03levelofservice.pdf

Level

of

Service

Measures

for

Biking:

A

Comparative

Analysis

of

Calculation

Methods

Matthew

M.

Day,

MRP

Abstract

Traditional

methods

for

computing

level

of

service

(LOS)

have

implicitly

favored

mobility at the

ex-pense

of

accessibility.

The

LOS

concept

was

developed

by highway

engineersinthe 1

950s

asa

method

of measuring

thelevel

of

mobility

provided

by

acertain facility

(FDOT,

2002). It

has

been

applied in

recent years to alternative transportation

modes

such

as

walking,

bicycling

and

public transit.

This

article

analyzes

and compares

the results

of applying

several

of

the

LOS

methods

that

have been

devel-oped

for alternative transportation

modes

to a

study

areain

Chapel

Hill,

North

Carolina.

Introduction

Traditional

methods of measuring

thelevel

of

service

(LOS)

focus narrowly

upon

mobility, asdetermined

by

therelationship

of

facility capacityto

volume

of

traffic, while ignoring accessibility. In the field

of

transportationplanning,mobilityhasbeen definedas

theability toget

from one

placetoanother(Hansen,

1959;

Handy.

1994). Accessibility,

by

contrast,has

been

definedas the potentialfor interaction. Inother words, mobility is a

measure of

how

easily a user can

move

throughafacility;accessibility,ontheother

hand,

measures

how

easily a user

can reach

a destinationusing afacility.

When

accessibilityis low. a person'sabilitytoreach

a destination is

compromised.

Traditional

LOS

measures

do

notcapturethiseffect.

Under

traditional

LOS

measures,corridorswith highlevelsof mobility

willscorehigh

on

traditional

LOS

methods,regardless

of

whether

theyofferaccessibility. In

many

cases a

facilitywillofferhigh mobilitybut

low

accessibility.

or vice versa.

For

example,

a

community

with

abundant

roads

and

little

congestion

but

with

relatively

few

destinations for

shopping

or other

activities displays

poor

accessibility but

good

mobility.

An

area featuringhighlevelsof congestion

butrelativelyshort distances

between

where

residents

live

and

all

needed and

desired destinationshas

good

accessibility butpoormobility.

A

more

accurate

measure of

level ofservice

would

considerboth themobility

and

accessibility offered

by

afacility(

Levine

and

Garb.

2002

). Recently,

new

LOS

methods emphasizing

accessibility have

been

developed.

These

new

measures

allow planners,

engineers,

and

others todetermine the accessibility

offered

by

a broad range

of

transportation facilities.

Matthew

M

Dav

is a graduate of the University of North

Carolina Departmentot Cityand RegionalPlanning.

He

is

currentlyworkingasa TransportationEngineerattheNorth

Carolina Department of Transportation in the Western

including roadways, transit facilities,

and

facilities

constructed forpedestrians

and

bicyclists.

This paperuses aselection ofaccessibility-based

LOS

measures

that

have been developed

for pedestrian,

bicycle,

and

transit facilities toanalyzethe

downtown

areaofChapelHill,NorthCarolina,(seeFigure1)

where

traffic levels are low and walking, biking, and public

transitarepopular

modes

of

moving from

placeto place.

Figure 1. Franklin Street in

Chapel

Hill,

NC

is a

pedestrian

and

bike-friendly

roadway.

Photo by

Helen

Chanev.

Capacity-based

LOS

standards

The

Highway

Capacity-

Manual

{HCM)

isthestandard

methodological guideintheUnitedStates for

computing

automobile levelofservice

(FDOT,

2002).

For

modes

of

transportation other than private automobiles (for

which

theHig/nvcn- Capacity-

Manual

method, though

flawed,isgenerallyused),thereislessagreement

among

transportationplannersandengineers astoanacceptable

approach

for

computing

level

of

service.

The

TransportationResearch

Board

(TRB)

hasdevelopeda

Transit

Capacity

and

Quality

of Senice

Manual

(TCOSM)

that outlines

many

different

methods of

computing

LOS

fortransitservicesbased

on

capacity/

mobility,accessibility,andqualitymeasures(Kittelson,

1999).

Some

authors suggestthatcharacteristics

of

the

built urban environment (Jaskiewicz,

no

date) or the

socialor policyenvironment

(Hoehner

ctal.,

2003)

are

also factorsthatinfluence thelevelofservice thataperson

perceives

on

aparticularnon-automobilefacility.

Non-capacity Level of Service

Models

Several accessibility-based

LOS

models have been

developedtoevaluatebicyclistandpedestrianperceived

safetywith respectto

motor

vehicletraffic

and

comfort

inusingthe

roadway

corridor.

The

most

popular

methods

fordeterminingPedestrian

Level ofService

(PLOS)

include the

PLOS

method, developed bySprinkleConsulting,

and

theFruin

PLOS

method, which

is included in the

Highway

Safety

Manual.

