A
Bicyclist Traversing a One and Two Lane Roundabout – Examples:
the 1-lane at St. Paul/Locust/Shelburne/S. Willard (Shelburne Street
Roundabout) as Designed; and the 2-lane at
Colchester/Barrett/Riverside (COLBARI Roundabout) Preliminary
Feasibility Design
How
bicycles traverse roundabouts safely and expeditiously has been
raised by Burlington and Chittenden County cyclists with the first
1-lane roundabout on a busy roadway is set for 2021 in Burlington and
discussion of the a 2-lane roundabout to benefit in part the
environment of the adjacent neighborhood at the Winooski Bridge
continues.
Since
the 1966 birth of modern roundabouts, design of 1-lane roundabouts
reached maturity in recent years as the 2000s emergence of bicycle
cycle track became incorporated at the design level. But 2-lane
designs have significantly evolved since the dawn of modern
roundabouts and for cyclists design concepts incorporating the
bicycle safety and cycle track continues to this day. From the start
roundabouts design assumed the cyclist as a vehicle, and in a sense
roundabouts were viewed and judged by the “vehicle cyclist”
population, a very small segment of the population, mostly young,
adult, white men, the same basic population utilizing the roadways in
urban America today along shared roadways and painted lanes and
shoulder spaces. In western Europe particularly a cross section of
the population survived the car invasion and with proper
infrastructure in place young and old of all skills and abilities
safely travel in urban areas today, a goal still far from being
realized anywhere in North America. Roundabouts are part and parcel
of the infrastructure now a given in order to reach safe “walkable
and bikable” downtowns, town centers and other built up areas.
First
a note of history. The short “bicycle age” extending from 1880s
to the first decade of the 20th century cycling really was
the new new mode reaching down from the upper classes of society to
become a democratic travel mode. Cycling groups helped obtain the
paved urban roadways bicyclists needed for quality travel surfaces.
But the advent of the automobile quickly drowned out cycling among a
cross section of the population, bicycling became almost extinct as
did roadway designs became the “wider faster smoother”
environment of the automobile, really still extent in some areas
today. The demarcation from bicycle to auto age is symbolized by
them passage of the first U.S. federal highway act in 1916. (A good
history of this age up through the cycle track emergence this decade
with emphasis on Burlington can be found in UVM Professor Luis
Vivanco's landmark book “Reconsidering the Bicycle.”)
While
Europeans never really ended wholly attention to walk and bike
modes--thankfully--the car except for sidewalks became king in
America until 1991 when change began. It was 1991 following public
referenda in Maine and California where voters overall rebelled
against vehicle mono-modalism, the “Intermodal Surface
Transportation Energy Act of 1991” referred to as ISTEA, passed
congress and was signed into law by President George H.W.
Bush—revolutionizing all federal highway programming to enable walk
and bike facilities as well as expanding financing of commuter and
intercity rail programming. Under previous law bikepaths, for
example, maximum allowed projects nationwide were held to under $20 a
year while under ISTEA the bounds were loosed and billions annually
were allowed (one state, Ohio, walked off with over half the funds
available nationally). Walk and bike facilities have increasingly
been built on the basis of need if not yet reaching complete equality
to this day. (The Maine referendum rejected an expansion of the
Maine Turnpike and in a companion approved vote for comprehensive
bottom up multi-modal transportation planning and programming. The
California referendum authorized sizable funding for rail projects
including the now successful commuter rail service services of
Caltrain. Note in spite of breaking up the carcentric aspect of U.S.
transportation, highway safety remained a neglected, fragmented
aspect of all U.S. transportation investment and that condition
remains mostly intact to this day.)
Roundabouts
The
first roundabout was built in Montpelier in 1995, the one laner was
the 19th built in America and first east of Vail, CO and
north of Maryland. There are over 10,000 roundabout roundabouts
today in the U.S. and Canada and just a single bicycle fatality and a
single pedestrian fatality recorded to date in the 29 years since the
first roundabout opened in a Las Vegas housing development in 1990.
For comparison to the 10,000 US/Canada roundabouts, in Burlington
since 1998, three pedestrians have died at intersections—two at the
75 signalized intersections—and one cyclist at a signalized
intersection. Burlington recorded (2013-2017) a bicycle or
pedestrian injury each week plus two car occupant injuries. Again,
for comparison, in a half century of data the five downtown
roundabouts in Vermont (Montpelier, Middlebury and Manchester Center)
recorded one non-serious pedestrian injury, 0 bicyclist injuries, and
four minor vehicle injuries. Base national research show roundabouts
reduce serious and fatal injures about 90%.
