National Academies Press: OpenBook
« Previous: Summary
Page 7
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 7
Page 8
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 8
Page 9
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 9
Page 10
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 10
Page 11
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 11
Page 12
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 12
Page 13
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 13
Page 14
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 14
Page 15
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 15
Page 16
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 16
Page 17
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 17
Page 18
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 18
Page 19
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 19
Page 20
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 20
Page 21
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 21
Page 22
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 22
Page 23
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 23
Page 24
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 24
Page 25
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 25
Page 26
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 26
Page 27
Suggested Citation:"Chapter 1 - Introduction." National Academies of Sciences, Engineering, and Medicine. 2011. Enhancing Internal Trip Capture Estimation for Mixed-Use Developments. Washington, DC: The National Academies Press. doi: 10.17226/14489.
×
Page 27

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

7Background Problem Statement NCHRP Project 8-51, “Enhancing Internal Trip Capture Estimation for Mixed-Use Developments,” was undertaken to improve the methodology(s) used to estimate the extent to which trips made within mixed-use developments are inter- nalized or satisfied with both origin and destination within the development. Such estimates are important in determin- ing the quantities of external trips generated by mixed-use developments. To fully understand the project, it is first necessary to under- stand some of the terms used in describing the project. Terms are defined as follows: • Mixed-Use Development: A mixed-use development, ac- cording to the Urban Land Institute (ULI), is a single phys- ically and functionally integrated development of three or more revenue-producing uses developed in conformance with a coherent plan (3, pp. 4–5). The Institute of Trans- portation Engineers (ITE) suggests two interacting land uses compose a mixed-use development (MXD) (2). MXDs have internal pedestrian connectivity and share parking among some or most uses. An example of a true MXD would be a galleria consisting of retail, hotel, office, restau- rant, and entertainment uses, possibly in separate build- ings, but interconnected and sharing parking facilities. For the purposes of this project, it has been deemed appropri- ate and necessary to expand this definition to include multi-use developments. A multi-use development is a real estate project of separate uses of differing and complemen- tary, interacting land uses that do not necessarily share parking and may not be internally interconnected except by public street and/or other public transportation facili- ties. A multi-use development example would be an activ- ity center such as Tysons Corner in northern Virginia, also with a variety of interactive land uses, but relying on the public road system and separate parking facilities for most of the interaction. • Activity Centers: An activity center is a well-defined, fo- cused concentration of development with high density and a high mix of land uses. An activity center usually meets the above expanded definition of an MXD. An activity center is generally very large compared with other MXDs in its urban area and usually occupies at least several blocks. Perimeter Center in Atlanta is a good example of an activity center. This is not to be confused with shopping centers (for which ITE has specific trip generation rates) (4, pp. 561–562); how- ever, for the purposes of this project, activity centers are not a focus of this research, but the estimation methodology may be adaptable for use in activity centers. • Neighborhoods and Subareas: ITE notes that any area that has a specific identity and generates large amounts of traffic could be considered an area or subarea with unique trans- portation issues (4, p. 561). For the purposes of this project, neighborhoods can be classified within this concept when they exhibit a mix of interactive uses. Neighborhoods and subareas are not specifically within the focus of this re- search; however, as with activity centers, the methodology developed by this research may be adaptable for use in neighborhoods and subareas. • Transit-Oriented Development: According to the Ameri- can Public Transportation Association (APTA), a transit- oriented development (TOD) is a compact, MXD near new or existing public transportation infrastructure that serves housing, transportation, and neighborhood goals. Its prox- imity to transit services and pedestrian-oriented design en- courages residents and workers to drive their cars less and ride mass transit more (5). For the purposes of this project, the research team stipulates that the development must be not only near transit, but the transit service must also be convenient to reach, the service must link the development C H A P T E R 1 Introduction

with other complementary locations, and the development must include land uses that generate activity that can be readily used by transit patrons. • Internal Trip: An internal trip, as defined by ITE, is one that is made without utilizing the major road system (2, p. 85). For the purposes of this project, the definition is expanded to include travel within a highly interactive area containing complementary land uses and convenient in- ternal on- or off-street connections that may use short segments of major streets. An example might be a one- block development consisting of residential, office, and retail buildings with convenient sidewalk connections be- tween them and a single parking facility serving all three land uses. • External Trip: An external trip is a trip made between land uses within the MXD and locations outside the boundaries of the development. This excludes internal trips. • Internal Trip Capture (Site) Rate: Internal trip capture for a development site is the percentage of total trips (nor- mally, but not always, vehicle trips when used for typical traffic impact studies) that are made internally to the de- velopment without using roads that are external to the site being analyzed. The internal trip capture is most frequently expressed in terms of a percentage or rate, but can be de- scribed in other forms such as equations. For example, if retail uses within an MXD generate 10 trips, 3 of which go to other land uses within the development and 7 of which go to external locations, the 3 internal trips are considered internally captured. The internal capture is 3 out of 10 trips, or 30%. MXDs addressed in this project may be a part of a major activity center. The level of internal connectivity and internalization of trips may be different for MXDs and ac- tivity centers. Only MXDs of less than 300 acres in size were examined in this project. • Internal Trip Capture (Area): This area can be defined to include all trips made internally to a defined area such that the trips do not use transportation facilities external to the area. For the purposes of estimating impact of such devel- opments and their internal trip capture on the transporta- tion, care must be taken when considering the impact of the internal trips on the (major) public road system pass- ing through the area. • Trip Generation: Trips to or from a specific land use or a group of land uses constitute trip generation. Trips are inbound, outbound, or total. • Transportation or Traffic Impact Analyses (TIAs) or Studies (TISs): TIAs are analyses of the impact of projected travel associated with existing or proposed land develop- ment and determination of needed access and transporta- tion system improvements to successfully accommodate the development without undue deterioration of travel conditions. Scope of Study Specifically, the project had three objectives: to develop 1. A classification system of MXDs that identifies site char- acteristics, features, and context likely to influence trips subject to internal capture; 2. A defensible improved methodology for estimating inter- nal trip capture with reasonable accuracy; and 3. A data-collection framework to quantify the magnitude of travel associated with MXDs to determine appropriate re- ductions below single-use trip generation estimates. To accomplish these objectives, several tasks were completed: • Compilation of a state-of-the-practice summary of meth- ods in use to estimate internal trip capture for use in TIS; • Development of a prototypes methodology to guide the subsequent work; • Analysis of internal capture relationships; • Determination of data needs; • Conduct of a pilot survey to test the data-collection method- ology and provide additional data; • Identification of data gaps and suggest data to be col- lected; and • Documentation of the findings, conclusions, and recom- mendations. Following a review of available methods, it was determined that there were few methods and little data available that could credibly be used to estimate internal capture for TIAs. As a result, emphasis shifted from analyzing existing data to expanding the database through an additional pilot study. Subsequently, a third pilot study was made possible through funding of a separate project by a different sponsor (Texas DOT). As a result, two additional tasks were added after the three pilot surveys: 1. Analysis and compilation of data in combination with data available from other sources, and 2. Refinement of the estimation methodology and factors and conduct a verification test. Past Research and Practice This portion of the chapter summarizes the state of the art as it was at the time the background work was completed. Land Use Synergy Interaction of land uses has probably existed since the first settlements had people who performed different types of work. Older towns and cities had all different types of uses within 8

