CALIFORNIA HEAVY DUTY
TRUCK TRAVEL SURVEY
ON SELECTED SITES
DECEMBER 2001
CALIFORNIA HEAVY DUTY
TRUCK TRAVEL SURVEY
ON SELECTED SITES
GOVERNOR GRAY DAVIS
STATE OF CALIFORNIA
MARIA CONTRERAS- SWEET, SECRETARY
BUSINESS, TRANSPORTATION AND HOUSING AGENCY
JEFF MORALES, DIRECTOR
DEPARTMENT OF TRANSPORTATION
DIVISION OF TRANSPORTATION SYSTEM INFORMATION
OFFICE OF TRAVEL FORECASTING AND ANALYSIS
STATEWIDE TRAVEL ANALYSIS BRANCH
DECEMBER 2001
TABLE OF CONTENTS
Page
Executive Summary i
Acknowledgments iii
Summary of Survey Data, Charts iv
Summary of Region to Region Daily Truck Trips, Maps v
1.0 INTRODUCTION 1
1.1 Purpose and Objectives 1
1.2 Project Organization 1
2.0 STUDY APPROACH AND SAMPLING STRATEGY 3
2.1 Overall Study Approach 3
2.2 Sample Size Requirements 3
2.3 Site Selection 4
2.4 Supplemental Data Sources 7
2.5 Data Collection Methods and Instruments 9
2.6 Geocoding 19
3.0 SURVEY STRENGTHS AND WEAKNESSES 19
3.1 Sampling Design and Survey Execution 19
3.2 Data Deficiencies 20
4.0 RECOMMENDATIONS 21
4.1 Collect Supplemental Survey Data 22
4.2 Increase Survey Productivity 23
i
- EXECUTIVE SUMMARY -
PURPOSE OF STUDY
The overall goal of the Heavy Duty Truck Travel ( HDT) Survey was to collect representative
truck travel data for selected sites in California. An integral aspect of the project was a
literature search and inventory study of freight data survey collection methods. Additional
study objectives included 1) Development of procedures to acquire basic truck travel data 2)
Actual conduct of data collection, 3) Collection of data useful to the identification of
relationships between economic activity and truck travel patterns, and 4) Collection of data
for analysis of commodity flow throughout California.
METHODOLOGY
To achieve the objectives outlined above; data was collected on the following items through
direct, in- person interviews with truck drivers.
 Freeway/ Route Traveled
 Direction of Travel
 Interview Facility Location
 Truck Type
 Truck Body Style
 Number of Axles
 Gross Vehicle Weight
 Percent Full
 Primary Commodity Carried
 Last Stop Location
 Last Stop Facility Type
 Last Stop Activity ( load/ unload)
 Last Stop Departure Time
 Next Stop Location
 Next Stop Facility Type
 Next Stop Activity ( load/ unload)
 Next Stop Arrival Time
 Distance From Last to Next Stop
 Total Number of Stops for Day
 Total Distance Traveled for Day
 Last State for Refueling
 Presence of Hazardous Material Sign
The survey data collected took place though assisted interviews of 8,287 truck drivers at
a total of 33 sites throughout the state. The sites included California Highway Patrol
weigh stations, agricultural inspection stations and roadside rest areas.
The interviews were conducted over the busiest periods of the day for each site, ranging
from eight to sixteen hours based on the site’s hours of operation and traffic volume. To
expand the data to a 24- hour time period, videotaping was conducted over a full 24 hours
at each of these locations. Since the amount of data that can be collected by video is
limited to what can be observed, the 24- hour video data was comprised solely of truck
type, truck body style, and number of axles for each observed vehicle.
A project team incorporating a wide range of expertise, practical experience and
resources was assembled to achieve the objectives of the study. Project team members
included Strategic Consulting & Research, Cambridge Systematics, K. T. Analytics, and
ATD Northwest.
ii
A Technical Advisory Committee comprised of members from various Caltrans
Divisions and District offices, the California Highway Patrol, several Metropolitan
Planning Organizations, the Air Resources Board, the California Energy Commission, the
Federal Highway Administration, and the California Trucking Association guided the
consultant team.
The project was conducted in five stages that included a literature search, sampling
design, survey instrument design, data collection, and reporting.
RESULTS
Overall, the results from the HDT survey addresses the objectives defined for the project.
These results are useful for regional and system planning and provide a foundation to
expand heavy- duty truck travel data collection activities for forecasting purposes. The
results also provide, specifically, the following statistical abstract information on selected
state highway routes between four major metropolitan regions in California:
 Descriptions of HDT travel activities by truck body type, geographic area, axle group
and other factors.
 Descriptions of travel patterns and trip definitions based on last/ next stop information,
trip time, and trip distance.
 Percent of heavy- duty truck travel between and within the four major metro regions.
 Percent of truck travel activity by axle groups.
 Percent of truck weight by cargo type.
 Percent of last/ next stop by type of facility ( e. g. port, rail facility, truck terminal, etc.).
 Distance traveled in and between most recent and next stop.
Collection of statewide heavy- duty truck travel activity by intercept survey represents
several challenges. California maintains about 15,200 miles of state highways with a
truck population in excess of 661,000 vehicles. Potential exists for double counting on
some routes, and missing counts on others. In order to obtain enough data to provide
complete origin/ destination tables for forecasting purposes, a more comprehensive survey
will be needed to enhance the results presented in this report.
iii
ACKNOWLEDGEMENTS
Strategic Consulting & Research served as the prime contractor for this project. Strategic
Consulting & Research managed all field data collection activities, survey design,
geocoding, data analysis and overall project management.