Emerging

national standards forevaluating

thebike-friendlinessofa

roadway

are theBicycle Level

of

Service

(BLOS)

method, developed

by

Sprinkle

Consulting,andtheBicycle CompatibilityIndex(BCI),

developed by

the Federal

Highway

Administration

(FHWA).

The

TransitLevelofService

(TLOS)

method,

developed bytheFloridaDepartment ofTransportation,

computes

levelofservicebased

on

availability

of

transit

within awalkingdistance.

Each

ofthesefive

methods

aredescribedindetailinthefollowingsections.

While

some

ofthese models focus

upon

travel

demand

and

facilitycapacity,othersarebased

upon

awider range

of

factors,suchas accessibility,environmentalquality,

and

safety.

LOS

methodolgies for pedestrian

and

bicyclist travel

can beuseful toplanners ina varietyofways.

These

tools can help planners to identify

weak

links in a

network ofsidewalks or bicyclefacilities,forexample.

Level

of

Service

Measures

for Biking

prioritizesitesneeding improvement. Plannerscan use

thebicycle

LOS

measurestodetermine

which

routesto

includeinthebicycle network.

They

may

alsouse the

measures to createa bicycle

map, which

can help the publicinchoosing

which

routesto take.

Pedestrian Level

of

Senice

Method, developed

by

SprinkleConsulting

The

FloridaDepartment ofTransportation

(FDOT)

uses

amethodologyforcomputingpedestrianlevelofservice

that

was

created

by

SprinkleConsultingandisbased

on

four

major

physical characteristics

of

the street

and

sidewalk space: presence

of

a sidewalk

and

lateral

separationfromstreet;motorvehiclevolume;trafficspeed;

and drivewaytraffic

volume

andaccessfrequency(Landis

etal.,

200

1 ).

The

creationofthe

model

involved asurvey

anda regressionanalysisofthesurveyresults.

The

firm

first conducted a pedestrian facility quality survey in

Pensacola. Florida, asking users about environmental

factors,includingwidthofsidewalk,width ofbikelane, presenceof sidewalkbuffer,

volume

and speed oftraffic,

and

number

oftrafficlanes,

among

others.

Second,thefirmanalyzedthe resultsofthesurveyusing

a regression analysis, in order to determine

which

environmental factors were

most

closely related tothe

users'perceivedqualityofthe facilities

(FDOT.

2002). Inalaterstudy,thefirmsoughttodeterminewhether

two

otherfactors

—

thepresenceofother pedestriansandthe presenceofbuildings againsttheedge of asidewalk

—

wererelatedtotheperceivedqualityofpedestrianfacilities,

buttheyfoundthatnosuchrelationship existed.

The

PLOS

method

isfocusedprimarily

upon

physical

characteristics

of

the

roadway

and

sidewalk

environment,

and

it provides a simple

method

for

computing

LOS

along a

segment of

the road/path

network. This

method

was

chosen becauseitisrelatively

objective

and

easily

converted

into a

uniformly-applicable levelofservicemeasure.

The

basicequationthatthis

PLOS

method

utilizes

is

(FDOT,

2002):

PLOS

=

-1.2276

InfWol

+

Wl

+

fp *

%OSP

+fb*

Wb

₊

_fsw *

_Ws)

+

0.0091(Vol15/L)

+

0.0004 *

_SPD2

+

6.0468

(forEnglishunits)

where,

Wol

=

widthofoutside laneoftraffic(including

on-streetparkingarea);

Wl

=

width of

marked

shoulderor

marked

bicycle lane;

fp

=

on-streetparkingcoefficientorfactor (0.2

usedinanalysis);

%OSP

=

percentof

segment

withon-street

parking;

fb

=

sidewalkbufferfactor;

Wb

=

width ofbuffer

between

streetand

sidewalk;

fsw

=

sidewalkcoefficientorfactor

(equals6-0.3*

_Ws);

Ws

=

width ofsidewalk;

Vol

15

=

volume

ofdirectional

motor

trafficin

peak 15-minuteperiod;

L

= number

ofdirectionalthroughlanes;

SPD =

average speed

of

motor

vehicletraffic.

Under

the

PLOS

method.

LOS

iscalculatedforboth

sidesof eachroad

segment

beingstudied;gradesare

LOS

Grade

PLOS/BLOS

Score

A

<

1.5

B

1.5-2.5

C

2.5-3.5

D

3.5-4.5

E

4.5-5.5

F

>5.5

Fruin Pedestrian Level

of

Service Method,

developed

by

Fruin

and

included

in the

Highway

Capacity

Manual

The

Fruinmethod,

which

requires the inputofpedestrian

count data, can provide useful information about the

capacityofthe sidewalksin high-traffic locations,and determine whetherthereisaneedforadditionalsidewalk

capacityin theselocations.