system (10). When evaluating internal trip capture for an area, site, or activity center, the presence of safe facilities for pedes- trians and bicyclists can be a factor in the ability for a project to attract and internalize higher percentages of trips. The importance of pedestrian-based design is emphasized in many studies promoting connections between land uses, but adding the transit component completes the overall pic- ture. TODs combine the MXD with good pedestrian connec- tions and direct access to transit. Portland’s Land Use, Trans- portation, and Air Quality (LUTRAQ ) approach to land use and transportation planning worked to reduce vehicle-miles of travel (VMT), increase transit usage, increase walking and biking, and reduce trips overall. Internal trip capture was assumed to explain a portion of the VMT reduction based on the design, proximity of uses, and overall accessibility (11). In a later study by the Oregon DOT, Reiff and Kim identified several similar characteristics that may influence internal trip capture including density; land use dissimilarity; urban form; proximity to complementary uses (specifically retail- residential); building coverage ratio (i.e., compactness); and local street connectivity (12). Ewing and Cervero identified a number of potential in- dependent variables that might be used to establish travel characteristics of MXD: land use mix, availability of conven- ience services, accessibility of services, perception of safety, and pleasing aesthetics (13). Much of their quantitative findings were derived from regional transportation models and may not be directly adaptable for individual sites and developments. Kittelson & Associates listed key characteristics to be ana- lyzed for MXDs when determining internal capture rate, which were as follows (14, p.7-1): • Site Characteristics – Development size; – Land uses and quantity of development for each use; – Parking spaces provided for each use; – Density of development for each use; and 9 walking distances since walking was the principal mode of transportation. When suburbanization started to occur in the late 1800s, there began to be separations of different land use types. By the mid 20th century, zoning and single-use areas had become the normal way to develop. However, a new type of development began to be seen: the major shopping center, followed by regional malls with restau- rants, theaters, and other uses. Next came the MXDs, which had combinations of uses. Developers found the mixed- or multi-use developments appealing because such developments offered a way to capture several types of development in one project that was larger than any single project they might cre- ate in the same place. Moreover, the interaction and sharing of facilities had the potential to reduce long-term develop- ment costs and increase profitability. Trends in MXDs have progressed through many phases—from early urban villages to downtown complexes, early mixed-use towers, atrium de- velopments, and open centers and, most recently, to town centers and urban villages (3, pp. 9–22). What made MXDs work then and now is the interaction and shared-use features. The key to success is synergy between the land uses. Table 1 shows what ULI considers to be major land use combinations that have the most synergy. Several other factors that affect internal trip capture have been suggested by Steele (6)—mixing uses in proximity, clus- tering, and siting buildings to promote interaction, connectiv- ity between buildings and parcels, and proper time-phasing. To those Cervero added density, diversity, and other factors in design such as accessibility and high-quality pedestrian con- venience and provisions (7). The Sacramento Transportation and Air Quality Collaborative lists land use balance as one of the most crucial factors in reducing off-site trips (8). Filion et al. found that the synergy works best if it is pedestrian- based to reduce the dependence on personal vehicle travel and internalize the trips (9, p. 427). In their evaluation of multimodal areas, Guttenplan et al. discuss the importance of the infrastructure for walking and biking when assessing the performance of the transportation Degree of Support/Synergy Land Use Residential Hotel1 Retail/ Entertainment2 Culture/Civic/ Recreation Office Residential Hotel Retail/Entertainment Cultural/Civic/Recreation Bullets: =very weak, =weak, =moderate, = strong, = very strong. 1 Synergy is strongest between high-end hotels and condominiums, less so for mid-priced hotels and residences. 2 Restaurants and food services are the main source of benefit for offices. Source: (3, p. 85.) Table 1. On-site support and synergy in mixed-use projects.

– Proximity of residential and non-residential develop- ments within the development. • Transit Characteristics – Bus or rail routes serving the development; – Proximity of transit stops to the development; – Transit assistance provided to workers by employers; and – On-site connectivity to transit stops. • Non-Motorized Transportation Characteristics – Internal connectivity among land uses (for pedestrians, bicyclists, and motorists); – Parking spaces designated for carpools or vanpools; – Fee charged for employee parking spaces; and – Availability of on-site bicycle amenities. Gordon and Peers noted that the jobs-to-housing balance was a crucial component to internal capture of trips. People liv- ing near where they work were more likely to stay within the development area for daily activities (15, p. 144). The Florida DOT (FDOT) cites the following factors to consider when eval- uating internal trip capture: remoteness from other develop- ments and areas, development phasing, income compatibility between residents and patrons, competing opportunities, and internal circulation (16). Other factors that have been discussed by the ITE Trip Generation Committee during development of ITE’s Trip Generation Handbook as affecting MXD synergy include competing opportunities and proximity, size of both the devel- opment and the individual land uses, maturity and viability of the development and its components, and compatibility of patron/employee income levels with the development’s uses. Trip Capture—Sites The research team reviewed websites and contacted repre- sentatives of a cross section of organizations and agencies that prepare or review traffic impact studies (TISs) to determine what surveys or other data may have been completed in recent years. Table 2 summarizes the responses. It had been expected that a significant amount of survey data would be available based on responses to a 2004 ITE member survey; however, it was determined that respondents misinterpreted a question regarding data in hand. Of the 77 persons interviewed, 12 were able to provide data either directly or indirectly related to internal trip capture. Some data had already been acquired by the research team. No additional new survey data was found. Some information related to regional travel modeling was discovered as was some general or limited findings that may be usable as supporting information. The interviews confirmed that the most frequently used resource for estimating internal trip capture is the ITE Trip Generation Handbook (2, p. V-39). It contains summaries of studies of internal trip capture for individual sites and devel- opments as available through 1998. With caveats, Chapter 7 of the report provides suggested capture rates and a recom- mended procedure for use in TIS for proposed developments. The recommended procedure permits estimates for several different land uses and includes a procedure for balancing internalization of trips based on the size of the component land uses. The handbook also contains unconstrained inter- nal capture rates (that assume sufficient quantity of comple- mentary land use to accept internal trips) for office, retail, and residential land uses. These rates are based on surveys that had been made available to ITE by 1998.Capture rates for origins within a multi-use development range between 0% and 53%; for destinations, they range between 0% and 37%. Tables 3 and 4 provide the unconstrained internal cap- ture rates used in the ITE internal trip capture procedure. The handbook also recommends procedures for data- collection including interview questions. The handbook in- cludes several summaries of key quantitative and qualitative findings from previous studies of trip generation characteris- tics at mixed-use sites. For each study, available data are pre- sented on the mix and sizes of land uses within the site, the level of internalization of trips within the site, overall trip gen- eration characteristics for the site, and the level of pass-by trips for the site. In most cases, the analyses use traditional 10 Sources Type Called Interviewed1 Have Completed Surveys or Other Information Suggested One or More Others Agency Rep. 35 34 3 9 TIA Preparer 44 35 8 5 Researcher 7 3 0 1 Other 5 5 1 2 Total 91 77 12 17 1 Sources not interviewed were called at least twice and either declined interview or did not return calls. Table 2. Summary of interview responses.

ITE independent variables. In several cases, new variables are introduced. Districtwide Trip Generation Study, FDOT, District IV, March 1995. This study sponsored by FDOT was to develop databases of internal capture rates for MXD sites and for pass- by capture rates. Table 5 presents a summary of the character- istics of six surveyed mixed-use sites (17). The sites range in area from 26 to 253 acres (with four of the sites being 72 acres or less). The office/commercial square footage ranges between 250,000 and 1.3 million sq. ft. (with three of the sites having less than 300,000 sq ft). Internal Trips. Table 6 lists the proportion of daily trips generated within the surveyed mixed-use sites, which were internal to the sites. The internal capture rates ranged be- tween 28% and 41% (average 36%). Three of the mixed-use sites were further evaluated to de- termine the internal capture rates for different types of trip- makers. As listed in Table 7, the internal capture rates for trips 11 Weekday Percent Trips Captured Internally1 From To Midday Peak Hour P.M. Peak Hour of Adjacent Street Traffic Daily Office 2% 1% 2% Retail 20% 23% 22% Office Residential 0% 2% 2% Office 3% 3% 3% Retail 29% 20% 30% Retail Residential 7% 12% 11% Office NA NA NA Retail 34% 53% 38% Residential Residential NA NA NA 1 Based on limited data; NA = not available. Source: (2, p. 93) Table 3. Unconstrained internal trip capture rates for trip origins within an MXD. Weekday Percent Trips Captured Internally1 From To Midday Peak Hour P.M. Peak Hour of Adjacent Street Traffic Daily Office 6% 6% 2% Retail 38% 31% 15% Office Residential 0% 0% NA Office 4% 2% 4% Retail 31% 20% 28% Retail Residential 5% 9% 9% Office 0% 2% 3% Retail 37% 31% 33% Residential Residential NA NA NA 1 Based on limited data; NA = not available. Source: (2, p. 94) Table 4. Unconstrained internal trip capture rates for trip destinations within an MXD. Mixed-Use Site Site Size (acres) Office (sq ft) Commercial (sq ft) Hotel (rooms) Residential (units) Crocker Center 26 209,000 87,000 256 0 Mizner Park 30 88,000 163,000 0 136 Galleria Area 165 137,000 1,150,000 229 722 Country Isles 61 59,000 193,000 0 368 Village Commons 72 293,000 231,000 0 317 Boca Del Mar 253 303,000 198,000 0 1,144 Table 5. Characteristics of mixed-use sites surveyed by FDOT.

