Enhancing Internal Trip Capture Estimation for Mixed-Use Developments (2011)

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.

S U M M A R Y The goal of NCHRP Project 8-51 is to improve the methodologies used to estimate the extent to which trips made within mixed-use developments (MXDs) are internalized or satisfied with both origin and destination within the development. Specifically, the project developed • A classification system of MXDs that identifies site characteristics, features, and context likely to influence trips subject to internal capture; • A defensible improved methodology for estimating internal trip capture with reason- able accuracy; • A data-collection framework and methodology to quantify the magnitude of travel asso- ciated with MXDs to determine appropriate reductions below single-use trip generation estimates; and • A spreadsheet estimation tool to facilitate computations (available at http://www.trb.org/ Main/Blurbs/165014.aspx). The improved estimation method was developed from existing survey data from prior studies plus three pilot data collection surveys from this project. The method is based on the existing Institute of Transportation Engineers (ITE) procedure, but expands it to cover both A.M. and P.M. peak periods, six primary land uses found at MXDs, and proximity of interacting land uses. This method was tested and found to reduce estimation error by about one-half compared with the existing ITE method and three-fourths compared with raw trip generation estimates. Summary of Findings, Conclusions, and Recommendations Existing Practice Internal capture for MXDs is of most interest to those who either prepare or review trans- portation impact analyses (TIAs) for such developments; however, transportation planners and developer consultants are also interested in internal capture and the resulting external trip generation. Some additional uses include planning for transit-oriented developments (TODs) and preparing environmental impact statements or assessments. ITE provides a recommended practice for estimating internal capture and associated exter- nal trip generation for such developments. The ITE method documented in the Trip Genera- tion Handbook (1) is the most widely used technical method. The other widely used approach is a policy determined flat percentage reduction in external trips. Such percentages are estab- lished by local planning, zoning, or transportation engineering officials for use in TIAs prepared Enhancing Internal Trip Capture Estimation for Mixed-Use Developments 1

2to support applications for zoning, subdivision, site plan approval, or access permits. The per- centages are usually arbitrarily selected for use throughout the jurisdiction. These percent- ages are most typically in the range of 10%, but were found to range between less than 5% and as much as 25%. The ITE method covers trips between only the three most frequent components of MXDs— office, retail, and residential. Data are available for the weekday P.M. peak hour; midday; and what is called “daily,” but which is drawn from data collected between noon and 6:30 P.M. The ITE method has nothing for the A.M. peak hour. The policy percentages mentioned above are applied to each analysis period used. There is some use of invalid applications for internal capture estimation—the two found most frequently were use of shared parking reduction percentages and metropol- itan area travel forecast model intrazonal trip percentages. Shared parking reductions apply only to parking accumulations in a parking facility serving multiple uses; the per- centage reduction applies only to parking accumulation, not trip generation. Intrazonal trips apply to complete traffic analysis zones used in regional travel forecast models. Zones may range from a block to a square mile. Intrazonal trips are for the complete zone and are not applicable to portions of a zone. Estimates are also accurate only to a regional level, not a development-site level. Neither method should be used for estimating internal capture for MXDs. Six land uses are the most frequently used components of MXDs—office, retail, restau- rant, residential, cinema, and hotel. Most major MXDs have all of these. Most other MXDs have at least four. Six MXDs analyzed in this project ranged in size between 7 and 300 acres. All were single developments from one master plan developed to fully inte- grate all uses. Available Data There are very limited data available that are capable of supporting internal capture rate estimation methodology that can use information that is available at the time of zoning. Three Florida surveys plus three more conducted for this and a parallel project were the only surveys with enough detail to develop internal capture methodology • For both A.M. and P.M. peak hours, • For use with information that is available at the time of zoning requests and can be reli- ably projected, • That provides the ability to analyze the effect of proximity of land uses to each other, and • That is sensitive to differences in land use mix. Some cordon counts have been completed for various periods and could be used for val- idation testing. More data are needed. Internal Capture Estimation Methodology This project expanded the database from three to six developments and after considering options, extended the ITE method to • Add the weekday A.M. peak hour; • Added restaurant, cinema, and hotel land uses; • Created a land use classification structure that would permit disaggregation of the six land uses to more detailed categories should enough data become available;

