Peak Hour Factor and Demand Flow Rate

Design for the busiest fifteen minutes, counted in passenger-car equivalents.

Before this Lesson

Between 7:00 and 8:00 a.m., a counter on a two-lane freeway direction logs 3,000 vehicles, or 1,500 per lane. That number hides two things. The vehicles didn't come evenly: the counter's 15-minute tallies read 650, 800, 900, and 650, so for a quarter of an hour they were arriving at a rate of 3,600 an hour. And one vehicle in ten was a truck, which the count scores the same as a hatchback. Put both back in, and at the peak each lane is carrying the equivalent of 1,980 cars an hour.

The busiest fifteen minutes

A freeway lane can pass only so many vehicles a minute. If they arrive faster than that, even for a few minutes, a queue forms, and it doesn't clear the moment arrivals ease off, because the lane is still working through the cars stacked up behind. The freeway never sees the hour. It sees the next car, and the one behind that. So an hour whose total looks comfortable can still break down in its busiest stretch, and traffic analysis is built around the busiest 15 minutes: long enough for a steady count, short enough to catch the surge.

The peak hour factor compares that stretch with the hour as a whole. The Handbook's Peak Hour Factor section prints it as

PHF=hourly volumehourly flow rate=V4V15\text{PHF} = \frac{\text{hourly volume}}{\text{hourly flow rate}} = \frac{V}{4V_{15}}

where VV is the hourly volume in veh/h and V15V_{15} is the volume in the peak 15 minutes. The 4 turns that 15-minute count into an hourly rate, the same move as multiplying a 5-minute count by 12 in Traffic Flow Fundamentals. For the morning in the Hook, V15=900V_{15} = 900:

PHF=3,0004(900)=3,0003,600=0.833\text{PHF} = \frac{3{,}000}{4(900)} = \frac{3{,}000}{3{,}600} = 0.833

Divide the top and bottom by 4 and the PHF reads another way: the average 15-minute count over the highest one, 750/900.

Vehicles per 15 min 650 800 V15 = 900 650 0 200 400 600 800 1,000 7:00–7:15 7:15–7:30 7:30–7:45 7:45–8:00 15-minute interval (a.m.) average = 750
Fifteen-minute counts for the 7:00 to 8:00 a.m. hour
The dashed line is the average count, 750. The PHF is the dashed line's height over the tallest bar's: 750/900 = 0.833.

Read the equation backward, and the PHF takes you from the hour to its peak. The peak 15-minute flow rate, which the Handbook calls the hourly flow rate, is

4V15=VPHF=3,0000.833≈3,600 veh/h4V_{15} = \frac{V}{\text{PHF}} = \frac{3{,}000}{0.833} \approx 3{,}600\ \text{veh/h}

and the peak 15-minute count is a quarter of that, V15=V/(4 PHF)=900V_{15} = V/(4\,\text{PHF}) = 900 vehicles. Dividing by the PHF marks the hourly volume up to the peak rate. Multiplying would mark it down, which is the wrong direction for a peak.

Checkpoint
1 of 3

Four consecutive 15-minute counts on a freeway direction are 1,050, 1,240, 1,310, and 1,100 vehicles. What is the peak hour factor?

A.
1.00
B.
0.90
C.
1.12
D.
0.28

Because the PHF is the average 15-minute count over the highest, its value alone tells you how the hour was shaped, even if you never see the counts.

Checkpoint

Which of these statements about the PHF are true?

A.
A PHF of 0.95 means a more sharply peaked hour than a PHF of 0.80.
B.
A PHF of 1.00 means all four 15-minute counts were equal.
C.
A PHF of 0.80 means the busiest 15 minutes ran at a rate 25% above the hourly volume.
D.
A PHF above 1.00 means traffic was unusually steady across the hour.
E.
A PHF of 0.25 means every vehicle in the hour came in a single 15-minute interval.

A truck counts as several cars

Now the second thing the count hides. A truck is several car lengths long, it gets up to speed slowly, and on an upgrade it slows while the cars around it don't, so the traffic behind it has to slow too or work its way around. So capacity analysis counts each truck as some number of passenger cars (pc), its passenger car equivalent ETE_T. The Handbook gives one value per terrain, under Basic Freeway Segment Highway Capacity: ET=2.0E_T = 2.0 on level terrain and 3.03.0 on rolling terrain. Hills cost a truck far more than they cost a car, so rolling terrain makes each truck worth more cars. The Handbook doesn't define the two terrains, so a question will say which one applies.