Cambridge Systematics conducted the Literature Search and Inventory study for this
project, as well as, the Site Selection Plan and Recommended Model Design Plan.
K. T. Analytics provided statistical expertise in the overall development of the Site
Selection Plan and in analysis of results.
The Project Team greatly appreciates the outstanding cooperation and support of
the California Highway Patrol in the conduct of this study.
This survey project was sponsored by: Division of Transportation System Information –
Office of Travel Forecasting and Analysis – Statewide Travel Analysis Branch.
Project funding was provided by the California Department of Transportation.
iv
Summary of Survey Data,
Charts
The following charts, A through M, will illustrate the survey data weighted and expanded
according to the total truck counts recorded by the video camera.
It should be noted that those trucks that travel on routes not covered by the interviewing
locations would not be included in these totals. Without a completely exhaustive
coverage of the state, it is likely that there is additional truck travel that has not been
included in this survey and consequently in the following charts.
A
Chart A: Truck Type
TRUCK TYPE
SINGLE- UNIT
WITH TRAILER
6%
SINGLE- UNIT
2%
SINGLE- TRAILER
84%
MULTIPLE-TRAILER
8%
B
Chart B: Truck Body
TRUCK BODY
FLATBED
15%
CONTAINER
5%
OPEN TRAILER
8%
TANKER
5%
SPECIALIZED/
OTHER
5%
HOPPER
2%
VAN
61%
C
Chart C: Number of Axles
4% 5%
90%
1%
0%
20%
40%
60%
80%
100%
3 AXLES 4 AXLES 5 AXLES 6+ AXLES
NUMBER OF AXLES
D
Chart D: Truck Weight
21%
6%
30%
44%
0%
10%
20%
30%
40%
50%
60%
EMPTY UNDER 33,000
POUNDS
33,000 TO 60,000
POUNDS
OVER 60,000 POUNDS
WEIGHT OF TRUCK INCLUDING CARGO
E
Chart E: Last Stop Facility Type
WHAT TYPE OF FACILITY DID YOU COME FROM ?
2% 1% 9%
2%
17%
2%
8%
23% 3%
7%
16%
8% MARINE PORT
RAIL FACILITY
TRUCK TERMINAL
RESIDENTIAL
MANUFACTURING
WHOLESALE
RETAIL
GOVERNMENT
DISTRIBUTION CENTER
AG PROC/ PACK
MOTEL/ REST AREA
OTHER
F
Chart F: Last Stop Activity
27%
3%
24%
47%
0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50%
NEITHER ( START OF DAY)
BOTH
UNLOAD
LOAD
DID YOU. . .? AT YOUR LAST STOP?
G
Chart G: Next Stop Facility Type
WHAT TYPE OF FACILITY ARE YOU GOING TO ?
2%
1%
1%
13%
3%
12%
2%
10%
24% 3%
6%
13%
9% 1%
MARINE PORT
RAIL FACILITY
AIR CARGO
TRUCK TERMINAL
RESIDENTIAL
MANUFACTURING
WHOLESALE
RETAIL
GOVERNMENT
DISTRIBUTION CENTER
AG PROC/ PACK
MOTEL/ REST AREA
OTHER
DON’T KNOW
H
Chart H: Next Stop Activity
WILL YOU . . .? AT YOUR NEXT STOP
3%
52%
16%
29%
0% 10% 20% 30% 40% 50% 60% 70%
NEITHER ( END OF DAY)
BOTH
UNLOAD
LOAD
I
Chart I: Distance Between Last and Next Stops
24%
18%
16%
8%
7%
5% 6%
16%
0%
5%
10%
15%
20%
25%
30%
< 75 75- 125 126- 200 201- 250 251- 300 301- 350 351- 400 > 400
DISTANCE ( IN MILES) BETWEEN MOST RECENT
AND NEXT STOPS
J
Chart J: Hazardous Materials Signage
DO YOU HAVE HAZARDOUS MATERIALS SIGNAGE?
YES
4%
NO
96%
K
Chart K: Total Number of Stops Today
0%
5%
10%
15%
20%
25%
30%
35%
2 3 4 5 MORE THAN 6
TOTAL STOPS FOR DAY
L
Chart L: Total Miles Driven Today
3%
6%
15%
10%
12% 9%
13%
31%
0%
5%
10%
15%
20%
25%
30%
35%
40%
< 75 75- 125 126- 200 201- 250 251- 300 301- 350 351- 400 > 400
TOTAL NUMBER OF MILES DRIVEN TODAY FROM
START TO FINISH
M
Chart M: State for Last Purchase of Fuel
STATE FOR LAST FUELING OF TRUCK
CALIFORNIA
78%
ARIZONA
10%
NEVADA
5%
OREGON
3%
OTHER
3%
MEXICO
1%
v
Summary of Region to Region Daily Truck Trips,
Maps
The following four maps, N through Q, illustrate the daily heavy duty truck trips between
the four major metropolitan regions in California. The region to region trip pairs as
shown in the following maps were derived from the actual truck driver surveys and were
expanded using the video tape counts.
