permitted for nonresidential construction. 2. Repetition of Same Floor Plan and Street Elevation The current requirement to skip at least 7 lots on the same side of the street has been identified as being difficult to implement for several reasons. First, the ordinance requirement for short curvilinear streets makes this requirement unnecessary in order to eliminate conflicts of similarity. Secondly, the residential building industry attempts to accomplish the intended diversity by skipping 3 or 4 lots, when allowed to do so by overt market pressures and model homes. Finally, the requirement to skip lots greatly reduces buyer choice in the final stages of development of a subdivision. 3. Village Residential Front Entry Cover The current requirement for a front porch to be 40 sq. ft. with a minimum depth of 4 ft. appears to limit the desired flexibility intended to result from the point system. Large porches introduce a standard or theme that reduces design creativity and variation. The larger dwellings of the Suburban and Country Residential Districts currently require only a 20 sq. ft. front porch. 4. Village Residential Garage Doors The requirement for an offset in the façade of the garage was never defined for Village Center dwellings. The other residential districts on larger lots have the offset defined as five feet minimum. The proposed offset of 12 inches is considered adequate to achieve the desired visual articulation as well as economically achievable without special foundation designs. Revisions to Residential District Requirements of the Zoning Ordinance Public Hearing Item 1, February 5, 2002 Page 3 of 3 Staff Recommendation: Staff is prepared to enforce the requirements as adopted until additional direction is provided by the Commission and City Council. Attachments: Survey of Requirements of Selected Neighboring Municipalities Table Comparison of Architectural Requirements Architectural Standards Survey 40 a J .G \i Qo o o 1r 4' , /4'0‘ o a ,,o 1 a � no no no no base req 1st flr 75% min 80% per 100% brick % Masonry 80% 75% response response response at this time 75% 50% per elev 750/0 elev or stone not as sub for Stucco yes yes n/a no masonry yes , yes no EFIS yes no n/a yes same as above yes yes no Hardy Plank no yes n/a yes same as above no no no Siding yes no n/a no same as above no no no Wood yes n/a yes except Alum yes n/a yes in H.D. Vinyl yes n/a yes Chimney Encl yes no n/a no yes yes yes yes Brick yes n/a yes yes yes yes Siding yes n/a no yes yes yes Roof Pitch n/a n/a n/a n/a n/a n/a n/a 6:12 Roof Matis Asphalt yes min class c n/a yes min class c yes yes yes Wood yes yes n/a no yes yes yes no Metal no yes n/a yes yes yes yes yes Textured yes yes n/a yes yes yes yes yes Tile/Slate yes yes n/a yes yes yes yes yes Lots Skipped n/a n/a n/a n/a n/a 5 & 5 n/a 7 & 5 Porch n/a n/a n/a n/a n/a n/a n/a 40 sf Mailbox Curbside only yes per P.O. regs n/a P.O. regs per P.O. regs yes yes yes Paired @ lot line yes n/a yes yes yes Att'd to house no n/a yes no no Garage Rear entry yes yes n/a yes yes yes yes yes Side entry yes yes n/a n/a yes yes yes yes Offset req. none yes none none none yes none yes 5 ft in C/R Offset distance n/a 18 ft n/a n/a n/a 20 ft n/a or S/R dist Note: n/a means that the subject city has no base standards for this category Architectural Requirements (For all Village Residential Districts) Element Base Standard Desirable Points 1. Exterior Facade Material ° none N/A 100%masonry on front facade and 80%masonry on remaining walls(masonry to include brick, stone and composite cementitious products as approved by Building Official) 6. Units with same floor plan a. 7 lots skipped both sides of 10 and same street elevation 4 lots skipped on same side of street street and 2 lots skipped on b. 9 lots skipped both sides of 20 opposite side of street no same street floor plan side by side 40-er 4-lets-opposite-side 20 9.Village Residential front a. Porch railing part of front 5 entry cover 4'minimum depth porch design 20 sf minimum front porch,5' 40 minimum widthminimum-depth b 40 sf minimum front porch, 10 5' minimum width 11.Village Residential Offset 12 inches minimum, Not on primary street elevation 20 garage doors maximum 50%of elevation Architectural Requirements (For all Suburban and Country Residential Districts) Element Base Standard Desirable Points 1. Exterior Facade Material ° none N/A 100%masonry on front facade and 80%masonry on remaining walls(masonry to include brick, stone and composite cementitious products as approved by Building Official) 6. Units with same floor plan a. 7 lots skipped both sides of 10 and same street elevation 4 lots skipped on same side of street street and 2 lots skipped on b. 9 lots skipped both sides of 20 opposite side of street no same street floor plan side by side 40-er and--different-stfeet elevation 4-lets-opposite-side 8. Garage doors on street Offset five feet minimum, a. Doors recessed from main 10 elevation maximum 50%of elevation front facade b. Not on primary street 20 elevation Townhouse District(TH/15) Lot Area 4000 Interior units that have no side yards,such as,zero lot line,attached single family, with 3 or more units—3000 sf minimum,all others 4000 sf minimum Minimum square footage 450G 1000 r . Oily Of Wylie ill Public Hearing Item No. 2 Revisions to Townhouse (TH/15) District Requirements Planning and Zoning Commission Meeting: February 5, 2002 Applicant: City of Wylie Summary: New Residential Districts and requirements were adopted in July, 2000 which combined Duplex (2F) and all other Single Family Attached (SF-A) residential types into one Townhouse (TH/15) District. However, these new requirements have not been tested. At the Work Session held on January 15, 2002, the Planning& Zoning Commission directed that a Public Hearing be called to consider the new Townhouse (TH/15) District requirements. The following proposed revisions were brought forward to consider for possible recommendation to the City Council for adoption. 1. Lot Area—Current requirements: 4000 square feet minimum Proposed: 3000 sq. ft. minimum for interior lots where 3 or more units are attached on zero lot lines and 4000 sq. ft. minimum for all others. 2. Minimum Unit Size—Current requirements: 1500 sq. ft. minimum Proposed: 1000 sq. ft. minimum. The previous Two-Family District (2F) lots were required to be platted in pairs of no less than 8500 sq. ft. or 4250 sq. ft. minimum per each dwelling unit. The minimum dwelling unit size was 900 sq. ft. The Single-family Attached District (SF-A) required units to be built on separately platted lots of at least 3000 sq. ft. with a minimum dwelling size of 1200 sq. ft. The new Townhouse District(TH/15) requires that each unit be on a separately platted lot of 4000 sq. ft. with a minimum dwelling unit size of 1500 sq. ft. By comparison the Multi-family District (MF) minimum dwelling sizes remained the same with a two bedroom unit of 900 sq. ft. and a three bedroom unit of 1000 sq. ft. Revisions to Townhouse(TH/15)District Requirements Public Hearing Item 2, February 5, 2002 Page 2 of 2 Issues: 1. The private development sector points out that duplex and attached dwellings differ from Single-Family detached dwellings in that they are more often purchased at least partially as investment properties and for rental occupancy. Therefore the dwelling size should correspond more closely to its multifamily competitor than to the owner occupied detached housing. The required size of multifamily units is 900 to 1000 sq. ft. compared to 1500 sq. ft. for attached units and 1700 sq. ft. for the smallest allowed detached housing. 