Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 1 MIXED-USE DEVELOPMENT TRIP GENERATION MODEL

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 1 MIXED-USE DEVELOPMENT TRIP GENERATION MODEL Original submission date: August 1, 2012 Revised submission date: November 13, 2012 Number of words in text: 6,661 Number of tables: 5 Number of figures: 1 Total length in words equivalent: 6,661 + 1,500 = 8,161 Corresponding Author: Jeff Gulden, PE, TE City and Metropolitan Planning University of Utah 375 S 1530 E Rm 235 AAC Salt Lake City, Utah 84112-0370 jeff.gulden@utah.edu Phone. 801.581.8255 Fax. 801.581.8217 Additional Authors: JP Goates, MCMP City and Metropolitan Planning University of Utah 375 S 1530 E Rm 235 AAC Salt Lake City, Utah 84112-0370 goatesjp@arch.utah.edu Phone. 801.581.8255 Fax. 801.581.8217 Reid Ewing, Ph.D. City and Metropolitan Planning University of Utah 375 S 1530 E Rm 235 AAC Salt Lake City, Utah 84112-0370 ewing@arch.utah.edu Phone. 801.581.8255 Fax. 801.581.8217

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 ABSTRACT The Mixed-Use Development Trip Generation model provides a mechanism for estimating vehicle, walking, and transit trips for a mixed-use development (MXD). The model applies trip modifications to standard single-use trip generation estimates developed by the Institute of Transportation Engineers (ITE). MXDs with diverse internal activities have been shown to capture internal trips at a rate higher than conventional suburban developments; therefore, the MXD Trip Generation model accounts for the internal capture of MXD sites by reducing the external trips produced and estimating the number of walking and transit trips that would typically be conducted by automobile. In addition, MXDs located in central areas have been shown to generate shorter vehicle trips, which has been accounted for in the model as internal and external vehicle miles traveled (VMT) are estimated based on published travel characteristics of MXDs. The MXD Trip Generation model provides a straightforward method of testing transportationrelated metrics of MXDs. The model uses ratios from a leading research-based MXD model to reduce ITE vehicle trip estimates and presents a summary of results that show the effects on VMT, internal capture, and mode split as a result of enhanced activity density and diversity of land uses within the MXD. The MXD trip reductions are based on the methodology described in Ewing et al. (2011), which analyzed datasets for 239 MXDs in six large and diverse metropolitan regions. Benefits of locallycalibrated vehicle ownership characteristics are also included in the mode split of the trip estimates of the MXD Trip Generation model. Number of words in Abstract text: 250

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 3 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 INTRODUCTION The Mixed-Use Development Trip Generation model estimates the daily vehicle, walking, and transit trips by purpose for a proposed development with a mix of land uses. The model, which has been developed to be incorporated as a component of the Envision Tomorrow Plus (ET+) planning tool, estimates a unique trip generation for planning scenarios and is based on land use amounts by type: residential, retail, office, and industrial. Daily trip generation is estimated based on rates and equations in Trip Generation, 8 th Edition from the Institute of Transportation Engineers (ITE), the nationally recognized source that is commonly used by traffic engineers and transportation planners across the United States (1). Trip reduction factors, based on mixed-use development (MXD) research, estimate reductions and trip distribution of a mixed-use area and are applied to the initial ITE trip estimates to determine MXD trip generation of internal and external travel. Singular land use variables, such as square feet of gross leasable area for retail, office, and industrial uses, have been used to develop trip generation estimates of stand-alone developments. Internal trips and a mode split which includes walking and transit trips can be expected when a combination of development types occur in a relatively small area. Therefore, trip estimates from ITE have been factored, based on the best research available in MXD applications, and local input variables to develop estimates of the walk, transit, and vehicle trips for travel that is both internal and external to the site. Estimates of the vehicle miles traveled (VMT) for the internal and external vehicle trips are also produced. Envision Tomorrow Plus Land use and transportation scenario planning allows users to preview impacts of development, rather than accept the trends frequently embodied in existing plans. Scenario planning is an approach that allows decision-makers, stakeholders, and the public to consider a wider range of possible futures than present in most traditional planning approaches. Scenario planning usually includes public outreach and a technical component that utilizes a sketch planning tool that enables scenarios to be drawn and quantified. Envision Tomorrow (ET), developed by Fregonese Associates, is being used for scenario planning in several locations throughout the US. ET creates building types via a "Prototype Builder" that tests the physical and financial feasibility of development. The tool allows users to examine land use regulations in relation to the current development market and to consider the impact of parking, height requirements, construction costs, rents, and subsidies. The tool can be used to examine if a development pencils. For example, it can assess how preferred forms of development, such as mixed-use retail with housing above, might become more financially feasible within an existing code. The subsequent part of ET is "Scenario Builder", which adds scenario-building functionality to ArcGIS. From prototypes generated in Prototype Builder, development type combinations are fed into Scenario Builder to paint a landscape by allocating different development types across a study area to create a land use scenario. Unlimited scenarios can be built to test against each other. The tool allows real-time evaluation of the impact of each scenario on land use, housing, sustainability, and economic conditions. In 2011, the University of Utah Metropolitan Research Center received funding as part of the Housing and Urban Development Sustainable Communities Program grant. In part, the grant included a joint effort with the Metropolitan Research Center and Fregonese Associates to expand ET tools using new research-based methods, resulting in Envision Tomorrow Plus (ET+). The expansion includes approximately 20 new models that range from energy and water consumption to transportation. The MXD Trip Generation model is one of the components of ET+ and outputs of the model are essential to the overall tool as many indicators for other models are dependent on its outputs. The authors have established procedures to generate variables to model trip generation, internal capture, VMT, and mode share of internal and external trips based on existing and proposed conditions.