1
1.0 Introduction
1.1 Purpose and Objectives
This report documents the findings and procedures of the California Heavy Duty Truck
Survey. The overall goal of this project was to collect representative truck travel data for
a subsequent forecasting model for statewide interregional heavy- duty truck travel in
California. An integral aspect of the project was a literature search and inventory study
of freight data survey collection methods. Findings from the search and inventory study
were instrumental in identifying a recommended model structure suited to meet
objectives of modeling and assessing impacts of line haul or inter- city/ inter- county
heavy duty truck movements on California state highways. Additional study objectives
included 1) Development of procedures to acquire basic truck travel data 2) Actual
conduct of data collection, 3) Collection of data useful to the identification of
relationships between economic activity and truck travel patterns, and 4) Collection of
data for analysis of commodity flow throughout California.
1.2 Project Organization
A) Team Approach
A project team incorporating a wide range of expertise, practical experience and
resources was assembled to achieve the objectives of the study. Project team members
included Strategic Consulting & Research ( SCR), Cambridge Systematics, K. T.
Analytics and ATD Northwest. SCR provided overall project management.
B) Model and Survey Design
Cambridge Systematics conducted model research and design efforts including the
Literature Search and Inventory Study and the resulting Recommended Model Design
Plan. Survey instrument design and procedures were developed by SCR with the active
participation of Cambridge Systematics, K. T. Analytics, and Caltrans.
C) Field Data Collection
Survey data collection took place through assisted interviews of 8,287 truck drivers at the
California Highway Patrol weight stations, agricultural inspection stations and roadside
rest areas throughout the state. These locations were selected for inclusion in the survey
through the Model Design Plan. Truck counts by truck body type were also collected at
these sites1 through videotape technology and used for the survey data expansion process.
Survey collection and management was conducted by SCR with video taping activities
conducted by ATD Northwest.
1 Videotaping was actually conducted on the freeway at the closest possible lighted location so that accurate
24- hour counts on the road could be conducted even when the inspection site is closed.
2
D) Data Base Management
Raw survey data was delivered to SCR and incorporated into the database. SCR
geocoded approximately 16,500 trip ends from 8,287 field surveys, as well as
information on truck type, body type, number of axles, truck weight, commodity carried
and other data, as requested by Caltrans. Additionally videotapes documenting truck
activity by truck- type, body- type and number of axles were obtained at every survey
location.
E) Data Analysis
The data for the spring wave was expanded from the in- person surveys to total daily truck
volume using the video counts for each survey location. Since the video coverage was
comprised of a one- tenth of each hour ( totaling four hours per day), the raw video counts
were multiplied by six to expand the data to a 24- hour day.
3
2.0 Study Approach and Sampling Strategy
2.1 Overall Study Approach
The primary objective of the project is to provide a database of heavy duty truck travel
that includes last stop and next stop data for the trip segment in question, and which is as
representative as possible of all heavy duty truck travel throughout the state. To
accomplish this, the methodology was comprised of two major components; in person
interviews of truck drivers, and video recording of truck traffic at the same locations over
a 24- hour period of time.
The in- person interviews of truck drivers covered between eight and sixteen hours based
on the relative volume of the sites with the objective of providing the broadest coverage
at the heaviest traffic locations. Interviewing at all sites was scheduled for the heaviest
volume portion of the day. The data collected in these interviews included last and next
stops, types of facilities at both ends, number of axles, truck body type and truck
configuration. There were also questions addressing the extent to which the truck was
full, cargo being carried, and where the truck last refueled.
The videotaped data provided the number of trucks that passed the target location over
the full 24- hour time period. It included only the number of trucks with each one
identified by number of axles, body type, and truck configuration.
The interviews provided the distribution of last stops and next stops for the data
collection point for the time period when the interviews were collected. This data was
then expanded to the full 24 hours using the video data truck counts. The truck counts,
which were comprised of 10 minutes per hour ( totaling 240 minutes or four hours per day
at each location), were expanded to the full day by multiplying by six. The subsequent
results that were used for both the last stop / next stop tables and the cross- tabulated data
were weighted using this total number of trucks per day ( typical Tuesday/ Wednesday/
Thursday). As a result, all data that is presented is in the form of total truck volume on a
typical Tuesday, Wednesday or Thursday weekday basis.
2.2 Sample Size Requirements
The sample size required for a sample of heavy- duty truck trips depends on the intended
use of the data and the parameter( s) being estimated from the sample. In a typical
statistical estimation exercise, the goal of the data collection is to estimate a population
parameter( s) with a given degree of precision at a given confidence level. If, for
example, the population to be measured is all regional truck trips on an average weekday,
variables of interest could include the average payload, or percent of trips that are empty.
Calculation of a sample size to estimate a population mean, such as average payload,
requires some knowledge of the expected mean and variation of the variable in question.
For the California heavy- duty truck survey, the primary intended use of the data is for
estimation of truck demand models. Rather than directly estimating a population
4
parameter, we will be estimating a relationship between the population and some
exogenous variable. For example, a trip generation rate is the relationship between
employment in a zone and the number of truck trips generating there. The chief data
collection concerns thus become collecting enough data to estimate statistically
significant model coefficients and avoiding bias in the sample.