2. When three or more Single Family Attached units are constructed together the interior units which do not have any side yards are forced to either increase lot depth or dwelling width in order to meet the minimum lot dwelling sizes. By reducing the minimum size of interior lots, consistency in size of lots and dwellings can be achieved. Staff Recommendation: Staff agrees that attached single family dwellings should be smaller than detached single family dwellings. However, staff believes that a minimum size of 1200 sq. ft. or 1300 sq. ft. both addresses the special qualities of single family attached dwellings and better serves to transition from the smaller multifamily units (1000 sq. ft.) to the detached single family units (1500 sq. ft. min.). Staff is prepared to enforce the requirements as adopted until additional direction is provided by the Commission and City Council. -5 1-. -04,7tii 6wlXi1 '- IA-114ii ,�,'- i4 Sprawl Cost Doess� t * rcove, UsAll? f.,,,,r,„,,..4.. 44,,,,,..._ iY . Patterns on Public the Effects of Housing tv4-' Water and Sewer Costs Cameron Speir and Kurt Stephenson S rrarrd is a term Frequently used to describe the spatial pattern of con- temporary residential development. Many proposed definitions char- This article assesses the public water acterize a spread-out,low-density housing pattern as the pri mar 'char- and sewer costs associated with alter- acceristic of1i7 aw°Crmcs o sprawl express concern on many evl •ls about native housing patterns. These pat- this housing pattern.The)'assert that s-irawl housing patterns mas'des ride terns are defined in terms of lot size, air and water quality,destroy open space ant. prime farn�ancl contrib- tract dispersion,and distance from ex- ute to the deterioration of urban centers'Burchell&Listokin, 1995). isting water and sewer service centers. One particular concern is the effect of sprawl on the fiscal costs to local The engineering cost model presented governments(Esseks&Sullivan, 1999).Researchers in economics and plan- here gives empirical evidence of how Wing assert that sprawl costs local governments more for infrastructure and sensitive local government service public services than more compact forms of growth (Burchell et al., 1992; costs are to the spatial pattern of sin- Downing&Gusteley, 1977; Frank, 1989).The effects of-spatial pattern on gle-family residential development. public costs have been analyzed for many services, including roads,water, The results show that more spread out dy toto sewers,parks,education,solid waste disposal,fire protection,and police pro- housing patterns are moree cosupply with public water and sewer tection(Transportation Research Board, 1998).Despite the continuing con- services,but that shifting a majority of cern,many studies report that the costs of providing some of these services these costs to the private sector may are relatively insensitive to spatial pattern.Schools,by Ear the largest corn- be a relatively simple matter. ponent of local government costs,are one such service(Burchell et al., 1992; Frank, 1989).Others include solid waste collection and fire and police pro- Speir is a regulatory pricing analyst at Paci- tection(Burchell&Listokin, 1995). fiCorp. He completed this work as a re- search associate in the Department of In contrast,a number of studies have found that the costs of providing Agricultural and Applied Economics atVir- infrastructure-intensive services,such as roads,water,and sewers,are much ginia Tech.Stephenson is an associate pro- more sensitive to spatial pattern (Burchell& Listokin, 1995; Downing& fessor in the Department of Agricultural Gusteley, 1977;Duncan et al., 1989; Peiser, 1984).Water and sewer services and Applied Economics at Virginia Tech. are of particular interest to local governments,since these governments bear Journal of the American Planning Association, the vast majority of these costs.Duncan et al.(1989)and Frank(1989)found Vol.68,No.1,Winter 2002.®American that water and sewer costs for compact,contiguous housing patterns are 60 Planning Association,Chicago,IL. and 66%,respectively,of chose for spread-out patterns. • • 56 APA Journal•Winter 2002•Vol.6S,No. 1 DOES SPRAWL COST US ALL? • While the literature indicates that water and sewer houses in a hypothetical medium-size town of 30,000 costs are sensitive to spatial pattern,previous cost stud- people.' Each house is assumed to be occupied by an av- ies have been limited in two ways. First,sprawl itself is erage of 3.5 people.The area is assumed to have basically not always precisely defined.In fact,"even the most cur- flat topography. Given these initial conditions, 60 dif rent literature on sprawl tends to describe its attributes ferent development scenarios were identified. Each de- rather than quantity,them[emphasis in original)"(Trans- velopment scenario settles the new population in a par- portation Research Board, 1998, p. 17). Often spatial titular spatial pattern based on different combinations patterns are described in general terms such as trend or of values for each of the three spatial attributes, where unplanned versus planned (Burchell et al., 1992; Peiser, the value of that attribute was allowed to vary across all 19S4). If spacial patterns are quantified, they include scenarios while the other two were held constant. only one or two spatial attributes,such as density or dis- A spreadsheet-based cost model was then used to tance(Downing&Gusteley, 1977;Frank, 1989).Second, calculate the cost of providing public water and sewer the influence of spatial pattern on costs is not always services for each scenario.First,the total water and sewer clearly separated from other factors such as population, system infrastructure requirements of each scenario service standards, and demographic characteristics. were estimated. Length of pipe; number of valves, hy- Windsor(1979),for example,suggests that much of the drants,and manholes;number of-booster pumps;and cost savings attributed co higher density is in fact due to pump energy requirements were determined according differences in the number of people, types of housing, CO the pattern ofeach scenario.Second,these system re- construction standards,and demographic assumptions. quirements were used in conjunction with unit cost data In particular, the water and sewer literature frequently from generally accepted industry sources to calculate fails to isolate the relationship between total water use costs.These costs include primarily the capital and en- and spatial pattern. ergy costs associated with water distribution, but not The present study aims to extend and refine the re- storage.3 Some scenarios assumed higher per-capita lationship between the costs of providing water and water use for larger lots. In these scenarios, the addi- sewer services and housing patterns by isolating three cional water treatment costs Were calculated. Costs for specific spatial attributes of these patterns. In the fol- water storage and wastewater treatment were not esti- lowing section,these attributes are described in terms of mated because they are thought to be relatively insensi- three types of-separation: • tare Co spatial pattern.' System dimensions for all sce- narios and sources of cost information are more fully I. lot size(separation between houses); explained in the Appendix. 