Design Speed & Speed Limits

#What design speed is

Design speed is a selected speed used to determine a roadway's geometric design features: curve radii, sight distance, lane width, and similar elements.¹ It is an input to the drawing board, not a number posted on a sign. A road's actual geometry, its curves, its lane widths, its sightlines, can support a much higher speed than the design speed on the plans, because geometric standards are written with a built-in safety margin for near-worst-case conditions.³ Engineers call the highest speed a road's geometry can actually support its inferred design speed, and it is frequently higher than the design speed the road was drawn to.⁵

#What a posted speed limit is

A posted speed limit is the legal maximum speed for a section of road, set by a state or local transportation authority. In conventional US practice, posted limits are commonly derived from the 85th-percentile operating speed: the speed at or below which 85 percent of drivers travel in free-flow conditions.² This method sets the limit to match how people already drive, not to match what the road's geometry was designed for.

#Why these two numbers diverge

Design speed, operating speed, and posted speed limit are three different measures, set by three different parties, for three different reasons: designers set design speed to establish geometric features, operators set posted limits they judge safe for the road, and drivers pick their own speed based on how safe the road feels to them. These three numbers are frequently incompatible with each other.⁴

The mismatch has a specific origin. A 1985 FHWA memorandum establishing the controlling criteria for federal-aid design stated that design speed should be selected to equal or exceed the posted speed limit. In 2001, AASHTO's Green Book dropped that link. The Green Book now defines design speed purely as a geometric design input and gives no guidance connecting it to posted speed limits.⁵ Since then, US roads are routinely built with generous, above-minimum geometry, then posted below what that geometry actually invites. The result: streets whose curves, lane widths, and sightlines all say "drive faster" while the sign says otherwise. Drivers overwhelmingly follow the road, not the sign. Documented cases show 85th-percentile operating speeds and posted speed limits both exceeding the road's own designated design speed, on roads whose geometry was never checked for consistency against the number on the plans.⁶

#Why it matters

A posted number does not change how a driver feels behind the wheel. Geometry does. A driver's comfortable, "natural" speed is set by what the road communicates: lane width, curvature, sightlines, and roadside friction, regardless of what the sign says.⁷ Design speed decoupled from posted speed limit is design speed decoupled from what actually happens on the road. A pattern book that treats posted speed as the safety mechanism, while leaving geometry free to invite something higher, is solving the wrong problem.

#The Commonway System's rule

Design speed always equals posted speed limit.

This is CS's foundational speed rule, and every Local, Collector, and Arterial pattern in this book is built to it. It follows FHWA's own self-enforcing roadway framework: the roadway's physical design should make the target speed the comfortable, natural driving speed, not just the legal one.⁷ A CS street is not "20 mph" or "30 mph" because a sign says so. It is that speed because its geometry makes any faster speed feel wrong to drive.

This is not a target. It is not aspirational. Where a CS pattern specifies a speed, that speed is the design speed, the posted speed, and the intended operating speed, all at once, by design.

#Applying the rule, by street type

#Local: 20 mph hard maximum, no exceptions

CS holds Local streets to 20 mph with no exceptions clause, a stricter standard than the Stockholm Declaration's own 30 km/h (≈20 mph) carve-out for areas where motor vehicles and vulnerable road users mix.⁸ The number is not arbitrary. Pedestrian severe-injury risk from a vehicle impact reaches 10 percent at 17.1 mph and death risk reaches 10 percent at 24.1 mph.⁹ Pedestrian survival exceeds 90 percent in collisions at or below 20 mph, and falls below 50 percent at 30 mph and above.¹⁰ Local street geometry must make 20 mph the comfortable driving speed. No Local pattern in this book is designed to a higher inferred speed than its posted number.

#Collector: 30 mph standard, 35 mph only with rigid separation

CS holds Collectors to 30 mph as the standard maximum. 35 mph is permitted only where bicycle facilities carry rigid, crash-protective separation, a concrete barrier or permanent curb, not flexible delineator posts. This distinction is FHWA's own: flexible delineators and rigid barriers are not treated as equivalent separation.¹¹ The 30 mph threshold matches Dutch distributor-road practice, where roads at this speed class require physical separation between cars and bicycles.¹² It is also a real safety cliff: bicyclist fatality risk in a collision roughly doubles at 30 mph relative to lower speeds.¹³ Seattle's own Streets Illustrated design guidance independently arrives at the same 30 mph threshold.¹⁴

#Arterial: density-tiered baseline, with two non-stacking modifiers

No existing framework prescribes Arterial design speed the way CS needs it prescribed, so this is original CS synthesis, built from FHWA's density principle, NACTO's speed guidance, Dutch noise policy, and VDOT's access-spacing logic. Every piece is cited below.

Baseline, set by density tier. FHWA is explicit that arterial design speed should fall as land-use density rises: an urban collector or arterial passing through a dense area should carry an appreciably lower design speed than the same functional class in a rural setting.¹⁵ NACTO's own Design Speed guidance goes further: design criteria should sit at or below a street's target speed, and higher speeds are inappropriate on urban streets, including urban arterials, with higher speeds reserved for limited-access freeways and highways.¹⁶ CS's six density tiers (see Density Tiers) give this principle a structured baseline:

Core's baseline meets Collector's ceiling by design. At CS's highest density tier, an Arterial's surrounding context is closer to a Collector's than to a rural Arterial's, even though it still carries a through-traffic function.