With respect to the first data collection concern, a rule of thumb often used in travel
demand modeling is that trip generation models require a minimum of 200 observations
per stratification to estimate statistically significant coefficients. For example, to
estimate a separate trip generation model for each of three vehicle classifications would
require at least 200 observations in each vehicle category. Similarly, separate models by
commodity category, or special generator type would require 200 observations in each
cell.
With respect to the second data collection concern, it is quite possible to estimate
“ statistically significant” ( as measured by the “ T” or other statistics) coefficients in
travel demand models that are wrong due to bias in the estimation data set. The bias
would not usually become apparent until the models are applied and give anomalous
results ( although one could compare the survey set to national data to assure that it does
not stray to far from the national norm). For the California truck survey, avoiding bias
means striving to capture and equally represent truck trips with variation in any variable
that we think might influence truck travel behavior: geographic location, time of day,
type of vehicle, season and so forth.
Because we do not yet know the exact form that the truck travel models will take, it is not
possible to explicitly calculate the required sample size on this basis. Typically, model
development is an iterative process starting with an initial form that is than influenced by
the limitations and potential of the data set. However the number of observations
projected for this survey ( about 7,500) should provide for a good range of model
development options.
2.3 Site Selection
Having ensured that the total sample size is likely to be sufficient for the intended
application of the data, the project team focused on developing a list of survey sites and
data collection procedures to minimize potential bias.
5
Figure 1: Interview Site Map
6
The basic criterion for site selection was to obtain the broadest geographic, time of day
and seasonal coverage within the constraints of the available study resources. The
constraints included the following:
 Potential sites for data collection- these were initially limited to CHP and agricultural
inspections stations but later expanded to include rest stops,
 Hours of operation- safety concerns dictate the CHP personnel be present at each
interview location and the CHP is only able to provide a certain number of hours for
assistance at each site; and
 The data collection budget.
7
Given the above criteria and constraints, the project team selected sites by first
identifying which locations would give general geographic coverage. Using visual
inspection, sites that would capture trips crossing screenlines, trips entering or exiting
major metropolitan areas, and trips at the state border were selected. The project team
also took into account input received from Technical Advisory Committee members and
Caltrans regarding locations that should be included.
These sites were then arrayed by truck AADT as a guide to relative importance and the
likely productivity at each site. Several alternative lists of sites were developed with the
aim of maximizing geographic coverage and second seasonal variation. Using
assumptions regarding labor productivity, it does not appear possible to obtain both
sufficient geographic coverage and total coverage of seasonal variation within the given
project budget. In other words, it would not be possible to visit each of the recommended
sites four times ( once for each season). The recommended approach involves obtaining a
comprehensive data set for a single season ( spring) with seasonal data collection at a
limited number of sites.
Three alternative surveying options were designed to assess the ability to provide optimal
coverage in the spring and offspring seasons. These included options with 75%, 56%,
and 38% for spring, with the balance being distributed to the offspring seasons. It was
only with the first option with 75% spring surveying that we were able to achieve a
representative sample of statewide heavy- duty truck travel. The other two options
resulted in reducing spring coverage beyond the point needed to provide representative
coverage. In addition, although options two and three increase off- spring surveying, it is
still not sufficient to provide the level of coverage that would be needed to develop
independent seasonal models.
2.4 Supplemental Data Sources
The project team has investigated the availability of truck origin- destination data that
might be used to supplement the data collected as part of this project, to date only data
collected by Caltrans District 11 has been identified as a promising supplemental data
source. These data include approximately 2000 survey records collected in Imperial
County at Calexico, and in San Diego County at Tecate, Otay Mesa, I- 15 at the Riverside
County border, and I- 5 at the Orange County border, from August through December of
1995. The data collected appear to be largely compatible with the questionnaire designed
for this project. With the exception of Tecate, these locations coincide with sites selected
for the statewide survey.
8
Figure 2
Reliance on the combination of data from different sources is a risky strategy for model
development. For this reason, the project team does not recommend substituting the
District 11 data for sites that would be surveyed as part of this project. Instead, the data
should be viewed as a supplemental data source. Since none of these southern California
sites are selected for offspring data collection, the District 11 data can provide some
information on variation between fall and spring characteristics at the sites.
Additional supplemental origin- destination data may be available from the neighboring
states of Arizona and Oregon. Oregon, in fact, conducted a truck survey at its ports of
entry in 1997. Data from the border crossings would be especially useful since this
project will only be able to intercept inbound traffic at the borders.
9
2.5 Data Collection Methods and Instruments
As previously discussed, data needs identified to support both truck generation and trip
distribution models included 1) vehicle classifications and counts to expand survey data
for a wide range of modeling purposes and 2) truck intercept surveys from approximately
7,500 trucks, maximizing geographical coverage and seasonal variation within project
constraints. Specifically the following data needs were identified for the truck intercept
survey element:
 Interstate/ state route
 Direction of travel,
 Date
 Time of travel
 Vehicle classification
 Truck body type
 Gross vehicle weight estimates
 Empty vehicle weight
 Cargo carrying status
 Commodity carried ( by type consistent with SCTG codes for origin and
destination)
 Origin and destination ( of surveyed trip segment)
 Number of loading/ unloading stops on the one- way journey and total day of travel
 Other travel behavior data
10
Geographically, 32 site locations were selected for surveying including 26 CHP
Inspection/ Platform stations, 4 Agricultural Inspection facilities and 2 Truck Rest Areas.