The

Fruin

methodology

isdefined

by

thefollowing

equation

(TRB,

2000):

Pedestrianunitflowrate

=

V

15/(15 *

We)

where,

VI

5

=

peak

15-minute

pedestrian traffic rate (persons per 15-minutcs);

We

=

effectivewidth ofsidewalk.

The

flowrategenerated

by

theequation

above

isused

todeterminea

LOS

gradefora pedestrianfacilitybased

onthestandardsbelow.

LOS

Grade

Flow

(persons/min/ft)

A

<5

B

5-7

C

7-

10

D

10-15

E

15-23

F

>23

The

Fruin

method

is a capacity-based

method

and

assumes

thattheprimary determinantofqualityservice

in the pedestrian environment is the ability to

move

through that environment with as little

impedance

as possible.

BicycleLevel

_ofSennce

Method, developed by

SprinkleConsulting

SprinkleConsultingdevelopeda

BLOS

method

forthe

Florida

Department

ofTransportation.Thismethod,like

Fruin's,isbased

upon

physicalcharacteristicsoftheroad

andbicyclefacilitiesbutfocusestoa greater extent than the Fruin

method

upon

the presence

and

quality

of

bicyclefacilities

and

the characteristicsof

motor

vehicle

traffic, including the volume, speed,

and

number

of

heavy

trucks (see Figure 2).

The

various data are

combined

intoa

LOS

scorebased

on

a regression

model

(FDOT,

2002).

Figure

2.

The

quality of

bike

facilities,

such

as

bike

lanes,

may

affectacyclist's

BLOS.

Photo

by

Level

of

Service

Measures

for Biking 7

This particular

method

includes a factoron

roadway

condition,

which

isa variable not included inthe

BCI

method

(discussed below). Otherfactorsinclude

motor

vehicletraffic

volume

andspeed,effectiveoutside lane width,and

amount

oftrucktraffic.

Bicycle Level

of

Service isdefined

by

this

model

as

(FDOT,

2002):

BLOS

=

0.507 ln(Voll5 /L)

+

0.199* SPt* (1

+

10.38*

HV)2

+

7.066 *

(1/PR5)2-0.005 *

We2

+

0.760(forEnglishunits)

where.

Vol

15

=

volume

ofdirectionaltrafficin 15-minute peakperiod;

L =

total

number

of throughlanes;

SPt

=

effective speed limit(1.1199 ln(SPp-20)

+

0.8103,

SPp

=

postedspeed);

HV

=

percent

heavy

trucks;

PR5

=

FHWA

5-point surfaceconditionrating;

We

=

averageeffectivewidthofoutside lane(lane

widthlessobstructions).

Levelofservice gradesareassigned forbothsides of

each road

segment

beingstudiedusingthe

same

scale

as forthe

PLOS

model

(see previous).

Bicycle CompatibilityIndex Method, developedbythe

Federal

Highway

Administration

The

Federal

Highway

Administration

(FHWA)

has developeda

BCI

thatservesasameasure ofquality for

differentroads in terms ofbicycletraffic.

The

BCI

is

similarto the aforementioned

FDOT

pedestrian and

bicycle level

of

service methods, in that it primarily

focuses

on

physicalcharacteristicsofthe road,suchas thepresence ofbicyclelanesorthe

volume

ofautomobile

andtrucktraffic,and

combines

them

intoameasure of

facility quality that is not based entirely

on

capacity

(FHWA,

1998).

While

the

FDOT

and

FHWA

methods

of

computing

BLOS

examine

similarcharacteristicsof

thebicyclist'senvironment,the

two

modelsusedifferent

criteriaweightsand could produce verydifferentresults.

The

FHWA

method

isdifferent

from

theSprinkle

BLOS

method

inthatitaccounts forthepresenceofabicycle

lane, thetraffic

volume

inlanesother than theoutside

lane, andthepresence,occupancy, andturnoverof

on-streetparking.

The

BCI

uses the following equationto

compute

levelofservice

(FHWA,

1998):

BCI

=

3.67-0.966 *

_BL-0.410

*

BLW

- _0.498 *

CLW

+

0.002 *

_CLV

+

0.0004 *

OLV

+

0.022 *

SPD

+

0.506*

PKG

-0.264*

AREA

+

AF

where,

BL

=

presenceofbike lane(no

=

0,yes

=

1);

BLW

=

bicycle lanewidth(meters);

CLW

=

curblanewidth(meters);

CLV

=

curb lane

volume

(peakhour);

OLV

=

otherlane(s)

volume

in

same

direction(peak

hour);

SPD

=

85thpercentile

of

speed;

PKG

=

presenceofparkinglaneoccupied

more

than

30%(no=0,yes=l);

AREA

=

type

of

development

(residential1,

other=0);

AF

=

truck

volume

factor

+

parkingturnoverfactor

+

rightturn

volume

factor.

