• Federal Highway Administration >
• Publications >
• Research Publications >
• 01159 >
• 3.Cfm >
• Model Development For National Assessment of Commercial Vehicle Parking

Model Development For National Assessment of Commercial Vehicle Parking

PDF files can be viewed with the Acrobat® Reader®

3.0 ESTIMATING TRUCK PARKING DEMAND

As part of the 1996 Study, a location-specific parking demand model was developed and calibrated to assess the demand for truck parking at individual rest areas.⁽¹⁾The structure of this model was not suitable, however, for incorporating the impact of parking provided by private truck stops, as data collected for the development of the model were obtained only from public rest areas. In addition, this model did not adequately account for the spatial distribution of parking opportunities and the effect of these opportunities on demand at a particular location. Therefore, a new model was formulated for this study—a model that bases parking demand on a segment of highway or corridor rather than an individual parking facility.

The corridor model developed for this study uses a somewhat different approach to estimating parking demand. Rather than basing the demand for parking on the characteristics of a parking facility, the model predicts truck parking demand for a highway segment based on total truck-hours of travel and the time and duration of stops.This approach is based on the theory that demand for parking is better explained by hours driving than by attributes of individual truck stops and rest areas. The model also considers the ratio of short-haul to long-haul trucks and the propensity to use public or private parking spaces for different parking purposes. Although the corridor model limits the conclusions that can be drawn about the spaces or amenities required at a specific truck stop or rest area (e.g., the need for lighting, additional parking, etc.), it was considered to be a more appropriate way to estimate truck parking demand for the purposes of this study. Building the modeling framework around this system-level approach also provided a basis to examine the influences of hours-of-service (HOS) regulations as well as driving time and distance on parking demand.

3.1 Model Development

The modeling framework begins by estimating the truck-hours of travel using annual average daily traffic (AADT), [1] percent trucks, length of the roadway segment being analyzed, and the speed limit or average truck speed.The key parameter in the model is the number of hours of parking required by drivers given the number of hours they travel.  Thus, Federal HOS regulations have an indirect, but very real, effect on parking demand; the more hours of parking required for a given period of time on the road (i.e., the higher the ratio of parking time to driving time), the higher the estimated parking demand.  The model produces a peak-hour estimate of parking spaces demanded for a highway segment.

Because short-haul drivers (i.e., those not making overnight trips) make relatively short stops, parking demand is based on minutes of parking time per hour on the road.For long-haul trips, when an overnight rest stop is required on the road, hours of parking demand are calculated using a ratio of parking time to driving time.This ratio of parking time to driving time takes into account HOS regulations and information from drivers regarding how they use their time throughout a typical week. Then, peak-parking factors are used to convert the 24-hour parking demand into peak-hour parking demand. Using the model, peak-parking demand can be estimated for different percentages of short-haul and long-haul trucks (e.g., 10 percent short-haul and 90 percent long-haul, 50 percent short-haul and 50 percent long-haul, etc.) and for different driver preferences for parking at public rest areas or private truck stops.

3.2 Data Requirements and Parameter Values

This section presents the data requirements for the model and a discussion of how the values for each model parameter were derived, whether based on field studies, driver survey results, professional judgment, or assumption.It is important to keep in mind that these “default” values represent national norms and do not necessarily reflect regional variations.When applying the parking demand model, users are encouraged to select parameter values that represent conditions within their local area.

The data requirements for the model are summarized in table 1. Most of these data are available through the Highway Performance Monitoring System (HPMS) or through a State’s own databases and information systems.⁽¹⁴⁾Model parameters and their values are shown in table 2.

Table 1. Data requirements for truck parking demand model.

Because travel demand is variable, traffic engineering analyses generally focus on the peak periods of travel (e.g., peak hour of the day, peak month of the year, etc.). Variation of traffic by month or season is primarily a function of the type of route and kinds of activities present in the area. For example, highways serving winter resort areas peak during winter months, while highways serving agricultural activities peak during the summer months.Due to proximity to a metropolitan area, urban routes tend to show less variation in traffic by season then do rural routes. For the model, a seasonal peaking factor of 15 percent (1.15) was used and represents the peaking characteristics of all vehicles (i.e., not specific to trucks).