Of these, 19 were surveyed in both directions of travel and 13 were surveyed in one
direction ( with the exception of Two Rock which was closed). One direction of travel
was surveyed when only one direction of travel was covered by CHP or Agricultural
inspection facilities. All 32 sites were surveyed during the spring survey wave, with the
exception of Cajon South and Two Rock, which were closed. Additionally, five sites
were surveyed during all four seasons, one site was surveyed three seasons, and two sites
were surveyed in one season.
11
Table 1: Survey Facilities and Completed Survey Counts
SPRING WAVE
FACILITY SURVEYS COLLECTED FACILITY SURVEYS COLLECTED
ANTELOPE - EAST 132 KEENE - EAST 104
ANTELOPE - WEST 137 LIVERMORE - EAST 242
BLYTHE - WEST 148 LIVERMORE - WEST 178
CACHE CREEK - WEST 89 NEEDLES - WEST 134
CAJON - NORTH 247 NIMITZ - NORTH 202
CAJON - SOUTH1 0 NIMITZ - SOUTH 113
CALEXICO - NORTH 132 OTAY MESA - NORTH 94
CARSON - NORTH 95 PERALTA - EAST 118
CARSON - SOUTH 79 PERALTA - WEST 141
CASTAIC - NORTH 260 RAINBOW - NORTH 160
CONEJO - NORTH 180 RAINBOW - SOUTH 134
CONEJO - SOUTH 164 RAINE - NORTH 97
CORDELIA - EAST 102 RAINE - SOUTH 120
CORDELIA - WEST 126 SAN ONOFRE - NORTH 155
COTTONWOOD - NORTH 113 SAN ONOFRE - SOUTH 180
COTTONWOOD - SOUTH 76 SANTA NELLA - NORTH 70
DESERT HILLS – EAST2 160 SANTA NELLA - SOUTH 63
DESERT HILLS - WEST 218 ST. VINCENTS - SOUTH 51
DONNER PASS - WEST 125 TERRA LINDA - NORTH 95
ENOCH - NORTH 117 WALNUT CREEK - NORTH 107
ENOCH - SOUTH 118 WALNUT CREEK - SOUTH 110
GILROY - NORTH 122 WHISKEYTOWN - EAST 99
GILROY - SOUTH 159 WILLITS - SOUTH 87
GRAPEVINE - SOUTH 137 WINTERHAVEN - WEST 140
HORNBROOK - SOUTH 95 YERMO - WEST 105
SUMMER WAVE
FACILITY SURVEYS COLLECTED FACILITY SURVEYS COLLECTED
CAJON - NORTH 122 DESERT HILLS - WEST 137
CAJON - SOUTH1 0 GRAPEVINE - SOUTH 122
CASTAIC - NORTH 117 LIVERMORE - EAST 128
DESERT HILLS - EAST2 0 LIVERMORE - WEST 131
FALL WAVE
FACILITY SURVEYS COLLECTED FACILITY SURVEYS COLLECTED
CAJON - NORTH 82 DESERT HILLS - WEST 111
CAJON - SOUTH 68 GRAPEVINE - SOUTH 145
CASTAIC - NORTH 104 LIVERMORE - EAST 162
DESERT HILLS - EAST2 12 LIVERMORE - WEST 63
WINTER WAVE
FACILITY SURVEYS COLLECTED FACILITY SURVEYS COLLECTED
CAJON - NORTH 64 DESERT HILLS - WEST 76
CAJON - SOUTH1 0 GRAPEVINE - SOUTH 74
CASTAIC - NORTH 80 LIVERMORE – EAST3 0
DESERT HILLS - EAST2 0 LIVERMORE - WEST 59
1 Facility closed with no viable alternative interviewing site
2 Unsafe interviewing conditions. One hundred sixty ( 160) surveys were collected during
the Spring wave to satisfy all interviewing waves.
3 Survey data quality was unacceptable. Re- interviewed during the Spring wave to account for shortfall.
12
Figure 3
13
Two methods of data collection were identified as best suited to obtain this information;
on- site interviews of truck drivers, and video camera recording. Logistics and
instruments for data collection under these methods are discussed below.
14
2.5.1 On- site Interviews
On- site interviewing of truck drivers at the selected survey site locations was easily
recognized as the only feasible approach of collecting the detailed information listed
above. The development of an efficient and productive survey form and adequate
training and procedural materials was more involved.
15
Two small- scale pre- tests or “ convenience” surveys were conducted prior to actual
surveying to test the value of preliminary survey instruments and procedures. The pre-tests
were conducted under conditions replicating those of the actual survey and
numerous changes to both the survey form and procedures were incorporated based on
their results.
Once site selections had been finalized, commanders at the selected facilities were
contacted to obtain permission for surveying and video taping activities. The
commanders were asked to designate a staff member to continuously direct trucks to a
designated area ( clearly marked by cones) where the surveyors would administer the
survey. Commanders were also asked to stagger staff break periods so that surveying
activities would not cease during staff lunches and breaks.
A team of two trained surveyors was assigned to administer surveys at each site using the
survey instrument shown in the pictures. The more experienced of the team members
was also charged with supervisory tasks such as periodically ( about every two hours)
reviewing survey forms for completeness and logic and dealing with any questions or
concerns of facility staff.
16
On the day of the survey, the survey team was sent to the facility to conduct the surveys.