• Include the effects of proximity (convenient walking distance) between interacting land uses to represent both compactness and design; and • Provide a method that could easily be put in spreadsheet form. The method uses the following inputs: • User-estimated A.M. and P.M. inbound and outbound vehicle trip generation for six land uses: office, retail, restaurant, residential, cinema, and hotel; • Mode split for MXD trips to/from each land use—percent by automobile, transit, non- motorized; • Vehicle occupancy by land use; and • Average walking distance between land use pairs. The following outputs are produced: • A.M. and P.M. peak-hour internal person trips by land use in origin-destination form; • A.M. and P.M. peak-hour percent internal capture (person trips); and • A.M. and P.M. peak-hour inbound, outbound, and total external trips (trips to and from the development being analyzed) by mode: – Person trips; – Vehicle trips; – Transit trips; and – Non-motorized trips. This method was tested for its ability to estimate external vehicle trip generation. The existing ITE method estimates produce about one-half as much error as do ITE trip gener- ation rates. The method developed in this project cuts the estimation error in half again or roughly to about one-fourth of the raw trip generation rates. The recommended method is described in Chapter 3. The researchers recommend its use for developments of up to 300 acres. Further testing could validate its use for larger developments, but that has not yet been attempted. Due to the difference in scales and reduced levels of inter- nal connectivity, the researchers do not recommend use of this method for suburban activity centers or new town types of development: the researchers do not believe it will be applicable. The recommended estimation method was validated by testing it against actual data from several MXDs. The recommended method was found to be more accurate for estimating external vehicle trips for MXDs than either the existing ITE method or unadjusted ITE trip generation rates and equations. Compared with peak-period cordon counts, the recom- mended method overestimates external trips by an average of about 1%. More telling is an absolute average of about 13% and a standard deviation of about 15%. Details and compar- isons with the other methods are discussed in Appendix F. Recommended Modifications to Existing ITE Procedures As mentioned previously, the recommended estimation method builds on the current ITE internal trip capture procedures contained in the second edition of the Trip Generation Handbook (1). Incorporation of this project’s recommendations could be accomplished by the following: 1. Expanding Tables 7.1 and 7.2 of the Trip Generation Handbook to include all six land uses covered in this report; 3

42. Adding the proximity adjustment to be made after the unconstrained internal capture estimates are performed but before the balancing process; and 3. Modifying the data-collection procedures to include those recommended in this project. Data-Collection Methodology A methodology and procedural instructions were developed for the selection of data- collection sites and for the data-collection itself. Those procedures were used to conduct surveys at three MXDs. The researchers recommend that additional data be collected. The researchers suggest that MXDs selected meet at least the following criteria: 1. Be representative of typical MXDs being developed or being planned so the data will be of use for future years; the area in which the MXD is located should also be representative; 2. Have at least four land uses; 3. Have owners or managers who will permit the needed surveys to be conducted; 4. Be easy to conduct a large enough sample for an affordable cost (in 2006 dollars, the three surveys each cost about $50,000 to set up, conduct the surveys, and summarize data); 5. Be generally in range of 300 to 500 acres or less; and 6. Be economically successful (by appearance) and mature (fully occupied for at least a year and in an area that is mostly developed). Organizations that collect additional internal capture data are encouraged to provide a copy of the data and analyses to ITE for further use and future refinement of what was pro- duced in this project. Application in Practice This research project developed an improved estimation methodology and data-collection framework for use in estimating internal trip capture in MXDs during weekday A.M. and P.M. peak periods. The estimation methodology is based on weekday A.M. and P.M. peak-period survey data from three MXDs in Texas and Georgia (part of this project) plus similar week- day P.M. peak-period data from three developments in Florida (prior project). The six developments surveyed ranged from about 7 to 300 acres in size and had between four and six primary land uses each. This report presents a technical advancement beyond the internal capture method pub- lished in the 2nd edition of the Trip Generation Handbook (1). The researchers believe that the limited validations conducted for the proposed estimation method confirm that the results provide accurate approximations of external trip generation for typical MXDs con- sisting of office, retail, restaurant, residential, cinema, and hotel land uses, consistent with the accuracy of trip generation estimates for single-use developments as portrayed in such references as the 8th edition of Trip Generation (2). User Instructions and Cautions At the time of publication of this report, the approach developed in this research had yet not been advanced through the ITE process for development of recommended practices and, therefore, should not yet be considered an ITE–approved methodology. This report presents information in Chapter 3 on how to use the proposed estimation pro- cedure; however, the researchers and the overseeing NCHRP project panel felt it is important

