2009 Highway Mobility Report: Vehicle Miles Traveled
2009 Highway Mobility Report: Vehicle Miles Traveled
Staff Draft ‐ June 2009
can be nearly two miles and extend travel time above the speed limit by roughly 15 minutes. As
noted above, the intersection with Georgia Avenue is scheduled for capacity improvements, and
ultimately, a grade‐separation.
• Veirs Mill Road (MD 586) from Georgia Avenue (MD 97) to First Street/Norbeck Road (MD 28).
This is one of the most congested corridors in the County, largely because it crosses several east‐
west arteries and runs in between the heavily traveled MD 97 and MD 355 corridors. The
intersections at Twinbrook Parkway (ranked eighth in this year’s most congested list) and First
Street both exceed their LATR standard. Travel time along this corridor is periodically marked by
delays at major intersections including University Boulevard, MD 185, Randolph Road, Aspen Hill
Road and First Street. As noted above, the intersection at First Street is part of a State Highway
Administration intersection improvement study. The Veirs Mill Road corridor is currently under
study as a candidate for Bus Rapid Transit.
III. TRANSPORTATION TRENDS
Vehicle Miles Traveled
National trend data collected by the Federal Highway Administration indicate that the nation is in the
midst of a decline in national vehicle miles traveled (VMT) that began in 2005 (U.S. DOT Traffic Volume
Trends, years 2006‐2009). February 2009 VMT data indicate a decline of 1.9 billion vehicle miles, or 0.9
percent, when compared with February 2008. A slumping economy and fluctuating gas prices are likely
causes of the national decline. This national decline in VMT is not uniform across the country, however.
Depending on the time of year, and seasonal flows, quadrants of the country experience different
changes in VMT. While the South Atlantic States (a group of eight states and the District of Columbia,
which includes Maryland) saw a VMT decline of 2.2 percent from February 2008 to February 2009, the
North Central states (12 states in the upper Midwest) saw an increase in VMT of 1.3 percent.
Both the state of Maryland and Montgomery County are affected by the recent downturn in the
economy, and VMT analysis for these areas also show recent declines. Job losses, a common indicator
of economic malaise, began to appear in Montgomery County in 2007. An estimated 4,800 jobs have
been lost in the County since the end of 2006 (MNCPPC Research Team estimates, April 2009). Figure
3.1 provides a comparison of the decline in VMT at the national, state and county level. Although
Maryland VMT has fluctuated somewhat over the past three years, Montgomery County VMT data
reveal decreases in the past two years. Figure 3.2 depicts Montgomery County VMT totals since the
year 2000 and illustrates the dramatic decrease in VMT since 2005.
Page | 48
20
009 Highway Mobilitty Report
Staff Draft ‐ June 20
009
1: Year to Yeaar Percent Ch
Figure 3.1 hange in Vehicle Miles Traaveled for the
e United State
es, Maryland
d and
Montgom mery County
Figure 3.2
2: Total Vehiccle Miles Travveled (VMT) o
on State High
hways in Mon
ntgomery County, MD
Pagee | 49
2009 Highway Mobility Report
Staff Draft ‐ June 2009
Pedestrian Counts
Traffic counts conducted for the 2009 HMR by MCV, Inc. included pedestrian counts as well as vehicular
counts where feasible. Pedestrian counts for 111 intersections have been loaded into the MNCPPC
intersection count database and serve as a base for pedestrian analysis, particularly in relation to other
transportation systems and trends. The highest 30 of these pedestrian counts are displayed in Table 3.1.
Data displayed here is for the combined morning and afternoon peak travel times (6:30 – 9:30 am and
4:00 – 7:00 pm). Pedestrian counts conducted as part of this study in 2009 were recorded at
intersections that were a high priority for the County from a vehicular standpoint (due to lack of
previous data, interest in ongoing development nearby, etc.), with secondary regard to pedestrian
connectivity or access. Therefore, pedestrian counts provided in Table 3.1 do not represent locations of
the highest pedestrian volume county‐wide or key pedestrian crossing locations. The pedestrian to
vehicle ratio is also represented in the table, to provide context for the total traffic volume at a given
intersection.
Of the 111 locations sampled for pedestrian data, 30 had counts over 100 during the peak hours on the
date of sampling, while just four had counts over 500 and two had counts greater than 1,000
pedestrians. The busiest intersection, from a pedestrian count standpoint was at where 1,841
pedestrians were counted. Veirs Mill Road (MD 586) at Randolph Road and East West Highway (MD
410) and 16th Street (MD 390) also had high pedestrian volumes. The Colesville Road (US 29) and East
West Highway (410) intersection also had the highest pedestrian to vehicle ratio of 0.098, followed by
the intersection of Carroll Avenue (MD 195) and Laurel Avenue with 0.066. Few counts were done
inside of central business districts were pedestrian volumes, and possibly ratios, would be higher.
The map in Figure 3.3 provides a picture of the locations of the 111 pedestrian counts, with grouping by
total pedestrian volume. Although a distinct pattern of higher pedestrian volumes in down county areas
is clearly displayed on the map, several intersections in Rockville and Gaithersburg have counts that are
above 150 pedestrians, and in the top 20 percent of county‐wide intersection counts.