The

gradingscale forthe

BCI

ispresentedlater,along

with a discussiononaproposedadjustmenttothegrading

scale as a result ofthe analysis conducted in Chapel

Transit Level

_of

Service

Method, developed

by the

Florida

Department of

Transportation

The

Florida

Department of

Transportation uses a

sophisticated

method

for determining transit level

of

service at thesystem, route,and stop levels.

FDOT's

method

isbuilt

upon

the

framework

setupintheFederal

TransitCapacity

and

Quality

_of

Service

Manual,

which

suggests

measuring

transit accessibility

by

service frequency,hoursofservice,andservicecoverage. This

TLOS

method

takesintoaccountthesefactors

and

uses a free

downloadable computer

program.

Geographic

InformationSystems(GIS),andspreadsheetsto

compute

LOS

based

on

availabilityoftransitwithin a walking

distance (based

on

awalking network),given vehicle

headways, andprojected waittimesforindividualroutes

andstops

(Ryus

etal.,2000). Itshouldbe notedthatthe

TLOS

doesnotaddresswhetherroutesconnectorigins

anddestinationswell,orwhethertransitcustomersare

comfortableandsafe

on

theirtrips.

The

transit level ofservice in the Chapel Hill

Town

Centerwillbe determinedusing a

form

oft£e Florida

Department

of Transportation's

TLOS

methodology.

The

fullversionofthe

TLOS

softwareisa

somewhat

burdensome program

touseandrequiresa greatdeal

of

data that is not always immediately available to the

public; however, theprogram,

which

is

downloadable

fromthe Internetat

no

cost,

comes

with a spreadsheet

thatallowsfora simplified calculationof

LOS

forroute

segments

and

stops.

The

spreadsheet has

few

data

requirements.

One

can obtain a complete output

by

inputtingonlythescheduledarrivalanddeparturetimes

of buses

—

information

which

can be easily obtained fromaschedule book.

routesoperatingalongthatsegment,

and

thetimesthat

buses are scheduled to stop at the stops along the

segment.

A

macro

built into the spreadsheet then

calculates the

number

of minutes duringthe

day

thata

stophasservice availableto it.based

upon

a

number

of

variables, including user-defined

maximum

wait

times at stops,

walking

distances,

environmental

characteristics,

and

theuseofstraight-lineor

network-basedbuffers.

Levelofservice can be

computed two

different

ways

usingthisspreadsheetbecausetheuser definesthetime

duration

of

the calculations. Iftheuseronlycalculates

TLOS

fora portion

of

the

day

(i.e.,duringthetimeof

service),the

program

definesalettergradebased

on

the

frequencyofserviceguidelinesintheTransitCapacity

and

Quality

_of

Service

Manual,

shown

below

(Kittelson,

2001 and

1999).

LOS

TLOS

Score

Headways*

(%

timeserved)

(TCQSM)

A

>

50

%

<

10 minutes

B

35.7%

-

50%

10-14 minutes

C

25%

-

35.7%

15-20 minutes

D

16.7%

-25%

21-30 minutes

E

8.3%-

16.7%

31-60minutes

F

< 8.3%

>

60

minutes

*assumes

5minute

maximum

waittime

See

Figure 3 fora graphic representation

of

TLOS

scores for the

Chapel

Hill transitsystem.

The

TLOS

route

segment

spreadsheetallowsauserto

Level

of

Service

Measures

for Biking

On

theotherhand,iftheuserdefines the calculationperiod asexactly24hours, the

LOS

gradeisdeterminedjointly

byfrequencyandhoursofservice. Thisisdone

by

simply

multiplyingthe

TLOS

score standards(intermsofpercent

time served) together to create ajoint standard. For

example

the

"A"

standardfor

headways

isless than 10

minutesandthestandardforhoursofservice isgreater

than19hours.

Headways

ofminutes,assuming5minute

waittimes,

mean

thata locationisserved50percentof

the time. Beingserved 19hours outof 24

means

being

served79percentofthe time. Seventy-nine percentof 50

percent is 39.6 percent, soany

TLOS

scoreover 39.6

percent

would

receivea grade of "A."

The

following

tablesummarizesthestandardsfor24-hour

TLOS

grading

(Kittclson,2001 and 1999).

LOS TLOS

Score

Headways

Hours

Served

(TCQSM)

(TCQSM)

A

> 39.6%<

1 minutes 19-24

B

25.3-

_39.6%

10-14 minutes 17-18

C

14.6-25.3%

15-20 minutes 14-16

D

8.4-

_14.6%

_{21-30 minutes} 12-13

E

1.4-8.4%

31-60minutes 4-11

F

<1.4%

>

60

minutes 0-3

O

Downtownbusstops

-Segments |EOperationTLOS

|DOperationTLOS

|COperationTLOS BOperationTLOS

AOperationTLOS

Figure

3.

