PE Exam (Civil) Traffic Flow Theory: Q = U·k, Time-mean vs Space-mean Speed
Last updated: May 2, 2026
Traffic Flow Theory: Q = U·k, Time-mean vs Space-mean Speed questions are one of the highest-leverage areas to study for the PE Exam (Civil). This guide breaks down the rule, the elements you need to recognize, the named traps that catch most students, and a memory aid that scales to test day. Read it once, then practice the same sub-topic adaptively in the app.
The rule
The fundamental traffic flow identity $q = u_s \cdot k$ links flow rate $q$ (veh/hr), space-mean speed $u_s$ (mi/hr), and density $k$ (veh/mi). The identity is exact only when $u_s$ is the space-mean speed (harmonic mean of individual vehicle speeds), NOT the time-mean speed (arithmetic mean). HCM Chapter 3 and the NCEES Reference Handbook §Transportation define $u_s = n / \sum (1/u_i)$ for individual speeds and $u_t = (1/n)\sum u_i$, with the relation $u_t \ge u_s$ always.
Elements breakdown
Flow ($q$)
Vehicles passing a fixed point per unit time.
- Counted at a point over time
- Units veh/hr or veh/sec
- Equal to $1/h$ where $h$ is mean headway
- Time-based observation
Common examples:
- Loop detector counts
- Pneumatic tube counts
Density ($k$)
Vehicles occupying a unit length of roadway at an instant.
- Counted along space at a snapshot
- Units veh/mi or veh/km
- Equal to $1/s$ where $s$ is mean spacing
- Space-based observation
Common examples:
- Aerial photo count
- Drone snapshot of segment
Space-mean speed ($u_s$)
Harmonic mean of individual vehicle speeds over a length of road.
- $u_s = \dfrac{n}{\sum_{i=1}^{n}(1/u_i)}$
- Equivalent to total distance / total travel time
- Always $\le u_t$
- Use this in $q = u_s k$
Common examples:
- Five vehicles traverse a 1-mi segment, average their travel times, divide 1 mi by the mean travel time
Time-mean speed ($u_t$)
Arithmetic mean of speeds measured at a fixed point.
- $u_t = \dfrac{1}{n}\sum u_i$
- Biased toward fast vehicles
- Always $\ge u_s$
- Do NOT plug into $q = u k$
Common examples:
- Radar gun samples at a single station
Headway and spacing
Per-vehicle time and distance gaps.
- Time headway $h_i$: front-to-front time
- Space headway $s_i$: front-to-front distance
- $\bar{h} = 1/q$ and $\bar{s} = 1/k$
- $u_s = \bar{s}/\bar{h}$
Wardrop relation
Mathematical link between $u_t$ and $u_s$.
- $u_t = u_s + \dfrac{\sigma_s^2}{u_s}$
- $\sigma_s^2$ is variance of space-mean speeds
- Equality only when all vehicles travel at same speed
- Use to convert when only $u_t$ is reported
Common patterns and traps
Arithmetic-Mean Trap
The problem hands you several spot speeds and asks for flow given a density. The tempting move is to average the speeds with $u_t = (1/n)\sum u_i$ and plug into $q = u k$. The correct move is the harmonic mean $u_s = n/\sum(1/u_i)$. The arithmetic answer is always larger than the harmonic answer, so the trap distractor sits above the correct value.
Two values close together, one slightly larger (time-mean × density), and the correct slightly smaller (space-mean × density).
Headway-vs-Spacing Swap
Mean headway $\bar{h} = 1/q$ has units of time per vehicle; mean spacing $\bar{s} = 1/k$ has units of length per vehicle. Candidates often invert by computing $1/\bar{h}$ and calling it density, or $1/\bar{s}$ and calling it flow. Tracking units (sec/veh vs ft/veh) catches the swap.
A choice with the correct number but units of veh/mi swapped with veh/hr.
Density-Flow Confusion
Density $k$ is an instantaneous-in-space count (veh/mi), while flow $q$ is an over-time count (veh/hr). A problem that gives a one-mile snapshot count tempts candidates to treat the count as flow, multiplying by speed unnecessarily. Read "how many vehicles are on the segment" as density; "how many cross the line per hour" as flow.