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 4 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 This report provides an overview of the research guidance and operations contained in the MXD Trip Generation model as well as a guide to the operation of the component within ET+. The MXD Trip Generation model is part of ET+ and will be available as an open-source software package. At time of publication of this paper, the complete software package was being developed and tested with real-world scenarios. It is anticipated that the software will be released soon. For additional information on the software status, please contact the authors. LITERATURE REVIEW Vehicle-based trip generation is a traffic engineering method that is used extensively across the US to analyze impacts of proposed developments on the surrounding transportation network. While no standard of trip estimation is universally agreed upon, the most common practice for traffic impact studies in the US is to rely upon the procedures from trip generation rates and adjustments published by ITE in its two reports: Trip Generation, 8 th Edition (ITE, 2008) and the Trip Generation Handbook, 2 nd Edition (ITE, 2004) (2). Limitations have been raised regarding the ITE MXD lookup tables in the Trip Generation Handbook. The ITE MXD analysis is based on data from a limited number of surveys; three multi-use sites in Florida (2). Accuracy of future forecasts is thus dependent on how closely the site and context being analyzed matches the three suburban Florida mixed-use sites. Therefore, in addition to ITE standard trip generation rates and equations, the MXD Trip Generation model of ET+ references studies that address the trip estimation characteristics of MXDs with respect to vehicular, transit, and pedestrian travel. To determine the most appropriate MXD trip generating method for the model and ET+, a literature review of MXD trip generation methods was conducted. After completion of the literature review, it was apparent that a combination of the two most applicable methodologies would be used for the MXD Trip Generation model: Trip Generation, 8 th Edition (ITE, 2008) and Traffic Generated by Mixed-Use Developments Six-Region Study Using Consistent Built Environmental Measures (Ewing et al., 2011) (3). The findings of the literature review are presented in this section, along with a detailed description of the methodologies selected for use. Evaluation of the Operation and Accuracy of Available Smart Growth Trip Generation Methodologies for Use in California A review of eight standard methods of MXD trip generation by Shafizadeh et al. (2011) found all to be more accurate than ITE procedures (4). The review found that, in estimating trip generation for multi-use sites, all of which are large-scale projects located outside the central business district, the most accurate estimate for the greatest number of sites was produced by Ewing et al. (2011) (4). For daily counts, the Ewing et al. (2011) study methodology produces the best estimate for seven of the 22 sites used for validation (4). Therefore, the Ewing et al. (2011) study methodology is the basis for MXD Trip Generation estimation in ET+. The study compared five of the trip generation methods by modeling a variety of hypothetical characteristics as well as a real-world assessment at 22 sites in California. Three additional methods were reviewed in the study but were not tested due to their methodology having been developed for sitespecific locations. The five methodologies selected for comparison by Shafizadeh et al. (2011) are included in the literature review in this section. National Cooperative Highway Research Program Report 684 One of the latest MXD trip generation studies is the National Cooperative Highway Research Program Report (NHCRP) 684, which summarizes the multi-year research effort of the NCHRP 8-51 project (5).