Example

of buffers

used

in

TLOS

software

that outlineareas of

Chapel

Hill's transit

Forsimplicityincalculating

TLOS

forthiscomparative

study, environmentalvalues fortheareas surrounding

stops

were

not calculated. Calculatingenvironmental

values

would

haverequiredthedata

on

pedestrianfacility

quality as well asjob and population density around

stops.

When

such environmental data arcincluded in

ananalysis, theresearcher

weighs

the stopsaccording

to the data.

When

such environmental data is not

includedinthe analysis,allstopsareweightedequally.

Methodology

This analysis involves

computing

level ofservice for

thevarious

modes

of

transportationinthe

Town

Center

area

of

Chapel Hill usingthe

methods

outlined in the

above literaturereview as a

means

of

discovering the

applicability

and

benefits

of

existing

LOS

methodologies.

Chapel Hill isa small city in the

Piedmont

region

of

centralNorthCarolina

and

isincluded intheResearch

Triangle region(Raleigh-Durham-ChapelHill). Chapel

Hill's

Town

Centeressentiallycentersaround

two

streets.

FranklinStreetand

Rosemary

Street,

which

runparallel

to

one

another.Traffic levelsare

low

inthe

downtown

area,

due

in large part toascarcityof parking anda

20

mph

posted speed limit. Walking, biking, and public

transitarcpopular formsoftransportation in this area.

The

University of NorthCarolina's

main

campus

isat

the eastern end ofthe

Town

Center.

The

Town

of

Carrboroliesdirectly tothewest of Chapel Hill.

Forthepurposesofthisanalysis,a study area

was

defined thatextended one block

from

thesouthsideofFranklin

Street

and

fromthenorthsideof

Rosemary

Street.

The

streets in this study area

were

broken into segments,

which

generally spanned

from one

intersection tothe

nextintersection,with a

few

exceptions. Data usedin

the

analyses

included

GIS

parcel data, aerial photographs, pedestrian andvehicletrafficcounts,and

GIS

busstop locationdata. Trafficcountsonindividual

streetsegments

were

estimatedbased

on

known

traffic

countsandestimatedtripendspersegment.

Inaneffortto

compare

different

methods

of

computing

levelofservice,

two

different

methods

have been used

foreach

mode

oftravel beingstudied. Forpedestrian

levelofservice,the

PLOS

method

developed bySprinkle

Consultingis

compared

withthe capacity-based Fruin

method,

which

isthe

method

presented inthe Higlma)-Capacity

Manual.

ForbicyclelevelofserviceSprinkle

Consulting's

BLOS

calculation is

compared

with the

Federal

Highway

Administration's

BCI

calculation.

Finally, fortransitlevelofservice,a simplified version

ofFlorida's

TLOS

method

isused.

The

TLOS

method

includes

two methods of

calculations

—

one

which

is

based

upon

frequencyofservice

and

another

which

is

based

on

frequencyand hours ofservice.Bothofthese

TLOS

methods

arc

employed

inthe analysis.

Findings

PedestrianFacilities

The two methods

utilizedforcalculatingpedestrianlevel

of

serviceyieldedwidelydivergentresults.

The

Fruin

method

paintsa picture

of

excellence inChapel Hill's

pedestrianenvironment. Allofthelocationsfor

which

theFruin

method was

applied received a

LOS

grade of

"A."

The

PLOS

model,

on

theotherhand, provides a

more

variedpicture.

PLOS

gradesforthe

town

center

rangedfrom

"A"

to "E." with

most

facilities fallingin

themiddleoftherange

("B"

or "C").

The

variation in scores

produced

by

the

two models

Level

of

Service

Measures

forBiking 11

Fruinmethod, beinga capacity-basedmethod,basesits

LOS

gradesentirely

on

the

volume

ofpedestriantraffic

and

the capacity ofa pedestrian facility.

The

Fruin

analysis,

which

was

conducted

on

the locationsin the

Town

Center study areafor

which

there

were

recent pedestrian

volume

counts,producedaresultofall"A's"

forpedestrianfacilities inthearea.

While

both

models

arehelpfulinevaluatingthe

LOS

of

pedestrians inChapel Hill, each

model

is based

upon

assumptions

which

are

somewhat

incompatiblewiththe

Chapel

Hill

environment.

The

Fruin

method

is a

capacity-based

method and assumes

that the primary

determinant of

quality service in the pedestrian

environment

is the ability to

move

through

that

environment with as little

impedance

aspossible. In

ChapelHill's

Town

Center,

where

pedestrianflowsare steady,butcertainly notcrush flows,everypedestrian

facilitywillscorean

"A"

(flowisuninterrupted). Clearly,

thishas littleutility fordetermining the quality

of

the

pedestrianenvironment in thissituation ofexamining

residential

and commercial

streetfronts.