A choice that equals the raw count without speed multiplication, masquerading as flow.
Wardrop-Conversion Skip
When only $u_t$ and $\sigma$ are given, you must convert: $u_s = u_t - \sigma^2/u_s$, solved iteratively, or rearranged to $u_s^2 - u_t u_s + \sigma^2 = 0$. Skipping the conversion and using $u_t$ directly in the fundamental relation produces a small-but-systematic over-estimate.
A distractor that uses $u_t$ directly; the correct answer is $\sim 2$–$8\%$ lower.
Unit-Mixing Failure
Mixing seconds with hours, feet with miles, or veh/km with veh/mi causes order-of-magnitude errors. A clean unit-cancellation check on every multiplication is mandatory: $\text{(veh/mi)} \times \text{(mi/hr)} = \text{veh/hr}$.
A distractor off by a factor of 60, 3600, or 5280.
How it works
Suppose three vehicles cross a $L = 1 \text{ mi}$ segment at $u_1 = 60 \text{ mph}$, $u_2 = 50 \text{ mph}$, and $u_3 = 30 \text{ mph}$. The arithmetic (time-mean) average is $u_t = (60+50+30)/3 = 46.7 \text{ mph}$, but the harmonic (space-mean) average is $u_s = 3 / (1/60 + 1/50 + 1/30) = 3/0.0700 = 42.9 \text{ mph}$. If a downstream observation reports $k = 35 \text{ veh/mi}$, the correct flow is $q = u_s k = 42.9 \times 35 = 1{,}500 \text{ veh/hr}$, NOT $46.7 \times 35 = 1{,}635 \text{ veh/hr}$. Plugging in $u_t$ overstates flow by ~9% here; the gap widens as speed variance grows. On exam day, when the problem gives spot speeds at a station, you must convert to space-mean speed (harmonic mean, or via Wardrop) before applying the fundamental relation.
Worked examples
On a level rural segment of the proposed Reyes Valley Highway, a study team records spot speeds for five passenger cars at a single radar station. The observed speeds are $u_1 = 68 \text{ mph}$, $u_2 = 62 \text{ mph}$, $u_3 = 55 \text{ mph}$, $u_4 = 48 \text{ mph}$, and $u_5 = 42 \text{ mph}$. A simultaneous aerial photo shows a density of $k = 28 \text{ veh/mi}$ on the same segment. The PM tells you to report flow using the fundamental traffic-flow relation. The reference notes from your firm cite NCEES Reference Handbook §Transportation and HCM Chapter 3.
Most nearly, what is the flow rate $q$ on the segment?
- A $1{,}480 \text{ veh/hr}$ ✓ Correct
- B $1{,}540 \text{ veh/hr}$
- C $1{,}610 \text{ veh/hr}$
- D $1{,}680 \text{ veh/hr}$
Why A is correct: You must use the space-mean (harmonic) speed: $u_s = 5 / (1/68 + 1/62 + 1/55 + 1/48 + 1/42) = 5 / 0.0945 = 52.9 \text{ mph}$. Then $q = u_s k = 52.9 \text{ mph} \times 28 \text{ veh/mi} = 1{,}481 \text{ veh/hr}$. Units cancel: $\text{(mi/hr)} \times \text{(veh/mi)} = \text{veh/hr}$. Choice A matches.
Why each wrong choice fails:
- B: Used the time-mean (arithmetic) speed $u_t = (68+62+55+48+42)/5 = 55.0 \text{ mph}$ in $q = u k$, giving $55.0 \times 28 = 1{,}540 \text{ veh/hr}$. The fundamental relation requires the harmonic mean. (Arithmetic-Mean Trap)
- C: Used the median speed of $57.5 \text{ mph}$ as a shortcut, giving $57.5 \times 28 = 1{,}610 \text{ veh/hr}$. The median has no role in $q = u k$; only $u_s$ does. (Arithmetic-Mean Trap)
- D: Took the speed of the fastest vehicle $u_1 = 60 \text{ mph}$ as a representative speed, giving $60 \times 28 = 1{,}680 \text{ veh/hr}$. Free-flow or maximum speeds are not equivalent to $u_s$.