to encourage users to adhere to the following instructions and cautions in using the proposed estimation methodology: • Identify specific land use components of the MXD and classify them into the six classifications—office, retail, restaurant, residential, cinema, and hotel—covered by the estimation methodology. Any component land uses that do not fit into those six classifi- cations or are too unique to be considered normal for a classification should be kept sep- arate. No internal capture is estimated in the proposed methodology for trips between uses within each of these categories (e.g., two or more different retail uses). • Estimate single-use trip generation individually for each land use within the MXD. Then, sum the individual estimates into the six aggregated classifications: office, retail, restaurant, residential, cinema, and hotel. Do not combine development units into the six classifications and then use one single-use trip generation rate or equation to estimate trip generation for the aggregated land use. • When applying the internal capture estimation methodology, use the percentages suggested in Chapter 3 unless local data are available from developments similar to the development being analyzed. Users are cautioned that data gathered in a method dif- ferent than the data-collection methods described in this report may not be applicable and could produce inaccurate internal capture estimates. • Do not apply the internal capture percentages for this report to other land uses. Inter- nal capture estimates were not developed for land uses beyond the six classifications pro- vided herein. The extent of the internal capture for other land use pairs has not been tested as part of this project. The results presented in this report are based on surveys of six MXDs, and validation was limited to seven such developments. As a result, some members of the project’s advisory panel strongly recommend that additional research, data collection, and validation testing be conducted before the method is adopted for use in TIAs. Furthermore, caution should be exercised in the application of this methodology. For example, it cannot be concluded that the methodology will be appropriate for MXDs that differ significantly from those surveyed in this project in terms of • Regional context, including competing opportunities outside the development; • Access and parking; • Scale of the development; • Complementary land uses, including specific pairs of business types; • Specific residence types, • Other component characteristics within each land use category; • Proximity and connectivity between each pair of land uses, especially the layout of the land uses relative to each other; • Other characteristics such as proximity to transit and pedestrian access within and around the site; and • Colder locations that might limit or constrain pedestrian traffic. Suggested Future Research Clearly, this project has made progress in estimation of internal capture; however, the data- base is still sparse and much that is thought to be logical about MXD travel characteristics is 5

6still unproven and even largely untested. Two of the research efforts recommended by the research team include the following: 1. Collect more data at MXDs—the researchers think data are needed from at least six more sites that have five to six land uses. 2. Independent of the additional data collection, test the applicability of the existing methodology for MXDs of different sizes, character, and land use components. Use val- idation tests similar to those used in this project. The only data needed are a complete directional cordon count for the A.M. and P.M. peak-hours plus development data and a good site plan from which to estimate proximities. Request for Additional Data Users are encouraged to collect and contribute additional data using the data-collection pro- cedures described in this report. Those data could be used to further enhance the accuracy of the proposed methodology and/or expand the number of land use classifications covered by the methodology. New data should be forwarded to the Institute of Transportation Engineers at 1627 I Street, Suite 610, Washington, D.C., 20006-4007, or by email to ite_staff@ite.org. Report Contents NCHRP Report 684 is composed of the following sections: • Chapter 1: Introduction is a summary of findings from a review of the state of the practice. • Chapter 2: Research Approach describes the objectives, approach, and work performed. • Chapter 3: Findings and Applications describes the work performed and the results, findings, and recommended estimation methodology. • Chapter 4: Conclusions, Recommendations, and Suggested Research includes those items plus lessons learned. Detailed procedures for computations are also included. • References lists the works cited in this report. • Appendix A: Trends in Mixed-Use Development is a description of past and expected trends and characteristics of MXD. • Appendix B: Land Use Classification System presents a framework for classifying land uses for analysis of internal capture. • Appendix C: Procedures for Internal Capture Surveys is a detailed description of how to prepare for and conduct surveys to obtain data for use in analyzing internal capture for MXDs. Sample forms are included. • Appendix D: Pilot Survey Experiences and Lessons Learned contains useful information for those who may be planning to conduct internal capture surveys. • Appendix E: Florida Survey Questionnaires presents the samples of questionnaires used for three Florida internal capture surveys that produced some data used in this project. • Appendix F: Validation of Estimation Procedure documents a test of seven MXDs for which recommended estimation methodology was tested for its ability to reproduce cor- don external vehicle trip volumes. Additionally, a spreadsheet estimation tool to facilitate computations is available at http://www.trb.org/Main/Blurbs/165014.aspx