The

method

seems

better suited to

determining

adequacy

of

pedestrianfacilities atairports,stadiums,

and

schools,

where one would

expect very large

crowds

at certain

peaktimes.

The

PLOS

method, bycontrast,calculatesscoresbased

upon

characteristicsofthepedestrianenvironment. Like

most

level ofservicemodels, the

PLOS

model was

developedprimarily foruse

on

arterial highways.

As

such,the assumptions

upon which

the

model

isbased

do

not logicallyapplyto local residential streets. For

example,

one

assumption ofthe

PLOS

isthatpedestrians

do

not

walk

inthestreet, but walk, instead,beside the

road

—

either

on

a sidewalkor

on

the grass. Experience

tellsus,however,that

many

peopleinChapelHill

walk

inthestreet

on low-volume

roads

which do

notfeature

sidewalks.

The

PLOS

model assumes

the carsalways

actas abuffer.

As

such, the

PLOS

model

gives high

gradestoside streets

where

on-streetparkingispresent.

In reality, streets in

Chapel

Hill featuring on-street

parkingand

no

sidewalksconstitutealess-safepedestrian

environment,as pedestriansareforced to

walk

further intothestreet.Thisproblemoccurs

on

severalstreetsin

the

Town

Center study area.

To

account for this

inconsistency,

we

must

assume

that for streets

where

thereis

no

sidewalkbutthereison-streetparking,both

sides

of

thestreetshouldhavea

LOS

gradethatisclose

to thatfound

on

the sideofthestreetthatdoesnot

have

on-streetparking.

BicycleFacilities

The

two methods

used forexaminingbicyclelevel

of

service

models

show

that bicycle level ofservice is

lowest

in the areas

immediately surrounding

the

Universityof NorthCarolinacampus.

The

resultsofthe Sprinkle Consulting

BLOS

method

portrayarelativelysafebicyclingenvironmentin

much

of

the

Chapel

Hill

Town

Center.

Most

areasnorth

and

west

of

theintersectionofFranklin

and

Columbia

Streets

(the

de

factocenteroftown)receiveda scoreofat least

"C."

Areas around

the

edge

of

the

UNC

campus,

however,received gradesof

"D"

and

"E"

forthelarge

part.

These

grades are given ineach direction, since

bicycle traffic flows in the

same

direction as

motor

vehicletraffic,

on

theright-handsideofthestreet.

The

resultsinthetestcase generally

show

lowerscores

on

roadswithhightraffic

volumes

and

narrowoutside

lanes. Locationswithon-streetparkingalsogenerally

have

lowerscores than those withouton-streetparking,

because this parking is an obstruction

and

potential

turnover.

None

oftheseroadsegmentscontainstriped

bicyclelanes,

which

also leadstothelowerscores.

The

Federal

Highway

Administration's

BCI

method

producedasimilarpatternofresultstothe

BLOS

method

but generally resulted in lowergrades.

The

BCI

also

found

the

most

deficient areas to be those near the

university campus, but found the

Town

Center tobe

more

deficient overall.

Only

Rosemary

Street

and

a

few

residentialstreetshaveconsistentlypassingscores,

arid

no

segmentsinthestudy areareceived a

BCI

grade

of'A."

The

BCI

resultsarcheavily influenced

by

theweighting

ofthe factors in the

BCI

model. This

model

includes

more

inputfactorsthanthe

BLOS

model,

which

would

suggestthatitmightbea

more

accurate representation

ofactual conditions.

The BCI

model, however,

seems

to

have

results that are very suspect. It

may

seem

surprisingthata

low-volume

sidestreetcouldreceive a

levelofservicegrade

of

"D." Thisresultinthetestcase

iscausedbytheheavy weightthatthe

BCI

model

gives tothewidthoftheroadway. Consideringthat

many

of

thesesidestreetsarenarrow,the

model

has ascribedto

them

a

low

score.

Thispointsatthe

same

issuenotedinthesection

above

pertaining to the

PLOS

model: these level

of

service

methods were

created primarily for use

on

arterial

highways,not

on

sidestreets.

The

results

of

the

BCI

model

still have

some

utility.

They

point to locations thatcouldcertainlybe

improved

intermsofthebicycling

environment.

However,

they are not as useful as the

BLOS

results for determining mitigation priorities

because ofthe

skewed

results

of

theanalysis.

Based

upon

the results calculated,the

BCI

gradingscale

seems

inadequateforexplaining bicyclelevelofservice

on

minor

sidestreets.