A municipal traffic engineer is calibrating a microscopic simulation of the Liu Boulevard arterial. From a 30-minute floating-car run, the team logs that the mean time headway between consecutive vehicles in the through lane is $\bar{h} = 2.4 \text{ sec/veh}$, and the mean spacing measured by drone snapshots is $\bar{s} = 165 \text{ ft/veh}$. The team needs to populate the simulation with consistent flow, density, and speed values for the through lane.
Most nearly, what space-mean speed should the team enter for this through lane?
- A $32 \text{ mph}$
- B $47 \text{ mph}$ ✓ Correct
- C $55 \text{ mph}$
- D $69 \text{ mph}$
Why B is correct: Flow $q = 1/\bar{h} = 1/2.4 = 0.4167 \text{ veh/sec} = 1{,}500 \text{ veh/hr}$. Density $k = 1/\bar{s} = 1/165 = 0.00606 \text{ veh/ft} \times 5280 \text{ ft/mi} = 32.0 \text{ veh/mi}$. From $q = u_s k$, $u_s = 1{,}500/32.0 = 46.9 \text{ mph}$. Units check: $\text{(veh/hr)}/\text{(veh/mi)} = \text{mi/hr}$. Choice B is closest.
Why each wrong choice fails:
- A: Computed $u_s = \bar{s}/\bar{h} = 165/2.4 = 68.75 \text{ ft/sec}$ but failed to convert ft/sec to mph (multiply by $0.682$, giving $46.9 \text{ mph}$); reading $68.75$ as a number close to $69$ leads to choice D, while clipping to $32$ comes from confusing density $32 \text{ veh/mi}$ with speed. (Density-Flow Confusion)
- C: Used flow $1{,}500 \text{ veh/hr}$ divided by an incorrect density of $27 \text{ veh/mi}$ (forgetting to multiply by $5280$ correctly), giving $\sim 55 \text{ mph}$. A unit-cancellation check on the density step would catch this. (Unit-Mixing Failure)
- D: Reported the raw $\bar{s}/\bar{h} = 68.75$ as if it were already in mph; it is actually ft/sec. Multiplying by $0.682 \text{ (mph per ft/sec)}$ converts properly to $46.9 \text{ mph}$. (Unit-Mixing Failure)
For the Carmichael County Travel Survey, a research crew posts at a single roadside station and reports the time-mean speed of free-flowing through traffic as $u_t = 58.0 \text{ mph}$ with a sample standard deviation of speeds $\sigma = 9.2 \text{ mph}$. Aerial reconnaissance during the same period gives a density of $k = 22 \text{ veh/mi}$. You are asked to apply Wardrop's relation $u_t = u_s + \sigma^2/u_s$ to obtain a flow estimate consistent with the fundamental relation $q = u_s k$.
Most nearly, what is the flow rate $q$ on the segment?
- A $1{,}210 \text{ veh/hr}$
- B $1{,}240 \text{ veh/hr}$ ✓ Correct
- C $1{,}280 \text{ veh/hr}$
- D $1{,}320 \text{ veh/hr}$
Why B is correct: Solve $u_s^2 - u_t u_s + \sigma^2 = 0$ for $u_s$: $u_s = \frac{u_t + \sqrt{u_t^2 - 4\sigma^2}}{2} = \frac{58.0 + \sqrt{3364 - 338.6}}{2} = \frac{58.0 + 55.0}{2} = 56.5 \text{ mph}$. Then $q = u_s k = 56.5 \times 22 = 1{,}243 \text{ veh/hr}$. Units: $\text{(mi/hr)} \times \text{(veh/mi)} = \text{veh/hr}$. Choice B matches.