On- route to the facility a team member would enter survey identification numbers on the
forms, identifying the order the survey was to be taken, the season, month, day, and
location. This was done to conserve time not always available during the survey day.
Upon arrival the team member designated for supervisor tasks would introduce
him/ herself to the commander and briefly review project forms and productivity targets
with the facility staff member assisting with the survey.
Once cones were set up to designate the survey area, the assisting CHP staff member
would direct trucks over to the surveyors. When the vehicle had come to a complete stop
the surveyors would first request the driver to set his brakes as a safety concern and
request the drivers participation in the surveys by reading the salutation on the survey
form. The driver would then be asked to shut off the vehicle to reduce the noise level.
During the next few moments the surveyor would enter visual information about the
truck onto the survey form such as, truck type, number of axles ( see figure 2). The
survey would then be conducted with the surveyor reading the question exactly as stated
on the form, with the following exceptions;
If the driver was uncertain about the vehicle's weight with cargo, the surveyor was
directed to prompt by asking if the weight was in certain ranges. The ranges used were
60,000 pounds+, 60,000- 33,000 pounds and under 33,000 pounds. These ranges were
selected because they are consistent with other major truck studies such as the SACOG
commercial vehicle study.
If a trucker would or could not provide an address or two cross streets he was asked for
the nearest landmark, a single cross street or a well- known building, and the town name
within a quarter mile of his next stop. If this was not offered any information offered was
taken down, such as town name.
Due to the uniformity of layouts and truck movements at CHP Inspection facilities the
survey area was always set up after the trucks had gone by the weigh scales where
vehicle inspections are typically performed. An exception was the Eastbound Desert
Hills facility. Space at this facility was very restricted and took place in a wide shoulder
adjacent to the exit lanes of the facility.
A very important element of this procedure was the use of a CHP or Agricultural
Inspection facility staff member to direct trucks to the surveyors. This was important
because it assured a safe process for directing the trucks and because the truck drivers
recognized the authority of the CHP or Agricultural Inspection staff. By design all of the
selected facilities had large and continuous flows of heavy trucks. It would not have been
feasible for anyone other than a uniformed professional staff member to safely or
effectively direct the trucks. Coincidentally the procedure may have aided survey
participation. Several truck drivers expressed relief that they had not been pulled over for
a vehicle inspection and were quite willing to participate in the survey.
17
Table 2: California Truck Travel Survey
1. Survey ID: _________________ 2. Date: ______ 3Time: ______ 4AM PM 5. Surveyor: ______________________
6. CHP/ AG 7Facility Name ____________________________________ 8Fwy/ Route: _______________________
Direction of Travel 9. N E S W 10Hazardous Materials Signage? Yes______ No _____ ( CHECK ONE)
11Truck Type: Single- Unit Single- Unit w/ Trailer Single- Trailer Multiple- Trailer 12Number of Axles: _____
( CIRCLE ONE) ( FILL IN)
13Truck body: Van Container Flatbed Tanker Hopper Open Trailer Specialized/ Other
( CIRCLE ONE)
- PLEASE SET YOUR BRAKES -
Good morning/ afternoon. We are conducting a survey of truck travel for the California Department of Transportation.
This information will be used to determine improvements to the California State Highway System. The interview
should take no longer than three minutes of your time.
14. Is the truck empty now? Yes_____ ( skip to # 18) No_____ ( CHECK ONE)
15What % of total capacity are you carrying now? _______% ( volume or weight)
16What is this truck’s weight including the cargo?_________( in pounds) If driver is uncertain prompt
17Would it be over 60,000 pounds? 60,000- 33,000 pounds? under 33,000 pounds?
What is the primary cargo being carried? ( normally carried if empty)_______________________ 18________
SCTG CODE
Where did the truck last stop to load, unload or start the day? 23Was this your Starting location? Yes ___ No ___
19_________________________________________________ 20 _________________________ 21______ 22_______
ADDRESS/ CROSS STREETS OR NEAREST INTERSECTION CITY STATE ZIP
24Did you load ____ or unload _____ at this location, or both? ____ neither ( start of day) ____? ( if start– skip arrive time)
What time did you arrive and depart that location? 25Arrive _________ 26AM PM 27Depart _________ 28AM PM
29What type of facility or terminal was that?___________________________________________( code letter or write in)
( port, rail yard, air cargo facility, truck terminal, residential, etc.)
Where will the truck stop next to load, unload, or end the day?
30_________________________________________________ 31________________________ 32_______ 33_____________
ADDRESS/ CROSS STREETS OR NEAREST INTERSECTION CITY STATE ZIP
34Will you load ____ or unload _____ cargo at this location, or both? ______ neither ( end of day)________?
35At what time will you arrive there?__________ 36AM PM
37What type of facility or terminal is this?_____________________________________________( code letter or write in)
( port, rail yard, air cargo facility, truck terminal, residential, etc.)
38 What is the distance between the most recent and next stops that we just identified?________( miles)
39 How many total stops will you make today for loading or unloading including your starting and ending points?___
40How many total miles will you drive the truck today from start to finish?________( miles)
41In which state did you last fuel your truck? CA_____ NV_____ AZ_____ OR_____ MX_____ Other_____
THANK- YOU FOR PARTICIPATING IN THE SURVEY!