12 Table 7. Internal trip capture rates by type of trip-maker at FDOT sites. Mixed-Use Development Site Internal Capture Rate Crocker Center 41% Mizner Park 40% Galleria Area 38% Country Isles 33% Village Commons 28% Boca Del Mar 33% Average 36% Trip-Maker Crocker Center Mizner Park Galleria Area Average Users 37% 38% 36% 37% Workers 46% 49% 46% 47% Total 41% 40% 38% 40% Table 6. Daily internal capture rates at FDOT sites. made by site workers is typically higher than rates found for visitors to the site (i.e., users of the mixed-use-site services). The rates by trip-maker are consistent across all three sites. On average, 37% of user trips are internal and 47% of worker trips are internal to the mixed-use site. Finally, three of the mixed-use sites were further evalu- ated to determine the internal capture rates of individual land uses. Table 8 lists the reported internal capture rates by land use/trip purpose. In general, the higher internal capture rates were reported for trips to/from banks and sit-down restaurants. Pass-By Trips. Table 9 lists the pass-by trip proportions as determined through intercept surveys for the six study sites. Pass-by trips are made as intermediate stops on the way along a street on the way from an origin to a primary trip des- tination (2, p. 29). Four of the six sites have pass-by rates be- tween 26% and 29%. These rates appear to be high given the size and composition of the developments. Future surveys should attempt to verify these rates. FDOT Trip Characteristics Study of MXDs, FDOT Dis- trict IV, December 1993. This study was the predecessor of the March 1995 FDOT trip generation study (18). Much of the data that were collected and many of the relationships derived in this first study are included in the 1995 study re- sults described previously. The 1995 study did not report on two relationships presented in the 1993 report: a procedure for estimating internal trips and internal trip capture by time of day. Internal Trip Estimation Method. Relationships were developed for estimating internal trips as a function of the Land Use/Trip Purpose Crocker Center Mizner Park Galleria Area Office (General) 11% 11% 7% Office (Medical) – 15% 12% Retail 36% 30% 42% Restaurant (Sit–Down) 54% 52% – Restaurant (Fast) 26% – 56% Hotel 30% – 29% Bank – 48% 62% Cinema – 23% – Multi–Family Housing – 11% 50% Retail Mall – – 39% Table 8. Internal trip capture rates by land use type at FDOT sites.

combination of two interacting land use types in terms of de- velopment units (e.g., residential dwelling units and office/ retail square footage). Good relationships were developed for two internal trip type categories: residential-retail and retail- retail. The office-retail relationship was less definitive. The study presented a working hypothesis that the number of internal trips from one land use type (A) to another land use (B) within a mixed-use site is directly proportional to the size of Land Use A and also proportional to the size of Land Use B. This suggests a functional relationship of the form where: Land Use A = total site land use of Type A in residential units or 1,000 sq ft; Land Use B = total site land use of Type B in residential units or 1,000 sq ft; and Constant = a value that is solely a function of the two land use types. In the equation shown above, the constant can be derived from information collected on person trips between different land use types and on the sizes of these different land uses. Table 10 shows the derived constants. Application of these coefficients was tested for the three MXDs. Table 11 shows the results (not included in ITE Trip Generation Handbook [1]) (16, p. V-39). Two of the three Person Trips between A and B Cons t Land= ×tan Use A Land Use B× estimates were within 15% of actual; the third differed from actual by about 25%. This study also collected information on internal capture rates by time of day. Table 12 shows the total internal capture rates for the three surveyed mixed-use sites. The estimated daily, midday, and evening peak period internal capture rates are quite similar. The mean values for the entire survey period shown in the table have a high degree of statistical validity; the maximum two-tailed errors calculated using the binomial distribution, with 90% confidence-level methodology, are all less than 5%. This report also identified the percentage of employees who are also residents and vice versa (18, p. V-27). Table 13 shows the findings for each of the three developments (not included in ITE report [1]). The 16% to 19% of employees being locally employed are possibly a major factor in the re- ported internal trip capture rates. Trip Generation for MXDs, Technical Committee Report, Colorado-Wyoming Section, ITE, January 1986. This study included interviews to determine whether persons entering and leaving mixed-use sites came there for multiple purposes (19). Table 14 lists the size and mix of land uses at the eight sites with interviews to ascertain internal trip-making. Internal Trips. A key piece of information collected was the number of trip purposes that a respondent accomplished on the particular trip to the mixed-use site. Overall, a major- ity (77%) of the interviewees indicated that their trip involved only a single stop within the mixed-use site, but this still left a significant proportion (23%) who indicated they were making 13 MXD Site Daily Pass-By Rate Crocker Center 26% Mizner Park 29% Galleria Area 40% Country Isles 28% Village Commons 14% Boca Del Mar 29% Overall Average 28% Table 9. Daily pass-by rates at FDOT sites. Paired Land Uses Midday Peak Period (12 noon–2 P.M.) Evening Peak Period (4 P.M.–6 P.M.) Daily Residential/Retail 0.00082 0.00103 0.00557 Retail/Retail 0.01219 0.00995 0.07407 Office/Retail 0.00087 0.00024 0.00232 Table 10. Internal trip coefficients for paired land use types. Trip Capture MXD Model Estimate Actual Country Isles 24.5% 33.0% Village Commons 31.9% 27.5% Boca Del Mar 35.0% 32.7% Source: (18, p. V-39) Table 11. Comparison of internal trip capture: estimation model vs. actual.

14 Time Period Average Recorded at Three Sites Range Recorded at Three Sites Daily 31% 28–33% Midday Peak Period (12 noon–2 P.M.) 32% 30–35% Evening Peak Period (4 P.M.–6 P.M.) 30% 28–32% Table 12. Internal person trip ends by time of day. MXD Country Isles Village Commons Boca Del Mar Residents employed within development 3.9% NA 0.9% Employees residing within development 16.1% 16.8% 18.9% Table 13. Percent locally employed residents and locally residing employees. Site Size (sq ft) Land Uses 1 240,917 Retail, General Office, Government Office, Restaurants, Health Club, Bank 2 731,846 Retail, Office, Restaurants, Hotel 3 500,000 Retail, Office, Restaurants, Motel, Theaters 4 115,000 Retail, Restaurants, Hardware Store, Supermarket 5 1,000,000 Regional Mall, Retail, Restaurants, Banks, Office, Theaters 6 110,000 Retail, Theaters, Restaurants, Banks 7 95,104 Retail, Restaurants, Supermarket, Medical Office, Savings and Loan 8 300,000 Retail, Hardware, Restaurants, Supermarkets, Post Office Table 14. Characteristics of mixed-use sites with interviews. two or more stops within the mixed-use site. Based on these interview results, the study authors estimated that 25% of an otherwise total number of trips generated by individual trips were eliminated with the linking of internal trips within the eight surveyed mixed-use sites. Table 15 presents the number of trip purposes/stops reported by survey respondents. The responses are arrayed according to the primary destination. Office buildings and a post office generated the greatest number of multi-stop trips. Theaters, restaurants, and banks tended to generate lower-than-average numbers of multi-stop trips within the mixed-use site. The Brandermill Planned Unit Developments Traffic Gen- eration Study, Technical Report, JHK & Associates, Alexan- dria, Virginia, June 1984. Brandermill is a large, planned MXD (and, in many respects, is a small town/village) located approximately 10 miles southwest of Richmond, Virginia. At the time of the study (20), there were approximately 2,300 occupied dwelling units, with 180 townhouse-style condo- miniums and 2,120 single-family detached units. Commercial development consisted of an 82,600–sq ft shopping center; a 63,000–sq ft business park; a 14,000–sq ft medical center; and a 4,400–sq ft restaurant. There were also recreational facilities including a golf course, tennis courts, swimming facilities, and several lakeside recreation facilities. Finally, there was a day-care center, a church, an elementary school, and a mid- dle school. The study had the overall goal of determining the onsite (internal) and off-site (external) traffic generation at Brandermill. Internal Trips. Based on the various data collected, the split between internal and external trips was estimated. As Table 16 shows, 51% of the daily trips, 55% of the P.M. peak- hour trips, and 45% of the A.M. peak-hour trips were inter- nal to (or captured within) the mixed-use site. Additionally, 46% of the persons employed in Brandermill also reside in Brandermill.