2. tract dispersion (separation between By examining how the total cost of providing water development tracts;sometimes called non- and sewer services to this new population changes across contiguous or`skipped-over development).and scenarios,the cost consequences of-each spatial attribute 3. distance(separation from existing water and can be isolated.For instance,one set of development sce- sewer centers). • narios varies lot size,while holding tract dispersion and This article aims to identify the relative contribu- distance constant.Any differences in costs among these Lions of these three spatial attributes to the cost of pro- scenarios can then be attributed co changes in lot size. riding water and sewer services while holding all other Development Scenarios: Definitions and cost-influencing factors constant.The methods section Values describes the cost simulation model that was used. It also describes alternative development scenarios that The following section describes the development were used to isolate the relative contributions of the scenarios for which costs were estimated.First,we define three spatial attributes.The following section presents the values used to quantify each of the three spatial at- the results.The article concludes with a discussion of the tributes. Second,we identify the specific combinations policy implications of the results. of these values in each development scenario. The smallest development tract consists of 750 housing lots arranged in a grid 27 lots long by 28 lots Methods wide.'This pattern applies regardless of the lot size.Fig- The capital and energy costs of providing water and ure 1 is a diagram of the arrangement of housing lots sewer services to different housing patterns were calcu- and the water and sewer service grid within a small Be- lated using a cost simulation model. The simulation velopment tract.The triple solid line running lengthwise began by assuming existing water and sewer services through the middle of the grid represents large-diameter would be expanded to 3,000 new single-family detached water transmission and sewer interceptor pipes that con- APA Journal•Winter 2002 •Vol.CS,No. I 5 CAMERON SPEIR AND KURT STEPHENSON Tract dispersion measures how far development tracts r are located from each other.They may be clustered to- i I , � r Ti T { I I T , i 1 i T 1 ' £ether to form a single large tract of as many as 3,000 II i t [.. .. 1.1111.1.1...1.1...1.1.1. M_1_ ._. 1.1..�.�.-!_,.1...._, lots or they may be smaller tracts arranged in separate � , l , jt , : ; 1f i I 1 � � , ; I ; t ; � , 1� ! , ! WI locations.in this model,tract dispersion can take any of � t , ; j ! I ! ,1 i ! � � j � � � ( ( four different values: 1,2 3,or 4 tracts,as shown in Fig- I l _._..1...t...1._...:._,...1. _ __.._ _ F......,.__..1 1 1 l ; 1 ,.- t tare 2.The term hie mitt dispersion describes four tracts ' 111while lour tractdispersion refers to one tract.The black rec- 115113."11411141i !.►._.1_�_i.1 , : tangle in each tract pattern represents the existing the 1 water and sewer service center.The white rectangles rep- resent the development tracts,with the number of hour- ; 1 1 1 l 1 i 1 1 1 t ; III _I ing lots in each. Moving from left to right,the develop- 4 1 1 f I t 1 1 1 1 1 !. 11 i._i 1 1 j t L I•._' i_L._ (hunt tracts become more dispersed as their number ' ' ,. i 1 ± t I i j ! i j i i i increases from One; CO four.Note that the tracts In each •1 ! development are separated by enough space so that each I ' I , ! i ! 1 1 ! ! _.. ..!._1._.:_..,...:..;.._.- .• -r f , l requires its own water transmission and sewer in- ! '. f !. I I 1 �: � `� 11; 1 i ' ! 1 ! I j j tract : : terse ptor pipes. ; :. 1 Ij _ j �. ._ ._...._1...__.;.. .-'.-i. �_�. ...=!-=-�-:-;- Distance measures the separation between develop 11 . ijj I !. 't ` 1 .i ' . ' l....11...1 I ' ' ' ' ' Halo tracts and existing water and sewer service centers. In all scenarios, the service center contains water and \wastewater treatment plants.The distance from the cen- ter, therefore, is the length of water transmission and LEGEND sewer interceptor pipes required to serve the new devel- '::! Housing lot opulent. For example, tracts located 5 miles from the ' eo Water transmission and sewer interceptor pipes center require 5-mile eater transmission and sewer in- - Water distribution and sewer collector mains tercepror pipe extensions. Distance may take any of — Water distribution mains only seven different values:0.25,0.5, 1,2,3,4,or 5 miles.The term near distance refers to tracts 0.25 mile from the cen- ter,while f irdist.unce refers to 5 miles. FIGURE 1. Diagram of water and sewer utilities in a small development tract. Development scenario sets.A total of 60 development scenarios were examined.The scenarios were grouped into sets based on the spatial attributes of the housing patterns. Lot size, tract dispersion, and distance were nett the tract to the service center.The single solid lines each analyzed by comparing four sets of scenarios,for a represent the water distribution and sewer collector ttotal or s cha12 sec .V'ithie n a sec ofr woscenarios, e olne attributensf mains that connect individual houses to the larger trans- mission and interceptor pipes. t. When the cost ()leach lot-size scenario was calculated, e for example, differences in the total cost of providing Lot size measures the separation between housing P ' lots in a development tract. in this analysis,lot size may water and sewer services could be attributed to different take any of four different values:0.25 acre,0.5 acre,0.75 capital and operating costs that result only from chang- acre, or 1 acre. For discussion purposes, the term small ing thet size.size.all four sets of scenarios, lot size varies lot lot is used to describe tracts with 0.25-acre lots,while large lot refers to 1-acre lots. Lot size determines the length of from 0.25 acre to 1 acre. in secs 1 and 2,tract dispersion frontage for each lot,which in turn affects the length of is 1 (low), and respective distances from the existing water distribution and sewer collector mains needed to water and sewer service centers are 0.25 miles (near) serve the development tract. For example, a square- and miles le s(fa`)_tIne sets s 3nand 4,ns are tract dispersion il p(near)n is 4 and shaped 1-acre lot has a 210-foot frontage. It is more (high), P' likely,however,that lots are rectangular,typically with a 5 miles(f r). ro+:. In all four sets ofscenarios,tract dis- sponds ratio, meaning that every foot of frontage corre- u f'a's to 3 feet of depth (New Jersey Office of State persion varies from 1 co 4.In sets 1 and 2,distance is 0.25 Planning, 1996).6 This ratio was used for each of the four an mile5 acre(small) d 1 acre(large),In sets 3 and 4,distance s 5 miles(far), lot-size values. 