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 5 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 The report studied the trip making characteristics of six MXD sites. The result was the determination of an amount of expected level of interactivity between land uses within a site, during both the morning and evening peak periods (5). The sites in the study contained the following land uses: office, retail, restaurant, residential, movie theater, and hotel. The report is intended to improve the MXD methodology presented in the ITE Trip Generation Handbook (at the time of publication of this paper, ITE has announced the release of Trip Generation, 9 th Edition and an updated Trip Generation Handbook; however, information in this paper is based on the Trip Generation, 8 th Edition.) Station Area Residents Survey The Metropolitan Transportation Commission, in the San Francisco Bay Area, conducted a demographic and travel characteristics survey of residents living within close proximity to rail/ferry terminals in the area. The study produced many travel-related findings, including those living (and working) close to rail/ferry transit use transit, walk and bike much more than people living farther from a rail/ferry stop (6). The finding indicates that the presence of transit is related to the walking and bicycling mode choice. The extensive data collection effort led to an MXD trip generating method that factors ITE Trip Generation rates based on location to rail/ferry terminals in urban areas (4). The method appears to only be applicable in urban areas with rail/ferry connection to the development site and its limited database (the San Francisco Bay Area) may be prohibitive for national use. Emissions Estimation for Land Use Development Projects (URBEMIS2007) The Emissions Estimation for Land Use Development Projects is a software package that estimates air emissions from land use sources and includes MXD trip generation estimations (7). The URBEMIS2007 is a stand-alone software package and was deemed incompatible with the ET+ tool being developed by Fregonese Associates. Emission estimates however, will be included in ET+, using outputs relating to the traffic estimated by the MXD Trip Generation model. The findings by Shafizadeh et al. (2011) related to the accuracy of the URBEMIS2007 software with respect to trip generation did not indicate that the product was superior to the other, more ET+ toolfriendly, methodologies (4). Traffic Generated by Mixed-Use Developments Six-Region Study Using Consistent Built Environmental Measures The analysis in Ewing et al. (2011) was based on travel data from 239 MXD sites in six diverse regions. In each region, consultations with local planners and traffic engineers were conducted to identify MXDs that met the ITE definition of MXD: A mixed-use development or district consists of two or more land uses between which trips can be made using local streets, without having to use major streets. The uses may include residential, retail, office, and/or entertainment. There may be walk trips between the uses (3). The 239 identified MXDs ranged from compact infill sites near the region s core to low-rise freewayoriented developments. The sites varied in population and employment densities, mix of jobs and housing, presence or absence of transit, and location within the region. The sites ranged in size from less than five acres to over 2,000 acres, with some sites containing over 15,000 residents and employees. Quantifiable relationships between a development site s external traffic generation and the following seven D characteristics; density, diversity, design, destination accessibility, distance from transit, development scale, and demographics were established. The statistical analysis found that several D factors influence internal trip capture, and that the effects vary by trip purpose. For home-based work trips, internal capture is related to two D variables, diversity and demographics. The odds of an internal

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 6 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 trip occurring decline with household size and vehicle ownership per capita, and increase with a jobpopulation balance in the MXD. Larger households have more complex activity patterns, which are more likely to take them beyond the bounds of an MXD. Households with higher vehicle ownership have fewer constraints on the use of household vehicles for long trips. A high job-population balance value translates into more opportunities to live and work on site. For home-based trips conducted for purposes other than commuting, the odds of internal capture decline with household size and vehicle ownership per capita and increase with the land area, job-population balance, and intersection density of an MXD. The model produced in this study enables a traffic analysis to account for the differences in trip generation that arise from infill developments within or adjacent to traditional downtowns or central business districts, implicitly accounting for the factors that differentiate downtown areas from conventional suburbs, including the presence or absence of rail and bus service, and the degree to which external trip generation is affected by the number of jobs in the area surrounding the development site. The model also reflects the extent to which the internal, walking, and transit trip generation of the site is affected by: the mixture of uses, the pedestrian interaction related to the scale of the area, the ease of access and proximity to a variety of activities, parking constraints, as well as higher automobile occupancy and more convenient transit service. Trip Generation, 8 th Edition Trip Generation, 8 th Edition (ITE, 2008 and referred to as Trip Generation in this document) is the standard for vehicle-based single-use trip generation estimates and is used by engineers and planners across the US (1). Trip Generation is developed by ITE and, at the time of development of ET+, is in its eighth edition. ITE compiles trip generation studies for various sites and isolated land uses, groups these studies into categories, and then develops rates and equations which can be applied to similar projects. The trip generation studies are summarized in ITE s Trip Generation with additional information provided in the Trip Generation Handbook, 2 nd Edition (ITE, 2004). Trip Generation Handbook, 2 nd Edition ITE produces a trip generation handbook to accompany the Trip Generation manual. The Trip Generation Handbook, 2 nd Edition (ITE, 2004 and referred to as Trip Generation Handbook in this document) provides guidance on proper application of the data summarized in the Trip Generation manual. Engineering and planning topics presented in the Trip Generation Handbook include guidelines for estimating project site trip generation, collection of local trip generation data, and development of local trip generation rates, among other traffic engineering-specific applications. The Trip Generation Handbook details a method for developing vehicle trip generation estimates for mixed-use developments. However, the methodology has several shortcomings, including: being based on a limited number of sites, providing only three land use types, and disregarding the scale of the development. Therefore, the methodology used in the ET+ model follows the trip estimation procedures described in the Ewing et al. (2011) study, primarily due to its ability to estimate walk trips and transit trips estimated at an MXD, thus reducing VMT. DEFINITION OF MIXED-USE DEVELOPMENT The study area boundary created for ET+ project sites should include only mixed-use type development. The boundary should be established as close to the proposed scenario as possible; however, it should also include any immediately adjacent existing land use that would be considered part of the MXD. Mixed-use is defined as development consisting of two or more land uses between which