Travel questionnaires were distributed to residences and used to measure the level of internal trip ends for home-based trips. As Table 17 shows, approximately 35% of the daily home-based trips from Brandermill residences are linked with trip ends within Brandermill. Over 39% of the daily trip ends to Brandermill residences began within Brandermill. For the shopping center trips within Brandermill, approximately two-thirds of the trips originate within Brandermill during the midday and evening peak hours. These internal percent- ages are higher than the Florida examples. Other Surveys. As previously mentioned, a study by the Colorado/Wyoming Section Technical Committee of ITE in- cluded surveys of eight MXDs ranging in size between about 95,000 and 1 million sq ft with varying combinations of com- ponent land uses (19). That study recommended that peak- hour trip generation rates be reduced by only 2.5% even though the surveys showed 25% internal trips. The reason is that driveway counts showed a lower reduction below esti- mates based on ITE rates. While one of the most ambitious of the early studies of internal trip capture, this study illustrates a key point: survey responses depend on how a question is 15 Number of Purposes/Stops Stated by Interviewee Primary Destination 1 Purpose 2 Purposes 3+ Purposes Bank/Savings and Loan 83% 8% 9% Hardware Store 76% 22% 2% Supermarket 77% 17% 6% Theater 93% 7% 0% Office/Work Site 68% 31% 1% Small Retail Shop 73% 14% 13% Restaurant 85% 12% 3% Health Club 71% 29% 0% Post Office 63% 24% 13% Total (Average) 77% 16% 7% Trips A.M. Peak Hour P.M. Peak Hour Daily Total Generated 2,570 2,935 33,540 External Trips 1,420 1,325 16,280 Internal Trips 1,150 (45%) 1,610 (55%) 17,260 (51%) Table 15. Percentages of persons within multi-sites by number of purposes (stops) and by primary destination. Table 16. Split between internal and external trip ends at Brandermill. worded, and asking how many trip purposes are being sat- isfied on one trip to a development may not yield the same responses as asking how many stops or how many differ- ent businesses were visited within the development or how many driving trips would have been needed otherwise. It also demonstrates that the effect of a successful (financially) de- velopment’s additional trips may overshadow internal trip capture (this is also one reason why trip generation data are so highly dispersed). For this project, the research team sought out developments that appeared to be active and had low vacancy rates. ITE recently conducted a member survey asking about avail- ability of additional studies on internal trip capture (21). The survey identified methods currently being used to estimate internal trip capture. Unfortunately, a question that inquired about trip capture data was misunderstood, and responses in- dicating 48 sources for additional information were incorrect. Other findings are described later in this section. In Transportation Research Record 1617, Steiner studied six shopping districts that were integrated within residential areas and found that in these districts walking was more prevalent, ranging from 24% to 41% of users studied (22, p. 29). Steiner

used the ITE rates for shopping centers, rather than for mixed use. Steiner compares trip rates from both ITE and NCHRP Report 187 (23) with the local daily trips that occurred in the six shopping districts studied and found situations where the ITE and NCHRP methods overestimate and underestimate trips when compared with the local data (22, p. 35). Kittelson & Associates conducted surveys for three mixed-use sites in Florida: the Crocker Center, Mizner Park, and the Galleria area. They found that the rate of internalization of trips ranged between 38% and 41% (14, pp. 5–7). Mehra and Keller reported relationships between the per- centage of internal trips and the ratio of office space to residen- tial units and the ratio of commercial space to residential units (24). Based on a Richmond Regional Planning District Com- mission Planned Unit Developments study they had reviewed, they reported finding that A.M. peak-period home-based work trips were internalized at rates between 0% and about 15% and that midday home-based other trip internal percentages ranged up to more than 40%. Both percentages increased as the ratio of office or other commercial space per dwelling unit increased in ranges of more than 80 sq ft/dwelling unit. JHK & Associates conducted a shared parking study for San Diego that included user surveys. Table 18 shows the re- sults of surveys of office worker trips to internal destinations in two MXDs (25). For both developments, 6% of the mid- day trips made by office workers are to onsite locations. Table 19 shows the percentage of internal trips to restau- rants and retail for five San Diego MXDs. Also shown are per- centages of trips made by walking. Trip Capture—Activity Centers In a comprehensive study of suburban activity centers, Hooper conducted interviews of employees, patrons, and vis- itors to office, retail, residential, and hotels within some of the largest U.S. suburban activity centers (SACs) (26). That re- search developed a comprehensive procedure for determining travel patterns, including trips internal to the activity centers. Data were collected at the six SACs listed in Table 20. In the following discussion, larger centers refer to the three centers having at least 15 million sq ft of office/retail space in each; smaller centers refer to the remaining three, which have less than 8 million sq ft. For activity center residents, Hooper found that 13% to 50% of employed residents work within the activity centers, with the average being 27% to 33% based on activity center size and whether they lived in owned or rented dwellings. An average of 50% of office employees was found to make mid- day trips outside their buildings; 20% to 33% of those trips were internal to the activity centers. Work-related, eating, and shopping trips were the most common midday trips for office employees. The study also examined stops to and from work during peak periods and found that such stops within the activity centers were made on an average of 13% to 15% of the trips. 16 Internal Trip Purpose Internal Trips Office work location to Marriott Mission Valley La Jolla Village Professional Center Business 6% – Shopping 14% 13% Eat Meal 29% – Health Club – – Other – – Total 6% 6% Table 18. Internal trips by office workers to onsite destinations. Hours Home-based trips with destinations within Brandermill Home-based trips with origins within Brandermill 7 A.M. to 9 A.M. 18% 51% 9 A.M. to 4 P.M. 44% 50% 4 P.M. to 6 P.M. 55% 34% 6 P.M. to 7 A.M. 41% 34% Daily 35% 39% Hours Shopping center trips with destinations within Brandermill Shopping center trips with origins within Brandermill 11 A.M. to 1 P.M. 66% 65% 4 P.M. to 6 P.M. 66% 52% Table 17. Internal trip ends linked with Brandermill residences and retail centers.

Hooper found that internal trips involving retail centers within activity centers were higher in larger activity centers. P.M. peak-hour internal trips averaged 24% (7% to 57% range) while midday trips averaged 37% (7% to 68% range). In the A.M. peak periods, hotel trips internal within the large and largest activity centers averaged 19% and 37%, respectively, and 27% and 36% in the P.M. peak period, respectively, with the internal percentage increasing with the amount of activity center office space. Table 21 presents a summary of some relevant relation- ships reported by Hooper in NCHRP Report 323. Many of the internal trip percentages resemble the 30% order of magni- tude reported in some of the studies previously mentioned. From the information provided, it appears that the larger SACs have higher percentages of internal capture. This is log- ical since larger activity centers (1) offer more opportunities to meet traveler needs and (2) similarly offer more choices to meet a given need. Zietsman and Joubert conducted extensive studies at three MXDs in South Africa (27, 28). They distinguished between internal trips made out of pure convenience and planned in- ternal trips that would have saved a trip on the external road network. Internal capture rates ranging from 5% to 33% were observed depending on factors such as center size, types of secondary land uses, and weekends versus weekdays. Cervero found that the existence of a retail component in office buildings in major activity centers was associated with an 8% reduction in vehicle trip rates per employee (29). Filion et al. found that over 40% of office building employees make restaurant trips outside their buildings, but internal to the activity center, averaging 2.2 such trips per week (9, pp. 420, 428–434). About one-third make similar trips for shopping, averaging about 1.6 trips per week. Four times as many retail customers said they shopped within the activity center due to location rather than because of specific retailers located there. About 55% of the internal trips are made on foot (compared with 26% driving and 19% by transit), with preference being given to “easy and pleasant” (pedestrian environment, no traf- fic conflicts) walking experiences. The researchers noted that more internalization of trips resulted from better balance, proximity, and pedestrian connectivity of interacting uses. Trip Capture—Neighborhoods, Small Communities, and Subareas Several studies have been conducted in neighborhoods and subareas to assess the amount of trip internalization as well as the differences in vehicle trip generation. Some have used regional travel modeling to compare characteristics of neigh- borhoods or areas with different design characteristics. The 17 Component Land Uses Origin Percent Internal Percent Walking Origin Percent Internal Percent WalkingMXD To Restaurants To Retail Retail Restau- rant Gen’l Office Medical Office Cinema Hotel Resi- dential La Jolla Village 23% 14% – – • • • University Square 15% 14% 2% 10% • • • Hazard Center 21% 6% 20% 18% • • • • • La Mesa Village 25% 21% 13% 17% • • • Point Loma Place 4% 25 – – • • • Table 19. Percentage of internal trips to restaurants and retail. Office Space Retail Space Hotel ResidentialSuburban Activity Center Gross Floor Area Employees Gross Leasable Area Employees Rooms Dwelling Units Bellevue (WA) 4.7 million 12,880 3 million 6,150 1,000 N/A South Coast Metro (Orange Co., CA) 3.5 million 10,465 4 million 6,865 1,800 2,300 Tysons Corner (Fairfax Co., VA) 17.0 million 35,020 7 million 13,355 3,100 15,000 Parkway Center (Dallas, TX) 13.0 million 39,000 2 million 3,430 1,800 206 Perimeter Center (Atlanta, GA) 13.0 million 32,500 3 million 5,150 910 2,000 Southdale (Minneapolis, MN) 4.0 million 13,700 3 million 6,155 2,200 3,000 Source: (2) Table 20. Characteristics of NCHRP Report 323 study sites.