58 APA Journal•Winter 2002•Vol.6S,No. 1 DOES SPRAWL COST US ALL? 1000 750 3000 1500 1000 750 I 750 MIN 1111111 H IN 1500 1000 750 1 2 3 4 Development Development Development Development Tract Tracts Tracts Tracts • LEGEND Water transmission/sewer interceptor pipes ® Water and sewer service center 750 Development tract with number of lots FIGURE 2. Diagram of tract dispersion types. and respective lot sizes are 0.25 acre (small) and 1 acre personal communication,August 17, 1999).Consump- (large). don of 100 gpd per person implies a need for 1.05 mil- Distance.In all four sets of scenarios,distance to the lion gallons per day under each scenario. existing service center varies from 0.25 mile to 5 miles. We also considered a situation in which water de- In secs 1 and 2,tract dispersion is 1 (low),and respective mand increases for larger lots because of landscape lot sizes arc 0.25 acre(small)and 1 acre(large). In sets 3 watering.We conducted a limited number of tests to de- and 4,tract dispersion is 4(high),and respective lot sizes [ermine the effects of this increase on costs. In these are 0.25 acre(small)and 1 acre(large). tests,water use per capita is increased 25%to 125 gpd for 1-acre lots.The magnitude of the relationship between Water Use lot size and water use is difficult to isolate,but the 25% Total water use for each household was initially held increase may be considered conservative (Hammer, constant across all scenarios. All 3,000 houses are as- 1996). The effects of this increase will be reflected in sumed to be single-family detached residences.Each res- higher costs due to additional infrastructure (larger idence is assumed to house 3.5 people,and each person pipes), higher pumping costs, and higher water treat- is assumed co consume 100 gallons of water per day merit costs. (gpd).This figure is based on several recent estimates. The U.S. Geological Survey (2001) estimates that the Results average person uses 80-100 gpd in the home.Virginia's Blacksburg-Christiansburg-VPI Water Authority(1999) The following results report the total annualized reports that the average person uses 125 gallons per day, costs to provide water and sewer services and the relative and other estimates range from 60 to 200 gpd U.Reilly, contribution of each spatial attribute. APA Journal•Winter 2002$Vol 68,No. 1 59 CAME:ON SI'EiR AND KURT STEPHENSON Effect of Lot Size on Cost of water transmission and sewer interceptor pipes, the number and size of water pump stations, and energy The results presented in Tables 1 and 2 clearly show costs. that i: creasing lot size can have a pronounced effect on the cost of providing water and sewer services.Table 1 Effect of Distance on Cost gives :he annualized water and sewer costs for each see -nario sec. Across all sets, smaller lots cost less to serve Table 5 presents the annualized costs of providing than larger lots when distance and tract dis version ark water and sewer services when development is located at le constan:.In lot-size scenario set 1(low tract disper- various distances from existing service centers.These re- si on and near distance),when lot size increases from 0.25 sults show that increasing these distances can increase co i acre,costs nearly double, a S560,000 increase. An- costs substantially. For example,in set 1,increasing the nual costs per household increase from S204 to 5392. distance of the development tract from 0.25 mile CO 4 The higher costs are due to the longer distribution mains miles from the service center increases costs 5186,500, required for larger lots. Friction head losses are also from S61 3,000 co 5799,500.This 16-fold increase in dis- high.er with longer lengths of pipe, which increase cance increases costs by 30%. pumping,costs. Table 6 shows how costs change as distance in- -able 2 shows how costs change as lot size increases creases.The absolute change for increases of 1 mile is in 0.25-acre increments. In secs 1,2,and 3, the size and nearly the same within each set of scenarios. For exam- rate of the increases in costs decreases as the lots become pie, in sec 1,each 1-mile increase in distance between 1 larger.This relationship is an artifact of the geometry of and 4 miles increases costs approximately S50,000.8 In the rectangular lots with a 1:3 frontage-to-depth ratio: sec 4, each 1-mile increase between 1 and 5 miles in- The increase in the length otfrontage is proportionally creases costs approximately S I22,000. less :pan the increase in acreage.Table 2 also shows that the increase in costs is fairly constant among sets for the Summary of the Effects of Spatial same increase in lot size. For example, as lot size in Attributes creases from 0.25 to 0.5 acre,the increase in costs is be- tween S233,046 and 5235,4I-I across all four sets ofsce- Using the results in Tables I through 5, Figure 3 tar:os.These small differences in the rate of increase are summarizes how the total cost of providing water and explained by differences in sewer interceptor pipe diam- sewer services changes when each spacial attribute is eters and small differences in pump energy costs. doubled and quadrupled while the others are held con- stant.The change in cost is calculated by averaging the Effect of Tract Dispersion on Cost percent change in costs across the different sets of sce- narios for lot size,tract dispersion,and distance.For in- Table 3 shows the annualized costs of providing stance,doubling,lot size from 0.25 to 0.5 acre increases wa:er and sewer services when tract dispersion is Varied. total water and sewer costs 3S,27,36,and 20%, respec- \\i:pin each scenario sec, differences in costs from tively(see Table 2).Thus,the average increase in cost for changes in tract dispersion are modest,but among the the sec of lot-size scenarios is 30%(see Figure 3).By com- sets of scenarios,costs are significantly different.This parison,doubling tract dispersion from 1 to 2 tracts and indicates that the other spatial attributes have more pro- distance from 0.25 to 0.5 mile increases costs an average flounced influences on costs than tract dispersion.In all of 6% and 3%, respectively. Quadrupling lot size from cases,costs increase as tract dispersion increases due to 0.25 to 1 acre,tract dispersion from 1 to 4 tracts,and dis- inc:eases in pipe length and pump capital costs. ta11ie from 0.25 to 1 mile supports the conclusion that Table 4 shows how costs change as tract dispersion the costs of providing,water and sewer services are most inc:eases. These results show that more development sensitise to eliai ges in�toc size. tracts increase the cost of providing water and sewer ser- vices,but the increases are not as large (in absolute or Composition of Costs percentage entage terms) as increases due to lot size. Not sur- prisingly, the additional cost of serving more tracts in- The primary reason total costs are most sensitive to creases with distance from the existing service center.For changes in lot size is that infrastructure within the de- la:distant tracts,as in secs 3 and 4,the increase in costs ve opulent tract—water distribution mains and sewer as :race dispersion increases is between 6 and 14%. For co ector mams—are the two arvestcom onents of total near distant tracts,as in secs 1 and 2,these costs increase cost.Ta e s lows each infrastructure component as a unit/ 1 to 2%. Cost differences among scenarios are at- proportion of total cost for selected secs of scenarios.For tri:,uted to changes in the length of pipes,the diameter example,in lot-size scenario sec 1 with 0.25-acre lots(the 60 Alva\Journal•Winter 2002 •Vol.6S,No. I DOES SPRAWL COST US ALL? TABLE 1. Lot size: Effect on annual water and sewer costs.