Almost

allsidestreets inthestudy areareceivedvery

low

gradesundertheinitialgrading

scale forthe

BCI

method,duetothe relatively

low

weight

the

BCI

places

on

traffic

volume

andthehighweightit

places

on

lanewidth and thepresenceofbicycle lanes

which

arcgenerally notfound

on minor

streets. Forthis

reason, a modified gradingscale presented

below

was

developedforlow-

volume

residentialstreets. Ingeneral,

the

low-volume

road

BCI

grading scale that

was

developed simply increases the acceptable

BCI

score

foreach correspondinglettergrade. This

was

determined

tobe asimpler, albeit a lessmethodologically-sound,

method

of modifyingthe

BCI

thanattemptingto

modify

the

BCI

equation itself. This grading scale

was

developed

somewhat

arbitrarily.

However,

with the

original dataused in developing the BCI, it might be

possibletogenerate alessarbitraryrevisedgradingscale for

low-volume

roads.

LOS

Grade

High-

volume

Low-volume

(original) (adjusted)

BCI

score

BCI

score

A

<

1.50

<2.0

B

1.51-2.30 2.01-3.0

C

2.31-3.40

3.01 -4.0

D

3.41 -4.40

4.01-5.0

E

4.41 -5.30 5.01 -6.0

F

>5.30

>6.0

TransitFacilities

As

a baseline determinationoflevelofservice,asimple

one-fourthmilebuffer analysisforeachbus stopinthe

Town

Center

was

performed.

The

entirestudy area

was

determinedtobewithin one-fourthmileofabusstop.

Level

of

Service

Measures

for Biking

13

serviceinanarea

would

stopatthispoint.

Based

onthis

simplespatial accessibility analysis alone, the Chapel

Hill

Town

Centerappearstohaveexcellenttransitservice

(referbacktoFigure3).

The

route

segment

worksheet in the

TLOS

software

offers

two methods

forcomputingtransitlevelofservice.

First,a24-hourlevelofservicecan be determinedbased

on

servicefrequency

and

hoursofservice standardsin

the

TCOSM.

Second, an operation-period level

of

service can

be determined based

only

on

service

frequency during the hours that a route is in service.

The

two methods

producesimilar,butslightly different

results. Usinga

GIS

program,itispossibletographically

display the results

and

find spatial patterns

and

differencesinthe resultsgenerated.

An

examination

of

the 24-hour

TLOS

accessibility

results forChapel Hillrevealsthatservicecoverage is

actually veiy

good

inthe

Town

Center.

While

some

corridors

may

not

have

good

service, there is

good

servicenearby

on

parallelcorridors.

Mapping

the

TLOS

spreadsheetresultsin

GIS

allowsrecoveryofthespatial

analysis that is lost

by

using the simple spreadsheet

insteadofthefull

TLOS

program

to

compute

levelof

service.

Almost

allofthe

Town

Centerstudy areafalls

within one-fourthmileofatransitstopwitha

TLOS

of

"B"

orbetter

—

onlythefarnorthwest

comer

ofthestudy

areahaspooraccessibility to

good

transitservice.

The

operation-time

TLOS

analysis

produced

similar

results.

The

corridor

and

stoplocations that

do

not

meet

a

minimum

standard of

TLOS

"C"

are identical

—

the

accessibility results are almost identical to the results

forthe24-hourTLOS/accessibility.

The main

difference

between

the

two methods

isindeterminingthe levelof

service alongFranklin Street,

which

isthe

main

street

through the study area.

The

operation-time analysis

shows

thatduringthetime busesoperatealongFranklin

Street, the frequency ofservice is not as

good

in the

westbound

direction as in the eastbound direction. Informationsuchasthiscould beusefulinshiftingbus schedules to

maximize

headway

efficiency in this corridor.

Conclusion

Pedestrian, bicyclist,

and

transit service quality vary widelyacrosstheChapel Hill

Town

Center. Levelsof service

vary

from

"A"

to

"E"

in all

modes

of

transportation. Thereiscertainlyanopportunityforthe

town

to improveconditions in low-scoring areas,

and

several potentialmitigationmeasures can be determined

based onthe factorvalues

and

datausedinthevarious

LOS

models.Potentialmitigationstrategiesincludethe

additionofsidewalks

and

bicyclelanes,theaddition or

removal ofon-street parking, thespatial

and

temporal addition

of

transit service,

and

other physical improvements.

Many

of these mitigation measures,

which

aredesignedtoallow a

segment

toreacha passing

gradein

one

ofthe level

of

servicemethods,areatodds

with mitigation measures suggested

by

otherlevel

of

seivice models. For example, a

PLOS

grade can be

improved

by

addingon-streetparking,buta

BLOS

grade

isimproved

by

removingtheparking. Whileitispossible tocontinue adjusting mitigationstrategiesineach

model

sothestrategy suggestedin

one

doesnotconflictwith

that of another model, it

would

be useful to have a

standard

method

for

combining

the various

models

across the different

modes

of

travel to ensure thatthe

needsofusersineach

mode

can be

met by

aproposed

mitigation measure. This type of

model

integration

would

alsoallowforaholisticapproachto prioritizing

improvement

projects, since automobile, bicycle,

and

pedestrian improvementsto

roadways

tendtobe

made

14

Day

Level

of

service

can

be

a very useful conceptual

techniqueforquantifyingthe qualityofatransportation

facility.