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 7 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 trips can be made using local streets (including walk trips), without having to use major streets. The uses may include residential, retail, office, and/or entertainment. Identification of local streets may use Census Feature Class Code (CFCC) local street criteria. However, streets that may not be classified as local by the CFCC may be included if exchanging trips as part of a contiguous development. For example, a main street in a smaller town may legally be classified as a US Highway but may function as a local street. Figure 1 provides aerial illustrations of several types of development areas and their applicability to the MXD definition: a) the type and scale of a typical MXD tested in ET+, b) an area that does not follow the MXD definition as it is divided by an arterial and has a distinct separation of land use, and c) a mixed-use district that meets the MXD definition but contains a larger traffic volume collector roadway which continues to function as a local street. A planning professional should be involved in the development of the study area boundary, and understand the research, to ensure that only MXDs are included and that land uses are exchanging trips on local streets. A study area boundary and scenario planning effort may include existing uses; however, the amount of existing uses should be carefully considered. Every effort should be made by the planning professional to develop the MXD study area boundary so that it encompasses only the mixed-use type development that follows the MXD definition. In addition, constraints with respect to scale have been established due to bounds of study locations and destination access of walk trips. An ET+ MXD study area boundary should be no larger than 960 acres (approximately 1.5 square miles). The area limit for the MXD was determined based on the areas used in the Ewing et al. (2011) study and by considering the upper bounds of a typical walking distance for pedestrians (0.5 mile) between land uses that may exchange trips. (The Federal Transit Administration has officially established the walkable catchment area around transit stops as 0.5 mile. In addition, a 2009 study commissioned by the Victoria Department of Transport found that the average walk trip in the US was 1.2 miles but that 50 percent of all walk trips were less than 0.5 mile.) Important Considerations All applications of the MXD Trip Generation model should meet the definition of an MXD. An MXD should contain local streets and not major streets (in most cases). The MXD should be 960 acres or less; results computed for sites larger are out of the bounds of the MXD Trip Generation model.

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Example of an area that follows the MXD definition. Example of an area that does not follow the MXD definition due to its large internal roadways. Page 8 Example of an area that follows the MXD definition despite the presence of larger roadways within the site. FIGURE 1 Examples of development areas and their applicability to the MXD definition.

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 9 251 252 253 254 255 256 257 258 259 INPUTS Inputs for MXD scenarios in ET+ are comprised of both manually collected inputs and GIS tool based auto-populated fields. Summary square footage by land use from scenarios and existing conditions are pushed to the model and generate variables used in model calculations. A general overview of the development of variables is provided in Table 1. Certain variables are specific to MXD-related transportation benefits and travel mode types, with different variables being used in different equations. TABLE 1 MXD Model Inputs Variable Inputs for MXD Trip Generation Estimates Description MXD Area Non-Residential Land Use Residential Units Intersections Employment in Region Employment in One Mile Median Household Income Employment in 20 Minute Automobile Commute Employment in 30 Minute Automobile Commute Employment in 30 Minute Transit Commute Household Income Transit Stop Proximity Gross land area in square miles and acres within the user-defined MXD boundary Existing and scenario lot area of retail, office, and industrial uses in MXD boundary plus one mile buffer (in square feet) Sum of single family dwelling units and other (multi-family) dwelling units from land use units in MXD (existing plus scenario) Number of intersections within the MXD boundary and within a one mile buffer Number of jobs in region from Metropolitan Planning Organization model Employment within a one mile buffer of the MXD boundary derived from non-residential land use (excluding the MXD study area) Median household income from the US Census American Community Survey five year estimates by MXD centroid tract converted to the Proportion of total employment in the region that is within a 20 minute automobile commute from the MXD Proportion of total employment in the region that is within a 30 minute automobile commute from the MXD Proportion of total employment in the region that is within a 30 minute transit commute from the MXD Derived from separate models contained in ET+ Stops within MXD, stops within 0.25 mile buffer of MXD and proportion of area within 0.25 mile buffer that has stop coverage in 260 261 262 Several of the input variables may require data from a variety of outside sources, such as the census, metropolitan planning organization, and transportation agencies, among others. The data should

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 10 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 always be collected from the best possible source at the finest grain, which could range from census tract to parcel level. EQUATIONS The MXD Trip Generation model requires a number of equations and inputs to produce a mixed-use trip generation estimate. Equations are based on research and are summarized in the literature review, with required inputs identified. Many of the inputs to the MXD Trip Generation model are outputs from ET and selected ET+ models, while others are regional outputs from large-scale travel estimates, such as a travel demand forecasting model. The equations are presented in four sections: 1) those required to determine an ITE Trip Generation calculation, 2) those inputs required for the 7D Household Vehicle Ownership model, 3) the separation of trips by purpose, and 4) those inputs required to apply the MXD trip reduction procedures to the initial trip estimates. 1. ITE Trip Generation The land use inputs required for the ITE Trip Generation estimates are provided by the scenario planning component of ET+ and existing geographic data. An estimate of the number of vehicle trips based on ITE Trip Generation rates and equations requires the following land use information: Existing and scenario square feet of retail, office, and industrial (if present) Existing and scenario dwelling units of single family residential, townhomes, multi-family residential, and mobile home park (if present) Equations and rates utilized in the MXD Trip Generation model are presented in Table 2 for single family residential, townhomes, multi-family residential, mobile home park, retail, office, and industrial.