* Set 1 Set 2 Set 3 Set 4 Low tract Low tract High tract High tract dispersion, dispersion, disperson, dispersion, Costs by lot size near distance far distance near distance far distance 0.25 acre Total $613,383 $878,158 5650,171 51,179,836 Per household 204.46 292.72 216.72 393.28 0.5 acre Total 848,797 1,112,120 883,21S 1,412,882 Per household 282.93 370.71 294.41 470.96 0.75 acre Total 1,035,967 1,299,290 1,069,655 1,653,783 Per household 345.32 433.10 356.55 551.26 1 acre Total 1,176,344 1,447,390 1,212,206 1,793,611 Per household 392.11 482.46 404.07 597.87 •AII costs are annualized over 30 years using a 7%discount rate. TABLE 2. Lot size: Effect of changes on total water and sewer costs." Set 1 Set 2 Set 3 Set 4 Low tract Low tract High tract High tract Cost changes by dispersion, dispersion, disperson, dispersion, change in lot size near distance far distance near distance far distance 0.25 to 0.5 acre Absolute 5235,414 5233,962 5233,047 $233,046 Per household $74.47 $77.99 $77.68 $77.68 Percentage 38% 27% 36% 20% 0.5 to 0.75 acre Absolute $187,170 $187,170 $186,437 $240,901 Per household 562.39 $62.39 $65.15 $80.30 Percentage 22% 17% 21% 17% 0.75 to 1 acre Absolute $140,377 S148,100 $142,551 $139,828 Per household $46.79 449.37 547.52 $46.61 Percentage 14% 11% 13% 8% •AII costs are annualized over 30 years using a 7%discount rate. APA Journal•Winter 2002•Vol.68,No. 1 61 CAM(ERON SPEIR AND KURT STEPHENSON TABLE 3.Tract dispersion: Effect on annual water and sewer costs.* Set 1 Set 2 Set 3 Set 4 Costs by Near distance, Near distance, Far distance, Far distance, number of tracts small lot large lot small lot large lot 1 tract Total $613,383 $1,176,344 5878,158 $1,447,390 Per household 204.46 392.11 292.72 482.46 2 tracts 1,004,707 1,565,121 Total 623,753 1,184,168 Per household 207.92 394.72 334.90 521.71 3 tracts Total 636,053 1,197,685 1,099,452 1,659,040 Per household 212.02 399.23 366.48 553.01 4 tracts Total 650,171 1,212,206 1,179,836 1,793,611 Per household 216.72 404.07 393.28 597.87 'All costs are annualized over 30 years using a 7%discount rate. TABLE 4.Tract dispersion: Effect on changes in total water and sewer costs.* Cost changes Set 1 Sec 2 Set 3 Set 4 by change in Near distance, Near distance, Far distance, Far distance, number of tracts small lot large lot small lot large lot 1 to 2 tract Absolute $10,370 $7,824 5126,549 $117,731 Per household $3.46 $2.61 542.18 $39.24 Percentage 2% 1% 14% 8% 2 to 3 tracts Absolute $12,300 $13,517 $94,745 $93,919 Per household $4.10 $4.51 $31.58 $31.31 Percentage 2% 1% 9% 6% 3 to 4 tracts Absolute $14,118 $14,521 $80,384 $134,571 Per household $4.71 $4.84 $26.79 $44.86 Percentage 2% 1% 7% 8% 'All costs are annualized over 30 years using a 7%discount race. 62 AI'A Journal•Winter 2002•Vol.68,No. I DOES SPRAWL COST US ALL? 80 N +' ■ ■ ■ ■ ■ 0 70 • Doubling one attribute: Lot size 0.25 acre to 0.5 acre, . ■ • . ■ L ••••••••• tract dispersion 1 to 2, distance 0.25 to 0.5 mile. d ■..... o 60 •..... _ - - . . . . .. . . . .• Q Quadrupling one attribute: Lot size 0.25 acre to 1 acre, •� • tract dispersion 1 to 4, distance 0.25 to 1 mile. ..... L. U . . . ■ 0 50 •:ti•:::•- - - - al a.... 3 . . . . . U.... ■ ■ ■ . ■ o . •• .. ••.... __ •..•.• 30 - - - •::::. •..... En •....■ C •a.... as •a.... •::::. o 20 - - - • • :• c ■ ■ ■ ■ ■ V••■••■•■••■•■ o •a. ::•:•:. ■ ■ ■ . . ■ . . . . ■ ■ ■ ■ Lot size Tract dispersion Distance Spatial attribute FIGURE 3. Summary of the sensitivity of water and sewer costs to changes in spatial attributes. most compact pattern, low tract dispersion, and near 48"i,of the total cost of S 1.179,836(see Table 1).On as- distance),the distribution and collector mains make up erage, the results presented in Table 7 show that to- 64%and 29%, respectively, of the total annualized cost ,-ether, water distribution and sewer collector mains ofS613,3S3(see Table 1),In other words,93%of all costs within the development tracts make up 7S°o of costs are attributable to internal tract infrastructure.Bycom- across all scenarios(54%and 24%on average of all costs, parison,the same infrastructure components in lot-size. respectively).Table 7 also shows that water distribution scenario set 4 for 1-acre lots comprise 44 and 19%, re- makes up a much greater proportion than wastewater spectively, or 63% of the total cost of S 1,793,611 (see collection.This is mainly because higher grade,more ex- Table 1). pensive piping is used for water distribution. These results show that the proportion of internal Effect of Higher Water Use tract infrastructure costs to total cost can vary widely ac- cording to spatial pattern. The proportion of costs for It is possible that households on larger lots will con- internal distribution and collector mains are lowest sume more water,primarily for landscaping.The effect when lot sizes are small,distance and tract dispersion is of increased water use on distribution costs was exam- high,and distance is far(lot-site scenario set 4 at 0.25 fined by recalculating the cost of delivering water to four acre). In this case,total internal infrastructure costs are scenarios with large (1-acre) lots.These recalculations APA Journal•Winter 2002•Vol.68,No. 1 63 CAMERON SPEIR AND KURT STEPHENSON TABLE 5. Distance from existing service centers: Effect on annual water and sewer costs." Set 1 Set 2 Set 3 Set 4 Low tract Low tract High tract High tract dispersion, dispersion, disperson, dispersion, Costs by distance small lot large lot small lot large lot 0.25 mile Total $613,383 $1,176,344 5650,171 $1,212,206 Per household 204.46 392.11 216.72 404.07 0.5 mile Total 625,790 1,190,203 678,048 1,242,806 Per household 208.60 396.73 226.02 414.27 1 mile Total 650,604 1,217,921 733,802 1,304,007 Per household 216.87 405.97 244.60 434.67 2 miles Total 700,233 1,273,358 845,311 1,426,408 Per household 233.41 424.45 281.77 475.47 3 miles Total 749,861 1,333,428 956,819 1,548,809 Per household 249.95 444.4S 318.94 516.27 4 miles Total 799,489 1,390,409 1,068,327 1,671,210 Per household 266.50 463.47 356.11 557.07 5 miles Total 878,158 1,447,390 1,179,836 1,793,611 Per household 292.72 482.46 393.28 597.87 'All costs are annualized over 30 years using a 796 discount rare. assume that water use is 125 gpd per person,instead of tion costs of treating the additional quantity of water 100 gpd.Higher daily water use should increase both the were considered. We assumed these costs of treatment size of certain components of the distribution system co be $0.75 per 1,000 gallons (B&B Consultants, Inc., and the total cost of water treatment. Transmission 1997). pipes and associated valves must be larger to accommo- Table S presents the additional costs of treating and date higher flow rates. Pump stations must be sized to delivering more water to larger-lot developments.The pump larger quantities of water,thus increasing capital last column shows the percent increases of total water costs,and energy costs will be higher.Collection costs in costs,which include both water distribution and treat- the sewer system should not be affected,as the increased ment. Total cost increased from $95,000 to almost water will be returned as groundwater. Changes in in- S 128,00 per year.This represents an 11 to 12%increase, frastructure needed as water use increases were calcu- which is primarily due to higher water treatment costs laced under the different assumptions about per-capita rather than higher distribution costs. water use. An increase in treatment costs was also included in Implications and Conclusions these calculations. If large lots use 125 gpd per person instead of 100, art additional 262,500 gpd must be The results show that smaller lots,shorter distances treated. It is assumed that the water treatment facilities between existing centers,and-lower tract dispersions re- have sufficient capacity to treat this water with no newduce water and sewer costs.These results confirm previ- capital investment.Therefore,only the variable opera- ous wort; on the costs of alternative housing patterns • • • 64 APA Journal•Winter 2002•Vol.68,No. 1 DOES SPRAWL COST US ALL? TABLE 6. Distance from existing service centers: Effect on changes in total water and sewer costs.