Urban planning community

+ Reply to thread
Results 1 to 6 of 6

Thread: TDM modeling

  1. #1
    Cyburbian Greenescapist's avatar
    Registered
    Apr 2002
    Location
    Massachusetts
    Posts
    1,169

    TDM modeling

    Has anyone used EPA's COMMUTER model to generate estimates for the success of various TDM programs/policies? Did you find this spreadsheet model helpful, reliable?

    I am trying to find a model that will help estimate the results of TDM program levels at a large university. The university does not have good data for mode splits, etc, so I'm using EPA's regional defaults. Any tips are much appreciated.

  2. #2
    Cyburbian Flying Monkeys's avatar
    Registered
    Sep 2005
    Location
    North Florida
    Posts
    608
    A couple of questions...

    When you say you are testing TDM policies...I am assuming you mean TDM as in Transportation Demand Management?

    Are you testing policies like the price of parking and headway on busses?

    Is there a full model available for the area you are in?


    If this is what you are testing, and it is a large university in a metro area, then the EPA spreadsheet will be a simplistic look at best. What it will give you will be related to air quality, no? And if there is a real demand model for the area (large university in a metro area, I am thinking yes) any results you generate may be challenged because you did not use the proper tool for the project.
    What’s in a name? – Your reputation….:)

  3. #3
    Cyburbian Greenescapist's avatar
    Registered
    Apr 2002
    Location
    Massachusetts
    Posts
    1,169
    Quote Originally posted by Flying Monkeys View post
    A couple of questions...

    When you say you are testing TDM policies...I am assuming you mean TDM as in Transportation Demand Management?

    Are you testing policies like the price of parking and headway on busses?

    Is there a full model available for the area you are in?


    If this is what you are testing, and it is a large university in a metro area, then the EPA spreadsheet will be a simplistic look at best. What it will give you will be related to air quality, no? And if there is a real demand model for the area (large university in a metro area, I am thinking yes) any results you generate may be challenged because you did not use the proper tool for the project.
    Yes, transportation demand management policies and physical improvements. I found an EPA model called COMMUTER that takes information on many things- cost of parking, transit improvements, program improvements (transit passes, TDM manager, etc) and compares those changes to current mode splits. The EPA model does give figures on air quality improvements, but I'm not interested in that right now.

    You're probably right that this model will give me a simplistic impression of TDM effectiveness. However, it appears there are few TDM models. Effectiveness of programs must be difficult to measure. BTW, the university I'm working for is a large state university in a fairly rural area and I don't believe they have any TDM modeling for the county. I got some mode splts from their MPO, but that's it.

  4. #4
    Cyburbian Flying Monkeys's avatar
    Registered
    Sep 2005
    Location
    North Florida
    Posts
    608
    Items like parking availability/fees policy can be modeled fairly well in a traditional model.

    I do not know how your spread sheet looks at intrazonal travel.... Does it just say 'you have 100 parking spots today, tomorrow you have 200, therefore you are inducing 200 trips a day' sort of simplistic logic...or does it incorporate some type of attraction model that relies on current and projected population data? (ex. traditional four-step model uses traffic analysis zones with pop and employment data and a gravity model routine.)

    Is there a link to this spreadsheet?

    I am thinking that this spread sheet does not look at intrazonal travel; therefore it might be in your best interest to explore where trips to this university will come from. Do they have some current and future pop and employment data in GIS?
    This could help you to kinda 'ground truth' your conclusions. For example, your model shows that for every .25 cents you reduce fares, the ridership will increase by X. It extrapolates X 15 years and states that ridership will be XX. Is XX reasonable give population projections and population distributions?

    Don't know if I am helping....or confusing...
    What’s in a name? – Your reputation….:)

  5. #5
    Cyburbian Greenescapist's avatar
    Registered
    Apr 2002
    Location
    Massachusetts
    Posts
    1,169
    Flying Monkeys- you are being helpful, thanks.

    I don't believe the model does take into account attractions, like a gravity model would. It's developed to assess effectiveness of TDM programs, for regions and specific employers. Here is a link to the EPA site with the model manual and the program (Excel).

    http://www.epa.gov/otaq/stateresourc..._transp.htm#cp

  6. #6
    Cyburbian Flying Monkeys's avatar
    Registered
    Sep 2005
    Location
    North Florida
    Posts
    608
    I looked at the model. It is very simplistic. It is based on a strict set of formulas (you can check those formulas out for yourself, just click on ‘disable macros’ when the program starts, then look at the tabs). So for a quick look (sketch planning level) based on what little info is available, you can present some general conclusions using the program. However, I would caution you to try to localize as many of the models coefficients as possible....the EPA manual states;

    “Mode choice models have been developed for many cities and regions nationwide, so you can often use coefficients that have already been developed to reflect local characteristics."

    (read the manual, it has good info) The EPA manual also states the following:

    “The COMMUTER Model is most appropriately applied to a single worksite, employment center, or subarea for sketch-level analysis purposes. Its most robust predictive power lies with measures that can be translated into time and cost strategies. The factors used for employer support and alternative work arrangements are based on the best research and professional judgment available. In most cases, though, the empirical evidence to support these factors is very limited. Furthermore, the impacts of any one type of program can vary substantially, depending upon the characteristics of the worksites being evaluated, the employees who work there, and how effectively a given program is implemented. For some planning purposes, therefore, it may be necessary to gather additional empirical data and/or conduct more detailed modeling, in order to increase the reliability of the estimate of a program’s impacts.”

    The ‘more detailed modeling’ is what I am thinking of. This program can tell you what ‘strategy’ to reduce VMT would appear to give the best result, given your inputs. I would recommend using a travel demand model, or developing your own ‘manual’ model, if you want to put a real number to the mode shift that you are looking to predict.

    Developing your own ‘manual’ model, if a calibrated model is not available, merely means looking at the surrounding SE data, projecting that data, then using a logical methodology to distribute that traffic. Then based on your mode shift, you could forecast some numbers.

    Uh….I made it sound simple…but if you are interested I can refer you to some TRB literature.
    What’s in a name? – Your reputation….:)

+ Reply to thread

More at Cyburbia

  1. Replies: 0
    Last post: 16 Aug 2012, 2:00 PM
  2. Modeling forums
    Transportation Planning
    Replies: 2
    Last post: 16 Jun 2010, 9:09 AM
  3. Transportation Demand Modeling
    Career Development and Advice
    Replies: 1
    Last post: 11 Mar 2005, 9:23 AM
  4. Transportation modeling for rural areas
    Transportation Planning
    Replies: 0
    Last post: 01 May 1997, 3:30 PM