As
important the roundabout performance in addition to safety is also
light years ahead of traffic signals in service (delay included) and
capacity. Delays of up about two minutes at a signalized
intersection at Colchester/Barrett/Riverside (COLBARI) drops by over
one half in a roundabout, with most of that drop benefitting vehicles
entering from Riverside and Barrett. For pedestrians, delay at a
roundabout crossing is about zero seconds. Maintenance costs are
lower for the roundabout and there are no electric bills.
There
are no busy street Burlington and Chittenden County roundabouts found
among the 14 in Vermont. There are five busy intersection 1-lane
roundabouts in three Vermont downtowns and the one 2-lane, Keene-Turn
Roundabout (I 91/US 5/VT 9) built in Brattleboro in 1999. Through
2015 involving over a half century of operation no cyclist injuries
have been recorded at the five downtown roundabout and apparently
none at the one 2-lane Brattleboro roundabout.
All
Cyclists versus the Vehicle Cyclist
With
the advent of safety designs for a cross section of cyclists (young,
old, experienced, non-experienced) there is some divergence of the
interest of the vehicle cyclist who wishes the speed and freedom of
the automobile. In built up urban areas like Burlington with less
than a handful of roadways with speed limits above 25 mph, vehicle
cyclist can function quite well and variation of speeds do not pose a
problem versus areas dominated by 35 mph roadways and above. That by
no means a reasonable level of safety provided to the vehicle cyclist
and for the less experienced cyclist no provisions exist, the still
challenged cycle track on part of the Unions notwithstanding.
Cycle
track which became almost overnight a universal design in US urban
areas during the last ten years does serve all cyclists but—like
busy highways—while providing maximum safety does get congested and
slows overall bicycle speeds compared to vehicle lane travel.
Result? Many vehicle cyclists will, depending on traffic
conditions, opt to use vehicle lanes even in the presence of cycle
track with its given superior safety rating. Vehicle cyclists do
properly express the demand to retain their “rights” to operate
in vehicle travel ways.
Bicyclists
at a One Lane Roundabout: Bicyclist Movements at St.
Paul/Locust/Shelburne/S. Willard
The
first 1-lane roundabout scheduled to be built in Chittenden County in
2021 occupies the “rotary” intersection known in the neighborhood
as the “intersection of death” at Shelburne Street/Locust
Street/South Willard Street/St. Paul Street (Shelburne Roundabout).
This intersection is in the top five of the worst crash records of
any in Vermont and is funded with 100% federal safety funds. This
one lane roundabout, about 130 feet in diameter, starts at the north
end of Shelburne Street, now a four lane roadway with no bicycle
lanes. The safety afforded by roundabouts occurs in great part by
slowing vehicle at entry and constraining speeds through a
curvature—“deflection” is the word used to describe this
design. Also an island,
“splitter
island” forms a median separating lanes at entry and exit. The
median areas provide a refuge for pedestrians so they only deal with
crossing one lane at a time. Entry and exit lanes are also only
about 10-15 feet in width, narrowness and curvature all inherent in
“deflection” design of roundabouts to manage entry speeds and
maintaining low speeds throughout the circular travelway.
Based
on some preliminary designs, the cyclist at Shelburne Roundabout will
either merge into the vehicle travelway about two-three car lengths
from the circular travelway of the roundabout, or “ramp off” to
the right onto what is essentially “share space” with pedestrian
and then move across one or more crosswalks before “ramping on”
to the street lanes beyond the intersection. Based on European
studies this “choice” for the cyclist is a critical element in
assuring a roundabout is safe for cyclists. The less able, less
skilled cyclists will generally choose to ramp off/ramp on.
Roundabout research clearly shows providing a cyclist “choice” at
a roundabout brings reductions in bicycle injuries approaching car
and bicyclist upwards of 90% compared to a signalize or sign
controlled intersection.
So,
the Shelburne Roundabout will provide off ramp/on ramp to serve
cyclists. The vehicle cyclist can choose to merge with traffic at
entry (not a problem most times of the day except peak times) or
switching to shared spaces before re-entering the destined roadway.
This integration from bike lanes/cycle track is the situation
routinely faced by cyclists at signalized intersections today where
lanes end in recognition in part that vehicles will be turning right
through the straight forward path of the cyclist.