LOS

does

have

its

drawbacks

as a quality

measure,though.

Depending on what

characteristicsare

used to determine level ofservice, the results can be

very biased orskewed. Traditionally,

LOS

has been

usedto describe the flow of

motor

vehicle trafficand

levelofcongestion

on

roads. Here,however,thisconcept

has

been

successfullyappliedina

way

thatdetermines

service quality for

modes

oftransportation otherthan

private

motor

vehicles.

These

methods

are not based

on

capacity

and

traffic flow, as the

highway

LOS

methods

are, but instead

on

environmental

characteristics, accessibility,

and

other diversemeasures

of

servicequality ratherthansimplyeaseofuse.

Whether

these

methods

adequatelycaptureallthevariablesthat

affectthe qualityofa transportationfacilityisdebatable,

but they

do

at least get

beyond

the simple traditional

notion

of

demand/capacity-basedlevel

of

service.

The

level

of

service

models

usedinthis analysis

were

developedfor

many

purposes.

The

BLOS,

BCI,

PLOS,

and

Fruin

methods were

developedlargely todetermine

the

adequacy

ofpedestrian

and

bicyclefacilitiesalong

arterial

highways

and

other

main

roads, similarto the

Highway

Capacity

Manual

method

of calculating

automobile

LOS

(whichis generally appliedtomajor

streetsas partofthemetropolitan planningprocess).

The

TLOS

route spreadsheet

method

determines

the

adequacy of

transit service frequency

and

hours

of

service (although thefull

TLOS

method

alsoaccounts

forenvironmentalfactorssuchasthesidewalknetwork

and

density ofresidents

and employees

in an area).

Necessarily, these

methods

arenot abletoaccount for

allfactors thatinfluencethe qualityofservice

on

agiven

transportation facility.

The

differencesin

LOS

scores

derived

from

the different

methods

usedin thisanalysis

show

thatthere is

some

need

forintegrationoffactors

and

methods

to

determine

a standard

method

for

computing

levelofservicefor alternativetransportation

facilities.

While beyond

thescopeofthispaper, future

researchcouldbe

done

todetermine whetheradditional

factorscouldbe

added

tothesecalculations,as wellas

determine

how

to integratethese various

methods and

thefactors usedineachmethod.

References

Federal

Highway

Administration.

"The

Bicycle

Compatibility Index:

A

Level

of

Service Concept." ImplementationManual. ReportNo.

FHWA-RD-98-095. 1998.

Florida

Department

ofTransportation.

2002

Quality/

Level

of

Service

Handbook.

FDOT:

Tallahassee,2002.

Hoehner,

Christine M., et al. "Opportunities for

Integrating

Public

Health

and

Urban

Planning

Approaches

to

Promote

Active

Community

Environments."

American

Journal

_of

Health Promotion.

Vol. 18,No. 1 (September/October2003).

Jaskiewicz, Frank. "PedestrianLevel

of

Service

Based

on

Trip Quality." Transportation

Research

Board

E-Circular

E-C019:

Urban

Street

Symposium.

Transportation Research Board:

Washington

D.C.,

no

date.

KittelsonandAssociates. TransitCapacity

and

Quality

of

Service Manual. Transportation Research Board:

Washington

D.C.,January 1999.

Kittelson

and

Associates,etal. TransitLevel

_of

Service

(TLOS)

Software Users Guide.

FDOT:

Tallahassee,

Level

of

Service

Measures

for Biking 1

5

Laiidis,Bruce,etal.

"Modeling

theRoadside

Walking

Environment:

Pedestrian

Level

of

Service."

Transportation Research Record,

No.

1773(

200

1 ):

82-88.

Levine.

Jonathan

and

Yaakov

Garb.

"Congestion

Pricing's

Conditional

Promise:

Promotion

of

Accessibilityor Mobility?" TransportPolicy.2002.

Meyer. Michael

D.

and

Eric J. Miller.

Urban

Transportation Planning.

Second

Edition.

McGraw:

Boston,2001.

Rodriguez,Daniel. "TransitService Characterization:

Key

Points." Lecture,January21.2004.

Ryus, Paul, et al.

"Development

ofFlorida's Transit Level-of-ServiceIndicator." TransportationResearch

Record

No. 1731 (2000):123-129.

Steiner,

Ruth

L. et al. "Refinement ofthe

Roadway

Segment

Level of

Service."

FDOT:

Tallahassee,

December

2002.

Transportation Research Board. Higinva}- Capacity