** FACILITY CODES Survey Revised 9/ 1/ 99
A. Marine port B. Rail facility C. Air cargo facility D. Truck terminal/ reload facility
E. Residential F. Manufacturing G. Wholesale H. Retail store
I. Hospital/ medical J. Public/ government K. Office services L. Distribution Center
M. Agricultural Processing/ Packaging N. Truckstop, Roadside rest area, or motel/ hotel
O. Other P. Don’t know ( Do not state)
18
2.5.2 Video Recording
Video recordings were taken at each site to document truck counts by body type. The tapes were
reviewed and truck counts documented for a total of four hours over each sites operational time.
For example, a 24- hour facility would have 10 minutes of truck counts by body type documented
for each of its 24 hours of operation ( 10 minutes x 24 hours/ 60 = 4hours). A 12- hour facility
would have 20 minutes of tape review for each hour of operation ( 20 minutes x 12 hours/ 60 = 4
hours). The recordings were conducted seasonally by site to be consistent with the on- site
survey schedule.
19
2.6 Geocoding
The last stop and next stop locations for all 8,287 surveys have been geocoded to latitude and
longitude using Microsoft Streets and Trips. There was an extensive cleaning process and the
57% of the data points were geocodable to an exact address, cross- streets, or a California border
location. An additional 26% were geocoded to cross- streets using the appropriate freeway or
highway and an exit street. Sixteen percent were geocoded to the center of a town or place name
location, and 1% could not be successfully geocoded. The cleaning and geocoding processes are
described in detail in Appendix I, which is Technical memorandum H. 1: Geocoding Process and
Results.
Using the geocoded data, each location was aggregated to the county level and subsequently to
the regional level. This is the data that was used to provide the county to county and region to
region last / next stop trip tables.
3.0 Survey Strengths and Weaknesses
This section presents a discussion of the California Heavy Duty Truck Survey methodology
including the strengths and weaknesses of the approach.
3.1 Sampling Design and Survey Execution
Strategic Consulting and Research designed and implemented truck intercept surveys using a
roadside interview method to obtain representative travel activity data from truckers using the
statewide transportation network. This survey was primarily designed to collect specific trip
segment travel behavior information for long- distance truck travel. Useful information for truck
travel modeling and planning were obtained using this survey method including last stop and
next stop locations, travel distance and times, commodity types, and truck types.
During the truck survey implementation phase, we refined this survey method to obtain
additional information about travel behavior in addition to the intercepted trip segment. For
example, questions were added to the survey instrument to better understand the truckers' entire
trip.
The strengths of this survey method included:
 Proven track record for obtaining representative truck travel behavior information;
 Traditionally have high response rates;
 Truckers are more amenable to responding to this survey method than other survey
methods; and
20
 Good truck travel information for travel modeling ( distribution modeling, model
validation) and transportation planning can be obtained using this survey method.
The weaknesses to using this survey method included:
 Additional data need to be collected from other sources to support modeling, and in
particular, truck trip generation; and
 Limited data were obtained to represent the characteristics of the entire ( or total) truck
trips.
Other truck data collection methods, primarily both the truck diary/ mail back method has proven
ineffective in obtaining representative truck travel behavior information compared to the method
used for this truck survey. Several areas including Sacramento, Southern California, and
Portland, Oregon, have failed to obtain quality and representative data using the truck diary/ mail
back method. In Sacramento, for example, the California Trucking Association ( CTA) led the
survey effort that proved to be costly, time consuming, and marginally successful. Many regions
throughout the United States have had similar experiences using the truck diary/ mail back
method.
3.2 Data Deficiencies
The project team considered proven survey methods to collect travel behavior information from
truck drivers. However, there were some weaknesses to the data collection methods, primarily
as a result of the following constraints:
 Survey locations were primarily restricted to Agriculture and CHP Inspection Station
Locations and in a small number of cases, rest stop locations were also used for
surveying;
 Survey resources were prioritized to collect data during a concentrated period during the
typical day ( 8 to 16 hour periods) rather than the entire 24- hour period; and
 Survey resources were also prioritized in the spring with less focus on other seasons
( summer, fall, and winter).
As a result, the full range of statewide truck travel behavior was not collected in this survey.
Many state routes and potentially many California counties are not being represented in the
survey process. Other data quality factors include the following:
 Only a limited number of surveys were collected on roadways entering and exiting
California during this survey. The external to internal California truck movement data is
currently limited and would need to be enriched with other data sources to support truck
modeling.
21
 A limited number of survey locations were collected on roadways into and out of major
metropolitan areas in California. This weakness has contributed to the need to collect
available data from other sources or to collect a fuller sample of inter- regional truck
movements as part of ongoing Caltrans activities.
 Truck last stop and next stop pairs surveyed at one location could have been surveyed at
other survey locations. In these instances, the resulting last stop and next stop trips
should be processed to identify the number of stations they would be expected to pass by
as part of the data modeling process.
 There will be numerous last stop and next stop pairs for which no crossings of survey
locations would be expected. In order to deal effectively with these potentially numerous
cases, it will be necessary to increase the survey sites and other survey data ( from
metropolitan areas, ports, etc.) to obtain estimated trip tables and/ or trip ends for
modeling.
 Because only trucks with three or more axles were included in this survey, and empty
vehicles were treated differently than loaded vehicles, the truck counts to be used for
survey expansion should only include the types of vehicles actually surveyed. Survey
expansion should be conducted separately by time of day and vehicle type ( and
potentially commodity type).