The short-term parking duration per hour traveled was assumed to be five minutes, which results in one 40-minute stop during an 8-hour shift.This assumption was based on professional judgment and information obtained from talking with drivers about their typical stopping patterns.

Table 2.  Demand model parameters.

A national survey of commercial truck drivers was undertaken in another task of this study and is documented in a separate report.⁽²⁾The survey was administered to over 2,000 truck drivers in select regions across the United States.Survey responses were used to determine truck drivers’ needs, preferences, and travel patterns (e.g., why, when, and where they park). This information was used to calibrate several of the parameter values in the model so that they would more accurately represent drivers’ behaviors and travel patterns.Driver survey results were used to determine values for the following parameters: average hours spent loading/unloading per week, average hours spent at home per week, average hours spent parking for rest at shipper/receiver per week, and the portion of demand for public rest area and private truck stop spaces.

The importance of determining the values of the loading/unloading time, at-home time, and time spent resting at shipper/receiver was to calculate the amount of time a driver demands parking on the road in a typical week.To determine this time, drivers’ daily activities including driving, on-duty non-driving time, and time spent at home, must be considered.

To begin, the hours that a driver spends on the road in a week are limited by the Federal HOS regulations.  Although there are different regulations for different types of carriers, the majority of long-haul drivers operate seven days a week. In this case, the Federal HOS regulations allow for no more than 70 hours on duty in any period of eight consecutive days.⁽¹⁵⁾

However, a driver’s time is not spent solely driving; they must also spend time at shippers and receivers loading and unloading their trailers.This is considered on-duty, non-driving time.While some drivers will load/unload several times per week, others may do so only once per week.The average hours spent loading/unloading the truck (whether the driver actually loads/unloads or waits for it to be done) was determined from a question that asked drivers how many hours, on average, per week they spend loading or unloading their trucks. The average response to this question was approximately 15 hours per week.

When drivers are off duty, they are sometimes able to return home.While some drivers are home every weekend, others may not make it home but a few days each month. The average hours spent at home per week was determined from a question on the driver survey that asked drivers how many days, on average, did they sleep at home each month.  The average response to this question was 6.7 days per month, which translated into approximately 42 hours in eight days.

Finally, in some situations, drivers are allowed to park for rest at a shipper/receiver prior to loading or unloading. Thus, they are not always looking for long-term parking along the highway.The average number of hours spent parking for rest at a shipper/receiver was determined from a question that asked drivers how many times on average in a typical week they park for long-term rest at a series of different locations. The average response to the “loading/unloading location” question was 2.6 times per week. Anecdotal information gathered from discussions with truck drivers suggests an average of six hours of rest per long-term stop.Using this information from drivers, 2.6 times per week translates into approximately 16 hours per week of rest at shippers/receivers.

From this, the amount of time a driver will demand parking along the highway in a week can be determined by taking the total number of hours in an eight-day period (192) and subtracting the time that drivers spend on-duty driving (70 hours), on-duty not driving (15 hours), off-duty (42 hours), and parking other places than along the road (16 hours). Therefore, the total hours of parking demanded per long-haul truck per week, used for this model, was 49 hours.

The proportions of total parking demand for rest area spaces and for truck stop spaces were derived based on responses to questions in the driver survey regarding where drivers prefer to stop for different activities (e.g., long-term rest, restroom, meal, etc.).Table 3 shows the data from the survey and illustrates how the data were used to derive the values for the proportion of demand for rest area and truck stop spaces. The values were derived as follows: 1) the number of driver responses for each preference category (i.e., rest area, truck stop, no preference) was weighted according to the average amount of time spent parking for each activity (thereby converting number of drivers into number of truck-hours of parking according to preference); 2) the truck-hours of parking were then summed for each preference category; 3) the truck-hours of parking in the “no preference” category were then divided evenly between the rest area and truck stop preference categories; and 4) the total truck-hours of parking for rest areas and truck stops were then divided into the overall total truck-hours of parking.This process resulted in values for the proportion of parking demand for rest area and truck stop spaces of 0.23 and 0.77, respectively.

The driver survey conducted for this study, as well as several of the State surveys reviewed at the beginning of this report, indicated that in general, drivers prefer to use rest areas when making a short stop (to make a phone call, get a snack, or use the restroom), because they are more convenient to the highway than truck stops.On the other hand, survey responses indicated that most long-haul drivers prefer to make their long-term rest stops in truck stops, because they provide more services (fuel, meal, showers) than rest areas. The proportion of total trucks that are short-haul and long-haul and the long-haul peak parking factor were calibrated using data from field surveys.The calibration of these parameters is discussed in detail in a later section titled, Truck Parking Demand Model Calibration.