Why
not a separate bike lane? Where there is sufficient right-of-way
certainly (new intersections along the Parkway where there is
sufficient room—Parkway/Home, Parkway/Flynn, for examples—a
separate and safe circular lane for cyclists along with one for
pedestrians meet fully the preferable “equality street” goal at
intersections. A separate circular bike lane is the regular design
approach where room is available, clearly not the case at Shelburne
Roundabout because of adjacent development constraints. The bike
lane is between the vehicle travelway and the pedestrian crossing,
about mid-way between, about 10 feet with pedestrian crossings are
about one car length or 25 feet from the circular travelway. (The
Dutch “Assen” design is preferred by some here in Burlington as
the design slows bikes at vehicle lane crossings—generally the
thinking so far is for cyclists to yield to traffic at such
crossing—yet to be determined. For graphics in this regard and
video see
http://www.aviewfromthecyclepath.com/2014/05/the-best-roundabout-design-for-cyclists.html
.)
Bicyclists
at a 2-Lane Roundabout: Colchester/Barrett/Riverside/Mill (COLBARI)
Vehicle
Cyclist North-South Movement Through a Roundabout on Colchester
Avenue at Barrett/Riverside Intersection onto Winooski
For
the basic movement for the vehicle cyclist north-south on Colchester
Avenue and the Winooski Bridge past Barrett/Riverside/Mill Street
intersections, the basic northerly move from the single entry
lane on Colchester is onto the outside lane straight through the
roundabout along the outside (east side) lane on the Winooski
Bridge to the Winooski Traffic Circulator. From Winooski Traffic
Circulator south onto Colchester Avenue, the vehicle cyclist
takes the inside lane onto the inside lane of the roundabout, then
passes the Riverside entry where entering vehicles are to yield to
the cyclist,and then, finally the cyclist exits onto the
Colchester southward.
The
2-lane roundabout presents a different set of issues for cyclists but
retains several basic elements of the 1-lane discussed above. First
and foremost where space allows, a separate circular bike lane is
provided. (For sake of discussion it is assume that all legs leading
to the roundabout have bicycle lanes.) And, where a separate bike
lane is not feasible as in the case of COLBARI. all entries and exits
like Shelburne Roundabout provide ramp off/ramp on cyclist “choice”
of either taking a vehicle lane or switching to shared space
pedestrian space and using the pedestrian crossings.
Before
considering specific pathways for cyclists, the base performance of a
roundabout cuts delay per vehicle at peak hour almost two thirds, 26
seconds average vehicle delay for the roundabout compared to 71
seconds, over a minute, for a signal. So before a bicycle or
vehicle enters a COLBARI roundabout it stands to save upwards of 45
seconds versus approaching a roundabout. Pedestrians would
experience based on the phasing length, likely in excess of 30 second
delay per crossing at a signal versus essentially zero seconds per
crossing in the roundabout. During off peak times vehicle delay
drops dramatically to a few seconds, a greater decline then the stop
control constrained systems.
Approaching
the Roundabout and Choice
Consider
each approach first, then how the cyclist would move to a set of
destinations from each approach leg. COLBARI both in signal and
roundabout design faces the physical constraints including south and
west grades, drop-off to the river, Mill Street access needs, etc.
(for example, the current two-lane eastbound approach on Riverside
Ave on the south side bordering a steep embankment has no sidewalk,
and a shared space multi-use path is sited on the north side). The
roundabout approach for cyclists westbound on Barrett remains but
there is no apparent “choice” ramp off onto the sidewalk
northerly—though this should not be complicated as there is a
parking area to the rear of large commercial at the northeast corner
of Colchester/Barrett. The roundabout northbound on Colchester plan
features the important “choice” addition,ramp off shared use
path/sidewalk with pedestrians, apparently from a point at or near
Chase. At this location a ramp off might also be considered in the
normal design just before the splitter island onto the shared space.
This other approach, the 2-lane southbound on Colchester from the
Winooski Bridge also appears to lack an off-ramp prior to narrowing
and the cross walk opposite Mill Street.
