Design hour volume vs. AADT, and why the gap matters for capacity work

Every traffic engineer has had the conversation where a planner pulls up AADT for a corridor and asks why that number can't just go straight into a capacity analysis. It can't, and the reason comes down to what AADT measures versus what a lane-group or signal-timing study needs.
What AADT tells you
Average Annual Daily Traffic is a 365-day average of two-way volume at a point, collected (or modeled) across every season, every day of the week, holidays included. It's the right number for pavement design life, functional classification, and long-range travel demand forecasts, because those decisions care about cumulative loading over years, not what happens on a Tuesday at 5 p.m. in October.
The problem is that AADT smooths out exactly the thing a capacity study needs to see: the peak. A corridor with 18,000 AADT could have a tame, evenly distributed load, or it could dump half that volume into two commute peaks with a big directional split. AADT alone can't tell you which one you're looking at, and designing a lane configuration or signal plan around the daily average risks undersizing the link for the hour that breaks it.
The 30th highest hour, and how DHV gets derived
Design Hour Volume is built from a different question: if you ranked every hour of the year by volume, from busiest to quietest, which hour should you design for? AASHTO and the HCM converged on the 30th highest hour as the practical answer. Design for that hour and you're sized for all but 29 hours a year, a threshold that holds up on all but a handful of unusually heavy hours while still covering the recurring peak that shows up week after week.
That 30th highest hour volume gets expressed as a K-factor, the ratio of DHV to AADT. It's typically in the range of 8 to 12 percent for a lot of urban and suburban arterials, though recreational routes and corridors with heavy seasonal swings can run well outside that band. Multiply AADT by K, apply the directional split (the D-factor), and you've got a directional design hour volume you can feed into an HCM capacity analysis.
Why this matters for signal timing and lane sizing
A capacity study or signal-timing study is sizing something, a lane, a phase, a cycle length, for a specific operating condition, and the design hour is that condition. Use AADT instead and you're effectively designing for an average day that never shows up on the ground. Use the single peak hour of the year and you're over-building for a condition that's gone before the warranty paperwork clears. The 30th highest hour balances those two failure modes, which is why DHV is the number that goes into the Highway Capacity Manual worksheets, signal warrant analysis, and turn-lane justification memos.
The catch: getting a real DHV reading on a link with no counter
All of this assumes you have count data to rank hours from, or at least a count you can factor using known seasonal and K-factor adjustments. On instrumented corridors with continuous count stations, that's straightforward. On the links that generate the capacity questions, the new collector tying into an arterial, the intersection approach nobody's put a tube counter on in a decade, the segment a developer's traffic study needs turned around in two weeks, there's often nothing to factor from.
That's usually the point where someone has to decide between scheduling a tube count crew, closing a lane to set a radar unit, or just estimating. If what you need is a defensible count for a specific link and direction during a known peak window, a single scheduled pass over the corridor can get you that tally without the field logistics, delivered as counts by link and direction plus the image it came from, something you can drop straight into the worksheet next to your K-factor math.
If the link in your capacity study still doesn't have a counter on it, that's worth a look before the next field season gets scheduled.