3.3 Parking Demand Model

This section presents the parking demand model in a step-by-step fashion, referring to table 1 and table 2 where data and parameter values are required as input to the equations.Table 4 lists and describes the terms calculated by each of the 12 equations in the step-by-step model process.Once again, analysis is done at the segment level.

Table 4. Terms calculated in step-by-step model process.

The first step in the parking demand estimation is to calculate the seasonal peak daily truck volume using the AADT, the percent trucks, and the seasonal peaking factor.  The seasonal peak daily truck volume, V_t, is expressed in trucks per day: The first step in the parking demand estimation is to calculate the seasonal peak daily truck volume using the AADT, the percent trucks, and the seasonal peaking factor. The seasonal peak daily truck volume, V_t, is expressed in trucks per day:

Vt = AADTX Pt X Fs

(1)

where	AADT	= annual average daily traffic (vehicles/day)
	Pt	= percent of total traffic that is trucks
	FS	= 1.15, seasonal peaking factor

Next, using the length of the analysis segment (defined by the analyst) and the speed limit or average truck speed, the average truck travel time, TT, for the segment is calculated in hours per truck:

(2)

where	L	= analysis segment length (km)
	S	=speed limit or average truck speed (kph)

Then, using the daily truck volume calculated in equation 1, the truck travel time calculated in equation 2, and the proportion of trucks that are short- and long-haul, the total daily truck-hours of travel, THT, for short-haul and long-haul trucks can be estimated:

THTSH = PSHXVt X TT

(3)

THTLH = PLHVt XTT

(4)

where P_SH = proportion of total trucks that are short-haul P_LH = proportion of total trucks that are long-haulAfter the total daily truck-hours of travel on the segment have been computed, the short-haul truck-hours of parking per day, THP_SH, can be estimated based on the time parking per time driving. Using the default value for duration of short-term stops (i.e., a driver will need to stop an average of five minutes for every hour of driving) and the truck-hours of short-haul travel calculated in equation 3, the daily short-haul truck-hours of parking demand, THP_SH, can be estimated:

(5)

where DST

= duration of short-term stops per hour traveled (min/hour)

For long-haul drivers, HOS regulations affect the number of hours they must spend parking in a given period of time. As previously stated, long-haul drivers must spend eight hours parking after ten hours of driving, and cannot be on duty more than 70 hours in eight consecutive days. The parking time per week can be computed by subtracting the maximum number of hours spent driving, the average number of hours spent at home, the average number of hours spent loading/unloading, and the average number of hours spent parked to rest at shippers/receivers from the 192 hours in a eight days. Referring to the default values for each of these items in table 2, the parked time per week would be approximately 49 hours (192-70-42-15-16 = 49), and the ratio of parking time to driving time is therefore 49 hours/70 hours (i.e., on average, a long-haul driver will stop for long-term rest approximately 49 hours for 70 hours of driving in a week). In addition, assuming that they will also park an average of five minutes for every hour of driving for purposes other than long-term rest, the daily truck-hours of long-haul travel, calculated in equation 4, can be used to estimate the daily long-haul truck-hours of parking demand, THP_LH, on the segment:

(6)

where DST

= duration of short-term stops per hour traveled (min/hour)

Now that the daily truck-hours of short- and long-haul parking demand for the segment have been estimated, the number of trucks demanding a parking space in the peak hour needs to be determined. This conversion can be made by considering the proportion of the daily truck-hours of parking demand that occurs during the peak hour. Assuming that all trucks occupy a space for at least one hour, the conversion from daily truck-hours of parking to truck-hours per hour can be made with a peak-hour parking factor (PPF). The units of the peak-hour parking demand (PHP) then become trucks or spaces. Default values for short-haul peak parking factor (PPF_SH) and long-haul peak parking factor (PPF_LH) have been set at 0.02 and 0.09, respectively (a discussion of the calibration is presented in the next section). Using these default values and the truck hours of parking, the peak-hour short-haul and long-haul parking demand, PHP_SH and PHP_LH, respectively, can be calculated:

PHPSH = PPFSHXTHPSH = 0.02 X THPSH

(7)

PHPLH = PPFLHXTHPLH = 0.09 X THPLH

(8)

where	PPFSH	= peak-parking factor for short-haul trucks, 0.02
	PPFLH	= peak-parking factor for long-haul trucks, 0.09

Finally, the total peak-hour parking space demand is distributed between public rest areas (RA) and private truck stops (TS) using preferences established from responses to the driver survey:

PHPSH, RA = PRAXPHPSH = 0.23 X PHPSH	(9)
PHPSH, TS = P TSXPHPSH = 0.77 X PHPSH	(10)
PHPLH, RA = PRAXPHPLH = 0.23 XPHPLH	(11)
PHPLH, TS = P TSXPHPLH = 0.77 X PHPLH	(12)

where	PRA	= proportion of demand for rest area spaces
	PTS	= proportion of demand for truck stop spaces

3.4 National Assessment of Commercial Vehicle Parking

The national assessment of commercial vehicle parking availability compared existing and projected supply to existing and projected demand along segments of the NHS. This process incorporated three steps that are listed in table 5. The national assessment is documented in a separate report.⁽³⁾

Table 5. National truck parking assessment process.

3.4.1 Step 1: Identify Major Trucking Corridors and Select Analysis Segments

The objective of this step was to develop a database consisting of highway analysis segments that make up the major trucking corridors along the NHS in the U.S. For this project, corridors that carried current truck traffic exceeding 1,000 trucks per day were considered to be major trucking corridors. Forecasts were for 20 years into the future (i.e., 2020).

Several sources were used to determine major trucking corridors: 1) the HPMS, 2) the Heavy Commercial Vehicle Flow Atlas of the U.S. NHS (derived from State department of transportation records and selected toll road authorities), and 3) a survey of State DOTs conducted as part of this study.^(14,16) After the corridors were identified, shorter analysis segments were selected under the following criteria:

• Truck volume along the segment was relatively uniform.
• For consistency, segments were between 100 kilometers (60 miles) and 320 kilometers (200 miles) in length.
• Logical end points for the analysis segments included: 1) urban-rural transition points; 2) rural interchanges with other major truck routes and bypasses; 3) major cities; and 4) areas with significant truck terminal facilities including ports, warehouses, rail yards, and other intermodal freight transfer points.

After the analysis segments were identified, information on each was entered into a database.

3.4.2 Step 2: Inventory Truck Parking Facilities

The objective of this step was to inventory public and private parking space supply for each analysis segment. A survey of State DOTs and Interstate America's Truck Stop Directory were used for this purpose.⁽¹⁷⁾ A survey of State DOTs was administered to obtain information on the number of rest areas operated by the State. Questions regarding the number of truck parking spaces available at each rest area, as well as AADT and percent trucks, were also included on the survey form. The Interstate America's Truck Stop Directory is a comprehensive database with location and amenity information of private truck stop/travel plaza facilities. It is updated on an annual basis and describes over 7,000 facilities in the U.S. and Canada that allow commercial vehicle parking.

3.4.3 Step 3: Apply Truck Parking Demand Model

The objective of this step was to estimate parking demand for each analysis segment using the parking demand model. A model application is presented here.

US 000 from Queenstown to Kingsville is a four-lane highway with a posted speed limit of 105 kph (65 mph). This segment of highway is 210 kilometers (130 miles) in length and carries approximately 17,500 vehicles per day, 18 percent of which are trucks. A segment between these two cities was selected for analysis for two reasons: 1) this section of US 000 is known to carry a high volume of truck traffic, and 2) the traffic volume along the segment is fairly uniform. There are three public rest areas and three private truck stops located along the segment.

What is the public/private parking demand along this segment? Is there a surplus or a shortage of public/private parking along this segment?

Problem Summary:

Using equations (1) through (12), the parameter values shown in table 2, and the values from the example shown above, the short- and long-haul parking space demand were calculated:

The total peak-hour parking demand for public rest areas is 1+75 = 76 trucks, and the total peak-hour parking demand for private truck stops is 3+252 = 255 trucks. Considering the supply of parking spaces on this segment, there is a shortage of public rest area parking of 51-76 = (-) 25 spaces, while there is a surplus of private truck stop parking of 275-255 = 20 spaces.