An
Aside: Access to Mill Street Southbound from Winooski
Access
to Mill Street remains problematic for a signal more than than a
roundabout, but still less than ideal. For cyclists and pedestrians
the easiest access southbound from Winooski appears to be the
sidewalk on the east side of the bridge—presumably expansive and
accommodating of both those on foot and bicycle in a new
bridge—perhaps even with a sidewalk level cycle track/sidewalk
tandem. In this case because of the overall geometry of the
intersection bicyclists southbound would move to pedestrian space
north of the bridge and then take the crosswalk to the east side of
the bridge, then to Mill Street. Northbound, again, all pedestrian
traffic would mostly use utilize shared space/sidewalk space from
Chase, across Barrett, then to Mill. For persistent vehicle cyclists
southbound from Winooski, the east lane is used and then a left turn
onto Mill (note the roundabout preliminary design does not
accommodate a left hand turn—this might be revisited in a more
detailed design process). Still the roundabout—unlike the
signal—does provide a pedestrian crosswalk at Mill and a ramp off
and ramp on for cyclists on the southbound Bridge lane on the west
side and northbound on the east side needs to be included in the
final design.
So
the roundabout option provides a crosswalk to access Mill Street.
So regardless of route approaches of cyclists and pedestrians, a
crosswalk—not provided by signal and a serious lack in the signal
choice. The lack of a signal at Mill Street arises simply from the
horrific delay throughout the intersections which would occur—again,
the signal not only provides poor safety but poor to unacceptable
service.
Finally
for both cyclists on the eastbound lanes of Riverside Ave and those
on the shared use path on the north side, one choice to access Mill
is shared space via the Riverside crosswalk, Colchester Ave
crosswalk, and Barrett crosswalk for pedestrians and cyclist in
shared mode. For cyclists (and pedestrians as well) approaching from
Riverside Ave the preferable second choice--the quickest and
easiest--access to Mill Street is via the roundabout crosswalk across
Colchester to Mill. The access to Mill Street from the shared use
path on Riverside in a signal context requires negotiating three
separate signals crossings—Riverside itself, Colchester, and
Barrett.
Cross Lane Movements
Note,
about one in five roundabouts in North America have partial or full
2-lane configurations—but the one cyclist fatality to date occurred
at a 1-lane roundabout in Truckee, CA (which in the process of
building another roundabout this year).
Clearly
the less experienced cyclist will likely choose to shift to
pedestrian mode rather attempt to deal with either bike lane to
vehicle travel lane at roundabout entry approach and/or lane changes
dictated for all vehicles where, essentially, left hand turns are
involved. For right hand turns and through movements, cyclists can
stay in the outside vehicle lane. First and foremost the cyclist
must shift from a bike lane either to a vehicle travel lane
approaching the entry line or ramp off. A decision to
merge
into the vehicle lane is familiar at just about every signalized
intersection in
Burlington
where there a bike lane approaching a signal—this is not an unusual
movement faced by the cyclists. Finally the task of moving from a
bike lane to a vehicle lane then moving to an inside lane where left
turns are dictated (see, for example, a cyclist if a bike lane were
south bound roundabout approach of Colchester Ave from the Bridge on
a bike lane facing both merging into the vehicle travel lane and
crossing the outside lane to the inside lane necessary to make a left
hand turn on the roundabout to Barrett Street.
There
are a number of pavement markings, signal control options, etc., to
ease cyclist movements at multi-lane signal entries to move across
lanes. A set of pavement marking at the end of a cycle lane into a
two-lane roundabout entry have also been devised by MTJ Engineering.
At the end of the bike lane the cyclist is given two choices—a
solid green ramp off to shared space, and second wide broken-block
green line indicator to vehicles taking the cyclist into the right
lane with continuance into the left entry lane. With roundabouts low
speed conditions exist compared to a cyclist coming to a green signal
and dealing with 20-25 mph or more traffic at stale green/early red
conditions—delay is dictated here (still a right turn on red an
additional contention for cyclists in many cases when abandoning a
bike lane at signals).
Barrett
Street Entry: For moving northbound, the cyclist enters the
right hand lane with first “exit” onto Mill Street, then if
continuing, being prepared to at a later point on the Winooski Bridge
making any lane adjustment required on the Winooski Traffic
Circulator approach. To go to Riverside or Colchester south from the
Colchester north entry, the cyclist enters directly into the inside
lane which takes the cyclist directly to Riverside and Colchester
south access accomplished by holding to the inside lane and exiting
via the “spiral” to the outside lane—traffic southbound from
Winooski to Riverside are to yield to the cyclist exiting the one
lane exit at Riverside—and a similar action by vehicles entering
the roundabout from Riverside eastbound entry. These maneuvers are
most challenging to cyclists at a multi-lane roundabout. Note that
all these destinations can be done via the ramp off/ramp on
pedestrian spaces and pedestrian crossings.