research team chose not to include those here since the level of detail is insufficient for use for development sites and the need is for primary data. In comparative surveys of Austin, Texas, neighborhoods, Handy found that walkable neighborhoods with neighbor- hood shopping could generate 6.3 walking trips per (adult) resident per month to internal neighborhood retail establish- ments and that 77% of those apparently substituted for driv- ing trips (30). This might correspond to a reduction in the residential vehicle trip rate of 3% to 5%. Steiner added that higher density puts destinations closer together, making it possible to walk for some trips, thereby reducing vehicle trip generation rates (31). She cautioned that other factors such as income, household size, and other fac- tors affect transportation choices and highlighted the impor- tance of separating the effects of those factors. Ewing et al. used regional travel surveys to identify internal travel within suburban communities in Florida that ranged in size between about 600 to more than 15,000 acres (32). Al- though this is not the development scale sought for this re- search, it is interesting to note that within complete suburban communities, internal trips averaged about 25% but ranged between 0% and 57%. Ewing et al. attributed the variation to two factors: (1) larger population communities had higher 18 Average Range OFFICE EMPLOYEES Percent who make an intermediate stop within SAC • on the way to work • on the way home from work Percent who make midday trips internal to the activity center • SACs with high level of professional employment1 • SACs with low level of professional employment 10% 11% — — 7% to 15% 6% to 16% 29% to 33% 20% to 23% OFFICE VISITORS—Percent from within activity center • A.M. peak period o all SACs o small SACs o large SACs • P.M. peak period o all SACs o small SACs o large SACs — 30% 54% — 33% 58% 15% to 59% — — 15% to 68% — — REGIONAL MALLS—Percent trips which are internal to SACs • Midday o all SACs o small SACs o large SACs • P.M. peak period o all SACs o small SACs o large SACs 37% 23% 47% 24% 14% 31% 7% to 68% — — 7% to 57% — — EMPLOYED RESIDENTS—Percent who work within SACs • all • small SACs • large SACs — 27% 33% 13% to 50% — — HOTEL TRIPS—Percent internal to SACs • A.M. peak period o all SACs o small SACs o large SACs • P.M. peak period o all SACs o small SACs o large SACs — 19% 37% — 27% 36% 13% to53% — — 15% to 46% — — 1 Sites with at least 60% of the work force in professional, technical, managerial, or administrative positions. Source: 2, 26. Table 21. Internal trip-making characteristics at NCHRP Report 323 study sites.

internal capture rates, and (2) lower regional accessibility re- sulted in higher internal trip capture. This finding is rele- vant when considering the relative attraction of an internal complimentary use destination given access to similar off- site opportunities of a similar type. According to this study, easy access to regional areas decreases the attraction of ful- filling several trip purposes without increasing trips on non- internal roadways. Rutherford et al. found that in multi-use neighborhoods, the total number of trips were about the same as for subur- ban single-use neighborhoods but walk trips accounted for about 8% more of the total trips (33). Vehicle availability did not seem to be a factor, but higher household income was associated with fewer walking trips. Over 70% of the walking trips were 1⁄2 mile or less, and about 40% were less than 1⁄4 mile. Less than 10% were over a mile. This confirms the importance of proximity and walkability in internaliz- ing trips. Gordon and Peers note in their research on pedestrian design for a mixed-use community in Sacramento County (Laguna West) that based on the correlation that the National Resources Defense Council has established between urban density and automobile usage, this development may have a reduction in VMT on the order of 20% to 25% (15, p. 144). Furthermore, they noted that the job capture rate in this area averaged between 15% to 20% of local residents holding jobs internal to the area, thus reducing trips and increasing the potential for walking (15, pp. 144–145). A 2003 cordon count of Celebration, Florida—a 10-year-old, self-contained MXD of 3,500 developable acres—compared a three-weekday cordon traffic count to estimated trip gen- eration for development existing at that time based on ITE trip generation rates. The comparison indicated that actual daily external trips were 27.7% less than ITE–based estimates. P.M. peak-hour counts were 31.8% less than ITE–based estimates (34). When analyzing the impact of smart growth site design using a travel modeling process for a project in Atlanta, Walters, Ewing, and Schroeer suggested that good site design using TOD and MXD principles conservatively resulted in a 14% to 52% reduction in travel. This evaluation utilized INDEX software in the modeling process, which is discussed later in this chapter (35). A study was conducted to compare trip-making character- istics between a traditional neighborhood development (TND) in Chapel Hill, North Carolina (Southern Village) and a con- ventional residential neighborhood in Carrboro, North Car- olina (36). The TND was comprised of 920 occupied dwelling units (611 single-family, 197 apartments, and 112 condomini- ums); 30,000 sq ft of retail (including a 5,800–sq ft grocery store and a four-screen movie theater); 95,000 sq ft of office; a 90,000–sq ft elementary school (with 606 students); a 6,000–sq ft daycare center; and a 27,000–sq ft church. A survey of TND residents found that TND households made about the same number of total trips, but made fewer automobile trips and fewer trips external to the site when compared with households in the conventional neighbor- hood. A survey of the TND businesses found that 5.2% of the employees live within the TND, 39.2% of the business customers/visitors live in the TND, and 18.1% of trips to TND businesses are by walking. Based on the survey results and vehicle counts taken at the neighborhood access points, the study estimated 20.2% inter- nal capture of all trips made to or from businesses and house- holds within the TND. The comparable surveys and counts at the conventional neighborhood measured 5.5% internal capture. The study postulated that the difference in internal capture (14.7%) is the product of the TND mixing of uses and spatial characteristics. Other Related Findings One of the trip characteristics that may be needed to esti- mate internal trip capture is trip purpose. The International Council of Shopping Centers conducted surveys in 2003 to ob- tain detailed information on typical office worker lunchtime activities and shopping habits during and after the workday (37). Based on about 500 completed interviews in both subur- ban and downtown locations, retail density is not a crucial fac- tor: employee mode of transportation was more important, with driving employees spending nearly 30% more per week on each category (shopping, food, and convenience items). On average, office workers bought lunch outside their offices three out of five days a week (more often downtown than in sub- urbs). Some 62% shopped before, during, or after work at least once a week (slightly more in suburban office locations), with an average of 2.6 shopping trips per week. Office workers were reported to make about twice as many shopping trips close to home than close to work. Of their shopping expenditures, al- most 60% were on dry goods and about 40% on convenience items. In addition, 32% of respondents socialize after work at least once per week with most stopping one or two times dur- ing the week. Those stopping after work for food and drinks were about twice more likely to stop closer to home than closer to work. TCRP Report 95, Chapter 15: Land Use and Site Design, Traveler Response to Transportation System Changes contains information related to analyzing transit ridership and other travel relationships to land use and site design features (38). This report is a compilation of a large number of sources, some of which are related to internal trip capture. This report concluded that transit mode choice and ridership are highly related to development density if it is coupled with a higher level of transit service. Density alone is not enough (38, p. 15-10). Similarly, non-motorized travel (primarily walking and biking) increases with density, but in conjunction with 19