Page | 50
2009 Highway Mobility Report
Staff Draft ‐ June 2009
Table 3.1: Total (AM + PM) Peak Period Observed Pedestrian Crossing Volumes
Page | 51
2009 Highway Mobility Report
Staff Draft ‐ June 2009
Figure 3.3: 2009 Peak Hour Pedestrian Counts
Page | 52
2009 Highway Mobility Report
Staff Draft ‐ June 2009
Ride‐On Bus Service
Montgomery County bus service, called Ride‐On, is comprised of an extensive network of local, feeder,
circulator and express routes that compliment the regional and line service provided by the Washington
Metropolitan Area Transit Authority (WMATA). Figures 3.4 and 3.5 illustrate the breadth of these
routes. These maps also depict fiscal year 2006 ridership (the most current data available) on these bus
lines. Figure 3.4 depicts average daily (weekday) ridership while Figure 3.5 depicts ridership per revenue
mile. In total, the Ride‐On system operates roughly 80 routes. Ride‐On service is a key factor in mobility
within the County and registered more daily (average) boardings in fiscal year 2008 than the
Montgomery County Metrorail stations. In fiscal year 2008, average daily ridership was approximately
95,000.
The County’s heaviest ridership is on routes that serve areas in a line running north‐west from Silver
Spring along the I‐270 corridor, stretching as far north as Germantown. Ridership is also heavy in the
south eastern portion of the County near White Oak. In order, the three routes with the highest
average daily ridership in fiscal year 2006 were Germantown to Rockville, Silver Spring to Langley Park,
and Shady Grove/Montgomery College to National Naval Medical Center. The proximity of these routes
to Metrorail stations underscores the importance of bus service beyond and between Metrorail stations
as an important part of the transit network. Figure 3.5 depicts a somewhat different picture of
ridership, with the additional consideration of cost. In this map it is clear that although ridership along
the I‐270 corridor is high, of these routes travel long distances and are not as productive on a rider per
revenue mile basis. Therefore, these longer routes are not as cost effective as some other routes.
Routes beyond the I‐270 corridor such as the route to Olney, carry the least riders per revenue mile,
while routes between Bethesda and Silver Spring are the most cost effective.
Frequency of Ride On bus service for 2008 is depicted in Figure 3.6. The majority of the County’s routes
provide peak hour service every 20 – 30 minutes. Busy routes from Germantown south along I‐270, and
routes serving the central business districts in and around I‐495, offer service with headways of 15
minutes or less.
Linkages and patterns between pedestrians and buses are important as riders must walk to and from
bus stops at either end of their trip. Intersections with heavy pedestrian activity reported in the
previous section, particularly those along Veirs Mill Road, in Silver Spring and Rockville are likely
connected to extensive bus service in those areas. Connections between bus and rail service in the
county are also documented in the number of lines near metro stations in Figure 3.6. Metrorail
ridership is discussed in the next section of the report.
Page | 53
2009 Highway Mobility Report
Staff Draft ‐ June 2009
Figure 3.4: Average Daily Ride On Ridership (2006)
Figure 3.5: Ride On Ridership by Revenue Mile (2006)
Page | 54
20
009 Highway Mobilitty Report
Staff Draft ‐ June 20
009
Figure 3.6
6: Peak Hour Ride On Bus Headways (2
2008)
Pagee | 55
2009 Highway Mobility Report
Staff Draft ‐ June 2009
Metrorail Analysis
Assessment of Metrorail ridership is another key component of mobility in Montgomery County. Similar
to road and intersection analysis presented in the previous sections of this document, measures of
volume and time of travel are also useful for Metrorail data. These two types of data are presented for
each of Montgomery County’s Red Line stations in terms of total ridership (entries and exits) in Figure
3.7 and ridership by time of day in Figure 3.8A (entries) and 3.8B (exits).
Unlike trends in vehicle miles traveled, ridership on Metrorail has increased, by approximately five
percent, between February 2006 and February 2009. Figure 3.7 depicts the varying levels of ridership
increases for average weekday ridership in Montgomery County. Two types of change are depicted in
this figure. First, ridership at the four most heavily used stations; Silver Spring, Shady Grove, Bethesda
and Friendship Heights, exhibit seasonal peaks (and troughs) with highest ridership in the summer
months of June, July and August and significantly lower volume in the winter months. At the Shady
Grove station, ridership increased from 28,933 in July 2006 to 30,952 in July 2009, a seven percent
change. This peak travel period is important because it represents maximum stress on the capacity of
the stations, and is ahead of the year to year increase in ridership of the entire system. A second change
depicted in Figure 3.7 is the uneven increase in ridership at each station. For example, average daily
ridership in Silver Spring was 27,285 in February 2006 and 27,708 in February 2009 while ridership at the
Rockville station was 8,594 and 10,000 for the same time periods. This 16 percent increase in ridership
in Rockville is roughly three times the percent change in total ridership for all Montgomery County
stations.
The time of day patterns in Figures 3.8A and 3.8B depict the predominantly residential nature of the
areas surrounding many Metrorail stations in Montgomery County. Metrorail stations located in
communities with a balance of residential and employment/commercial uses have an equal pattern of
entries and exits throughout the day. A good example of uneven ridership is the Medical Center station
where the adjacent major employment sites of the National Institute of Health and Naval Medical
Center draw large numbers of riders from elsewhere in the system who exit at the station in the
morning and then head onto Metrorail in the afternoon to return home. Shady Grove is a station that
exhibits ridership that is nearly the opposite of that at Medical Center. Roughly three quarters of entries
to the Shady Grove Station occur during the morning peak hours, while approximately two thirds of the
exits from the station occur in the afternoon peak hours. Entry and exit data are similar for the majority
of the Red Line stations in the County with a few exceptions. The central business districts of Silver
Spring, Bethesda and Friendship Heights exhibit more balanced ridership due to their large employment
base.
Page | 56
20
009 Highway Mobilitty Report
Staff Draft ‐ June 20
009
Figure 3.7:
: Metrorail Riidership 2006
6 to 2009
Pagee | 57
20
009 Highway Mobilitty Report
Staff Draft ‐ June 20
009
Figure 3.8A: Montgomeryy County Metrorail Entrie
es (February 2
2008)
Figure 3.8B: Montgomeryy County Mettrorail Exits ((February 200
08)
Pagee | 58