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 11 290 TABLE 2 ITE Equations Utilized in the MXD Trip Generation Model Land Use Type Equation Variables Single Family Residential (ITE Code 210) Townhomes (ITE Code 230) Multi-Family Residential (ITE Code 220) Mobile Home (ITE Code 240) Retail (ITE Code 820) Office (ITE Code 710) Industrial (ITE Code 130) Ln (T) = 0.92 * Ln (X) + 2.71 Ln (T) = 0.87 * Ln (X) + 2.46 T = 6.06 * X + 123.56 T = 3.52 * X + 277.51 Ln (T) = 0.65 * Ln (X) + 5.83 Ln (T) = 0.77 * Ln (X) + 3.65 T = 6.96 * X T = number of daily vehicle trips X = number of dwelling units in MXD T = number of daily vehicle trips X = number of dwelling units in MXD T = number of daily vehicle trips X = number of dwelling units in MXD T = number of daily vehicle trips X = number of dwelling units in MXD T = number of daily vehicle trips X = 1,000 square feet of gross leasable area in MXD T = number of daily vehicle trips X = 1,000 square feet of gross leasable area in MXD T = number of daily vehicle trips X = 1,000 square feet of gross leasable area in MXD 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 For additional information on ITE trip generation estimating equations, rates, and procedures, refer to Trip Generation Handbook and Trip Generation. Commercial vehicle trips, such as truck and delivery trips, were not isolated in the MXD Trip Generation model as they represent a relatively minor portion of trips associated with retail land uses. In addition, commercial trips would not experience many of the transportation-related benefits of an MXD. Important Considerations Development scenarios that include land use amounts greater than the following upper limits are not applicable to these trip generation estimates: Retail greater than 1,500,000 square feet Office greater than 1,300,000 square feet Industrial greater than 2,300,000 square feet Single family residential greater than 3,000 dwelling units Townhomes greater than 1,250 dwelling units Multi-family residential greater than 1,000 dwelling units Mobile home park greater than 810 dwelling units

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 12 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 2. 7D Household Vehicle Ownership Model The 7D Household Vehicle Ownership model was developed to estimate the number of vehicles owned by households and is based on local land use and geographic characteristics. The model has been recently developed by researchers at the University of Utah and publication of the model is pending. The MXD Trip Generation model utilizes inputs which are generated by the 7D Household Vehicle Ownership model. Single family households and multi-family households (referred to as Other Households in the model) have distinct vehicle ownership characteristics and are represented by separate models. Each vehicle ownership model is presented in this section with inputs identified. Several of the vehicle ownership model inputs require sources outside of the ET+ package. Single Family Households Ln (V) = -1.42478 + [0.33331 * Ln (HHSIZE)] + [0.220001 * Ln (SFINCOME)] + [- 0.09165 * Ln (INTDEN)] + [-0.01382 * Ln (EMP30T)] V = vehicles HHSIZE = single family household size SFINCOME = single family median household income (modified to 1982 consumer price index) INTDEN = intersection density within the MXD EMP30T = employment within a 30 minute transit commute Other Households Ln (V) = -2.27444 + [0.390903 * Ln (OHHSIZE)] + [0.315116 * Ln (OINCOME)] + [- 0.1513 * Ln (INTDEN1)] + [-0.02201 * Ln (TMILE)] V = vehicles OHHSIZE = other household size OINCOME = other median household income (modified to 1982 consumer price index) INTDEN1 = intersection density within a one mile buffer from every separate other household TMILE = transit stops within a quarter mile buffer from every separate other household Important Considerations All income is converted to the 1982 Consumer Price Index. ET+ produces income in current dollars. The variable TMILE represents an aggregate of the number of transit stops within a one mile buffer of every household. Although this value is often impractical to obtain in practice, the model requires this input for proper functionality. As a substitute, the number of transit stops within a one mile buffer from the edge of the MXD study area can be used to provide an approximate result.