* Set 1 Set 2 Set 3 Set 4 Low tract Low tract High tract High tract dispersion, dispersion, disperson, dispersion, Cost changes by P a lot small lot large lot large change in distance small lot $ 0.25 to 0.5 mile 527,877 $30,600 Absolute $12,407 $13,859 $2.61 $9.29 $10.20 Per household $4.144% 3% Percentage 2% 1% 0.5 to 1 mile $55,754 $61,201 Absolute 524,814 527,718 58 $20.40 518. Per household $8.27 59.24 8% 5% Percentage 4% 2% 1 to 2 miles 5111,509 $122,401 Absolute 549,628 555,437 Per household $16.54 518.48 $37.17 $40.80 Percentage 8% 5% 15% 9% 2 to 3 miles 5111,508 $122,401 Absolute $49,628 $60,070 Per household $16.54 520.02 537.17 540.80 Percentage 7% 5% 13% 9% 3 to 4 miles Absolute $49,628 556,981 5111,508 $122,401 $37.17 $40.80 Per household $16.54 518.99 12% 8% 7% 4% Percentage 4 to 5 miles 5111,503 5122,401 Absolute 578,669 556,981 Per household $26.22 518.99 537.17 $40.80 Percentage 10%�o 4% 10% 7% *All costs are annualized over 30 years using a 7%discount rate. (including l3urchell et al., 1992;Duncan et al., 1989;Es- base fees on water usage using average cost methods.If seks et al., 1999; Frank, 1939).The present study,how- average cost pricing for water and sewer services is used ever, extends these works by explicitly isolating the of to recover costs,users in less compact spatial patterns fects of three spatial attributes on water and sewer costs. will pay less than their true cost of service,while users in \X'hile the term sprawl may be used to describe a variety of more compact patterns will pay more.In these instances, spatial patterns of housing developments,this analysis sprawl may indeed"cost us all"by subsidizing residents suggests that the lot-size aspect of sprawl is its most cost- in dispersed develo ment atterns and giving up the sensitive spatial attribute(see Figure 3). total cost of provi ing.water and sewer services. One public policy concern about differences in costs The results presented here show that lot size(or den- is who pays. Depending on the particular financing ar- sity)is the spatial attribute that has the most impact on rangements, residents in denser,more conti uous de- water and sewer costs. Costs related to lot size,includ- velopments could end u subs n in awater and sewer ser- example, ing water and sewerlines, are peasily attributable to vices or t ose in more g patterns. specific housingunits and development tracts.This sug- some public utilities install and pay for all intrastate- gests that the most simple policy solution for overcom- ture. Many are also self-financing entities that are re- ing a"sprawl subsidy" may also be the most effective. quired—legally,in some cases—to recover their costs of The results of this study indicate that on average 78%of service through fees. Most utilities charge a flat rate or the costs associated with spatial pattern are for infra- APA Journal•Winter 2002•Vol.68,No. 1 CAMERON SPEIR AND KURT STEPHENSON TABLE 7.Contribution of infrastructure components to annual water and sewer costs. %of water costs %of sewer costs Distribution Transmission Pump Pump Collector Interceptor main mains capital energy mains mains Scenario set (internal) (external) costs costs (internal) (external) Lot size 0.25 acre Set 1 64 1 4 1 29 1 Set 4 33 28 4 <1 15 19 1 acre Set 1 66 1 2 1 30 <1 Set 4 44 18 3 <1 19 16 Tract dispersion 1 tract Set 1 64 1 4 1 29 1 Set 4 54 12 2 1 24 8 4 tracts Set 1 61 3 7 1 27 2 Set 4 44 18 3 <1 19 16 Distance 0.25 mile Set 1 64 1 4 1 29 1 Set 4 64 1 4 <1 29 1 5 miles Set 1 45 19 3 1 20 12 Set4 44 18 3 <1 19 16 Average 54 11 3 1 24 7 TABLE 8. Effect of increased water use on annual water distribution system costs.* Lot size Distance No. Distribution Treatment Total %of total (acres) (miles) tracts cost cost water cost water cost 1 0.25 1 $23,649 $71,859 $95,508 12% 1 0.25 4 42,888 71,859 114,747 12 1 5 1 22,379 71,859 94,238 11 1 5 4 56,072 71,859 127,931 11 "Daily water use increased from 100 gpd to 125 gpd per person. • 66 APA Journal •\''inter 2002•Vol.68,No. 1 DOES SPRAWL COST US ALL? • structure associated with installing water distribution sprawling developments (a maximum of S 17.24 per and sewer collector mains. Requiring developers to in- month or 52.567 over 30 years), reducing the subsidy stall and pay for water and sewer lines located within a would be unlikely to have a large impact on housing tract during its construction would therefore shift a choices. Given the relative insensitivity of other local large percentage of these spatially sensitive costs to those government services(primarily education)to spatial pat- settling there. Most public utilities currently use such terns, this analysts also su-,-ests that the most signth- strategies.In Roanoke County,Virginia,for example,de- cant consequences of housing patterns may not e oca velopers must install and pay for any infrastructure oc- fiscal outcomes, but_t n t ter enviromental 1n soda curring within a development tract (G. Robertson, per- ones. Further research on the environmental and social sonal communication,November 1999). In other cases, impacts of sprawl should pay closer attention to which however,local government may pay for most water and spatial attributes of housing patterns may be responsible sewer infrastructure required by new development,per- for any negative impacts. haps as an economic growth strategy (Esseks et al., Finally,the analysis presented here suggests a num- 1999).9 her of possible extensions and research needs.This anal- The remaining infrastructure costs associated with ysis assumed an existing medium-size public water and tract dispersion and distance could be also paid for by sewer system serving 3,000 new households. How in- use of proffers and variable impact fees.The results pre- creasing the dimensions of the system(scale economies) stinted here disaggregate the spatial patterns of housing will change the relative influence of spatial pattern is un- developments into discrete attributes.This type of dis- clear. This seems to be a particularly pressing issue in aggregation should be considered when developing var- some fast-growing counties where annual noncom- iable impact fees to recover the costs of capital facilities. pounded growth rates exceed 7 or 8%.Another interest- Variable impact fees,or system development charges,are ing question char is nor addressed is how costs for re- one-time charges designed to recover capital costs asso- placement and maintenance can be equitably financed elated with water or sewer system expansion. More im- once initial facilities outlive their useful lives.Typically, pot-candy,they are structured to reflect differences in set.