To convert a whole stream, take 100 vehicles, a proportion PTP_T of them trucks. The cars count once each, and the trucks count ETE_T times each:

100(1−PT)+100PTET=100[1+PT(ET−1)] pc100(1 - P_T) + 100P_TE_T = 100\left[1 + P_T(E_T - 1)\right]\ \text{pc}

Every truck is already one vehicle, so converting it adds ET−1E_T - 1 more. The bracket is the number of passenger cars per vehicle, and the Handbook packages it upside down as the heavy-vehicle adjustment factor:

fHV=11+PT(ET−1)f_{HV} = \frac{1}{1 + P_T(E_T - 1)}

PTP_T is a proportion, so 10% trucks is 0.10. For the morning count, on level terrain, the bracket is 1+0.10(2.0−1)=1.101 + 0.10(2.0 - 1) = 1.10 and fHV=0.909f_{HV} = 0.909: ten vehicles take up the road of eleven cars. Written as a factor below 1, it can sit in the same denominator as the PHF, and dividing by it multiplies by the bracket.

Actual stream: 10 vehicles Level terrain, ET = 2.0: 11 pc 1 truck = 2.0 pc Rolling terrain, ET = 3.0: 12 pc 1 truck = 3.0 pc
Ten vehicles, counted as passenger cars
Not to scale. The one truck in ten counts as 2.0 passenger cars on level terrain and 3.0 on rolling terrain.
Checkpoint

Traffic on a freeway in rolling terrain is 15% trucks. What is the heavy-vehicle adjustment factor?

A.
0.87
B.
0.69
C.
0.77
D.
0.03

Three corrections in one fraction

Put both corrections together and spread the result across the lanes, and you have the number a freeway is analyzed on. The Handbook's Basic Freeway Segment Highway Capacity section gives the demand flow rate:

vp=VPHF×N×fHVv_p = \frac{V}{\text{PHF} \times N \times f_{HV}}

Here VV is the hourly volume in veh/h, NN is the number of lanes in the analysis direction, and vpv_p is in passenger cars per hour per lane, pc/h/ln. Each factor in the denominator fixes one thing about the count. The PHF moves it from the hour to the busiest 15 minutes, NN from a whole direction to one lane, and fHVf_{HV} from vehicles to passenger cars. It's one fraction, so the order doesn't matter, but following the units one step at a time keeps each factor honest.

3,000 veh/h hourly volume, V 3,600 veh/h peak 15-min flow rate 1,800 veh/h/ln per lane 1,980 pc/h/ln demand flow rate, vp ÷ PHF = 0.833 to the busiest 15 min ÷ N = 2 to one lane ÷ fHV = 0.909 to passenger cars
The morning count, one correction at a time

The hourly count said 1,500 vehicles per lane. In its busiest stretch, each lane is really carrying the equivalent of 1,980 passenger cars an hour, 32% more.

Worked Example Demand flow rate on rolling terrain
1 of 2

A three-lane freeway direction carries 4,400 veh/h in its peak hour, with a PHF of 0.88 and 8% trucks on rolling terrain. Find the demand flow rate.

Two-way counts

Counts often cover both directions together, with a directional split: the share of the volume heading the busier way. The analysis is for one direction, so take that direction's share first, then divide by its lanes alone. A four-lane freeway has two lanes each way, so N=2N = 2.

Checkpoint

A six-lane freeway carries 5,600 veh/h in both directions combined during its peak hour, with 60% of it in the peak direction. The PHF is 0.90, and traffic is 8% trucks on level terrain. What is the peak-direction demand flow rate?

A.
896 pc/h/ln
B.
2,240 pc/h/ln
C.
1,344 pc/h/ln
D.
672 pc/h/ln

Judging a change

Every adjustment sits in the denominator, so a factor that shrinks makes vpv_p grow, and one that grows makes it shrink. To judge a change, trace it to the factor it moves.

Checkpoint

A freeway direction's hourly volume stays the same. Which of these statements are true?

A.
Raising the truck share from 5% to 15% raises the demand flow rate.
B.
Moving the same traffic from level to rolling terrain lowers the demand flow rate.
C.
A PHF of 0.85 instead of 0.95 raises the demand flow rate.
D.
Because fHVf_{HV} is below 1, dividing by it lowers the flow.
E.
Adding a third lane to a two-lane direction cuts the demand flow rate by a third.

The demand flow rate is where Freeway Level of Service picks up: divide vpv_p by the speed, and you have the density a freeway is graded on.

Recap

  • Design for the busiest 15 minutes. PHF=V/(4V15)\text{PHF} = V/(4V_{15}), the average 15-minute count over the highest, and the peak flow rate is V/PHFV/\text{PHF}.
  • The PHF runs from 0.25, with the whole hour in one interval, to 1.00, with a perfectly even hour. The lower it is, the sharper the peak.
  • Each truck counts as ETE_T passenger cars, 2.0 on level terrain and 3.0 on rolling, and fHV=1/[1+PT(ET−1)]f_{HV} = 1/[1 + P_T(E_T - 1)] with PTP_T as a decimal.
  • vp=V/(PHF×N×fHV)v_p = V/(\text{PHF} \times N \times f_{HV}) in pc/h/ln, with NN the lanes in one direction and VV that direction's share of a two-way count.
  • More trucks, rolling terrain, or a lower PHF raise the demand flow rate, and an added lane lowers it.