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 13 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 Household vehicle ownership estimates for MXDs may be higher than generally accepted and national household vehicle ownership rates due to local land use and income distributions which may be higher in the MXD. 3. Trips by Purpose Information from the Nationwide Personal Transportation Survey and the National Highway Cooperative Research Program 365 report was used to determine the percentage of trips based on purpose for each land use type used in the MXD Trip Generation model of ET+ (8, 9). Vehicle trips estimated by the ITE Trip Generation rates and equations are separated by purpose prior to the application of MXD reduction factors. To separate the vehicle trip estimates into home-based work (HBW), home-based other (HBO), and non-home-based (NHB) trips, a unique set of trips by purpose percentages were used. A distinct trip percentage is provided for each land use type as the trip making characteristics differ. The following unique percentages were used in the MXD Trip Generation model of ET+: TABLE 3 Trips by Purpose (8, 9) Land Use HBW HBO NHB Residential 25% 75% 0% Retail 5% 45% 50% Office / Industrial 65% 5% 30% 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 Data in Table 3 is used to assign trips by purpose to land use types. As an example, residential trips generated by a development are assumed to the following purposes: 25 percent will travel between the home and work, 75 percent will travel between the home and an other destination, and zero percent are non-home-based. For additional information on MXD trip generation estimating equations, rates, and procedures, refer to Ewing et al. (2011). 4. Mixed-Use Development Model An estimation of the trips associated with MXDs is based on Ewing et al. (2011), which uses variables from national data which were evaluated with a best fit regression model that generated the coefficients for the equations. The MXD equations predict the internal and external trips generated by a mix of land use based on the vehicle trip estimation. Internal and external trips are further divided into walk trips, transit trips, and MXD-related vehicle trips. The specific indicators calculated by the model are: internal share of total trips, internal automobile trip length, external automobile trip length, internal walk trips, external walk trips, and external transit trips. All trip lengths are in miles. For each trip type, there is a separate equation for HBW, HBO, and NHB trips. An example of the equations and rates utilized in the MXD model portion of the MXD Trip Generation model are presented in this section. The following set of equations detail the probability that a trip will be an internal trip.

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 14 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 Internal Share of Total Trips - HBW Ln (P) / [1 + Ln (P)] = -1.75 + [0.389 * Ln (JOBPOP)] [1.33 * Ln (HHSIZE)] [0.99 * Ln (VEHCAP)] P = percentage of HBW internal share of total trips JOBPOP = proportion of employees in study are to population of study area HHSIZE = single family household size in the MXD VEHCAP = vehicles per capita in the MXD Internal Share of Total Trips - HBO Ln (P) / [1 + Ln (P)] = -2.43 + [0.486 * Ln (AREA)] + [0.399 * Ln (JOBPOP)] + [0.385 * Ln (INTDEN)] [0.867 * Ln (HHSIZE)] [0.59 * Ln (VEHCAP)] P = percentage of HBO internal share of total trips AREA = area in square miles of the MXD JOBPOP = proportion of employees in study area to population of study area INTDEN = intersection density in the MXD HHSIZE = single family household size in the MXD VEHCAP = vehicles per capita in the MXD Internal Share of Total Trips - NHB Ln (P) / [1 + Ln (P)] = -5.32 + [0.208 * Ln (EMP)] + [0.468 * Ln (AREA)] + [0.638 * Ln (INTDEN)] [0.237 * Ln (HHSIZE)] [0.163 * Ln (VEHCAP)] P = percentage of NHB internal share of total trips EMP = employees in study area AREA = area in square miles of the MXD INTDEN = intersection density in the MXD HHSIZE = single family household size in the MXD VEHCAP = vehicles per capita in the MXD The following set of equations is an example of trip length calculations for external vehicle trips. External Automobile Trip Length - HBW L = 6.54 + [1.07 * Ln (AREA)] [0.298 * Ln (JOBPOP)] [1.19 * Ln (EMP30A)] + [2.76 * Ln (HHSIZE)] + [2.76 * Ln (VPC)] L = external automobile trip length for HBW trips of the MXD AREA = area in square miles of the MXD JOBPOP = proportion of employees in study area to population of study area EMP30A = employees within a 30 minute automobile commute as a portion of the employees in the region