- all users in the system share these costs, regardless of vice costs by charging higher fees to users in areas that whether lower-density development is responsible for a are more costly to serve. This study has identified as higher portion of the replacement costs.Methods ofpre- higher cost those areas with larger housing lots and dieting aid policy mechanisms for allocating future re- more dispersed development tracts farther from exist- placement and maintenance costs to users based on their ing service centers.The differences in costs reported here marginal contribution,however,could be further devel- could also be used to compute appropriate impact fees oped.Additional research could also investigate the sen- according to the spacial attributes ofa new development. sitivityofcosts to different housing patterns within ade- in practice,widespread adoption of these subsidy- velopment tract. Better quantitative measures of the reducing policy reforms has proven difficult. Yet, this spatial attribute we label tract dispersion would also ap- analysis suggests that once internal infrastructure costs pear warranted. Such a measure (perhaps developed are paid for by those creating these costs,the gains from using graph theoretic tools)would prove useful in con- instituting proffer and impact fee systems are more ducting statistical analyses that attempt to isolate the modest because external infrastructure(water transmis- impact of spatial patterns on fiscal,social,and environ- sion pipes,pump stations,and sewer interceptor pipes) mental indicators. accounts for a relatively small proportion of total costs (5 to 52%,or 22%on average). Ifcosts external to the tract(pipes and pumps)can- ACKNOWLEDGEMENTS not be shifted,that part of the sprawl subsidy is relatively The authors would like to thank the three anonymous referees small and should produce only a modest impact on for their helpful comments. Funding and support for this housing choices. For example,Table 7 shows that costs research was provided by the Virginia Chesapeake Bay Local for external infrastructure comprise 7%of all costs for Assistance Department and the Virginia Water Resources Re- the most compact development pattern and 37%ofcosts search Center. in the most dispersed one.These percentages translate into monthly costs of S 1.19 and 518.43 per household, respectively. The present value of these monthly pay- NOTES ments over a 30-year period would he S178 and S2,745. 1. The new residents would settle the area over a 30-year pe- \\'hile these figures show that residents in compact riod.This represents 100 new households per year,or an developments would be subsidizing residents in more annual growth race ofoyer 1.1%(not compounded)and is APA Journal•Winter 2002•Vol.6S,No. 1 67 CA\MERON SPEiR AND KURT STEPHENSON not an unreasonable forecast for such a town(Town of REFERENCES Blacksburg, 1996).This rate ofgrowth is relatively modest Blacksburg-Christiansburg\'PI Water Authority.(1999).Fas- compared to the national rate of 0.9%in 1998-1999 and civating utter fats [On-line]. Available: <http://www. to the fastest growing counties in the nation, such as h2o4u.org/facts.htnll>. Loudon County,Virginia,which grew by 8.1%(U.S.Cen B&B Consultants, Inc. (1997). Comprehensive water and sewer sus Bureau,2000).It should also be noted that costs were study for Brunswick County,Virginia.South Hill,VA:Author. calculated as if all development costs are incurred in year Bouthillier, P. H. (1981). Hydnudic tables for water supply and one and planners have perfect foresight about which drainage. Ann Arbor, MI: Ann Arbor Science Publishers, spatial patterns and how much development will occur. inc Holding the timing of development constant in this way Burchell,R.W.,&Liscokin,D.(1995).Land,infrastructure,haus- allows for the effects of spatial pattern on costs CO be bet irtoca>ts,and fiscal impacts associated a itbgrou'tb:The literature ter isolated. on the impacts of sprawl versus managed growth. Cambridge, 2. Sewage pumping costs are not considered in this general- ized model because their placement is mainly a function Ni:\:Lincoln Institute of Land Policy. Downing, P. B.,& Gustely, R. D. (1977).The public service of topography' costs of alternative development patterns:A review of the 3. Alaintenance costs are also excluded from this model.Ide evidcote. in P. B. Downing(Ed.),Local service pricing poll ally,these costs would be included;however,they are dif tiesand their effiictort urban spatial structure(pp.63-86).\'an- ticult to estimate and derive. Estimates of maintenance costs by spatial attributes could not be found in the exist- couver:University of British Columbia Press. Duncan,J., and Associates;\'an Horn-Gray Associates;Ivey, in literature,engineering cost studies,or local govern- l;ennect.Harris,and Walls,Inc.;&Wade-Trim,Inc.(19S9). merit budgets. Local government maintenance costs for The sea rih fi,rcrficirrttnrb:rngruu'th patterns:A study of the irrt water and sewer services are typically embedded in salaries pacts nfclerrlopnrrrtt in Florida.Tallahassee:Florida Depart- and general overhead.Also,attributing maintenance ac- memC or-Community Affair's. tiviries to specific spatial attributes would add another [:ngineering News Rrcor.{.(19991.E:\'R'siustinde.re On linr1. complication. 4. Water storage costs,however,could increase it-one spatial Available: •http://www.enr.com/cost/cost I.asp'. lacks,I. 1)..&Sullivan,K. L.(1999).Scattered development. pattern generates higher water use than another. The l , ' „rum fir Applied Research and Public i oticy, H O).24-18. tn.t liitudes of the potential increases in water use Fromm LSSikS, I. I).. Schmidt. H. E.,&Sullivan,K.(1999). Fiscal costs different patterns are fairly small relative to total water nand prrl l;is,rfih•risks urlurr density residential drvrinpnrenturt use,so we assumed that there was enough excess storage r.n•ndart,l: Findings from three diverse locations on the urban in the system to deal with these changes. ( ` \\'I'C_t\nt i u t!c Chic,r{ p n mina nlit. r area E/ 98-I) \1'ash- 5. Changing the layout of houses within a development tract h f '' will certainly affect the cost of frontage facilities such as ingcore DC::American farmland Trust,Centel for Agri- culture and the Environment. water and sewer pipe.The grid pattern described here 'en frank,J. i . (1989). The costs of alternative development patterns: crates a conservative estimate of int•- Intrasttits:cure A review of-the lkcr•:tture. Washington, DC: Urban Land requirements. Conventional cul-de-sac layouts rrtay i- Institute. crease these costs, while clustered lavou s_mav reduce Hammer.M.J.(1996).Irate,.and waste-water technology(3rd ed.). them(Zielinski,2000). New York:John Wiley&Sons. 6. A 1-acre lot,therefore,has 120 feet of frontage.Recall that Hauser. B. A. (1996). I'r.titii,tl hydraulics handbook (2nd ed.). the area ofa rectangle is equal to its length times its width: - Boca Eaton,FL:CRC Lewis. A= L x W. in the case of a 1:3 ratio,43,560 feet=3L x L. New Jersey Office ofState Planning.(1996).Projecting municipal Solving for L gives 120.5 feet. road costs under various growth scenarios:The relationship be- 7. Set 4 shows a similar pattern,except that as lot size in tu•een municipal residential density,municipal road density and creases from 0.5 to 0.75 acre,the rate of change increases density change(Document#109).Trenton:Author. slightly.This occurs because the sewer interceptor diam Ne1v Jersey Office of State Planning.