more land use mixing, compactness involving interacting uses, and pedestrian connections. This report concluded that density was not found to be significant by itself in some cases. This report also reports more walking in traditional neighborhoods (mixed use) than in late 20th-century planned unit develop- ments. This report also contains a finding that transit rider- ship declines with distance of housing to transit, falling 1% to 2% per 100-ft increase in walking distance (38, p. 15-31). A California DOT (Caltrans) funded study confirmed that residential density is insignificant (correlation −0.025) in affecting transit ridership within a 1-mile radius of a tran- sit station (36). Street connectivity was found to have the highest correlation (+0.373). Walking distance to the transit station was found to have a significant affect, as Figure 1 shows. The number of walking conflicts is more influential (–0.11 correlation) as is presence of sidewalks on one or both sides of the street (+0.171 and +0.150, respectively). That research concluded that sidewalk width, landscaping, and number of intersections have insignificant influence on transit ridership. TCRP Report 95, Chapter 15 also reports that vehicle trip generation is 1% to 3% less when improved pedestrian access is provided at regional shopping centers and 6% to 8% less for office employee vehicle trips at the edge city office build- ings containing retail (38, p. 15-12). This source also reported that Steiner found decreased vehicle use in higher-density residential areas because of closeness, safety in numbers, and attraction of supportive lifestyles that support walking (38, p. 15-18). The report contains elasticities of −0.10 for total VMT related to density and −0.05 for vehicle trips related to density, but (1) those elasticities reflect other urban area con- ditions and (2) the elasticities are derived from regional travel forecasting zonal databases and may not be directly transfer- able for this internal trip capture research (38, p. 15-23). The same report shows that good pedestrian environment and transit versus bad results in about 21% less trips per house- hold and 46% less household VMT (38, p. 15-28). TCRP Report 95, Chapter 15 also examined the relationship between jobs/housing balance and trip making. Most find- ings showed significantly better balance results in shorter trips, but not fewer trips (38, p. 15-41). The quantified results reported in this report varied widely, but one finding was that the “best new communities in the United States” are estimated to achieve 31% to 37% internal commutes (38, p. 15-41). Job balance was also reported to result in employees taking jobs closer to home, although the quantification relates to inside or outside city of residence rather than distance per se (38, pp. 15-44 through 15-45). The same report indicates that land use balance/mix has an elasticity of −0.10 related to household VMT and that land use balance/mix has an elas- ticity of +0.23 related to walk/bike trip elasticity (38, pp. 15-47 through 15-51). Another source quoted in this report indi- cates that local land use balance/diversity has an elasticity of –0.03 related to vehicle trips (38, p. 15-48). The same report contains information on residence and shopping land use mix in traditional neighborhoods—those with shopping in or adjacent to and well connected with hous- ing areas. Table 22 shows the relationship between the percent- age of survey respondents living within 1⁄2 mile of shopping and the number who reported walking to shop (38, pp. 15-52 20 Source: 39, p. 101. Figure 1. Percentage of transit commutes by walking distance from station.

through 15-53). This table shows a very close relationship between residential location and the percentage of residents who walk. Hooper showed in activity center surveys that an inte- grated development—the Dallas Galleria—had a midday walking trip share of 17% while other suburban activity cen- ters with nearby, but mostly auto-accessible, complementary uses had walk shares of only 2% to 7% (38, p. 15-61). TCRP Report 95, Chapter 15 reports that land use mix in activity centers reduce midday vehicle shares, at least to major retail, and that land use mix influences choice of vehicle or walk access, with greater mix associated with less vehicle use and more walk access (but not transit access) (38, p. 15-55). Another researcher found that vehicle trip generation rates at office buildings in suburban activity centers were 6% to 8% lower than normal and transit trips were about 3% higher than normal. The same source reported vehicle occupancy rates for 1 million–sq ft office buildings averaged 0.8 more passengers per work trip than for buildings half that size (38, p. 15-62). For activity centers with major office concentra- tions, for every 10% addition of retail or commercial uses, there was a 3% increase in non-single occupant vehicle com- muting (+0.30 elasticity) (38, p. 15-64). Similarly, it was re- ported for Seattle that walking is about twice as prevalent in mixed-use neighborhoods than for suburban-type neighbor- hoods, although walk percentages varied by location in the region (38, p. 15-72). The same report shows that household income has more effect on mode choice and on total trips per household than does whether the development is a traditional or conven- tional suburban neighborhood (38, p. 15-78). Table 23 shows results of a survey in Orange County, California. Similar walk 21 Percent Walking Trips Traditional Shopping Area Residents Living within 1/2 Mile of Shopping Area Weekday Saturday Rockridge—Market Hall (full array, restaurants) 24% 26% 28% Rockridge—Alcatraz (grocery, specialty) 40% 38% 41% Elmwood (convenience, specialty) 33% 28% 36% El Cerrito Plaza (full array) 12% 10% 10% Hopkins Specialty (food) 32% 23% 29% Kensington (convenience, services) 58% 20% 27% All Areas 32% 24% 28% Table 22. Comparison of shoppers who walk to shopping with percentage of residents within one-half mile of shopping. Neighborhood Type Travel Parameter Income Traditional neighborhood development Planned unit development All types Low 6.4 7.2 6.5 Medium 8.8 10.7 9.9 High 10.8 12.3 12.5 Mean daily trips per household All 8.2 10.9 9.6 Low 5.1 6.6 5.6 Medium 8.0 9.7 8.8 High 10.2 11.3 11.6 Mean daily vehicle trips per household All 7.0 9.8 8.5 Low 80% 91% 86% Medium 91% 91% 90% High 94% 92% 92% Percent by vehicle All 86% 91% 89% Low 6% 3% 3% Medium 2% 2% 2% High 1% 1% 15 Percent by transit All 4% 3% 3% Low 15% 11% 11% Medium 7% 7% 7% High 5% 7% 7% Percent by walk All 9% 8% 8% Table 23. Trip rates and mode share in different neighborhood types, Orange County, California.

mode results were reported for a pair of neighborhoods in northern California (38, p. 15-79). Transit mode shares at work sites vary based on different land use characteristics. As Table 24 shows, transit ridership is higher—approximately double—with substantial land use and services mixes than without (38, p. 15-86). Providing safety and aesthetics also produce greater willingness to use transit. Table 25 shows some bottom line elasticities contained in TCRP Report 95, Chapter 15 (38, p. 15-117). Local density, diversity, and design all have modest impacts on both vehicle trips and VMT. A Portland, Oregon, METRO report found a source that concluded that residents of mixed-use, gridded neighborhoods in the San Francisco area made 15% fewer au- tomobile trips and 22% more walking trips than did residents of typical suburban neighborhoods (40). It is not clear if other factors were kept constant. Current Practice When using TIS became more widespread during the 1970s and 1980s and developers took more interest in mixed- and multi-use development during the same period, traffic study preparers and reviewers began to focus on internal trip capture. In a 1993 survey of 15 Texas cities that required TIS, 11 permitted reductions for MXDs (41). One had a set reduc- tion percentage and a minimum development size; the oth- ers required justification, and what constituted acceptable justification varied. A national survey in 1994 indicated that 17% of responding agencies that required TISs permitted re- ductions for mixed use (42). Permitted reductions reported averaged 10%. Procedures vary significantly—for example, Destin, Florida, states that any claim for internal capture rate must be justified by the applicant based on empirical data for similar land uses located in similar urban environments. Data are to be from a source generally acceptable to the transportation planning profession. Any internal capture rate exceeding 25% must be justified and approved by the city (43). The City of Tempe, Arizona, simply requires that capture rates and sources of information be documented and limits internal capture to no more than 15% (44). The City of San Diego uses a simple method. It stipulates internal capture reductions to be used, providing a table of reductions by land use type (i.e., residen- tial, industrial, office, or retail) by time of day (i.e., daily, A.M. peak, and P.M. peak) (45). Table 26 is a reproduction of San Diego’s table. Retail reductions are permitted only if the re- tail is neighborhood oriented and more than 100,000 sq ft. All three approaches are used in a variety of cities. San Jose, Cal- ifornia, limits internal capture to a maximum of 10%, but provides a bonus if there is a commitment to travel demand management programs and if nearby transit is available in ad- dition to the site being mixed use (46). In California, Caltrans indicates that internal trip capture rates may exceed a 5% re- duction, but requires approval and review with transporta- tion staff (47). Table 27 was compiled by the research team 22 Percent Trips By Transit Conditions Principal Land Use Characteristic1 With Land Use Characteristic Without Land Use Characteristic Offices, residential, retail, personal services, parks within mile of site Substantial land use mix 6.4% 2.9% Four or more services, service frequency, sidewalks, transit, transit stops Accessibility to services 6.3% 3.4% Restaurant, bank, child care, dry cleaner, drug store, post office Availability of convenience services 7.1% 3.4% Sidewalks, street lighting, pedestrian activity, no vacant lots Perception of safety 5.4% 3.6% Trees, shrubs in sidewalk zone, wide sidewalks, small building setbacks, no graffiti Aesthetic setting 8.3% 4.2% 1 Sites also have TDM programs. Table 24. Transit share at work sites with alternative land use characteristics. ElasticityCharacteristic Description Vehicle Trips VMT Local density (residents + employees)/ land area 0.05 0.05 Local diversity (land use mix) Jobs/population balance 0.03 0.05 Local design Sidewalk completeness, route directness, street network density 0.05 0.03 Table 25. Typical travel elasticities related to land use density, diversity, and design.