Noise modifier: −5 mph. Where an Arterial segment runs adjacent to noise-sensitive land uses (residential, schools, healthcare), subtract 5 mph from the baseline. This has direct precedent on both sides of the Atlantic. The Netherlands' Noise Pollution Act sets statutory road-noise limits and mandates abatement where they're exceeded,²¹ and the N325 Pleyroute corridor had its speed limit set specifically for noise and environmental reasons under that framework.²² In the US, speed reduction is a federally recognized highway noise abatement measure in its own right: FHWA's traffic noise regulation, 23 CFR Part 772, which governs every state DOT's noise policy, lists reducing speed limits among the traffic management measures available to control highway noise, noting that roughly a 20 mph reduction produces a perceptible (5 dBA) decrease.²³ AASHTO's own environmental guidance repeats the same finding.²⁴ CS's 5 mph step is smaller than FHWA's ~20 mph figure for a perceptible reduction because it is one of two contributing modifiers in a cumulative framework, not a standalone noise-abatement measure sized to solve noise on its own.

"Noise-sensitive adjacency" is a distance, defined the way the Dutch define it, not the way FHWA defines it. Article 74 of the Dutch Noise Pollution Act gives every road a statutory noise zone, measured from the inside edge of its outermost travel lane on both sides, sized by lane count and by whether the road sits in an urban or rural area:²⁶

A noise-sensitive land use inside this zone qualifies for the modifier; one outside it doesn't. Roads posted at 30 km/h and woonerven carry no zone at all under Dutch law, since they're already slow enough that noise isn't the binding constraint,²⁶ the same logic behind CS restricting this modifier to Arterials only. CS adopts this table directly, converted to feet and keyed to the Arterial segment's own lane count.

FHWA's approach is not a distance rule at all. Under 23 CFR 772, highway agencies must model predicted noise levels in decibels at specific receptor points using the Traffic Noise Model, and abatement is only triggered once a receptor's predicted level approaches or exceeds the federal Noise Abatement Criteria.²³ That's a case-by-case standard requiring an acoustic model for each project, not a bright-line rule a pattern-book user can check with a tape measure.

CS deliberately takes the Dutch rule over the FHWA standard here, and the choice has real backing beyond convenience. The foundational law-and-economics treatment of this exact tradeoff concludes that standards are more costly for individuals to interpret when deciding how to act and for an adjudicator to apply after the fact, while rules make outcomes predictable without requiring case-by-case precedent or specialized interpretation each time.²⁷ The same conclusion holds specifically for noise regulation: quantitative, rule-based noise ordinances are capable of providing non-discretionary, objective, and predictable standards, while qualitative, discretionary approaches are more prone to inconsistent enforcement and vagueness.²⁸ Enforcement research reaches the same conclusion in general terms: rules are cheaper to enforce than standards precisely because they don't require case-specific information to apply.²⁹ A pattern book meant to be usable without commissioning an acoustic study for every Arterial segment is exactly the situation this literature describes. CS chooses the version of this rule that produces the same answer every time it's applied.

Access modifier: −5 mph. Where local streets tie into the Arterial more frequently than a defined spacing threshold, subtract 5 mph from the baseline. This inverts VDOT's own logic tying minimum intersection spacing to speed class: VDOT requires wider spacing at higher speeds because more conflict points at speed raise crash risk.²⁵ CS applies the same relationship in reverse: tighter local-street spacing is itself evidence that the speed should come down.

The two modifiers do not stack. If a segment qualifies for both, only the single larger reduction applies, not both added together. This is not an arbitrary simplification. It follows the same logic used across three independent fields that all study what happens when overlapping risk-reduction factors are combined:

• NACTO's City Limits methodology evaluates conflict density and activity level jointly, in a risk matrix, rather than summing them into a single additive number.¹⁷
• AASHTO's Green Book treats design speed as one of a small set of governing controlling criteria, with deviations handled through a documented design-exception process rather than compounding penalties across every applicable factor.¹⁸
• FHWA's own Highway Safety Manual guidance on combining multiple Crash Modification Factors cautions that unless countermeasures act completely independently, multiplying several together is likely to overestimate their combined effect, with the risk of overestimation growing as more factors are combined.¹⁹ The standard correction for exactly this situation is the dominant effect method: apply only the single most effective (largest) factor, the simplest and most conservative approach, used when the applicable factors overlap rather than act independently.²⁰

CS's noise and access modifiers are not independent, both are proxies for the same underlying thing: how much a road's context, not its through-traffic function, should be pulling its design speed down. Applying the dominant one, not both, follows the same conservative logic transportation safety practice already uses when two risk factors point the same direction.

Floor: 30 mph, with a documented-exception process. The baseline minus the applicable modifier never exceeds the Step 1 baseline, and the result is not permitted below 30 mph except through a documented exception, following the same default-plus-documented-exception architecture CS already uses for roundabout-first intersection control and the crash-investigation lifecycle.