Why each wrong choice fails:
- A: Approximated $u_s \approx u_t - \sigma^2/u_t = 58.0 - 84.64/58.0 = 56.54$, then rounded down aggressively to $55 \text{ mph}$, giving $55 \times 22 = 1{,}210 \text{ veh/hr}$. The full quadratic gives $u_s = 56.5$, not $55$. (Wardrop-Conversion Skip)
- C: Skipped the Wardrop conversion entirely and used $u_t$ directly: $58.0 \times 22 = 1{,}276 \text{ veh/hr}$. The fundamental relation needs $u_s$, which is always less than $u_t$ when $\sigma > 0$. (Wardrop-Conversion Skip)
- D: Added the variance correction in the wrong direction, computing $u_s \approx u_t + \sigma^2/u_t = 59.5 \text{ mph}$, then $59.5 \times 22 = 1{,}320 \text{ veh/hr}$. The Wardrop relation always gives $u_s < u_t$. (Arithmetic-Mean Trap)
Memory aid
"FLOW = SPACE × DENSITY." Flow is at-a-point, density is in-space, and the speed that bridges them must also be the in-space (harmonic) one. If radar gave you the speeds, harmonic-mean them first.
Key distinction
Time-mean speed is the arithmetic mean of speeds at a point; space-mean speed is the harmonic mean over a length. Only $u_s$ satisfies $q = u k$, and $u_t \ge u_s$ with equality only when speed variance is zero.
Summary
Use the harmonic-mean (space-mean) speed in $q = u_s k$; the arithmetic-mean (time-mean) speed always overstates flow.
Practice traffic flow theory: q = u·k, time-mean vs space-mean speed adaptively
Reading the rule is the start. Working PE Exam (Civil)-format questions on this sub-topic with adaptive selection, watching your mastery score climb in real time, and seeing the items you missed return on a spaced-repetition schedule — that's where score lift actually happens. Free for seven days. No credit card required.
Start your free 7-day trialFrequently asked questions
What is traffic flow theory: q = u·k, time-mean vs space-mean speed on the PE Exam (Civil)?
The fundamental traffic flow identity $q = u_s \cdot k$ links flow rate $q$ (veh/hr), space-mean speed $u_s$ (mi/hr), and density $k$ (veh/mi). The identity is exact only when $u_s$ is the space-mean speed (harmonic mean of individual vehicle speeds), NOT the time-mean speed (arithmetic mean). HCM Chapter 3 and the NCEES Reference Handbook §Transportation define $u_s = n / \sum (1/u_i)$ for individual speeds and $u_t = (1/n)\sum u_i$, with the relation $u_t \ge u_s$ always.
How do I practice traffic flow theory: q = u·k, time-mean vs space-mean speed questions?
The fastest way to improve on traffic flow theory: q = u·k, time-mean vs space-mean speed is targeted, adaptive practice — working questions that focus on your specific weak spots within this sub-topic, getting immediate feedback, and revisiting items you missed on a spaced-repetition schedule. Neureto's adaptive engine does this automatically across the PE Exam (Civil); start a free 7-day trial to see your sub-topic mastery climb in real time.
What's the most important distinction to remember for traffic flow theory: q = u·k, time-mean vs space-mean speed?
Time-mean speed is the arithmetic mean of speeds at a point; space-mean speed is the harmonic mean over a length. Only $u_s$ satisfies $q = u k$, and $u_t \ge u_s$ with equality only when speed variance is zero.
Is there a memory aid for traffic flow theory: q = u·k, time-mean vs space-mean speed questions?
"FLOW = SPACE × DENSITY." Flow is at-a-point, density is in-space, and the speed that bridges them must also be the in-space (harmonic) one. If radar gave you the speeds, harmonic-mean them first.
What's a common trap on traffic flow theory: q = u·k, time-mean vs space-mean speed questions?
Plugging time-mean speed into $q = u k$
What's a common trap on traffic flow theory: q = u·k, time-mean vs space-mean speed questions?
Confusing density (veh/mi) with flow (veh/hr)
Ready to drill these patterns?
Take a free PE Exam (Civil) assessment — about 35 minutes and Neureto will route more traffic flow theory: q = u·k, time-mean vs space-mean speed questions your way until your sub-topic mastery score reflects real improvement, not luck. Free for seven days. No credit card required.
Start your free 7-day trial