and lists a total of 21 agencies and their requirements for accounting for internal capture for MXDs. For the U.S. Environmental Protection Agency, Criterion Partners developed geographic information systems (GIS) based software, INDEX. The INDEX software assists in deter- mining the impact of a variety of community design charac- teristics on vehicle trip generation and VMT (48). As inputs to vehicle trips and VMT, the procedure uses population and employment density; population and employment balance (as an indicator of mixed land uses); street network and sidewalk densities; distance to transit; and travel times. The methodol- ogy is calibrated and applied at the traffic-analysis-zone level. It uses zone-level regional travel model trip generation as a base and applies elasticities associated with the factors listed above. It does not directly use specific land use trip generation rates or equations of the type typically used in TISs. ITE’s Trip Generation Handbook includes a detailed method for estimating internal trip capture (1, Ch. 7). It is based on 23 Percent weekday internal trip reductions for MXDs that include predominantly neighborhood-oriented commercial retail Land use within MXD A.M. Peak Hour P.M. Peak Hour Daily Residential 8% 10% 10% Industrial 5% 5% 4% Commercial Office 5% 4% 3% Commercial Retail * * * * Commercial retail reduction equals the sum of the total mixed-use reduction in residential, industrial, and commercial office. Source: (45) Table 26. Permitted internal trip capture reductions, City of San Diego. Internal Trip Capture Procedure State Agency Max or flat % Justify/ agency approval for higher rate Agency approval ITE Trip Generation Handbook procedure Verify with survey Formula or table Other AZ Phoenix (10–15%) Tempe (15%) Tucson CA Caltrans (5%) L.A. County Newport Beach (10%) Pasadena San Diego San Jose (10%) TDM bonus CO Boulder FL Destin (25%) FDOT Additional considerations Gainesville Orlando GA GRTA (modified) IN Indianapolis NM NMDOT Prescribed by city TX Austin Other approved sources Plano Or citywide study WA Seattle D.C. Washington Documented alternative Source: Texas Transportation Institute. Table 27. Internal trip capture rates for selected agencies.

complementary land use by number of development units, trip generation rates, and trip capture percentages for any given pair of land use classifications for which data are avail- able and provides a balancing computation to ensure the ori- gin and destination land uses can send and receive the same number of internal trips. It assumes convenient internal connectivity. It depends on empirical data supplied from surveys; data in the handbook are from studies transmitted to ITE. The Georgia Regional Transportation Authority (GRTA) requirements represent a more specific approach now more commonly used (49): it requires use of the ITE–recommended practice as documented in ITE’s Trip Generation Handbook (1). However, GRTA modifies the procedure in accordance with a table that reduces the adjustments according to a com- bination of distance between complementary uses and whether bicycle/pedestrian facilities are provided (see Table 28). Any other claims for internal trip reductions must be approved by GRTA in advance. A survey conducted in 2004 by ITE indicated that 64% of the respondents use the method provided in the Trip Generation Handbook (50). The responses were from a combination of preparers and reviewers, so the percentages should not be in- terpreted as representing the portion of agencies that require a given method. Multiple responses were permitted, so the total does not add to 100%. A total of 12% reported they use locally established methods; 34% reported they use rule of thumb (usually specific percentage) methods; and 19% reported they use other detailed methods. The locally established and other methods include engineering judgment, specific considera- tions, state DOTs or other guidance, distance-based method, ULI shared parking rates, results from surveys, and travel fore- cast model. Land uses for which internal capture estimates are desired were most frequently reported to include retail, resi- dential, office, hotel, health club, theater, and conference cen- ter, but several other uses were also mentioned. Those that col- lected new data usually have done so mainly through interview surveys, although several other methods were reported includ- ing traffic and turning movement counts, parking durations/ turnover, and field observations. Additionally, Kittelson & Associates note that it is not advis- able to apply internal capture rate reductions in very-high- density MXDs that generate activity that exceeds suburban development because the rates developed by ITE were based on suburban vehicle-oriented travel patterns and may be lower than the same land uses in high-density MXDs (14, p. 7-3). URBEMIS2002, a national model for calculating air-quality impacts of projects, contains adjustments to reflect the effects of several land use and design factors discussed earlier in this chapter. Internal trip capture-related factors specifically in- cluded in formulas that compose the adjustment factors are as follows (51, 52): • Net residential density (households per net acre; excludes land consumed by arterial right-of-way); • Mix of land uses (based on number of study area [0.5 mile radius] households and employment—a jobs/housing balance—with a 2% bonus for inclusion of retail within the study area); • Transit service index (function of buses stopping within 1⁄4 mile of site, number of rail or bus rapid stops within 1⁄2 mile of site, number of dedicated daily shuttle trips); • Pedestrian/bike score (function of intersection density, and sidewalk and bike lane completeness); and • Parking supply (function of parking provided/ITE parking generation rate). Formulas are provided for each of the reductions, but the documentation does not provide complete explanations of how the formulas were derived, and it appears that at least one formula (reflecting residential density) is based on assump- tions that are not supported. Nevertheless, URBEMIS2002 provides for air-quality emissions estimation trip reductions of up to the amounts shown in Table 29. The numerical in- formation was developed using a variety of sources including some referenced above. Further review of additional support- ing documentation would be needed before the formulas should be considered for use in this project’s improved estima- tion method. The reports’ text states that redundancy has been removed by using reduction factors within the equations. Ewing slightly deviated from the standard classification of trips in the modeling process when studying communities in Palm Beach County. Ewing treated trips as part of tours rather than home-based or non-home-based (53). Assessing trips as part of a multistop and multipurpose tour or activity-based traffic 24 Percent of full reductions allowed by distance between complementary uses Bicycle/pedestrian facilities provided ¼ mile or less ¼ – ½ mile ½ – ¾ mile > ¾ mile Yes 100% 67% 33% None No 67% 33% None None Source: (49) Table 28. Adjustments to ITE Trip Generation Handbook mixed-use internal trip capture rates.

modeling is an enhancement to standard modeling that may address internal capture rates more effectively. Some have tried to adapt the ULI shared parking method for use in estimating trip generations for MXDs. While the ULI shared parking method is applicable to MXDs, it is valid only for estimating parking accumulation and not for trip generation estimation (54); however, it is apparent that some preparers are using it to estimate internal trip capture. Trip Capture Variables Travel is affected by a myriad of factors ranging from trav- elers’ own demographic characteristics to characteristics of the trip destination. Extensive research has been conducted related to travel behavior. For example, it is widely accepted that income levels and vehicle ownership affect the magni- tude of a person’s and household’s travel. Travel time, travel distance, available travel modes, residential development density, and other factors have all been shown to influence travel characteristics. Table 30 lists a wide range of variables that could influence internal trip capture. Also listed are con- siderations that are applicable in selecting a smaller set of variables for consideration in developing an improved esti- mation procedure. Table 30 also lists (in the first column) the final candidate variables selected by the research team for consideration in developing an improved estimation method. These variables were selected based on causal relationship to internal trip cap- ture, ease of quantification in the field and from preliminary site plans, potential data availability, data collection complex- ity, and likelihood of acceptance by the user community. Chap- ter 3 addresses these variables more fully. Trends in MXD and classification of land uses found in MXDs are covered in Appendixes A and B. Summary These findings revealed several estimation techniques and a lot of related data and research findings, but detailed sur- veys of only seven MXDs (six in two Florida studies and one in Virginia). Hard-copy survey data were acquired for the six Florida sites. All were completed by the mid-1990s, prior to the time that ITE published the first edition of its Trip Gener- ation Handbook in 1999 (55), which as an ITE–recommended 25 Land Use Type Physical Measures Residential Non-Residential Net residential density Up to 55% Not applicable Mix of uses Up to 9% Up to 9% Local-serving retail 2% 2% Transit service Up to 15% Up to 15% Pedestrian/bicycle friendliness Up to 9% Up to 9% Total Up to 90% Up to 90% Source: (51, p. 3) Table 29. URBEMIS2002 trip reduction credits related to internal trip capture factors. Use Variable Anticipated Sensitivity Comments No Density/compactness High Proximity High No Connectivity High Combine as a single independent variable (proximity) No Parking Moderate Reflect instead in mode of access that may be considered similar in effect. Parking-supply constraints reduce total trip generation but may not significantly change internal trip capture percentage. Normally only a factor in central business districts (CBDs), TODs; such sites may require special study anyway. Add parking garage “access time” to impedance used for “competing external opportunity” model component. Land use synergy High Use as “yes/no” variable to match users among site land uses Balance of land use quantities High Use as control check No Principal trip purpose to site High Covered largely by land uses and time of day Mode of access Moderate Driver trips can be associated with mode of access to site for primary trip. Primary trip purpose strongly influences mode of access. Will be a significant factor where good transit service exists. Time-of-day Moderate Day of week, season Moderate Provide one trip capture table for each time period of interest (e.g., weekday A.M. peak hour, P.M. peak hour, midday; Saturday peak hour) No Competing external opportunities High Attempt to quantify if data can be found. Data expensive to collect. Table 30. Candidate independent variables.

