Hi Rusdi. I can see two questions:

1) How to calculate ADT based on 1-hour survey?

It is practice to use adjustment factors to "factor up" short count into ADT. More for example here - http://en.wikipedia.org/wiki/Annual_average_daily_traffic

But it is critical to choose the duration so that the result ADT is precise enough. In the Czech Republic we are using minimaly 4-hour period (7-11am or 1-5pm, to omit off-peak hours). So I am not sure whether 1-hours count is sufficient. It should be based on some traffic engineering practice in your jurisdiction. Or use some international sources such as ITE handbooks (I have some in the office, not here at home, I can find it there for you on Monday, if you need)..

2) The division of road is no problem until there is no intersection in between, since intersections may "add" further volume. On segments between intersections ADT should be the same (number of entering vehicles = number of leaving vehicles, with the exception of parking vehicles).

Maybe I am explaining in a silly way, sorry for that :-)

Hi Jiri,

Thanks Jiri, it is very helpful for me.

This study focus to the three major federal highway connection between main town. We already decided to divided them to 322 segments and inside them, a few big and small junctions. I think we should separate the road based on the big junctions - now we call it as section. Small junctions are not to be considered because low entering and leaving vehicle. After that, conduct traffic census in appropriate hours (have to confirm) for every section.

Intersection volume should be counted separately.  Road mid block should be counted separately.  Even one hour data is enough to understand if collected during peak hour time.then you can always convert to ADT.  Though ADT requires data collector over different season of year to cover randomness.

Thanks Ravindra,

I got ADT data  for the whole road from local authorities. I have to recheck back either this ADT consider season effect. Do you have any formula to convert one hour traffic volume data to ADT?

Hi Rusdi;

for converting a 1 hour counting result to an ADT-value, you need to know the variation patterns of the traffic volume on the road in question. That means, that you have to know the seasonal variations, weekday variations and the hourly variations. The question is of course, where you can get these patterns, if there has not been continuous countings on the road in question. Usually the road administrations and city transportation offices have information of typical variation patterns collected from different types of roads in different areas. So, try to find such a locations with variation data, which correspond your own counting sites as good as possible and use those as a reference locations and basis for your calculations.

Further on, it is also important to choose the 1 hour counting period carefully. Choose a time of day with a representative volume (not too small, but not necessarily the peak hour either). In general the accuracy of the estimation increases with volume, but it is possible, that certain daytime hours between peak times are more stable than the peak volumes. Choose also a normal weekday (no special events or other irregularities in flow). If possile, repeat your count some other day.

To get the ADT form one hour traffuc, you first have to count the daily estimate from your one hour value by the knowledge of the share of this hour from the whole day in your reference location. After that, you convert the value to correspond an average weekday based on the weekday variations and finally you convert your average weekday to yearly average.

In a nutshel, what you need is a reference location for each of your counting sites and based on the variation patterns of the reference locations, you can estimate the ADTs for your counting points.

For example the general text in HCM gives examles of the calculations and variation patterns.

Yours,

Matti Pursula

Rusli,

I suggest that since you have an ADT for the whole road divided into 24-hour values, you may use the same percentage for each hour of the day to convert any hour to an equivalent ADT.

Another idea, if the one hour count is during the peak hour then you may use the k-factor to convert the peak hour volume to the equivalent ADT. Typical k-factor can be obtained from variety of sources, AASHTO, etc.

Best of luck

Khaled Al-Sahili

Thank you all for your answers. I'm working on that right now with consider all the recommendation.

Dear Rusdi,

In accordance with the gentlemen above, you can work with expansion factors from continuous count stations. In order to determine such factors is necessary to have counts of 24 hours, hour by hour. With this information you will get hourly expansion factors. But also, with enough data (month and year counts) it's possible to calculate daily factors and monthly factors. You will find a better explanation in Nicholas Garber's book "Traffic & Highway Engineering", page 130, available at link below.

I hope this information will be useful to you.

Lenin Bulla

https://books.google.com.co/books?id=x8TaPEE0CZkC&printsec=frontcover&dq=nicholas+garber&hl=es&sa=X&ei=bzXcVOq4JYemgwS344CoBA&ved=0CCIQ6AEwAQ#v=onepage&q=nicholas%20garber&f=false

Hi Lenin,

Thanks for your information..It is helpful to me.

what was the best solution after all?

if you have previous old data you can extrapolate current one hour according to that.

i do not understand the extrapolation thing here.

for example i have the traffic counts for a full day but some of hours are missed due to a technical problem  in the counter itself so may i use the extrapolation to get estimations for the missing numbers? how? please check the attached sample data

thanks

Plz chk this doc

(i) For the prediction of the day's volume from an hour's volume,

the maximum consistency occurs between 9 a.m. and 7 p.m.

(ii) For transition from the day's to the week's volume, any day of

the week except Sunday is reasonably consistent, with Monday

showing the minimum variation.

(iii) For an estimate of annual volume, expansion based on counts

in April and September give the best results.

It can be seen that the most reliable prediction of the annual traffic

based on an hour's count is made if the count is taken between 9 a.m.

and 7 p.m. on a Monday in April or September.

Thanks everyone for the answers and suggestions.

Hi Rusdi how u use AADT in accident pridiction....As you take average of 24hrs or a segment while linking it with accident spot AADT?

if you have 7 day flow for 24 hrs for each segment then, how u will correlate it? as an AADT with accident spot variables like severity index ?

Hi Syyed,

My problem solved after read a book shared by Lenin in the previous comment. In this book, they suggest to calculate Hourly Expansion Factors (HEF) for the selected road using 24-hours data from existing census (in my case, Road Traffic Volume Malaysia - RTVM). In RTVM, census was conducted in two seasons; (Jan - June) and (July - December). Using these seasonal HEFs, the two-hurs vehicle counts of each segment were converted to ADT.

I used ADT and segment length to represent exposure variables in my crash count model.

Regards,

Rusdi

Hourly counts can be expanded into ADT using hourly expansion factors. However, you still need to apply the daily and monthly factors to adjust for seasonality. Good Luck Engr Rusdi

I have 15 minutes traffic volume mode wise at different location of a 175 Km long highway stretch. I want to calculate ADT and AADT with this volume. Is it right, to multiply 25 minute volume by a factor of 4 to get, hourly volume. And then from this hourly volume using hourly expansion factors, I will get ADT. Subhi M Bazlamit Matti Pursula , Request you to help. Thanks in advance!