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 15 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 HHSIZE = single family household size in the MXD VEHCAP = vehicles per capita in the MXD External Automobile Trip Length - HBO L = 4.33 [0.356 * Ln (JOBPOP)] [0.697 * Ln (EMP20A)] + [0.772 * Ln (HHSIZE)] + [1.48 * Ln (VEHCAP)] L = external automobile trip length for HBO trips of the MXD JOBPOP = proportion of employees in study area to population of study area EMP20A = employees within a 20 minute automobile commute as a portion of the employees in the region HHSIZE = single family household size in the MXD VEHCAP = vehicles per capita in the MXD External Automobile Trip Length - NHB L = 8.99 [0.282 * Ln (JOBPOP)] [0.832 * Ln (INTDEN)] [0.823 * Ln (EMP20A)] + [0.52 * Ln (HHSIZE)] + [1.06 * Ln (VEHCAP)] L = external automobile trip length for NHB trips of the MXD JOBPOP = proportion of employees in study area to population of study area INTDEN = intersection density in the MXD EMP20A = employees within a 20 minute automobile commute as a portion of the employees in the region HHSIZE = single family household size in the MXD VEHCAP = vehicles per capita in the MXD Additional equations are used in the MXD Trip Generation model to determine the trip characteristics for the following types of trips: internal walk trips, external walk trips, and external transit trips. Internal automobile trip length is also determined by a separate equation. For additional information on MXD trip generation estimating equations, rates, and procedures, refer to the Ewing et al. (2011) study. Important Considerations Acquiring variables for modeling a complete MXD model analysis requires several data collection and processing techniques. Accuracy of the model and ITE trip modification may not be considered valid without proper inputs. MXD boundaries must be defined according to the definition of the MXD section and the following variable details must be considered: Developed land ratio considers scenario totals as built, and is combined with existing land use totals. Intersections must include existing and any proposed intersections. Employment information is estimated by square feet of non-residential use per employee according to ET estimates.

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 16 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 PROCESS The trip generation estimates are developed through a combination of ET+ scenario and user inputs, empirically based coefficients, and standard transportation rates and equations. An overview of the process for developing MXD trip generation estimates is presented in the following, in which steps are summarized in the order they are completed in the ET+ MXD Trip Generation model: 1. Calculate an ITE Trip Generation estimate of total vehicle trips to be generated by the development types. 2. Separate the vehicle trips by purpose (HBW, HBO, and NHB) based on land use type. 3. Employ the 7D Household Vehicle Ownership model to predict the number of vehicles per household, to be used in the MXD equations. 4. Apply the MXD equations to factor the estimated vehicle trips into internal and external vehicle trips. 5. Apply the MXD equations to factor the estimated internal and external vehicle trips into walking trips, transit trips, and vehicle trips (for both internal and external situations). 6. Apply the MXD equations to factor the estimated internal and external vehicle trips into VMT (for both internal and external situations). Each step of the process is described in greater detail in this section. The intent of describing each individual process is to provide the user insight into the general operations of the MXD Trip Generation model. The model has been developed to operate transparently, allowing the user to examine and follow each step, if desired. Land use amounts from ET are input into the MXD Trip Generation model and ITE Trip Generation rates and equations are used to estimate the number of vehicle trips expected to be generated if the land uses were stand alone developments. The trip generation estimates are produced according to standard traffic engineering protocol. A significant portion of the process involves separating trips into trip types: HBW, HBO, and NHB. The division of vehicle trips by purpose has significant effects on the ultimate MXD trip generation of the development; each of the vehicle trip generation estimates from ITE are factored by the nationwide HBW, HBO, and NHB percentages before being factored by the site-specific MXD ratios. One of the inputs for the MXD equations, vehicle ownership, requires a separate set of models with unique inputs. The 7D Household Vehicle Ownership model predicts the number of vehicles per household and is embedded within the MXD Trip Generation model. The model operates as part of the module and predicts the number of vehicles owned by single family and other households (multi-family households) based on local land use and economic conditions. The next step of the MXD Trip Generation model uses site-specific ratios from the MXD equations to transform the vehicle trips estimated by ITE into internal and external trips. The site-specific ratios factor the total vehicle trips into internal trips and external trips. The internal vehicle trip estimates are further separated into the HBW, HBO, and NHB purpose categories by land use type to determine the amount of internalization of vehicle trips. For example, all office vehicle trips are separated into HBW, HBO, and NHB vehicle trips. The vehicle trips are then grouped by purpose and the land use productions and attractions are compared. The smaller of the two, production or attraction, is used as the governing internalization maximum amount and the internal trip generation of the associated land uses are set at that amount. The next step of the model applies the MXD equations to factor the estimated internal and external vehicle trips into walking trips, transit trips, and vehicle trips. Once the internal trips are set for each land use type, the site-specific MXD ratios are applied to factor the trips into internal walking and vehicle trips. The internal trips are subtracted from the total vehicle trips to arrive at the estimated number of total external trips. Separate equations from the MXD model are used, for both the internal and external trips, to factor the trips into walking trips, transit trips, and vehicle trips. The resulting estimates