 Since commodity data and last stop / next stop data was only collected during the more
heavily traveled hours of the day, it does not provide a representative sample of how
these factors fall out in the low volume, late PM and early AM hours. Since they may be
significantly different than peak travel hours, some commodities and stops may be ( for
example agricultural products) might be under stated, and other categories overstated.
4.0 Recommendations
The truck survey intercept approach using roadside interviews implemented for this survey was
the most appropriate method to obtain representative long- distance truck travel in the state of
California. This method has proven effective in many metropolitan areas and states including
Oregon, southern California, and Ohio.
In addition, it was recognized early on in this project that additional, supplementary data would
be needed to support this survey effort, and in particular, to support truck travel model
development. For example, this intercept data, in addition to the following supplementary data,
were specified in the Model Design Plan as requirements to support statewide truck travel
modeling and planning in California:
 Commodity flow data by origin and destination collected by other public agencies or the
private sector;
22
 Truck classification counts collected at interview locations through videotaping; and
 Available truck counts and travel behavior information at Port facilities ( Oakland, Long
Beach) obtained from Port representatives.
4.1 Collect Supplemental Survey Data
Supplementary information could be obtained to enrich the vehicle intercept, commodity flow,
and truck classification data already available to Caltrans. The following additional data
collection efforts are recommended to support this survey and future truck modeling for
California:
 Collect Intercept Data at External Cordons2. Conduct additional surveys, using the same
intercept survey method on roadways entering and exiting the state of California. This is a
weakness of the data collection plan for this project. For example, locations surveyed in this
project were limited to CHP and Agriculture Inspection and Rest Stop locations that do not
encompass an entire cordon of truck movements entering and exiting the state. This data
would be very useful in developing external- to- internal truck models for California. In order
to use this method, Caltrans should explore the possibility of using Rest Stop locations and
other methods ( establish temporary survey stations) to survey truckers coming into and out of
the state.
 Collect Intercept Data at Regional Cordons. Collect truck vehicle intercept data using this
project's survey method representing regional cordon areas to capture truck trips to and from
large regional centers, i. e., Bay Area to/ from Sacramento, Sacramento to/ from
Fresno/ Bakersfield. The limitations of this project regarding survey locations has contributed
to the need to collect a fuller sample of inter- regional truck movements. As with the external
surveying above, we should explore the possibility of using Rest Stop locations and other
methods ( establish temporary survey stations) to survey truckers coming into and out of the
state.
 Design and Collect Freight Distribution Center Surveys. Commodity flow data collected by
the U. S. Bureau of the Census includes good information on the ultimate origin and
destination of activity but, generally does not obtain commodity flow data related to re- load
facility activity. For example, truck activity and commodity flow data at key distribution
centers, such as in Stockton, are not represented in the commodity flow databases available
in other Divisions in Caltrans. A supplemental survey targeted at key distribution centers in
California ( Stockton, San Bernardino, Riverside, Otay Mesa) is recommended to better
understand truck activities at these re- load facilities that are traditionally under- represented.
These surveys can be designed to have dispatchers distribute and collect surveys on travel
behavior to truckers entering and exiting these facilities or by having survey interview teams
2 Although this was an original objective of the project, this approach was not employed in this study because there
was no feasible way to safely conduct interviews at many of these locations where there was no CHP or Agricultural
inspection facility or rest area where trucks could safely pull over. If this approach is to be pursued in future data
collection efforts, some new method will have to be developed to collect the data in a safer manner.
23
conduct the surveys at each location. This method has worked in Monterey, California and
other locations in the United States.
At the conclusion of this project, we completed over 8,000 truck intercept surveys. These
surveys in conjunction with additional data, to be collected in future, can be used to develop
statewide truck models for California. The three supplemental surveys identified above are
designed to provide Caltrans with additional truck activity to represent truck movements on the
entire transportation system including into and out of major distribution truck centers.
Supplemental survey questionnaires and sampling plans should use the current heavy duty truck
travel survey methods as its foundation with some revision required associated with the Freight
Distribution Center Surveys.
4.2 Increase Survey Productivity
Survey productivity was negatively impacted by the following factors:
 Inadequate space at CHP Platform stations
 Facility closures due to freeway congestion and Hazardous Materials alarms
 Unannounced facility closures due to facility staff reassignments and construction being
performed at the facilities
Several CHP Platform facilities have severe space restrictions, which can greatly limit the
frequency that surveys can be conducted. As there is not a uniform site layout it is difficult to
predetermine which stations will present these challenges or how much they will impede the
survey process. If Platform stations are to be utilized for survey purposes they should be visited
prior to final planning efforts to realistically assess productivity expectations.
Most facility closures due to freeway congestion and other factors are unforeseen and
unavoidable, however, they are predictable. While most “ freeway congestion” closures last only
15- 25 minutes, they are frequent enough to impact productivity. Hazardous Material alarms can
last 2 hours or more and severely effect productivity. A 10- 15% reduction factor should be
applied to overall productivity expectations to reflect these circumstances.
There were instances where surveyors arrived at a facility that was closed for the day, even
though the survey day had been scheduled in advance and confirmed with the station no more
than a week preceding the survey date. More coordination with the CHP is recommended to
increase the survey productivity. This will be possible if a CHP officer is assigned to guide the
survey crews and their activities at each site.