(1990).Estimating co>ts for eter(and unit cost)must increase as a result of increased wastewater collection under various growth scenarios (OSP infiltration and inflow into the sewer system. Technical Report).Trenton:Author. S. The increase in costs between 4 and 5 miles in distance Peiser. scenario set 1 is attributed to a necessary increase in sewer R.B.(19S4).Does it pay to plan suburban growth?Jour- rr,tl of the American P1:ntrurigAssociation,50,419-433. interceptor pipe diameter from 21 to 24 inches. R.S. Nleans Company, Inc.(1999).Site work and Lcndscape cost 9. The local government may also pay the cost to connect existing houses to the existing water and sewer systems data 1999(1Sth annual ed.).Kingston,MA:Author. because of failing wells and septic tanks.Many recent ex Rutgers University Center for Urban Policy Research.(1992). Impact assessment of the New Jersey interim state development arnpies are available,but the cost consequences of this and redevelopment plan,Report 11:Research findings.Trenton: possibility are beyond the scope of this article. New Jersey Office of State Planning. Sanks,R.L.(Ed.).(199S).Pumping station design(2nd ed.).Bos- ton:Butterworth-Heinemann. 68 A}'A Journal •Winter 2002 •Vol.68,No. 1 DOES SPRAWL COST US ALL? Town of Blacksburg.(1996). Town of Blacksburg 1996 compre- U.S.Geological Survey. (2001). Water Qe;A:Water use at home bemire plan.Blacksburg,VA:Author. [Online]. Available: ,http://ga.water.usgs.gov/edu/ Transportation Research Board. (1998). The costs of sprawl— dahome.html=HDii3'. revisited(TCRP Report 39). Washington, DC: National \X'indsor.D.(1979).A critique of the costs olsprawl.Journalof Academy Press. the American l'Lznning;lssoci,rtion,45,279-292. U.S.Census Bureau.(2000).County population estimates forfuly 1, Zielinski,J.A.(2000).The benefits of better site design in resi- 1999andpopulation change forJuy1, 1998toJulyl,1999(CO- dential subdivisions. lf'atersbed Protection Techniques, 3, 99-1, Population Estimates Program, Population Divi- 623-632. sion).Washington,DC:Author. APPENDIX The length of pipe inside a development tract is de- pipe S inches in diameter is used for collector mains.All termined by its grid dimensions.in our simulation,a tract capital,materials,and labor costs per foot are taken from with 750 housing lots has 2S smaller-diameter lateral :\/urns'Site Work and L:rn lscaptCostData 1999(12.S.Means water mains and sewer mains on either side of the larger Company, Inc., 1999). water transmission and sewer interceptor pipes(see Fig- The length of water transmission and sewer inter- tire 1). Each pipe runs underneath a 30-foot wide road- ceptor pipes is the distance between the service center \va .The dimensions oldie pipe grid dilfer depending on and the development tract. Diameters for water trans- the size oldie lots. For this study,they are as follows: mission mains change across scenarios, as shown in Table A-2,and are approximated from a table equating Lot size Grid dimensions daily average flow to appropriate pipe size (New Jersey (acres) (L x W in feet) Office of Stare Planning. 1990). Diameters for intercep- tor pipes arc based on now,using Table 14 in Bouthillier 0.25 4860 x 1680 (19S I),which is based on the Manning formula for pipe 0.50 6SS5 x 2380 Clow. I)ianieter size may vary slightly among scenarios 0.75 8505 x 2940 because of different storm water runoffyolumes associ- 1 9720 x 3360 aced with different lot sizes. Individual communities have their own standards Table A-I shows the lengths of pipe used in the four for locating fire hydrants. This simulation places hy- lot-size scenarios. For water distribution, class 50, ce- drants as is done in Blacksburg,Virginia,according to mend-lined ductile iron pipe using push-on joints and interviews with public officials.All hydrants are assumed purchased in 1 S-foot sections was assumed.Pipe dia me- to be three-way units with a 4.5-inch valve,located 4 feet ter is always S inches for distribution mains,a generally deep.Valves are placed every S00 feet along transmission accepted standard. Note chat there is more water pipe mains,a generally accepted engineering practice accord- than sewer pipe within each development tract,because ing to manuals and interviews with engineers. Per unit only the water grid includes two pipes that run the costs for hydrants and valves are reported in R.S.iMeans length oldie tract to complete loops in the water distri- Company,Inc. (1999). bution system. Wastewater flows toward the larger in- Table A-2 shows the number of pumps and flows terceptor pipe (middle of the grid). Schedule 35 PVC used to size the pump for the four degrees of tract dis- TABLE A-1. Pipe length specifications that vary with lot size. Length (feet) per lot size(acres) Type of pipe 0.25 0.5 0.75 1 Water distribution main 134,520 189,920 234,240 267,480 Sewer collector main 94,200 133,400 164,760 188,280 Internal water transmission pipe 20,880 28,980 35,460 40,320 Internal sewer interceptor pipe 20,880 28,980 35,460 40,320 APA Journal •Winter 2002•Vol. 68,No. 1 69 CAMERON SPE[R AND KURT STEPHENSON persion. One pump is used for each development tract. by the Hazen-Williams formula(see Hauser, 1996),and To size the pumps, average flow in gallons per minute pressure added to the system to maintain adequate ser- (gpm)is multiplied by a peaking factor of 2.5 to find the vice(30 pounds per square inch or 69.3 feet of head).It is flow per minute under peak demand conditions.To en- assumed that in-line booster pumps operate 18 hours sure that the pump is able to provide adequate water to per day,and energy costs are assumed to be 50.039 per fight fire,a minimum fire flow of S00 gpm is added to kilowatt-hour, the average electricity rate paid by the the peak flow.The sum is used to determine pump ca- Roanoke County (Virginia) Utility Department (G. pacity. Capital costs (equipment and installation) for Robertson,personal communication,November 1999). water pumping stations are obtained from cost curves This model places a manhole every 400 feet along distri- by Sanks(1998) and updated to September 1999 price bution mains, a common rule for municipalities (G. levels using the Engineering News-Record Construction Robertson,personal communication,November 1999). Cost Index(Engineering News-Record, 1999). Manholes with a 4-foot inside diameter are used on sew- Pump energy requirements are calculated by deriv- ers with pipes up to 24 inches in diameter. Manholes ing power requirements using the average flow and the with a 5-foot inside diameter are used on sewers larger total discharge head(TDH)for each pump station.Total than 24 inches.Unit costs for manholes are reported in discharge head consists ofthe pressure required to over- R.S.Means Company,Inc.(1999). come friction in the distribution system,as computed TABLE A-2. System component dimensions that vary with tract dispersion. Number of tracts System component 1 2 3 4 Water transmission pipe diameter 1 S in. 14 in. 12 in. 10 in. Sewer interceptor pipe diameter 21-27 in. 18 in. 12-15 in. 12-15 in. Number of pump stations 1 2 3 4 Capacity in gallons per minute 2322 gpm 1413 gpm 1 10S gpm 955 gpm Sources:New Jersey Office of State Planning(1990),Bouthillier(19S 1),and Sanks(1998). 70 APA Journal•Winter 2002•Vol.68,No. 1