practice (approved by its International Board of Direction) contained the first endorsed internal trip capture estimation technique for use in TISs for MXDs. Most public agencies and preparers of TISs use the ITE method (or a locally developed variation of the ITE method). The two other approaches that are also commonly used are (1) a local agency accepted or established internal trip capture reduction percentage to apply to estimated site vehicle trip generation and (2) negotiations between the study preparer and agency reviewer. Developers, through payment for TIS, have typically funded most previous site trip generation research; however, since the appearance of the ITE Trip Generation Handbook that endorsed an estimation method and provided some data on capture rates for the most frequent mixed uses, a combination of high cost of internal trip capture data collection and an existing accepted method have resulted in no new comprehensive data. Since the late 1980s, there have been numerous studies of various census and regional travel survey databases, limited site data collection, and studies and surveys of related travel and development characteristics that could contribute use- ful material for developing an improved estimation tech- nique. Many studies were related to mode of access and find- ing ways to promote transit usage, including through use of land use and development tools such as TODs. Internal trip capture rates found in the research vary widely depending on conditions and land uses, but for developments with major commercial components, capture rates (percentage of trips made from internal points to internal destinations) typically ranged up to more than 30%. For mixed-use neighborhoods and small communities, internal capture reached 50% and even higher. Interaction between individual pairs of land uses, in the proper balance, also was found in similar ranges; however, it appears from the available data that few develop- ments (all uses combined) completed by about 2000 can typ- ically be expected to have internal capture rates much above 30%, and that percentage requires the right mixes and bal- ances of land use mix. Besides land use mix, other factors were found to affect in- ternal trip capture. These include connectivity and proxim- ity between interacting land uses and location within an urban area (thought to reflect both competing opportunities and modal options). Conflicting information was found on the effects of development density. Modal impacts found were attributed to proximity to transit (with good service). Trip generation rates and mode split were found to be af- fected by such traveler characteristics as income and vehicle availability. However, no site-internal travel data have been collected that included those characteristics, and they would be hard or impossible to accurately project for a proposed development at the zoning stage. Conclusions Based on this review of past work and the personal experi- ence of the research team, the following were selected as being a reasonable starting place for NCHRP Project 8-51 to de- velop an improved internal trip capture estimation method: • To be of value, the project should address both mixed-use and multi-use developments (hereafter referred to in com- bination as MXDs). • Activity synergy between the different uses within an MXD is what captures trips internally. Other factors contribute to making this synergy and interaction both possible and more or less attractive compared with other opportunities. • Land uses that are most frequently identified as having synergy of the type that affects trip making and that are commonly included in MXDs include residential; retail (especially convenience); office; hotel; restaurant; and en- tertainment (theater). However, within each general land use classification, there will be a need for subclassifications if a method is to be easily and accurately applied. Chapter 3 addresses land use categories. • The research team identified other characteristics most likely to influence internal trip capture and be most readily devel- oped in actual practice. Table 30 lists these characteristics. • Trip capture has been studied at essentially three develop- ment levels: single-site project, larger multi-site develop- ment and activity centers, and neighborhoods and subareas. The issues and challenges are similar, but some implications of internal trip capture are different and the extent and com- plexity of data collection will be different. Findings at each level may not be directly transferable, at least quantitatively. • Specifically, there are more different scales of mixed devel- opment that may act somewhat differently or have to be treated or have data collected in different ways: – Single developments; – Blocks of separate interactive developments; – Small areas of blocks containing interactive uses; – Neighborhoods and districts with multiple interactive uses; – Mixed- and multi-use subdivisions; – Multi-use activity centers; and – Small communities. • The sites for which travel data were used to develop the recommendations in this study are all single master-planned developments. Mockingbird Station is a single block. Atlantic Station and Legacy Town Center are multiple block districts containing fully integrated and adjacent complementary uses. Boca del Mar, Country Isles, and Village Commons all contain pod-type mixtures of single-use development within a single development to provide the mixed-use interaction. 26

• Trip capture percentages vary greatly among land uses and development types. They also vary by time of day and prob- ably to some extent by the day of week and by season. Var- ious studies have found internal trips make up as little as 0 and as much as more than 60% of total trips generated. Sev- eral studies included multiple developments or areas and were able to compute averages. • The extent to which trips are captured internally may also be influenced by other factors, such as – Availability of personal vehicle during the stay at the pri- mary destination (accounted for by mode of access); – Match between traveler characteristics and characteris- tics of potential destinations (e.g., market position ver- sus income levels); – Availability of competing onsite and off-site opportuni- ties; and – Internal and external accessibility (including such fac- tors as proximity, connectivity, cost, comfort, attractive- ness, convenience, parking availability, etc.) to desired activities. • Local data or more diverse and representative data points regarding internal trips associated with the different MXDs and multi-use-development types is needed to improve the accuracy of predicting trips for MXDs. • Despite the availability of the method provided in the Trip Generation Handbook, several other methods are being used. Some are arbitrary (e.g., set or maximum percentages), and a few are incorrect for application to transportation or TIA or studies (e.g., ULI shared parking percentages). It appears that only the ITE method balances internal trips based on the amount of each interacting land use. • Two methods are most currently used for estimating inter- nal trip capture: The ITE method contained in the Trip Generation Handbook, 2nd edition (1), and percentages that local agencies establish as acceptable. In many cases, these methods are specified in local agency TIS require- ments or even ordinances. Both approaches are easy to use and require minimal data. • Since the advent of the first edition of the Trip Generation Handbook in 1999 (55), there has been wide acceptance of internal trip capture percentages contained in the hand- book or lower values accepted by review agencies. The cost of internal trip data collection is high compared with other TIS components, which has resulted in little incentive for developers to fund collection of new data. Obtaining devel- oper commitments to fund additional data collection may be a challenge unless there is expectation of major increases in internal trip capture credit. • Little detail was found in the literature on data collection methods. The research team’s familiarity with data collec- tion for internal trips has revealed a relatively high cost necessitated by interviews, a low return rate on intercept mail-back surveys, and, most crucially, significant variabil- ity in questions and the way they were asked—which affects data stability and accuracy. A standard, low-cost method for collecting data is needed. • Travel forecast models have been used to provide the basis for internal trip estimation and even directly to estimate internal trips. Given the absence of intrazonal trips on the model network and limits to traffic analysis zones, these travel models are not usable for estimating internal trips for TIS or traffic impact fee use. In conclusion, the estimation and data-collection meth- ods developed by NCHRP Project 8-51 should be easily used, explained, and understood so that they can be used in zoning cases and other TIS applications as well as for other more sophisticated uses. They should also be as economical as possible while supplying enough data to be reasonably reliable. 27

Next: Chapter 2 - Research Approach »
Enhancing Internal Trip Capture Estimation for Mixed-Use Developments Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s National Cooperative Highway Research Program (NCHRP) Report 684: Enhancing Internal Trip Capture Estimation for Mixed-Use Developments explores an improved methodology to estimate how many internal trips will be generated in mixed-use developments—trips for which both the origin and destination are within the development.

The methodology estimates morning and afternoon peak–period trips to and from six specific land use categories: office, retail, restaurant, residential, cinema, and hotel. The research team analyzed existing data from prior surveys and collected new data at three mixed-use development sites. The resulting methodology is incorporated into a spreadsheet model, which is available online for download.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!