Colchester
Northbound Entry: The roundabout northbound entry, 1-lane, on
Colchester presents the cyclists with two choices: a. outside lane
for right turn at Barrett or Mill Street and straight ahead in right
exit lane at 2-lane north roundabout exit to the Winooski traffic
circulator at the Winooski end of the bridge; and (b) enter the
inside lane destined to inside lane at 2-lane exit north exit onto
the bridge to the Winooski traffic circulator or continuing on the
lane which through spiral marking becomes “outside” opposite the
north splitter island and onto the 1-lane exit onto Riverside.
Colchester
Southbound Entry: The
Colchester southbound 2-lane entry outside lane is for Riverside
destination only at the 1-lane exit onto Riverside. The inside lane
following the pavement markings exits onto Colchester and Barrett,
both 1-lane exits.
Riverside
Entry: The Riverside entry is
2-lane. The outside lane accommodates all destinations: the 1-lane
exits at Colchester, Barrett , Mill and outside lane at the north
2-lane exit as well as by moving to the inside lane after the
Colchester exit onto the inside lane at the north 2-lane exit toward
the Winooski traffic circulator. The inside lane entry at Riverside
serves only vehicles northbound to the Winooski and vehicles are to
use the west exit lane at the north roundabout exit.
Reversing
Direction: Reversing direction
is a natural service all roundabout provide. Generally the reverse
direction move is straightforward, enter to the left onto the inside
lane and remain on the inside lane until the exit on the entry one
started form—this may or may not involve yield by vehicles on an
outside lane where there is an outside lane at the exit point.
The
COLBARI design is a preliminary design and changes may be made—but
the types of pathways described here pretty much cover all the
possibilities a cyclist would encounter at COLBARI or any other
multi-lane roundabout (three lane roundabouts do exist but are rare
and present an additional set of challenges). Pavement markings and
some signing will aid all in decision making at a 2-lane roundabout.
All
roundabouts experience startup hiccups as drivers, pedestrians and
cyclists learn the ropes. But roundabouts easily accommodate the
learning curve with little or no crashes as the the caution of users
at an already low speed environment remains forgiving. The COLBARI
roundabout will be significantly smaller than the Brattleboro Keene
Turn roundabout, 172 feet in diameter. COLBARI will be about 150
feet in diameter which insures lower speeds which are more compatible
with urban levels of walk and bike users as well as, most
importantly, reduced injuries for all modes. It is noteworthy that
Oakland County, MI, a two-million population center adjacent to
Detroit has a large number of roundabouts (Waterloo, ON does also)
including 2-laners and already is showing injury and fatality rates
far lower than state and national averages. Vermont is about the
national average in serious and fatal highway injuries. Roundabout
are in fact beginning to reflect reduced area rates of serious and
fatal injuries where their numbers are dense.
Finally,
there are now several videos and educational materials (and more all
the time) on “how to” negotiate 1-lane and 2-lane roundabouts.
So as we approach the first roundabouts in Chittenden County, these
AV materials can be provided both the to the public—and as was done
at Montpelier High School driver ed and private driver ed training
in 1994 and forward the education of future drivers will cover the
coming dominance of roundabouts on our northeastern Vermont roadways.
The roundabout education materials will be incorporate on how those
on foot, bicycle and motor vehicle can work together to their
personal safety and reduce the needless carnage on our streets and
highways, now 22,000 fatals a year over the low rates of the leading
four nations as the U.S. has fallen from first to 20th
since 1990.
Two
Lane Roundabouts and Pedestrian Signals
The
U.S. is the only nation where two-lane roundabouts likely will
include pedestrian signals, ostensibly because because recent
regulations and policies under the American Disabilities Act of 1990
(ADA). There are no ADA policies and regulations in regard to single
lane roundabouts. There is not evidence that signals, like the
Rectangular Rapid Flashing Beacon (RRFB) familiar to most in
Burlington, or other types of signals will increase or decease safety
for pedestrians at 2-lane roundabouts. That said it can be expected
that actuated signals at each pedestrian crossing (and at splitter
islands between lanes) will be equipped with this technology.
Sources
1.
Colchester/Barrett/Riverside 2018 final draft report
Roundabout
alternative page 49; signal “Alternative 1” page 42
2.
St. Paul/Locust/Shelburne/S. Willard Website Addresses
3.
CCRPC Shelburne Street Rotary
https://studiesandreports.ccrpcvt.org/wp-content/uploads/2017/01/ShelburneSt_Rotary_200203.pdf
Tony
Redington
TonyRVT99@gmail.com