#Freeway: governed by existing systems

CS does not set a speed rule for Freeways. An extensive federal and state regulatory system, spanning AASHTO's own Interstate design standards and each state DOT's speed-setting authority, already governs freeway speed comprehensively. CS respects that existing system rather than duplicating or second-guessing it.

#References

1. Legal / regulatory AASHTO. A Policy on Geometric Design of Highways and Streets, 7th ed. Washington, DC: American Association of State Highway and Transportation Officials, 2018.
2. Legal / regulatory Federal Highway Administration. "Relationship Between Design Speed, Operating Speed, and Posted Speed." Available at: https://www.fhwa.dot.gov/planning/css/resources/speed(opens in new tab)
3. Evidence-based Texas Transportation Institute / Federal Highway Administration. Compatibility of Design Speed and Posted Speed for Texas Highways. Report FHWA/TX-95/1465-1.
4. Legal / regulatory Toole, Joseph S. Foreword to Speed Concepts: Informational Guide. FHWA-SA-10-001. Washington, DC: Federal Highway Administration, 2009.
5. Legal / regulatory Federal Highway Administration. Memorandum from E. Dean Carlson establishing the thirteen controlling criteria for design, April 15, 1985, and subsequent guidance on the 2001 AASHTO Green Book redefinition of design speed. Available at: https://www.fhwa.dot.gov/design/standards/151007.cfm(opens in new tab)
6. Legal / regulatory Federal Highway Administration. "Chapter 6: Government Roles Related to Traffic Speeds." Speed Concepts: Informational Guide. Available at: https://highways.dot.gov/safety/speed-management/speed-concepts-informational-guide/chapter-6-government-roles-related(opens in new tab)
7. Legal / regulatory Federal Highway Administration. Self-Enforcing Roadways: A Guidance Report. FHWA-HRT-17-098, 2018.
8. Precedent-based Third Global Ministerial Conference on Road Safety. Stockholm Declaration. Stockholm, 2020.
9. Evidence-based Tefft, Brian C. Impact Speed and a Pedestrian's Risk of Severe Injury or Death. Washington, DC: AAA Foundation for Traffic Safety, 2011.
10. Evidence-based Peden, Margie, et al. World Report on Road Traffic Injury Prevention. Geneva: World Health Organization, 2004, as cited by the Federal Highway Administration.
11. Legal / regulatory Federal Highway Administration. Separated Bike Lane Planning and Design Guide.
12. Precedent-based CROW. Dutch Sustainable Safety design guidance on distributor-road separation requirements.
13. Evidence-based Federal Highway Administration / National Transportation Safety Board. Bicyclist fatality risk data by impact speed.
14. Precedent-based Seattle Department of Transportation. Streets Illustrated.
15. Legal / regulatory Federal Highway Administration. "Design Speed." Available at: https://www.fhwa.dot.gov/planning/css/resources/dspeed(opens in new tab)
16. Precedent-based National Association of City Transportation Officials. "Design Speed." Urban Street Design Guide. Available at: https://nacto.org/publication/urban-street-design-guide/design-controls/design-speed/(opens in new tab)
17. Precedent-based National Association of City Transportation Officials. City Limits: Setting Safe Speed Limits on City Streets, 2020.
18. Legal / regulatory AASHTO. A Policy on Geometric Design of Highways and Streets, 7th ed., controlling criteria and design exception guidance.
19. Evidence-based Federal Highway Administration, CMF Clearinghouse. Combining Multiple CMFs. Available at: https://www.cmfclearinghouse.fhwa.dot.gov(opens in new tab)
20. Evidence-based Texas Department of Transportation. "5.4 Crash Modification Factors (CMFs)," dominant effect method. Roadway Design Manual.
21. Legal / regulatory Netherlands. Wet geluidhinder [Noise Pollution Act].
22. Precedent-based N325 Pleyroute corridor speed-setting precedent, as documented in Dutch noise policy literature.
23. Legal / regulatory Federal Highway Administration. 23 CFR Part 772, Procedures for Abatement of Highway Traffic Noise and Construction Noise.
24. Precedent-based AASHTO Center for Environmental Excellence. "Noise Overview." Available at: https://environment.transportation.org/focus-areas/noise/noise-overview/(opens in new tab)
25. Precedent-based Virginia Department of Transportation. Intersection and access spacing standards by design speed class.
26. Legal / regulatory Netherlands. Wet geluidhinder [Noise Pollution Act], Article 74 (statutory noise zone widths by lane count and urban/rural context).
27. Evidence-based Kaplow, Louis. "Rules Versus Standards: An Economic Analysis." Duke Law Journal 42, no. 3 (1992): 557-629.
28. Evidence-based Malone, Linda A. Environmental Regulation of Land Use, §11.06[2], as summarized in "Guidelines for Drafting Municipal Noise Control Ordinances," FindLaw.
29. Evidence-based "Enforcement and Learning Under Rules and Standards." Cambridge University Press, 2026.

Suggest changes to this page (opens in new tab) · Last updated September 2, 2026 · Share: