Reference
From a seven-day traffic count to AADT
A seven-day count gives the average day of one week. AADT is the average day of a year, so the week has to be corrected for the season it was counted in. Below is the method as the current Indian guideline sets it out, with its own worked example, followed by the three different things IRC codes mean by “peak hour factor”.
IRC:108-2015, B.1.2 and C.1
AADT = ADT × seasonal factor
The documents
Which editions apply, and which you can read
MoRTH’s standard RFP for DPR consultancy, revised 29 September 2025, states the requirement: “The traffic volume count per day shall be averaged to show a weekly average daily traffic (ADT) by vehicle type. The annual average daily traffic (AADT) shall be worked out by applying seasonal factors.” It does not say where the factors come from; its report checklist asks for “Traffic projection methodology as per IRC:108”.
IRC:108-2015 defines ADT as traffic over “a given time period (more than a day and less than a year)” divided by its days, and AADT as traffic over “one full year” divided by the days in the year.
- IRC:108, traffic forecast on highways2015
- Definitions, ADT from seven days, sources of seasonal factors and a worked example. Public.
- IRC:SP:19, survey and preparation of road projects2020
- The 2001 text converts ADT “using conversion factors from any continuous traffic count in the region”. Only 2001 is public.
- IRC:9, traffic census on non-urban roads2022
- The 1972 text: seven days × 24 hours, twice a year, ending at the week’s average daily traffic. Only 1972 is public.
- IRC:37, flexible pavement design2018
- Present traffic from “the seven-day 24-hour traffic volume count”; no mention of AADT. Public.
IRC:9-1972 contains no seasonal factor, and the 2022 revision is not public, so we cannot say what it adds. A free file labelled “IRC:9-2022” online is in fact a different code, IRC:131-2022 on blackspots.
Step 1
Average daily traffic from seven days
IRC:108-2015 asks for classified counts “in both directions for successive 15 minute periods”, run for one week “to account for the daily variation of demand which is predominant on such roads”. MoRTH asks for seven days, “continuous, direction-wise”. Seven consecutive days already cover every weekday, so no day-of-week factor is needed.
The guideline’s example: a control station counted 21,000; 23,000; 23,500; 24,000; 20,000; 18,000; 16,000 PCU, Monday to Sunday. ADT = 1,45,500 ÷ 7 = 20,786 PCU/day.
The count will be checked. IRC:108-2015 notes that a manual count cannot be validated after the fact, so “it is desirable to record vehicular movements using videography and extract data using manual/semi-automatic methods”. MoRTH’s RFP penalises traffic data “found to be varying by more than 25% on resurvey at a later date, unless there are justifiable reasons” at 10% of the performance bank guarantee.
Record the count in the 15-minute format the rest of the method expects, with PCU per IRC:64-1990, which IRC:108-2015 names for expressing ADT.
Step 2
The seasonal factor, and which way round it goes
Seasonal factors “consider variation of travel demand during different months or seasons in a year”, and are generally estimated from permanent count stations. Without one, IRC:108-2015 accepts “monthly fuel sale data for at least three years, authentic vehicle category wise traffic data from any nearby toll plaza, or data from any permanent traffic count station in the area”.
Split the year into three or four seasons suited to the location, and derive separate factors “for different vehicle categories, say, passenger and goods vehicles”. The count cannot supply its own factor: the factor describes a year, and a week shows only itself.
Check the convention before you apply a factor table. IRC:108-2015 multiplies: the factor is the year’s average seasonal index over the index of the months counted. The US Traffic Monitoring Guide’s monthly factor is also applied as a multiplier, but it notes that “many States use the inverse … by dividing”.
Getting the direction wrong in the example below gives 1,482 ÷ 1.07 = 1,385 instead of 1,586: 201 vehicles a day, or 12.7%.
Worked example
IRC:108-2015’s example, step by step
Two-axle heavy commercial vehicles, ADT 1,482, counted in June–July. The guideline calls its values “purely hypothetical”: this is the method, not survey data.
| Year | WinterNov–Feb | SummerMar–Jun | MonsoonJul–Oct |
|---|---|---|---|
| 2010 | 1,389 | 1,325 | 1,480 |
| 2011 | 1,600 | 1,430 | 951 |
| 2012 | 1,703 | 1,362 | 1,182 |
| Average | 1,564 | 1,372 | 1,204 |
- Seasonal index, winter = 100. Summer: 100 × 1,372 ÷ 1,564 = 87.72. Monsoon: 100 × 1,204 ÷ 1,564 = 76.98.
- The year’s average index. Each season is 4 months, so (100.00 × 4 + 87.72 × 4 + 76.98 × 4) ÷ 12 = 88.23.
- The index of the count months. June is summer and July is monsoon: (87.72 + 76.98) ÷ 2 = 82.35.
- The seasonal factor. 88.23 ÷ 82.35 = 1.07.
- AADT. 1,482 × 1.07 = 1,586.
A factor can be below 1. On the same index, a count taken wholly in winter gets 88.23 ÷ 100 = 0.88, so a busy-season week is scaled down; one taken wholly in the monsoon gets 1.15.
Peak hour factor
Three IRC codes, three meanings
Ask which one you are being given. Only the first can be calculated from a week of 15-minute counts.
- IRC:108-2015cl. A.2
- The hourly volume during the analysis hour divided by four times the peak 15-minute flow in that hour
- Needs: 15-minute counts
- IRC:64-1990cl. 3.9
- Peak-hour volume as a percentage of AADT, the peak hour being the thirtieth highest hour of the year
- Needs: A year of hourly data
- IRC:106-1990cl. 4.9
- Peak-hour volume as a percentage of AADT, the peak hour being the highest hourly volume in actual counts
- Needs: Counts and an AADT
IRC:108-2015’s version is PHF = Vhour ÷ (4 × Vpeak 15 min). It runs from 0.25, when the hour’s traffic all arrives in one quarter-hour, to 1.00, when the four quarter-hours are equal. The guideline says its use “is remote in case of highway traffic analysis, unless intersections are considered”.
What it does ask you to report with the ADT is the peak hour and the peak hour proportion: the highest hour as a percentage of the 24-hour volume.
Design hourly volume
The design hour comes from a year, not a week
The design hourly volume is “normally taken as 30th highest hourly volume, which means that only for 29 hours in a year the traffic volume can exceed the design volume”. IRC:64-1990 puts it around 8–10% of AADT on trunk routes, and bases its capacity figures on that. A seven-day count observes 168 hours; it cannot tell you which hour of the year is the thirtieth highest.
Rural roads are designed on the daily figure: IRC:64-1990 says it is “usual to adopt daily traffic volumes for design instead of hourly volumes”, and tables its design service volumes in PCU per day. That is where the AADT, in PCU, is used.
Urban divided carriageways are different. IRC:106-1990 designs them on “peak hour traffic in the busier direction of travel”, and counts 16 hours to find the peaks.
Where Playcraft fits
What we ship, and what we build on demand
Playcraft ATCC counts traffic from recorded video and reports the 19-class schedule by direction, in 15-minute intervals, with PCU per IRC:64-1990, a per-vehicle log and an annotated verification video. That count is Step 1’s input, and its 15-minute rows are what the IRC:108-2015 peak hour factor and peak hour proportion are calculated from.
AADT with seasonal correction is not part of the shipped product. We build it on demand for your counts.
What we cannot supply is the seasonal factor itself. It has to come from a year-round source for your corridor — fuel sales, toll-plaza data by category or a permanent count station — and seven days of video do not contain one.
Sources
- IRC:108-2015, Guidelines for Traffic Forecast on Highways (First Revision)Indian Roads Congress · A.2, B.1.1, B.1.2, C.1, C.5 · January 2015
- Revised RFP for consultancy services for preparation of Detailed Project ReportsMoRTH · Circular 130.247, Terms of Reference 4.9.2 · 29 September 2025
- IRC:64-1990, Guidelines for Capacity of Roads in Rural AreasIndian Roads Congress · cl. 3.5, 3.9, 6.3; Tables 2–4 · 1990
- IRC:106-1990, Guidelines for Capacity of Urban Roads in Plain AreasIndian Roads Congress · cl. 4.9, 6.1, 6.2 · 1990
- IRC:37-2018, Guidelines for the Design of Flexible PavementsIndian Roads Congress · cl. 4.1.3 · 2018
- IRC:SP:19-2001, Manual for Survey, Investigation and Preparation of Road ProjectsIndian Roads Congress · cl. 6.2; superseded by the 2020 edition, not public · 2001
- IRC:9-1972, Traffic Census on Non-Urban RoadsIndian Roads Congress · cl. 4.1, 6.2; superseded by IRC:9-2022, not public · 1972
- Traffic Monitoring Guide, chapters 3 and 5US Federal Highway Administration · US practice, cited for conventions only · 2022
Next step
See it run on a call
- The tool running on footage you choose, or ours
- Every output file opened and explained
- A plain answer on fit before commercials