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 17 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 are walk trips and vehicles trips that are internal and external to the study area. In addition, transit trips are estimated and are entirely external and are represented in person-trips. The final step summarizes trip type by mode to produce the total trips estimated to be generated by the MXD. The MXD Trip Generation model uses ratios from the MXD models to reduce the ITE vehicle trip estimates. Therefore, the resulting vehicle trips estimated by the described process will always produce a trip generation estimate less than ITE. An MXD ratio is applied to the vehicle trips to estimate the VMT for those internal and external vehicle trips. The site-specific VMT ratio is developed from the MXD models and is dependent on external data, as previously described. Total internal and external VMT is summarized for the study area. CASE STUDY In 2012, the MXD Trip Generation model was used in community-based scenario planning efforts in Utah. Envision Utah, a community visioning organization in Salt Lake City, provided inputs and scenarios for the MXD Trip Generation model to estimate the transportation benefits of MXDs in proposed redevelopment areas in Salt Lake County and Utah County. One development site in Salt Lake County has been described in this section as a case study for the MXD Trip Generation model. Case Study Background The township of Magna is located in western Salt Lake County, Utah (the left image in Figure 1 of this document is the study area for the Magna, Utah case study). Envision Utah worked with the community to undertake a scenario planning process for redevelopment of part of downtown Magna. As part of the planning process, the MXD Trip Generation model was used to provide the community and decision makers with the transportation benefits related to the proposed MXD scenarios. The scenario planning process initially evaluated three distinct development types. The developments varied in their mix of land uses and overall size. In all scenarios, the location was the same, as it was centered on downtown Magna. Case Study Inputs Table 4 summarizes the inputs used in the MXD Trip Generation model for the Magna, Utah case study. For all variables, the best available data was selected for use in the model. The inputs were obtained from a variety of local, regional, and national sources, including: Wasatch Front Regional Council (the local metropolitan planning organization) US Census Salt Lake County Assessor data Utah Transit Authority ESRI 2007 roadway data Envision Utah These data were processed using ESRI ArcMap and ET+ tool extensions and input into the ET+ Scenario Builder tools.

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 18 582 TABLE 4 Inputs for MXD Trip Generation Model for Magna, Utah Case Study Inputs Scenario 1 Scenario 2 Scenario 3 Employees in Study Area 926 1,204 1,999 Population in Study Area 4,724 5,669 6,005 Employees in Region 1,177,000 1,177,000 1,177,000 Employment within 1 Mile 1,240 1,240 1,240 Area (square miles) 0.66 0.66 0.66 Intersections 99 99 99 Transit Stop Proportion 1.00 1.00 1.00 Vehicles Per Capita 0.62 0.62 0.62 Total Residential (du) 1,583 2,152 2,521 Residential Land Use (du) 2,092,763 2,638,954 2,866,216 Retail Land Use (ksf) 171,484 136,639 192,736 Office Land Use (ksf) 147,515 263,546 485,199 Industrial Land Use (ksf) 47,170 31,336 7,160 Total Land Use (ksf) 2,458,932 3,070,475 3,551,311 Median Household Income 20,017 21,670 22,719 Employment within 20 Minute Commute 25,3580 25,3580 25,3580 Employment within 30 Minute Commute 68,7958 68,7958 68,7958 Employment within 30 Minute Transit Ride 3,588 3,588 3,588 Notes: du dwelling units ksf thousand square feet 583 584 585 586 587 588 589 Case Study Process The inputs listed in Table 4 were processed through the six steps of the MXD Trip Generation model listed on page 16. Each step of the process is summarized below. Trip generation estimates were calculated based on the land use amounts of the proposed MXD and the rates and equations from ITE Trip Generation.

Jeff Gulden, PE, TE, JP Goates, MCMP, and Reid Ewing, Ph.D. Page 19 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 The estimated vehicle trips were separated into trip purpose, either HBW, HBO, or NHB, based on the land use type and the percentages from Table 3. The number of vehicles per household was determined by using the 7D Household Vehicle Ownership model. The number of vehicles per household is an input into the equations that determine the transportation benefits of the MXD. The MXD equations were used to factor the ITE Trip Generation-estimated vehicle trips into internal and external vehicle trips. The next step of the process used MXD equations to factor the internal vehicle trips into walking and vehicle trips (internal) and to factor external vehicle trips into walking trips, transit trips, and vehicle trips (external). The final step of the model factored the internal and external vehicles trips into VMT. Case Study Results The results of the MXD Trip Generation model for three planning scenarios from Magna, Utah are summarized in Table 5. The MXD Trip Generation model provided results that allowed the community and decision makers to better understand the transportation-related benefits of the proposed MXD scenarios. The results of MXD Trip Generation model summarize the internal and external trips by mode (walking, vehicle, and transit) as well as the total number of trips by mode. To provide a measure of comparison for situations where scenarios of significantly different land use amounts are proposed, the number of trips per dwelling unit has been calculated. The mode split for internal trips, between walking and vehicle use, and external trips, between walking, vehicle, and transit use, is shown to better understand the character of transportation in the proposed MXD. The results show between a 6.3 percent and 7.9 percent reduction in external vehicle trips when compared to a typical traffic study relying on summarizing the vehicle trips generated by the individual land use amounts contained in the proposed development. It is reasonable that the MXD reductions aren t larger than 7.9 percent for this case study as patrons and residents would likely drive to destinations since the location is in suburban Salt Lake County. Higher percentages of MXD trip reduction benefits would be expected in other areas in the US.